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/W4310695031	https://zenodo.org/record/7397344/files/4..pdf	Marathi (Marāṭhī)	null	पंढरपूर तालुक्यातील सहकारी चळवळ	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	3,098	चळवळ 2सांशोधक तिद्याथी, इततहास तिभाग, पु. अ. होळकर, सोलापूर तिद्यापीठ, सोलापूर Corresponding Author- डॉ. तानाजी जनाधधन फुलारी प्रस्तािना : ्र ्र् प्राचीन काळापासून सहकार हा मानिी सांस्कृतीचा एक महत्त्िाचा भाग बनला आहे. मानिाच्या तशकारी अिस्थेपासून आजच्या प्रगत अिस्थेपयंतचा तिचार केला तर असे आढळते की, एकतितपणे तिचार करणे ि काम करणे ही मानिाची सहज आतण नैसर्गगक प्रिृत्ती आहे. त्यातून त्याला सुखी आयुष्य जगण्याची प्रेरणा तमळते. त्यातूनच त्याच्या आर्गथक आतण सामातजक आयुष्यात पुढे क्ाांतीकारक बदल घडून आले. सहकार ही मानिी सहजीिनाच्या कल्पनेची एक प्रगत अिस्था आहे. यामध्ये मानिास आर्गथक ि सामातजक तिकासासाठी आतण अन्याय तनिारण्यासाठी एकि येण्याची प्रखर इच्छा तनमाधण होते. मानिाच्या जीिनप्रणालीचा इततहास हा सहकाराचा इततहास आहे. त्यामुळे आधुतनक अथधव्यिस्थेत सहकाराला तिशेष महत्त्ि प्राप्त झाले आहे. सहकार म्हणजे एकमेकाांना मदत करणे, कौटुांतबक ि सामातजक जीिनात एकतित राहणे ि काम करणे. सहकार आतण मानिी सहजीिन याांचा जिळचा सांबांध आहे. सहकारामध्ये व्यक्तिंनी एकतित राहण्याबरोबरच एकमेकाांना मदत करण्यासाठी केलेल्या सामूतहक प्रयत्ाांचाही समािेश होतो. सिांनी तमळून एकमेकाांच्या मदतीने सिांच्या तहतासाठी, एकतित काम करणे याला सहकारात तिशेष महत्त्ि आहे. ३) भारतीय सहकार कायदा (१९१२) - "सहकारी तत्त्िाांना अनुसरुन आपल्या सभासदाांचे आर्गथक तहतसांबांध िृद्धींगत करण्याचे ज्या सांस्थेचे उद्दिष् असते ततला सहकारी सांस्था असे म्हणतात". सहकाराला इांग्रजीमध्ये Co-operation असे म्हणतात. Co-operation हा मूळ शब्द लॅटटन भाषेतील Cooperari या शब्दापासून आला आहे. या शब्दात co म्हणजे 'सह' ककिा एकि ि operari म्हणजे कायध करणे असा या शब्दाांचा अथध आहे. त्यामुळे Co-operation म्हणजे एकतितपणे काम करणे होय. कोणतेही काम एकतितपणे कराियाचे असल्यास त्यासाठी इतर व्यक्तिंची मदत घेणे आिश्यक असते. लोकाांना स्ित:च्या ज्या गरजा िैयतिंकटरत्या पूणध करता येत नाहीत अशा गरजा एकमेकाांच्या साहाय्याने सहकारात पूणध केल्या जातात. सहकार म्हणजे समाजातील आर्गथकदृष्या दुबधल व्यक्तिंनी एकतित येऊन एकमेकाांच्या मदतीने आपल्या तितिध गरजा पूणध करण्यासाठी स्थापन केलेली सांस्था होय. ्ह ४) सहकार तनयोजन सतमती (१९४६) -"सहकार हा व्यिसाय सांघटनेचा एक प्रकार असून व्यतिं आपल्या तहतरक्षणासाठी समानतेच्या तत्त्िािर स्िेच्छेने एकि येतात." ५) प्रा. पॉल लॅम्बटध - "स्ित:च्या उपयोगासाठी व्यक्तिंच्या समूहाने तनमाधण केलेली आतण द्ददग्दर्गशत केलेली, लोकशाही तनयमानुसार चालतिलेली, सभासद ि एकूण समाज याांच्या सेिेसाठी स्थापन झालेली व्यिहार सांस्था म्हणजे सहकारी सांस्था होय." ं् ्ह हं् ह पांढरपुरचे भौगोतलक स्थान : पांढरपुरचे भौगोतलक स्थान 17° 41° ते 39° उत्तर अक्षाांश आतण 75° ते 19' 49" पूिध रेखाक्षाांशी सीतमत आहे. पांढरपुरचे क्षेिफळ साधारणता १,३०३.६ चौ. द्दकमी आहे. समुद्र सपाटीपासून पांढरपूरची उांची ४६५.१२ मी आहे. पांढरपुरातील महसुली गािाांची सांख्या १०१ आहे. व्याख्या: अनेक तिचारिांताांनी सहकाराच्या ककिा सहकारी सांस्थेच्या केलेल्या व्याख्या पुढीलप्रमाणे. व्याख्या: अनेक तिचारिांताांनी सहकाराच्या ककिा सहकारी सांस्थेच्या केलेल्या व्याख्या पुढीलप्रमाणे. ं््््य् ु्र १) एच. कॅल््हटध- "मानिी भूतमकेतून व्यक्तिंनी स्िेच्छेने एकि येऊन आपल्या िृद्धीसाठी समानतेच्या तत्त्िािर स्थापन केलेली सांघटना म्हणजे सहकारी सांस्था होय." ं््््य् ु्र १) एच. कॅल््हटध- "मानिी भूतमकेतून व्यक्तिंनी स्िेच्छेने एकि येऊन आपल्या िृद्धीसाठी समानतेच्या तत्त्िािर स्थापन केलेली सांघटना म्हणजे सहकारी सांस्था होय." International Journal of Advance and Applied Research www.ijaar.co.in ISSN – 2347-7075 Impact Factor – 7.328 Peer Reviewed Bi-Monthly Vol.3 No.5 Nov – Dec 2022 चळवळ . 1 . 2 1इततहास तिभाग प्रमुख ,तिज्ञान महातिद्यालय,साांगोला, 2सांशोधक तिद्याथी, इततहास तिभाग, पु. अ. होळकर, सोलापूर तिद्यापीठ, सोलापूर Corresponding Author- डॉ. तानाजी जनाधधन फुलारी D ्र् ािना International Journal of Advance and Applied Research www.ijaar.co.in ISSN – 2347-7075 Impact Factor – 7.328 Peer Reviewed Bi-Monthly Vol.3 No.5 Nov – Dec 2022 चळवळ . 1 . 2 1इततहास तिभाग प्रमुख ,तिज्ञान महातिद्यालय,साांगोला, 2सांशोधक तिद्याथी, इततहास तिभाग, पु. अ. होळकर, सोलापूर तिद्यापीठ, सोलापूर Corresponding Author- डॉ. तानाजी जनाधधन फुलारी D ्र् ािना International Journal of Advance and Applied Research www.ijaar.co.in ISSN – 2347-7075 Vol.3 No.5 ंं भारतीय सहकारी चळिळ : ्यं्ं् ्ं् ह हं् ह २) श्री. बैकुांठलाल मेहता (भारतीय सहकारी चळिळीचे प्रणेते) -"समान गरजा असलेल्या व्यक्तिंनी एकि येऊन आपल्या समान आर्गथक उद्दिष्ाांच्या पूतधतेसाठी स्िेच्छेने स्थापन केलेला सांघटन प्रकार म्हणजे सहकारी सांस्था होय. ्ं् ह हं् ह २) श्री. बैकुांठलाल मेहता (भारतीय सहकारी चळिळीचे प्रणेते) -"समान गरजा असलेल्या व्यक्तिंनी एकि येऊन आपल्या समान आर्गथक उद्दिष्ाांच्या पूतधतेसाठी स्िेच्छेने स्थापन केलेला सांघटन प्रकार म्हणजे सहकारी सांस्था होय. १८ व्या शतकात इांग्लांडमध्ये झालेल्या औद्योतगक क्ाांतीमुळे िस्तूांचे मोठया प्रमाणािर उत्पादन होऊ लागले. पक्कक्कया मालाची हक्काची बाजारपेठ ि कच्च्या मालासाठी भारताच्या 22 IJAAR पटरणामी शेतीिर अिलांबून असणाऱयाांची सांख्या िाढली. शेतीिरील अततटरिं भार, दुष्काळ यामुळे शेतकऱयाांची आर्गथक तस्थती दुबधल झाली. शेतकऱयाांचा कजधबाजारीपणा िाढला. शेतकऱयाांिरील कजाधचा बोजा िाढून ते सािकारी पाशात अडकले. शेतकऱयाांना सािकारी पाशातून मुिं करण्यासाठी सरकारने शेतकऱयाांना पतपुरिठा करण्याच्या दृष्ीने कायदेशीर उपाय योजले. मुांबई प्राांतातील बडोदा येथे भारतातील पतहली सहकारी पतपुरिठासांस्था ५ फेिुिारी, १८८९ रोजी स्थापन झाली. प्रा. तिठ्ठल लक्ष्मण किठेकर याांनी अन्योन्य सहाय्यकारी मांडळी' या नािाने पतहली पतपुरिठा सांस्था स्थापन केली. तिटटश सरकारने भारतातील सहकारी चळिळीस मूतध स्िरूप देणारा पतहला सहकार कायदा सन १९०४ मध्ये सांमत केला. भारतीय शेतकऱयाांच्या आर्गथक ि सामातजक जीिनास निे िळण देणारा हा कायदा होता. हा कायदा फिं पतसांस्था स्थापन करण्यापुरताच मयाधद्ददत होता. सन १९०४ च्या कायद्यातील िुटी ि उतणिा दूर करण्यासाठी सन १९१२ मध्ये दुसरा सहकार कायदा सांमत केला. या सहकार कायद्यामुळे सहकारी सांस्थाांची स्थापना सिधक्षेिात होऊ लागली. भारतातील सहकारी चळिळीची प्रगती ि िुटी अभ्यासण्यासाठी सन १९१४ साली तिटीश सरकारने एडिडध मॅक्कलेगन याांच्या अध्यक्षतेखाली एक सतमती नेमली. या सतमतीने सहकारी चळिळीच्या भतिष्यातील िाटचालीबाबत मौतलक सूचना केल्या. पतहल्या महायुद्धानांतर सन १९१९ मध्येतिटटश सरकारने ‘मॉन्टफोडध सुधारणा कायदा' मांजूर केला. या कायद्यामुळे सहकार हा तिषय केंद्र सरकारकडून प्राांततक सरकाराांच्या अखत्याटरत आला. त्यानुसार सिधप्रथम मुांबई प्राांताने सन १९२५ मध्ये सहकार कायदा सांमत केला. सहकारी चळिळीची पुनरधचना करण्यासाठी ज्या तितिध सतमत्या तनयुिं केल्या होत्या त्याांच्या तशफारशीिरुन भारतात टरझ्हध बँकेची स्थापना १ एतप्रल, १९३५ रोजी करण्यात आली. या बँकेने शेती पतपुरिठा तिभाग सुरू केला. यानांतरच्या काळात केंद्र सरकारने सहकाराच्या तिकासासाठी प्रा.धनांजयराि गाडगीळ याांच्या अध्यक्षतेखाली 'कृषी अथधपुरिठा सतमती' (१९४४) ि श्री.आर.जी.सरैय्या याांच्या अध्यक्षतेखाली 'सहकार तनयोजन सतमती' (१९४५) तनयुिं केली. बँककग सोयी ि सुधारणासाठी सरकारने श्री. पुरुषोत्तमदास ठाकुरदास याांच्या अध्यक्षतेखाली ‘ग्रामीण ि बँककग चौकशी सतमती' (१९४९) नेमली. सन १९५१ मध्येग्रामीण कजधपुरिठ्याचा अभ्यास करून तशफारस करण्यासाठी 'अतखल भारतीय ग्रामीण पतपुरिठा पाहणी सतमतीची' डॉ. एडीगोरिालायाांच्याअध्यक्षतेखालीतनयिंीकेलीहोती IJAAR भूप्रदेशाचा िापर तिटटशाांनी केला. औद्योतगक क्ाांतीचा दुष्पटरणाम ि तिटटशाांच्या कूटनीतीमुळे भारतातील लघु ि कुटीरोद्योग बांद पडले. कारागीर ि कामगार बेकार झाले. रोजगारासाठी ते शेतीकडे िळले. पटरणामी शेतीिर अिलांबून असणाऱयाांची सांख्या िाढली. शेतीिरील अततटरिं भार, दुष्काळ यामुळे शेतकऱयाांची आर्गथक तस्थती दुबधल झाली. शेतकऱयाांचा कजधबाजारीपणा िाढला. शेतकऱयाांिरील कजाधचा बोजा िाढून ते सािकारी पाशात अडकले. शेतकऱयाांना सािकारी पाशातून मुिं करण्यासाठी सरकारने शेतकऱयाांना पतपुरिठा करण्याच्या दृष्ीने कायदेशीर उपाय योजले. मुांबई प्राांतातील बडोदा येथे भारतातील पतहली सहकारी पतपुरिठासांस्था ५ फेिुिारी, १८८९ रोजी स्थापन झाली. प्रा. तिठ्ठल लक्ष्मण किठेकर याांनी अन्योन्य सहाय्यकारी मांडळी' या नािाने पतहली पतपुरिठा सांस्था स्थापन केली. तिटटश सरकारने भारतातील सहकारी चळिळीस मूतध स्िरूप देणारा पतहला सहकार कायदा सन १९०४ मध्ये सांमत केला. भारतीय शेतकऱयाांच्या आर्गथक ि सामातजक जीिनास निे िळण देणारा हा कायदा होता. हा कायदा फिं पतसांस्था स्थापन करण्यापुरताच मयाधद्ददत होता. सन १९०४ च्या कायद्यातील िुटी ि उतणिा दूर करण्यासाठी सन १९१२ मध्ये दुसरा सहकार कायदा सांमत केला. या सहकार कायद्यामुळे सहकारी सांस्थाांची स्थापना सिधक्षेिात होऊ लागली. भारतातील सहकारी चळिळीची प्रगती ि िुटी अभ्यासण्यासाठी सन १९१४ साली तिटीश सरकारने एडिडध मॅक्कलेगन याांच्या अध्यक्षतेखाली एक सतमती नेमली. या सतमतीने सहकारी चळिळीच्या भतिष्यातील िाटचालीबाबत मौतलक सूचना केल्या. पतहल्या महायुद्धानांतर सन १९१९ मध्येतिटटश सरकारने ‘मॉन्टफोडध सुधारणा कायदा' मांजूर केला. या कायद्यामुळे सहकार हा तिषय केंद्र सरकारकडून प्राांततक सरकाराांच्या अखत्याटरत आला. त्यानुसार सिधप्रथम मुांबई प्राांताने सन १९२५ मध्ये सहकार कायदा सांमत केला. सहकारी चळिळीची पुनरधचना करण्यासाठी ज्या तितिध सतमत्या तनयुिं केल्या होत्या त्याांच्या तशफारशीिरुन भारतात टरझ्हध बँकेची स्थापना १ एतप्रल, १९३५ रोजी करण्यात आली. या बँकेने शेती पतपुरिठा तिभाग सुरू केला. यानांतरच्या काळात केंद्र सरकारने सहकाराच्या तिकासासाठी प्रा.धनांजयराि गाडगीळ याांच्या अध्यक्षतेखाली 'कृषी अथधपुरिठा सतमती' (१९४४) ि श्री.आर.जी.सरैय्या याांच्या अध्यक्षतेखाली 'सहकार तनयोजन सतमती' (१९४५) तनयुिं केली. बँककग सोयी ि सुधारणासाठी सरकारने श्री. पुरुषोत्तमदास ठाकुरदास याांच्या अध्यक्षतेखाली ‘ग्रामीण ि बँककग चौकशी सतमती' (१९४९) नेमली. सन १९५१ मध्येग्रामीण कजधपुरिठ्याचा अभ्यास करून तशफारस करण्यासाठी 'अतखल भारतीय ग्रामीण पतपुरिठा पाहणी सतमतीची' डॉ. ी ों्््ष े ीतं ी े ी ोी सन १९५४ मध्ये या सतमतीने आपला अह्िाल सरकारकडे सादर केला. या सतमतीने आपल्या अहिालात असे मत नोंदतिले की, ‘सहकार अयशस्िी ठरला आहे तथातप तो यशस्िी झालाच पातहजे.' स्िाांतत्र्य प्राप्तीनांतर भारत सरकारने सन १९५१ पासून पांचिार्गषक योजनाांच्या माध्यमातून देशाचा आर्गथक तिकास घडिून आणण्याचे धोरण स्िीकारले. पांचिार्गषक योजना दरम्यान सहकारी चळिळीचा तिकास ि सहकारी चळिळीत लोकाांचा सहभाग िाढतिणे यािर भर देण्यात आला. ्ट्र भूप्रदशाचा िापर तिटटशांनी कला. औद्योतगक क्ांतीचा दुष्पटरणाम ि तिटटशाांच्या कूटनीतीमुळे भारतातील लघु ि कुटीरोद्योग बांद पडले. कारागीर ि कामगार बेकार झाले. रोजगारासाठी ते शेतीकडे िळले. IJAAR IJAAR Vol.3 No.5 दृष्ीने तिचार केला असता असे द्ददसून येईल द्दक, सहकारी सोसायटी अतस्तत्िाचा कायदा १९०४ साली करण्यात आला. १९०५ साली बाशी तालुक्कयामध्ये 'तडिळे' या गािी पतहली सहकरी सोसायटी स्थापन झाली. सोलापूर तजल्यातील सहकारी चळिळीचा तिस्तार पांचिार्गषक योजनेच्या काळात मोठ्या प्रमाणािर झाला पतहली पांचिार्गषक योजना सुरू होण्याच्या पूिी या तजल्यात 471 सहकारी सोसायटी होत्या आतण त्याांच्या सभासदाांची सांख्या 83 हजार 700 होती ि खेळते भाांडिल अडीच कोटी रुपयाांचे होते इसिी सन 59 60 अखेर तजल्यातील सहकारी सांस्थाांची सांख्या 1294 झाली, तसेच या सोसायटीचे खेळते भाांडिल रुपये 6.91 कोटी होते. बदल होऊन ऊस, कापूस, डाक्तळब, द्राक्षे यासारखी नगदी तपके घेतली जाऊ लागली. या बदलत्या पीक पद्धतीमुळे सहकारी चळिळीच्या माध्यमातून तितिध सहकारी सांस्थाांची स्थापना झाली. सहकार क्षेिाने पांढरपूर तालुक्कयाच्या ग्रामीण अथधव्यिस्थेमध्ये आमूलाग्र बदल केला, क्ाांती घडतिली, सामान्य माणसाला उभे करण्याचे काम केले. पांढरपूर तालुक्कयात तितिध प्रकारच्या सहकारी सांस्थाांची स्थापना झालेली आहे. त्यामध्ये प्रामुख्याने ति. का. सेिा सहकारी सांस्था, सहकारी साखर कारखाने, नागरी सहकारी बँका, पतसांस्था, पगारदार सेिकाांच्या पतसांस्था, अबधन बँका, पणन सांस्था, शेतीमाल प्रद्दक्या सांस्था, ग्राहक भाांडार, मजूर सांस्था, बलुतेदार सांस्था, औद्योतगक िसाहती, सूततगरण्या, हातमाग तिणकर सांस्था, यांिमाग तिणकर सांस्था, देखरेख सांघ, शेती साधन पुरिठा सांस्था, गृहतनमाधण सांस्था, िाहतूक सांस्था इत्यादी तितिध प्रकारच्या सहकारी सह. सांस्थाांची स्थापना झाली आहे. सोलापूर तजल्यातील या चळिळीच्या तिकासाची खऱया अथाधने सुरुिात ती म्हणजे सोलापूर तजल्हा मध्यिती सहकारी बँक मयाध सोलापूर या बँकेच्या स्थापनेने झाली या बँकेची स्थापना द्ददनाांक 8 माचध 1818 रोजी झाली या बँकेच्या िाढीस प्रारांभी सोलापूर शहरातील रा ब जी के साठी ्ही एस जोग एम आर गोडबोले जी डी पुांडे जे इकुलकणी जी आर कुांभकोणी या मान्यिराांनी योगदान द्ददले बँकेच्या सभासदात्िास ग्रामीण भागात िाि तमळू लागला ते्हापासून शेतकऱयाांमधील ि तजल्यातील नेतृत्ि या बँकेस तमळू लागले त्यामध्ये गणपतराि साठी सहकार महषी शांकरराि मोतहते पाटील नामदेिराि जगताप याांनी या बँकेच्या उिेशास शेतकऱयाांच्या कारातगराांच्या घरापयंत पोचतिण्यास ि तजल्यामध्ये सहकारी साखर कारखान्याच्या स्थापनेस ि िाढीस क्तसहाचा क्तसहाचा िाटा उचललेला होता सामान्य शेतकरी आतण सहकारी सांस्था बळकट करण्यासाठी या बँकेची स्थापना झाली स्थापनेपासून ही बँक शेतकऱयाांच्या उन्नतीचे कायध करते गाि पातळीिर शाखा स्थापन करून शेतकऱयाांच्या दारापयंत अथधसाहाय्य घेऊन जाणारी ही महत्िपूणध बँक आहे सोलापूर तजल्यात साधारणता इस 1904 ते 1918 या कालािधीत सुमारे 100 पेढे स्थापन झाल्या होत्या या पतपेढ्ाांना मुांबई सेंट्रल बँकेकडून कजध पुरिठा होत असेल त्याकाळी तजल्हा गॅरांटी युतनयनसमाफधत कजध तमळत असे अशा युतनयन्स कुई म्हैसगाि टेंभुणी ि पापरी येथे स्थापन झाल्या होत्या. ं्क हं्ं् ह पांढरपूर तालुक्कयातील पतहली प्राथतमक शेती पतपुरिठा सहकारी सांस्था (ति.का.सेिा.सह.सांस्था) द्ददनाांक ३ नो्हेंबर १९११ रोजी भाळिणी ता. ् महाराष्ट्रातील सहकारी चळिळ; महाराष्ट्रातील सहकारी चळिळ; े्ष महाराष्ट्र हे सहकार क्षेिातील पुढारलेले राज्य असून राज्यभरात सहकारी सांस्थाांचे मोठे जाळे पसरले आहे. सहकारी पतपुरिठा सांस्था, तबगर कृषी पतपुरिठा सांस्था, तिपणन सहकारी सांस्था, ग्राहक सहकारी सांस्था, प्रद्दक्या सहकारी सांस्था, सेिा सहकारी सांस्था, मजूर सहकारी सांस्था, गृहतनमाधण सहकारी सांस्था अशा तितिध प्रकारच्या सहकारी सांस्थाांची स्थापना झाली. त्या यशस्िीपणे कायधरत असल्यामुळे महाराष्ट्रातील सामान्य नागटरक, शेतकरी या ना त्या कारणामुळे सहकारी सांस्थेच्या सांपकाधत आला. महाराष्ट्रातील सहकारी चळिळीला सुमारे शांभर िषाधहून अतधक काळाचा इततहास आहे. मुांबई तजल्हा मध्यिती सहकारी सांस्थेची स्थापना सन १९२३ मध्ये झाली. या बॅंकेचे पुढे राज्य तशखर बँकेत रूपाांतर होऊन ती सध्या ‘महाराष्ट्र राज्य सहकारी बँक तलतमटेड' म्हणून कायधरत आहे. महाराष्ट्रामध्ये सहकारी पतपुरिठ्याबाबत तिस्तरीय रचना तस्िकारलेली आहे. प्राथतमक स्तरािर प्राथतमक सहकारी पतपुरिठा सांस्था, तजल्हा स्तरािर तजल्हा मध्यिती सहकारी बँका, तर राज्य स्तरािर राज्य सहकारी बँक (तशखर बँक) कायध करते. स्िातांत्र्यानांतर सन १९५१ मध्ये प्रा.धनांजयराि गाडगीळ ि पद्मश्री डॉ. तिठ्ठलराि तिखेपाटील याांच्या मागधदशधनामुळे प्रिरा सहकारी साखर कारखाना तल., प्रिरानगर तज. अहमदनगर येथे महाराष्ट्रातील पतहला सहकारी साखर कारखाना स्थापन झाला. महाराष्ट्रातील सहकारी साखर कारखान्याांनी सहकारी चळिळीचे एक निे पिध सुरू केले. १ मे १९६० रोजी महाराष्ट्र राज्याची तनर्गमती झाली. महाराष्ट्र सहकारी सांस्था कायदा १९६० या कायद्यानुसार सहकारी सांस्थाांचे तनयमन ि तनयांिण केले जाते. सन १९६१ मध्ये महाराष्ट्रात तितिध प्रकारच्या ३१,५६५ सहकारी सांस्था होत्या. माचध २०१८ अखेर ही सांख्या १,९८,२५२ पयंत पोहोचली. सहकारी सांस्थाांच्या सांख्येत जसजशी िाढ झाली तसतशी सहकारी सांस्थाांच्या सभासद सांख्येत, भागभाांडिल, कजे, ठेिी यामध्ये िाढ झाली. सोलापरतजल्यातीलसहकारीचळिळ: सोलापूर तजल्यातील सहकारी चळिळ : सोलापूर तजल्यात सहकारी चळिळीचा तिकास मोठ्या प्रमाणात झाला आहे. या चळिळीचा या तजल्याच्या 23 डॉ. तानाजी जनाधधन फुलारी IJAAR पांढरपूर या टठकाणी स्थापन झाली. या सहकारी सांस्थेचे नाि भाळिणी ति. का. सेिा. सांस्था मयाध. भाळिणी असे असून ततचा नोंदणी क्माांक एसयुआर / पीआरआर / आरएसआर/ एसआर / २७६३ असा आहे. दुसरी ति. का. सेिा सहकारी सांस्था द्ददनाांक १५ जानेिारी १९१३ रोजी स्थापन झाली. या सहकारी सांस्थेचे नाि करकांब नां. १ ति. का. सेिा सह सांस्था मयाध, करकांब असे आहे. तर ततचा नोंदणी क्माांक एसयुआर / पीआरआर / आरएसआर / एसआर / ५३७ असा आहे. इ.स. १९११ ते इ.स. २००४ या कालखांडात एकूण १२८ ति. का. सेिा, सह. सांस्था. पांढरपूर तालुक्कयात कायधरत होत्या. पांढरपूर तालुक्कयातील दुसरी सहकारी सांस्था ि पतहली बँक द्ददनाांक १९ नो्हेंबर १९१२ रोजी पांढरपूर अबधन को-ऑप बँक तल पांढरपूर या नािाने स्थापन झाली. या सांस्थेचा नोंदणी क्माांक एसयुआर / बीएनके / (ओ) /५१० असा आहे. ' नोकरदाराांच्या पत पुरिठा सहकारी सांस्थेमध्ये पांढरपूर नगरपटरषद नां.१ कमधचारी सहकारी पत सांस्था मयाध; पांढरपूर ही सांस्था द्ददनाांक २० माचध १९१८ मध्ये स्थापन झाली. इ.स. २००४ पयंत या प्रकारच्या सहकारी सांस्थाांची सांख्या २७ इतकी होती. IJAAR IJAAR Vol.3 No.5 रोजी झाली. इ.स. २००४ पयंत या प्रकारच्या सहकारी सांस्थाांची सांख्या ११ एिढी होती. द्ददनाांक ९ जुलै १९५१ रोजी पांढरपूर तालुका सहकारी खरेदी-तिक्ी सांघ मयाध पांढरपूर या सांस्थेची स्थापना झाली. असून ततचा नोंदणी क्माांक २४०६० इतका आहे. या सांस्थेच्या माध्यमातून शेतकऱयाांच्या मालाला बाजारपेठ उपलब्ध झालेली आहे. द्ददनाांक १७ एतप्रल १९८९ रोजी श्री तिठ्ठल फळ उत्पादक सहकारी सांस्था मयाध. पांढरपूर या सांस्थेची स्थापना झाली. पांढरपूर तालुक्कयात दोन सहकारी साखर कारखाने कायधरत आहेत त्यापैकी श्री तिठ्ठल सहकारी साखर कारखाना तल. िेणुनगर गुरसाळे या कारखान्याची स्थापना ०५ माचध १९७४ रोजी झाली असून त्याचा नोंदणी क्माांक एसयुआर/पीआरजी (ओ) / ४ असा आहे तर चांद्रभागा (स.तश.ि. काळे) सहकारी साखर कारखाना भाळिणी या सहकारी साखर कारखान्याची पीआरआर/पीआरजी / (अ)/ २९ / (एस) असा आहे. ं्ं््दंं झाली. तर नागरी भागात द्ददनाांक १४ नो्हेंबर १९५६ रोजी तसद्धाथध बैकिडध क्कलास सह. गृहतनमाधण सांस्था मयाध पांढरपूर या सांस्थेची स्थापना झाली मजूर सहकारी सांस्थामध्ये द्ददनाांक २७ ऑगस्ट १९५७ रोजी पांढरपूर िडार समाज मजूर सहकारी सांस्था मयाध पांढरपूर या सांस्थेची स्थापना झाली असून ततचा नोंदणी क्माांक एसयुआर / पीआरआर /१२४ पीआरडी / (एल) ३९३० असा आहे. इ.स. २००४ पयंत अशा प्रकारच्या सहकारी सांस्थाांची सांख्या ७८ इतकी होती, िाहतूक सहकारी सांस्थाांमध्ये द्ददनाांक २९ सप्टेंबर १९९८ सुधाकरपांत पटरचारक िाहतूक सहकारी सांस्था मयाध पांढरपूर तर द्ददनाांक २२ माचध २००१ रोजी श्री चांद्रभागा कृतष उद्योग ऊस तोडणी िाहतूक सहकारी सांस्था मयाध भाळिणी या इ.स. सांस्थेची स्थापना झाली. पाणी िापर सहकारी सांस्थामध्ये द्ददनाांक ३० माचध १९७९ रोजी श्री. तिठ्ठल कालिा पाणी िापर सह. सांस्था मयाध, शांकरगाांिनळी या सांस्थेची स्थापना झाली. २००४ पयंत पांढरपूर तालुक्कयातील अशा प्रकारच्या सहकारी सांस्थाांची सांख्या ३२ इतकी होती. उपसा जलक्तसचन प्रकारच्या सांस्थामध्ये द्ददनाांक २७ सप्टेंबर १९९१ रोजी श्री गणेश पाणी पुरिठा सहकारी सांस्था मयाध करकांब या सांस्थेची स्थापना झाली. द्ददनाांक १६ जून १९९४ रोजी पांढरपूर तालुका कृतष पदिीधर सहकारी सांस्था मयाध, पांढरपूर या सांस्थेची स्थापना झाली. शेती सेिा साधन पुरिठा सहकारी सांस्थामध्ये द्ददनाांक १ फेिुिारी १९९३ रोजी ज्योततर्ललग रासायतनक खते तब-तबयाणे सह.सांस्था मयाध, अजनसोंड या सांस्थेची स्थापना झाली. सहकारी औद्योतगक िसाहती सांस्थेमध्ये द्ददनाांक २९ नो्हेंबर २००० रोजी श्री रूतक्कमणी सहकारी औद्योतगक िसाहत मयाध पांढरपूर या सांस्थेची स्थापना झाली असून ततचा नोंदणी क्माांक एसयुआर / पीआरआर / आरएसआर / १०७१ असा आहे. इ.स. १९०४ ते इ.स. २००४ पयंतच्या कालखांडात पांढरपूर तालुक्कयातील तितिध प्रकारच्या सहकारी सांस्थाांची सांख्या ७७२ एिढी होती. ्ं्् पांढरपूर तालुक्कयातील सहकार चळिळ : ं् पांढरपूर हे सोलापूर तजल्यामधील महत्त्िाचे असे तालुक्कयाचे टठकाण आहे. स्िातांत्र्योत्तर काळानांतर भारतात कृतष तिकास, औद्योतगक तिकास आतण शैक्षतणक तिकासासाठी पांचिार्गषक योजनाांच्या माध्यमातून ि सहकारी चळिळीच्या माध्यमातून स्थातनक भागाच्या तिकासास सुरूिात झाली. अत्यांत कमी पाऊस पडत असलेल्या या क्षेिामध्ये स्िातांत्र्यपूिध काळात पारांपटरक पद्धतीने शेती ि उद्योग केले जात होते. स्िातांत्र्योत्तरकाळानांतर ि सहकारी चळिळीच्या तिकासानांतर तालुक्कयातील पीक पद्धतीत मोठया प्रमाणात इतर नागरी पत पुरिठा सहकारी सांस्थामध्ये (ग्रामीण) यशिांतराि च्हाण ग्रामीण तबगर शेती सह. पत सांस्था मयाध, गुरसाळे या सांस्थेची स्थापना १६ जानेिारी १९८६ रोजी झाली. या प्रकारच्या सहकारी सांस्थाांची सांख्या इ.स. २००४ पयंत २५ इतकी होती. इतर नागरी पत पुरिठा सहकारी सांस्था (नागरी) या प्रकारच्या सहकारी सांस्थामध्ये श्री तशिप्रताप नागरी सहकारी पत सांस्था मयाध. पांढरपूर या सांस्थेची स्थापना द्ददनाांक १४ जानेिारी १९८९ 24 डॉ. तानाजी जनाधधन फुलारी ISSN – 2347-7075 ISSN – 2347-7075 ISSN – 2347-7075 IJAAR ां ध सामुदातयक शेती सांस्थाांमध्ये द्ददनाांक ७ तडसेंबर १९६४ रोजी क्तशगला तगरीजन मागासिगीय सामुदाई शेती सहकारी सांस्था मयाध, बादलकोट नेमतिाडी या सांस्थेची स्थापना झाली तर २१ ऑगस्ट १९६९ रोजी शेगाांि दुमाला मागासिगीय सामुदाई शेती सहकारी सांस्था मयाध; शेगाांि दुमाला या सांस्थेची स्थापना झाली. तेल घाण्या सहकारी सांस्थाांमध्ये द्ददनाांक १८ तडसेंबर १९५६ रोजी करकांब तेल उत्पादक सहकारी सांस्था मयाध. करकांब या सांस्थेची स्थापना झाली. ताडगुळ सहकारी सांस्थामध्ये द्ददनाांक १२ ऑक्कटोबर १९५६ रोजी पांढरपूर तिभाग तनरा ताड पदाथध उत्पादक सहकारी सांस्था मयाध; पांढरपूर या सांस्थेची स्थापना झाली. द्ददनाांक ८ ऑगस्ट १९९१ रोजी पांढरपूर गटपातळीिरील बहुउिेशीय औद्योतगक सहकारी सांस्था मयाध पांढरपूर या सांस्थेची स्थापना झाली. द्ददनाांक ७ जानेिारी १९६० रोजी मेिा तमठाई उत्पादक सहकारी सांस्था मयाध, पांढरपूर या सांस्थेची स्थापना झाली असून ततचा नोंदणी क्माांक एसयुआर / पीआरआर / पीआरडी / (आय) / ५४३० इतका आहे. द्ददनाांक ०७ माचध १९९३ रोजी श्री. तिठ्ठल खत उत्पादक सहकारी सांस्था मयाध; पांढरपूर या सांस्थेची स्थापना झाली. द्ददनाांक ५ तडसेंबर १९९० रोजी श्री, तिठ्ठल सहकारी सांस्था सूत तगरणी मयाध; पांढरपूर या सहकारी सूत तगरणीची स्थापना झाली. द्ददनाांक २९ नो्हेंबर २००० रोजी श्री रूतक्कमणी सहकारी औद्योतगक िसाहत मयाध, पांढरपूर या सांस्थेची स्थापना झाली, प्राथतमक ग्राहक भाांडार सहकारी सांस्थामध्ये पांढरपूर द्दकराणा भुसार सहकारी ग्राहक भाांडार सांस्था मयाध. पांढरपूर या सांस्थेची द्ददनाांक १ ऑक्कटोबर १९४८ रोजी स्थापन झाली असून ततचा नोंदणी क्माांक एसयुआर / पीआरआर / सीओएन / ११५५ असा आहे. इ.स. २००४ पयंत अशा प्रकारच्या सहकारी सांस्थाची सांख्या ८ इतकी होती. गृहतनमाधण सहकारी सांस्था (ग्रामीण) मध्ये द्ददनाांक १२ ऑगस्ट १९९७ रोजी श्री तिश्वकमाध गुमास्ता सह गृहतनमाधण सांख्या मयाध कासेगाांि या सांस्थेची स्थापना 1. कामत (डॉ.) गो. स. सहकार तत्त्ि, व्यिहार आतण व्यिस्थापन, समाज प्रबोधन सां पुणे, १९६४ 2) महाराष्ट्राचे तजल्हे, सोलापूर, प्रतसद्धी तिभाग महाराष्ट्र सरकारम, मुांबई १९६० 3) डॉ. रणनिरे जयश्री 2. ज्ञानदेि, पांढरपूरचे साांस्कृततक िैभि, अक्षरलेण प्रकाशन, सोलापूर, २००६ 4) बतहरट भा.प., प्र. ज्ञा. भालेराि, िारकरी सांप्रदाय उदय ि तिकास, ्हीनस प्रकाशन, पुणे, तितीयािृत्ती जुलै १९८८ 3. पारेकर दशरथ, सहकारी साखर उद्योग उत्कषध आतण िाताहत, तेजस पतब्लकेशन कोल्हापूर, प्रथमािृत्ती, जानेिारी २०१५ 25 डॉ. तानाजी जनाधधन फुलारी डॉ. तानाजी जनाधधन फुलारी IJAAR 4. देशमुख गोपाळ, सोलापूर तजल्याचा इततहास (इ.स. १६०० ते १८५०), मराठा कालखांड, रेिू प्रकाशन, पांढरपूर, प्रथमािृत्ती, फेिु २००९ 5. जोशी सी. जे. सहकाराचा तिकास तत्त्ि ि व्यिहार, फडके प्रकाशन, कोल्हापूर, १९९१ 8) च्हाण यशिांतराि, युगाांतर तनिडक भाषणे, कॉतन्टनेन्टल प्रकाशन, पुणे, १९९९ 6. डॉ. चौधरी दत्ता पांढरपूर नगरीचा इततहास ( प्राचीन ते इसिी १९९०), मुिंा पतब्लक्तशग हाऊस, कोल्हापूर, प्रथमािृत्ती फेिु. २०१५ 7. किडे सुधाकर; झुांज, सहकार तशरोमणी िसांतराि काळे याांचे चटरि, तिठाई प्रकाशन, पांढरपूर प्रथमािृत्ती, फेिुिारी २०१४ 8. Benare Gopal; Lord Vitthal and Pandharpur, Pune, 2000 9. Mathur B.S.; Co-operation in India, Sahitya Bhavwan, Agra, 1980 7. Noel Deen, History of sugar, vol-l, chapman and Hall Ltd. 100001 26 डॉ. तानाजी जनाधधन फुलारी
https://openalex.org/W4288052762	https://www.researchsquare.com/article/rs-1891190/latest.pdf	English	null	Drainage Density and Its Application in Soil Erosion Assessment in Sarbari Khad Watershed of Himachal Pradesh, India	Research Square (Research Square)	2,022	cc-by	3,322	Drainage Density and Its Application in Soil Erosion Assessment in Sarbari Khad Watershed of Himachal Pradesh, India Sunil Kumar (  sunildse@gmail.com ) Amity University https://orcid.org/0000-0002-6853-4100 Research Article Page 1/11 Abstract Precise measurement of geomorphometric parameters is significant for assessing watershed conditions, nevertheless these are very difficult to measure accurately in context of the mountain ecosystems and the extensive areas. In the present research work drainage density has been used to assess the soil erosion by employing the Geographical Information System (GIS) for the Sarbari Khad watershed located in the lesser Himalayan region of the Himachal Pradesh, India. For assessing drainage density along with remotely sensed digital data, direct survey based large scale (1:50,000) topographical map have been used. The drainage density analysis unveiled that the Sarbari Khad watershed is vulnerable to surface erosion as well as soil degradation. Further, depending on the drainage density values the watershed has been classified under low, moderate, severe and very severe erosion categories. This classification revealed that about 60 percent area of the watershed suffers from very severe erosion. In this study, the drainage density values have been used for estimating erosion status of watershed. These soil erosion differentials in the watershed can be used for watershed areas prioritization for ensuring soil and water conservation measures to make the watershed sustainable. Introduction India is a rich country in terms of the total water resource potential. India has 4 percent of world water resource (Saroha, 2017). The spatial and temporal concentration of monsoon is strongly associated with high rates of soil erosion. The highland areas like the Himalayan watersheds are highly prone to soil erosion and land degradation, especially during the monsoon season. Soil erosion in India is widespread and a serious threat to survival and well-being. The most important agent of soil erosion in India is running water. Though the process of soil erosion is natural and occurring since time immemorial but it has become a serious problem due to increased anthropogenic interferences over the period of time (Saroha, 2017). Gurmel Singh, Ram Babu, Narain and others (1990) estimated that the average annual soil loss is about 15.2 tonnes per hectare and at national level it amounts to about 5 billion tonnes annually. Sehgal and Abrol (1994) have estimated that water and wind erosion in India extends over about 162.4 million-hectare area and out of this about 91 percent is water eroded area. Thus, approximately 50 percent of the geographical area of India is prone to various degrees of soil erosion and other forms of land degradation. This problem is most intense in the Himalayan watersheds. This has impacted not only in terms of decline in per capita availability of land but also in terms of qualitative loss and reduced productivity of soils in these watersheds where primary activities are the main source of livelihood. Therefore, the need of the hour to assess the degrees of soil erosion in different parts of these watersheds. Integrated watershed management approach is required to control and prevent soil erosion as soil conservation and management is pre requisite to achieve the goal of sustainable development. India is a rich country in terms of the total water resource potential. India has 4 percent of world water resource (Saroha, 2017). The spatial and temporal concentration of monsoon is strongly associated with high rates of soil erosion. The highland areas like the Himalayan watersheds are highly prone to soil erosion and land degradation, especially during the monsoon season. Soil erosion in India is widespread and a serious threat to survival and well-being. The most important agent of soil erosion in India is running water. Introduction Though the process of soil erosion is natural and occurring since time immemorial but it has become a serious problem due to increased anthropogenic interferences over the period of time (Saroha, 2017). Gurmel Singh, Ram Babu, Narain and others (1990) estimated that the average annual soil loss is about 15.2 tonnes per hectare and at national level it amounts to about 5 billion tonnes annually. Sehgal and Abrol (1994) have estimated that water and wind erosion in India extends over about 162.4 million-hectare area and out of this about 91 percent is water eroded area. Thus, Drainage density is a quantitative measure used to understand the physical parameters of the morphology of a watershed or catchment. Horton (1945) defined drainage density as the stream length per unit area in the drainage basin or watershed. It is calculated by dividing the total length of channels by the total area of that hydrological unit or watershed. Thus, it represents average length of channels Page 2/11 per unit area of the watershed. It is an efficient element for drainage analysis and it provides a better quantitative expression to the dissection and analysis of morphology the physical landscape. It is a fundamental characteristic of physical landscape with interface with other elements of the watershed such as – structure or local geology, climate, topography, edaphic factors (soil type and permeability) and extent of vegetative cover. Drainage density has been used by scholars to assess and predict soil erosion potential and to predict flooding behavior in the hydrological units like watersheds and catchments. These predictions vary depending upon the diverse geological, topographic and climatic conditions prevailing in various watersheds. It has been observed (Nag, 1998) that, generally, areas with low relative relief, highly resistant or permeable structure and dense vegetative cover have low drainage density. On the other hand, the high drainage density has been found associated with high absolute and high relative relief, weak or impermeable structures and limited vegetative cover (Nag, 1998). High drainage density prevails during the youth stage of landscape development. According to Tucker, Catani and others (2001) drainage density reflects the intensity to which a given landscape is dissected. Berger and Entekhabi (2001) concluded that high drainage density is strong association with highly dissected terrain and torrential runoff resulting into intense floods and intense soil erosion. Introduction Therefore, it is noteworthy that drainage density is a good indicator and measure of physical and hydrological characteristics of a watershed. The torrential, with occasional cloud bursts, nature of rainfall along with seasonal concentration (about 75 percent of the annual precipitation during monsoon months) makes the Himalayan watersheds highly vulnerable and prone to soil erosion, landslides and flash floods. Anthropogenic interventions in the form of infrastructure expansion, land use/land cover changes and increasing population pressure further intensify these problems in the fragile watershed ecosystems of the Himalayas. The main objective of the present research work to assess soil erosion on the basis of drainage density analysis of the Sarbari Khad watershed of Himachal Himalaya. Study Area The Sarbari Khad Watershed of the Beas River Basin has been selected as the study area. It mainly extends between 31°54' 30” N latitudes to 32° 6' 30” N latitudes and 76° 56' E to 77° 7' 40” E longitudes. It occupies an area of 930.30 Km2.The Sarbari Khad Watershed covers western part of the Kullu district. The area is an example of natural region topographically and hydrologically bounded by Pir-Panjal Range and Beas River. The altitude varies from 1500 to 4000 meter above mean sea level that increases from south to north. The area can be broadly categorized into snow cover, barren land, pastures, forested valley.This watershed comprises of 13 villages in the Kullu and Nagar revenue blocks. The villages are - Pichhbhar, Bath, Bastori, Bhalyani, Dughilag, DunkhriGahar, Gahar, Gramang, Majhat, Mashna, Peej, Phallan and Pichhli. The watershed also includes some parts of Kullu town. Total population of the watershed is 15247 people. The main economic activity of the people here is agriculture. Tourism is also an important economic activity in the study area (Gardner et al., 2002 & Kumar et al., 2022). Page 3/11 Page 3/11 The watershed experiences warm subtropical climate in summer season and cold temperate climate in winter season. In between during monsoon season torrential rainfall occurs and average annual rainfall is about 1000 m. It is experienced during the south west monsoon (June to September) in winter season some precipitation takes place due to the western disturbances. Average annual snowfall in the region amounts to about 345 mm confined to upper reaches and during winter season only. The average monthly temperature is in the range from a minimum of 8.7°C in January to maximum of 28.3°C in the month of June. The minimum and maximum relative humidity in May and August is 63.3 %and 78.7% respectively. In the coldest months of December and January evaporation was observed minimum of 36.1 mm and 38.7 mm, respectively. Evaporation is maximum during June (165.0 mm) which is considered to be the warmest month of the year. Following method of Horton (1932) first of all the drainage density was calculated as the ratio of the total length of the stream channels of all orders to the total area of the watershed. The steps mentioned in the flow chart (Figure 2) were followed for the calculation of drainage density (Figures 3-6). Study Area Therefore, in the present research work focus is exclusively on determination of soil erosion intensities based on drainage density variations in the Sarbari Khad watershed. Study Area Following the generalization of many research studies the soil erosion assessment was performed on the basis of drainage density differentials. Area under variousdegrees of erosion was evaluated as per the methodology suggested by Bucko and Mazurova (1958) and Mikhailov (1972). Rai, Mohan and Mishra and others(2017) applied a GIS-based approach in drainage morphometric analysis in Kanhar River Basin of India and concluded that drainage density is strongly correlated with valley density, channel head catchment area, topography, soil and structural characteristics, climate, vegetation and landscape development . Likewise, Pal and Saha (2017) have concluded that a strong positive correlation prevails between drainage density and the parameters such as: the number of stream merging points in the catchment, length of overland flow, the infiltration coefficient and drainage texture. Earlier Strahler (1964) also concluded that areas of high drainage density have high runoff, coarse drainage texture and higher erosion potential. Omvir Singh (2010) applied the drainage density based methodology of Bucko and Mazurova and Mikhailov for assessing the soil erosion intensities in the Lesser Himalayan region watersheds. In view of all these studies, in the present study focus has been kept exclusively on relationship of drainage density with soil erosion. Although soil erosion is result of interplay of many complex factors such as rainfall erosivity factor, the soil erodibility factor, the length and degree of slope and land use/land cover characteristics, but keeping in view the limited variability of climatic, edaphic and vegetative factors in this small watershed, drainage density has been used as the main determinant of soil erosion. Drainage density parameter indirectly incorporates other factors also to a significant level in this scale of study. Therefore, in the present research work focus is exclusively on determination of soil erosion intensities based on drainage density variations in the Sarbari Khad watershed. In view of all these studies, in the present study focus has been kept exclusively on relationship of drainage density with soil erosion. Although soil erosion is result of interplay of many complex factors such as rainfall erosivity factor, the soil erodibility factor, the length and degree of slope and land use/land cover characteristics, but keeping in view the limited variability of climatic, edaphic and vegetative factors in this small watershed, drainage density has been used as the main determinant of soil erosion. Drainage density parameter indirectly incorporates other factors also to a significant level in this scale of study. Data Used And Methodology This work is mainly based on assessing the drainage density and its implication on soil erodibility. The boundary of watershed and drainage network was extracted through ASTER-DEM data and Survey of Page 4/11 Page 4/11 India toposheets. This data base has also been used to calculate slope and digital elevation of the area. The remote sensing data were acquired from www.earthexplorere.usgs.gov. The development of stream networks relies on rainfall, slope and geology of that area. The watershed demarcation and the hydrological analysis of streams have been carried out in QGIS 3.16 software. Stream ordering has been successfully applied with a view to characterize watershed and drainage network. The parameters were sought from the streams extracted through ASTER-DEM data employing QGIS hydrological analysis tool. Drainage density, bifurcation ratio, drainage frequency etc. has been calculated through the already developed formulae. Results And Discussion The total drainage or channels length of the watershed is 2137km and out of this 1228 km is constituted by first order streams. The average drainage density for the watershed as a whole is 0.5 km/ km². Within the watershed the drainage density varies from 1.07 to 3.30 km/km².The drainage density has been noticed in the category of high in the upper-central part of the Sarbari Khad watershed. This is strongly corrected with greater rainfall intensity in this part of the watershed. The maximum drainage density has been noticed in the southern part of the watershed and lowest drainage density was observed in northern part of the area which is mostly covered by snow. The watershed has extremely resistant or permeable sub soil material with dense vegetative cover which may have advantaged low drainage density in north-eastern part of the region. The distribution of area under different classes of erosion has been given in Table 1 for the study area. The drainage density indicates that majority area of the watershed comes in the category of severe to catastrophic rates of soil erosion. Table 1 Distribution of area under different categories of erosion Category Drainage Density Watershed Area Cumulative area Vertical assessment I < 0.1 - - No erosion II 0.1–0.5 - - Slight III 0.5-1.0 - - Moderate IV 1.0–2.0 408.4 45.1 Severe V 2.0–3.0 453.8 50.4 Very severe VI > 3.0 48.2 5.2 catastrophic Table 1 About 45.1 percent area of the total watershed area falls in the drainage density range of 1–2 km/km² and indicates severe level of soil erosion. The area withdrainage density value of 2–3 km/km² indicates very severe erosion and constitutes about 50 percent of the total area of the watershed. About 5 percent area of the watershed falls in the highest drainage density category of above 3km/km² and is associated Page 5/11 with catastrophic soil erosion potential. The severity of soil erosion in the Sabari Khad watershed is directly related to local geology, intense monsoon rains, high relative relief, land use/land cover changes, deforestation and expansion of infrastructure. This study indicates that the watershed is facing severe to catastrophic intensity of soil erosion. Splash erosion, sheet erosion, rill and gully erosion can be easily observed in different parts of the watershed during monsoon season. Results And Discussion For the sustainable development of the watershed immediate application of different soil and water conservation techniques is required on priority basis. Afforestation, check on deforestation, overgrazing and faulty agriculture practices, contour farming, check dams, terrace farming, stabilizing of slopes and land use according to land capability classes should be operationalized at the earliest to avoid further intensification of problems. Conclusion The Himalayan watersheds are geo-dynamically unstable due to high seismic sensitivity of the area, presence of faults and nappes, high absolute height and high relative relief and steep slopes. These watersheds as ecosystems are very sensitive and fragile. The orographic and seasonal monsoon rainfall over weak structures contributes in high rates of soil erosion and landslides. Geospatial technologies especially remote sensing and GIS play a significant role in making assessment related to hydrological and morphometric parameters especially of watersheds in these highland areas. These data acquisition and processing techniques help not only in the calculation of drainage density but also the assessment of the degree of soil erosion potential in different parts of the watershed. As drainage density is function of geology, topography, climate, edaphic, vegetation and anthropogenic factors it is easy to overlay these themes and do modeling in GIS environment. In Sarbari Khad watershed the drainage density is in the range of high to very high and consequently the soil erosion degree ranges mainly from severe to catastrophic. These levels of soil erosion call for watershed management approach in this study area. Watershed management implies a rational utilization of local resources especially soil, water and forest for optimum production that too with minimum hazard to natural resources or the environment. References Saroha, J (2017). Water Resource Potential, Utilization and Conservation in India, IJRAR, 4 (3):27-35. Saroha J (2017) Soil Erosion: Causes Extent and Management in India 5 (4): 1321-1330 Saroha, J (2017). Soil Erosion: Causes, Extent and Management in India, 5 (4): 1321-1330. Gurmel Singh, Ram Babu, Narian, P., Bhusan, L.S. and Abrol, I.P. (1990). Soil Erosion Rates in India, Journal of Soil and Water Conservation, 47(1):97-99. Gurmel Singh, Ram Babu, Narian, P., Bhusan, L.S. and Abrol, I.P. (1990). Soil Erosion Rates in India, Journal of Soil and Water Conservation, 47(1):97-99. Gurmel Singh, Ram Babu, Narian, P., Bhusan, L.S. and Abrol, I.P. (1990). Soil Erosion Rates in India, Journal of Soil and Water Conservation, 47(1):97-99. Sehgal, J and Abrol, I.P. (1994) Soil Degradation in India: Status and Impact, Oxford and IBH, New Delhi. Sehgal, J and Abrol, I.P. (1994) Soil Degradation in India: Status and Impact, Oxford and IBH, New Delhi. Horton, R.E. (1945). Erosional Development of Streams and their Drainage Basins: Hydrological Approach to Quantitative Morphology, Bulletin of Geological Society of America, 56: 257-370. Horton, R.E. (1945). Erosional Development of Streams and their Drainage Basins: Hydrological Approach to Quantitative Morphology, Bulletin of Geological Society of America, 56: 257-370. Page 6/11 Nag S, K. (1998). Morphometric analysis using remote sensing techniques in the Chaka sub basin, Purulia, West Bengal.Jour of Indian Society of Remote Sensing 26: 69-76. Tucker, G.E., Catani, F., Rinaldo, A and Bras, R. L. (2001). Statistical analysis of drainage density from digital terrain data. Geomorphology. 36: 187-202. Berger, K. P., and Entekhabi, D., (2001).Basin hydrologic response in relation to distributed physiographic descriptors and climate. Journal of Hydrology. 247: 169-182. Gardner, J. S. (2002). Natural Hazards Risk in the Kullu District, Himachal Pradesh, India.Geographical Review. 92(2): 282-306. Kumar, S., Singh, V., & Saroha, J. (2022). Geospatial technology for sustainable development: A strategic framework for Sarbari Khad Watershed of Himachal Pradesh. International Journal of Health Sciences, 6(S2). https://doi.org/10.53730/ijhs.v6nS2.8258. Horton, R. E. (1932). Drainage basin characteristics. Transaction of American Geophysical Union. 13: 350-361. Bucko, S., and Mazurova, V.,(1958). Gully erosion in Slovakia, in; Water erosion in Slovakia (in Slovak). SAV publishers, Bratislava. Mikhailov, T., (1972). Certain particularities des processus di erosion contemporians on Bulgarie.Acta Geographica Devercina 10: 41-50. Rai, P.K., Mohan, K., Mishra, S., Ahmad, A. & Mishra, V. N. (2017). A GIS-based approach in drainage morphometric analysis of Kanhar River Basin, India. Applied Water Science. 7:217-232. References Pal, S., and Saha, T.K. (2017). Exploring drainage/relief-scape sub-units in Atreyee River basin of India and Bangladesh, Spat.Inf.Res. 25:685-692. Strahler, A. N. (1964).Quantitative geomorphology of drainage basin and channel networks. In; V.T. Chow (ed.) Handbook of Applied Hydrology.MacGraw Hill, New York, PP-439-476. Singh, O. (2010).Spatial variation of drainage density and its application in erosion assessment in the lesser Himalayan region using Geographical, Information System. Indian Journal of Geomorphology. 15(1+2): 29-38. Declarations Competing interests: The authors declare no competing interests. Page 7/11 Figures Figure 1 Study Area Location Page 8/11 Page 8/11 Figure 2 Flow Chart of Hydrological process of Sarbari Khad Watershed Figure 2 Figure 2 Flow Chart of Hydrological process of Sarbari Khad Watershed Flow Chart of Hydrological process of Sarbari Khad Watershe Page 9/11 Figure 3 Figure 3 Page 9/11 3D view of the Sarbari Khad Watershed Figure 5 Stream order of the drainage Stream order of the drainage Figure 4 Digital Elevation Model of the Study Area Page 10/11 Page 10/11 Page 11/11 Figure 5 Stream order of the drainage Figure 6 Stream order wise stream length (length in kilometers) Figure 5 Figure 6 Stream order wise stream length (length in kilometers) Page 11/11
https://openalex.org/W4221144382	https://iris.unitn.it/bitstream/11572/378968/1/sayantan.pdf	English	null	Continual Learning from Demonstration of Robotics Skills	arXiv (Cornell University)	2,022	cc-by	22,543	a b s t r a c t Article history: Received 4 November 2022 Received in revised form 17 March 2023 Accepted 7 April 2023 Available online 22 April 2023 Keywords: Learning from demonstration Continual learning Hypernetwork Neural ordinary differential equation solver Article history: Received 4 November 2022 Received in revised form 17 March 2023 Accepted 7 April 2023 Available online 22 April 2023 Keywords: Learning from demonstration Continual learning Hypernetwork Neural ordinary differential equation solver Methods for teaching motion skills to robots focus on training for a single skill at a time. Robots capable of learning from demonstration can considerably benefit from the added ability to learn new movement skills without forgetting what was learned in the past. To this end, we propose an approach for continual learning from demonstration using hypernetworks and neural ordinary differential equation solvers. We empirically demonstrate the effectiveness of this approach in remembering long sequences of trajectory learning tasks without the need to store any data from past demonstrations. Our results show that hypernetworks outperform other state-of-the-art continual learning approaches for learning from demonstration. In our experiments, we use the popular LASA benchmark, and two new datasets of kinesthetic demonstrations collected with a real robot that we introduce in this paper called the HelloWorld and RoboTasks datasets. We evaluate our approach on a physical robot and demonstrate its effectiveness in learning real-world robotic tasks involving changing positions as well as orientations. We report both trajectory error metrics and continual learning metrics, and we propose two new continual learning metrics. Our code, along with the newly collected datasets, is available at https://github.com/sayantanauddy/clfd. © 2023 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/). teaching, where a human user teaches new skills by guiding the robot. A trend that recently received increased attention in LfD is encoding observations into a vector field [7–12]. These methods, like many other works in the field, focus on learning a single motion. To naively learn multiple motion skills, one would need to train a different model for each skill, or jointly train on the demonstrations for all skills. Sayantan Auddy a,∗, Jakob Hollenstein a, Matteo Saveriano b, Antonio Rodríguez-Sánchez a, Justus Piater a,c Sayantan Auddy a,∗, Jakob Hollenstein a, Matteo Saveriano b, Antonio Rodríguez-Sánchez Justus Piater a,c a Department of Computer Science, University of Innsbruck, Austria b Department of Industrial Engineering, University of Trento, Italy c Digital Science Center (DiSC), University of Innsbruck, Austria ∗Corresponding author. E-mail address: sayantan.auddy@uibk.ac.at (S. Auddy). Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 p // g/ /j 0921-8890/© 2023 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/). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0 https://doi.org/10.1016/j.robot.2023.104427 0921-8890/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://cre https://doi.org/10.1016/j.robot.2023.104427 0921-8890/© 2023 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 Robots deployed in unstructured real-world environments will face new tasks and challenges over time, requiring capabilities that cannot be fully anticipated at the beginning. These robots need to learn continually, which implies that they should be able to acquire new capabilities without forgetting the previously learned ones. Furthermore, a continual learning robot should be able to do this without the need to store and retrain on the training data of all the previously learned skills. In this paper, we propose an approach to continual learning from demonstration in which a robot learns individual motions sequentially without retraining on past demonstrations. The mo- tion demonstrations are recorded in the robot’s task space and consist of either trajectories of the end-effector position, or tra- jectories of the end-effector position and orientation (the robot is free to rotate in all 3 rotation axes). The skills learned from demonstrations of different tasks are incorporated into a common unified model, after which our robot can reproduce all the trajectories it has learned in the past (Fig. 1). To the best of our knowledge, this is the first work on continual trajectory learning from demonstrations. Continual learning can be effective in expanding a robot’s repertoire of skills and in increasing the ease of use for non- expert human users. However, apart from a few approaches for robotics [1,2], the current continual learning research mostly focuses on vision-based tasks such as incrementally learning clas- sification of new image categories [3–5]. Continually acquiring perceptive skills is important for a robot that interacts with its environment, but equally important is the ability to incrementally learn new movement skills. Learn- ing from demonstration (LfD) [6] is a popular and tangible way to impart motion skills to robots, for instance via kinesthetic More specifically, we show that a Hypernetwork [5,13], that generates the parameters of a Neural Ordinary Differential Equa- tion (NODE) solver [14], remembers a long sequence of motion skills equally well as when learning each task with a separate NODE. The number of parameters of the hypernetwork grows S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 1. A robot, continually trained using learning from demonstration to perform real-world tasks involving changing positions and orientations of the end-effector, can reproduce all the tasks that it has learned in the past with negligible parameter growth and without retraining on past tasks. • We release 2 new datasets suitable for continual LfD: a dataset containing 7 tasks of planar motions, and a dataset of 4 tasks of motions involving both position and orien- tation. Both datasets are collected with a real robot using kinesthetic teaching. • We release 2 new datasets suitable for continual LfD: a dataset containing 7 tasks of planar motions, and a dataset of 4 tasks of motions involving both position and orien- tation. Both datasets are collected with a real robot using kinesthetic teaching. by a negligible amount for each new task, making it suitable for potential deployment on resource-constrained, non-networked robotic platforms. We also demonstrate the effectiveness of chun- ked hypernetworks [5] which can be even smaller in size than the NODEs they generate. Our results show how using a time index as an additional, direct input to a NODE increases its prediction accuracy for complex trajectories. We first evaluate our approach on the popular LASA trajectory learning benchmark [8], where our model learns a long sequence of 26 tasks without forgetting. Due to the lack of datasets containing real robot data useful for contin- ual learning from demonstration, we introduce two new datasets, named HelloWorld and RoboTasks. The HelloWorld dataset con- sists of two-dimensional trajectories of the robot’s position col- lected with a Franka Emika Panda robot. RoboTasks is a dataset of real-world robot tasks containing trajectories of the robot’s position as well as its orientation in 3D space (Fig. 1). These two datasets serve as additional benchmarks to evaluate our approach, both quantitatively and qualitatively on a real robot. For all three datasets, we extensively compare our hypernetwork- based continual LfD approach against methods from all major continual learning families (dynamic architecture, replay, and reg- ularization [15]) and report multiple metrics, both for accuracy of the reproduced motion skills, as well as the continual learning performance. Finally, we introduce two new easily-computable metrics that offer additional insights into the continual learning performance: Time Efficiency (TE) and Final Model Size (FS). TE captures training time changes while learning multiple tasks, and FS measures the relative model size after learning all tasks. 1. Introduction The order of tasks during sequential training is 0: box opening, 1: bottle shelving, 2: plate stacking, 3: pouring. Video of the robot performing these tasks is available at https://youtu.be/0gdIImIBnXc. Further details can be found in Section 5.3.3. Fig. 1. A robot, continually trained using learning from demonstration to perform real-world tasks involving changing positions and orientations of the end-effector, can reproduce all the tasks that it has learned in the past with negligible parameter growth and without retraining on past tasks. The order of tasks during sequential training is 0: box opening, 1: bottle shelving, 2: plate stacking, 3: pouring. Video of the robot performing these tasks is available at https://youtu.be/0gdIImIBnXc. Further details can be found in Section 5.3.3. 3.1.1. Neural ODE solver Consider a set of N observed trajectories D = {y (0) 0:T−1, . . . , y (N−1) 0:T−1}, where each trajectory y (i) 0:T−1 is a sequence of T obser- vations y (i) t ∈Rd. Each observation y (i) t is a perturbation of an unknown true state x (i) t generated by an unknown underlying vector field ftrue [33]: xt = x0 + ∫t 0 ftrue(xτ ) dτ, (1) (1) Continual learning has also been successfully applied to robotics, though the number of such studies is relatively few. Gao et al. [1] present an approach for continual imitation learning that relies on deep generative replay [3] and action-conditioned video prediction to generate state and action trajectories of past tasks. This pseudo-data is interleaved with demonstrations of the current task to train a policy network that controls the robot’s actions. The authors note that the generation of high-quality video frames can be problematic for a long sequence of tasks. where x0 is the true starting state of the trajectory. The goal of a Neural Ordinary Differential Equation (NODE) solver [14] is to learn a neural network fθ parameterized by θ that approximates the true underlying dynamics of the observed system such that fθ ≈ftrue. As we do not have access to ftrue but only to the noisy observed trajectories, we compute the loss L based on the difference of the forward simulated states of the NODE ˆyt and the observations yt: L = 1 2 ∑ t ∥yt −ˆyt ∥2 2 (2) where ˆyt = ˆy0 + ∫t 0 fθ(ˆyτ ) dτ L = 1 2 ∑ t ∥yt −ˆyt ∥2 2 (2) Xie and Finn [32] propose a method for lifelong robotic re- inforcement learning that seeks to improve the forward transfer performance while learning a new current task by pre-training on the entire experience collected from all previous tasks. The problem of catastrophic forgetting is not considered. (2) where ˆyt = ˆy0 + ∫t 0 fθ(ˆyτ ) dτ Our continual LfD approach shares some similarities with the work of Huang et al. [2] who continually learn a dynamics model for reinforcement learning. In their work, a task-conditioned hy- pernetwork generates the parameters of the dynamics model for reinforcement learning tasks such as opening doors or pushing blocks. 2. Related work Robot learning from demonstration (LfD) is a means for humans to teach motion skills to robots without explicitly programming them [6], which allows even users without expertise in robotics to train robots. It is also known by other names such as programming by demonstration or imitation learning [6,16]. The demonstrations required for training robots via LfD can be obtained by different means, some of which are: (i) using a motion-tracking system to record human motions, (ii) using teleoperation to operate the robot while recording the robot’s state, or (iii) using kinesthetic teaching where a human user physically guides the robot to perform a motion task [6,17–19]. Once the demonstrations are available, there exist several different algorithmic approaches for learning from this data [6]. Supervised learning has been used to learn from either a single demonstration [20] or a collection of demonstrations [21]. LfD has also been used in conjunction with reinforcement learning (RL) [22] where RL is used to refine the skills learned with LfD. Another approach is to learn a cost function from demonstrations and then to train a model predic- tive controller to reproduce the skills through inverse RL [23] or through constrained optimization [24]. In addition to demonstra- tions which show the robot the motion it has to perform, negative demonstrations have also been shown to be advantageous [25]. We refer the reader to [6,17,18] for a comprehensive overview of LfD. FS measures the relative model size after learning all tasks. In summary, our primary contributions are: In summary, our primary contributions are: In summary, our primary contributions are: • We propose an approach to learning from demonstration with hypernetworks and NODEs for continually learning new tasks without reusing training data of previous tasks. We show that this approach can be used for learning robot tasks in the real world. In this paper, we focus on a subfield of LfD: trajectory-based learning methods that use a supervised approach for acquiring motion skills. These methods can be broadly categorized into two 2 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 groups [19]: some methods use generative models to fit a distri- bution from the training data [7–9], while other methods exploit function approximators like neural networks to fit the training data [10,11]. 2. Related work In both groups, training data can be used to learn a static mapping (time input →desired position) or a dynamic mapping (input position →desired velocity). A dynamic mapping generates vector fields where input quantities are transformed into their time derivatives, and different strategies have been proposed to ensure convergence of the vector field to a given target [9,10,26,27]. Among others, the Imitation Flow (iFlow) approach of Urain et al. [9] leverages the representational power of neural networks and normalizing flows to learn vector fields from demonstrations. Another neural network based approach for learning vector fields is Neural Ordinary Differential Equation solvers (NODEs) [14]. Although NODEs have not been exploited for LfD, in our experiments we found that their empirical perfor- mance is comparable to that of iFlow. Crucially, the time needed to train a NODE to convergence is significantly less than that required for an iFlow model with an equivalent parameter size. a more challenging problem than typical supervised regression or classification scenarios where the amount of training data is usually much greater. To continually learn such vector fields of different kinds of motion demonstrations, we need to tackle the challenges of LfD as well as the problem of catastrophic forgetting associated with neural networks. As far as we know, ours is the first work that demonstrates that continual LfD is feasible for sequences of real-world robot tasks. 3. Background In this paper, we utilize Neural Ordinary Differential Equation (NODE) solvers [14] to learn from demonstrations and different state-of-the-art continual learning approaches [4,5,30] to allevi- ate catastrophic forgetting [15] when the NODEs learn a sequence of multiple tasks. 3.1. Learning from demonstration In the simplest case, task space demonstrations provided to a robot can consist of trajectories of only the end-effector positions. Since these position trajectories reside in Euclidean space, we can directly use a NODE to learn them using maximum likelihood estimation. However, the more general situation is when demon- strations contain both the position and orientation (which are commonly expressed in terms of unit quaternions). Trajectories of unit quaternions do not reside in Euclidean space and require additional processing, as discussed in Section 3.1.2. Trajectory-based LfD is a mature research field, but most methods assume that different tasks are encoded in different representations, i.e., one has to fit a new model for each task the robot has to execute [6]. In this paper, we take the continual learning perspective on learning by demonstration and propose an approach capable of continuously learning new tasks without needing to store and use the training data from past tasks. Continual Learning approaches in the current literature mostly address the problem of continual image classification. Popular strategies include growing the network architecture [28], replay- ing data from past tasks, or regularizing trainable parameters to avoid catastrophic forgetting [15]. Replay-based methods cache samples of real data from past tasks [29], or use generative models to create pseudo-samples of past data [3], which are interleaved with the current task’s data during training. Regu- larization-based methods [4,30] add a regularization term to the learning objective to minimize changes to parameters important for solving previous tasks. Refer to [15,31] for in-depth surveys on continual learning methods. 3.1.1. Neural ODE solver 3.1.1. Neural ODE solver In contrast, we use hypernetworks for generating param- eters for a trajectory learning NODE in a setup for learning from demonstration. We follow a supervised approach and do not need to rely on robot simulators. Compared to [2], we evaluate on much longer sequences of tasks and also investigate the effective- ness of chunked hypernetworks [5]. In addition, we qualitatively evaluate our approach on a physical robot. When the set of observed trajectories D consists of only positions, then the demonstrations can be learned directly with the help of Eq. (2). When the set of observed trajectories D consists of only positions, then the demonstrations can be learned directly with the help of Eq. (2). 3.2.3. Regularization MAS differs from SI in the way Ωm k is computed: the importance of a trainable parameter depends on the gradient of the squared L2 norm of the network’s output, i.e., ˆq (i) t = q (i) T−1 ∗Exp(ˆr (i) t ) (5) (5) where Exp(·) : R3 ↦→S3 is the exponential map [36], defined as Exp(r) = ⎧ ⎨ ⎩ [ cos(∥r∥), sin(∥r∥) rT ∥r∥ ]T if ∥r∥> 0, [1, 0, 0, 0]T otherwise. (6) Ωm k = 1 N N ∑ n=1 \|\|gk(xn)\|\| = 1 N N ∑ n=1  ∂L2 2 (fθ (xn)) ∂θk . (9) (6) (9) Once we obtain ˆDq, it can be directly compared against the ground truth demonstrations Dq. However, we assume that the input domain of Log(·) is restricted to S3 except for [1, 0, 0, 0]T and the input domain of Exp(ζ) is constrained to satisfy ∥ζ∥< π [34]. The above summation is performed over N input data points. p p p Hypernetworks: A hypernetwork [5,13] is a meta-model that generates the parameters of a target network that solves the task we are interested in. It uses a trainable task embedding vector as an input to generate the network parameters for a task. Though the parameters h of the hypernetwork fh are regularized, the parameters θm+1 produced by a hypernetwork for the (m + 1)th task can be arbitrarily far away in parameter space from the pa- rameters θm produced for the previous mth task. Intuitively, this gives a hypernetwork more freedom to find good solutions for both the mth and (m+1)th tasks than other regularization-based approaches [4,30]. 3.2.1. Dynamic architectures i k y Progressive Networks: One of the early continual learning ap- proaches, proposed by Rusu et al. [28], involves the addition of a new network for each new task, while reusing feature- mapping knowledge from previous tasks through lateral layer- wise connections from the networks of previous tasks. Although this approach eliminates catastrophic forgetting by design, it does not scale well to a large number of tasks due to the unconstrained growth of parameters and it also has the problem of a gradual slowing down of inference due to the increasing number of lateral connections. Lm = Lm(θm, ym) where θm = fh(em, h) (10) (10) Here em is the task embedding vector and ym is the data for the mth task. Next, ∆h is considered to be fixed and the actual change for the hypernetwork parameters h is learned [5] by minimizing the regularized loss ˜Lm with respect to θm = fh(em, h): Here em is the task embedding vector and ym is the data for the mth task. Next, ∆h is considered to be fixed and the actual change for the hypernetwork parameters h is learned [5] by minimizing the regularized loss ˜Lm with respect to θm = fh(em, h): ˜Lm = Lm( θm, ym) + β m −1 m−1 ∑ l=0 ⏐⏐⏐⏐fh(el, h∗) −fh(el, h + ∆h) ⏐⏐⏐⏐2 (11) (11) 3.2.3. Regularization Synaptic Intelligence: Synaptic Intelligence (SI) [30] is a regular- ization-based continual learning approach. Each neural network parameter is assigned an importance measure based on its con- tribution to the change in the loss. The loss for the mth task is defined as: ˜Lm = Lm + c ∑ k Ωm k ( θ∗ k −θk )2 , (7) (7) r (i) t = Log(¯q (i) t ∗q (i) T−1) r (i) t = Log(¯q (i) t ∗q (i) T−1) (3) (3) where ¯q indicates the conjugate of the quaternion q and the ∗ operator denotes quaternion multiplication, and Log(·) : S3 ↦→R3 is the logarithmic map [36], defined as where c is the regularization constant which trades off between learning a new task and remembering previously learned tasks, θ∗ k denotes the value of the kth parameter before starting to learn the mth task, and θk is the current value of the kth parameter. The per-parameter regularization strength Ωm k [30] is given by Log(q) = { arccos(v) u ∥u∥if ∥u∥> 0, [0, 0, 0]T otherwise. (4) (4) Ωm k = ∑ l<m ωl k (∆l k)2 + ξ , (8) (8) Here, q (i) T−1 is the final quaternion in the sequence q (i) 0:T−1 and ¯q (i) t is the conjugate of q (i) t . By applying Eq. (3) on the trajectories in Dq, we obtain a set of Euclidean trajectories Dr, which can then be learned using Eq. (2). After learning is complete, the trajectories ˆDr = {ˆr (0) 0:T−1, . . . , ˆr (N−1) 0:T−1} predicted by the NODE can be converted into quaternion trajectories ˆDq using where ωl k is the importance of parameter k for learning task l, ∆l k is the change in parameter k, and ξ is a damping constant. Memory Aware Synapses: Memory Aware Synapses (MAS) [4] is also a regularization-based continual learning approach. The loss for the mth task for MAS has the same form as SI (Eq. (7)). 3.2. Continual learning Existing continual learning approaches can be broadly cate- gorized into a few groups, the most prominent among which are methods based on dynamic architectures, methods based on replaying (or pseudo-rehearsing on) training data of past tasks, and methods based on regularization [15]. In this paper, we consider continual learning methods (CL methods) from all of these categories. A two-step optimization process is used for training a hyper- network [5]. First, a candidate change ∆h for the hypernetwork parameters is computed which minimizes the task-specific loss Lm for the (current) mth task with respect to θm: 3.1.2. Learning orientation trajectories 3.1.2. Learning orientation trajectories When demonstration trajectories of the robot’s orientation are expressed as unit quaternions, it is not possible to directly utilize Eq. (2) due to the unit norm constraint [34]. Following Ude et al. [35] and Huang et al. [34], we project quaternions into the tangent space which we can consider a local Euclidean space. These transformed trajectories can then be learned with Eq. (2). For inference, the Euclidean trajectories predicted by the NODE are transformed back into quaternions and then passed to the robot. LfD involves learning the entire vector field of a robot’s motion from only a handful of demonstration trajectories. This makes it 3 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Consider a set of N observed orientation trajectories Dq = {q (0) 0:T−1, . . . , q (N−1) 0:T−1}, where the ith trajectory q (i) 0:T−1 is a sequence of T quaternions q (i) t = [ v (i) t u (i) t ] , where q (i) t ∈S3, v (i) t ∈R and u (i) t ∈R3. We convert q (i) t into a rotation vector r (i) t ∈R3 using Consider a set of N observed orientation trajectories Dq = {q (0) 0:T−1, . . . , q (N−1) 0:T−1}, where the ith trajectory q (i) 0:T−1 is a sequence 3.2.3. Regularization 3.2.2. Replay Here h∗denotes the hypernetwork parameters before learning the mth task, and β is a hyperparameter that controls the regu- larization strength. To calculate the second part of Eq. (11), the stored task embedding vectors {e0, e1, . . . , el, . . . , em−1} for all tasks before the mth task are used. In each learning step, the current task embedding vector em is also updated to minimize the task-specific loss Lm [5]. Note that the parameters of the target network θm for the mth task are simply the hypernetwork outputs and are not directly trainable. Some continual learning approaches are based on the idea of replaying the training data (or some part of the data, or a compressed version of the data) of previous tasks while learning a new task [29]. While learning task k, the most naive way to do this is to combine the data from older tasks {D0, . . . , Dk−1} with the data of the current task Dk to get a combined dataset {D0, . . . , Dk}. Thus, the network has access to all the training data from every task. However, this approach is not scalable due to the linear storage requirements with the number of tasks. 4 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 2. Overview of a continual learning system. The task learner performs the task we are interested in (purple dashed box) and the continual learning mechanism (green dashed box) mitigates the effect of catastrophic forgetting. The continual learning (CL) mechanism can be one of many alternatives: dynamic architecture-based, replay-based, or regularization-based. For learning from demonstration (LfD), the task learner can be any trajectory learning approach, but we use NODEs for LfD (reasons are detailed in Section 4.1). Methods depicted with gray text are not considered for continual LfD in this paper. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) There exist many approaches for LfD [9,26], and in principle, any such method can be used as the task learner (in Fig. 2). In this paper we focus on neural network-based continual learning. 4. Methods As depicted in Fig. 2, any continual learning system in general consists of 2 components: (i) a task learner, and (ii) a continual learning mechanism. The task learner performs the task we are interested in; e.g. for continual learning of image classification, the task learner can be a CNN classifier. In this paper the task that we are interested in is learning from demonstration (LfD), and the task learner in our case is a NODE as described in Section 3.1.1. The continual learning mechanism mitigates the effects of catas- trophic forgetting as the task learner is trained on a sequence of tasks. This mechanism can act directly on the parameters of the task learner and can also contain other sub-components to achieve this goal. In this section, we describe in detail our proposed approach for continual learning from demonstration, as well as the other models used in our experiments. We also discuss the design choices made in this paper regarding the task learner and the continual learning mechanism. 4.2.2. Optimizing only for the current task 4.2.2. Optimizing only for the current task Finetuning (FT): A single NODE is sequentially finetuned on M tasks. To tell the NODE which task it should reproduce, i.e. to make it task-conditioned, we use an additional input in the form of a trainable vector known as a task embedding vector. This is similar to the approach followed for hypernetworks. After the mth task is learned, the trained task embedding vector em for that task is saved. To reproduce the trajectory for the mth task, we pick the corresponding task embedding vector and use it as the additional network input. The NODE parameters are finetuned to minimize the loss on the current task without any mechanism for avoiding catastrophic forgetting. In this setting (Fig. 3(b)), we would expect the NODE to only remember the latest task and so this acts as a lower-performance baseline. 3.2.2. Replay Hence methods such as Stable Estimator of Dynamical Systems (SEDS) [26] are not considered as the constrained parameters (covariance matrices in case of SEDS) required by such methods cannot be produced by a neural network in a straightforward way. Alternative options for the task learner can be deep learn- ing methods such as Imitation Flow (iFlow) [9] or NODEs [14]. We performed an experimental comparison between NODEs and iFLow (details in Appendix A.5) and found that the empirical performance of NODEs is better than iFlow and their training converges much quicker. Consequently, we use NODEs as the task learner in all the models used in our experiments. Fig. 2. Overview of a continual learning system. The task learner performs the task we are interested in (purple dashed box) and the continual learning mechanism (green dashed box) mitigates the effect of catastrophic forgetting. The continual learning (CL) mechanism can be one of many alternatives: dynamic architecture-based, replay-based, or regularization-based. For learning from demonstration (LfD), the task learner can be any trajectory learning approach, but we use NODEs for LfD (reasons are detailed in Section 4.1). Methods depicted with gray text are not considered for continual LfD in this paper. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 4.2.1. Dynamic architecture i l k ( y Single NODE per task (SG): A simple way to learn M tasks is to use a dedicated, newly-initialized NODE to learn a task and to freeze it afterwards. At the end we get M NODEs, from which we can pick one at prediction time to reproduce the desired trajectory (Fig. 3(a)). In this setting, which acts as an upper-performance baseline, catastrophic forgetting is eliminated because the parameters of a NODE trained on a task are not affected when a new NODE is trained on the next task. However, this also means that we end up with M times the number of parameters of a single NODE. This setup acts as a simplified version of the Progressive Networks approach [28], but without the layer-wise lateral connections to the networks of previous tasks. This keeps the inference time constant as new tasks are added. 4.2. Baselines and continual learning models We enable continual learning for NODEs by adapting sev- eral state-of-the-art continual learning approaches for LfD. We consider all major families of continual learning (dynamic ar- chitecture, replay, and regularization) and describe the details of each continual learning method (CL method) used in our experiments (Fig. 3). Chunked hypernetworks [5] produce the parameters of the target network in segments known as chunks. A regular hy- pernetwork has a very high-dimensional output, but a chunked hypernetwork’s output is of a much smaller dimension, leading to a lower hypernetwork parameter size. A chunked hypernetwork requires additional inputs in the form of trainable chunk embed- ding vectors. While each task has its dedicated task embedding vector, chunk embedding vectors are shared across tasks and are regularized in the same way as the hypernetwork parameters. The task embedding vector and all the chunk embedding vec- tors are combined in a batch and fed into the hypernetwork to produce the target network parameters for a task in one forward pass [5]. 4.1. Learning from demonstration 4.2.3. Replaying training data Replay (REP): For most continual learning scenarios replaying training data from previous tasks is a trivial exercise of combining the data from disparate tasks and then training a model with the composite dataset. However, for learning from demonstration, such a simple approach does not suffice as our NODE models need to be task conditioned. If they are trained using a jumbled up combination of demonstration trajectories from different tasks, they will be unable to learn any of the tasks. To overcome this issue we propose to use a task selector, as shown in Fig. 3(c). REP stores the dataset of each task it learns, and similar to FT, also maintains separate task embeddings for each task. In Along with the basic NODE fθ(ˆyt) described in Section 3.1.1, we use another variant where the NODE neural network fθ(ˆyt, t) is a function of both the state ˆyt and the normalized time t of the state (t scales linearly from t = 0.0 for the starting state to t = N/f for the Nth state of the trajectory, where f is the recording frequency). This explicit time input results in the NODE learning a time-evolving vector field. We show empirically that this improves the accuracy of predicted trajectories, especially for those containing loops. We refer to this time-dependent NODE as NODET, and to the time-independent one as NODEI. 5 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 3. Continual learning models used in our experiments are depicted in (a)–(f). Non-regularized trainable parameters that are initialized before and saved after each task are shown with . Non-regularized parameters that are initialized only once before learning the first task are shown with . Regularized trainable parameters that are protected from catastrophic forgetting while learning a sequence of tasks are shown with . Other inputs and outputs are not trainable (shown with ). Given a start state and time steps, a NODE generates the state trajectory for the time steps. (a) Single NODE per task (SG): A separate NODE learns only a single task. This forms the upper-performance baseline. (b) Finetuning (FT): The NODE from the previous task is optimized for the current task by finetuning. (c) Replay (REP): The training data from previous tasks is combined with the data of the current task. 4.1. Learning from demonstration (d) Synaptic Intelligence (SI) and Memory Aware Synapses (MAS): NODE parameters are regularized to prevent changes to parameters important for solving previous tasks. (e) Hypernetworks (HN): A hypernetwork produces all the NODE parameters using a task embedding vector. (f) Chunked Hypernetworks (CHN): Chunk embedding vectors together with a task embedding vector are used to produce the NODE parameters in segments called chunks. HN and CHN (highlighted in turquoise) are our proposed solutions for continual learning from demonstration. Fig. 3. Continual learning models used in our experiments are depicted in (a)–(f). Non-regularized trainable parameters that are initialized before and saved after each task are shown with . Non-regularized parameters that are initialized only once before learning the first task are shown with . Regularized trainable parameters that are protected from catastrophic forgetting while learning a sequence of tasks are shown with . Other inputs and outputs are not trainable (shown with ). Given a start state and time steps, a NODE generates the state trajectory for the time steps. (a) Single NODE per task (SG): A separate NODE learns only a single task. This forms the upper-performance baseline. (b) Finetuning (FT): The NODE from the previous task is optimized for the current task by finetuning. (c) Replay (REP): The training data from previous tasks is combined with the data of the current task. (d) Synaptic Intelligence (SI) and Memory Aware Synapses (MAS): NODE parameters are regularized to prevent changes to parameters important for solving previous tasks. (e) Hypernetworks (HN): A hypernetwork produces all the NODE parameters using a task embedding vector. (f) Chunked Hypernetworks (CHN): Chunk embedding vectors together with a task embedding vector are used to produce the NODE parameters in segments called chunks. HN and CHN (highlighted in turquoise) are our proposed solutions for continual learning from demonstration. where θm = fh(em, h), ˆym t = ˆym 0 + ∫t 0 fθm(ˆym τ ) dτ each training iteration during learning task m, the task selector randomly selects 1 out of the m previous tasks, and then passes the selected dataset and task embedding to the NODE, which optimizes its parameters for the selected task in that training iteration. When the REP model is trained for a moderately high number of iterations, each past task is sampled approximately uniformly and the NODE is able to learn continually. 4.1. Learning from demonstration For learning new tasks, instead of scaling the number of iterations linearly with the number of tasks, we use a fixed number of training iterations, irrespective of the task being learned so that the run time does not explode while learning a long sequence of tasks. As in Eqs. (2) and (10), ym t and ˆym t denote the ground truth observation and the prediction for the tth timestep of task m respectively, and θm denotes the NODE parameters for task m generated by the hypernetwork fh (with parameters h) using the embedding vector em as input. We use Eq. (11) for training the hypernetwork in the second optimization step. The structure of HN is shown in Fig. 3(e). As in Eqs. (2) and (10), ym t and ˆym t denote the ground truth observation and the prediction for the tth timestep of task m respectively, and θm denotes the NODE parameters for task m generated by the hypernetwork fh (with parameters h) using the embedding vector em as input. We use Eq. (11) for training the hypernetwork in the second optimization step. The structure of HN is shown in Fig. 3(e). Chunked Hypernetworks (CHN): We use a chunked hypernet- work to generate the parameters of a NODE (Fig. 3(f)). For this, Eqs. (12) and (11) are employed as the loss functions in the 2-step optimization process. 4.2.4. Regularization i lli Hollenstein, M. Saveriano et al. S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 the contents of the cup are emptied into a container. Thus, the RoboTasks dataset DRobot = {D0:3} consists of 4 tasks, where each task contains 9 demonstrations, each of 1000 steps. In each step we record the position p ∈R3 and the orientation as a unit quaternion q ∈S3. Compared to DLASA and DHW, DRobot contains a lot more variability in the demonstrations and presents a more difficult learning challenge. 5.1.1. LASA LASA [8] is a widely-used benchmark [9,27,37,38] for eval- uating motion generation algorithms. It contains 30 patterns of handwritten motions, each with 7 similar demonstrations (see Fig. 7 for examples). We refer to each pattern Dm as a task. Of the 30 tasks, we use the first 26 tasks: DLASA = {D0:25}. We omit the last 4 tasks, each of which contains 2 or 3 dissimilar patterns merged together. Each demonstration of a task is a sequence of 1000 2-D points. We arrange the 26 tasks alphabetically in the order of their names and train sequentially models of SG, FT, REP, SI, MAS, CHN and HN on each task. During training only REP has access to the data of older tasks while the other methods are trained with the data of only the current task. We evaluate on the LASA dataset because of its wide adoption by the LfD community [9,27,37,38], and because it contains a large number of diverse tasks which can be used to gauge the continual learning ability of our models. eq(q1, q2) = 2Log(q1 ∗¯q2) ∈R3 (13) (13) where q1 and q2 are quaternions. Given a ground truth trajec- tory q0:T−1 and a predicted trajectory ˆq0:T−1 (both containing T quaternions), we compute the error between the two trajectories using Eq = 1 3T T−1 ∑ t=0 ∥eq(qt, ˆqt)∥1 (14) (14) 5.2.2. Continual learning metrics 5.2.2. Continual learning metrics We report Accuracy (ACC), Remembering (REM), Model Size Efficiency (MS) and Sample Storage Size Efficiency (SSS) [42]. ACC is a measure of the average accuracy for the current and past tasks. REM measures how well past tasks are remembered. MS measures how much the size of a model grows compared to its size after learning the first task. SSS measures how the storage requirements of a model grows due to the storage of training data from older tasks. 5. Experiments the contents of the cup are emptied into a container. Thus, the RoboTasks dataset DRobot = {D0:3} consists of 4 tasks, where each task contains 9 demonstrations, each of 1000 steps. In each step we record the position p ∈R3 and the orientation as a unit quaternion q ∈S3. Compared to DLASA and DHW, DRobot contains a lot more variability in the demonstrations and presents a more difficult learning challenge. We first describe the datasets used in our experiments and the metrics we report. This is followed by the details of our experimental results. 5.2.1. Trajectory metrics To evaluate the end-effector position trajectories, we report the following widely used metrics: Dynamic Time Warping error (DTW) [9,41], Swept Area error [26], and Frechet distance [9], which measure how close the predicted trajectories are to the ground-truth demonstrations. For orientation, we report the com- monly used quaternion error [35,36] defined as 5.1.2. HelloWorld We further evaluate our approach on a dataset of demonstra- tions we collected using the Franka Emika Panda robot. The x and y coordinates of the robot’s end-effector were recorded while a human user guided it kinesthetically to write the 7 lower-case letters h,e,l,o,w,r,d one at a time on a horizontal surface. The Hel- loWorld dataset DHW = {D0:6} consists of 7 tasks, each containing 8 slightly varying demonstrations of a letter. Each demonstration is a sequence of 1000 2-D points. After training on all the tasks, the objective is to make the robot write the words ‘‘hello world’’. Our motivation for using this dataset is to test our approach on trajectories containing loops and to show that it also works on kinesthetically recorded demonstrations using a real robot. This dataset is available at https://github.com/sayantanauddy/clfd. Additionally, we introduce two new easy-to-compute contin- ual learning metrics: Time Efficiency (TE) and Final Model Size (FS). TE measures the increase in training duration with the number of tasks, relative to the training time for the first task. TE only needs the training times to be logged, and it reflects the extra effort needed in the training loop (e.g. due to extra regularization steps) with an increase in the number of tasks. TE is similar to the Computational Efficiency metric (CE) proposed in [42], but while CE is based on the number of addition and multiplication operations, TE is based on the observed wall clock time. When a neural network is trained on a GPU (as is most often the case), many neural network operations are performed in parallel. Hence, simply counting the number of operations does not provide an accurate estimate of the increase in training effort with increasing number of tasks. A time-based measure such as TE is more suited to this task. For a meaningful interpretation of this metric, all experiments need to be run on near-identical hardware. 5.1. Datasets We use this dataset because we want to evaluate our approach in a real-world setup involving changing positions and orienta- tions, and also because we did not find any existing datasets that are suitable for our purpose. Datasets such as RoboNet [39] and Meta-World [40] are designed for reinforcement learning and do not contain the demonstrations in terms of task-space trajecto- ries. See Fig. 1 for a visual example of the tasks in DRobot. This dataset is also available at https://github.com/sayantanauddy/ clfd. We evaluate the performance of all models described in Sec- tion 4 on three different datasets with different numbers of tasks and degrees of difficulty: (i) LASA [8]: A dataset that is used frequently for comparisons and benchmarking in the area of LfD [9,37,38]; (ii) HelloWorld: A dataset of trajectories recorded while a robot was shown how to write letters containing loops; (iii) RoboTasks: A dataset of trajectories collected while the robot was shown how to perform tasks which need both the positions as well as orientations of the end-effector to be varied. Of these three datasets, HelloWorld and RoboTasks are introduced by us in this paper. 4.2.4. Regularization i lli Synaptic Intelligence (SI): To learn M tasks, the NODE parame- ters are regularized with the SI loss ˜Lm (Eq. (7)). The task-specific part (Lm) of ˜Lm corresponds to the NODE loss (Eq. (2)). Similar to FT, we make the SI NODE task-conditioned using a trainable task embedding vector, as shown in Fig. 3(d). To reproduce a particular task with HN or CHN at test time, we input the corresponding task embedding vector to the hyper- network which generates the NODE for that task. Thereafter, we simply provide the NODE with the starting state of the trajectory (and the timesteps in the case of NODET) and it predicts the entire trajectory for the desired task. Memory Aware Synapses (MAS): For MAS [4], we also follow the architecture in Fig. 3(d). The NODE parameters are learned using Eq. (7) and we use Eq. (9) to compute Ωm k . As before, we apply Eq. (2) as the task-specific loss Lm. The MAS NODE is also made task-conditioned with a trainable task embedding vector. j y For the comparative baselines (SG, REP, FT, SI, MAS), we in- cluded representative methods from all the major groups of con- tinual learning methods. The choice of hypernetworks for our proposed approach to continual LfD (HN, CHN) was motivated by the following: (i) Hypernetworks have exhibited remarkable performance in multiple domains [2,5]. (ii) A robot capable of learning multiple tasks from demonstration needs to be told which task it should execute. Hypernetworks provide a natural way of task-conditioning using task embedding vectors, and thus are a very good fit for the LfD setup. (iii) Hypernetworks do not need to retrain on the data of past tasks. (iv) The chunked version of hypernetworks can be smaller than the NODEs they generate (model compression). Both SI and MAS seek to change the plasticity/elasticity of the parameters of the NODE in an attempt to prevent changes to parameters that are important for solving past tasks. Hypernetworks (HN): We use a hypernetwork to generate the parameters of a NODE. We first compute the candidate change ∆h for the hypernetwork parameters by minimizing the NODE loss (Eq. (2)), which acts as our task-specific loss Lm: Lm = Lm(θm, ym) = 1 2 ∑ t ∥ym t −ˆym t ∥2 2 (12) Lm = Lm(θm, ym) = 1 2 ∑ t ∥ym t −ˆym t ∥2 2 (12) (12) 6 S. Auddy, J. 5.3.1. LASA dataset Fig. 5. DTW errors (lower is better) of all trajectory predictions during the learn- ing of all tasks of the LASA dataset. Results are obtained using 5 independent seeds. The hypernetwork approach (HN, CHN) performs substantially better than regularization-based continual learning methods (SI, MAS) and on par with the upper baseline SG and REP. We train each model (SG, FT, REP, SI, MAS, CHN and HN) on the 26 tasks of DLASA sequentially. Fig. 4 shows the median DTW errors of the predictions for tasks D0–Dm after training on task Dm (using NODET) for m = 0, 1, . . . , 25. For example, the value for task 10 denotes the median of the evaluation errors for all the predicted trajectories from tasks 0 to 10 after training on task 10. We provide the same information using the other trajectory error metrics (Frechet distance and Swept Area in Fig. A.2 in Appendix A). Fig. 4 (top) shows a drastic increase in the error for FT as the number of learned tasks increases, since FT optimizes its parameters only for the current task. After the first task, the errors for SI and MAS also increase steeply. Fig. 4 (bottom) shows a zoomed-in version showing the methods which perform well in more detail (SG, REP, CHN and HN). The performance of REP for task 0 is the same as FT, but the performance does not deteriorate as more tasks are learned, since REP has access to the data from prior tasks. Till task 7, CHN performs better than REP, after which it starts exhibiting catastrophic forgetting as can be seen from the upward trend in its error plot. Overall, SG and HN exhibit the best performance. They do not suffer from catastrophic forgetting, and their error plots overlap with each other (red and blue lines in Fig. 4 (bottom)). Fig. 5. DTW errors (lower is better) of all trajectory predictions during the learn- ing of all tasks of the LASA dataset. Results are obtained using 5 independent seeds. The hypernetwork approach (HN, CHN) performs substantially better than regularization-based continual learning methods (SI, MAS) and on par with the upper baseline SG and REP. target network parameters. 5.1.3. RoboTasks To evaluate our approach on robotic tasks that can be expected in the real world, we collect a dataset of 4 tasks. For each task, a human user kinsethetically guides the robot’s end-effector while varying both the position in 3D space and also the orientation in all three rotation axes. The tasks of this dataset are: (i) box opening: the lid of a box is lifted to an open position; (ii) bottle shelving: a bottle in a vertical position is transferred to a hori- zontal position on a shelf; (iii) plate stacking: a plate in a vertical position is transferred to a horizontal position on an elevated platform while orienting the arm so as to avoid the blocks used for holding the plate in its initial vertical position; (iv) pouring: a cup full of coffee beans is taken from an elevated platform and For M tasks, TE is defined as For M tasks, TE is defined as TE = min { 1, T0 M M−1 ∑ i=0 1 Ti } , (15) (15) where Ti is the time required to learn task i. where Ti is the time required to learn task i. 7 Robotics and Autonomous Systems 165 (2023) 104427 Fig. 4. DTW errors of trajectories predicted for the LASA dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. HN performs as well as SG but with negligible growth in parameter size. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 FS is a measure of the absolute parameter size, which contrasts with MS which only measures the parameter growth relative to the size after learning the first task. 5.3. Results Fig. 4. DTW errors of trajectories predicted for the LASA dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. HN performs as well as SG but with negligible growth in parameter size. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) We present next the results of our experiments on three different datasets: LASA, HelloWorld, and RoboTasks. Details of our hardware setup (Appendix A.1), experiment hyperparameters (Appendix A.4), results of the experimental comparison of NODEs and iFlow (Appendix A.5), and additional results (Appendix A.6) can be found in Appendix A. 5.1.3. RoboTasks A model which has a large number of parameters for the first task and adds a relatively small number of parameters for subsequent tasks will achieve a high score for MS, but will fare worse in terms of FS if models of other compared methods have a smaller absolute size. FS is defined as FS = 1 −Memnorm(θM−1) (16) (16) Memnorm(θM−1) is the parameter size after learning M tasks, normalized by the size of the largest compared model among SG, FT, REP, SI, MAS, HN and CHN. With these 6 metrics, we compute the overall continual learning metrics [42]: CLscore = n(C) ∑ ci, (17) CLstability = 1 − n(C) ∑ STDEV(ci), (18) (17) (18) where C = {ACC, REM, MS, TE, FS, SSS}. All the continual learning metrics lie in the range 0 (worst) to 1 (best). where C = {ACC, REM, MS, TE, FS, SSS}. All the continual learning metrics lie in the range 0 (worst) to 1 (best). Table 2 Table 2 Robustness to changes in regularization hyperparameters for the LASA dataset (5 configurations for each method). Table 2 Robustness to changes in regularization hyperparameters for the LASA dataset (5 configurations for each method). Method CLscore CLstability Median IQR Median IQR HN 0.8578 0.0011 0.8324 0.0050 CHN 0.7939 0.0098 0.8126 0.0022 MAS 0.7104 0.0019 0.6562 0.0062 SI 0.6047 0.0065 0.6403 0.0011 Fig. 7. Example of trajectories predicted by SG, CHN and HN using NODET for a selection of LASA tasks after learning the last task. HN can remember all past tasks, while CHN exhibits some forgetting. each for SI, MAS, HN and CHN by drawing independently and uniformly from the following ranges: (SI) c ∈[0.1, 0.5], ξ ∈ [0.1, 0.5], (MAS) c ∈[0.1, 0.5], (CHN) β ∈[10−3, 10−2], (HN) β ∈[10−3, 10−2] resulting in 20 different configurations. We then repeat the LASA experiment with NODET for all these con- figurations. In terms of CLscore we observe that all configurations of HN outperform all configurations of CHN, which in turn are better than all configurations of MAS, followed by SI. This trend is reflected in the medians and inter-quartile ranges (IQR) of the overall continual learning metrics CLscore and CLstability for each method (over its 5 configurations) shown in Table 2. It can be seen that HN and CHN perform better than the other methods and the variability in terms of IQR is very small, thereby showing that they are robust to changes in the regularization hyperparameter β. For calculating CLscore and CLstability in this experiment, we do not consider the SSS metric since none of the regularization-based methods need to cache training data from prior tasks. by SG, CHN, and HN for a selection of past tasks after learning the last task. To compute the continual learning metrics [42], each predicted trajectory needs to be marked as accurate or inaccurate based on its difference to the ground truth. Since there is no preexisting procedure for this, we adopt the following approach: We set a threshold on the DTW error, such that predictions with an error less than the threshold are considered accurate. As each task has multiple ground truth demonstrations, we first compute the DTW error between all pairs of demonstrations for each task. We then find the maximum value from this list and multiply it by 3. Table 1 Table 1 Continual learning metrics for the LASA dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). Continual learning metrics for the LASA dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). (a) NODET MET ACC REM MS TE FS SSS CLsco CLstab SG 0.87 1.00 0.15 0.85 0.00 1.00 0.64 0.55 FT 0.06 0.20 1.00 0.89 0.96 1.00 0.68 0.57 REP 0.67 1.00 1.00 0.88 0.96 0.48 0.83 0.79 SI 0.04 0.38 1.00 0.98 0.78 1.00 0.70 0.60 MAS 0.02 0.86 1.00 0.83 0.83 1.00 0.76 0.63 HN 0.86 0.97 1.00 0.51 0.92 1.00 0.88 0.81 CHN 0.51 0.80 1.00 0.53 0.96 1.00 0.80 0.77 (b) NODEI MET ACC REM MS TE FS SSS CLsco CLstab SG 0.81 1.00 0.15 0.85 0.00 1.00 0.64 0.56 FT 0.06 0.22 1.00 0.88 0.96 1.00 0.69 0.57 REP 0.65 1.00 1.00 0.83 0.96 0.48 0.82 0.79 SI 0.04 0.38 1.00 0.91 0.78 1.00 0.69 0.61 MAS 0.03 0.84 1.00 0.87 0.83 1.00 0.76 0.63 HN 0.76 0.97 1.00 0.55 0.92 1.00 0.87 0.82 CHN 0.57 0.86 1.00 0.52 0.96 1.00 0.82 0.78 Fig. 6. Growth of parameter size with new tasks for the LASA dataset (using NODET). SG has a high rate of growth since it uses a separate network for each task. All other models grow by only 256 parameters for each new task. Plots for CHN and FT overlap with each other. The plot for REP is not shown since its size is identical to FT. Fig. 7. Example of trajectories predicted by SG, CHN and HN using NODET for a selection of LASA tasks after learning the last task. HN can remember all past tasks, while CHN exhibits some forgetting. 5.3.1. LASA dataset Values range from 0 (worst) to 1 (best). (a) NODET MET ACC REM MS TE FS SSS CLsco CLstab SG 0.87 1.00 0.15 0.85 0.00 1.00 0.64 0.55 FT 0.06 0.20 1.00 0.89 0.96 1.00 0.68 0.57 REP 0.67 1.00 1.00 0.88 0.96 0.48 0.83 0.79 SI 0.04 0.38 1.00 0.98 0.78 1.00 0.70 0.60 MAS 0.02 0.86 1.00 0.83 0.83 1.00 0.76 0.63 HN 0.86 0.97 1.00 0.51 0.92 1.00 0.88 0.81 CHN 0.51 0.80 1.00 0.53 0.96 1.00 0.80 0.77 (b) NODEI MET ACC REM MS TE FS SSS CLsco CLstab SG 0.81 1.00 0.15 0.85 0.00 1.00 0.64 0.56 FT 0.06 0.22 1.00 0.88 0.96 1.00 0.69 0.57 REP 0.65 1.00 1.00 0.83 0.96 0.48 0.82 0.79 SI 0.04 0.38 1.00 0.91 0.78 1.00 0.69 0.61 MAS 0.03 0.84 1.00 0.87 0.83 1.00 0.76 0.63 HN 0.76 0.97 1.00 0.55 0.92 1.00 0.87 0.82 CHN 0.57 0.86 1.00 0.52 0.96 1.00 0.82 0.78 Table 2 Robustness to changes in regularization hyperparameters for the LASA dataset (5 configurations for each method). Method CLscore CLstability Median IQR Median IQR HN 0.8578 0.0011 0.8324 0.0050 CHN 0.7939 0.0098 0.8126 0.0022 MAS 0.7104 0.0019 0.6562 0.0062 SI 0.6047 0.0065 0.6403 0.0011 each for SI, MAS, HN and CHN by drawing independently and uniformly from the following ranges: (SI) c ∈[0.1, 0.5], ξ ∈ [0.1, 0.5], (MAS) c ∈[0.1, 0.5], (CHN) β ∈[10−3, 10−2], (HN) β ∈[10−3, 10−2] resulting in 20 different configurations. We then repeat the LASA experiment with NODET for all these con- figurations. In terms of CLscore we observe that all configurations of HN outperform all configurations of CHN, which in turn are better than all configurations of MAS, followed by SI. This trend Table 1 Continual learning metrics for the LASA dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). 5.3.1. LASA dataset To keep the overall size of the HN comparable to the other models, the target NODE for HN has one- tenth the number of units in each layer as the NODEs used by the other models (SG, FT, REP, SI, MAS, CHN). We refer the reader to Table A.3 in Appendix A for details. An overall picture of the continual learning performance of the different methods may be obtained from Fig. 5, which shows the overall errors of the predicted trajectories during the course of learning all the tasks. It can be seen that SG, REP, CHN and HN perform much better than FT, SI, and MAS, and that HN and SG are the best performers. Note that in Fig. 5 the trajectory metrics are plotted in the log10 scale to accommodate the high errors for FT, SI, and MAS in the same plot as SG, REP, HN and CHN. Although HN’s performance after learning 26 tasks is very similar to that of SG, its parameter size is 4.3 × 106 compared to SG’s combined size of 52.2 × 106 parameters, as shown in Fig. 6. CHN’s final parameter size is 1.9×106. Also, the parameter count for SG grows by 2.1 × 106 per task, whereas CHN and HN grow at a much smaller rate of 256 parameters per task, same as SI, MAS, FT and REP (Fig. 6 does not include REP as it has the same model size as FT). Thus, CHN and especially HN perform similar to the upper baseline SG, while their parameter size is close to the lower baseline FT. Fig. 7 shows examples of trajectories predicted Note that a regular HN model produces all the parameters of the target NODE directly from its final layer. This implies a much larger number of HN parameters compared to the produced 8 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Fig. 6. Growth of parameter size with new tasks for the LASA dataset (using NODET). SG has a high rate of growth since it uses a separate network for each task. All other models grow by only 256 parameters for each new task. Plots for CHN and FT overlap with each other. The plot for REP is not shown since its size is identical to FT. Table 1 Continual learning metrics for the LASA dataset (median over 5 seeds). Robotics and Autonomous Systems 165 (2023) 104427 DTW errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. The performance of SG, REP, HN and CHN are nearly identical and much better than FT, SI and MAS. Fig. 10. Example of trajectories predicted by SG, CHN and HN using NODET for all past HelloWorld tasks after learning the last task. With NODET, even trajectories with loops can be learned. HN and CHN (with NODET) can remember all past tasks. Fig. 10. Example of trajectories predicted by SG, CHN and HN using NODET for all past HelloWorld tasks after learning the last task. With NODET, even trajectories with loops can be learned. HN and CHN (with NODET) can remember all past tasks. Fig. 10. Example of trajectories predicted by SG, CHN and HN using NODET for all past HelloWorld tasks after learning the last task. With NODET, even trajectories with loops can be learned. HN and CHN (with NODET) can remember all past tasks. Fig. 10. Example of trajectories predicted by SG, CHN and HN using NODET for all past HelloWorld tasks after learning the last task. With NODET, even trajectories with loops can be learned. HN and CHN (with NODET) can remember all past tasks. Fig. 9. DTW errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. The performance of SG, REP, HN and CHN are nearly identical and much better than FT, SI and MAS. Fig. 11. DTW errors (lower is better) of all trajectory predictions during the learning of all tasks of the HelloWorld dataset. Results are obtained using 5 independent seeds. HN, CHN perform as well as the upper baseline SG. SG is unable to learn trajectories with loops. Robotics and Autonomous Systems 165 (2023) 104427 The advantage of using NODEs with a time input is clear from the higher values of ACC for NODET compared to NODEI for all the methods. In terms of ACC or REM, there is very little to choose between SG, REP, HN and CHN. However, when all the CL metrics are considered together (see the CLsco column in Table 3), CHN exhibits the best performance on account of its small size and the fact that it does not need to cache training data from past tasks. HN shows the second-best overall performance. Fig. 12 illustrates that the hypernetwork methods (HN and CHN) perform well in all continual learning metrics, while the other methods perform well in only some aspects of continual learning but fail in others. Fig. 10. Example of trajectories predicted by SG, CHN and HN using NODET for all past HelloWorld tasks after learning the last task. With NODET, even trajectories with loops can be learned. HN and CHN (with NODET) can remember all past tasks. Fig. 11. DTW errors (lower is better) of all trajectory predictions during the learning of all tasks of the HelloWorld dataset. Results are obtained using 5 independent seeds. HN, CHN perform as well as the upper baseline SG. We qualitatively evaluate how the trajectories predicted by HN can be reproduced with a physical robot. For this, we use the same Franka Emika Panda robot that was used for recording the demonstrations for DHW. The HN model trained on the 7 tasks of DHW is queried to produce the letters h, e, l, l, o, w, o, r, l, d by using the appropriate task embedding vectors in sequence. The trajectory of each letter is scaled and translated by a constant amount and provided to the robot, which then follows this path with its end-effector. The z-coordinate and orientation of the Fig. 9. DTW errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. The performance of SG, REP, HN and CHN are nearly identical and much better than FT, SI and MAS. Fig. 9. Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 8. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the LASA dataset. In (a)–(e), each of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. HN shows good performance for almost all metrics (apart from TE, in which it performs moderately). CHN is similar to HN in terms of TE but scores lower in REM and ACC. REP also performs well but has a low score for SSS, and the other methods fail in multiple metrics. Fig. 8. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the LASA dataset. In (a)–(e), each of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. HN shows good performance for almost all metrics (apart from TE, in which it performs moderately). CHN is similar to HN in terms of TE but scores lower in REM and ACC. REP also performs well but has a low score for SSS, and the other methods fail in multiple metrics Fig. 9. DTW errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The DTW errors for all methods. (bottom) A zoomed-in view of the methods that perform well. The performance of SG, REP, HN and CHN are nearly identical and much better than FT, SI and MAS. SG is unable to learn trajectories with loops. This can be seen from the errors for the letters e, r and d in Fig. 10(b). This is also evident in Fig. 11 which shows the errors for all predictions during the course of learning all the tasks. Apart from FT, all methods have higher median errors when using NODEI (Fig. 11(b)) compared to NODET (Fig. 11(a)). Using the same threshold computation approach we followed for DLASA, we compute a DTW threshold value of 1821 for DHW. With this, we compute the continual learning metrics shown in Table 3. Table 2 The multiplicative factor 3 was chosen empirically to allow some room for error such that a predicted trajectory with the same general shape as its demonstration (discerned visually) is considered accurate. Doing so, we arrive at a DTW threshold of 2191 for DLASA, and use it to evaluate the continual learning metrics shown in Table 1. For both NODE variants, HN outperforms all the compared models in terms of CLscore. For NODET, HN performs close to the upper baseline SG in terms of both ACC and REM. The addi- tional regularization needed for training hypernetworks leads to a comparatively lower score for the time efficiency metric TE for CHN and HN (actual wall-clock training times can be found in Table A.1(a) in Appendix A). A very high parameter growth rate for SG results in poor scores for MS and FS. The replay method REP achieves a low score for SSS since it is the only method that needs to store the training data from previous tasks. The direct time input in NODET also leads to better ACC scores for SG, REP and HN, compared to NODEI. Fig. 8 shows how the different methods measure up against each other in the different aspects of continual learning performance. For DHW, which comprises 7 tasks, we perform the same experiments as DLASA. Fig. 9 shows the errors in the predicted tra- jectories for all past and current tasks, as new tasks are learned. The median errors for CHN, HN and REP stay nearly unchanged and are similar to the upper baseline SG, although CHN and HN do not need to store the training data of past tasks like REP. As before, FT, SI, and MAS exhibit severe catastrophic forgetting. Due to fewer tasks in DHW, CHN’s performance does not deteriorate even after learning all tasks. Fig. A.3 in Appendix A shows the other trajectory error metrics (Frechet distance and Swept Area), which exhibit the same trend as DTW. Fig. 10 shows examples of trajectories predicted by SG, CHN and HN for past tasks after being trained sequentially on all DHW tasks. All models exhibit superior performance when using the additional time input in NODET (Fig. 10(a)), without which even To test the sensitivity of the regularization-based continual learning methods (SI, MAS, HN and CHN) to changes in the regu- larization hyperparameters, we create sets of 5 hyperparameters 9 Table 3 (a) NODET MET ACC REM MS TE FS SSS CLsco CLstab SG 0.99 1.00 0.37 0.94 0.00 1.00 0.72 0.57 FT 0.26 0.07 1.00 0.97 0.84 1.00 0.69 0.59 REP 0.94 0.96 1.00 0.98 0.84 0.43 0.86 0.78 SI 0.25 0.04 1.00 0.95 0.19 1.00 0.57 0.55 MAS 0.38 0.25 1.00 0.86 0.36 1.00 0.64 0.65 HN 0.97 0.98 1.00 0.76 0.69 1.00 0.90 0.86 CHN 0.89 0.95 1.00 0.78 0.87 1.00 0.92 0.91 (b) NODEI MET ACC REM MS TE FS SSS CLsco CLstab SG 0.71 1.00 0.37 0.93 0.00 1.00 0.67 0.59 FT 0.17 0.24 1.00 0.93 0.84 1.00 0.70 0.62 REP 0.71 1.00 1.00 0.94 0.84 0.43 0.82 0.78 SI 0.19 0.29 1.00 0.94 0.19 1.00 0.60 0.59 MAS 0.25 0.47 1.00 0.86 0.36 1.00 0.66 0.66 HN 0.74 0.99 1.00 0.75 0.69 1.00 0.86 0.85 CHN 0.68 0.96 1.00 0.76 0.87 1.00 0.88 0.87 Fig. 13. After being continually trained using learning from demonstration to write single letters, the robot can reproduce all the trajectories that it has learned in the past with a single HN network and without having access to training data from past tasks. Video is available at https://youtu.be/0gdIImIBnXc. (a) NODET MET ACC REM MS TE FS SSS CLsco CLstab SG 0.99 1.00 0.37 0.94 0.00 1.00 0.72 0.57 FT 0.26 0.07 1.00 0.97 0.84 1.00 0.69 0.59 REP 0.94 0.96 1.00 0.98 0.84 0.43 0.86 0.78 SI 0.25 0.04 1.00 0.95 0.19 1.00 0.57 0.55 MAS 0.38 0.25 1.00 0.86 0.36 1.00 0.64 0.65 HN 0.97 0.98 1.00 0.76 0.69 1.00 0.90 0.86 CHN 0.89 0.95 1.00 0.78 0.87 1.00 0.92 0.91 (b) NODEI MET ACC REM MS TE FS SSS CLsco CLstab SG 0.71 1.00 0.37 0.93 0.00 1.00 0.67 0.59 FT 0.17 0.24 1.00 0.93 0.84 1.00 0.70 0.62 REP 0.71 1.00 1.00 0.94 0.84 0.43 0.82 0.78 SI 0.19 0.29 1.00 0.94 0.19 1.00 0.60 0.59 MAS 0.25 0.47 1.00 0.86 0.36 1.00 0.66 0.66 HN 0.74 0.99 1.00 0.75 0.69 1.00 0.86 0.85 CHN 0.68 0.96 1.00 0.76 0.87 1.00 0.88 0.87 p g Fig. 14 shows the position and orientation errors of the pre- dicted trajectories for all past and current tasks, as new tasks are learned. For both positions and orientations, FT is unable to predict correct trajectories for past tasks as soon as the second task is learned. Table 3 The performance of the other methods is com- parable to each other. For the end-effector position trajectories (Fig. 14(a) (bottom)), the performance of SG, REP and HN are nearly identical, while CHN is slightly worse. However this dif- ference does not result in significant qualitative differences in the robot’s movement. SG, REP, CHN and HN are equally good at predicting the orientation trajectories, as can be seen in Fig. 14(b) (bottom). The slight upward trend in the plots of SG, REP, CHN and HN in Fig. 14 can be attributed to the fact that task 2 (plate stacking) is more difficult than the other tasks due to very diverse ground truth demonstrations. It can be observed that CHN shows a more pronounced upward trend in the position error plot (Fig. 14(a) (bottom)). This is probably due to the fact that CHN has far fewer parameters than HN and thus has lower representation and remembrance capabilities. Fig. 13. After being continually trained using learning from demonstration to write single letters, the robot can reproduce all the trajectories that it has learned in the past with a single HN network and without having access to training data from past tasks. Video is available at https://youtu.be/0gdIImIBnXc. Similar to the results for DLASA and DHW, it can be seen that for continually learning real-world robot tasks involving both position and orientation, the hypernetwork-based methods CHN and HN perform similar to SG and REP, but without needing significant growth in parameters and without requiring to be retrained on the data from prior tasks. Fig. 15 shows an example of the position trajectories and Fig. 16 shows an example of the orientation trajectories predicted by SG, CHN and HN for tasks 0–4 after the last task has been learned (here we show the orientation in terms of Euler angles and the corresponding results in terms of quaternion elements can be seen in Fig. A.4). In both these qualitative examples, CHN and HN produce trajectories that closely mimic the ground truth demonstration. Fig. 17 shows the overall prediction errors during the course of training. Images of the robot performing the 4 tasks of DRobot after learning all the tasks sequentially with CHN can be seen in Fig. 1. Fig. 13. Table 3 After being continually trained using learning from demonstration to write single letters, the robot can reproduce all the trajectories that it has learned in the past with a single HN network and without having access to training data from past tasks. Video is available at https://youtu.be/0gdIImIBnXc. Next, we report the continual learning metrics for DRobot in Table 4. For this we again need to set thresholds on the posi- tion and orientation errors to categorize predicted trajectories as accurate or inaccurate. This time we follow a slightly different approach for setting these thresholds compared to DLASA and DHW. For positions, we first compute the DTW error between all pairs of demonstrations for each task. Since the ground truth demonstrations are already quite diverse, we use the maximum value from this list as the position error threshold. This threshold DTW value is 7131. For setting the orientation error threshold, we simply use a threshold of 10 degrees, since an absolute error of 10 degrees (averaged across the 3 rotation axes) results in roughly the same orientation as the desired one. end-effector are fixed. Fig. 13 shows the letters written by the robot. Robotics and Autonomous Systems 165 (2023) 104427 This can be seen from the errors for the letters e, r and d in Fig. 10(b). This is also evident in Fig. 11 which shows the errors for all predictions during the course of learning all the tasks. Apart from FT, all methods have higher median errors when using NODEI (Fig. 11(b)) compared to NODET (Fig. 11(a)). ( g ( )) Using the same threshold computation approach we followed for DLASA, we compute a DTW threshold value of 1821 for DHW. With this, we compute the continual learning metrics shown in Table 3. The advantage of using NODEs with a time input is clear from the higher values of ACC for NODET compared to NODEI for all the methods. In terms of ACC or REM, there is very little to choose between SG, REP, HN and CHN. However, when all the CL metrics are considered together (see the CLsco column in Table 3), CHN exhibits the best performance on account of its small size and the fact that it does not need to cache training data from past tasks. HN shows the second-best overall performance. Fig. 12 illustrates that the hypernetwork methods (HN and CHN) perform well in all continual learning metrics, while the other methods perform well in only some aspects of continual learning but fail in others. Fig. 11. DTW errors (lower is better) of all trajectory predictions during the learning of all tasks of the HelloWorld dataset. Results are obtained using 5 independent seeds. HN, CHN perform as well as the upper baseline SG. We qualitatively evaluate how the trajectories predicted by HN can be reproduced with a physical robot. For this, we use the same Franka Emika Panda robot that was used for recording the demonstrations for DHW. The HN model trained on the 7 tasks of DHW is queried to produce the letters h, e, l, l, o, w, o, r, l, d by using the appropriate task embedding vectors in sequence. The trajectory of each letter is scaled and translated by a constant amount and provided to the robot, which then follows this path with its end-effector. The z-coordinate and orientation of the 10 Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 12. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the HelloWorld dataset. In (a)–(e), each of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. The hypernetwork methods (HN,CHN) show good performance for all metrics whereas the other methods perform well only in some metrics and poorly for others. Fig. 12. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the HelloWorld dataset. In (a)–(e), each of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. The hypernetwork methods (HN,CHN) show good performance for all metrics whereas the other methods perform well only in some metrics and poorly for others. Table 3 Continual learning metrics for the HelloWorld dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). Table 3 Continual learning metrics for the HelloWorld dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). Table 3 baseline. For each method, we train separate models for learning the position and orientation trajectories. The order of tasks for this experiment is: box opening, bottle shelving, plate stacking, and pouring. Continual learning metrics for the HelloWorld dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). (a) NODET MET ACC REM MS TE FS SSS CLsco CLstab SG 0.99 1.00 0.37 0.94 0.00 1.00 0.72 0.57 FT 0.26 0.07 1.00 0.97 0.84 1.00 0.69 0.59 REP 0.94 0.96 1.00 0.98 0.84 0.43 0.86 0.78 SI 0.25 0.04 1.00 0.95 0.19 1.00 0.57 0.55 MAS 0.38 0.25 1.00 0.86 0.36 1.00 0.64 0.65 HN 0.97 0.98 1.00 0.76 0.69 1.00 0.90 0.86 CHN 0.89 0.95 1.00 0.78 0.87 1.00 0.92 0.91 (b) NODEI MET ACC REM MS TE FS SSS CLsco CLstab SG 0.71 1.00 0.37 0.93 0.00 1.00 0.67 0.59 FT 0.17 0.24 1.00 0.93 0.84 1.00 0.70 0.62 REP 0.71 1.00 1.00 0.94 0.84 0.43 0.82 0.78 SI 0.19 0.29 1.00 0.94 0.19 1.00 0.60 0.59 MAS 0.25 0.47 1.00 0.86 0.36 1.00 0.66 0.66 HN 0.74 0.99 1.00 0.75 0.69 1.00 0.86 0.85 CHN 0.68 0.96 1.00 0.76 0.87 1.00 0.88 0.87 Fig. 13. After being continually trained using learning from demonstration to write single letters, the robot can reproduce all the trajectories that it has learned in the past with a single HN network and without having access to training data from past tasks. Video is available at https://youtu.be/0gdIImIBnXc. Values range from 0 (worst) to 1 (best). 5.3.3. RoboTasks DRobot comprises 4 tasks, where each task involves learning the vector fields of the position as well as the orientation of the end-effector of the robot. For this experiment, the methods we evaluate are SG, FT, REP, CHN and HN. We omit SI and MAS because of their poor performance on the previous datasets. We use only NODET for this experiment since in the previous exper- iments it has been shown to be better than NODEI. As before, SG acts as an upper baseline and FT serves as a lower performance 11 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 14. Position (a) and orientation (b) errors of trajectories predicted for the RoboTasks dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top row) The errors for all methods are plotted. (bottom row) A detailed view of the methods that perform well. The performance of HN is comparable to the upper baseline SG. Fig. 14. Position (a) and orientation (b) errors of trajectories predicted for the RoboTasks dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top row) The errors for all methods are plotted. (bottom row) A detailed view of the methods that perform well. The performance of HN is comparable to the upper baseline SG. Fig. 15. Example of end-effector position trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. Hypernetworks (HN and CHN) remember all past tasks and mimic the demonstrations accurately. Fig. 15. Example of end-effector position trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. Hypernetworks (HN and CHN) remember all past tasks and mimic the demonstrations accurately. Fig. 16. Table 4 Table 4 Continual learning metrics for the RoboTasks dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). Table 4 Continual learning metrics for the RoboTasks dataset (median over 5 seeds). Values range from 0 (worst) to 1 (best). g ( ) ( ) (a) Position MET ACC REM MS TE FS SSS CLsco CLstab SG 0.95 1.00 0.52 0.93 0.00 1.00 0.73 0.60 FT 0.39 0.01 1.00 0.84 0.69 1.00 0.66 0.61 REP 0.94 0.94 1.00 0.84 0.69 0.38 0.80 0.77 HN 0.88 0.98 1.00 0.72 0.45 1.00 0.84 0.78 CHN 0.90 0.90 1.00 0.75 0.74 1.00 0.88 0.89 (b) Orientation MET ACC REM MS TE FS SSS CLsco CLstab SG 0.98 1.00 0.52 0.99 0.00 1.00 0.75 0.59 FT 0.49 0.15 1.00 0.99 0.66 1.00 0.68 0.62 REP 0.99 1.00 1.00 1.00 0.66 0.38 0.84 0.74 HN 0.99 1.00 1.00 0.77 0.20 1.00 0.83 0.68 CHN 0.98 0.97 1.00 0.78 0.79 1.00 0.92 0.90 Fig. 18. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the RoboTasks dataset. In each individual plot, one of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. The first column (a)–(c) shows the results for position, and the second column (d)–(f) shows the results for orientation. Across all metrics, CHN shows good performance for both position and orientation. HN exhibits a similar performance but due to its bigger size, achieves a low score for FS for orientation. REP’s score for SSS is low due to its requirement of storing training data of all tasks. SG and FT achieve low scores in multiple metrics. parameter size and its non-dependence on the demonstrations of previous tasks. This is also corroborated by Fig. 18, which shows that CHN performs well in all the continual learning metrics, while the other methods perform poorly in one or more metrics. (HelloWorld and RoboTasks). Compared to the growth of SG for each new task, the hypernetwork models (HN, CHN) scale much more efficiently as the growth per task of HN and CHN is negligible (Section 5.3). We also evaluate the robustness of the predicted trajectories when the robot starts from novel initial positions that are not present in the demonstrations used for training. 5.3.3. RoboTasks Example of end-effector orientation trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. The ground truth demonstrations as well as the predictions of the orientation consist of a sequence of unit quaternions. Here, we convert each quaternion in the sequence to its corresponding Euler angle representation (z −y −x convention) for easier interpretation. Columns show the predictions for each task, and rows show the different Euler angles in degrees. A figure with the original quaternion sequences can be seen in Fig. A.4. Hypernetworks (HN and CHN) remember all past tasks and mimic the ground truth orientations accurately. Fig. 16. Example of end-effector orientation trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. The ground truth demonstrations as well as the predictions of the orientation consist of a sequence of unit quaternions. Here, we convert each quaternion in the sequence to its corresponding Euler angle representation (z −y −x convention) for easier interpretation. Columns show the predictions for each task, and rows show the different Euler angles in degrees. A figure with the original quaternion sequences can be seen in Fig. A.4. Hypernetworks (HN and CHN) remember all past tasks and mimic the ground truth orientations accurately. For position as well as orientation, SG, REP, CHN and HN are comparable to each other in terms of ACC and REM. However, when all the metrics are considered together, CHN is the best amongst all the methods on account of its small near-constant when all the metrics are considered together, CHN is the best amongst all the methods on account of its small near-constant 12 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. 17. Prediction errors (lower is better) showing DTW position errors (a) and Quaternion orientation errors (b) of all trajectory predictions during the learning of all tasks of the RoboTasks dataset. Results are obtained using 5 independent seeds. Overall, HN and CHN produce low errors which are comparable to the performance of the upper baseline SG. Fig. 18. Continual learning metrics (0: worst, 1: best) for the different methods (with NODET) on the RoboTasks dataset. Table 4 To this end, we compare the predictions made by upper-baseline SG and CHN for 100 randomly selected starting positions on each of the 4 tasks of the RoboTasks dataset. Please refer to Appendix A.7 for details. Our results indicate that SG is much more sensitive to small changes in the starting position despite being the best at predicting trajectories that have the same starting position as the demonstrations. The continual learning process for CHN appears to reduce overfitting to the demonstrations and makes the model more robust to changes in the initial state of the robot. Using continual learning metrics (e.g. accuracy of predictions, how well past tasks are remembered, model size growth, stor- age of training data from past tasks, etc.), we compared the hypernetwork-based approach to methods from all continual learning families and found that empirically, hypernetworks per- form best. Specifically, HN is the best among the compared mod- els for the LASA dataset. On the remaining 2 robot datasets, the hypernetwork-based methods outperform the other approaches. CHN, due to its memory efficiency, is empirically the best, fol- lowed by HN. We also improve the trajectory learning NODE by introduc- ing an additional time input to create the NODET model. In Section 5.3.2, we show how NODET improves performance and enables the models to learn trajectories with loops. 5.3.3. RoboTasks In each individual plot, one of the baseline methods is compared against HN and CHN to illustrate the different tradeoffs made by the methods. The first column (a)–(c) shows the results for position, and the second column (d)–(f) shows the results for orientation. Across all metrics, CHN shows good performance for both position and orientation. HN exhibits a similar performance but due to its bigger size, achieves a low score for FS for orientation. REP’s score for SSS is low due to its requirement of storing training data of all tasks. SG and FT achieve low scores in multiple metrics. Fig. 17. Prediction errors (lower is better) showing DTW position errors (a) and Quaternion orientation errors (b) of all trajectory predictions during the learning of all tasks of the RoboTasks dataset. Results are obtained using 5 independent seeds. Overall, HN and CHN produce low errors which are comparable to the performance of the upper baseline SG. A.3. Inference times Further interesting research directions in continual LfD may include: improving the ability of the chunked hypernetworks to remember long sequences of tasks (e.g., via architectural changes), using continual reinforcement learning [2] to refine predicted trajectories [22], and the inclusion of visual cues to improve and adapt trajectories [43]. The time taken by a network to produce predictions is criti- cally important in a robotics scenario. The hypernetwork-based solutions for learning from demonstration are well suited for this task, as can be seen by the speed of making predictions reported in Table A.2. For this evaluation, the models which were originally trained on the GPU, were loaded either into GPU memory (Ta- ble A.2 (a)) or into the CPU memory (Table A.2 (b)). After loading the hypernetworks (HN or CHN), one of the task embedding vec- tors was used as an input to generate the parameters for a NODE and the speed of this forward pass is noted (column fh →fθ in Table A.2 (a) and (b)). Note that this NODE generation step occurs only when the robot is required to perform a new task (task switch). Once a NODE for a task is generated, it can be reused for multiple predictions of that task. The generated NODE is then made to predict a trajectory and the speed of each step (column fθ(x) = ˙x), as well as the time to integrate over a 1000 steps (column ∫ fθ(xτ )dτ) is noted. This process is repeated 100 times and the median times and speeds are reported in milliseconds and Hz. As expected, using the GPU generally results in faster predictions. The time taken to integrate over the entire trajectory of 1000 steps is more than a thousand times the time for a single step (due to the integration operation) and in the majority of cases, this takes much less than 1 s. It can be seen from Table A.2 that hypernetworks are able to produce predictions very fast on both the GPU as well as the CPU. This makes them suitable for use in a robotics application. A.2. Training times In Table A.1, we show the time (in seconds) needed to train each of the 7 continual learning methods presented in this paper (median values over 5 independent seeds). The hypernetwork- based methods (HN and CHN) show a small gradual increase in the training time as the number of tasks increase. This is due to the extra regularization that needs to be performed to prevent catastrophic forgetting. A possible way to attain a near constant training time (as proposed by von Oswald et al. [5]) is to randomly sample a fixed set of task embedding vectors to regularize with instead of using the embeddings of all tasks in each training iteration. We do not apply this process in this paper, but leave it for future work. In this paper, our focus is on continual learning and we use NODEs as a simple trajectory learning method because of its good empirical performance and fast convergence during train- ing, which facilitates running multiple experiments on long se- quences of tasks (Section 4.1). However, NODEs do not have a mechanism for enforcing the stability of the predicted trajec- tories. To overcome this, it is possible to substitute the simple NODE with a stable alternative in the future [9,10]. Interestingly, we found that even without stability guarantees, NODEs gen- erated by a chunked hypernetwork (CHN) produce trajectories that are much more robust to novel initial conditions than the upper-baseline SG, which uses separate networks to learn each individual task (refer to Appendix A.7 for details). 6. Discussion To be functional in the real world, a robot must be able to acquire multiple skills over time, without forgetting the skills that were learned in the past and also without needing to be retrained on past skills. In this paper, we have taken a further step in this direction by presenting the first work on continual trajectory learning from demonstrations. To facilitate future research in the area of continual-LfD, we release 2 new datasets of trajectories collected kinesthetically with a real robot: HelloWorld (Section 5.3.2) and RoboTasks (Sec- tion 5.3.3). The latter includes tasks involving changing positions and orientations. After continually learning all the tasks from these datasets using our hypernetwork-based approach, we ver- ified that the real robot can successfully perform all past tasks. Further, we verified that the order of tasks during training does The results from our experiments on 3 different datasets show that the hypernetwork model (HN) performs on par with the upper baseline (SG) for all datasets, and the smaller chunked hy- pernetwork (CHN) performs on par with SG on the robot datasets 13 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 not affect the performance of these models (refer to Appendix A.8 for details). not affect the performance of these models (refer to Appendix A.8 for details). 7. Conclusion In this paper, we presented the first study on continual learn- ing of trajectories from demonstrations, thereby expanding the field of continual learning in robotics to LfD. We adapted con- tinual learning (CL) methods from all CL families, and compared these against our hypernetwork-based approaches on three dif- ferent datasets. Two of these three datasets were collected using kinesthetic teaching with the real robot and are released by us in this paper. We have also verified that our approach works for real-world tasks involving changing positions and orientations by evaluating it on a real robot. Compared to other continual learn- ing approaches considered in this paper, hypernetworks achieve equal or better accuracy, but scale much better in terms of other criteria such as growth in model size, and storage of past training data. Our results suggest that a regular hypernetwork (HN) is capable of learning many LfD tasks and is empirically the method of choice for a long sequence of tasks. For a shorter task sequence, the smaller chunked hypernetwork (CHN) is empirically the best choice for continual-LfD. Data availability The code used in this paper is made publicly available to aid reproducibility. Although the HelloWorld and RoboTasks datasets used in our experiments were recorded kinesthetically, our im- plementation only requires the training data in terms of the robot’s trajectories and does not depend on how this data is collected. Similar to [19], our implementation can potentially also be used for trajectories recorded using other means such as shadowing or teleoperation [17]. The link to our open source git repository containing our code and datasets has been shared in the paper. Appendix A A.1. Hardware setup A.1. Hardware setup We run all our experiments on a shared computing cluster with 125 GB of RAM and 16 nodes, with each node having 1 AMD Ryzen 2950X 16-Core processor and 4 GeForce RTX 2070 GPUs. Each experiment uses a limited amount of RAM and only a single GPU and can be easily run on a single GPU or even on a CPU-only system. A limitation of our current hypernetwork-based approach is the increase in training time with an increasing number of tasks because an additional regularization term is added for each new task (e.g. HN needs 620 s to learn task 0 and 1609 s to learn task 25 of the LASA dataset, further details in Table A.1). As proposed by von Oswald et al. [5], this can be overcome by using an approximate sample-based regularization instead of the full regularization process. We leave the experimental verification of this solution to future work. A.4. Hyperparameters The authors declare that they have no known competing finan- cial interests or personal relationships that could have appeared to influence the work reported in this paper. The hyperparameters used for the LASA, HelloWorld and Rob- oTasks datasets are listed in Tables A.3, A.4 and A.5 respectively. 14 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Table A.1 Wall clock training times (in seconds) of the models for continual learning from demonstration (median over 5 independent seeds). Table A.1 Wall clock training times (in seconds) of the models for continual learning from demonstration (median over 5 independent seeds). Table A.1 Wall clock training times (in seconds) of the models for continual learning from demonstration (median over 5 independent seeds). independent seeds). (a) LASA (15 × 103 iterations/task) Method Task 0 Task 5 Task 9 Task 13 Task 17 Task 21 Task 25 SG 586 691 714 728 677 721 737 FT 665 722 767 752 673 747 735 REP 792 909 900 913 904 934 936 SI 694 699 707 702 647 713 686 MAS 603 705 753 697 712 723 705 HN 620 1087 1186 1253 1305 1510 1609 CHN 604 991 1119 1257 1384 1624 1743 (b) HelloWorld (40 × 103 iterations/task) Method Task 0 Task 1 Task 2 Task 3 Task 4 Task 5 Task 6 SG 1896 2012 1872 1960 2030 2030 2281 FT 1954 2022 1973 2016 2050 2086 2385 REP 1926 1939 1950 1942 1957 1975 2016 SI 2019 2128 2093 2091 2110 2132 2268 MAS 1965 2281 2252 2302 2316 2362 2500 HN 1968 2337 2508 2702 2843 2930 3234 CHN 1948 2254 2341 2578 2786 2858 3150 (c) RoboTasks-position (50 × 103 iterations/task) Method Task 0 Task 1 Task 2 Task 3 SG 1869 2400 2328 1684 FT 1483 1985 1923 1732 REP 1486 1952 1901 1797 HN 1508 2387 2585 2214 CHN 1591 2484 2662 2141 (d) RoboTasks-orientation (50 × 103 iterations/task) Method Task 0 Task 1 Task 2 Task 3 SG 1809 1800 1800 1783 FT 2026 2068 2061 2058 REP 1396 1409 1409 1409 HN 1461 1903 2116 2303 CHN 1504 1840 2204 2511 (a) LASA (15 × 103 iterations/task) Table A.2 Inference speeds of trained hypernetworks (HN and CHN) in milliseconds and Hz (median over 100 forward passes). A.4. Hyperparameters fh →fθ denotes the generation of a NODE by a hypernetwork, fθ(x) = ˙x denotes a single step taken by the generated NODE, and ∫ fθ(xτ )dτ denotes the generation of an entire trajectory from a NODE by integrating over 1000 steps. Table A.5 Table A.5 Hyperparameters for the RoboTasks dataset. (a) Position (b) Orientation Hyperparameter SG FT REP HN CHN Hyperparameter SG FT REP HN CHN Train iterations 50 000 50 000 50 000 50 000 50 000 Train iterations 50 000 50 000 50 000 50 000 50 000 Learning rate 0.00001 0.00001 0.00001 0.00001 0.00001 Learning rate 0.0001 0.0001 0.0001 0.0001 0.0001 NODE output dim. 3 3 3 3 3 NODE output dim. 3 3 3 3 3 NODE hidden layers [1000] × 3 [1000] × 3 [1000] × 3 [100] × 3 [1000] × 3 NODE hidden layers [1500] × 3 [1500] × 3 [1500] × 3 150,150,150 [1500] × 3 NODE activation elu elu elu elu elu NODE activation elu elu elu elu elu Task Emb. dim. – 512 512 512 512 Task Emb. dim. – 1024 1024 1024 1024 HN hidden layers – – – [200] × 3 [200] × 3 HN hidden layers – – – 300,300,300 300,300,300 β (HN) – – – 0.0005 0.0005 β (HN) – – – 0.0005 0.0005 Chunk Emb. dim. – – – – 512 Chunk Emb. dim. – – – – 1024 Chunk dim. – – – – 8192 Chunk dim. – – – – 8192 (b) Orientation Hyperparameter SG FT REP HN CHN Train iterations 50 000 50 000 50 000 50 000 50 000 Learning rate 0.0001 0.0001 0.0001 0.0001 0.0001 NODE output dim. 3 3 3 3 3 NODE hidden layers [1500] × 3 [1500] × 3 [1500] × 3 150,150,150 [1500] × 3 NODE activation elu elu elu elu elu Task Emb. dim. – 1024 1024 1024 1024 HN hidden layers – – – 300,300,300 300,300,300 β (HN) – – – 0.0005 0.0005 Chunk Emb. dim. – – – – 1024 Chunk dim. – – – – 8192 Fig. A.1. Comparison of NODE with iFlow. (left) DTW errors (lower is better) of all trajectory predictions for all tasks of the LASA dataset. Standalone models of iFlow, NODET, and NODEI are used to learn each of the 26 LASA tasks. Results are obtained using 5 independent seeds. (right) Hyperparameters for iFlow. Hyperparameters for NODE are given in Table A.3. NODEs show better empirical performance and converge faster than iFlow. The regularization hyperparameters for SI (c, ξ), MAS (λ) and HN/CHN (β) are based on values proposed by [4,5,30] respec- tively. Table A.5 Earlier, in Section 5.3 we have shown that HN and CHN are robust to changes in regularization hyperparameters. In all cases, NODEs use the smooth ELU activation function, and hyper- networks have ReLU activations. We use the Adam optimizer in all our experiments. Fig. A.1. Comparison of NODE with iFlow. (left) DTW errors (lower is better) of all trajectory predictions for all tasks of the LASA dataset. Standalone models of iFlow, NODET, and NODEI are used to learn each of the 26 LASA tasks. Results are obtained using 5 independent seeds. (right) Hyperparameters for iFlow. Hyperparameters for NODE are given in Table A.3. NODEs show better empirical performance and converge faster than iFlow. A.5. Imitation flow comparison To compare NODEs against an alternative trajectory learning approach, we performed a small non-continual learning experi- ment using the LASA dataset. For each task, we trained a model of iFlow [9], NODEI, and NODET and measured the DTW errors (Fig. A.1 shows the errors for all tasks taken together). The iFlow hyperparameters are shown in Fig. A.1 (right) and the NODE hyperparameters are given in Table A.3. Although for iFlow we used almost twice the number of training iterations as NODEs, the empirical performance of NODEs is still better. An iFlow model needs around 60 min to be trained to convergence for a single LASA task on our setup, whereas a NODE takes around 10 min. Table A.2 Table A.2 Inference speeds of trained hypernetworks (HN and CHN) in milliseconds and Hz (median over 100 forward passes). fh →fθ denotes the generation of a NODE by a hypernetwork, fθ(x) = ˙x denotes a single step taken by the generated NODE, and ∫ fθ(xτ )dτ denotes the generation of an entire trajectory from a NODE by integrating over 1000 steps. (a) GPU Dataset Method fh →fθ fθ(x) = ˙x ∫ fθ(xτ )dτ (ms) (Hz) (ms) (Hz) (ms) (Hz) LASA HN 0.73 1373.83 0.27 3765.08 584.27 1.71 LASA CHN 0.61 1642.25 0.27 3771.86 569.71 1.76 HelloWorld HN 0.72 1391.15 0.26 3856.83 562.79 1.78 HelloWorld CHN 0.60 1662.43 0.27 3731.59 569.51 1.76 RoboTasks (pos) HN 0.74 1360.24 0.27 3761.71 570.53 1.75 RoboTasks (pos) CHN 0.60 1660.45 0.27 3728.27 570.28 1.75 RoboTasks (ori) HN 0.74 1348.00 0.27 3685.68 595.44 1.68 RoboTasks (ori) CHN 0.63 1597.83 0.28 3548.48 574.78 1.74 (b) CPU Dataset Method fh →fθ fθ(x) = ˙x ∫ fθ(xτ )dτ (ms) (Hz) (ms) (Hz) (ms) (Hz) LASA HN 3.01 332.56 0.12 8422.30 261.13 3.83 LASA CHN 5.23 191.04 0.61 1634.89 730.18 1.37 HelloWorld HN 2.94 339.99 0.12 8456.26 262.61 3.81 HelloWorld CHN 5.29 189.14 0.66 1511.73 750.89 1.33 RoboTasks (pos) HN 2.87 348.16 0.17 5833.52 257.85 3.88 RoboTasks (pos) CHN 5.78 172.94 0.57 1749.81 738.15 1.35 RoboTasks (ori) HN 7.51 133.10 0.15 6579.30 265.23 3.77 RoboTasks (ori) CHN 17.68 56.55 1.22 820.88 1499.81 0.67 15 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Table A.3 Hyperparameters for the LASA dataset. Hyperparameter SG FT REP SI MAS HN CHN Train iterations 15 000 15 000 15 000 15 000 15 000 15 000 15 000 Learning rate 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 NODE output dim. 2 2 2 2 2 2 2 NODE hidden layers [1000] × 3 [1000] × 3 [1000] × 3 [1000] × 3 [1000] × 3 [100] × 3 [1000] × 3 NODE activation elu elu elu elu elu elu elu Task Emb. dim. – 256 256 256 256 256 256 c (SI) – – – 0.3 – – – ξ (SI) – – – 0.3 – – – λ (MAS) – – – – 0.1 – – HN hidden layers – – – – – [200] × 3 [200] × 3 β (HN) – – – – – 0.005 0.005 Chunk Emb. dim. Table A.2 – – – – – – 256 Chunk dim. – – – – – – 8192 Table A.4 Hyperparameters for the HelloWorld dataset. Hyperparameter SG FT REP SI MAS HN CHN Train iterations 40 000 40 000 40 000 40 000 40 000 40 000 40 000 Learning rate 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 NODE output dim. 2 2 2 2 2 2 2 NODE hidden layers [1000] × 3 [1000] × 3 [1000] × 3 [1000] × 3 [1000] × 3 [100] × 3 [1000] × 3 NODE activation elu elu elu elu elu elu elu Task Emb. dim. – 256 256 256 256 256 256 c (SI) – – – 0.3 – – – ξ (SI) – – – 0.3 – – – λ (MAS) – – – – 0.1 – – HN hidden layers – – – – – [200] × 3 [200] × 3 β (HN) – – – – – 0.005 0.005 Chunk Emb. dim. – – – – – – 256 Chunk dim. – – – – – – 8192 A.6. Additional results Figs. A.2 and A.3 report additional trajectory error metrics (Frechet distance and Swept Area error) for the LASA and Hel- loWorld datasets respectively. Fig. A.4 shows a qualitative ex- ample of orientation prediction in terms of quaternions for the RoboTasks dataset. 16 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. A.2. DTW, Frechet distance and Swept Area errors of trajectories predicted for the LASA dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Plots for SG and HN overlap with each other. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The errors for all methods. (bottom) A zoomed-in view of the methods that perform well. S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. A.2. DTW, Frechet distance and Swept Area errors of trajectories predicted for the LASA dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Plots for SG and HN overlap with each other. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The errors for all methods. (bottom) A zoomed-in view of the methods that perform well. Fig. A.3. DTW, Frechet distance and Swept Area errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The errors for all methods. (bottom) A zoomed-in view of the methods that perform well. Fig. A.3. DTW, Frechet distance and Swept Area errors of trajectories predicted for the HelloWorld dataset (lower is better). The x-axis shows the current task. After learning a task (using NODET), all current and previous tasks are evaluated. Lines show medians and shaded regions denote the lower and upper quartiles of the errors over 5 independent seeds. (top) The errors for all methods. (bottom) A zoomed-in view of the methods that perform well. Fig. A.4. A.7. Robustness to starting position of the demonstration. The results are shown in Fig. A.5. In the ideal case, the end position ∆would always be 0 irrespective of the start position ∆(every trajectory, irrespective of the starting position would always converge at the goal). In the absence of any stability guarantees, if the starting positions of the robot are very different from the ground truth, the predicted trajectory can be expected to diverge far from the ground truth goal, leading to a large end position ∆. To test the robustness of the different CL methods to changes in the initial positions of the robot (i.e., when the robot starts at a position that is different from the start position of the demon- strated trajectory), we perform an analysis using the RoboTasks dataset. For each of the four tasks of this dataset, we randomly select one ground truth demonstration. We then create a sphere with a radius of 20 cm (the maximum reach of the Franka Emika Panda robot is 85.5 cm) centered at the starting position of the ground truth trajectory. From within the volume of this sphere, we uniformly sample 100 random starting points. We then take the trained models of SG and CHN (trained on all the tasks of the dataset), and use these to predict trajectories for each of the 4 tasks and each of the 100 random starting points with the same timesteps as the ground truth demonstration. By sampling points from a sphere of radius 20 cm, we are asking the models to extrapolate from the demonstrations and start from areas for which no demonstration has been provided. g p It can be seen that for all tasks, when the random starting position is up to 6 cm from the GT source, the CHN model pro- duces trajectories that end up very close to the GT goal. This is not true for SG. Although SG performs almost perfectly when starting from the GT source position (see the × symbols in column (b) of Fig. A.5, and also Figs. 14 and 15), it produces trajectories that diverge far from the GT goal even when the random starting positions are quite close to the GT source position. A.6. Additional results Example of end-effector orientation trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. The ground truth demonstrations as well as the predictions of the orientation consist of a sequence of unit quaternions. Columns show the predictions for each task, and rows show the different elements of the quaternions (scalar first convention). Fig. A.4. Example of end-effector orientation trajectories predicted for all past tasks after sequentially learning all the tasks of the RoboTasks dataset. All methods employ NODET as the trajectory learning method. The ground truth demonstrations as well as the predictions of the orientation consist of a sequence of unit quaternions. Columns show the predictions for each task, and rows show the different elements of the quaternions (scalar first convention). 17 Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Fig. A.5. Robustness of SG and CHN on the RoboTasks dataset. Each row shows the results for 1 of the 4 tasks of this dataset. Columns (a,b) show that for small values of start position ∆, SG’s trajectories often diverge far from the goal position but CHN’s trajectories do not. For larger values of start position ∆, CHN gets worse but much less than SG. In column (a), values greater than end position ∆= 100 cm are grouped together and shown in the shaded region marked ‘‘Large’’. Column (b) shows the divergence of SG and CHN by grouping the results based on the start position ∆. Here, the symbol × shows the value of end position ∆when start position ∆= 0. Column (c) shows overall divergence of SG and CHN for novel initial positions, where CHN shows much less divergence. All models use NODE-T for learning. Fig. A.5. Robustness of SG and CHN on the RoboTasks dataset. Each row shows the results for 1 of the 4 tasks of this dataset. Columns (a,b) show that for small values of start position ∆, SG’s trajectories often diverge far from the goal position but CHN’s trajectories do not. For larger values of start position ∆, CHN gets worse but much less than SG. In column (a), values greater than end position ∆= 100 cm are grouped together and shown in the shaded region marked ‘‘Large’’. A.6. Additional results Column (b) shows the divergence of SG and CHN by grouping the results based on the start position ∆. Here, the symbol × shows the value of end position ∆when start position ∆= 0. Column (c) shows overall divergence of SG and CHN for novel initial positions, where CHN shows much less divergence. All models use NODE-T for learning. A.7. Robustness to starting position For larger start position differences, the performance of CHN does degrade, but much less than that for SG, which most of the time produces trajectories that end up far beyond the reachable task space of the robot (in the scatter plots we club together all predictions that end up more than 100 cm from the goal). This shows that the continual learning process for CHN has a regularization ef- fect which prevents the strong overfitting to demonstrations as exhibited by SG. For each predicted trajectory, we measure (i) start position ∆: the distance between the random starting position and the start- ing position of the demonstration; (ii) end position ∆: the distance between the end point of the predicted trajectory (that starts from a random starting position) and the end position (goal) 18 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Fig. A.6. Robustness to task order during training. The models for SG, REP, CHN and HN are trained and evaluated with 3 different task orders for each dataset (each column of plots shows the result of 1 order). Overall the same trends are visible for all task orders. For LASA, HN is almost identical to SG, and CHN performs well for a small number of tasks but degrades afterwards. For HelloWorld, HN and CHN perform as well as the other baselines. For RoboTasks, again HN and CHN perform comparably to the other baselines. This shows that the continual learning performance is not affected by the order in which tasks are learned. The x-axis in the plots denotes the task number. To see the name of a task from a given order, please refer to Table A.6 (e.g., for LASA, task 0 from Order 0 is ‘Angle’, while task 0 from Order 1 is ‘Spoon’). Also, note that the DTW metric is dependent on the dimension of the trajectory points. Thus, the DTW values for a 2D dataset such as HelloWorld are lower than those for RoboTasks. In all cases, the upper baseline SG provides a reference for comparison. For each model, each dataset and each task order, we train and evaluate with 3 independent seeds. Similar to Figs. A.8. Robustness to task order A.6), the perfor- mance of HN is almost identical to that of the upper baseline SG, while REP and CHN perform worse for all orders. CHN performs well up to a small number of tasks but its performance degrades as more tasks are learned. We train and evaluate models for SG, REP, CHN, and HN (as they perform the best) each with 3 independent seeds on each task order of each dataset. The results for this evaluation are shown in Fig. A.6. Although the results are marginally different for the different task orders, the overall trends are similar across task orders for the LASA, HelloWorld and RoboTasks datasets. For LASA (see the first row of plots in Fig. A.6), the perfor- mance of HN is almost identical to that of the upper baseline SG, while REP and CHN perform worse for all orders. CHN performs well up to a small number of tasks but its performance degrades as more tasks are learned. A.7. Robustness to starting position 4, 9 and 14, in this plot lines show median values and shaded areas represent the IQR of the errors (lower is better). S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (202 Fig. A.6. Robustness to task order during training. The models for SG, REP, CHN and HN are trained and evaluated with 3 different task orders for ea (each column of plots shows the result of 1 order). Overall the same trends are visible for all task orders. For LASA, HN is almost identical to SG, and CHN well for a small number of tasks but degrades afterwards. For HelloWorld, HN and CHN perform as well as the other baselines. For RoboTasks, again HN Fig. A.6. Robustness to task order during training. The models for SG, REP, CHN and HN are trained and evaluated with 3 different task orders for each dataset (each column of plots shows the result of 1 order). Overall the same trends are visible for all task orders. For LASA, HN is almost identical to SG, and CHN performs well for a small number of tasks but degrades afterwards. For HelloWorld, HN and CHN perform as well as the other baselines. For RoboTasks, again HN and CHN perform comparably to the other baselines. This shows that the continual learning performance is not affected by the order in which tasks are learned. The x-axis in the plots denotes the task number. To see the name of a task from a given order, please refer to Table A.6 (e.g., for LASA, task 0 from Order 0 is ‘Angle’, while task 0 from Order 1 is ‘Spoon’). Also, note that the DTW metric is dependent on the dimension of the trajectory points. Thus, the DTW values for a 2D dataset such as HelloWorld are lower than those for RoboTasks. In all cases, the upper baseline SG provides a reference for comparison. For each model, each dataset and each task order, we train and evaluate with 3 independent seeds. Similar to Figs. 4, 9 and 14, in this plot lines show median values and shaded areas represent the IQR of the errors (lower is better). A.8. Robustness to task order The results for HelloWorld (second row of plots in Fig. A.6) also show similar trends for all 3 task orders: CHN performs similarly to SG, REP and HN, in spite of being much smaller in size than SG and HN, and without needing to store training data of past tasks like REP. The results for HelloWorld (second row of plots in Fig. A.6) also show similar trends for all 3 task orders: CHN performs similarly to SG, REP and HN, in spite of being much smaller in size than SG and HN, and without needing to store training data of past tasks like REP. We test the robustness of the best performing models (SG, REP, CHN and HN) against changes in the order in which tasks are presented during training. For each of the 3 datasets used in this paper, we create 3 different orders of tasks: Order 0, Order 1, and Order 2. The first order (Order 0) is the same as that used for the results reported in Section 5.3. The next 2 task orders for each dataset are created by randomly shuffling the first task order of the respective datasets. The task orders and the corresponding data file names used during training are listed in Table A.6. Similarly, for RoboTasks (last two rows of plots in Fig. A.6), the task order does not seem to affect the continual learning performance. Learning a difficult task is equally challenging for all the methods (e.g. the plateslyingtostanding task, which is task 2 in Order 0 and task 0 in Order 1, appears to be a difficult task judging from the relatively higher errors). However this does not affect the relative performance of the different methods. CHN performs almost equally well as the other methods, although other methods have more parameters (SG, HN) or store past training data (REP). We train and evaluate models for SG, REP, CHN, and HN (as they perform the best) each with 3 independent seeds on each task order of each dataset. The results for this evaluation are shown in Fig. A.6. Although the results are marginally different for the different task orders, the overall trends are similar across task orders for the LASA, HelloWorld and RoboTasks datasets. For LASA (see the first row of plots in Fig. References [28] A.A. Rusu, N.C. Rabinowitz, G. Desjardins, H. Soyer, J. Kirkpatrick, K. Kavukcuoglu, R. Pascanu, R. Hadsell, Progressive neural networks, 2016, arXiv preprint arXiv:1606.04671. [1] C. Gao, H. Gao, S. Guo, T. Zhang, F. Chen, CRIL: Continual robot imitation learning via generative and prediction model, in: 2021 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), 2021, pp. 6747–5754, http://dx.doi.org/10.1109/IROS51168.2021.9636069. [1] C. Gao, H. Gao, S. Guo, T. Zhang, F. Chen, CRIL: Continual robot imitation learning via generative and prediction model, in: 2021 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), 2021, pp. 6747–5754, http://dx.doi.org/10.1109/IROS51168.2021.9636069. [29] S.-A. Rebuffi, A. Kolesnikov, G. Sperl, C.H. Lampert, Icarl: Incremental clas- sifier and representation learning, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 2001–2010. [2] Y. Huang, K. Xie, H. Bharadhwaj, F. Shkurti, Continual model-based re- inforcement learning with hypernetworks, in: 2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 799–805. [30] F. Zenke, B. Poole, S. Ganguli, Continual learning through synaptic intelli- gence, in: International Conference on Machine Learning, PMLR, 2017, pp. 3987–3995. [3] H. Shin, J.K. Lee, J. Kim, J. Kim, Continual learning with deep generative replay, in: Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017, pp. 2994–3003. [31] M. Delange, R. Aljundi, M. Masana, S. Parisot, X. Jia, A. Leonardis, G. Slabaugh, T. Tuytelaars, A continual learning survey: Defying forgetting in classification tasks, IEEE Trans. Pattern Anal. Mach. Intell. (2021). [32] A. Xie, C. Finn, Lifelong robotic reinforcement learning by retaining experiences, 2021, arXiv preprint arXiv:2109.09180. [4] R. Aljundi, F. Babiloni, M. Elhoseiny, M. Rohrbach, T. Tuytelaars, Memory aware synapses: Learning what (not) to forget, in: Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 139–154. [33] M. Heinonen, C. Yildiz, H. Mannerström, J. Intosalmi, H. Lähdesmäki, Learning unknown ODE models with Gaussian processes, in: International Conference on Machine Learning, PMLR, 2018, pp. 1959–1968. [5] J. von Oswald, C. Henning, J. Sacramento, B.F. Grewe, Continual learn- ing with hypernetworks, in: International Conference on Learning Representations (ICLR), 2019. [34] Y. Huang, F.J. Abu-Dakka, J. Silvério, D.G. Caldwell, Toward orientation learning and adaptation in cartesian space, IEEE Trans. Robot. 37 (1) (2020) 82–98. [6] A. Billard, S. Calinon, R. Dillmann, Learning from humans, in: Springer Handbook of Robotics, second ed., 2016. [35] A. Ude, B. Nemec, T. Petrić, J. Appendix B. Supplementary data Supplementary material related to this article can be found online at https://doi.org/10.1016/j.robot.2023.104427. 19 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Table A.6 Task orders. (a) LASA Task ID Order 0 file Order 1 file Order 2 file 0 Angle Spoon BendedLine 1 BendedLine Sine RShape 2 CShape Zshape Trapezoid 3 DoubleBendedLine NShape WShape 4 GShape PShape PShape 5 heee heee heee 6 JShape_2 Angle Sshape 7 JShape Khamesh Line 8 Khamesh Line JShape 9 Leaf_1 LShape DoubleBendedLine 10 Leaf_2 JShape_2 Sine 11 Line WShape GShape 12 LShape Leaf_2 Sharpc 13 NShape RShape LShape 14 PShape JShape Saeghe 15 RShape Sharpc Khamesh 16 Saeghe Snake CShape 17 Sharpc DoubleBendedLine Zshape 18 Sine Saeghe JShape_2 19 Snake Worm Snake 20 Spoon Leaf_1 Angle 21 Sshape GShape Worm 22 Trapezoid BendedLine Leaf_2 23 Worm Sshape Spoon 24 WShape CShape NShape 25 Zshape Trapezoid Leaf_1 (b) HelloWorld Task ID Order 0 file Order 1 file Order 2 file 0 h l o 1 e d w 2 l h l 3 o o h 4 w w e 5 r r d 6 d e r (c) RoboTasks Task ID Order 0 file Order 1 file Order 2 file 0 openbox platestandingtolying pouring 1 bottle2shelf openbox openbox 2 platestandingtolying bottle2shelf platestandingtolying 3 pouring pouring bottle2shelf International Conference on Intelligent Robots and Systems (IROS), IEEE, 2020, pp. 5231–5237. International Conference on Intelligent Robots and Systems (IROS), IEEE, 2020, pp. 5231–5237. [10] J.Z. Kolter, G. Manek, Learning stable deep dynamics models, Advances in Neural Information Processing Systems 32 (2019) 11128–11136. [11] A.J. Ijspeert, J. Nakanishi, S. Schaal, Movement imitation with nonlinear dynamical systems in humanoid robots, in: International Conference on Robotics and Automation (ICRA), 2002, pp. 1398–1403. [12] M. Saveriano, F.J. Abu-Dakka, A. Kramberger, L. Peternel, Dynamic move- ment primitives in robotics: A tutorial survey, 2021, arXiv preprint arXiv: 2102.03861. [13] D. Ha, A.M. Dai, Q.V. Le, HyperNetworks, in: International Conference on Learning Representations, 2017, URL https://openreview.net/forum?id= rkpACe1lx. [14] R.T. Chen, Y. Rubanova, J. Bettencourt, D. Duvenaud, Neural ordinary differential equations, in: Proceedings of the 32nd International Conference on Neural Information Processing Systems, 2018, pp. 6572–6583. [15] G.I. Parisi, R. Kemker, J.L. Part, C. Kanan, S. Wermter, Continual lifelong learning with neural networks: A review, Neural Netw. 113 (2019) 54–71. [16] S. Calinon, Learning from demonstration (programming by demonstration), Encyclopedia Robot. (2018) 1–8. [17] B.D. Argall, S. Appendix B. Supplementary data Chernova, M. Veloso, B. Browning, A survey of robot learning from demonstration, Robot. Auton. Syst. 57 (5) (2009) 469–483. [18] H. Ravichandar, A. Polydoros, S. Chernova, A. Billard, Robot learning from demonstration: A review of recent advances, Annu. Rev. Control Robot. Auton. Syst. (2019). [19] S.R. Ahmadzadeh, S. Chernova, Trajectory-based skill learning using generalized cylinders, Front. Robot. AI 5 (2018) 132. [20] Y. Wu, Y. Demiris, Towards one shot learning by imitation for hu- manoid robots, in: 2010 IEEE International Conference on Robotics and Automation, IEEE, 2010, pp. 2889–2894. [21] B.D. Argall, B. Browning, M.M. Veloso, Teacher feedback to scaffold and refine demonstrated motion primitives on a mobile robot, Robot. Auton. Syst. 59 (3–4) (2011) 243–255. [22] S. Calinon, P. Kormushev, D.G. Caldwell, Compliant skills acquisition and multi-optima policy search with EM-based reinforcement learning, Robot. Auton. Syst. 61 (4) (2013) 369–379. [23] N. Das, S. Bechtle, T. Davchev, D. Jayaraman, A. Rai, F. Meier, Model-based inverse reinforcement learning from visual demonstrations, in: Conference on Robot Learning, PMLR, 2021, pp. 1930–1942. [24] P. Englert, N.A. Vien, M. Toussaint, Inverse KKT: Learning cost functions of manipulation tasks from demonstrations, Int. J. Robot. Res. 36 (13–14) (2017) 1474–1488. [25] A. Kalinowska, A. Prabhakar, K. Fitzsimons, T. Murphey, Ergodic imitation: Learning from what to do and what not to do, in: 2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 3648–3654. [26] S.M. Khansari-Zadeh, A. Billard, Learning control Lyapunov function to ensure stability of dynamical system-based robot reaching motions, Robot. Auton. Syst. 62 (6) (2014) 752–765. [27] M. Saveriano, An energy-based approach to ensure the stability of learned dynamical systems, in: IEEE International Conference on Robotics and Automation (ICRA), 2020, pp. 4407–4413. References Maltoni, Don’t forget, there is more than forgetting: new metrics for Continual Learning, 2018, arXiv preprint arXiv:1810.13166. [43] R. Rahmatizadeh, P. Abolghasemi, L. Bölöni, S. Levine, Vision-based multi- task manipulation for inexpensive robots using end-to-end learning from demonstration, in: 2018 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2018, pp. 3758–3765. Antonio José Rodríguez Sánchez is currently an As- sociate Professor in the Intelligent and Interactive Systems group of the department of Computer Science at the Universität Innsbruck (Austria). He completed his Ph.D. at the Center for Vision Research (York University, Toronto, Canada) on modeling attention and intermedi- ate areas of the visual cortex under the supervision of Prof. John K. Tsotsos (Canada Research Chair) in 2010. He obtained the degree of M.Sc. in Computer Science at the Universidade da Coruña (Spain) in 1998. He received his B.Sc. in Computer Science at Universidad de Córdoba (Spain) with Honors and did his Bachelor Thesis at the Université de La Rochelle (France). His current research interests include different areas of artificial intelligence: Explainable AI, computational neuroscience, (deep) neural networks, computer vision, machine learning and robotics. Sayantan Auddy is a Ph.D. student in the Intelli- gent and Interactive Systems group of the department of Computer Science at the University of Innsbruck (Austria). He received his M.Sc. degree in Intelligent Adaptive Systems from the University of Hamburg (Germany) in 2018, and his B.Tech. degree in Com- puter Science and Engineering from Haldia Institute of Technology (India) in 2009. Prior to starting his masters studies, he has worked as a Software Engineer in the Information Technology industry for 5 years. His research interests include continual learning for robotics, reinforcement learning and deep learning. Justus Piater is a professor of computer science at the University of Innsbruck, Austria, where he leads the Intelligent and Interactive Systems group and serves as the founding director of the interdisciplinary Digital Science Center. He earned his Ph.D. degree at the University of Massachusetts Amherst, USA, and was a professor of computer science at the University of Liège, Belgium, and a visiting researcher at the Max Planck Institute for Biological Cybernetics in Tübingen, Germany. His research interests focus on learning and inference in sensorimotor systems. He has published papers, several of which have received best-paper awards. References Morimoto, Orientation in cartesian space dynamic movement primitives, in: 2014 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2014, pp. 2997–3004. [7] M. Hersch, F. Guenter, S. Calinon, A. Billard, Dynamical system modulation for robot learning via kinesthetic demonstrations, IEEE Trans. Robot. 24 (6) (2008) 1463–1467. [8] S.M. Khansari-Zadeh, A. Billard, Learning stable nonlinear dynamical sys- tems with Gaussian mixture models, IEEE Trans. Robot. 27 (5) (2011) 943–957. [36] M. Saveriano, F. Franzel, D. Lee, Merging position and orientation motion primitives, in: 2019 International Conference on Robotics and Automation (ICRA), IEEE, 2019, pp. 7041–7047. [9] J. Urain, M. Ginesi, D. Tateo, J. Peters, ImitationFlow: Learning deep stable stochastic dynamic systems by normalizing flows, in: 2020 IEEE/RSJ [37] H.C. Ravichandar, I. Salehi, A.P. Dani, Learning partially contracting dynamical systems from demonstrations, in: CoRL, 2017, pp. 369–378. 20 S. Auddy, J. Hollenstein, M. Saveriano et al. Robotics and Autonomous Systems 165 (2023) 104427 Robotics and Autonomous Systems 165 (2023) 104427 [38] C. Blocher, M. Saveriano, D. Lee, Learning stable dynamical systems using contraction theory, in: 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI), IEEE, 2017, pp. 124–129. Matteo Saveriano received his B.Sc. and M.Sc. de- gree in automatic control engineering from University of Naples, Italy, in 2008 and 2011, respectively. He received his Ph.D. from the Technical University of Munich in 2017. Currently, he is an assistant profes- sor at the Department of Industrial Engineering (DII), University of Trento, Italy. Previously, he was an as- sistant professor at the University of Innsbruck and a post-doctoral researcher at the German Aerospace Center (DLR). He is an Associate Editor for RA-L. His research activities include robot learning, human–robot interaction, understanding and interpreting human activities. Webpage: https: //matteosaveriano.weebly.com/ [39] S. Dasari, F. Ebert, S. Tian, S. Nair, B. Bucher, K. Schmeckpeper, S. Singh, S. Levine, C. Finn, RoboNet: Large-scale multi-robot learning, 2019, http://dx. doi.org/10.48550/ARXIV.1910.11215, URL https://arxiv.org/abs/1910.11215. [40] T. Yu, D. Quillen, Z. He, R. Julian, A. Narayan, H. Shively, A. Bellathur, K. Hausman, C. Finn, S. Levine, Meta-world: A benchmark and evaluation for multi-task and meta reinforcement learning, 2019, http://dx.doi.org/10. 48550/ARXIV.1910.10897, URL https://arxiv.org/abs/1910.10897. [41] C.F. Jekel, G. Venter, M.P. Venter, N. Stander, R.T. Haftka, Similarity measures for identifying material parameters from hysteresis loops using inverse analysis, Int. J. Mater. Form. 12 (3) (2019) 355–378. [42] N. Díaz-Rodríguez, V. Lomonaco, D. Filliat, D. References Justus Piater is a professor of computer science at the University of Innsbruck, Austria, where he leads the Intelligent and Interactive Systems group and serves as the founding director of the interdisciplinary Digital Science Center. He earned his Ph.D. degree at the University of Massachusetts Amherst, USA, and was a professor of computer science at the University of Liège, Belgium, and a visiting researcher at the Max Planck Institute for Biological Cybernetics in Tübingen, Germany. His research interests focus on learning and inference in sensorimotor systems. He has published more than 200 papers, several of which have received best-paper awards. Jakob Hollenstein is a Ph.D. student in the Intelli- gent and Interactive Systems group of the department of Computer Science at the University of Innsbruck (Austria). He received his B.Sc. degree in Computer Engineering and his Dipl.-Ing. degree in Computational Intelligence from the Vienna University of Technology (Austria) in 2010 and 2014. He worked as a software engineer in the Information Technology industry for 3 years before joining the Intelligent Interactive Systems group. His research interests include robot learning, deep learning and reinforcement learning. 21
https://openalex.org/W2685648398	https://f1000research.com/articles/6-999/v1/pdf	English	null	2017 ISCB Overton Prize: Christoph Bock	F1000Research	2,017	cc-by	1,748	2017 ISCB Overton Prize: Christoph Bock [version 1; peer review: not peer reviewed] 1Freelance Writer, Kensington, MD, USA First published: 26 Jun 2017, 6(ISCB Comm J):999 https://doi.org/10.12688/f1000research.11586.1 Latest published: 26 Jun 2017, 6(ISCB Comm J):999 https://doi.org/10.12688/f1000research.11586.1 v1 Abstract The International Society for Computational Biology (ISCB) each year recognizes the achievements of an early to mid-career scientist with the Overton Prize. This prize honors the untimely death of Dr. G. Christian Overton, an admired computational biologist and founding ISCB Board member. Winners of the Overton Prize are independent investigators who are in the early to middle phases of their careers and are selected because of their significant contributions to computational biology through research, teaching, and service. ISCB is pleased to recognize Dr. Christoph Bock, Principal Investigator at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences in Vienna, Austria, as the 2017 winner of the Overton Prize. Bock will be presenting a keynote presentation at the 2017 International Conference on Intelligent Systems for Molecular Biology/European Conference on Computational Biology (ISMB/ECCB) in Prague, Czech Republic being held during July 21-25, 2017. Keywords Editorial F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 Keywords Editorial This article is included in the Bioinformatics gateway. This article is included in the Bioinformatics gateway. Page 1 of 5 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 Corresponding author: Diane E. Kovats (dkovats@iscb.org) Competing interests: No competing interests were disclosed. Grant information: The author(s) declared that no grants were involved in supporting this work. Copyright: © 2017 Fogg CN 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 work is properly cited. How to cite this article: Fogg CN, Kovats DE and Berger B. 2017 ISCB Overton Prize: Christoph Bock [version 1; peer review: not peer reviewed] F1000Research 2017, 6(ISCB Comm J):999 https://doi.org/10.12688/f1000research.11586.1 First published: 26 Jun 2017, 6(ISCB Comm J):999 https://doi.org/10.12688/f1000research.11586.1 Page 2 of 5 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 Christoph Bock: At home in the epigenome Christoph Bock: At home in the epigenome Christoph Bock’s scientific curiosity was nurtured from a young age. His parents were math and science teachers, and while they did not push him to pursue these areas of study, he sees how this intel- lectually stimulating environment cultivated his natural curiosity and provided a critical foundation to his career as a scientist. Bock started exploring computer programming from the age of twelve, and he realizes in retrospect how learning to code was a valuable tool for practicing problem solving and scientific thinking. During high school, Bock specialized in physics and maths. His undergraduate studies at the University of Mannheim focused on computer science and business information systems, emphasizing machine learning and artificial intelligence. Toward the end of his studies, Bock yearned to tackle questions with broader relevance than the “toy problems” he encountered in his course work. Bock recalled, “Human biology seemed the biggest challenge and also most societally relevant. I was lucky that Jürgen Hesser offered a bioinformatics lecture and agreed to supervise my Master’s thesis at the University of Mannheim”. His Master’s research work focused on protein structure prediction and homology modeling. Bock pursued his PhD studies in bioinformatics under the supervi- sion of Thomas Lengauer at the Max Planck Institute for Informat- ics, studying epigenetic regulation of the genome. Keywords Editorial “Moving into bioinformatics and epigenetics, I had to catch up on a lot of impor- tant biological knowledge”, Bock recalled. “Reading papers and collaborating was key, but it also helped that my research focused on a field that was quite young, with ample opportunity to try out something new.” Bock developed several software tools as part of his PhD, includ- ing BiQ Analyzer for processing DNA methylation data and Epi- GRAPH for analyzing and predicting epigenome profiles in their genomic context. Bock went on to pursue postdoctoral studies under Alexander Meissner at the Broad Institute. There, Bock was exposed to the world of wet-lab biology, and he discovered the thrill and power of jointly developing new laboratory techniques and computational methods, which he used to study the epigenome of pluripotent and hematopoietic stem cells. Bock developed several software tools as part of his PhD, includ- ing BiQ Analyzer for processing DNA methylation data and Epi- GRAPH for analyzing and predicting epigenome profiles in their genomic context. Bock went on to pursue postdoctoral studies under Alexander Meissner at the Broad Institute. There, Bock was exposed to the world of wet-lab biology, and he discovered the thrill and power of jointly developing new laboratory techniques and computational methods, which he used to study the epigenome of pluripotent and hematopoietic stem cells. He attributes much of his bioinformatics training to the time spent in the research group of Thomas Lengauer, and he has been grate- ful for his mentor’s continued support and collaboration throughout his early career. Bock also acknowledges the important guidance and feedback on his research provided by Jörn Walter, who co-supervised his PhD dissertation and introduced Bock to the international epigenetics community. In 2012, Bock started his own research group at CeMM, an insti- tute dedicated to advancing precision medicine through basic and translational research. He was hired by Giulio Superti-Furga, Scientific Director of CeMM, who, as Bock said, “Provided ample encouragement and let me try things that were initially quite far outside of my comfort zone, such as starting a wet lab and leading a next generation sequencing technology platform.” Bock has thrived at CeMM, where he has been able to work with many passionate researchers within the institute and at the neighboring Medical University of Vienna. Keywords Editorial Bock’s first encounter with epigenetics data transformed his scien- tific career path, and he has been one of the first bioinformaticians that dedicated their work to epigenetic data. “When I started my PhD studies in 2004, the largest epigenetic dataset consisted of just over 100 data points, and one of my first papers established epige- nome prediction as a means of inferring what was still very difficult and costly to measure experimentally.” At CeMM, Bock has also developed his personal style of being a PI and mentor, acting as a catalyst of ideas and projects for an interdis- ciplinary team. He explained, “Our lab combines computational and wet-lab biology on roughly equal terms, with a good dose of tech- nology development – including single-cell sequencing, CRISPR, epigenome editing, machine learning, and more. There is also an extensive network of collaborations, ranging from fundamental biology to immediate clinical applications in the area of personal- ized and precision medicine. It is a great privilege to work with such an interdisciplinary and creative group of smart people.” In the following years, next generation sequencing transformed the field, and it became possible to collect several billion data points in a single epigenome mapping experiment. This development created a strong demand for bioinformatic methods. “Working at the forefront of the epigenome revolution has been the highlight of my scientific research so far. But the most exciting times may still be ahead as epigenome research is starting to become broadly relevant for medicine, and I am looking forward to contributing to this development.” Page 3 of 5 Page 3 of 5 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 Bock considers the success of his students and postdocs as a key measure of his achievement as a PI. He explained, “I work hard to maintain an environment in which every group member can build a great CV and learns what he or she needs to advance in their scientific career. So far, we have a 100% success rate of postdocs moving on to attractive PI jobs, which is great for young lab. Keywords Editorial But it is clear that helping others succeed in their career is not an easy task, and you need to create room for success and failure, and a safety net that encourages risk taking.” Bock is deeply gratified to be honored with the Overton Prize, especially since he will receive his award this year in Prague. He said, “Ten years ago, I attended ISMB 2007 in Vienna – one of the first conferences where I presented my PhD project on epigenome prediction. That year, Eran Segal won the Overton Prize, and his keynote lecture about DNA’s regulatory code reinforced my inter- est in understanding the role of epigenome regulation in biology and medicine. ISMB 2007 was also my first time in Vienna, and the great impressions from that visit surely contributed to the fact that a job ad from Vienna caught my attention a few years later. This year, it will be my pleasure to give the Overton Prize lecture at ISMB 2017 in Prague, ten years and just a few hundred kilometers away from a truly career-defining ISMB 2007.” Bock is still excited about epigenetics and what it can teach us about a cell’s past, present and future. He hopes that epigenomic data can be used to understand the regulatory logic of cells and to determine what goes awry in diseases like cancer. Bock said, “We are pursuing an engineering-inspired ‘build it to understand it’ approach to cancer biology, where we combine CRISPR epigenome editing and computationally designed drug combinations to ration- ally reprogram normal cells into cancer cells and vice versa. Build- ing upon a breakthrough technology for pooled CRISPR screening with single-cell sequencing, we seek to decipher complex biologi- cal pathways and gene regulatory networks in high throughput, in order to overcome the classical ‘one gene, one postdoc’ paradigm of functional (epi-)genomics.” Competing interests No competing interests were disclosed. Grant information The author(s) declared that no grants were involved in supporting this work. Grant information Page 4 of 5 F1000Research 2017, 6(ISCB Comm J):999 Last updated: 12 MAR 2024 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 5 of 5
https://openalex.org/W3136494418	https://digibug.ugr.es/bitstream/10481/68690/1/ijerph-18-03211-v3.pdf	English	null	Is It Possible to Be Happy during the COVID-19 Lockdown? A Longitudinal Study of the Role of Emotional Regulation Strategies and Pleasant Activities in Happiness	International journal of environmental research and public health/International journal of environmental research and public health	2,021	cc-by	7,714	Citation: Gutiérrez-Cobo, M.J.; Megías-Robles, A.; Gómez-Leal, R.; Cabello, R.; Fernández-Berrocal, P. Is It Possible to Be Happy during the COVID-19 Lockdown? A Longitudinal Study of the Role of Emotional Regulation Strategies and Pleasant Activities in Happiness. Int. J. Environ. Res. Public Health 2021, 18, 3211. https://doi.org/10.3390/ ijerph18063211 Keywords: cognitive reappraisal; expressive suppression; pleasant activities; happiness; COVID-19 Academic Editors: Fraser Carson and Paul B. Tchounwou Received: 19 February 2021 Accepted: 18 March 2021 Published: 19 March 2021 é Gutiérrez-Cobo 1,* , Alberto Megías-Robles 2 , Raquel Gómez-Leal 2, Rosario Cabello 3 1 Department of Developmental and Educational Psychology, Faculty of Psychology, University of Málaga, 29071 Málaga, Spain 2 Department of Basic Psychology, Faculty of Psychology, University of Málaga, 29071 Málaga, Spain; amegias@uma.es (A.M.-R.); raqgomlea@uma.es (R.G.-L.); berrocal@uma.es (P.F.-B.) 3 Department of Developmental and Educational Psychology, Faculty of Psychology, University of Granada, 18071 Granada, Spain; rcabello@ugr.es * Correspondence: mjgc@uma.es 1 Department of Developmental and Educational Psychology, Faculty of Psychology, University of Málag 29071 Málaga, Spain 1 Department of Developmental and Educational Psychology, Faculty of Psychology, University of Málaga, 29071 Málaga, Spain 2 Department of Basic Psychology, Faculty of Psychology, University of Málaga, 29071 Málaga, Spain; amegias@uma.es (A.M.-R.); raqgomlea@uma.es (R.G.-L.); berrocal@uma.es (P.F.-B.) 3 Department of Developmental and Educational Psychology, Faculty of Psychology, University of Granada, 18071 Granada, Spain; rcabello@ugr.es * Correspondence: mjgc@uma es Abstract: This study aimed to longitudinally analyze the role played by two emotional regulation strategies (cognitive reappraisal and expressive suppression), through the mediating effect of en- gagement in pleasant activities during lockdown, in changes in affective and cognitive happiness in comparison with pre-pandemic levels. Eighty-eight participants from a community sample were evaluated at two timepoints. At timepoint 1 (before the COVID-19 pandemic), participants were evaluated on emotional regulation and cognitive and affective happiness. At timepoint 2 (during the COVID-19 lockdown), participants were evaluated on cognitive and affective happiness and the frequency with which they engaged in pleasant activities. We found an optimal ﬁt of the proposed model in which cognitive reappraisal was signiﬁcantly related to engagement in more pleasant activities during the lockdown. In turn, these pleasant activities were related to more affective happiness during the lockdown (compared with pre-pandemic levels), and this affective happiness was associated with greater cognitive happiness. In conclusion, cognitive reappraisal was a protective factor for affective and cognitive happiness through the mediating role of engagement in pleasant activities during lockdown. Limitations and future lines of investigation are discussed. Int. J. Environ. Res. Public Health 2021, 18, 3211. https://doi.org/10.3390/ijerph18063211 International Journal of Environmental Research and Public Health International Journal of Environmental Research and Public Health Is It Possible to Be Happy during the COVID-19 Lockdown? A Longitudinal Study of the Role of Emotional Regulation Strategies and Pleasant Activities in Happiness María José Gutiérrez-Cobo 1,* , Alberto Megías-Robles 2 , Raquel Gómez-Leal 2, Rosario Cabello 3 and Pablo Fernández-Berrocal 2 1. Introduction Since the worldwide emergence of COVID-19, an extensive body of research has been conducted to clarify the impact of the pandemic on mental health and well-being [1–5]. The results of all these studies seem to indicate that there has been an overall decline in the mental health of the general population, with high percentages of depression, anxiety, and stress, particularly in risk populations. Of particular relevance are the consequences of the special measures that the authorities put in place during the lockdown periods where the freedom of citizens was severely restricted [6,7]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. y The previous literature on the mental health consequences of COVID-19 has mainly employed cross-sectional methodologies. These methodologies, however, do not allow for capturing changes in the psychological impact and predictors of the COVID-19 lockdown. Therefore, longitudinal methodologies should be used to address this gap in order to identify the real changes that have occurred during the pandemic. In addition, previous studies have focused on psychopathology, with rather less attention being paid to positive psychology and the protective factors of people’s happiness during the COVID-19 lock- down. Happiness is understood as a balance between pleasant and unpleasant experiences, with many of the former and relatively few of the latter [8]. A happy life is dependent on Copyright: © 2021 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/). Int. J. Environ. Res. Public Health 2021, 18, 3211. https://doi.org/10.3390/ijerph18063211 https://www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2021, 18, 3211 2 of 10 2 of 10 an affective and cognitive component [9], with these components being interrelated [10]. Following the time-sequential model of well-being [11], the affective component has a sig- niﬁcant impact on the cognitive component, given that individuals use the former to judge life satisfaction. To evaluate the affective component, individuals judge their positive and negative affects whilst analysis of the cognitive component involves a subjective evaluation of their life in general. Leading a happy life has been related to positive outcomes such as longevity or good physical and mental health [12–14]. One of the central aims of positive psychology has been to discover how to improve happiness. 1. Introduction Following the Sustainable Happiness Model [15], happiness is inﬂuenced by the following three components: genetics, life circumstances, and intentional activities. Even though some of these components leave little space for potential modiﬁcation, individuals can play a role in improving their happiness levels by managing their intentional activities. Such activities could account for approximately 15% of the variance in our happiness [16]. The COVID-19 conﬁnement period in Spain (15 March to 4 May 2020) was an uncontrollable situation and one of the strictest lockdowns worldwide. Nonetheless, individuals may have intentionally acted to protect their happiness by frequently carrying out pleasant activities such as exercising at home, reading, and watching TV. Individual differences in the frequency with which these pleasant activities are carried out have previously been linked to emotional regulation strategies. Emotional regulation refers to the individual efforts to inﬂuence emotions [17]. There are a variety of emotional regulation strategies, with expressive suppression and cognitive reappraisal being the most widely studied. Following the model of emotional regulation [18], cognitive reappraisal implies a cognitive change in the reinterpretation of a situation or goal (e.g., not interpreting the lockdown as a punishment, but as a way of avoiding the spread of the disease to both ourselves and our loved ones) and it is regarded as an adaptative strategy. Expressive suppression implies a response modulation where the individual avoids the external display of an internal emotional state (e.g., avoiding the expression of the anxiety caused by the lockdown). Cognitive reappraisal, unlike expressive suppression, has previously been linked to better mental health (e.g., fewer depressive symptoms) and—of particular relevance for the purposes of the present study—to happier individuals [17–24]. In this regard, those individuals employing cognitive reappraisal seem to engage in more pleasant activities such as practicing sport regularly [25]. In contrast, expressive suppression has a negative impact on emotions and can thus be regarded as a risk factor for happiness [20]. The main objective of the present study was to analyze the potential role of two emo- tional regulation strategies (cognitive reappraisal and expressive suppression) as protective or risk factors for alterations in affective and cognitive happiness through the mediating role of engagement in pleasant activities during lockdown. 2.2. Instruments The Emotion Regulation Questionnaire (ERQ) [22]. This instrument is a 10-item self-report questionnaire developed to assess two emotion regulation strategies based on Gross’s model of emotion regulation [27]: cognitive reappraisal and expressive suppression. Four items are related to expressive suppression and six items to cognitive reappraisal. Participants must indicate the extent to which they agree with each item statement on a 7-point Likert scale (ranging from 1 “strongly disagree” to 7 “strongly agree”). In the present study, the Spanish version of the ERQ [22] was used, which has been shown to have adequate internal consistency (Cronbach’s α = 0.75 for expressive suppression, Cronbach’s α = 0.79 for cognitive reappraisal). The internal consistency for the sample of this study was also adequate (Cronbach’s α = 0.78 for expressive suppression, Cronbach’s α = 0.79 for cognitive reappraisal). g pp The Positive and Negative Affect Schedule (PANAS) [28]. This instrument was em- ployed as a measure of affective happiness [29], following previous studies. This is a widely used and well-validated self-report scale for measuring both positive affect (PA) and negative affect (NA). The 10-item short form version of the scale was used [30], in which ﬁve items assess PA and another ﬁve items assess NA. Participants must respond to each item on a 5-point Likert scale ranging from 1 “not at all” to 5 “strongly” where they must indicate the extent to which they have felt a particular emotion during the two previ- ous weeks. In the current study, the measure of affective happiness was the affect balance score [31], computed as the difference between mean PA score and mean NA score (i.e., PA–NA). The internal consistency for the sample of this study was acceptable (timepoint 1: Cronbach’s α = 0.72 for PA; Cronbach’s α = 0.71 for NA; timepoint 2: Cronbach’s α = 0.73 for PA; Cronbach’s α = 0.74 for NA). Cognitive happiness was assessed using one of the items of the mental health subscale of the SF-36 health survey [32] (adapted to Spanish by Alonso, Prieto, and Anto [33]). This item was as follows: “During the past two weeks, how often have you felt happy?” (In Spanish: “Durante las últimas dos semanas ¿con que frecuencia te sentiste feliz?”). Participants had to respond on a 6-point Likert scale ranging from 1 “none of the time” to 6 “all of the time”. 2.1. Participants Eighty-eight participants from a community sample voluntarily took part in this study. Their ages ranged from 19 to 67 years (M = 29.09; SD = 12.77) and 59 were women (67.05%). The participants were recruited by snowball sampling with the help of undergraduate students from the University of Málaga, Spain. Participants were assured about the conﬁdentiality and anonymity of the collected data, and they were treated in accordance with the Helsinki declaration [26]. 1. Introduction We employed a longitudinal methodology by analyzing affective and cognitive happiness at two timepoints: before the COVID-19 outbreak, and during the lockdown in Spain, where very strict conﬁne- ment measures were applied from 15 March to 4 May 2020. The COVID-19 pandemic has changed the way we interact with our environment in the short, medium, and long- term [20]. There is still a high degree of uncertainty about how this situation will evolve, which makes it necessary to focus on the psychological mechanisms that will make it easier for individuals to deal with this crisis. In this regard, the present longitudinal study could offer relevant information for facing situations of restricted freedom in an optimal way as well as elucidating the mechanisms that could potentiate affective and cognitive happiness. g p g pp On the basis of results reported in the previous literature, we proposed the follow- ing hypotheses: Hypothesis 1 (H1). Engaging in pleasant activities is positively correlated with affective happiness. Hypothesis 2 (H2). Engaging in pleasant activities is positively correlated with cognitive happiness. Hypothesis 3 (H3). Cognitive reappraisal is positively correlated with engaging in pleasant activities. Hypothesis 1 (H1). Engaging in pleasant activities is positively correlated with affective happiness. Hypothesis 3 (H3). Cognitive reappraisal is positively correlated with engaging in pleasant activities. 3 of 10 Int. J. Environ. Res. Public Health 2021, 18, 3211 Hypothesis 4 (H4). Cognitive reappraisal is positively correlated with affective happiness. Hypothesis 5 (H5). Cognitive reappraisal is positively correlated with cognitive happiness. Hypothesis 6 (H6). Expressive suppression is negatively correlated with engaging in pleasant activities. Hypothesis 7 (H7). Expressive suppression is negatively correlated with affective happiness. Hypothesis 8 (H8). Expressive suppression is negatively correlated with cognitive happiness. Hypothesis 8 (H8). Expressive suppression is negatively correlated with cognitive happiness. Hypothesis 9 (H9). Engaging in pleasant activities mediates the relationship between emotional regulation strategies and affective and cognitive happiness. 2.2. Instruments The frequency with which pleasant activities were carried out during the lockdown was assessed by the following statement: “In the last two weeks, I have done activities at home that I like and that make me feel good. For example, exercise, reading, watching TV Int. J. Environ. Res. Public Health 2021, 18, 3211 4 of 10 shows, listening to music, cooking, etc.” (In Spanish: “En estas dos últimas semanas he hecho actividades en casa que me gustan y me hacen sentir bien. Por ejemplo: ejercicio, ver series, leer, escuchar música, cocinar, etc.”). Participants responded on a 7-point Likert-scale in which 1 indicates “totally disagree” and 7 indicates “totally agree”. 2.3. Procedure The study consisted of two assessment phases. A ﬁrst phase (hereinafter timepoint 1), which was conducted before the COVID-19 outbreak (11–26 November 2019), and a second phase (hereinafter timepoint 2), which was conducted four months later during the Spanish lockdown imposed in response to the COVID-19 pandemic (1–16 April 2020). The assessment at timepoint 1 was part of a larger project concerned with emotional abilities and well-being (project UMA18-FEDERJA-114). In this ﬁrst assessment, partici- pants completed a battery of questionnaires that included the ERQ, the PANAS, the item on cognitive happiness belonging to the SF-36 health survey, and questions on demographic characteristics. Seven hundred and eighty-three participants completed the questionnaires in this ﬁrst phase. These participants were recruited by snowball sampling and completed the questionnaires through online platforms (LimeSurvey and Google forms). To avoid missing data, blank responses were not allowed in the questionnaires. g p q For the assessment carried out at timepoint 2, we invited all the people who had completed the ﬁrst assessment to participate in this second phase. In this second assessment, participants had to complete again the PANAS and the item on cognitive happiness, in order to explore possible changes between timepoint 1 and 2. Moreover, they had to respond to a set of questions on their experience during the COVID-19 lockdown, among which was our question of interest about the frequency with which they engaged in pleasant activities. For participating in this second assessment phase, the respondents were compensated with the chance to win a €40 Amazon gift card. A total number of 88 participants completed these questionnaires, which represents the ﬁnal sample of the study that completed both the timepoint 1 and timepoint 2 assessment. As in the assessment at timepoint 1, all the questionnaires were administered online, and blank responses were not allowed in order to avoid missing data. 2.4. Statistical Analysis Before conducting the analyses, we computed differential scores for affective happi- ness (from PANAS) and for cognitive happiness by subtracting the scores obtained before the COVID-19 outbreak from those obtained during the COVID-19 lockdown (hereinafter, affective happiness change and cognitive happiness change). First, descriptive analyses (means and standard deviations) were conducted to explore the scores obtained for each measure. Second, gender and age-related differences in the study variables were examined using independent two-sample t-tests and linear and quadratic regression analysis, respectively. Moreover, changes in levels of affective and cognitive happiness between timepoint 1 and timepoint 2 were analyzed by paired sample t-tests. Third, the relationships between the variables included in the study were evaluated by bivariate Pearson’s correlations. The statistical signiﬁcance threshold for t-tests, regression analyses, and Pearson’s correlations was set at p < 0.05. Fourth, and as the main objective of this research, a path analysis was conducted to evaluate the proposed model integrating emotion regulation strategies, engagement in pleasant activities during the COVID-19 lockdown, and changes in the levels of affective happiness and cognitive happiness during lockdown in comparison with pre-pandemic levels. Maximum likelihood estimation was used to estimate the model parameters, and the model ﬁt to the data was tested by the following indices: χ2 statistic, comparative ﬁt index (CFI), and the root mean square error of approximation (RMSEA) [34,35]. Indirect effects were computed with the bias-corrected bootstrapping method (5000 samples) and the level of statistical signiﬁcance was set at 95% conﬁdence intervals (CIs). All analyses were conducted using IBM SPSS 23 (IBM corp., Armonk, NY, USA) and IBM AMOS 21.0 (IBM corp., Armonk, NY, USA) software. 5 of 10 Int. J. Environ. Res. Public Health 2021, 18, 3211 3. Results Descriptive statistics for both the total sample and separated by gender are presented in Table 1. Gender differences were not observed for any variable (all ps > 0.05). Age was only found to be linearly related to the variable of affective happiness at timepoint 1 (β = 0.25, p = 0.02). The older the age, the higher the level of affective happiness at timepoint 1. Paired sample t-tests comparing the levels of cognitive happiness between timepoint 1 and timepoint 2 revealed higher levels of cognitive happiness at timepoint 1 (t (87) = 3.93; p < 0.001; Cohen’s d = 0.42). Differences in affective happiness between timepoint 1 and timepoint 2 were not signiﬁcant (p > 0.05). Table 1. Descriptive statistics (mean and standard deviation (SD)) of the total sample and separated by gender for the variables included in this study. Gender Comparisons x (SD) Total Sample x (SD) Men x (SD) Women t-Value p-Value Expressive suppression 3.40 (1.41) 3.52 (1.35) 3.35 (1.45) 0.53 0.60 Cognitive reappraisal 4.66 (1.18) 4.62 (1.32) 4.68 (1.12) 0.20 0.84 Pleasant activities 5.95 (1.25) 5.86 (1.13) 6.00 (1.31) 0.48 0.63 Affective happiness timepoint 1 0.82 (1.04) 0.86 (1.06) 0.80 (1.05) 0.22 0.83 Affective happiness timepoint 2 0.77 (1.19) 0.83 (1.33) 0.75 (1.13) 0.29 0.77 Affective happiness change −0.05 (1.26) −0.03 (1.10) −0.05 (1.34) 0.09 0.93 Cognitive happiness timepoint 1 3.95 (1.08) 3.76 (1.27) 4.05 (0.97) 1.19 0.24 Cognitive happiness timepoint 2 3.53 (1.14) 3.59 (1.38) 3.51 (1.02) 0.30 0.77 Cognitive happiness change −0.42 (1.00) −0.17 (1.10) −0.54 (0.93) 1.64 0.10 Table 1. Descriptive statistics (mean and standard deviation (SD)) of the total sample and separated by gender for the variables included in this study. ptive statistics (mean and standard deviation (SD)) of the total sample and separated by gender for the ed in this study. ics (mean and standard deviation (SD)) of the total sample and separated by gender for the d Pearson’s correlation analyses are presented in Table 2. Focusing on the correlations of interest, we observed that cognitive reappraisal was positively correlated with engage- ment in pleasant activities (r = 0.35, p < 0.001), affective happiness at timepoint 1 (r = 0.41, p < 0.001) and timepoint 2 (r = 0.35, p < 0.001), and cognitive happiness at timepoint 1 (r = 0.42, p < 0.001) and timepoint 2 (r = 0.27, p = 0.01). 3. Results Expressive suppression was nega- tively correlated with affective happiness at timepoint 1 (r = −0.25, p = 0.02) and timepoint 2 (r = −0.24, p = 0.02). It is worth noting that, although non-signiﬁcant, expressive suppres- sion showed a correlation near to the statistical signiﬁcance threshold with engagement in pleasant activities (r = −0.19, p = 0.07) and cognitive happiness at timepoint 1 (r = −0.19, p = 0.07) and timepoint 2 (r = −0.19, p = 0.07). With respect to engagement in pleasant activities, in addition to the relationships previously described, this variable showed a positive correlation with affective happiness at timepoint 1 (r = 0.28, p < 0.01) and timepoint 2 (r = 0.58, p < 0.001), affective happiness change (r = 0.32, p < 0.01), and cognitive happiness at timepoint 1 (r = 0.40, p < 0.001) and timepoint 2 (r = 0.51, p < 0.001). Finally, a positive correlation was observed between the affective happiness change and cognitive happiness change (r = 0.29, p < 0.01). g p The results of the path analysis revealed an optimal ﬁt of the proposed model (χ2(5) = 4.92, p = 0.42; CFI = 1.00; RMSEA < 0.001). Figure 1 shows the model with the standardized path coefﬁcients and proportion of explained variance. According to these results, cognitive reappraisal was signiﬁcantly associated with engagement in pleasant activities during the COVID-19 lockdown period. In turn, these pleasant activities were related to more positive affective happiness during the COVID-19 lockdown (in compari- son with pre-pandemic levels), which was related to higher levels of cognitive happiness. The emotion regulation strategy of expressive suppression was not signiﬁcantly related to engagement in pleasant activities; however, the resulting model after removing the path between expressive suppression and pleasant activities from the original model showed slightly worse ﬁt indices (χ2(3) = 4.84, p = 0.18; CFI = 0.93; RMSEA = 0.08), although differences between models were not signiﬁcant (χ2 diff = 1.87, df diff = 1, p > 0.05). Int. J. Environ. Res. Public Health 2021, 18, 3211 (9) Cognitive happiness change − 6 of 10 0 Table 2. Pearson’s correlation matrix for the study variables. 3. Results 1 2 3 4 5 6 7 8 (1) Expressive suppression - (2) Cognitive reappraisal −0.16 - (3) Engagement in pleasant activities −0.19 0.35 ** - (4) Affective happiness timepoint 1 −0.25 * 0.41 0.28 - (5) Affective happiness timepoint 2 −0.24 * 0.35 0.58 0.37 - (6) Affective happiness change −0.02 −0.01 0.32 −0.48 0.64 - (7) Cognitive happiness timepoint 1 −0.19 0.42 0.40 . 0.51 0.31 −0.12 - (8) Cognitive happiness timepoint 2 −0.19 0.27 * 0.51 0.38 0.48 0.14 0.60 - (9) Cognitive happiness change −0.00 −0.14 0.15 −0.11 0.20 0.29 −0.40 0.50 * p < 0.05, ** p < 0.01. The results of the path analysis revealed an optimal fit of the proposed model (χ2(5) = 4.92, p = 0.42; CFI = 1.00; RMSEA < 0.001). Figure 1 shows the model with the standard- ized path coefficients and proportion of explained variance. According to these results, cognitive reappraisal was significantly associated with engagement in pleasant activities during the COVID-19 lockdown period. In turn, these pleasant activities were related to more positive affective happiness during the COVID-19 lockdown (in comparison with pre-pandemic levels), which was related to higher levels of cognitive happiness. The emo- tion regulation strategy of expressive suppression was not significantly related to engage- ment in pleasant activities; however, the resulting model after removing the path between expressive suppression and pleasant activities from the original model showed slightly worse fit indices (χ2(3) = 4.84, p = 0.18; CFI = 0.93; RMSEA = 0.08), although differences between models were not significant (χ2 diff = 1.87, df diff = 1, p > 0.05). Figure 1. Graphical representation of the model including standardized path coefﬁcients and explained variance (R2). An asterisk indicates signiﬁcance at the p < 0.05 level. Table 2. Pearson’s correlation matrix for the study variables. p y p .92, p = 0.42; CFI = 1.00; RMSEA < 0.001). Figure 1 shows d ath oeffi ie t a d o o tio of e lai ed a ia Figure 1. Graphical representation of the model including standardized path coefﬁcients and explained variance (R2). An asterisk indicates signiﬁcance at the p < 0.05 level. entation of the model including standardized path coefficients and explained variance (R2). An nce at the p < 0.05 level. 3. Results Finally, analysis of the indirect effects, as tested by the bootstrapping method (see Table 3) revealed that cognitive reappraisal was positively related to the cognitive happiness change through the effect on engagement in pleasant activities and the affective happiness change (standardized indirect effect = 0.030; 95% CI: 0.004, 0.080). Engagement in pleasant activities was also positively indirectly related to the cognitive happiness change through the affective happiness change (standardized indirect effect = 0.092; 95% CI: 0.011, 0.206). Expressive suppression did not show an indirect effect on the cognitive happiness change via engagement in pleasant activities and the affective happiness change (standardized indirect effect = −0.013; 95% CI: −0.057, 0.002). Table 3. Standardized indirect effects and conﬁdence intervals for the path model. Path Indirect Effect Bias-Corrected 95% CI Expressive suppression →Pleasant activities →Affective happiness change → Cognitive happiness change −0.013 [−0.057, 0.002] Cognitive reappraisal →Pleasant activities →Affective happiness change → Cognitive happiness change 0.030 * [0.004, 0.080] Pleasant activities →Affective happiness change →Cognitive happiness change 0.092 * [0.011, 0.206] Note: * signiﬁcant indirect effect. Table 3. Standardized indirect effects and conﬁdence intervals for the path model. Table 3. Standardized indirect effects and conﬁdence intervals for the path model. 4. Discussion The aim of the present study was to longitudinally explore the role of two emotional regulation strategies (cognitive reappraisal and expressive suppression) and engagement in pleasant activities as protective and risk factors of affective and cognitive happiness during the Spanish COVID-19 lockdown. More speciﬁcally, we were interested in analyzing the effect of these emotion regulation strategies on the ﬂuctuations in affective and cognitive happiness during the COVID-19 lockdown (in comparison with pre-pandemic levels) through the mediating role of engagement in pleasant activities during the lockdown. 7 of 10 Int. J. Environ. Res. Public Health 2021, 18, 3211 7 of 10 Consistent with the results of previous studies where the COVID-19 lockdown has been demonstrated to have a signiﬁcant psychological impact [6,7], the present study showed how cognitive happiness diminished during the lockdown in comparison with the period prior to the outbreak of the pandemic. Unexpectedly, no differences were found in affective happiness between the two timepoints. However, this ﬁnding is congruent with the results of a previous metanalysis showing that life events appear to have a greater impact on cognitive happiness than affective happiness [36]. Further, and in accordance with the Sustainable Happiness Model [15], the results of the present study revealed that those individuals that engaged more frequently in pleasant activities during the lockdown showed a smaller decrease in affective happiness. This provides support for H1 of the present study. In contrast, a similar correlation was not found with the cognitive happiness change (H2). However, we observed an indirect effect of engaging in pleasant activities on cognitive happiness, through the affective happiness change, which is consistent with the time-sequential model of well-being [11]. This issue will be discussed later. q g In addition, and congruent with H3, those participants that employed the adaptative strategy of cognitive reappraisal appeared to more frequently engage in pleasant activities during the lockdown and had higher levels of affective (H4) and cognitive happiness (H5) both at timepoint 1 and timepoint 2. Previous studies have highlighted the relevance of this strategy as an adequate way of regulating emotions and achieving more happiness [17,19–24]. This result is also consistent with the ﬁndings reported in other studies conducted before the pandemic in which the cognitive reappraisal strategy was related to exercising more frequently [25]. 4. Discussion With respect to the expressive suppression strategy, H6 and H8 were not supported, since the negative correlation between engaging in pleasant activities and cognitive happiness was not signiﬁcant (although this approached signiﬁcance). Future research should aim to provide a more in-depth analysis of how maladaptive strategies could impact both the extent to which individuals engage in these activities, and their cognitive happiness. Congruent with H7 and the results of previous studies—and contrary to the results found for the cognitive reappraisal strategy—expressive suppression was negatively correlated with affective happiness at timepoint 1 and timepoint 2. This ﬁnding was expected, given that expressive suppression is an emotional regulation ability that is usually regarded as inadequate [18,20–22]. Finally, and concerning the main focus of this study, a model has been offered for protecting people’s happiness during the COVID-19 lockdown period through the use of adaptative emotion regulation strategies and engagement in pleasant activities (H9). Speciﬁcally, the results showed how those participants that employed the cognitive reap- praisal strategy engaged in more pleasant activities during the COVID-19 lockdown period. Subsequently, carrying out these pleasant activities during this period was related to higher levels of affective happiness, which, in turn, was related to higher levels of cognitive happiness, as predicted by the time-sequential model of well-being [11]. Thus, these results indicate that cognitive reappraisal could be an indirect protective factor of happiness in uncontrollable circumstances such as the COVID-19 lockdown, a ﬁnding that is consistent with the previous literature [17–22]. This makes sense, given that the aim of the reappraisal emotion strategy is to reevaluate a given situation, thus allowing people to look for alterna- tives to cope with the situation in a more adaptive way, which, in this case, was achieved by engaging in pleasant activities while conﬁned at home [25]. Although the correlation between the expressive suppression strategy and engagement in pleasant activities did not reach signiﬁcance (p = 0.07), this variable was included in the model as it slightly improves the ﬁt of the proposed model. Our results have important implications for the well-being of people during situations of conﬁnement. From the three factors that inﬂuence happiness following the Sustainable Happiness Model [15], only one is in the hands of the individual, that is, intentional activities. 4. Discussion In an uncontrollable situation such as that presented by the strict COVID-19 lockdown imposed in Spain where a reduction in the cognitive happiness of the participants was clearly evident, engaging in pleasant activities at home was one of the few things Int. J. Environ. Res. Public Health 2021, 18, 3211 8 of 10 they could manage in order to improve their happiness. Tools such as the cognitive reappraisal strategy—which allows for increasing the frequency with which people engage in these activities [25]—could offer an adaptive way of improving people’s happiness during difﬁcult times. g The present study has a number of strengths. First, this work focused on positive psychology, as opposed to psychopathology. As already mentioned, the previous literature has primarily focused on studying the latter, paying less attention to the protective factors that could help to address the emotional consequences of the pandemic and maintain happiness. This will be particularly useful for dealing with future lockdowns in an adaptive way. Second, this study employed a longitudinal design, evaluating how cognitive and affective happiness were altered during the COVID-19 lockdown by comparing the scores obtained during that period with those obtained prior to the pandemic. This allowed for measuring changes in happiness associated with the COVID-19 lockdown. Finally, our results have clear practical implications. The current limitations that people are experiencing as a result of COVID-19 are expected to continue for the foreseeable future [37], and it is, therefore, essential that people are able to draw upon resources that have the potential to improve their happiness. In this regard, the ﬁndings of this study suggest that promoting engagement in pleasant activities during periods of lockdown could help to mitigate some of the negative consequences of the pandemic. In order to achieve this, interventions aimed at improving happiness should focus on developing the most adaptive emotional regulation strategies, and according to the results reported here, training in the cognitive reappraisal strategy could be a ﬁrst step towards achieving this goal. Therefore, brief interventions should focus on improving adaptive emotional regulation strategies such as emotional reappraisal [38]. In addition to this strategy, there are others (e.g., problem solving) that have also shown to be relevant for mental health [39]. 4. Discussion Nonetheless, this model revealed a positive indirect relationship between these variables through the frequency of engagement in pleasant activities. and affective happiness. Nonetheless, this model revealed a positive indirect relationship between these variables through the frequency of engagement in pleasant activities. 5. Conclusions In conclusion, the present study contributes towards understanding the mechanism(s) by which individuals can improve their happiness in stressful and restricted times such as the COVID-19 conﬁnement. The theoretical implication is that happiness is not predeter- mined, but instead depends on adaptive emotional regulation strategies and intentional activities. In this work, a model is presented for maintaining happiness in potential future lockdown periods. The results of testing this model suggest that improving the cognitive reappraisal strategy could be an adaptive way of increasing the frequency with which people engage in pleasant activities during conﬁnement, which will subsequently increase affective happiness and ultimately, cognitive happiness. Author Contributions: Conceptualization, M.J.G.-C., A.M.-R., and P.F.-B.; data curation, A.M.-R. and R.G.-L.; formal analysis, A.M.-R. and R.G.-L.; funding acquisition, A.M.-R., R.G.-L., R.C., and P.F.-B.; methodology, M.J.G.-C., R.C., and P.F.-B.; software, M.J.G.-C. and R.G.-L.; writing—original draft, M.J.G.-C. and A.M.-R.; writing—review and editing, R.G.-L., R.C., and P.F.-B. All authors have read and agreed to the published version of the manuscript. Author Contributions: Conceptualization, M.J.G.-C., A.M.-R., and P.F.-B.; data curation, A.M.-R. and R.G.-L.; formal analysis, A.M.-R. and R.G.-L.; funding acquisition, A.M.-R., R.G.-L., R.C., and P.F.-B.; methodology, M.J.G.-C., R.C., and P.F.-B.; software, M.J.G.-C. and R.G.-L.; writing—original draft, M.J.G.-C. and A.M.-R.; writing—review and editing, R.G.-L., R.C., and P.F.-B. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Spanish Ministry OF Economy, Industry and Competitive- ness, grant number “PSI2017-84170-R”, the Junta de Andalucía, grant number “UMA18-FEDERJA- 137” to A.M.-R. and grant number “UMA18-FEDERJA-114” to P.F.-B. and R.C. and by the Spanish Ministry of Education, grant number “FPU15/05179” to R.G.-L. (FPU). Institutional Review Board Statement: The research ethics committee of the University of Málaga approved the study protocol (14-2019-H) as part of the project PSI2017-84170-R. Participants gave informed consent and were assessed in accordance with the Helsinki declaration. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: Data are available upon reasonable request by emailing M.J.G.-C. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. 4. Discussion The joint use of these strategies could help individuals to increase their happiness during these challenging times, although future lines of investigation should focus on empirically investigating the relevance of these alternative regulation strategies within the framework of the proposed model. Finally, together with an intervention based on emotional regulation strategies, it would be interesting to help individuals to identify the pleasant activities that are most enjoyable for them in order to facilitate the implementation of these strategies when necessary. Moreover, in the future, it might be interesting to identify what could be the most useful pleasant activities depending on sociodemographic and individual differences (e.g., people living alone, adolescents, marital status, or risk population). g p p g p p Despite the strengths of this study, it is important to note some limitations. First, the size of the study sample was small. Only a small percentage of the participants involved in timepoint 1 decided to participate at timepoint 2. Here, it is important to note that when participants completed the assessment at timepoint 1, the second phase of the study had still not been planned, given that COVID-19 had not yet emerged. Thus, participants had not made any commitment to being involved in the second assessment. Although a larger number of participants would have helped to improve the representativeness of the sample, working with the present sample made it possible to employ a longitudinal design by taking advantage of data that had been collected prior to the pandemic, which has provided a unique way of examining the psychological consequences of COVID-19 and lockdown. Another limitation concerns the gender imbalance of the sample (with fewer men than women), which could potentially affect the generalizability of the ﬁndings. It should also be noted that the variance explained by the model was somewhat small. However, this is consistent with previous studies based on the Sustainable Happiness Model [16]. This model proposes that intentional activities explain only 15% of the variance in happiness. Therefore, in this study, the 8% of the variance explained by engagement in pleasant activities during the COVID-19 lockdown represents more than half of that predicted by the Sustainable Happiness Model. Finally, contrary to our expectations, no signiﬁcant correlation was found between cognitive reappraisal and the change in cognitive Int. J. Environ. Res. Public Health 2021, 18, 3211 9 of 10 9 of 10 and affective happiness. References [CrossRef] 10 of 10 10 of 10 Int. J. Environ. Res. Public Health 2021, 18, 3211 10. Schimmack, U. The structure of subjective well-being. In The Science of Subjective Well-Being; Eid, M., Larsen, R.J., Eds.; Guilford Press: New York, NY, USA, 2008; pp. 97–123. , , , ; pp 11. Kim-Prieto, C.; Diener, E.; Tamir, M.; Scollon, C.; Diener, M. Integrating the Diverse Deﬁnitions of Happiness: A Time-Sequential Framework of Subjective Well-Being. J. Happiness Stud. 2005, 6, 261–300. [CrossRef] , , , ; pp 11. Kim-Prieto, C.; Diener, E.; Tamir, M.; Scollon, C.; Diener, M. Integrating the Diverse Deﬁnitions of Happiness: A Time-Sequential F k f S bj i W ll B i J H i St d 2005 6 261 300 [C R f] 11. Kim-Prieto, C.; Diener, E.; Tamir, M.; Scollon, C.; Diener, M. Integrating the Diverse Deﬁnitions of Happ Framework of Subjective Well-Being. J. Happiness Stud. 2005, 6, 261–300. [CrossRef] j g pp 12. Lombardo, P.; Jones, W.; Wang, L.; Shen, X.; Goldner, E.M. The fundamental association between men Results from successive waves of a Canadian national survey. BMC Public Health 2018, 18, 1–9. [Cr j g pp ardo, P.; Jones, W.; Wang, L.; Shen, X.; Goldner, E.M. The fundamental association between mental health and l 12. Lombardo, P.; Jones, W.; Wang, L.; Shen, X.; Goldner, E.M. The fundamental association between mental h R lt f i f C di ti l BMC P bli H lth 2018 18 1 9 [C R 12. Lombardo, P.; Jones, W.; Wang, L.; Shen, X.; Goldner, E.M. The fundamental association between mental health and life satisfaction: Results from successive waves of a Canadian national survey. BMC Public Health 2018, 18, 1–9. [CrossRef] [PubMed] 12. Lombardo, P.; Jones, W.; Wang, L.; Shen, X.; Goldner, E.M. The fundamental association between mental h Results from successive waves of a Canadian national survey. BMC Public Health 2018, 18, 1–9. [CrossR y 13. Sirgy, M.J. Positive balance: A hierarchical perspective of positive mental health. Qual. Life Res. 2019, 28, 1921–1930. [CrossRef] 14. Diener, E.; Chan, M.Y. Happy People Live Longer: Subjective Well-Being Contributes to Health and Longevity. Appl. Psychol. Heal. Well Being 2011, 3, 1–43. [CrossRef] 13. Sirgy, M.J. Positive balance: A hierarchical perspective of positive mental health. Qual. Life Res. References 1. Xiong, J.; Lipsitz, O.; Nasri, F.; Lui, L.M.W.; Gill, H.; Phan, L.; Chen-Li, D.; Iacobucci, M.; Ho, R.; Majeed, A.; et al. Impact of COVID-19 pandemic on mental health in the general population: A systematic review. J. Affect. Disord. 2020, 277, 55–64. [CrossRef] [ ] 2. Fiorillo, A.; Gorwood, P. The consequences of the COVID-19 pandemic on mental health and implications for clinical practice. Eur. Psychiatry 2020, 63, e32. [CrossRef] [PubMed] 3. McCracken, L.M.; Badinlou, F.; Buhrman, M.; Brocki, K.C. Psychological impact of COVID-19 in the Swedish population: Depression, anxiety, and insomnia and their associations to risk and vulnerability factors. Eur. Psychiatry 2020, 63, e81. [CrossRef] [PubMed] 4. Wang, S.; Zhang, Y.; Ding, W.; Meng, Y.; Hu, H.; Liu, Z.; Zeng, X.; Wang, M. Psychological distress and sleep problems when people are under interpersonal isolation during an epidemic: A nationwide multicenter cross-sectional study. Eur. Psychiatry 2020, 63, e77. [CrossRef] 5. Salari, N.; Hosseinian-Far, A.; Jalali, R.; Vaisi-Raygani, A.; Rasoulpoor, S.; Mohammadi, M.; Rasoulpoor, S.; Khaledi-Paveh, B. Prevalence of stress, anxiety, depression among the general population during the COVID-19 pandemic: A systematic review and meta-analysis. Glob. Health 2020, 16, 57. [CrossRef] y 6. Fernández, R.S.; Crivelli, L.; Guimet, N.M.; Allegri, R.F.; Pedreira, M.E. Psychological distress associated with COVID-19 quarantine: Latent proﬁle analysis, outcome prediction and mediation analysis. J. Affect. Disord. 2020, 277, 75–84. [CrossRef] [PubMed] 7. Brooks, S.K.; Webster, R.K.; Smith, L.E.; Woodland, L.; Wessely, S.; Greenberg, N.; Rubin, G.J. The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet 2020, 395, 912–920. [CrossRef] 8. Schimmack, U.; Oishi, S.; Furr, R.M.; Funder, D.C. Personality and Life Satisfaction: A Facet-Level An Bull. 2004, 30, 1062–1075. [CrossRef] [PubMed] 8. Schimmack, U.; Oishi, S.; Furr, R.M.; Funder, D.C. Personality and Life Satisfaction: A Facet-Level Analysis. Pers. Soc. Psychol. Bull. 2004, 30, 1062–1075. [CrossRef] [PubMed] 9 Diener E ; Oishi S ; Lucas R E Personality Culture and Subjective Well Being: Emotional and Cognitive Evaluations of Life 8. Schimmack, U.; Oishi, S.; Furr, R.M.; Funder, D.C. Personality and Life Satisfaction: A Facet Level Analysis. Pers. Soc. Psychol. Bull. 2004, 30, 1062–1075. [CrossRef] [PubMed] 9. Diener, E.; Oishi, S.; Lucas, R.E. Personality, Culture, and Subjective Well-Being: Emotional and Cognitive Evaluations of Life. 9. Diener, E.; Oishi, S.; Lucas, R.E. Personality, Culture, and Subjective Well-Being: Emotional and Cogn Annu. Rev. Psychol. 2003, 54, 403–425. References 201 3 S gy, J os e ce e c c pe spec e o pos e e e Q f , , 9 930 [C oss e ] 14. Diener, E.; Chan, M.Y. Happy People Live Longer: Subjective Well-Being Contributes to Health and Longevity. Appl. Psychol. Heal. Well Being 2011, 3, 1–43. [CrossRef] g 15. Sheldon, K.M.; Lyubomirsky, S. Revisiting the Sustainable Happiness Model and Pie Chart: Can Happiness Be Successfully Pursued? J. Posit. Psychol. 2019, 1–10. [CrossRef] Pursued? J. Posit. Psychol. 2019, 1–10. [CrossRef] 16. Klug, H.J.P.; Maier, G.W. Linking Goal Progress and Subjective Well-Being: A Meta-analysis. J. Happiness Stud. 2014, 16, 37–65. [CrossRef] [ ] 17. McRae, K.; Gross, J.J. Emotion regulation. Emotion 2020, 20, 1–9. [CrossRef] [PubMed] 18. Gross, J.J. The Extended Process Model of Emotion Regulation: Elaborations, Applications, and Future Directions. Psychol. Inq. 2015, 26, 130–137. [CrossRef] 19. Koval, P.; Kalokerinos, E.; Greenaway, K.; Medland, H.; Kuppens, P.; Nezlek, J.B.; Hinton, J.; Gross, J. Emotion Regulation in Everyday Life: What Can We Learn from Global Self-Reports? arXiv 2020, 1–59. [CrossRef] 20. Ecutuli, D. Cognitive reappraisal and expressive suppression strategies role in the emotion regulation: An overview on their modulatory effects and neural correlates. Front. Syst. Neurosci. 2014, 8, 175. [CrossRef] y y 21. Uusberg, A.; Taxer, J.L.; Yih, J.; Uusberg, H.; Gross, J.J. Reappraising Reappraisal. Emot. Rev. 2019, 11, 2 21. Uusberg, A.; Taxer, J.L.; Yih, J.; Uusberg, H.; Gross, J.J. Reappraising Reappraisal. Emot. Rev. 2019, 11, 267–282. [CrossRef] 22. Cabello, R.; Salguero, J.M.; Fernández-Berrocal, P.; Gross, J.J. A Spanish Adaptation of the Emotion Regulation Questionnaire. Eur. J. Psychol. Assess. 2013, 29, 234–240. [CrossRef] 23. Moore, S.A.; Zoellner, L.A.; Mollenholt, N. Are expressive suppression and cognitive reappraisal associated with stress-related symptoms? Behav. Res. Ther. 2008, 46, 993–1000. [CrossRef] 24. John, O.P.; Gross, J.J. Healthy and Unhealthy Emotion Regulation: Personality Processes, Individual D Development. J. Pers. 2004, 72, 1301–1334. [CrossRef] [PubMed] 24. John, O.P.; Gross, J.J. Healthy and Unhealthy Emotion Regulation: Personality Processes, Individual Differences, and Life Span Development. J. Pers. 2004, 72, 1301–1334. [CrossRef] [PubMed] 25. Giles, G.E.; Cantelon, J.A.; Eddy, M.D.; Brunyé, T.T.; Urry, H.L.; Mahoney, C.R.; Kanarek, R.B. Habitual e cognitive control and cognitive reappraisal success. Exp. Brain Res. 2017, 235, 3785–3797. [CrossRef] Giles, G.E.; Cantelon, J.A.; Eddy, M.D.; Brunyé, T.T.; Urry, H.L.; Mahoney, C.R.; Kanarek, R.B. References Habitual exer cognitive control and cognitive reappraisal success Exp Brain Res 2017 235 3785 3797 [CrossRef] Giles, G.E.; Cantelon, J.A.; Eddy, M.D.; Brunyé, T.T.; Urry, H.L.; Mahoney, C.R.; Kanarek, R.B. Habitual exer cognitive control and cognitive reappraisal success. Exp. Brain Res. 2017, 235, 3785–3797. [CrossRef] 26. World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects; WMA: Seoul, Korea, 2008. j 27. Gross, J.J. The emerging ﬁeld of emotion regulation: An integrative review. Rev. Gen. Psychol. 1998, 2, 271–299. [CrossRef] 28. Watson, D.; Clark, L.A.; Tellegen, A. Development and validation of brief measures of positive and neg scales. J. Pers. Soc. Psychol. 1988, 54, 1063–1070. [CrossRef] 29. Ryan, R.M.; Deci, E.L. On Happiness and Human Potentials: A Review of Research on Hedonic and Eudaimonic Well-Being. Annu. Rev. Psychol. 2001, 52, 141–166. [CrossRef] y 30. Thompson, E.R. Development and Validation of an Internationally Reliable Short-Form of the Positive and Negative Affect Schedule (PANAS). J. Cross Cultural Psychol. 2007, 38, 227–242. [CrossRef] 31. Diener, E.; Wirtz, D.; Biswas-Diener, R.; Tov, W.; Kim-Prieto, C.; Choi, D.W.; Oishi, S. New Measures of Well-Being; Springer: Heidelberg, Germany, 2009. [CrossRef] 32. Ware, J.E., Jr.; Sherbourne, C.D. The MOS 36-item short-form health survey (SF-36): I. Conceptual framework and item selection. Med Care 1992, 30, 473–483. [CrossRef] [PubMed] 33. Alonso, J.; Prieto, L.; Anto, J.M. La versión española del SF-36 Health Survey (Cuestionario de Salud SF- la medida de los resultados clínicos. Med. Clin. 1995, 104, 771–776. , , 34. Arbuckle, J.L. IBM SPSS Amos 21 User’s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL, USA, 2012. 34. Arbuckle, J.L. IBM SPSS Amos 21 User’s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL, USA, 2012. 35 B rne B M Structural Equation Modeling with EQS Basic Concepts Applications and Programming Routledge London UK 2013 , J.L. IBM SPSS Amos 21 User’s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL, USA, 2012. M. Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming; Routledge: London 34. Arbuckle, J.L. IBM SPSS Amos 21 User’s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL SPSS Amos 21 User’s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL, USA, 2012. l E i M d li i h EQS B i C A li i d P i R l d L d U , J.L. References IBM SPSS Amos 21 User s Guide; Amos Development Corporation, SPSS Inc.: Chicago, IL, USA, 2012. M. Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming; Routledge: London 35. Byrne, B.M. Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming; Routl 36. Luhmann, M.; Hofmann, W.; Eid, M.; Lucas, R.E. Subjective well-being and adaptation to life events: A meta-analysis. J. Pers. Soc. Psychol. 2012, 102, 592–615. [CrossRef] [PubMed] y 37. Skegg, D.; Gluckman, P.; Boulton, G.; Hackmann, H.; Karim, S.S.A.; Piot, P.; Woopen, C. Future scenarios for the COVID-19 pandemic. Lancet 2021, 397, 777–778. [CrossRef] 38. Ranney, R.M.; Bruehlman-Senecal, E.; Ayduk, O. Comparing the Effects of Three Online Cognitive Reappraisal Trainings on Well-Being. J. Happiness Stud. 2016, 18, 1319–1338. [CrossRef] 39. Schäfer, J.Ö.; Naumann, E.; Holmes, E.A.; Tuschen-Cafﬁer, B.; Samson, A.C. Emotion Regulation Strategies in Depressive and Anxiety Symptoms in Youth: A Meta-Analytic Review. J. Youth Adolesc. 2017, 46, 261–276. [CrossRef] [PubMed]
https://openalex.org/W4360583826	https://periodicos.unemat.br/index.php/jhnpeps/article/download/4307/3609	Portuguese	null	null	Journal Health NPEPS	2,016	cc-by	7,870	Uso de bebidas alcoólicas por estudantes de medicina Izete Soares da Silva Dantas Pereira1, Rodrigo Alves de Melo2, Alison Jesissai Pinto Gurgel3, Ana Gabriela Miranda Barbosa4, Indira Coan Zanatta5, Sarah Lis Araújo Melo6, João Dantas Pereira7 RESUMO SU O Objetivo: identificar a frequência do consumo de álcool por estudantes do curso de medicina. Método: estudo transversal, com abordagem quantitativa. Utilizou-se para a coleta de dados o questionário AUDIT aplicado a 128 estudantes que frequentam do primeiro ao oitavo períodos do curso de graduação em medicina. Resultados: a maioria não apresentou consumo elevado de álcool. O uso foi referido apenas em ocasiões festivas, sem que isso acarretasse consequências negativas na rotina dos acadêmicos. Não se verificou relação das variáveis sociodemográficas com a quantidade e intensidade do consumo de bebidas alcoólicas. Considerações finais: Apesar de não constituir ainda um problema na realidade dos participantes, torna-se de fundamental importância investir em ações/orientações preventivas sobre os riscos do uso de bebidas alcoólicas e os malefícios que podem trazer para o exercício desses futuros profissionais. Descritores: Bebidas Alcoólicas; Estudantes de Medicina; Prevenção q p p p Descritores: Bebidas Alcoólicas; Estudantes de Medicina; Prevenção. 1Assistente social/Bacharel em Direito. Doutora em Saúde Pública. Docente do curso de Medicina da Universidade do Estado do Rio Grande do Norte (UERN). Natal, Rio Grande do Norte, Brasil. E-mail: izetedantas@hotmail.com ORCID ID: https://orcid.org/0000-0002-2239-6582 Autor principal – Endereço para correspondência: Rua das Tílias, 2114, Cidade Satélite-Pitimbú-Natal, RN, Brasil. CEP 59067-690. 2Acadêmico de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: rodrigoamelo3@gmail.com ORCID ID: https://orcid.org/0000-0002-8070-6614 3Acadêmico de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: ajesissai@gmail.com ORCID ID: https://orcid.org/0000-0003-1105-2407 4Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: belinha0623@gmail.com ORCID ID: https://orcid.org/0000-0002-5130-8086 5Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: indiracoan@hotmail.com ORCID ID: https://orcid.org/0000-0002-4884-1381 6Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: sarah.lis.araujo@hotmail.com ORCID ID: https://orcid.org/0000-0001-8238-9993 7Assistente Social/Bacharel em Direito. Doutor em Sociologia. Docente do curso de Direito da Universidade Federal do Rio Grande do Norte (UFRN). Natal, Rio Grande do Norte, Brasil. E-mail: jotarn@hotmail.com ORCID ID: https://orcid.org/0000-0001-9855-827X Este artigo está licenciado sob forma de uma licença Creative Commons Atribuição 4.0 Internacional, que permite uso irrestrito, distribuição e reprodução em qualquer meio, desde que a publicação original seja corretamente citada. 1Assistente social/Bacharel em Direito. Doutora em Saúde Pública. http://dx.doi.org/10.30681/252610104307 ARTIGO ORIGINAL Uso de bebidas alcoólicas por estudantes de medicina Use of alcoholic drinks by medicine students Uso de bebidas alcohólicas por estudiantes de medicina Izete Soares da Silva Dantas Pereira1, Rodrigo Alves de Melo2, Alison Jesissai Pinto Gurgel3, Ana Gabriela Miranda Barbosa4, Indira Coan Zanatta5, Sarah Lis Araújo Melo6, João Dantas Pereira7 http://dx.doi.org/10.30681/252610104307 ARTIGO ORIGINAL Uso de bebidas alcoólicas por estudantes de medicina Use of alcoholic drinks by medicine students Uso de bebidas alcohólicas por estudiantes de medicina Izete Soares da Silva Dantas Pereira1, Rodrigo Alves de Melo2, Alison Jesissai Pinto Gurgel3, Ana Gabriela Miranda Barbosa4, Indira Coan Zanatta5, Sarah Lis Araújo Melo6, João Dantas Pereira7 RESUMO Docente do curso de Medicina da Universidade do Estado do Rio Grande do Norte (UERN). Natal, Rio Grande do Norte, Brasil. E-mail: izetedantas@hotmail.com ORCID ID: https://orcid.org/0000-0002-2239-6582 Autor principal – Endereço para correspondência: Rua das Tílias, 2114, Cidade Satélite-Pitimbú-Natal, RN, Brasil. CEP 59067-690. 2Acadêmico de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: rodrigoamelo3@gmail.com ORCID ID: https://orcid.org/0000-0002-8070-6614 3Acadêmico de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: ajesissai@gmail.com ORCID ID: https://orcid.org/0000-0003-1105-2407 4Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: belinha0623@gmail.com ORCID ID: https://orcid.org/0000-0002-5130-8086 5Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: indiracoan@hotmail.com ORCID ID: https://orcid.org/0000-0002-4884-1381 6Acadêmica de Medicina. Universidade do Estado do Rio Grande do Norte (UERN). Mossoró, Rio Grande do Norte, Brasil. E-mail: sarah.lis.araujo@hotmail.com ORCID ID: https://orcid.org/0000-0001-8238-9993 7Assistente Social/Bacharel em Direito. Doutor em Sociologia. Docente do curso de Direito da Universidade Federal do Rio Grande do Norte (UFRN). Natal, Rio Grande do Norte, Brasil. E-mail: jotarn@hotmail.com ORCID ID: https://orcid.org/0000-0001-9855-827X Este artigo está licenciado sob forma de uma licença Creative Commons Atribuição 4.0 Internacional, que permite uso irrestrito, distribuição e reprodução em qualquer meio, desde que a publicação original seja corretamente citada. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. ISSN 2526-1010 Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. ISSN 2526-1010 242 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Uso de bebidas alcoólicas… Uso de bebidas alcoólicas… ABSTRACT Objective: to identify the frequency of alcohol consumption by medical students. Method: cross-sectional study with a quantitative approach. The AUDIT questionnaire applied to 128 students attending the first to eighth periods of the undergraduate medical course was used for data collection. Results: most did not have a high alcohol consumption. The use was mentioned only on festive occasions, without this having negative consequences on the academic routine. There was no relationship between sociodemographic variables and the amount and intensity of alcohol consumption. RESUMEN Objetivo: identificar la frecuencia del consumo de alcohol por parte de los estudiantes de medicina. Método: estudio transversal con enfoque cuantitativo. El cuestionario AUDIT aplicado a 128 estudiantes que asistieron a los períodos primero a octavo del curso de medicina de pregrado se utilizó para la recopilación de datos. Resultados: la mayoría no tenían un alto consumo de alcohol. El uso se mencionó solo en ocasiones festivas, sin que esto tenga consecuencias negativas en la rutina académica. No hubo relación entre las variables sociodemográficas y la cantidad e intensidad del consumo de alcohol. Consideraciones finales: Aunque todavía no es un problema en la realidad de los participantes, es de fundamental importancia invertir en acciones / pautas preventivas sobre los riesgos del uso de bebidas alcohólicas y el daño que pueden ocasionar al ejercicio de estos futuros profesionales. p f Descriptores: Bebidas Alcohólicas; Estudiantes de Medicina; Prevención RESUMO Final considerations: although it is not yet a problem in the reality of the participants, it is of fundamental importance to invest in preventive actions / guidance on the risks of using alcoholic beverages and the harm they can bring to the exercise of these future professionals. f f p f ptors: Alcoholic beverage; Medicine Students; Prevention. g f f p f Descriptors: Alcoholic beverage; Medicine Students; Prevention. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. INTRODUÇÃO drogas pode acarretar inúmeros prejuízos ao indivíduo e à sociedade, gerando problemas políticos, econômicos e sociais, afetando a qualidade de vida de todos os envolvidos2-4. O consumo de drogas, sejam elas lícitas ou ilícitas, tem aumentado nas últimas décadas, se caracterizando como uma questão de saúde pública no mundo1. O referido consumo é responsável por cerca de 4% das mortes, 5% dos casos de doenças e 90% das internações hospitalares por dependência química e outras comorbidades. O uso de Entre maio e outubro de 2015 foi realizada uma pesquisa em todo o Brasil com cerca de 17 mil pessoas com idades entre os 12 e os 65 anos, tendo como objetivo estimar e avaliar os parâmetros epidemiológicos do uso Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 243 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… jovens. Nesses se constatam mais complicações relacionadas às interações com professores, colegas e famíliares3,5,6. de drogas. Os resultados revelaram que 3,2% dos brasileiros usaram drogas ilícitas nos 12 meses anteriores à pesquisa. Esse dado equivale a 4,9 milhões de pessoas, sendo maior entre homens (5%) e menor nas mulheres (1,5%)3. O consumo afeta o rendimento acadêmico desses estudantes de forma negativa levando a ausência às aulas e avaliações, desconcentração durante as atividades acadêmicas, problemas de relacionamentos com professores, colegas e familiares. Ocasiona episódios de agressividade, níveis baixos de convívio social, risco de gravidez indesejável e doenças sexualmente transmissíveis, além de incitar a violência doméstica e no trânsito, provocando lesões e acidentes7-10. O uso abusivo de álcool e o alcoolismo, por exemplo, são, direta e indiretamente pelo surgimento de danos físicos, mentais e sociais11. O impacto que o uso de álcool tem na saúde das pessoas tende a ser cada vez maior pelo fato de ser uma droga lícita, de custo reduzido, de fácil acesso e ter como suporte campanhas publicitárias que estimulam o seu consumo. Observa-se um aumento significativo nas diferentes classes sociais e faixas etárias com iniciação nestas últimas, cada vez mais precoce2,4. Em 2015, a Organização Mundial da Saúde (OMS) fez uma projeção, por país, sobre o consumo de álcool puro e a estimativa para o Brasil fora de 9,1 litros por pessoa com 15 anos ou mais. No país, cerca de 70% da população faz uso das bebidas alcoólicas1. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. INTRODUÇÃO Segundo essa instituição, o álcool é a substância psicoativa mais consumida no mundo (2 bilhões de pessoas). Nesse contexto, há um público representado pelos jovens universitários. Há diferenças nos problemas decorrentes do consumo de álcool entre adultos e A expressão, alcoolismo tem sido empregada para designar o consumo regular e inadequado de bebidas alcoólicas ou o conjunto de consequências relacionadas ao consumo excessivo e prolongado do álcool, vontade incontrolável de beber, falta de controle ao tentar parar a ingestão, tolerância (doses cada vez maiores para sentir os mesmos efeitos da bebida)11. A Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 244 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… dependência física é uma manifestação de sintomas físicos e psíquicos nas situações de abstinência alcoólica. O entendimento sobre a etiologia do alcoolismo é amplo, incluindo fatores biológicos, psicológicos e sociais3-5,10,11. do público universitário entrevistado na pesquisa alegaram já ter consumido algum tipo de bebida alcoólica e 1/3 afirmou já ter realizado o consumo em binge em pelo menos uma ocasião. Relativamente ao álcool são identificados quatro padrões de consumo: 1) o moderado, sem risco; 2) o arriscado, que tem o potencial de produzir danos; 3) o nocivo, que se define por um padrão constante de uso associado a danos à saúde e 4) o consumo em binge, que diz respeito ao uso eventual de álcool em grande quantidade14,15,16. O ato do binge drinking vem crescendo no âmbito estudantil. O termo é utilizado para designar episódios nos quais há um consumo excessivo de bebida alcoólica. A quantidade que corresponde a esta prática seria de cinco ou mais doses de bebidas alcoólicas em uma única ocasião (no caso dos homens) ou quatro ou mais (para mulheres), sendo irrelevante a frequência dos episódios. A relevância dessa abordagem se prende ao fato de ser essa prática muito utilizada como “trote” para os calouros, provocando, muitas vezes, coma alcoólico e já ter levado à morte em casos graves14,17,18. Para fazer frente a essa realidade em algumas universidades já existem projeto ou programa de prevenção e/ou orientação para os estudantes envolvidos com o consumo problemático de bebidas alcoólicas11. Os jovens universitários são considerados mais vulneráveis ao consumo de bebidas alcoólicas, sendo estes uma das maiores preocupações referentes à saúde e ao comportamento do estudante cuja prática pode acarretar o alcoolismo9,12,13. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. INTRODUÇÃO Para o que interessa reter desta pesquisa sobre o uso do álcool entre estudantes de medicina são elucidativos os dados de um levantamento nacional referente ao uso de drogas por universitários brasileiros de instituições públicas e privadas, em todas as 27 capitais do país11. Esse estudo nacional apresentou os seguintes resultados: o álcool é a substância mais consumida nos últimos 12 meses. Além disso, 80% Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 245 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… O ingresso na universidade tem sido considerado um período crítico, principalmente para os calouros, com maior vulnerabilidade para o início e manutenção do consumo de álcool e outras drogas. Por outro lado, os calouros apresentam um risco maior de vivenciar as consequências relacionadas ao consumo de álcool. Nesse sentido, quanto mais tarde for o início do consumo de álcool, menores serão as probabilidades de se desenvolver problemas decorrentes desse comportamento, incluindo níveis de consumo abusivo ou, até mesmo, uma dependência20. Dentre as variáveis sócio culturais, a prática religiosa é um fator importante em relação à prevenção do alcoolismo. Estudo envolvendo jovens mostrou que a religião é um fator protetor ao consumo de drogas lícitas e ilícitas e ajuda o ser humano a se auto- conhecer, amenizando as angústias e dramas da vida e comportamentos degradantes. Os indivíduos com uma menor taxa de espiritualidade têm 3,3 vezes mais risco de um consumo de álcool problemático. Essa variável, a religião, pode agir como fator protetor para o uso de drogas entre estudantes adolescentes e jovens, possivelmente em razão das normas de conduta disseminadas pela maioria das religiões19. Alguns fatores estão relacionados com o consumo de álcool entre os estudantes como, por exemplo, ingresso precoce na universidade, área de formação universitária (humanas, tecnológico, biomédica, etc.), distância da família, falta de maturidade emocional, facilidade de acesso ao álcool, entre outros6,17,21. Ao ingressar na universidade muitos estudantes vivenciam novas experiências como se distanciar da família, residir com outros estudantes e passar a maior parte do tempo no ambiente universitário, além de adquirirem maior liberdade e independência para a tomada de decisões. Essas novas experiências podem favorecer um maior consumo de álcool, assim como os riscos associados a tal consumo2,5. Além dos fatores já mencionados, no caso dos estudantes de medicina, um dos fatores que contribuem para a iniciação e o uso constante e/ou abusivo de álcool é a grande exigência do rendimento Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 246 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… acadêmico. A competitividade existente entre eles, as extensas e cansativas práticas acadêmicas, longas horas de dedicação exclusiva ao curso, também devem ser consideradas. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Há uma quantidade excessiva de conteúdos que se traduz numa carga horária que, em muitas faculdades, ultrapassam as 9.000 horas. Além disso, tem-se a cobrança por um bom desempenho estudantil, culminando numa preocupação diária em conseguir notas adequadas para aprovação. Todos esses representam fatores estressores na rotina desses estudantes, o qual pode utilizar o consumo de álcool com o intuito de enfrentar os obstáculos existentes7,8,18,22,23. Do exposto, as universidades que registram baixo consumo de bebidas alcoólicas podem se constituir em espaços sociais estratégicos para a promoção de saúde desses grupos específicos, com consequente impacto positivo sobre a população em geral. Essa promoção da saúde no âmbito acadêmico implica em proporcionar condições necessárias aos estudantes para melhorar e exercer o controle sobre sua própria saúde15. A preocupação com a temática motivou o presente estudo que tem como objetivo identificar a frequência do consumo de álcool por estudantes do curso de medicina. MÉTODO Frente a esse cenário, as questões relacionadas ao consumo de álcool têm-se configurado como como um problema de saúde pública, visto a sua magnitude, vulnerabilidade e transcendência. Esse último termo é aqui entendido como o custo pessoal e social dos agravos à população, em que tudo que se relaciona ao processo saúde-doença-intervenção afeta diretamente as relações sociais, econômicas, profissionais e culturais20. Tratou-se de um estudo transversal, com abordagem quantitativa, realizada com 128 acadêmicos do 1º, 2º, 3º, 5º e 7º do curso de graduação em medicina de uma Universidade do Estado do Rio Grande do Norte. A escolha desse local ocorreu por ser uma Universidade pública e pela oportunidade da pesquisa em abordar uma temática ainda não estudada, nesse lócus. e Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 247 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoóli Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… presença do professor. Utilizou-se para a coleta de dados o questionário AUDIT (Alcohol Use Disorders Identification Test) e um formulário de coleta de informações elaborado pelos pesquisadores contendo informações sociodemográficas: sexo, faixa etária, estado civil, procedência, vínculo laboral e o semestre em que estavam matriculados. O tempo de aplicação do instrumento variou de 10 a 20 minutos, dependendo do número de estudantes na turma. No momento da pesquisa, o curso contava com 196 estudantes (100%), regularmente matriculados do 1º ao º12º períodos. No entanto, os estudantes do 4º, 6º e 8º períodos não participaram da pesquisa, devido à incompatibilidade e/ou indisponibilidade das respectivas turmas. Os estudantes do 9º ao 12º estavam no internato, o que dificultou o contato com os mesmos. presença do professor. Utilizou-se para a coleta de dados o questionário AUDIT (Alcohol Use Disorders Identification Test) e um formulário de coleta de informações elaborado pelos pesquisadores contendo informações sociodemográficas: sexo, faixa etária, estado civil, procedência, vínculo laboral e o semestre em que estavam matriculados. O tempo de aplicação do instrumento variou de 10 a 20 minutos, dependendo do número de estudantes na turma. Assim sendo, do total previsto, ou seja, o universo, apenas 128 (65,3%) acadêmicos, de ambos os sexos, responderam aos instrumentos. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. MÉTODO O AUDIT é um modelo de questionário desenvolvido pela Organização Mundial de Saúde (OMS), adaptado para uso no Brasil. Esse instrumento inclui perguntas sobre as consequências do consumo, quantidade e frequência de ingestão de álcool. Tem como característica ser um recurso autoaplicável e de fácil preenchimento, não requisitando um treinamento prévio especializado. É composto por 10 itens capazes de avaliar o consumo de álcool nos últimos 12 meses. Cada um dos itens tem 5 respostas possíveis, pontuadas de 0 a 4 (exceto os itens 9 e 10, que pontuam apenas 0, 1 e 2), variando o escore total de 0 a 36. De acordo com o resultado obtido, é possível O AUDIT é um modelo de questionário desenvolvido pela Organização Mundial de Saúde (OMS), adaptado para uso no Brasil. Esse instrumento inclui perguntas sobre as consequências do consumo, quantidade e frequência de ingestão de álcool. Tem como característica ser um recurso autoaplicável e de fácil preenchimento, não requisitando um treinamento prévio especializado. É composto por 10 itens capazes de avaliar o consumo de álcool nos últimos 12 meses. Cada um dos itens tem 5 respostas possíveis, pontuadas de 0 a 4 (exceto os itens 9 e 10, que pontuam apenas 0, 1 e 2), variando o escore total de 0 a 36. De acordo com o resultado obtido, é possível De acordo com os critérios estabelecidos para este estudo, foram incluídos todos os estudantes regularmente matriculados no 1º ao 12º período do curso de medicina e excluídos aqueles que frequentavam disciplinas especiais no curso de medicina, bem como os da Pós- Graduação e os acadêmicos que por algum motivo não estiveram presente no momento da coleta de dados, como, por exemplo, por estarem afastados da sala de aula por licença médica/maternidade. Os dados foram coletados nas salas de aula teóricas durante o primeiro semestre de 2018, antes do início ou após o término, sem a o Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 248 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… classificar o respondente dentro de 4 padrões de consumo de álcool ou zonas de risco, sendo elas: 1) uso de baixo risco (escore de 0 a 7 pontos); 2) uso de risco (8 a 15 pontos); 3) risco nocivo (16 a 19 pontos) e 4) provável dependência (20 ou mais pontos)15,19,24. RESULTADOS E DISCUSSÃO Os dados a seguir apresentados incluem as informações sobre o perfil sociodemográfico dos acadêmicos e os resultados dos questionários (AUDIT) aplicados. Após a coleta, os dados foram armazenados em planilhas do Excel versão 20.0 para posterior processamento. Para a análise, utilizou-se a estatística descritiva com frequências absolutas e relativas, apresentados em forma de gráfico e tabela. Os acadêmicos de medicina estiveram distribuídos nos seguintes períodos: 21,0% do primeiro; 24,0% do segundo; 15,0% do terceiro; 24,0% do quinto e 16,0% do sétimo, totalizando 128 estudantes (65,3%). Na Tabela 1, observa-se que do total de estudantes do curso de medicina que participaram deste estudo, 56,25% são do sexo feminino. A faixa etária predominante foi representada por aqueles com idade compreendida entre os 18 e os 25 anos. Em relação ao estado civil 92,97% são solteiros. A maioria, 82,03%, é procedente de outros estados da federação. Quanto ao quesito ocupação, apenas, 7,03% desempenham alguma atividade remunerada. Os demais se dedicam, exclusivamente, aos estudos. A pesquisa foi submetida ao Comitê de Ética em Pesquisa da Universidade do Estado do Rio Grande do Norte-CEP/UERN, sob o número CAAE 25736619.7.0000.5294 e aprovada pelo Parecer nº 1.981.033. Foram respeitados, rigorosamente, todos os princípios e as diretrizes éticas de pesquisa envolvendo seres humanos, em atendimento à Resolução 466/2012 do Conselho Nacional de Saúde e suas complementares. Todos os estudantes que aceitaram participar do estudo assinaram o Termo de Consentimento Livre e Esclarecido (TCLE). Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 249 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Tabela 1 - Perfil sociodemográfico dos estudantes. Mossoró - RN, Brasil. 2018. N % Sexo Feminino 72 56,25 Masculino 56 43,75 Faixa etária 18 a 25 anos 96 75,00 25 a 35 anos 30 23,44 35 a 45 anos 01 0,78 45 a 55 anos 01 0,78 Estado civil Solteiro 119 92,97 Casado 09 7,03 Procedência Mossoró 23 19,97 Outros estados 105 82,03 Ocupação Sim 09 7,03 Não 119 92,97 Total 128 100,00 Tabela 1 - Perfil sociodemográfico dos estudantes. Mossoró - RN, Brasil. 2018. faixa etária dos 18 aos 34 anos, consumiram álcool e tabaco nos doze meses anteriores à pesquisa. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. RESULTADOS E DISCUSSÃO Desses, cerca de 400 mil fizeram uso do álcool e, pelo menos, uma substância ilícita. Em relação ao sexo, a dependência do álcool foi 3,4 vezes mais frequente entre os homens do que entre as mulheres, no mesmo período. Contudo, as mulheres apresentaram prevalência superior à dos homens no consumo de álcool quando aliado a, pelo menos, um medicamento não prescrito. O uso de álcool associado a uma droga ilícita ou a um medicamento não prescrito pareceu estar diretamente relacionado ao nível de escolaridade, ou seja, à medida que o nível de escolaridade se A associação de variáveis sociodemográficos com o consumo de bebidas alcoólicas tem sido considerada em vários estudos realizados no Brasil e em outros países. Estudos têm mostrado que a quantidade e o padrão de consumo de álcool variam na população de acordo com idade, etnia/cor da pele, condição socioeconômica e/ou educação, trabalho, estado civil e características da vizinhança, sendo que essas podem atuar de maneira diferente entre homens e mulheres e apresentar resultados divergentes2,6,14,15,18,25. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. divergentes2,6,14,15,18,25. 251 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Tabela 2 - Consumo de bebidas alcoólicas pelos estudantes de medicina. Mossoró - RN, Brasil. 2018. (n=128) Consumo N (%) Nunca 43 33,60 Uma vez por mês ou menos 49 38,28 Duas a quatro vezes por mês 34 26,56 Duas a três vezes por semana 02 1,56 Total 128 100,00 Tabela 2 - Consumo de bebidas alcoólicas pelos estudantes de medicina. Mossoró - RN, Brasil. 2018. (n=128) entre desempenho acadêmico e consumo de álcool entre os alunos de sete faculdades da Universidade de Turku, envolvendo 1177 estudantes de graduação. Os resultados mostraram que os alunos relataram um consumo de longa duração e grande quantidade de bebida em torno de 50%, alta frequência de bebida (41%), consumo episódico pesado em 66%, problema de consumo de álcool (29%) e possível dependência de álcool para 9% dos participantes27. Ao comparar uma pesquisa realizada sobre o uso de drogas entre estudantes de medicina com universitários de outros cursos, os resultados mostraram que aqueles consumiam mais bebidas alcoólicas do que os dos outros cursos25. Portanto, mesmo conhecendo os danos potenciais e diretos que essa substância pode causar ao próprio organismo e à sociedade em geral, os estudantes de Medicina ainda buscam alguma utilidade no uso dessa droga, ficando sujeitos e vulneráveis às consequências desse consumo26. A preocupação com o uso nocivo de álcool por parte da população estudantil nos vários países ganhou destaque no relatório da OMS que apresentou resultados de estudos americanos, mostrando que 2/3 da população de universitários dos Estados Unidos consumiam álcool e que cerca de metade o fazia em um padrão excessivo1. Apesar de ser lícito, o consumo de álcool pode ser uma “porta de entrada” para as drogas ilícitas. Além disso, o uso A literatura consultada mostrou que essa não é apenas a realidade do Brasil. Um estudo similar realizado na Universidade da Cantábria (Espanha) encontrou um padrão de consumo do álcool de pouco mais da metade da população entre acadêmicos de medicina e enfermagem17. A literatura consultada mostrou que essa não é apenas a realidade do Brasil. divergentes2,6,14,15,18,25. O 3º Levantamento Nacional sobre o uso de drogas pela população brasileira destacou que mais de um milhão de adolescentes e jovens, na Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 250 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… eleva, são maiores as prevalências de consumo múltiplo dessas drogas3. consumo de álcool e outras drogas. Um ambiente de maior liberdade pode facilitar o acesso e o consumo dessas drogas2,21,23. Esses resultados sinalizam no sentido de que o consumo de drogas, entre elas o álcool, está disseminado na população jovem e que independe do gênero. Em relação à variável ocupação, a Pesquisa Nacional de Saúde (2013), mostrou que os estudantes que tinham algum vínculo laboral apresentaram um percentual de baixo risco para o consumo de bebidas alcoólicas quando comparados àqueles que não exerciam atividade laboral6. Os resultados apresentados pela Pesquisa Nacional de Saúde realizada em 2013 corroboram que uma união estável constitui um dos fatores protetores/inibidores do consumo de álcool entre as mulheres6. Estudo desenvolvido na Colômbia mostrou que o consumo de álcool pelos jovens universitários era maior entre as pessoas solteiras que se encontram nos primeiros anos da faculdade. Nos homens, essa conduta vai se mantendo ao longo do curso e aumentando a prevalência19. No que diz respeito à frequência com que os estudantes de medicina consomem bebidas alcoólicas, neste estudo, 33,60% afirmaram ser abstêmios, enquanto 66,41% referiram beber de uma a quatro vezes por mês (Tabela 2). O valor identificado nesta pesquisa é semelhante aos números encontrados em um estudo realizado na capital do Distrito Federal, onde 33,60% dos participantes também se declararam abstêmios9. Esse percentual representa um valor, consideravelmente, inferior ao de outros estudos feitos entre estudantes de cursos universitários no Brasil, nos quais a prevalência do consumo de álcool variou de 75 a 93,7%2,7,8,18,21. Neste estudo a procedência dos participantes é majoritariamente de outros estados da federação, o que significa que moram sozinhos, com amigos ou ainda em pensionatos, longe de casa, sem o apoio da família. Diversas pesquisas realizadas no Brasil entre estudantes mostraram que morar sozinho ou com amigos, longe da vigilância dos pais ou de outros familiares pode aumentar, consideravelmente, os riscos para o Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. divergentes2,6,14,15,18,25. Um estudo similar realizado na Universidade da Cantábria (Espanha) encontrou um padrão de consumo do álcool de pouco mais da metade da população entre acadêmicos de medicina e enfermagem17. Na Finlândia, realizou-se uma pesquisa para conhecer a relação Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 252 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… associado do álcool com drogas ilícitas pode intensificar os efeitos deletérios dessas drogas quando comparado ao consumo isolado20,28,29. para que seja possível alcançar a meta previamente estipulada de redução relativa de 10% no consumo nocivo mundial de álcool em 2025, segundo a OMS1. Na Etiópia um estudo transversal realizado entre 725 estudantes universitários em 2015, constatou o uso problemático de álcool esteve associado à fobia social, baixo desempenho acadêmico, uso na vida e uso de outras drogas30. A relação entre o uso de bebida alcoólica e o sexo apresentado na tabela 3, mostra que neste estudo houve um ligeiro predomínio do sexo feminino em relação ao masculino, o que vem divergindo da maior parte dos estudos, onde essa relação (mulheres vs consumo de álcool) não é estatisticamente significativa. O uso do álcool tem sido mais frequente entre os estudantes do sexo masculino, embora esse aspecto não seja consensual entre os autores2,10,15. As bebidas alcoólicas foram aquelas que em vários estudos apresentaram um maior número de usuários em todos os níveis de risco, independentemente dos instrumentos de coleta de dados utilizados e dos tipos de drogas lícitas e/ou ilícitas consumidas1,5,11,23. A prevenção e a redução do uso nocivo do álcool devem ser tratadas como prioridade, Tabela 3- Relação entre gênero e risco de acordo com a classificação AUDIT. Mossoró - RN, 2018. (n=128) Sexo Risco I Risco II Risco III Risco IV Total Feminino 63 07 02 00 72 Masculino 47 08 01 00 56 Total 110 15 03 00 128 entre gênero e risco de acordo com a classificação AUDIT. Mossoró - RN, apresentavam uso regular (um ou mais vezes por semana) de consumo de álcool foram 27%, aumentando para 38% em 2012, enquanto que os homens aumentaram de 54% para 63%13. Esses números chamam atenção Ainda em relação ao sexo, o II Levantamento Nacional de Álcool e Drogas - 2012 comparou resultados do consumo de álcool entre homens e mulheres em dois períodos, 2006 e 2012. No ano de 2006, mulheres que Journal Health NPEPS. divergentes2,6,14,15,18,25. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 253 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Um estudo realizado na Colômbia mostrou que para cada oito homens que consomem bebidas alcoólicas há três mulheres19. Têm-se observado que nos últimos anos, no Brasil, registra-se uma tendência gradual, por parte das mulheres, em aumentar o consumo e de forma mais precoce3,11. para um crescimento do etilismo no sexo feminino, hábito que, em décadas passadas, era observado apenas no público masculino3. Os resultados de um estudo realizado com estudantes do curso de medicina de uma Universidade do Sul Fluminense, utilizando a mesma metodologia do AUDIT mostrou que o hábito de beber no primeiro semestre, em ambos os gêneros, aumentou com o decorrer do curso. A prática de beber pesado foi superior nos homens em comparação às mulheres28. Foi realizado um estudo analítico transversal, utilizando teste de triagem de envolvimento de álcool, tabagismo e outras substâncias (ASSIST) em dois grupos populacionais com o objetivo de calcular a prevalência de risco de uso de substâncias psicoativas entre estudantes, professores e trabalhadores de uma universidade pública do Equador. Os resultados mostraram a prevalência de consumo de alto risco para álcool de 5,3% para estudantes e 4,9% para trabalhadores e professores. O tabaco e a cannabis foi alta nos grupos estudados23. Embora não se enquadre nos critérios de dependência, existe uma probabilidade maior das pessoas que fazem uso de consumo nocivo de álcool não cumprirem as tarefas habituais e de se envolverem em atos de violência17. Outros estudos nacionais e internacionais concluíram que pessoas que aderem a comportamentos saudáveis apresentam maior longevidade, redução da vulnerabilidade ao adoecer e da morbimortalidade geral, além da diminuição do impacto financeiro sobre o sistema de saúde e de efeitos adversos em sua qualidade de vida15,17,19,20,23,27,30,31. Ao se comparar estudos realizados no Brasil com outros similares em países como a Espanha, Colômbia, Equador, Finlândia, Turquia já citados, anteriormente, que utilizaram método semelhante, constataram-se resultados parecidos no consumo de álcool pelos Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 254 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. divergentes2,6,14,15,18,25. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… sobre o álcool e priorizar os grupos de estudantes em risco, envidando esforços mais eficientes15,20. estudantes17,19,23,27,30. Naqueles países o consumo de álcool também é considerado alto com Riscos III e IV na escala AUDIT, ao contrário do que se verificou no presente estudo24. O uso problemático de álcool entre estudantes de medicina é preocupante, pois há a possibilidade desse uso permanecer na atuação profissional desses futuros médicos. Essa prática pode interferir na habilidade em fazer diagnóstico precoce, encaminhamento e tratamento de seus pacientes dependentes ou não do álcool3,6,11,16,28. Essa preocupação é agravada pelo fato de que o médico servirá de modelo para seus pacientes. Além disso, a ideia de que o médico apenas indicaria o caminho a ser seguido sem, necessariamente, o fazer é contraditória e pode tornar o tratamento com baixa adesão por parte dos pacientes. O consumo de bebidas alcoólicas, por parte de qualquer pessoa cria risco para dependência dessas substâncias, o que se caracteriza como um grande problema social e de saúde3,6,11,16. Neste estudo, nenhum dos participantes, apresentou padrão de consumo associado ao nível IV, ou seja, indicativo de provável dependência de álcool. Esse critério de dependência pode ser constatado quando existe um padrão de uso em que a pessoa desenvolve mudanças fisiológicas, cognitivas e comportamentais como desejo intenso de consumir bebidas alcóolicas; falta de controle sobre o consumo; continuar consumindo apesar das consequências; dar maior prioridade ao consumo do que às outras atividades; aumento da tolerância e presença da síndrome de abstinência com a descontinuidade do consumo21. Os autores são consensuais que as universidades precisam avaliar os problemas de bebida e distúrbios provocados pelo uso de álcool entre os estudantes. São necessárias medidas para definir e implementar estratégias de prevenção para reduzir os riscos decorrentes dessa prática. Os esforços de promoção da saúde podem se concentrar nas crenças e expectativas As consequências do uso de álcool, além de todo o impacto negativo para o indivíduo, família e amigos acabam por onerar a sociedade. De forma direta e indireta, potencializam os custos em hospitais e Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 255 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… outros dispositivos do sistema de saúde, sistema judiciário, previdenciário, perda de produtividade do trabalho, absenteísmo, desemprego, entre outros. divergentes2,6,14,15,18,25. Em todo o mundo, nota-se que as faixas etárias mais jovens são as principais afetadas em relação às mortes associadas ao uso do álcool, traduzindo como uma maior perda de pessoas na fase mais produtiva em que estão economicamente ativas31. através da literatura, que o consumo de álcool e de outras drogas tem crescido no país, entre estudantes da área da saúde, sobretudo, entre os de medicina. Apesar de ainda não se constituir um problema na realidade dos participantes, torna-se de fundamental importância investir em ações/orientações preventivas sobre riscos do uso de bebidas alcoólicas e os malefícios que podem trazer para o exercício desses futuros profissionais. Essas estratégias se tornam pertinentes, principalmente, pelo fato de esses indivíduos serem futuros especialistas na área da saúde e, portanto, profissionais que orientarão outras pessoas acerca deste assunto. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. CONSIDERAÇÕES FINAIS A pesquisa mostrou que a maioria dos participantes faz uso de álcool esporadicamente, sendo considerado de baixo risco (Risco I) e um terço se declarou abstêmio. O consumo de bebidas alcoólicas ocorre de forma pontual, em festas ou comemorações, com frequência média de duas a quatro vezes por mês. Tal padrão de uso, configura-se como oportunidade para que o contexto em que esses estudantes estão inseridos, seja um espaço de promoção e prevenção de problemas e agravos à saúde, relacionados ao uso e abuso de álcool. Essa constatação é muito importante e positiva para a realidade estudada na medida em que se sabe, Como limitação da pesquisa, pode-se citar a ausência de alguns estudantes de outros períodos letivos durante a coleta dos dados. Isso pode estar relacionado ao fato de que a coleta ocorreu durante o período em que a instituição esteve com grande parte das suas atividades paralisadas por reivindicação salarial e melhores condições de trabalho dos servidores. Desta forma, salienta-se que o estudo pode ter subestimado a prevalência daqueles que fazem o consumo de álcool nas categorias de Risco III e IV. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 256 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Álcool: uso por estudantes de medicina da Universidade Federal do Tocantins. Rev Cereus. 2015; 7(1):26-39. Outra limitação a ser considerada é o fato de que o formulário sociodemográfico utilizado não incluiu a religião como uma variável a ser identificada, uma vez que essa é um importante fator de proteção ao consumo de álcool e de outras drogas lícitas e ilícitas. Muitas pessoas que não fazem uso de bebidas alcóolicas associam essa decisão à sua prática religiosa. 3. Bastos FIPM, Vasconcellos MT, De Boni RB, Reis NB, Coutinho CFS. 3º Levantamento Nacional sobre o uso de drogas pela população brasileira. Rio de Janeiro: FIOCRUZ/ICICT; 2017. 4. Souza LGS, Menandro MCS, Menandro PRM. O alcoolismo, suas causas e tratamento nas representações sociais de profissionais de Saúde da Família. Physis. 2015; 25(4):1335-1360. Sugerem-se a realização de outros estudos, com metodologias quantitativas em que sejam avaliados os índices de consumo do álcool e sua associação com outras drogas entre acadêmicos de outros cursos e outras Instituições de Ensino Superior. Esses estudos poderão ampliar os conhecimentos sobre os padrões de uso dessas drogas. 5. Vecchia MD. Notas sobre a conjuntura da política de drogas no Brasil. CONSIDERAÇÕES FINAIS J Health NPEPS. 2018; 3(2):298-303. 6. Machado IE, Monteiro MG, Malta DC, Lana FCF. Pesquisa Nacional de Saúde 2013: relação entre uso de álcool e características sociodemográficas segundo o sexo no Brasil. Rev bras epidemiol. 2017; 20(3):408-422. REFERÊNCIAS 1. Organização Mundial da Saúde. Relatório Global sobre Álcool e Saúde 2018. Genebra, Suiça, 2018. Disponível em: Disponível em: https://cisa.org.br/index.php/pesq uisa/dados-oficiais. Acesso em: 12 jun 2019. 7. Pinheiro MA, Torres LF, Bezerra MS, Cavalcante RC, Alencar RD, Donato AC, et al. Prevalência e fatores associados ao consumo de álcool e tabaco entre estudantes de medicina no Nordeste do Brasil. Rev 2. Ribeiro GFF, França VM, Faria RLBC, Cuellar PMG, Martins MLB. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 257 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… bras educ med. 2017; 41(2):231- 250. Correia DS, Freitas DA. Uso de Bebidas Alcoólicas entre Acadêmicos da Área de Saúde. Rev Bras Educ Med. 2016; 40(3):446- 45. 8. Nascimento MI, Costa JS, Pereira MA, Kiepper MS, Keher NB, Moraes RFS. Uso de álcool por estudantes de medicina segundo características de cursos e escolas médicas: uma revisão da literatura. Rev bras educ med. 2019; 43(1 supl.):98-107. 8. Nascimento MI, Costa JS, Pereira MA, Kiepper MS, Keher NB, Moraes RFS. Uso de álcool por estudantes de medicina segundo características de cursos e escolas médicas: uma revisão da literatura. Rev bras educ med. 2019; 43(1 supl.):98-107. 13. Laranjeira R, Madruga CS, Pinsky I, Caetano R, Ribeiro M, Mitsuhiro S. II Levantamento Nacional de Álcool e Drogas- Consumo de Álcool no Brasil: Tendências entre 2006/2012. São Paulo: INPAD; 2013. 13. Laranjeira R, Madruga CS, Pinsky I, Caetano R, Ribeiro M, Mitsuhiro S. II Levantamento Nacional de Álcool e Drogas- Consumo de Álcool no Brasil: Tendências entre 2006/2012. São Paulo: INPAD; 2013. 9.Trindade BPA, Diniz AV, Sá-Júnior AR. Uso de drogas entre estudantes universitários: uma perspectiva nacional. Rev Med Saude Brasilia. 2018; 7(1):52-60. 14. Pelicioli M, Barelli C, Gonçalves CBC, Hahn S, Scherer JI. Perfil do consumo de álcool e prática do beber pesado episódico entre universitários brasileiros da área da saúde. J Bras Psiquiatr. 2017; 66(3):150-156. 10. Fernandes TF, Monteiro BMM, Silva JBM, Oliveira KM, Viana NAO, Gama CAP, et al. Uso de substâncias psicoativas entre universitários brasileiros: perfil epidemiológico, contextos de uso e limitações metodológicas dos estudos. Cad saúde colet. 2017; 25(4):498-507. 15. Raposo JCS, Costa ACQ, Valença PAM, Zarzar PM, Diniz AS, Colares V, et al. Uso de drogas ilícitas e binge drinking entre estudantes adolescentes. Rev Saúde Pública. 2017; 4(51):51-83. 11. Brasil. Presidência da República. Secretaria Nacional de Políticas sobre Drogas. REFERÊNCIAS I Levantamento Nacional sobre o Uso de Álcool, Tabaco e Outras Drogas entre Universitários das 27 Capitais Brasileiras. Brasília: SENAD; 2010. 16. Mendonça AKRH, Nogueira MSN, Andrade RLB, Azevedo DX, Jesus CVF, Lima SO. Binge drinking entre estudantes de medicina de universidades particular e pública de um estado do nordeste brasileiro. Temas saúde. 2018; 18(3):30-53. 12. Oliveira SKM, Sousa ÁAD, Cavalcanti SL, Taveira MGMM, Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. 258 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Rev Eletrônica Saúde Mental Álcool Drog. 2017; 13(1): 22-9 17. Herrero-Montes M, Alonso-Blanco C, Paz-Zulueta M, Sarabia-Cobo C, Ruiz-Azcona L, Parás-Bravo P. Binge Drinking in Spanish University Students: Associated Factors and Repercussions: A Preliminary Study. Int J Environ Res Public Health. 2019; 16:4822. 22. Silva CF, Oliveira AP, Mundim FL, Alves TA, Silva NA, Silva LLC, et al. Consumo de álcool e fatores de risco associados entre estudantes de uma Universidade Federal brasileira, 2014. Rev Bras Geo Med Saúde. 2016; 12(23):1-10. 18. Maia DAM, Marques RB, Maia Filho AM. Consumo de bebidas alcoólicas e a prática do binge drinking em acadêmicos de medicina. Rev Interdisciplin. 2017; 10(1):139-146. 23. Piedra S, Narváez A, Jácome P, Terán R, Barreto D, Aguirre R, et al. Consumo de riesgo de drogas utilizando la herramienta ASSIST modificado en estudiantes, docentes y personal administrativo de una universidad pública. Rev Med Vozandes. 2019; 30(2):19–25. 19. Yañez-Castillo BG, VIllar-Luis MA, Alonso-Castillo MM. Espiritualidad, autotrascendencia y consumo de alcohol en jóvenes universitarios. J Health NPEPS. 2018; 3(1):5-21. 24. MorettI-Pires RO, Corradi-Webster CM. Adaptação e validação do Alcohol Use Disorder Identification Test (AUDIT) para população ribeirinha do interior da Amazônia, Brasil. Rio de Janeiro. Cad Saúde Pública. 2011; 27(3):497-509. 20. Porto NT, Ferreira DS, Figueiredo GLA. Comportamento de universitários em relação ao consumo de álcool, tabaco e outras drogas: subsídios para ações promotoras de saúde. Rev Educ Cult Contemp. 2019; 16(42):104-121. 25. Monteiro LZ, Varela AR, Carneiro MLA, Alves LR, Góis RFG, Lima TB. Uso de tabaco e álcool entre acadêmicos da saúde. Rev Bras Promoç Saúde. 2018; 31(1):1-9. 21. Morales GA, Galera SAF, Reyes AT, Aguila SRG, Arroyo MLA, Castillo FAM. Factores de riesgo para el consumo de alcohol en adolescentes estudiantes. SMAD, 26. Toledo L, Coutinho C, Bastos FI. Panorama sobre a política de Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. Journal Health NPEPS. 2020 jan-jun; 5(1):242-260. REFERÊNCIAS Participação dos autores:  Concepção: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Desenvolvimento: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Redação e revisão: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD. Participação dos autores:  Concepção: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Desenvolvimento: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Redação e revisão: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD. REFERÊNCIAS 259 Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas… Uso de bebidas alcoólicas… drogas e saúde mental no Brasil contemporâneo: prevenção e tratamento. Rio de Janeiro: Fundação Oswaldo Cruz; 2020. 29. FinellI LAC, Santos ARS, Ruas LF, Soares WD. Consumo de álcool por estudantes de engenharia biomédica. Rev Bionorte. 2017; 6(2):1-12. 27. El-Ansari W. Abdul Salam A, Suominen S. Is Alcohol Consumption Associated with Poor Perceived Academic Performance? Survey of Undergraduates in Finland. J Environ Res Public Health. 2020; 17(4):1369. 30. Mekonen T, Fekadu W, Chane T, Bitew S. Problematic Alcohol Use among University Students. Front Psychiatry. 2017; 8(86):1-5. 31. Silva EC, Tucci AM. Padrão de consumo de álcool em estudantes universitários (calouros) e diferença entre os gêneros. Temas Psicol. 2016; 24(1):313-323. 28. Gomes LS, Barroso CRD, Silvestre VA, Baylão ACP, Garcia SCM, Pacheco SJB. Consumo de álcool entre estudantes de medicina do Sul Fluminense - RJ. Rev Med (São Paulo). 2018; 97(3):260-6. 28. Gomes LS, Barroso CRD, Silvestre VA, Baylão ACP, Garcia SCM, Pacheco SJB. Consumo de álcool entre estudantes de medicina do Sul Fluminense - RJ. Rev Med (São Paulo). 2018; 97(3):260-6. Conflito de interesses: Os autores declaram não haver conflito de interesses. Conflito de interesses: Os autores declaram não haver conflito de interesses. Participação dos autores:  Concepção: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Desenvolvimento: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Redação e revisão: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD. Participação dos autores:  Concepção: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Desenvolvimento: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Redação e revisão: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD. Participação dos autores:  Concepção: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Desenvolvimento: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD.  Redação e revisão: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, Pereira JD. Participação dos autores: Como citar este artigo: Pereira ISSD, Melo RA, Gurgel AJP, Barbosa AGM, Zanatta IC, Melo SLA, et al. Uso de bebidas alcoólicas por estudantes de medicina. J Health NPEPS. 2020; 5(1):242-260. Submissão: 27/02/2020 Aceito: 15/05/2020 Publicado: 01/06/2020 Submissão: 27/02/2020 Aceito: 15/05/2020 Publicado: 01/06/2020 Submissão: 27/02/2020 Aceito: 15/05/2020 Publicado: 01/06/2020 260 Journal Health NPEPS. 2020 jan-jun; 5(1):242-260.
https://openalex.org/W2084494533	https://eejournal.ktu.lt/index.php/elt/article/download/1373/1290	English	null	Using Higher Order Nonlinear Operators for SVM Classification of EEG Data	Elektronika ir elektrotechnika	2,012	cc-by	3,557	Introduction classification such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), Bayesian Linear Discriminant Analysis (BLDA), Support Vector Machine (SVM), Fisher Linear Discriminant Analysis (FLDA), etc. [3]. Usually, the raw EEG data is pre- processed using the DSP methods such as Fourier analysis or wavelet transform for denoising and the filtered signal is used as input data for the classification method. Brain-computer interface (BCI) is a communication system that translates brain activity into commands for a computer or other digital devices [1]. The major goal of BCI research is to develop systems that allow disabled users to communicate with other persons, to control artificial limbs, or to control their environment. Other applications include multimedia communication, Augmented Reality applications, robot control and game development. Recently, the non-linear operators such as the Teager- Kaiser energy operator (TKEO) [4, 5] have attracted the attention of researchers in the BCI domain. The TKEO so far has been applied in speech recognition [6], for image enhancement [7] and in EEG data analysis for detecting high frequency oscillations [8], mental task classification [9], and sleep spindle detection [10]. p The majority of BCI systems work by reading and interpreting cortically evoked electro-potentials across the scalp via an electro-encephalogram (EEG). The EEG signal has become the main data source of BCI study due to its low cost and non-invasive nature. The EEG data is inherently complex and difficult to analyze. Oscillatory activity in the EEG is classified into different frequency bands or rhythms: delta (0.5–3.5 Hz), theta (4–8 Hz), alpha 1 (8–10.5 Hz), alpha 2 (10.5–13 Hz), beta 1 (13–21 Hz), beta 2 (20–32 Hz), and gamma (36–44 Hz) [1]. Because EEG signals are non-stationary and nonlinear, and normally interfered by eye movements and muscle noises, it is difficult to differentiate the classes of mental tasks from EEG [2]. Different features can be extracted from the EEG data such as: time domain features related to changes in the amplitude of neurophysiologic signals, occurring time-locked to the presentation of stimuli or time-locked to actions of the user of a BCI, frequency domain features related to changes in oscillatory activity, and spatial domain features extracted and combined from several electrodes [1]. The structure of the remaining parts of the paper is as follows. Section II discusses the Teager-Kaiser energy operator. Section III proposes and describes the proposed higher order nonlinear operator. Using Higher Order Nonlinear Operators for SVM Classification of EEG Data I. Martisius, R. Damasevicius, V. Jusas, D. Birvinskas Software Engineering Department, Kaunas University of Technology, e-mails: ignas.martisius@stud.ktu.lt, robertas.damasevicius@ktu.lt, vacius.jusas@ktu.lt, darius.birvinskas@stud.ktu.lt I. Martisius, R. Damasevicius, V. Jusas, D. Birvinskas Software Engineering Department, Kaunas University of Technology, e-mails: ignas.martisius@stud.ktu.lt, robertas.damasevicius@ktu.lt, vacius.jusas@ktu.lt, darius.birvinskas@stud.ktu.lt http://dx.doi.org/10.5755/j01.eee.119.3.1373 classification such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), Bayesian Linear Discriminant Analysis (BLDA), Support Vector Machine (SVM), Fisher Linear Discriminant Analysis (FLDA), etc. [3]. Usually, the raw EEG data is pre- processed using the DSP methods such as Fourier analysis or wavelet transform for denoising and the filtered signal is used as input data for the classification method. Introduction Section IV describes data classification using Support Vector Machine. Section V presents a case study. Section VI presents conclusions and outlines future work. Classification using Support Vector Machine           ш k k i n x n x 1  . (6) (6) Support Vector Machines (SVM) [15] is a binary classification algorithm based on structural risk minimization. First, the SVM implicitly maps the training data into a (usually higher-dimensional) feature space. A hyper-plane (decision surface) is then constructed in this feature space that bisects the two categories and maximizes the margin of separation between itself and those points lying nearest to it (the support vectors). This decision surface can then be used as a basis for classifying vectors of unknown classification. A combination of (4) and (5) operators is proposed in [13]            m m m i i n x i n x n x n x / 1 / 2      . (7) (7) A generalization of the continuous TKEO as the higher-order energy operator (HOEO) k  is proposed in [14] Consider an input space X with input vectors X xi  , a target space   1 ,1   Y with Y yi  and a training set       N N y x y x T , ,..., , 1 1  . In SVM classification, separation of the two classes   1 ,1   Y is done by the maximum margin hyper-plane, i.e. the hyper-plane that maximizes the distance to the closest data points and guarantees the best generalization on new examples. In order to classify a new point j x , the classification function  j x g is used     k k k k k t x t x t x t x t x             1 1 . (8) (8) For discrete-time series, the HOEO can be rewritten as the discrete energy operator (DEO) [Maragos]:         m k n x m n x k n x n x n x km        . (9) (9) The advantage of the TKEO family of operators over the traditional DSP analysis methods such as Fourier Transform or wavelet analysis is the ability of the TKEO to discover high-frequency low-amplitude components [8] in analyzed data. Teager-Kaiser Energy Operator Nonlinear models are systems where either the additivity or the scalability properties do not hold in general           t g H t f H t g t f H    ,       t f H t f H      , (1) where t g t f , are signals, H is a nonlinear model (operator), and is a constant. BCI systems require correct classification of signals interpreted from the brain for useful operation. After acquiring the EEG data, the pre-processing (filtering, de- noising), feature extraction, and dimensionality reduction is performed, before machine learning algorithms can be applied to learn from a training dataset how to classify the signals into classes, where each class corresponds to a specific action of the user. A variety of machine learning and artificial intelligence methods are used for EEG data The Teager-Kaiser Energy Operator (TKEO), proposed by Teager [4] and further investigated by Kaiser [5], is a special case of nonlinear models. For a continuous real-valued signal t x , the TKEO    t x  is defined as follows     2 2 2 t x t x t x t x               . (2) (2) 99       n x n x 2 3    , where              0 0 2 / 1 0 1 0 2 / 1 0 0 A . (12) An approximation of the derivatives by one-sample differences provides the definition of the TKEO for the discrete-time signal [5]       1 1 2      n x n x n x n x . (3) (3) Such operator also can be seen as a special case of the 2D Volterra system as noted by Kvedalen [13]. Moore et al. Teager-Kaiser Energy Operator [11] propose a generalization of the Teager operator as 1-D Volterra filter The properties of the    n x k m  operator are as follows:            m m m n x n x n x n x / 1 / 2 1 1      . (4) (4)  Symmetry. Reversing the signal in time does not change the resulting value; Tomar et al. [12] introduce two generalizations of TKEO. A variable length TKEO (VTEO) is defined as  Robustness. The operator is robust even if the signal passes through zero, i.e.,  0  n x , i.e. there is no division operation;         i n x i n x n x n x i      2 . (5) (5)  Complexity. Complexity of the operator is   k m  . The Summed-over Variable length Teager Energy Operator (S-VTEO) is defined as The Summed-over Variable length Teager Energy Operator (S-VTEO) is defined as Classification using Support Vector Machine The TKEO unlike conventional energy takes into account the frequency component of the signal as well as the signal amplitude.               SV x j i i i j i b x x K y x g , sgn  , (13) (13) where SV are the support vectors,   j i x x K , is the kernel function, i  are weights, and b is the offset parameter.  Proposed Nonlinear Operators If  1   j x g , j x belongs to the Positive class, if  1   j x g , j x belongs to the Negative class, if  0  j x g , j x lies on the decision boundary and can not be classified. In a general case, the TKEO operator can be generalized to the Homogeneous Multivariate Polynomial Operator (HMPO)    n x k m  , where the 2nd order HMPO is defined as Therefore, here we have a binary classification problem in which the outcomes are labelled either as positive (P) or negative (N) class. There are four possible outcomes from a binary classifier. If the outcome from a prediction is P and the actual value is also P, then we have a true positive (TP); however if the actual value is N then we have a false positive (FP). Conversely, a true negative (TN) has occurred when both the prediction outcome and the actual value are N, and false negative (FN) is when the prediction outcome is N while the actual value is P. To evaluate the precision of classification the following metrics are commonly used:                z z i z z j ij m j n x i n x A n x 2 , (10) (10) where   2 / m z  , and A is the coefficient matrix. The 3rd order HMPO is defined as follows                    z z i z z j z z k ijk m k n x j n x i n x A n x 3 . (11) 1) Precision is a measure of how well a binary classification test correctly identifies the true positives against all the positive results; For example, TKEO (2) can be written as follows For example, TKEO (2) can be written as follows 2) Recall is a measure of how well a binary 100 results of experiments are summarized in Table 1 (best results in each category are shown in bold). Proposed Nonlinear Operators classification test correctly identifies the positive cases; classification test correctly identifies the positive cases; y p 3) Accuracy is the proportion of true results (both true positives and true negatives) in the test data; Evaluating the experimental results, we can claim that the 3rd order nonlinear operators such as the proposed HMPO can demonstrate better results for feature identification of the EEG data than traditional 2nd order operators such as TKEO or its generalizations as demonstrated by Table 1. Even visually inspecting the graphs of positive and negative series in Figure 1, one can see that the HMPO operator allows for better identification of significant features (slow cortical potential signals of ~ 1 Hz frequency). p g ) 4) F-measure (F) evaluates the accuracy of classification as the harmonic mean of specificity and recall; 5) Area Under Curve (AUC) is the probability that a classifier will rank a randomly chosen positive instance higher than a randomly chosen negative one. 5) Area Under Curve (AUC) is the probability that a classifier will rank a randomly chosen positive instance higher than a randomly chosen negative one. 5) Area Under Curve (AUC) is the probability that a classifier will rank a randomly chosen positive instance higher than a randomly chosen negative one. Case study For experiments, Data set Ia (Tübingen, ‹self- regulation of SCPs›, subject 1) [16] from the BBCI competition datasets (http://bbci.de/competition/) was used. The datasets were taken from a healthy subject. The subject was asked to move a cursor up and down on a computer screen, while his cortical potentials were taken. During the recording, the subject received visual feedback of his slow cortical potentials (SCPs). The dataset consists of 135 trials belonging to class 0 and 133 trials belonging to class 1. Each trial consists of 896 samples from each of 6 channels. The sampling rate of 256 Hz and the recording length is 3.5s. The dataset was randomly partitioned into 5 parts, and 5-fold cross-validation was used to evaluate the classification results. a) b) Fig. 1. Samples of raw EEG data after application of HMPO operator: a – positive instance; b – negative instance a) The following nonlinear operators were applied to the raw EEG data: TKEO    t x  (3) [5], TKEO-Volterra    t x 2  (4) [11], VTEO    t x 2  (5) [12], VTEO- Volterra    t x 2 2  (7) [13], and the proposed HMPO    t x 3 3  (11), where the non-zero elements of the 3D matrix A are: 1 1 ,1 ,1     a , 3 0,1 ,1    a , 1 1,1 ,1     a , 2 0,0,1   a , 2 1,0,1    a . The matrix values were set based on the results of grid-based search using all possible combinations of integer numbers from the range {-3, 3} a) b) a) Table 1. Experimental results Operator applied Classification metric Acc. Prec. Rec. F AUC None (original EEG data) 0.7800 0.7462 0.8501 0.7891 0.9018 TKEO    t x  (3) 0.4740 0.4805 0.7527 0.5849 0.4670 TKEO-Volterra    t x 2  (4) 0.5931 0.5657 0.7889 0.6583 0.5428 VTEO    t x 2  (5) 0.5635 0.5415 0.8051 0.6466 0.6500 VTEO-Volterra    t x 2 2  (7) 0.4813 0.4207 0.5330 0.4214 0.1381 DEO    t x 1,1  (9) 0.5262 0.5160 0.7202 0.6063 0.5851 HMPO    t x 3 3  (11) 0.8283 0.8042 0.8727 0.8349 0.8450 b) ) Fig. 1. Case study Samples of raw EEG data after application of HMPO operator: a – positive instance; b – negative instance References Birbaumer N., et al. A Brain–Controlled Spelling Device for the Completely Paralyzed // Nature magazine, 2000. – No. 398. – P. 297–298. 17. Joachims T. A Support Vector Method for Multivariate Performance Measures // 22nd Int. Conf. on Machine Learning (ICML), 2005. – P. 377–384. 8. Nelson R., et al. Detection of High Frequency Oscillations with Teager Energy in an Animal Model of Limbic Epilepsy // 28th Int. Conf. of the IEEE Engineering in Medicine and Biology Society (EMBS'06), 2006. – P. 2578–2580. 18. Damaševičius R. Optimization of SVM Parameters for Recognition of Regulatory DNA Sequences // TOP, 2010. – Springer, 2010. – Vol. 18. – No. 2. – P. 339–353 9. Kaleem M. F., et al Application of Empirical Mode Decomposition and Teager energy operator to EEG signals Received 2011 12 13 Accepted after revision 2012 01 14 I. Martisius, R. Damasevicius, V. Jusas, D. Birvinskas. Using Higher Order Nonlinear Operators for SVM Classification of EEG Data // Electronics and Electrical Engineering. – Kaunas: Technologija, 2012. – No. 3(119). – P. 99–102. Brain-Computer Interface (BCI) systems require application of complex analysis, signal processing, denoising, feature extraction, dimensionality reduction and classification methods on acquired raw electroencephalogram (EEG) data to allow for useful operation. In this paper, we consider application of nonlinear operators such as Taeger-Kaiser Energy Operator (TKEO) and its multiple generalizations on the EEG signals and evaluate the efficiency of the operators using a Support Vector Machine (SVM) classifier with linear kernel. We propose a new generalization of TKEO, called Homogeneous Multivariate Polynomial Operator (HMPO), and compare the efficiency of the 3rd order HMPO with other nonlinear operators. Experimental results show that the 3rd order HMPO operator allows for better identification of significant features representing slow cortical potentials in the EEG data. Ill. 1, bibl. 18, tabl. 1 (in English; abstracts in English and Lithuanian). I. Martišius, R. Damaševičius, V. Jusas, D. Birvinskas. Netiesinių aukštesnės eilės operatorių taikymas EEG duomenims klasifikuoti naudojant SVM klasifikatorių // Elektronika ir elektrotechnika. – Kaunas: Technologija, 2012. – Nr. 3(119). – P. 99– 102. Tinkamam rezultatui užtikrinti smegenų ir kompiuterio sąsajoms reikalingi sudėtingi elektroencefalogramos (EEG) signalų apdorojimo, triukšmų šalinimo, požymių išskyrimo ir klasifikavimo algoritmai. Šiame straipsnyje nagrinėjamas netiesinių operatorių, tokių kaip Taeger-Kaiser energijos operatorius (TKEO) ir keletas jo atmainų, naudojimas EEG signalams apdoroti. Šių operatorių efektyvumas vertinamas klasifikavimui taikant atraminių vektorių metodą (SVM) su tiesine funkcija. References for mental task classification // Int. Conf. of the IEEE Engineering in Medicine and Biology Society (EMBC’10), 2010. – P. 4590–4593. 1. Hoffmann U., Vesin J. M., Ebrahimi T. Recent advances in brain–computer interfaces // IEEE 9th Workshop on Multimedia Signal Processing (MMSP’2007), 2007. – P. 17–19. 10. Ahmed B., Redissi A., Tafreshi R. An automatic sleep spindle detector based on wavelets and the Teager energy operator // Annual Int. Conf. of the IEEE Engineering in Medicine and Biology Society (EMBC’2009), 2009. – P. 2596–2599. 2. Guo L., Wu Y., Zhao L., Cao T., Yan W., Shen X. Classification of Mental Task From EEG Signals Using Immune Feature Weighted Support Vector Machines // IEEE Trans. on Magnetics, 2011. – Vol. 47. – No. 5. – P. 866–869. 11. Moore M., Mitra S., Bernstein R. A Generalization of the Teager Algorithm // IEEE Workshop on Nonlinear Signal Porcessing. – Ann Arbor, Michigan, 1997. 3. Selim A. E., Wahed M. A., Kadah Y. M. Machine Learning Methodologies in Brain–Computer Interface Systems // Proc. of Cairo Int. Biomedical Engineering Conf. (CIBEC’2008). – P. 1–5. 12. Tomar V., Patil H. A. On the development of variable length Teager energy operator (VTEO) // 9th Annual Conf. of the Int. Speech Communication Association (ISCA’08), 2008. – P. 1056–1059. 4. Teager H. M. Some observations on oral air flow during phonation // IEEE Trans. Acoust., Speech, Signal Processing, 1980. – Vol. 28. – P. 599–601. 13. Kvedalen E. Signal processing using the Teager energy operator and other nonlinear operators. – MSc thesis. – University of Oslo, 2003. – 20 p. g 5. Kaiser J. F. On a simple algorithm to calculate the 'energy' of a signal // Int. Conf. on Acoustics, Speech, and Signal Processing (ICASSP’90), 1990. – Vol. 1. – P. 381–384. 14. Maragos P., Potamianos A. Higher order differential energy operators // IEEE Signal Processing Lett., 1995. – Vol. 2. – No. 8. – P. 152–154. 6. Nehe N. S., Holambe R. S. Power Spectrum Difference Teager Energy Features for Speech Recognition in Noisy Environment // Third Int. Conf. on Industrial and Information Systems (ICIIS’2008), 2008. – P. 1–5. 15. Vapnik V. Statistical Learning Theory. – Wiley– Interscience, New York, 1998. 7. Huang M.–Y., Tseng D.–C., Liu M. S. C. Wavelet image enhancement based on Teager energy operator // Proc. 16th Int. Conf. on Pattern Recognition, 2002. – Vol. 2. – P. 993– 996. 16. Conclusions and future work In this paper, we proposed a novel nonlinear operator based on the generalization of the Teager-Kaiser Energy Operator, called Homogeneous Multivariate Polynomial Operator (HMPO). The applicability of the proposed operator is demonstrated for classification of the EEG signals. The experimental results obtained using a Support Vector Machine demonstrate an improvement of the classification results. The proposed operator can be used for developing new EEG signal processing algorithms, which can be used in Brain-Computer Interface applications, e.g., for robot control in the noisy environment. Classification of data was performed using the SVMPerf [17] implementation of Support Vector Machine (available at http://svmlight.joachims.org/) with linear kernel. Kernel parameters were optimized using the Nelder- Mead algorithm based method described in [18]. The Future work will focus on the integration of higher- order nonlinear operators with the DSP-based filtering techniques to improve the classification accuracy of the EEG data. 101 References References Siūlomas naujas TKEO algoritmo apibendrinimas, pavadintas multivariantiniu homogeniniu polinomų operatoriumi (HMPO), ir 3-ios eilės HMPO lyginamas su kitais netiesiniais operatoriais. Eksperimento rezultatai rodo, kad 3-ios eilės HMPO operatorius leidžia lengviau atpažinti lėtų žievinių potencialų pasikeitimo požymius. Il. 1, bibl. 18, lent. 1 (anglų kalba; santraukos anglų ir lietuvių k.). 102
https://openalex.org/W4237388920	https://acp.copernicus.org/articles/19/13957/2019/acp-19-13957-2019.pdf	English	null	Diurnal variation of high-level clouds from the synergy of AIRS and IASI space-borne infrared sounders	null	2,019	cc-by	13,948	1 Introduction Due to the importance of clouds for the Earth’s energy bud- get, global satellite observations of cloud properties and their diurnal variations are essential for climate studies, for con- straining climate models, and for evaluating cloud param- eterizations. The diurnal variation of clouds modulates the radiative cooling and heating of the atmosphere and of the surface. Both the clouds embedded in the planetary bound- ary layer and the clouds connected with the surface through deep convection exhibit systematic diurnal variations related to the daily cycle of surface solar heating. The International Satellite Cloud Climatology Project (IS- CCP, Rossow and Schiffer, 1999) uses multi-spectral imager data from a combination of polar orbiting and geostationary weather satellites for a globally complete long-term cloud data record at spatial and temporal scales consistent with cloud dynamical processes (approximately 3 h and 25 km). Although many regional studies were done earlier, the ﬁrst global analyses of diurnal cloud variations were based on the ISCCP products (Cairns, 1995; Rossow and Cairns, 1995). Based on these results, the most notable features of the cloud diurnal cycle are signiﬁcant differences between the phase of diurnal low-level cloud variations over ocean and land and between the phase of diurnal low-level and high-level cloud variations: Diurnal variation of high-level clouds from the synergy of AIRS and IASI space-borne infrared sounders Artem G. Feoﬁlov and Claudia J. Stubenrauch LMD/IPSL, Sorbonne Université, UPMC Univ Paris 06, CNRS, École polytechnique, Palaiseau, 91128, France We are providing a global monthly database of detected diurnal cycle amplitude and phase for each of these three high-level cloud types. Atmos. Chem. Phys., 19, 13957–13972, 2019 https://doi.org/10.5194/acp-19-13957-2019 © Author(s) 2019. This work is distributed under the Creative Commons Attribution 4.0 License. Diurnal variation of high-level clouds from the synergy of AIRS and IASI space-borne infrared sounders Artem G. Feoﬁlov and Claudia J. Stubenrauch LMD/IPSL, Sorbonne Université, UPMC Univ Paris 06, CNRS, École polytechnique, Palaiseau, 91128, France Correspondence: Artem G. Feoﬁlov (artem.feoﬁlov@lmd.polytechnique.fr) Received: 17 February 2019 – Discussion started: 13 March 2019 Revised: 31 August 2019 – Accepted: 4 October 2019 – Published: 20 November 2019 Received: 17 February 2019 – Discussion started: 13 March 2019 Revised: 31 August 2019 – Accepted: 4 October 2019 – Published: 20 November 2019 Abstract. By covering about 30 % of the Earth and by ex- erting a strong greenhouse effect, high-level clouds play an important role in the energy balance of our planet. Their warming and cooling effects within the atmosphere strongly depend on their emissivity. The combination of cloud data from two space-borne infrared sounders, the Atmospheric In- fraRed Sounder, AIRS, and the Infrared Atmospheric Sound- ing Interferometer, IASI, which observe the Earth four times per day, allows us to investigate the diurnal variation of these high-level clouds by distinguishing between high opaque, cirrus, and thin cirrus clouds. We demonstrate that the di- urnal phase and amplitude of high-level clouds can be esti- mated from these measurements with an uncertainty of 1.5 h and 20 %, respectively. By applying the developed method- ology to AIRS and IASI cloud observations for the period of 2008–2015, we obtained monthly geographical distribu- tions of diurnal phase and amplitude at a spatial resolution of 1◦latitude ×1◦longitude. In agreement with other studies, the diurnal cycle of high-level clouds is the largest over land in the tropics. At higher latitudes, their diurnal cycle is the largest during the summer. For selected continental regions we found diurnal amplitudes of cloud amount of about 7 % for high opaque clouds and for thin cirrus, and 9 % for cirrus. Over ocean, these values are 2 to 3 times smaller. The diurnal cycle of tropical thin cirrus seems to be similar over land and over ocean, with a minimum in the morning (09:00 LT) and a maximum during the night (01:00 LT). Tropical high opaque clouds have a maximum in the evening (21:00 LT over land), a few hours after the peak of convective rain. This lag can be explained by the fact that this cloud type includes not only the convective cores, but also part of the thicker anvils. Tropical cirrus show maximum coverage during the night (01:00 LT over land). This lag indicates that they are part of the deep convective cloud systems. However, the peak local times also vary regionally. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds – Low-level clouds over ocean have a maximum coverage early morning, while over land the maximum is in the early afternoon. Sounding Interferometer (IASI; Hilton et al., 2012), now covering 15 and 10 years, respectively, have been presented by Stubenrauch et al. (2017). In this article, we use the syn- ergy of these two instruments, observing each point of the Earth at least four times, to build a database of amplitude and phase of the diurnal cycle of the coverage and emissivity of high-level clouds, which can be further used for regional and global climate studies and for climate model evaluation. – High-level clouds have a maximum coverage in early to late evening. – Mid-level clouds have a maximum coverage late at night or early in the morning. The structure of the article is as follows. In Sect. 2, we brieﬂy describe the AIRS and IASI cloud data as well as the environmental data used for this study. Then we present the newly developed approach to estimating the diurnal cycle of cloud coverage from a combination of AIRS and IASI obser- vations. Section 3 ﬁrst presents a comparison of the diurnal variation of high-level cloud coverage from our method with results from other datasets (Sect. 3.1). Then we introduce the diurnal variation of high opaque cloud, cirrus, and thin cirrus coverage. For speciﬁc land regions, we try to estab- lish the links and temporal lags between the different high- level cloud types, the surface temperature, and relative hu- midity (Sect. 3.3). Section 3.4 presents another application: as the combined dataset covers a period of 8 years we analyse the geographical patterns of amplitude change as a function of the global surface temperature change. Conclusions are drawn in Sect. 4. However, the combination of a maximum of two atmospheric window channels, one IR and one visible (the latter not work- ing during the night), leads to a low sensitivity of ISCCP to thin cirrus at night and when low-level clouds are underneath. g By contrast, the high spectral resolution of the IR verti- cal sounders used in this study allows the selection of the spectral channels with the contribution functions centred at different heights: the radiances measured near the centre of the 15 µm CO2 absorption band are sensitive to the upper at- mospheric layers while the radiances in the absorption band wings are used to probe successively lower levels. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Compared to other passive remote sensing instruments, IR sounders are sensitive to cirrus with emissivity as low as 0.1, day and night (Stubenrauch et al., 2010, 2017; Menzel et al., 2016). TIROS-N Operational Vertical Sounder (TOVS) data (TOVS Path-B, Scott et al., 1999) have been used by Stubenrauch et al. (2006) to identify high opaque clouds, cirrus, and thin cirrus according to their emissivity. By exploiting the time drifting of the afternoon polar orbiting NOAA satellites, in combination with the non-drifting morning orbits, Stuben- rauch et al. (2006) have determined the diurnal cycle of high opaque clouds, cirrus, and thin cirrus to be largest over land in the tropics and mid-latitude summertime. From TIROS-N Operational Vertical Sounder (TOVS) data (TOVS Path-B, Scott et al., 1999) they also concluded the following: Published by Copernicus Publications on behalf of the European Geosciences Union. 13958 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 2.1 Cloud properties from AIRS and IASI Since 2002, the AIRS cross-track scanning instrument aboard the polar orbiting Aqua satellite has been provid- ing very high spectral resolution measurements of atmo- spheric radiation in 2378 spectral bands in the thermal infrared (3.74–15.40 µm), at a spatial resolution of about 13.5 km ×21 km at nadir to 41 km ×21 km at the scan ex- tremes (Chahine et al., 2006). Local observation times are 01:30 and 13:30 LT. – High opaque clouds have a maximum coverage in the evening. – Thin cirrus increase during the afternoon and persist during the night. IASI aboard the polar orbiting Metop-A platform is a Fourier transform spectrometer based on a Michelson inter- ferometer, which covers the IR spectral domain from 3.62 to 15.5 µm. As a cross-track scanner, the swath corresponds to 30 ground ﬁelds per scan, and each of these measures a 2×2 array of footprints. The geometry of IASI observations is similar to that of the AIRS instrument: ±48.3◦ground cov- erage, 12 km resolution at nadir, with observations at 09:30 and 21:30 LT, since 2007. – The varying proportions of thinner and thicker cirrus imply a gradual thickening of the cirrus clouds from late afternoon into the night-time. – Mid-level cloud coverage exhibits a small increase dur- ing night-time. As passive instruments are only able to provide information on the uppermost cloud layer in the case of multi-level cloud ﬁelds, the results on the lower cloud diurnal cycle will be in- evitably modulated by the clouds above. Therefore, we con- centrate on the diurnal variation of high-level clouds. The CIRS cloud property retrieval package (Feoﬁlov and Stubenrauch, 2017; Stubenrauch et al., 2017) is based on a weighted χ2 method using eight channels along the 15 µm CO2 absorption band (Stubenrauch et al., 1999b). It provides cloud pressure (pcld), cloud emissivity (εcld), cloud temper- ature (Tcld), and cloud height (zcld), as well as their uncer- tainties. We deﬁne the cloud types according to pcld and εcld: The CIRS (Clouds from Infrared Sounders) cloud clima- tologies established from the Atmospheric InfraRed Sounder (AIRS; Chahine et al., 2006) and the Infrared Atmospheric www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13959 Figure 1. 2.1 Cloud properties from AIRS and IASI Latitudinal distribution of high opaque, cirrus, and thin cirrus cloud coverage estimated from AIRS (01:30 and 13:30 LT) and IASI (09:30 and 21:30 LT) by the CIRS retrieval: (a) January; (b) July. Climatological averages over 2008 to 2015. high-level clouds are the ones with pcld < 440 hPa, and these are further divided into high opaque (εcld > 0.95), cirrus (0.95 > εcld > 0.5), and thin cirrus (0.5 > εcld > 0.1). Ancil- lary data (surface temperature, atmospheric temperature, and water vapour) are used in the radiative transfer calculations of the retrieval. While the sensitivity of the retrieved cloud properties to ancillary data is small for high-level clouds, the low-level cloud coverage is sensitive to surface temperatures used in the retrieval (Stubenrauch et al., 2017). To avoid po- tential retrieval problems associated with inconsistent ancil- lary data between AIRS and IASI, we have adopted the same ancillary dataset for both, namely, the ERA-Interim meteo- rological reanalysis (Dee et al., 2011) by the European Cen- tre for Medium-Range Weather Forecasts (ECMWF), given every 6 h in universal time. For the cloud retrieval, the ERA- Interim surface temperature and pressure as well as atmo- spheric temperature and water vapour proﬁles have been in- terpolated towards the local observation times of AIRS and IASI. Figure 1. Latitudinal distribution of high opaque, cirrus, and thin cirrus cloud coverage estimated from AIRS (01:30 and 13:30 LT) and IASI (09:30 and 21:30 LT) by the CIRS retrieval: (a) January; (b) July. Climatological averages over 2008 to 2015. From Fig. 1, presenting latitudinal distributions of high opaque, cirrus, and thin cirrus coverage for January and for July, averaged from 2008 to 2015, separately at 01:30, 09:30, 13:30, and 21:30 LT, we deduce that (i) in the tropics (30◦S– 30◦N) high-level clouds are present in more than half of the observations (the sum of the three cloud types reaches 60 %); (ii) the tropical high-level cloud coverage maximum moves seasonally towards the summer hemisphere, while the cover- age of high-level clouds in the mid-latitudes is larger in win- ter, due to storm tracks; and (iii) all cloud types demonstrate a diurnal variation, but its zonal amplitude is small compared to the zonal mean of the corresponding cloud coverage. It is the largest around the peak of the Intertropical Convergence Zone (ITCZ), with about 5 % for cirrus. 2.2 Estimating the diurnal cycle amplitude and phase A(t) = A24 · sin 2π 24 t + ϕ24 + A12 · sin 2π 12 t + ϕ24 + 1ϕ = A24 · sin 2π 24 t + ϕ24 + 0.28 · sin 2π 12 t + ϕ24 + 1ϕ , (1) In this section, we develop an approach to identify both the amplitude and the phase (or the “peak time”) of the diurnal variation of cloud coverage, using a combination of AIRS and IASI cloud data, with four measurements per day. Both amplitude and phase depend on the cloud type, region, and season (Cairns, 1995; Soden, 2000; Tian et al., 2004; Stuben- rauch et al., 2006; Eastman and Warren, 2014, and references therein), so for each location they should be determined in- dividually. where the indices “24” and “12” correspond to diurnal and semi-diurnal harmonics, respectively, t is time in hours, 1φ is the phase shift between semi-diurnal and diurnal harmon- ics, and the numeric parameters are estimated from Fig. 1 of Cairns (1995). It is interesting to note that a similar mix- ture of diurnal and semi-diurnal harmonics describes trop- ical precipitation (Bowman et al., 2005). Since the ratio of A12/A24 ≈0.28 obtained from Cairns (1995) does not change much with the type of the cloud, we simplify the equation to a form shown in the second part of Eq. (1) and use this ratio throughout the analysis assuming that it is A24, which dominates the diurnal variation. By analysing Fig. 1 of Cairns (1995) using least-square ﬁtting of Eq. (1) we found where the indices “24” and “12” correspond to diurnal and semi-diurnal harmonics, respectively, t is time in hours, 1φ is the phase shift between semi-diurnal and diurnal harmon- ics, and the numeric parameters are estimated from Fig. 1 of Cairns (1995). It is interesting to note that a similar mix- ture of diurnal and semi-diurnal harmonics describes trop- ical precipitation (Bowman et al., 2005). Since the ratio of A12/A24 ≈0.28 obtained from Cairns (1995) does not change much with the type of the cloud, we simplify the equation to a form shown in the second part of Eq. (1) and use this ratio throughout the analysis assuming that it is A24, which dominates the diurnal variation. By analysing Fig. 1 of Cairns (1995) using least-square ﬁtting of Eq. 2.1 Cloud properties from AIRS and IASI Figure 1 will serve as a reference when considering the diurnal amplitudes dis- cussed below. are known to include variations on both diurnal and semi- diurnal timescales (e.g. Cairns, 1995), which is clearly be- yond the detection limit. Moreover, the sampling of AIRS and IASI measurements is not equidistant in time with its 8 and 4 h intervals. Correspondingly, one has to use an external source of information to ensure an unambiguous detection of the diurnal cycle and estimate its phase ϕ and amplitude A. We found this missing piece of the puzzle in the function describing the general behaviour of the diurnal cloud cov- erage variation as a mixture of two harmonics, diurnal and semi-diurnal, as demonstrated by the analysis of ISCCP ob- servations (Cairns, 1995). Accordingly the diurnal cycles of high-, middle-, and low-level clouds are well represented by a mixture of two harmonic functions of the following form: 2.2 Estimating the diurnal cycle amplitude and phase (1) we found The Nyquist–Shannon–Kotelnikov sampling theorem says: “if a function x(t) contains no frequencies higher than B hertz, it is completely determined by giving its ordinates at a series of points spaced 1/(2B) seconds apart”. In ap- plication to diurnal variation analysis this means that four measurements per day are just on the edge of the diurnal cy- cle detectability. In addition, the condition of the theorem is not completely fulﬁlled since variations in cloud coverage www.atmos-chem-phys.net/19/13957/2019/ To test the sensitivity of the approach to the uncertainties of the amplitudes related to un- certainties of AIRS and IASI cloud coverage, we also con- sider the case when a 20 % random “noise” is added to the amplitudes at the sampled observation times. This uncertainty was estimated from the most recent CIRS-AIRS and CIRS-IASI cloud products (Stubenrauch et al., 2017), which were compared with active lidar cloud mea- surements of the CALIPSO mission (Winker et al., 2009). The comparison showed a “hit rate” for individual measure- ments in the tropics and mid-latitudes of the order of 88 % over oceans and of 82 % over land (hit rate stands for the ra- tio of number of cases for which AIRS cloud detection agrees with CALIOP to the total number of cases). Figure 2. Illustration of the approach to estimate the diurnal vari- ation phase and amplitude from four measurements, taking advan- tage of a known form of the variation: (a) “true” proﬁle measured at four points and moving guess proﬁles; (b) Pearson’s correlation co- efﬁcient calculated for guess proﬁles for noise-free and noisy simu- lations; (c) determination of the amplitude with the phase known. estimate of the uncertainty of our method of phase (or peak local time) determination to be ±1.5 h. The ﬁrst step in the analysis is to build the function in ac- cordance with Eq. (1) with an arbitrary amplitude A24. Then we numerically solve the system for φ24 as follows. The φ24 is changed in ﬁne increments, each time calculating A(t) at the four local observation times of AIRS and IASI (blue cir- cles on grey curves in Fig. 2a). A set of obtained values is compared to a reference “measurement” (red circles), and a Pearson’s correlation coefﬁcient kcorr is calculated for each phase shift (Fig. 2b, phase converted to peak time for the sake of visualization). In the noise-free self-consistency study the maximum of kcorr should exactly match the phase reproduc- ing the original function. The tests show that even 20 % ran- dom noise added to the “reference” points does not spoil the phase determination by more than half an hour. Since the peak of the kcorr curve is not sharp, we make a conservative With the phase known, we estimate the A24 amplitude (Fig. www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13960 Figure 2. Illustration of the approach to estimate the diurnal vari- ation phase and amplitude from four measurements, taking advan- tage of a known form of the variation: (a) “true” proﬁle measured at four points and moving guess proﬁles; (b) Pearson’s correlation co- efﬁcient calculated for guess proﬁles for noise-free and noisy simu- lations; (c) determination of the amplitude with the phase known. that the phase shift 1φ is equal to −2 h for the low- and mid-level clouds and to 0 h for high-level clouds. With the A12/A24 set to 0.28 and 1φ set to 0, the diurnal “shape” of Eq. (1) is ﬁxed (see the grey line in Fig. 2a) and the prob- lem is reduced to one of determining the amplitude A24 and φ24. Two satellite instruments provide us with measurements four times a day, and we determine the best-ﬁt amplitude and phase using a minimization technique based on the “sliding proﬁle” approach as depicted in Fig. 2 and described be- low. This approach is similar to the one used in Goldberg et al. (2013) where it was applied to determine the phase and period of interhemispheric coupling. Later in the text we will show that the shape given by Eq. (1) represents the diurnal cycle in clouds better than a simple harmonic ﬁt, but prior to validation of the shape one has to introduce a general diur- nal cycle estimation approach itself. The examples shown in Fig. 2 utilize the A(t) given by Eq. (1), though the approach will remain valid for any periodic function. Figure 2 explains the approach for estimating the diurnal variation phase and amplitude: let us imagine that a real di- urnal variation for a given type of cloud at a given location is deﬁned by Eq. (1), with A24 and φ24 known (black curve in Fig. 2a). For this case, the red circles in Fig. 2a corre- spond to the values obtained at the local observation times of AIRS and IASI, which are passed to the diurnal phase and amplitude estimation algorithm. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Searching a better-ﬁtting shape of the diurnal varia- tion is out of the scope of this study, but the approach to esti- mate phase and amplitude of the diurnal variation, under the assumption of a known and ﬁxed shape of the diurnal vari- ation, remains valid for any periodic function (e.g. see the surface temperature variation ﬁtting in Appendix B). Differences in the amplitudes between AIRS/IASI and CATS may be mostly explained by the small sampling of CATS, as only a few regions per latitude band are sampled each time. With increasing sampling towards higher latitudes (to be checked in the CATS paper), the agreement increases: over mid-latitude land, both curves agree very well. ISCCP also agrees well with AIRS/IASI in general, in particular in the tropics and the subtropics, except during the night where the high-level cloud coverage is slightly underestimated by ISCCP due to misidentiﬁcation of thin cirrus as mid-level cloud (Stubenrauch et al., 1999a). Over summer mid-latitude land, the lower sensitivity of ISCCP to thin cirrus also leads to a slight diurnal amplitude underestimation in addition to a slight phase shift (see also Fig. 4). Summarizing this section, the “sliding proﬁle approach” allows for the estimation of the phase and amplitude of the diurnal variation from four measurements per day performed at arbitrary times with respect to peak time. The uncertainty of the estimated peak time for the combination of four AIRS and IASI monthly averages over 1◦latitude ×1◦longitude is ±1.5 h while the diurnal cycle amplitude is estimated with ∼20 % uncertainty. g p g Concerning the phase, the agreement between AIRS/IASI and CATS and between AIRS/IASI and ISCCP is indicated in Fig. 3 by Pearson’s linear correlation coefﬁcients, together with the local peak times. The values are in general high, when the diurnal amplitudes are large (above noise), which is the case in the summer hemisphere and tropics over land. The peak local times (marked in red and blue) of AIRS/IASI and CATS do agree to within 2 h that is comparable to the es- timated uncertainty range of 1.5 h. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds servations are more sensitive to thin cirrus than IR sounders, ISCCP is less sensitive to thin cirrus (e.g. Stubenrauch et al., 2013). The much larger statistics of several years of global AIRS/IASI and ISCCP data compared to only nadir track statistics of 3 years of CATS leads to a smoother behaviour of the diurnal variations presented in Fig. 3 for boreal sum- mer. The shape assumption in the diurnal cycle estimation used for AIRS/IASI also contributes to a smoother behaviour and is compensating well for the better temporal resolution of ISCCP. All three datasets show larger diurnal cycle am- plitudes over land than over ocean, in agreement with many other studies (e.g. Soden, 2000; Tian et al., 2004; Stuben- rauch et al., 2006; Zhang et al., 2008). They are the largest in the tropics, and towards higher latitudes they decrease in winter (see also Fig. S1 for Southern Hemisphere in the Sup- plement), as shown for AIRS/IASI. methods to be equivalent, but the one involving more points should be less noisy and, therefore, more reliable. The noise in the measured points affects the uncertainty on A24, with 20 % noise in the source data leading to about 20 % uncer- tainty in the estimated amplitude. y p To justify the choice of the relationship in Eq. (1) for the ﬁtting, we have performed the following numerical experi- ment using real data: 1 year of AIRS and IASI cloud data using the methodology outlined above for two different hy- potheses on the ﬁtting functions: a simple harmonic one with a 24 h period and a mixture of diurnal and semi-diurnal de- scribed by Eq. (1). For each tested hypothesis, we have built a histogram of the best correlation coefﬁcient values, sepa- rately for high- and low-level cloud coverage diurnal varia- tion. We found that using Eq. (1) for the ﬁtting of real-life observations one achieves ∼8 % and ∼18 % higher correla- tion coefﬁcients for the diurnal cycle of high- and low-level clouds, respectively, than with a simple harmonic function ﬁtting. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Concerning ISCCP, the lo- cal peak time in the tropics and subtropics has been systemat- ically determined earlier, because of the slight underestima- tion of high-level cloud coverage during the night, while over summer mid-latitude land the local peak time is estimated there hours later as by AIRS/IASI and CATS, because of missing thin cirrus during the daytime. The latter are respon- sible for the local peak time at 17:00 LT, as seen in Fig. 4, which presents the contributions of the different cloud types (high opaque, cirrus, and thin cirrus) to the total diurnal vari- ation of high clouds for the same latitude bands as in Fig. 3, during boreal summer in the Northern Hemisphere (Fig. S2 for austral summer in the Southern Hemisphere). Again, am- plitudes are larger over land than over ocean, and in general the amplitudes of the individual cloud types are larger than of all high-level clouds mixed together, as the phases of these cloud types differ. 3.1 Zonal averages We apply the diurnal cycle estimation algorithm to the cov- erage of all high-level clouds and separately on high opaque, cirrus, and thin cirrus cloud coverage from the CIRS-AIRS and CIRS-IASI cloud climatologies. For the following anal- yses we determined A24 and φ24 for each month and each 1◦latitude ×1◦longitude grid box, calculated A(t) in accor- dance with Eq. (1), and averaged the resulting shapes for a given latitude band or region. To demonstrate the feasibility of our methodology, we compare diurnal variations of high-level cloud coverage av- eraged over three latitudinal bands, separately for ocean and land, to those presented by Noel et al. (2018), obtained from new lidar measurements of the Cloud-Aerosol Transport Sys- tem (CATS; Palm et al., 2018; Yorks et al., 2016) aboard the International Space Station (ISS), and to those from ISCCP presented by Rossow and Schiffer (1999). While lidar ob- Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ www.atmos-chem-phys.net/19/13957/2019/ 2c) as follows: we draw a virtual “zero line” at the level corresponding to a mean of all four points and calcu- late the magnitudes at these points with respect to this “zero line”. Then we compare the obtained values with those esti- mated from the Eq. (1) in the same way (dashed black line in Fig. 2c represents a zero line of Eq. 1). The mean ratio of amplitudes gives A24. Since using the values close to the “zero line” might lead to zero-over-zero-type errors and to an increase in the A24 uncertainty, we pick up only the A(t) points with the amplitude greater than 0.2 of the maximal \|A(t)\| value (the threshold is marked by green dashed lines in Fig. 2c). Another way of estimating A24 is to compare the maximal spans of the reference proﬁle sampled at four ob- servation times and that of the measured one. We ﬁnd these Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ www.atmos-chem-phys.net/19/13957/2019/ 13961 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds The similar diurnal cycle of thin cirrus over ocean and over land is another indication that some of these thin cirrus are formed in situ, having no direct relation to the convective systems. The difference in phase of convec- tion between ocean and land and the much broader peak of convection over ocean has been associated with differences in the vertical structure of land and ocean convection by So- den (2000). In contrast to land, over open ocean the thermal properties of the ocean surface undergo a relatively weak di- urnal cycle. Chen and Houze Jr. (1997) have shown that the life cycle of convective cloud systems also plays a role in affecting the diurnal cycle of cloudiness. The formation of longer lived oceanic convective cloud systems may introduce a bi-diurnal cycle, with large systems occurring at the same location only every other day. 3 2 Geographical distributions y We apply the methodology described in Sect. 2.2 to cloud type coverage statistics gathered over grid cells of 1◦lati- tude ×1◦longitude for the period 2008–2015, separately for January and July. Average diurnal amplitude and phase are represented by vectors as suggested by Cairns (1995) and also utilized in Soden (2000) and Tian et al. (2004), where the vector’s length corresponds to the amplitude of the diur- nal variation and the phase is converted to a local peak time, given as the direction of the vector. Since different cloud types are characterized by different diurnal amplitudes, a unit vector is added to the lower right corner of each panel. We consider the diurnal cycle to be reliably detected at a given latitude and longitude, if the Pearson’s correlation coefﬁcient for the corresponding diurnal curves (see Sect. 2.2) is greater than 0.85, an empirical threshold based on examining numer- ous diurnal variation curves. Whenever we average the diurnal cycle parameters, we calculate the mean phase (or peak time) using the corre- sponding amplitudes as weights. To avoid errors caused by averaging the phases in the vicinity of 24:00–00:00 LT transi- tion (for example, direct averaging of 23:00 and 01:00 LT re- turns noon instead of midnight), we apply a “resulting force” algorithm (Appendix A). g pp Figures 5 and 6 present detected diurnal variations for the three different high-level cloud type coverage, for January and July, respectively. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13962 explore the links to their atmospheric environment. The latter includes surface temperature (Tsurf) and upper tropospheric (UT) relative humidity (RH), the parameters, which are both linked to their formation and then affected by the clouds. The distinction between the cloud types also allows a deeper interpretation of the diurnal cycle. The diurnal cycle over tropical land, which is already largest for all high-level clouds together, as illustrated in Fig. 3, can be mainly un- derstood as the diurnal cycle of deep convective cloud sys- tems, with a peak of precipitation around 18:00 LT (Tian et al., 2004; Zhang et al., 2008), followed by high opaque cloud coverage, including convective cores as well as some thick anvil cirrus, with a peak around 21:00 LT and developing cir- rus anvil amount with a maximum around 01:00 LT. The di- urnal cycle of thin cirrus is smaller than the one of cirrus, be- cause part of the thin cirrus corresponds to dissipating con- vective cloud systems and part to cirrus formed in situ by large-scale forcing (e.g. Luo and Rossow, 2004; Riihimaki et al., 2012). While the phase of cirrus and thin cirrus is lagged in the tropics, with a minimum of thin cirrus in the morn- ing and a minimum of cirrus around 13:00 LT, their phase gets more similar towards higher latitudes. This can probably be explained by the fact that in situ freezing TTL cirrus do not exist at higher latitudes. In the mid-latitudes cirrus and thin cirrus have a maximum in the afternoon and are most probably linked to synoptic situations of fronts. Some cirrus may be orographic, generated by ascent of air within large- amplitude vertically propagating waves over mountains and even over hills (e.g. Queney, 1948; Ludlam, 1952). When comparing tropical land with tropical ocean, we observe a difference of about 11 h for the high opaque clouds, with a broader maximum coverage occurring in midmorning, again a few hours later than the convection and precipitation of early morning (e.g. Tian et al., 2004; Zipser et al., 2006; Zhang et al., 2008). Cirrus and thin cirrus follow with max- ima in the evening and during the night. Their minima are again shifted, with the minimum of cirrus just before the maximum of high opaque clouds, as over land. The mini- mum of thin cirrus occurs in midmorning, similar to the min- imum over land. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds A common feature of all of these maps is that the amplitudes of the diurnal variations maximize over the tropical belt and in the summer hemisphere mid-latitudes. This is an expected behaviour consistent with other obser- vations (e.g. Rossow and Schiffer, 1999; Wylie and Woolf, 2002; Tian et al., 2004; Hong et al., 2006; Stubenrauch et al., 2006). Figures 5 and 6 of Hong et al. (2006) present peak times of precipitation, of very cold cloud coverage (IR brightness temperature T IR B < 210 K), and of cold-cloud cov- erage (T IR B < 235 K) for different tropical regions, with lags between the three, corresponding to the development of deep convective systems. Compared to these results, our results are consistent when associating high opaque clouds with very cold clouds and cirrus with the warmer clouds. Another ex- pected feature is the magnitude of diurnal change over land being generally larger than that over ocean, as the ocean sur- face temperature has a much smaller diurnal cycle (Fig. B1 in the Appendix B). In general, the high opaque cloud coverage with about 5 % is much smaller than the cirrus and thin cirrus cov- erage (Fig. 1). The clouds identiﬁed by the CIRS retrieval as high opaque ones, for which a diurnal variation is de- tected, have an average diurnal amplitude of about 5 %, but certain regions (Figs. 5a, 6a) demonstrate amplitudes reach- ing 10 %. High opaque clouds, often associated with deep convective cores in the tropics, have a large regional vari- www.atmos-chem-phys.net/19/13957/2019/ A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 3.2 Geographical distributions We ﬁrst present geographical maps of phase and amplitude of the diurnal cycle of the high-level cloud types and then Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ 13963 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Figure 3. Comparison of average diurnal cycles of high-level cloud coverage over three latitudinal bands estimated from CATS lidar ob- servations (Noel et al., 2018), from AIRS/IASI (this work), and from ISCCP (Rossow and Schiffer, 1999). CATS statistics only includes June, July, and August (JJA, red curves), whereas AIRS/IASI results are shown both for boreal summer (JJA, blue curves) and boreal winter (December, January, and February, DJF, green curves) in 2008–2015. The ISCCP JJA data cover the period of 1989–1991. The correlation coefﬁcients are given for AIRS/IASI vs. CATS and AIRS/IASI vs. ISCCP sequences for JJA in the upper left part of each panel; the peak local times for CATS are marked in red, for AIRS/IASI in blue, and for ISCCP in black in the upper right part of each panel. (a) 30–60◦N, ocean; (b) 30–60◦N, land; (c) 15–30◦N, ocean; (d) 15–30◦N, land; (e) 15◦S–15◦N, ocean; (f) 15◦S–15◦N, land. Figure 3. Comparison of average diurnal cycles of high-level cloud coverage over three latitudinal bands estimated from CATS lidar ob- servations (Noel et al., 2018), from AIRS/IASI (this work), and from ISCCP (Rossow and Schiffer, 1999). CATS statistics only includes June, July, and August (JJA, red curves), whereas AIRS/IASI results are shown both for boreal summer (JJA, blue curves) and boreal winter (December, January, and February, DJF, green curves) in 2008–2015. The ISCCP JJA data cover the period of 1989–1991. The correlation coefﬁcients are given for AIRS/IASI vs. CATS and AIRS/IASI vs. ISCCP sequences for JJA in the upper left part of each panel; the peak local times for CATS are marked in red, for AIRS/IASI in blue, and for ISCCP in black in the upper right part of each panel. (a) 30–60◦N, ocean; (b) 30–60◦N, land; (c) 15–30◦N, ocean; (d) 15–30◦N, land; (e) 15◦S–15◦N, ocean; (f) 15◦S–15◦N, land. Figure 4. Diurnal cycle of high opaque, cirrus, and thin cirrus coverage in NH mid-latitudes, NH subtropics and tropics in boreal summer: (a) 30–60◦N, ocean; (b) 30–60◦N, land; (c) 15–30◦N, ocean; (d) 15–30◦N, land; (e) 15◦S–15◦N, ocean; (f) 15◦S–15◦N, land. The AIRS/IASI statistics is averaged for 2008–2015. Figure 4. 3.2 Geographical distributions Diurnal cycle of high opaque, cirrus, and thin cirrus coverage in NH mid-latitudes, NH subtropics and tropics in boreal summer: (a) 30–60◦N, ocean; (b) 30–60◦N, land; (c) 15–30◦N, ocean; (d) 15–30◦N, land; (e) 15◦S–15◦N, ocean; (f) 15◦S–15◦N, land. The AIRS/IASI statistics is averaged for 2008–2015. Atmos. Chem. Phys., 19, 13957–13972, 2019 Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ 7, seems to be maximum in the convective regions just a few hours after the maxi- mum of cirrus and thin cirrus, as observed by Horvath and Soden (2008). Figure 5. Diurnal variation of (a) high opaque clouds, (b) cirrus clouds, and (c) thin cirrus clouds estimated from AIRS/IASI for January. Statistics is averaged over the period of 2008–2015. Vec- tor length is proportional to the amplitude and its orientation de- ﬁnes the local time of the peak: downwards for 00:00 LT, left for 06:00 LT, upwards for 12:00 LT, and right for 18:00 LT. For the sake of readability, the arrows are also coloured in tints of red for day (06:00–18:00 LT) and in shades of blue for night (18:00–06:00 LT). While relatively small cloud coverage (like the one of high opaque clouds) or relatively similar cloud coverage during the day (like over ocean) provide a “noisier” input for the diurnal cycle estimation, large “blank” areas in the winter hemisphere with no detected diurnal variation assure us that the algorithm is stable against false triggering provoked by noise, and the diurnal cycle of cirrus clouds detected over the tropical ocean in July (Fig. 6b) is very close to that re- ported in Soden (2000) for their averaged June–August 1987 (10 %, 18:00 LT). This allows us to say that larger ocean zones, which demonstrate a consistent diurnal cycle, deserve attention. However, the global maps have been established with one single detection threshold using a Pearson’s corre- lation coefﬁcient of 0.85, so if one wants to focus on speciﬁc ocean regions one might want to reﬁne the threshold on the correlation coefﬁcient. 7 are in the southern subtropics–tropics. The regional ampli- tudes and peak times of cloud type coverage, Tsurf, and UT relative humidity are summarized in Tables 1 and 2 and il- lustrated in Fig. 9, which presents circular histograms of the peak amplitude local times, separately for January and July. 7 are in the southern subtropics–tropics. The regional ampli- tudes and peak times of cloud type coverage, Tsurf, and UT relative humidity are summarized in Tables 1 and 2 and il- lustrated in Fig. 9, which presents circular histograms of the peak amplitude local times, separately for January and July. www.atmos-chem-phys.net/19/13957/2019/ d C. J. Stubenrauch: Diurnal variation of high-level clouds Figure 5. Diurnal variation of (a) high opaque clouds, (b) cirrus clouds, and (c) thin cirrus clouds estimated from AIRS/IASI for January. Statistics is averaged over the period of 2008–2015. Vec- tor length is proportional to the amplitude and its orientation de- ﬁnes the local time of the peak: downwards for 00:00 LT, left for 06:00 LT, upwards for 12:00 LT, and right for 18:00 LT. For the sake of readability, the arrows are also coloured in tints of red for day (06:00–18:00 LT) and in shades of blue for night (18:00–06:00 LT). A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13964 ability, which may be inﬂuenced by orography. Maximum high opaque cloud coverage moves with the ITCZ towards the summer hemisphere, with peak times between 18:00 LT and midnight. In July, there is a large contrast between con- tinental opaque clouds and nearby oceanic regions. Over the continents, the peak typically occurs in the evening around 20:00 LT as compared with oceanic areas near the conti- nents with peaks closer to noon. Nonetheless, in some other ocean locations opaque clouds also peak in the evening or overnight (e.g. tropical longitudes −115 to −135◦). The di- urnal variation of cirrus clouds (Figs. 5b, 6b) is larger than that of high opaque clouds: the average amplitude is ∼8 %, with some regions reaching the values of 12 % and individ- ual grid cells showing amplitudes of up to 20 %. There is less contrast between land and ocean for the cirrus and thin cirrus peak times, with continental peak times between mid- night and 02:00 LT and oceanic ones around 18:00 LT. Over land, a lag of about 3 h can be identiﬁed for two-thirds of the cases (see also Sect. 3.3). It seems to be more complicated to identify the lags over ocean, with a maximum of cirrus in the evening. Thin cirrus have a maximum at midnight over ocean, lagging behind cirrus, which can be interpreted as the thinning of convective systems during the night. Over land the situation is a bit more complicated, with a maximum af- ter noon in large mountain areas (Rocky Mountains, South America, Africa, and Asia) and some lagging behind cirrus, again to be interpreted as thinning of convective systems. UT relative humidity, presented in Fig. www.atmos-chem-phys.net/19/13957/2019/ From these tables we deduce that the two regions in the northern mid-latitudes (1 and 2) show a large difference in the Tsurf amplitude, with a larger one in summer. The regions which are affected by the ITCZ (4, 5, and 6) have a slightly larger Tsurf amplitude in summer, while the two regions in the subtropics (3 and 7) show a large Tsurf amplitude both in January and in July. In general, a strong convective activity is revealed by a very large diurnal cycle in the occurrence of high opaque clouds (more than 10 %). This is the case for regions 6 and 7 in January and for regions 3 and 4 in July. The large diurnal cycle in Tsurf in summer (about 10 K) of the two regions in the northern mid-latitudes (1 and 2) leads to some convective activity (diurnal cycle of high opaque clouds of about 7 % and 5 %, respectively) in the afternoon. However, multiple peak amplitude local times may indicate the effects of orography. The signals for cirrus and espe- cially thin cirrus are more evident, with similar peak ampli- tudes and more concise peak amplitude local times in the early afternoon. The thin cirrus may be orographic cirrus. 3.3 Speciﬁc regions Since different geographical regions are characterized by dif- ferent cloud regimes (e.g. Rossow et al., 2002), we deﬁne seven regions over land, presented in Fig. 8, which we anal- yse in more detail. Regions 1, 2, 3, and 4 are in the northern mid-latitudes and subtropics–tropics, while regions 5, 6, and www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of Figure 6. Same as Fig. 5, but for averaged July (2008 through 2015). 13965 Figure 6. Same as Fig. 5, but for averaged July (2008 through 2015). Figure 8. Regions selected for a more detailed analysis of the cloud diurnal cycle over land: 1 – North America, 2 – north-east Asia, 3 – northern Africa, 4 – Indian subcontinent, 5 – South America, 6 – southern Africa, 7 – Australia. Figure 8. Regions selected for a more detailed analysis of the cloud diurnal cycle over land: 1 – North America, 2 – north-east Asia, 3 – northern Africa, 4 – Indian subcontinent, 5 – South America, 6 – southern Africa, 7 – Australia. ple peaks while thin cirrus has a peak around 13:00 LT and a smaller one in the morning (07:00 LT), while over north-east Asia there is one peak detected in cirrus around 14:00 LT. It is difﬁcult to relate the UT humidity clearly to the cirrus during winter. The two other regions in the Northern Hemisphere are sub- tropical (3) and tropical (4). While the diurnal Tsurf cycle is large both in January and in July (11 K) over north-western Africa, it is large (10 K) during the summer monsoon over the Indian subcontinent, while it is smaller (4 K) in January, when the ITCZ moves southwards. The diurnal behaviour of UT clouds is quite different between these regions, whereas the peak amplitude times in UT relative humidity are similar (17:00 LT in winter and 05:00 LT in summer). In the sub- tropics, in summer there seems to be some convective ac- tivity, with a peak time in the evening (21:00–24:00 LT) fol- lowed by the development of cirrus and thin cirrus during the night (02:00 LT), whereas in winter no diurnal cycle of high opaque clouds could be detected, and cirrus and especially thin cirrus have multiple peak times. 3.3 Speciﬁc regions The summer monsoon over the Indian subcontinent leads to a peak amplitude time of high opaque clouds in the early evening, indicating con- vective activity, followed by the development of cirrus anvils in the night and thin cirrus until noon. In January, the ampli- tude of the diurnal cycle is much smaller, with a peak of high opaque clouds later during the night. Thin cirrus develop in the morning and thicken towards cirrus in the afternoon. Figure 6. Same as Fig. 5, but for averaged July (2008 through 2015). Figure 7. Diurnal variation of average relative humidity in a layer 150 hPa below the tropopause, estimated from ERA-Interim reanal- yses, (a) for January and (b) for July. The vector representation is consistent with the caption of Fig. 5. The tropopause level was de- termined in accordance with Reichler et al. (2003). Figure 7. Diurnal variation of average relative humidity in a layer 150 hPa below the tropopause, estimated from ERA-Interim reanal- yses, (a) for January and (b) for July. The vector representation is consistent with the caption of Fig. 5. The tropopause level was de- termined in accordance with Reichler et al. (2003). Comparing the tropical regions in the Southern Hemi- sphere, South America has a slightly smaller diurnal ampli- tude of Tsurf than the African region (5 K compared to 7 K), as the latter also includes subtropical land and the former is part of the “Amazonian green ocean”. In summer (January), maximum convective activity in South America seems to be from 15:00 LT onwards, with cirrus and thin cirrus during the night. Thin cirrus also has another peak amplitude in paral- lel with high opaque cloud between noon and 18:00 LT. Over Africa the diurnal behaviour is similar. In winter (July), the peak amplitude times of high opaque clouds also occur later, with cirrus during the night and again thin cirrus at multiple times. UT relative humidity has two peaks, one in the early morn- ing (05:00 LT) and one in the afternoon (17:00 LT), in both cases the afternoon peak lagging shortly behind cirrus and thin cirrus and the early morning peak lagging behind high opaque clouds during the night. During winter, the diurnal Tsurf amplitude is much smaller (about 5 K), and no diurnal cycle could be detected in high opaque clouds. 3.3 Speciﬁc regions The diurnal cycle in UT relative humidity has opposite peaks, again at 05:00 and 17:00 LT. Over North America, cirrus has multi- A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13966 Table 1. Diurnal cycle amplitude and local time of the peak for Tsurf, high opaque clouds, thick cirrus, thin cirrus, and relative humidity for average January. Long dash symbol (–) means that no reliable cycle was detected while a star symbol () corresponds to multiple peaks in the circular histogram (see Fig. 9). Table 1. Diurnal cycle amplitude and local time of the peak for Tsurf, high opaque clouds, thick cirrus, thin cirrus, and relative humidity for average January. Long dash symbol (–) means that no reliable cycle was detected while a star symbol () corresponds to multiple peaks in the circular histogram (see Fig. 9). Variable Tsurf High opaque Cirrus UT relative humidity Thin cirrus Geographic area Ampl. (K) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) North America 4.6 12.3 – – 5.8 6–12 3.8 5; 17 5.2 7–13 North-east Asia 6.3 14.0 – – 7.4 15.1 3.3 5; 17 – – North-west Africa 11.0 12.3 – – 5.9 – 2.2 16.4 5.2 * Indian subcontinent 9.7 12.0 5.7 22–2 5.3 15.2 1.8 17.9 4.8 9–15 South America 4.9 12.5 7.5 16.1 12.8 23.3 5.2 5.0 7.3 13–19 Africa 6.9 12.7 7.9 17–22 12.2 1.3 4.4 4.8 7.8 * Australia 11.3 12.1 7.5 20.3 8.2 19; 1 2.0 5.5 6.2 15.5 Table 2. Same as Table 1, but for average July. In general, UT relative humidity has a peak in the early morning for regions of convection, which may be explained by UT humidiﬁcation from the dissipation of the anvils (e.g. Horvath and Soden, 2008). In winter, there seems to be a peak in UT relative humidity in the afternoon, often also just after the appearance of thin cirrus, which this time seem to be formed in situ and are not offspring of a convective system. p g y For the regions where diurnal variation was reliably de- tected both for high opaque clouds and for cirrus, the diur- nal cycle amplitudes of these two cloud types are correlated with kcorr = 0.75. This can be probably attributed to the cloud types being part of the same cloud system. Even though peak times for cirrus clouds and for high opaque clouds vary al- most in the same range, an average lag of ∼3 h can be iden- tiﬁed for two-thirds of the cases. www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 13966 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Table 1. Diurnal cycle amplitude and local time of the peak for Tsurf, high opaque clouds, thick cirrus, thin cirrus, and relative humidity for average January. Long dash symbol (–) means that no reliable cycle was detected while a star symbol () corresponds to multiple peaks in the circular histogram (see Fig. 9). Variable Tsurf High opaque Cirrus UT relative humidity Thin cirrus Geographic area Ampl. (K) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) North America 4.6 12.3 – – 5.8 6–12 3.8 5; 17 5.2 7–13 North-east Asia 6.3 14.0 – – 7.4 15.1 3.3 5; 17 – – North-west Africa 11.0 12.3 – – 5.9 – 2.2 16.4 5.2 Indian subcontinent 9.7 12.0 5.7 22–2 5.3 15.2 1.8 17.9 4.8 9–15 South America 4.9 12.5 7.5 16.1 12.8 23.3 5.2 5.0 7.3 13–19 Africa 6.9 12.7 7.9 17–22 12.2 1.3 4.4 4.8 7.8 * Australia 11.3 12.1 7.5 20.3 8.2 19; 1 2.0 5.5 6.2 15.5 Table 2. Same as Table 1, but for average July. Variable Tsurf High opaque Cirrus UT relative humidity Thin cirrus Geographic area Ampl. (K) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) Ampl. (%) Peak (h) North America 10.5 11.8 7.2 14–2 7.7 16–24 2.9 5; 17 7.5 14.8 North-east Asia 8.3 12.0 5.2 * 6.8 15.7 2.8 5; 17 6.4 15.9 North-west Africa 11.0 12.5 6.9 22.7 8.1 1.4 2.7 4.9 7.3 2.3 Indian subcontinent 4.3 11.9 8.3 19.5 11.7 22.7 4 5.3 8.0 * South America 5.6 13.2 8.1 1.0 8.8 21.2 2.9 5.1 6.3 18.9 Africa 7.8 13.4 7.7 22.1 11.2 1.2 3.5 5; 17 7.0 * Australia 8.5 12.6 – – – – 1.5 5; 17 – – Figure 9. Circular histograms of local time of peak amplitude for Tsurf, high opaque clouds, thick cirrus, relative humidity (RH), and thin cirrus for January and July averages. The convention for vec- tor direction representation is consistent with that described in the caption of Fig. 6 while the amplitude is proportional to normalized histogram count number. “N/D” marks the cases for which the diur- nal cycle could not be reliably detected. www.atmos-chem-phys.net/19/13957/2019/ For Australia, a diurnal cycle was only detectable in sum- mer (January), with two peaks in high opaque clouds and in cirrus (19:00 and 01:00 LT) and a peak in thin cirrus in the early afternoon. In general, UT relative humidity has a peak in the early morning for regions of convection, which may be explained by UT humidiﬁcation from the dissipation of the anvils (e.g. Horvath and Soden, 2008). In winter, there seems to be a peak in UT relative humidity in the afternoon, often also just after the appearance of thin cirrus, which this time seem to be formed in situ and are not offspring of a convective system. For the regions where diurnal variation was reliably de- tected both for high opaque clouds and for cirrus, the diur- nal cycle amplitudes of these two cloud types are correlated with kcorr = 0.75. This can be probably attributed to the cloud types being part of the same cloud system. Even though peak times for cirrus clouds and for high opaque clouds vary al- most in the same range, an average lag of ∼3 h can be iden- tiﬁed for two-thirds of the cases. We have to note that in cer- tain cases the deﬁnition of the lag becomes ambiguous due to a possible 24 h phase shift, which is not detectable in our ap- proach. For example, the 15 h peak in the January histogram for thick clouds over South Asia (Fig. 9) should be caused both by in situ formation of high clouds 4 h after the peak of local insolation and by an outﬂow of the high opaque col- umn with a characteristic time of ∼16 h (the lag between high opaque and cirrus cloud peaks). Another presumable element of the cloud system life cycle is the dissipation of the anvil. This can be manifested both in the RH change (Fig. 7) and in thin cirrus variation (Figs. 5c, 6c). The peak time of RH “release” from the cloud oscillates between late afternoon and early morning maxima (Kottayil www.atmos-chem-phys.net/19/13957/2019/ A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13967 Figure 10. Geographical maps of linear regression slopes between de-seasonalized monthly mean diurnal amplitude anomalies from combined AIRS-CIRS and IASI-CIRS cloud data and global mean Tsurf anomalies from ERA-Interim for (a) high opaque cloud cov- erage, (b) cirrus, and (c) thin cirrus. et al., 2016) and its lag with respect to cirrus can be estimated as ∼3.5 h. Finally, thin cirrus show good coupling with high opaque clouds and the cirrus, and the amplitudes of their di- urnal cycle are comparable with those of the cirrus. The peak times for these clouds have broad distributions that make it difﬁcult to estimate the lags. We assign this to long character- istic times of anvil dissipation and to the mixing the effects of in situ and convective outﬂow cloud formation mechanisms: air parcels saturated with water vapour released in the pro- cess of anvil dissipation may travel to other areas and form clouds there. For the cases when the peak time distribution is narrow, the average lag with respect to anvil can be estimated as ∼10 h, but the individual values vary from almost zero to 20 h. et al., 2016) and its lag with respect to cirrus can be estimated as ∼3.5 h. Finally, thin cirrus show good coupling with high opaque clouds and the cirrus, and the amplitudes of their di- urnal cycle are comparable with those of the cirrus. The peak times for these clouds have broad distributions that make it difﬁcult to estimate the lags. We assign this to long character- istic times of anvil dissipation and to the mixing the effects of in situ and convective outﬂow cloud formation mechanisms: air parcels saturated with water vapour released in the pro- cess of anvil dissipation may travel to other areas and form clouds there. For the cases when the peak time distribution is narrow, the average lag with respect to anvil can be estimated as ∼10 h, but the individual values vary from almost zero to 20 h. 3.4 Relating diurnal amplitudes to climate ﬂuctuations Figure 10. Geographical maps of linear regression slopes between de-seasonalized monthly mean diurnal amplitude anomalies from combined AIRS-CIRS and IASI-CIRS cloud data and global mean Tsurf anomalies from ERA-Interim for (a) high opaque cloud cov- erage, (b) cirrus, and (c) thin cirrus. As the period of 8 years of combined AIRS and IASI ob- servations is too short to directly study long-term variabil- ity of diurnal phases of the different cloud type coverage, we present in this section geographical patterns of diurnal amplitude variability in relation to climate variability, given by de-seasonalized monthly mean global surface temperature (Tsurf) anomalies. Within this short time period, changes in global Tsurf (or tropical, as both are strongly correlated) re- ﬂect the El Niño–Southern Oscillation (ENSO). ENSO is the most dominant mode of interannual variability in the Earth’s climate system and has often been used to study cloud feed- backs (e.g. Lloyd et al., 2012; Liu et al., 2017; Stephens et al., 2018). In general, a positive global Tsurf anomaly cor- responds to El Niño, with maximum convection over the central Paciﬁc, and a negative Tsurf anomaly corresponds to La Niña, with maximum convection over the West Paciﬁc. Stubenrauch et al. (2017) have shown with 15 years of AIRS- CIRS cloud data that changes in the relative coverage of trop- ical high opaque and thin cirrus with respect to increasing global mean Tsurf have different geographical patterns (see their Fig. 12): while the high opaque clouds, often linked to strong precipitation (Protopapadaki et al., 2017), relative to all clouds, increase in a narrow band in the tropics, there is a large increase in relative thin cirrus coverage around these regions. of 1◦latitude ×1◦longitude, we obtained the spatial patterns presented in Fig. 10. Indeed, we also observe an ENSO pat- tern for the change in diurnal amplitude, with an increas- ing amplitude in regions of deep convection (the band near the Equator, in the Central Paciﬁc for El Niño, and for the West Paciﬁc and South America for La Niña). The strongest changes are observed for high opaque and cirrus cloud cov- erage, but one can also distinguish them for the thin cirrus, which seem to be directly linked to convective cloud systems. Considering the relative errors of the determined slopes, also presented as geographical maps in the Supplement (Fig. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds These lags and correlations indicate that the convective cloud life cycle might be described as follows: (a) the con- vective cloud peak time precedes the cirrus anvil formation; (b) the cirrus anvil dissipates, releasing water vapour and turning to thin cirrus; (c) both the cirrus anvil and thin cirrus are strongly coupled with the high opaque core; (d) relative humidity is strongly coupled with the cirrus and lags behind, which may be associated with upper tropospheric humidiﬁ- cation by cirrus outﬂow. Figure 10. Geographical maps of linear regression slopes between de-seasonalized monthly mean diurnal amplitude anomalies from combined AIRS-CIRS and IASI-CIRS cloud data and global mean Tsurf anomalies from ERA-Interim for (a) high opaque cloud cov- erage, (b) cirrus, and (c) thin cirrus. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds We have to note that in cer- tain cases the deﬁnition of the lag becomes ambiguous due to a possible 24 h phase shift, which is not detectable in our ap- proach. For example, the 15 h peak in the January histogram for thick clouds over South Asia (Fig. 9) should be caused both by in situ formation of high clouds 4 h after the peak of local insolation and by an outﬂow of the high opaque col- umn with a characteristic time of ∼16 h (the lag between high opaque and cirrus cloud peaks). Figure 9. Circular histograms of local time of peak amplitude for Tsurf, high opaque clouds, thick cirrus, relative humidity (RH), and thin cirrus for January and July averages. The convention for vec- tor direction representation is consistent with that described in the caption of Fig. 6 while the amplitude is proportional to normalized histogram count number. “N/D” marks the cases for which the diur- nal cycle could not be reliably detected. Figure 9. Circular histograms of local time of peak amplitude for Tsurf, high opaque clouds, thick cirrus, relative humidity (RH), and thin cirrus for January and July averages. The convention for vec- tor direction representation is consistent with that described in the caption of Fig. 6 while the amplitude is proportional to normalized histogram count number. “N/D” marks the cases for which the diur- nal cycle could not be reliably detected. Another presumable element of the cloud system life cycle is the dissipation of the anvil. This can be manifested both in the RH change (Fig. 7) and in thin cirrus variation (Figs. 5c, 6c). The peak time of RH “release” from the cloud oscillates between late afternoon and early morning maxima (Kottayil For Australia, a diurnal cycle was only detectable in sum- mer (January), with two peaks in high opaque clouds and in cirrus (19:00 and 01:00 LT) and a peak in thin cirrus in the early afternoon. www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds www.atmos-chem-phys.net/19/13957/2019/ 3.4 Relating diurnal amplitudes to climate ﬂuctuations S3), though in general quite noisy, we see nevertheless that the strongest signal over the Central Paciﬁc has the smallest er- ror (of about 30 %). With this analysis we have shown in yet another way that high-level clouds linked to deep convec- tive systems have the largest diurnal amplitudes. To study the change in phase is more delicate due to possible misin- terpretation of peak amplitude around midnight and also as other factors such as winds may play a role. By applying the same technique, namely by determining a change in diurnal amplitude of cloud type coverage as a function of change in global mean Tsurf by a linear regres- sion of their monthly time anomalies, at a spatial resolution Atmos. Chem. Phys., 19, 13957–13972, 2019 4 Conclusions Multi-spectral infrared sounders are advantageous for the re- trieval of the high-level cloud properties. Their good spectral resolution allows a reliable cirrus identiﬁcation down to an IR optical depth of 0.1, day and night. However, these instru- ments are mostly aboard polar orbiting satellites, providing only observations twice per day. In this article, we presented a methodology to use the synergy of AIRS and IASI cloud observations to address the diurnal variation not only of to- tal high-level cloud coverage, but also separately for high opaque, cirrus, and thin cirrus clouds. Data availability. The monthly database of detected diurnal cy- cle amplitude and phase for UT clouds (high opaque, cir- rus, and thin cirrus), at a spatial resolution of 1◦latitude ×1◦ longitude, from the AIRS-IASI synergy can be downloaded from the ResearchGate repository using the following informa- tion: https://doi.org/10.13140/RG.2.2.13038.15681 (Feoﬁlov and Stubenrauch, 2019). Based on previous studies (Stubenrauch et al., 2017), we needed to implement the same set of surface and atmospheric ancillary data (from the meteorological ERA-Interim reanal- ysis) to extract a reliable diurnal cycle from the cloud re- trievals performed on different satellite instrument measure- ments. We demonstrated the feasibility to determine the diur- nal cycle amplitude and phase from just four measurements per day using the “sliding proﬁle approach”, which is based on the correlation of a measured variation with an assumed shape of the diurnal cycle. For the combination of AIRS and IASI, this approach allows the diurnal variation phase to be estimated with an accuracy of ±1.5 h while the amplitude is determined with ∼20 % accuracy. The zonally averaged diurnal cycle of high-level cloud coverage estimated from AIRS and IASI cloud data com- pares relatively well with other datasets, such as the one de- termined from CATS lidar observations and the one from ISCCP multispectral imager observations. Slight differences can be understood by much more limited statistics of CATS, in particular in the tropics, and misidentiﬁcation of thin cirrus by ISCCP. Considering diurnal variations of high opaque, cirrus, and thin cirrus coverage separately leads to a better understand- ing, as one can also study the lags between the different cloud types, which also have different radiative effects. In general the amplitude of the diurnal cycle is larger over land than over ocean and largest in the tropics, followed by summer mid-latitudes, in agreement with other analyses. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Lags and correlation coefﬁcients between high opaque, cir- rus, thin cirrus and UT relative humidity indicate a life cy- cle of continental tropical convective systems as (a) the con- vective cloud peak time precedes the cirrus anvil formation; (b) the cirrus anvil dissipates, releasing water vapour and turning to thin cirrus; and (c) both the cirrus anvils and thin cirrus are strongly coupled with the high opaque core. www.atmos-chem-phys.net/19/13957/2019/ A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13968 www.atmos-chem-phys.net/19/13957/2019/ 4 Conclusions By using the time series of the 8 years of combined AIRS-CIRS and IASI-CIRS data, we could relate diurnal amplitude changes to climate ﬂuctuations, given by de-seasonalized global Tsurf anomalies from ERA-Interim, and found that largest diurnal amplitudes of high-level cloud coverage are linked to deep convective systems. A more detailed analysis of speciﬁc regions over land has shown that the largest diurnal cycle seems to prevail dur- ing the summer monsoon over the Indian subcontinent, with 8.3 % for high opaque clouds, 11.7 % for cirrus, and 8.0 % for thin cirrus clouds. The local peak times vary with cloud type, season, and location, but in general AIRS alone with its observations at 01:30 and 13:30 LT is closer to capturing the maximum and minimum of cloud cover over tropical land than IASI observing the atmosphere at 09:30 and 21:30 LT. www.atmos-chem-phys.net/19/13957/2019/ Appendix B: Diurnal variation of surface temperature To avoid error caused by averaging the phases in the vicin- ity of 24:00–00:00 LT transition (for example, direct aver- aging of 23:00 and 01:00 LT returns noon instead of mid- night), we apply a “resulting force” algorithm, which resem- bles the calculation of the tilt of a disc with masses on its edges (Fig. A1): the positions of the “masses” on the disc’s perimeter correspond to phase values while the masses them- selves are proportional to the amplitudes. When everything is set up and the disc is “released”, the direction of the tilt de- ﬁnes the average phase (arrow in Fig. A1b). We searched the parameters of diurnal variation of surface temperature using the shape provided by Aires et al. (2004), which we approximate as follows: A(t) = 0.95 · 1 t2 + 15 + e−0.08·(t−13.2)2 + 0.04 · t , (B1) (B1) where the amplitude A(t) is in kelvin (K), time t is in hours (h), and the coefﬁcients come from the best-ﬁt approxima- tion. As one can see (Fig. B1 and Tables 1 and 2), the ampli- tude of Tsurf variation over land is large in tropical areas and in the summer hemisphere, reaching 20 K in the deserts (the doubled amplitude corresponds to max(Tsurf) −min(Tsurf) temperature span) while the variation over ocean is mostly negligible. The absolute values of the Tsurf variation agree with those reported in Göttsche and Olesen (2001), Pinker et al. (2007), Duan et al. (2014), Holmes et al. (2015), and Ruz- maikin et al. (2017). The spreading of the diurnal cycle from coastal regions out to surrounding oceans has been noted al- ready in Yang and Slingo (2000), who suggested complex land–sea-breeze effects as an explanation. The local time of the peak is quite stable in all areas where the diurnal cycle was detected, both in January and July, and we estimate the interannual variability for the Tsurf peak time for each zone to be ∼0.5 h. Depending on the geographical area, Tsurf peak time changes within 11.8–14.0 h limits, and possible mecha- nisms of the lag with respect to peak of solar insulation are discussed in Ait-Mesbach et al. (2015). According to their simulations, the soil thermal inertia “impacts directly the am- plitude of Tsurf variation with lower thermal inertia inducing higher amplitude (and vice versa)”. Their study shows that thermal inertia also impacts the turbulent heat ﬂuxes. www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 13969 Appendix B: Diurnal variation of surface temperature Figure A1. Illustration of “resulting force” algorithm for calculat- ing the average of phase values: (a) diurnal cycle amplitudes are used as “weights” (black dots of different sizes), which are placed on the perimeter of the disc in accordance with the corresponding phase values; (b) when “released”, the disc tilts in the direction of the most frequent phases with the strongest amplitudes. Figure A1. Illustration of “resulting force” algorithm for calculat- ing the average of phase values: (a) diurnal cycle amplitudes are used as “weights” (black dots of different sizes), which are placed on the perimeter of the disc in accordance with the corresponding phase values; (b) when “released”, the disc tilts in the direction of the most frequent phases with the strongest amplitudes. Figure B1. Diurnal variation of surface temperatures, estimated from ERA-Interim reanalyses, in January (a) and in July (b). The vector representation is consistent with the caption of Fig. 5. Figure B1. Diurnal variation of surface temperatures, estimated from ERA-Interim reanalyses, in January (a) and in July (b). The vector representation is consistent with the caption of Fig. 5. Atmos. Chem. Phys., 19, 13957–13972, 2019 Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13970 Supplement. The supplement related to this article is available on- line at: https://doi.org/10.5194/acp-19-13957-2019-supplement. Supplement. The supplement related to this article is available on- line at: https://doi.org/10.5194/acp-19-13957-2019-supplement. Supplement. The supplement related to this article is available on- line at: https://doi.org/10.5194/acp-19-13957-2019-supplement. Meteorol. Soc., 87, 911–926, https://doi.org/10.1175/BAMS-87- 7-911, 2006. Meteorol. Soc., 87, 911–926, https://doi.org/10.1175/BAMS-87- 7-911, 2006. Chen, S. S. and Houze Jr., R. A.: Diurnal variation and life-cycle of deep convective systems over the tropical Paciﬁc warm pool, Q. J. Roy. Meteor. Soc., 123, 357–388, 1997. Author contributions. AGF performed the cloud property retrievals from AIRS and IASI observations, developed the methodology of diurnal variation retrieval from the combination of two infrared sounders, and analysed the data. CJS compared the retrieved diur- nal variations with those retrieved from CATS lidar, performed the zonal analysis, and established the links between the cloud system elements. Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A. C. M., van de Berg, L., Bid- lot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Geer, A. J., Haimberger, L., Healy, S. B., Hersbach, H., Hólm, E. V., Isaksen, L., Kållberg, P., Köhler, M., Matricardi, M., McNally, A. P., Monge-Sanz, B. M., Morcrette, J.-J., Park, B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thépaut, J.-N., and Vitart, F.: The ERA-Interim reanalysis: conﬁguration and performance of the data assimilation system, Q. J. Roy. Meteor. Soc., 137, 553–597, https://doi.org/10.1002/qj.828, 2011. Competing interests. The authors declare that they have no conﬂict of interest. Duan, S.-B., Li, Z.-L., Tang, B.-H., Wu, H., Tang, R., Bi, Y., and Zhou, G.: Estimation of diurnal cycle of land sur- face temperature at high temporal and spatial resolution from clear-sky MODIS Data, Remote Sensing, 6, 3247–3262, https://doi.org/10.3390/rs6043247, 2014. Acknowledgements. This research was supported by ESRIN within the framework of ESA Climate Change Initiative (ESA CCI), Phase I, and by CNRS. Eastman, R. and Warren, S. G.: Diurnal cycles of cumulus, cu- mulonimbus, stratus, stratocumulus, and fog from surface ob- servations over land and ocean, J. Climate, 27, 2386–2404, https://doi.org/10.1175/JCLI-D-13-00352.1, 2014. Financial support. This research has been supported by the ESRIN (grant no. 4000101773/10/I-LG). Financial support. This research has been supported by the ESRIN (grant no. 4000101773/10/I-LG). Feoﬁlov, A. and Stubenrauch, C.: LMD Cloud Re- trieval using IR sounders. References Goldberg, R. A., Feoﬁlov, A. G., Pesnell, W. D., and Kutepov, A. A.: Inter-hemispheric Coupling during Northern Po- lar Summer Periods of 2002–2010 using TIMED/SABER Measurements, J. Atmos. Sol.-Terr. Phys., 104, 277–284, https://doi.org/10.1016/j.jastp.2012.11.018, 2013. Aires, F., Prigent, C., and Rossow, W. B.: Temporal interpolation of global surface skin temperature diurnal cycle over land un- der clear and cloudy conditions, J. Geophys. Res., 109, D04313, https://doi.org/10.1029/2003JD003527, 2004. Göttsche, F.-M. and Olesen, F. S.: Modelling of diurnal cycles of brightness temperature extracted from ME- TEOSAT data, Remote Sens. Environ., 76, 337–348, https://doi.org/10.1016/S0034-4257(00)00214-5, 2001. Ait-Mesbah, S., Dufresne, J. L., Cheruy, F., and Hourdin, F.: The role of thermal inertia in the representation of mean and diurnal range of surface temperature in semi- arid and arid regions, Geophys. Res. Lett., 42, 7572–7580, https://doi.org/10.1002/2015gl065553, 2015. Hilton, F., Armante, R., August, T., Barnet, C., Bouchard, A., Camy-Peyret, C., Capelle, V., Clarisse, L., Clerbaux, C., Co- heur, P., Collard, A., Crevoisier, C., Dufour, G., Edwards, D., Faijan, F., Fourrié, N., Gambacorta, A., Goldberg, M., Guidard, V., Hurtmans, D., Illingworth, S., Jacquinet-Husson, N., Kerzen- macher, T., Klaes, D., Lavanant, L., Masiello, G., Matricardi, M., McNally, A., Newman, S., Pavelin, E., Payan, S., Péquignot, E., Peyridieu, S., Phulpin, T., Remedios, J., Schlüssel, P., Serio, C., Strow, L., Stubenrauch, C., Taylor, J., Tobin, D., Wolf, W., and Zhou, D.: Hyperspectral Earth Observation from IASI: Five Years of Accomplishments, B. Am. Meteorol. Soc., 93, 347–370, https://doi.org/10.1175/BAMS-D-11-00027.1, 2012. Bowman, K. P., Collier, J. C., North, G. R., Wu, Q., Ha, E., and Hardin, J.: Diurnal cycle of tropical precipitation in Tropical Rainfall Measuring Mission (TRMM) satellite and ocean buoy rain gauge data, J. Geophys. Res., 110, D21104, https://doi.org/10.1029/2005JD005763, 2005. Cairns, B.: Diurnal variations of cloud from ISCCP data, Atmos. Res., 37, 133–146, https://doi.org/10.1016/0169- 8095(94)00074-N, 1995. Chahine, M. T., Pagano, T. S., Aumann, H. H., Atlas, R., Barnet, C., Blaisdell, J., Chen, L., Divakarla, M., Fetzer, E. J., Goldberg, M., Gautier, C., Granger, S., Hannon, S., Irion, F. W., Kakar, R., Kalnay, E., Lambrigtsen, B. H., Lee, S., Le Marshall, J., McMillan, W. W., McMillin, L., Olsen, E. T., Revercomb, H., Rosenkranz, P., Smith, W. L., Staelin, D., Strow, L. L., Susskind, J., Tobin, D., Wolf, W., and Zhou, L.: AIRS: Improving weather forecasting and providing new data on greenhouse gases, B. Am. Holmes, T. R. H., Crow, W. T., Hain, C., Anderson, M. C., and Kus- tas, W. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds Algorithm Theoretical Basis, CIRS-LMD software package V2, 19 pp., https://doi.org/10.13140/RG.2.2.15812.63361, 2017. Review statement. This paper was edited by Martina Krämer and reviewed by three anonymous referees. Review statement. This paper was edited by Martina Krämer and reviewed by three anonymous referees. Feoﬁlov, A. and Stubenrauch, C.: Diurnal cycle of high clouds retrieved from the synergy of AIRS and IASI infrared sounders for 2008–2015, ResearchGate, https://doi.org/10.13140/RG.2.2.13038.15681, 2019. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13971 mation mechanisms, J. Geophys. Res., 117, D06218, https://doi.org/10.1029/2011JD016832, 2012. Hong, G., Heygster, G., and Rodriguez, C. A. M.: Ef- fect of cirrus clouds on the diurnal cycle of tropical deep convective clouds, J. Geophys. Res., 111, D06209, https://doi.org/10.1029/2005JD006208, 2006. Rossow, W. B. and Cairns, B.: Monitoring changes of clouds, Climatic Change, 31, 175–217, https://doi.org/10.1007/BF01095151, 1995. Horvath, A. and Soden, B.: Lagrangian Diagnostics of Tropical Deep Convection and its Effect upon Upper- Tropospheric Humidity, J. Climate, 21, 1013–1028, https://doi.org/10.1175/2007JCL1786.1, 2008. Rossow, W. B. and Schiffer, R. A.: Advances in un- derstanding clouds from ISCCP, B. Am. Meteorol. Soc., 80, 2261–2288, https://doi.org/10.1175/1520- 0477(1999)080<2261:AIUCFI>2.0.CO;2, 1999. Kottayil, A., John, V. O., Buehler, S. A., and Mohanakumar, K.: Evaluating the Diurnal Cycle of Upper Tropospheric Humidity in Two Different Climate Models Using Satellite Observations, Remote Sensing, 8, 325–338, https://doi.org/10.3390/rs8040325, 2016. Rossow, W. B., Delo, C., and Cairns, B., Implications of the Ob- served Mesoscale Variations of Clouds for the Earth’s Radiation Budget, J. Climate, 15, 557–585, https://doi.org/10.1175/1520- 0442(2002)015<0557:IOTOMV>2.0.CO;2, 2002. Ruzmaikin, A., Aumann, H. H., Lee, J., and Susskind, J.: Di- urnal cycle variability of surface temperature inferred from AIRS data, J. Geophys. Res.-Atmos., 122, 10928–10938, https://doi.org/10.1002/2016JD026265, 2017. Liu, R., Liou, K.-N., Su, H., Gu, Y., Zhao, B., Jiang, J. H., and Liu, S. C.: High cloud variations with surface temperature from 2002 to 2015: Contributions to atmospheric radiative cooling rate and precipitation changes, J. Geophys. Res.-Atmos., 122, 5457– 5471, https://doi.org/10.1002/2016JD026303, 2017. Scott, N. A., Chédin, A., Armante, R., Francis, J., Stubenrauch, C.,Chaboureau, J.-P., Chevallier, F., Claud, C., and Chéruy, F.: Characteristics of the TOVS Pathﬁnder Path-B data set, B. Am. Meteorol. Soc., 80, 2679–2701, https://doi.org/10.1175/1520- 0477(1999)080<2679:COTTPP>2.0.CO;2, 1999. Lloyd, J., Guilyardi, E., and Weller, H.: The Role of Atmo- sphere Feedbacks during ENSO in the CMIP3 Models. Part III: The Shortwave Flux Feedback, J. Climate, 25, 4275–4293, https://doi.org/10.1175/JCLI-D-11-00178.1, 2012. Ludlam, F. H.: Orographic cirrus clouds, Q. J. Roy. Meteor. Soc., 78, 552–562, https://doi.org/10.1002/qj.49707833806, 1952. Soden, B. J.: The diurnal cycle of convection, clouds, and water vapour in the tropical upper troposphere, Geophys. Res. Lett., 27, 2173–2176, https://doi.org/10.1029/2000GL011436, 2000. Luo, Z. and Rossow, W. B.: Characterizing tropical cirrus life cycle, evolution, and interaction with upper-tropospheric water vapor using Lagrangian trajectory analysis of satellite observations, J. Climate, 17, 4541–4563, https://doi.org/10.1175/3222.1, 2004. Stephens, G. L., Hakuba, M. Z., Webb, M. J., Lebsock, M., Yue, Q., Kahn, B. H., Hristova-Veleva, S., Rapp, A. D., Stubenrauch, C. J., Elsaesser, G. A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds S., and Slingo, J.: Regional Intensiﬁcation of the Tropical Hydro-logical Cycle During ENSO, Geophys. Res. Lett., 45, 4361–4370, https://doi.org/10.1029/2018GL077598, 2018. Menzel, W. P., Frey, R. A., Borbas, E. E., Baum, B. A., Cureton, G., and Bearson, N.: Reprocessing of HIRS Satellite Measure- ments from 1980 to 2015: Development towards a consistent decadal cloud record, J. Appl. Meteorol. Clim., 55, 2397–2410, https://doi.org/10.1175/JAMC-D-16-0129.1, 2016. Stubenrauch, C. J., Rossow, W. B., Chéruy, F., Scott, N. A., and Chédin, A.: Clouds as seen by Infrared Sounders (3I) and Imagers (ISCCP): Part I) Evaluation of cloud parame- ters, J. Climate, 12, 2189–2213, https://doi.org/10.1175/1520- 0442(1999)012<2189:CASBSS>2.0.CO;2, 1999a. Noel, V., Chepfer, H., Chiriaco, M., and Yorks, J.: The diur- nal cycle of cloud proﬁles over land and ocean between 51° S and 51° N, seen by the CATS spaceborne lidar from the In- ternational Space Station, Atmos. Chem. Phys., 18, 9457–9473, https://doi.org/10.5194/acp-18-9457-2018, 2018. Stubenrauch, C. J., Chédin, A., Armante, R., and Scott, N. A.: Clouds as Seen by Satellite Sounders (3I) and Im- agers (ISCCP). Part II: A New Approach for Cloud Parameter Determination in the 3I Algorithms, J. Cli- mate, 12, 2214–2223, https://doi.org/10.1175/1520- 0442(1999)012<2214:CASBSS>2.0.CO;2, 1999b. Palm, S. P., Hlavka, D. L., Selmer, P., and Pauly, R.: the Cloud Aerosol Transport System (CATS), Data Product Catalog release 7.0, available at: https://cats.gsfc.nasa.gov/media/docs/CATS_ Data_Products_Catalog.pdf (last access: 14 November 2019), 2018. Stubenrauch, C. J., Chédin, A., Rädel, G., Scott, N. A., and Serrar, S.: Cloud properties and their seasonal and diurnal variability from TOVS Path-B, J. Climate, 19, 5531–5553, https://doi.org/10.1175/JCLI3929.1, 2006. Pinker, R. T., Sun, D., Miller, M., and Robinson, G. J.: Diurnal cy- cle of land surface temperature in a desert encroachment zone as observed from satellites, Geophys. Res. Lett., 34, L11809, https://doi.org/10.1029/2007GL030186, 2007. Stubenrauch, C. J., Cros, S., Guignard, A., and Lamquin, N.: A 6-year global cloud climatology from the Atmospheric In- fraRed Sounder AIRS and a statistical analysis in synergy with CALIPSO and CloudSat, Atmos. Chem. Phys., 10, 7197–7214, https://doi.org/10.5194/acp-10-7197-2010, 2010. Protopapadaki, S. E., Stubenrauch, C. J., and Feoﬁlov, A. G.: Upper tropospheric cloud systems derived from IR sounders: properties of cirrus anvils in the tropics, Atmos. Chem. Phys., 17, 3845– 3859, https://doi.org/10.5194/acp-17-3845-2017, 2017. Queney, P.: The problem of airﬂow over mountains: a summary of theoretical studies, B. Am. Meteorol. Soc., 29, 16–26, 1948. Stubenrauch, C. J., Rossow, W. References P.: Diurnal temperature cycle as observed by thermal in- frared and microwave radiometers, Remote Sens. Environ., 158, 110–125, https://doi.org/10.1016/j.rse.2014.10.031, 2015. www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds www.atmos-chem-phys.net/19/13957/2019/ A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds B., Kinne, S., Ackerman, S., Ce- sana, G., Chepfer, H., Di Girolamo, L., Getzewich, B., Guig- nard, A, Heidinger, A., Maddux, B. C., Menzel, W. P., Min- nis, P., Pearl, C., Platnick, S., Poulsen, C., Riedi, J., Sun-Mack, S., Walther, A., Winker, D., Zeng, S., and Zhao, G.: Assess- ment of Global Cloud Datasets from Satellites: Project and Database Initiated by the GEWEX Radiation Panel, B. Am. Me- Reichler, T., Dameris, M., and Sausen, R.: Determining the tropopause height from gridded data, Geophys. Res. Lett., 30, 2042, https://doi.org/10.1029/2003GL018240, 2003. Riihimaki, L. D., McFarlane, S. A., Liang, C., Massie, S. T., Beagley, N., and Toth, T. D.: Comparison of meth- ods to determine tropical tropopause layer cirrus for- Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/ Atmos. Chem. Phys., 19, 13957–13972, 2019 A. G. Feoﬁlov and C. J. Stubenrauch: Diurnal variation of high-level clouds 13972 teorol. Soc., 94, 1031–1049, https://doi.org/10.1175/BAMS-D- 12-00117.1, 2013. Yang, G. Y. and Slingo, J. M.: The diurnal cycle in the Tropics, Mon. Weather Rev., 129, 784–801, 2001. Yorks, J. E., McGill, M. J., Palm, S. P., Hlavka, D. L., Selmer, P. A., Nowottnick, E. P., Vaughan, M. A., Rodier, S. D., and Hart, W. D.: An overview of the CATS level 1 processing algo- rithms and data products, Geophys. Res. Lett., 43, 4632–4639, https://doi.org/10.1002/2016GL068006, 2016. Stubenrauch, C. J., Feoﬁlov, A. G., Protopapadaki, S. E., and Armante, R.: Cloud climatologies from the infrared sounders AIRS and IASI: strengths and applications, Atmos. Chem. Phys., 17, 13625–13644, https://doi.org/10.5194/acp-17-13625- 2017, 2017. Tian, B., Soden, B. J., and Wu, X.: Diurnal cycle of convection, clouds, and water vapour in the tropical upper troposphere: Satel- lites versus a general circulation model, J. Geophys. Res., 109, D10101, https://doi.org/10.1029/2003JD004117, 2004. Zhang, Y., Klein, S. A., Liu, C., Tian, B., Marchand, R. T., Haynes, J. M., McCoy, R. B., Zhang, Y., and Acker- man, T. P.: On the diurnal cycle of deep convection, high- level cloud, and upper troposphere water vapor in the Multi- scale Modeling Framework, J. Geophys. Res., 113, D16105, https://doi.org/10.1029/2008JD009905, 2008. Winker, D. M., Vaughan, M. A., Omar, A. H., Hu, Y., Pow- ell, K. A., Liu, Z., Hunt, W. H., and Young, S. A.: Overview of the CALIPSO Mission and CALIOP Data Pro- cessing Algorithms, J. Atmos. Ocean. Tech., 26, 2310–2323, https://doi.org/10.1175/2009JTECHA1281.1, 2009. Zipser, E. J., Liu, C., Cecil, D. C., Nesbitt, S. W., and Yorty, D. P.: Where are the most intense thunderstorms on earth?, B. Am. Meteorol. Soc., 87, 1057–1071, https://doi.org/10.1175/BAMS- 87-8-1057, 2006. Wylie, D. P. and Woolf, H. M.: The Diurnal Cycle of Upper- Tropospheric Clouds Measured by GOES-VAS and the ISCCP, Mon. Weather Rev., 130, 171–179, https://doi.org/10.1175/1520- 0493(2002)130<0171:TDCOUT>2.0.CO;2, 2002. Atmos. Chem. Phys., 19, 13957–13972, 2019 www.atmos-chem-phys.net/19/13957/2019/
https://openalex.org/W3156209974	https://www.frontiersin.org/articles/10.3389/fpsyg.2021.636109/pdf	English	null	Rebound and Spillovers: Prosumers in Transition	Frontiers in psychology	2,021	cc-by	11,512	Rebound and Spillovers: Prosumers in Transition Elisabeth Dütschke 1, Ray Galvin 2 and Iska Brunzema 1 1 Fraunhofer Institute for Systems and Innovation Research ISI, Karlsruhe, Germany, 2 Institute for Future Energy Consumer Needs and Behavior, Aachen, Germany Generating energy by renewable sources like wind, sun or water has led to the emergence of “clean” energy that is generally available at low cost to the environment and is generated from seemingly unbounded resources. Many countries have implemented schemes to support the diffusion of renewable energies. The diffusion of micro-generation technologies like roof-top photovoltaics is one of the success stories within the energy transition and has been signiﬁcantly driven—at least in countries such as Germany—by households. As these households usually not only generate energy but also consume it they are often called “prosumers.” How does it inﬂuence the energy behavior of households if they become prosumers? Are these behavioral changes in line with further goals of the energy transition, e.g., reducing demand? What shapes individual behaviors of prosumers? The paper introduces a conceptual framework based on the existing literature on rebound and spillover effects. It systematizes possible behavioral consequences as well as mechanisms behind them. This framework is then used to code and analyze data from 48 in-depth interviews with prosumer households. These interviews reveal a broad variety of behavioral responses which have their roots in economic conditions and their evaluation by the prosumers, psychological mechanisms like central guiding principles and a clear conscience as well as sociotechnical context and legislative frameworks. Edited by: Edited by: Sebastian Bamberg, Bielefeld University of Applied Sciences, Germany Reviewed by: Jed J. Cohen, Energy Institute at Johannes Kepler University, Austria Anke Blöbaum, Otto von Guericke University Magdeburg, Germany Correspondence: Elisabeth Dütschke elisabeth.duetschke@isi.fraunhofer.de Edited by: Sebastian Bamberg, Bielefeld University of Applied Sciences, Germany Reviewed by: Jed J. Cohen, Energy Institute at Johannes Kepler University, Austria Anke Blöbaum, Otto von Guericke University Magdeburg, Germany Keywords: prosuming, rebound, spillover, psychological and economic drivers, socio-technical context HIGHLIGHTS - Private energy prosumers are a relevant group of active agents in the energy system - To support the energy transition their behavior needs to align with demand reduction goals - This interview-based study explores self-reported behaviors and how it emerges - Behavioral response is heterogeneous and driven by individual and systemic factors. - Private energy prosumers are a relevant group of active agents in the energy system - To support the energy transition their behavior needs to align with demand reduction goals - This interview-based study explores self-reported behaviors and how it emerges - Behavioral response is heterogeneous and driven by individual and systemic factors. Specialty section: This article was submitted to Environmental Psychology, a section of the journal Frontiers in Psychology INTRODUCTION Received: 30 November 2020 Accepted: 01 March 2021 Published: 15 April 2021 The transformation of conventional energy systems that heavily rely on fossil fuels is a crucial element in strategies to solve humanity’s current major challenge of achieving climate change mitigation goals and enhancing sustainability in order to stay within the limits of planetary boundaries. Increasing the shares of renewable energy, i.e., energy that is gained from resources like wind, water, and sun is one of the main pathways in the energy system transformation. Most prominent so far is the transition of the electricity sector by installing windfarms, biomass power plants, hydroelectric power stations, and photovoltaic (PV) panels. However, in addition to such ORIGINAL RESEARCH published: 15 April 2021 doi: 10.3389/fpsyg.2021.636109 Behavioral Consequences of PV Use Behavioral Consequences of PV Use Energy behavior refers to broad categories, ranging from everyday routines which are usually mainly shaped by habits, social practices, learned schemata and situational cues and performed without much cognitive eﬀort (e.g., turning on the lights) to conscious decision making processes of much lower frequency that involve more extensive evaluation of potential risks, beneﬁts, and probable outcomes (e.g., buying a home, installing a PV). In comparison to the habitual daily behaviors such investment behavior is sometimes called one-shot behavior. Potential behavioral consequences of PV use refer to all these diﬀerent types of behavior. With regard to the energy transition, all these behavior types could be beneﬁcial in the sense that, for example, they could contribute to reducing energy consumption or the level of demand management by synchronizing supply with demand, or have adverse eﬀects by increasing consumption. The main topic of this paper is to take a closer look at the interplay between households’ understanding of their role in the energy system and their experiences and perceptions. Therefore, this paper takes a close look at prosuming households, i.e., households owning a photovoltaic (PV) system and their energy lifestyle. More speciﬁcally, we analyze how being a prosumer inﬂuences households’ energy-related behaviors. As a frame of reference to address this question we draw on current streams of literature that analyze rebound and spillover eﬀects. While mainstream rebound eﬀects literature describes unexpected shortfalls in reductions in energy demand following an increase in energy eﬃciency of an energy service (Chitnis et al., 2014), the literature on spillover refers to broader behavioral changes when an environmental behavior triggers further changes in other behaviors (Nash et al., 2017). Thus, the two concepts describe two sides of the same coin as they both account for how prior behavior—in our case becoming a prosumer—inﬂuences later behaviors. From a normative perspective, spillover refers to the positive side of further increases in environmental behaviors, while rebound captures the downside of more demand and resource-use. In a ﬁrst step, this paper investigates behavioral consequences using the rebound- spillover dimension as a normative anchor. Furthermore, as outlined in more detail below, both literatures have identiﬁed possible mechanisms underlying such behavioral consequences. Citation: Dütschke E, Galvin R and Brunzema I (2021) Rebound and Spillovers: Prosumers in Transition. Front. Psychol. 12:636109. doi: 10.3389/fpsyg.2021.636109 April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org Prosumers in Transition Dütschke et al. a supply side oriented approach, all prominent scenarios for the transformation of the energy system also encompass the reduction of the demand for current energy services by increasing energy eﬃciency (e.g., IEA and IRENA, 2017). A well-known example for such a combined strategy is the 20-20-20-goals of the European Union (EU) which foresaw a 20% cut in greenhouse gas emissions (from 1990 levels), 20% of EU energy from renewables, and a 20% improvement in energy eﬃciency by 2020; the EU goals for 2030 were again deﬁned in a similar way. Citizens and their investment decisions as well as their daily behaviors play an important role for the success of these scenarios. Micro- generation technologies have become available at decreasing prices and have found considerable support from policies like Feed-In-Tariﬀs (FITs) which made them a safe and proﬁtable investment for many. Consequently, private investors including households are playing a signiﬁcant role in this ﬁeld and these households have become so-called “prosumers” who generate and use their own electricity in addition to feeding it into the grid. Thus, the role of households in the energy system has been enlarged and at the same time is subject to expectations with regard to system contributions, i.e., keeping their demand stable or reducing it while contributing to supply. 2020). To build on this new stream this paper ﬁrstly advances a conceptual framework within which prosumer energy behavior can be evaluated. Secondly, it applies this empirically by drawing on 48 in-depth interviews with prosumer households in Germany. The interview data is analyzed with respect to (i) behavioral consequences of being a prosumer and (ii) underlying mechanisms to these behaviors. The next sections ﬁrst further develop the conceptual background by deﬁning relevant terms and describing possible outcomes of being a prosumer. This includes a categorization of possible underlying mechanisms. We then present the empirical data, describing the methods for data collection and analysis before presenting ﬁndings. In the concluding discussion we refer back to the broader embeddedness of prosumers in the energy system as a system under transition. Behavioral Consequences of PV Use Traditionally, economic approaches emphasizing changes in prices and available income have featured prominently in the literature (Dimitropoulos et al., 2018), but further researchers have also emphasized psychological mechanisms (Peters and Dütschke, 2016; Dütschke et al., 2018; Seebauer, 2018) and socio- technical conﬁgurations (Galvin, 2020). Frontiers in Psychology \| www.frontiersin.org Mechanisms Behind Rebound and Spillover Mechanisms Behind Rebound and Spillover Economists have often associated rebound eﬀect with price eﬀects, i.e., if the usage of a service gets cheaper due to lower energy demand, then the demand for this service will increase (Dimitropoulos et al., 2018). These approaches usually do not consider the upfront investment but focus on the costs for obtaining the energy service. Applying this to the case of electricity generation with rooftop PV without considering the initial investment, the economics for the lifetime of the PV system strongly depend on the policy framework. In Germany, where our empirical case studies are situated, payments for renewable energy are governed by the Renewable Energy Law [Erneuerbare-Energien-Gesetz (EEG)]; the EEG has been revised several times and now incentivizes households to use the electricity from their PV themselves as this is cheaper than buying it from the grid. This can be maximized by households if they shift their consumption to times of (higher) generation. Thus, similar to the case of eﬃciency rebounds, prosumer households pay less for electricity services compared to non- prosumer households. May also occur in an opposite direction, and this is also supported by the literature reporting further reduction in demand or more broadly rising eﬀorts of environmental behaviors (Truelove et al., 2014). The term (positive)1 spillover is used for eﬀects in diﬀerent domains (Galizzi and Whitmarsh, 2019), e.g., if the installation of a more eﬃcient heating system is followed by electricity saving measures or triggers the purchase of a more eﬃcient car. The rebound and spillover literatures have developed independently of each other, but have acknowledged each other’s respective phenomena. For example, rebound literature has deﬁned terms like reverse rebound (Chenavaz et al., 2021), prebound (Sunikka-Blank and Galvin, 2012), or super-conservation (Saunders, 2008; Li et al., 2020) to refer to situations where the actual energy demand falls below the expected. Similarly, research also refers to “permitting” or “negative” spillover to describe rebound-type eﬀects (Galizzi and Whitmarsh, 2019). Taking the learnings from these literatures together, this paper combines the notion of rebound and spillover to describe the two sides of the same coin. To diﬀerentiate between eﬀects in the same or other domains analogously to direct and indirect rebound eﬀects, we will use the term conservation for eﬀects in the same domain and spillovers for eﬀects in other behavioral domains (see Figure 1 for an overview on the terms). 1Sometimes the literature diﬀerentiates between positive and negative spillover. Negative spillover eﬀects are conceptually identical to the concept of indirect rebound eﬀects (Nash et al., 2017). Mechanisms Behind Rebound and Spillover In addition to economic inﬂuences on behavior, the literature also suggests that psychological factors can foster or limit the emergence of rebound eﬀects. This has to do with the degree to which needs are already satisﬁed (Hofstetter et al., 2006; Wörsdorfer, 2010), and norms and attitudes toward the relevant behavior and toward the environment (Haan et al., 2007; Matiaske et al., 2012). Peters and Dütschke (2016) proposed and empirically explored a conceptual model covering these concepts. Recently, moral licensing and consistency as explanatory factors have emerged in the literature (Dütschke et al., 2018). The moral licensing concept assumes that past morally positive behavior increases the probability that people will subsequently show potentially less moral behavior (Mazar and Zhong, 2010; Mullen and Monin, 2016). For behavioral spillovers, social and environmental identity have also been investigated (Elf et al., 2018; van der Werﬀand Steg, 2018; Verfuerth et al., 2019). Overall empirical research on these types of factors is rare so far, even more so in respect of renewable energy. From a conceptual point of view, all of the concepts under discussion seem highly applicable to also trigger rebound or spillovers in the case of renewables or more speciﬁcally the installation of a PV system. For example, studies have shown that investments in PV are likely to be regarded as environmental behaviors (Palm and Tengvard, 2011; Korcaj et al., 2015). These investments could thus provide a basis for a moral license, i.e., less environmentally friendly behavior and therefore lead to higher consumption. Alternatively, they could trigger consistent behavior, i.e., curtailment of consumption, by making an environmental identity or energy-related topics more salient. Transferring the deﬁnition of rebound eﬀects to the ﬁeld of renewable energy, a direct rebound eﬀect in renewable energy use occurs if there is a higher demand for the same energy carrier when renewable energy is involved, compared to when no or less renewable energy is involved. In the case of household prosumers this would mean that the demand for electricity increases after installing a PV system, for example by buying additional appliances or using existing appliances more extensively. An indirect rebound eﬀect of renewable energy use would occur if the demand for energy or other resources increases in other domains, e.g., an increase in travel or heating after installing a PV system. Deﬁning Rebound and Spillover Research on rebound eﬀects has traditionally mainly emerged from studying the eﬀects of increases in energy eﬃciency. It refers to the phenomenon that often the implementation of an energy eﬃciency measure does not lead to the expected level of energy savings but these remain at lower levels (Sorrell, 2015). Quantiﬁcations of rebound eﬀects are usually estimated by subtracting the ratio of actual savings to expected savings from one, or alternately expressed: they are the ratio between the shortfall in savings and the expected savings. Psychological approaches to the rebound eﬀect agree with this deﬁnition in principle, but emphasize behavioral aspects and determinants (Dütschke et al., 2018). From their perspective the increase in energy eﬃciency is understood “as an intervention that interrupts previous routines and thereby leads to behavioral change in how the relevant product or service is used” (Dütschke et al., 2018, p. 5). If this behavioral change intensiﬁes the use of an energy service, this is observed as a rebound eﬀect. Often authors diﬀerentiate between direct and indirect rebounds depending on whether the increase in demand occurs in the same or another behavioral domain (Chitnis et al., 2014). A body of literature that investigates behavioral consequences of using renewable energy sources or more speciﬁcally installing PV systems has recently begun to emerge (Wittenberg and Matthies, 2016; Oberst et al., 2019; Qiu et al., 2019; Li et al., April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 2 Dütschke et al. Prosumers in Transition e.g., thinking about and actually implementing home insulation after installing a rooftop-PV. FIGURE 1 \| Overview of potential behavioral consequences of the way how energy is used or supplied. FIGURE 1 \| Overview of potential behavioral consequences of the way how energy is used or supplied. Figure 1 takes up the notion that either increases in energy eﬃciency or a change to renewable energy supply could trigger behavioral responses, and summarizes the diﬀerent eﬀects. Mechanisms Behind Rebound and Spillover In a similar vein, the concepts of spillover and conservation can also be transferred to the area of renewable energy use. The change to renewable energy would be said to trigger conservation if the demand is lower than before, e.g., if, after installing a PV, everyday usage behavior is changed such that lower electricity demand results (for example by turning lights oﬀmore frequently). Finally, there could be spillover to other domains, While psychological approaches put a strong emphasis on individual control, they partly neglect the socio-cultural habitual April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 3 Dütschke et al. Prosumers in Transition embedding of behavior (e.g., learned behavioral patterns) (Galvin and Gubernat, 2016; Sonnberger and Gross, 2018) as well as socio-structural factors. Galvin (2013) elaborates on this for the example of windows: Most German windows are very badly designed for eﬃcient manual ventilation by opening inwards in combination with the cultural habit of decorating window sills—thus this limits behavior that is ideal for energy eﬃcient ventilation. In case of solar PV, the inﬂuence of such socio-technical structures is highly relevant in relation to the synchronization of supply and demand. Technical devices and ICT can support the synchronization which is otherwise limited to everyday heuristics by weather observations. However, such supporting technology is also likely to bring very speciﬁc conditions regarding practicalities that encourage or impede certain behaviors (cf. for example Wittenberg and Matthies, 2016 on the visibility in everyday life). The current German regulation on peak load prevention is an example of such a conﬁguration for the case under study: it sets an incentive to use this electricity that prosumers otherwise perceive to be wasted. This could trigger households to make investments in, for example, electric mobility (bikes, cars) to make use of this electricity, and this could lead to higher demand overall (Galvin, 2020). blog. Participants were interviewed twice and reported their consumption data based on their entries in the web user interfaces of their electricity retailers. The researchers observed no major changes in consumption and hardly any indications of shifting demand according to electricity generation, but increasing energy awareness. Qiu et al. (2019) obtained data from a utility company in the US including electricity meter data and survey data. Mechanisms Behind Rebound and Spillover In contrast to the studies cited before they estimated a solar rebound as high as 18 % by comparing the energy consumption of prosumer households with non-prosumers, and of 15% by comparing pre- and post-installation consumption (Qiu et al., 2019). The study also found eﬀects of moderating variables, i.e., consumers from a neighborhood with more green/left wing voters showed smaller rebound eﬀects. In a recent study, Li et al. (2020) who also combine metering and survey data ﬁnd a small conservation eﬀect for US PV prosumers who are ﬁnancially incentivized to feed as much of their self-produced electricity into the grid as possible. Finally, Oberst et al. (2019) investigate energy use more broadly by analyzing self-reported heating costs for PV prosumer households and ﬁnd no diﬀerences to non-prosumers. Thus, the overall literature gives little or no consistent indication as to what (quantitative) extent the issue of rebound or spillover eﬀects is relevant to PV households. The few studies available point out that there is variety among prosumers, and this appears to lead back to the categories of factors as identiﬁed in Figure 2. To enhance the state of knowledge we therefore explore the topic further through an analysis of 48 interviews with German prosumer households. Other sources of rebound include lack of knowledge and technical or design failures. For example if PV modules are not set at an optimal angle, system components not optimally combined or settings of control units are wrong, this could lead to other energy demand patterns than anticipated. This can be due to lack of knowledge by users or installers, as well as the complexity of systems. A qualitative study by Peters and Dütschke (2016) found some evidence in this direction with regard to heating systems but also for lighting. Contextual Background The study presented here is situated in Germany. PV panels are the dominant technology in this country for private self- generation of electricity. In 2016, around 8% of residential buildings in Germany were already equipped with a PV system, with the proportion particularly high for newer buildings, detached houses and buildings in southern Germany (Cischinsky and Diefenbach, 2018). PV generation is overall ﬁnancially attractive for households, with high investments initially but very low running costs (Haar, 2020). DATA AND METHODS Figure 2 summarizes the list of mechanisms identiﬁed from the literature. Frontiers in Psychology \| www.frontiersin.org Description of Database grid than to consume one’s own electricity for households who installed PV up to 2012. For households who have installed PV since then, self-consumption is ﬁnancially more attractive as the guaranteed FIT per kWh from then on became lower than the (average) price households pay for using electricity from the grid. This diﬀerence has constantly increased since then (Wirth, 2020). Thus, depending on when the PV system was installed, it is either more attractive for households to feed all electricity to the grid or more proﬁtable to use it themselves, though with some diﬀerences regarding the precise economic beneﬁt. Four series of interviews serve as the database for this study. The total of 48 interviews were conducted in Germany between July 2017 and March 2019. They were obtained in four regional clusters and through a variety of recruitment procedures: (1) State of Hesse: The ﬁrst series of interviewees was conducted between July and September 2017. The homes of the 13 respondents were mainly situated around the city of Darmstadt in the southern part of the State of Hesse which is at the center of Germany. Interviewers contacted potential participants by ringing at the door if PV systems were visible from outside or via internet maps. An earlier paper based on these interviews investigated the motivation to adopt a PV system (Köhler et al., 2019). Under current legislation that was valid for the most recent interviewees as well as at the time of writing this paper, consuming self-generated PV electricity is free from electricity taxes and levies for PV installations below 10 kWp;2 but at higher capacities 40% of the regular EEG levy of a few cents must be paid per kWh consumed (EEG, 2017)3. However, the ﬁnancial beneﬁt from self-consumed electricity is subject to income taxes. Additionally, the current legislation guarantees a FIT of around 10 ct/kWh to households for the electricity they still feed to the grid if they do not use it themselves. In any case, consuming self-generated electricity is still cheaper than obtaining electricity from the grid, where prices are around 30 ct/kWh (BMWi, 2020). Thus, consuming electricity from a PV system installed by a household after 2012 leads to lower costs. 2Although the marginal cost of producing each extra kWh is zero (Haar, 2020), there are still maintenance costs for PV and its electrical circuit technology. Most of these are usually very low, but storage batteries and DC-AC converters are expensive to replace if they fail. 3In 2020, this was 2,7 ect / kWh, cf. Bundesnetzagentur (2020). 4As the real output is rarely higher due to weather and technical conditions, the actual resulting loss is only about 2-5%, cf. Wirth (2020). Empirical Findings in the Literature FIGURE 2 \| Schematic overview of potential behavioral effects and underlying mechanisms of efﬁciency increases. Own ﬁgure further developed from Dütschke et al. (2018) Description of Database Furthermore, to prevent grid overloads at peak generating times (e.g., midday in summer), PV system owners in Germany are obliged to allow grid operators to regulate their system (receiving lump-sum compensation for revenue lost); alternatively, smaller systems below 30 kW can limit their feed-in to 70% of their maximum eﬀective power (EEG, 2017). Thus, for most households with recent PV installations, there is a limit to the amount that households can feed to the grid4. (2) (2) Wüstenrot: This small cluster was recruited at a citizen assembly and focused on inhabitants of an innovative building site at the small town of Wüstenrot, which is located in a rural area between the agglomerations of the cities Stuttgart and Heilbronn. In order to stand out and become attractive for potential citizens, the municipality has been pursuing local energy projects for some time. All houses on this newly developed housing estate were obliged to be equipped with a PV system. The homes are heated by an innovative heat network based on near-surface geothermal energy. Four households participated in the interviews which were conducted in March 2018. These were recruited at a citizen assembly for inhabitants of the housing estate. (2) Wüstenrot: This small cluster was recruited at a citizen assembly and focused on inhabitants of an innovative building site at the small town of Wüstenrot, which is located in a rural area between the agglomerations of the cities Stuttgart and Heilbronn. In order to stand out and become attractive for potential citizens, the municipality has been pursuing local energy projects for some time. All houses on this newly developed housing estate were obliged to be equipped with a PV system. The homes are heated by an innovative heat network based on near-surface geothermal energy. Four households participated in the interviews which were conducted in March 2018. These were recruited at a citizen assembly for inhabitants of the housing estate. (3) (3) Lower Franconia: 16 interviews with prosumer households in rural villages and towns around Schweinfurt in Lower Franconia, in the northern part of Bavaria, were conducted in February and March 2019. Interviewees were recruited through municipal newsletters, through a staﬀmember of Schweinfurt County’s energy support team and ﬁnally through local contacts of the authors. A paper focusing on other questions than those of the current paper is published by Galvin (2020) employing these interviews as a data source. Empirical Findings in the Literature g The body of literature that examines potential consequences of small-scale PV on individual energy demand or more broadly of renewable energy use has only recently been emerging (cf. Luthander et al., 2015 for a review on earlier literature). The ﬁndings published so far cover a variety of samples studied by qualitative and quantitative approaches in diﬀerent contexts and, thus, heterogeneous political and contextual factors. Consequently, the results vary substantially: Studies by Wittenberg and colleagues (Wittenberg and Matthies, 2016, 2018) used a German sample of more than 400 PV owners recruited by spreading the questionnaire through dedicated webpages. They obtained self-reported meter readings as well as questionnaire data. However, the quantitative analyses were limited in places due to missing data and small size of subgroups. Overall these two studies do not detect signiﬁcant diﬀerences in consumption compared to general consumption in the population as reported in oﬃcial statistics, but reveal support for a relationship between self-reported energy saving behaviors and positive environmental attitudes. Palm et al. (2018) interviewed 44 prosumer households in Sweden. These were recruited through a variety of sources, e.g., contacts from the energy agency, solar installers and advertisement on a In 2018, around 20% of the renewable electricity generated in Germany was produced by PV, including large PV ﬁeld arrays, and this contributed 7.7% to gross electricity consumption (ZSW and UBA, 2019). Of the installed capacity of German PV systems 15% falls into the category of up to 10 kWp and 34% in the range from 10 to 100 kWp (Wirth, 2020). After years of strong growth between 2005 and 2012 growth rates have slowed down (ZSW and UBA, 2019) since the policy and regulatory framework has changed. German legislation mandates that the level of FIT at the time of installation of PV applies for 20 years. While the FIT for small-scale PV was around 57 ct/kWh for PV installed in 2004 it has constantly decreased since then and was around 11 ct/kWh for units installed in 2019 (Kelm et al., 2019). It is ﬁnancially more attractive to feed PV electricity into the April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org Frontiers in Psychology \| www.frontiersin.org 4 Dütschke et al. Prosumers in Transition FIGURE 2 \| Schematic overview of potential behavioral effects and underlying mechanisms of efﬁciency increases. Own ﬁgure further developed from Dütschke et al. (2018). 4As the real output is rarely higher due to weather and technical conditions, the actual resulting loss is only about 2-5%, cf. Wirth (2020). Description of Database First, the interviews from series 1 and 2 were coded by the ﬁrst author using a simpliﬁed coding scheme on behavioral consequences, extracting (1) further (intended) investments, (2) behavioral changes regarding electricity use, (3) synchronicity of consumption behavior with the sun, and (4) behavioral changes in other domains like water, transport, In this ﬁrst analysis the lines of arguments by which households explained their respective behaviors and the behavioral outcomes were not separated from each other but subsumed using the same main codes. In a next step, the third author of the paper coded all interviews from all four series with a focus on the behavioral consequences. The main codes in this step were the same as above excluding quotes on underlying mechanisms and extending the behavior change category also to explicit statements that behaviors have not changed. In this step the code assignment in the ﬁrst cluster was also checked for diverging interpretations, and high levels of agreement emerged. Finally, the ﬁrst author reﬁned the coding on behavioral consequences by going through all interviews again and additionally coding the underlying mechanisms. In a next step, the quotes on the subcodes were extracted by the ﬁrst author and densiﬁed according to themes to allow for counting frequencies where applicable, e.g., regarding the technologies the households invested in. The motivation behind combining these diﬀerent clusters was to acquire a broad sample which is heterogeneous, for example with regard to local history and context including local discourses on renewable energy. This rationale was fueled by the aim that a qualitative study is appropriate when the goal is to further develop theory and enhance the in-depth knowledge on a topic. Thus, the main goal for sample composition is to make sure the full variety of the subject under study is captured. For this reason we also combined a variety of recruitment strategies, e.g., trying to acquire both more and less eager participants. The speciﬁc recruitment strategies were outlined above. Due to their heterogeneity it is not possible to estimate response rates. g y p p The interviews were on the household level, i.e., in some cases more than one household member participated. More speciﬁcally the 48 households were represented by 32 men, three women, eleven couples and two women with their adult sons living in the same home. The average age of interviewees was 56, ranging from 27 to 82. Description of Database (3) Lower Franconia: 16 interviews with prosumer households in rural villages and towns around Schweinfurt in Lower Franconia, in the northern part of Bavaria, were conducted in February and March 2019. Interviewees were recruited through municipal newsletters, through a staﬀmember of Schweinfurt County’s energy support team and ﬁnally through local contacts of the authors. A paper focusing on other questions than those of the current paper is published by Galvin (2020) employing these interviews as a data source. (4) (4) Markgräﬂerland around Freiburg: Finally another 15 participating households were recruited in the rural area around Freiburg in the southwest of Germany near the borders to France and Switzerland. Again, municipal newsletters were used to ﬁnd interviewees, this was April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 5 Dütschke et al. Prosumers in Transition complemented by pre-identifying relevant homes through internet maps and ringing doorbells. The interviews were held in March 2019. the PV; and questions about the local context. All interviews were accompanied by a short written questionnaire to assemble some key data about the household and its composition, electricity consumption, and PV system. The interview conversations were digitally recorded and transcribed verbatim. The text corpus of these transcripts adds up to an amount of 280,310 words overall. The interviews in cluster 1 were part of a psychology student research project and conducted under close supervision of the corresponding author. In Clusters 2 and 4 interviews were conducted by experienced interviewers from the corresponding author’s institution. In cluster 3 the second author, who is also an experienced interviewer, social scientist and former electrical engineer, did the interviews. Originally, series 3 and 4 also included a few additional households who owned solar-thermal panels to heat water but no PV; they were excluded for reasons of consistency in the current paper. p p As indicated above, parts of the interviews have been analyzed for diﬀerent research questions. For the research interests under study in the present paper a coding frame was developed including main codes and subcodes (see Table 1) by applying content analysis starting with a theory driven deductive approach and reﬁning the coding scheme inductively where necessary (Mayring, 2015). Description of Database Average household size was three with a range from one to seven. One third of the homes were situated in a town or city, two thirds in a rural area. We asked for self-ratings regarding income: one household saw themselves as below average, 20 as average, and 27 as above average. The solar panels were installed between 1999 and 2018 and thus cover the full range of the various FITs in this period. Nineteen households solely feed their electricity into the grid while a majority of 28 combine feeding into the grid with self-consumption and one household was not sure about this. For a detailed overview of the interview partners, see Appendix 2. The main codes are based on the concepts included in Figure 1 and displayed in Table 1. The coding process and the interpretation of results was also checked by the second author who was the interviewer of the (relatively large) Franconia study for consistency and plausibility. For a fuller account of issues that arise in coding to a high degree of reliability together with reviews of recent literature on this see O’Connor and Joﬀe (2020). In the following, where quotes are provided from the interviews they are given by letters symbolizing the region, i.e., HE, Hesse; WÜ, Wüstenrot; FRAN, Franconia; FR, Freiburg, and a number identifying the relevant interview in the sample. Interview Topics and Analyses The interviews were semi-structured, based on an interview guideline which was highly similar for clusters 1 and 2 and for clusters 3 and 4. The main diﬀerence in the interview contents of clusters 1/2 vs. 3/4 is that in the ﬁrst two clusters a larger part of the interviews focused more extensively on the adoption process and how the decision for the PV system evolved; these interviews were on average also longer than in the second two clusters. The interview guideline for the second clusters is given in the Appendix to this paper. Besides the adoption process, the guideline featured details about the PV system and technologies, investments and systems for monitoring connected with it; the motivation and aims for the installation and discussions in the household around it; energy behavior before and after installing RESULTS Behavioral Consequences Frontiers in Psychology \| www.frontiersin.org Behavioral Consequences Our analysis on behavioral consequences will start by outlining the ﬁndings on energy system investments. In this category we summarize investments in addition to the PV, that households made to save energy, to make better use of the electricity from the PV or replace the use of less sustainable energy sources, e.g., buying an electric car instead of a conventional one. This will be followed by an analysis of daily behaviors starting with (i) April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 6 Prosumers in Transition Dütschke et al. TABLE 1 \| Overview on main codes and sub-codes applied to the interview data. Main code Sub-code Behavioral consequences Energy system investments Realized further investments Future options Denied investments Daily behaviors: Synchronizing electricity demand with supply Daily behaviors: Behavioral change in electricity consumption Reduction of demand Increased demand No change More conscious consumption Daily behaviors: Behavioral change in other domains Reduction of demand Increased demand No change More conscious consumption Mechanisms behind behavioral change Individual level mechanisms Economic, psychological Socio-technical mechanisms Other themes. For those feeding all their electricity to the grid, the next anticipated step is to move to self-consumption. Further ideas are similar to those already implemented and include sustainable heating systems, smart home elements, and micro wind turbines. The reasons these ideas have not yet been implemented are heterogeneous—in about a third of cases intentions are still vague and more in the stage of ﬁrst ideas. That the necessary investment is considered as too high also plays a role. In some cases, the intention is ﬁrm but households are waiting for the right point in time, i.e., when the current car gets too old, the heating system breaks down or their guaranteed FIT is about to end. Fourteen interviewees also excluded certain investments: ﬁve had turned down the option of buying electric cars due to restricted range, environmental reasons, or high prices; four were generally skeptical about battery storage, again due to high prices or an insuﬃcient economic rationale as well as doubts that decentralized storage is beneﬁcial to the energy system. Further ideas that were turned down by one of the households included more sustainable heating systems or home renovations, the main reason being too high initial investments. Daily Behaviors: Synchronizing Electricity Demand With Supply issues around synchronizing demand with supply, (ii) electricity use more generally and ﬁnally (iii) behaviors around energy and resource use more broadly. Many of the interviewees reported some degree of synchronizing electricity demand with sunshine. However, the majority of these are from the subgroup that is engaged in self-consumption. Of those fully feeding to the grid only two out of 19 households engage in synchronizing behaviors compared to 22 out of the 28 who do not fully feed into the grid. The main synchronizing activity is to aim at using basic household appliances like washing machines, driers, and dishwashers when the sun is shining or at least during daytime. Very few combine this with setting timers or some sort of home automation, i.e., these activities are mainly performed manually and the women in the households are often the ones implementing it, with the men often presenting themselves as the ones pushing in this direction: Energy System Investments In many respondent households the PV system is not the only step toward active integration into the energy production system or the uptake of relevant innovations. Overall, three quarters of the interviewed households (36 out of 48) have made additional investments in further technologies. On average this encompasses two further investments per household, ranging from 1 to 5. Most prominent is the use of a solar thermal system (15 households), battery storage (9), or a heat pump (8). Seven report that they implemented high insulation standards, including passive house standard in some cases; seven interviewees state that they use a sustainable heating system, e.g., running on wood or as a combined heat and power unit. Overall, ten use alternative drives for their vehicles, most prominently full electric cars (5). Further investments include highly eﬃcient household appliances and lighting, water re-use systems, and smart home equipment. The timing of these investments and how they relate to owning the PV is often not fully clear in the interviews. Many interviewees describe them as diﬀerent stations of a longer journey: I: So it is also in her blood that she [his wife] will turn on the washing machine or dishwasher in four hours or something like that? I: So it is also in her blood that she [his wife] will turn on the washing machine or dishwasher in four hours or something like that? FR11: Yes, she does that. Because that’s just a requirement of the boss [i.e. the interviewee]. I: Do you urge her or does that come from her? FR11: No, no. She already realizes that it makes sense. (...) [However,] if it doesn’t ﬁt and [she] just wants to have it done in the evening so that it is clean in the morning (...) then it must be possible to do that without the sun shining. HE4: As you said, it has always been important to us that we are aware of energy issues and we enjoyed having this possibility that we can contribute to exploiting the sun. The quote also points out limitations that are repeatedly mentioned, i.e., that synchronicity ends where it puts too much strain on comfort or interrupts necessary activities. This also refers to activities which interviewees do not consider shifting, such as cooking. I: Has that changed over time or increased? I: Has that changed over time or increased? Energy System Investments HE4: Well, it expanded into other areas and we are now driving a hybrid car Daily Behaviors: Changes in Electricity Consumption Codings around possible changes in daily behaviors regarding electricity consumption fall into four groups: (i) respondents reporting that they have reduced their electricity consumption Even more interviewees (34) elaborate on future investments. On average interviewees had two further ideas as to what such investments could be. Among these, adding battery storage (18) and/or buying an electric vehicle (15) are the dominant April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 7 Dütschke et al. Prosumers in Transition due to or following the installation of the PV, (ii) households reporting increased consumption, (iii) statements indicating no change in consumption and ﬁnally, (iv) interviewees talking about an increase in awareness without stating or knowing the inﬂuence on actual consumption. These categories are not necessarily exclusive, i.e., the sample includes eleven people each making statements that fall into more than one group. For example, households explained about using more electricity in one case and less in another. Twenty one households only gave statements from just one of these categories. Hardly any of the interviewees were able to provide precise quantitative observations comparing the development before owning the PV and since then. Some had incidental data about yearly consumption, but, major changes were mainly due to children moving out. Some households started to constantly monitor their consumption since they own the PV. Twenty reported they used less electricity now and described themselves as frugal. appliances and gadgets like garden lights, a fountain, or a solarium to get tanned. For some, the investment in PV was a response to high demand: FR3: So we were angry about our [electricity] bill. (...) We have a swimming pool inside and sand ﬁlter and that was close to 2000 Euro per year. And then we said: Well, that doesn’t have to be. We wanted to reduce that. Daily Behaviors: Behavioral Change in Other Domains Statements on further behavioral change in other domains than electricity were also given. One topic that repeatedly came up was travel behavior and more speciﬁcally ﬂying. Several interviewees were very conscious about this topic and brought it up themselves. Mechanisms Behind Behavioral Change gy y I: Yes. Do you think that over time, over the last 20 years, you have become more energy eﬃcient, or about the same? Mechanisms Behind Behavioral Change The underlying mechanisms that interviewees refer to are broad and heterogeneous. For some the investment in the PV system is already described as one step that was logical from what they had thought and experienced earlier and which also led them on to further investments and/or consistency in their daily behaviors (cf. quote from HE4 above). As outlined before, the PV investment is sometimes followed by behavioral changes, sometimes the PV is installed in response to behavioral change or high demand (e.g., maintaining a swimming pool FRA4). This will be described in more details in the remainder of this section. When the coding scheme was developed it also included the category “other” (cf. Figure 1 and Table 1), however, this subcode did not turn out to become relevant. FRA16: I think I have actually always been. R: Always? R: Always? R: Always? FRA16: Yes, I think strangely enough yes. Energy System Investments A small group made statements that they had given up ﬂying a long time ago and do not intend to do so now, or explain about very speciﬁc exemptions from this principle (e.g., a couple working for the church ﬂying to Israel for once in their life). Others claimed to make very conscious decisions regarding ﬂying. However, there was also some variety as to what “ﬂying rarely” means: HE12: And to always check where you can save more, or where you could use an energy saving lamp, or where you can replace a device with something that uses less energy. Of course always in a reasonable manner. You also need energy to produce the device, so to buy a new refrigerator for one kilowatt, that would be nonsense. FRA5-wife: Or, we also take a lot of vacations by bike. And often we go there by car. And if we deliberately go to vacation apartments, we have contact to the landlords. And, but we do take an airplane trip in winter (laughs). However, many of these statements remained very general, sometimes alluding to turning oﬀ lights or reducing standby consumption. FRA5-husband: Rarely. Every 2 years on average. But not a long distance trip, but sometimes to the Canaries or FRA5-husband: Rarely. Every 2 years on average. But not a long distance trip, but sometimes to the Canaries or FRA5-wife: Still little FRA5-husband: We want to go to Crete now. Or we went to Sicily now last year. Thirteen made statements describing perceptions that nothing has changed: FR13: I think nothing has changed. It is not that I now for example produce electricity and say, I can then waste all the more somewhere else. (I: Yes.) My behavior has not really changed because I now produce electricity myself and do not store anything. (I: Yes.) Nothing has changed. (I: Exactly.) Deﬁnitely not. FR13: I think nothing has changed. It is not that I now for example produce electricity and say, I can then waste all the more somewhere else. (I: Yes.) My behavior has not really changed because I now produce electricity myself and do not store anything. (I: Yes.) Nothing has changed. (I: Exactly.) Deﬁnitely not. In a similar vein, ambivalence about modes of travel extends to the choice of transport mode in daily life or the extent of car use. Energy System Investments Another area of resource use that is repeatedly mentioned is the use of water, with some households reporting about their installations for using rain-water or re-using e.g., water from showering for the toilet. Another topic is sparse or very conscious consumption when buying goods, reduced number of appliances, recycling, or reducing waste. Overall, interviewees give more examples of reduced or very conscious use of resources and fewer examples of high resource use levels. Of those who did speak of high resource use levels, two households reported heating over-generously. In some cases further explanations about this lack of change go in diﬀerent directions—either pointing out why reductions are not perceived to be necessary or, contrastingly, how the household just continued their always frugal lifestyle: FRA6: I say that you have a certain quality of life, and you don’t really need to restrict it because the sun makes enough energy, yes. Some (7) explain that the PV has increased their awareness: FR5: You just perceive it much more consciously. Because I get feedback on my energy consumption every day, I am much more aware of it. And I also realize what consumes energy at all and what doesn’t. Finally, a smaller group (5) outline that their demand has increased. This is mainly bound to the acquisition of additional April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 8 Prosumers in Transition Dütschke et al. 5As the real output is rarely higher due to weather and technical conditions, the actual resulting loss is only about 2-5%, cf. Wirth (2020). Individual Level: Economic and Psychological Mechanisms FRA7: I have now already told my wife that it is a completely diﬀerent feeling to shower with solar heat, with solar thermal water. (...) Not a lot of oil runs down over your head but solar heat runs over your head. Economic vs. environmental motivations are the dominant areas of discussion (often contrasted by interviewees). Some state a clear dominance of the one or the other or emphasize both, in other cases motives and how they actually inﬂuence decision making and daily behaviors seem less clear. For some, saving money is an important mechanism that drives them to synchronize their consumption patterns with the sunshine. Others that emphasize ecological motives focus on increasing awareness as an ongoing process as pointed out above. Others that emphasize ecological motives focus on increasing awareness as an ongoing process as pointed out above. Finally, in one case, frugality per se is described as the guiding principle. A diﬀerent psychological mechanism in addition to the leitmotif that emerges in several interviews is the idea of having a clear conscience due to using solar energy. In some cases this clear conscience is then used to justify behaviors that are not fully sustainable like traveling or using more energy/electricity. In some cases, the economic outcome was not clear at the point of time of decision making, with high initial investments, and was evaluated positively when people realized that later. However, some directly reject economic thinking: FRA2: When you buy a Mercedes with leather seats, do you ask if the Mercedes with leather seats is proﬁtable or if a Golf with cloth seats is proﬁtable? Does anyone ask, if one builds a dormer, if this dormer is proﬁtable or if a roof window would be suﬃcient? Socio-Technical Mechanisms The legislative framework also plays a role in shaping the behaviors of PV households. As pointed out earlier, in few cases further investments are currently held back as households still enjoy a high FIT and do not want to change the conﬁguration before it ends. In one case, the household chose a smaller PV to stay beyond a certain limit in the regulations. In these cases, the decision on the investment for the PV and further technologies depended on the aﬀordability, but not on anticipated ﬁnancial gains. Furthermore, the interviews indicate a variety of guiding motivations (“Leitmotif”) that some interviewees refer too, often repeatedly, during the interview and connecting diﬀerent behaviors and decisions following this Leitmotif. One of them is autarchy, i.e., some interviewees explain their investment in PV and also additional investments like storage by their desire to become independent of the energy system, and also of changing prices. Regulatory and sociotechnical inﬂuences can sometimes be closely interwoven. One of the peculiarities of the German legislation on renewable energy is that to prevent peak loads, PV system owners in Germany are obliged to allow grid operators to regulate their system (receiving compensation for it); alternatively, smaller systems below 30 kW can limit their feed-in to 70% of their maximum eﬀective power (EEG, 2017). Thus, there is a limit to the amount that households can feed to the grid5. This is only relevant to newer systems as this rule is relatively new. Those aﬀected by it in our sample often refer to it and some are deeply concerned to ﬁnd ways to use the relevant electricity and prevent it from being “wasted.” FRA7: I think there is a high vulnerability of our systems that we are not aware of today and the idea that I can get an emergency power supply from my own - my own energy storage and my photovoltaic system - is already a motivation to invest even more money. Another topic at the interplay between technology and household behaviors is how the actual supply with electricity is monitored, if at all. Some “monitor” the system only scarcely by checking if the light of the control unit is still on when they pass by. Socio-Technical Mechanisms WÜ1: I don’t care how much the oil costs (...) I always have mine somewhere and as I said, I can inﬂuence it myself, just very well by simply orienting myself a little towards the sun, so that’s a great thing. FRA19: So technology is—I must say—I am from the humanities. I’m really not interested in technology. Not very, huh? (...) I look at my equipment working in my basement. I can see whether the green light is on or not (laughing). In some cases the themes of sustainability and/or environmentalism are playing an important role across diﬀerent situations: In some cases the themes of sustainability and/or environmentalism are playing an important role across diﬀerent situations: HE7: So, as I said, I wanted to do something for the environment. And of course that’s one aspect, decentralized energy generation. There are many other environmental things you can do (...). Not driving a car, for example, is one. [laughs] Well, I’m also a cyclist, just by the way. Others are in the position to access real-time information about current supply and battery status (if applicable) through smartphones and similar devices and also report that they observe this closely and also use it to educate other household members. Others have established a paper- and-pencil monitoring, often on a monthly basis to detect larger deviations. Some households are proud and enjoy what they achieved in this regard: Some households are proud and enjoy what they achieved in this regard: HE12: So the feeling is that the electricity I consume here, it is also fun with such an attitude as mine to consume as little as possible. HE3: For me, I do it in my book, in which I enter my consumption and production every month, just like with water This goes as far that the enjoyment in everyday life is described in vivid pictures: April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 9 Prosumers in Transition Dütschke et al. and gas, and I have sensitized my children to the point that they are happy if they consume less themselves. The link to the energy transition shows up most via socio- structural mechanisms, and these relate most strongly in our analysis to the embeddedness of prosumers and their PV in the electricity system and its regulatory framework. DISCUSSION AND CONCLUSIONS This paper started out to take a closer look at how prosuming changes energy-related behavior of households. As a conceptual framework we drew on the literatures on rebound and spillover and as an empirical basis on 48 interviews with prosumers from four regional clusters. The focus of prosumers is motivated by the fact that prosuming households are an example how the energy transition as part of the great transformation toward sustainability manifests on the individual level. By generating electricity, households change their role in the electricity system and leave behind being passive consumers. It is against this background that we take a detailed look at how prosumers describe their interactions with the PV system, if, how and why it changes their energy behaviors. Pointing to the limitations of the paper, it seems highly likely that diﬀerent ﬁndings would emerge in diﬀerent national contexts, e.g., where ﬁnancial incentives and regulatory contexts diﬀer. In our case this is mirrored by the diﬀerences between people fully feeding into the grid and those who consume some of electricity themselves. Another limitation is that due to the semi- structured guideline there is variation between the interviews as to which topics came to the fore and which did not. Thus, it is possible that some issues or mechanisms play a role for further households but did not enter the discussion during any of the interviews. This is especially likely to apply to behavioral consequences beyond electricity use where it might have been diﬃcult for interviewers and interviewees to touch upon all possible topics. What we found in the interviews is a broad variety of behavioral responses. Further investments, already realized or planned for the future, play a prominent role. Many households have already combined their PV with further technologies or have thought a lot about how to do so in the future. To some extent this resonates with the ﬁnding of Cohen et al. (2019) as there appears to be “q-complementarity” between investment in PV and in certain other electrical goods. Q-complementarity is said to occur when the welfare gain from adopting one good is increased by the welfare gain from adopting another good and vice versa. Becoming a prosumer is for many of our respondents not an end in itself but just one step in a longer journey. Socio-Technical Mechanisms We ﬁnd diﬀerent types of eﬀects of the legislative context which give important signals to prosumers; however, the perceived inﬂuence of regulations seems sometimes higher than their actual relevance. For example, some of the households are concerned about the energy they are not allowed to feed to the grid due to a recent cut-oﬀrule. However, technical estimations indicate that the actual loss is likely to be small (Wirth, 2020). Thus, it seems likely that many households cannot draw on exact economic or technical estimations (due to lack of knowledge, interest and/or data). Rather, the regulative structure provides rules of thumb which are then translated into behavioral heuristics. DISCUSSION AND CONCLUSIONS In this vein, the investment in the PV system not only impacts future investment decisions, but was often triggered by earlier experiences. This paper adds to the literature by giving a very detailed and thereby innovative account of the behavioral consequences of adopting PV and why these emerge. Some of the ﬁndings are in line with earlier literature that pointed to increases in awareness (Palm et al., 2018). Furthermore, the broad variety of behavioral responses also ﬁts with the heterogeneity of past ﬁndings regarding the emergence and size of potential rebound or conservation/rebound eﬀects (Oberst et al., 2019; Qiu et al., 2019; Li et al., 2020): Households brought forward a variety of logics and descriptions to explain their behaviors. This in-depth account of qualitative ﬁndings can inform the design of future quantitative studies that build on our ﬁndings. Large samples would also allow for subgroups, so that the full context could be better grasped and considered via rigorous statistical analysis. The behavioral responses in daily routines are also heterogeneous within the households. Some quite clear cases of consistent environmental concern and motivation throughout emerge, and others where environmental concern and environmentally supportive behavior were gradually ampliﬁed through their experience of having PV. Furthermore, for some the PV is a kind of compensatory investment as they perceive their consumption as exceptionally high. Finally, for some the PV is also a means to justify increases in demand or luxury investments, one of the impressive examples is probably the household that added a solarium. DATA AVAILABILITY STATEMENT The mechanisms that trigger the behavioral responses are also broad and heterogeneous, and economic, psychological and socio-technical drivers were sometimes closely interwoven. At the same time, drivers do not seem to unfold homogeneously or consistently. For example, economic mechanisms act as an important driver to some, but others highlight the relevance of aﬀordability rather than economic viability). Psychological issues were mainly revealed in the form of guiding principles (leitmotif), and less as speciﬁc relationships between psychological variables like norms or attitudes. Having a “good conscience” was emphasized in some interviews and points to the relevance of moral issues, i.e., licensing or consistency behaviors. The datasets presented in this article are not readily available because the interview data is impossible to be anonymized. Requests to access the datasets should be directed to Elisabeth Dütschke, elisabeth.duetschke@isi.fraunhofer.de. Frontiers in Psychology \| www.frontiersin.org REFERENCES Galvin, R. (2013). Impediments to energy-eﬃcient ventilation of German dwellings: a case study in Aachen. Energy Build. 56, 32–40. doi: 10.1016/j.enbuild.2012.10.020 BMWi (2020). Energiepreise und Transparenz für Verbraucher: Der Strompreis. Available online at: https://www.bmwi.de/Redaktion/DE/Artikel/Energie/ strompreise-bestandteile.html. Galvin, R. (2020). I’ll follow the sun: Geo-sociotechnical constraints on prosumer households in Germany. Energy Res. Soc. Sci. 65:101455. doi: 10.1016/j.erss.2020.101455 Bundesnetzagentur (2020). EEG-Umlage: Was ist die EEG-Umlage und Wie Funktioniert Sie? Available online at: https://www.bundesnetzagentur.de/ SharedDocs/FAQs/DE/Sachgebiete/Energie/Verbraucher/Energielexikon/ EEGUmlage.html. Galvin, R., and Gubernat, A. (2016). The rebound eﬀect and Schatzki’s social theory: reassessing the socio-materiality of energy consumption via a German case study. Energy Res. Soc. Sci. 22, 183–193. doi: 10.1016/j.erss.2016.08.024 Chenavaz, R. Y., Dimitrov, S., and Figge, F. (2021). When does eco-eﬃciency rebound or backﬁre? An analytical model. Eur. J. Operational Res. 290, 687–700. doi: 10.1016/j.ejor.2020.08.039 Haan, P., de, Peters, A., and Scholz, R. W. (2007). Reducing energy consumption in road transport through hybrid vehicles: Investigation of rebound eﬀects, and possible eﬀects of tax rebates. J. Cleaner Product. 15, 1076–1084. doi: 10.1016/j.jclepro.2006.05.025 Chitnis, M., Sorrell, S., Druckman, A., Firth, S. K., and Jackson, T. (2014). Who rebounds most? Estimating direct and indirect rebound eﬀects for diﬀerent UK socioeconomic groups. Ecol. Econom. 106, 12–32. doi: 10.1016/j.ecolecon.2014.07.003 Haar, L. (2020). “Inequality and renewable electricity support in the European Union,” in Inequality and Energy: How Extremes of Wealth and Poverty in High Income Countries Aﬀect CO2 Emissions and Access to Energy, ed R. Galvin (London: Academic Press), 189–220. doi: 10.1016/B978-0-12-817674-0.00009-6 Cischinsky, H., and Diefenbach, N. (2018). Datenerhebung Wohngebäudebestand 2016: Datenerhebung zu den energetischen Merkmalen und Modernisierungsraten im deutschen und hessischen Wohngebäudebestand [Data collection residential building stock 2016: Data collection on the energy characteristics and modernisation rates of the German and Hessian residential building stock]. Darmstadt. Available online at: https://www.iwu.de/ﬁleadmin/ user_upload/dateien/gebaeudebestand/prj/Endbericht_Datenerhebung_ Wohngeb%C3%A4udebestand_2016.pdf. Hofstetter, P., Madjar, M., and Ozawa, T. (2006). Happiness and sustainable consumption: psychological and physical rebound eﬀects at work in a tool for sustainable design. Int. J. Life Cycle Assessment 11, 105–115. doi: 10.1065/lca2006.04.018 IEA and IRENA (2017). Perspectives for the Energy Transition. Investment needs for a low-carbon energy system. Cohen, J., Azarova, V., Kollmann, A., and Reichl, J. (2019). Q-complementarity in household adoption of photovoltaics and electricity-intensive goods: the case of electric vehicles. Energy Econom. 83, 567–577. doi: 10.1016/j.eneco.2019.08.004 Kelm, T., Metzger, J., Jachmann, H., Günnewig, D., Püschel, M., Schicketanz, S., et al. (2019). ETHICS STATEMENT Ethical review and approval was not required for the study on human participants in accordance with the local legislation and April 2021 \| Volume 12 \| Article 636109 10 Prosumers in Transition Dütschke et al. project EE-Rebound]. Parts of the data collection was supported by a grant from the German Federal Ministry for Economic Aﬀairs and Energy [O3SIN125, project: C/sells]. institutional requirements. The patients/participants provided their written informed consent to participate in this study. ACKNOWLEDGMENTS ED wrote a ﬁrst draft of the paper and revised it based on co-authors comments and was strongly involved in the data analysis. RG conducted one of the interview series, commented extensively on earlier versions of the paper and edited the text. IB was involved in coding the data and provided comments on the paper. All authors contributed to the article and approved the submitted version. We thank Julius Wesche and Florian Emsmann for their support as well as the student project group around Daniel Hanss at Hochschule Darmstadt for their support in conducting the interviews. In addition we are grateful to Julika Weiß for valuable input on conceptual considerations. FUNDING The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2021.636109/full#supplementary-material This work was supported by the German Federal Ministry of Education and Research [grant number FKZ 01UT1705 A-C, REFERENCES doi: 10.1016/j.gloenvcha.2014.09.004 Mayring, P. (2015). Qualitative Inhaltsanalyse. Grundlagen und Techniken [Qualitative content analysis] (Vol. 12). Beltz Verlag. Mazar, N., and Zhong, C.-B. (2010). Do green products make us better people? Psychol. Sci. 21, 494–498. doi: 10.1177/0956797610363538 van der Werﬀ, E., and Steg, L. (2018). Spillover beneﬁts: emphasizing diﬀerent beneﬁts of environmental behavior and its eﬀects on spillover. Front. Psychol. 9:2347. doi: 10.3389/fpsyg.2018.02347 Mullen, E., and Monin, B. (2016). Consistency versus licensing eﬀects of past moral behavior. Annual Rev. Psychol. 67, 363–385. doi: 10.1146/annurev-psych-010213-115120 Nash, N., Whitmarsh, L., Capstick, S., Hargreaves, T., Poortinga, W., Thomas, G., et al. (2017). Climate-relevant behavioral spillover and the potential contribution of social practice theory. Wiley Interdiscipl. Rev. Climate Change 8:e481. doi: 10.1002/wcc.481 Verfuerth, C., Jones, C. R., Gregory-Smith, D., and Oates, C. (2019). Understanding contextual spillover: using identity process theory as a lens for analyzing behavioral responses to a workplace dietary choice intervention. Front. Psychol. 10:345. doi: 10.3389/fpsyg.2019.00345 Oberst, C. A., Schmitz, H., and Madlener, R. (2019). Are prosumer households that much diﬀerent? Evidence from stated residential energy consumption in Germany. Ecol. Econom. 158, 101–115. doi: 10.1016/j.ecolecon.2018.12.014 Wirth, H. (2020). Aktuelle Fakten zur Photovoltaik in Deutschland. Freiburg. Availabe online at: https://www.ise.fraunhofer.de/content/dam/ise/de/ documents/publications/studies/aktuelle-fakten-zur-photovoltaik-in- deutschland.pdf. O’Connor, C., and Joﬀe, H. (2020). Intercoder reliability in qualitative research: debates and practical guidelines. Int. J. Qualitat. Methods 19:160940691989922. doi: 10.1177/1609406919899220 Wittenberg, I., and Matthies, E. (2016). Solar policy and practice in Germany: how do residential households with solar panels use electricity? Energy Res. Soc. Sci. 21, 199–211. doi: 10.1016/j.erss.2016.07.008 Palm, J., Eidenskog, M., and Luthander, R. (2018). Suﬃciency, change, and ﬂexibility: critically examining the energy consumption proﬁles of solar PV prosumers in Sweden. Energy Res. Soc. Sci. 39, 12–18. doi: 10.1016/j.erss.2017.10.006 Wittenberg, I., and Matthies, E. (2018). How do PV households use their PV system and how is this related to their energy use? Renew. Energy 122, 291–300. doi: 10.1016/j.renene.2018.01.091 Palm, J., and Tengvard, M. (2011). Motives for and barriers to household adoption of small-scale production of electricity: examples from Sweden. Sustainability 7, 6–15. doi: 10.1080/15487733.2011.11908061 Wörsdorfer, J. S. (2010). Consumer Needs and Their Satiation Properties as Drivers of the Rebound Eﬀect. The case of energy-eﬃcient washing machines (Papers on Economics and Evolution). Jena. Max-Planck-Institute of Economics. ZSW and UBA (2019). Erneuerbare Energien in Zahlen: Nationale und internationale Entwicklung im Jahr 2018. Berlin. Available online at: https:// www.erneuerbare-energien.de/EE/Redaktion/DE/Downloads/Berichte/ erneuerbare-energien-in-zahlen-2018.pdf?__blob=publicationFile&v=6. Peters, A., and Dütschke, E. (2016). REFERENCES Vorbereitung und Begleitung bei der Erstellung eines Erfahrungsberichts gemäß § 97 Erneuerbare-Energien-Gesetz: Teilvorhaben II c: Solare Strahlungsenergie. Abschlussbericht. Stuttgart, Hannover. Available online at: https://www.erneuerbare-energien.de/EE/Redaktion/ DE/Downloads/bmwi_de/zsv-boschundpartner-vorbereitung-begleitung-eeg. pdf?__blob=publicationFile&v=7. Dimitropoulos, A., Oueslati, W., and Sintek, C. (2018). The rebound eﬀect in road transport: a meta-analysis of empirical studies. Energy Econom. 75, 163–179. doi: 10.1016/j.eneco.2018.07.021 Dütschke, E., Frondel, M., Schleich, J., and Vance, C. (2018). Moral Licensing—Another Source of Rebound? Frontiers Energy Res. 6:393. doi: 10.3389/fenrg.2018.00038 Köhler, J., Lange, J., Lohmann, V., Lemki, V., Dütschke, E., and Hanss, D. (2019). Warum investieren Haushalte (nicht) in Photovoltaik-Anlagen? Eine empirische Exploration von Motiven, Barrieren und Erwartungen. Umweltpsychologie 23, 12–37. EEG (2017). Gesetz für den Ausbau Erneuerbarer Energien (Erneuerbare-Energien- Gesetz - EEG 2017). Available online at: https://www.gesetze-im-internet.de/ eeg_2014/. Korcaj, L., Hahnel, U. J., and Spada, H. (2015). Intentions to adopt photovoltaic systems depend on homeowners’ expected personal gains and behavior of peers. Renew. Energy 75, 407–415. doi: 10.1016/j.renene.2014. 10.007 Elf, P., Gatersleben, B., and Christie, I. (2018). Facilitating positive spillover eﬀects: new insights from a mixed-methods approach exploring factors enabling people to live more sustainable lifestyles. Front. Psychol. 9:2699. doi: 10.3389/fpsyg.2018.02699 Li, X., Lim, M. K., Du, N.i, Zhong, B., Xiao, Z., and Hao, H. (2020). Sustainability or continuous damage: a behavior study of prosumers’ electricity consumption after installing household distributed energy resources. J. Cleaner Product. 264:121471. doi: 10.1016/j.jclepro.2020.121471 Galizzi, M. M., and Whitmarsh, L. (2019). How to measure behavioral spillovers: a methodological review and checklist. Front. Psychol. 10:342. doi: 10.3389/fpsyg.2019.00342 April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 11 Dütschke et al. Prosumers in Transition Luthander, R., Widén, J., Nilsson, D., and Palm, J. (2015). Photovoltaic self-consumption in buildings: a review. Appl. Energy 142, 80–94. doi: 10.1016/j.apenergy.2014.12.028 Sorrell, S. (2015). Reducing energy demand: a review of issues, challenges and approaches. Renew. Sustain. Energy Rev. 47, 74–82. doi: 10.1016/j.rser.2015.03.002 Matiaske, W., Menges, R., and Spiess, M. (2012). Modifying the rebound: It depends! Explaining mobility behavior on the basis of the German socio- economic panel. Energy Policy 41, 29–35. doi: 10.1016/j.enpol.2010.11.044 Sunikka-Blank, M., and Galvin, R. (2012). Introducing the prebound eﬀect: the gap between performance and actual energy consumption. Build. Res. Informat. 40, 260–273. doi: 10.1080/09613218.2012.690952 Truelove, H. B., Carrico, A. R., Weber, E. U., Raimi, K. T., and Vandenbergh, M. P. (2014). Positive and negative spillover of pro-environmental behavior: an integrative review and theoretical framework. Global Environm. Change 29, 127–138. Frontiers in Psychology \| www.frontiersin.org REFERENCES “Exploring rebound eﬀects from a psychological perspective,” in Rethinking Climate and Energy Policies: New Perspectives on the Rebound Phenomenon, eds T. Santarius, H. J. Walnum, and C. Aall (Cham: Springer), 89–105. doi: 10.1007/978-3-319-38807-6_6 Qiu, Y., Kahn, M. E., and Xing, B. (2019). Quantifying the rebound eﬀects of residential solar panel adoption. J. Environ. Econom. Manage. 96, 310–341. doi: 10.1016/j.jeem.2019.06.003 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. 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. Saunders, H. D. (2008). Fuel conserving (and using) production functions. Energy Econom. 30, 2184–2235. doi: 10.1016/j.eneco.2007.11.006 Seebauer, S. (2018). The psychology of rebound eﬀects: Explaining energy eﬃciency rebound behaviours with electric vehicles and building insulation in Austria. Energy Res. Soc. Sci. 46, 311–320. doi: 10.1016/j.erss.2018. 08.006 Copyright © 2021 Dütschke, Galvin and Brunzema. 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. Sonnberger, M., and Gross, M. (2018). Rebound eﬀects in practice: an invitation to consider rebound from a practice theory perspective. Ecol. Econom. 154, 14–21. doi: 10.1016/j.ecolecon.2018.07.013 April 2021 \| Volume 12 \| Article 636109 Frontiers in Psychology \| www.frontiersin.org 12
https://openalex.org/W2989684857	https://www.nature.com/articles/s41598-019-53943-3.pdf	English	null	pH-depended protein shell dis- and reassembly of ferritin nanoparticles revealed by atomic force microscopy	Scientific reports	2,019	cc-by	7,110	pH-depended protein shell dis- and reassembly of ferritin nanoparticles revealed by atomic force microscopy Lukas Stühn1,2, Julia Auernhammer1,2 & Christian Dietz1 Ferritin, a protein that is present in the human body for a controlled iron storage and release, consists of a ferrihydrite core and a protein shell. Apoferritin, the empty shell of ferritin, can be modified to carry tailored properties exploitable for targeted and direct drug delivery. This protein shell has the ability to dis- and reassemble depending on the pH value of the liquid environment and can thus be filled with the desired substance. Here we observed the dis- and reassembly process of the protein shell of ferritin and apoferritin in situ and in real space using atomic force microscopy. Ferritin and apoferritin nanoparticles adsorbed on a mica substrate exhibited a change in their size by varying the pH value of the surrounding medium. Lowering the pH value of the solution led to a decrease in size of the nanoparticles whereas a successive increase of the pH value increased the particle size again. The pH dependent change in size could be related to the dis- and reassembling of the protein shell of ferritin and apoferritin. Supplementary imaging by bimodal magnetic force microscopy of ferritin molecules accomplished in air revealed a polygonal shape of the core and a three-fold symmetry of the protein shell providing valuable information about the substructure of the nanoparticles. Ferritin is an iron-storage core-shell-protein present in almost all living organisms. Ferritin has an outer diam- eter of approximately 12 nm containing iron in the form of ferrihydrite in its approximately 8 nm large core. In the body of animals, ferritin is utilized to store, transport and release iron in a controlled manner for the oxy- gen transport, indispensable to life, and can be exploited for different medical1,2 and electronic applications3. Depending on the environment and the ferritin subtype, ferritin features a high thermal stability up to 77 °C4. As other proteins5, apoferritin, the empty protein shell of ferritin, can be used to transport various substances, such as drugs for medical applications6–8. The empty shell of ferritin can be used as a platform for the synthesis of nanoscale materials9. This protein shell is known to dis- and reassemble depending on the pH value of the sur- rounding medium10: the protein shell is stable in acidic environments down to pH values of 3.4, whereas the core is stable down to pH 2.111. Below pH 3, the protein shell starts to disassemble. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports 1Physics of Surfaces, Institute of Materials Science, Technische Universität Darmstadt, Alarich-Weiss-Str. 2, 64287, Darmstadt, Germany. 2These authors contributed equally: Lukas Stühn and Julia Auernhammer email: dietz@pos. tu-darmstadt.de Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 Results Electrosteric interactions. The isoelectric point is the pH value of a liquid solution, at which the surface of an immersed material or molecule does not carry an electrical net charge. For proteins, this is also the pH value for best solubility. In Fig. 1a the sign of the surface net charge of silicon dioxide17, muscovite mica18 and apoferri- tin19 depending on the pH-value of the medium is depicted. The resulting charge interactions between the AFM tip, nanoparticle and SiO2 or mica for solutions in different pH-values are summarized in Fig. 1b,c, respectively. p p 2f p g p y As can be deduced from Fig. 1 attractive forces between the ferritin protein shell and the silicon substrate are only present in solution with pH values between approximately 2.5–4.8. For a mica substrate, attractive forces between ferritin nanoparticles and the substrate are also present for more acidic solutions. It can be concluded that a movement of the ferritin particles on a mica substrate by moderate tip-ferritin interactions is strongly inhibited in the above-mentioned pH ranges. The forces involved in the tip-ferritin interactions are net repulsive at pH values above 4.8 and below 2.5 and net attractive between 2.5 < pH < 4.8. The attractive interaction between the tip and the protein shell causes the z-piezo to move downward while scanning across an (apo)ferritin mol- ecule whereas the actual size of the molecule forces the z-piezo to move up simply because of the topographical feedback that keeps the applied setpoint force constant. Thus, it is reasonable to believe that the height and hence the size of the ferritin nanoparticles at pH values in the range of 2.5 < pH < 4.8 is slightly underestimated by the apparent topography. If we detect a decrease in particle size in this range while lowering the pH value, then this decrease is an upper limit and might be even stronger in reality, compared to height changing values measured at pH < 2.5 or > 4.8 caused by the variation of the electrostatic interactions. This effect has to be taken into account when evaluating the development of the size of nanoparticles when changing the pH value of the solution. Dis- and reassembling of ferritin nanoparticles. Figure 2 shows a series of topography images taken at the same spot of a mica substrate covered with ferritin nanoparticles. Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports tificreports/ Figure 1. Surface charges of SiO2, mica and apoferritin nanoparticles at different pH values. (a) Net charge of the different surfaces. (b) Electrostatic interactions between an AFM tip, apoferritin and a SiO2 substrate or (c) muscovite mica in solutions of different pH values (pH 2, 4, 6 and 9), respectively. Figure 1. Surface charges of SiO2, mica and apoferritin nanoparticles at different pH values. (a) Net charge of the different surfaces. (b) Electrostatic interactions between an AFM tip, apoferritin and a SiO2 substrate or (c) muscovite mica in solutions of different pH values (pH 2, 4, 6 and 9), respectively. tip-shell forces and the magnetic tip-core interactions with bimodal magnetic force microscopy14–16 allowed us to resolve substructures of the molecules, such as the three-fold symmetry of the protein shell and a polygonal shape of the core indicating its crystalline condensed state. pH-depended protein shell dis- and reassembly of ferritin nanoparticles revealed by atomic force microscopy However, this process is reversible by the subsequent increase to higher pH values turning apoferritin to an almost ideal candidate as carrier for drug delivery, i.e. loading with a substance and subsequently release it. This loading occurs in a random fashion: the substance to be enclosed inside the protein shell is step-wise surrounded by the reassembling shell. It is possible to adsorb ferritin onto a substrate12 to study dis- and reassembling of the protein shell using sedimentation-velocity measurements10 and small angle x-ray scattering11,13. However, these studies do not provide a direct observation of the dis- and reassembling process and hence direct evidence, instead the conclusions had to be drawn based on subsequent transformation of the data. In particular for scattering methods, the observed data reflects the protein state in the reciprocal space and must be subsequently transferred to real space by means of simulations based on specific assumptions. Here, we report on the direct in situ observation of the pH dependent dis- and reassembly process of ferritin and apoferritin nanoparticles using atomic force microscopy (AFM) based on images that pro- vide the size of individual particles in real space in liquid solutions of various pH values. Additionally, we examine the electrosteric interactions between the tip of an AFM cantilever, the ferritin nanoparticle and the substrate at different pH-values to validate the size interpretation of the molecules. Exploiting the coexistence of mechanical 1Physics of Surfaces, Institute of Materials Science, Technische Universität Darmstadt, Alarich-Weiss-Str. 2, 64287, Darmstadt, Germany. 2These authors contributed equally: Lukas Stühn and Julia Auernhammer email: dietz@pos. tu-darmstadt.de Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 Results During the image acquisition, the pH value was varied in the following order: pH 6 (a), pH 2 (b), pH 4 (c), pH 6 (d), pH 8 (e), pH 2 (f), pH 4 (g), pH 6 (h). After each image acquisition, the adjusted pH value within the fluid cell was confirmed at the outlet of the cell and a new measurement at the next pH value was accomplished. In Figure 2 a homogenous distribution of ferritin nanoparticles is apparent at the surface of the substrate. Individual particles could be identified, however their distribution in size was found to be polydisperse and the particles appeared slightly inhomogeneous in their shapes. In addition, laterally small and low objects in height assumed to be fragments of the protein shell were already found in the initial solution at rare locations. Particles that disappeared in the next step are enclosed by squares. We interpret the disappearance as complete dissem- bling of the protein shell accompanied by a release of the ferrihydrite core. Circles mark particles that are present throughout the entire measurement, yet varying in size (see number given above the particles in Fig. 2). Arrows indicate particles that were not apparent in the previous pH step and started to assemble. The maximum size of the particles changed during the course of the measurement series. The initial change from pH 6 (Fig. 2a) to pH 2 (Fig. 2b) caused a decrease in size of all nanoparticles marked with circles. The topmost particle among them changed its height from 9 nm to 6 nm, corresponding to the size of the ferrihydrite core. The particles marked Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ with the squares are no longer visible in Fig. 2b. In the further course of the experiment when the pH value was Figure 2. AFM topography images of ferritin nanoparticles adsorbed on a mica substrate measured in deionized water with adjusted pH value in the following order: (a) pH 6 (b) pH 2 (c) pH 4 (d) pH 6 (e) pH 8 (f) pH 2 (g) pH 4 (h) pH 6. Particles that vanished in the next step are enclosed by a square. Particles that assembled and did not exist in the previous step are marked by an arrow. Particles present through the whole experiment but vary in size are marked with a circle. Their sizes in nm are designated by numbers. Results Scale bar is 100 nm. Figure 2. AFM topography images of ferritin nanoparticles adsorbed on a mica substrate measured in deionized water with adjusted pH value in the following order: (a) pH 6 (b) pH 2 (c) pH 4 (d) pH 6 (e) pH 8 (f) pH 2 (g) pH 4 (h) pH 6. Particles that vanished in the next step are enclosed by a square. Particles that assembled and did not exist in the previous step are marked by an arrow. Particles present through the whole experiment but vary in size are marked with a circle. Their sizes in nm are designated by numbers. Scale bar is 100 nm. with the squares are no longer visible in Fig. 2b. In the further course of the experiment when the pH value was step-by-step increased until pH 8 was adjusted (Fig. 2e), the size of the particles marked with the circles increased again, however they did not reach the initial value exactly, as measured in Fig. 2a. During the change of the pH with the squares are no longer visible in Fig. 2b. In the further course of the experiment when the pH value was step-by-step increased until pH 8 was adjusted (Fig. 2e), the size of the particles marked with the circles increased again, however they did not reach the initial value exactly, as measured in Fig. 2a. During the change of the pH with the squares are no longer visible in Fig. 2b. In the further course of the experiment when the pH value was step-by-step increased until pH 8 was adjusted (Fig. 2e), the size of the particles marked with the circles increased again, however they did not reach the initial value exactly, as measured in Fig. 2a. During the change of the pH Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ Figure 3. AFM topography images of apoferritin nanoparticles adsorbed on a mica substrate in solutions of different pH values: (a) pH 6 (b) pH 2 (c) pH 3 (d) pH 4 (e) pH 6. Scale bar is 100 nm. (f) Development of the average height (crosses) of the apoferritin nanoparticles adsorbed on a mica substrate in solution at different pH values analysed from the topography data. Error bars correspond to the standard deviation. Histograms show the height deviation at each pH value step. Figure 3. Results AFM topography images of apoferritin nanoparticles adsorbed on a mica substrate in solutions of different pH values: (a) pH 6 (b) pH 2 (c) pH 3 (d) pH 4 (e) pH 6. Scale bar is 100 nm. (f) Development of the average height (crosses) of the apoferritin nanoparticles adsorbed on a mica substrate in solution at different pH values analysed from the topography data. Error bars correspond to the standard deviation. Histograms show the height deviation at each pH value step. value from pH 2 (Fig. 2b) to pH 8 (Fig. 2e), several new particles have formed as indicated by arrows. Similar phenomena were approved during a repeated pH change from pH 2 (Fig. 2e) to pH 6 (Fig. 2h). Summarizing the observations of the experiment: Decreasing the pH value of the solution causes a shrinkage of the nanoparticles in size whereas the subsequent increase of the pH value of the solution results in a growth of the particles and the formation of new particles. Dis- and reassembling of apoferritin nanoparticles. To study the pH dependent size behaviour of apoferritin nanoparticles, topographical analysis by AFM in solutions of different pH values was performed in the same manner as for ferritin described in the previous section. The Apoferritin nanoparticles show a higher density on the substrate and a higher initial height deviation, than compared to the ferritin samples. The pH value of the solution was adjusted in the following order: pH 6, pH 2, pH 3, pH 4 und pH 6. Figure 3 shows the corre- sponding topography images of the measurement series. In contrast to the series accomplished on ferritin, the apoferritin sample exhibited a higher particle density on the surface for the prepared samples. The pattern of the nanoparticles and hence the particle distribution on the substrate partially changed with every step of varying pH value. This becomes particularly evident when the pH value of the solution was changed from pH 2 (Fig. 3b) to pH 3 (Fig. 3c). In addition to the increased number of nanoparticles on the surface compared to the experiment performed on the ferritin sample, the rearrangement of the particles was another reason why a statistical evalu- ation was preferred in that case. Hence, it was difficult to follow the presence of a particular nanoparticle during the course of the experiment. Results Scale bar is 10 nm. (e) 3d topography with color-coded first eigenmode phase, (f) 3d topography with color-coded second eigenmode phase. using atomic force microscopy. First high resolution studies on the protein shell of ferritin were performed by Ohnishi using contact mode AFM20. However, the core and shell substructure of ferritin molecules could not be resolved by this technique. In the following we show that by using bimodal magnetic force microscopy15,16, we can separate magnetic from mechanical cantilever responses of the mixed tip-sample interactions. This allowed us to image substructure features of the shell and the core of a single ferritin nanoparticle in great detail. g g p g Figure 4 shows the five available observables of the bimodal MFM measurement, i.e. first (a) and sec- ond (b) eigenmode amplitude images as well as the first (c) and second (d) eigenmode phase images and a three-dimensional representation of the topography rendered by the colour code of the first and second eigen- mode phase (Fig. 4e,f). Because of the use of a magnetic cantilever with a platinum/iridium coated tip exhibiting a larger tip radius as compared to standard cantilevers, the resulting diameter of the nanoparticle of almost 50 nm is significantly larger than its real diameter. The cores of the ferritin particles are superparamagnetic and can be polarized by the presence of an external magnetic field and/or the field originated by the oscillating magnetic tip. In the first eigenmode amplitude (Fig. 4a) and phase (Fig. 4c) images a ring around a centre area of constant amplitude and phase value, respectively, is apparent. In particular in the phase image of the first eigenmode it is obvious that this centre area is separated from the outer ring through linear edges intersecting each other by vari- ous angles forming a polygonal structure. Both second eigenmode observables (amplitude Fig. 4b, phase Fig. 4d) show a threefold symmetry in the centre area of the nanoparticle.h y y p Further analysis of the polygonal structure is shown in Fig. 5a. The angles between the intersecting lines were measured to be α = 126°,β = 150°, γ = 142°, δ = 105°, ε = 126°, ζ = 114° and η = 138°. Based on the variation of existing angles we conclude that the ferrihydrite core condenses in a complex morphologic state. Results p A general change in size of the apoferritin nanoparticles by reducing the pH value is clearly apparent when comparing Fig. 3a,b: All apoferritin nanoparticles present exhibited a smaller size in topography in Fig. 3b com- pared to Fig. 3a, derivable from the generally darker colour value of the particles at the lower pH value. In the sub- sequent increase of the pH value from pH 3 to pH 6 (Fig. 3c–e), the initial size of the nanoparticles was restored. A statistical analysis of the size distribution of apoferritin at the different pH values based on the topography data is shown as histograms in Fig. 3f. g g As for ferritin nanoparticles, it can be deduced that with decreasing pH value, the average height of the nano- particles decreases, from approximately 11 nm at the first step of the measurement (pH 6) to approximately 8 nm in the second step (pH 2). The standard deviations of the size distribution displayed in Fig. 3f decreased too, from 10 nm at pH 6, to 7 nm at pH 2. The subsequent increase to pH 3 caused nearly no increase of the average height but an increase of the standard deviation from 4 nm to 8 nm. Further increase of the pH value led to an increase of the apoferritin nanoparticles in size to a value of 10 nm and standard deviation back to the initial values found at the beginning of the experiment at pH 6. We interpret the pH dependent decrease and increase of the ferritin and apoferritin molecules in size as pH-triggered dis- and resembling process of the protein shell. High resolution bimodal magnetic AFM measurements of ferritin. As demonstrated in the previous section, the pH dependent dis- and reassembling of ferritin and apoferritin particles can be directly visualized Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ Figure 4. High-resolution bimodal MFM image of a single ferritin nanoparticle on a silicon substrate performed in air. (a) first and (b) second eigenmode amplitude, as well as (c) first and (d) second eigenmode phase images. Scale bar is 10 nm. (e) 3d topography with color-coded first eigenmode phase, (f) 3d topography with color-coded second eigenmode phase. Figure 4. High-resolution bimodal MFM image of a single ferritin nanoparticle on a silicon substrate performed in air. (a) first and (b) second eigenmode amplitude, as well as (c) first and (d) second eigenmode phase images. Results The core can consist of different iron oxide phases (ferrihydrite, hematite, magnetite)21, which presents itself in our study as a (common) polygonal geometric structure. The analysis of the symmetry visible in the second eigenmode ampli- tude image was conducted in Fig. 5b. The bimodal excitation of the cantilever and the observation of the ampli- tude and phase shifts of the first two flexural eigenmodes in a single-pass technique allows for the simultaneous Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ www.nature.com/scientificreports e.com/scientificreports/ Figure 5. Analysis of the substructure features visible by bimodal MFM. (a) High-resolution first eigenmode phase image of a single ferritin nanoparticle on a silicon substrate performed in air. The separation lines between the outer ring and the inner central area as well as the corresponding angles at the intersections are highlighted. (b) Second eigenmode amplitude image exhibiting a threefold symmetry in the inner centre area. (c) Three-dimensional crystal structure of ferritin showing a threefold symmetry, calculated and created with Jmol32 (Jmol: an open-source Java viewer for chemical structures in 3D. http://www.jmol.org/, version 14) using data provided in refs. 22,23. Figure 5. Analysis of the substructure features visible by bimodal MFM. (a) High-resolution first eigenmode phase image of a single ferritin nanoparticle on a silicon substrate performed in air. The separation lines between the outer ring and the inner central area as well as the corresponding angles at the intersections are highlighted. (b) Second eigenmode amplitude image exhibiting a threefold symmetry in the inner centre area. (c) Three-dimensional crystal structure of ferritin showing a threefold symmetry, calculated and created with Jmol32 (Jmol: an open-source Java viewer for chemical structures in 3D. http://www.jmol.org/, version 14) using data provided in refs. 22,23. detection of mechanical and magnetic tip sample interactions at the nanometre scale20. By the choice of the experimental parameters (oscillation amplitudes A1, A2, amplitude ratio A1/A2, setpoint amplitude Asp/A1) it is determined whether the first or the second eigenmode channels are mainly dominated by magnetic or mechan- ical forces. Hence, the polygonal structure visible in Fig. 5a and the threefold symmetry in Fig. 5b is the result of a magnetomechanical contrast. This measured symmetry is in good agreement with the calculated structure visible in Fig. 5c. Figure 5c shows the symmetry of the protein shell of ferritin, calculated from X-ray data22,23. A threefold symmetry of the protein shell exposed to the topside, i.e. Conclusions In conclusion, the pH dependent dis- and reassembly process of ferritin and apoferritin proteins can be directly visualized and studied in detail by an in situ experiment performed in an electrochemical cell of an atomic force microscope. The more acidic the environment of the molecules became, the lower was their measured size on average. The height of the molecules increased again when changing the solution to a neutral or basic milieu. A consideration of the surface charges of the AFM tip, protein shell and substrates allowed us to exclude artefacts in the observed trends arising from different electrostatic interactions. However, the attractive interaction in the range 2.5 < pH < 4.8 between tip and protein shell led to a generally reduced apparent height of ferritin and apoferritin molecules deduced from the topography images due to feedback restrictions. Thus, it is reasonable to believe that the real size of the molecules is slightly larger in that range. Strikingly, we observed an emergence/ appearance of protein clusters at locations where no particle was present before. As the pH value was increased, new (apo)ferritin particles formed most likely developing from surrounding protein fragments, which remained in the liquid from previously disassembled protein shells. Due to the lack of a constant liquid flow in the electro- chemistry cell, there is no reason to assume that the dissembled protein shell fragments were flushed away from the vicinity of the sample surface. Several authors11,26 assume that the higher-order structure of the protein shell is disassembled upon pH value decrease of the environment of the particles. However, the individual protein chains should remain intact. In the current study, conclusions about the structure of individual proteins cannot be drawn. Using atomic force microscopic measurements and further improving the resolution, it might pave the way to visualise the behaviour of individual proteins during the formation and degradation of the protein shell. Th bi d l i i f i il ll d i l l h h f f i i l The bimodal excitation of a magnetic cantilever allowed us to simultaneously map the shape of a ferritin mole- cule, its mechanical as well as magnetic properties. In our experiments, these two properties could be clearly sepa- rated in the amplitude and phase channels of the different eigenmodes. Results opposite to the substrate, is clearly visible in the image taken by MFM (Fig. 5b). Such symmetries arise at channels that form the protein shell24,25 based on the alignment of the protein molecules around the shell. In this study, we chose a relatively large first eigenmode amplitude enabling the first mode to be sensitive to magnetic interactions caused by the magnetic coating of the tip and the ferrihydrite core and a relatively high second eigenmode amplitude (compared to ref. 20) amplifying the mechanical contrast between the tip and the ferritin shell. For the measurement parameters used in this study such as free amplitude of the first eigenmode (A1 = 23 nm) and relative setpoint (ASP = 16 nm) as well as the rela- tively large free amplitude of the second eigenmode (A2 = 16 nm) the observables of the first eigenmode are more prone for magnetic forces and the observables of the second mode are dominated by mechanical forces, which is in contrast to previous studies19,20 where substantially different parameters were applied, in particular a small second eigenmode amplitude. Thus, the first eigenmode (Figs. 4a,c and 5a) reveals the magnetic image of the particles generated by the ferritin core, the second eigenmode (Figs. 4b,d and 5b) the mechanical contrast given by the structure of the protein shell. A detailed study of the origin of the magnetomechanical contrast, however, is beyond the scope of this paper. Strikingly, as a result of this combination, a three-fold symmetry on the shell and the polycrystalline state of the core of this biomolecule became apparent in a single measurement. Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 Experimental Section p Sample preparation. To prepare ferritin samples, a two-step dilution of the stock solution was conducted. 10 µL of the purchased ferritin solution (10 mg/mL in 0.15 M NaCl) cationized from horse spleen (MFCD00081593, Merck KGaA, Darmstadt, Germany) were diluted into 1 mL deionized water (Milli-Q, Merck KGaA, Darmstadt, Germany) and homogenized for 5 min at 800 rpm at 22 °C (ThermoMixer, Eppendorf, Wessling-Berzdorf, Germany). 10 µL of the prepared solution were again diluted into 1 mL deionized water and homogenized repeat- ing the procedure. 200 µL of the final solution were drop casted onto a freshly cleaved muscovite mica (50-D-10, NanoAndMore GmbH, Wetzlar, Germany) substrate or a 1 × 1 cm2 large piece of a Si(100) wafer (LamResearch Corporation, Fremont, CA). Size measurements with AFM were initiated 5 minutes after the samples preparation to provide sufficient time for the nanoparticles to settle on the substrate. For the apoferritin sample preparation, 10 µL of the apoferritin stock solution (MFC00081365, from equine spleen, 50% glycerol + 0.075 M NaCl), purchased from Merck KGaA as well, were drop casted onto the substrates (mica or Si(100)) and 200 µl deionized water were dropped above for further dilution. Size measurements by AFM were initiated after 5 minutes settling time. To observe the dis- and reassembly process of the molecules, particular attention was paid to avoid exsiccation of the sample after preparation to keep the tertiary structure of the protein shell intact. Size analysis by AFM. For the size analysis of the nanoparticles, PeakForce Tapping AFM measurements were carried out in a closed, heated electrochemistry cell (DIM EC cell, 932–012–300, Bruker, Santa Barbara, CA) mounted in an ICON AFM from the same manufacturer. Topographical images were taken using an oscillation amplitude of 300 nm at a driving frequency of 500 Hz, far less than resonance frequency of the cantilever in the solu- tion to avoid resonance effects. The maximum force exerted to the sample was adjusted to 500 pN. The force applied is rather high and it can affect the analysis based on height measurements. Since the applied force was kept constant for all images such an effect applies to the whole imaging series and there is no change for the outcome of the experi- ment. Experimental Section Based on our observations we exclude particle migration or particle movement during the pH change because no traces that are typical for particle movement28 could be observed with the low force we adjusted. A structural change of e.g. the protein shell that would alter the observed height by changing the pH value seems unlikely to be the major origin of the size increase/decrease because of the relatively large size change we observe. The cantilevers used were SNL-A (Bruker AXS, Santa Barbara, CA), exhibiting a typical force constant of 0.35 N/m as determined by the thermal noise method29 and a nominal resonance frequency of 65 kHz in air.tilh Topography images were tilt and drift corrected by first order flattening. The temperature was set to 37 °C through the electrochemistry cell. The pH value of the measurement solution was adjusted to the desired values either by adding hydrochloric acid (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) or ammonia (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) to deionized water before filling the solution into a syringe. No additional ferritin or apoferritin nanoparticles are in the syringe-solutions. Using pre-heated (37 °C) syringes filled with the measuring solution of different pH values and a syringe pump Aladdin-1000 (World Precision Instruments, Sarasota, Florida), the pH value inside the electrochemistry cell was changed after each measuring step through connection tubes attached to the inlet of the cell. The pH stability of the solution in the cell during the measure- ment was confirmed at the outlet of the cell. The samples were kept in a liquid environment at all times. Due to the low number of particles, the size of the ferritin could be measured by drawing a cross-sectional profile through the particle and extracting the average height over a 3-pixel line with respect to the substrate level. The size of the apoferritin nanoparticles was analysed by an automated particle analysis tool30. Height profiles of individual par- ticles can be found in the supplementary material (Figs. S1 and S3) and a trend of the change in height (Fig. S2). Substructure detection by bimodal MFM. To study the substructure of ferritin nanoparticles bimodal force microscopy31 was combined with magnetic force microscopy in a single-pass technique. Using a mag- netic tip, it is feasible to separate long-range magnetic from short-range mechanical tip-sample interactions in air19 and in liquid environments20. Conclusions The results suggest that the contrast found in the phase image of the first eigenmode mainly results from the magnetic interactions between the ferrihydrite www.nature.com/scientificreports/ www.nature.com/scientificreports/ core of the ferritin particle and the tip providing a defined polygonal structure as a result of the polycrystalline condensed state. Additionally, the contrast in the amplitude image deduced from the second eigenmode revealed a three-fold symmetry in the center of the ferritin nanoparticle arising from the mechanical interaction of the tip with the shell. This symmetry is in very good agreement with the calculated ferritin structure22 and the appear- ance demonstrates that the protein capsule was mainly intact. p p y In future, bimodal magnetic force microscopy in combination with a liquid environment of changeable pH value would allow for the observation of the dis- and reassembling process in great structural detail due to the simultaneous interplay of magnetic and mechanical forces between tip and sample. The methodology could open up new pathways to study drug release for targeted drug delivery in cancer research. However, magnetic force microscopy performed in liquids20,27 has been a great challenge so far due to the low-Q environment and unstable magnetic coatings, in particular on tiny magnetic structures such as present on nanoparticles. Acknowledgementsh g The authors thank the Deutsche Forschungsgemeinschaft (Sachbeihilfe DI 2176/2-1) for financial support. We acknowledge support by the German Research Foundation and the Open Access Publishing Fund of Technische Universität Darmstadt. g The authors thank the Deutsche Forschungsgemeinschaft (Sachbeihilfe DI 2176/2-1) for financial support. We acknowledge support by the German Research Foundation and the Open Access Publishing Fund of Technische g pp y Universität Darmstadt. References The Protein Data Bank. Nucleic Acids Res 28, 235–242, https://doi.org/10.1093/nar/28.1.235 (2000).h h 24. Harrison, P. M. & Arosio, P. The ferritins: molecular properties, iron storage function and cellular regulation. Biochim Biophy 1275, 161–203, https://doi.org/10.1016/0005-2728(96)00022-9 (1996). p g 5. Takahashi, T. & Kuyucak, S. Functional properties of threefold and fourfold channels in ferritin deduced from electrostatic calculations. Biophys J 84, 2256–2263, https://doi.org/10.1016/S0006-3495(03)75031-0 (2003). 6. Carmona, F. et al. Study of ferritin self-assembly and heteropolymer formation by the use of Fluorescence Resonance Energy Transfer (FRET) technology. Biochim. Biophys. Acta-Gen. Subj. 1861, 522–532, https://doi.org/10.1016/j.bbagen.2016.12.011 (2017) 27. Ares, P., Jaafar, M., Gil, A., Gomez-Herrero, J. & Asenjo, A. Magnetic Force Microscopy in Liquids. Small 11, 4731–4736, https:/ org/10.1002/smll.201500874 (2015). g 28. Schiwek, S., Heim, L.-O., Stark, R. W. & Dietz, C. Manipulation of polystyrene nanoparticles on a silicon wafer in the peak force tapping mode in water: pH-dependent friction and adhesion force. J. Appl. Phys. 117, 104303, https://doi.org/10.1063/1.4914354 (2015). 29. Butt, H. J. & Jaschke, M. Calculation of Thermal Noise in Atomic-Force Microscopy. Nanotechnology 6, 1–7, https:/ org/10.1088/0957-4484/6/1/001 (1995).t g 30. Necas, D. & Klapetek, P. Gwyddion: an open-source software for SPM data analysis. Cent Eur J Phys 10, 181–188, https:/ org/10.2478/s11534-011-0096-2 (2012). g 1. Rodriguez, T. R. & Garcia, R. Compositional mapping of surfaces in atomic force microscopy by excitation of the second norma mode of the microcantilever. Applied Physics Letters 84, 449–451, https://doi.org/10.1063/1.1642273 (2004). pp y p g 32. Jmol: an open-source Java viewer for chemical structures in 3D, http://www.jmol.org/. Experimental Section To this end, we measured a ferritin sample in air using a Cypher S (Oxford Instruments, Asylum Research, Santa Barbara, CA) AFM with a magnetic Multi75-G cantilever (BudgetSensors, Sofia, Bulgaria) exhibiting a nominal force constant of 3 N/m and a nominal resonance frequency of 75 kHz of the fundamental flexural eigenmode in air. First and second cantilever eigenmodes were simultaneously excited by a photothermal excitation system and the respective amplitude and phase shifts of both modes were mapped. We used a scan speed of 1 Hz, free oscillation amplitudes of A1 = 23 nm and A2 = 16 nm and a setpoint amplitude ratio of Asp/A1 = 16 nm/23 nm = 70% to obtain images. Received: 29 August 2019; Accepted: 7 November 2019; Published: xx xx xxxx Received: 29 August 2019; Accepted: 7 November 2019; Published: xx xx xxxx Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ References 1. Wilhelm, S. et al. Analysis of nanoparticle delivery to tumours. Nat Rev Mater 1, https://doi.org/10.1038/natrevmats.2016.14 (2016). l l f f h ( ) 1. Wilhelm, S. et al. Analysis of nanoparticle delivery to tumours. Nat Rev Mater 1, https://doi.org/10.1038/natrevmats.2016.14 (2016). 1. Wilhelm, S. et al. Analysis of nanoparticle delivery to tumours. Nat Rev Mater 1, https://doi.org/10.1038/natrevmats.2016. 2. Peer, D. et al. Nanocarriers as an emerging platform for cancer therapy. Nat Nano 2, 751–760 (2007). y p y p g 2. Peer, D. et al. Nanocarriers as an emerging platform for cancer therapy. Nat Nano 2, 751–760 (2007). g g y 3. Jutz, G., van Rijn, P., Santos Miranda, B. & Böker, A. Ferritin: A Versatile Building Block for Bionanotechnology. Chemical Review 115, 1653–1701, https://doi.org/10.1021/cr400011b (2015).h p g 4. Stefanini, S. et al. Thermal stability of horse spleen apoferritin and human recombinant H apoferritin. Archives of Biochemistry and Biophysics 325, 58–64, https://doi.org/10.1006/abbi.1996.0007 (1996). p y p g ( ) 5. MaHam, A., Tang, Z. W., Wu, H., Wang, J. & Lin, Y. H. Protein-Based Nanomedicine Platforms for Drug Delivery. Small 5 1706–1721, https://doi.org/10.1002/smll.200801602 (2009).fi p g 6. Truffi, M. et al. Ferritin nanocages: A biological platform for drug delivery, imaging and theranostics in cancer. Pharmacol Res 107 57–65, https://doi.org/10.1016/j.phrs.2016.03.002 (2016). p g j p ( ) 7. Zhen, Z. P., Tang, W., Todd, T. & Xie, J. Ferritins as nanoplatforms for imaging and drug delivery. Expert Opin Drug Del 11 1913–1922, https://doi.org/10.1517/17425247.2014.941354 (2014). p g 8. Bhaskar, S. & Lim, S. Engineering protein nanocages as carriers for biomedical applications. NPG Asia Mater. 9, 18, https://doi org/10.1038/am.2016.128 (2017). g ( ) 9. Uchida, M. et al. Biological containers: Protein cages as multifunctional nanoplatforms. Advanced Materials 19, 1025–1042, https:/ doi.org/10.1002/adma.200601168 (2007). g 10. Crichton, R. R. & Bryce, C. F. A. Subunit Interactions in Horse Spleen Apoferritin - Dissociation by Extremes of Ph. Biochem J 133, 289–299, https://doi.org/10.1042/bj1330289 (1973). p g j 1. Kim, M. et al. pH-dependent structures of ferritin and apoferritin in solution: disassembly and reassembly. Biomacromolecules 12 1629–1640, https://doi.org/10.1021/bm200026v (2011). p g 2. Caruso, F., Furlong, D. N. & Kingshott, P. Characterization of ferritin adsorption onto gold. J Colloid Interf Sci 186, 129–140, https:/ doi.org/10.1006/jcis.1996.4625 (1997).h g j ( ) 13. Cornell, T. A., Srivastava, Y., Jauch, R., Fan, R. L. & Orner, B. P. Author contributions J.A. performed the AFM experiments in liquids; L.S. performed the high-resolution AFM experiments; L.S. has written the manuscript; C.D. revised the manuscript; C.D. and L.S. designed the experiments. L.S. and J.A. contributed equally to this work. References The Crystal Structure of a Maxi/Mini-Ferritin Chimera Reveals Guiding Principles for the Assembly of Protein Cages. Biochemistry-Us 56, 3894–3899, https://doi.org/10.1021/acs.biochem.7b00312 (2017). ( ) 14. Dufrene, Y. F. et al. Imaging modes of atomic force microscopy for application in molecular and cell biology. Nat Nano 12, 295–307, https://doi.org/10.1038/nnano.2017.45 (2017). g 15. Li, J. W., Cleveland, J. P. & Proksch, R. Bimodal magnetic force microscopy: Separation of short and long range forces. Appl. Lett. 94, 163118, https://doi.org/10.1063/1.3126521 (2009). p g 16. Dietz, C., Herruzo, E. T., Lozano, J. R. & Garcia, R. Nanomechanical coupling enables detection and imaging of 5 superparamagnetic particles in liquid. Nanotechnology 22, https://doi.org/10.1088/0957-4484/22/12/125708 (2011).h 17. Kosmulski, M. The pH-dependent surface charging and points of zero charge V. Update. J Colloid Interf Sci 353, 1–15, https://doi. org/10.1016/j.jcis.2010.08.023 (2011).h 18. Kosmulski, M. The pH dependent surface charging and points of zero charge. VII. Update. Adv Colloid Interfac 251, 115–138, https://doi.org/10.1016/j.cis.2017.10.005 (2018). p g j 9. Nerurkar, A. V., Vaidya, N. R., Advani, S. H. & Gothoskar, B. P. Characterization of ferritin from spleens of patients with Hodgkin’ disease (HD). Neoplasma 40, 31–34 (1993). p 0. Ohnishi, S., Hara, M., Furuno, T., Okada, T. & Sasabe, H. Direct Visualization of Polypeptide Shell of Ferritin Molecule by Atomic Force Microscopy. Biophys J 65, 573–577, https://doi.org/10.1016/S0006-3495(93)81125-1 (1993). py p y p g ( ) ( ) 21. Narayanan, S. et al. On the structure and chemistry of iron oxide cores in human heart and human spleen ferrit liquid cell electron microscopy Nanoscale 11 16868–16878 https://doi org/10 1039/C9NR01541H (2019) py p y p g 21. Narayanan, S. et al. On the structure and chemistry of iron oxide cores in human heart and human spleen ferritins using grap l d ll l l h d ( ) 21. Narayanan, S. et al. On the structure and chemistry of iron oxide cores in human heart and human spleen ferritins using graphene liquid cell electron microscopy. Nanoscale 11, 16868–16878, https://doi.org/10.1039/C9NR01541H (2019). 21. Narayanan, S. et al. On the structure and chemistry of iron oxide cores in human heart and human spleen ferritins using graphene liquid cell electron microscopy. Nanoscale 11, 16868–16878, https://doi.org/10.1039/C9NR01541H (2019). q y g 22. Granier, T., Gallois, B., Dautant, A., DEstaintot, B. L. & Precigoux, G. Comparison of the structures of the cubic and tetragonal f of horse-spleen apoferritin. Acta Crystallogr D 53, 580–587, https://doi.org/10.1107/S0907444997003314 (1997).h 23. Berman, H. M. et al. Competing interestsh p g The authors declare no competing interests. Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3 www.nature.com/scientificreports/ Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-53943-3. Correspondence and requests for materials should be addressed to C.D. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 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 Cre- ative 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 per- mitted 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/. © The Author(s) 2019 Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-53943-3. Correspondence and requests for materials should be addressed to C.D. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 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 Cre- ative 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 per- mitted 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/. © The Author(s) 2019 Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-019-53943-3. Correspondence and requests for materials should be addressed to C.D. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 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 Cre- ative 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 per- mitted 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/. © The Author(s) 2019 Scientific Reports \| (2019) 9:17755 \| https://doi.org/10.1038/s41598-019-53943-3
https://openalex.org/W4294662279	http://estocastica.azc.uam.mx/index.php/re/article/download/65/61	Spanish; Castilian	null	A Multifractal Analysis Application of Hölder Exponents in Mexican Financial Markets: Mexican Stock Index and Foreign Exchange USD/MXN	Estocástica finanzas y riesgo (en línea)/Estocástica finanzas y riesgo	2,014	cc-by	7,007	* Universidad Nacional Autónoma de México Facultad de Contaduría y Administración. stretomx@gmail.com Estocástica: finanzas y riesgo. Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 . ISSN: 2007-5383 Estocástica FINANZAS Y RIESGO Estocástica FINANZAS Y RIESGO Estocástica FINANZAS Y RIESGO Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica: finanzas y riesgo. . ISSN: 2007-5383 RESUMEN El modelo multifractal ha demostrado que es posible modelar sistemas eco- !" # $ %& ' $( ) % &!* propuesta de este trabajo es demostrar que a través de la evolución de los expo- +, ) ' . # !/ ) .- terminaron los exponentes locales y puntuales de las series de tiempo históricas % 0/123.) 4 3 56678$9 & ) . ) % '% ! " ' ) una introducción, en la primera parte se elabora una revisión de la literatura, ' 9 '% . ) 9 ) . ( ! 1 ;"18.1<8.1.=> Palabras clave:"3 +, .? #2 .@ 2'3% 0%,!%4,!%5(!6(7 Keywords:89'" ! : !& # Aplicación de análisis multifractal de exponentes de Hölder en mercados financieros mexicanos: índice accionario IPC y tipo de cambio USD/MXN 191 191 Introducción E 3 +, ) 3 !"3 ( ) ( %$ &) # !"3 - ' A ' B' del supuesto de que antes de que ocurra un evento crítico la dinámica intrínse- .' ( ' !" explica cualitativamente como una reducción universal de los horizontes de inversión al aproximarse un punto crítico, es decir, los operadores o inversio- 9 ' 9 # - $ A A &# A el mercado se hunda, es decir, que el espectro de los horizontes temporales se ) ' !@ ' - . % % $C D @.&. . ! E * % - 9 F GH caídas dramáticas de precios en los mercados, precisamente porque las pro- piedades estadísticas de sus datos altamente no estacionarios son muy di- % !" . modelo basado en datos obtenidos en tiempos estables y estacionarios no es ! ABSTRACT through the evolution of Hölder Exponents and the detection of their irregulari- ! " " # $% & '" - !())* +(,! '- ! series studied, through the methodology stated, is proposed. 0 ! ! ! 1! last part conclusions are drawn. 2'3% 0%,!%4,!%5(!6(7 Keywords:89'" ! : !& # 192 Volumen 4, número 2, julio - diciembre, 2014 Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO '% ' 9 !? temperatura, los spins forman un campo de vector aleatorio, con media cero ) A ' ! / .% )A - dística es la misma a diferentes escalas, sin importar las características de la !* A # A ' % ' 9. %- B .' # A5 9' !* % ) ! I 3. A $IG'.& A# A - sultado de la incorporación de información nueva sobre las empresas y el !" . A crean; en tanto los componentes reaccionan, el todo cambia, y en tanto el !* ' ' . ' !@ ) 9 A dictan los índices de los mercados, de la misma forma que la frecuencia y la naturaleza de las partículas en la arena dictan la estadística de las avalanchas B $2 - .6>. .6JJ&!* #. y la economía es importante porque los sistemas dinámicos muestran que la ) ' 3 ! Introducción * ' %- A )A # A $IG'.J&! sistemas complejos en equilibrio pueden conducir a estados críticos y mos- trar estructuras fractales aleatorias en los que sus datos estadísticos poseen !/ . ' ) . A ) '- 193 1. Revisión de la literatura L A . ' % $@ 6)6>.66&. ' A lineal de la estructura de los mercados no puede explicar con exactitud las % B . # !* ' ) 9 - # % %A #- cos complejos revelan sus propiedades mejor bajo condiciones de estrés que $; .@ D.&! 194 Volumen 4, número 2, julio - diciembre, 2014 Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... ' ( 5 !" 66> ) 2 ' )@ ) A - ))A ' A incluye leyes de potencia, es decir encontraron evidencia de escalamiento en ' !" 66>.2 . - $2 . )166>&B / $66&)+ $66& ) %- ' ) ) ! 1. Revisión de la literatura 1)M $8& ) + - cado brasileño a partir de un modelo multifractal para pronosticar crisis en dicho mercado; Kuperin et !9 3 +, - 9 "O 6J>!1 )P $& 9 - 6J>) ) A) # A ! Q @ $8& - ) 3 +, !; . @ D$& A '7 - A %' .) A # ! L @ .; DI$66& mercados accionarios antes de una crisis se relaciona con la conducta volátil A A !" ) $) &% A A 3 - # ! R %) # + 2 " $+2"&)+ 2 $+2& Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Tabla 1: Diferencias más notables entre HME y HMF: Tabla 1: Diferencias más notables entre HME y HMF: Fuente de elaboración propia. Estadística Gaussiana. Retornos indepen- dientes normalmente distribuidos. Estadística No Gaussiana. Los cambios en los precios no están distribuidos de forma normal. Pueden caer más rápido de lo que se elevan. Movimiento browniano geométrico: 0.50 o Raíz cuadrada del tiempo para escalar volati- lidades a diferentes periodos de tiempo. Movimiento browniano fraccionario: Leyes de potencia o escala. Procesos estacionarios. Procesos no estacionarios. No hay correlaciones históricas. Hay correlaciones históricas. No hay memoria. Eventos pasados no in- Existe memoria (los eventos están interco- - ciados por lo ocurrido. Sus expectativas están basadas en experiencias anteriores. No se repiten (o muy raramente) los patrones a ninguna escala. Se repiten patrones en todas las escalas de tiempo (minutos, días, años). Continuas y estables en todas las escalas. Se presentan discontinuidades en cualquier escala. (Por ej. Vuelos de Lévy o eventos tipo cisnes negros) Cada individuo interpreta la información de diferentes formas y en distintos momentos. (En ocasiones la información no se asimila de forma inmediata ni en un solo periodo de tiempo, a veces la información del presente tiene efectos claros en el futuro). Los inversionistas son adversos al riesgo, y son racionales. Los inversionistas pueden actuar de forma irracional, y pueden buscar riesgo cuando hay peligro de perder valor. Análisis multifractal El análisis multifractal es relativamente reciente en el campo de estudio de #$ .6JJ&!? % . - 195 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO Fuente de elaboración propia. Estadística No Gaussiana. Los cambios en los precios no están distribuidos de forma normal. Pueden caer más rápido de lo que se elevan. Estadística Gaussiana. Retornos indepen- dientes normalmente distribuidos. Movimiento browniano fraccionario: Leyes de potencia o escala. Existe memoria (los eventos están interco- - ciados por lo ocurrido. Sus expectativas están basadas en experiencias anteriores. Cada individuo interpreta la información de diferentes formas y en distintos momentos. (En ocasiones la información no se asimila de forma inmediata ni en un solo periodo de tiempo, a veces la información del presente tiene efectos claros en el futuro). Los inversionistas pueden actuar de forma irracional, y pueden buscar riesgo cuando hay peligro de perder valor. Fuente de elaboración propia. # %. # 'S%. . #) %! * ' A% incremento de susceptibilidad y señales precursoras similares a la inestabili- % #%. # 3 ! 196 Volumen 4, número 2, julio - diciembre, 2014 Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... 2. Metodología Existen diversas técnicas que han sido desarrolladas para estimar los expo- ) +, . - !" ' %!"9 3 ) # !@9 ' 2?R?I$ & 3 - ) +, ! Exponentes de Hölder "3 ' !# 3 +, . 3 3 +, ! O 3 +, ' 5 y se usa para detectar la presencia de discontinuidad y cuándo ocurre di- 5 #!?%. ' 5 # ( A 5 !/ - . ' 3 +, # T!1 . 9 3 +, !'. ' 3 +, $@ .P D .&! 3 ( 5 #' .# 9!? . %3 +, 9 - 9 5 # ! * ' ' - mente poseen información esencial en su señal; para caracterizar estas es- ' 5 ! Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Exponente puntual * 3 ).A 3 9 ' . )3 ' - . ! 197 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 O ' # 9 3 +, . A '- res y suavizadas de una función presentan valores más altos, es decir, este 3 ' ' #!1 - . ' 5.A $'.P D@ .&! Sea f : ... y... ² tal que f no es diferenciable en . Sea f : ... y... ² tal que f no es diferenciable en . "3 +, f en ( "3 +, f en ( p lim inf log \| - \|\|h\|\| log \|\|h\|\| (1) (1) donde \|\| . \|\| denota la norma euclidiana en ²!" ' ' AA. A >0.' # f , ' A5.# p ) - ± \| x - \| (2) (2) / p ± \| x - \| (3) (3) * % 5 % alrededor de un punto ' B(,n)..? - )' . A A5; cercana al valor de cero se traduce en una señal altamente variable, mientras que una ; cercana al valor . 5#! Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 199 Exponente puntual @'( R .' .Q'))7M$& para estimar ;sería usar los incrementos y directamente aplicar la ecuación .3 . ' ) - !@' 3 AA9 198 Volumen 4, número 2, julio - diciembre, 2014 Aplicación de análisis multifractal de exponentes de Hölder... Estocástica FINANZAS Y RIESGO Figura 1: Envolvente de Hölder en una señal no diferenciable f en un punto . Figura 1: Envolvente de Hölder en una señal no diferenciable f en un punto . Fuente: Elaboración propia. Esta gráfica no contiene datos reales es únicamente para explicar de forma gráfica el comportamiento de una señal a partir del valor de ;. Fuente: Elaboración propia. Esta gráfica no contiene datos reales es únicamente para explicar de forma gráfica el comportamiento de una señal a partir del valor de ;. * f en una bola centrada en t osc (t) = sup f(t’) - inf f(t’) -t - (4) (4) @ 3 3 +, $ &! $ & " . 3 +, f (t) = lim inf log osc (t) log (5) (5) Para estimar (t) con esta fórmula, el límite inferior es de hecho un límite y ' < A5 Para estimar (t) con esta fórmula, el límite inferior es de hecho un límite y ' < A5 log osc (t) f (t) log + k (6) (6) Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 199 199 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO Donde k !@ f (t) ' % ' ' 5< * # - 5% )#3 ' <, y la base de incremento <entre sus 3 ! Exponente puntual O 3 ' ) ) ' ! Volumen 4, número 2, julio - diciembre, 2014 0 Volumen 4, número 2, julio - diciembre, 2014 Exponente local / 3 +, . A$?'. C <& / 3 +, . A$?'. C <& " +, ;!si se (7) \| f (t+ h) - f (t)\| < consts × h ; t, h, (7) O +, AA 3 +, ! En el caso de que ;dependa de t, , entonces ; 3 +, ! Cuando ; = 1, coincide con la clase diferenciable de funciones; cuando ; = 0, coincide con la clase de funciones discontinuas, por lo tanto, la clase +, # ) .)3 +, ' ! Para obtener un exponente local, se parte de información expresada en . Sea Y(t) .' ' Y - ln Y (t ))² (8) (8) 200 Volumen 4, número 2, julio - diciembre, 2014 Aplicación de análisis multifractal de exponentes de Hölder... Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... Estocástica FINANZAS Y RIESGO ' [0,T] entre N intervalos del tamaño de > ) N-1 Z \| X i = 0 X (9) (9) (9) " . ' (r)= lim ln Z( T, t) t (r)= (10) (10) (10) " ' % . X(t,1) " ' % . Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 201 Exponente local X(t,1) D= ! r-1 D= (11) D= ! r-1 D= (11) (11) . 3 +, . 3 +, = d (r) dr = (12) (12) " !@ D "# para r real, entonces X (t,1) !/ monofractales la función de escala (r) ! 9 3 +, (t) $t) (13) (13) En el modelo clásico de precios de activos0 (t) = 1 En el modelo clásico de precios de activos0 (t) = 1 O 3 A ) ' A 5 !? . #3 ' ) ' 5! 201 3. Resultados Estocástica FINANZAS Y RIESGO Cuando se obtienen exponentes locales y puntuales con valores entre 0 y 1, ' A5 - B # A5 3 ' - A5# ' ) # 3# ' ! " 3 . ' A .) !L '%?')C $<&. 9 ) ' ' - ' ' !".A $) & 3 '3 ) '! + > # Aplicación de análisis multifractal de exponentes de Hölder... Aplicación de análisis multifractal de exponentes de Hölder... Estocástica FINANZAS Y RIESGO Cuando se obtienen exponentes locales y puntuales con valores entre 0 y 1, ' A5 - B # A5 3 ' - A5# ' ) # 3# ' ! " 3 . ' A .) !L '%?')C $<&. 9 ) ' ' - ' ' !".A $) & 3 '3 ) '! +)> ' ' #3 3 . '.< 3 - ) '3 ) A % ! 3. Resultados Se analizaron datos de cierres diarios de mercado, tanto del índice accionario 230/1 4 3 !* 9 66 8. ! Se analizaron datos de cierres diarios de mercado, tanto del índice accionario 230/1 4 3 !* 9 66 8. ! En estos resultados se muestra que existen patrones que pueden modelar correctamente los momentos de crisis o caídas e incluso como alertas para anticipar una crisis, y que dichos patrones han mostrado funcionar bien en 9! En estos resultados se muestra que existen patrones que pueden modelar correctamente los momentos de crisis o caídas e incluso como alertas para anticipar una crisis, y que dichos patrones han mostrado funcionar bien en 9! ' . ' #. 5 % - les punteadas continuas las épocas de crisis que fueron detectadas exitosa- 3 +, !% y el círculo discontinuo son caídas detectadas por los exponentes puntuales, A 3 ! * ' # 4 3 !" ' ' # 3 +,- !" ' # 3 A ! 202 Volumen 4, número 2, julio - diciembre, 2014 Figura 2: Tipo de cambio USD/MXN por el periodo de 1994-2013, exponentes locales de Hölder y exponentes puntuales de Hölder Figura 2: Tipo de cambio USD/MXN por el periodo de 1994-2013, exponentes locales de Hölder y exponentes puntuales de Hölder Figura 2: Tipo de cambio USD/MXN por el periodo de 1994-2013, exponentes locales de Hölder y exponentes puntuales de Hölder 202 Volumen 4, número 2, julio - diciembre, 2014 Volumen 4, número 2, julio - diciembre, 2014 Aplicación de análisis multifractal de exponentes de Hölder... Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 3. Resultados 1 .% ' #- 3 A $ 3 )'3 &!L .3 3- % . 66.66<. 66J) 67 !O 3 A ) 5 ! " )' 566 566<B 3 \"RA] $2.&) ! "' 566J. - ? * #!" 5)A'! 203 " 56) - A9 ) ' )A - "O ) ! / 3 ' - ' 3 . . ' A3 ) cercano a cero, coinciden con los momentos en los que sucedieron las caídas 566.66<.66J.)6B '.3 - tes puntuales revelan otro patrón de caída que ocurre a mediados de 2012 y que no es detectado por los exponentes locales, y que corresponde a una % !" ' # . A% A3 ! ' 8.# ' #B 9% 230/1.' ' # 3 ) 3 ! % A baja fuerte o crisis en el mercado y que fue detectada exitosamente por los 3 !% % 3 ! Volumen 4, número 2, julio - diciembre, 2014 3. Resultados @ 3 8 % 3 - !3 . # 3 A ' 9# ' !@ % 66>7 666.' 56) ! " 566) 3 ) ' pero es muy alto, e indica un cambio en la dinámica del comportamiento de .A 66 23B A 663 ' .A A ' ese año y hubo muchos movimientos muy bruscos en las subidas y bajadas de precios; los exponentes puntuales también detectan este cambio y aunque no ' ) . ' 6666<! ? 566>.3 #)A 3 ' .)' ).A !" #A' 66J!3 204 Volumen 4, número 2, julio - diciembre, 2014 Aplicación de análisis multifractal de exponentes de Hölder... Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... Figura 3: Índice de la Bolsa Mexicana de Valores: IPC 1992-2013, exponentes locales de Hölder y exponentes puntuales de Hölder Figura 3: Índice de la Bolsa Mexicana de Valores: IPC 1992-2013, exponentes locales de Hölder y exponentes puntuales de Hölder locales de Hölder y exponentes puntuales de Hölder ' 566>. - 3 ! 3 % 6) - % A % # 0/1! Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO 4 3 . #3 +, !?#. . - A 9 3 - nentes, tanto el puntual como el local, ya que existió un caso en el que los que los exponentes puntuales detectaron una caída que los exponentes locales .A A3 ) +, # ! Q . # 3 ) +, . ' 4 ' !2## 3%a posteriori, los exponen- +, % A % )A ' .%- tectaron de forma previa y un instrumento útil en el modelado y posible de- % ( ' ' '' ! Volumen 4, número 2, julio - diciembre, 2014 4. Conclusiones * A o aplicaciones anteriores de este tipo de análisis multifractal para analizar % 3 . % 0/1) 205 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO En este estudio, se encontró que los exponentes locales y puntuales de +, ( ) % .)A - cientemente bien ciertos eventos críticos en las series de tiempo del tipo de 4 3 ) % 230/1!" #- 9 % 230/1 O@L42_`. 6678 A 3 )+, )%.) A ! 3 +, distintas dinámicas como los mercados cambiarios y los accionarios y muy posiblemente serían útiles en otros mercados como los de commodities, deu- ) ! %3 #- 9 ! * ' '- ?')C $<&.'.P )@ - $&)C )@$& A #- tifractal utilizando los exponentes para detectar patrones de crisis en otros ! 206 Volumen 4, número 2, julio - diciembre, 2014 Estocástica FINANZAS Y RIESGO Estocástica FINANZAS Y RIESGO Aplicación de análisis multifractal de exponentes de Hölder... Volumen 4, número 2, julio - diciembre, 2014 Bibliografía ?'.0!?!.C g!?!$<&!2 ? ) +, "3 - ? L '@G2 G1 . ?)?!.)7M.;!$&!FQ T 3 ' 9 I T H # Processes and Their Applications.M!$&. !67<! IG'.;!2!.$J&!F? 2 G+) 2 - R@ ? )H!#:# ' # Processing.M!$&! $&!F@) PG ? ' ` 7 ) - L) @2 ? )" @' !H#:# % # .M!$&. !>k6! 1 .?!.P.g!$&!F2 @ ? )6J>@G 2 G 1 H! C @ O ). 44<!<!8!4 / 4<J<<>84 7 7 )7776J>7G7 G7 ! 1P!.M =!!$8&!F 'P '1 ` ) I 9 @G2 GH. Physica ?.86. !87J! "!! $6>&! F" 1 2 G PT R ) " w G?2 & .<! $66&!F" 1 2 G0!2 & H!M! $<&! .;!$6JJ&! “ H!Plenumm Press.`Tg G.* . !<! '?!.P 2!.@ ."!/!$&!FP' )? )2 - G0 3 J" 1 H!@ B 0C Aplicaciones.M!6$&. !<7>J! + $66&!% D F !Q3 O )/ ! ; .@$&!w A )!0 =- ) ? ?;I 2 .$ # in Pure Mathematics.M!>. !67<! ; .?!.@ L!.Q!$&!F1 1 / ”. Interna- 2 G: & .M!8$&. !67<<! C g!?.@P!P!$&!2+ "3 ? / O@?@G2 G1 / 1 ”. Saint Peters- '@O ).P! 207 Volumen 4, número 2, julio - diciembre 2014, pp. 191–208 Estocástica FINANZAS Y RIESGO Estocástica FINANZAS Y RIESGO Estocástica FINANZAS Y RIESGO La geometría fractal de la naturaleza ! " #$ % &" ' ( ' ") - +. ./0 1 2 3 Physica A 45 1 * 2 3 Physica A 45 4 An Introduction to Econophysics, Correlations and Complex- # 1 6$ 7 (! " 44 &8( 9 :( ( 0;<+ 101 44 3 $ 2(! 7 Chaos, order in the capital markets = ! > ? 2 @( AB C D 2 7"/0 7 ' ! 7 "( 7 ( ( * - ly. Industrial Management Review / 2 ! 3 * " A # * 44 &21 E( 61 3F 8 + !"# "$ % - "$ 7 2 ! G 2 E " = ! = (! 7 / &2 (D D # ! 7 ( - *(&' / /;< >! 2 (D D # ! Physics Reports H 5I ; >! 2 (D D # ! Physics Reports H 5I ; J % %1 7 K1 L %M$;HM! = 44 “$ $1 ' ! 7 B 3F 8 B! L( 7 1 1+ %$ "- ences $ H 4 /; 208
https://openalex.org/W4385547024	https://journal.formosapublisher.org/index.php/esa/article/download/4840/5063	English	null	Demographic Characteristics and Reasons for Progestin Subdermal Implant (PSI) Use: Inputs for Alternative Family Planning Method	Indonesian Journal of Applied and Industrial Sciences	2,023	cc-by	6,397	Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Demographic Characteristics and Reasons for Progestin Subdermal Implant (PSI) Use: Inputs for Alternative Family Planning Method Ma. Sandra S. Rivera1, John Mark R. Asio2* College of Allied Health Studies, Gordon College, Olongapo City Corresponding Author: John Mark R. Asio asio.johnmark@gmail.com Ma. Sandra S. Rivera1, John Mark R. Asio2* College of Allied Health Studies, Gordon College, Olongapo City Corresponding Author: John Mark R. Asio asio.johnmark@gmail.com A R T I C L E I N F O Keywords: Family Planning, Progestin Subdermal Implant (PSI), Reproductive Health, Community Health Development Received: 22 May Revised: 23 June Accepted: 28 July ©2023 Rivera, Asio: This is an open-access article distributed under the terms of the Creative Commons Atribusi 4.0 Internasional. A R T I C L E I N F O A B S T R A C T Keywords: Family Planning, Progestin Subdermal Implant (PSI), Reproductive Health, Community Health Development Received: 22 May Revised: 23 June Accepted: 28 July ©2023 Rivera, Asio: This is an open-access article distributed under the terms of the Creative Commons Atribusi 4.0 Internasional. This study aims to analyze the demographic characteristics and reasons for Progestin Subdermal Implant (PSI) use among selected women in Central Luzon, Philippines. The study used descriptive-correlation research on 75 volunteer women using purposive sampling. The study observed significant differences in age, occupation, educational attainment, number of living children, type of family planning used before PSI use, reasons for practicing family planning, type of birth delivery, and plan to have another baby. Lastly, significant relationships were found in the same demographic characteristics and reason for the PSI use. The study concluded that there exists variation in the use of PSI and association for the benefit of PSI among the respondents. A B S T R A C T This study aims to analyze the demographic characteristics and reasons for Progestin Subdermal Implant (PSI) use among selected women in Central Luzon, Philippines. The study used descriptive-correlation research on 75 volunteer women using purposive sampling. The study observed significant differences in age, occupation, educational attainment, number of living children, type of family planning used before PSI use, reasons for practicing family planning, type of birth delivery, and plan to have another baby. Lastly, significant relationships were found in the same demographic characteristics and reason for the PSI use. The study concluded that there exists variation in the use of PSI and association for the benefit of PSI among the respondents. INTRODUCTION Progestins are synthetic hormone drugs that mimic progesterone's endogenous hormone (Edwards & Can, 2023). In contraception, progestin is available in multiple forms (Liu et al., 2022), including an implant. Progestin subdermal implants (PSI) are a highly effective long-acting reversible contraceptive method that has gained popularity in recent years across the globe (Kolawole et al., 2018). PSI is a small, flexible rod inserted under the upper arm's skin and releases a synthetic hormone called progestin, which prevents pregnancy by thickening cervical mucus and thinning the lining of the uterus. PSI offers a convenient and discreet form of contraception that lasts 3-5 years, making it an attractive option for women who want to avoid daily pill-taking or other short-acting methods. PSI is safe and effective in numerous clinical trials and has been endorsed by the World Health Organization (WHO) as a recommended contraceptive method. For instance, in sub-Saharan African countries, 10 out of 12 countries have an implant contraceptive prevalence rate of around six percent or higher (Jacobstein, 2018; Krogstad et al., 2019). This finding is a significant improvement from the previous study by Fiato in 2016. As access to family planning services expands globally, PSI is becoming an increasingly popular choice for women seeking reliable, long-term contraception. Although a decade ago, only less than one percent of used hormone-releasing subdermal implants worldwide (Rademacher et al., 2013). However, the use of PSI in the Philippines has been seen as a positive development in the country's efforts to expand access to family planning services. PSI's convenience, effectiveness, and safety make it an attractive option for many women seeking reliable contraception, and continued efforts to increase access and awareness of this method can help further improve reproductive health outcomes in the Philippines. Nevertheless, since the PSI is still new here in the locality of Olongapo City, the proponents proposed this study to lay down some preliminary information which can be helpful for all healthcare workers. This study presents the following research questions, which the current udy intends to answer at the end: This study presents the following research questions, which the current study intends to answer at the end: 1) What are the demographic characteristics of the respondents of the study? 2) Is there a significant difference in the reasons for Progestin Subdermal Implant (PSI) use when grouped according to the demographic characteristics of the respondents? ( DOI: https://doi.org/10.55927/esa.v2i4.4840 E-ISSN: 2985-5055 https://journal.formosapublisher.org/index.php/esa ( DOI: https://doi.org/10.55927/esa.v2i4.4840 E-ISSN: 2985-5055 https://journal.formosapublisher.org/index.php/esa 361 Rivera, Asio LITERATURE REVIEW There are several advantages of progestin subdermal implants (PSI) over other contraceptive methods like its long-acting (Sah et al., 2018), highly effective use, reversible (Britton et al., 2020), low-maintenance, discreet, and safe. A recent study by Waris and Fatima (2022) provided evidence of such a claim. Although, there will always be some unaccounted effects on the individuals, like physiological and psychobehavioral side effects (Mitchell & Welling, 2020) y g y g Overall, PSI offers highly effective, long-lasting, and low-maintenance contraception that is safe and reversible. These advantages make it an attractive option for many women seeking reliable contraception. p y g p In the Philippines, progestin subdermal implants (PSI) have become an increasingly popular contraceptive method in recent years. PSI provides a highly effective, long-lasting, and low-maintenance form of contraception suitable for women of all ages, especially those who have difficulty accessing or remembering to use other contraceptive methods (Santiago & Pastrana, 2022). The Philippine government has recognized the importance of family planning and has made efforts to increase access to modern contraceptives, including PSI. The passage of the Responsible Parenthood and Family Planning Law in 2012 marked the government's initiative. Nevertheless, the law's implementation was slow and fragmented, leading to a sluggish improvement in reproductive health (Van et al., 2021). The Department of Health has included PSI in its national family planning program and has worked to train healthcare providers in the proper insertion and removal of the implant. However, based on several searches on the world-wide-web, only a few significant pieces of literature mainly discussed PSI use in the Philippines. Therefore, the proponents conceived writing out an article for the healthcare field as a source of essential baseline data for everyone else to use and to be a basis for other vital projects that involve family planning. Despite these efforts, access to PSI remains limited in some areas, particularly in rural and remote areas where healthcare services are scarce. For instance, in sub-Saharan Africa, only 1% of women use implants (Jacobstein & Polis, 2014). There are also some cultural and religious barriers to using modern contraceptives, making it difficult for women to access and use PSI. p g Figure 1 shows the study's conceptual framework on which the basic premise presented the independent variable (IV) – dependent variable (DV) model. One can decipher that the dependent variable includes the demographic characteristics of the respondents. INTRODUCTION 3) Is there a significant relationship between the reasons for Progestin Subdermal Implant (PSI) use and the demographic characteristics of the respondents? The main objective of this study is to gather information, especially the demographic characteristics of the chosen sample, and try to analyze any variations or relationships that may be present within the gathered data. The results of this study can benefit healthcare workers, especially midwives, nurses, and medical doctors. At the same time, this study can also become a basis and reference to the growing literature about family planning and contribute to 362 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 promoting responsible parenting and safe motherhood to all Filipino women in the country. promoting responsible parenting and safe motherhood to all Filipino women in the country. Design The study's proponents used a descriptive-correlation research design with a survey as the primary data-gathering tool. Since the main objective of this study is to gain vital information regarding the demographic characteristics of the respondents, the said design is applicable and appropriate. LITERATURE REVIEW They comprise the respondents' age, occupation, civil status, educational attainment, religion, location, monthly income, number of living children, type of family planning prior to PSI use, menstrual flow, how they heard about PSI, reasons for practicing family planning, plan to have baby and type of delivery. On the other hand, the study's independent variable is the reason for PSI use by the respondents. Since the study aims to determine the association between the demographic characteristics and 363 Rivera, Asio the reason for PSI use among the respondents, the IV-DV model suits the framework. Figure 1. Conceptual Framework Reasons for Progestin Subdermal Implant Use Demographic Characteristics of Respondents Dependent Variables Independent Variables Rivera, Asio the reason for PSI use among the respondents, the IV-DV model suits th framework. Figure 1. Conceptual Framework Reasons for Progestin Subdermal Implant Use Demographic Characteristics of Respondents Dependent Variables Independent Variables Independent Variables Dependent Variables Reasons for Progestin Subdermal Implant Use Demographic Characteristics of Respondents Figure 1. Conceptual Framework Respondents The study's respondents were selected individuals from a community extension service from November 2018 to February 2019 sponsored by a higher education institution in coordination with the Population Commission ( the Citof y of Olongapo. Seven, Philippinesty-five voluntary respondents participated with the aid of the purposive sampling technique. The inclusion criteria were: a) a participant in the said community extension service; b) a new or old family planning user; c) a woman or a mother; d) willing to try the PSI as a family planning method. Exclusion criteria, on the other hand, include a) passerby or not participant in the community extension; b) not in the family planning method; c) a male or a father; and d) not willing to use PSI. ) ) g The proponents of this survey provided and explained the PSI method to the participants. After a thorough discussion, they asked for informed consent prior to the administration of the PSI. Participation in the said family planning method was purely voluntary, and there was no threat or harm to those who opted not to join or participate. Data Analysis The gathered data from the survey underwent a normality check to determine whether the data was standard in distribution. At the same time, upon determination of its normality, the data analyst can use the appropriate statistical treatment for the data. The normality test yielded an abnormal data distribution. Therefore, the data analyst opted to use a non-parametric test. The data analyst performed a Mann-Whitney U test, Kruskal Wallis test, and a Chi-Square test in the study. In order to calculate the said statistical treatments, the study used a statistical software, Statistical Package for Social Sciences (SPSS) version 23, for the said purpose. Instrument The instrument used in the survey was the Family Planning Client Assessment Record form by the Department of Health, which contained the respondents' basic profiles and other essential information checklists. The form does not need any validation anymore since it came from a reliable government agency which already assured its contents are viable and reliable for data gathering. 364 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 RESEARCH RESULT The study's main objective is to analyze the demographic characteristics and the reasons for the Progestin Subdermal Implant (PSI) use by selected women in the city of Olongapo. It also determined the differences and relationships between the demographic characteristics and the reasons for PSI use. The succeeding tables below illustrate the results of the study. Table 1. Demographic Characteristics of the Respondents Variables Frequency Percentage Age 17-20 years old 20-25 years old 26-30 years old 31-35 years old 36-40 years old 41 years old and above 5 11 17 25 14 3 6.67 14.67 22.67 33.33 18.67 4.00 Occupation Teacher Housekeeping Student Vendor SBMA Employee 5 39 11 5 15 6.67 52.00 14.67 6.67 20.00 Educ. Attainment Elementary Graduate High School Level High School Graduate College Level College Graduate 8 8 26 17 16 10.67 10.67 34.67 22.67 21.33 Religion Roman Catholic Born Again Iglesia ni Cristo Mormons 36 17 19 3 48.00 22.67 25.33 4.00 Location Banicain Bataan East Bajac Bajac Gordon Heights 6 15 3 7 8.00 20.00 4.00 9.33 Table 1. Demographic Characteristics of the Respondents Variables Frequency Percentage 365 Rivera, Asio New Cabalan Old Cabalan Pampanga Sta. Rita West Bajac Bajac Zambales 6 11 2 9 3 13 8.00 14.67 2.67 12.00 4.00 17.33 Monthly Family Income Less than 5,000 pesos 5,000-9,000pesos 10,000-14,000 pesos 15,000-19,000 pesos 23 24 19 9 30.67 32.00 25.33 12.00 No. of Living Children One Two Three Four Five Six Seven 12 11 9 17 14 8 4 16.00 14.67 12.00 22.67 18.67 10.67 5.33 Type of FPFP Prior to PSI COC Condom DMPA LAM POP None 11 9 9 10 10 26 14.67 12.00 12.00 13.33 13.33 34.67 Menstrual Flow Scanty Moderate Heavy 29 33 13 38.67 44.00 17.33 How did you hear about PSI? BHW Mother’s Class Friends Hospitals RHUs Social Media TVTV and Radio 15 19 14 5 7 7 8 20.00 25.33 18.67 6.67 9.33 9.33 10.67 Civil Status Single Married 53 22 70.67 29.33 Reason for Practicing FPFP Limiting Spacing 26 49 34.67 65.33 Type of Delivery NSD CS 63 12 84.00 16.00 Plan to have another baby Yes No 32 43 42.67 57.33 Total 75 100 366 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 ( ) Vol. 2, No. RESEARCH RESULT 4 2023 : 361 - 376 Table 1 provides information on the frequency and percentage of responses to various variables among the survey participants. The variables include age, occupation, educational attainment, religion, location, monthly family income, number of living children, type of family planning (FP) prior to PSI, menstrual flow, how they heard about PSI, civil status, the reason for practicing FPFP, type of delivery, and plan to have another baby. The table shows that most survey participants were between the ages of 26 and 35, with 33.33% falling within this age range. The most common occupation among the participants was housekeeping, with 52% of the respondents indicating this. Regarding educational attainment, 34.67% of the participants were college graduates. Most participants were Roman Catholic, with 48% indicating this as their religion. The most common location was Bataan, with 20% of the participants residing there. Most participants had a monthly family income of between 5,000 to 9,000 pesos, with 32% of the respondents falling within this range. Regarding family planning, 65.33% of the participants indicated practising FPFP for spacing purposes. Most participants had a standard delivery, with 84% indicating that they had a normal spontaneous delivery (NSD). Over half of the participants (57.33%) did not plan to have another baby. Most participants were Roman Catholic, with 48% indicating this as their religion. The most common location was Bataan, with 20% of the participants residing there. Most participants had a monthly family income of between 5,000 to 9,000 pesos, with 32% of the respondents falling within this range. Regarding family planning, 65.33% of the participants indicated practising FPFP for spacing purposes. Most participants had a standard delivery, with 84% indicating that they had a normal spontaneous delivery (NSD). Over half of the participants (57.33%) did not plan to have another baby. Overall, the table provides a summary of the demographic and reproductive health characteristics of the survey participants, which can help understand the population and tailor health interventions to their specific needs. Table 2. Differences for the Reason for PSI Use When Grouped According to Demographic Characteristics Variables H df p-value Age 19.170* 5 .002 Occupation 15.531* 4 .004 Educ. Attain. 13.059* 4 .011 Religion 0.889 3 .828 Location 6.580 9 .681 Monthly Family Income 1.447 3 .694 Note: p < .05 Table 2. Differences for the Reason for PSI Use When Grouped According to Demographic Characteristics able 2. RESEARCH RESULT Differences for the Reason for PSI Use When Grouped According to Demographic Characteristics Table 2 displays the result of the Kruskal Wallis test of difference when grouped according to their demographic characteristics. One can decipher, based on the presentation, that there were significant variations in the responses when the respondents were grouped according to age (H[5]= 19.170, p= .002), occupation (H[4]= 15.531, p= .004); and educational attainment (H[4]= 13.059, p= .011). All of the mentioned probability values were lower than the alpha significance level of .05. There is a significant difference in the respondents' responses, and we reject the null hypothesis on these particular characteristics. These results only mean that the reason for PSI usage depends on demographic characteristics, which tend to affect their decision-making, especially in the family planning method. On the other hand, there were no significant differences in the answers of the respondents when grouped according to their religion 367 Rivera, Asio (H[3]= 0.889, p= .828), location (H[9]= 6.580, p= .681); and monthly family income (H[3]= 1.447, p= .694). Again, one can notice that the p-values were more greater than the .05 alpha significance level. Thus, no significant differences were observed, and we accept the null hypothesis on these particular characteristics. This finding shows that regardless of religious beliefs, location, and income, the respondents do not seem to have any discrepancies in their perspectives about PSI. Table 3. Differences for the Reason for PSI use When Grouped According to Family Characteristics Variables H df p-value No. of Living Children 34.393 6 .000 Type of FP Prior to PSI 20.822* 5 .001 Menstrual Flow .196 2 .907 How did you hear about PSI? 3.198 6 .784 Note: p < .05 able 3. Differences for the Reason for PSI use When Grouped According to Family Characteristics For table 3, the study presents the result of the Kruskal Wallis test of significant difference for the reason that PSI was used when the respondents were grouped according to their family characteristics. As seen from the table, some characteristics yielded significant results. There were significant differences for the number of living children wherein the study produced H(6)= 34.393, p= .000, and the type of family planning method prior to PSI usage generated H(5)= 20.822, p= .001. The p-value of each characteristic is less than the alpha significance level of .05. RESEARCH RESULT This result means a significant difference exists, and we reject the null hypothesis for these particular characteristics. However, in the case of menstrual flow (H[2]=.196, p= .907); and to whom did they hear about PSI (H[6]=3.198, p= .784 got no significant result based on the Kruskal Wallis test. The study grounded this generalization based on the obtained probability values, which were higher than the alpha significance level of .05. Therefore, there were no significant differences in the reason for PSI usage when grouped according to menstrual flow and their source of information about PSI. Table 4. Difference in the Reason for PSI use When Grouped According to Civil Status and Family Planning Strategies Variables Mann- Whitney U p- value Z Civil Status 520.000 .394 - .853 Reason for Practicing FP 353.500 .000 -3.671 Type of Delivery 555.000* .003 2.975 Plan to have another baby 16.000* .000 -8.372 Note: p < .05 Table 4. Difference in the Reason for PSI use When Grouped According to Civil Status and Family Planning Strategies In order to determine if there exist significant differences in the reason for PSI use when grouped according to Civil Status and Family Planning Strategies, the study performed a Mann-Whitney U test. Table 4 presents the result of the 368 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 test. As one can deduce, there was a significant result obtained by the calculation. The study revealed no significant difference between the single and married respondents since U=520.000, p= .394. The probability value was more significant than the alpha significance level of .05. Therefore, we reject the null hypothesis for this particular characteristic. On the other hand, the reason for practicing family planning generated a significant difference between limiting and spacing since the study obtained U= 353.500, p= .000. In addition, the type of delivery also yielded significant result between normal spontaneous delivery (NSD) and Cesarean Section (CSCS) with U= 555.000, p= .003; and plan to have another baby, have U= 16.000, p= .000. All of the mentioned probability values were less than the alpha significance level of .05. This result means that there exists significant difference and we reject the null hypothesis for these particular set of characteristics. Table 5. DISCUSSION The main objective of this study is to analyze the demographic characteristics and the use of Progestin Subdermal Implant (PSI) among selected women in Olongapo City. It also tried to elicit variations and associations between the demographic characteristics and the reason for using the PSI. Based on the presented results and relevant computations, the study did find some interesting findings. There were significant differences among the respondents when grouped according to their age, occupation, educational attainment, number of living children, type of family planning method prior to PSI, the reason for practicing family planning, type of delivery, and plan to have another baby. A local study by Quijencio, Jr. (2021) supports the findings of the current study, wherein there is variation in the PSI type of clients in Pasig City. One possible explanation for differences in the use of the progestin subdermal implant among different age groups is related to reproductive goals and preferences. Older women may prefer other methods that are easier to discontinue if they decide to become pregnant (Abasiattai et al., 2014). Occupation may also play a role in the use of the progestin subdermal implant. Women in certain professions, such as healthcare or education, may have greater access to information about different contraceptive methods and may be more likely to choose the implant because of its convenience and effectiveness. However, in the current study, more than half of the respondents were full-time housewives, therefore, disagreeing with the previous study by Madugu et al. in 2015. In terms of Educational attainment may also be a factor in the use of the progestin subdermal implant. Women with higher levels of education may be more likely to have access to healthcare providers who are knowledgeable about different contraceptive methods and may be more likely to have the financial resources to afford this method. The result of the study also coincides with the result of Madugu et al., 2015), wherein more than 1/3 of their respondents were in their tertiary education. Women who have already had children may be more likely to choose the progestin subdermal implant because they are looking for a long-acting and effective form of contraception that does not require daily attention (Abasiattai et al., 2014). Women who have had multiple children may also be looking for a method that is more reliable than other forms of contraception they may have used in the past. RESEARCH RESULT Chi-Square Result Between Demographic Characteristics and Reasons for PSI Profile χ2 φ df p-value Age 19.429 .509 5 .002 Occupation 15.741* .458 4 .003 Educ. Attainment 13.235* .420 4 .010 Religion 0.901 .110 3 .825 Location 6.669 .298 9 .671 Ave. Monthly Income 1.467 .140 3 .690 No. of Living Children 34.857* .682 6 .000 Type of FPFP Prior to PSI 22.072* .542 4 .000 Menstrual Flow 0.198 .051 2 .906 How did you hear about PSI 3.241 .208 6 .778 Civil Status 0.391 .099 1 .391 Reason for Practicing FP 13.655* .427 1 .000 Plan to have a baby 71.036* .973 1 .000 Type of Delivery 8.970* .346 1 .003 Note: *p < .05 Table 5. Chi-Square Result Between Demographic Characteristics and Reasons for PSI Table 4 presents the result of Chi-square computation for possible relationship between the demographic characteristics and the reasons for the progestin subdermal implant (PSI) usage among the respondents. As seen from the table, there were several demographic characteristics that obtained significant results based on the calculation of the study. The following results produced evidence of significant relationships: since age got χ2 (5, N=75) = 19.429, p= .002; occupation generated χ2 (4, N=75) = 15.741, p= .003; and for educational attainment garnered χ2 (4, N= 75) = 13.235, p= 0.010. In addition, for the family characteristics, the number of living children yielded χ2 (6, N= 75) = 34.857, p= .000 and type of family planning prior to PSI garnered χ2 (4, N= 75) = 22.072, p= .000. And for the family planning aspects of the study, the reason for practicing FPFP got χ2(1, N= 75) = 13.655, p= .000; plan to have a baby generated χ2 (1, N= 75) = 71.036, p= .000, and type of delivery obtained χ2 (1, N= 75) = 8.970, p= .003. 369 Rivera, Asio All of the mentioned probability values were less than the .05 alpha significance level. These results only mean that all of these demographic characteristics got a significant relationship with the reason for using progestin subdermal implant (PSI) among the respondents. Therefore, the null hypothesis in these characteristics is rejected. The rest of the demographic characteristics did not generate enough to sustain a significant relationship with the use of PSI among the respondents. DISCUSSION In the study, about 57% of the respondent have more than three children already, which is more than expected as compared to that of Balogun et al. (2014), where 33% of the respondents have 370 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 ( ) Vol. 2, No. 4 2023 : 361 - 376 two children already. On the other hand, women who have not had children may be less likely to choose the progestin subdermal implant because they may be concerned about the impact of the implant on their fertility in the future. They may also be more likely to choose other forms of contraception that are less invasive or easier to discontinue if they decide to become pregnant. Another aspect to consider would be the side effects that the user will experience (Olaifa et al., 2022) The type of family planning used prior to the progestin subdermal implant may also play a role in its use. Rademacher et al. (2013) mentioned that only less than one percent of women use hormone-releasing implants. Women who have used other forms of contraception in the past, such as oral contraceptives or condoms, may be more likely to choose the implant because they are looking for a more reliable form of contraception. Women who have not used any form of contraception in the past may be less likely to choose the implant because they may be less familiar with long-acting reversible contraception and its benefits. In a previous study by Mayoche (2022), the trend in the method of contraception among teenagers showed an increased increasing use of male condoms and oral contraceptives. Furthermore, alternative birth control options for teenagers use injectable contraception. p g j p The Chi-square computation of the study also revealed a significant relationship between the demographic characteristics and the reason for PSI use among the respondents. There was substantial evidence of an association between the selected demographic characteristics of the respondents and the reason for PSI use. Rivera, Asio Also, women who are planning to have a baby in the near future may be less likely to choose the PSI because of its long-acting nature. Women who have had a cesarean section delivery may be more likely to choose the PSI because it can be inserted immediately after delivery. This idea is parallel to the perspective of Olaifa et al. (2022), where the desire to conceive is relevant among the participants of their study. p p y It is important to remember that the choice of contraceptive method is a personal decision that should be made in consultation with a healthcare provider based on individual needs and preferences. While these factors may influence the use of the PSI, each woman's situation is unique and should be evaluated on a case-by-case basis. CONCLUSIONS AND RECOMMENDATIONS Based on the result of the study, the proponents hereby concluded that the characteristics of the respondents were aged 31-35 years old, housewives, high school graduate, Roman Catholic in faith, located in Bataan area, with monthly income between 5,000-9,000 pesos, with four living children, with no contraception used, with moderate menstrual flow, attended a mother’s class session, single, trying birth spacing, with normal spontaneous delivery, and no longer want to have a baby. There were significant differences in the reason for respondents' use of Progestin Subdermal Implant (PSI) in terms of age, occupation, educational attainment, number of living children, the type of family planning prior to PSI, the reason for practicing family planning, type of delivery, and plan to have another baby. There was a significant relationship between the reason respondents use Progestin Subdermal Implants (PSI) with age, occupation, educational attainment, number of living children, type of family planning prior to PSI, the reason for practicing family planning, type of delivery and plan to have another baby. y Based on the results and conclusion of the study, the researchers hereby recommend that healthcare workers (e.g. midwives or nurses) provide relevant and timely health education among couples regarding the appropriate and suitable family planning method for them. Also, users should follow check-up routines and monitoring of untoward or unpleasant side-effects to those women using PSI. Healthcare providers should offer health teaching related to management of side-effects and warning signs related to PSI use. At the same time, users should be acquainted with reminders and assessments of using PSI for optimal results. Finally, healthcare providers should encourage health promotion for PSI use and other relevant measures pertaining to the use of the implant. DISCUSSION The study further found that age, occupation, educational attainment, number of living children, type of family planning used prior to the implant, the reason for practicing family planning, plans to have a baby and type of birth delivery were associated with the reason for PSI use. The current study coincides with the past study of Abasiattai et al. (2014); the age group of women that accepted implants were 30-34 years old. In terms of occupation and educational attainment, Madugu et al. (2015) provided both disagreement and support to the current result of the study. Women who have not had children may be less likely to choose the PSI because they may be concerned about the impact of the implant on their fertility in the future. A meta-analysis study by Riedel et al. in 2020 concluded that PSIs have the potential to increase contraception among women. Women who have used other forms of contraception in the past, such as oral contraceptives or condoms, may be more likely to choose the PSI because they are looking for a more reliable form of contraception as well as long duration, reversibility and efficacy (Bachorik et al., 2015). Women who are practicing family planning to prevent pregnancy may be more likely to choose the PSI because of its high effectiveness rate. Since PSI is a modern family planning method, increased use was high among women (Abdalla, 2020). Women who are practising family planning for other reasons, such as to regulate their menstrual cycle, may be less likely to choose the PSI. However, early discontinuation due to unacceptable bleeding needs pre- insertion counselling (Weisberg et al., 2014). 371 ADVANCED RESEARCH In the process of writing this research, the proponents foresee several limitations, just like other research articles available out there in the field. First, the respondents for the study are pretty challenging to achieve since family planning is a choice done by a couple. Second, the duration of gathering enough data is painstakingly long and needs a lot of patience and following-up. Another one was the proximity of the study, wherein some neighbouring provinces also 372 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 participated in the study. The study is part of a community extension service program; thus, information dissemination is also equally important and extensive. Additionally, in terms of technical writing, there are still some flaws within the manuscript that may be overseen by a professional proofreader. Thus, a constructive criticism is very much welcomed by the proponents for future elaboration and improvement of the entire manuscript. ACKNOWLEDGMENT The proponents express their sincerest gratitude to all of the respondents who voluntarily participated in the study. Also, to the Population Commission (POPCOM) of the City of Olongapo for the support to the said activity to Gordon College Community Extension Service Unit. 373 Rivera, Asio REFERENCES Abasiattai, A.M., Utuk, N.M., & Inyang-Etoh, E.C. (2014). Subdermal contraceptive implants: Profile of acceptors in a tertiary hospital in Southern Nigeria. International Journal of Gynecology, Obstetrics and Neonatal Care, 1(1), 9-13. https://cosmosscholars.com/phms/index.php/ijgonc/article/view/72 Abdalla, K. K. (2020). Prevalence of and factors associated with modern contraceptive use among female sex workers in Dar Es Salaam, Tanzania. Muhimbili University of Health and Allied Sciences Institutional Repository. http://dspace.muhas.ac.tz:8080/xmlui/handle/123456789/2850 Bachorik A, Friedman J, Fox A, Nucci AT, Horowitz CR, & Diaz A. (2015). Adolescent and young adult women's knowledge of and attitudes toward etonogestrel implants. Journal of Pediatric and Adolescent Gynecology, 28, 229– 233. https://doi.org/10.1016/j.jpag.2014.08.002 Balogun, O.R., Olaomo, N., Adeniran, A.S., & Fawole, A.A. (2014). Implanon sub- dermal implant: an emerging method of contraception in Ilorin, Nigeria. Journal of Medical and Biomedical Sciences, 3(1), 1-5. https://doi.org/10.4314/jmbs.v3i1.1 Britton, L. E., Alspaugh, A., Greene, M. Z., & McLemore, M. R. (2020). CECE: An Evidence-Based Update on Contraception. The American Journal of Nursing, 120(2), 22–33. https://doi.org/10.1097/01.NAJ.0000654304.29632.a7 Edwards, M., & Can, ASAS (2023). Progestin. StatPearls. https://www.ncbi.nlm.nih.gov/books/NBK563211/ Fiato, Natalie (2016) Family planning in Sub-Saharan Africa: a review of interventions in the promotion of long-acting reversible contraception. Master's Thesis, University of Pittsburgh. (Unpublished). http://d- scholarship.pitt.edu/id/eprint/27371 Inthavong, S., Pantasri, T., Morakote, N., Muangmool, T., Piyamongkol, W., Pongsatham S., & Chaovisitseree, S. (2022). Change of contraceptive preference after the free LARC program for Thai teenagers. BMC Women's Health, 22, 211. https://doi.org/10.1186/s12905-022-01797-9 Jacobstein, R. (2018). Liftoff: The blossoming of contraceptive implant use in Africa. Global Health: Science and Practice, 6(1), 17-39. https://doi.org/10.9745/GHSP-D-17-00396 374 Indonesian Journal of Applied and Industrial Sciences (ESA) Vol. 2, No. 4 2023 : 361 - 376 Jacobstein, R., & Polis, C.B. (2014). Progestin-only contraception: Injectables and implants. Best Practice & Research Clinical Obstetrics & Gynaecology, 28(6), 795-806. https://doi.org/10.1016/j.bpobgyn.2014.05.003 Kolawole, O.O., Sowemimo, O.O., Ojo, OOOO, & Fasubaa, O.B. (2018) Contraceptive implants: A review and current perspective in southwest Nigeria. Tropical Journal of Obstetrics and Gynaecology, 35, 108-12. https://doi.org/10.4103/TJOG.TJOG_6_18 Krogstad, E.A., Odhiambo, O.K., Ayallo, M., Bailey, V.C., Rees, H., & van der Straten, A. (2019). Contraceptive implant uptake in Kenya versus South Africa: Lessons for new implantable technologies. Contraception, 101, 220- 225. https://doi.org/10.1016/j.contraception.2020.01.001 Liu, S., Kciuk, O., Frank, M., & Tyson, N. (2022). Progestin today and tomorrow. Current Opinion in Obstetrics and Gynecology, 34(6), 344-350. https://doi.org/10.1097/GCO.0000000000000819 Madugu, N., Abdul, M., Bawa, U., & Kolawole, B. (2015). Uptake of hormonal implants contraceptive in Zaria, Northern Nigeria. REFERENCES Open Journal of Obstetrics and Gynecology, 5, 268-273. https://doi.org/10.4236/ojog.2015.55039. Mayoche, Justin. (2022). A study of barriers and facilitators of access to family planning services by adolescent girls in selected secondary schools in Lukulu District. Cavendish Repository. http://155.0.3.194:8080/jspui/ Mitchell, V.E., & Welling, L.L.M. (2020). Not all Progestins are created equally: Considering unique Progestins individually in psychobehavioral research. Adaptive Human Behavior and Physiology, 6, 381–412. https://doi.org/10.1007/s40750-020-00137-1 Olaifa, B. T., Okonta, H. I., Mpinda, J. B., & Govender, I. (2022). Reasons were given by women for discontinuing the use of progestogen implants at Koster Hospital, North West province. South African Family Practice: Official Journal of the South African Academy of Family Practice/Primary Care, 64(1), e1– e7. https://doi.org/10.4102/safp.v64i1.5471 Quijencio, Jr., W.D. (2021). Variations in the Clinico-demographic profile among progestin subdermal implant acceptors. Globus International Journal of Medical Science, Engineering & Technology, 10(2), 11-16. https://doi.org/10.46360/globus.met.320212003 Rademacher, K. H., Vahdat, H. L., Dorflinger, L., Owen, D. H., & Steiner, M. J. (2013). Global introduction of a low-cost contraceptive implant. Critical Issues in Reproductive Health, 33, 285–306. https://doi.org/10.1007/978-94- 007-6722-5_14 375 Rivera, Asio Riedel, E.M.C., Turner, D.T., Kobeissi, L.H., Karyotaki, E., Say, L., & Cuijpers, P. (2020). The impact of psychosocial interventions on condom and contraceptive use in LMICs: Meta-analysis of randomized controlled trials, Global Public Health, 15(8), 1182- 1199. https://doi.org/10.1080/17441692.2020.1744679 Sah S., Jaiswal., A., & Paul, P. (2018). Current status of contraceptive use. World Journal of Pharmaceutical Research, 7(14), 320-331. https://wjpr.s3.ap-south- 1.amazonaws.com/article_issue/1531725186.pdf Santiago, M.S., Pastrana, M.L. (2022). Prevalence and determinants of long-acting reversible contraception initiation among teenage mothers in a tertiary hospital. Philippine Journal of Obstetrics and Gynecology, 46, 210-217. https://doi.org/10.4103/pjog.pjog_39_22 Van, V.T.S., Uy, J., Bagas, J., & Ulep, V.G.T. (2021). Trends in national-level governance and implementation of the Philippines’ Responsible Parenthood and Reproductive Health Law from 2014 to 2020. Global Health: Science and Practice, 9(3), 5348-564. https://doi.org/10.9745/GHSP-D-21- 00184 Waris, N., & Fatima, R.M. (2022). Frequency of various side effects and early removal of progesterone implant among females attending family planning clinics. Pakistan Journal of Medical & Health Sciences, 16(1), 637-638. https://doi.org/10.53350/pjmhs22161637 Weisberg, E., Bateson, D., McGeechan, K., & Mohapatra, L. (2014) A three-year comparative study of continuation rates, bleeding patterns and satisfaction in Australian women using a subdermal contraceptive implant or progestogen releasing-intrauterine system. The European Journal of Contraception & Reproductive Health Care, 19(1), 5-14. https://doi.org/10.3109/13625187.2013.853034 376
https://openalex.org/W3134229349	https://www.relea.ufscar.br/index.php/relea/article/download/492/453	Portuguese	null	ATIVIDADES INVESTIGATIVAS NA FORMAÇÃO INICIAL DE PROFESSORES EM ASTRONOMIA: O MOVIMENTO APARENTE DO SOL NO CÉU E A DURAÇÃO DOS DIAS E NOITES	Revista latino-americana de educação em astronomia	2,021	cc-by	17,578	DOI: 10.37156/RELEA/2020.30.083 DOI: 10.37156/RELEA/2020.30.083 ACTIVIDADES DE INVESTIGACIÓN EN LA FORMACIÓN INICIAL DE PROFESORES EN ASTRONOMÍA: EL MOVIMIENTO APARENTE DEL SOL EN EL CIELO Y LA DURACIÓN DE LOS DÍAS Y LAS NOCHES Resumen: En el presente trabajo, reportamos la aplicación de dos actividades de investigación sobre el movimiento aparente del Sol en el cielo, durante un día y durante todo el año, que se llevó a cabo con estudiantes de pregrado en Ciencias Biológicas - futuros maestros de Ciencias de la Escuela Primaria - que participaron en un curso corto de entrenamiento inicial en Astronomía. La primera actividad, sobre el movimiento diario del Sol, se implementó a través de una demostración de investigación, realizada por el maestro, utilizando el software Stellarium. La segunda, sobre la variación de la duración del día a lo largo del año, se realizó en forma de un laboratorio abierto, en el que los propios estudiantes recopilaron los datos, manipulando el Stellarium. Las actividades presentaron una buena participación e interacción de los estudiantes entre ellos, con el maestro y con la herramienta Stellarium, y parecen haber promovido no solo el aprendizaje de conceptos, sino también la experiencia de procedimientos y actitudes características de una investigación científica. Palabras clave: Actividad Investigativa; Enseñanza por Investigación; Enseñanza de la Astronomía; Movimiento Aparente del Sol; Días y Noches; Stellarium. ATIVIDADES INVESTIGATIVAS NA FORMAÇÃO INICIAL DE PROFESSORES EM ASTRONOMIA: O MOVIMENTO APARENTE DO SOL NO CÉU E A DURAÇÃO DOS DIAS E NOITES Wesley Quintiliano Vidigal 1 Sérgio Mascarello Bisch 2 Resumo: No presente trabalho, relatamos a aplicação de duas atividades investigativas sobre o movimento aparente do Sol no céu, durante um dia e ao longo do ano, que foi realizada com estudantes de licenciatura em Ciências Biológicas – futuros professores de Ciências do Ensino Fundamental – que participaram de um minicurso de formação inicial em Astronomia. A primeira atividade, sobre o movimento diário do Sol, foi implementada por meio de uma demonstração investigativa, conduzida pelo professor, utilizando o software Stellarium. A segunda, sobre a variação da duração do dia ao longo do ano, foi conduzida na forma de um laboratório aberto, em que os próprios estudantes coletaram os dados, manipulando o Stellarium. As atividades apresentaram uma boa participação e interação dos alunos entre si, com o professor e com a ferramenta Stellarium, e os dados indicam que além da aprendizagem de conceitos, estas proporcionaram aos estudantes uma vivência de procedimentos e atitudes característicos de uma investigação científica. Palavras-chave: Atividade Investigativa; Ensino por Investigação; Ensino de Astronomia; Movimento Aparente do Sol; Dias e Noites; Stellarium. p y@ 2 Universidade Federal do Espírito Santo (UFES), Espírito Santo, Brasil. E-mail: sergiobisch@gmail.com 1 Secretaria de Estado da Educação, Espírito Santo, Brasil. E-mail: fisica.wesley@hotmail.com. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 1 Introdução A inserção de atividades investigativas no ensino de Ciências na Educação Básica como forma de promover um ensino em que o estudante tenha um papel ativo, de sujeito na construção de seu conhecimento, praticando a argumentação, a formulação e teste de hipóteses, a comunicação e sistematização de conhecimentos, vem sendo abordada e defendida por diversos pesquisadores da área da Educação em Ciência (Azevedo, 2006; Zômpero & Laburú, 2011, 2012; Carvalho, 2013, 2018; Carvalho & Gil-Pérez, 2011). Segundo esses autores, esse tipo de atividade promoveria não apenas uma aprendizagem de conceitos, mas também o desenvolvimento de habilidades cognitivas, de atitudes e procedimentos científicos. Essa inserção também é fortemente recomendada em documentos oficiais, como nas Diretrizes Curriculares Nacionais da Educação Básica (Brasil, 2013) e na Base Nacional Comum Curricular (BNCC) (Brasil, 2018), segundo a qual é imprescindível que os alunos: [...] sejam progressivamente estimulados e apoiados no planejamento e na realização cooperativa de atividades investigativas, bem como no compartilhamento dos resultados dessas investigações. Isso não significa realizar atividades seguindo, necessariamente, um conjunto de etapas predefinidas, tampouco se restringir à mera manipulação de objetos ou realização de experimentos em laboratório. Ao contrário, pressupõe organizar as situações de aprendizagem partindo de questões que sejam desafiadoras e, reconhecendo a diversidade cultural, estimulem o interesse e a curiosidade científica dos alunos e possibilitem definir problemas, levantar, analisar e representar resultados; comunicar conclusões e propor intervenções (Brasil, 2018, p. 322). [...] sejam progressivamente estimulados e apoiados no planejamento e na realização cooperativa de atividades investigativas, bem como no compartilhamento dos resultados dessas investigações. Isso não significa realizar atividades seguindo, necessariamente, um conjunto de etapas predefinidas, tampouco se restringir à mera manipulação de objetos ou realização de experimentos em laboratório. Ao contrário, pressupõe organizar as situações de aprendizagem partindo de questões que sejam desafiadoras e, reconhecendo a diversidade cultural, estimulem o interesse e a curiosidade científica dos alunos e possibilitem definir problemas, levantar, analisar e representar resultados; comunicar conclusões e propor intervenções (Brasil, 2018, p. 322). De igual modo, a abordagem de temas e conteúdos de Astronomia na Educação Básica, associada à área das Ciências da Natureza, também é sugerida por documentos oficiais. INVESTIGATIVE ACTIVITIES IN THE INITIAL TRAINING OFTEACHERS IN ASTRONOMY: THE APPARENT MOVEMENT OF THE SUN IN THE SKY AND THE DURATION OF DAYS AND NIGHTS Abstract: In the present work, we report the application of two investigative activities on the apparent movement of the Sun in the sky, during one day and throughout the year, which was carried out with 1 Secretaria de Estado da Educação, Espírito Santo, Brasil. E-mail: fisica.wesley@hotmail.com. 83 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch undergraduate students in Biological Sciences - future teachers of Elementary School Sciences - who participated in a short course of initial training in Astronomy. The first activity, on the daily movement of the Sun, was implemented through an investigative demonstration, conducted by the teacher, using the Stellarium software. The second, on the variation of the length of the day throughout the year, was conducted in the form of an open laboratory, in which the students themselves collected the data, manipulating the Stellarium. The activities presented a good participation and interaction of the students among themselves, with the teacher and with the Stellarium tool, and seem to have promoted not only the learning of concepts, but also the experience of procedures and attitudes characteristic of a scientific investigation. Keywords: Investigative Activity; Inquiry-based Learning; Astronomy Teaching; Apparent Movement of the Sun; Days and Nights; Stellarium. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 1 Introdução A BNCC especifica alguns dos temas básicos de Astronomia a serem abordados ao longo do Ensino Fundamental, desde os anos iniciais, tais como: o movimento aparente do Sol no céu, a observação do céu, características da Terra (como seu formato esférico) e pontos cardeais (Brasil, 2018, p. 334-338). Cabe ressaltar que, nos Parâmetros Curriculares Nacionais (PCN) (Brasil, 1998), a temática “Terra e Universo” já era considerada como um dos quatro eixos temáticos para o ensino das Ciências Naturais e, atualmente, compõe uma das três unidades temáticas da BNCC, a partir das quais deve se desenvolver o ensino de Ciências na Educação Básica. 84 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Nesse sentido, durante a realização de um minicurso de formação inicial em Astronomia para estudantes de licenciatura em Ciências Biológicas – futuros professores de Ciências do Ensino Fundamental – da Universidade Federal do Espírito Santo, foram propostas e implementadas duas atividades de natureza investigativa sobre o movimento aparente do Sol com relação ao horizonte local, a variação desse movimento conforme a latitude do observador, os pontos cardeais e a variação da duração do dia e da noite ao longo do ano. Ao incorporar essas atividades num minicurso de formação inicial, além de promover uma aprendizagem conceitual, a intenção foi contribuir para que os futuros professores tenham maior conhecimento e experiência acerca das características, possibilidades e potencialidades desse tipo de abordagem, de um ensino por investigação, e a utilizem, futuramente, com seus próprios alunos do Ensino Fundamental. Essas atividades fizeram parte da aplicação piloto de uma proposta didática desenvolvida num projeto de mestrado profissional em Ensino de Física, que teve como objetivo a elaboração, aplicação e avaliação de um minicurso para a formação inicial de professores de Ciências do Ensino Fundamental em Astronomia (Vidigal, 2019), que buscou incluir diversas atividades investigativas. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 1 Introdução A primeira das duas atividades aqui relatadas – sobre o movimento aparente do Sol com relação ao horizonte local – pode ser caracterizada como sendo do tipo “demonstração investigativa” (Azevedo, 2006), na qual um problema, em geral apresentado na forma de uma pergunta, é inicialmente proposto à classe pelo professor e, a seguir, é dada liberdade para que os alunos explicitem suas concepções e hipóteses iniciais, sejam intuitivas ou de senso comum, acerca do mesmo, sendo realizada uma discussão em grupo e/ou com a classe toda, na qual os alunos possam praticar suas habilidades de reflexão, argumentação e comunicação, típicas de procedimentos científicos. Posteriormente, por meio da apresentação, pelo professor, de modelos ou experimentos, eventualmente subsidiados por consultas ou pesquisas bibliográficas, é conduzida uma investigação a respeito do fenômeno e de sua explicação científica, cabendo ao professor atuar como um guia, auxiliando na construção da passagem do saber cotidiano ao saber científico. A segunda atividade – sobre a variação da duração do dia e da noite ao longo do ano –, por sua vez, teve características semelhantes às de um “laboratório aberto” (Azevedo, 2006), no qual também tudo se inicia pela proposição de um problema, seguida de uma etapa de expressão e discussão de concepções e hipóteses iniciais dos alunos acerca dele, com a posterior elaboração de um plano de trabalho para a realização do experimento, seguida da realização do experimento, pelos próprios alunos, e a posterior análise dos dados colhidos, possibilitando um confronto com as hipóteses inicialmente levantadas pelos alunos e a obtenção de conclusões. Em ambas as atividades, a ferramenta utilizada para realizar a demonstração acerca do movimento do Sol com relação ao horizonte, no primeiro caso, e de um experimento virtual sobre a variação na duração do dia e da noite durante o ano, no segundo caso, foi o software livre Stellarium (Stellarium.org, 2020), que permite simulações realistas destes fenômenos. Com ele se pode simular o nascimento e ocaso do Sol e sua trajetória diurna no céu para qualquer posição do observador na superfície da Terra, bem como se observar a posição de nascimento e ocaso do Sol com relação aos pontos cardeais leste e oeste e o horário em que eles ocorrem. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Ressaltamos que os temas astronômicos abordados nessas atividades se referem a fenômenos que fazem parte do cotidiano dos alunos e são observados a partir de um referencial local, topocêntrico, ou seja, se referem ao movimento do Sol no céu conforme observado da superfície da Terra. Os problemas propostos nas atividades aqui relatadas se referem a uma descrição fenomenológica de como se dá esse movimento a partir deste referencial e sua relação com a posição do observador na superfície da Terra. Não foram abordadas, diretamente, explicações que utilizam um referencial heliocêntrico, envolvendo uma representação da Terra como ela seria vista a partir do espaço, num referencial em repouso em relação ao Sol. A abordagem de fenômenos astronômicos a partir de um referencial local é a mais indicada para alunos dos anos iniciais do Ensino Fundamental, uma vez que se acha mais próxima do que pode ser diretamente observado, do vivencial e concreto, enquanto que a abordagem por meio da explicação desses mesmos fenômenos em termos dos movimentos da Terra, conforme observados a partir de um referencial heliocêntrico, bem mais abstrata e conceitual (Leite, 2006), é mais indicada para os anos finais do Ensino Fundamental, nos quais um pensamento mais descentrado e abstrato já se faz presente e é mais acessível aos alunos (Bisch, 1998; Langhi & Nardi, 2010). O ponto de vista a partir de um referencial local se acharia próximo do que poderíamos denominar de um polo vivencial, realista ingênuo, já o ponto de vista heliocêntrico, que implica numa visão da Terra a partir do espaço, estaria mais próximo de um polo conceitual (Bisch, 1998). De maneira coerente com essas observações, notamos que, na BNCC, os temas “movimento aparente do Sol no céu”, “observação do céu” e “pontos cardeais” são indicados para serem abordados, respectivamente, no 2º, 3º e 4º anos (Brasil, 2018, p. 334-338), enquanto que os temas “Movimento de rotação da Terra”, “Forma, estrutura e movimentos da Terra” e “Sistema Sol, Terra e Lua”, que envolvem a utilização de um referencial heliocêntrico, só são indicados, respectivamente, para o 5º, 6º e 8º anos (Brasil, 2018, p. 340-348). Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 85 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch 2 Embora seja habitual denominar o movimento do Sol, e dos demais astros, com relação a um referencial local, baseado no plano do horizonte, como sendo um movimento “aparente”, é importante notar que, conforme se aprende em Física, todo movimento sempre é relativo, isso é depende do referencial utilizado para observá-lo. Nesse sentido, o movimento observado do Sol, com relação a um referencial local, nada tem de aparente no sentido de algo ilusório, mas sim corresponde ao movimento real descrito pelo Sol neste referencial. Apenas ele não é um referencial inercial. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Algumas pesquisas que discutem a questão da espacialidade no ensino de Astronomia e na formação continuada de professores (Bisch, 1998; Leite, 2006; Ferreira & Leite, 2015) indicam que a coordenação entre esses dois pontos de vista, o geocêntrico e o heliocêntrico, o vivencial e o conceitual, não é nada trivial, e que isso deve ser levado em consideração em seu ensino. Em atividades de formação inicial de professores, portanto, também será relevante que ambas as perspectivas – a descrição do que é observado no céu, a partir de um referencial local, e a sua explicação em termos do um modelo heliocêntrico – sejam abordadas e sua articulação e complementaridade sejam investigadas e compreendidas. Na proposta didática de um minicurso de formação inicial, da qual as duas atividades investigativas aqui relatadas fizeram parte (Vidigal, 2019), o ponto de vista heliocêntrico foi trabalhado em outras atividades, nas quais foram utilizados modelos tridimensionais. A relação dessa perspectiva com o ponto de vista geocêntrico foi parcialmente abordada, devendo ainda ser aperfeiçoada e aprofundada em aplicações futuras da proposta. Acreditamos que, idealmente, as atividades aqui relatadas seriam adequadas para inclusão numa primeira etapa de uma proposta de ensino que partisse do vivencial, do que é diretamente observado no céu, a partir de um referencial local, que 86 86 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites prosseguiria e seria complementada com o estudo dos movimentos da Terra vistos de um referencial heliocêntrico, a partir do qual seria possível explicar os movimentos “aparentes” do Sol com relação ao horizonte, abordados na primeira etapa. Após uma descrição razoavelmente detalhada acerca de como ocorre o movimento aparente do Sol no céu, apresentada na próxima seção, nas seguintes é apresentado um relato da elaboração e aplicação das duas atividades investigativas no referido minicurso, seguido de uma análise dos dados colhidos quanto ao envolvimento, participação e aprendizagem dos estudantes. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 2 O movimento aparente do Sol no céu Para um observador situado num ponto da superfície da Terra, o movimento que o Sol descreve no céu com relação ao referencial local, no qual este observador se encontra, delimitado pelo seu plano do horizonte, pode ser decomposto em dois movimentos principais: um movimento diário e um movimento anual do Sol. Esses movimentos costumam ser denominados “aparentes” como forma de contrastá-los aos movimentos considerados “reais”, descritos com relação a um referencial inercial situado no espaço, em repouso com relação ao Sol, ou seja, em um referencial heliocêntrico, aos quais esses dois movimentos aparentes se acham, respectivamente, associados, a saber: o movimento de rotação da Terra em torno de um eixo que passa pelo seu centro e seu movimento de translação em torno do Sol2. Para a explicação que será dada a seguir, acerca de como ocorrem esses dois movimentos observados de um referencial local, será importante representar o céu utilizando o conceito de esfera celeste. Ele é um conceito que possui raízes antigas, com origem numa concepção geocêntrica de universo, na qual, como os antigos gregos imaginavam, se considerava que as estrelas se achavam incrustadas, em posições fixas, em uma grande superfície esférica cristalina que circundava a Terra (Bisch, 2012; Oliveira Filho & Saraiva, 2014). Atualmente, sabemos muito bem que não existe essa superfície esférica, que o céu tem profundidade, que as estrelas se situam a diferentes distâncias, mas, ainda assim, a versão atual do conceito de esfera celeste é extremamente útil e muito usada para indicar a posição (direção) em que se encontram os astros no céu, independentemente de sua distância à Terra, por meio de sistemas de coordenadas definidos sobre a esfera celeste. De fato, a esfera celeste é um dos conceitos centrais utilizados na área de estudos denominada “Astronomia de Posição” ou “Astrometria” (Caniato, 1993; Boczko, 1984), dedicada à medida precisa da posição e movimento dos astros, área que pode ser considerada a mais antiga da Astronomia. Segundo a concepção atual, a esfera celeste é definida como sendo uma superfície esférica abstrata, de raio arbitrariamente grande, centrada na Terra, ou no 87 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch observador em sua superfície3, na qual podemos imaginar projetados todos os objetos celestes, em relação à qual as estrelas mais distantes estão paradas e cujo movimento, em relação à superfície da Terra, corresponde ao movimento destas estrelas fixas (Bisch, 2012). 2 O movimento aparente do Sol no céu Conforme indicado na Figura 1, os pontos em que o prolongamento do eixo de rotação da Terra “fura” a esfera celeste, que também podem ser pensados com a projeção dos polos geográficos da Terra no céu, ou seja, na esfera celeste, correspondem aos polos celestes, sul e norte. A 90o dos polos celestes temos o equador celeste, que também pode ser pensado como a projeção do equador da Terra na esfera celeste. Mais detalhes sobre o conceito de esfera celeste e sua aplicação na descrição da posição e movimentos dos astros podem ser consultados em Oliveira Filho e Saraiva (2014) e Boczko (1987). Modelos físicos didáticos para representar a esfera celeste e como se pode, com eles, simular os movimentos das estrelas e do Sol no céu, conforme vistos de um referencial local, são apresentados por Caniato (1993) e Bedaque e Bretones (2016). Figura 1 - A esfera celeste segundo o conceito atual: uma esfera imaginária, centrada na Terra, na qual as estrelas distantes parecem fixas e que gira em torno de um eixo de rotação que é o prolongamento do eixo de rotação da Terra. Os polos celestes e o equador celeste são as projeções, na esfera celeste, dos polos e do equador terrestres. Na figura, a esfera celeste, esquematicamente representada por uma esfera de cor cinza, apresenta um tamanho um pouco maior que o da Terra, em azul, mas, conceitualmente, a esfera celeste possui um raio arbitrariamente grande, que tende ao infinito, de modo que a Terra e um observador em sua superfície podem ser considerados apenas como pontos em seu centro. Fonte: Saraiva, Oliveira Filho e Muller (2014). Figura 1 - A esfera celeste segundo o conceito atual: uma esfera imaginária, centrada na Terra, na qual as estrelas distantes parecem fixas e que gira em torno de um eixo de rotação que é o prolongamento do eixo de rotação da Terra. Os polos celestes e o equador celeste são as projeções, na esfera celeste, dos polos e do equador terrestres. Na figura, a esfera celeste, esquematicamente representada por uma esfera de cor cinza, apresenta um tamanho um pouco maior que o da Terra, em azul, mas, conceitualmente, a esfera celeste possui um raio arbitrariamente grande, que tende ao infinito, de modo que a Terra e um observador em sua superfície podem ser considerados apenas como pontos em seu centro. Fonte: Saraiva, Oliveira Filho e Muller (2014). 3 Na maioria das situações, como o tamanho do raio terrestre pode ser considerado desprezível em relação às distâncias até os astros, pode-se considerar que a posição em que os astros são vistos no céu, por um observador na superfície da Terra, é a mesma que seria vista a partir do seu centro: ambas as posições podem ser, portanto, consideradas equivalentes e localizadas no centro da esfera celeste. Isso só não é inteiramente válido no caso de objetos mais próximos, dentro do sistema solar, como a Lua. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 2 O movimento aparente do Sol no céu Devido à rotação da Terra em torno de seu próprio eixo, no sentido de oeste para leste, um observador na superfície da Terra vê a esfera celeste girar em relação ao plano do horizonte, de leste para oeste, com as estrelas nascendo no lado leste e se pondo no lado oeste, mantendo suas posições relativas fixas, descrevendo o chamado movimento diário (ou diurno) da esfera celeste. Algumas estrelas, mais próximas ao 88 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites polo celeste visível, denominadas estrelas circumpolares (Boczko, 1984, p. 38), não nascem nem se põem, por estarem muito próximas deste polo, girando em torno dele, sem jamais atingirem o horizonte de observação4 de uma pessoa na superfície da Terra, conforme ilustrado na Figura 2. Nessa figura é apresentada uma foto de longa exposição tirada a noite, próximo da cúpula do Telescópio Anglo-Australiano, localizado no Siding Spring Observatory, na Austrália, com uma câmera fotográfica apontada para o céu junto ao horizonte sul, mantendo seu obturador aberto durante várias horas. A imagem pode ser considerada um registro do movimento diário da esfera celeste. Os arcos luminosos circulares e concêntricos, que aparecem na foto, são as trilhas luminosas deixadas pelas estrelas em seus movimentos circulares diários, em torno do polo sul celeste, cuja posição no céu corresponde à da direção para onde aponta o eixo de rotação da Terra. Trilhas deixadas por diversas estrelas circumpolares podem ser percebidas na imagem. Figura 2 - Um registro do movimento diário da esfera celeste. Fotografia de longa exposição. Fonte: David Malin, Austrália. Disponível em: https://images.datacentral.org.au/malin/AAO2/001 Figura 2 - Um registro do movimento diário da esfera celeste. Fotografia de longa exposição. Fonte: David Malin, Austrália. Disponível em: https://images.datacentral.org.au/malin/AAO2/001 Fonte: David Malin, Austrália. Disponível em: https://images.datacentral.org.au/malin/AAO2/001 O movimento diário do Sol consiste em acompanhar o movimento diário da esfera celeste, produzido pela rotação da Terra, nascendo a leste, descrevendo um arco sobre nossas cabeças, e se pondo a oeste, gerando assim os dias e noites, fenômeno que dita o ritmo de nossas vidas. 4 No presente texto, estamos sempre considerando um horizonte ideal, totalmente desimpedido, como o seria para um observador situado no meio de um mar calmo, ou de uma grande planície. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 2 O movimento aparente do Sol no céu Contudo, diferentemente das estrelas – que aparentam estarem fixas umas em relação às outras, devido à sua imensa distância até nós – o Sol também descreve um movimento anual, apresentando, a cada dia, um pequeno deslocamento angular, de 4 No presente texto, estamos sempre considerando um horizonte ideal, totalmente desimpedido, como o seria para um observador situado no meio de um mar calmo, ou de uma grande planície. 5 Como o Sol leva um ano para completar uma volta na esfera celeste, a uma velocidade angular aproximadamente constante, o ano corresponde a aproximadamente 365 dias e uma volta a um deslocamento angular de 360o, a cada dia o Sol percorre, portanto, cerca de 1o na esfera celeste. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 89 90 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Conforme indicado na Figura 4, nos instantes em que o Sol cruza o equador celeste, por volta dos dias 21 março e 23 de setembro, ocorrem os chamados equinócios. Nos pontos em que ele atinge o máximo afastamento, ao norte ou ao sul do equador, por volta de 22 de junho e 22 de dezembro, respectivamente, ocorrem os solstícios. Esses eventos astronômicos marcam o início das estações do ano, de maneira alternada, nos hemisférios norte e sul da Terra. Figura 4 - A linha da eclíptica – caminho descrito pelo Sol na esfera celeste – não coincide com o equador celeste. Metade desse caminho se situa ao norte, metade ao sul do plano do equador, o que acaba produzindo as estações do ano. Fonte: Saraiva, Oliveira Filho e Muller (2014). Figura 4 - A linha da eclíptica – caminho descrito pelo Sol na esfera celeste – não coincide com o equador celeste. Metade desse caminho se situa ao norte, metade ao sul do plano do equador, o que acaba produzindo as estações do ano. Fonte: Saraiva, Oliveira Filho e Muller (2014). Outra consequência do fato de a eclíptica apresentar trechos ao norte e ao sul do equador celeste é que, embora o deslocamento anual do Sol na esfera celeste ocorra predominantemente na direção leste-oeste, no sentido de oeste para leste, como já foi dito anteriormente, ele também apresentará um deslocamento na direção norte-sul: durante cerca de metade do ano o Sol se desloca de sul para o norte (desde o solstício de dezembro até o de junho) e, na outra metade (do solstício de junho até o de dezembro), do norte para o sul, num ciclo anual. Ele atinge um deslocamento angular máximo ao norte do equador no solstício de junho e, ao sul do equador, no solstício de dezembro. Ambos de cerca de 23,5o, o que significa uma variação total, entre um extremo e outro, na direção norte-sul da esfera celeste, de cerca de 47o. O que não é pouco e pode ser facilmente percebido se observarmos a natureza a nossa volta com atenção: você já notou como mudam, ao longo do ano, as posições de nascimento e ocaso do Sol junto ao horizonte? 89 89 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch cerca de um grau5, com relação às estrelas. A trajetória anual descrita pelo Sol na esfera celeste passa pelas constelações zodiacais e ocorre ao longo da chamada linha de eclíptica, ou simplesmente eclíptica, contida no plano da órbita da Terra em torno do Sol (Figura 3). A cada dia, devido ao seu movimento anual, o Sol se “atrasa” um pouco em relação às estrelas, movendo-se, predominantemente, na direção leste-oeste, no sentido de oeste para leste. Figura 3 - Movimento anual do Sol na esfera celeste ao longo da linha da eclíptica, passando pelas constelações zodiacais, visto de uma perspectiva em que o norte está para cima. Fonte: Varella, 2003. Figura 3 - Movimento anual do Sol na esfera celeste ao longo da linha da eclíptica, passando pelas constelações zodiacais, visto de uma perspectiva em que o norte está para cima. Fonte: Varella, 2003. Outro fato muito importante, que tem uma influência decisiva na forma como o Sol descreve seu movimento anual na esfera celeste, é o de que a Terra, assim como outros planetas do sistema solar, apresenta o seu eixo de rotação inclinado em relação à direção perpendicular ao plano da sua órbita em torno do Sol. No caso da Terra, inclinado cerca de 23,5o. Consequentemente, o plano do equador celeste também é inclinado em relação ao plano de eclíptica, os mesmos cerca de 23,5o, o que significa que a linha da eclíptica não coincide com o equador celeste, mas apresenta uma metade situada ao norte, outra ao sul do equador celeste (Figura 4). Isso gera outro fenômeno astronômico importantíssimo, que também regula a vida na Terra: as estações do ano. Podemos perceber que o movimento do Sol na eclíptica explica as estações do ano porque, se durante cerca de metade do ano ele se situa ao norte do plano do equador, nesta época ele iluminará e aquecerá mais o hemisfério norte terrestre, e menos o sul – ocorrerão, então, as estações da primavera e verão no hemisfério norte e de outono e inverno no hemisfério sul. Na outra metade do ano, o Sol estará ao sul do equador e iluminará e aquecerá mais o hemisfério sul da Terra, e menos o norte, alternando as estações do ano produzidas em cada hemisfério (Boczko, 1984, p. 128- 129). Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 89 Na Figura 5 é apresentada uma imagem ilustrando esse efeito. Conforme discutido por Bedaque e Bretones (2016), a variação da posição de nascimento e ocaso do Sol no horizonte, entre os extremos, nos solstícios de junho e dezembro, será de cerca de 47 o para observadores situados sobre o equador da Terra, e será maior do que isso em qualquer outra latitude. Em latitudes muito elevadas, acima dos círculos polares, o Sol, inclusive, pode nem nascer ou se pôr em determinados períodos do ano. 6 O hemisfério visível, acima do horizonte, é denominado hemisfério zenital, e, o que fica abaixo, hemisfério nadiral (Boczko, 1984, p. 40), em referência aos pontos zênite e nadir, que são as intersecções da vertical do lugar com a esfera celeste, o zênite sendo o ponto diretamente acima de nossas cabeças, o nadir o seu oposto, abaixo de nossos pés (Boczko, 1984, p. 28-29). 7 A altura de um ponto no céu corresponde ao ângulo entre a direção em que se encontra este ponto e o plano do horizonte do observador, medido sobre um plano vertical (Boczko, 1984, p. 41). Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 p , p ( , , p ) A altura de um ponto no céu corresponde ao ângulo entre a direção em que se encontra este ponto e o ano do horizonte do observador, medido sobre um plano vertical (Boczko, 1984, p. 41). 8 Devido à posição da eclíptica na esfera celeste, esse deslocamento, na direção norte-sul, será maior nas épocas próximas dos equinócios e praticamente nulo nos dias próximos aos solstícios. 91 92 92 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Para um observador situado no equador terrestre, como dito anteriormente, os polos celestes estarão localizados sobre o horizonte. O equador celeste, por sua vez, estará situado sobre um círculo vertical, que contém a linha leste-oeste (Fig. 6b). O movimento diário do Sol ocorrerá, então, de maneira semelhante ao das estrelas, sobre círculos verticais, perpendiculares ao plano do horizonte. A melhor representação da trajetória do Sol, após seu nascer no horizonte leste, corresponderá, portanto, a uma reta vertical, perpendicular ao horizonte leste, conforme indicado na Fig. 6c. Em verdade, rigorosamente falando, devido ao seu movimento anual na esfera celeste, percorrendo a linha da eclíptica, o Sol também apresentará, diariamente, um pequeno deslocamento com relação à esfera celeste, predominantemente na direção leste-oeste, mas também um pouco na direção norte-sul8. Contudo, esse último será praticamente desprezível durante um intervalo de tempo relativamente pequeno, como o de algumas poucas horas após o nascimento do Sol. Pode-se, assim, considerar que a melhor representação de sua trajetória com relação ao horizonte leste, ao nascer, será, efetivamente, a de uma reta perpendicular a este horizonte, conforme representado na Figura 6c. A faixa amarela mostrada nessa figura busca justamente representar a variação, ao longo do ano, na posição de nascimento do Sol e de sua trajetória diária junto ao horizonte leste, devido ao seu movimento anual na esfera celeste, que apresenta uma componente na direção norte-sul. Figura 6 - Representação do hemisfério celeste visível para um observador situado em um ponto do equador terrestre (Fig. 6a) e o arco descrito pela trajetória do Sol em seu movimento diário, para este observador, em uma perspectiva tridimensional (Fig. 6b), bem como sua representação bidimensional (Fig. 6c), na qual se considera o observador de frente para o horizonte leste. A letra “Z”, nas Figuras 6a e 6b, representa o zênite: ponto de intersecção entre a vertical local e a esfera celeste, exatamente acima da cabeça do observador. Figura 6 - Representação do hemisfério celeste visível para um observador situado em um ponto do equador terrestre (Fig. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 91 6a) e o arco descrito pela trajetória do Sol em seu movimento diário, para este observador, em uma perspectiva tridimensional (Fig. 6b), bem como sua representação bidimensional (Fig. 6c), na qual se considera o observador de frente para o horizonte leste. A letra “Z”, nas Figuras 6a e 6b, representa o zênite: ponto de intersecção entre a vertical local e a esfera celeste, exatamente acima da cabeça do observador. É importante notar também que, para esse observador no equador da Terra, metade da trajetória diária descrita pelo Sol (bem como pelas estrelas), em relação ao seu referencial local, ficará acima do horizonte e metade abaixo, em qualquer época do ano, pois o plano do horizonte dividirá exatamente ao meio esta trajetória circular aparente, o que significa que a duração do dia (Sol acima do horizonte) será igual à da noite (Sol abaixo do horizonte): 12 horas, ao longo de todo o ano. Além disso, à medida que a esfera celeste descreve seu giro diário, toda ela será visível para esse observador no equador. Ele terá acesso visual a todas as direções do Universo, para ele não haverá estrelas circumpolares, pois todas nascem e se põem (Bedaque & Bretones, 2016). 91 91 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Figura 5 - Variação da posição onde o Sol nasce ao longo do ano. Fonte: Bedaque e Bretones, 2016. Figura 5 - Variação da posição onde o Sol nasce ao longo do ano. Fonte: Bedaque e Bretones, 2016. Por fim, para subsidiar a discussão das atividades investigativas que serão relatadas na próxima seção, é importante considerar a relação existente entre a trajetória diária descrita pelo Sol no céu, em especial a sua inclinação em relação a uma direção normal ao horizonte, e a latitude do observador. Para tanto, primeiramente é fundamental notar que o plano do horizonte de um observador situado na superfície da Terra é sempre um plano tangente a esta superfície no ponto em que este observador se encontra. Esse plano corta a esfera celeste ao meio, dividindo-a em um hemisfério celeste visível e outro invisível, que fica abaixo da linha do horizonte6 (Figuras 6, 7 e 8). Devido a esse fato, é importante notar que, para um observador que estiver situado sobre o equador terrestre, os polos celestes estarão situados sobre a linha do horizonte (Figuras 6a, 6b), enquanto que, para qualquer observador situado em um dos hemisférios da Terra, apenas um dos polos celestes ficará acima do horizonte, o outro ficará abaixo: um observador no hemisfério norte verá apenas o polo celeste norte, enquanto o observador no hemisfério sul verá apenas o polo sul celeste acima do seu horizonte (Figuras 7a, 7b). Na fotografia apresentada na Figura 2, por exemplo, tirada na Austrália, que fica no hemisfério sul da Terra, o polo celeste visível, em torno do qual as estrelas aparecem girando (no sentido horário) é o sul. Também é possível demonstrar que a altura7 do polo celeste visível para um observador situado a uma latitude , ao norte ou ao sul do equador, será um ângulo exatamente igual a esta própria latitude (Boczko, 1984, p. 54), conforme indicado na Figura 7b. Vamos considerar três casos particulares, representados nas Figuras 6, 7 e 8, por meio dos quais buscamos sintetizar qual será a trajetória diária descrita pelo Sol no céu e como ocorre a variação na duração dos dias e noites para observadores situados em diversas localizações (latitudes) no planeta Terra. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Por outro lado, para um observador situado em um dos hemisférios, sul ou norte, da Terra, onde a latitude se situa entre 0o e 90o, o movimento do Sol se dará sob uma perspectiva diferente. No exemplo apresentado na Figura 7, para um observador localizado no hemisfério sul da Terra, numa latitude , o Sol descreverá um arco inclinado em relação ao plano do horizonte (Fig. 7b). Neste caso, uma representação bidimensional deste movimento pode ser descrita a partir de um segmento de reta inclinado em relação à normal ao plano do horizonte (Fig. 7c), cuja inclinação, assim como a altura do polo celeste visível, será igual ao valor da latitude do observador. Essa será a trajetória do Sol no céu para um observador situado, por exemplo, na cidade de Vitória, ES, uma vez que esta cidade está situada, aproximadamente, na latitude de 20° sul. Novamente, a faixa amarela na Figura 7c representa a variação na posição da trajetória diária do Sol ao longo do ano, devido ao seu movimento anual na esfera celeste. Figura 7 - Representação da esfera celeste visível para observador em algum ponto situado no hemisfério sul (Fig. 7a) e o arco descrito pela trajetória do Sol, para este observador, em uma perspectiva tridimensional (Fig. 7b), bem como sua representação bidimensional (Fig. 7c), na qual se considera o observador de frente para o horizonte leste. A letra “Z” representa o zênite e “PS” o polo sul celeste. Figura 7 - Representação da esfera celeste visível para observador em algum ponto situado no hemisfério sul (Fig. 7a) e o arco descrito pela trajetória do Sol, para este observador, em uma perspectiva tridimensional (Fig. 7b), bem como sua representação bidimensional (Fig. 7c), na qual se considera o observador de frente para o horizonte leste. A letra “Z” representa o zênite e “PS” o polo sul celeste. Quanto à duração dos dias e noites, note que agora o plano do horizonte não cortará os arcos diários descritos pelos astros na esfera celeste da mesma maneira, deixando metade acima, metade abaixo do horizonte, como no caso de um observador situado no equador. No exemplo apresentado na Figura 7, podemos perceber que, para um observador situado no hemisfério sul da Terra, astros situados ao sul do equador celeste terão mais de metade da sua trajetória diária acima do horizonte (Fig. 7b). 93 93 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Quanto mais ao sul, maior a porção da trajetória acima do horizonte. As estrelas bem próximas ao polo sul celeste, inclusive, terão suas trajetórias inteiramente acima do horizonte, ou seja, serão circumpolares. Por outro lado, quanto mais ao norte do equador, menor será a porção acima do horizonte. As estrelas mais próximas do polo norte celeste jamais serão visíveis, permanecendo sempre abaixo do horizonte. Para um observador que esteja no hemisfério norte da Terra, algo equivalente acontecerá: quanto mais ao norte o astro se encontrar na esfera celeste, maior será a fração da sua trajetória diária que ficará acima do horizonte, e, quanto mais ao sul, menor ela será. Apenas um astro que esteja exatamente sobre o equador celeste terá metade de sua trajetória diária acima, metade abaixo do horizonte, para observadores de qualquer latitude (à exceção dos polos!), nos hemisférios sul ou norte, pois o plano do horizonte sempre cortará o equador celeste ao meio, não importando a latitude do observador. 94 94 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites E o que isso significa se consideramos a duração dos dias e noites ao longo do ano? Ora, como vimos, devido ao seu movimento anual na eclíptica, o Sol se encontra cerca de metade do ano ao norte e metade ao sul do equador celeste. Portanto, para um observador no hemisfério sul da Terra, quando o Sol estiver ao sul do equador celeste – o que acontece a partir do equinócio de setembro até o de março –, mais de metade da sua trajetória diária ficará acima do horizonte, o que significa que o dia terá uma duração maior do que 12 horas, ou seja, os dias serão maiores que as noites! Como já foi dito anteriormente, isso também quer dizer que o Sol iluminará e aquecerá mais o hemisfério sul terrestre, e teremos as estações da primavera e verão neste hemisfério. Por outro lado, quando o Sol estiver ao norte do equador celeste, nas estações de outono e inverno para esse observador no hemisfério sul da Terra, menos de metade de sua trajetória diária ficará acima do horizonte e, portanto, a duração do dia será menor que 12 horas, ou seja, as noites serão maiores que os dias! Apenas nos equinócios, quando o Sol estiver passando exatamente sobre o equador celeste, metade de sua trajetória diária será acima, metade abaixo do horizonte, e o dia e a noite terão a mesma duração, de 12 h, em qualquer ponto, seja do hemisfério sul ou norte da Terra. Nos dias dos solstícios, quando o Sol atinge o máximo afastamento ao sul ou ao norte do equador, ocorrerão os dias em que a duração do dia, ou da noite, será a maior de todo o ano. Por fim, para um observador situado em um dos polos da Terra, o plano do horizonte cortará a esfera celeste exatamente sobre o equador celeste, o seja, o equador celeste estará sobre o círculo do horizonte. Esse observador verá apenas o hemisfério celeste correspondente ao terrestre em que ele se encontra (Fig. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Quanto à duração dos dias e noites, esse observador em um dos polos presenciará algo sui generis: o dia (Sol acima do horizonte) durará metade do ano – todo o tempo em que o Sol estiver situado no hemisfério celeste correspondente ao terrestre desse observador – e a noite durará a outra metade, ou seja: haverá seis meses de dia e seis meses de noite! Após essas explicações sobre como se dá o movimento aparente do Sol no céu, na próxima seção será apresentado o relato das duas atividades investigativas abordadas neste artigo, nas quais foram propostas questões aos estudantes sobre a trajetória diária do Sol e a duração dos dias e noites ao longo do ano. As explicações aqui expostas ajudarão no relato e discussão dessas atividades e no esclarecimento de quais seriam as respostas corretas às questões propostas aos estudantes. 9 Por se tratar da concessão de imagem e dados dos alunos, antes de se iniciar o registro das atividades foi repassado, a cada um deles, um termo de consentimento livre e esclarecido (TCLE), contendo instruções para participação no minicurso, bem como um termo para assinatura do aluno e devolução ao pesquisador. Na primeira aula, todos os alunos espontaneamente assinaram e devolveram o TCLE. 10 Feitas a cada “pausa” do professor na condução das atividades (ex.: em momentos onde era necessário aguardar os alunos registrarem suas respostas na folha de atividades e/ou discutirem acerca destas em grupo) e ao final de cada aula. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 94 8a), o hemisfério oposto permanecerá sempre abaixo do seu horizonte, invisível a ele. A trajetória diária do Sol corresponderá, então, a círculos paralelos ao horizonte (Figura 8). Isso quando ele estiver acima do horizonte, ou seja, durante o período do ano em que o Sol se encontra no hemisfério celeste correspondente ao hemisfério terrestre em cujo polo o observador se encontra. No restante do ano, quando o Sol estiver no hemisfério celeste oposto, que será invisível a este observador, o Sol nunca ficará acima do horizonte. Na Figura 8c, a faixa amarela novamente representa a variação na posição da trajetória diária do Sol devido ao seu movimento anual, durante a metade do ano em que permanecerá acima do horizonte. Figura 8 - Representação da esfera celeste visível para observador em um dos polos da Terra (Fig. 8a) e a trajetória do Sol, para este observador, em uma perspectiva tridimensional (Fig. 8b), bem como sua representação bidimensional (Fig. 8c). A letra “Z” representa o zênite e “P” o polo celeste visível para o observador, os quais, neste caso, coincidem. Figura 8 - Representação da esfera celeste visível para observador em um dos polos da Terra (Fig. 8a) e a trajetória do Sol, para este observador, em uma perspectiva tridimensional (Fig. 8b), bem como sua representação bidimensional (Fig. 8c). A letra “Z” representa o zênite e “P” o polo celeste visível para o observador, os quais, neste caso, coincidem. 95 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 95 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch 3 Elaboração e Aplicação das Atividades Investigativas sobre o movimento aparente do Sol e a duração dos dias e noites O minicurso que incluiu as atividades investigativas acerca do movimento diário do Sol e sua variação ao longo do ano foi aplicado no mês de outubro de 2017, com carga horária de dezesseis horas, a um grupo de trinta alunos do curso de graduação em Licenciatura em Ciências Biológicas da UFES. As atividades foram realizadas em uma sala de aula com capacidade para quarenta alunos, contendo um quadro branco com um projetor de vídeo integrado. A coleta de dados e o registro do desenvolvimento das atividades foram feitos por meio de vídeo, áudio, da aplicação de questionários9 e de anotações feitas pelo professor em um diário de aula. O registro em vídeo foi feito por meio de uma câmera digital posicionada junto ao quadro branco, em um ângulo que possibilitasse a visualização do maior número de alunos da sala. Para o registro em áudio, foi utilizado um aparelho celular, posicionado próximo ao único notebook que havia na sala, utilizado pelo professor para realizar as projeções com o software Stellarium e que, durante a segunda atividade, também foi utilizado pelos próprios alunos para manipularem o Stellarium. Contudo, a principal forma de registro e coleta de dados ocorreu por meio das respostas dos alunos – na forma de textos e/ou desenhos – aos questionários contidos nas folhas de atividades e as anotações do professor10 em um diário de aula, composto por um bloco de anotações específico para cada atividade aplicada. Nesse diário eram registrados fatos como: diálogos, expressões e reações dos alunos e o envolvimento deles em cada etapa da atividade, bem como sobre a própria mediação do professor na condução das atividades investigativas, fazendo uma autoanálise, buscando avaliar se esta mediação estava ocorrendo conforme as orientações metodológicas de um ensino por investigação, em especial as propostas por 96 96 96 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Azevedo (2006), no que se refere à aplicação de atividades do tipo “demonstração investigativa” e “laboratório aberto”. Azevedo (2006), no que se refere à aplicação de atividades do tipo “demonstração investigativa” e “laboratório aberto”. As atividades aqui relatadas foram desenvolvidas em um primeiro minicurso, realizado em 2017, que serviu de piloto para um segundo minicurso, aplicado em 2018, como parte do mesmo projeto de mestrado (Vidigal, 2019). 3 Elaboração e Aplicação das Atividades Investigativas sobre o movimento aparente do Sol e a duração dos dias e noites Os alunos novamente apontaram em direções distintas e mais uma vez começaram a rir uns dos outros (Diário de aula, 04/10/2017). Esse diálogo, em virtude da dúvida e envolvimento dos alunos ao tentar responder a simples pergunta de onde nasce o Sol, fez com que o professor percebesse a pertinência de propor uma atividade investigativa a esse respeito, que abordasse justamente a questão da posição de nascimento do Sol, de sua relação com o ponto cardeal leste, que também poderia ser associada à observação do céu ou, mais especificamente, à questão de qual seria o movimento do Sol no céu, com relação ao horizonte do observador, após o seu nascimento e antes do seu ocaso. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 3 Elaboração e Aplicação das Atividades Investigativas sobre o movimento aparente do Sol e a duração dos dias e noites Nesse primeiro minicurso, estruturado em cinco módulos, foi adotada uma organização de conteúdos “do macro para o micro” (Longhini, 2009), de acordo com sua natureza mais ou menos abrangente, iniciando com uma abordagem sobre o Universo como um todo, por considerá-lo um macroconteúdo, até a Terra e suas características físicas, considerando-a um conteúdo menos abrangente. De acordo com essa organização, a abordagem do movimento aparente do Sol, conforme visto da superfície da Terra, num referencial local, seria abordada na penúltima aula de uma forma apenas expositiva. Contudo, na primeira aula houve aplicação de um questionário inicial de autoavaliação em que se solicitava que cada estudante atribuísse uma nota de 0 a 10 quanto ao seu grau de conhecimento sobre os principais tópicos de Astronomia, indicados pela BNCC, a serem abordados na disciplina de Ciências, no Ensino Fundamental, sendo que o primeiro tópico indicado neste questionário foi justamente o “Movimento do Sol no céu” (Vidigal, 2019, p. 187). No momento seguinte à aplicação do questionário, ocorreu uma conversa entre o professor (P) e os alunos (A) acerca das respostas destes à autoavaliação e seus conhecimentos iniciais sobre Astronomia. Em especial, numa discussão quanto ao grau de conhecimento deles acerca do tópico “Movimento do Sol no céu”, ocorreu o seguinte diálogo, registrado no diário de aula do professor: Então eu disse: P: - “Ok! No leste! E para onde fica o leste?” Neste momento, respondendo à pergunta, praticamente cada aluno apontou para uma direção distinta. E daí ocorreu uma cena curiosa: no segundo seguinte em que cada aluno apontou em uma direção distinta, querendo indicar a direção leste, eles riram uns dos outros pela confusão ao descreverem esta referência. De repente uma das alunas argumentou com a turma: A: - “Gente, o leste fica na direção da praia!”. A: - “Gente, o leste fica na direção da praia!”. Neste momento, novamente fiz uma intervenção, dizendo: , ç , P: - “Ok! Para nós, observadores situados em Vitória, no Espírito Santo, de fato o leste fica na direção da praia. Mas para qual direção está à praia?”. Os alunos novamente apontaram em direções distintas e mais uma vez começaram a rir uns dos outros (Diário de aula, 04/10/2017). P: - “Ok! Para nós, observadores situados em Vitória, no Espírito Santo, de fato o leste fica na direção da praia. Mas para qual direção está à praia?”. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch A atividade então elaborada e aplicada na penúltima aula consistiu numa “demonstração investigativa” (Azevedo, 2006), a qual teve como objetivo abordar e investigar qual é o movimento diário do Sol no céu, buscando descrever, por meio de desenho, qual seria sua trajetória ao nascer e se pôr, tendo como referência os pontos cardeais leste (sendo identificado por um horizonte sem objetos, representando o mar) e oeste (contendo diversos objetos, representando a cidade, local onde ocorre o ocaso para um observador situado em Vitória) (Figuras 9 e 10). Inicialmente, como forma de proposição do problema e registro das concepções iniciais dos estudantes, foi solicitado a eles que formassem grupos, de aproximadamente três alunos. Ao se agruparem, totalizando dez grupos de alunos, foi então distribuída uma folha de atividades (atividade 3, Figura 9) a cada um deles. Nessa folha eram propostas as seguintes questões: Nessa folha eram propostas as seguintes questões: Nessa folha eram propostas as seguintes questões: 1. Desenhar a trajetória do Sol ao nascer, em relação ao horizonte leste, considerando um observador situado na cidade de Vitória; 1. Desenhar a trajetória do Sol ao nascer, em relação ao horizonte leste, considerando um observador situado na cidade de Vitória; 2. Desenhar a trajetória do Sol no ocaso, com relação ao horizonte oeste, para este mesmo observador. 2. Desenhar a trajetória do Sol no ocaso, com relação ao horizonte oeste, para este mesmo observador. Na Figura 10, destacamos duas respostas (duas folhas de atividades) dos alunos, indicando suas concepções com relação à trajetória ascendente do Sol no horizonte leste, caracterizado pela ausência de árvores, uma vez que, para um observador situado no Espírito Santo, a praia pode ser uma referência para a identificação da direção leste; assim como a trajetória do Sol até o seu ocaso no horizonte oeste, este último caracterizado por um horizonte contendo uma paisagem, que tipifica o horizonte oeste para este mesmo observador. 98 98 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Figura 9 - Folha da Atividade 3, distribuída aos alunos para registro de suas concepções iniciais acerca da posição de nascimento e ocaso do Sol e sua trajetória com relação ao horizonte. Figura 9 - Folha da Atividade 3, distribuída aos alunos para registro de suas concepções iniciais acerca da posição de nascimento e ocaso do Sol e sua trajetória com relação ao horizonte. Figura 9 - Folha da Atividade 3, distribuída aos alunos para registro de suas concepções iniciais acerca da posição de nascimento e ocaso do Sol e sua trajetória com relação ao horizonte. 99 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Figura 10 - Exemplo de respostas iniciais dos estudantes para as posições de nascimento e ocaso do Sol e de sua trajetória com relação ao horizonte. Figura 10 - Exemplo de respostas iniciais dos estudantes para as posições de nascimento e ocaso do Sol e de sua trajetória com relação ao horizonte. Inicialmente, cada aluno registrou suas concepções prévias acerca deste movimento em sua própria folha. Em seguida, foi solicitado que discutissem acerca de suas representações em seus respectivos grupos. Por meio da troca de ideias e argumentação, eles deveriam elaborar uma única representação que fornecesse uma ideia geral da resposta daquele grupo à atividade proposta, seja por meio de um desenho já feito por um dos integrantes do grupo, ou mediante um novo registro em desenho de uma possível ideia inédita surgida no grupo. Enquanto todos os grupos de alunos discutiam e registravam suas respostas, o professor preparava o projetor multimídia para projetar no quadro branco, que havia na sala, o software Stellarium (Stellarium.Org, 2020) para a etapa de demonstração investigativa e apresentação de uma simulação do movimento diário do Sol ao nascer e se pôr. Em seguida, simulamos o céu, utilizando o Stellarium. No momento exatamente antes da ocorrência do nascimento do Sol no dia em que a atividade estava sendo realizada (23 de outubro de 2017), pausamos a simulação para discussão com os alunos acerca das respostas indicadas por eles. Como forma de debater e problematizar as respostas iniciais dos estudantes, a primeira pergunta que fizemos foi acerca de onde (no horizonte leste) ocorreria o nascimento do Sol. As respostas dos grupos de alunos foram, então, marcadas no quadro branco (onde estava sendo projetado o Stellarium), conforme os apontamentos feitos por eles. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Observamos que a maioria indicou o nascimento do Sol exatamente sobre o ponto cardeal leste (Figura 11) – resultado semelhante ao indicado por outros autores, que relatam ser muito comum, entre as pessoas, incluindo estudantes e professores, a concepção alternativa de que o Sol nasce exatamente no ponto cardeal leste e se põe no ponto cardeal oeste todos os dias (Langhi, 2011; Bedaque & Bretones, 2016). Conforme 100 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites exposto na seção anterior deste artigo, isso não acontece. Devido ao movimento anual do Sol na eclíptica, há uma grande variação na sua posição de nascimento e ocaso junto ao horizonte ao longo do ano (Figura 5). exposto na seção anterior deste artigo, isso não acontece. Devido ao movimento anual do Sol na eclíptica, há uma grande variação na sua posição de nascimento e ocaso junto ao horizonte ao longo do ano (Figura 5). Figura 11 - Classificação das respostas (I, II, III, IV e V) dos dez grupos de alunos (representados pelas letras A, B, C, ..., J) acerca da trajetória de ascensão do Sol no horizonte leste. Figura 11 - Classificação das respostas (I, II, III, IV e V) dos dez grupos de alunos (representados pelas letras A, B, C, ..., J) acerca da trajetória de ascensão do Sol no horizonte leste. Demos, então, continuidade à simulação do nascer do Sol e a pausamos novamente no momento exato após o Sol nascer, sobre o horizonte leste, mas este não nasceu exatamente sobre o ponto cardeal. Ao se darem conta deste fato, houve uma série de discussões entre os alunos. Acreditamos que, neste momento, tenha ocorrido um conflito cognitivo acerca deste fenômeno, uma vez que parte dos alunos olhou para seu próprio registro, comparando com o que estava sendo projetado, e, ao tentar explicar, de alguma forma, a inconsistência entre o que previram e o fenômeno observado, não souberam descrever, de imediato, o porquê desta inconsistência. Outros alunos, por sua vez, conversavam entre si, buscando uma justificativa para o fato de o Sol não nascer exatamente sobre o ponto cardeal. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Foi necessária uma intervenção do professor, solicitando que os alunos, e grupos, se manifestassem a todos os alunos da turma – uma vez que estavam iniciando uma discussão apenas entre seus pares – para valorizar o debate com a turma acerca das inconsistências observadas e anotassem suas hipóteses acerca da razão de o Sol não ter nascido exatamente sobre o ponto cardeal, conforme haviam previsto. Ainda sem dar respostas à indagação anterior, prosseguimos com a demonstração investigativa, com o software ainda pausado, perguntando aos alunos qual seria a trajetória (de ascensão) descrita pelo Sol a partir de seu nascimento, conforme indicaram na folha da atividade. A partir das respostas dos alunos, classificamos os resultados conforme indicado no Quadro 1. 101 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Quadro 1 - Categorização das respostas dos alunos à Atividade 3 (Figura 9), acerca da trajetória de ascensão do Sol junto ao horizonte leste, para um observador situado na cidade de Vitória, ES, voltado de frente para este horizonte. Quadro 1 - Categorização das respostas dos alunos à Atividade 3 (Figura 9), acerca da trajetória de ascensão do Sol junto ao horizonte leste, para um observador situado na cidade de Vitória, ES, voltado de frente para este horizonte. Após todos os grupos apontarem a possível trajetória do Sol, retomamos a simulação e “movemos” lentamente os astros do céu em uma velocidade que nos permitiu demarcar no quadro branco a trajetória do Sol, utilizando um pincel, à medida que este ascendia no horizonte leste. Os alunos ficaram surpresos com o fato de o Sol não ascender paralelamente à normal ao horizonte leste, conforme previsto pela maioria deles (Figura 11), mas em uma trajetória inclinada em relação à normal ao horizonte. Pausamos novamente a simulação e questionamos os alunos acerca deste fato. Eles não souberam responder o porquê deste fenômeno. Conforme explicado na seção 2, isso se deve ao fato de o plano de horizonte ser sempre tangente à superfície da Terra, o que faz com que apenas no equador as trajetórias diárias das estrelas e do Sol, ao nascer e se pôr, sejam na direção perpendicular à linha do horizonte (Figura 6). Nos polos as trajetórias são paralelas ao horizonte (Figura 8) e, em todas as outras latitudes, a trajetória será sempre inclinada. No caso da cidade de Vitória, que fica no hemisfério sul, a inclinação será no sentido do norte (Figura 7). É interessante notar que, de acordo com os dados apresentados no Quadro 1, apenas o grupo A indicou corretamente a inclinação e seu sentido em sua resposta inicial. A indicação do ponto de nascimento do Sol feita por esse grupo, à direita do leste, também é a correta considerando-se que a resposta tenha sido dada com base na data de realização da atividade, que foi no dia 23 de outubro. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 101 Essa data é posterior ao equinócio de primavera para o hemisfério sul, que em geral ocorre no dia 23 de setembro, o que significa que o Sol, em sua trajetória anual na eclíptica, na data em que a atividade foi realizada, se encontrava no hemisfério sul celeste, ou seja, ao sul do equador celeste (Figura 4) e, portanto, nascendo à direita do ponto cardeal leste para um observador que esteja de frente para este ponto, lembrando que o equador celeste sempre corta o círculo do horizonte exatamente nos pontos cardeais leste e oeste. Contudo, como o enunciado da questão não se referia explicitamente à data em que a indicação da trajetória do Sol ao nascer deveria ser representada, respostas que indicassem o Sol nascendo à esquerda, ou no próprio ponto leste, não poderiam ser 102 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites consideradas erradas, apenas corresponderiam a outras épocas do ano, mas, em todas elas – é importante ressaltar – a trajetória diária do Sol ao nascer sempre terá uma inclinação para o norte (para a esquerda, no caso de um observador de frente para o leste), pois o observador se encontra no hemisfério sul da Terra. Para esse observador, será o polo sul celeste que estará visível, acima do horizonte, e o polo norte celeste estará sempre invisível, abaixo do horizonte. A fim de mediar a resposta a esse problema, perguntamos aos alunos se em todos os locais no planeta Terra a trajetória do Sol no céu seria a mesma. Um dos alunos, então, argumentou que no polo Sul deveria ser diferente, uma vez que nesta região há um longo período de incidência solar, ou de ausência dessa luz, durante o ano. Como forma de investigar essa afirmação, abrimos a janela de localização no Stellarium e simulamos o movimento aparente do Sol no céu para um observador situado no polo sul da Terra. Os alunos puderam, então, constatar o fato, apontado pelo colega, de que, para um observador situado bem próximo aos polos, a trajetória diária descrita pelo Sol é praticamente paralela ao horizonte, em geral sem nascer ou se pôr no mesmo dia. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 101 Em seguida, ainda com o Sol movendo-se lentamente na simulação, encerramos a demonstração abrindo a janela de localização novamente, mudando gradativamente a posição da latitude do observador e retornando sua localização para Vitória, ES, mantendo-se a longitude, e perguntamos aos alunos se a trajetória do Sol teria alguma relação com a posição do observador na Terra. Por meio da simulação, aparentemente todos perceberam que existia essa relação: para uma localização próxima ao polo, a trajetória era quase paralela ao horizonte e, à medida que a latitude se reduzia e o ponto de observação se aproximava do equador, a trajetória inclinada tendia a se aproximar cada vez mais da perpendicular ao horizonte. Assim, foi possível relacionar, qualitativamente, a posição do observador na Terra e a inclinação, com relação ao horizonte, da trajetória diária descrita pelo Sol no céu, devido a uma mudança na orientação do plano do horizonte do observador, sempre tangente à superfície da Terra. Isso pode ser bem visualizado por meio de um modelo tridimensional em que a Terra é representada por uma bola de isopor, como proposto por Canalle (1999), modelo com o qual já havíamos trabalhado com a turma, no mesmo dia, na aula anterior à da realização dessa demonstração investigativa. Por fim, foi solicitado que os alunos anotassem suas observações acerca deste fenômeno e então encerramos a atividade. Destacamos, na Figura 12, a resposta de um grupo que utilizou a mesma folha da atividade para registrar suas anotações após a realização da demonstração investigativa. 103 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Figura 12 - Respostas acerca do nascimento e ocaso do Sol, após a demonstração investigativa. Figura 12 - Respostas acerca do nascimento e ocaso do Sol, após a demonstração investigativa Após a realização dessa demonstração investigativa, apresentamos à turma as principais funcionalidades do software Stellarium a fim de realizarmos uma nova atividade, numa forma semelhante à de um “laboratório aberto” (Azevedo, 2006), na qual os próprios alunos planejariam e realizariam o experimento. No nosso caso um experimento virtual, por meio de uma simulação do movimento do Sol em diferentes dias do ano, utilizando o Stellarium. Após a realização da primeira atividade, foi feito um intervalo de aproximadamente vinte minutos. Nesta segunda atividade, solicitamos que os alunos novamente se organizassem em grupos. 101 Eles então se organizaram, espontaneamente, em seis grupos, porém com quantidades diferentes de alunos, sendo: grupo A, cinco alunos; grupo B, onze alunos; grupo C, três alunos; grupo D e F, dois alunos; e um único aluno, que não quis participar de nenhum grupo, ao qual denominamos de Grupo E. Embora a divisão dos grupos pudesse ter sido mais homogênea, limitada ao número de até cinco alunos, todos eles participaram ativamente da proposta. A ideia da nova atividade era dar continuidade à investigação acerca do movimento aparente do Sol no céu, buscando abordar não apenas como se dá este movimento num dado dia específico, mas como ele variaria ao longo do ano, produzindo fenômenos como o da variação na duração dos dias e noites, mudança da posição de nascimento e ocaso do Sol no horizonte e a sucessão de estações do ano. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 104 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Buscando abordar mais diretamente o primeiro desses fenômenos, a questão proposta à turma pelo professor foi: 1. A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? 1. A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? Solicitamos que eles, antes de utilizarem o software, realizassem as etapas de apresentação de hipóteses e plano de trabalho. Na folha da atividade distribuída para cada grupo, havia uma tabela (Figura 13) para que fossem anotados os horários de: nascimento do Sol; passagem pelo zênite (ou o mais próximo dele); ocaso do Sol; duração do dia e duração da noite num dado dia fixo do mês (por exemplo: 15/01; 15/02; 15/03 etc..), ao longo de todos os meses do ano, para um observador situado em Vitória, ES. Os dados a serem preenchidos na tabela deveriam ser colhidos pelos próprios grupos de alunos, por meio do Stellarium. Figura 13 - Planilha, distribuída aos grupos, para preenchimento com os horários de nascimento e ocaso do Sol, de sua passagem próximo ao zênite e da duração do dia e da noite, numa mesma data fixa do mês (p. ex., dia 15), ao longo dos 12 meses do ano. Figura 13 - Planilha, distribuída aos grupos, para preenchimento com os horários de nascimento e ocaso do Sol, de sua passagem próximo ao zênite e da duração do dia e da noite, numa mesma data fixa do mês (p. ex., dia 15), ao longo dos 12 meses do ano. Durante a manipulação do software, alguns alunos fizeram perguntas ao professor, como acerca do fato de o Sol não passar pelo zênite em todos os dias e sobre a variação dos dias e das noites no decorrer do ano. Porém, nesta etapa, o professor evitou dar respostas, deixando que os alunos discutissem entre eles suas dúvidas, apenas atuando na orientação dos alunos quanto à operação do software e a coleta de dados. 104 Notamos que, praticamente, não houve dificuldade, por parte dos alunos, na manipulação do Stellarium para obtenção dos horários de nascimento e ocaso do Sol. Contudo, devido à limitação de contarmos com apenas um computador (notebook) para ser manipulado pelos estudantes, foi necessário efetivar um rodízio entre os grupos para a coleta de dados, o que demandou um tempo maior que o desejável. Um dos grupos não teve tempo de preencher completamente a tabela de dados e houve um comprometimento da etapa final, de discussão entre os grupos e busca de uma síntese acerca do fenômeno, verificando a validade, ou não, das hipóteses iniciais, explorando também algumas de suas consequências e outros fenômenos associados, como o da mudança da posição do Sol ao nascer e se pôr e as estações do ano. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Sim, pois pudemos notar que o sol muda sua rota no decorrer do ano Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch ✓ A análise reafirma a hipótese ✓ Descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. Grupo F A inclinação da Terra atua veementemente para estas singularidades. No verão há maior propensão à radiação solar e no inverno ao contrário. Não respondeu Não respondeu ✓ Indicação da possível causa do fenômeno. Quadro 2 - Respostas dos alunos registradas na folha da atividade e observações dos autores acerca das hipóteses, análises e conclusões de cada grupo. Problema proposto: A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? A partir da manipulação do software Stellarium, foi possível chegar aos resultados da proposta do problema? Observações dos autores desta pesquisa: um comparativo entre as hipóteses apresentadas pelos grupos e a análise e conclusão destes, após a realização da atividade investigativa. ✓ A análise reafirma a hipótese ✓ Associação a estações do ano ✓ Maior descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. ✓ Associação a estações do ano ✓ Identificação da ocorrência do fenômeno ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno ✓ Concepção alternativa quanto aos equinócios ✓ Indicação da possível causa do fenômeno. ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno ✓ A análise reafirma a hipótese ✓ Descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. ✓ Indicação da possível causa do fenômeno. e e observações Analise a tabela e descreva algumas observações acerca dos dados. Hipóteses Iniciais No decorrer do ano a duração do dia e noite variam, os dias se tornam mais curtos a partir de março, mas em agosto já volta a ser longo. Isso se deve a mudança de inclinação do eixo da Terra que muda a incidência dos raios solares. Sim, pois pudemos notar que o sol muda sua rota no decorrer do ano No inverno há dias mais curtos e dias mais longos no verão. Em março e setembro o dia dura aproximadamente 12 horas. Não respondeu Sim, de acordo com os resultados, os dias mais longos são dos meses de verão e os mais curtos, ou seja, com as noites mais longas, são dos meses de inverno. Não há nenhum mês em que o dia e a noite possuem durações iguais. O dia é mais longo nos meses de dezembro, janeiro, fevereiro, março, e novembro, isto é, nos meses de verão. Já a noite é mais longa nos meses de maio, junho, julho e agosto, os meses de inverno. Sim, pois ficou evidente a correlação entre a duração e a posição da terra em relação ao sol (estações do ano relacionadas ao movimento de translação da terra). Dias longos de outubro a março Noites longas de abril a agosto Dias de duração mediana nos demais Os dias são maiores em dezembro, novembro e janeiro. Já em maio, junho, julho e agosto ocorre oposto, a noite tem maior duração. Em setembro e outubro o dia e a noite têm quase a mesma duração No inverno o arco está mais próximo do horizonte, nasce e logo se põe. No verão o arco descreve uma trajetória maior, por isso o dia é mais longo. Quadro 2 - Respostas dos alunos registradas na folha da atividade e observações dos autores acerca das hipóteses, análises e conclusões de cada grupo. A partir dessas anotações, foi possível perceber que, dos seis grupos, três consideraram que os dados colhidos confirmaram sua hipótese inicial, na qual consideravam que haveria associação entre a duração do dia e as estações do ano (dias mais longos no verão, mais curtos no inverno e igual duração nos equinócios) ou com o movimento de translação da Terra e a sua “inclinação do eixo” ou a “inclinação do planeta” (sem, contudo, indicar claramente que a inclinação seria do eixo de rotação da Terra com relação ao plano de sua órbita). Um dos grupos pareceu fazer uma associação correta, tanto em sua hipótese inicial como na conclusão, entre a variação na duração do dia e a trajetória diária aparente do Sol, que se situava mais, ou menos, próxima ao zênite ou ao horizonte, ao longo do ano. Um dos grupos não anotou sua conclusão, embora, na hipótese inicial, tenha indicado haver uma relação entre a duração do dia e Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch No Quadro 2, a seguir, sintetizam da atividade e observações feitas pelos conclusões de cada grupo. Problema proposto: A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? Analise a tabela e desc algumas observações ac dos dados. Hipóteses Iniciais Análise Grupo A Os dias mais longos são no verão, já os mais curtos são inverno e os de igual duração são os equinócios. No decorrer do ano a du do dia e noite variam, os d tornam mais curtos a part março, mas em agosto já v ser longo. Isso se dev mudança de inclinação do da Terra que muda a incid dos raios solares. Grupo B Não respondeu No inverno há dias mais c e dias mais longos no v Em março e setembro o dura aproximadamente horas. Grupo C Por causa do movimento de translação da Terra e da inclinação do planeta, nós recebemos intensidades diferentes de luz ao longo do ano, tendo dias mais longos em alguns meses e noites mais longas em outros. O dia é mais longo nos m de dezembro, janeiro, feve março, e novembro, isto é meses de verão. Já a no mais longa nos meses de junho, julho e agosto, os m de inverno. Grupo D Os dias são mais longos no verão, mais curtos no inverno, e de aproximada igual duração nos equinócios (primavera e outono ). Dias longos de outubro a m Noites longas de abril a ag Dias de duração mediana demais Grupo E Por causa da passagem pelo zênite Os dias são maiores dezembro, novembro e jan Já em maio, junho, jul agosto ocorre oposto, a tem maior duração. setembro e outubro o dia noite têm quase a m duração Grupo F A inclinação da Terra atua veementemente para estas singularidades. No verão há maior propensão à radiação solar e no inverno ao contrário. Não respondeu Quadro 2 - Respostas dos alunos regis dos autores acerca das hipóteses, No Quadro 2, a seguir, sintetizamos as anotações feitas pelos grupos na folha da atividade e observações feitas pelos autores acerca das hipóteses, análises e conclusões de cada grupo. No Quadro 2, a seguir, sintetizamos as anotações feitas pelos grupos na folha da atividade e observações feitas pelos autores acerca das hipóteses, análises e conclusões de cada grupo. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch ç p p , conclusões de cada grupo. Problema proposto: A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? Analise a tabela e descreva algumas observações acerca dos dados. A partir da manipulação do software Stellarium, foi possível chegar aos resultados da proposta do problema? Observações dos autores desta pesquisa: um comparativo entre as hipóteses apresentadas pelos grupos e a análise e conclusão destes, após a realização da atividade investigativa. Hipóteses Iniciais Análise Conclusão Grupo A Os dias mais longos são no verão, já os mais curtos são inverno e os de igual duração são os equinócios. No decorrer do ano a duração do dia e noite variam, os dias se tornam mais curtos a partir de março, mas em agosto já volta a ser longo. Isso se deve a mudança de inclinação do eixo da Terra que muda a incidência dos raios solares. Sim, pois pudemos notar que o sol muda sua rota no decorrer do ano ✓ A análise reafirma a hipótese ✓ Associação a estações do ano ✓ Maior descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. Grupo B Não respondeu No inverno há dias mais curtos e dias mais longos no verão. Em março e setembro o dia dura aproximadamente 12 horas. Não respondeu ✓ Associação a estações do ano ✓ Identificação da ocorrência do fenômeno Grupo C Por causa do movimento de translação da Terra e da inclinação do planeta, nós recebemos intensidades diferentes de luz ao longo do ano, tendo dias mais longos em alguns meses e noites mais longas em outros. O dia é mais longo nos meses de dezembro, janeiro, fevereiro, março, e novembro, isto é, nos meses de verão. Já a noite é mais longa nos meses de maio, junho, julho e agosto, os meses de inverno. Sim, de acordo com os resultados, os dias mais longos são dos meses de verão e os mais curtos, ou seja, com as noites mais longas, são dos meses de inverno. Não há nenhum mês em que o dia e a noite possuem durações iguais. ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno ✓ Concepção alternativa quanto aos equinócios ✓ Indicação da possível causa do fenômeno. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Grupo D Os dias são mais longos no verão, mais curtos no inverno, e de aproximada igual duração nos equinócios (primavera e outono ). Dias longos de outubro a março Noites longas de abril a agosto Dias de duração mediana nos demais Sim, pois ficou evidente a correlação entre a duração e a posição da terra em relação ao sol (estações do ano relacionadas ao movimento de translação da terra). ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno Grupo E Por causa da passagem pelo zênite Os dias são maiores em dezembro, novembro e janeiro. Já em maio, junho, julho e agosto ocorre oposto, a noite tem maior duração. Em setembro e outubro o dia e a noite têm quase a mesma duração No inverno o arco está mais próximo do horizonte, nasce e logo se põe. No verão o arco descreve uma trajetória maior, por isso o dia é mais longo. ✓ A análise reafirma a hipótese ✓ Descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. Grupo F A inclinação da Terra atua veementemente para estas singularidades. No verão há maior propensão à radiação solar e no inverno ao contrário. Não respondeu Não respondeu ✓ Indicação da possível causa do fenômeno. Quadro 2 - Respostas dos alunos registradas na folha da atividade e observações dos autores acerca das hipóteses, análises e conclusões de cada grupo. Problema proposto: A partir da manipulação do software Stellarium, em que época do ano os dias são mais longos? E mais curtos? E quando o dia tem igual duração que a noite? Analise a tabela e descreva algumas observações acerca dos dados. A partir da manipulação do software Stellarium, foi possível chegar aos resultados da proposta do problema? Observações dos autores desta pesquisa: um comparativo entre as hipóteses apresentadas pelos grupos e a análise e conclusão destes, após a realização da atividade investigativa. Hipóteses Iniciais Análise Conclusão Grupo A Os dias mais longos são no verão, já os mais curtos são inverno e os de igual duração são os equinócios. No decorrer do ano a duração do dia e noite variam, os dias se tornam mais curtos a partir de março, mas em agosto já volta a ser longo. Isso se deve a mudança de inclinação do eixo da Terra que muda a incidência dos raios solares. Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Sim, pois pudemos notar que o sol muda sua rota no decorrer do ano ✓ A análise reafirma a hipótese ✓ Associação a estações do ano ✓ Maior descrição da ocorrência do fenômeno ✓ Indicação da possível causa do fenômeno. Grupo B Não respondeu No inverno há dias mais curtos e dias mais longos no verão. Em março e setembro o dia dura aproximadamente 12 horas. Não respondeu ✓ Associação a estações do ano ✓ Identificação da ocorrência do fenômeno Grupo C Por causa do movimento de translação da Terra e da inclinação do planeta, nós recebemos intensidades diferentes de luz ao longo do ano, tendo dias mais longos em alguns meses e noites mais longas em outros. O dia é mais longo nos meses de dezembro, janeiro, fevereiro, março, e novembro, isto é, nos meses de verão. Já a noite é mais longa nos meses de maio, junho, julho e agosto, os meses de inverno. Sim, de acordo com os resultados, os dias mais longos são dos meses de verão e os mais curtos, ou seja, com as noites mais longas, são dos meses de inverno. Não há nenhum mês em que o dia e a noite possuem durações iguais. ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno ✓ Concepção alternativa quanto aos equinócios ✓ Indicação da possível causa do fenômeno. Grupo D Os dias são mais longos no verão, mais curtos no inverno, e de aproximada igual duração nos equinócios (primavera e outono ). Dias longos de outubro a março Noites longas de abril a agosto Dias de duração mediana nos demais Sim, pois ficou evidente a correlação entre a duração e a posição da terra em relação ao sol (estações do ano relacionadas ao movimento de translação da terra). ✓ A análise reafirma a hipótese ✓ Maior descrição da ocorrência do fenômeno Grupo E Por causa da passagem pelo zênite Os dias são maiores em dezembro, novembro e janeiro. Já em maio, junho, julho e agosto ocorre oposto, a noite tem maior duração. Em setembro e outubro o dia e a noite têm quase a mesma duração No inverno o arco está mais próximo do horizonte, nasce e logo se põe. No verão o arco descreve uma trajetória maior, por isso o dia é mais longo. 105 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch A partir da manipulação do software Stellarium, foi possível chegar aos resultados da proposta do problema? Análise Conclusão Conclusão Hipóteses Iniciais A Os dias mais longos são no verão, já os mais curtos são inverno e os de igual duração são os equinócios. Observações dos autores desta pesquisa: um comparativo entre as hipóteses apresentadas pelos grupos e a análise e conclusão destes, após a realização da atividade investigativa. 4 Conclusões Na primeira atividade, que foi do tipo “demonstração investigativa”, sobre a trajetória do Sol em relação ao horizonte numa dia específico do ano, observamos que, antes da demonstração, a maioria dos alunos evidenciou uma concepção alternativa de que o Sol nasce exatamente sobre o ponto cardeal leste, descrevendo uma trajetória ascendente vertical, paralela à normal ao horizonte leste, até passar para o lado oeste do céu, descendo também verticalmente com relação ao horizonte, até o seu ocaso exatamente sobre o ponto cardeal oeste. Esse resultado parece corroborar a constatação, feita por diversos autores, de que é muito comum, tanto entre os estudantes como entre professores da Educação Básica, a concepção alternativa de que o Sol nasce exatamente no ponto cardeal leste e se põe no ponto cardeal oeste todos os dias (Langhi, 2011; Bedaque & Bretones, 2016). Quanto ao conteúdo, após a realização dessa primeira atividade, desenhos realizados por alguns dos grupos, como o indicado na Figura 12, parecem sugerir que houve uma apropriação conceitual acerca de como ocorre o movimento diário do Sol, bem como sobre a relação desse movimento com a posição do observador na superfície da Terra e o fato de ele não necessariamente nascer no ponto cardeal leste e se pôr no ponto cardeal oeste. Quanto à metodologia de ensino investigativo utilizada nessa atividade, os resultados obtidos estão em consonância com alguns apontamentos de Azevedo (2006) acerca das contribuições da aplicação de uma demonstração investigativa, uma vez que: evidenciou-se a percepção de concepções espontâneas dos alunos por meio da participação destes nas diversas etapas da resolução de problemas; aproximou-se a atividade de um procedimento de investigação científica, por meio da observação atenta de um fenômeno natural; proporcionou-se maior participação e interação dos alunos em sala de aula; valorizou-se a interação do aluno com o objeto de estudo; valorizou-se a aprendizagem de atitudes e não apenas de conteúdos e possibilitou a criação de conflitos cognitivos em sala de aula. Já a proposta da segunda atividade, sobre a variação da duração dos dias e noites ao longo do ano, revelou-se acertada com relação às suas possibilidades de participação ativa dos estudantes na obtenção de dados e teste de suas próprias hipóteses, seguindo uma metodologia semelhante à de um “laboratório aberto” (Azevedo, 2006), tornando possível um aprofundamento da investigação do movimento diário do Sol, dessa vez acerca de como este movimento varia ao longo do ano. 106 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites as estações do ano (duração maior no verão, menor no inverno) e a “inclinação da Terra”, que parece se referir à inclinação do eixo de rotação do nosso planeta em relação ao plano de sua órbita ao redor do Sol. Um último grupo não anotou nem sua hipótese inicial, nem sua conclusão, mas indicou, no item dedicado à análise dos dados, que há dias mais curtos no inverno, dias mais longos no verão e que, em março e setembro, o dia dura aproximadamente 12 horas. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 4 Conclusões Contudo, na implementação dessa aplicação piloto, conforme citado anteriormente, houve uma limitação em termos de recursos e de tempo disponível para sua aplicação. A possível discussão de outras questões relacionadas ao movimento anual do Sol, além da duração do dia e da noite, que poderiam ter sido melhor exploradas e relacionadas a 107 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch essa variação, como a mudança na posição de nascimento e ocaso do Sol no horizonte, sua passagem mais, ou menos, próxima ao zênite e as estações do ano, infelizmente não pôde ser feita com toda a turma, entre os grupos e com a participação do professor como mediador, apenas dentro dos grupos houve discussões nesse sentido. Contudo, a atividade parece ter contribuído para que alguns estudantes, por meio da visualização oferecida pelo Stellarium, compreendessem bem como se dá a variação do movimento diário do Sol no céu ao longo do ano e sua relação com a duração do dia, como parece ser o caso do estudante que fez a seguinte anotação: “No inverno o arco [trajetória do Sol] está mais próximo do horizonte, nasce e logo se põe. No verão o arco descreve uma trajetória maior, por isso o dia é mais longo”. Conforme apresentado na seção 2 do artigo, numa latitude como a da cidade de Vitória (20o S), é exatamente isso que ocorre: no verão, próximo ao solstício deste estação, a porção do arco diário descrito pelo Sol acima do horizonte é maior, enquanto que, no inverno, no solstício, esse arco é menor e se situa mais baixo, mais próximo do horizonte. Com relação à metodologia investigativa, essa segunda atividade, apesar de algumas falhas na sua implementação (excesso de alunos em um dos grupos e apenas um computador disponível para uso), também conseguiu promover uma participação ativa dos estudantes, com a prática e aprendizagem de atitudes e não apenas de conteúdos. Os próprios estudantes colheram os dados acerca da duração do dia, buscaram verificar se as hipóteses, levantadas por alguns deles, de que haveria uma variação dessa duração ao longo do ano associada às estações do ano, aos movimentos da Terra ou à passagem mais, ou menos, próxima ao zênite, estariam corretas ou não e realizaram discussões no interior dos grupos sobre esses resultados a fim de extrair conclusões. Algo que merece ser destacado no desenvolvimento das duas atividades foi o papel essencial que a ferramenta Stellarium desempenhou. No momento em que abrimos e exibimos esse software pela primeira vez, foi possível observar que os alunos se surpreenderam com seus recursos e elogiaram sua aparência (design). 107 Além de simular o céu e a posição e movimento do Sol, conforme vistos de qualquer ponto da superfície da Terra, em qualquer dia e horário, ele também permite pausar, acelerar ou retroceder a passagem do tempo. Somente graças a esses recursos, típicos de um planetário, foi possível conduzir as investigações propostas, acerca da posição de nascimento do Sol, seu subsequente movimento com relação ao horizonte e a variação desse movimento ao longo do ano. Nossos resultados corroboram o que é dito por Longhini e Menezes (2010), que identificam o Stellarium como sendo um “objeto virtual de aprendizagem”: As possibilidades de exploração desse software são inúmeras, tornando-o um valoroso objeto de ensino e de aprendizagem para o ensino de Ciências, Geografia e, mais especificamente, no campo da Astronomia [...] Ao se constituir como uma ferramenta aberta e de múltiplas possibilidades, propicia ao professor criar desafios ou questões para explorar temáticas relativas à Astronomia (Longhini & Menezes, 2010, p. 436). De fato, foi notório o engajamento dos alunos, participando ativamente da produção e análise dos dados obtidos, por eles próprios, ao realizarem as atividades, apoiadas no uso do Stellarium, o que também já foi observado em outras pesquisas na área de Ensino de Astronomia junto a alunos da Educação Básica, como em Neres Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 5 Considerações Finais As atividades relatadas no presente artigo fazem parte de uma aplicação piloto de uma proposta didática para formação inicial de professores de Ciências do Ensino Fundamental em Astronomia (Vidigal, 2019) que busca incorporar diversas atividades baseadas na abordagem do Ensino por Investigação (Zômpero & Laburú, 2011). Isso é feito como forma de familiarizar e estimular os futuros docentes à adoção, com seus futuros alunos, de uma perspectiva de ensino que parece adequada aos dias atuais, envolvendo a participação ativa dos estudantes, a aprendizagem não só de conceitos, mas também de procedimentos característicos da ciência, o desenvolvimento de habilidades cognitivas e senso crítico. Conforme relatado, embora a segunda atividade tenha apresentado algumas falhas em sua implementação, que resultaram num prejuízo na etapa final de discussão entre os grupos e busca de uma sistematização das conclusões, o conjunto das duas atividades parece ter evidenciado o bom potencial da união entre a abordagem metodológica – o ensino por investigação – e a ferramenta – software Stellarium – utilizadas para o ensino do tema do movimento aparente do Sol no céu. Uma descrição mais detalhada de cada uma dessas atividades, incluindo o material utilizado (plano de aula e folhas de questões) se acham disponíveis na dissertação de Vidigal (2019) e em seu produto, a apostila “Astronomia: atividades investigativas para a formação inicial de professores”, que segue como apêndice à dissertação. Ressaltamos que o conteúdo das duas atividades, com foco na visão topocêntrica, mais ligada à vivência e ao cotidiano das pessoas, conforme já mencionado na introdução, é a mais indicada para o trabalho com as séries iniciais do Ensino Fundamental. Contudo, numa proposta de ensino para formação inicial de professores, é necessário ir além e buscar promover uma articulação desta visão com a visão heliocêntrica. Uma complementa a outra. A partir da consideração dos movimentos da Terra, o modelo heliocêntrico fornece uma excelente explicação científica do que é observado e vivenciado a partir de uma visão da superfície da Terra, e será bem importante que o professor, mesmo ensinando sobre o movimento do Sol no céu a partir de um referencial local, saiba como essas observações se articulam e podem ser bem explicadas pelo modelo científico heliocêntrico. Essa coordenação de pontos de vista, como já mencionamos, não é trivial (Bisch, 1998; Leite, 2006) e deve ser um dos focos de uma formação inicial. 108 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites (2017) e Silva (2015). Isso nos leva a concluir sobre o potencial desse software como um ótimo recurso de apoio ao desenvolvimento de atividades investigativas, possibilitando simular eventos astronômicos – que na natureza levariam horas, dias ou anos – e assim testar eventuais hipóteses levantadas e fazer uma crítica e revisão das concepções iniciais dos estudantes acerca desses fenômenos. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 5 Considerações Finais Por fim, considerando a atual popularidade e difusão de concepções “terraplanistas” na mídia, especialmente na internet, que tem motivado análises por parte de pesquisadores da área de ensino (Silveira, 2017; Pivaro, 2019), uma outra possibilidade muito pertinente de desenvolvimento e extensão da discussão associada à primeira atividade aqui relatada, que analisa a trajetória aparente do Sol com relação ao horizonte em seu movimento diário, seria abordar como a variação da inclinação dessa 109 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch trajetória do Sol, à medida que se altera a posição (latitude) do observador, pode ser usada como um forte argumento a favor da esfericidade da Terra. Esse ponto poderia ser incluído como uma continuidade da elaboração do problema da atividade: após a demonstração de que a inclinação da trajetória varia com a latitude, poderia se perguntar acerca da causa desta variação, aguardar as hipóteses levantadas pelos alunos, promover um debate sobre elas e, então, dar continuidade à demonstração investigativa utilizando, por exemplo, modelos tridimensionais, um de uma Terra esférica, outro de uma Terra plana. Usando um modelo tridimensional, com a Terra sendo representada, por exemplo, por uma bola de isopor, seria possível demonstrar que a variação da direção do movimento diário do Sol com relação ao horizonte – paralela ao horizonte nos polos e perpendicular ao horizonte no equador – pode ser bem explicada pelo fato de a orientação do plano do horizonte, sempre tangente à superfície esférica da Terra, ir variando à medida que o ponto de observação se desloca sobre a Terra, ao longo de um meridiano, variando sua latitude. Nesse modelo, o observador poderia ser representado por um alfinete, o plano do horizonte por um pequeno pedaço plano de folha de papel, preso pelo alfinete e tangente à esfera, e o Sol por uma lâmpada. Esse efeito, de variação da inclinação da trajetória do Sol em relação ao horizonte de acordo com a latitude do observador, jamais ocorreria se a Terra fosse plana, pois nela o plano do horizonte e sua orientação seriam sempre os mesmos. Agradecimentos Aos estudantes do curso de Ciências Biológicas da Universidade Federal do Espírito Santo (UFES) que gentilmente aceitaram participar das atividades do projeto de pesquisa em ensino desenvolvido pelos autores. Ao professor Carlos Augusto Cardoso Passos, da UFES, por ter gentilmente colaborado e cedido parte do tempo de suas aulas de Física para alunos de Ciências Biológicas para que o projeto fosse aplicado. Referências Azevedo, M. C. P. S. (2006). Ensino por investigação: problematizando as atividades em sala de aula. In Carvalho, A. M. P. (Org). Ensino de Ciências: unindo a pesquisa e a prática (2a ed., pp. 19-33). São Paulo: Pioneira Thomson Learning. Azevedo, M. C. P. S. (2006). Ensino por investigação: problematizando as atividades em sala de aula. In Carvalho, A. M. P. (Org). Ensino de Ciências: unindo a pesquisa e a prática (2a ed., pp. 19-33). São Paulo: Pioneira Thomson Learning. Bedaque, P., & Bretones, P. S. (2016). Variação da posição de nascimento do Sol em função da latitude. Revista Brasileira de Ensino de Física, 38(3), 1-11. Recuperado em 08 jul. 2020, de http://dx.doi.org/10.1590/1806-9126-RBEF-2015-0023 Bedaque, P., & Bretones, P. S. (2020). O Sol está sempre a pino ao meio-dia?. Revista Brasileira de Ensino de Física, 42. Recuperado em 05 ago. 2020, de https://doi.org/10.1590/1806-9126-rbef-2019-0025 Bedaque, P., & Bretones, P. S. (2020). O Sol está sempre a pino ao meio-dia?. Revista Brasileira de Ensino de Física, 42. Recuperado em 05 ago. 2020, de https://doi.org/10.1590/1806-9126-rbef-2019-0025 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 110 110 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Bisch, S. M. (1998). Astronomia no Ensino Fundamental: natureza e conteúdo do conhecimento de estudantes e professores. Tese de Doutorado, Faculdade de Educação, USP, São Paulo. Bisch, S. M. (2012). Introdução à Astronomia. Vitória: UFES, Núcleo de Educação Aberta e a Distância. Recuperado em 26 jul. 2020, de http://acervo.sead.ufes.br/arquivos /introducao-a-astronomia.pdf Boczko, R. (1984). Conceitos de Astronomia. São Paulo: Edgard Blücher. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Boczko, R. (1984). Conceitos de Astronomia. São Paulo: Edgard Blücher. Brasil. Ministério da Educação, Secretaria da Educação Fundamental. (1998). Parâmetros Curriculares Nacionais: Ciências Naturais (5ª a 8ª Série). Brasília: MEC. Recuperado em 26 jul. 2020, de http://portal.mec.gov.br/seb/arquivos/pdf/ ciencias.pdf Brasil. Ministério da Educação, Secretaria de Educação Básica. (2013). Diretrizes Curriculares Nacionais da Educação Básica. Brasília: MEC, 2013. Recuperado em 26 jul. 2020, de http://portal.mec.gov.br/index.php?option=com_docman&view=download &alias=13448-diretrizes-curiculares-nacionais-2013-pdf&Itemid=30192 Brasil. Ministério da Educação, Secretaria de Educação Básica. (2018). Base Nacional Comum Curricular. Brasília: MEC. Recuperado em 26 jul. 2020, de http://basenacional comum.mec.gov.br Canalle, J. B. G. (1999). Explicando Astronomia básica com uma bola de isopor. Caderno Catarinense de Ensino de Física, 16(3). Recuperado em 05 jul. 2018, de https://periodicos.ufsc.br/index.php/fisica/article/view/6797/13485 Caniato, R. (1993). O Céu. São Paulo: Ática. Caniato, R. (1993). O Céu. São Paulo: Ática. Carvalho, A. M. P. (2018). Fundamentos teóricos e metodológicos do ensino por inves- tigação. Revista Brasileira de Pesquisa em Educação em Ciências, 18(3), 765-794. Re- cuperado em 26 jul. 2020, de https://doi.org/10.28976/1984-2686rbpec2018183765 Carvalho, A. M. P. (Org.). (2013). Ensino de Ciências por Investigação: condições para implementação em sala de aula. São Paulo: Cengage Learning. Carvalho, A. M. P., & Gil-Pérez, D. (2011). Formação de professores de Ciências: tendências e inovações. São Paulo: Cortez. Ferreira, F. P., & Leite, C. (2015). A forma e os movimentos da Terra: percepções de professores acerca das relações entre observação cotidiana e os modelos científicos. Revista Latino-Americana de Educação em Astronomia, 19, 123-146. Recuperado em 26 jul. 2020, de https://doi.org/10.37156/RELEA/2015.19.123 Langhi, R. (2011). Educação em Astronomia: da revisão bibliográfica sobre concepções alternativas à necessidade de uma articulação nacional. Caderno Brasileiro de Ensino de Física, 28 (2), 373-399. Recuperado em 26 jul. 2020, de https://doi.org/10.5007/ 2175-7941.2011v28n2p373 111 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Wesley Quintiliano Vidigal e Sérgio Mascarello Bisch Langhi, R., & Nardi, R. (2007). Ensino de Astronomia: erros conceituais mais comuns presentes em livros didáticos de Ciências. Caderno Brasileiro de Ensino de Física, 24(1), 87-111. Recuperado em 26 jul. 2020, de https://periodicos.ufsc.br/index.php/ fisica/article/view/6055 Langhi, R., & Nardi, R. (2010). Formação de professores e seus saberes disciplinares em Astronomia essencial nos anos iniciais do Ensino Fundamental. Ensaio: Pesquisa em Educação em Ciências, 12(2), 205-224. Recuperado em 26 jul. 2020, de https://doi.org/10.1590/1983-21172010120213 Leite, C. M. (2006). Formação do professor de Ciências em Astronomia: uma proposta com enfoque na espacialidade. Tese de doutorado, Faculdade de Educação, USP, São Paulo. Longhini, M. D. (2009). Boczko, R. (1984). Conceitos de Astronomia. São Paulo: Edgard Blücher. O universo representado em uma caixa: introdução ao estudo da astronomia na formação inicial de professores de física. Revista Latino-Americana de Educação em Astronomia, 7, 31-42. Recuperado em 26 jul. 2020, de https://doi.org/10.37156/RELEA/2009.07.031 Longhini, M. D., & Menezes, L. D. D. (2010). Objeto virtual de aprendizagem no ensino de Astronomia: algumas situações-problema propostas a partir do software Stellarium. Caderno Brasileiro de Ensino de Física, 27(3), 433-448. Recuperado em 25 abr. 2020, de https://doi.org/10.5007/2175-7941.2010v27n3p433 Neres, L. B. (2017). O Stellarium como Estratégia para o Ensino de Astronomia. Dis- sertação de Mestrado, Universidade Estadual de Santa Cruz, Ilhéus. Oliveira Filho, K. S., & Saraiva, M. F. O. (2014). Astronomia e Astrofísica. São Paulo: Livraria da Física. Pivaro, G. F. (2019). A crença numa Terra plana e os ambientes virtuais: identificando relações e construções de conhecimento. Anais do Encontro Nacional de Pesquisa em Educação em Ciências, Natal: ABRAPEC, 12. Recuperado em 26 jul. 2020, de http://abrapecnet.org.br/enpec/xii-enpec/anais/resumos/1/R2128-1.pdf Saraiva, M. F. O., Oliveira Filho, K. S., & Muller, A. M. (2014). Fundamentos de As- tronomia e Astrofísica para EAD. Porto Alegre: UFRGS. Recuperado em 06 ago. 2020, de www.if.ufrgs.br/~fatima/faad.htm Silva, T. P. (2015). Nossa posição no Universo: uma proposta de sequência didática para o ensino de Astronomia no Ensino Médio. Dissertação de Mestrado, Universidade Federal do Espírito Santo, Vitória. Silveira, F. L. (2017). Sobre a forma da Terra. Física na Escola, 15(2), 4-14. Recupera- do em 24 abr. 2020, de www1.fisica.org.br/fne/phocadownload/Vol15-Num2/a02- low.pdf Stellarium.org. (2020). Stellarium 0.20.1: software de código aberto do tipo planetário. Recuperado em 24 abr. 2020, de www.stellarium.org/pt_BR/ Artigo recebido em 14/05/2020. Aceito em 14/05/2020. Aceito em 14/05/2020. Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020 Artigo recebido em 14/05/2020. 112 Atividades investigativas na formação inicial de professores da educação básica em Astronomia: o movimento aparente do Sol no céu e a duração dos dias e noites Varella, I. G. (2003). Constelações do Zodíaco. Recuperado em 08 ago. 2020, de www.uranometrianova.pro.br/astronomia/AA001/zodiaco.htm Vidigal, W. Q. (2019). Elaboração e aplicação de atividades investigativas na forma- ção inicial de professores da Educação Básica em Astronomia. Dissertação de Mestra- do, Universidade Federal do Espírito Santo, Vitória. Recuperado em 13 maio 2020, de www.ensinodefisica.ufes.br/pt-br/pos-graduacao/PPGEnFis/detalhes-da-tese?id=13609 Vidigal, W. Q. (2019). Elaboração e aplicação de atividades investigativas na forma- ção inicial de professores da Educação Básica em Astronomia. Dissertação de Mestra- do, Universidade Federal do Espírito Santo, Vitória. Recuperado em 13 maio 2020, de www.ensinodefisica.ufes.br/pt-br/pos-graduacao/PPGEnFis/detalhes-da-tese?id=13609 Zômpero, A. F., & Laburú, C. E. (2011). Atividades investigativas no ensino de ciên- cias: aspectos históricos e diferentes abordagens. Ensaio: Pesquisa em Educação em Ciências, 13(3), 67-80. Recuperado em 26 jul. 2020, de https://doi.org/10.1590/1983- 21172011130305 Zômpero, A. F., & Laburú, C. E. (2011). Atividades investigativas no ensino de ciên- cias: aspectos históricos e diferentes abordagens. Ensaio: Pesquisa em Educação em Ciências, 13(3), 67-80. Recuperado em 26 jul. 2020, de https://doi.org/10.1590/1983- 21172011130305 Zômpero, A. F., & Laburú, C. E. (2012). Implementação de atividades investigativas na disciplina de Ciências em escola pública: uma experiência didática. Investigações em Ensino de Ciências, 17(3), 675-684. Recuperado em 26 jul. 2020, de www.if.ufrgs.br/ cref/ojs/index.php/ienci/article/view/181 Artigo recebido em 14/05/2020. Aceito em 14/05/2020. 113 Revista Latino-Americana de Educação em Astronomia - RELEA, n. 30, p. 83-113, 2020
https://openalex.org/W1607635299	https://www.frontiersin.org/articles/10.3389/fphy.2015.00040/pdf	English	null	Broad current sheets, current bifurcation, and collisionless reconnectionâ€”An Opinion on â€œOnset of fast magnetic reconnection via subcritical bifurcationâ€ by Z. Guo and X. Wang	Frontiers in physics	2,015	cc-by	3,043	Edited by: Hermann Lühr, Deutsches GeoForschungsZentrum GFZ, Germany In the absence of resistive friction and Ohmic losses, the agent transforming the stored energy is the current-related magnetic ﬁeld. Contacting ﬁelds of opposite direction may reorganize, locally annihilate the antiparallel components and provide the liberated energy for direct plasma heating and particle/plasma acceleration. In this case reconnection may become very fast proceeding on electron-plasma time scales. Its onset is subject to the scales of the current layer in that it goes on. Distributed currents ﬁlling the entire bulk plasma do barely support reconnection. Also ﬁeld- aligned currents ﬂowing along strong magnetic ﬁelds experience negligible reconnection even when restricted to ﬂux tubes, unless the current is so strong that it winds the magnetic ﬁeld up into spirals. This is believed to occur in stellar magnetic loops like on the sun during solar eruptions. Reviewed by: Stephan Christoph Buchert, Swedish Institute of Space Physics, Sweden Correspondence: Rudolf A. Treumann, artt@issibern.ch The relevant set-ups for collisionless reconnection are currents ﬂowing across moderately strong magnetic ﬁelds when two plasmas carrying ﬁelds of opposite directions encounter each other, the classical model case of reconnection. Such cases are abundant in space and all kinds of astrophysical magnetized objects. Oppositely polarized solar magnetic loops impacting on each other are but one example. The currents produced in the contact region of the two plasmas are fed by the mechanical plasma motion and are restricted to a ﬁnite range across the magnetic ﬁeld only. Specialty section: This article was submitted to Space Physics, a section of the journal Frontiers in Physics Received: 20 April 2015 Accepted: 26 May 2015 Published: 08 June 2015 Specialty section: This article was submitted to Space Physics, a section of the journal Frontiers in Physics The important parameter for collisionless reconnection to occur is the transverse scale of the currents (see Figure 1). In a broad current layer this scale exceeds the ion gyroradius. Here an mhd-ﬂuid description is appropriate since all particles contributing to the current are magnetized. Such layers are subject to “slow reconnection” only if plasma instabilities give rise to anomalous diﬀusion. Narrower layers of widths in the intermediate region of scales between ion and electron gyroradii require distinguishing between electron and ion dynamics. Ions become non-magnetic while electrons maintain their magnetization. Finally, on scales below the electron gyroradius electrons also become non-magnetic. OPINION published: 08 June 2015 doi: 10.3389/fphy.2015.00040 Broad current sheets, current bifurcation, and collisionless reconnection—An Opinion on “Onset of fast magnetic reconnection via subcritical bifurcation” by Z. Guo and X. Wang Rudolf A. Treumann and Wolfgang Baumjohann Space Research Institute, Austrian Academy of Sciences, Graz, Austria Space Research Institute, Austrian Academy of Sciences, Graz, Austria Keywords: collisionless reconnection, current bifurcation, turbulence, magnetotail reconnection, free energy of current layer Magnetic reconnection in current carrying plasmas, in particular when the plasma is collisionless, is an eﬃcient channel of dissipating the energy temporarily stored in electric currents that are generated in one or the other way in the mechanical plasma ﬂow. When collisions (including anomalous) dominate, reconnection is a slow diﬀusion process, proceeding on resistive time-scales and properly described by ﬂuid theory [1]. Received: 20 April 2015 Accepted: 26 May 2015 Published: 08 June 2015 Received: 20 April 2015 Accepted: 26 May 2015 Published: 08 June 2015 Current layers that narrow that their transverse widths drop below the electron gyroradius are completely determined by electron inertia and undergo prompt collisionless reconnection for the reason that electrons non- diﬀusively pass across the layer, crossing the transition region from magnetic ﬁelds of one polarity Citation: Electron-gradient-drift current (green Je∇of magnetized electrons in current-density gradient are the second current source. In such sheets not ion-acoustic or Buneman instabilities can be generated. They are unstable with respect to the Lower-hybrid-drift instability (LHDI) which maximizes in the steepest density gradient (red vertical bar) and may cause current bifurcation shown as the blue current proﬁle corresponding to a weak deformation of magnetic ﬁeld strength (dashed blue line) and a weak antiparallel magnetic component only. Brown J: Narrow current sheets of widths comparable to λe have very steep current density proﬁle and are carried mainly by non-magnetized electrons. Such sheets are subject to several kinds of electron inertia driven instabilities causing electron viscosity. Magnetic ﬁelds in such sheets are antiparallel across the sheet and thus unstable to prompt collisionless reconnection. Observations in space show that electrons, in addition to being magnetic, are indeed cooler than ions by several times but less than up to an order of magnitude in temperature [5]. The important conclusion is that for these reasons neither Buneman nor ion acoustic waves can be excited here as their instability conditions are not met. The relevant waves are modiﬁed-two stream (lower-hybrid-drift) waves with frequency near the lower hybrid frequency, and possibly also weak oblique whistlers with, however, energy ﬂow along the ﬁeld. In this region no reconnection takes place. There are no oppositely directed magnetic ﬁeld components which could support reconnection nor any reason for liberating electrons from magnetic slavery. Just slow ﬁeld diﬀusion becomes possible. Reconnection is strictly restricted to narrow current layers inside the electron inertial scale distance. This has been conﬁrmed many times [6] by numerical PIC simulations with and without guide ﬁelds. to the opposite. Here electrons escape magnetic slavery, allowing the opposite magnetic ﬁelds to get in touch, annihilate and quickly though not instantaneously to reorganize. This process, present in collisionless reconnection, is based on generation of electron viscosity [2, 3] in the thin current layers. The timescale of this fast local reconnection process is determined by the plasma being a dielectric. Thus, it is at the electron plasma time scale. The authors of the paper in question [7] make a very interesting and also very important case of the dynamics of a current layer of width in this intermediate region referring to inertial scales rather than gyro-radii, which is well justiﬁed for plasma-β ∼1. Citation: Treumann RA and Baumjohann W (2015) Broad current sheets, current bifurcation, and collisionless reconnection—An Opinion on “Onset of fast magnetic reconnection via subcritical bifurcation” by Z. Guo and X. Wang. Front. Phys. 3:40. doi: 10.3389/fphy.2015.00040 June 2015 \| Volume 3 \| Article 40 Frontiers in Physics \| www.frontiersin.org 1 Current sheets, bifurcation, collisionless reconnection Treumann and Baumjohann Collisionless currents of widths on the intermediate scale between the ion and electron gyro-radii consist of three parts: the primary unmagnetized ion current which is driven by the convection electric ﬁeld of the ﬂow of the two approaching oppositely magnetized plasma streams. Pushing the nonmagnetic ions along this ﬁeld carries the current while, at the same time, accelerates the ions over one ion inertial length thereby heating them up. The second contribution is the magnetized electron- pressure-drift current which pushes the electrons in the direction opposite to the ion ﬂow thus as well contributing to the main current. Finally one has the Hall currents [4] that are carried by magnetized electrons in the crossed convection electric and magnetic ﬁelds. The Hall currents stir the electrons around, causing a weak quadrupolar Hall magnetic ﬁeld. They close most eﬃciently along the magnetic ﬁeld to regions away from the current center, while giving rise to (kinetic) Alfvén waves. The Alfvén waves are kinetic for the reason that ions are nonmagnetic and the transverse size of their source region is of the order of the ion inertial scale thus matching the required transverse wavelength of kinetic Alfvén waves. They carry the ﬁeld-aligned currents, transport hot electrons away from the current sheet toward the footpoints of the magnetic ﬁeld and replace them by cool electrons from the environment thereby cooling the electron component. FIGURE 1 \| Schematic of the three cases of current sheets J of FIGURE 1 \| Schematic of the three cases of current sheets J of different widths but same strength. Open J: Broad current sheet, width > ion inertial length λi. The current is distributed, density gradients are very weak. Such current sheets are either stable or undergo mechanical turbulence when evolving large-scale eddies. Black J: Current sheet of intermediate width λe < width < λi between electron and ion inertial scales (green background). In such layers ions are non-magnetic, while electrons are magnetized. The current is carried by nonmagnetic ions (red Ji) which are accelerated in the convection electric ﬁeld (E). Frontiers in Physics \| www.frontiersin.org Citation: The ﬁnding of the authors is that in the domain of scales of interest the current layer is unstable with respect to bifurcation/splitting. The problem with that paper is, however, that the authors insist on referring to reconnection. As explained above, in the intermediate region there is no reconnection. Moreover reference to ion-acoustic or Buneman modes is unjustiﬁed for the conditions are not satisﬁed for these waves to become unstable. Pressure-gradient drifts are slow. In addition pressure gradient- current driven whistlers are very weak with group velocity parallel only to the magnetic ﬁeld. This, however, does not imply the proposed mechanism to be wrong. It is only misplaced. It refers to current layers which may undergo reconnection somewhere while reconnection is not the central point of what the authors really demonstrate. At the contrary, since electrons are magnetic while ions are not, the intermediate scale current is clearly unstable though neither to Buneman nor ion-acoustic waves. It is unstable to the modiﬁed-two stream (lower-hybrid drift) instability which has a very low threshold such that one expects the current necessarily to bifurcate once the author’s mechanism is applied to lower-hybrid-drift waves. The growth rate of the modiﬁed two- stream/lower-hybrid-drift instability is roughly the lower-hybrid frequency [9] which it can even slightly exceed. Using it instead of ion-acoustic waves should lead to similar results. Asking under which conditions current layer bifurcation by the originally proposed mechanism [7] could indeed take place, causing multiple break-oﬀof the current, one is led to consider current layers that become driven extremely hard by persistent external forces like relativistic convection in relativistic plasma encounters. In this case the pressure gradients on the edges of the layer could possibly become comparable to the size of the layer causing large gradient drifts. However, in such cases other eﬀects like shock waves will come into play as well modifying the evolution of the current layer. In summary, the proposed mechanism of current bifurcation [7] resulting from phase transition is intriguing in particular as such bifurcations have indeed been observed in near- Earth space suggesting that they are a common feature of moderately broad current layers. It has always been a problem of understanding what happens to broad current layers and how they may reorganize into narrow once which may then undergo reconnection and dissipate their energy. Citation: Their ﬁnding is, using a combination of ion- acoustic/Buneman waves and whistlers, that such moderately broad current sheets are nonlinearly unstable. They should break oﬀ, as the authors suggest, into many narrower current ﬁlaments and thus cause fast (collisionless) reconnection. Electron viscosity has three sources: the ﬁrst source is the anisotropy of the electron thermal pressure which under collisionless conditions in the local magnetic frame is diagonal, while in the frame of the central current layer it assumes all six independent components for the reason that the magnetic ﬁeld is locally inclined against the current frame. The second source is electron inertia leading to an inertial contribution to the electron current. The third source is due to so-called Speiser orbits of electrons meandering in the center of the current layer by experiencing mirror reﬂection from the oppositely directed magnetic ﬁelds outside the electron inertial scale width of the current sheet center. The physics of reconnection takes place in this region. Reference to spatial bifurcation of the current is most interesting as it strongly supports observations of current layer bifurcation [8] respectively splitting of the main current layer, an important and so far barely understood phenomenon to which the present paper of the authors indeed makes an important contribution. The author’s suggestion is based on the derivation of a Landau–Ginzburg equation for the electron current speed June 2015 \| Volume 3 \| Article 40 Frontiers in Physics \| www.frontiersin.org 2 Current sheets, bifurcation, collisionless reconnection Treumann and Baumjohann conﬁrmed [11] of being caused by the three-dimensionality of the reconnection process in the main current layer center below the electron inertial scale. Here, under 3d-conditions, collisionless reconnection evolves turbulently for a number of various reasons, in particular for the freedom gained by including the third dimension. In three dimensions it is reconnection in the electron inertial scale current layers that causes the turbulent break-oﬀof the current. Another reason for the appearance of many electron scale current ﬁlaments in collisionless plasma is genuine plasma turbulence resulting from decaying and inertial range cascading of large eddies down to the electron scale [10]. This has primarily nothing in common with reconnection though, however, reconnection becomes the ultimate dissipation process of turbulence once the current eddy ﬁlaments approach the electron inertial scale. which allows constructing the thermodynamic free energy. Discussion of this energy, as usual, allows distinguishing between stable and unstable regions. Citation: Classical electrodynamics tells that parallel currents attract each other until merging into a single broad current. The usual way of approaching the problem of production of narrow current ﬁlaments is via the assumption that the plasma undergoes the transition to turbulence, generating large mechanical eddies in the ﬂow which cascade down to electron scale structures. Examples are found in solar wind turbulence for which it has been shown that the dissipation is via reconnection [10]. This, however, requires a source of turbulence which is thought to be found in the generation of mhd-waves in the bulk plasma. The value of this investigation thus lies not in a contribution to reconnection, a process that today is quite well understood playing an important role in collisionless dissipation of magnetic turbulence [10, 11]. The theory rather provides a consistent view of bifurcation of moderately thick current layers, an observational fact in near-Earth space [8]. The location of strongest lower-hybrid-drift instability is at the pressure-gradient turning point on the current proﬁle. In the vicinity of this place the current will break oﬀ, according to the theory in Guo and Wang [7], splitting into two current maxima separated by, say, one ion inertial scale or a larger fraction of it. This splitting produces weak antiparallel magnetic ﬁeld components only which locally act to weaken the main magnetic ﬁeld a bit. Reconnection among these weak components may occur in the centers of the two bifurcation branches on scales less than the electron inertial scale, if only electron viscosity is generated there and becomes large enough. This is barely the case because the weak antiparallel ﬁeld components produced in the bifurcation are insuﬃcient for demagnetizing the electrons in this region. Observations in the Earth’s magnetic tail [8] do not support such an eﬀect though clearly ﬁnd current bifurcation. In any case any possible annihilation of antiparallel magnetic ﬁelds will be weak only, dissipating little of the main current energy. Any mechanism is of large interest demonstrating the splitting of currents into few or many narrow layers. Guo and Wang [7] show a way out of this dilemma, though under barely realized conditions in the currents sheet which, in contrast to turbulence, do not necessarily lead to reconnection. It will be very interesting to apply this mechanism to the more realistic conditions that are prevalent in broader than electron scale current layers. Frontiers in Physics \| www.frontiersin.org Citation: It is also interesting to learn that phase transitions can, in principle, take place in high temperature plasmas as well which is something which was suspected but never demonstrated suﬃciently convincingly. Breaking oﬀ into many small current ﬁlaments in reconnection is observationally found and in 3d-simulations June 2015 \| Volume 3 \| Article 40 3 Current sheets, bifurcation, collisionless reconnection Treumann and Baumjohann Frontiers in Physics \| www.frontiersin.org June 2015 \| Volume 3 \| Article 40 References 8. Runov A, Nakamura R, Baumjohann W, Zhang TL, Volwerk M, Eichelberger H-U, et al. Cluster observations of a bifurcated current sheet. Geophys Res Lett. (2003) 30:1036. doi: 10.1029/2002GL016136 1. Biskamp D. Magnetic Reconnection. Cambridge: Cambridge University Press (2000). 9. Davidson RD. Quasi-linear stabilization of lower-hybrid-drift instability. Phys Fluids (1978) 21:1373. doi: 10.1063/1.862379 2. Hesse M, Schindler K, Birn J, Kuznetsova M. The diﬀusion region in collisionless magnetic reconnection. Phys Plasmas (1999) 6:1781. doi: 10.1063/1.873436 10. Treumann RA, Baumjohann W. Ideal MHD turbulence: the inertial range spectrum with collisionless dissipation, Front Phys. (2015) 03:00022. doi: 10.3389/fphys.2015.00022 3. Treumann RA, Baumjohann W. Superdiﬀusion revisited in view of collisionless reconnection, Ann Geophys. (2014) 32:643. doi: 10.5194/angeo- 32-643-2014 11. Karimabadi H, Roytershteyn V, Daughton W, Liu Y-H. Recent evolution in the theory of magnetic reconnection and its connection with turbulence. Space Sci Rev. (2013) 178:307. doi: 10.1007/s11214-013-0021-7 11. Karimabadi H, Roytershteyn V, Daughton W, Liu Y-H. Recent evolution in the theory of magnetic reconnection and its connection with turbulence. Space Sci Rev. (2013) 178:307. doi: 10.1007/s11214-013-0021-7 4. Øieroset M, Phan TD, Fujimoto M, Lin RP, Lepping RP. In situ detection of collisionless reconnection in the Earth’s magnetotail. Nature (2001) 412:414. doi: 10.1038/35086520 Conﬂict of Interest Statement: 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. 5. Artemyev AV, Baumjohann W, Petrukovich AA, Nakamura R, Dandouras I, Fazakerley A, et al. Proton/electron temperature ratio in the magnetotail. Ann Geophys. (2011) 29:2253 doi: 10.5194/angeo-29-2253-2011 6. Pritchett PL. Onset and saturation of guide-ﬁeld magnetic reconnection. Phys Plasmas (2005) 12:062301. doi:10.1063/1. 1914309 6. Pritchett PL. Onset and saturation of guide-ﬁeld magnetic reconnection. Phys Plasmas (2005) 12:062301. doi:10.1063/1. 1914309 Copyright © 2015 Treumann and Baumjohann. 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) or licensor 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. 7. Guo Z, Wang X. Onset of fast magnetic reconnection via subcritical bifurcation. Front Phys. (2015) 03:00018. doi: 10.3389/fphys.2015. 00018 June 2015 \| Volume 3 \| Article 40 Frontiers in Physics \| www.frontiersin.org 4
https://openalex.org/W2012269418	https://pure.eur.nl/ws/files/47590224/REPUB_88012-OA.PDF	English	null	Levels of Alpha-Toxin Correlate with Distinct Phenotypic Response Profiles of Blood Mononuclear Cells and with agr Background of Community-Associated Staphylococcus aureus Isolates	PloS one	2,014	cc-by	8,720	Abstract Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data a Supporting Information files. Funding: This work was supported by grants from Karolinska Institutet (ANT), Swedish Research Council (521-2011-2540, ANT; 521-2009-3900 MS), The Swedish Research Links Programme (348-2009-6501; ANT & GA), Knut and Alice Wallenberg Foundation, European Union FP7 INFECT-project, NPP grant New INDIGO ERA Net, Hospices Civils de Lyon (GL, FV) and INSERM (GL, FV). 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. * Email: anna.norrby-teglund@ki.se * Email: anna.norrby-teglund@ki.se The pathogenesis of invasive staphylococcal infections involves a variety of virulence factors. In severe invasive infections, several exotoxins have been implicated in disease pathogenesis, such as superantigens [5–7], as well as the cytotoxins Panton Valentine Leukocidin (PVL) [8–10], alpha-toxin (a-toxin) [8,10], and phenol soluble modulins (PSMs) [10,11]. There are to date 24 distinct superantigens identified in S. aureus, including the well charac- terized TSST-1 and the staphylococcal enterotoxins (SE) A–C. Superantigens have been attributed a central role in mediating the systemic toxicity and shock by virtue of their ability to induce hyper-inflammatory responses [6,7], and they have also been implicated in severe pulmonary infections [5]. Cytotoxins have been associated with cell and tissue injury as well as inflammation Srikanth Mairpady Shambat1, Axana Haggar1, Francois Vandenesch2, Gerard Lina2, Willem J. B. van Wamel3, Gayathri Arakere4, Mattias Svensson1, Anna Norrby-Teglund1* 1 Karolinska Institutet, Centre for Infectious Medicine, Stockholm, Sweden, 2 CIRI, International Center for Infectiology Research, LabEx Ecofect, Universite´ Lyon 1, Inserm U1111, Ecole Normale Supe´rieure de Lyon, CNRS UMR5308, Centre National de Re´fe´rence des Staphylocoques, Hospices civils de Lyon, Lyon, France, 3 Department of Medical Microbiology and Infectious Diseases, Erasmus Medical Centre, Rotterdam, Netherlands, 4 Society for Innovation and Development, Indian Institute of Science Campus, Bangalore, India Citation: Mairpady Shambat S, Haggar A, Vandenesch F, Lina G, van Wamel WJB, et al. (2014) Levels of Alpha-Toxin Correlate with Distinct Phenotypic Response Profiles of Blood Mononuclear Cells and with agr Background of Community-Associated Staphylococcus aureus Isolates. PLoS ONE 9(8): e106107. doi:10.1371/ journal.pone.0106107 Abstract Epidemiological studies of Staphylococcus aureus have shown a relation between certain clones and the presence of specific virulence genes, but how this translates into virulence-associated functional responses is not fully elucidated. Here we addressed this issue by analyses of community-acquired S. aureus strains characterized with respect to antibiotic resistance, ST types, agr types, and virulence gene profiles. Supernatants containing exotoxins were prepared from overnight bacterial cultures, and tested in proliferation assays using human peripheral blood mononuclear cells (PBMC). The strains displayed stable phenotypic response profiles, defined by either a proliferative or cytotoxic response. Although, virtually all strains elicited superantigen-mediated proliferative responses, the strains with a cytotoxic profile induced proliferation only in cultures with the most diluted supernatants. This indicated that the superantigen-response was masked by a cytotoxic effect which was also confirmed by flow cytometry analysis. The cytotoxic supernatants contained significantly higher levels of a-toxin than did the proliferative supernatants. Addition of a-toxin to supernatants characterized as proliferative switched the response into cytotoxic profiles. In contrast, no effect of Panton Valentine Leukocidin, d-toxin or phenol soluble modulin a-3 was noted in the proliferative assay. Furthermore, a significant association between agr type and phenotypic profile was found, where agrII and agrIII strains had predominantly a proliferative profile whereas agrI and IV strains had a predominantly cytotoxic profile. The differential response profiles associated with specific S. aureus strains with varying toxin production could possibly have an impact on disease manifestations, and as such may reflect specific pathotypes. Citation: Mairpady Shambat S, Haggar A, Vandenesch F, Lina G, van Wamel WJB, et al. (2014) Levels of Alpha-Toxin Correlate with Distinct Phenotypic Response Profiles of Blood Mononuclear Cells and with agr Background of Community-Associated Staphylococcus aureus Isolates. PLoS ONE 9(8): e106107. doi:10.1371/ journal.pone.0106107 Editor: J. Ross Fitzgerald, University of Edinburgh, United Kingdom Editor: J. Ross Fitzgerald, University of Edinburgh, United Kingdom Editor: J. Ross Fitzgerald, University of Edinburgh, United Kingdom Received May 9, 2014; Accepted July 28, 2014; Published August 28, 2014 Copyright: 2014 Mairpady Shambat 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. authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its les. Levels of Alpha-Toxin Correlate with Distinct Phenotypic Response Profiles of Blood Mononuclear Cells and with agr Background of Community-Associated Staphylococcus aureus Isolates Srikanth Mairpady Shambat1, Axana Haggar1, Francois Vandenesch2, Gerard Lina2, Willem J. B. van Wamel3, Gayathri Arakere4, Mattias Svensson1, Anna Norrby-Teglund1* Flow cytometry analysis Stimulated PBMC were washed and incubated for 30 minutes on ice with directly conjugated antibodies (CD3: SK7, R&D Systems; CD45: T29/33, BD Biosciences; HLA-DR: TU36, Life technologies) in combination with a dead cell marker (Live/Dead Fixable near IR; Molecular Probes). Analyses were done using a Beckton Dickinson LSRII SORP flow cytometer and FlowJo 9.5.3. A clinical USA300 strain (LUG2012) from a patient from South-Ouest of France and its isogenic mutants deficient for either a-toxin (LUG2209) or PVL (LUG2040) were also included to confirm toxin-mediated effects. The deletion mutants were obtained by using pMAD [27] carrying the replacement cassette that was electroporated to RN4220 recipient strain and then to LUG2012. Growth at non-permissive temperature (44uC) was followed by several subcultures at 30uC and 37uC to favor double crossing over as previously described [28]. Validation of deletion Toxin determination Bacterial supernatants were assessed for the presence of selected exotoxins, including a-toxin, staphylococcal enterotoxin B (SEB), and toxic shock syndrome toxin 1 (TSST-1), using a multiplex competition immunoassay based on Luminex technology [29], and the amounts of a-toxin and PVL in the bacterial supernatants were determined by specific ELISA as previously described [30,31] using specific antibodies kindly provided, respectively, by GSK Biologicals Inc. (USA) and bioMerieux R&D Immunodiag- nostic (France). S. aureus infected patients have been shown to develop antibodies against both superantigens and cytotoxins [20–24]; thus demonstrating that patients are exposed to a combination of exotoxins secreted by the strains during infection. Exactly how these different combinations of toxins affect virulence and disease outcome is, as of yet, not fully elucidated. One example of interactions between superantigens and cytotoxins was reported by Broshnan et al [25] who showed that cytolysins promoted increased penetrance of superantigens in mucosa. Here we set out to explore how diversity in exotoxin profiles among CA S. aureus strains translates into virulence-associated functional responses. To this end, we determined the effect on peripheral blood mononuclear cells (PBMC) elicited by CA S. aureus strains of different sequence types (ST) with distinct agr allelic types and toxin gene profiles. Our data revealed that the strains induced differential responses in PBMC, characterized by either cytotoxic or proliferative responses, which were linked to specific agr types and levels of a-toxin expression. PBMC proliferation assay Human PBMC were isolated from healthy donors using Lymphoprep density centrifugation. The cells were cultured in RPMI-1640 medium supplemented with 10% FCS, 10 mM of L- glutamine, Penicillin (100 U/ml)/Streptomycin (100mg/ml) and 25 mmol/L HEPES (all from Thermo Scientific HyClone, USA). PBMC were seeded at 26105 cells/well and stimulated at 37uC with serial dilutions of bacterial supernatants. After 72 hours, the cells were pulsed for 6 hours with 1 mCi/well of 3H-thymidine (Perkin-Elmer) after which 3H-uptake was measured in a beta- scintillation counter. Phytohemagglutinin-L (PHA) (1 mg/ml) (Sigma-Aldrich, St. Louis, USA) was used as a positive control for polyclonal T cell activation. The cytotoxic/inhibitory effect was tested by the addition of bacterial supernatants in combination with PHA in the proliferation assay. The bacterial culture medium CCY was included as a negative control, and was found to have negligible effect on proliferation (mean CPM 3088, 1785, 1719 for 1:50; 1:100 and 1:1000 dilution, resp.) and no inhibitory effect on PHA-induced proliferation. PBMC were also stimulated with bacterial supernatants or PHA in combination with purified a- toxin (Sigma-Aldrich, St. Louis, USA), recombinant PVL, recombinant d-toxin and purified PSM-a3 (all from IBT Bioservices, Gaithersburg, USA) and proliferation assessed. Ethics Statement This study includes blood samples from buffy coats of blood provided by the blood bank at the Karolinska University Hospital. The buffy coats were provided anonymously; hence informed consent was not required. The ethical research committee at Huddinge University Hospital (Forskningskommitte Syd) ap- proved the study. Preparation of bacterial culture supernatants The strains were cultured overnight at 37uC in 25 ml casein hydrolysate and yeast extract (CCY) medium. Cell-free superna- tants were prepared through centrifugation at 3350 g followed by filter sterilization. Introduction Staphylococcus aureus (S. aureus) is a significant cause of human infections and an emerging health problem globally due to its increasing resistance to beta-lactams (methicillin-resistant S. aureus, MRSA). A special cause of concern is the rise in community-acquired (CA) S. aureus strains, and particularly concerning are reports of specific CA-MRSA clones associated with highly aggressive infections, including necrotizing fasciitis and pneumonia in otherwise healthy individuals [1–3]. Although there are large geographical differences, epidemiological studies have shown that more than 20 distinct CA-MRSA lineages are present globally [4]. PLOS ONE \| www.plosone.org August 2014 \| Volume 9 \| Issue 8 \| e106107 1 August 2014 \| Volume 9 \| Issue 8 \| e106107 Staphylococcal Phenotypic Response Profiles of the hla (LUG2209) or lukSF-PV (LUG2040) genes were done by PCR and the production of a-toxin and PVL were determined with previously described toxin-specific ELISA (see below). in severe invasive infections, including necrotising infections of the skin and lung [10]. Several of the toxins are encoded by mobile genetic elements resulting in great diversity in toxin gene content among S. aureus strains [12,13] Many of these virulence factors are regulated and controlled by a global regulator system called the accessory gene regulator (agr) system encoded by the agr locus [14–17]. The agr locus has diverged among different S. aureus strains with polymorphism in agrBDC region, resulting in four major allelic types of agr, i.e. agrI–IV [15,17,18]. A recent study demonstrated the impact of allelic variation on agr induction dynamics, which translated into significant differences in expres- sion of several virulence factors [19]. IVIG inhibition assay PBMC were stimulated with bacterial supernatants, a-toxin or PHA in the presence or absence of different concentrations of IVIG (Gammagard S/D, Baxter). Proliferative responses were determined after 72 hours as described above. Clinical isolates Strains (n = 38) were selected from a heterogeneous cohort of CA S. aureus representing a diverse collection of isolates with different ST and toxin profiles collected from colonized individuals (nasal swabs) or patients with varying S. aureus infections in India (Table 1). The isolate collection has previously been characterized with respect to antibiotic resistance profile, molecular typing including ST, agr-types, and toxin profile as determined by either PCR for specific genes or by a DNA microarray covering 185 S. aureus genes [26]. The study also included a confirmatory cohort of 31 isolates collected from patients with CA S. aureus pneumonia included in a prospective cohort study in France [9] or from cases referred to the French national reference laboratory for staphylococci. Statistical evaluation Data were analyzed by GraphPad Prism version 4.0 for Windows (GraphPad software). Two-sided Mann-Whitney U test PLOS ONE \| www.plosone.org August 2014 \| Volume 9 \| Issue 8 \| e106107 2 Staphylococcal Phenotypic Response Profiles Table 1. Characteristics of S. aureus strains with respect to antibiotic resistance, ST and agr types, toxin genotype and production in vitro. Statistical evaluation Strains Clinical Diagnosis MSSA/MRSA agr type SPA type CC/ST type pvl lukD/E se/tsst-1/egc Etd/edinB Response Profile# Luminex* Colonizing strains Sa559 MSSA I t005 ST22/CC22 + ND 2/2/+ 2 C (96.6) a-toxin+++ Sa115 MRSA-IV I t852 ST22/CC22 + ND 2/2/+ 2 P (12.3) a-toxin + Sa165 MRSA-IV I t852 ST22/CC22 + ND 2/2/+ 2 C (95.7) a-toxin +++ Sa95 MSSA I t3840 ST672 2 + 2/2/+ 2 C (98.0) a-toxin +++ Sa134 MRSA-V I t064 ST1208/CC8 + + se-A, B, K, Q/2/2 2 P (76.5) a-toxin ++/SEB++ Sa180 MSSA I t4285 ST6 2 ND se-L/tsst-1/2 2 C (97.5) a-toxin +++/TSST+++ Sa168 MSSA I t937 ST291 2 2/+ 2/2/2 + C (96.7) a-toxin +++ Sa337 MSSA I t3096 ST291 2 ND 2/2/2 + C (96.1) a-toxin +++ Sa18 MRSA-V II t657 ST772/CC1 + 2 se-A, C, L/2/+ 2 P (27.0) a-toxin ++ Sa289 MSSA II t1839 ST772/CC1 + 2 se-A/2/+ 2 P (28.5) a-toxin ++ Sa1437 MSSA II t345 ST772/CC1 + ND se-A, C, L/tsst-1/+ 2 P (23.7) a-toxin + Sa1446 MRSA-V II t657 ST772/CC1 + ND se - A, E/2/+ 2 P (25.4) a-toxin + Sa233 MRSA-V II t657 ST772/CC1 + ND se-A, C, E, L/2/+ 2 P (30.0) a-toxin ++ Sa159 MSSA II t774 ST199/CC15 2 + 2/2/2 2 Ambigious a-toxin + Sa160 MSSA II t774 ST199/CC15 2 + 2/2/2 2 C (69.6) a-toxin + Sa32 MRSA-IV III t021 CC30 + 2 2/2/+ 2 P (21.2) a-toxin +/2 Sa368 MSSA IV t1999 CC121 + ND 2/2/+ 2 C (93.7) a-toxin +++ Sa14 MSSA IV t3204 CC121 + 2/+ se-B/2/+ 2 C (95.0) a-toxin +++/SEB+++ Disease strains Sa37 SSTI MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 P (12.8) a-toxin + Sa08 SSTI MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 C (98.0) a-toxin +++ Sa113 Brain abscess MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 C (95.0) a-toxin +++ Sa114 Cerebral abscess MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 C (95.8) a-toxin +++ SaN08 Meningitis MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 C (96.5) a-toxin +++ Sa1 Invasive infection MRSA-IV I t1309 ST672 2 + 2/2/+ 2 C (94.6) a-toxin +++ Sa754 Invasive infection MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 C/P (87.3) a-toxin +++ Sa755 Invasive infection MSSA I ND ST22/CC22 + 2 2/2/+ 2 P (3.31) a-toxin ++ Sa118 Pyomyositis MRSA-V II t657 ST772/CC1 + 2 se-A, C, L/2/+ 2 P (30.7) a-toxin ++ Sa3957 Breast abscess MRSA-V II t1387 ST772/CC1 + ND se-A, E/2/+ 2 P (21.5) a-toxin +/2 Sa3989 Pneumonia MRSA-V II t3596 ST772/CC1 + ND se-A, E/2/+ 2 P (24.8) a-toxin ++ Sa120/1 Cerebral abscess (pus) MRSA-V II t657 ST772/CC1 + ND ND ND P (24.7) a-toxin ++ Sa2332 Pleural Empyema MSSA III t021 CC30 + 2 2/2/+ 2 P (36.2) a-toxin +/2 PLOS ONE \| www.plosone.org 3 August 2014 \| Volume 9 \| Issue 8 \| e10610 Table 1. Distinct functional phenotypic profiles of clinical S. aureus strains A proliferation assay was employed in which PBMC were exposed to bacterial supernatants prepared from CA S. aureus strains. The strain collection included CA MRSA and MSSA strains of varying ST types, agr types and toxin gene profile isolated from patients or colonized individuals (Table 1). All strains, except two (strains Sa159 and Sa160), harbored superan- tigen encoding genes (Table 1). To confirm that the genes were expressed during in vitro culture, the bacterial supernatants were analyzed for toxin content using a customized luminex assay. Due to technical limitations only a selected set of exotoxins, i.e. a-toxin, SEB and TSST-1, were included for which the protein detection concurred with the toxin gene profile of respective strain (Table 1). g p p ( ) PBMC from four different donors were stimulated with serial dilutions of the bacterial supernatants and proliferative responses assessed (Figure 1A). This assay is commonly used to functionally assess superantigen-mediated proliferation, but the response will be influenced by the presence of cytotoxins targeting PBMC. Accordingly, supernatants prepared from the different S. aureus strains induced starkly different (p,0.0001) and highly reproduc- ible response profiles (Figures 1A, 1B). While some strains induced consistently high proliferative responses at all dilutions (1:50, 1:100 and 1:1000) of the bacterial supernatants (denoted as a prolifer- ative profile) (Figures 1A, 1B), other strains induced proliferation only at the highest dilution (1:1000) of bacterial supernatants (Figures 1A, 1B). Thus, demonstrating that the supernatants contain superantigens that trigger proliferation, but this activity is masked in more concentrated supernatants from certain strains, potentially through toxin-mediated cytotoxicity. To investigate this further, proliferation was assessed following stimulations of PBMC with the polyclonal T cell activator PHA in combination with supernatants displaying a proliferative or cytotoxic profile (Figure 1C). Using the 1:50 dilutions of bacterial supernatants, the PHA-response was completely abolished by the cytotoxic supernatants, whereas augmented by the proliferative supernatants (Figure 1D). To directly assess whether bacterial supernatants were cytotoxic, flow cytometry analysis was applied on cells stimulated with supernatants and subsequently stained with a dead cell marker in combination with antibodies directed towards defined cell markers. PBMC stimulated with a prolifer- ative supernatant showed a relative expansion of the T cell population similar to that seen for PHA (Figure 1E). Notably, stimulation with a cytotoxic supernatant resulted in extensive cell death as almost all cells stained positive for the dead cell marker (Figure 1E). Staphylococcal Phenotypic Response Profiles Cel stimulated with supernatants (1:50 dilution) were also analyzed by flow cytometry and the value in parenthesis shows % cells stained positive for dead cell marker. Analyses of selected exotoxins including a-toxin (alpha-toxin), SEB (staphylococcal enterotoxin B) and TSST-1 (toxic shock syndrome toxin 1) are assessed in overnight bacterial culture supernatants by luminex. Only posit results are indicated and based on the dilution series response a semi quantitative measure is given as +/2, +, ++, +++. Staphylococcal Phenotypic Response Profiles Staphylococcal Phenotypic Response Profiles ant S. aureus; MSSA, methicillin-susceptible S. aureus; ND, not determined; se, staphylococcal enterotoxin;P, proliferative; C, cytotoxic; SSTI, skin and soft tissue infection. s determined by the pattern of proliferative responses elicited by different dilutions of bacterial supernatants. A proliferative profile was denoted if all dilutions elicited a proliferative response. A cytotoxic a) proliferation was only noted at the highest dilution of the supernatants, and b) the supernatants (1:50 dilution) resulted in inhibition of PHA-induced responses in co-stimulation experiments. Cells atants (1:50 dilution) were also analyzed by flow cytometry and the value in parenthesis shows % cells stained positive for dead cell marker. xotoxins including a-toxin (alpha-toxin), SEB (staphylococcal enterotoxin B) and TSST-1 (toxic shock syndrome toxin 1) are assessed in overnight bacterial culture supernatants by luminex. Only positive d based on the dilution series response a semi quantitative measure is given as +/ + ++ +++ or Fisher’s exact test were used for comparison between two groups. Comparisons between multiple groups were done using ANOVA and Dunn’s multiple comparison test. Differences were considered significant when p,0.05. Strains Clinical Diagnosis MSSA/MRSA agr type SPA type CC/ST type pvl lukD/E se/tsst-1/egc Etd/edinB Response Profile# Luminex SaP1 Suture induced infiltrate MSSA III t127 ST1 2 ND se-D, E/2/2 2 P (80.5) ND SaP3 Keratitis MSSA III t8078 ST1 2 ND se-D/2/2 2 P (27.1) ND SaP6 Keratitis MSSA III t127 ST1 2 ND se-D, E/2/2 2 P (60.1) ND SaP7 Orbital abscess MRSA-V III t2526 ST88 + ND se-D/2/+ 2 C (97.3) ND Sa753 Necrotizing pneumonia MSSA IV t159 ST121 + + 2/2/+ 2 C (98.3) a-toxin +++ Sa796 Necrotizing pneumonia MSSA IV t159 ST121 + + 2/2/+ 2 C (97.4) a-toxin +++ Sa1059 Invasive inf. MSSA IV t159 ST121 + + 2/2/+ 2 C (98.5) a-toxin +++ MRSA, methicillin-resistant S. aureus; MSSA, methicillin-susceptible S. aureus; ND, not determined; se, staphylococcal enterotoxin;P, proliferative; C, cytotoxic; SSTI, skin and soft tissue infection. #The response profile is determined by the pattern of proliferative responses elicited by different dilutions of bacterial supernatants. A proliferative profile was denoted if all dilutions elicited a proliferative response. A cytoto pattern was denoted if a) proliferation was only noted at the highest dilution of the supernatants, and b) the supernatants (1:50 dilution) resulted in inhibition of PHA-induced responses in co-stimulation experiments. August 2014 \| Volume 9 \| Issue 8 \| e106107 Statistical evaluation Characteristics of S. aureus strains with respect to antibiotic resistance, ST and agr types, toxin genotype and production in vitro. Strains Clinical Diagnosis MSSA/MRSA agr type SPA type CC/ST type pvl lukD/E se/tsst-1/egc Etd/edinB Response Profile# Luminex* Colonizing strains Sa559 MSSA I t005 ST22/CC22 + ND 2/2/+ 2 C (96.6) a-toxin+++ Sa115 MRSA-IV I t852 ST22/CC22 + ND 2/2/+ 2 P (12.3) a-toxin + Sa165 MRSA-IV I t852 ST22/CC22 + ND 2/2/+ 2 C (95.7) a-toxin +++ Sa95 MSSA I t3840 ST672 2 + 2/2/+ 2 C (98.0) a-toxin +++ Sa134 MRSA-V I t064 ST1208/CC8 + + se-A, B, K, Q/2/2 2 P (76.5) a-toxin ++/SEB++ Sa180 MSSA I t4285 ST6 2 ND se-L/tsst-1/2 2 C (97.5) a-toxin +++/TSST+++ Sa168 MSSA I t937 ST291 2 2/+ 2/2/2 + C (96.7) a-toxin +++ Sa337 MSSA I t3096 ST291 2 ND 2/2/2 + C (96.1) a-toxin +++ Sa18 MRSA-V II t657 ST772/CC1 + 2 se-A, C, L/2/+ 2 P (27.0) a-toxin ++ Sa289 MSSA II t1839 ST772/CC1 + 2 se-A/2/+ 2 P (28.5) a-toxin ++ Sa1437 MSSA II t345 ST772/CC1 + ND se-A, C, L/tsst-1/+ 2 P (23.7) a-toxin + Sa1446 MRSA-V II t657 ST772/CC1 + ND se - A, E/2/+ 2 P (25.4) a-toxin + Sa233 MRSA-V II t657 ST772/CC1 + ND se-A, C, E, L/2/+ 2 P (30.0) a-toxin ++ Sa159 MSSA II t774 ST199/CC15 2 + 2/2/2 2 Ambigious a-toxin + Sa160 MSSA II t774 ST199/CC15 2 + 2/2/2 2 C (69.6) a-toxin + Sa32 MRSA-IV III t021 CC30 + 2 2/2/+ 2 P (21.2) a-toxin +/2 Sa368 MSSA IV t1999 CC121 + ND 2/2/+ 2 C (93.7) a-toxin +++ Sa14 MSSA IV t3204 CC121 + 2/+ se-B/2/+ 2 C (95.0) a-toxin +++/SEB+++ Disease strains Sa37 SSTI MRSA-IV I t852 ST22/CC22 + 2 2/2/+ 2 P (12.8) a-toxin + PLOS ONE \| www.plosone.org 3 PLOS ONE \| www.plosone.org 3 Staphylococcal Phenotypic Response Profiles Staphylococcal Phenotypic Response Profiles Figure 1. Proliferative or cytotoxic responses elicited by S. aureus strains. Human peripheral blood mononuclear cells (PBMC) isolated from healthy donors were stimulated with dilution series (1:1000 (open), 1:100 (stripped) and 1:50 (filled) bars) of bacterial supernatants prepared from overnight cultures of S. aureus strains. Proliferative responses were determined by 3H-thymidine uptake and are presented as mean counts per minute 6 SD. The supernatants induced either a proliferative or cytotoxic response profile. A cytotoxic profile was assigned when strains did not elicit a proliferative response in the more concentrated supernatants but only in the most diluted samples. A) Shows the response of a representative proliferative and cytotoxic supernatant in separate experiments using cells from four healthy donors. B) Scatter plot of proliferative responses induced by bacterial supernatants with a proliferative or a cytotoxic profile. Mean values of four different experiments are shown for dilutions 1:1000 (filled symbols) and 1:50 (open symbols). C) Bacterial supernatants with proliferative or cytotoxic profiles were found among both patients (n = 20) and colonized individuals (n = 17). The figure shows proliferative responses (mean 6 SD) of one representative, out of five, experiments using cells from different donors stimulated with dilution series (1:1000 (open), 1:100 (stripped) and 1:50 (filled) bars) of bacterial supernatants. D) To confirm a cytotoxic effect, proliferative responses were assessed in cells stimulated with PHA alone (the dashed line shows the average PHA response) or in combination with supernatants (proliferative and cytotoxic; dilution 1:50 (filled) bars). The figure shows one representative of three experiments using cells from different donors. E) Flow cytometry analysis on PBMC stimulated with PHA and bacterial supernatants (1:50 dilution). Total PBMC were gated based on CD45 expression and CD45 positive cells were further analysed for dead cell marker (DCM, Y axis) and CD3 (X axis) positivity. The figure shows one representative of three individual experiments using cells from different donors. F) Relation between agr types and proliferative (black bars) or cytotoxic (white bars) community S. aureus strains collected from colonizing individuals (n = 17) or patients with infections (n = 20) in India. G) Relation between agr type and proliferative or cytotoxic profiles elicited by community-acquired (CA) pneumonia patients (n = 31). Statistical significant differences were determined by use of the two-sided Mann Whitney test and Fisher’s exact test with two sided p value and p values are indicated in the figure. High a-toxin expression is linked to the cytotoxic profiles High a-toxin expression is linked to the cytotoxic profiles Quantitation of the cytotoxins a-toxin and PVL revealed that the cytotoxic supernatants had significantly higher levels of a-toxin than the proliferative, regardless of whether the strains were collected from patients, colonized individuals, or CA pneumonia (p,0.002) (Figure 2A). In contrast, there was no correlation between the response profile and PVL expression (Figure 2B) which is in line with the reported lack of susceptibility of PBMC to PVL [32]. In agreement with the data in Figure 1F demonstrating that different agr types are linked to either proliferative or cytotoxic profiles, significantly higher amounts of a-toxin were found in strains belonging to the cytotoxic agr type I and IV, as compared to the proliferative agr type II and III strains (p,0.05) (Figure 2C). In this context, no association between PVL levels and specific agr types was seen (Figure 2D). Also, expression data on the psm-a gene revealed no difference between strains eliciting either response profile (data not shown). IVIG inhibition of toxin mediated responses in PBMC Studies have demonstrated the presence of antibodies against defined S. aureus virulence factors, including a-toxin, PVL and superantigens, in intravenous polyclonal immunoglobulin (IVIG) preparations [33–37]. Here we tested whether IVIG could inhibit the toxin-mediated functional response profiles, in particular the cytotoxic effect associated with high a-toxin levels. To this end, PBMC were stimulated with PHA and different concentrations of purified a-toxin in the presence or absence of IVIG. We found a clear dose response of IVIG-mediated inhibition of a-toxin cytotoxicity, evident by increased proliferative responses (Fig- ure 4A). At the lowest concentration of a-toxin (60 ng/ml), a significant inhibition was achieved with 0.1 mg/ml IVIG (p = 0.05) while at the highest concentration of 480 ng/ml of a- toxin significant neutralization was achieved only at concentra- tions .1.0 mg/ml of IVIG (p = 0.034) (Figure 4A). Similarly we found that IVIG was able to inhibit both the proliferative as well as the cytotoxic responses elicited by superantigens and cytotoxins present in S. aureus culture supernatants (Figure 4B). These data implied that a-toxin is a main mediator of the noted cell death in PBMC. To further test this, PBMC were stimulated with PHA in combination with purified a-toxin revealing a clear dose response pattern where increasing amounts of a-toxin resulted in reduced proliferation (Figure 3A). Furthermore, supplementing a proliferative (Sa1437) or a cytotoxic (Sa368) S. High a-toxin expression is linked to the cytotoxic profiles aureus supernatant with purified a-toxin, resulted in a switch to a cytotoxic response both at the 1:1000 and 1:50 dilutions for the proliferative strain and an increased cytotoxic response at the 1:1000 dilution for the cytotoxic strain (Figure 3B). Further support for an a-toxin-mediated effect was provided by using a strain of the USA300 lineage (LUG2012), which belongs to the cytotoxic agr type I and produces high levels of a-toxin (a-toxin 83 mg/ml, PVL 14 mg/ml), and its isogenic mutants for PVL (a- toxin 54 mg/ml, PVL 0 mg/ml) and a-toxin (a-toxin 0 mg/ml, PVL 11 mg/ml). The USA300 and the PVL-deficient mutant both showed a cytotoxic profile, whereas the a-toxin mutant showed a proliferative response profile (Figure 3C). Also when PBMC were stimulated with PHA in combination with these supernatants a reduction of PHA-induced responses was noted in the presence of supernatants from USA300 and the PVL-deficient mutant (Figure 3D). In contrast, the a-toxin mutant supernatant resulted Distinct functional phenotypic profiles of clinical S. aureus strains As shown in Table 1, cell death was significantly more pronounced among cells exposed to cytotoxic, as compared to proliferative, supernatants (mean % of cells staining positive for dead cell marker: 96.3% and 33.2% for cytotoxic and proliferative supernatants, resp.; p,0.001). The two functional response profiles were found in both colonizing as well as patient isolates and there was no significant association between functional profile and ST-type/clonal com- plex or MRSA/MSSA type (Table 1). In contrast, an association with agr type was evident, and agr II and agr III strains had predominantly a proliferative profile whereas agr I and IV strains were cytotoxic (p,0.015) (Figure 1F). As these analyses were August 2014 \| Volume 9 \| Issue 8 \| e106107 August 2014 \| Volume 9 \| Issue 8 \| e106107 4 Staphylococcal Phenotypic Response Profiles Staphylococcal Phenotypic Response Profiles PLOS ONE \| www.plosone.org 5 August 2014 \| Volume 9 \| Issue 8 \| e106107 August 2014 \| Volume 9 \| Issue 8 \| e106107 August 2014 \| Volume 9 \| Issue 8 \| e106107 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org Staphylococcal Phenotypic Response Profiles d i 10 1371/j l 0106107 001 g doi:10.1371/journal.pone.0106107.g001 in an augmentation of the PHA-induced response (Figure 3D). Flow cytometry analysis further confirmed a potent cytotoxic effect of USA300 and the PVL-deficient mutant supernatants (97.2% and 85.8% dead cell marker positivity, resp.), whereas the a-toxin mutant supernatant had limited cytotoxicity (20.4% dead cell marker positivity). It should be noted that although the superna- tant of the a-toxin mutant elicited a proliferative response at all dilutions tested, the proliferative response increased with more diluted supernatants (Figure 3C); thus, indicating that there are inhibitory factors, other than a-toxin, present in the USA300 supernatant. Other cytotoxins tested, including purified PVL, PSM-a3 and d-toxin alone or in combination with PHA did not elicited any cytotoxic or proliferative responses in PBMC (Figure 3E). conducted on a highly heterogeneous strain cohort, we expanded the analyses to include a more homogenous strain cohort collected from patients with CA S. aureus pneumonia. Also in this cohort, a similar association with agr I and IV being significantly more cytotoxic than agr II or III was noted (p,0.034) (Figure 1G). August 2014 \| Volume 9 \| Issue 8 \| e106107 Discussion Statistical significant differences were determined by use of the two-sided Mann Whitney test or with ANOVA and Dunn’s multiple comparison test for comparisons of multiple groups and p values are indicated in the figure. doi:10.1371/journal.pone.0106107.g002 proliferative profile in contrast to the wildtype USA300 strain that was distinctly cytotoxic. As a vast majority (29/33) of the cytotoxic supernatants contained $225 ng/ml a-toxin, whereas all prolif- erative supernatants had levels below 221 ng/ml (in the 1:50 dilutions), we speculate that there might be a critical a-toxin concentration determining cytotoxic responses in PBMC. This underscores the value of quantifying the levels of toxins produced and secreted by the strains as this seems to dramatically influence the cellular responses and potentially virulence. response was masked in cytotoxic strains due inhibitory/cytotoxic factors present in the supernatants. This illustrates the relevance of using such a mixture of secreted factors from clinical isolates, which allows simultaneous analyses of the combined activity of toxins and which more closely reflects what the patients are exposed to. The PBMC assay provided an efficient tool to assess both superantigenic and cytotoxic effects; thereby allowing for identification of distinct phenotypic response profiles among CA S. aureus isolates. However, it should be noted that this assay involves only PBMC and hence, excludes effects of pore-forming toxins, such as PVL, LukDE and LukAB, that targets cell populations which are either missing or represent a minor subset in PBMCs [32,38,39]. Therefore, future studies should include also other clinically relevant cells, such as neutrophils, that are highly susceptible to the above mentioned cytotoxins. Similarly, a significant association between response profiles, a- toxin levels and the agr type of the strain was noted. Significantly increased a-toxin levels were found in the cytotoxic agrI and IV types, as compared to the proliferative agrII and III strains. The complexity of agr regulation of S. aureus virulence factors was emphasized in the report by Geisinger et al [19] in which protein A, a-toxin, PVL and TSST-1 were tested in congenic strains each harboring a unique agr allele (I–IV). They found significant variation in the kinetics and degree of the agr signal resulting in differential induction of specific virulence factors. Of special interest, is their finding that agr I and IV are the earliest and strongest, followed by agr II and III, which is the same hierarchical order as we report here in regards to cytotoxicity. Discussion In this study we demonstrate that CA S. aureus strains with distinct toxin profiles exhibit stable robust phenotypic profiles evident by their ability to elicit either a proliferative or cytotoxic response profile in human PBMC. All experiments were done using bacterial supernatants containing superantigens and cyto- toxins secreted by the strains. The data shows that the vast majority of strains elicit superantigenic activity as demonstrated by the induction of proliferative responses in PBMC. However, this August 2014 \| Volume 9 \| Issue 8 \| e106107 PLOS ONE \| www.plosone.org August 2014 \| Volume 9 \| Issue 8 \| e106107 6 Staphylococcal Phenotypic Response Profiles Figure 2. Levels of alpha-toxin expression correlate with cytotoxic profile and agr types. Amounts of alpha-toxin (a-toxin) (A) and Panton Valentine Leukocidin (PVL) (B) were determined by ELISA in the S. aureus bacterial supernatants (1:50 dilutions) having either a cytotoxic (C; open symbols) or a proliferative (P; filled symbols) response profile. Supernatants prepared from isolates collected from colonized individuals (n = 17), patients (n = 17), or a confirmatory cohort of CA pneumonia (n = 31) are shown. Association between amount of a-toxin (C) and PVL (D) in the S. aureus bacterial supernatants (1:50 dilutions; n = 65) with their respective agr types (agr I and IV, open symbol; agr II and III, filled symbol). Statistical significant differences were determined by use of the two-sided Mann Whitney test or with ANOVA and Dunn’s multiple comparison test for comparisons of multiple groups and p values are indicated in the figure. doi:10.1371/journal.pone.0106107.g002 Figure 2. Levels of alpha-toxin expression correlate with cytotoxic profile and agr types. Amounts of alpha-toxin (a-toxin) (A) and Panton Valentine Leukocidin (PVL) (B) were determined by ELISA in the S. aureus bacterial supernatants (1:50 dilutions) having either a cytotoxic (C; open symbols) or a proliferative (P; filled symbols) response profile. Supernatants prepared from isolates collected from colonized individuals (n = 17), patients (n = 17), or a confirmatory cohort of CA pneumonia (n = 31) are shown. Association between amount of a-toxin (C) and PVL (D) in the S. aureus bacterial supernatants (1:50 dilutions; n = 65) with their respective agr types (agr I and IV, open symbol; agr II and III, filled symbol). August 2014 \| Volume 9 \| Issue 8 \| e106107 Discussion The molecular basis for the noted association between toxin Our findings demonstrated a striking association between high a-toxin levels and a cytotoxic phenotypic profile of the S. aureus strains. Supernatants that displayed a cytotoxic profile had in average 20-fold higher a-toxin levels than the proliferative supernatants (p,0.0001). Furthermore addition of purified a- toxin to either bacterial supernatants or to PHA resulted in a dose- dependent transition from proliferative to cytotoxic response profiles. Also, a USA300 a-toxin deficient mutant revealed a August 2014 \| Volume 9 \| Issue 8 \| e106107 7 PLOS ONE \| www.plosone.org Figure 3. Alpha-toxin mediates cytotoxicity in PBMC. A) Proliferation assay using human PBMC stimulated with PHA alone or in the presence of different concentrations of a-toxin. Proliferative responses were determined by 3H-thymidine uptake and are presented as mean counts per minute 6 SD. Mean values of four different experiments from different donors are shown. B) Inhibition of PHA-induced and bacterial supernatant- induced proliferation of PBMCs by addition of increasing concentrations of a-toxin (60 to 480 ng/ml). The figure shows one representative of two experiments using cells from different donors. C) Proliferation assay using cells from two donors stimulated with different dilutions (1:1000 (open), 1:100 (stripped) and 1:50 (filled) bars) of supernatants prepared from USA300 (LUG2012), and its isogenic mutants of PVL (DlukSF-PV) and a-toxin (Dhla). D) Proliferative responses assessed in cells stimulated with PHA alone or in combination with indicated supernatants at dilutions 1:1000 (open) and 1:50 (filled) bars. The figure shows experiments using cells from two donors. E) PBMC were stimulated with increasing concentrations of PVL, PSM a3 and d-toxin at indicated concentrations (mg/ml) alone and in combination with PHA. Statistical significant differences were determined by Fisher’s exact test with two sided p value and p value is indicated in the figure. doi:10.1371/journal.pone.0106107.g003 Staphylococcal Phenotypic Response Profiles Staphylococcal Phenotypic Response Profiles Figure 3. Alpha-toxin mediates cytotoxicity in PBMC. A) Proliferation assay using human PBMC stimulated with PHA alone or in the presence of different concentrations of a-toxin. Proliferative responses were determined by 3H-thymidine uptake and are presented as mean counts per minute 6 SD. Mean values of four different experiments from different donors are shown. B) Inhibition of PHA-induced and bacterial supernatant- induced proliferation of PBMCs by addition of increasing concentrations of a-toxin (60 to 480 ng/ml). The figure shows one representative of two experiments using cells from different donors. Discussion C) Proliferation assay using cells from two donors stimulated with different dilutions (1:1000 (open), 1:100 (stripped) and 1:50 (filled) bars) of supernatants prepared from USA300 (LUG2012), and its isogenic mutants of PVL (DlukSF-PV) and a-toxin (Dhla). D) Proliferative responses assessed in cells stimulated with PHA alone or in combination with indicated supernatants at dilutions 1:1000 (open) and 1:50 (filled) bars. The figure shows experiments using cells from two donors. E) PBMC were stimulated with increasing concentrations of PVL, PSM a3 and d-toxin at indicated concentrations (mg/ml) alone and in combination with PHA. Statistical significant differences were determined by Fisher’s exact test with two sided p value and p value is indicated in the figure. doi:10.1371/journal.pone.0106107.g003 whereas agr II and III AIPs show greater variations in their primary sequence. production and agr types translating into distinct response profiles has yet to be determined but could possibly be linked to varying levels of secreted auto inducing peptides (AIP). Another possibility is that the AIP-AgrC interaction kinetics may vary between the proliferative and cytotoxic isolates, as previous studies [15,17,18,40] have shown that agr I and IV AIPs share almost identical primary sequence with only one amino acid difference Another interesting aspect is the potential link between genetic background, agr type and disease manifestation [18,41]. Although the link is far from exclusive, agr IV are often emphasized as a disease-related isolate [42,43]. Furthermore Jarraud et al [41] reported a predominance of agr II and III in TSST-1 mediated TSS and scarlet fever (90–94%, respectively). Here we report PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org August 2014 \| Volume 9 \| Issue 8 \| e106107 8 Staphylococcal Phenotypic Response Profiles gure 4. Inhibition of a-toxin and staphylococcal supernatant mediated cytotoxicity and proliferation of PBMC by IVIG. A) utralization of a-toxin induced cytotoxicity of PBMC by different concentrations of IVIG. PBMC were stimulated with PHA with increasing ncentrations of a-toxin (60 to 480 ng/ml) in the presence or absence of IVIG (0.1 to 2.5 mg/ml, as indicated). The dashed line indicates the mean A-induced proliferative response. B) Proliferative responses induced by bacterial supernatants from both proliferative and cytotoxic strains lutions 1:1000 and 1:50) in the presence or absence of different concentrations of IVIG (0.1 to 2.5 mg/ml, as indicated). 3H-thymidine uptake after hours of culture is presented as mean counts per minute 6 SD. Discussion The figure shows one representative of two experiments using cells from different nors. i:10.1371/journal.pone.0106107.g004 Figure 4. Inhibition of a-toxin and staphylococcal supernatant mediated cytotoxicity and proliferation of PBMC by IVIG. A) Neutralization of a-toxin induced cytotoxicity of PBMC by different concentrations of IVIG. PBMC were stimulated with PHA with increasing concentrations of a-toxin (60 to 480 ng/ml) in the presence or absence of IVIG (0.1 to 2.5 mg/ml, as indicated). The dashed line indicates the mean PHA-induced proliferative response. B) Proliferative responses induced by bacterial supernatants from both proliferative and cytotoxic strains (dilutions 1:1000 and 1:50) in the presence or absence of different concentrations of IVIG (0.1 to 2.5 mg/ml, as indicated). 3H-thymidine uptake after 72 hours of culture is presented as mean counts per minute 6 SD. The figure shows one representative of two experiments using cells from different donors. doi:10.1371/journal.pone.0106107.g004 Figure 4. Inhibition of a-toxin and staphylococcal supernatant mediated cytotoxicity and proliferation of PBMC by IVIG. A) Neutralization of a-toxin induced cytotoxicity of PBMC by different concentrations of IVIG. PBMC were stimulated with PHA with increasing concentrations of a-toxin (60 to 480 ng/ml) in the presence or absence of IVIG (0.1 to 2.5 mg/ml, as indicated). The dashed line indicates the mean PHA-induced proliferative response. B) Proliferative responses induced by bacterial supernatants from both proliferative and cytotoxic strains (dilutions 1:1000 and 1:50) in the presence or absence of different concentrations of IVIG (0.1 to 2.5 mg/ml, as indicated). 3H-thymidine uptake after 72 hours of culture is presented as mean counts per minute 6 SD. The figure shows one representative of two experiments using cells from different donors. doi:10 1371/journal pone 0106107 g004 distinct phenotypic response profiles elicited by clinical CA S. aureus isolates, which are determined by the combined action of a-toxin and superantigens, and the data shows a striking association to specific agr types, such as the agr II and III that demonstrated a superantigen-mediated proliferative profile where- as agr I and IV were cytotoxic. Whether these phenotypic response profiles reflect specific pathotypes that contribute to distinct disease manifestations can at present time only be speculated upon and can only be addressed by the use of large well-defined clinical cohorts. August 2014 \| Volume 9 \| Issue 8 \| e106107 References 24. Verkaik NJ, Dauwalder O, Antri K, Boubekri I, de Vogel CP, et al. (2010) Immunogenicity of toxins during Staphylococcus aureus infection. Clin Infect Dis 50: 61–68. 1. Vandenesch F, Naimi T, Enright MC, Lina G, Nimmo GR, et al. (2003) Community-acquired methicillin-resistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg Infect Dis 9: 978–984. 1. Vandenesch F, Naimi T, Enright MC, Lina G, Nimmo GR, et al. (2003) Community-acquired methicillin-resistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg Infect Dis 9: 978–984. 25. Brosnahan AJ, Mantz MJ, Squier CA, Peterson ML, Schlievert PM (2009) Cytolysins augment superantigen penetration of stratified mucosa. J Immunol 182: 2364–2373. 2. (1999) Four pediatric deaths from community-acquired methicillin-resistant Staphylococcus aureus - Minnesota and North Dakota, 1997–1999. Morb Mortal Wkly Rep 48: 707–710. 26. Shambat S, Nadig S, Prabhakara S, Bes M, Etienne J, et al. (2012) Clonal complexes and virulence factors of Staphylococcus aureus from several cities in India. BMC Microbiol 12: 64. 3. Miller LG, Perdreau-Remington F, Rieg G, Mehdi S, Perlroth J, et al. (2005) Necrotizing fasciitis caused by community-associated methicillin-resistant Staphylococcus aureus in Los Angeles. N Engl J Med 352: 1445–1453. 27. Arnaud M, Chastanet A, Debarbouille M (2004) New vector for efficient allelic replacement in naturally nontransformable, low-GC-content, gram-positive bacteria. Appl Environ Microbiol 70: 6887–6891. 4. Mediavilla JR, Chen L, Mathema B, Kreiswirth BN (2012) Global epidemiology of community-associated methicillin resistant Staphylococcus aureus (CA- MRSA). Curr Opin Microbiol 15: 588–595. 28. Boisset S, Geissmann T, Huntzinger E, Fechter P, Bendridi N, et al. (2007) Staphylococcus aureus RNAIII coordinately represses the synthesis of virulence factors and the transcription regulator Rot by an antisense mechanism. Genes Dev 21: 1353–1366. 5. Strandberg KL, Rotschafer JH, Vetter SM, Buonpane RA, Kranz DM, et al. (2010) Staphylococcal superantigens cause lethal pulmonary disease in rabbits. J Infect Dis 202: 1690–1697. 29. Hansenova Manaskova S, Bikker FJ, Veerman EC, van Belkum A, van Wamel WJ (2013) Rapid detection and semi-quantification of IgG-accessible Staphy- lococcus aureus surface-associated antigens using a multiplex competitive Luminex assay. J Immunol Methods. 6. Xu SX, McCormick JK (2012) Staphylococcal superantigens in colonization and disease. Front Cell Infect Microbiol 2: 52. 7. Spaulding AR, Salgado-Pabon W, Kohler PL, Horswill AR, Leung DY, et al. (2013) Staphylococcal and streptococcal superantigen exotoxins. Clin Microbiol Rev 26: 422–447. 30. Badiou C, Dumitrescu O, Croze M, Gillet Y, Dohin B, et al. References (2008) Panton- Valentine leukocidin is expressed at toxic levels in human skin abscesses. Clin Microbiol Infect 14: 1180–1183. 8. Bubeck Wardenburg J, Bae T, Otto M, Deleo FR, Schneewind O (2007) Poring over pores: alpha-hemolysin and Panton-Valentine leukocidin in Staphylococcus aureus pneumonia. Nat Med 13: 1405–1406. 31. Rasigade JP, Trouillet-Assant S, Ferry T, Diep BA, Sapin A, et al. (2013) PSMs of hypervirulent Staphylococcus aureus act as intracellular toxins that kill infected osteoblasts. PLoS One 8: e63176. 9. Gillet Y, Issartel B, Vanhems P, Fournet JC, Lina G, et al. (2002) Association between Staphylococcus aureus strains carrying gene for Panton-Valentine leukocidin and highly lethal necrotising pneumonia in young immunocompetent patients. Lancet 359: 753–759. 32. Spaan AN, Henry T, van Rooijen WJ, Perret M, Badiou C, et al. (2013) The staphylococcal toxin Panton-Valentine Leukocidin targets human C5a receptors. Cell Host Microbe 13: 584–594. p 10. Vandenesch F, Lina G, Henry T (2012) Staphylococcus aureus hemolysins, bi- component leukocidins, and cytolytic peptides: a redundant arsenal of membrane-damaging virulence factors? Front Cell Infect Microbiol 2: 12. 33. Darenberg J, Soderquist B, Normark BH, Norrby-Teglund A (2004) Differences in potency of intravenous polyspecific immunoglobulin G against streptococcal and staphylococcal superantigens: implications for therapy of toxic shock syndrome. Clin Infect Dis 38: 836–842. 11. Wang R, Braughton KR, Kretschmer D, Bach TH, Queck SY, et al. (2007) Identification of novel cytolytic peptides as key virulence determinants for community-associated MRSA. Nat Med 13: 1510–1514. 34. Farag N, Mahran L, Abou-Aisha K, El-Azizi M (2013) Assessment of the efficacy of polyclonal intravenous immunoglobulin G (IVIG) against the infectivity of clinical isolates of methicillin-resistant Staphylococcus aureus (MRSA) in vitro and in vivo. Eur J Clin Microbiol Infect Dis 32: 1149–1160. 12. Deghorain M, Van Melderen L (2012) The Staphylococci phages family: an overview. Viruses 4: 3316–3335. and in vivo. Eur J Clin Microbiol Infect Dis 32: 1149–1160. 13. Malachowa N, DeLeo FR (2010) Mobile genetic elements of Staphylococcus aureus. Cell Mol Life Sci 67: 3057–3071. 35. Nishi JI, Kanekura S, Takei S, Kitajima I, Nakajima T, et al. (1997) B cell epitope mapping of the bacterial superantigen staphylococcal enterotoxin B: the dominant epitope region recognized by intravenous IgG. J Immunol 158: 247– 254. 14. Foster TJ (2005) Immune evasion by staphylococci. Nat Rev Microbiol 3: 948– 958. 36. References Takei S, Arora YK, Walker SM (1993) Intravenous immunoglobulin contains specific antibodies inhibitory to activation of T cells by staphylococcal toxin superantigens. J Clin Invest 91: 602–607. 15. Ji G, Beavis R, Novick RP (1997) Bacterial interference caused by autoinducing peptide variants. Science 276: 2027–2030. 16. Lindsay JA, Holden MT (2004) Staphylococcus aureus: superbug, super genome? Trends Microbiol 12: 378–385. p g J 37. Gauduchon V, Cozon G, Vandenesch F, Genestier AL, Eyssade N, et al. (2004) Neutralization of Staphylococcus aureus Panton Valentine leukocidin by intravenous immunoglobulin in vitro. J Infect Dis 189: 346–353. 17. Novick RP (2003) Autoinduction and signal transduction in the regulation of staphylococcal virulence. Mol Microbiol 48: 1429–1449. 38. Alonzo F 3rd, Kozhaya L, Rawlings SA, Reyes-Robles T, DuMont AL, et al. (2013) CCR5 is a receptor for Staphylococcus aureus leukotoxin ED. Nature 493: 51–55. 18. Jarraud S, Lyon GJ, Figueiredo AM, Lina G, Vandenesch F, et al. (2000) Exfoliatin-producing strains define a fourth agr specificity group in Staphylo- coccus aureus. J Bacteriol 182: 6517–6522. 39. DuMont AL, Torres VJ (2014) Cell targeting by the Staphylococcus aureus pore-forming toxins: it’s not just about lipids. Trends Microbiol 22: 21–27. 19. Geisinger E, Chen J, Novick RP (2012) Allele-dependent differences in quorum- sensing dynamics result in variant expression of virulence genes in Staphylo- coccus aureus. J Bacteriol 194: 2854–2864. 40. Novick RP, Geisinger E (2008) Quorum sensing in staphylococci. Annu Rev Genet 42: 541–564. 20. Croze M, Dauwalder O, Dumitrescu O, Badiou C, Gillet Y, et al. (2009) Serum antibodies against Panton-Valentine leukocidin in a normal population and during Staphylococcus aureus infection. Clin Microbiol Infect 15: 144–148. 41. Jarraud S, Mougel C, Thioulouse J, Lina G, Meugnier H, et al. (2002) Relationships between Staphylococcus aureus genetic background, virulence factors, agr groups (alleles), and human disease. Infect Immun 70: 631–641. 21. Jacobsson G, Colque-Navarro P, Gustafsson E, Andersson R, Mollby R (2010) Antibody responses in patients with invasive Staphylococcus aureus infections. Eur J Clin Microbiol Infect Dis 29: 715–725. 42. Collery MM, Smyth DS, Twohig JM, Shore AC, Coleman DC, et al. (2008) Molecular typing of nasal carriage isolates of Staphylococcus aureus from an Irish university student population based on toxin gene PCR, agr locus types and multiple locus, variable number tandem repeat analysis. J Med Microbiol 57: 348–358. 22. Discussion Thus, our data support a beneficial role of IVIG that through its broad August 2014 \| Volume 9 \| Issue 8 \| e106107 PLOS ONE \| www.plosone.org 9 Staphylococcal Phenotypic Response Profiles Vaccines for providing the antibodies required for PVL ELISA and a-toxin ELISA, respectively. spectrum of antibodies can prevent toxin-mediated damage, including both excessive proliferation as well as cytolytic events. Discussion For instance, the course of bacter- emia, toxic shock syndrome and/or necrotizing infections of the skin and lung might be connected to a specific ratio of superantigenic activity versus cytotoxicity, which would determine the overall biological effect on patients’ immune cells. However, we fully appreciate the complexity of identification of pathotypes dictated by toxin-mediated cellular responses. Clinical manifesta- tions and outcome will depend on a variety of factors including distinct phenotypic response profiles elicited by clinical CA S. aureus isolates, which are determined by the combined action of a-toxin and superantigens, and the data shows a striking association to specific agr types, such as the agr II and III that demonstrated a superantigen-mediated proliferative profile where- as agr I and IV were cytotoxic. Whether these phenotypic response profiles reflect specific pathotypes that contribute to distinct disease manifestations can at present time only be speculated upon and can only be addressed by the use of large well-defined clinical cohorts. For instance, the course of bacter- emia, toxic shock syndrome and/or necrotizing infections of the skin and lung might be connected to a specific ratio of superantigenic activity versus cytotoxicity, which would determine the overall biological effect on patients’ immune cells. However, we fully appreciate the complexity of identification of pathotypes dictated by toxin-mediated cellular responses. Clinical manifesta- tions and outcome will depend on a variety of factors including among others varying host susceptibility at both the cellular and genetic level as well as site of infection. In addition, patients can present with both necrotizing infections and severe sepsis/septic shock, which raises the currently unexplored question of expres- sion levels and effects of superantigens and cytotoxins systemically and locally. Nevertheless, identification of specific pathotypes would be of great clinical value as it enables recognition of risk patients and allows for epidemiologic surveillance. Although clinical data are limited, IVIG has been proposed as adjunctive therapy in fulminant invasive S. aureus infections and the mechanistic action suggested to involve neutralizing antibodies against superantigens [33,35], PVL [37] and a-toxin [34]. Here we found that addition of IVIG to PBMC cultures efficiently inhibited both a-toxin-mediated cytotoxicity as well as superan- tigen-mediated proliferation at physiological concentrations. Acknowledgments Conceived and designed the experiments: SMS ANT MS. Performed the experiments: SMS AH FV WvW. Analyzed the data: SMS AH FV WvW GA MS ANT. Contributed reagents/materials/analysis tools: FV GL GA. Contributed to the writing of the manuscript: SMS FV GA MS ANT. Conceived and designed the experiments: SMS ANT MS. Performed the experiments: SMS AH FV WvW. Analyzed the data: SMS AH FV WvW GA MS ANT. Contributed reagents/materials/analysis tools: FV GL GA. Contributed to the writing of the manuscript: SMS FV GA MS ANT. The skillful technical assistance provided by Sushma Prabhakara, Yvonne Benito, Michele Bes, Florence Couzon and Cedric Badiou is gratefully acknowledged. We are grateful to bioMe´rieux and GlaxoSmithKline August 2014 \| Volume 9 \| Issue 8 \| e106107 References Kanclerski K, Soderquist B, Kjellgren M, Holmberg H, Mollby R (1996) Serum antibody response to Staphylococcus aureus enterotoxins and TSST-1 in patients with septicaemia. J Med Microbiol 44: 171–177. 43. Szkaradkiewicz A, Karpinski TM, Zeidler A, Szkaradkiewicz AK, Masiuk H, et al. (2012) Cytokine response in patients with chronic infections caused by Staphylococcus aureus strains and diversification of their Agr system classes. Eur J Clin Microbiol Infect Dis 31: 2809–2815. p p 23. Ide F, Matsubara T, Kaneko M, Ichiyama T, Mukouyama T, et al. (2004) Staphylococcal enterotoxin-specific IgE antibodies in atopic dermatitis. Pediat Int 46: 337–341. August 2014 \| Volume 9 \| Issue 8 \| e106107 PLOS ONE \| www.plosone.org 10
https://openalex.org/W4319166236	https://www.researchsquare.com/article/rs-1874540/latest.pdf	English	null	Primary and potential secondary risks of landslide outburst floods	Natural hazards	2,023	cc-by	13,772	Primary and potential secondary risks of landslide outburst §oods Yunjian Gao (  gaoyunjiansu@163.com ) Sichuan University Chack Fan Lee Zhejiang University Jianhui Deng Sichuan University Siyuan Zhao Sichuan University Wen Dai Nanjing University of Information Science and Technology Jun Zhu Institute of Mountain Hazards and Environment Chinese Academy of Sciences Zhiliang He Sichuan University Zhongkang Yang Sichuan University Stuart N. Lane University of Lausanne Faculty of Geosciences and Environment: Universite de Lausanne Faculte des Geosciences et de l'environnement Research Article Keywords: Outburst §ood, Baige landslide, BASEMENT model, remote sensing interpretation, §ood disasters assessment Posted Date: July 29th, 2022 DOI: https://doi.org/10.21203/rs.3.rs-1874540/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Primary and potential secondary risks of landslide outburst §oods Yunjian Gao (  gaoyunjiansu@163.com ) Sichuan University Chack Fan Lee Zhejiang University Jianhui Deng Sichuan University Siyuan Zhao Sichuan University Wen Dai Nanjing University of Information Science and Technology Jun Zhu Institute of Mountain Hazards and Environment Chinese Academy of Sciences Zhiliang He Sichuan University Zhongkang Yang Sichuan University Stuart N. Lane University of Lausanne Faculty of Geosciences and Environment: Universite de Lausanne Faculte des Geosciences et de l'environnement Research Article Keywords: Outburst §ood, Baige landslide, BASEMENT model, remote sensing interpretation, §ood disasters assessment Posted Date: July 29th, 2022 DOI: https://doi.org/10.21203/rs.3.rs-1874540/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Primary and potential secondary risks of landslide outburst §oods Yunjian Gao (  gaoyunjiansu@163.com ) Sichuan University Chack Fan Lee Zhejiang University Jianhui Deng Sichuan University Siyuan Zhao Sichuan University Wen Dai Nanjing University of Information Science and Technology Jun Zhu Institute of Mountain Hazards and Environment Chinese Academy of Sciences Zhiliang He Sichuan University Zhongkang Yang Sichuan University Stuart N. Lane University of Lausanne Faculty of Geosciences and Environment: Universite de Lausanne Faculte des Geosciences et de l'environnement Primary and potential secondary risks of landslide outburst floods 1 5. Key laboratory of mountain hazards and earth surface process, Institute of Mountain Hazards and Environment, 10 Chinese Academy of Sciences (CAS), Chengdu 610041, China. 11 Abstract: Outburst floods triggered by breaching of landslide dams may cause severe loss of life 14 and property downstream. Accurate identification and assessment of such floods, especially 15 when they leading to secondary impacts, is critical. In 2018, the Baige landslide in the Tibetan 16 Plateau twice blocked the Jinsha River, eventually resulting in a severe outburst flood. The Baige 17 landslide remains active and it is possible that a breach happens again. Based on numerical 18 simulation using a hydrodynamic model, remote sensing, and field investigation, we reproduce 19 the outburst flood process and assess the hazard associated with future floods. The results show 20 that the hydrodynamic model could accurately simulate the outburst flood process, with overall 21 accuracy and kappa accuracy for the flood extent of 0.956 and 0.911. Three future dam break 22 scenarios were considered with landslide dams of heights 30m, 35m, and 51m. The potential 23 storage capacity / back-flow reaches 142x106m3/32km, 182x106m3/40km, and 331x106m3/50km 24 in the valley upstream. Failure of these three dams leads to maximum inundation in the 25 downstream of study reach of 1.674 km2, 1.833 km2, and 1.927 km2. These results demonstrate 26 the serious secondary hazard associated with this region. 27 Keyword: Outburst flood, Baige landslide, BASEMENT model, remote sensing interpretation, 28 flood disasters assessment 29 Research Article Posted Date: July 29th, 2022 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Primary and potential secondary risks of landslide outburst floods 1 Yunjian Gao1,2,, Lee Chack Fan3, Jianhui Deng1, Siyuan Zhao1, Wen Dai4, Jun Zhu5, Zhiliang He1, Zhongkang Yang1, 2 Stuart N. Lane2 3 1. State Key Laboratory of Hydraulics and Mountain River Engineering, College of Water Resource & Hydropower, 4 Sichuan University, Chengdu 610065, China, 5 2. Institute of Earth Surface Dynamics (IDYST), University of Lausanne, Lausanne 1015, Switzerland, 6 3. Ningbo Institute of Technology, Zhejiang University, Ningbo 315100, China, 7 4. School of Geographical Sciences, Nanjing University of Information Science & Technology, Nanjing 210044, 8 China, 9 5. Key laboratory of mountain hazards and earth surface process, Institute of Mountain Hazards and Environment, 10 Chinese Academy of Sciences (CAS), Chengdu 610041, China. 11 Corresponding author: gaoyunjiansu@163.com; Tel.: +8618328654971 12 13 Abstract: Outburst floods triggered by breaching of landslide dams may cause severe loss of life 14 and property downstream. Accurate identification and assessment of such floods, especially 15 when they leading to secondary impacts, is critical. In 2018, the Baige landslide in the Tibetan 16 Plateau twice blocked the Jinsha River, eventually resulting in a severe outburst flood. The Baige 17 landslide remains active and it is possible that a breach happens again. Based on numerical 18 simulation using a hydrodynamic model, remote sensing, and field investigation, we reproduce 19 the outburst flood process and assess the hazard associated with future floods. The results show 20 that the hydrodynamic model could accurately simulate the outburst flood process, with overall 21 accuracy and kappa accuracy for the flood extent of 0.956 and 0.911. Three future dam break 22 scenarios were considered with landslide dams of heights 30m, 35m, and 51m. The potential 23 storage capacity / back-flow reaches 142x106m3/32km, 182x106m3/40km, and 331x106m3/50km 24 in the valley upstream. Failure of these three dams leads to maximum inundation in the 25 downstream of study reach of 1.674 km2, 1.833 km2, and 1.927 km2. These results demonstrate 26 the serious secondary hazard associated with this region. Research Article 27 Keyword: Outburst flood, Baige landslide, BASEMENT model, remote sensing interpretation, 28 flood disasters assessment 29 0 Introduction 30 Landslide damming of laterally-confined rivers is reported with increasing frequency (Knapp et al. 31 2018; Fan et al. 2019; Ermini et al. 2003; Chen et al. 2013; Guo et al. 2021). If these dams breach, 32 the subsequently outburst floods may induce catastrophic casualties and major damage to 33 d (F l 2020 B d 2006 W l 2021) Th Tib Pl 34 0 Introduction 30 Landslide damming of laterally-confined rivers is reported with increasing frequency (Knapp et al. 31 2018; Fan et al. 2019; Ermini et al. 2003; Chen et al. 2013; Guo et al. 2021). If these dams breach, 32 the subsequently outburst floods may induce catastrophic casualties and major damage to 33 property downstream (Fan et al. 2020; Bonnard 2006; Wang et al. 2021). The Tibetan Plateau, 34 European Alps, and Western North American mountainous regions are those where landslide 35 dam breaches are most commonly reported (Knapp et al. 2018; Fan et al. 2019). Examination of 36 previous dam failures shows that globally about 80 to 90% fail within the first year of formation 37 (Costa and Schuster 1988; Fan et al. 2020). Fan et al. (2020) record more than 410 landslide dams 38 around the world, of which about 80% failed in less than one year. Ermini and Casagli (2003) 39 recorded more than 350 landslides dams worldwide, either rainfall or earthquake triggered with 40 80% and 57% of them breaching and triggering a flood disaster respectively. In China, giant 41 landslide dammed events are focused in mountainous areas. The Yigong landslide is a good 42 example, feeding sediment into the Zhamunong tributary and blocking the Yigong River with a 43 barrier of about 50m to 80m height in 1901 and 2000 (Zhou et al. 2016; Shang et al. 2003; Xu et 44 al. 2012). Subsequently, the outburst flood traveled over 500 km south into India (Delaney and 45 Evans 2015), it threatened about 4000 people downstream, and the economic loss from the 46 flooding was about $169’000 US dollars (Shang et al. 2005; Evans and Delaney 2010). 47 On October 10th and November 3rd 2018 two successive landslides occurred at Baige Village, 48 Jiangda County, Tibetan Autonomous Region, P.R. China (Wu et al. 2019; Yang et al. 2019; Wang 49 et al. 2020; Li et al. 2020; Ding et al. 2021; An et al. 2021; Zhang et al. 2020). They delivered 50 about 18.7x106m3 and 6.3x106m3 of sediment, blocking the Jinsha River and forming landslide- 51 dammed lakes (Chen et al. 2021; Ouyang et al. 2019). After each dam formed, the barrier lake 52 breached and the outburst floods caused flood disasters downstream. The landslide dams 53 blocked the Jinsha River for 44 hours and 13 days respectively. 0 Introduction 30 The dam heights were 47m and 54 72m, and the subsequent lake that formed extended 45km and 70km and the peak storage 55 capacity reached 0.29x109m3 and 0.58x109m3. There was significant infrastructure damage and 56 the whole of Boluo Town was inundated (Chen et al. 2021; Gao et al. 2021; Xiong et al. 2020). 57 With the dam break, the river discharge increased to 10000 m3s-1 and 31000 m3s-1 respectively. 58 The outburst flood reached 670 km downstream, with roads, houses, farms, and bridges 59 damaged; for example 18 bridges were affected and collapsed (Zhang et al. 2021; Liu et al. 2020). 60 Fortunately, the two floods events did not induce casualties because of sufficient time between 61 breaches and inundation for warnings to be issued. 62 The rear edge of the Baige landslide has continued to deform since the first two landslide events 63 and there is a high probability of further secondary hazards involving dam formation and 64 breaching (Chen et al. 2021; Fan et al. 2020). Although future outburst flood hazard assessment 65 for the Baige landslide is essential, most of the attention has been on post-outburst floods. This 66 work has struggled to quantify post-flood impacts using remote sensing (Liu et al. 2020; Yang et 67 al. 2021), notably because of poor image quality. The one-dimensional HEC-RAS hydraulic model 68 has been used to simulate the flood (Fan et al. 2020; Gao et al. 2021), but its precision is 69 insufficient because it relies on cross-section density and struggles to reproduce spatial patterns 70 of inundation when there is substantial flux of water on the floodplain rather than in the channel. 71 Improvements in hydraulic modelling are needed, not least to allow better assessment of 72 inundation patterns in two dimensions. The aim of this paper is to develop a two-dimensional 73 hydraulic model for inundation downstream of the Baige landslide and to use this to expel future 74 flood hazard. 75 1 Overview of the landslide and the study area 76 The Baige landslide is located at Boluo Town, Jiangda County, in the Tibet Autonomous Region, 77 China (31.082336°N, 98.704722°E). It is periodically dams the Jinsha River of the Tibetan Plateau. 78 It is in the Jinsha River suture zone and is associated with strong tectonic activity; the tectonic 79 setting of the area is quite complex, faults and folds are well developed. Proterozoic, 80 Carboniferous, and Triassic strata are the major rocks in the study area. The major faults in the 81 area generally strike NW-SE, and the Boluo-Muxie Fault is the nearest one to the slides which 82 strike N30° W and dip 50 to 70° to SW. The fault is 146 km long, and its fault belt is 100 to 300 m 83 wide. The Jinsha River undercuts the plateau forming a V-shaped valley which allows for the 84 Baige landslide to readily block the river. 85 According to field investigations and remote sensing, a total of 41 towns were affected by two 86 Baige outburst floods, with a maximum inundated area of about 101.75 km2 in the second flood. 87 This paper focuses on a 9 km reach at Zhubalong Town in the Jinsha River in Batang County 88 (Figure 1). Zhubalong Town was selected for the focus as is a city of considerable importance in 89 terms of infrastructure. It was 180km downstream of the dam but still experienced serious flood 90 losses. The study reach is representative of other reaches impacted by the floods in that it 91 comprises a wide floodplain. 92 93 Fig. 1 The location of the Baige landslide and the study area. (a)/(c) the Baige landslide flood impact extent. (b) 94 the study area satellite image. (d) Baige landslide aerial image 95 93 Fig. 1 The location of the Baige landslide and the study area. (a)/(c) the Baige landslide flood impact extent. (b) 94 the study area satellite image. (d) Baige landslide aerial image 95 96 Fig. 2 The flow chart of the outburst flood analysis 97 Fig. 2 The flow chart of the outburst flood analysis 97 2 Overall methodology and data sources 98 Figure 2 provides an overview of the outburst flood analysis presented in the paper. It is based 99 upon applying a two-dimensional hydrodynamic model to the Zhubalong Town reach for the 100 larger of the 2018 events, which provides for validation of the analysis. Then, possible future dam 101 breach scenarios are applied to the hydrodynamic model to determine the inundation that would 102 result. 103 2.1 Data sources 104 The flood hazard assessment in this work was based upon numerical simulation, satellite imagery, 105 digital elevation model (DEM) data and hydrological information. First, the PlanetScope satellite 106 imagery data were obtained from the “Planet Developer Resource Center (Planet Explorer)”. 107 PlanetScope data are based upon approximately 130 satellites, and can image the entire land 108 surface of the Earth daily. PlanetScope images are approximately 3 m or 3.7 m per pixel 109 resolution and are generally divided into 4 bands (B, G, R, and NIR) (Table 1). These data were 110 used to identify maximum flood extent during the event (see below). The flood has a clear 111 signature in terms of maximum extent due to extensive fine sediment deposition. Second, the 112 ALOS (Advanced Land Observing Satellite) DEM data, produced using ALOS panchromatic 113 three-line images, were used for the hydrodynamic model. These data have a spatial resolution 114 The flood hazard assessment in this work was based upon numerical simulation, satellite imagery, 105 digital elevation model (DEM) data and hydrological information. First, the PlanetScope satellite 106 imagery data were obtained from the “Planet Developer Resource Center (Planet Explorer)”. 107 PlanetScope data are based upon approximately 130 satellites, and can image the entire land 108 surface of the Earth daily. PlanetScope images are approximately 3 m or 3.7 m per pixel 109 resolution and are generally divided into 4 bands (B, G, R, and NIR) (Table 1). These data were 110 used to identify maximum flood extent during the event (see below). The flood has a clear 111 signature in terms of maximum extent due to extensive fine sediment deposition. Second, the 112 ALOS (Advanced Land Observing Satellite) DEM data, produced using ALOS panchromatic 113 three-line images, were used for the hydrodynamic model. These data have a spatial resolution 114 of 5 m and were obtained from the Geographical Survey Institute of Japan. Finally, discharge data 115 were provided from the Batang gauging station in the Jinsha River (Figure 3), providing reliable 116 real-time monitoring data (Gao et al. 2021; Chen et al. 2021). A peak discharge of 20900 m3s-1 117 was recorded during the second flood (Figure 3). 118 of 5 m and were obtained from the Geographical Survey Institute of Japan. 2.1 Data sources 104 Finally, discharge data 115 were provided from the Batang gauging station in the Jinsha River (Figure 3), providing reliable 116 real-time monitoring data (Gao et al. 2021; Chen et al. 2021). A peak discharge of 20900 m3s-1 117 was recorded during the second flood (Figure 3). 118 of 5 m and were obtained from the Geographical Survey Institute of Japan. Finally, discharge data 115 were provided from the Batang gauging station in the Jinsha River (Figure 3), providing reliable 116 real-time monitoring data (Gao et al. 2021; Chen et al. 2021). A peak discharge of 20900 m3s-1 117 was recorded during the second flood (Figure 3). 118 Table 1 The satellite image and DEM data source 119 Data Acquisition time Spatial resolution Data source PlanetScope (Satellite) 2018.10.07 3 m Planet Developer Resource Center (https://account.planet.com/) 2018.10.25 3 m 2018.11.15 3 m DEM --- 5 m ALOS 120 Fig. 3 The discharge hydrograph measured for the Batang County reach 121 Table 1 The satellite image and DEM data source 119 Data Acquisition time Spatial resolution Data source PlanetScope (Satellite) 2018.10.07 3 m Planet Developer Resource Center (https://account.planet.com/) 2018.10.25 3 m 2018.11.15 3 m DEM --- 5 m ALOS 120 Fig. 3 The discharge hydrograph measured for the Batang County reach 121 Table 1 The satellite image and DEM data source Fig. 3 The discharge hydrograph measured for the Batang County reach 2.2.1 Remote sensing interpretation 123 The outburst floods occurred on 12 October and 13 November 2018. The 7 October 2018 and 15 124 November 2018 were selected dates with cloud-free imagery from before and after the two 125 events; as the second flood was larger, the 15 November data is likely representative of this event. 126 It is recorded in the image in terms of sediment deposition and vegetation removal and so the 127 image can be used to distinguish between inundated and non-inundated zones (vegetation 128 covered zone). The maximum inundated extent was mapped by comparing imagery from before 129 and after the flood. The eCogition Developer version 9.0 software was used to conduct the 130 remote sensing interpretation. First, the RGB bands were displayed (Figure 4a). Second, a 131 multi-resolution segmentation algorithm (MESA; Benz et al., 2004) was applied at the pixel scale. 132 This algorithm starts with individual pixels and merges them with adjacent pixels according to 133 their homogeneity. The MRSA was applied with a scale parameter of 25m (Figure 4b). Third, the 134 Normalized Difference Vegetation Index (NDVI) was determined (Figure 4c): 135 This algorithm starts with individual pixels and merges them with adjacent pixels according to 133 their homogeneity. The MRSA was applied with a scale parameter of 25m (Figure 4b). Third, the 134 Normalized Difference Vegetation Index (NDVI) was determined (Figure 4c): 135 (1) 136 R NIR R NIR mean mean mean mean NDVI    (1) The NDVI is used to detect vegetation growth state, vegetation coverage and eliminate partial 137 radiation error, which can well reflect the characteristics of plant distribution and is easy to 138 distinguish from inundated zones. Fourth, the imagery was classified into a vegetation zone 139 (non-inundated zone) and an inundated zone based upon a threshold value of the NDVI. In this 140 study, the NDVI of the vegetation zone was 0.16 ~ 1.0 (Figure 4d). Fifth, a manual classification 141 was undertaken to provide validation data compared with raw imagery, error classification was 142 reclassified by manual operation (Figure 4e). Sixth, the classes were merged to produce 143 non-inundated zones (vegetation covered zone) and inundated zones (Figure 4f). 144 145 Fig. 4 Method to delineate flood inundation extent using remote sensing 146 Fig. 4 Method to delineate flood inundation extent using remote sensing 2.2.2 Determination of potential landslide dam height 147 2.2.2 Determination of potential landslide dam height 147 2.2.2 Determination of potential landslide dam height 147 Secondary failure and landslide dammed lake risk of the Baige landslide is high with the 148 continuous deformation of the landslide edge and rear (Chen et al. 2021; Zhang et al. 2021; 149 Zhou et al. 2022). Crack zones have been identified in several detailed field investigations, which 150 implies that the area of landslide potential high-risk zones can be determined (Figure 5 (a)-(e)). A 151 deep displacement monitoring instrument (SINCO) was used to detect slope deformation and to 152 identify the landslide sliding surface with manual measurement, once per month on average. The 153 total monitoring time was from June 2019 to December 2021. The potential landslide volume 154 was calculated based upon surface area and the estimated depth to the shear plane. 155 156 Fig. 5 Deformation characteristics of the Baige landslide. (a) the deformation area and crack distribution of the 157 Baige landslide edge and rear. (b) the boundary crack of the the Baige landslide rear. (c) historical deformation 158 scarp. (d) crack propagation in engineering cutting area. (e) severe deformation zone 159 Fig. 5 Deformation characteristics of the Baige landslide. (a) the deformation area and crack distribution of the 157 Baige landslide edge and rear. (b) the boundary crack of the the Baige landslide rear. (c) historical deformation 158 scarp. (d) crack propagation in engineering cutting area. (e) severe deformation zone 159 The potential landslide dam height was determined based upon the landslide volume, dam shape, 160 and the shape of the deposit (dam breach deposit) that formed downstream from the breach. 161 For the two 2018 breaches, distinctive dam breach deposit shapes were identified (Figure 6a). 162 These deposit then determine the required landslide mass needed to create a landslide dam 163 during future landslide failures (Figure 6a). The landslide dam shape, especially the back and 164 front slope, could be extracted to restrict the crest of the dam assuming that the landslide 165 accumulation process operates in a similar way (Figure 6b-c). Using this approach it was possible 166 to identify to possible landslide dam heights for future breaching. 167 168 Fig. 6 (a) Landslide dam cross sections and dam breach deltas for the two flood events. (b) longitudinal profile 169 for the first 2018 event. (c) longitudinal profile for the second 2018 event 170 Fig. 6 (a) Landslide dam cross sections and dam breach deltas for the two flood events. 2.2.2 Determination of potential landslide dam height 147 There are four aspects that must be considered: (1) 174 the form of the terms use to describe conservation of mass and momentum; (2) numerical 175 solution of the associated partial differential equations; (3) parameterization, notably of flow 176 resistance which is generally the parameter that has most impact on flood inundation prediction; 177 and (4) the description of channel and floodplain geometry (Lane, 1998; Worni et al. 2013). 178 The two-dimensional numerical model BASEMENT (Zischg et al. 2018), was used to investigate 179 and simulate flood inundation in the study reach. BASEMENT has been designed as a tool for 180 application to the study of breaching processes for non-cohesive earthen dam structures and 181 water-sediment processes (Lala et al. 2018). The 2D shallow water Equations (2) through (4), are 182 solved with an explicit finite-volume method on an unstructured mesh (Worni et al. 2013). The 183 water depth h and specific discharges (q = uh, r = vh) are the primary variables in the coordinate 184 directions. 185 (2) 186 (3) 187 (4) 188 0          y h) v ( x h) u ( t h  Bx B x Z gh ) huv ( y ) gh hu ( x ) hu ( t               2 2 2 1  By B y Z gh ) gh hv ( y ) huv ( x ) hv ( t               2 2 1 2 186 (2) (3) (4) 188 Where u is the velocity in x direction, v is velocity in y direction, g is the gravitational acceleration, 189 ρ is the fluid density, ZB is the bottom elevation. The bed shear stresses (τBx, τBy) act in the 190 direction of depth-averaged velocities and are determined using the quadratic resistance law 191 with cf being the dimensionless friction factor as (5). Flow resistance is generally described by the 192 empirical Strickler (Range: 7~40 m1/3s-1) or Manning coefficients (Range: 0.025~0.143 m-1/3s) 193 (Vanzo et al. 2021 ). 194 Where u is the velocity in x direction, v is velocity in y direction, g is the gravitational acceleration, 189 ρ is the fluid density, ZB is the bottom elevation. 2.2.2 Determination of potential landslide dam height 147 (b) longitudinal profile 169 for the first 2018 event. (c) longitudinal profile for the second 2018 event 170 2.2.3 Hydrodynamic model 171 2.2.3 Hydrodynamic model 171 The simulation of inundation was based upon the shallow water equations which solve for 172 continuity (conservation of mass or volume) and momentum to calculate the propagation of 173 shallow water flows over natural topography. There are four aspects that must be considered: (1) 174 the form of the terms use to describe conservation of mass and momentum; (2) numerical 175 solution of the associated partial differential equations; (3) parameterization, notably of flow 176 resistance which is generally the parameter that has most impact on flood inundation prediction; 177 and (4) the description of channel and floodplain geometry (Lane, 1998; Worni et al. 2013). 178 The two-dimensional numerical model BASEMENT (Zischg et al. 2018), was used to investigate 179 and simulate flood inundation in the study reach. BASEMENT has been designed as a tool for 180 application to the study of breaching processes for non-cohesive earthen dam structures and 181 water-sediment processes (Lala et al. 2018). The 2D shallow water Equations (2) through (4), are 182 solved with an explicit finite-volume method on an unstructured mesh (Worni et al. 2013). The 183 water depth h and specific discharges (q = uh, r = vh) are the primary variables in the coordinate 184 directions. 185 The simulation of inundation was based upon the shallow water equations which solve for 172 continuity (conservation of mass or volume) and momentum to calculate the propagation of 173 shallow water flows over natural topography. There are four aspects that must be considered: (1) 174 the form of the terms use to describe conservation of mass and momentum; (2) numerical 175 solution of the associated partial differential equations; (3) parameterization, notably of flow 176 resistance which is generally the parameter that has most impact on flood inundation prediction; 177 and (4) the description of channel and floodplain geometry (Lane, 1998; Worni et al. 2013). 178 The simulation of inundation was based upon the shallow water equations which solve for 172 continuity (conservation of mass or volume) and momentum to calculate the propagation of 173 shallow water flows over natural topography. 2.2.2 Determination of potential landslide dam height 147 The bed shear stresses (τBx, τBy) act in the 190 direction of depth-averaged velocities and are determined using the quadratic resistance law 191 with cf being the dimensionless friction factor as (5). Flow resistance is generally described by the 192 empirical Strickler (Range: 7~40 m1/3s-1) or Manning coefficients (Range: 0.025~0.143 m-1/3s) 193 (Vanzo et al. 2021 ). 194 , (5) 2 2 2 f Bx c / u v u    2 2 2 f By c / v v u    195 (5) (5) Generating an irregular triangle mesh is a fundamental step in the process. DEM data were used 196 to generate the nodes, each with an elevation value. Triangular meshes were generated by using 197 the pre-processing plug-in in QGIS software as shown in Figure 7 Finally execution of the 198 Generating an irregular triangle mesh is a fundamental step in the process. DEM data we 196 Generating an irregular triangle mesh is a fundamental step in the process. DEM data were used 196 to generate the nodes, each with an elevation value. Triangular meshes were generated by using 197 the pre-processing plug-in in QGIS software, as shown in Figure 7. Finally, execution of the 198 BASEMENT model, the flood result such as the water depth, velocity, and water surface elevation 199 (WSE) were obtained. 200 to generate the nodes, each with an elevation value. Triangular meshes were generated by using 197 the pre-processing plug-in in QGIS software, as shown in Figure 7. Finally, execution of the 198 BASEMENT model, the flood result such as the water depth, velocity, and water surface elevation 199 (WSE) were obtained. 200 to generate the nodes, each with an elevation value. Triangular meshes were generated by using 197 the pre-processing plug-in in QGIS software, as shown in Figure 7. Finally, execution of the 198 BASEMENT model, the flood result such as the water depth, velocity, and water surface elevation 199 (WSE) were obtained. 200 201 Fig. 7 The irregular triangulation mesh for the BASEMENT flood simulation of the study reach 202 Fig. 7 The irregular triangulation mesh for the BASEMENT flood simulation of the study reach The study area has low forest coverage and housing density, and so the Manning’s n friction was 203 set as 0.03 (Chow et al. 1988). 2.2.4 Accuracy analysis of model predictions 211 Kappa-based contingency analysis (Cohen 1960) was used to validate the accuracy of the 212 BASEMENT simulation flood result using predicted and observed inundation patterns. The 213 foundation of all accuracy assessments is an error matrix (Figure 8). With two classes, inundated 214 or not, the error matrix is a 2 × 2 space (Figure 8). Here, n11 and n22 are represented the numbers 215 of cells of consistent wet and consistent dry classes in both the simulation and the actual results. 216 n12 and n21 are represented the number of cells of the ‘wet versus dry’ and the ‘dry versus wet’ 217 between the simulation and the actual results. 218 Error matrix Actual result (Interpretation + field investigation) Wet dry ... Class Simulation result wet n11 n12 n1j 1 dry n21 n22 n2j 2 ... ni1 ni2 nij k Class 1 2 k ... Kappa value degrees: slight (0.0 - 0.2), fair (0.21 - 0.4), moderate (0.41 - 0.6), substantial (0.61 - 0.8), almost perfect (0.81 - 1.0) (Cohen 1960; Landis and Koch 1977) Fig. 8 The error matrix for the Kappa accuracy analysis 219 Kappa value degrees: slight (0.0 - 0.2), fair (0.21 - 0.4), moderate (0.41 - 0.6), substantial (0.61 - 0.8), almost perfect (0.81 - 1.0) (Cohen 1960; Landis and Koch 1977) 219 Fig. 8 The error matrix for the Kappa accuracy analysis 2.2.2 Determination of potential landslide dam height 147 The run time of the flood simulation was set as 46,800 s, noting 204 the peak discharge occurred at 10,800 s, and that the maximum inundation extent was reached 205 well before the end of this time for the second flood. The upstream boundary condition used the 206 stage-discharge curve for Batang County and the outlet used a normal depth condition, being set 207 downstream from the zone of interest to avoid any influence on the region of interest. The 208 detailed parameters as shown in Table 2. 209 Table 2 The parameters of the BASEMENT model simulation outburst flood 210 Gravity (ms-2) Viscosity (m-2s) Reynolds number(kg/m3) Boundary condition Friction (sm-1/3) Total run time (s) 9.8 0.000001 5000 Inlet Outlet 0.03 46800 (actual process) Data: discharge Type: hydrograph Data: the normal depth Type: hqrelation The study area has low forest coverage and housing density, and so the Manning’s n friction was 203 set as 0.03 (Chow et al. 1988). The run time of the flood simulation was set as 46,800 s, noting 204 the peak discharge occurred at 10,800 s, and that the maximum inundation extent was reached 205 well before the end of this time for the second flood. The upstream boundary condition used the 206 stage-discharge curve for Batang County and the outlet used a normal depth condition, being set 207 downstream from the zone of interest to avoid any influence on the region of interest. The 208 detailed parameters as shown in Table 2. 209 Table 2 The parameters of the BASEMENT model simulation outburst flood 210 Gravity (ms-2) Viscosity (m-2s) Reynolds number(kg/m3) Boundary condition Friction (sm-1/3) Total run time (s) 9.8 0.000001 5000 Inlet Outlet 0.03 46800 (actual process) Data: discharge Type: hydrograph Data: the normal depth Type: hqrelation Table 2 The parameters of the BASEMENT model simulation outburst flood Slope:3 (per mill) Slope:3 (per mill) Fig. 8 The error matrix for the Kappa accuracy analysis The overall accuracy is identified from Equation (6) as the sum of the correctly predicted cells 220 divided by the total number of cells (Yu and Lane 2006). However, this statistic take no account of 221 random level of agreement in the contingency table. 222 The overall accuracy is identified from Equation (6) as the sum of the correctly predicted cells 220 divided by the total number of cells (Yu and Lane 2006). However, this statistic take no account of 221 random level of agreement in the contingency table. 222 (6) 223 n n accuracy Overall k i ii    1 (6) Following Cohen (1960), we therefore used Kappa analysis, a discrete multivariate technique that 224 is used to allow statistical analysis of model performance, and to test if one error matrix is 225 different from the actual result (Figure 6) (Bishop et al. 1975). 226 (7) 227 ) n n ( ) n n ( n ) n n ( ) n n ( n n K k j j k i i k j j k i i k j j k i i k j j k i i k i ii                            1 2 1 2 1 1 1 1 2 1 2 1 2 1 1 1 1 1 227 (7) Equation (7) is essentially expressing the ratio of the observed excess over chance agreement to 228 the maximum possible excess over chance, with K=1.0 at perfect agreement and K=0.0 when 229 observed agreement equals chance agreement (Everitt 1998). Data required for this assessment 230 were based upon the spatial extent of the modelled domain. 231 Equation (7) is essentially expressing the ratio of the observed excess over chance agreement to 228 the maximum possible excess over chance, with K=1.0 at perfect agreement and K=0.0 when 229 observed agreement equals chance agreement (Everitt 1998). Data required for this assessment 230 were based upon the spatial extent of the modelled domain. 231 2.2.5 Calculation of the potential outburst flood peak discharge 232 In the study, we aimed to quantify how inundation changes as a function of different magnitudes 233 of river valley blockage by the landslides. This needed simulations of outburst flood magnitudes 234 for different dam heights and their attenuation with distance downstream. The peak discharge is 235 the key parameter that determines the submergence range, scour intensity, and arrival time of 236 flood downstream. Empirical models and hydrological experiments are the main two ways to 237 determine the peak discharge for potential dam breaching (Costa 1985; Costa and Schuster 1988; 238 Costa and Schuster 1991; Walder and O’Connor 1997; Fread 1988; Peng and Zhang 2012; Habib 239 et al. 2014; Li 2020). Empirical models vary in their complexity. For example, Costa’s (1985) 240 empirical model only considered dam height (Hd) for calculating peak flow (Qp), and Costa and 241 Schuster (1988, 1991) only considered dam height (Hd) and barrier lake storage capacity (Vl). 242 In the study, we aimed to quantify how inundation changes as a function of different magnitudes 233 of river valley blockage by the landslides. This needed simulations of outburst flood magnitudes 234 for different dam heights and their attenuation with distance downstream. The peak discharge is 235 the key parameter that determines the submergence range, scour intensity, and arrival time of 236 flood downstream. Empirical models and hydrological experiments are the main two ways to 237 determine the peak discharge for potential dam breaching (Costa 1985; Costa and Schuster 1988; 238 Costa and Schuster 1991; Walder and O’Connor 1997; Fread 1988; Peng and Zhang 2012; Habib 239 et al. 2014; Li 2020). Empirical models vary in their complexity. For example, Costa’s (1985) 240 empirical model only considered dam height (Hd) for calculating peak flow (Qp), and Costa and 241 Schuster (1988, 1991) only considered dam height (Hd) and barrier lake storage capacity (Vl). 242 Peng and Zhang (2012) proposed an empirical model to calculate peak discharge. The method 243 was based upon a database of 1239 landslide dams, including 257 dams formed during the 244 Wenchuan earthquake on May 12, 2008. The model considered dam height, dam width, dam 245 volume, reservoir storage capacity of barrier lake, and dam erodibility. The calculation formula is 246 shown in (7): 247 Peng and Zhang (2012) proposed an empirical model to calculate peak discharge. 2.2.5 Calculation of the potential outburst flood peak discharge 232 The method 243 was based upon a database of 1239 landslide dams, including 257 dams formed during the 244 Wenchuan earthquake on May 12, 2008. The model considered dam height, dam width, dam 245 volume, reservoir storage capacity of barrier lake, and dam erodibility. The calculation formula is 246 shown in (7): 247 (7) 248 a . d / l . r d / d / p e H V h H H g Q 536 1 3 1 371 1 2 5 2 1               (7) Where, Qp is the peak discharge at dam burst (m3s-1), Hd is dam height (m), Hr = 1 m (Peng and 249 Zhang 2012), Vl is storage capacity of barrier lake (m3), g = 9.8 ms-1, e = 2.718, ɑ is the dam 250 erodibility parameter which can be divided into three grades: high (1.236), medium (-0.380) and 251 low (-1.615) (Peng and Zhang 2012; Briaud 2008; Hanson and Simon 2001). 252 Table 3 Peng and Zhang (2012) model validation for potential peak discharge of the dam breaching 253 Dam height (Hd, m) g Storage capacity (Vd, m3) e a Peak discharge (m3/s-1) Measured outburst flood peak discharge for the first and second 2018 events 47 9.8 290000000 2.718 -1.615 2794 47 9.8 290000000 2.718 -0.38 9590 47 9.8 290000000 2.718 1.236 48331 72 9.8 580000000 2.718 -1.615 3300 72 9.8 580000000 2.718 -0.38 11347 72 9.8 580000000 2.718 1.236 57088 Table 3 Peng and Zhang (2012) model validation for potential peak discharge of the dam breaching Actual peak discharge: 10000m3/s-1(first event) and 31000m3/s-1(second event) Potential outburst flood peak discharge 30 9.8 142000000 2.718 -1.615 2326 35 9.8 182000000 2.718 -1.615 2467 51 9.8 331000000 2.718 -1.615 2918 30 9.8 142000000 2.718 -0.38 7928 35 9.8 182000000 2.718 -0.38 8440 51 9.8 331000000 2.718 -0.38 9965 30 9.8 142000000 2.718 1.236 40262 35 9.8 182000000 2.718 1.236 42668 51 9.8 331000000 2.718 1.236 50468 The spatial analysis tool in Arcgis 10.2 was used to calculate the storage capacity (Vl) and as a 254 by-product, the inundation area upstream dams of different height. The SRTM 30 m spatial 255 resolution DEM data was combined with the dam height to determine the water level height and 256 its intersection with valley topography upstream. 2.2.5 Calculation of the potential outburst flood peak discharge 232 This was the overlain on the DEM and, using 257 the cut fill tool of the Arcgis 10.2 allowed the volume of water stored for each dam to be 258 determined. Table 3 illustrated example calculations and also the sensitivity to the dam 259 erodibility parameter ɑ. In this case, we could also back-validate the Peng and Zhang (2012) 260 model against the measured peak discharge calculation for the Baige landslide. For the first 261 outburst flood, and using ɑ = -0.38, the landslide dam height was 47m corresponding with an 262 actual outburst flood peak discharge of 10000 m3s-1. Equation (7) gives 9590 m3s-1 for that dam 263 height (47m). The model result of peak discharge is very close to the actual result, which means 264 Peng and Zhang (2012) model is effective. For the second outburst flood, the landslide dam 265 height was 72m corresponding with an actual outburst flood peak discharge of 31000 m3s-1. 266 Equation (7) gives 11347 m3s-1 for that dam height (72m) with using ɑ = -0.38. This model result 267 appears distorted compared with the second peak discharge with the dam height significantly 268 increasing. 269 Whilst this analysis gives the peak discharge at the dam breach, it was also necessary to quantify 270 the evolution of the peak discharge with distance downstream due to attenuation. Data 271 suggested that the shape of the hydrograph retained similarity as the flood wave moved 272 downstream. This allowed the shape of the simulated hydrographs to be scaled to the estimated 273 peak discharge (Figure 9 and Figure 10). 274 275 Fig. 9 Dam downstream outburst hydrographs. (a) first dam breach, October 2018. (b) second dam breach, 276 November 2018 277 275 Fig. 9 Dam downstream outburst hydrographs. (a) first dam breach, October 2018. (b) second dam breach, 276 November 2018 277 278 Fig. 10 Potential peak discharge attenuation downstream of the dam 279 Fig. 9 Dam downstream outburst hydrographs. (a) first dam breach, October 2018. (b) second dam breach, 276 November 2018 277 November 2018 277 278 Fig. 10 Potential peak discharge attenuation downstream of the dam 279 278 Fig. 10 Potential peak discharge attenuation downstream of the dam 279 Fig. 10 Potential peak discharge attenuation downstream of the dam 3.1 Validation of flood inundation simulations 281 Figure 11 shows the estimated active river bed before the second flood and then after, the latter 282 likely indicating the maximum inundation extent. Remote sensing is reliable for validating flood 283 simulation. The remote sensing result suggest that the river was inundated to an extent of 1.5 284 km2 (i.e. in the first flood, before the second flood), and that this area increased by 1.6 km2 in the 285 second flood, such that the total flooded area in the study area is 3.1 km2. 286 Figure 11 shows the estimated active river bed before the second flood and then after, the latter 282 likely indicating the maximum inundation extent. Remote sensing is reliable for validating flood 283 simulation. The remote sensing result suggest that the river was inundated to an extent of 1.5 284 km2 (i.e. in the first flood, before the second flood), and that this area increased by 1.6 km2 in the 285 second flood, such that the total flooded area in the study area is 3.1 km2. 286 Figure 11 shows the estimated active river bed before the second flood and then after, the latter 282 likely indicating the maximum inundation extent. Remote sensing is reliable for validating flood 283 simulation. The remote sensing result suggest that the river was inundated to an extent of 1.5 284 km2 (i.e. in the first flood, before the second flood), and that this area increased by 1.6 km2 in the 285 second flood, such that the total flooded area in the study area is 3.1 km2. 286 287 Fig. 11 The November dam outburst flood illustrating the active zone of the river before the flood (a), after the 288 flood (b) and the estimated flood area that results (c) 289 287 Fig. 11 The November dam outburst flood illustrating the active zone of the river before the flood (a), after the 288 flood (b) and the estimated flood area that results (c) 289 Figure 12 superimposes the modeled and predicted inundation and is visually encouraging. This 290 was confirmed quantitatively with an overall accuracy reaches of to 0.956 and a Kappa value of 291 0.911. Figure 13 confirms a qualitatively good agreement between the measured and the 292 modelled water depths. Mismatches in this figure are likely to reflect inaccuracies in the DEM 293 data and local inconsistencies relating to minor topographical details. 3.1 Validation of flood inundation simulations 281 294 Figure 12 superimposes the modeled and predicted inundation and is visually encouraging. This 290 was confirmed quantitatively with an overall accuracy reaches of to 0.956 and a Kappa value of 291 0.911. Figure 13 confirms a qualitatively good agreement between the measured and the 292 modelled water depths. Mismatches in this figure are likely to reflect inaccuracies in the DEM 293 data and local inconsistencies relating to minor topographical details. 294 295 Fig. 12 The flood extent measured using remote sensing (a), modelled (b) and difference between the two (c) 296 Fig. 12 The flood extent measured using remote sensing (a), modelled (b) and difference between the two (c) Fig. 12 The flood extent measured using remote sensing (a), modelled (b) and difference between the two (c) 296 297 g g g ( ), ( ) ( ) 297 Fig.13 The flooded depth difference between the interpretation and the simulation result 298 Fig.13 The flooded depth difference between the interpretation and the simulation result 298 Fig.13 The flooded depth difference between the interpretation and the simulation result 3.2 Potential landslide dam heights 299 3.2 Potential landslide dam heights 9 Investigation and monitoring results show that CZ1-1, CZ1-2, and CZ2-1 were at the highest risk of 300 potential failure (Figure 5a). Their volumes were about 0.8×106m3, 0.8×106m3, and 3.3×106m3. In 301 addition, there is about 1.0×106m3 landslide debris residue in the landslide groove created from 302 the last landslide event. Although the total volume of potential high-risk zones was merely 303 5.9×106m3, the potential hazards and losses will likely be similar serious. Notably, CZ1-1, CZ1-2, 304 and CZ2-1 are the highest risk zones now. The crack zone boundary is further up-slope and the 305 extent of potential slope instability may be bigger (Figure 5). 306 Based on the shape of the 2018 dams, the dam breach deposit shape, and potential landslide 307 volume, the potential dam height can be calculated. Figure 14 shows three scenarios for potential 308 landslide failures in terms of the high-risk zone covering ‘CZ1-1’, ‘CZ1-1+CZ2-1’, and 309 ‘CZ1-1+CZ1-2+CZ2-1’. The three scenarios are as follows: ‘CZ1-1’ is likely the first failure because 310 of the highest risk; CZ1-1 and CZ2-1 (‘CZ1-1+CZ2-1’) are likely simultaneously start-up considering 311 Investigation and monitoring results show that CZ1-1, CZ1-2, and CZ2-1 were at the highest risk of 300 potential failure (Figure 5a). Their volumes were about 0.8×106m3, 0.8×106m3, and 3.3×106m3. 3.1 Validation of flood inundation simulations 281 In 301 addition, there is about 1.0×106m3 landslide debris residue in the landslide groove created from 302 the last landslide event. Although the total volume of potential high-risk zones was merely 303 5.9×106m3, the potential hazards and losses will likely be similar serious. Notably, CZ1-1, CZ1-2, 304 and CZ2-1 are the highest risk zones now. The crack zone boundary is further up-slope and the 305 extent of potential slope instability may be bigger (Figure 5). 306 the most severe deformation; CZ1-1, CZ1-2, and CZ2-1 (‘CZ1-1+CZ1-2+CZ2-1’) are likely 312 simultaneously failures because the crack zones are interconnected (Figure 5). The associated 313 landslide failure volumes are 1.3×106m3, 2.6×106m3, 5.9×106m3, and form landslide dams 30m, 314 35m, and 51m high, respectively (Figure 14). 315 316 Fig. 14 Potential dam heights caused by future blockage by the Baige landslide 317 Fig. 14 Potential dam heights caused by future blockage by the Baige landslide 317 3.3 Flood simulations for the November 2018 event 318 The main three simulation outputs were depths, velocities, and the maximum flood inundated 319 area of the outburst flood. The total simulation time is 46,800s, which has been divided into four 320 steps that included 10,800s, 21,600s, 324,00s, and 43,200s for visualisation. Because the 321 landslide dam breaching process was slow, the flood process lasted a long time; whilst the study 322 area was only 9km long, the flood lasted more than 13 hours. The peak discharge time was at 323 10800s (Figures 15 and 16). As shown in Figure 15, the maximum flood depth was about 27m at 324 10,800s, it then has sharply decreased to 15m by 43,200s. As shown in Figure 16, most of the 325 flood velocities were lower than 12 ms-1 at peak discharge (10,800s). However, there were locally 326 exceptionally high velocities (19 ms-1 at peak, declining to 13s-1 by 43,200s) due to a bedrock 327 valley constraint. The maximum flood inundation at the 10,800s stage is shown in Figures 15 and 328 16. 329 The main three simulation outputs were depths, velocities, and the maximum flood inundated 319 area of the outburst flood. The total simulation time is 46,800s, which has been divided into four 320 steps that included 10,800s, 21,600s, 324,00s, and 43,200s for visualisation. Because the 321 landslide dam breaching process was slow, the flood process lasted a long time; whilst the study 322 area was only 9km long, the flood lasted more than 13 hours. The peak discharge time was at 323 10800s (Figures 15 and 16). As shown in Figure 15, the maximum flood depth was about 27m at 324 10,800s, it then has sharply decreased to 15m by 43,200s. As shown in Figure 16, most of the 325 flood velocities were lower than 12 ms-1 at peak discharge (10,800s). However, there were locally 326 exceptionally high velocities (19 ms-1 at peak, declining to 13s-1 by 43,200s) due to a bedrock 327 valley constraint. The maximum flood inundation at the 10,800s stage is shown in Figures 15 and 328 16. 329 330 Fig. 15 The outburst flood depth evolution process based on the BASEMENT model 331 330 Fig. 15 The outburst flood depth evolution process based on the BASEMENT model 331 331 332 Fig. 16 The outburst flood velocity evolution process based on the BASEMENT model 333 332 Fig. 3.3 Flood simulations for the November 2018 event 318 16 The outburst flood velocity evolution process based on the BASEMENT model 333 3.4 Upstream impacts of different landslide dam scenarios 334 The 30m, 35m, and 51m dam height scenarios induce the storage capacity up to 142×106 m3, 335 182×106 m3, and 331×106 m3, with the back-flow distances of 32 km, 40 km, and 50 km, 336 respectively (Table 4 and Figure 17 ). The potential flooded area upstream of these dams is 5.01 337 km2, 6.12 km2, and 9.1 km2, respectively (Table 4). Compared with the two landslide dam flood 338 events of October and November 2018, the flooded area and back-flow distance of the potential 339 events are lower, but the disaster is no less serious. Take Boluo Town as an example, located 20 340 km upstream of the landslide dam, the 2018 floods inundated the whole town. The scenarios will 341 also inundate the whole town (Figure 17). There is some non-linearity in both storage volume 342 and upstream inundated area reflecting the form of the valley upstream of the Baige landslide. 343 The storage volume begins to increase more rapidly with dam heights greater than 50m (Figure 344 18). This increase occurs slightly earlier for dam heights greater than 45 m (Figure 18) but then it 345 slows for dam heights greater than 55 m. 346 and upstream inundated area reflecting the form of the valley upstream of the Baige landslide. 343 The storage volume begins to increase more rapidly with dam heights greater than 50m (Figure 344 18). This increase occurs slightly earlier for dam heights greater than 45 m (Figure 18) but then it 345 slows for dam heights greater than 55 m. 346 Table 4 Dam upstream potential landslide dam disasters. The table shows data for the measured 2018 events 347 and the three simulated dam heights 348 Dam height (breach, m) Dam volume (×104m3) Storage capacity (×106m3) Flooded area (km2) Back-flow distance (km) 30 130 142 5.01 32 35 260 182 6.12 40 47 measured 1800 290 9.1 45 51 590 331 10.75 50 72 measured 630 580 16.43 70 Table 4 Dam upstream potential landslide dam disasters. The table shows data for the measured 2018 events 347 and the three simulated dam heights 348 Dam height (breach, m) Dam volume (×104m3) Storage capacity (×106m3) Flooded area (km2) Back-flow distance (km) 30 130 142 5.01 32 35 260 182 6.12 40 47 measured 1800 290 9.1 45 51 590 331 10.75 50 72 measured 630 580 16.43 70 349 Fig. 3.4 Upstream impacts of different landslide dam scenarios 334 17 Upstream inundation extent for different dam heights 350 Table 4 Dam upstream potential landslide dam disasters. The table shows data for the measured 2018 events and the three simulated dam heights Table 4 Dam upstream potential landslide dam disasters. The table shows data for the measured 2018 eve 7 72 measured 349 Fig. 17 Upstream inundation extent for different dam heights 350 Fig. 17 Upstream inundation extent for different dam heights 351 Fig. 18 Dam height, storage capacity, and dam volume relation upstream 352 351 Fig. 18 Dam height, storage capacity, and dam volume relation upstream 352 Fig. 18 Dam height, storage capacity, and dam volume relation upstream 352 Fig. 18 Dam height, storage capacity, and dam volume relation upstream 3.5 Outburst flood impacts downstream of the simulated dams 353 Under the peak discharge calculation by the attenuation trend, the peak discharges are 5300 354 m3s-1, 5644 m3s-1, and 6664 m3s-1 corresponding to the dam heights 30 m, 35 m, and 51m at the 355 study area located at Batang county reach, respectively. The BASEMENT simulation results show 356 that the potential flood areas are 1.674 km2, 1.833 km2, and 1.927 km2 under the three different 357 dam height scenarios (Figure 19). These compare with a proportionately much higher flooded 358 area for the inundated area of the 72 m height November 2018 inundation event and this reflects 359 the fact that the inundatable width of the floodplain increases markedly for the discharges 360 associated with the higher case (Figure 20). 361 362 Fig. 19 The flooded area and water depth downstream for failure at different dam heights. Note that only one 363 of the measured events is simulated 364 362 Fig. 19 The flooded area and water depth downstream for failure at different dam heights. Note th 363 Fig. 19 The flooded area and water depth downstream for failure at different dam heights. Note that only one Fig. 19 The flooded area and water depth downstream for failure at different dam heights. Note that only one 363 of the measured events is simulated 364 of the measured events is simulated 364 364 of the measured events is simulated 365 Fig. 20 The relation between dam height and potential inundated area for the study reach 366 Fig. 20 The relation between dam height and potential inundated area for the study reach 4 Discussion 367 408 Table 5 Empirical models of the peak discharge calculation of the Baige first two landslides 409 Source Empirical equation Peak discharge (m3/s-1) Dam height 47m Dam height 72m First, the potential dam heights were impacted by the dam shape, dam breach deposit, and 380 landslide volume. Based upon the landslide accumulation process, the dam shape and dam 381 breach deposit size are easily determined after the breach. However, accurately determining 382 landslide volume before the breach is harder. The volume calculation appeared distinct 383 differences between different scholars even in the failed Baige landslide events (Fan et al. 2020; 384 Xu et al. 2018; Chen et al. 2021; Gao et al. 2021). Fan and Xu et al. (Fan et al. 2020; Xu et al. 2018) 385 calculated the Baige landslide volume to be ~ 30.0×106 m3 in the first and 12.0×106m3 in the 386 second by superposing estimated topography before and after failure; Chen et al. (Chen et al. 387 2021) calculated the volumes to be 27.0×106 m3 in the first and 8.0×106m3 in the second also 388 using topographic superposition methods; Chen and Gao et al. (Chen et al. 2021; Gao et al. 2021) 389 calculated the volumes to be 18.7×106 m3 in the first and 6.3×106m3 in the second using field 390 measurements and before and after topographic superposition methods. The potential 391 deformation range and unstable zones of the Baige landslide themselves have also differed 392 according to the methods adopted (Fan et al. 2020; Chen et al. 2021). This work shows that the 393 dam height has an important and non-liner impact upon both upstream inundation (Figure 18) 394 and notably downstream outburst flood extent (Figure 20), emphasizing that determining the 395 potential landslide volume and the form of the dam it produces is a key challenge for flood 396 hazard assessment. 397 Second, the peak discharge is a critical diver of the outburst flood simulation. Although there is a 398 long history in the application of empirical determinations of peak discharge, it is not clear that a 399 universal model exists that is suitable for all regions lets alone all landslide within a region (Costa 400 and Schuster 1988; Walder and O’Connor 1997; Fread 1988). 4 Discussion 367 The Baige landslide is continuously deforming due to gravity and this resulted in landslide dams 368 that breached on two closely spaced dates in 2018 (Fan et al. 2020; Chen et al. 2021; Liu et al. 369 2021; Yang et al. 2021). The potentially unstable slope continues to erode is likely already 370 The Baige landslide is continuously deforming due to gravity and this resulted in landslide dams 368 that breached on two closely spaced dates in 2018 (Fan et al. 2020; Chen et al. 2021; Liu et al. 369 2021; Yang et al. 2021). The potentially unstable slope continues to erode is likely already 370 blocking the Jinsha River again. This paper, using the Baige landslide as a case, focused on 371 quantifying how the dam drives inundation both upstream and downstream of the dam, the 372 former due to storage of water upstream, the latter due to an outburst flood with impacts 373 downstream. Remote sensing was combined with the hydrodynamic model BASEMENT to 374 estimate flood extent, and with a good agreement between measurements and model 375 predictions (Figure 12, Figure 13) reflected in Kappa value >0.9. This approach was aided by high 376 quality remotely-sensed imagery for validation. The BASEMENT modelling, however, is crucially 377 dependent upon estimates of the size of the flood discharge, and hence dam heights, the 378 liberation of water during failure and attenuation of the discharge wave as it moves downstream. 379 estimate flood extent, and with a good agreement between measurements and model 375 predictions (Figure 12, Figure 13) reflected in Kappa value >0.9. This approach was aided by high 376 quality remotely-sensed imagery for validation. The BASEMENT modelling, however, is crucially 377 dependent upon estimates of the size of the flood discharge, and hence dam heights, the 378 liberation of water during failure and attenuation of the discharge wave as it moves downstream. 379 First, the potential dam heights were impacted by the dam shape, dam breach deposit, and 380 landslide volume. Based upon the landslide accumulation process, the dam shape and dam 381 breach deposit size are easily determined after the breach. However, accurately determining 382 landslide volume before the breach is harder. The volume calculation appeared distinct 383 differences between different scholars even in the failed Baige landslide events (Fan et al. 2020; 384 Xu et al. 2018; Chen et al. 2021; Gao et al. 2021). 4 Discussion 367 Fan and Xu et al. (Fan et al. 2020; Xu et al. 2018) 385 calculated the Baige landslide volume to be ~ 30.0×106 m3 in the first and 12.0×106m3 in the 386 second by superposing estimated topography before and after failure; Chen et al. (Chen et al. 387 2021) calculated the volumes to be 27.0×106 m3 in the first and 8.0×106m3 in the second also 388 using topographic superposition methods; Chen and Gao et al. (Chen et al. 2021; Gao et al. 2021) 389 calculated the volumes to be 18.7×106 m3 in the first and 6.3×106m3 in the second using field 390 measurements and before and after topographic superposition methods. The potential 391 deformation range and unstable zones of the Baige landslide themselves have also differed 392 according to the methods adopted (Fan et al. 2020; Chen et al. 2021). This work shows that the 393 dam height has an important and non-liner impact upon both upstream inundation (Figure 18) 394 and notably downstream outburst flood extent (Figure 20), emphasizing that determining the 395 potential landslide volume and the form of the dam it produces is a key challenge for flood 396 hazard assessment. 397 Second, the peak discharge is a critical diver of the outburst flood simulation. Although there is a 398 long history in the application of empirical determinations of peak discharge, it is not clear that a 399 universal model exists that is suitable for all regions lets alone all landslide within a region (Costa 400 and Schuster 1988; Walder and O’Connor 1997; Fread 1988). Most empirical models only 401 undertake simple fitting of geometrical parameters for different dams in a particular area and use 402 few parameters (Costa 1985; Evans 1986; Cenderelli 2000; Pierce et al. 2010; Habib et al. 2014; Li 403 2020). The estimations of peak discharge can be highly sensitive to the model used, as shown for 404 the Baige landslide (Table 5). It implies that specific models need to be selected according to the 405 specific study area. More challenging is the fact that the correct discharge-height model may 406 depend on landslide height. With only two landslide events available to calibrate such models 407 this uncertainty is extreme and a major weakness with the work presented in this paper. 4 Discussion 367 However, there is some evidence 413 that the form of these decay curves (Figure 10, 11) depends on the dam height with a more 414 non-liner decay in peak discharge with distance for the first and smaller of the two dam breach 415 events. This suggests that hydraulic modelling of flood wave attenuation should be explored. 416 Finally, the BASEMENT model validation was surprisingly good given that the model had not been 417 optimized against remotely-sensed inundation extent. That said, there are a range of known 418 impacts of hydraulic model parameters on flood extent, notably roughness (e.g. Yu and Lane, 419 2006) as respresented here by the Manning friction coefficient, and also the quality of floodplain 420 topography. The mesh used for model solution may also have a significant impact (Lala et al. 421 2018). Testing of mesh effects may be worthwhile, especially as these interact with model 422 parameters and notably friction parameterization. A crucial finding of this work was that the 423 relationship between dam height and downstream flood extent is likely to be non-liner (Figure 20) 424 emphasizing the importance of using hydrodynamic models for evaluating the downstream 425 impacts of outburst flood events. 426 4 Discussion 367 Most empirical models only 401 undertake simple fitting of geometrical parameters for different dams in a particular area and use 402 few parameters (Costa 1985; Evans 1986; Cenderelli 2000; Pierce et al. 2010; Habib et al. 2014; Li 403 2020). The estimations of peak discharge can be highly sensitive to the model used, as shown for 404 the Baige landslide (Table 5). It implies that specific models need to be selected according to the 405 specific study area. More challenging is the fact that the correct discharge-height model may 406 depend on landslide height. With only two landslide events available to calibrate such models 407 this uncertainty is extreme and a major weakness with the work presented in this paper. 408 409 Table 5 Empirical models of the peak discharge calculation of the Baige first two landslides 9 Source Empirical equation Peak discharge (m3/s-1) Dam height 47m Dam height 72m Table 5 Empirical models of the peak discharge calculation of the Baige first two landslides Costa (1985) 17707 28339 Froehlich (1995)   24 1 295 0 607 0 . w . l p H V . Q  22540 46924 Evans (1986) 21999 31765 USBR(1982) 85 1 1 19 . w p H . Q  23681 52131 McDonald and Landridge Monopolis (1984)   412 0 154 1 . w l p H V . Q  17282 27413 Cenderelli (2000) 26557 36530 Pierce et al. (2010) 21505 33655 24379 78050 Peng and Zhang (2012) 9590 11347 Measured 10000 31000 Note: Hw = water depth (m); Hd = dam height (m); Vl = storage capacity (×106m3)(as shown in Table 3); Hr = Reference height (1.0); a is erodibility parameter (Peng and Zhang (2012)), here a = 0.38. 0.42 l d P V H . Q ）（ 981 0  53 0 72 0 . l P V . Q  46 0 4 3 . l p V . Q  40 0 9 1 . l d P V H ( . Q ）  606 0 0176 0 . l d p ) V H ( . Q  a . d / l . 4 Discussion 367 r d / d / p e ) H / V ( ) H / H ( ) H g /( Q 536 1 3 1 371 1 2 5 2 1   Note: Hw = water depth (m); Hd = dam height (m); Vl = storage capacity (×106m3)(as shown in Table 3); Hr = Reference height (1.0); a is erodibility parameter (Peng and Zhang (2012)), here a = 0.38. Third, the inundation extent downstream is not only a function of the magnitude of the dam 410 breach but also the attenuation of the flood wave. In this case, empirical evidence suggested a 411 relatively simply decline in peak discharge with distance downstream (Figure 11) and little change 412 in flood wave shape (Figure 10). This facilitated the analysis. However, there is some evidence 413 that the form of these decay curves (Figure 10, 11) depends on the dam height with a more 414 non-liner decay in peak discharge with distance for the first and smaller of the two dam breach 415 events. This suggests that hydraulic modelling of flood wave attenuation should be explored. 416 Finally, the BASEMENT model validation was surprisingly good given that the model had not been 417 optimized against remotely-sensed inundation extent. That said, there are a range of known 418 impacts of hydraulic model parameters on flood extent, notably roughness (e.g. Yu and Lane, 419 2006) as respresented here by the Manning friction coefficient, and also the quality of floodplain 420 topography. The mesh used for model solution may also have a significant impact (Lala et al. 421 2018). Testing of mesh effects may be worthwhile, especially as these interact with model 422 parameters and notably friction parameterization. A crucial finding of this work was that the 423 relationship between dam height and downstream flood extent is likely to be non-liner (Figure 20) 424 emphasizing the importance of using hydrodynamic models for evaluating the downstream 425 impacts of outburst flood events. 426 Third, the inundation extent downstream is not only a function of the magnitude of the dam 410 breach but also the attenuation of the flood wave. In this case, empirical evidence suggested a 411 relatively simply decline in peak discharge with distance downstream (Figure 11) and little change 412 in flood wave shape (Figure 10). This facilitated the analysis. 5 Conclusion 427 The Baige landslide dam has failed and twice-triggered severe outburst floods. Unfortunately, the 428 Baige landslide is continuously deforming and continues to be a significant potential hazard. 429 BASEMENT is an accurate and effective hydrodynamic model for simulating the outburst flooding 430 process, and associated scenarios can be used to assess future flood hazard. This was confirmed 431 here by comparison with remotely sensed data. Three key conclusions follow. First, remote 432 sensing showed that the second time the Baige dam breached, it triggered an inundated area 433 reaching up to 3.1 km2 within a length 9 km of downstream river valley. Second, the Baige 434 landslide is continuously deforming which implies a high probability of a future outburst flood 435 events. Third, there is some evidence that the relationships between landslide height and both 436 upstream inundation extent and downstream outburst flood extent are non-liner. This 437 emphasizes that the likely consequences of an outburst flood will depend upon the size of the 438 landslide dam and also the time since the last failure event. 439 440 Funding: This study was financially supported by The Second Tibetan Plateau Scientific Expedition 441 and Research Program (No.2019QZKK0905). 442 443 Data Availability Statement: The satellite data and DEM data present in this study are 444 non-openly available (https://account.planet.com/). 445 446 Ethical statements：This manuscript has not been published or presented elsewhere in part or in 447 entirety and is not under consideration by another journal. I have read and understood your 448 journal’s policies, and we believe that neither the manuscript nor the study violates any of these. 449 There are no conflicts of interest to declare. 450 451 Conflicts of Interest: The authors declare no conflict of interest. 452 453 Reference 454 An HC, Ouyang CJ, Zhou S (2021) Dynamic process analysis of the Baige landslide by the 455 combination of DEM and long-period seismic waves. Landslides 18(5): 1625-1639. 456 Bonnard C (2006) Technical and human aspects of historic rockslide dammed lakes and landslide 457 dam breaches. Italian Journal of Engineering Geology and Environment 1: 21-30. 458 Bishop Y, Fienberg S, Holland P (1975) Discrete Multivariate Analysis: Theory and Practice. MIT 459 Press: Cambridge, MA. 460 Briaud JL (2008) Case histories in soil and rock erosion: woodrow wilson bridge, Brazos River 461 Meander, Normandy Cliffs, and New Orleans Levees. Journal Geotechnical and 462 Geoenvironmental Engineering 134(10):1425-1447. 5 Conclusion 427 463 Benz UC, Hofmann P, Willhauck G, Lingenfelder L (2004) Multi-resolution, object-oriented fuzzy 464 analysis of remote sensing data for GIS-ready information. ISPRS Journal of Photogrammetry 465 and Remote Sensing 58:239–258. 466 Chen F, Gao YJ, Zhao SY, Deng JH, Li ZL, Ba RJ, Yu ZQ, Yang ZK, Wang S (2021) Kinematic process 467 and mechanism of the two slope failures at Baige Village in the upper reaches of the Jinsha 468 River, China. Bulletin of Engineering Geology and the Environment 80(4): 3475-3493. 469 Chow VT, Maidment DR, Mays LW (2011) Applied Hydrology. New York: McGraw-Hill . Costa JE, Schuster RL (1988) The formation and failure of natural dams. Geol. Soc. Am. Bull. 100 471 (7), 1054-1068. 472 Chen Z, Zhou HF, Ye F, Liu B, Fu WX (2021) The characteristics, induced factors, and formation 473 mechanism of the 2018 Baige landslide in Jinsha River, Southwest China. Catena 203: 474 105337. 475 Costa JE (1985) Floods from dam failures. US Geological Survey, Open-File Report 85-560, pp 54. 476 Costa JE, Schuster RL (1988) The formation and failure of natural dams. Geological Society of 477 America Bulletin 100:1054-1068. 478 Costa JE, Schuster RL (1991) Documented historical landslide dams from around the world. US 479 Geological Survey Open-File Report 91-239, pp 486. 480 Cohen JA (1960) coefficient of agreement for nominal scales. Educational and psychological 481 measurement 20(1): 37-46. 482 Cenderelli DA (2000) Floods from natural and artificial dam failures, in: Inland Flood Hazards, 483 edited by: Wohl, E. E.. Cambridge University Press, New York 73-103. 484 Delaney KB, Evans SG (2015) The 2000 Yigong landslide (Tibetan Plateau), rockslide-dammed lake 485 and outburst flood: Review, remote sensing analysis, and process modelling. 486 Geomorphology 246: 377-393. 487 Ding C, Feng GG, Liao MS, Tao PG, Zhang L, Xu Q (2021) Displacement history and potential 488 triggering factors of Baige landslides, China revealed by optical imagery time series. Remote 489 Sensing of Environment 254: 112253. 490 Evans SG, Delaney KB (2011) Characterization of the 2000 Yigong Zangbo River (Tibet) landslide 491 dam and impoundment by remote sensing. Natural and artificial rockslide dams 543-559. 492 Evans SG (1986) The maximum discharge of outburst floods caused by the breaching of 493 man-made and natural dams. Can.Geotech. J. 23:385-387. 494 Evans SG (1986) The maximum discharge of outburst floods caused by the breaching of 493 man-made and natural dams. Can.Geotech. J. 23:385-387. 5 Conclusion 427 494 Everitt BS (1998) The Cambridge Dictionary of Statistics. Cambridge University Press: Cambridge. 495 Ermini L, Casagli N (2003) Prediction of the behaviour of landslide dams using a 496 geomorphological dimensionless index. Earth Surface Processes and Landforms: The Journal 497 of the British Geomorphological Research Group 28(1): 31-47. 498 Fan XM, Dufresne A, Subramanian SS, Alexander S, Hermannsd R, Stefanelli CT, Hewitt K, Yunus 499 AP, Dunning S, Capra L, Geertsema M, Miller B, Casagli N, John DJ, Xu Q (2020) The 500 formation and impact of landslide dams–State of the art. Earth-Science Reviews 203: 501 103116. 502 Fan XM, Scaringi G, Korup O, West AJ, Westen Cees JV, Tanyas H, Hovius N, Hales TC, Jibson RW, 503 Allstadt KE, Zhang LM, Evans SG, Xu C, Li G, Pei XG, Xu Q, Huang RQ (2019) Earthquake฀ 504 induced chains of geologic hazards: Patterns, mechanisms, and impacts. Reviews of 505 geophysics 57(2): 421-503. 506 Fan XM, Yang F, Subramanian SS, Xu Q, Feng ZT, Olga M, Peng M, Ouyang CJ, John DJ, Huang RQ 507 (2020) Prediction of a multi-hazard chain by an integrated numerical simulation approach: 508 the Baige landslide, Jinsha River, China. Landslides 17(1): 147-164. 509 Fread DL (1988) BREACH: an erosion model for earth dam failures. National Weather Service 510 (NWS) Report, NOAA, Silver Spring, Maryland, USA. 511 Froehlich, DC (1995) Peak outflow from breached embankment dam. J. Water Resour. Pl. 512 Manage. 121: 90–97. 513 Gao YJ, Zhao SY, Deng JH, Yu ZQ, Mahfuzur R (2021) Flood assessment and early warning of the 514 reoccurrence of river blockage at the Baige landslide. Journal of Geographical Sciences 515 31(11): 1694-1712. 516 Guo CB, Yan YQ, Zhang YS, Zhang XJ, Zheng YZ, Li X, Yang ZH, Wu RA (2021) Study on the 517 Creep-Sliding Mechanism of the Giant Xiongba Ancient Landslide Based on the SBAS-InSAR 518 Method, Tibetan Plateau, China. Remote Sensing 13(17): 3365. 519 Habib H, Vahid N, Alireza BA (2014) Genetic programming simulation of dam breach hydrograph 520 and peak outflow discharge. Journal of Hydrologic Engineering 757-768. 521 Hanson GJ, Simon A (2001) Erodibility of cohesive streambeds in the loess area of the 522 midwestern USA. Hydrological processes 15(1): 23-38. 523 Knapp S, Gilli A, Anselmetti FS, Krautblatter M, Hajdas L (2018) Multistage rock฀slope failures 524 revealed in lake sediments in a seismically active Alpine region (Lake Oeschinen, 525 Switzerland). 5 Conclusion 427 Journal of Geophysical Research: Earth Surface 123(4): 658-677. 526 Köpfli P, Grämiger LM, Moore JR, Christof V, Susan LO (2018) The Oeschinensee rock avalanche, 527 Bernese Alps, Switzerland: a co-seismic failure 2300 years ago? Swiss J. Geosci. 111: 205– 528 219. 529 Lane SN (1998) Hydraulic modelling in hydrology and geomorphology: a review of high resolution 530 approaches. Hydrological Processes 12(8): 1131-1150. 531 Liu XJ, Zhao CY, Zhang Q, Lu Z, Li ZH (2020) Deformation of the Baige landslide, Tibet, China, 532 revealed through the integration of cross฀platform ALOS/PALSAR฀1 and ALOS/PALSAR฀2 533 SAR observations. Geophysical Research Letters 47(3): e2019GL086142. 534 Li YS, Jiao QS, Hu XH, Li ZL, Li BQ, Zhang JF, Jiang WL, Luo Y, Li Q, Ba RJ (2020) Detecting the slope 535 movement after the 2018 Baige Landslides based on ground-based and space-borne radar 536 observations. International Journal of Applied Earth Observation and Geoinformation 84: 537 101949. 538 Liu W, Ju NP, Zhang Z, Chen Z, He SM (2020) Simulating the process of the Jinshajiang 539 landslide-caused disaster chain in October 2018. Bulletin of Engineering Geology and the 540 Environment 1-11. 541 Lala JM, Rounce DR, McKinney DC (2018) Modeling the glacial lake outburst flood process chain 542 in the Nepal Himalaya: reassessing Imja Tsho's hazard. Hydrology and Earth System Sciences 543 22(7): 3721-3737. 544 Li J, Cao ZX, Cui YF, Fan XM, Yang WJ, Huang W, Alistair B. (2021) 545 Hydro-sediment-morphodynamic processes of the baige landslide-induced barrier Lake, 546 Jinsha River, China. Journal of Hydrology 596: 126134. 547 Liu D, Cui Y, Wang H, Jin W, Wu CH, Nazir AB, Zhang GT, Paul AC, Chen HY (2021) Assessment of 548 local outburst flood risk from successive landslides: Case study of Baige landslide-dammed 549 lake, upper Jinsha river, eastern Tibet. Journal of Hydrology 599: 126294. 550 Li YL, Chen AK, Wen LF, Bu P, Li KP (2020) Numerical simulation of non-cohesive homogeneous 551 dam breaching due to overtopping considering the seepage effect. European journal of 552 environmental and civil engineering 1-15. 553 Lala JM, Rounce DR, McKinney DC (2018) Modeling the glacial lake outburst flood process chain 554 in the Nepal Himalaya: reassessing Imja Tsho's hazard. Hydrology and Earth System Sciences 555 22(7): 3721-3737. 556 McDonald T C and Langridge-Monopolis J (1984) Breaching characteristics of dam failures. J. 557 Hydraul. Eng. 110:567–586. 5 Conclusion 427 558 Ouyang CJ, An HC, Zhou S, Wang ZW, Su PC, Wang DP, Cheng DX, She JX (2019) Insights from the 559 failure and dynamic characteristics of two sequential landslides at Baige village along the 560 Jinsha River, China. Landslides 16(7): 1397-1414. 561 Peng M, Zhang LM (2012) Breaching parameters of landslide dams. Landslides 9(1): 13-31. 562 Pierce MW, Thornton CL, Abt SR (2010) Predicting peak outflow from breached embankment 563 dams. J. Hydrol. Eng. 15: 338-349. 564 Shang YJ, Yang ZF, Li LH, Liu DA, Liao QL, Wang YC (2003) A super-large landslide in Tibet in 2000: 565 background, occurrence, disaster, and origin. Geomorphology 54(3-4): 225-243. 566 Shang YJ, Park HD, Yang ZF, Yang J (2005) Distribution of landslides adjacent to the northern side 567 of the Yarlu Tsangpo Grand Canyon in Tibet, China. Environmental geology 48(6): 721-741. 568 USBR – US Bureau of Reclamation (1982) Guidelines for defining inundated areas downstream 569 from Bureau of Reclamation dams, Reclamation Planning Instruction 82-11. Water 570 Resources ResearchLaboratory, Denver, CO. 571 Vanzo D, Peter S, Vonwiller L, Bürgler M, Weberndorfer M, Siviglia A, Conde D, Vetsch DF (2021) 572 BASEMENT v3: A modular freeware for river process modelling over multiple computational 573 backends. Environmental Modelling & Software 143: 105102. 574 Wu YB, Niu RQ, Lu Z (2019) A fast monitor and real time early warning system for landslides in 575 the Baige landslide damming event, Tibet, China. Natural Hazards and Earth System Sciences 576 Discussions 1-20. 577 Wang HJ, Zhou Y, Wang SX, Wang FT (2020) Coupled model constructed to simulate the landslide 578 dam flood discharge: a case study of Baige landslide dam, Jinsha River. Frontiers of Earth 579 Science 14(1): 63-76. 580 Worni R, Huggel C, Stoffel M (2013) Glacial lakes in the Indian Himalayas—From an area-wide 581 glacial lake inventory to on-site and modeling based risk assessment of critical glacial lakes. 582 Science of the Total Environment 468: S71-S84. 583 Worni R, Stoffel M, Huggel C, Volz C, Casteller A, Luckman B (2012) Analysis and dynamic 584 modeling of a moraine failure and glacier lake outburst flood at Ventisquero Negro, 585 Patagonian Andes (Argentina). Journal of Hydrology 444: 134-145. 586 Wang SJ, Yang YD, Gong WY, Che YJ, Ma XG, Xie J (2021) Reason Analysis of the Jiwenco Glacial 587 Lake Outburst Flood (GLOF) and Potential Hazard on the Qinghai-Tibetan Plateau. Remote 588 Sensing 13(16): 3114. 5 Conclusion 427 589 Walder JS, O’Connor JE (1997) Methods for predicting peak discharge of floods caused by failure 590 of natural and constructed earthen dams. Water Resources Research 33(10):2337-2348. 591 Xu Q, Shang YJ, van Asch T, Wang ST, Zhang ZY, Dong XJ (2012) Observations from the large, rapid 592 Yigong rock slide–debris avalanche, southeast Tibet. Canadian Geotechnical Journal 49(5): 593 589-606. 594 Xu Q, Zheng G, Li WL, He MC, Dong XJ, Guo C, Feng WK (2018) Study on successive landslide 595 damming events of Jinsha River in Baige Village on October 11 and November 3, 2018. 596 Journal of engineering geology 26(6):1534-1551. 597 Xiong ZQ, Feng GC, Feng ZX, Miao L, Wang YD, Yang DJ, Luo SR (2020) Pre-and post-failure 598 spatial-temporal deformation pattern of the Baige landslide retrieved from multiple radar 599 and optical satellite images. Engineering Geology 279: 105880. 600 Yang ZK, Wei JB, Deng JH, Gao YJ, Zhao SY, He ZL (2021) Mapping Outburst Floods Using a 601 Collaborative Learning Method Based on Temporally Dense Optical and SAR Data: A Case 602 Study with the Baige Landslide Dam on the Jinsha River, Tibet. Remote Sensing 13(11): 2205. 603 Yu D, Lane SN (2006) Urban fluvial flood modelling using a two฀dimensional diffusion฀wave 604 treatment, part 1: mesh resolution effects. Hydrological Processes: An International Journal 605 606 20(7): 1541-1565. Yang WT, Wang YQ, Sun S, Wang YJ, Ma C (2019) Using Sentinel-2 time series to detect slope 607 movement before the Jinsha River landslide. Landslides 16(7): 1313-1324. 608 Zhang SL, Yin YP, Hu XW, Wang WP, Zhang N, Zhu SN, Wang LQ (2020) Dynamics and 609 emplacement mechanisms of the successive Baige landslides on the Upper Reaches of the 610 Jinsha River, China. Engineering Geology 278: 105819. 611 Zhang SL, Yin YP, Hu XW, Wang WP, Li ZL, Wu XM, Luo G, Zhu SN (2021) Geo-structures and 612 deformation-failure characteristics of rockslide areas near the Baige landslide scar in the 613 Jinsha River tectonic suture zone. Landslides 18(11): 3577-3597. 614 Zischg AP, Galatioto N, Deplazes S, Weingartner R, Mazzorana B (2018) Modelling spatiotemporal 615 dynamics of large wood recruitment, transport, and deposition at the river reach scale 616 during extreme floods. Water 10(9): 1134. 617 Zhou JW, Cui P, Hao MH (2016) Comprehensive analyses of the initiation and entrainment 618 processes of the 2000 Yigong catastrophic landslide in Tibet, China. Landslides 13(1): 39-54. 5 Conclusion 427 619 Zhou S, Ouyang CJ, Huang Y (2022) An InSAR and depth-integrated coupled model for potential 620 landslide hazard assessment. Acta Geotechnica 1-20. 621 Yang WT, Wang YQ, Sun S, Wang YJ, Ma C (2019) Using Sentinel-2 time series to detect slope 607 movement before the Jinsha River landslide. Landslides 16(7): 1313-1324. 608 Zhang SL, Yin YP, Hu XW, Wang WP, Zhang N, Zhu SN, Wang LQ (2020) Dynamics and 609 emplacement mechanisms of the successive Baige landslides on the Upper Reaches of the 610 Jinsha River, China. Engineering Geology 278: 105819. 611 Zhang SL, Yin YP, Hu XW, Wang WP, Li ZL, Wu XM, Luo G, Zhu SN (2021) Geo-structures and 612 deformation-failure characteristics of rockslide areas near the Baige landslide scar in the 613 Jinsha River tectonic suture zone. Landslides 18(11): 3577-3597. 614 Zischg AP, Galatioto N, Deplazes S, Weingartner R, Mazzorana B (2018) Modelling spatiotemporal 615 dynamics of large wood recruitment, transport, and deposition at the river reach scale 616 during extreme floods. Water 10(9): 1134. 617 Zhou JW, Cui P, Hao MH (2016) Comprehensive analyses of the initiation and entrainment 618 processes of the 2000 Yigong catastrophic landslide in Tibet, China. Landslides 13(1): 39-54. 619 Zhou S, Ouyang CJ, Huang Y (2022) An InSAR and depth-integrated coupled model for potential 620 landslide hazard assessment. Acta Geotechnica 1-20. 621
https://openalex.org/W2076955029	https://zenodo.org/records/1849688/files/article.pdf	English	null	Ticks in Transmission of Disease—Specimens Wanted.	Journal of the American Medical Association	1,908	public-domain	1,525	Hypersusceptibility to Serum in Human Beings. g Riverdale-on-Hudson, N. Y., Jan. 24, 1908. Riverdale-on-Hudson, N. Y., Jan. 24, 1908 To the Editor:\p=m-\Dr.Wiley's case of sudden death after the injection of antitoxin, in The Journal, Jan. 11, 1908, page 137, recalls a personal experience which may be of interest. In August, 1904, I was called to a case of tonsillar diphtheria. I injected 3,000 units of antitoxin into the muscles of the patient's back without any untoward effects. At the same time, as I was suffering with painful and swollen tonsils, I injected 1,000 units of antitoxin into myself as a precaution- ary measure, plunging the needle deeply into the muscles of the forearm. The dose for the patient had only the usual beneficial result: rapid recovery, and no disagreeable after- effects. On the other hand, I suffered a most violent attack of sneezing, which became violent and uncontrollable, even though firm pressure was made against the upper lip and the sides of the nose, and cold compresses were placed against the face. I left the sickroom in haste, and on my way toAvard the drug store the nose became intensely congested and respiration through the larynx and lungs more difficult every moment. The irritation in the trachea and bronchi caused a continuous cough. On reaching the drug store I took some brandy, think¬ ing that the increased circulation throughout the body would relieve the congestion of the nose, threat and lungs; but on account of the SAvollen condition of the tongue and throat, I had the greatest difficulty in SAvalloAving it, even though it Avas greatly diluted. I Avas enduring intense pain in the nose, eyes and throat, due to the enormous congestion and irritation of the parts, and there Avere also burning and SAvelling of the scalp. When I reached home the eyelids were SAvollen to im¬ mense proportions, making vision almost impossible; the tongue was SAvollen, making speech thick; the scalp, forehead, nose, lips and cheeks were SAvollen and erythematous; and the conjunctiA'oe were injected. During the afternoon and eA'ening the symptoms abated, as the result of a saline purge, douch¬ ing of the nose and eyes Avith a detergent solution, and the drinking of quantities of an alkaline water, Avhich I continued throughout the attack. After midnight I was aroused by an intense itching and burning of the entire surface of the body caused by urticaria. Ticks in Transmission of Disease\p=m-\SpecimensWanted. Washington, D. C., Feb. 12, 1908. as gto , , , To the Editor:\p=m-\Theagency of ticks in the transmission of diseases of man as well as of the lower animals is becoming of such importance that the Bureau of Entomology, in connec- tion with the investigation of the cattle tick, the transmitter of splenetic or Texas fever of cattle, is also taking up the study of the biology of the other ticks found in this country. study b o ogy country. The importance of such an investigation into the life history and habits of ticks has been apparent in South Africa and in Europe for nearly a decade, but not until within the past year (with the exception of the North American fever tick), has it been the case in this country. After failure by several scien- tists to connect the tick with the Rocky Mountain spotted fever, Dr. H. T. Ricketts, following careful and extensive experi- ments, has succeeded in demonstrating that what may now be known as the "spotted fever tick," Dermacentor occidentalis, is the active agent in the transmission of the disease. The details of these experiments he has presented through The Journal. In this brilliant demonstration he has rendered great service to the human family. But there is still much in connection with the source of the disease that remains to be discovered. What is the source of the disease? What ani¬ mal harbors the protozoan (if such it be), in its blood, etc.? The elucidation of the information needed will come with a further experimental study of the life history and habits of the ticks. We are glad to learn that Dr. Ricketts will con¬ tinue his research in Montana during the coming spring and trust that his already brilliant discoveries will be added to. y The importance of the valuable studies of Dr. L. O. Howard, chief of the Bureau of Entomology; Dr. J. B. Smith of New Jersey, Dr. E. P. Felt of New York and others on the biology of the mosquitoes is now fully appreciated. In order to im¬ press on the minds of the entomologie fraternity the impor¬ tance of the study of the ticks I prepared a paper for presen¬ tation at the meeting of the Association of Economic En¬ tomologists, held at Chicago, Dec. 27-28, 1907, in which the agency of ticks in the transmission of disease was reviewed. Hypersusceptibility to Serum in Human Beings. Applications of a solution of bicarbonate of soda did no good, and being restless, I went out of doors, thinking that the cool night air might give relief. After walking about in the open, thinly clad, for some time, I felt exhausted but little relieved. I then resorted to mopping the body Avith sponges dipped in alcohol, which was only partially successful, but eventually I found relief by the application of a spray of chloroform. A hypodermic injection of morphin only aggravated the condition. On the following day I called a fellow practitioner and he ordered a solution of sodium bi¬ carbonate as an application to the urticarial Avheals; this, hoAvever, increased the eruption as well as the itching; both Avere controlled, hoAvever, by the use of a chloroform spray. On the third day my appearance Avas again normal, and I Avas able to attend to my practice, though in a weakened condi¬ tion. On returning home I was again compelled to go to bed, resume the chloroform spray and treat the injected conjnnc- tivse and congested nasal mucous membrane. On the fourth day the hands and feet swelled greatly, and a troublesome edema of the glottis threatened, so that preparations AA-ere made for a tracheotomy in case breathing should become en¬ tirely obstructed. By means of a hot mustard footbath, with cold compresses to the neck and the internal administration of ergot, the symptoms were relieved. Nevertheless, for two weeks after the initial attack, fresh outbreaks occurred, con¬ sisting of urticaria and swellings of various parts of the body, including the scalp. There Avas also purpura covering ths entire body. The reaction at the point of injection in my arm was very severe, causing- a painful arm and forearm, but Ticks in Transmission of Disease\p=m-\SpecimensWanted. Washington, D. C., Feb. 12, 1908. able. The time is past when the owners of great city dailies can take the position that there is no connection between the editorial and advertising departments, or can with any degree of consistency or self-respect advocate civic righteousness on editorial pages ana advertise easy and safe means of commit¬ ting felonies on the advertising pages. It is to be hoped that the entire medical profession Of Mississippi will lend its support and endorsement to the Morning World, and that this example will lead the other newspapers of New Orleans to follow the World into paths of decency. of this bureau at Dallas. Texas, which is in charge of Mr. W. D. Hunter, under \vh:>" direction I am carrying on the Avork on the biology of ticks. Any further information as to collecting, etc., will be gladly furnished. W. A. Hooker, , Bureau of Entomology, U. S. Department of Agriculture. Ticks in Transmission of Disease\p=m-\SpecimensWanted. Washington, D. C., Feb. 12, 1908. g y The studies now being made by this bureau on the biology of the ticks, particularly as related to host relationship and geographical distribution require the cooperation of collectors from all parts of the country. The Bureau of Entomology will be glad to receive ticks from any physician, will identify them and will furnish any information desired. Because of their easy access the domestic animals have in the past fur¬ nished the greater part of the collections. While further col¬ lections from these animals are needed, extensive collections from game birds and mammals are also required. In connec¬ tion with the hunting trip ticks can be collected in a pill box or in vials carried in the pocket. Some ticks leave the host as soon as the blood commences to coagulate and to capture these it has been found necessary to examine the host at once over a white surface, such as a newspaper. The ticke collected will be of the greatest value if forwarded immedi¬ ately on collection, alive in small boxes, sufficient moisture being furnished by the use of green leaves or moist absorbent cotton. Ticks from different hosts should be kept separate, and data as to the locality, host, date and collector should ac¬ company the sending. On birds, the ticks are usually found attached to the head or neck, protected by the feathers. p y All material should be sent to the southern field laboratory Downloaded From: http://jama.jamanetwork.com/ by a University of Michigan User on 06/15/2015
https://openalex.org/W2004302519	https://projecteuclid.org/journals/abstract-and-applied-analysis/volume-2010/issue-none/On-the-Nevanlinnas-Theory-for-Vector-Valued-Mappings/10.1155/2010/864539.pdf	English	null	On the Nevanlinna's Theory for Vector-Valued Mappings	Abstract and applied analysis	2,010	cc-by	6,633	Hindawi Publishing Corporation Abstract and Applied Analysis Volume 2010, Article ID 864539, 15 pages doi:10.1155/2010/864539 Hindawi Publishing Corporation Abstract and Applied Analysis Volume 2010, Article ID 864539, 15 pages doi:10.1155/2010/864539 1. Introduction In 1980s, Ziegler 1 established Nevanlinna’s theory for the vector-valued meromorphic functions in ﬁnite dimensional spaces. After Ziegler some works in ﬁnite dimensional spaces were done in 1990s 2–4. In 2006, C. Hu and Q. Hu 5 considered the case of inﬁnite dimensional spaces and they investigated the E-valued meromorphic mappings deﬁned in the disk Cr {z : \|z\| < r}. In this article, by using Green function technique, we will consider this theory deﬁned in generic domain D ⊆C see Section 2. In Section 3, motivated by the work of 6–8, we will prove Chuang’s inequality, which is to compare the relationship between Tr, f and Tr, f′. Consequently, we obtain that the order and the lower order of fz and its derivative f′z are the same. This is an extension of an important result for meromorphic functions. Correspondence should be addressed to Zu-Xing Xuan, xuanzuxing@ss.buaa.edu.cn Correspondence should be addressed to Zu-Xing Xuan, xuanzuxing@ss.buaa.edu.cn Received 7 September 2009; Revised 25 January 2010; Accepted 17 April 2010 Received 7 September 2009; Revised 25 January 2010; Accepted 17 April 2010 Academic Editor: Jean Pierre Gossez Copyright q 2010 Z.-X. Xuan and N. Wu. 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 q 2010 Z.-X. Xuan and N. Wu. 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 purpose of this paper is to establish the ﬁrst and second fundamental theorems for an E- valued meromorphic mapping from a generic domain D ⊂C to an inﬁnite dimensional complex Banach space E with a Schauder basis. It is a continuation of the work of C. Hu and Q. Hu. For fz deﬁned in the disk, we will prove Chuang’s inequality, which is to compare the relationship between Tr, f and Tr, f′. Consequently, we obtain that the order and the lower order of fz and its derivative f′z are the same. Zu-Xing Xuan1, 2 and Nan Wu1 1 Department of Mathematical Sciences, Tsinghua University, Beijing 100084, China 2 Department of Mathematical Sciences, Tsinghua University, Beijing 100084, China 2 Basic Department, Beijing Union University, No.97, Bei Si Huan Dong Road, Chaoyang District, Beijing 100101, China 2 Basic Department, Beijing Union University, No.97, Bei Si Huan Dong Road, Chaoyang District, Beijing 100101, China 2. First and Second Fundamental Theorem in Generic Domains Let E, ∥· ∥ be a complex Banach space with a Schauder basis {ei} and the norm ∥· ∥. Thus an E-valued meromorphic mapping fz deﬁned in a domain D ⊆C can be written as fz 2 Abstract and Applied Analysis f1z, f2z, . . . , fkz, . . .. The elements of E are called vectors and are usually denoted by letters from the alphabet: a, b, c, d, . . .. The symbol 0 denotes the zero vector of E. We denote vector inﬁnity, complex number inﬁnity, and the norm inﬁnity by ∞, ∞, and ∞, respectively. A vector-valued mapping is called holomorphicmeromorphic if all fjz are holomorphic meromorphic. The jth derivative j 1, 2, . . . and the integration of fz are deﬁned by fjz f j 1 z, f j 2 z, . . . , f j k z, . . . , z fζdζ z f1ζdζ, z f2ζdζ, . . . , z fkζdζ, . . . , 2.1 2.1 respectively. We assume that f0z fz. A point z0 ∈D is called a “pole” or “ ∞-point” of fz f1z, . . . , fkz, . . . if z0 is a pole of at least one of the component functions fkz k 1, 2, . . .. We deﬁne ∥fz0∥ ∞when z0 is a pole. A point z0 ∈D is called “zero” of fz if all the component functions fkz k 1, 2, . . . have zeros at z0. mark 2.1. The integrals are well deﬁned because the set of singularities making ∞− ∞ eaningless is zero measurable. Remark 2.1. The integrals are well deﬁned because the set of singularities making ∞− ∞ meaningless is zero measurable. In order to make our statement clear, we ﬁrst recall some knowledge of Green functions. Deﬁnition 2.2. Let D be a domain surrounded by ﬁnitely many piecewise analytic curves. 2. First and Second Fundamental Theorem in Generic Domains Then for any a ∈D, there exists a Green function, denoted by GDz, a, for D with singularity at a ∈D which is uniquely determined by the following: 1 GDz, a is harmonic in D \ {a}; 1 GDz, a is harmonic in D \ {a}; 2 in a neighborhood of a, GDz, a −log \|z −a\| wz, a for some function wz, a harmonic in D; 2 in a neighborhood of a, GDz, a −log \|z −a\| wz, a for some function wz, a harmonic in D; 2 in a neighborhood of a, GDz, a −log \|z −a\| wz, a for some function wz, a harmonic in D; 3 GDz, a ≡0, on the boundary of D. 3 GDz, a ≡0, on the boundary of D. By ∂D we denote the boundary of D and ⃗n the inner normal of ∂D with respect to D. Using Green function we can establish the following general Poisson formula for the E- valued meromorphic mapping, which is similar with 5, Lemma 2.2 see 9, Theorem 2.1, or 10, Theorem 2.1.1. We do not give the details here. Theorem 2.3. Let f : D⊂C →E be an E-valued meromorphic mapping, which does not reduce to the constant zero element 0 ∈E. Then log ∥fz∥ 1 2π ∂D log ∥fζ∥∂GDζ, z ∂⃗n ds − am∈D GDam, z bn∈D GDbn, z −1 2π D GDζ, zΔ log ∥fζ∥dx ∧dy, 2.2 2π ∂D log ∥fζ∥ ζ ∂⃗n ds − am∈D GDam, z bn∈D GDbn, z −1 2π D GDζ, zΔ log ∥fζ∥dx ∧dy, 2.2 2.2 where ζ x iy, {am} are the zeros of fz and {bn} are the poles of fz according to their multiplicities. where ζ x iy, {am} are the zeros of fz and {bn} are the poles of fz according to their multiplicities. Abstract and Applied Analysis 3 3 Remark 2.4. A simple inspection to the C2-valued case shows that log ∥fz∥is not harmonic for a holomorphic or meromorphic E-valued function. Therefore we have an additional term in formula 2.2. Following Theorem 2.3, we introduce some notations. 2. First and Second Fundamental Theorem in Generic Domains N D, a, f bn∈D GDbn, a, m D, a, f 1 2π ∂D log∥fζ∥∂GDζ, a ∂⃗n ds, V D, a, f 1 2π D Gζ, aΔ log ∥fζ∥dx ∧dy, 2.3 2.3 where a is a point in D and {bn} are the poles of fz in D appearing according to their multiplicities, logx log max{x, 1}. Deﬁne where a is a point in D and {bn} are the poles of fz in D appearing according to their multiplicities, logx log max{x, 1}. Deﬁne T D, a, f m D, a, f N D, a, f . 2.4 2.4 TD, a, f is called the Nevanlinna characteristic function of fz with the center a ∈D. Next, we give the ﬁrst FFT and the second SFT fundamental theorems for fz. TD, a, f is called the Nevanlinna characteristic function of fz with the center a ∈D. Next, we give the ﬁrst FFT and the second SFT fundamental theorems for fz. Theorem 2.5 FFT. Let fz be an E-valued meromorphic mapping on D. Then for a ﬁxed vector b ∈E and for any a ∈D such that fa / b, one has Theorem 2.5 FFT. Let fz be an E-valued meromorphic mapping on D. Then for a ﬁxed vector b ∈E and for any a ∈D such that fa / b, one has T D, a, 1 f −b T D, a, f −V D, a, f −b −log ∥fa −b∥ εb, D, 2.5 2.5 where where where \|εb, D\| ≤ ⎧ ⎨ ⎩ log∥b∥ log 2, b / 0, 0, b 0. 2.6 \|εb, D\| ≤ ⎧ ⎨ ⎩ log∥b∥ log 2, b / 0, 0, b 0. 2.6 2.6 \|εb, D\| ≤ ⎨ ⎩ log ∥b∥ log 2, b / 0, 0, b 0. 2.6 \|εb, D\| ≤ ⎨ ⎩0, b 0. 2 oof. We can rewrite Theorem 2.3 as follows: T D, a, f T D, a, 1 f V D, a, f log ∥fa∥. 2.7 2.7 Applying this formula to the function fz −b, we can prove the theorem. Applying this formula to the function fz −b, we can prove the theorem. Abstract and Applied Analysis 4 Theorem 2.6 SFT. Let fz be an E-valued meromorphic mapping on D, let ajj 1, 2, . . 2. First and Second Fundamental Theorem in Generic Domains . , q ∈ E { ∞} be q distinct vectors, and let fa / aj. Then, Theorem 2.6 SFT. Let fz be an E-valued meromorphic mapping on D, let ajj 1, 2, . . . , q ∈ E { ∞} be q distinct vectors, and let fa / aj. Then, q −2 T D, a, f ≤ q j1 N D, a, f aj V D, a, f −aj −V D, a, f′ −N1 D, a, f S D, a, f , 2.8 2.8 where S D, a, f 1 2π ∂D log ⎡ ⎣ q j1 ∥f′ζ∥ ∥fζ −aj∥ ⎤ ⎦∂GDζ, a ∂⃗n ds −log ∥f′a∥ q log 2q δ q i1 log ∥fa −ai∥, N1 D, a, f 2N D, a, f −N D, a, f′ N D, a, 1 f′ , δ min i / j ∥ai −aj∥> 0. 2.9 −log ∥f′a∥ q log 2q δ q i1 log ∥fa −ai∥, 2N D, a, f −N D, a, f′ N D, a, 1 f′ , i j 2.9 2.9 N1 D, a, f 2N D, a, f −N D, a, f′ N D, a, 1 f′ , δ min i / j ∥ai −aj∥> 0. Furthermore, one has the following form: Furthermore, one has the following form: q −2 T D, a, f ≤ q j1 N D, a, 1 f −aj V D, a, f −aj −V D, a, f′ S D, a, f . 2.10 2.10 Proof. Set Fζ q j1 1 ∥fζ −aj∥. 2.11 Proof. Set Fζ q j1 1 ∥fζ −aj∥. 2.11 2.11 According to the property of the logarithm function, we get 1 2π ∂D logFζ∂GDζ, a ∂⃗n ds ≤m D, a, 1 f′ 1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds. 2.12 2.12 Denote δ mini / j∥ai −aj∥, and ﬁx μ ∈{1, 2, . . . , q}. Then we obtain Denote δ mini / j∥ai −aj∥, and ﬁx μ ∈{1, 2, . . . , q}. 2. First and Second Fundamental Theorem in Generic Domains Then we obtain ∥fz −av∥≥∥aμ −av∥−∥aμ −fz∥> 3δ 4 2.13 2.13 Abstract and Applied Analysis for μ / v by Abstract and Applied Analysis Abstract and Applied Analysis 5 for μ / v by 5 for μ / v by ∥fz −aμ∥< δ 2q ≤δ 4 . 2.14 2.14 So either the set of points on ∂D which is determined by 2.14 is empty or any two of some sets for diﬀerent μ have intersection. In any case, on ∂D we have 1 2π ∂D logFζ∂GDζ, a ∂⃗n ds ≥1 2π q μ1 ∥f−aμ∥<δ/2q logFζ∂GDζ, a ∂⃗n ds ≥1 2π q μ1 ∥f−aμ∥<δ/2q log 1 ∥fζ −aμ∥ ∂GDζ, a ∂⃗n ds. 2.15 2.15 μ ≥1 2π q μ1 ∥f−aμ∥<δ/2q log 1 ∥fζ −aμ∥ ∂GDζ, a ∂⃗n ds. 2.15 Since 1 2π ∥f−aμ∥<δ/2q log 1 ∥fζ −aμ∥ ∂GDζ, a ∂⃗n ds m D, a, aμ −1 2π ∥f−aμ∥>δ/2q log 1 ∥fζ −aμ∥ ∂GDζ, a ∂⃗n ds ≥m D, a, aμ −log 2q δ , 2.16 2.16 it follows that 1 2π ∂D logFζ∂GDζ, a ∂⃗n ds ≥ q μ1 m D, a, aμ −q log 2q δ . 2.17 2.17 From 2.12, we get From 2.12, we get From 2.12, we get m D, a, 1 f′ ≥ q μ1 m D, a, aμ −q log 2q δ −1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds. m D, a, 1 f′ ≥ q μ1 m D, a, aμ −q log 2q δ −1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds. 2.18 2.18 μ 2.18 Since fz is nonconstant vector, f′z does not reduce to the constant zero element 0. Applying FFT to f′z, we can obtain Since fz is nonconstant vector, f′z does not reduce to the constant zero element 0. Applying FFT to f′z, we can obtain T D, a, f′ N D, a, 1 f′ m D, a, 1 f′ V D, a, f′ log ∥f′a∥. 2. First and Second Fundamental Theorem in Generic Domains 2.19 2.19 6 Abstract and Applied An 6 Abstract and Applied Analy Abstract and Applied Analysis 6 Abstract and Applied Analysis Using this formula, we have 6 Using this formula, we have Using this formula, we have Using this formula, we have Using this formula, we have T D, a, f′ ≥ q μ1 m D, a, aμ −q log 2q δ −1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds N D, a, 1 f′ V D, a, f′ log ∥f′a∥. 2.20 2.20 On the other hand, we have On the other hand, we have On the other hand, we have T D, a, f′ m D, a, f′ N D, a, f′ ≤m D, a, f N D, a, f′ 1 2π ∂D log ∥f′ζ∥ ∥fζ∥ ∂GDζ, a ∂⃗n ds. 2.21 2.21 e two inequalities above give The two inequalities above give q μ1 m D, a, aμ V D, a, f′ q μ1 m D, a, aμ V D, a, f′ m D, a, aμ V D, a, f′ ≤m D, a, f N D, a, f′ −N D, a, 1 f′ 1 2π logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds 1 2π log ∥f′ζ∥ ∥fζ∥ ∂GDζ, a ∂⃗n ds 2.22 ≤m D, a, f N D, a, f′ −N D, a, 1 f′ 2.22 ≤m D, a, f N D, a, f′ −N D, a, 1 f′ 1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds 1 2π ∂D log ∥f′ζ∥ ∥fζ∥ ∂GDζ, a ∂⃗n ds 2.22 2.22 1 2π ∂D logFζ∥f′ζ∥∂GDζ, a ∂⃗n ds 1 2π ∂D log ∥f′ζ∥ ∥fζ∥ ∂GDζ, a ∂⃗n ds −log ∥f′a∥ q log 2q δ . −log ∥f′a∥ q log 2q δ . o say, That is to say, That is to say, q μ1 m D, a, aμ V D, a, f′ ≤m D, a, f N D, a, f′ −N D, a, 1 f′ S D, a, f . 2. First and Second Fundamental Theorem in Generic Domains 2.23 q μ1 m D, a, aμ V D, a, f′ 2 23 2.23 Adding q μ1 ND, a, f aμ to the above inequality and applying FFT, we can formulate Adding q μ1 ND, a, f aμ to the above inequality and applying FFT, we can formulate q −1 T D, a, f < N D, a, f q j1 N D, a, f aj V D, a, f −aj −N1 D, a, f S D, a, f , 2.24 2.24 bstract and Applied Analysis 7 and Applied Analysis 7 Abstract and Applied Analysis 7 where Abstract and Applied Analysis Abstract and Applied Analysis Abstract and Applied Analysis 7 where ere where where S D, a, f 1 2π ∂D log ⎡ ⎣ q j0 ∥f′ζ∥ ∥fζ −aj∥ ⎤ ⎦∂GDζ, a ∂⃗n ds −log ∥f′a∥ q log 2q δ q i0 log ∥fa −ai∥, a0 0. 2.25 S D, a, f 1 2π ∂D log ⎡ ⎣ q j0 ∥f′ζ∥ ∥fζ −aj∥ ⎤ ⎦∂GDζ, a ∂⃗n ds 2.25 −log ∥f′a∥ q log 2q δ q i0 log ∥fa −ai∥, a0 0. Since ND, a, f ≤TD, a, f, 2.24 can be written as Since ND, a, f ≤TD, a, f, 2.24 can be written as q −2 T D, a, f < q j1 N D, a, f aj V D, a, f −aj −N1 D, a, f S D, a, f . 2.26 2.26 If {aj} contains ∞, 2.26 also holds. Let aq1 ∞, and substitute q with q 1; then we have 2.26, where aq ∞, and S D, a, f 1 2π ∂D log ⎡ ⎣ q−1 j0 ∥f′ζ∥ ∥fζ −aj∥ ⎤ ⎦∂GDζ, a ∂⃗n ds −log ∥f′a∥ q log 2q δ q−1 i0 log ∥fa −ai∥. 2.27 S D, a, f 1 2π ∂D log ⎡ ⎣ q−1 j0 ∥f′ζ∥ ∥fζ −aj∥ ⎤ ⎦∂GDζ, a ∂⃗n ds 2 27 2.27 Next we establish Hiong King-Lai’s inequality for fz. Next we establish Hiong King-Lai’s inequality for fz. Theorem 2.7. 2. First and Second Fundamental Theorem in Generic Domains Let fz be an E-valued meromorphic mapping on D, l ∈D, let a, b, c ∈E be three ﬁnite vectors, and let b / 0, c / 0, b / c, fkl / 0, b, c. Then one has T D, l, f < N D, l, f a N D, l, fk b N D, l, fk c V D, l, fk V D, l, fk −b V D, l, fk −c −N D, l, 1 fk1 S D, l, fk . 2 28 2.28 2.28 2.28 2.28 Proof. First, we have First, we have Proof. First, we have Proof. First, we have Proof. First, we have 1 2π ∂D log 1 fζ −a ∂GDζ, l ∂⃗n ds ≤1 2π ∂D log 1 fkζ ∂GDζ, l ∂⃗n ds 1 2π ∂D log fkζ fζ −a ∂GDζ, l ∂⃗n ds. 2.29 2.29 8 Abstract and Applied Analy Abstract and Applied Analysis Abstract and Applied Analysis 8 Applying FFT to fz and fkz, respectively, we have Applying FFT to fz and fkz, respectively, we have Applying FFT to fz and fkz, respectively, we have pplying FFT to fz and fkz, respectively, we have 1 2π ∂D log 1 fζ −a ∂GDζ, l ∂⃗n ds T D, l, f −N D, l, f a −V D, l, f −a −log ∥fl −a∥ εa, D, 1 2π ∂D log 1 fkζ ∂GDζ, l ∂⃗n ds T D, l, fk −N D, l, 1 fk −V D, l, fk −log ∥fkl∥. 2.30 2.30 Thus we have Thus we have T D, l, f ≤T D, l, fk N D, l, f a V D, l, f −a −N D, l, 1 fk −V D, l, fk log ∥fl −a∥ ∥fkl∥−εa, D. 2. First and Second Fundamental Theorem in Generic Domains 2.31 2.31 Applying SFT to fk with 0, b, c, we have Applying SFT to fk with 0, b, c, we have Applying SFT to fk with 0, b, c, we have T D, l, fk ≤N D, l, 1 fk N D, l, fk b N D, l, fk c −N D, l, fk1 V D, l, fk V D, l, fk −b V D, l, fk −c −V D, l, fk1 S D, l, fk . 2 32 S D, l, f . 2.32 2.32 2.32 Combining 2.31 with 2.32, we have Combining 2.31 with 2.32, we have mbining 2.31 with 2.32, we have T D, l, f ≤N D, l, f a N D, l, fk b N D, l, fk c −N D, l, fk1 V D, l, f −a V D, l, fk −b V D, l, fk −c −V D, l, fk1 S D, l, fk . 2.33 9 Abstract and Applied Analysis 3. The Vector-Valued Mapping and Its Derivative In this section, we will discuss the value distribution theory of fz deﬁned in the disk Cr {z : \|z\| < r}. We will prove Chuang’s inequality. According to 2.3, we have the following terms: m r, f 1 2π 2π 0 log∥f reiθ ∥dθ, N r, f r 0 n t, f −n 0, f t dt n 0, f log r, V r, f 1 2π Cr log r ζ Δ log ∥fζ∥dx ∧dy, ζ x iy, T r, f m r, f N r, f , 3.1 N r, f r 0 n t, f −n 0, f t dt n 0, f log r, 3.1 3.1 V r, f 1 2π Cr log r ζ Δ log ∥fζ∥dx ∧dy, ζ x iy, T r, f m r, f N r, f , T r, f m r, f N r, f , where nr, f denotes the number of poles of fz in {z : \|z\| < r}. The order and the lower order of an E-valued meromorphic mapping fz are deﬁned by λ f lim sup r →∞ log T r, f log r , μ f lim inf r →∞ log T r, f log r . 3.2 λ f lim sup r →∞ log T r, f log r , 3.2 μ f lim inf r →∞ log T r, f log r . The following lemma is well known. The following lemma is well known. Lemma 3.1 see 11, Boutroux-Cartan Theorem. Let {aj}n j1 be n complex numbers. Then the set of the point z satisfying n j1 z −aj < hn 3.3 n j1 z −aj < hn 3.3 3.3 can be contained in several disks, denoted by γ; the total sum of its radius does not exceed 2eh. can be contained in several disks, denoted by γ; the total sum of its radius does not exceed 2eh. The next lemma is a special case of Theorem 2.3. The next lemma is a special case of Theorem 2.3. The next lemma is a special case of Theorem 2.3. 10 Abstract and Applied Analysis Abstract and Applied Analysis Lemma 3.2 see 5. 3. The Vector-Valued Mapping and Its Derivative Let f : Cr →E be an E-valued meromorphic mapping, which does not reduce to the constant zero element 0 ∈E. Then, for a z ∈Cr, one has Lemma 3.2 see 5. Let f : Cr →E be an E-valued meromorphic mapping, which does not reduce to the constant zero element 0 ∈E. Then, for a z ∈Cr, one has log ∥fz∥ 1 2π 2π 0 log ∥f reiθ ∥ r2 −t2 r2 −2rt cos θ −φ t2 dφ − zj0∈Cr log r2 −zj0z r z −zj0 zj ∞∈Cr log r2 −zj ∞z r z −zj ∞ −1 2π Cr log r2 −ξz rz −ξ Δ log fξ dx ∧dy. 3.4 3.4 −1 2π Cr log r2 −ξz rz −ξ Δ log fξ dx ∧dy. Here zj0 and zj ∞ are all the zeros and poles counting their multiplies of f in D. Here zj0 and zj ∞ are all the zeros and poles counting their multiplies of f in D. In order to obtain the relationship between Tr, f and Tr, f′, we should ﬁrst establish the following two lemmas. Lemma 3.3. Let f : C →E be a nonzero E-valued meromorphic mapping, and f0 / ∞. If R and R′ are two positive numbers, and R < R′, then there exists a θ0 ∈0, 2π, such that for any 0 ≤r ≤R one has log∥f reiθ0 ∥≤R′ R R′ −Rm R′, f n R′, f log 4 N R′, f . 3.5 3.5 For z reiθ, 0 ≤r ≤R. By Lemma 3.2 we have Proof. For z reiθ, 0 ≤r ≤R. By Lemma 3.2 we have log ∥fz∥≤1 2π 2π 0 log ∥f R′eiθ ∥ R ′2 −r2 R ′2 −2R′r cos θ −φ r2 dφ n j1 log R ′2 −bjz R′ z −bj , 3.6 3.6 where {bj}n j1 are the poles of fz in \|z\| ≤R′. Then where {bj}n j1 are the poles of fz in \|z\| ≤R′. Then where {bj}n j1 are the poles of fz in \|z\| ≤R′. Then log∥fz∥≤R′ r R′ −r m R′, f n j1 log 2R′ z −bj ≤R′ r R′ −r m R′, f log 2R′n n j1 z −bj . 3. The Vector-Valued Mapping and Its Derivative 3.7 3.7 ≤R′ r R′ −r m R′, f log 2R′n n j1 z −bj . 3.7 Writing bj \|bj\|eiφj, we have Writing bj \|bj\|eiφj, we have Writing bj \|bj\|eiφj, we have Writing bj \|bj\|eiφj, we have ng bj \|bj\|eiφj, we have reiθ− bj eiφj ≥ bj∥sin θ −φj . 3.8 3.8 Abstract and Applied Analysis 11 Thus Thus n j1 z −bj ≥ ⎛ ⎝ n j1 bj ⎞ ⎠ ⎛ ⎝ n j1 sin θ −φj ⎞ ⎠. 3.9 3.9 π 0 log n j1 sin θ −φj \| dθ n π 0 log\|sin θ\|dθ −nπ log 2. 3.10 3.10 Hence there exists a real number θ0 such that Hence there exists a real number θ0 such that Hence there exists a real number θ0 such that n j1 sin θ0 −φj > 1 2n . 3.11 3.11 Combining 3.7 and 3.9 with 3.11, we have Combining 3.7 and 3.9 with 3.11, we have Combining 3.7 and 3.9 with 3.11, we have log∥f reiθ0 ∥≤R′ R R′ −Rm R′, f n log 4 n j1 log R′ bj ≤R′ R R′ −Rm R′, f n log 4 N R′, f . 3.12 log∥f reiθ0 ∥≤R′ R R′ −Rm R′, f n log 4 n j1 log R′ bj 3.12 ≤R′ R R′ −Rm R′, f n log 4 N R′, f . Lemma 3.4. Let f : C →E be a nonzero E-valued meromorphic mapping, and let R < R′ < R′′ be three positive numbers. Then there exists a positive number R ≤ρ ≤R′, and for \|z\| ρ, one has Lemma 3.4. Let f : C →E be a nonzero E-valued meromorphic mapping, and let R < R′ < R′′ be three positive numbers. Then there exists a positive number R ≤ρ ≤R′, and for \|z\| ρ, one has log∥fz∥≤R′′ R′ R′′ −R′ m R′′, f n R′′, f log 8eR′′ R′ −R. 3.13 3.13 Proof. Let {bj}n j1 be the poles of fz in \|z\| ≤R′′. By Boutroux-Cartan Theorem, we have Proof. Let {bj}n j1 be the poles of fz in \|z\| ≤R′′. 3. The Vector-Valued Mapping and Its Derivative By Boutroux-Cartan Theorem, we have n j1 z −bj ≥ R′ −R 4e n , 3.14 3.14 except for some points contained in a pack of disks whose radius does not exceed R′ −R/2. Then there exists a circle {z : \|z\| ρ} such that R ≤ρ ≤R′ and {\|z\| ρ} $γ ∅. except for some points contained in a pack of disks whose radius does not exceed R′ −R/2. Then there exists a circle {z : \|z\| ρ} such that R ≤ρ ≤R′ and {\|z\| ρ} $γ ∅. 12 Abstract and Applied Analy Abstract and Applied Analysis 12 Thus 3.14 holds on {\|z\| ρ}. For any z ∈{z : \|z\| ρ}, we have Thus 3.14 holds on {\|z\| ρ}. For any z ∈{z : \|z\| ρ}, we have us 3.14 holds on {\|z\| ρ}. For any z ∈{z : \|z\| ρ}, we have log∥f reiθ0 ∥≤R′′ ρ R′′ −ρm R′′, f n j1 log R′′2 −bjz R′′ z −bj ≤R′′ R′ R′′ −R′ m R′′, f n log 8eR′′ R′ −R. 3.15 3.15 Now we are in the position to establish the following Chuang’s inequality. Now we are in the position to establish the following Chuang’s inequality. Theorem 3.5. Let f : C →E be a nonzero E-valued meromorphic mapping and f0 / ∞. Then for τ > 1 and 0 < r < R, one has Theorem 3.5. Let f : C →E be a nonzero E-valued meromorphic mapping and f0 / ∞. Then for τ > 1 and 0 < r < R, one has T r, f < CτT τr, f′ logτr 4 log∥f0∥, 3.16 3.16 ere Cτ is a positive constant. where Cτ is a positive constant. Proof. Take a σ such that σ3 τ and denote r1 σr, r2 σr1, r3 σr2. Applying Lemma 3.3 to f′z, we can ﬁnd a real number θ0 such that 0 ≤t ≤r1, and we have log∥f′ teiθ0 ∥≤r2 r1 r2 −r1 m r2, f′ n r2, f′ log 4 N r2, f′ . 3. The Vector-Valued Mapping and Its Derivative 3.17 3.17 In view of Lemma 3.4, for a ﬁxed ρ ∈r, r1 we have log∥f′z∥≤r2 r1 r2 −r1 m r2, f′ n r2, f′ log 8er2 r1 −r , 3.18 3.18 on {z : \|z\| ρ}. on {z : \|z\| ρ}. From the origin along the segment arg z θ0 to ρeiθ0, and along {z : \|z\| ρ} turn a rotation to ρeiθ0. We denote this curve by L, and its length is 2π 1ρ. on {z : \|z\| ρ}. From the origin along the segment arg z θ0 to ρeiθ0, and along {z : \|z\| ρ} turn a rotation to ρeiθ0. We denote this curve by L, and its length is 2π 1ρ. We notice that ϕz ∥fz∥is continuous on L. As in 5, En is an n-dimensional projective space of E with a basis {ei}n i1. The projection operator Pn : E →En is a realization of En associated to the basis and Pnfz f1z, f2z, . . . , fnz. We have Pnf′z Pnfz′ n i1 f′ izei and Pnfz Pnf0 n i1 %z 0 f′ iζdζei. Therefore, since En is We notice that ϕz ∥fz∥is continuous on L. As in 5, En is an n-dimensional projective space of E with a basis {ei}n i1. The projection operator Pn : E →En is a realization of En associated to the basis and Pnfz f1z, f2z, . . . , fnz. We have Pnf′z Pnfz′ n i1 f′ izei and Pnfz Pnf0 n i1 %z 0 f′ iζdζei. Therefore, since En is Abstract and Applied Analysis 13 ﬁnite dimensional, there exists K > 0 appearing in the inequality ∥· ∥1 ≤K∥· ∥2, where ∥· ∥1 and ∥· ∥2 are any two norms on En such that ∥Pn fz ∥≤∥Pn f0 ∥ n i1 z 0 f′ iξdξ ei ≤∥Pn f0 ∥ 1 K n i1 z 0 f′ iξdξ 21/2 ≤∥Pn f0 ∥ 1 K n i1 max ξϵL f′ iξ 2 1/2 2π 1ρ ≤∥Pn f0 ∥ K′ K Mn2π 1ρ, 3.19 3.19 ≤∥Pn f0 ∥ K′ K Mn2π 1ρ, where Mn maxz∈L∥Pnf′z∥. Thus, we have where Mn maxz∈L∥Pnf′z∥. Thus, we have where Mn maxz∈L∥Pnf′z∥. 3. The Vector-Valued Mapping and Its Derivative Thus, we have ∥Pn fz ∥≤∥Pn f0 ∥ Mn2π 1ρ O1, \|z\| ρ. 3.20 3.20 In virtue of 6–8, every meromorphic mapping fz with values in a Banach space E with a Schauder basis and the projections Pnf are convergent in its natural topology; that is, they converge uniformly to f in any compact subset W of C \ Pf Pf being the set of poles the f in C. Thus for n large enough, we have ∥Pn fz ∥ ∥fz∥ O1, for any z ∈W ⊆C \ Pf. 3.21 3.21 A similar argument to f′z implies that for n large enough A similar argument to f′z implies that for n large enough ∥Pn f′z ∥ ∥f′z∥ O1, Mn ≤M O1 for any z ∈W ⊆C \ Pf′, 3.22 3.22 where M maxz∈L∥f′z∥. Combining 3.20, 3.21, and 3.22 and the fact that the compact set {z : \|z\| ρ} ⊆ L ⊆C \ Pf, we get where M maxz∈L∥f′z∥. Combining 3.20, 3.21, and 3.22 and the fact that the compact set {z : \|z\| ρ} ⊆ L ⊆C \ Pf, we get ∥fz∥≤∥f0∥ M2π 1ρ O1. 3.23 log∥fz∥≤log∥f0∥ logM logρ log 8eπ O1. 3.24 14 Abstract and Applied Analy Abstract and Applied Analysis Abstract and Applied Analysis 14 tue of 3.13 and 3.17, we have In virtue of 3.13 and 3.17, we have In virtue of 3.13 and 3.17, we have In virtue of 3.13 and 3.17, we have logM ≤r2 r1 r2 −r1 m r2, f′ n r2, f′ log 8er2 r1 −r N r2, f′ ≤ &log8er2/r1 −r logr3/r2 r2 r1 r2 −r1 ' T r3, f′ C′ τT r3, f′ . 3.25 logM ≤r2 r1 r2 −r1 m r2, f′ n r2, f′ log 8er2 r1 −r N r2, f′ 3.25 Therefore, Therefore, Therefore, m ρ, f < C′ τT r3, f′ logτr 4 log∥f0∥. 3.26 m ρ, f < C′ τT r3, f′ logτr 4 log∥f0∥. 3.26 3.26 Thus we have T r, f ≤T ρ, f < C′ τ 1 T r3, f′ logτr 4 log∥f0∥ Thus we have Thus we have Thus we have Thus we have T r, f ≤T ρ, f < C′ τ 1 T r3, f′ logτr 4 log∥f0∥ CτT τr, f′ logτr 4 log∥f0∥. 3.27 3.27 3.27 The following result says that we can also control the Tr, f′ by Tr, f. Theorem 3.6. Let fz(z ∈C) be a nonconstant E-valued meromorphic mapping. Then one has T r, f′ ≤2T r, f O log r logT r, f . 3.28 The following result says that we can also control the Tr, f′ by Tr, f. Theorem 3.6. Let fz(z ∈C) be a nonconstant E-valued meromorphic mapping. A similar argument to f′z implies that for n large enough Then one has T r, f′ ≤2T r, f O log r logT r, f . 3.28 3.28 Proof. One has Proof. One has T r, f′ m r, f′ N r, f′ ≤m r, f N r, f′ 1 2π 2π 0 log ∥f′ reiφ ∥ ∥f reiφ ∥dφ m r, f N r, f N r, f 1 2π 2π 0 log ∥f′ reiφ ∥ ∥f reiφ ∥dφ ≤2T r, f 1 2π 2π 0 log ∥f′ reiφ ∥ ∥f reiφ ∥dφ 2T r, f O log r logT r, f . 3.29 T r, f′ m r, f′ N r, f′ m r, f N r, f N r, f 1 2π 2π 0 log ∥f′ reiφ ∥ ∥f reiφ ∥dφ 3.29 3.29 ≤2T r, f 1 2π 2π 0 log ∥f′ reiφ ∥ ∥f reiφ ∥dφ 2T r, f O log r logT r, f . From Theorems 3.5 and 3.6, we have the following. From Theorems 3.5 and 3.6, we have the following. From Theorems 3.5 and 3.6, we have the following. Corollary 3.7. For a nonconstant E-valued meromorphic mapping fzz ∈C, One has λf λf′, μf μf′. 15 Abstract and Applied Analysis Abstract and Applied Analysis Acknowledgments The authors would like to thank the referee for his/her many helpful comments and suggestions on an early version of the manuscript. The work is supported by NSF of China Grant no. 10871108. The ﬁrst author is also supported partially by NSF of China Grant no. 10926049. References 1 H. Ziegler, Vector Valued Nevanlinna Theory, vol. 73 of Research Notes in Mathematics, Pitman, Boston, Mass, USA, 1982. 2 I. Lahiri, “Milloux theorem and deﬁciency of vector-valued meromorphic functions,” The Journal of the Indian Mathematical Society, vol. 55, no. 1–4, pp. 235–250, 1990. y pp 3 I. Lahiri, “Generalisation of an inequality of C. T. Chuang to vector meromorphic functions,” Bulle of the Australian Mathematical Society, vol. 46, no. 2, pp. 317–333, 1992. f y pp 4 I. Lahiri, “Milloux theorem, deﬁciency and ﬁx-points for vector-valued meromorphic functions,” The Journal of the Indian Mathematical Society, vol. 59, no. 1–4, pp. 45–60, 1993. f y pp 5 C.-G. Hu and Q. Hu, “The Nevanlinna’s theorems for a class,” Complex Variables and Elliptic Equations, vol. 51, no. 8–11, pp. 777–791, 2006. 6 W. Arendt and N. Nikolski, “Vector-valued holomorphic functions revisited,” Mathematische Zeitschrift, vol. 234, no. 4, pp. 777–805, 2000. f pp 7 K.-G. Grosse-Erdmann, “The locally convex topology on the space of meromorphic function Australian Mathematical Society A, vol. 59, no. 3, pp. 287–303, 1995. y pp 8 E. Jord, “Topologies on spaces of vector-valued meromorphic functions,” Journal of the Australian Mathematical Society, vol. 78, no. 2, pp. 273–290, 2005. y pp 9 A. A. Goldberg and I. V. Ostrovskii, Value Distribution of Meromorphic Functions, vol. 236 of Translatio of Mathematical Monographs, American Mathematical Society, Providence, RI, USA, 2008. f g p y 10 J. Zheng, Value Distribution of Meromorphic Functions, Tsinghua University Press, Beijing, China, 20 f g p y 10 J. Zheng, Value Distribution of Meromorphic Functions, Tsinghua University Press, Beijing, China, 2010. 11 L. Yang, Value Distribution And New Research, Springer, Berlin, Germany, 1993. 11 L. Yang, Value Distribution And New Research, Springer, Berlin, Germany, 1993. 11 L. Yang, Value Distribution And New Research, Springer, Berlin, Germany, 1993.
https://openalex.org/W2330087126	https://zenodo.org/records/1968861/files/article.pdf	English	null	CHEMICAL METHODS FOR ASCERTAINING THE LIME REQUIREMENT OF SOILS.	Journal of the American Chemical Society	1,900	public-domain	11,331	By H. J. WHEELER, B. L. HARTWELL, AND C. L. SARGENT. Received February 20. xgoa. By H. J. WHEELER, B. L. HARTWELL, AND C. L. SARGENT. Received February 20. xgoa. H AVING recognized the occurrence of an injurious degree of H acidity in the upland soils of Rhode Island and other states, even upon steep hillsides, it became important to ascertain by chemical means if possible the relative lime requirement of different sails. If such could be done in a simple and effective way, then one would be in a position to recommend satisfactorily what amount of lime to use per acre without resort to field tests. 1 Anleitung zur wissenschaftlichen Bodenuntersuchungen, Berlin (1887), p. 56. 2 ” Treatise on Chemistry,” I, h’ew York (1888). p. 638. 1 Anleitung zur wissenschaftlichen Bodenuntersuchungen, Berlin (1887), p. 56. 2 ” Treatise on Chemistry,” I, h’ew York (1888). p. 638. 8 Vol. I, London (1890). p. 6gj. 2 ” Treatise on Chemistry,” I, h’ew York (1888). p. 638. 8 Vol. I, London (1890). p. 6gj. 1 Anleitung zur wissenschaftlichen Bodenuntersuchungen 2 ” Treatise on Chemistry,” I, h’ew York (1888). p. 638. TESTS BY BLUE LITMUS PAPER. Perhaps the simplest, though nevertheless a very satisfactory test, is that made by bringing a piece of blue litmus paper in contact with the moistened soil and observing the intensity and rapidity of the reddening thus produced. Though unsatisfactory, in view of its not being a quantitative test, much valuable infor- mation regarding the amounts of lime which should probably be used, may thus be obtained, after some experience, by one who is a careful observer. Wahnschaffe’ in speaking of this reaction says, “ since the free carbonic acid of the soil reddens the paper, one must allow it to dry and observe whether or not the red color is still visible.” Roscoe and Schorlemmer* state that “ carbonic acid readily decomposes into water and carbon dioxide. In consequence of this, litmus paper which has been turned red in the aqueous acid, becomes blue on drying.’’ They then immediately add that the reddening of litmus solution may be brought about by conducting into it carbon dioxide, but that upon heating the solution the carbon dioxide is expelled, and after boiling a few seconds the liquid again becomes blue. In Watts’ Dictionary of Chemistry3 it is stated that an ‘‘ aqueous solution of CO, turns blue litmus paper wine-red ; the blue color returns on exposure to air.” It seems probable that on account of the well-known reaction of litmus solution whereby, when reddened by carbon dioxide gas, it returns to a blue color, it has 154 H. J. WHEELER, B. I,. HARTWELL, AND C. L. SARGENT. 154 been assumed that a similar reaction would necessarily take place when operating with litmus paper. This conclusion seems exceedingly probable since the writers have been unable to find references in the literature to any definite tests bearing upon this point. I t is possible also that tests, if made, were carried out by operating with carbonic acid upon small sections of a piece of litmus paper, in which case, sufficient alkali might diffuse from the surrounding paper into the section acted upon, to neutralize the carbonic acid still remaining. That such a return of the blue color may result when the entire portion of the litmus paper is not subjected to treatment with acid, has been demonstrated by us upon commercial paper in the case of carbonic acid and of very dilute citric acid. 2 The third method. Sutton's ' * Handbook of Volumetric Analysis", London (ISSZ), p. jl. 1 Bockmaun's '' Chermsch-techuische ~-iitersuchungsmethodeu ', Vol. I, Berlin (1888). p . g i TESTS BY BLUE LITMUS PAPER. Owing to the action of the fingers upon blue litmus paper and the bleaching action of the sun, the test would be far simpler and more reliable in the hands of practical farmers if it were not necessary to wait for the paper to dry before drawing conclusions as to the probable needs of the soil. In view of the fact that some of the conditions prevailing when working with litmus paper are so different from those existing when employing the solution, it seemed probable that blue litmus paper, if entirely reddened by carbonic acid, and dried with the promptness that would ordinarily be expected, would not return to its original blue color. Dr. H. E. Stockbridge told one of us that he was under the impression that he once found this to be the case. In order to test this matter thoroughly, trials were made with ordinary commercial blue litmus paper and also with lots of paper prepared from solutions made according to Classen' and Sutton.' The carbonic acid was prepared by expelling the carbon dioxide from an aqueous solution of chetnicallv pure sodium bicarbonate by simply boiling without the use of acid. A return condenser was employed and the carbon dioxide was washed in distilled water before its collection. Steam, generated from the distilled water which was used subsequently to dissolve the sodium bicarbonate, was first passed through the entire apparatus for some time and the distillate tested to insure the absence of 1 Bockmaun's '' Chermsch-techuische ~-iitersuchungsmethodeu ', Vol. I, Berlin (1888). p . g i 1 Bockmaun's '' Chermsch-techuische ~-iitersuchungsmethodeu ', Vol. I, Berlin (1888). p . g i L I N E REQUIREMENT O F SOILS. 155 volatile organic acids. Treated with carbonic acid prepared in this way all three lots of blue litmus paper were permanently reddened when dried naturally in the open air or in a room free from acids and ammonia. It is evident therefore that when dried naturally the carbon dioxide is not lost to any great extent or else the red, acid coloring-matter of the litmus is too firmly fixed upon the fiber of the paper, to permit of its reunion with the base, a change which would have resulted in the case of a solution, after the expulsion of the carbon dioxide. TESTS BY BLUE LITMUS PAPER. Coupling with these results the fact that so long as a liberal amount of calcium car- bonate is present in a soil, the presence of carbonic acid tends to increase rather than lessen the tendency toward an alkaline reaction’ (a point proved by us experimentally in the laboratory) it must be evident that the indication of acidity afforded by a piece of blue litmus paper is sufficient to show a deficiency of calcium carbonate or other basic ingredients in soils, without first allowing the paper to dry, as suggested by Wahnschaffe. 1 Compare Detmer Die naturwissenschaftlichen Grundlagen d. allg. landw. Boden- E l di Chi i 4 P 4 kuude (1876), p. 4jI; also Stockhardt, Jahrbuch fur Agriculturchemie (1866). p. IO. 2 Encyclopedie Chimique, 4, Pt. 4, p. 18. 8 Bulletin No. 56, Chemical Division, United States Department of Agriculture, Wash- ington, D. C., 1899. 2 Encyclopedie Chimique, 4, Pt. 4, p. 18. g TESTS BY EXTRACTION WITH AMMONIA. Miintz‘ states that it is easy to distinguish whether or not a soil is acid, since if such is the case it imparts a black color to ammonium hydroxide when the two are brought in contact. He adds that in the case of soils which are not acid, the organic matter is in an insoluble combination with lime and consequently yields nothing to ammonium hydroxide. The same has been found by us to be the case with a sample of fertile adobe soil kindly forwarded by Professor GOSS, from New Mexico. Similar observations have also been made on one Massachusetts and two Rhode Island soils where wood ashes or calcium carbonate in some form had been applied to the land with such frequency as to entirely overcome its acid reaction upon blue litmus paper, Snyder,3 however, calls attention to a soil which he says con- tained 3.15 per cent. of lime (CaO) and 4.19 per cent. of carbon dioxide, and yet when treated with ammonium hydroxide for the determination of the so-called ‘ I free humic acid ” yielded a dark brown extract amounting to from 0.3 to 0.4 of a H . J. WHEELER, B. L. HARTWELL, AND C. L. SARGENT. 156 per cent. This would tend to throw some doubt upon the reliability of the treatment with ammonium hydroxide as a means of arriving at the lime requirements of all soils, though it is not improbable that for soils of certain sections of the United States which are possessed of definite characteristics, the treatment may be of considerable value as was the case in the particular acid soils, for which Muntz employed the method. The details of the method as proposed by hIuntz’ are as follows : To 20 grams of powdered soil, add jocc. of concentrated aminoniuui hydroxide. Allow it to stand for from two to three days in a warn1 place. Agitate frequently and at the end of the period make the volume up to I liter. It is then shaken and allowed to subside for twenty-four hours. To joo cc. of the liquid add sufficient hydrochloric acid to give a strong acid reaction in order to pre- cipitate the humus, which carries with it some mineral matter. TESTS BY EXTRACTION WITH AMMONIA. After the precipitate has subsided, decant the supernatant liquid, collect on a filter, dry and weigh in a weighing bottle, incinerate, deduct the ash, and the “free humic acids ’’ are thus obtained by difference. In order to determine what influence, if any, the use of concentrated ammonium hydroxide might have upon the results compared with dilute solutions, a sample of upland soil was tested by the above method and yielded 1.00 per cent. Treated for the same time with the same amount of ammonium hydroxide previously diluted to joo cc. (containing I . 13 per cent. of ammonia) it also yielded 1.00 per cent. The same soil treated for a like time with joo cc. of the same solvent containing 2.4 per cent. of ammonia yielded 1.00 per cent. Employing sufficient concentrated ammonium hydroxide so that after diluting to joo cc. it would have the same strength as in the last- mentioned instance, it yielded, treated for the same time, 1.03 per cent. It appears therefore that in the case of ordinary soils the strength of the solvent within reasonable limits is not an essential feature, and that the solution usually employed in total hunius determinations may be used here with equal propriety. T H E METHOD O F HOLLELIIIAS FOR DETERXINING THE LIBIE KEQUIREXENTS O F STIFF CLAY SOILS. T H E METHOD O F HOLLELIIIAS FOR DETERXINING THE LIBIE KEQUIREXENTS O F STIFF CLAY SOILS. Holleman2 in searching for the cause of the occasional 1 Encyclop6die Chimique, 4, Pt 4, p ISZ 2 Die latldw Versuchs-stationen 41 (I892), p. 38 T H E METHOD O F HOLLELIIIAS FOR DETERXINING THE LIBIE KEQUIREXENTS O F STIFF CLAY SOILS. Holleman2 in searching for the cause of the occasional LIME REQUIREMENT OF SOILS. I57 ineffectiveness of gas-lime upon certain stiff clays in Holland concluded that it might be due to the presence in them of con- siderable calcium carbonate. H e thereupon made determinations of the lime removable by water and by water saturated with carbon dioxide. The water-soluble lime was determined as follows: Twenty grams of air-dried soil were digested with frequent shaking, at ordinary temperature, for forty-eight hours with I liter of distilled water. TESTS BY EXTRACTION WITH AMMONIA. Five hundred grams of the filtrate to which a few drops of acetic acid had been added were evaporated, the residue taken up in water; and first a little acetic acid and then a measured quantity of FT solution of ammonium oxalate were added and the volume made up to IOO cc. After filtration, the excess of oxalic acid is estimated by titrating with a FT solution of potassium permanganate. For determining the lime soluble in carbonic acid 20 grams of air-dried soil are taken and sufficient distilled water added to make up I liter ; it is then saturated with carbon dioxide and after twenty-four hours filtered and treated as described above. From the data obtained by Holleman, he concluded that where extraction with carbonic acid removed 0.15 per cent., or less, of calcium oxide, stiff clay soils would be benefited by liming. y g In order to test the applicability of this method to the acid sandy and clayey loam soils of our State, two samples were selected, the first being from a farm in Warren, R. I., where, with the same manurial and other treatment, the yield of beets upon the unlimed plot was 272 pounds against 329 pounds upon the limed plot. The second soil was from Foster Center, R. I., where under the same conditions the yield of beets upon the unlinied plot had been 36.6 pounds and upon the limedone 143.4 pounds, In making the test of the soil from Warren, 80 grams of soil were employed, and in that from Foster Center, 40 grams. The soil was placed in each case in a flask to which I liter of distilled water, saturated with carbon dioxide, was added. The flask was agitated frequently and the treatment continued for forty hours. Four hundred cc. of the extract were evaporated to dry- ness, taken up in water and a little acetic acid, :* ammonium oxalate was added, and after filtration, the whole was made up to 200 cc. Aliquot parts of this solution were titrated with & potas- sium permanganate, after acidulation with sulphuric acid. In 8 H. J. WHEELER, B . I,. HARTWELL, AND c. L. SARGENT. 158 one instance the small amount of acetic acid was removed by evaporation to dryness before adding sulphuric acid and titra- ting without practically affecting the results. TESTS BY EXTRACTION WITH AMMONIA. Allowing for the blank found, the amount of calciurri oxide thus determined amounted in the Foster Center soil to 0.0360 Fer cent. In the case of the Warren soil, which was treated in the same manner, the amount of calcium oxide removed from the soil was found to be 0.0287. The Warren soil extracted with distilled water, in a similar manner for forty-eight hours, yielded 0.0120 per cent. of calcium oxide. From these tests it will be seen that the soil which was least benefited by lime, yielded less lime than the other. It is possible that the use of double the amount of soil in the last instance may have depressed the result somewhat as compared with the other, though it is not likely that the difference caused by this factor would, under the circumstances, have been readily appreciable. At all events the amount of lime removed in the highest case was only about one-fifth of that below which Holleman found lime useful on stiff clays. It appears, therefore, in view of the marked degree of acidity of those of our soils which it seemed desirable to test as to their lime requirements, that this method is probably not of great value. It is regretted, nevertheless, that the pressure of routine work has prevented making further tests of the applicability of the method to our special needs, particularly with greater attention to maintaining the saturation of the carbonated water. 1 Chem Zfg S o . 20, 1Sy7, p 174 T H E METHOD O F TACKE FOR DETERNIKIXG T H E F R E E HUMUS ACIDS I N THE PEAT (MOORBODEN) SOILS O F NORTH GERRIANT. Tacke' states that no method is known for determining quantitatively the acidity of soils, and calls attention to the desirability of possessing a satisfactorily exact method for the purpose. H e discusses the difficulties in the way of determining the acidity by various ways which would naturally suggest themselves, and finally gives a method which yielded good results. It is impossible to give the method satisfactorily in LIME REQUIREMENT OF SOILS. I59 detail without reproducing the illustration of the apparatus which he employed. Essentially the method consists in expel- ling the carbon dioxide from calcium carbonate brought in con- tact with the soil at ordinary room temperature. Hydrogen is passed over the soil for some time in the apparatus, then finely divided calcium carbonate, shaken up with water, is introduced and allowed to act upon the soil for two and a half hours or longer as desired. The carbon dioxide disengaged is taken as a measure of the acidity of the soil, the amount being determined by passing it through a Pettenkofer tube containing 4 or & sodium hydroxide. The liquid is finally removed from the tube, taking care in regard to the absorption of carbon dioxide from the air, barium chloride then added and the solution titrated with or Tn hydrochloric acid using phenolphthalein as an indicator with attention to the points brought out by the investi- gations by Kiister’ and by Lunge.* Determined in this way, using soil in its natural state, the amount of acid reckoned as carbonic acid, represented by IOO parts of dry material, amounted in the case of various peat soils to from 1 . 7 2 1 to 2.240 per cent. The method gives promise of much value for such soils as a guide to the amount of lime which they should receive. In April, 1895, experiments were begun by one of us for the pur- pose of ascertaining if the amount of carbon dioxide liberated from calcium carbonate upon heating it with upland acid soils in the presence of water, could be used as a guide to the amount of lime to apply to them. Owing to the pressure of routine matters, this work was several times taken up and then soon discontinued. Not having satisfactorily perfected a method at the time Tacke’s results were published. it was decided to see if his method was applicable to our upland soils. 1 Chem. Ztg. Repevt, 29, 282 (1896) ; Zfschr. anovg. Chem., 13, 127 pp. (1896). 2 Bestimmung von Caustischen und Kohlensauren Alkalien, G. Lunge, Ztschv. angezu. Chem.. 41 (1897). 1 Wilej s Priiiclpies arid Practice of Agricultural htialysis, To1 I 52- j 2 9 p 325 T H E METHOD O F TACKE FOR DETERNIKIXG T H E F R E E HUMUS ACIDS I N THE PEAT (MOORBODEN) SOILS O F NORTH GERRIANT. Two unlimed soils were selected for the trial, the one from land where sodium nitrate had been applied for several years and the other from land where the manuring had been identical except that the same quantity of nitrogen had been applied in ammonium sulphate. The yields of most crops had been for several years far better upon the former 160 60 H. J. WHEELER, B. L. HARTWELL, A N D c. L. SARGENT. The percentage of carbon dioxide obtained by treat- ment for three hours, calculated to dry soil, was 0.039 for the soil which had received sodium nitrate and 0.067 for the one which had received ammonium sulphate. 111 other words the results stood in the same relation as the yields. Taking into account the fact that the soils contained nearly -1. per cent. of humus and from 1 . 2 j to I . j j per cent. of humus immediately soluble in ammonium hydroxide, it is evident that the action of this class of soils upon calcium carbonate, at ordinary tempera- ture, is far less vigorous than that of the peat soils tested by Tacke. The percentages obtained by this method are so small, and the element of error proportionately so great, that the indi- cations are not particularly favorable to its use with soils of the character to which it was hoped the test would apply. DETERMINATION O F THE AMOUNT OF LIME REMOVABLE BY HYDROCHLORIC ACID. I n soils containing some calcium carbonate, a number of methods have been proposed for arriving at their probable need of lime. Among others, Mondesir' gives a method by which he proposes to ascertain the active calcium carbonate in soils, since, as he says, soils may contain considerable of this substance, per- haps so surrounded by particles of clay that for practical purposes it is of little or no use, and nevertheless stand in need of lime. In our own soils which have been shown to contain practically no calcium carbonate, it must be evident that such methods can be of no avail. If the need of lime, in soils of the character under consideration, is determinable by the quantity of lime present, regardless of its form, then extraction with a strong mineral acid ought to serve as a fair basis for arriving at the lime requirement of the soil. Various methods of treatment, with different kinds, strengths, and volumes of acids have been proposed. For the purpose in hand the method* of the Association of Official Agri- cultural Chemists of the United States was followed, the diges- tion of the soil ( I O grams) having been carried on for ten hours with IOO cc. hydrochloric acid ( I I I 15 sp. gr.) at the temperature of boiling water with the employment of a return condenser. The data thus secured, with soils tested in a practical way with lime, in the field, are given later in tabular form in which con- nection they will be discussed. DETERRIINATION OF TOTAL HGMUS. The determination of total humus could only be expected to throw light upon the lime requirement of soils where marked acidity of the soil exists, and while it might even then bear no relation to the benefit to be derived from liming, it was deemed of interest to ascertain by actual tests what relation, if any, would be found. Combined with the tests to determine the amount of humus soluble in ammonium hydroxide, without previous extraction of the soil with hydrochloric acid, it was also possible to learn if there appeared to be any relation between the gain from liming and the percentage of the total humus, which was directly soluble in ammonium hydroxide. The method employed was essentially that of Grandeau as modified by Huston and McBride.’ Twenty grams of soil were first extracted with hydrochloric acid as proposed by Hilgardl with the employment of an automatic washing apparatus, until no further reaction for lime wasobtainable in the wash solution. After the removal of the acid by successive washings with dis- tilled water, the material was extracted with amnioniuni hydrox- ide, aliquot portions of the extract evaporated to dryness, incin- erated, the ash deducted and the humus thus obtained by differ- ence. The relation between the results obtained and the data LIME REQUIREMENT OF SOILS. 161 secured in actual experiments with lime upon several soils will be discussed later. DETERMINATION O F THE AMOUNT OF LIME REMOVABLE BY HYDROCHLORIC ACID. 1 Annales de la Science .4gron., z, zio (1887). 2 Bull. 46, Div. of Chemistry, U. S. Dept. of Agriculture, pp, 39, 40. ACIDITY BY TITRATION. If means could be devised for determining the relative acidity, or the lime requirements of soils by titration, the matter would be rendered extremely simple and the tests could be made in great numbers at small expense. Owing to the extreme insolu- bility of the acids, or acid compounds of the soil, the titration of a watery extract of the soil could not be expected to furnish any very valuable data. In order to obtain, if possible, results which might indicate in some measure the probable affinity for lime 162 H. J. WHEELER, B. L. HARTWELL, AND c. I,. SARGENT. (the substance which would usually be applied in the neutraliza- tion of soils), a study was made of the action of lime-water, the strength of the supernatant liquid, after contact with the soil for some time, being determined by titration. Air-dried soil was employed so as to eliminate any possible influence which might arise from drying at higher temperatures. Owing to the fact that these trials were only preliminary, and on account of the slight variation in the moisture of the soils examined which varied but little from 2 . 2 j per cent., no correction has been applied to the results for the moisture contained in the soil itself. Ten grams of air-dried soil were brought in contact with jo cc. of lime-water in a small flask. The flask was securely corked and occasionally agitated. At the end of about twenty hours the supernatant liquid was passed rapidly through a dry filter, an excess of hydrochloric acid of known strength (3 cc. neutral- izing 36. I cc. of the lime-water) added and the amount of unneu- tralized acid determined by titration with lime-water. In this way the volume of lime-water neutralized by I O grams of the air-dried soil was indirectly determined. With the same soil more lime-water was required, without exception, where it had remained unlimed, than where lime had been applied. In the comparison of two unlimed soils manured alike excepting that one had received its nitrogen for several years in sodium iiitrate and the other in ammonium sulphate, the differences in the amounts of lime-water neutralized were not always in harmony with, nor were they marked enough to accord at all with the experience obtained in tests of the soils made by plants themselves. ACIDITY BY TITRATION. In a comparison of limed and unlimed soils it was found that by employing a greater amount of lime-water (7 j cc.) greater differences were obtained than by the use of but 50 cc. Possibly the use of even a somewhat greater amount would have given even better comparative results. Trials were also made, employing hydrochloric acid twenty times as dilute as that first tried. Coralline was used as an indicator, and though far from satisfactory in some respects, it was the only one tried which proved at all effective. The preliminary trials not having indicated any particular (the substance which would usually be applied in the neutraliza- tion of soils), a study was made of the action of lime-water, the strength of the supernatant liquid, after contact with the soil for some time, being determined by titration. Air-dried soil was employed so as to eliminate any possible influence which might arise from drying at higher temperatures. Owing to the fact that these trials were only preliminary, and on account of the slight variation in the moisture of the soils examined which varied but little from 2 . 2 j per cent., no correction has been applied to the results for the moisture contained in the soil itself. Ten grams of air-dried soil were brought in contact with jo cc. of lime-water in a small flask. The flask was securely corked and occasionally agitated. At the end of about twenty hours the supernatant liquid was passed rapidly through a dry filter, an excess of hydrochloric acid of known strength (3 cc. neutral- izing 36. I cc. of the lime-water) added and the amount of unneu- tralized acid determined by titration with lime-water. In this way the volume of lime-water neutralized by I O grams of the air-dried soil was indirectly determined. With the same soil more lime-water was required, without exception, where it had remained unlimed, than where lime had been applied. In the comparison of two unlimed soils manured alike excepting that one had received its nitrogen for several years in sodium iiitrate and the other in ammonium sulphate, the differences in the amounts of lime-water neutralized were not always in harmony with, nor were they marked enough to accord at all with the experience obtained in tests of the soils made by plants themselves. 1 Berthelot aud Andre (Ann. chim. fhys., 7th Series, I , 286 (1894)) in operating upon organic residues from soils treated with hydrofluoric acid, state that the amount of po- tassium oxide held by the organic matter varied with the strength of the potassium hy- droxide with which it was treated. ACIDITY BY TITRATION. In a comparison of limed and unlimed soils it was found that by employing a greater amount of lime-water (7 j cc.) greater differences were obtained than by the use of but 50 cc. Possibly the use of even a somewhat greater amount would have given even better comparative results. Trials were also made, employing hydrochloric acid twenty times as dilute as that first tried. Coralline was used as an indicator, and though far from satisfactory in some respects, it was the only one tried which proved at all effective. Th li i t i l t h i i di t d ti l The preliminary trials not having indicated any particular general value of such a method, a trial of ammonium hydroxide was next made. 163 LIME REQUIREMENT OF SOILS. TITRATION TESTS BY MEANS OF AMNONIUM HYDROXIDE. A number of preliminary tests gave indications that the use of ammonium hydroxide in place of lime-water, was capable of giv- ing results more nearly in accord with those obtained by plant tests, The method finally used was as follows : Fifteen grams of airrdried soil were placed in a glass flask and 100 cc. of about & ammonium hydroxide and IOO cc. of distilled water added. This was allowed to stand at the ordinary temperature of the laboratory for about forty-two hours, with frequent agitation during the earlier portion of the time. After it had finally been allowed to settle, a definite amount of the supernatant liquid was treated with hydrochloric acid (about F) at the rate of five cc. for each 20 cc. of the liquid, and made up to a given volume, After the precipitated humus had subsided, aliquot portions of the solution (50 cc.) were titrated with ammonium hydrox- ide, using coralline as an indicator. Since by this treatment some portion of the organic acids may be dissolved by the ammo- nia, the method cannot give an absolute idea of the acidity of a soil. On the other hand, a method which determines the humus inimediately soluble in animonium hydroxide before extraction with hydrochloric or other strong acid, gives no indication of the acidity represented by acid inorganic compounds. Possibly if the amount of humus soluble in ammonium hydroxide were determined, and likewise the amount of ammonia held at the same time in combination by the insoluble residue, the combined data might furnish a better guide than either one alone. Unfor- tunately the occasional opportunities available for conducting such investigations have not yet permitted a study of this question. I n order to ascertain what effect a variation in the strength and volume of the ammonium hydroxide and the volume of soil might have, a given' soil was tested in various ways. ( I ) Twenty grams of soil were treated with IOO cc. of ammo- nium hydroxide diluted with a like quantity of water. ( I ) Twenty grams of soil were treated with IOO cc. of ammo- nium hydroxide diluted with a like quantity of water. (2) Twenty grams of soil were treated with 50 cc. of ammo- nium hydroxide diluted with 150 cc. of water. (2) Twenty grams of soil were treated with 50 cc. of ammo- nium hydroxide diluted with 150 cc. 164 H. J. WHEELER, B . L. HARTWELL, A N D c. L. SARGENT, 164 H. J. WHEELER, B . L. HARTWELL, A N D c. L. SARGENT, 164 (3) Twenty grams of soil were treated with 200 cc. of ammo- nium hydroxide. (4) Twenty grams of soil were treated with 150 cc. of ammo- ni u m hydroxide. ( j ) Fifteen grams of soil were treated with IOO cc. of ammo- nium hydroxide. The following were the amounts of ammonium hydroxide retained in combination by the soil calculated in each case to IO grams of substance. Ammonium hydrox cc. ....................................... (I). 14.3 ( 2 ) ........................................ 11.1 ( 3 ) ........................................ r;.j (4). ...................................... 14.0 (j) 13.7 ........................................ It will be seen from a careful examination of the data that they give some indication that only the weakest solution materially affected the results. The trials of a number of Rhode Island soils by this method as described on page 170 are given in a subsequent table where the data may he compared with those obtained by other chemical methods and with the actual gains in )yield of beets, obtained by the employment of lime. TITRATION TESTS BY MEANS OF AMNONIUM HYDROXIDE. of water. METHOD FOR DETERRIINING T H E LIME REQUIREMENT O F SOILS BY COMPAKISG, UNDER IDENTICAL CONDITIONS, T H E COLOR O F AN AMiIfONIUJI HYDROXIDE EXTRACT O F SOILS O F UNKNOTVN CHARACTER WITH O S E OF KNOWN CHARACTER. The method employed was to treat in a warm room like amounts of soil with like volumes of ammonium hydroxide (containing about 2.4 per cent. of ammonia) for two days with occasional agitation. The small amount of water contained in the air-dried soil itself has been shown to exert no appreciable effect upon the result so that for approximate comparisons, drying a portion of the soil at a high temperature and making allowances for the water in the soil, is unnecessary. The colors of the solutions are finally compared and if darker than the standard, they are diluted with water to match it, or if lighter in color some of the standard solution prepared from the LIME REQUIREMENT OF SOILS. 165 soil of known character, is similarly diluted until it matches the one with which the comparison is being made. The amount of dilution in either case measures the excess or deficiency of color as compared with the standard. It was thought that certain weighable matters might be extracted by ammonium hydroxide from unhumified plant residues which might not color the solution to the same extent as the matter which is actually humified. In such a case, this method might perhaps give better indications than that as employed by Miintz (see page I j j). The possible variation in color of the individual ingredients of the complex, which is classed as humus, may in some cases seriously interfere with the use of the method. The rapidity with which a large number of comparative determinations can be made after the material has settled, is a strong point in its favor as compared with the method of Muntz. The results obtained by both methods may be compared below in connection with those obtained in actual plant tests upon limed and unlimed soil. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. This work was begun in April 1895, at which time Tacke's method of treating peat soils at ordinary temperatures with calcium carbonate, had not been published. No notice of similar attempts. to arrive at the lime requirement of soils having been found, the line of work was apparently wholly new and had to be conducted without the aid of any previous experience. The first soil selected for this work was from the permanent experimental plot No. 23? which had been manured annually beginning with the year 1893, witk the following commercial materials ; viz., dissolved bone-black, potassium chloride and ammonium sulphate. The same arrangement was employed for the absorption of the carbon dioxide as is customary in ordinary determinations of carbon. As a preliminary trial 40 grams of soil, containing 24 per cent. of water capable of being driven off at 100' C., were placed in a Wolff flask and heated, in the 166 166 H. J. WHEELER, B. L. HARTWELL, A N D C . I.. SARGEST. presence of water, with 4 grams of calcium carbonate. On account of the rapid evolution of carbon dioxide, the heating had to be conducted with the utmost care, as the boiling-point was approached, in order that the gas might be entirely absorbed. A blank test with a definite amount of water and 80 grams of the soil (passed through a sieve containing holes & of an inch in diameter) gave 0.0124 gram of carbon dioxide. A blank test with the water and 8 grams of calcium carbonate yielded 0.002 gram of carbon dioxide. Assuming that TIO carbon dioxide came from the calcium carbonate, 80 grams of the soil must have yielded upon heating in themanner described but (o.o~z~-o.oo 0.0104 gram of carbon dioxide. Eighty grams of soil were heated with the same amount of water and calcirim carbonate, and kept at the boiling-point until the evolution of gas seemed to have become practically constant and but very limited in amount. The carbon dioxide yielded, amounted to 0.338 gram. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. Assuming again that the entire blank from the water and calcium carbonate was due to the water and deducting therefore the blank of 0.0124 gram, it is seen that the amount of carbon dioxide fonnd to have been liberated as a result of the action of the calcium carbonate upon the 80 grams of soil, was 0.3256 gram. Owing to rapid evolution of gas on approaching the boiling-point, which necessitated the fre- quent removal of the lamp, it was feared that some carbon dioxide must have passed through the entire apparatus. h subsequent test with 40 grams of soil and 8 grams of calcium carbonate was then made. Deducting a blank of 0.0052 gram for the soil and o.002 gram for the water and calcium carbonate, the aniount of carbon dioxide expelled by the soil was found to be 0.1762 gram. This corresponds to 0.3J24 gram for 80 grams of soil as compared with 0.3256 gram in the first test, but in the last test great care was exercised that no carbon dioxide should be lost. X third test with 40 grams of soil and 4 grams of calcium carbonate yielded 0.178 gram of carbon dioxide. Deducting, as above, a blank of 0.0072, the amount of carbon dioxide found to have been liberated by the soil was equivalent to 0.3416 gram for 80 grams of soil. In order to ascertain if differences between soils corresponding LIME REQUIREMENT OF SOILS. '67 to those shown by the growth of plants could be detected, a soil (from permanent Plot No. 27) was selected which had been manured the same as the other excepting that it had received its nitrogen in sodium nitrate instead of in ammonium sulphate. A blank test with 40 grams of soil and with the usual amount of water yielded 0.015 gram of carbon dioxide. A test with 40 grams of soil gave, after deduction of the blank just mentioned, 0.1402 gram of carbon dioxide or the equivalent of 0.2804 gram for 80 grams of soil, or much less than was shown by the soil first tested. This accorded with the results secured in plant tests and indicated much greater acidity where nitrogen in ammonium sulphate had been employed than where nitrogen in sodium nitrate was used, or where a basic instead of an acid residue had been left in the soil. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. In all of the tests thus far made, the operation was stopped after the chief evolution of gas was apparently over and when only a constant and very slight evolution was still noticeable. To ascertain the effect of longer heating, 40 grams of soil and 4 grams of calcium carbonate were treated for a much longer time than heretofore. The amount of calcium oxide eliminated, allowing for the blank of 0.015 gram was equivalent for 80 grams of soil to 0.3126 gram as compared with 0.2804 gram where the period of heating was less prolonged. Another test was made under the same conditions, continuing the heating for about the same period as in the preceding instance. The net amount of carbon dioxide driven off, per 80 grams of soil, amounted to 0.322 gram. The same material brought to boiling once more and heated for some time yielded, per 80 grams of soil, 0.032 gram of additional carbon dioxide. Brought to boil- ing and heated for a third time the further loss for the same amount of soil was equivalent to 0.0256 gram. From these tests it appeared probable that beyond a certain point the attack upon the calcium carbonate became very limited or that a gradual oxida- tion or else a splitting up of the organic matter was taking place. To ascertain if this would go on at an appreciable rate at ordinary temperatures, a current of air was next passed for two and one- half hours over the material which had been thus successively heated. As a result it was found that no weighable amount of carbon dioxide had been eliminated. Further heating, as 168 H. J. WHEELER, B. L. HARTWELL, AND C. L. SARGENT. before, resulted in a still further elimination of carbon dioxide and an increase in the weight of the absorption tubes. From this it was evident that there was apparently no reasonable time limit within which the elimination of carbon dioxide would be ended, and hence to obtain satisfactory comparative tests with different soils, the period of treatment, temperature, and other conditions must all be kept as uniform as possible. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. One apparent drawback to the method as it was used, was the necessity of conducting the heating so cautiously, and in con- sequence, for so long a time, in order that the carbon dioxide might all surely be absorbed, for by doing so the opportunity for the gradual breaking upof theorganic matter was much increased. To overcome this objection, in a measure, the use of soda-lime was resorted to as an absorbent for carbon dioxide. By this means, less care was required to insure the absorption of the gas and the period of heating could be materially reduced. Never- theless the method was still unsatisfactory, and in consequence, it was proposed to operate so as to subject the material to the action of heat a minimum length of time and either measure the volume of the carbon dioxide disengaged or to determine it indirectly by titration with the employment of barium hydroxide, or by estimating the amount of carbonate formed from it and so indirectly the carbon dioxide. Preliminary to this work it was thought wise to ascertain if possible the minimum time that it was necessary to heat the calcium carbonate and soil, in the presence of water, to insure that all of the humus had united with calcium oxide. I t is well known that in many cases where soil contains naturally considerable quantities of lime and magnesia, ammonium hydroxide extracts practically no humus until the lime and magnesia have first been removed by acids. Such has been found to be the case in soil from S e w Mexico and with certain soils from Massachusetts and Rhode Island, which had been treated successively with calcium carbonate. The idea therefore suggested itself that by heating a number of soils with calcium carbonate and water for varying intervals of time, and subsequently testing the residues with ammonium hydroxide, it could be ascertained how long it was necessary to continue the heating in order to get practically all of the humus into an insoluble combination with lime. It was thought that this time LIME REQUIREMENT OF SOILS. 169 might be safely taken as the limit for continuing the heating and that the amount of carbon dioxide thus disengaged would serve as a suitable measure of the lime requirement of soils. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. It was also proposed to test the method with soils which do not react acid, in order that further light might be thrown upon the probable applicability of such a method to the purpose in hand. Unfortunately, routine work has prevented the further carrying on of this investigation. It is a question of too great scientific and practical interest to be laid aside at this point, and it is hoped that opportunity for further pursuit of the question will be afforded in the near future. COMPARISON OF THE GAIN I N BEETS, RESULTING FROM T H E USE OF LIME, WITH T H E DETERMINATION O F T H E LIME REQUIREMENT OF SOILS MADE BY VARIOUS CHEMICAL METHODS. The following table shows the amounts of beets produced upon limed and unlimed soil in different sections of Rhode Island. Each plot was manured alike with commercial nitrate of soda, muriate of potash and dissolved bone-black. The table gives the percentage gain from liming, also the amount of humus immediately soluble in ammonium hydroxide, the percentage of " total " humus, the amount of calcium oxide removable by extracting with strong mineral acid, the number of milligrams of nitrogen in the ammonia held in combination by IO grams of soil, the comparative dilution of the ammonium hydroxide extract of a soil of known character to give a degree of coloration corresponding to that exhibited by the extracts from the different soils tested, the amount of calcium oxide removable by carbonated water and the percentage of material in the respective soils which was so coarse as not to pass a sieve with openings 21g of an inch i n diameter. Only such material as passed circular openings of this size, was employed in the various tests. The analytical data in the table were obtained with soil pass- ing a sieve with hole & of an inch in diameter. The results are calculated to dry soil. 1 7 0 H. J. WHEELER, B. I,. HARTWELL, XXD c. L. SXRGENT. 1 7 0 H. J. WHEELER, B. I,. HARTWELL, XXD c. L. SXRGENT. 'aD1 w ... . - - - - . Y u u) N d LIME KEQUIREMENT OF SOILS. 1 Compf. rend., 115, 317, August (1892). ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. 171 It will be seen that in the foregoing table the Warren soil which was least helped by lime, is placed at the top, the others being arranged in the order of the benefit noted, from top to bottom, with the Slocumville soil, representing a gain of 10,080 per cent. as a result of liming, placed at the foot of the list. Comparing the percentages of material extracted immediately by treatment with ammonium hydroxide, with the gains from liming, it will be seen that there is no regular correspondence. The tests show a successive decrease in the first three soils while the crop increased. The Niantic soil shows a similar decrease. Of these soils, those at Niantic and Jamestown were exceedingly sandy, while those at Harrisville and Summit were quite gravelly. The result with the Warren soil was unexpectedly high. In the case of the balance of the “ heavier’’ or more compact soils, a successive increase in humus immediately soluble in ammonium hydroxide was noticed, corresponding well, with the exception of the Slocumville soil, with the crop results. In fact this was the only method tried which indicated differences between the So. Portsmouth and Hamilton soils corresponding with those afforded by the crops. It will be noted that the Slocumville soil showed more benefit from liming than that at ,Kingston, though the latter gave a higher test by this method than the Slocumville soil. For the heavier classes of soils containing naturally considerable quantities of humus, this method gives much promise of usefulness. In the case of the more silicious soils containing but moderate amounts of humus, the method may perhaps indicate with considerable truthfulness the relative lime requirements within the particular class, but when it comes to comparisons of individuals of the one class with those of the other, it is possible that it fails to indicate the truth with sufficient accuracy on account of a considerable amount of the acidity being due more to inorganic, than to organic matter. In support of this, it may be cited that, many of our subsoils, which contain only minimum quantities of organic matter, are decidedly acid, indicating the presence of acid silicates, or other acid inorganic compounds. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. Mondesir’ found that upon the destruction of all of the organic matter in a soil by means of “permanganate” at a temperature below 100’ C., or by combus- In support of this, it may be cited that, many of our subsoils, which contain only minimum quantities of organic matter, are decidedly acid, indicating the presence of acid silicates, or other acid inorganic compounds. Mondesir’ found that upon the destruction of all of the organic matter in a soil by means of “permanganate” at a temperature below 100’ C., or by combus- 1 Compf. rend., 115, 317, August (1892). H. J. WHEELER, B. L. HARTWELL, A N D C. L. SARGEXT I72 tion at dull red heat, and after subsequent treatment at ordinary temperature with very dilute hydrochloric or nitric acid, followed by complete removal of these by washing with distilled water and subsequent heating for several hours at from 160' to 170' C., the soil residue was still acid. It decomposed calcium carbonate to some extent at ordinary temperature, and still further upon heating with water to the boiling-point. I t is of interest to note that in the case of the Slocumville soil, where beets almost absolutely refuse to grow without the aid of lime, only 27.76 per cent. of the total humus was immediately soluble in ammonium hydroxide, and in the Kingston soil which stood second in its degree of benefit from liming, but 38.64 per cent. was soluble or much less than in many of the soils wheYe lime proved far less beneficial. The percentage of the total hu- mus which is thus dissolved is not then in any sense a true meas- ure of the lime requirement, though the absolute amounts of humus thus dissolved, have seemed. particularly in our quite heavy soils containing considerable total humus, to correspond with considerable accuracy to the tests with beets. Though there seems to be no reason to think that there would be any relation between the lime requirement of soils and their total humus on soils of widely divergent character, yet upon soils of analogous mineral origin, it seemed probable that some such relation might exist. The results obtained when plotted in a curve, show, with a few exceptions, the same general trend as those obtained by direct extraction with ani monium hydroxide. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. In the case of the Jamestown and Harrisville soils the total humus was higher in the former while the amount of humus immediately extracted, was greater in the latter. In the Hamil- ton soil, the total humus runs too low to accord with the crop test, while the amount of humus immediately soluble in amnionium hydroxide is greater in the case of that soil than in the one at So. Portsmouth, or in agreement with the results secured with beets. In regard to the Kingston and Slocumville soils, the total humus is slightly less in the latter than in the former, but the results come far nearer agreeing with the crop test. than t h ~ s e obtained by immediate extraction of the humus. In the case of the determination of the lime (calcium oxide) removable from the soil by digestion with hydrochloric acid, the I,IRIE REQUIREMENT OF SOILS. I73 results are of marked interest, since they show in the most striking manner that the data thus secured with our Rhode Island soils furnished no reliable guide to their lime requirements. As marked illustrations, it will be seen that the Foster Center soil which contained 0.558 per cent. of calcium oxide, or the greatest quantity of any of the soils examined, gave a gain in crop of 291.8 per cent. as compared with a gain from liming of only 21 per cent. in the Warren soil, which yielded only 0.382 per cent. of calcium oxide upon extraction with hydrochloric acid. Again the Slocumville soil showed by extraction with the mineral acid 0.553 per cent. of calcium oxide, or but 0.005 per cent. less than the Foster Center soil, and yet it was benefited by lime more than any other soil tested giving a gain from its use amounting to 10,080 per cent. The third highest content of calcium oxide as determined by extraction with hydrochloric acid was shown by the Kingston soil, which yielded 0.412 per cent., and yet the only soil which showed a greater benefit from liming was that at Slocumville where the percentage of calcium oxide was even much higher than at Kingston. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. These examples are sufficient to show that even if by this method of analysis considerable lime, even amounting to more than 0.5 per cent., is shown to be present, it may nevertheless be in such combinations that the soil will almost absolutely refuse to produce certain crops until subjected to liming or other alkaline treatment. The number of milligrams of nitrogen represented by the ammonia retaitled by the soil, as determined by the method of titration, accords better with the crop results in the case of the Jamestown and Harrisville soils, than those obtained by immediate extraction with ammonium hydroxide. With the Hamilton soil, the result by the latter method agreed more closely with the crop tests, than those by the former, while in the case of the Kingston and Slocumville soils, the agreement with the crop results was more nearly perfect than those by any other method employed. Only two tests were made by treatment of the soil with carbonated water. The soil which was the more in need of liniing showed by this method a greater quantity of calcium oxide removable by the treatment than the other, or just the reverse of what should have been the case if the method were to 174 H. J. WHEELER, B. I,. HARTWELL, ,4ND C . L. SARGBN’I’. 174 prove of value for the purpose in hand. This fact, together with some experience in the same line in connection with other soils, coupled with the small quantities of calcium oxide removable, gave little reason to hope that the method was applicable for ascertaining the true lime requirements of our acid Rhode Island soils. The tests by means of the comparison of the color of the ammoniuni hydroxide extract in the case of the four upper soils in the table, viz., those from Warren, Summit, Jamestown, and Harrisville indicate a gradation in the opposite direction from those with beets. The amount of total humus and the ammonia held by the soil as determined by the method of titration, agree, in the case of these soils, more nearly with the crop results. The results with the Foster Center and South Portsmouth soils. though not showing relative differences in the line of those exhibited by the crop, point, when compared with the first four soils in the series, in the same general way in relation to the soil requirements as the tests with beets. ATTEMPTS TO DETERMINE T H E LIME REQUIREMENT OF SOILS BY MEASURING T H E CARBON DIOXIDE EXPELLED UPON HEATING A GIVEN QUANTITY O F SOIL, I N T H E PRESENCE O F WATER, WITH CALCIUM CARBONATE OR WITH SODIUM CARBONATE. The results with the South Portsmouth and Hamilton soils fail to accord between themselves with the beet test, though the result with the latter soil, when cornpared with the first four of the series, points in the right general direction. Making a general comparison of the data secured with the Kingston and Slocumville soils n-ith those in connection with the soils standing above them in the table, the tests point to the truth as concerns the needs of the soils. Between themselves the tests would indicate a slightly greater need of lime in the Kingston soil than in the Slocumville soil, or just the opposite of the actual experience. I n closing this consideration of these results, it should be stated that it was recognized at the outstart that unfavorable physical conditions might exert a very disturbing influence, and that in consequence a method which takes into account only cheniical defects could not be expected to agree in all cases with the crop results, I t was hoped, however, to combine physical analyses of the soil with the chemical, with the idea that possibly by a study of both, one might be helped very materially in arriving at a more correct judgment than by the use of either one singly. The resources at disposal have not as yet made it possible for these determinations to be made. There seemed no reason to LIME REQUIREMENT OF SOILS. I 7 5 conclude that a physical analysis of the soils only, could come as near revealing their litne requirements as chemical tests, particularly as it had already been abundantly demonstrated' that the beneficial action of lime upon our soils is, in a large degree, due to its correcting their prevailing acidity. 1 Eighth Annual Report, Rhode Island Agricultural Experiment Station, pp. 232-280 (1895) ; also Ninth Annual Report, pp. 294-318 (1896). SUMMARY (4) The method of Tacke for determining the relative acidity of peat (moor) soils of North Germany based upon the amount of carbon dioxide which the acid ingredients of the soil can expel (4) The method of Tacke for determining the relative acidity of peat (moor) soils of North Germany based upon the amount of carbon dioxide which the acid ingredients of the soil can expel LIME REQUIREMEST O F SOILS. 176 from calcium carbonate at ordinary temperatures in an atmos- phere of hydrogen. This method has not seemed to be applicable to our acid upland soils. from calcium carbonate at ordinary temperatures in an atmos- phere of hydrogen. This method has not seemed to be applicable to our acid upland soils. ( 5 ) The determination of the total humus by a modification of the Grandeau method, which consists in removing the lime and magnesia from their combinations with the humus by treatment with dilute hydrochloric acid, and then dissolving the humus in ammonium hydroxide (animonia water) and estimating the amount dissolved. In our granitic soils containing considerable quantities of humus, this method gives results standing in some- what definite relation to the lime requirements. (6) Determinations of the amount of lime (calcium oxide) which can be dissolved by weak hydrochloric acid by digestion at a high temperature for several hours (official method of the A. 0. A. C.). This method furnishes no reliable basis for arriving at the lime requirement of our acid upland soils. (6) Determinations of the amount of lime (calcium oxide) which can be dissolved by weak hydrochloric acid by digestion at a high temperature for several hours (official method of the A. 0. A. C.). This method furnishes no reliable basis for arriving at the lime requirement of our acid upland soils. ( 7 ) Method by titration, based upon measuring the number of milligrams of nitrogen combined as animonium salts and held by the soil when it is treated in a given way and for a given time with a very dilute solution of ammonium hydroxide. ( 7 ) Method by titration, based upon measuring the number of milligrams of nitrogen combined as animonium salts and held by the soil when it is treated in a given way and for a given time with a very dilute solution of ammonium hydroxide. SUMMARY In comsequence of the recognition of the acidity of many of the upland soils of Rhode Island and their consequent need of lime, it seemed desirable to find some chemical means for satisfactorily revealing their lime requirements, so as to avoid resorting to the slower and more expensive field experiments. A number of methods have been tried, as follows : A number of methods have been tried, as follows : ( I ) Moistening the soil with water, then bringing it in con- tact with blue litmus paper and noting with what rapidity and to what extent it reddens the paper as compared with some soil whose character is already well known. This method is highly effective in the hands of a close observer who has had much experience in testing soils of known character. ( 2 ) Treatment of the soil with water to which a little ammonium hydroxide (ammonia water) has been added and noting whether, after settling, the liquid has assumed a dark brown or black appearance. This test applies only where the acidity is due in a considerable measure to acid organic substances. It may not apply in all sections of the country, as shown by Snyder, in Minnesota, but it has thus far proved a very useful and reliable test with Rhode Island and certain other New England and New York soils containing considerable quantities of humus. Gravimetrically applied this is the method of Miintz. (3) The method employed by Holleman for determining the lime requirement of stiff clay soils in Holland, based upon the amount of lime (calcium oxide) removable by extracting with carbonated water. This method has given little promise of use- fulness in testing our upland acid soils. (3) The method employed by Holleman for determining the lime requirement of stiff clay soils in Holland, based upon the amount of lime (calcium oxide) removable by extracting with carbonated water. This method has given little promise of use- fulness in testing our upland acid soils. SUMMARY has thus far prevented further attempts to carry out a satisfactory study of this method including the perfection of its details. LABORATORY OF THE RHODE ISLAND AGRICULTURAL EXPPRIMENT STATION SUMMARY The results by this method agree more closely with the crop tests than by any other method tried, though it is closely approximated to, by the gravimetric determination of the humus dissolved from the soil by extraction with amnioniuni hydroxide without the previous removal of the lime and magnesia. ( 8 ) Method based upon the comparison of the color of the extract made from soils by treating directly with ammonia water, with the color of such an extract prepared from a like quantity of soil of known character. It seems probable that where the acidity is due largely to acid organic substances, this method may give fairly good results. The rapidity with which tests may be made by it is an important practical point in its favor as compared with the method as proposed by Muntz. ( 8 ) Method based upon the comparison of the color of the extract made from soils by treating directly with ammonia water, with the color of such an extract prepared from a like quantity of soil of known character. It seems probable that where the acidity is due largely to acid organic substances, this method may give fairly good results. The rapidity with which tests may be made by it is an important practical point in its favor as compared with the method as proposed by Muntz. ( 9 ) Xethod based upon the amount of carbon dioxide expelled by a given amount of soil from calcium carbonate when the two are heated together at the boiling-point in the presence of water. This method gives promise of much value if the period of heat- ing is made uniform and very brief, and if the carbon dioxide liberated is easily and accurately determined. Routine work ( 9 ) Xethod based upon the amount of carbon dioxide expelled by a given amount of soil from calcium carbonate when the two are heated together at the boiling-point in the presence of water. This method gives promise of much value if the period of heat- ing is made uniform and very brief, and if the carbon dioxide liberated is easily and accurately determined. Routine work NEW BOOKS. ‘77 has thus far prevented further attempts to carry out a satisfactory study of this method including the perfection of its details. NEW BOOKS. ASALYSIS OF Tt‘HITE PAINTS. 4 Collection of Notes on the Chemical Analysis of White and Tinted Paints. BY GEORGE H. ELLIS, B.Sc. Evanston, Ills.: T h e Technical Press. 8vo. pp. 57. The notes under review were originally published in the Paint, OiZ and Drag Review, of Chicago, and were prepared and edited by oiie who has had long practical experience in the technical examination of the products treated. They consist really of a collection of methods for analysis of paints and pigments long used by the author in his daily practice largely devoted to the products of the paint industry, and as the methods have been thoroughly tested and are described in the fullest detail they will be particularly useful to those chemists whose work has given them but limited practice with such products and methods. With the methods of analysis in each case are presented results obtained by the author in the use of the methods, showing the average composition of the commercial product treated, and thus offers much useful information not easily found elsewhere. The book is commended to those whose duties lead them occasionally in the lines of work here treated and indeed to all needing guid- ance in the technical examination of pigments and paints. WM. MCMURTRIE. THE CHEMISTRY O F SOILS AND FERTILIZERS. B Y HARRY SNYDER, B.S., Professor of Agricultural Chemistry, University of Minnesota, and Chemist of the Minnesota Agricultural Experiment Station. Easton, Pa. : T h e Chemical Publishing Co. 1899. IZ mo. ix f 277 pp. Price $1.50. We are told in the preface that this book was intended primar- ily to be used as a text-book in the author’s classes, This fact explains the ‘ experiments” and ‘ ( review questions” which are given at the end of the book, as well as the condensed manner in which it is written. It should answer admirably the purpose for which it was intended. The subject-matter is well chosen, well arranged, and clearly and concisely stated.
https://openalex.org/W3176790427	https://www.e3s-conferences.org/articles/e3sconf/pdf/2021/51/e3sconf_eilcd2021_03024.pdf	English	null	Practice and Reflection on the Application of Blockchain Technology in the Guangdong-Hong Kong-Macao Greater Bay Area	E3S web of conferences	2,021	cc-by	4,237	Practice and Reflection on the Application of Blockchain Technology in the Guangdong-Hong Kong-Macao Greater Bay Area LIAO Qian1, ZENG LiJuan2* 1School of Economics and Trade, GuangzhouXinhua University; Guangzhou,Guangdong,China; 2C di A th S h l f E i d T d G h Xi h U i it G h 1School of Economics and Trade, GuangzhouXinhua University; Guangzhou,Guangdong,China; 2Corresponding Author, School of Economics and Trade, GuangzhouXinhua University;Guangzhou 1School of Economics and Trade, GuangzhouXinhua University; Guangzhou,Guangdong,China; 2Corresponding Author, School of Economics and Trade, GuangzhouXinhua University;Guangzhou, Guangdong,China; School of Economics and Trade, GuangzhouXinhua University; Guangzhou,Guangdong,China; 2Corresponding Author, School of Economics and Trade, GuangzhouXinhua University;Guangzhou, Guangdong,C g y g g g sponding Author, School of Economics and Trade, GuangzhouXinhua University;Guangzhou, Guangdong,China; Abstract: Blockchain technology has now entered the 3.0 era with solutions of“Blockchain Plus Industry”. As a technological innovation and industry-leading region, the Guangdong-Hong Kong-Macao Greater Bay area has inherent advantages in the development of blockchain technology. How blockchain can lead the development of industry ecology in the future is a question worthy of our in-depth discussion. By analyzing the policy support and basic status of the Guangdong-Hong Kong-Macao Greater Bay Area, this paper sorted out the structure of the four-level characteristic development mode of blockchain technology, summarized the impossible triangle of technology, incompatibility issues, standard construction, supervision and other issues, and putforward relevant suggestions for the development of the blockchain in the Guangdong-Hong Kong-Macao Greater Bay area. 1 Introduction property rights, high cost and difficulty in rights protection, and low utilization rate in the authorization. The application scenario proposed in this paper is to use the central CA, root CA and sub-CA for identity authentication, use the alliance chain for authorization, and store the ownership in the public chain. After the alliance chain authentication and authorization, the AI intelligent patent value evaluation system issues SMART coins to the patent owner to confirm the patent value and ownership. SMART currency is endowed with patent value and cannot be artificially hyped. It is a right affirmation tool with stability and security. In the 18th collective study of the Political Bureau of the CPC Central Committee, General Secretary Xi Jinping emphasized that the integrated application of blockchain technology played an important role in new technological innovation and industrial change, which means that the development of blockchain technology has become a national strategy. In the development plan of the Guangdong-Hong Kong-Macao Greater Bay Area, blockchain technology, as an important breakthrough in the independent innovation of core technologies, is the driving force for the economic development and scientific and technological progress of the Greater Bay Area. Wang Y. (2020)[2] pointed out that China Construction Bank has applied blockchain technology in domestic letter of credit, fortification, international factoring and other fields, which can improve timeliness, reduce operational risks and ensure the security of financial information. *Email: selenazlj@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 275, 03024 (2021) EILCD 2021 E3S Web of Conferences 275, 03024 (2021) EILCD 2021 https://doi.org/10.1051/e3sconf/202127503024 EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 icenses/by/4.0/). s. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 0/) 2 Blockchain application scenarios Blockchain has the characteristics of distributed, shared and immutable, which can enhance trust and improve cooperation efficiency in multi-party cooperation. These advantages make blockchain highly concerned in industrial application scenarios. Many scholars analyze the application scenarios of blockchain mainly from the following aspects. Ye Y.Q, Wei M. (2020)[3] analyzed and pointed out that the Finttech industry incubation center of Guangdong financial high-tech zone of blockchain,including smart retail, supply chain finance, supply chain management, and blockchain security monitoring. Yang X.M., Li X., Wu H.Q. and Zhao K.Y. (2017)[4] proposed that the characteristics of blockchain could improve the deficiencies in the regional collaborative innovation mechanism of the Guangdong-Hong Kong- Macao Greater Bay area, and solve the problems of system, trust and information flow in the development process. Shenzhen Special Zone News (2020)[1] proposed to promote the construction of the intellectual property interconnection chain of universities in the Guangdong- Hong Kong-Macao Greater Bay Area through the construction of public-owned chain combined with alliance chain, realize the application of blockchain in intellectual property, and solve the problems of long time and low efficiency in the confirmation of intellectual From the government's policy level, industry level, education level of scientific research about big bay, Hong Kong and Macao to blockchain development. E3S Web of Conferences 275, 03024 (2021) EILCD 2021 https://doi.org/10.1051/e3sconf/202127503024 Figure 1. Inventory and increment of blockchain registered enterprises in Guangdong Province in 2019 Data source: collated according to interchain pulse statistics 33.10% 66.90% 20602 Guangdong Blockchain registered enterprises Increment in 2019 33.10% 66.90% 20602 Guangdong Blockchain registered enterprises Increment in 2019 Zhong W. (2019)[5] present situation that the current large bay area standard system has not yet formed. The core technology needs to be overcome, relevant laws and regulations is not perfect, imperfect market mature, block chain lack of talent, etc. The research proposes that the development of blockchain in the Guangdong-Hong Kong-Macao Greater Bay area, its related industrial policies, planning and layout, research and development direction and application scenarios should be based on scientific research and demonstration. 3 The development status of blockchain in the Guangdong-Hong Kong-Macao Greater Bay area Figure 1. Inventory and increment of blockchain registered enterprises in Guangdong Province in 2019 Data source: collated according to interchain pulse statistics Blockchain application scenarios in the Guangdong-Hong Kong-Macao Greater Bay area have covered all aspects of the industry application layer, including government administration, standard construction, education and scientific research, industry level, intellectual property rights, industry supervision, real estate developer lottery room selection application and so on. The application of the industrial map of blockchain in the Guangdong-Hong Kong-Macao Greater Bay area is in a leading position nationwide and has its own unique advantages in the world. Due to the different positioning of cities within the Bay area, the development of blockchain varies from city to city. Blockchain projects are concentrated in Hong Kong, Shenzhen and Guangzhou, and activities are concentrated in Macao. The development pattern of blockchain is coordinated and matched with the development outline of the Guangdong-Hong Kong- Macao Greater Bay area. As the core engine of regional development, Hong Kong, Macao, Guangzhou and Shenzhen concentrate on their scientific and technological innovation ability, complement each other's advantages, promote industrial upgrading by point and area, and promote the development of surrounding areas with technology. 3.1 Policy support Since the launch of blockchain planning in the Guangdong-Hong Kong-Macao Greater Bay Area in 2016, the state has formulated multiple policies to support the development of the Greater Bay Area, support the development of blockchain technology in the region, and strengthen the supervision of digital currency transactions. Thanks to the support of numerous policies, favorable conditions have been created for the development of blockchain in the Guangdong-Hong Kong-Macao Greater Bay area. November 2016, Shenzhen Municipal Financial Office released the 13th Five-Year Plan for the Development of Shenzhen Financial Industry, which mentioned that it supports financial institutions to strengthen the research and exploration of emerging technologies such as blockchain and digital currency. In December 2016, the People's Government of Guangzhou issued the notice of Guangzhou City Accelerates the Development of IAB Industry 5-year Plan (2018-2022)[7] , which mentioned blockchain technology as a key field. In February 2018, the Securities and Futures Commission of Hong Kong issued the Securities and Futures Commission Warns Investors against Cryptocurrency Risks, which showed that digital currencies involving ICOs would be considered as securities and would be subject to supervision. In December 2019, Shenzhen Pilot Demonstration Zone Construction Action Plan (2019- 2025) was officially issued[8] . Shenzhen Futian District will speed up the construction of the country's first digital currency building; In May 2020, the People's Bank of China, the Banking and Insurance Regulatory Commission, the Securities Regulatory Commission and the SAFE jointly issued the Opinions on Financial Support for the Construction of the Guangdong-Hong Kong-Macao Greater Bay area, proposing to support research on the promotion of innovative technologies such as blockchain, big data and artificial intelligence in risk prevention and financial supervision. With the establishment of the Guangdong-Hong Kong-Macao Greater Bay area blockchain alliance in Guangzhou in 2018, the number of registered blockchain enterprises has grown rapidly. As shown in Figure 1, by the end of 2019, there are a total of 20,602 registered blockchain enterprises in Guangdong Province, among which 6,820 are newly increased in 2019[6] , which is enough to reflect that the blockchain technology will usher in a more rational and scientific development under the support of supporting policies and institutions. 2 2 E3S Web of Conferences 275, 03024 (2021) EILCD 2021 E3S Web of Conferences 275, 03024 (2021) https://doi.org/10.1051/e3sconf/202127503024 3.2.1 Hardware and software infrastructure The Guangdong-Hong Kong-Macao Greater Bay area has dual advantages in software and hardware. A large bay area of Guangdong has the largest distribution center - Shenzhen global electronic market, as well as a leading manufacturing in Dongguan, international finance, shipping and trade center and international aviation hub of Hong Kong, has a perfect infrastructure construction and traffic configuration, more famous airport and fine port, the transportation is convenient, is a large bay area of Guangdong infrastructure construction has provided the important safeguard. A large bay area of Guangdong is also created the world's first article millions of TPS thunderbolt chain, a large bay area of Guangdong mill enterprise deployed 1.5 million nodes play, such as guest cloud block chain and improvement of the hardware and software infrastructure development complement each other, infrastructure promoted the innovation of chain block, block chain applications to promote the process of the construction of infrastructure. Financial service is the first application field of block chain technology in China, and the construction of financial infrastructure is relatively perfect. On this basis, the technology expansion, general applications and industrial applications of the Greater Bay area are more expansible and adaptable. 3.2.4 Industry application layer Blockchain in the Guangdong-Hong Kong-Macao Greater Bay Area has an extremely rich range of applications. The technology has infiltrated into government management, education and scientific research, finance, logistics and other industrial fields, intellectual property, industry supervision and so on. In the aspect of government affairs management, in August 2018, Shenzhen Municipal Tax Bureau, Tencent and Kingdee cooperated to release the first research results of the application of blockchain withinvoice ecosystem in Chinaand set up the application scenario of the whole invoice process management of WeChat payment -- invoicing -- reimbursement. In industry regulation, blockchain plays a crucial role in the financial industry. Blockchain, as a regulatory technology, can improve the accuracy and efficiency of internal compliance procedures of financial institutions, accelerate the construction of cooperative regulatory models, and acquire a certain degree of risk assessment ability. 3.2 Four-level characteristic development mode of blockchain with industry technologies, the cross-innovation of blockchain and new generation technologies such as artificial intelligence, cloud computing, big data and the Internet of Things has laid a foundation for the future distributed business system. According to the unique geographical advantages of the Guangdong-Hong Kong-Macao Greater Bay area, it is planned to develop in four key directions: blockchain with industrial manufacturing upgrading, blockchain with digital justice and digital government, blockchain with big data circulation and cross-border data financing, and blockchain with industrial think tank and talent base. With the implementation of more and more blockchain projects, the blockchain industry enters the scene of technology landing, and the application effect of blockchain development extends from the financial field to the real economy becomes more and more significant. 3.2.3 General application layer The geographical and institutional advantages of the Guangdong-Hong Kong-Macao Greater Bay area enable the general application of blockchain technology to be realized in Hong Kong and Macao, such as distributed exchanges, TokenFund, STO, gambling, e-sports, etc. The Guangdong-Hong Kong-Macao Greater Bay area shows diversity and extension in terms of blockchain ecological services, such as capital institutions such as Coin Capital and Huobi Capital; Incubators Guangzhou Yimi Incubator, Guangzhou Blockchain International Innovation Center Capital Machine, and successively produced rating agencies, training institutions, industry associations and other organizations serving the blockchain industry. 4 Blockchain boosts the development prospects of the Guangdong-Hong Kong-Macao Greater Bay Area The technology extension layer includes technological iterations of blockchain technology itself attributes, such as consensus mechanism, smart contract, distributed ledger, etc., as well as cross-disciplinary applications combined with modern technologies, such as big data, artificial intelligence, Internet of Things, etc. The Guangdong-Hong Kong-Macao Greater Bay area has the largest concentration of BaaS/SaaS technology extension layer service enterprises in the world, including Huawei Cloud Platform and Tencent Cloud Platform. As a new technology of the infrastructure of digital economy, its characteristics of decentralization and difficulty in tampering make it capable of reconstructing models in many fields and have great potential to create a new ecology. Regardless of the development trend of blockchain itself or the trend of integration with other In the development outline of the Guangdong-Hong Kong-Macao Greater Bay area, it is necessary to give full play to the regional advantages of the Bay area, explore the characteristics of local economic development, and activate new potential energy of innovation and entrepreneurship. The implementation of the plan needs to rely on the innovation and application of technology. The value transfer of blockchain technology and the trust mechanism can meet the landing of the planning. Its applications are mainly summarized in the following aspects: 3 3 https://doi.org/10.1051/e3sconf/202127503024 E3S Web of Conferences 275, 03024 (2021) EILCD 2021 E3S Web of Conferences 275, 03024 (2021) 4.3 Blockchain with smart manufacturing: smart contracts give new impetus to the real economy One of the goals of the application of blockchain in the real economy is the intelligent manufacturing that can be customized to meet the increasing personalized needs. The manufacturing in the Guangdong-Hong Kong-Macao Bay area has always led the world. As the world factory in China, made in China looks at Dongguan, the Pearl River Delta area has a long history of resource advantages in the manufacturing field. Under the new normal manufacturing from the unity of the traditional factory production to change due to the need to become flexible production model, from the supply market dominance to demand led production mode transformation of supply and demand, from the supermarket stores marketing mode to public domain, the live and shop club deals such as flowers, blossoming operation way to reform the innovation, Not only in the public domain traffic to focus on, but also in the private domain traffic to explore the potential. The programmable feature of smart contract can meet the needs of various intelligence in the new real economy era. Self and consensus are contradictory and unified. Blockchain technology can solve this problem of social governance. Blockchain technology, as a decentralized, distributed ledger of smart contracts, can create individual sections for each individual, which can intelligentially record and track individual behaviors in real time. The links formed by each node can restrict each other, and a two-way incentive mode is formed based on Proof of Work (PoW) or Proof of Stake (PoS) to promote the consensus of the group. For example, in the initiative of low-carbon travel, the application of blockchain will record the daily travel behavior of each individual. Each record can only be automatically generated in the node according to the use of application scenarios. After the node is generated, it will be broadcast to all nodes, which cannot be tampered with and cannot be reversed. Given the corresponding material incentives for each method, each record generated will be able to accumulate digital currency or points, which can be used for consumption or issued with a certificate of entitlement. This virtuous circle mechanism will reshape the governance system of intelligent service society. 4.1 Blockchain with social governance: intelligent services reshape the governance system of a diversified society peer payment method. According to the forecast of McKinsey & Company, P2P technology in B2B cross- border business settlement application can reduce the cost of each transaction by a total of $11, of which 75% is the intermediary bank network cost and 25% is the compliance investigation and exchange cost. In addition to reducing transaction costs, the application of block chain technology will provide financial industry into science and technology strength, establish effective communication between stakeholder body mechanism, coordinate with all concerned groups, with tamper- resistant, decentralized characteristics such as books and set up the stranger mutual checks and balances between the self-consistent trust mechanism, establish the platform of financial innovation. The "9+2" institutional characteristics of the Guangdong- Hong Kong-Macao Bay area and the regional combination with their respective advantages have added more uncontrollable factors to the increasingly diversified social governance. Hong Kong and Macao society demands more personality, more dispersed, the history of Guangzhou culture deep, deep philosophy and traditional redundancy procedure is more, Shenzhen strong innovation consciousness, the concept of Shenzhen girl identity is strong, but crash culture is popular, Dongguan is a manufacturing hub, the characteristics of line for public health, Common problems, such as pollution, are difficult. Huizhou grows under the spillover effect of Shenzhen. The back garden style of social development has too much external influence and insufficient internal force. 4.2 Blockchain with fintech: a self-consistent decentralized trust mechanism to build a financial innovation platform The 2020 novel coronavirus attack on the global epidemic has led to a historical revolution in online education. Blockchain provides technical support for the change of online education mode, and also creates new educational demand for blockchain technical talents. It is predicted that by 2023, 10% of the world's GDP will be stored in blockchain or its technology. The application of blockchain to promote the development of online and offline education lies in credit management, learning certification and the development of online teaching platform. In addition to the application of blockchain in the field of education, it also creates an educational and training demand for blockchain talents. Blockchain talent cultivation is closely related to education. Among the world's four major bay areas, San Francisco Bay Area and New York Bay Area have taken the lead in opening blockchain-related courses, followed by Tokyo Bay Area Technological innovation is the core strength of the Guangdong-Hong Kong-Macao Greater Bay area, and also the potential of the Greater Bay area to compete with the other three bay areas in the world. Financial value as the transfer of industry has been in the application of the restriction of the trust factor to innovation practice, set up in the center of the block chain as value passed after the bridge of financial technology is endowed with the trust of the label decentralization, can high efficiency, low transaction costs to realize financial revenue and expenditure, payment and settlement, capital allocation, cross-border settlement of international trade and other business. The core technology of blockchain technology applied in financial technology is P2P (Point to Point) peer-to- 4 E3S Web of Conferences 275, 03024 (2021) EILCD 2021 https://doi.org/10.1051/e3sconf/202127503024 in Japan and Guangdong-Hong Kong-Macao Greater Bay area. According to the QS World 100, 15 schools in the New York Bay area are currently among the world's top 100. in Japan and Guangdong-Hong Kong-Macao Greater Bay area. According to the QS World 100, 15 schools in the New York Bay area are currently among the world's top 100. and the development roadmap and time node table of blockchain with industry should be formulated. In the development of a large bay area of Guangdong blockchain, we can build a data sharing platform, promote the mechanism innovation, system integration. 5 Conclusions and Suggestion As blockchain technology becomes more and more mature, the combination of application scenarios will become more and more close. This paper attempts to put forward solutions and construction suggestions from the following three aspects: Reference: 1. Shenzhen Special Zone News (2020) Building Intellectual Property Interconnection Chain of Universities in the Guangdong-Hong Kong-Macao Greater Bay Area http://sztqb.sznews.com/PC/content/202006/30/cont ent 880867.html 5.3 Give full play to the first-mover advantage of the Guangdong-Hong Kong-Macao Bay area, stimulate the ability of technological innovation, and reserve talents of blockchain with industry The Guangdong-Hong Kong-Macao Greater Bay area is blessed with unique geographical advantages, and the pilot demonstration zone in Shenzhen provides infinite possibilities for the innovative economy of the Bay area. The open, trusting and fair environment provided by blockchain technology is conducive to reaching institutional consensus within the Guangdong-Hong Kong-Macao Greater Bay area, realizing coordinated regional development, and leading the rapid development of blockchain with industry. The Guangdong-Hong Kong-Macao Greater Bay Area should make good use of this opportunity to strengthen infrastructure construction, enhance inter-city exchanges within the Greater Bay area, train local professionals through universities, enterprises and associations, introduce international outstanding talents, and establish a talent pool for the development of blockchain with industry. Acknowledgments: Thank you for the support of the following fund projects: [1]Guangdong Provincial Key Discipline Public Administration Construction Project, No. F2017STSZD01 In the rapid development of blockchain application scenarios, it is necessary to accelerate the legislative process of blockchain with industry, improve legal construction, improve regulatory means, establish regulatory responsibilities of various institutions, and establish a sound regulatory governance system. In regulation, we can try traditional with innovative methods, such as establishing negative list with consensus mechanism and economic incentive mechanism. Setting up a negative list refers to setting up a negative list of industry codes of conduct, and companies that do not comply with the agreement will be blacklisted. [2]Blockchain with Cross-border E-commerce: Application and Future of New Technology Enabling E- commerce, a university-level scientific research project of Xinhua College of Sun Yat-Sen University （2019KYYB09） 4.2 Blockchain with fintech: a self-consistent decentralized trust mechanism to build a financial innovation platform On the other hand, we can reduce the information asymmetry of inefficient communication, promotingHong Kong and Macao large bay area with the international communication between interconnected, enhancing information transparency and trust. All can provide blockchain with industry build non-barrier development platform. Under the favorable policies, the Guangdong-Hong Kong-Macao Greater Bay area, relying on its superior geographical location, has a good economic foundation and a good hardware foundation. Now the blockchain with industry ecology has been basically formed, which has brought strong support for the subsequent development of blockchain. The current research question of blockchain is no longer about whether it can be applied to a specific industry, but how to apply blockchain to the industry. As a benchmark region for high-speed development in China, the Guangdong-Hong Kong- Macao Greater Bay area should not only focus on close exchanges among cities within the region, promote resource sharing and mutual help among all parties, but also apply technological innovation to a wider range of industries and regions to promote the spillover effect of blockchain technology. In the case of multiple factors tending to favorable development, based on policies, research strength and rich application scenarios, the landing and improvement of the blockchain with industry will bring a better and faster development trend than that of other Bay area. 5.3 Give full play to the first-mover advantage of the Guangdong-Hong Kong-Macao Bay area, stimulate the ability of technological innovation, and reserve talents of blockchain with industry 8. Qu S.N.. (2020) Blockchain to boost the high quality development of real economy: mode, vector and path , Reform, (01): 39-47. 4. Yang X.M., Li X., Wu H.Q. and Zhao K.Y. (2017) Application modes and practical challenges of blockchain technology in the field of education., Modern Distance Education Research, (02):34-45. 5.2 Pay attention to top-level design, build a sharing platform, and formulate the overall development plan of blockchain with industry 2. Wang Y. (2020) Enablement of science and technology to help the development of Guangdong- Hong Kong-Macao Greater Bay Area Financial Technology Times, 01:24-26. The development of blockchain technology has been elevated to national strategy. Under the favorable conditions of this major decision, the top-level design and planning of blockchain development should be done well, 5 E3S Web of Conferences 275, 03024 (2021) E3S Web of Conferences 275, 03024 (2021) EILCD 2021 https://doi.org/10.1051/e3sconf/202127503024 3. Ye Y.Q, Wei M. (2020) Guangzhou Blockchain International Innovation Center: Deep Cultivation of Incubation to Promote the Development of Blockchain Industry in Guangdong-Hong Kong- Macao Greater Bay Area, Guangdong Science and Technology, 29(07):36-40. 6. Chen L., Lin S.J. (2020). Making Blockchain Better Enable Wisdom and Benefit People's Ecology -- Interview with Dr. Zou Jun, General Manager of Guangzhou Yuntong Dajin Service Technology Co., Ltd, Guangdong Science and Technology, 29(07):24- 26. 4. Yang X.M., Li X., Wu H.Q. and Zhao K.Y. (2017) Application modes and practical challenges of blockchain technology in the field of education., Modern Distance Education Research, (02):34-45. 4. Yang X.M., Li X., Wu H.Q. and Zhao K.Y. (2017) Application modes and practical challenges of blockchain technology in the field of education., Modern Distance Education Research, (02):34-45. 7. Xu Z., Zou C.W. (2018) What can and can't blockchain do? Journal of Financial Research, (11):1- 16. 8. Qu S.N.. (2020) Blockchain to boost the high quality development of real economy: mode, vector and path , Reform, (01): 39-47. 5. Zhong W. (2019) A Brief Analysis on the Development of Blockchain Industry in Guangdong- Hong Kong-Macao Greater Bay Area , Guangdong Economy, (11):26-31. 5. Zhong W. (2019) A Brief Analysis on the Development of Blockchain Industry in Guangdong- Hong Kong-Macao Greater Bay Area , Guangdong Economy, (11):26-31. 6 6
https://openalex.org/W4231622384	https://europepmc.org/articles/pmc6069730?pdf=render	English	null	Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico	F1000Research	2,018	cc-by	7,456	RESEARCH ARTICLE Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico [version 3; referees: 2 approved] Dhanya Haridas , Justin C. Biffinger , Thomas J. Boyd , Preston A. Fulmer , Leila J. Hamdan , Lisa A. Fitzgerald3 American Society for Engineering Education, Washington, DC, USA Chemistry Department, University of Dayton, Dayton, OH, USA Naval Research Laboratory, Washington, DC, USA University of Southern Mississippi, Hattiesburg, MS, USA 1 2,3 3 3 4 3 1 2 3 4 Keywords shipwreck, denitrification, microbe, 16S, Gulf of Mexico Referee Status: Invited Referees version 3 published 06 Aug 2018 version 2 published 12 Jun 2018 version 1 published 13 Oct 2017 1 2 report report report report , Max Planck Maxim Ru-bin Blum Institute for Marine Microbiology , Germany Israel Oceanographic and Limnological Research, Israel 1 , Université Laval, Dimitri Kalenitchenko Canada 2 13 Oct 2017, :1834 (doi: ) First published: 6 10.12688/f1000research.12713.1 12 Jun 2018, :1834 (doi: ) Second version: 6 10.12688/f1000research.12713.2 06 Aug 2018, :1834 (doi: ) Latest published: 6 10.12688/f1000research.12713.3 v3 Open Peer Review Referee Status: Invited Referees version 3 published 06 Aug 2018 version 2 published 12 Jun 2018 version 1 published 13 Oct 2017 1 2 report report report report 13 Oct 2017, :1834 (doi: ) First published: 6 10.12688/f1000research.12713.1 12 Jun 2018, :1834 (doi: ) Second version: 6 10.12688/f1000research.12713.2 06 Aug 2018, :1834 (doi: ) Latest published: 6 10.12688/f1000research.12713.3 v3 Abstract : Shipwrecks serve as a rich source for novel microbial Background populations that have largely remained undiscovered. Low temperatures, lack of sunlight, and the availability of substrates derived from the shipwreck’s hull and cargo may provide an environment in which microbes can develop unique metabolic adaptations. : To test our hypothesis that shipwrecks could influence the microbial Methods population involved in denitrification when a consortium is grown in the laboratory, we collected samples proximate to two steel shipwrecks in the northern Gulf of Mexico. Then under laboratory conditions, we grew two independent denitrifying microbial consortia. Each consortium was grown by using the BART assay system and analyzed based on growth kinetics, ion chromatography and 16S amplicon sequencing. : Both denitrifying consortia were different from each other based on Results varied growth profiles, rates of nitrate utilization and 16S amplicon sequencing. : Our observations conclude that the laboratory grown water Conclusions column microbial consortia from deep-sea shipwrecks in the Gulf of Mexico are able to undergo aggressive denitrification. , Max Planck Maxim Ru-bin Blum Institute for Marine Microbiology , Germany Israel Oceanographic and Limnological Research, Israel 1 Keywords shipwreck, denitrification, microbe, 16S, Gulf of Mexico RESEARCH ARTICLE Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico [version 3; referees: 2 approved] Dhanya Haridas , Justin C. Biffinger , Thomas J. Boyd , Preston A. Fulmer , Leila J. Hamdan , Lisa A. Fitzgerald3 American Society for Engineering Education, Washington, DC, USA Chemistry Department, University of Dayton, Dayton, OH, USA Naval Research Laboratory, Washington, DC, USA University of Southern Mississippi, Hattiesburg, MS, USA 1 2,3 3 3 4 3 1 2 3 4 RESEARCH ARTICLE Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico [version 3; referees: 2 approved] Dhanya Haridas , Justin C. Biffinger , Thomas J. Boyd , Preston A. Fulmer , Leila J. Hamdan , Lisa A. Fitzgerald3 American Society for Engineering Education, Washington, DC, USA Chemistry Department, University of Dayton, Dayton, OH, USA Naval Research Laboratory, Washington, DC, USA University of Southern Mississippi, Hattiesburg, MS, USA 1 2,3 3 3 4 3 1 2 3 4 American Society for Engineering Education, Washington, DC, USA Chemistry Department, University of Dayton, Dayton, OH, USA Naval Research Laboratory, Washington, DC, USA University of Southern Mississippi, Hattiesburg, MS, USA 1 2 3 4 Open Peer Review Discuss this article (0) Comments Abstract : Shipwrecks serve as a rich source for novel microbial Background populations that have largely remained undiscovered. Low temperatures, lack of sunlight, and the availability of substrates derived from the shipwreck’s hull and cargo may provide an environment in which microbes can develop unique metabolic adaptations. : To test our hypothesis that shipwrecks could influence the microbial Methods population involved in denitrification when a consortium is grown in the laboratory, we collected samples proximate to two steel shipwrecks in the northern Gulf of Mexico. Then under laboratory conditions, we grew two independent denitrifying microbial consortia. Each consortium was grown by using the BART assay system and analyzed based on growth kinetics, ion chromatography and 16S amplicon sequencing. : Both denitrifying consortia were different from each other based on Results varied growth profiles, rates of nitrate utilization and 16S amplicon sequencing. : Our observations conclude that the laboratory grown water Conclusions column microbial consortia from deep-sea shipwrecks in the Gulf of Mexico are able to undergo aggressive denitrification. F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Keywords Page 1 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Lisa A. Fitzgerald ( ) Corresponding author: lisa.fitzgerald@nrl.navy.mil : Conceptualization, Investigation, Writing – Original Draft Preparation, Writing – Review & Editing; : Author roles: Haridas D Biffinger JC Conceptualization, Supervision, Writing – Review & Editing; : Investigation, Writing – Review & Editing; : Resources, Writing – Boyd TJ Fulmer PA Review & Editing; : Resources, Writing – Review & Editing; : Funding Acquisition, Resources, Supervision, Writing – Hamdan LJ Fitzgerald LA Review & Editing No competing interests were disclosed. Competing interests: Funding was provided by the Office of Naval Research (ONR) through the Naval Research Laboratory (PE# 61153N), the Grant information: Bureau of Ocean Energy Management (BOEM) No. M13PG00020, and the Navy Platform Support Program. © 2018 Haridas D . This is an open access article distributed under the terms of the , which Copyright: et al Creative Commons Attribution Licence permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Data associated with the article are available under the terms of the (CC0 1.0 Public domain dedication). Creative Commons Zero "No rights reserved" data waiver Haridas D, Biffinger JC, Boyd TJ How to cite this article: et al. Laboratory growth of denitrifying water column microbial consortia from 2018, :1834 (doi: deep-sea shipwrecks in the northern Gulf of Mexico [version 3; referees: 2 approved] F1000Research 6 ) 10.12688/f1000research.12713.3 13 Oct 2017, :1834 (doi: ) First published: 6 10.12688/f1000research.12713.1 Lisa A. Fitzgerald ( ) Corresponding author: lisa.fitzgerald@nrl.navy.mil Lisa A. Fitzgerald ( ) Corresponding author: lisa.fitzgerald@nrl.navy.mil : Conceptualization, Investigation, Writing – Original Draft Preparation, Writing – Review & Editing; : Author roles: Haridas D Biffinger JC Conceptualization, Supervision, Writing – Review & Editing; : Investigation, Writing – Review & Editing; : Resources, Writing – Boyd TJ Fulmer PA Review & Editing; : Resources, Writing – Review & Editing; : Funding Acquisition, Resources, Supervision, Writing – Hamdan LJ Fitzgerald LA Review & Editing No competing interests were disclosed ompeting interests: Haridas D, Biffinger JC, Boyd TJ How to cite this article: et al. Laboratory growth of denitrifying water column microbial consortia from 2018, :1834 (doi: deep-sea shipwrecks in the northern Gulf of Mexico [version 3; referees: 2 approved] F1000Research 6 ) 10.12688/f1000research.12713.3 13 Oct 2017, :1834 (doi: ) First published: 6 10.12688/f1000research.12713.1 Page 2 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Enrichment of denitrifying microbial consortium Amendments from Version 2 REVISED y g The commercially available denitrifying Biological Activity Reaction Test (DN-BART) assay (HACH, Colorado, USA) was used to enrich for DN bacteria. Briefly, the lyophilized media in the DN-BART was solubilized with 15 mL water sample from either the Halo or U-166 shipwreck site. The assay was per- formed based on manufacturer’s instructions, with the exception that the assay was incubated for 30 days instead of the suggested 4 days. The enriched microbial consortium obtained from the DN-BART assay was used as the inoculum to perform the growth curve and Ion Chromatographic (IC) studies. Version 3 has a revised Discussion section. The reviewer recommended rewording the statement “Each shipwreck site...” to “The water mass passing by each shipwreck site...” to reflect the fact that the DN consortium identified at each site was obtained from the water sample isolated from a region surrounding the shipwreck and not the shipwreck itself. See referee reports See referee reports Growth curve studies Nunc tubes (Chemglass, NJ, USA) containing 10 mL of modi- fied Indole Nitrite medium (pancreatic digest of casein 20 g/L, disodium phosphate 2 g/L, dextrose 1 g/L, potassium nitrate 1 g/L) were used for all assays. Sterile nitrogen gas was bubbled through the media for 15 min prior to inoculation to de-gas and establish an anaerobic environment. The enriched Halo and U-166 DN consortium derived from the DN-BART assay was used as the inoculum to perform the growth curve studies. Each tube was inoculated with 100 µL of DN-BART consortium. The inoculated Nunc tubes were analyzed for a period of 24 h at 30°C (Excella E25, Fisher Scientific, MA, USA). The optical density of the samples was measured at 600 nm (OD600) every 2 h post-inoculation using the Spectronic 200 Spectropho- tometer (Thermo Scientific, PA, USA) over a 24 h period after inoculation. All inoculated samples were done in triplicates. In recent years, shipwrecks have been identified as areas from which novel microbial species have been isolated, because of the introduction of foreign material to the area6. Thus, they would be an ideal location to discover unique microorganisms and metabolic activity, as these areas are known to be diverse habitats for macroorganisms in the marine environment7. The goal of this research was to prospect for novel DN microbial consortia near deep-sea shipwrecks in the Gulf of Mexico, culture the consor- tia under laboratory conditions and determine their DN activity. In this study, we collected water samples proximal to two steel shipwreck sites located in the northern part of the Gulf of Mexico, and analyzed the denitrifying and culturing potential of the microbial consortia obtained from the two sites. Ion chromatography studies Th i / i i i The nitrate/nitrite concentrations were tracked by IC using a Dionex IC-3000 IC fitted with an IonPac AS16 column. The mobile phase was 9 mM Na2CO3 at 1.0 mL/min and a Dionex 7 anion standard mix was used for calibration before and half way through sample runs. The DN microbial con- sortium was cultured similarly to the growth curve assay (i.e. anaerobically to enable the denitrifying conditions to develop). The Halo and U-166 cultures were sampled (1 mL) using a ster- ile syringe and needle (BD, NJ, USA) every 2 h over a period of 24 h from the inoculated Nunc tubes. The sample was cen- trifuged at 12,000 × g for 3 min in a sterile 1.5 mL eppendorf tube (Eppendorf, NY, USA). The supernatant was additionally syringe filtered (0.2 µm filter Syringefilter, SC, USA) and stored in 1 mL IC vials (Thermo Scientific, Pittsburg, PA). All ana- lyzes were performed in duplicate or triplicate. Nitrate concen- trations were often above the calibration level (100 mg/L) and are annotated as estimated values (J). Standard errors for rep- licate measurements ranged from 0 to 2.26% with an average of 0.31% for the aggregate runs. Introduction The biogeochemical process that transforms dissolved inorganic nitrogen to nitrogen gas is known as denitrification (DN). This metabolic pathway impacts the nitrogen (N) cycle by returning elemental N to the atmosphere1,2. It can alternatively be defined as the reduction of more oxidized forms of nitrogen (NO3 -, NO2 -, NO and N2O) to N2 gas, where it can be linked to the oxidation of iron, sulfur and reduced carbon species3. It is primarily per- formed by facultative heterotrophic or chemolithoautotrophic bacteria under anoxic or very low-oxygen conditions3, where microorganisms utilize nitrate or nitrite as the terminal electron acceptor4. DN, alongside other biogeochemical processes (carbon and sulfur cycles), plays a key role in maintaining the nutrient balance in marine habitats5. Environmental sampling p g We obtained water samples ~600 m down current from two steel-hulled shipwrecks investigated as part of the Shipwreck Corrosion, Hydrocarbon Exposure, Microbiology and Archaeology (SCHEMA) study, which addresses the effect of the 2010 Deepwater Horizon spill on deep-sea shipwrecks in the north- ern Gulf of Mexico (http://www.boem.gov/GOM-SCHEMA/). Samples were collected onboard the R/V Pelican using a CTD-rosette during the PE15-22 expedition in May 2015. The shipwreck Halo, is a steel-hulled steam tanker, resting in ~140 m of water, and ~50 miles west of the Mississippi River’s South- west Pass. The double steel-hulled German U-Boat U-166 shipwreck rests in ~1400 m of water within 10 km of the Macondo wellhead, the epicenter of the 2010 Deepwater Horizon spill. The water samples were stored in sterile plastic bottles at 4°C until further use. Results Hydrographic conditions of the water samples The depth, temperature, salinity and dissolved oxygen (DO) of the water samples obtained from the region proximal to the Halo and U-166 shipwreck sites were obtained with the CTD (Table 1). The water sample obtained proximal to the U-166 shipwreck site had a higher DO content (6.6 mg/L) and lower temperature (4.3°C), as compared to the Halo shipwreck site that had 4.1 mg/L DO and 17.7°C. This indicates that both the water samples have varied hydrographic conditions. Figure 1. Growth curves from Halo and U-166 denitrifying population. The growth profiles of both DN consortia were analyzed for a period of 24 h at 0D600. The U-166 DN consortia grew at a much faster rate compared to the Halo DN consortia. Growth profile of the DN microbial consortium Growth profile of the DN microbial consortium To determine the growth profile of the DN consortia, the growth curve assay was performed. It was observed that the DN consortia from Halo grew (OD600 = 0.980) much slower than the U-166 consortia (OD600 = 2.448) over the 24 h period (Figure 1; Dataset 1). Thus providing the first evidence that both the DN consortia are different from each other. All analyses were performed in triplicate (Figure 1). 16S amplicon sequencing To determine the DN phylotypes present in the DN consortia, genomic DNA was isolated using DNA isolation solutions I, II and III (bioWORLD, Dublin, OH) from the Halo and U-166 denitrifying microbial consortia after 24 h of growth and 16S amplicon sequencing of the V4 region and bioinformatics analysis Page 3 of 11 Page 3 of 11 Page 3 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Figure 1. Growth curves from Halo and U-166 denitrifying population. The growth profiles of both DN consortia were analyzed for a period of 24 h at 0D600. The U-166 DN consortia grew at a much faster rate compared to the Halo DN consortia. was performed by Seqmatic (Fremont, CA). High throughput NGS was performed using the Illumina MiSeq platform using 2x250bp reads and the FASTQ data was processed using the Qiime pipeline. Denitrifying microbial consortia Both water samples were also analyzed for the presence of denitrifying microbial consortia using the commercially available DN-BART assay. Upon performing the assay, it was observed that Halo and U-166 water samples did not produce foam or bubbles around the ball or in the tube after 4 days, the recommended dura- tion for developing a positive reaction. Hence, due to the nature of the unique water samples, the assay was continued for 30 days. Following the 30 days, foam was detected around the ball, providing evidence for potential DN microbial consortia from the Halo and U-166 shipwreck sites. The DN consortia that were enriched using the DN-BART assay were further analyzed for microbial growth, nitrate/nitrite media concentrations and microbial composition over 24 h. 16S amplicon sequencing Since the growth curve and IC studies indicated that the DN consortia from Halo and U-166 are mutually exclusive, we wanted to determine the microbial composition of both Halo and U-166 DN consortia using 16S amplicon sequencing. The Halo DN consortium primarily consisted of the Pseudomonas genus (98.1%), while the U-166 DN consortium was dominated by the Citrobacter genus (72.6%). At the species level, P. tropicalis and P. aeruginosa for Halo, and C. werkmanii and C. freundii for U-166 were primarily detected (Figure 3; Dataset 3)), thus indicating that both DN consortia are mutually exclusive. Table 1. Hydrographic conditions of water column samples collected proximate to two steel shipwrecks Halo and U-166 in the northern part of Gulf of Mexico. The depth, temperature, salinity and dissolved oxygen (DO) for the water samples collected from Halo and U-166 are listed below. Halo U-166 Depth (m) 141.3 1448.9 Temp (°C) 17.7 4.3 Salinity (PSU) 36.3 35.0 DO (mg/L) 4.1 6.6 Table 1. Hydrographic conditions of water column samples collected proximate to two steel shipwrecks Halo and U-166 in the northern part of Gulf of Mexico. The depth, temperature, salinity and dissolved oxygen (DO) for the water samples collected from Halo and U-166 are listed below. Halo U-166 Depth (m) 141.3 1448.9 Temp (°C) 17.7 4.3 Salinity (PSU) 36.3 35.0 DO (mg/L) 4.1 6.6 Table 1. Hydrographic conditions of water column samples collected proximate to two steel shipwrecks Halo and U-166 in the northern part of Gulf of Mexico. The depth, temperature, salinity and dissolved oxygen (DO) for the water samples collected from Halo and U-166 are listed below. Halo U-166 Depth (m) 141.3 1448.9 Temp (°C) 17.7 4.3 Salinity (PSU) 36.3 35.0 DO (mg/L) 4.1 6.6 Dataset 1. Growth curve assay: Growth curve studies were performed over a 24 h period for both the Halo and U-166 DN consortia http://dx.doi.org/10.5256/f1000research.12713.d179765 The optical density of the cultures were measured every 2 h at 600nm. Dataset 2. Ion chromatography studies: Nitrate and nitrite levels of the DN consortia isolated from Halo and U-166 sites were determined every 2 h for a period of 24 h using ion chromatography http://dx.doi.org/10.5256/f1000research.12713.d179766 Dataset 1. Growth curve assay: Growth curve studies were performed over a 24 h period for both the Halo and U-166 DN consortia http://dx.doi.org/10.5256/f1000research.12713.d179765 The optical density of the cultures were measured every 2 h at 600nm. Nitrate and nitrite ion chromatography analysis Ion chromatography (IC) studies were performed to identify the denitrifying potential of the isolated microbial consortia. The Halo microbial supernatants showed a steady decline in nitrate concentration (734 mg/L to 0.7 mg/L) as microbial growth entered into the logarithmic growth phase. As the nitrate concentration decreased, there was an increase in nitrite concentration from 1.4 mg/L, to a maximum of 130 mg/L and tapered down to 4.3 mg/L at 24 h. The U-166 microbial consortium rapidly converted nitrate into nitrite, as shown with a decrease in nitrate concentration (730 mg/L to 2.5 mg/L) followed by an increase in nitrite concentration (0 to 240 mg/L), which was later followed by a subsequent decrease in nitrite levels to 2.2 mg/L (Figure 2; Dataset 2). Table 1. Hydrographic conditions of water column samples collected proximate to two steel shipwrecks Halo and U-166 in the northern part of Gulf of Mexico. The depth, temperature, salinity and dissolved oxygen (DO) for the water samples collected from Halo and U-166 are listed below. 16S amplicon sequencing 16S amplicon sequencing: Genomic DNA was isolated from DN consortia after 24 h of growth and the V4 region was analyzed using the 16S metagenomics sequencing http://dx.doi.org/10.5256/f1000research.12713.d179767 Dataset 3. 16S amplicon sequencing: Genomic DNA was isolated from DN consortia after 24 h of growth and the V4 region was analyzed using the 16S metagenomics sequencing Dataset 3. 16S amplicon sequencing: Genomic DNA was isolated from DN consortia after 24 h of growth and the V4 region was analyzed using the 16S metagenomics sequencing The water sample from the Halo shipwreck site was able to grow under the conditions set forth in this study, but at a much slower rate when compared to U-166. When the Halo microbial consor- tium began its logarithmic growth, there was a steady decline in the nitrate concentration and a subsequent increase in the nitrite concentration in the supernatant. The U-166 DN consortium also grew and the turbidity of the culture was greater as compared to Halo DN consortium. The IC studies further corroborated the decrease in nitrate levels and a concurrent increase in the nitrite concentration during the logarithmic phase of growth. Further, the U-166 consortia consumed nitrite, most likely as nitrate was completely consumed, at a rate 2 times slower than that observed in the Halo DN consortium (30 mg/L compared to 65 mg/L respectively). To determine the DN phylotypes present in the DN consortia, 16S amplicon sequencing was performed, and it was observed that at the species level P. tropicalis and P. aeruginosa for Halo and C. werkmanii and C. freundii for U-166 were the most dominant and are known denitrifiers9,10. It was also observed that the Citrobacter dominating the U-166 DN consortia consumed nitrate at a rate that was faster than other industrial microbial consortia containing Citrobacter adapted for denitrification11. 16S amplicon sequencing The optical density of the cultures were measured every 2 h at 600nm. Dataset 2. Ion chromatography studies: Nitrate and nitrite levels of the DN consortia isolated from Halo and U-166 sites were determined every 2 h for a period of 24 h using ion chromatography Dataset 2. Ion chromatography studies: Nitrate and nitrite levels of the DN consortia isolated from Halo and U-166 sites were determined every 2 h for a period of 24 h using ion chromatography http://dx.doi.org/10.5256/f1000research.12713.d179766 Page 4 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Figure 2. Ion chromatographic studies of the DN microbial consortium isolated from Halo and U-166 shipwreck sites. Samples were collected every 2 h for a period of 24 h and nitrate and nitrite levels were determined. Note: Nitrate values were above the calibration level (100 mg/L) and are thus estimates (but proportional). Figure 2. Ion chromatographic studies of the DN microbial consortium isolated from Halo and U-166 shipwreck sites. Samples were collected every 2 h for a period of 24 h and nitrate and nitrite levels were determined. Note: Nitrate values were above the calibration level (100 mg/L) and are thus estimates (but proportional). Figure 3. 16S amplicon sequencing on the Halo and U-166 DN consortium after 24 h culturing in Indole Nitrite medium. Genomic DNA was isolated from the Halo and the U-166 DN consortium and the V4 region of the 16S was analyzed. Figure 3. 16S amplicon sequencing on the Halo and U-166 DN consortium after 24 h culturing in Indole Nitrite medium. Genomic DNA was isolated from the Halo and the U-166 DN consortium and the V4 region of the 16S was analyzed. Figure 3. 16S amplicon sequencing on the Halo and U-166 DN consortium after 24 h culturing in Indole was isolated from the Halo and the U-166 DN consortium and the V4 region of the 16S was analyzed. quencing on the Halo and U-166 DN consortium after 24 h culturing in Indole Nitrite medium. Genomic DNA and the U-166 DN consortium and the V4 region of the 16S was analyzed. consortia, the turbidity of the media was monitored and the nitrate/ nitrite concentrations were examined every 2 h over a 24 h time period. Dataset 3. References Haridas D, Biffinger J, Boyd T, et al.: Dataset 3 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179767 1. Rivera-Monroy VH, Lenaker P, Twilley RR, et al.: Denitrification in coastal Louisiana: A spatial assessment and research needs. J Sea Res. 2010; 63(3–4): 157–172. Publisher Full Text 2. Seitzinger S, Harrison JA, Böhlke JK, et al.: Denitrification across landscapes and waterscapes: a synthesis. Ecol Appl. 2006; 16(6): 2064–2090. PubMed Abstract \| Publisher Full Text 3. Hulth S, Aller RC, Canfield DE, et al.: Nitrogen removal in marine environments: recent findings and future research challenges. Mar Chem. 2005; 94(1–4): 125–145. Publisher Full Text 4. Seitzinger SP: Denitrification in freshwater and coastal marine ecosystems: Ecological and geochemical significance. Limnol Oceanogr. 1988; 33(4part2): 702–724. Publisher Full Text 5. Falkowski PG: Evolution of the nitrogen cycle and its influence on the biological sequestration of CO2 in the ocean. Nature. 1997; 387: 272–275. Publisher Full Text 6. Sánchez-Porro C, Kaur B, Mann H, et al.: Halomonas titanicae sp. nov., a halophilic bacterium isolated from the RMS Titanic. Int J Syst Evol Microbiol. 2010; 60(Pt 12): 2768–2774. PubMed Abstract \| Publisher Full Text 7. Perkol-Finkel S, Shashar N, Benayahu Y: Can artificial reefs mimic natural reef communities? The roles of structural features and age. Mar Environ Res. 2006; 61(2): 121–135. PubMed Abstract \| Publisher Full Text 8. Danovaro R, Snelgrove PV, Tyler P: Challenging the paradigms of deep-sea ecology. Trends Ecol Evol. 2014; 29(8): 465–475. PubMed Abstract \| Publisher Full Text 9. Hernandez D, Rowe JJ: Oxygen regulation of nitrate uptake in denitrifying Pseudomonas aeruginosa. Appl Environ Microbiol. 1987; 53(4): 745–750. PubMed Abstract \| Free Full Text 10. Rehr B, Klemme JH: Formate dependent nitrate and nitrite reduction to ammonia by Citrobacter freundii and competition with denitrifying bacteria. Antonie Van Leeuwenhoek. 1989; 56(4): 311–321. PubMed Abstract \| Publisher Full Text 11. Cyplik P, Juzwa W, Marecik R, et al.: Denitrification of industrial wastewater: Influence of glycerol addition on metabolic activity and community shifts in a microbial consortium. Chemosphere. 2013; 93(11): 2823–2831. PubMed Abstract \| Publisher Full Text 12. Haridas D, Biffinger J, Boyd T, et al.: Dataset 1 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179765 13. Data availability Dataset 1: Growth curve assay: Growth curve studies were performed over a 24 h period for both the Halo and U-166 DN consortia. The optical density of the cultures were measured every 2 h at 600nm. DOI, 10.5256/f1000research.12713.d17976512 The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Dataset 2: Ion chromatography studies: Nitrate and nitrite levels of the DN consortia isolated from Halo and U-166 sites were determined every 2 h for a period of 24 h using ion chromatography. DOI, 10.5256/f1000research.12713.d17976613 Acknowledgements We thank the officers, crew and scientific party on board the R/V Pelican during the April-May 2015 expedition (PE15-22). Grant information di id Grant information Funding was provided by the Office of Naval Research (ONR) through the Naval Research Laboratory (PE# 61153N), the Bureau of Ocean Energy Management (BOEM) No. M13PG00020, and the Navy Platform Support Program. Publisher Full Text 2. Seitzinger S, Harrison JA, Böhlke JK, et al.: Denitrification across landscapes and waterscapes: a synthesis. Ecol Appl. 2006; 16(6): 2064–2090. PubMed Abstract \| Publisher Full Text 3. Hulth S, Aller RC, Canfield DE, et al.: Nitrogen removal in marine environments: recent findings and future research challenges. Mar Chem. 2005; 94(1–4): 125–145. 1. Rivera-Monroy VH, Lenaker P, Twilley RR, et al.: Denitrification in coastal Louisiana: A spatial assessment and research needs. J Sea Res. 2010; 63(3–4): 157–172. Publisher Full Text Discussion The deep sea, identified with shelf depths greater than 200 m, has been documented to be the largest hypoxic and anoxic environ- ment present on earth8. The varied living conditions mentioned earlier induces microbes to adopt unique metabolic adaptations. Hence, the marine dark biosphere has been recognized as a rich resource of unique microbial populations. Apart from the unique microbial life detected in the marine dark biosphere, shipwreck sites located in the deep sea also serve as a rich source of dis- tinct flora and fauna6. Using two different shipwrecks at varying depth and material allows for the comparison of the metabolic activity of DN microbial consortia isolated from steel shipwreck sites. One of the biggest challenges in characterizing new microbes from the deep-sea is the ability to successfully culture them in the laboratory. The initial approach to identifying a DN consortium was to assess growth using the commercially available DN-BART assay. The DN-BART assay provided the necessary nutrients in a modified nitrate medium and the presence of a potential DN microbial consortium from both Halo and U-166 shipwreck sites was confirmed. To further characterize the DN microbial This study indicates that Halo and U-166 were good pros- pecting sites for novel microbial consortia related to Page 5 of 11 Page 5 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 denitrification. The water mass passing by each shipwreck site has a distinct DN consortium which can be grown under laboratory settings. The U-166 DN microbial consortium performs denitrification at a much faster rate than the Halo DN microbial consortium and most known industrial microbial consortia. This elevated DN activity could be the result of local hydrodynamic conditions or the proximity to the shipwreck, but additional studies are needed to identify the exact parameters. In conclusion, both DN consortia isolated from novel prospecting sites (shipwrecks) in the Gulf of Mexico can be cultured in the laboratory and can utilize a DN metabolic pathway for growth. Dataset 3: 16S amplicon sequencing: Genomic DNA was isolated from DN consortia after 24 h of growth and the V4 region was analyzed using the 16S metagenomics sequencing. DOI, 10.5256/f1000research.12713.d17976714 Competing interests No competing interests were disclosed. Competing interests No competing interests were disclosed. References Haridas D, Biffinger J, Boyd T, et al.: Dataset 2 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179766 14. Haridas D, Biffinger J, Boyd T, et al.: Dataset 3 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179767 References 1. Rivera-Monroy VH, Lenaker P, Twilley RR, et al.: Denitrification in coastal Louisiana: A spatial assessment and research needs. J Sea Res. 2010; 63(3–4): 157–172. Publisher Full Text 2. Seitzinger S, Harrison JA, Böhlke JK, et al.: Denitrification across landscapes and waterscapes: a synthesis. Ecol Appl. 2006; 16(6): 2064–2090. PubMed Abstract \| Publisher Full Text 3. Hulth S, Aller RC, Canfield DE, et al.: Nitrogen removal in marine environments: recent findings and future research challenges. Mar Chem. 2005; 94(1–4): 125–145. Publisher Full Text 4. Seitzinger SP: Denitrification in freshwater and coastal marine ecosystems: Ecological and geochemical significance. Limnol Oceanogr. 1988; 33(4part2): 702–724. Publisher Full Text 5. Falkowski PG: Evolution of the nitrogen cycle and its influence on the biological sequestration of CO2 in the ocean. Nature. 1997; 387: 272–275. Publisher Full Text 6. Sánchez-Porro C, Kaur B, Mann H, et al.: Halomonas titanicae sp. nov., a halophilic bacterium isolated from the RMS Titanic. Int J Syst Evol Microbiol. 2010; 60(Pt 12): 2768–2774. PubMed Abstract \| Publisher Full Text 7. Perkol-Finkel S, Shashar N, Benayahu Y: Can artificial reefs mimic natural reef communities? The roles of structural features and age. Mar Environ Res. 2006; 61(2): 121–135. PubMed Abstract \| Publisher Full Text 8. Danovaro R, Snelgrove PV, Tyler P: Challenging the paradigms of deep-sea ecology. Trends Ecol Evol. 2014; 29(8): 465–475. PubMed Abstract \| Publisher Full Text 9. Hernandez D, Rowe JJ: Oxygen regulation of nitrate uptake in denitrifying Pseudomonas aeruginosa. Appl Environ Microbiol. 1987; 53(4): 745–750. PubMed Abstract \| Free Full Text 10. Rehr B, Klemme JH: Formate dependent nitrate and nitrite reduction to ammonia by Citrobacter freundii and competition with denitrifying bacteria. Antonie Van Leeuwenhoek. 1989; 56(4): 311–321. PubMed Abstract \| Publisher Full Text 11. Cyplik P, Juzwa W, Marecik R, et al.: Denitrification of industrial wastewater: Influence of glycerol addition on metabolic activity and community shifts in a microbial consortium. Chemosphere. 2013; 93(11): 2823–2831. PubMed Abstract \| Publisher Full Text 12. Haridas D, Biffinger J, Boyd T, et al.: Dataset 1 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179765 13. Haridas D, Biffinger J, Boyd T, et al.: Dataset 2 in: Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico. F1000Research. 2017. http://www.doi.org/10.5256/f1000research.12713.d179766 14. Open Peer Review Current Referee Status: Current Referee Status: No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. Version 2 27 July 2018 Referee Report doi:10.5256/f1000research.16633.r34921 doi:10.5256/f1000research.16633.r34921 Publisher Full Text Page 6 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Dimitri Kalenitchenko Université Laval, Quebec City, QC, Canada The revised version of "Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico", allows a better interpretation of the data thanks to the new mat and met section. I still have concern about the fact that the method used by the author can't really test that each shipwreck has a particular DN consortium. I can understand that samples are difficult to obtain but I would recommend to change a little bit the last paragraph to temper your words. For example instead of saying "Each shipwreck site" say "The Water mass passing by each shipwreck site". It is a small difference but I think a reader that only read the last section of the discussion would think that the shipwreck itself has its own DN consortium. It might be true but it is not what you test for. No competing interests were disclosed. Competing Interests: Version 3 09 August 2018 Referee Report doi:10.5256/f1000research.17320.r36863 Dimitri Kalenitchenko Université Laval, Quebec City, QC, Canada The manuscript version that has been submitted answered all the comments made on previous version. No competing interests were disclosed. Competing Interests: Author Response 03 Aug 2018 Dimitri Kalenitchenko Université Laval, Quebec City, QC, Canada The paper entitled “Laboratory growth of denitrifying water column microbial consortia from deep-sea shipwrecks in the northern Gulf of Mexico” by Haridas . explores the potential of steel shipwrecks to et al host a specialized community with a focus on denitrifying bacteria. Overall the paper is well written and easy to follow, their results are relevant but before suggesting the acceptance of the manuscript, I suggest that authors resubmit a modified version of the manuscript that include a detailed material and method section. 16S amplicon sequencing section : 16S amplicon sequencing section : Please provide the DNA extraction kit details Please provide the name of the sequencing platform and primers Please explain the bioinformatics treatment including the taxonomic affiliation used Denitrifying microbial consortia : Denitrifying microbial consortia : Why did the authors extend the incubation time up to 30 days, I am wondering if they checked for contamination from external source. 16S amplicon sequencing : 16S amplicon sequencing : I am very surprised about the very low diversity they obtained, I agree that they enriched the community in DN bacteria but with only two tubes for two conditions I am not convinced they do not just select for a contaminant or a bacterium in each tube. How did the authors obtained these percentages, BLAST ? p g , US Naval Research Laboratory, USA Dhanya Haridas , US Naval Research Laboratory, USA Dhanya Haridas , US Naval Research Laboratory, USA Dhanya Haridas Dear Reviewer, Thank you for taking the time and reviewing the paper again. We have included the edit suggested by you in the discussion. Respectfully, Dhanya Haridas Dear Reviewer, Thank you for taking the time and reviewing the paper again. We have included the edit suggested by you in the discussion. Respectfully, Dhanya Haridas No competing interests were disclosed. Competing Interests: Version 1 29 March 2018 Referee Report doi:10.5256/f1000research.13771.r31968 No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. Author Response 03 Aug 2018 , US Naval Research Laboratory, USA Dhanya Haridas Page 7 of 11 I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. Author Response 03 Aug 2018 Page 7 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Author Response 03 Aug 2018 , US Naval Research Laboratory, USA Dhanya Haridas Dear Reviewer, Thank you for taking the time and reviewing the paper again. We have included the edit suggested by you in the discussion. Respectfully, Dhanya Haridas No competing interests were disclosed. Competing Interests: Author Response 03 Aug 2018 , US Naval Research Laboratory, USA Dhanya Haridas Dear Reviewer, Thank you for taking the time and reviewing the paper again. We have included the edit suggested by you in the discussion. Respectfully, Dhanya Haridas No competing interests were disclosed. Competing Interests: Are the conclusions drawn adequately supported by the results? Partly No competing interests were disclosed. Competing Interests: Author Response 06 Jun 2018 Author Response 06 Jun 2018 , US Naval Research Laboratory, USA Dhanya Haridas No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Discussion I think the conclusion that shipwreck are good prospecting area for novel microbial consortia is highly speculative based on these data especially without a reference point before the water mass flow across the wreck. Page 8 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Partly Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Not applicable Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Partly No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confirm t it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Author Response 06 Jun 2018 , US Naval Research Laboratory, USA Dhanya Haridas Dear Reviewer, Thank you for taking the time to review and suggest comments/edits to the manuscript. We have included information about the DNA isolation kit, the sequencing platform on which the 16S metagenomics studies were performed and the bioinformatics pipeline used. We hope the reviewer will find the responses satisfactory. You expressed your concern on the need for us to extend the incubation time from 4 days to 30 days for the BART assay We appreciate the reviewer raising this concern and the reason we Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Partly Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Not applicable Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Partly No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confir it is of an acceptable scientific standard, however I have significant reservations, as outlin above. , US Naval Research Laboratory, USA Dhanya Haridas , US Naval Research Laboratory, USA Dhanya Haridas Dear Reviewer, Thank you for taking the time to review and suggest comments/edits to the manuscript. We have included information about the DNA isolation kit, the sequencing platform on which the 16S metagenomics studies were performed and the bioinformatics pipeline used. We hope the reviewer will find the responses satisfactory. You expressed your concern on the need for us to extend the incubation time from 4 days to 30 days for the BART assay. We appreciate the reviewer raising this concern and the reason we decided to extend the incubation time was due to the uncharacterized marine environmental samples used and it is known the concentration of viable microbes is less (as determined by a 3H-Leucine incorporation assay) than when grown in a rich medium. Hence extending the incubation time in this study resulted in an increased concentration of the viable microbial culture which helped determine if the assay was positive for microbes which could perform denitrification. You mentioned another concern about the study selecting for a contaminant. The raw data (excel file) containing all the sequencing tags obtained from the 16S metagenomics sequencing has been provided. The authors would like to mention that the microbial population obtained from the BART assay served as the starting inoculum for all the assays mentioned in the paper and therefore any potential aerobic or facultative aerobic bacteria cultured in the BART assay medium would not have propagated in the Nunc tube due to the anaerobic nature of the environment, thus further diminishing the diversity and potential for an aerobic contaminant. The percentages were part of the bioinformatics report submitted to us by Seqmatic. For example, in the Halo sample the total number of reads at the species level was 169,918 of which 59,117 reads were classified as Page 9 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 (By Seqmatic) thus making it 34.72% of the total species population. Pseudomonas tropicalis We hope we addressed all your concerns in this response and thank you once again for reviewing the manuscript. Respectively, Dhanya Haridas No competing interests were disclosed. Competing Interests: 26 March 2018 Referee Report doi:10.5256/f1000research.13771.r31967 doi:10.5256/f1000research.13771.r31967 Maxim Rubin Blum 1,2 Department of Symbiosis, Max Planck Institute for Marine Microbiology , Bremen, G 1 Israel Oceanographic and Limnological Research, Haifa, Israel 2 Israel Oceanographic and Limnological Research, Haifa, Israel 2 Israel Oceanographic and Limnological Research, Haifa, Israel 2 This study aims to examine the cultivable denitrifying microbial consortia from marine environment near deep-sea shipwrecks in the Gulf of Mexico. The authors hypothesized that the shipwrecks could influence the microbial population involved in denitrification. The methods used in this study, however, have not adequately tested this hypothesis, as no control samples were obtained. On the other hand, the authors were able to report enrichment of two interesting bacterial consortia, one of which appeared to perform denitrification at a much faster rate than that measured for most known industrial microbial consortia, providing an opportunity to study these bacteria in more detail in the future. If I understood correctly, only a single water sample from each location was used for the denitrifying Biological Activity Reaction Test, allowing to isolate a denitrifying consortium per sampling site. It is likely that these bacteria occur in a patchy manner in the environment, thus further sampling may reveal very different denitrifying consortia. Broadening the sampling effort is needed to test if either environmental factors or presence of shipwrecks affects the diversity of denitrifying bacteria. Is the work clearly and accurately presented and does it cite the current literature? Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Partly Are sufficient details of methods and analysis provided to allow replication by others? Yes If applicable, is the statistical analysis and its interpretation appropriate? Yes Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes N ti i t t di l d C ti I t t Is the study design appropriate and is the work technically sound? P l Page 10 of 11 F1000Research 2018, 6:1834 Last updated: 09 AUG 2018 No competing interests were disclosed. Competing Interests: Referee Expertise: Marine Microbiology I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. 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 No competing interests were disclosed. Competing Interests: Referee Expertise: Marine Microbiology I have read this submission. I believe that I have an appropriate level of expertise to confirm tha it is of an acceptable scientific standard. 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 Page 11 of 11
https://openalex.org/W4293073231	https://jes.utm.md/wp-content/uploads/sites/20/2022/06/JES-2-2022_123-132.pdf	English	null	UPDATING THE STATISTICAL REGISTER OF HOUSING IN THE REPUBLIC OF MOLDOVA USING OPEN-SOURCE GIS TECHNOLOGIES	Journal of Engineering Science	2,022	cc-by	4,062	ournal of Engineering Science Vol. XXIX, no. 2 (2022), pp. 123 - 132 ascicle Architecture, Civil and Environmental Engineering ISSN 2587-3474 Topic Civil Engineering and Management eISSN 2587-3482 ttps://doi.org/10.52326/jes.utm.2022.29(2).12 DC 332.87:311.21:04.6+312] (478) UPDATING THE STATISTICAL REGISTER OF HOUSING IN THE REPUBLIC OF MOLDOVA USING OPEN-SOURCE GIS TECHNOLOGIES Mihail Spataru1, ORCID ID: 0000-0003-2547-6171, Ana Vlasenco2, ORCID ID: 0000-0002-5017-6392, Livia Nistor-Lopatenco2, ORCID ID: 0000-0003-3509-648X, Vasile Grama2, ORCID ID: 0000-0002-2100-2756 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md Received: 02. 18. 2022 A t d 04 23 2022 ournal of Engineering Science Vol. XXIX, no. 2 (2022), pp. 123 - 132 ascicle Architecture, Civil and Environmental Engineering ISSN 2587-3474 Topic Civil Engineering and Management eISSN 2587-3482 ttps://doi.org/10.52326/jes.utm.2022.29(2).12 DC 332.87:311.21:04.6+312] (478) UPDATING THE STATISTICAL REGISTER OF HOUSING IN THE REPUBLIC OF MOLDOVA USING OPEN-SOURCE GIS TECHNOLOGIES Mihail Spataru1, ORCID ID: 0000-0003-2547-6171, Ana Vlasenco2, ORCID ID: 0000-0002-5017-6392, Livia Nistor-Lopatenco2, ORCID ID: 0000-0003-3509-648X, Vasile Grama2, ORCID ID: 0000-0002-2100-2756 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md Received: 02. 18. 2022 A t d 04 23 2022 ournal of Engineering Science Vol. XXIX, no. 2 (2022), pp. 123 - 132 ascicle Architecture, Civil and Environmental Engineering ISSN 2587-3474 Topic Civil Engineering and Management eISSN 2587-3482 ttps://doi.org/10.52326/jes.utm.2022.29(2).12 DC 332.87:311.21:04.6+312] (478) UPDATING THE STATISTICAL REGISTER OF HOUSING IN THE REPUBLIC OF MOLDOVA USING OPEN-SOURCE GIS TECHNOLOGIES Mihail Spataru1, ORCID ID: 0000-0003-2547-6171, Ana Vlasenco2, ORCID ID: 0000-0002-5017-6392, Livia Nistor-Lopatenco2, ORCID ID: 0000-0003-3509-648X, Vasile Grama2, ORCID ID: 0000-0002-2100-2756 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md Received: 02. 18. 2022 Accepted: 04. 23. 2022 Vol. XXIX, no. 2 (2022), pp. 123 - 132 g ISSN 2587-3474 eISSN 2587-3482 Journal of Engineering Science Civil Engineering and Management Topic https://doi.org/10.52326/jes.utm.2022.29(2).12 UDC 332.87:311.21:04.6+312] (478) 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md Received: 02. 18. 2022 Accepted: 04. 23. 2022 Received: 02. 18. 2022 Accepted: 04. 23. 2022 Abstract. This article describes the possibilities of promoting and implementing open-source geoinformation technologies in the field of statistical activity in the Republic of Moldova, which will simplify and streamline the process of analysis, collection, systematization of data, as well as their dissemination. Introduction Numerous real-world examples, with clear explanations of the methodology, demonstrate the usefulness of geographic information systems (GIS) for solving practical problems in various fields of activity [1, 2]. In the field of statistics, high quality data are of particular importance for a contemporary country. They are essential for solid decision-making, which will affect the daily lives of citizens. The implementation of geo-informational technologies in statistics will simplify and streamline the data collection process as well as their dissemination [3, 4]. Given that the update of the Statistical Register of Housing (SRH) in the Republic of Moldova is a very important database for the future census in 2023, the use of GIS technologies, especially open-source, are welcome. SRH is a database of all dwellings, regardless of the form of ownership, including specialized houses (dormitories, boarding houses, penitentiaries and others), apartments, service rooms and other houses in other useful constructions for living. Until 2019 there was no such register, an alternative used being the Lists of Buildings and Housing (LBH), prepared by the Local Public Authorities (LPA) together with the Territorial Statistics Offices (TSOs). These lists were difficult to use because not all dwellings had an address, and even if they did, they were not standardized [5]. According to the recommendations of the United Nations, our country has proposed the creation of the geo-coded Statistical Register of Housing (each dwelling must have x and y coordinates). In this sense, in order to achieve the proposed goal, the following objectives have been outlined: In this sense, in order to achieve the proposed goal, the following objectives have been outlined: − analysis, collection and systematization of housing data, from official sources of administrative data; − presenting the structure of the SRH layers and explaining the variables; − choosing and establishing methods for collecting and storing data in a database, using the open-source applications QGIS [6,7] and QField [8]; − populating, validating and processing housing data in SRH. Cuvinte cheie: Registrul Statistic al Locuințelor (SRH), Sisteme Informaționale Geografice (GIS), Recensământul Populației și Locuințelor (PHC), QGIS, QField. Cuvinte cheie: Registrul Statistic al Locuințelor (SRH), Sisteme Informaționale Geografice (GIS), Recensământul Populației și Locuințelor (PHC), QGIS, QField. Cuvinte cheie: Registrul Statistic al Locuințelor (SRH), Sisteme Informaționale Geografice (GIS), Recensământul Populației și Locuințelor (PHC), QGIS, QField. ournal of Engineering Science Vol. XXIX, no. 2 (2022), pp. 123 - 132 ascicle Architecture, Civil and Environmental Engineering ISSN 2587-3474 Topic Civil Engineering and Management eISSN 2587-3482 ttps://doi.org/10.52326/jes.utm.2022.29(2).12 DC 332.87:311.21:04.6+312] (478) UPDATING THE STATISTICAL REGISTER OF HOUSING IN THE REPUBLIC OF MOLDOVA USING OPEN-SOURCE GIS TECHNOLOGIES Mihail Spataru1, ORCID ID: 0000-0003-2547-6171, Ana Vlasenco2, ORCID ID: 0000-0002-5017-6392, Livia Nistor-Lopatenco2, ORCID ID: 0000-0003-3509-648X, Vasile Grama2, ORCID ID: 0000-0002-2100-2756 1National Bureau of Statistcis of the Republic of Moldova, 106 Grenoble street, Chisinau, Republic of Moldova 2Technical University of Moldova, 168 Stefan cel Mare Boulevard, Chisinau, Republic of Moldova Corresponding author: Ana Vlasenco, ana.vlasenco@gcg.utm.md Received: 02. 18. 2022 A t d 04 23 2022 Since 2019, the National Bureau of Statistics of the Republic of Moldova, according to the recommendations of the United Nations, has proposed the creation of the geo-coded Statistical Register of Housing (SRH). With the creation of the National Spatial Data Infrastructure (NSDI), this goal can be achieved, as SRH will form the basis of the Population and Housing Census (PHC) from the 2020 round. The geospatial data from SRH will be used to delimit the census enumeration areas, and the reviewed population will be attached to the geographical coordinates of the dwelling. These enumeration areas will cover the entire territory of the Republic of Moldova and will serve to plan human resources, organize field activities and ensure the full coverage of census units (for people, households and dwellings). Keywords: Statistical Register of Housing (SRH), Geographic Information Systems (GIS), Population and Housing Census (PHC), QGIS, QField. Keywords: Statistical Register of Housing (SRH), Geographic Information Systems (GIS), Population and Housing Census (PHC), QGIS, QField. (GIS), Rezumat. Articolul descrie posibilitățile de promovare și implementare a tehnologiilor de geoinformație open-source în domeniul activității statistice în Republica Moldova, care vor simplifica și eficientiza procesul de analiză, colectare, sistematizare a datelor, precum și difuzarea acestora. Din 2019, Biroul Național de Statistică al Republicii Moldova, conform recomandărilor Națiunilor Unite, a propus crearea Registrului Statistic al Locuințelor (SRH) geocodat. Odată cu crearea Infrastructurii Naționale de Date Spațiale (NSDI), acest obiectiv poate fi atins, întrucât SSR va sta la baza Recensământului Populației și Locuințelor (PHC) din runda 2020. Datele geospațiale de la SRH vor fi utilizate pentru delimitarea zonelor de enumerare a recensământului, iar populația revizuită va fi atașată la coordonatele geografice ale locuinței. Aceste zone de enumerare vor acoperi întreg teritoriul Republicii Moldova și vor servi pentru planificarea resurselor umane, organizarea activităților de teren și June, 2022, Vol. XXIX (2) Journal of Engineering Science pdating the statistical register of housing in the Republic of Moldova using Open-Source GIS technologie 124 asigurarea acoperirii integrale a unităților de recensământ (pentru persoane, gospodării și locuințe). Journal of Engineering Science Data collected and structuring of SRH layers When updating the SRH for the study area were served the following sources of eospatial data: When updating the SRH for the study area were served the following sources of geospatial data: Real Estate Register (RER), managed by the Public Services Agency (ASP) [9]; − Real Estate Register (RER), managed by the Public Services Agency (ASP) [9]; − State Register of Administrative-Territorial Units and Addresses (SRATUA), Public Services Agency (ASP) [9]; − State Register of Administrative-Territorial Units and Addresses (SRATUA), Public Services Agency (ASP) [9]; − Linear Map 2017, State Enterprise Institute of Geodesy, Engineering Research and Cadastre "INGEOCAD" under the coordination of the Agency for Land Relations and Cadastre of the Republic of Moldova (ARFC) [10]; − Orthoimages (Orthophoto) 2007/2016/2020, State Enterprise Institute of Geodesy, Engineering Research and Cadastre "INGEOCAD" under the coordination of the Agency for Land Relations and Cadastre of the Republic of Moldova (ARFC) [10]; June, 2022, Vol. XXIX (2) June, 2022, Vol. XXIX (2) Journal of Engineering Science M. Spataru, A. Vlasenco, L. Nistor-Lopatenco, V. Grama 125 For areas that were not covered by the geospatial data listed above or whose source is obsolete, field data were collected. For areas that were not covered by the geospatial data listed above or whose source is obsolete, field data were collected. The database that will contain the homes which we are going to update will be of the SpatiaLite type [11] and will be called “ActRSL_Ialoveni”. It will consist of the “Buildings” layers, of point type (Table 1) and “Dwellings”, of table type (Table 2) [5, 12]. Table 1 Structure of the "Buildings" layer No. Name Type Description Values 1 Commune Text, 150 char. Commune name 2 Locality Text, 150 char. Locality name 3 ID_ea Text, 12 char. Enumeration area code Address data 4 ID_Entry Integer, 3 char. Building entry ID 5 CODE_ Building Integer, 3 char. Building code 6 JID_ Entry Text, 20 char. Building-entrance code 7 Street_ type Text, 15 char. Street type 1) Street; 2) Stradella; 3) Boulevard; 4) Road; 5) City Square; 6) Blind alley; 7) Other. 8 Street Text, 150 char. Street name 9 Number Integer, 3 char. Building number 10 Slash Text, 10 char. Slash 11 Stair Integre, 2 char. Stair number Building data 12 Build_Cond. Text, 150 char. Journal of Engineering Science Data collected and structuring of SRH layers Building condition 1) With all structural components 2) Under (re)construction 3) Damaged / ruined 4) Demolished Structure of the "Buildings" layer June, 2022, Vol. XXIX (2) 126 Updating the statistical register of housing in the Republic of Moldova using Open-Source GIS technologies 126 Continued Table 1 13 Build_Use. Text, 150 char. Building use 1) For residential purposes only 2) Residential and other purpose 3) Non-residential 14 Build_Type. Text, 150 char. Building type 1) Detached house 2) Row house 3) Apartment building 4) Collective living quarter 5) Accessory building for the main building 6) Industrial 7) Commercial 8) Educational 9) Health 10) Institutional 11) Others 15 Nr_ Floors Integer, 2 char. Total number of floors in the building 16 Nr_ Dwell. Integer, 3 char. Total number of dwellings in the building 17 List_ Dwell. Text, 254 char. Dwelling list 18 Notes Text, 254 char. Notes Table 2 Structure of the "Dwellings" layer No. Name Type Description Values 1 ID_ Dwell. Integer, 3 char. Dwelling ID 2 CODE_ Entry Text, 20 char. Building-entrance code 3 Floor Integer, 2 char. Floor number 4 Apartment Text, 10 char. Apartment number 5 DirectAccess Text, 2 char. Presence of a direct access from outside the building 1) Yes 2) No Journal of Engineering Science M. Spataru, A. Vlasenco, L. Nistor-Lopatenco, V. Grama 127 Continued Table 2 6 Type Dwelling Text, 15 char. Dwelling type 1) Conventional dwelling 2) Unconventional dwelling 3) Collective living quarter 4) Business unit 5) Unit for Institutional or cultural use 6) Unit used as accessory of the building 7) Ruined unit 8) Others 9) Non-existent 7 Notes Text, 254 char. Notes Continued Table 2 6 Type Dwelling Text, 15 char. Dwelling type 1) Conventional dwelling 2) Unconventional dwelling 3) Collective living quarter 4) Business unit 5) Unit for Institutional or cultural use 6) Unit used as accessory of the building 7) Ruined unit 8) Others 9) Non-existent 7 Notes Text, 254 char. Notes 6 7 Notes Text, 254 char. Notes Results - Case study Results - Case study For the case study, two pilot enumeration area from the Republic of Moldova were chosen, namely from Ialoveni district: − 552000002061 from the village of Milestii Mici [13] (Figure 1 (a)), where it took place massive primary real estate registration, and in the Population and Housing Census (PHC-2014) 126 dwellings were identified [14]; − 552000002061 from the village of Milestii Mici [13] (Figure 1 (a)), where it took place massive primary real estate registration, and in the Population and Housing Census (PHC-2014) 126 dwellings were identified [14]; [ ] − 552300004023 from the village of Pojareni [13] (Figure 1 (b)), where it did not take place massive primary real estate registration, and at PHC-2014, 87 dwellings were identified [14]. − 552300004023 from the village of Pojareni [13] (Figure 1 (b)), where it did not take place massive primary real estate registration, and at PHC-2014, 87 dwellings were identified [14]. b) a) b) Figure 1. Excerpt from the enumeration area boundaries: a) 552000002061 Milestii Mici; b) 552300004023 Pojareni Source: [5]. a) b) ) ) Figure 1. Excerpt from the enumeration area boundaries: a) 552000002061 Milestii Mici; b) 552300004023 Pojaren Source: [5]. The process of updating the SRH will include three stages: population of the register; data validation; data processing. June, 2022, Vol. XXIX (2) 128 Updating the statistical register of housing in the Republic of Moldova using Open-Source GIS technologies Population of the register. Population of the register. p f g the first stage, using the open-source application QGIS, SRH was populated based istrative data and orthophoto images as follows: In the first stage, using the open-source application QGIS, SRH was populated based on administrative data and orthophoto images, as follows: on administrative data and orthophoto images, as follows: − Pilot enumeration area of the Population and Housing Census (PHC) - 2014, defined the work area [14]; − Pilot enumeration area of the Population and Housing Census (PHC) - 2014, defined the work area [14]; − Address point of the building from SRATUA, was used to complete the variables fo the address; − Land and buildings from Real Estate Register, were needed to define the boundaries between the properties and to identify the buildings and their use; − Roof Area from the Linear Map 2017, completed the Buildings layer, especially in the areas where it did not take place massive primary real estate registration; − Orthophoto images have completed the Buildings layer and Roof Area, especially in high-speed construction areas. The database in QGIS will contain the dwellings to be updated and will be named "ActRSL_Ialoveni", which will consist of the Buildings layer, of point type and Dwellings of table type (Tables 1 and 2). Following this stage, there were obtained a total of 917 buildings: 613 in the enumeration area 552000002061 from Milestii Mici (Figure 2 (a)) and 304 in the enumeration area 552300004023 from Pojareni (Figure 2 (b)). Following this stage, there were obtained a total of 917 buildings: 613 in the enumeration area 552000002061 from Milestii Mici (Figure 2 (a)) and 304 in the enumeration area 552300004023 from Pojareni (Figure 2 (b)). a) a) b) Figure 2. Excerpt from the distribution and classification of buildings after the population of the SRH from administrative data and orthophoto images: a) 552000002061 Milestii Mici; b) 552300004023 Pojareni Source: [5]. b) a) Figure 2. Excerpt from the distribution and classification of buildings after the population of the SRH from administrative data and orthophoto images: a) 552000002061 Milestii Mici; b) 552300004023 Pojareni Source: [5]. In the village of Pojareni 97% of the buildings are with all the structural components, of which 38% are used only for residential purposes and another 59% for non-residential purposes. 100% of the residential buildings are of the individual house type, and in the June, 2022, Vol. Population of the register. XXIX (2) Journal of Engineering Science 129 M. Spataru, A. Vlasenco, L. Nistor-Lopatenco, V. Grama category of non-residential buildings 53% are auxiliary buildings of the household and another 40% are buildings without capital walls. In the village of Milestii Mici 94% of the buildings are with all the structural components, of which 38% are used only for residential purposes and another 56% for non- residential purposes. 99% of the residential buildings are of the individual house type, and in the category of non-residential buildings 62% are auxiliary buildings of the household and another 32% are buildings without capital walls. Data validation. Given that administrative data, used in the SRH population, is not always current and it is necessary to go to the field and validate the data obtained directly from the source. The QField open-source mobile application was used to accomplish this step. The QField application uses the same graphical editing controls as the QGIS desktop [8]. To create a project in the QField application, we will use the "QFieldSync" plugin, which will help us to prepare and package QGIS projects for QField. The QField project will be based on the QGIS project "ActRSL_Ialoveni" In this case, a copy of this project will be created with the name "ActRSL_Ialoveni_QField", where only the necessary layers from the field will be kept. Thus, in the QGIS platform, this project was created for data collection in the QField application (Figure 3). Figure 3. Creating the project for collecting data from the field. Source: [5]. Figure 3. Creating the project for collecting data from the field. Source: [5]. This project is exported and uploaded to the mobile device via a USB. To open the project in the QField application, will be performed the following steps from Figure 4: (1) „Open local file”; (2) choosing the location of the file; (3) opening the project with the terminations "qgs". With the help of the tools for enlarging, reducing and finding the user's location, the pilot area will be navigated. So, when we are in the field, we will visit each building and check the correctness of the classification using the QField application (Figure 5). To view the data of a building, is accessed the geometry of the building (1), and from the "Features" window (2) is selected from the list the desired building, based on the code "pkuid" (3). Journal of Engineering Science Population of the register. To edit the existing data, go to the "Enable editing" button (4), and from the list (5) you will find the move buttons "Move Feature" (6), move duplication "Duplicate Feature" (7) or June, 2022, Vol. XXIX (2) Journal of Engineering Science ing the statistical register of housing in the Republic of Moldova using Open-Source GIS technologies 130 delete the object "Delete Feature" (8). And by accessing the check mark (9) we will save the created changes. Figure 4. Opening of the "ActRSL_Ialoveni_QField" project. Source: [5] Figure 4. Opening of the "ActRSL_Ialoveni_QField" project. Source: [5] Figure 5. Visualization, editing and saved an object in QField. Source: [5]. Figure 4. Opening of the "ActRSL_Ialoveni_QField" project. Source: [5] Figure 5. Visualization, editing and saved an object in QField. Source: [5]. Figure 5. Visualization, editing and saved an object in QField. Source: [5]. Data processing. After the field verification, we move on to the third stage, that of data processing. The data processing step involves downloading the data collected in the field, from the mobile device to the computer, with their repeated verification and uploading as a final variant in the SRH database. Following the data analysis, were obtained a total of 916 buildings: 615 in the 552000002061 enumeration area in Milestii Mici (Figure 6 (a)) and 301 in the 552300004023 enumeration area in Pojareni (Figure 6 (b)). In Pojareni village 92% of the buildings are with all the structural components, of which 39% are used only for residential purposes and another 53% for non-residential purposes. 100% of the residential buildings are of the individual house type, and in the category of non-residential buildings 54% are auxiliary buildings of the household and June, 2022, Vol. XXIX (2) M. Spataru, A. Vlasenco, L. Nistor-Lopatenco, V. Grama 131 another 40% are buildings without capital walls. At the same time, 1% of the buildings are in (re) construction, 6% are damaged and 1% are demolished. a) a) b) Figure 6. Excerpt from the distribution and classification of buildings according to the field data validation process: a) 552000002061 Milestii Mici; b) 552300004023 Pojareni. Source: [5]. b) a) ) ) Figure 6. Excerpt from the distribution and classification of buildings according to the field data validation process: a) 552000002061 Milestii Mici; b) 552300004023 Pojareni. Source: [5]. Population of the register. In the village of Milestii Mici, 93% of the buildings are with all the structural components, of which 37% are used only for residential purposes and another 55% for non- residential purposes. 99% of the residential buildings are of the individual house type, and in the category of non-residential buildings 74% are auxiliary buildings of the household and another 20% are buildings without capital walls. At the same time, 4% of the buildings are in (re) construction, 2% are damaged and 2% are demolished. Conclusions Comparing the number of dwellings registered at PHC-2014 with those collected in this paper, there is an increase of 26% in the case of Pojareni village, from 87 dwellings registered at PHC-2014 to 118 recently identified. And in the case of the village of Milestii Mici, an increase of 45%, from 126 dwellings in the PHC-2014 to 229 recently identified. The increase in the number of dwellings is largely influenced by two factors: the first factor would be the spatial mismatch of the dwellings in the census enumeration area in the spatial base with those in the database collected at the census. And the second reason is due to a positive dynamics of various constructions in the localities near the municipality of Chisinau. The implementation of open-source GIS technologies are recommended in the field of statistical activity in the Republic of Moldova. Their use will lead to an increase in the quality of data collected and disseminated, will reduce the team involved in data collection and monitoring, while reducing the cost of statistical work. June, 2022, Vol. XXIX (2) Journal of Engineering Science Journal of Engineering Science 132 Updating the statistical register of housing in the Republic of Moldova using Open-Source GIS technologies Journal of Engineering Science References 1. Peter A. Burrough, Rachael A. Mcdonnel, Cristopher D. LLOYD. Principles of Greogrphical Information Systems, Oxford, University Press, 1998. 2. Davis D. E. GIS for everyone. Third Edition, ESRI Press, Redlands, California, 2003, 152 p. 3. Spătaru M., Iacovlev A. Reflectarea distribuției grafice a populației pe localități în proces și la finalizarea recensământului populației. In: Materials of the Technical-Scientific Conference of Students, Master and Doctoral Students TUM March 23-25 2021 Chisinau Vol II ISBN 978-9975-45-701-9 pp 35-38 Doctoral Students TUM, March 23-25, 2021, Chisinau, Vol. II. ISBN 978-9975-45-701-9, pp. 3 4. Grama V., Dilan V., Cepoi E. Tehnologii geoinformaționale avansate. TUM, 2013, 260 p. 5. Spătaru M. Utilizarea tehnologiilor GIS open-source pentru actualizarea Registrului Statistic al Locuințelor din Republica Moldova. Master Thesis. Chisinau, TUM, 2022. p 6. QGIS User Guide. [online], [access 14.12.2021]. Available: https://docs.qgis.org/2.18/pdf/en/QGIS-2.18- UserGuide-en.pdf. 6 7. User Manual - QGIS. Version 2.0. [online], [access 16.12.2021]. Available: http://adrnord.md/public/files /publication /GIS_manual_dhv_final.pdf 8. QField Documentation. [online], [access 20.12.2021]. Available: https://qfield.org/docs/ro/ 9. Geoserver. [online], [access 15.10.2021]. Available: https://moldova-map.md/geoserver/web/ 9. Geoserver. [online], [access 15.10.2021]. Available: https://moldova-map.md/geoserver/web/ 10. Geoportal INDS. [online], [access 15.10.2021]. Available: http://www.geoportalinds.gov.md/ 11. SpatiaLite. [online], [access 21.12.2021]. Available: https://www.gaia-gis.it/fossil/libspatialite/ 11. SpatiaLite. [online], [access 21.12.2021]. Available: https://www.gaia-gis.it/fossil/libspatialite/home 12. Galer L. Manual metodologic privind efectuarea procedurii de actualizare a hărților (colectare date georeferențiate 12. Galer L. Manual metodologic privind efectuarea procedurii de actualizare a hărților (colectare date g referitor la clădiri și locuințe). Chisinau: NBS, 2021. 13. National Bureau of Statistics of the Republic of Moldova: Clasificări şi nomenclatoare. [online], [access 31.10.2021]. Available: https://statistica.gov.md/pageview.php?l=ro&idc=385 13. National Bureau of Statistics of the Republic of Moldova: Clasificări şi nomenclatoare. [online], [access 31.10.2021]. Available: https://statistica.gov.md/pageview.php?l=ro&idc=385 14. National Bureau of Statistics of the Republic of Moldova: Solicitare de informații statistice. [online], [access 31.10.2021]. Available: https://statistica.gov.md/solicitare_informatii_statistice.php?l=ro 14. National Bureau of Statistics of the Republic of Moldova: Solicitare de informații statistice. [online], [access 31.10.2021]. Available: https://statistica.gov.md/solicitare_informatii_statistice.php?l=ro June, 2022, Vol. XXIX (2)
https://openalex.org/W3188628170	https://www.ijhpm.com/article_4086_535e9b6e476ce5075b491ec19a0970fb.pdf	English	null	Unbalanced Treatment Costs of Breast Cancer in China: Implications From the Direct Costs of Inpatient and Outpatient Care in Liaoning Province	International journal of health policy and management	2,021	cc-by	7,637	Article History: Received: 28 October 2020 Accepted: 23 June 2021 ePublished: 1 August 2021 Implications for policy makers p p y • This study provides a comprehensive description of breast cancer healthcare cost based on the System of Health Accounts 2011 (SHA 2011), which will better understand the disease burden on the healthcare system, facilitate a more meaningful discussion of cost containment and allocation of resources. • This study provides a comprehensive description of breast cancer healthcare cost based on the System of Health Accounts 2011 (SHA 2011), which will better understand the disease burden on the healthcare system, facilitate a more meaningful discussion of cost containment and allocation of resources. • In 2017, the curative care expenditure (CCE) for breast cancer was about 830.19 million CNY in Liaoning province (0.7% of the total health expenditure and 9.9% of cancer-related healthcare costs), with out-of-pocket (OOP) payments accounting for 36.4% of total breast cancer care cost. • In 2017, the curative care expenditure (CCE) for breast cancer was about 830.19 million CNY in Liaoning province (0.7% of the total health expenditure and 9.9% of cancer-related healthcare costs), with out-of-pocket (OOP) payments accounting for 36.4% of total breast cancer care cost. Inpatient cost was the main cost driver (74.4%), almost three times as large as outpatient costs (25.6%). • Inpatient cost was the main cost driver (74.4%), almost three times as large as outpatient costs (25.6%). https://ijhpm.com Int J Health Policy Manag 2022, 11(9), 1735–1743 https://ijhpm.com Int J Health Policy Manag 2022, 11(9), 1735–1743 doi 10.34172/ijhpm.2021.75 Unbalanced Treatment Costs of Breast Cancer in China: Implications From the Direct Costs of Inpatient and Outpatient Care in Liaoning Province Unbalanced Treatment Costs of Breast Cancer in China: Implications From the Direct Costs of Inpatient and Outpatient Care in Liaoning Province Zihua Ma1 ID , Gongman Deng2, Zhaolin Meng3, Yanan Ma4* ID , Huazhang Wu1* ID Zihua Ma1 ID , Gongman Deng2, Zhaolin Meng3, Yanan Ma4* ID , Huazhang Wu1* ID *Correspondence to: Yanan Ma Email: ynma@cmu.edu.cn Huazhang Wu Email: hzwu@cmu.edu.cn Abstract This is an open-access article b d d h f h C C A b (h // /l / py g ( ); y y p distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/ by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. p p y Citation: Ma Z, Deng G, Meng Z, Ma Y, Wu H. Unbalanced treatment costs of breast cancer in China: implications from the direct costs of inpatient and outpatient care in Liaoning province. Int J Health Policy Manag. 2022;11(9):1735– 1743. doi:10.34172/ijhpm.2021.75 Key Messages Implications for policy makers Abstract Background: The increasing incidence of breast cancer and its financial burden highlights the need for controlling treatment costs. This study aimed to assess the direct costs of inpatient and outpatient care for breast cancer patients in Liaoning Province to provide a policy reference for cost containment. Methods: Based on the System of Health Accounts 2011 (SHA 2011), systematic data collection was conducted via multistage stratified cluster random sampling. A total of 1160 health institutions, including 83 hospitals, 16 public health institutions, 120 primary health institutions, and 941 outpatient institutions were enrolled in 2017. A database was established containing 20 035 patient-level medical records from the information system of these institutions. Curative care expenditure (CCE)was calculated, and generalized linear modeling was performed to determine cost-related factors. Results: In 2017, the CCE for breast cancer was approximately CNY 830.19 million (US$122.96 million) in Liaoning province (0.7% of the total health expenditure and 9.9% of cancer-related healthcare costs). Inpatient care costs were estimated to be CNY 617.27 million (US$91.42 million), accounting for 74.4% of the CCE for breast cancer, almost three times as large as outpatient costs (25.6%). The average inpatient and outpatient costs for breast cancer were estimated to be CNY 12 108 (US$1793) and CNY 829 (US$123) per visit. Medication cost was the main cost driver, which comprised 84.0% of the average outpatient cost and 37.2% of the mean inpatient cost. Conclusion: Breast cancer imposes a large economic burden on patients and the social health insurance system. Results show an irrational cost pattern of inpatient and outpatient services, with patients relying excessively on inpatient services for treatment. Promoting outpatient care whenever relevant is conducive to cost containment and rational utilization of resources. Keywords: Breast Cancer, Healthcare Expenditure, Medical Service Utilization, Cost Containment, China Keywords: Breast Cancer, Healthcare Expenditure, Medical Service Utilization, Cost Containment, China Copyright: © 2022 The Author(s); Published by Kerman University of Medical Sciences. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/ by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Keywords: Breast Cancer, Healthcare Expenditure, Medical Service Utilization, Cost Containment, China Copyright: © 2022 The Author(s); Published by Kerman University of Medical Sciences. This is an open-access article opyright: © 2022 The Author(s); Published by Kerman University of Medical Sciences. Data Sources h For this population-based study, we selected patients in Liaoning Province as the sample population, and the macro data and demographic data were derived from official statistics and field investigations, respectively. There were about 44 million inhabitants and the gross domestic product was 234 090 billion in Liaoning Province. The total health expenditure was extracted from the Liaoning Health Statistical Yearbook (2017) and Liaoning Health Financial Yearbook (2017). The case data were gathered from sample institutions, which were selected via a multistage stratified cluster random sampling approach. In the first stage, sample cities were selected from Liaoning Province based on the economic development level and population density, and Shenyang, Dalian, Fushun, Jinzhou, Panjin, and Tieling were identified and selected as appropriate cities. Figure 1 indicated the locations of the province and cities where the study was done. In the second stage, one district and two counties were selected from each included prefecture-level city based on the quality of available data and the completeness of their health information systems. After determining the sample areas, in the third stage, a random sample was selected, ensuring that a representative sample of health agencies and hospital levels were included. A total of 1160 health institutions, including 83 hospitals, 16 public health institutions, 120 primary health institutions, and 941 outpatient institutions were enrolled. A database was established containing patient-level medical records from the information system of these institutions, g The process of transforming cancer care from inpatient to outpatient settings has progressed far in many developed countries and produces significant savings in treatment.4 However, due to the policies to expand insurance coverage in China, the utilization of patient medical services has been greatly promoted, especially for inpatient services. The inpatient utilization rate has increased rapidly from 8.7% in 2008 to 17.6% in 2017.5 As a result, China has a higher inpatient utilization rate compared to the global rate (0.14 vs. 0.10 admissions per capita), while the outpatient utilization rate is still lower than the global rate (5.17 vs. 5.42 visits per capita).6 Since different patient service utilization patterns have a significant impact on the treatment cost of breast cancer, an analysis of inpatient and outpatient treatment costs associated with breast cancer will facilitate a more meaningful discussion of cost containment and allocation of resources. Ma et al on the healthcare system: the provider interface, the consumer health interface, and the financing interface.15 International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1736 Background Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death among women worldwide. Globally, there were about 2.1 million cases of newly diagnosed female breast cancer in 2018, accounting for almost a quarter of cancer cases among women.1 In China, breast cancer is the most frequently diagnosed cancer (19.2%) and the fourth leading cause of cancer-related deaths among women (9.1%).2 The incidence of breast cancer among Chinese women has increased more than twice as fast as global rates. In 2018, with 367 900 new cases of invasive breast cancer and 97 972 related deaths, China accounted for 17.6% of global cases and 15.6% of related deaths, which is much higher than the proportion in 2008 (12.2% and 9.6%, respectively).2,3 Due to the high prevalence of breast cancer and the rapid increase in incidence, the costs related to the treatment of breast cancer will comprise a larger proportion of healthcare costs. Trends toward greater intensity of healthcare service use also lead to a greater economic burden of cancer in the future. This study aimed to estimate the curative care expenditure (CCE) for breast cancer in Liaoning Province, Northeast China, using the SHA 2011 framework, and assess the average inpatient and outpatient costs per visit based on the sample data. Combining the results of macro and micro data, we further evaluated the distribution of CCE and the cost patterns of inpatient and outpatient services to provide a policy reference for cost containment in China. Implications for the public Providing affordable healthcare services so that anyone can get the required services is in accordance with the core idea of universal health coverage and the agenda of the Sustainable Development Goals. We are looking for options. This study indicates that there is an irrational cost pattern of inpatient and outpatient services in China, with patients relying excessively on inpatient services for treatment. This may increase the financial burden on patients and prolong the length of hospital stay. Thus, recommendations from international experience are provided to promote the transfer of cancer care from inpatient to outpatient settings whenever appropriate, which is conducive to cost containment and rational utilization of resources. Full list of authors’ affiliations is available at the end of the article. Full list of authors’ affiliations is available at the end of the article. Ma et al Data Sources h Several studies have assessed the direct medical costs of breast cancer in China,7-9 but to our knowledge none have studied the cost patterns of inpatient and outpatient services for breast cancer patients. Moreover, these studies are limited in that they focus merely on costs for inpatients or patients in specialized hospitals and lack the support of a systematic accounting framework and representative samples, thus decreasing the accuracy of expenditure calculation results and comparability with different countries. Figure 1. Locations of the Province and Cities Where the Study Was Done. Measuring expenditure on breast cancer in a comparative and standard manner is a critical first step in understanding the economic burden of breast cancer. The System of Health Accounts 2011 (SHA 2011) is the most widely used approach for estimating health expenditures.10-13 It was revised and launched by several international organizations including the Organization for Economic Cooperation and Development, Eurostat, and the World Health Organization (WHO) and commonly used as the global standard in guiding systematic data collection or routine tracking health expenditures, which makes the calculated result comparable across countries.14 Furthermore, analyses based on the SHA 2011 can provide more comprehensive insights on tracking health expenditure through three interfaces that help explain the disease burdens Figure 1. Locations of the Province and Cities Where the Study Was Done. Ma et al expenses was calculated from the sample data. Finally, the CI was calculated after deducting the total prevention expenses from the total hospital income. The formula for CI was as follows: expenses was calculated from the sample data. Finally, the CI was calculated after deducting the total prevention expenses from the total hospital income. The formula for CI was as follows: including age, sex, date of hospital visit, primary diagnosis, length of stay (LOS), insurance type, and inpatient and outpatient costs. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1737 Measurement Estimating CCE for Breast Cancer Measurement Estimating CCE for Breast Cancer i p The outpatient basic expenditure subsidy (OBES) was also calculated in 3 steps. First, the total BES, the total number of inpatient bed days (b), and the total number of outpatient visits (c) were extracted from the statistical yearbook and financial annual reports. Second, to separate preventive outpatient visits from the total outpatient visits (c), the total number of curative outpatient visits (d) was calculated based on the number of preventative outpatient visits (e) and total outpatient visits (f) in the sample data. Third, the hospitalization sharing coefficient (β) was calculated based on the total number of inpatient bed days (b) and the total number of curative outpatient visits (d) from the sample data. Finally, the OBES was calculated after deducting the inpatient BESs from the total BESs. The formulas for the BES are as follows: Based on the SHA 2011, a top-down allocation method was used to calculate the CCE for breast cancer in Liaoning province. After extracting the total amount of current health expenditures from the macro data, the sharing coefficient calculated with the data from the field investigation was used to allocate the total health expenditures into various dimensions (such as age and institution type) and calculate the provincial breast cancer CCE. Age was defined in increments of 5-year periods, and institution type was defined as general hospitals, specialized hospitals, or traditional Chinese medicine (TCM) hospitals. In China, there are two major compensation sources for public hospitals, which are government subsidies and service charges. The government’s subsidies for public hospitals mainly includes two types: recurrent investment and special item investment. The recurrent investment is to subsidize the medical service, mainly determined according to the number of hospital staff and the number of beds provided, while the special item investment is to subsidize infrastructure, scientific research and the purchase of large-scale equipment. Therefore, the calculation of CCE mainly includes the regular investment of the government subsidies-that is, basic expenditure subsidies (BES), and the service charges of hospitals—that is, curative income (CI). The specific calculation formula is as follows: SOBES = STOBES × (1 – β) b b d K β = + ∗ Considering the conversion relationship between the workload of a physician for 1 hospital day and 1 outpatient visit, the value of K is set to 0.1 according to the recommendation of China National Health Development Research Center. Study Sample 1 p OCI TOMI S S α α   = −     1 p OCI TOMI S S α α   = −     The study population comprised all patients with primary diagnosis as breast cancer of the facilities included in the study within the Liaoning province between January 1 and December 31, 2017. The International Classification of Disease Tenth Revision (ICD-10) was used to identify primary breast cancer, coded as C50. A total of 23 014 breast cancer- related medical records were selected from the database. After excluding records with missing values, suspected errors, and outliers (with the total cost below the 1st percentile or above the 99th percentile of the total cost distribution for inpatient and outpatient care), the final valid sample size for the overall breast cancer records was 20 035. And CI for a certain dimension was calculated according to the following formula: 1( n i OCI OCI i p a S S α α = ′ = × − ∑ In the above two formulas, CI per visit from the sample data is denoted as ai, and total income of the sample hospitals is denoted as a. Moreover, the sample’s total preventive costs, denoted as ap, was calculated by adding all samples that contained preventive services based on the ICD-10 codes. Thus, the sharing coefficient for each patient is denoted by ai /(a-ap). Measurement Estimating CCE for Breast Cancer Outpatient costs included prescription medication, imaging, laboratory examinations, therapy, surgery, materials, registration, and consultation fees. Curative Care Expenditure for Breast Cancer Curative Care Expenditure for Breast Cancer p In 2017, the CCE for breast cancer was about CNY 830.19 million (US$122.96 million) in Liaoning province, which accounted for 0.7% of the total health-care expenditure and 9.9% of cancer-related healthcare costs. Inpatient care costs were estimated to be CNY 617.27 million (US$91.42 million), accounting for 74.4% of the CCE for breast cancer. Outpatient care costs were calculated to be CNY 212.92 million (US$31.54 million), accounting for 25.6% of the CCE for breast cancer. Of the three types of financing schemes, the government financing scheme provided the most, with CNY 429.28 million (51.7%). Of this, most funds came from basic social health insurance (41.9%), and government subsidies only accounted for 9.8%. The OOP payment for breast cancer patients was CNY 302.12 million (US$44.75 million), which accounted for 36.4% of the total CCE among breast cancer patients. Regarding different types of institutions, over 60% of the expenditure on treatment for breast cancer was spent in general hospitals, followed by specialist hospitals (22.8%) and TCM hospitals (10.6%). Following the descriptive analysis, the Mann-Whitney rank-sum test was used to determine the significance of observed differences in costs, and the chi-square statistic was used to compare proportions. To determine the cost- related factors of inpatient and outpatient costs, generalized linear modeling with a gamma distribution and a log link was performed. The generalized linear modeling model accounts for skewness in the distribution of medical costs without requiring retransformation of the results from the log scale back to levels.19 The cost-related factors in this study included age, institution type, insurance type, LOS, and surgery. Cost estimates that were different from zero at the 95% confidence level were considered to be statistically significant. All expense data were reported in Chinese yuan (CNY). The exchange rate between US dollar and CNY was: US$1.00 = CNY 6.7518 in 2017. The analyses were conducted using STATA 14.0 and SPSS 22.0 statistical software. Similar to the total CCE pattern, the money paid by social health insurance and OOP payments went more to inpatient services (73.0%), whereas the proportion paid by Table 1. Measurement Estimating CCE for Breast Cancer (1 ) e d c f = × − Th i f The outpatient BESs for a certain dimension were calculated according to the following formula: SCCE = SOCI + SOBES + SICI + SIBES 1 ( ) n i OBES OBES i p S S α α α ′ = = × − ∑ In the above formula, SOCI and SOBES represent the CI and BES for outpatient service, and SICI and SIBES represent the CI and BES for inpatient service. We also calculated the CCE by three types of financing schemes: government financing schemes (eg, government subsidies, basic social health insurance), voluntary financing schemes (eg, non-government social insurance contributions, commercial health insurance payments, and donations from non-governmental organizations), and out-of-pocket (OOP) payments.17 Since the calculation method of inpatient CCE is similar to that of outpatient CCE, we take outpatient CCE as an example to introduce the specific measures of accounting. To calculate the outpatient CI of the breast cancer, first, the total outpatient medical institution income was extracted through the summary of relevant data in the statistical yearbook and financial annual reports. Second, the proportion of prevention 1737 Ma et al Estimating Average Cost Per Inpatient and Outpatient Visit for and 32.2% in general hospitals. Moreover, the proportion of self-paying outpatients was high (20.1%), which is much higher than the proportion of self-paying inpatients (7.3%). Table 1 shows the characteristics of inpatient admissions and outpatient visits. Breast Cancer The charges of each inpatient admission and outpatient visit associated with breast cancer were calculated based on the sample data from the field investigation. Descriptive statistics were used to determine demographic variables, healthcare resource utilization, and costs, including frequencies, percentage, means and standard deviation (SD), median, and 95% confidence intervals. Age at diagnosis was divided into 4 groups (<40, 40-54, 55-69, and ≥70 years) based on the age range of the target population for cancer screening in China (between 40 and 69 years of age).18 Moreover, patients’ insurance status was divided into four categories: self-paid and three main social health insurance schemes, including the Urban Employee Basic Medical Insurance (UEBMI), the Urban Resident Basic Medical Insurance (URBMI), and the New Cooperative Medical Scheme (NCMs). The inpatient costs consisted of the following components: prescription medication, imaging, laboratory examinations, therapy, surgery, materials, beds, diagnoses, and nursing care costs. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1738 Measurement Estimating CCE for Breast Cancer Characteristics of Breast Cancer Inpatient Admission and Outpatient Visit Variable Inpatient Admission Outpatient Visit (n = 7416) (n = 12 619) Age, years (Mean, SD) 54.2 (10.4) 58.1 (11.0) Age group (n, %) <40 582 (7.8) 657 (5.2) 40-54 3322 (44.8) 3770 (29.9) 55-69 3040 (41.0) 6565 (52.0) ≥70 472 (6.4) 1627 (12.9) LOS (Mean, SD) 9.8 (8.3) NA Surgery 2704 (36.5) 88 (0.7) Payment type (n, %) UEBMI 2561 (34.5) 2641 (20.9) URBMI 2726 (36.8) 7170 (56.8) NCMs 1589 (21.4) 277 (2.2) Self-paid 540 (7.3) 2531 (20.1) Hospital type (n, %) TCM hospital 1257 (16.9) 648 (5.1) Specialized hospital 1144 (15.4) 7913 (62.7) General hospital 5015 (67.6) 4058 (32.2) Abbreviations: SD, standard deviation; UEBMI, Urban Employee Basic Medical Insurance; URBMI, Urban Resident Basic Medical Insurance; NCMs, New Rural Cooperative Medical Scheme; LOS, length of stay; TCM, Traditional Chinese Medicine. Table 1. Characteristics of Breast Cancer Inpatient Admission and Outpatient Visit Characteristics of Inpatient and Outpatient Visits Characteristics of Inpatient and Outpatient Visits A total of 7416 inpatient admissions associated with breast cancer were included in the analyses. The mean ± SD age of the patients was 54.2 ± 10.4 years, and most of the admitted patients were 40-69 years old (85.8%). The average LOS was 9.8 ± 8.3 days. A total of 2704 patients (36.5%) underwent surgery. Most patients (92.7%) were supported by social health insurance. Patients were mainly admitted to general hospitals (67.6%), followed by TCM hospitals (16.9%) and specialist hospitals (15.4%). A total of 12 619 outpatient visits associated with breast cancer were assessed. The mean ± SD age of the patients was 58.1 ± 11.0 years. Similar to the age distribution of inpatients, 81.9% of outpatients were 40-69 years old. However, in contrast to the institution distribution of inpatients, 62.7% of cases were diagnosed or treated in specialized hospitals, Abbreviations: SD, standard deviation; UEBMI, Urban Employee Basic Medical Insurance; URBMI, Urban Resident Basic Medical Insurance; NCMs, New Rural Cooperative Medical Scheme; LOS, length of stay; TCM, Traditional Chinese Medicine. Abbreviations: SD, standard deviation; UEBMI, Urban Employee Basic Medical Insurance; URBMI, Urban Resident Basic Medical Insurance; NCMs, New Rural Cooperative Medical Scheme; LOS, length of stay; TCM, Traditional Chinese Medicine. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 38 Ma et al government subsidies was directed slightly less to inpatient care (57.0%). Among different types of hospitals, the share of inpatient and outpatient care costs was slightly different. TCM hospitals and general hospitals spent significantly more on inpatient services (82.2% vs. 77.9%), while inpatient and outpatient costs in specialized hospitals were more balanced (59.9% vs. 40.1%). Figure 2 shows how the CCE for breast cancer was simultaneously split by financing schemes, service use (inpatient admissions and outpatient visits), and type of institution. and 60-64 years old, accounting for 17.5% and 17.8% of the total CCE, respectively. The breast cancer CCE for inpatient services was mostly from people aged 50–54 years, accounting for 18.3% of the total inpatient CCE. The 60-64 year old age group accounted for the highest proportion (18.6%) of the total outpatient CCE. In addition, inpatient CCE always accounted for the majority of the total costs in all age groups (over 50.0%), but as age increased, its proportion gradually decreased. Figure 3 displays the age distribution of breast cancer CCE for inpatient admissions and outpatient visits. Characteristics of Inpatient and Outpatient Visits The CCE for breast cancer was dominated by the population aged 40-69 years, accounting for 82.9%, with a small portion of costs among patients under 40 years or over 70 years. The total CCE for breast cancer, including outpatient and inpatient CCEs, all showed a trend of rising first and then falling with age. The age distribution had two peaks, at 50-54 years old Charges From Each Inpatient Admission and Outpatient Visit Resource utilization for both inpatient care and outpatient care per visit is presented in Table 2. The average inpatient and outpatient costs for breast cancer were estimated to be CNY 12 108 (US$1793) and CNY 829 (US$123) per visit, respectively. The average inpatient cost per admission was about 14.6 times the average outpatient cost per visit. The main cost drivers for inpatient cost were medication Figure 2. CCE Associated With Breast Cancer by Financing Scheme, Service Function, and Type of Institution. Abbreviation: CCE, Curative Care Expenditure. Government subsidies Social basic health insurance Voluntary financing scheme Out-of-pocket payments Outpatient care lnpatient care General hospital Specialist hospital TCM hospital Financing scheme Service function Type of institution 387.22 302.12 617.27 212.92 552.69 189.48 Figure 2. CCE Associated With Breast Cancer by Financing Scheme, Service Function, and Type of Institution. Abbreviation: CCE, Curative Care Expenditure. Figure 3. The Age Distribution of Breast CCE for Inpatient Admissions and Outpatient Visits. Abbreviation: CCE, Curative Care Expenditure. Figure 3. The Age Distribution of Breast CCE for Inpatient Admissions and Outpatient Visits. Abbreviation: CCE, Curative Care Expenditure. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1739 Ma et al Table 2. Cost Category of Mean Inpatient and Outpatient Cost Per Visit (CNY) Table 2. Characteristics of Inpatient and Outpatient Visits Cost Category of Mean Inpatient and Outpatient Cost Per Visit (CNY) Cost Category Inpatient Outpatient Mean ± SD Percentage of Cost (%) Mean ± SD Percentage of Cost (%) Total 12108 ± 10629 100.0 829 ± 505 100.0 Medication 4502 ± 4326 37.2 696 ± 545 84.0 Imaging 1605 ± 2356 13.3 58 ± 198 7.0 Laboratory test 1029 ± 899 8.5 25 ± 105 3.0 Therapy 826 ± 1893 6.8 12 ± 114 1.4 Surgery 1209 ± 2036 10.0 3 ± 37 0.4 Material 2312 ± 4647 19.1 20 ± 73 2.4 Bed 335 ± 630 2.8 NA NA Diagnose 111 ± 133 0.9 NA NA Nursing 180 ± 228 1.5 NA NA Registration NA NA 10 ± 43 1.2 Consultation NA NA 5 ± 49 0.6 Abbreviations: CNY, Chinese Yuan; SD, standard deviation. The exchange rate between US dollar and CNY was: US$1.00 = CNY 6.7518 in 2017. cases (C00-C97) in China, corresponding a share of 8.6%. Moreover, the high prevalence of breast cancer and its rapid increase in incidence have brought a heavy economic burden to the health insurance system and individual patients. Our results show that the funds for breast cancer treatment mainly came from basic social health insurance (46.0%) and OOP payments (36.4%), with little voluntary financing and government subsidies. Despite China’s commitment and progress in reducing the medical financial burden on households and ensuring access to healthcare, our results show that the financial burden on the individual for breast cancer is still much higher than the government’s preset goal, which is to reduce the percentage of the OOP payments in total health expenditure to below 28% by 2020.17 Moreover, the mean inpatient cost per visit for breast cancer was 12 108 CNY, accounting for 46.6% of the average annual disposable income per capital in Liaoning in 2017, further suggesting that the financial burden on the individual for breast cancer care is excessive. Relying too much on OOP payments may result in catastrophic health expenditure and impoverishment from medical expenses. Therefore, relevant policies are urgently needed to control the growth of breast cancer costs and reduce the personal medical burden of patients. expense and material consumption, which accounted for 37.2% and 19.1%, respectively, while the proportions for surgical treatment (10.0%) and therapy expenses (6.8%) were relatively low. Characteristics of Inpatient and Outpatient Visits The major source of expenses for outpatient services was medication, which contributed to the bulk of the total outpatient costs (84.0%), while other cost categories only comprised small portions of the total costs. Predictors of Inpatient and Outpatient Costs The factors related to inpatient and outpatient costs per visit for breast cancer, are presented in Supplementary file 1. For inpatient services, univariate analysis shows that age, surgery, insurance type, and hospital type are highly correlated the cost per admission (all P < .001, Table S1). The multivariable analysis adjusted for these factors indicates that young age, undergoing surgery, having insurance, and admission in a TCM hospital were independently associated with high costs (Table S2). Moreover, both univariate and multivariate analyses showed that patients supported by insurance with better benefits such as UEBMI (with higher funding criteria and reimbursement rates) had higher costs. Slightly different from inpatient costs, although the factors related to the outpatient costs per visit were age, insurance status, and hospital type (all P < .001), surgery was non- significant in the univariate analysis (Table S1). Furthermore, multivariate analysis showed that older age, visit to a specialist or TCM hospital, and having insurance (excluding NCMs) were independently associated with high costs (Table S2). Our results show that of the total CCE for breast cancer, inpatient cost was the main cost driver (74.4%). These results are different from those reported in other countries, which show that the major cost component of breast cancer care is outpatient cost, accounting for 71% of the total costs in the United States and 62% in France.21-23 The relatively high proportion of inpatient CCE may indicate irrational utilization patterns of inpatient and outpatient services for breast cancer care, and patients relying too much on inpatient services. There are two possible reasons for the over- utilization of inpatient services. First, this may be related to the reimbursement system of medical insurance. According to Zhang et al, China’s medical insurance schemes are mainly intended to provide financial protection for inpatients, while the coverage of outpatient services is very limited or not covered at all.24 Our results show that the proportion of self-paid outpatients was significantly higher than that of International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1740 Discussion Our study provides much-needed data on the scale and distribution of CCEs associated with breast cancer. We found that the direct expenditure of breast cancer is burdensome and cost patterns of inpatient and outpatient services are irrational compared to those in developed countries. The CCE for breast cancer in Liaoning Province in 2017 accounted for 9.9% of cancer-related healthcare costs, which is a higher percentage than the national average level. According to Global Cancer Statistics 2018, breast cancer incidence cases account for 367 900 cases out of 4 285 033 cancer-related International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 40 Ma et al self-paid inpatients (20.1% vs. 7.3%). As patients mainly rely on inpatient services for reimbursement, patients are more inclined to choose inpatient services to receive treatment, thus further exacerbating the imbalance between the proportion of outpatients and inpatients. Second, this may be related to the relatively smaller number of service items provided in outpatient settings. Our results showed that medication expenses comprised about 84.0% of the outpatient costs, which suggests that the service function of outpatient care is relatively simple, mainly for prescription drugs, while more complex treatment services are generally obtained through hospitalization. between Chinese and American cancer survivors confirmed this point. The study showed that Chinese cancer survivors are more likely than American cancer survivors to retire earlier than planned (37% vs. 9%) and reported more hindrance to work abilities due to cancer treatment (84% vs. 21%).30 With outpatient treatment, patients can return to work more easily during treatment with fewer side effects.4 Therefore, the transition from inpatient to outpatient treatment should be strengthened, especially for younger patients. Our results show that the CCE for breast cancer is mainly spent in general hospitals (66.6%), other than specialist hospitals (22.8%), which may be associated with the insufficient and uneven distribution of specialist medical resources. Although more targeted in cancer treatment, specialized hospitals have limited beds and are mostly concentrated in provincial capitals. Additionally, our results show that inpatient and outpatient costs in specialized hospitals are more balanced (59.9% and 40.1%) compared to general hospitals and TCM hospitals. One of the possible reasons is that the inpatient reception capacity is limited by the number of beds in a specialist hospital, so many non-essential inpatients are transferred to outpatient treatment. Discussion Conversely, outpatient reception capacity is not restricted by the number of beds, so specialist hospitals can receive more outpatients, which is verified by the fact that a higher percentage of outpatients visit specialist hospitals (62.7%), and a higher percentage of inpatients visit general hospitals (67.6%) (Table 1). The effective use of outpatient services for specialist hospitals is conducive to controlling the total medical expenses in the region. Therefore, as the main provider of breast cancer care, the transfer of inpatient services to outpatient settings should be focused on general hospitals to control the total cost. Moreover, the average hospital stay for breast cancer in China (9.8 days) is much longer compared to the United States and many European Union countries, with an average LOS of 5.7 days and 4.2-5.0 days, respectively.25,26 The magnitude of inpatient costs and longer hospital stays highlight the importance of considering shifting the medical services from inpatient to outpatient settings. Chemotherapy is one of the main treatments for breast cancer. In China, approximately 81.4% of breast cancer patients receive chemotherapy, and most of them are treated in an inpatient setting,3 which is different from in many developed countries. Recently, outpatient chemotherapy has become a relatively popular international practice. Patient-centered daily chemotherapy wards have been widely implemented in many countries to improve the efficiency of medical resources and reduce costs. A previous study showed that the transition of chemotherapy to an outpatient setting led to a greater than 20% reduction in inpatient oncology costs.27 The benefits of this shift also include reducing the need for inpatient medical resources, decreasing hospital stays, reducing infection rates, and improving quality of life.25 According to Joo et al, patients who begin chemotherapy administration in an outpatient setting report greater satisfaction with their treatment compared with those treated in an inpatient setting.28 Therefore, outpatient chemotherapy and surgery should be promoted to facilitate more non-essential hospitalization patients transferring to outpatient settings, thereby reducing the economic burden of patients. g We found that several factors influence the health expenditure of patients based on the univariate and multivariate analyses, including age, surgery, LOS, type of hospital, and insurance status. The impact of some factors on costs is difficult to control through policies, while the impact of certain factors, such as the LOS and insurance status, can be regulated. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1741 Discussion As in other studies,31 we found that costs were significantly higher for patients supported by health insurance than self-paying ones. In addition, patients under an insurance scheme with higher funding criteria and reimbursement rates (such as UEBMI) had higher costs. The reason may be related to the fact that medical insurance promotes healthcare utilization. Based on the fee-for-service payment, both doctors and patients are motivated to use more medical services, thus resulting in higher healthcare costs. Therefore, China has launched a series of reforms by moving from fee-for-service to diagnosis-related groups, which was initially proven to be effective in reducing hospital expenditures and LOS.32 Furthermore, the diagnosis-related group policy can effectively control hospitalization expenses and promote the transfer of medical services to outpatient settings. Therefore, the wider application of diagnosis-related groups should be strengthened. We found that breast cancer costs were concentrated among people aged 40–69, which is mainly related to the age of onset of breast cancer. Although inpatient costs consistently contributed to the majority of the total costs in all age groups, younger patients had a higher proportion of resource consumption for inpatient services than the older groups. Furthermore, per capita, the cost was significantly higher for younger inpatients and older outpatients. These results may indicate that younger patients are more dependent on inpatient services for aggressive treatment than older patients. However, excessive hospitalization services could impose a higher financial burden and lead to more loss of working time. In China, the median age of breast cancer diagnosis is much younger than in western countries, with over half of cases diagnosed before 50 years of age.3,29 This means that the majority of women who are diagnosed are still of working age, which may suggest that the loss of productivity in China is greater than that in western countries. A comparative study This study had several limitations. First, since the individual outpatient and inpatient records were reported by the sample hospital with the patient identification anonymized, we could 1741 Ma et al not determine the number of cancer patients, the number of outpatient visits and inpatients admissions per patient in 2017, and the annual per capital cost of patients with breast cancer. Second, our database had no further diagnostic information, and we could not determine the cancer stage of the patients. References 1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394-424. doi:10.3322/caac.21492 2. Feng RM, Zong YN, Cao SM, Xu RH. Current cancer situation in China: good or bad news from the 2018 Global Cancer Statistics? Cancer Commun (Lond). 2019;39(1):22. doi:10.1186/s40880-019-0368-6 3. Fan L, Strasser-Weippl K, Li JJ, et al. Breast cancer in China. Lancet Oncol. 2014;15(7):e279-289. doi:10.1016/s1470-2045(13)70567-9 4. Hofmarcher T, Lindgren P, Wilking N, Jönsson B. The cost of cancer in Europe 2018. Eur J Cancer. 2020;129:41-49. doi:10.1016/j. ejca.2020.01.011 5. The 2017 Statistical Bulletin of China’s Health Development. National Health Commission of the People’s Republic of China website. http:// www.nhc.gov.cn/guihuaxxs/s10743/201806/44e3cdfe11fa4c7f928c879d4 35b6a18.shtml. Accessed October 23, 2020. 5. The 2017 Statistical Bulletin of China’s Health Development. National Health Commission of the People’s Republic of China website. http:// www.nhc.gov.cn/guihuaxxs/s10743/201806/44e3cdfe11fa4c7f928c879d4 35b6a18.shtml. Accessed October 23, 2020. 6. Moses MW, Pedroza P, Baral R, et al. Funding and services needed to achieve universal health coverage: applications of global, regional, and national estimates of utilisation of outpatient visits and inpatient admissions from 1990 to 2016, and unit costs from 1995 to 2016. Lancet Public Health. 2019;4(1):e49-e73. doi:10.1016/s2468-2667(18)30213-5 Funding This paper was supported by the Education Department of Liaoning Province under Grant No: LZDR201701. 18. Guo LW, Huang HY, Shi JF, et al. Medical expenditure for esophageal cancer in China: a 10-year multicenter retrospective survey (2002-2011). Chin J Cancer. 2017;36(1):73. doi:10.1186/s40880-017-0242-3 Competing interests Authors declare that they have no competing interests. 14. Maeda A, Harrit M, Mabuchi S, Siadat B, Nagpal S. Creating Evidence for Better Health Financing Decisions: A Strategic Guide for the Institutionalization of National Health Accounts. Washington, DC: The World Bank; 2012. Authors’ contributions The contributions made by the individual authors are as follows. ZHM, YNM, and HZW designed the study; HZW coordinated the data collection; ZHM and GMD conducted data analysis and interpretation; ZHM wrote the first draft; ZLM, YNM, and HZW provided data analysis recommendations and revised the ﬁnal manuscript. All authors read and approved the manuscript. 15. Mueller M, Morgan D. New insights into health financing: first results of the international data collection under the System of Health Accounts 2011 framework. Health Policy. 2017;121(7):764-769. doi:10.1016/j. healthpol.2017.04.008 Disclaimer 16. He J, Yin Z, Duan W, Wang Y, Wang X. Factors of hospitalization expenditure of the genitourinary system diseases in the aged based on “System of Health Account 2011” and neural network model. J Glob Health. 2018;8(2):020504. doi:10.7189/jogh.08.020504 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. 17. Fu W, Zhao S, Zhang Y, Chai P, Goss J. Research in health policy making in China: out-of-pocket payments in Healthy China 2030. BMJ. 2018;360:k234. doi:10.1136/bmj.k234 International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1742 Conclusion Breast cancer imposes a significant economic burden on both patients and the social health insurance system in Liaoning Province. In addition, there are irrational cost patterns of inpatient and outpatient service use, with an overreliance on inpatient services. Specific policies are needed to promote the transfer of cancer care from inpatient to outpatient settings, which is conducive to cost containment and rational utilization of resources. Lessons learned, recommendations from international experience, and evidence from the systematic data can inform decisions about the allocation of resources to service provision and support policy-makers in reducing the financial burden on society and patients in China and other countries with similar situations. 7. Lu JJ, Li HA, Xiong Y. Breast cancer inpatients undergoing mastectomy from a hospital in Guangzhou, China: a retrospective analysis 2004- 2013. Asian Pac J Cancer Prev. 2015;16(11):4577-4581. doi:10.7314/ apjcp.2015.16.11.4577 7. Lu JJ, Li HA, Xiong Y. Breast cancer inpatients undergoing mastectomy from a hospital in Guangzhou, China: a retrospective analysis 2004- 2013. Asian Pac J Cancer Prev. 2015;16(11):4577-4581. doi:10.7314/ apjcp.2015.16.11.4577 8. Li R, Zhang L, Yang J, et al. Analysis of inpatient payments of breast cancer patients with different medical insurance coverages in China (mainland) in 2011-2015. Chin J Cancer Res. 2017;29(5):419-425. doi:10.21147/j.issn.1000-9604.2017.05.06 8. Li R, Zhang L, Yang J, et al. Analysis of inpatient payments of breast cancer patients with different medical insurance coverages in China (mainland) in 2011-2015. Chin J Cancer Res. 2017;29(5):419-425. doi:10.21147/j.issn.1000-9604.2017.05.06 9. Liao XZ, Shi JF, Liu JS, et al. Medical and non-medical expenditure for breast cancer diagnosis and treatment in China: a multicenter cross- sectional study. Asia Pac J Clin Oncol. 2018;14(3):167-178. doi:10.1111/ ajco.12703 9. Liao XZ, Shi JF, Liu JS, et al. Medical and non-medical expenditure for breast cancer diagnosis and treatment in China: a multicenter cross- sectional study. Asia Pac J Clin Oncol. 2018;14(3):167-178. doi:10.1111/ ajco.12703 10. Organization for Economic Cooperation and Development, Eurostat, World Health Organization. A System of Health Accounts. Paris: OECD Publishing; 2011. 11. Dieleman JL, Graves CM, Templin T, et al. Global health development assistance remained steady in 2013 but did not align with recipients’ disease burden. Health Aff (Millwood). 2014;33(5):878-886. doi:10.1377/ hlthaff.2013.1432 Ethical issues Ethical issues Ethical issues The need for ethics approval was waived by the Ethics Committee of China Medical University, Shenyang, China, on the basis that the data used in this study is anonymized and de-identified. No identified or potentially identifiable human information was collected or generated in this study. 12. Bui AL, Lavado RF, Johnson EK, et al. National health accounts data from 1996 to 2010: a systematic review. Bull World Health Organ. 2015; 93(8):566-576d. doi:10.2471/blt.14.145235 13. World Health Organizationn. Health accounts methodology. http://www. who.int/health-accounts/methodology/en/. Accessed October 23, 2020. Discussion Although cancer staging has a significant impact on the cost of breast cancer care, we were unable to conduct a stratified analysis of patients with different cancer stages. Third, due to the lack of comprehensive data on indirect costs, our analysis is limited to direct costs. Due to the above reasons, our estimates may underestimate the overall medical service cost of cancer treatment. Despite these limitations, healthcare costs were obtained from the SHA, which includes a representative sample used to provide a comprehensive and comparative cost analysis for breast cancer care. Authors’ affiliations 1Department of Health Service Management, China Medical University, Shenyang, China. 2The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China. 3School of Nursing, Capital Medical University, Beijing, China. 4Department of Biostatistics and Epidemiology, School of Public Health, China Medical University, Shenyang, China. 19. Manning WG, Mullahy J. Estimating log models: to transform or not to transform? J Health Econ. 2001;20(4):461-494. doi:10.1016/s0167- 6296(01)00086-8 20. Luengo-Fernandez R, Leal J, Gray A, Sullivan R. Economic burden of cancer across the European Union: a population-based cost analysis. Lancet Oncol. 2013;14(12):1165-1174. doi:10.1016/s1470- 2045(13)70442-x International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1743 Supplementary files 21. Allaire BT, Ekwueme DU, Guy GP Jr, et al. Medical care costs of breast 21. Allaire BT, Ekwueme DU, Guy GP Jr, et al. Medical care costs of breast Supplementary file 1 contains Tables S1 and S2. International Journal of Health Policy and Management, 2022, 11(9), 1735–1743 1742 Ma et al cancer in privately insured women aged 18-44 years. Am J Prev Med. 2016;50(2):270-277. doi:10.1016/j.amepre.2015.08.035 cancer. Accessed August 21, 2020. 27. McBride A, Campen CJ, Camamo J, et al. Implementation of a pharmacy- managed program for the transition of chemotherapy to the outpatient setting. Am J Health Syst Pharm. 2018;75(9):e246-e258. doi:10.2146/ ajhp170138 22. Bonastre J, Jan P, Barthe Y, Koscielny S. Metastatic breast cancer: we do need primary cost data. Breast. 2012;21(3):384-388. doi:10.1016/j. breast.2012.03.012 28. Joo EH, Rha SY, Ahn JB, Kang HY. Economic and patient-reported outcomes of outpatient home-based versus inpatient hospital-based chemotherapy for patients with colorectal cancer. Support Care Cancer. 2011;19(7):971-978. doi:10.1007/s00520-010-0917-7 23. Ekwueme DU, Allaire BT, Guy GP Jr, Arnold S, Trogdon JG. Treatment costs of breast cancer among younger women aged 19-44 years enrolled in Medicaid. Am J Prev Med. 2016;50(2):278-285. doi:10.1016/j. amepre.2015.10.017 29. Sung H, Rosenberg PS, Chen WQ, et al. Female breast cancer incidence among Asian and Western populations: more similar than expected. J Natl Cancer Inst. 2015;107(7):djv107. doi:10.1093/jnci/djv107 24. Zhang A, Nikoloski Z, Mossialos E. Does health insurance reduce out-of-pocket expenditure? Heterogeneity among China’s middle- aged and elderly. Soc Sci Med. 2017;190:11-19. doi:10.1016/j. socscimed.2017.08.005 30. Yao N, Zhang N, Wang J, et al. Experiences with cancer survey in China. Cancer. 2019;125(17):3068-3078. doi:10.1002/cncr.32164 25. Pathak R, Giri S, Aryal MR, Karmacharya P, Bhatt VR, Martin MG. Mortality, length of stay, and health care costs of febrile neutropenia- related hospitalizations among patients with breast cancer in the United States. Support Care Cancer. 2015;23(3):615-617. doi:10.1007/s00520- 014-2553-0 31. Wang Z, Li X, Chen M, Si L. Social health insurance, healthcare utilization, and costs in middle-aged and elderly community-dwelling adults in China. Int J Equity Health. 2018;17(1):17. doi:10.1186/s12939-018-0733-0 32. Meng Z, Ma Y, Song S, et al. Economic implications of Chinese diagnosis- related group-based payment systems for critically ill patients in ICUs. Crit Care Med. 2020;48(7):e565-e573. doi:10.1097/ccm.0000000000004355 26. EUROSTAT. Cancer statistics. https://ec.europa.eu/eurostat/statistics- explained/index.php/Cancer_statistics_-_specific_cancers#Breast_
https://openalex.org/W2884023570	https://europepmc.org/articles/pmc6222791?pdf=render	English	null	(−)-Asarinin from the Roots of Asarum sieboldii Induces Apoptotic Cell Death via Caspase Activation in Human Ovarian Cancer Cells	Molecules/Molecules online/Molecules annual	2,018	cc-by	9,252	Received: 30 May 2018; Accepted: 24 July 2018; Published: 25 July 2018 Abstract: Two tetrahydrofurofurano lignans (1 and 2), four phenylpropanoids (3–6), and two alkamides (7 and 8) were isolated from the EtOAc-soluble fraction of the roots of Asarum sieboldii. The chemical structures of the isolates were identiﬁed by analysis of spectroscopic data measurements, and by a comparison of their data with published values. The isolates (1, 2, 4–8) were evaluated for their cytotoxicity against human ovarian cancer cells (A2780 and SKOV3) and immortalized ovarian surface epithelial cells (IOSE80PC) using a MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay. Of the isolates, (−)-asarinin (1) exhibited the most potent cytotoxicity to both A2780 and SKOV3 cells. A propidium iodide/annexin V-ﬂuorescein isothiocyanate (V-FITC) double staining assay showed that (−)-asarinin (1) induces apoptotic cell death in ovarian cancer cells. In addition, (−)-asarinin (1) increased the activation of caspase-3, caspase-8, and caspase-9 in ovarian cancer cells. Pretreatment with caspase inhibitors attenuated the cell death induced by (−)-asarinin (1). In conclusion, our ﬁndings show that (−)-asarinin (1) from the roots of A. sieboldii may induce caspase-dependent apoptotic cell death in human cancer cells. Keywords: Asarum sieboldii; (−)-asarinin; cytotoxicity; ovarian cancer cells; apoptosis; caspase Article (−)-Asarinin from the Roots of Asarum sieboldii Induces Apoptotic Cell Death via Caspase Activation in Human Ovarian Cancer Cells Miran Jeong 1,†, Hye Mi Kim 1,†, Jin Su Lee 1,2, Jung-Hye Choi 1,2,* and Dae Sik Jang 1,2,* 1 College of Pharmacy, Kyung Hee University, Seoul 02447, Korea; jeongmiran@hanmail.net (M.J.); hyemi586@gmail.com (H.M.K.); lee2649318@naver.com (J.S.L.) 2 Department of Life & Nanopharmaceutical Sciences, Kyung Hee University, Seoul 02447, Korea * Correspondence: jchoi@khu.ac.kr (J.-H.C.); dsjang@khu.ac.kr (D.S.J.); Tel.: +82-2-961-2172 (J.-H.C.); +82-2-961-0719 (D.S.J.) † These authors contributed equally to this work. † These authors contributed equally to this work. www.mdpi.com/journal/molecules Molecules 2018, 23, 1849; doi:10.3390/molecules23081849 molecules molecules 1. Introduction Asarum sieboldii Miq. (Aristolochiaceae) is a herbal plant that is widely distributed in Korea, China, and Japan. In traditional medicine, the roots of A. sieboldii are used as treatment for all types of colds, fever, chills, headaches, acute toothaches, sinusitis, cough, and dyspnoea, due to retention of phlegm, pharyngitis, chronic gastritis, and rheumatoid arthritis [1,2]. It has been demonstrated that A. sieboldii has a broad range of biological activity, such as anti-allergic [3], antitussive [4], anti-inﬂammatory [5], anti-nociceptive [6], anti-fungal [7,8], neuroprotective [9], and anticancer activity [10,11]. Previous phytochemical investigations on the roots of A. sieboldii resulted in the isolation of essential oils [12, 13], alkamides [5,14], lignans [5,14], and alkaloids [15,16]. A 70% EtOH extract from Asiasari radix induced apoptosis preceded by a tight cell cycle arrest in the G2/M phase, suggesting that the extract prevented the growth of HCT-116 human colon cancer cells. The extract increased the expression of Bax/Bcl-2 and p53, and activated caspases, including Caspase-9 and Caspase-8 [10]. In addition, Asiasari radix extract signiﬁcantly enhanced the sensitivity of HeLa cells to paclitaxel [17]. Moreover, Park et al. have demonstrated that the EtOAc-soluble fraction of Asiasari radix exhibits cytotoxic activity against human cancer cell lines such as A549 (human lung cancer), SKOV3 (human ovarian Molecules 2018, 23, 1849; doi:10.3390/molecules23081849 www.mdpi.com/journal/molecules 2 of 11 Molecules 2018, 23, 1849 cancer), and SKMEL-2 (human melanoma) [11]. As part of an our ongoing project to search for novel, plant-derived anti-cancer agents [18], we found that the EtOAc-soluble fraction of the 70% EtOH extract of A. sieboldii roots exhibited signiﬁcant cytotoxicity in the human ovarian cancer cells A2780 and SKOV3. Fractionation of the active EtOAc-soluble fraction resulted in the isolation and identiﬁcation of eight known compounds consisting of tetrahydrofurofurano lignans (1 and 2), phenylpropanoids (3–6), and alkamides (7 and 8). The structures of the isolates were determined by spectroscopic analyses, including 1H-NMR, 13C-NMR, 2D-NMR, and MS spectra, and via a comparison of the data with published values. The isolates (1, 2, 4–8) were evaluated for their cytotoxicity against human ovarian cancer cells (A2780 and SKOV3) and immortalized ovarian surface epithelium cells (IOSE80PC), using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assays. Of the isolates, a tetrahydrofurofurano lignin (−)-asarinin (1), which displayed potent cytotoxicity against both A2780 and SKOV3 cells, was investigated for the molecular mechanism of its cytotoxic activity. Here, we describe the isolation and identiﬁcation of the isolates (1–8) from the roots of A. 1. Introduction sieboldii, as well as their cytotoxicity to human ovarian cancer cells. This paper also deals with a mechanism study for cytotoxicity of (−)-asarinin (1) in human cancer cells. 2.1. Cytotoxicity of the Extract and Solvent Fractions against Human Ovarian Cancer Cells The 70% EtOH extract of the roots of A. sieboldii and two solvent partitions (EtOAc- and water-soluble fractions) from the 70% EtOH extract were investigated for their cytotoxicity against human ovarian cancer cells (A2780 and SKOV3) using MTT assays (Table 1). The 70% EtOH extract showed a signiﬁcant cytotoxicity against A2780, with an observed IC50 value of 31.5 ± 16.83 µg/mL. Of the solvent partitions, the EtOAc-soluble fraction exhibited more potent cytotoxicity than the water-soluble fraction, against both ovarian cancer cells (IC50 values were 19.89 and 118.47 µg/mL in A2780 and SKOV3, respectively). Thus, we attempted to identify the cytotoxic constituents in the EtOAc-soluble fraction. Our data on the EtOAc-soluble fraction were consistent with previous results [11] on the cytotoxic activity of the EtOAc-soluble fraction of Asiasari radix against several human cancer cell lines including SKOV3 cells. Table 1. Cytotoxicity of the 70% EtOH extract of A. sieboldii roots and its solvent fractions in human ovarian cancer cells A2780 and SKOV3. Extract or Fractions IC50 (µg/mL) * Human Ovarian Cancer Cells A2780 SKOV3 70% EtOH extract 31.5 ± 16.83 >200 EtOAc fraction 19.89 ± 4.20 118.47 ± 19.78 Water fraction 107.20 ± 15.80 139.30 ± 29.13 * The IC50 value is deﬁned as being a concentration that lowers cell viability by 50%. The value shows an average of the results of three independent experiments having similar patterns. Table 1. Cytotoxicity of the 70% EtOH extract of A. sieboldii roots and its solvent fractions in human ovarian cancer cells A2780 and SKOV3. * The IC50 value is deﬁned as being a concentration that lowers cell viability by 50%. The value shows an average of the results of three independent experiments having similar patterns. * The IC50 value is deﬁned as being a concentration that lowers cell viability by 50%. The value shows an average of the results of three independent experiments having similar patterns. 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells To identify compounds with cytotoxic activity against human cancer cells from the A. sieboldii roots, we investigated the effect of the isolates (1, 2, 4–8) obtained from the EtOAc-soluble fraction of the roots of A. sieboldii in human ovarian cancer cells A2780 and SKOV3. The effects of the isolates were assessed using IC50 values and they are summarized in Table 2. Of these, a tetrahydrofurofurano lignin, (−)-asarinin (1), exhibited the most potent cytotoxicity on both A2780 and SKOV3 cells, with observed IC50 values of 38.45 ± 2.78 and 60.87 ± 5.01 μM, respectively. Interestingly, (−)-asarinin (1) did not show any cytotoxicity against immortalized ovarian surface epithelial IOSE80PC cells, which were used as a normal counterpart of ovarian cancer cells. On the other hand, another lignan (−)- pluviatilol (2) showed a mild cytotoxicity against all the three cells tested (A2780 (IC50 value of 101.85 ± 13.55 μM), SKOV3 (IC50 value of 173.82 ± 9.42 μM), and IOSE80PC cells (IC50 value of 178.92 ± 3.30 μM)). An alkamide, (2E,4E,8Z,10E/Z)-N-isobutyldodeca-2,4,8,10-tetraenamide (8), exhibited a mild cytotoxicity only against A2780 cells (IC50 value of 101.20 ± 10.35 μM), but not against SKOV3 cells. These results implicate that (−)-asarinin (1), one of major constituents of the roots of A. sieboldii, seems to contribute to the cytotoxicity of the EtOAc-soluble fraction against cancer cells. (−)-Sesamin, which is an epimer of asarinin and is also found in this plant [5], exerts cytotoxic activity toward cancer cells and induces apoptosis in vitro [25–28]. An anticancer effect of (−)-sesamin in athymic mice transplanted with human MCF-7 breast cancer cells has also been reported [25]. Furthermore, the dietary supplementation of (−)-sesamin on 7,12-dimethylbenz[a]anthracene (DMBA)-induced breast carcinogenesis led to a significantly reduced appearance of breast cancers, compared to mice that were not supplemented with (−)-sesamin [26]. In contrast, (−)-asarinin (1) has been suggested to have a significant inhibitory effect on EBV-EA activation and preserved the high viability of the Raji cells [29,30]. In addition, (−)-asarinin (1) inhibited DNA synthesis in human leukemia HL-60 cells [31]. However the effect of (−)-asarinin (1) on human solid cancer cell growth and its underlying molecular mechanism have not been reported to date. To identify compounds with cytotoxic activity against human cancer cells from the A. sieboldii roots, we investigated the effect of the isolates (1, 2, 4–8) obtained from the EtOAc-soluble fraction of the roots of A. sieboldii in human ovarian cancer cells A2780 and SKOV3. 2.2. Identiﬁcation of Compounds 1–8 from the Roots of A. sieboldii Two tetrahydrofurofurano lignans (1 and 2), four phenylpropanoids (3–6), and two alkamides (7 and 8) were isolated from the active EtOAc-soluble fraction of the roots of A. sieboldii. The chemical structures of the isolates were identified to be (−)-asarinin (1) [19], (−)-pluviatilol (2) [20], kakuol (3) [21], methylkakuol (4) [21], asaricin (5) [22], methyleugenol (6) [23], (2E,4E,8Z)-N-isobutyldeca-2,4,8-trienamide (7) [23], and (2E,4E,8Z,10E/Z)-N-isobutyldodeca-2,4,8,10-tetraenamide (8) [24], by analysis of spectroscopic data (1H-NMR, 13C-NMR, 2D-NMR, and MS) measurements, and by a comparison of their data with published values (Figure 1). 3 of 11 Molecules 2018, 23, 1849 es 2018, 23, x Figure 1. Compounds 1–8 isolated from the EtOAc-soluble fraction of the roots of A. sielboldii. Figure 1. Compounds 1–8 isolated from the EtOAc-soluble fraction of the roots of A. sielboldii. Figure 1. Compounds 1–8 isolated from the EtOAc-soluble fraction of the roots of A. sielboldii. Figure 1. Compounds 1–8 isolated from the EtOAc-soluble fraction of the roots of A. sielboldii. 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells The effects of the isolates were assessed using IC50 values and they are summarized in Table 2. Of these, a tetrahydrofurofurano lignin, (−)-asarinin (1), exhibited the most potent cytotoxicity on both A2780 and SKOV3 cells, with observed IC50 values of 38.45 ± 2.78 and 60.87 ± 5.01 µM, respectively. Interestingly, (−)-asarinin (1) did not show any cytotoxicity against immortalized ovarian surface epithelial IOSE80PC cells, which were used as a normal counterpart of ovarian cancer cells. On the other hand, another lignan (−)-pluviatilol (2) showed a mild cytotoxicity against all the three cells tested (A2780 (IC50 value of 101.85 ± 13.55 µM), SKOV3 (IC50 value of 173.82 ± 9.42 µM), and IOSE80PC cells (IC50 value of 178.92 ± 3.30 µM)). An alkamide, (2E,4E,8Z,10E/Z)-N-isobutyldodeca-2,4,8,10-tetraenamide (8), exhibited a mild cytotoxicity only against A2780 cells (IC50 value of 101.20 ± 10.35 µM), but not against SKOV3 cells. These results implicate that (−)-asarinin (1), one of major constituents of the roots of A. sieboldii, seems to contribute to the cytotoxicity of the EtOAc-soluble fraction against cancer cells. (−)-Sesamin, which is an epimer of asarinin and is also found in this plant [5], exerts cytotoxic activity toward cancer cells and induces apoptosis in vitro [25–28]. An anticancer effect of (−)-sesamin in athymic mice transplanted with human MCF-7 breast cancer cells has also been reported [25]. Furthermore, the dietary supplementation of (−)-sesamin on 7,12-dimethylbenz[a]anthracene (DMBA)-induced breast carcinogenesis led to a signiﬁcantly reduced appearance of breast cancers, compared to mice that were not supplemented with (−)-sesamin [26]. In contrast, (−)-asarinin (1) has been suggested to have a signiﬁcant inhibitory effect on EBV-EA activation and preserved the high viability of the Raji cells [29,30]. In addition, (−)-asarinin (1) inhibited DNA synthesis in human leukemia HL-60 cells [31]. However the effect of (−)-asarinin (1) on human solid cancer cell growth and its underlying molecular mechanism have not been reported to date. 4 of 11 Molecules 2018, 23, 1849 Table 2. Cytotoxicity of compounds 1–8 from the EtOAc-soluble fraction of the roots of A. sieboldii in human ovarian cancer cells (A2780 and SKOV3) and immortalized ovarian surface epithelial cells (IOSE80PC). Table 2. Cytotoxicity of compounds 1–8 from the EtOAc-soluble fraction of the roots of A. sieboldii in human ovarian cancer cells (A2780 and SKOV3) and immortalized ovarian surface epithelial cells (IOSE80PC). 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells 2.3. Cytotoxicity of Compounds 1–8 against Human Ovarian Cancer Cells Compound IC50 (µM) * Human Ovarian Cancer Cells Immortalized Ovarian Surface Epithelial Cells A2780 SKOV3 IOSE80PC 1 38.45 ± 2.78 60.87 ± 5.01 >200 2 101.85 ± 13.55 173.82 ± 9.42 178.92 ± 3.30 3 >200 >200 >200 4 >200 >200 >200 5 ND ND ND 6 >200 >200 >200 7 >200 >200 >200 8 101.20 ± 10.35 >200 >200 Cisplatin * 8.77 ± 0.48 24.18 ± 0.28 45.62 ± 0.30 * IC50 value is deﬁned as the concentration that results in a 50% decrease in the proliferation of the cells. The values represent the means of the results from three independent experiments with similar patterns. Not determined since the amount of available compound was insufﬁcient. * Cisplatin was used as an assay positive control. Compound IC50 (μM) * Human Ovarian Cancer Cells Immortalized Ovarian Surface Epithelial Cells A2780 SKOV3 IOSE80PC 1 38.45 ± 2.78 60.87 ± 5.01 >200 2 101.85 ± 13.55 173.82 ± 9.42 178.92 ± 3.30 3 >200 >200 >200 4 >200 >200 >200 5 ND ND ND 6 >200 >200 >200 7 >200 >200 >200 8 101.20 ± 10.35 >200 >200 Cisplatin * 8.77 ± 0.48 24.18 ± 0.28 45.62 ± 0.30 * IC50 value is defined as the concentration that results in a 50% decrease in the proliferation of the cells. The values represent the means of the results from three independent experiments with similar patterns. Not determined since the amount of available compound was insufficient. * Cisplatin * IC50 value is deﬁned as the concentration that results in a 50% decrease in the proliferation of the cells. The values represent the means of the results from three independent experiments with similar patterns. Not determined since the amount of available compound was insufﬁcient. * Cisplatin was used as an assay positive control. * IC50 value is defined as the concentration that results in a 50% decrease in the proliferation of the cells. The values represent the means of the results from three independent experiments with similar patterns. Not determined since the amount of available compound was insufficient. * Cisplatin 2.4. (−)-Asarinin (1)-Induced Apoptotic Cell Death in Human Ovarian Cancer Cells was used as an assay positive control. ( ) A ( ) I d d A ll h O ll 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells The activation mechanisms that trigger apoptosis are often referred to as the extrinsic and intrinsic pathways [33,34]. The intrinsic pathway (also called the mitochondria-mediated pathway) is activated directly by a variety of intracellular death stimuli. On the other hand, the extrinsic pathway (also called the receptor-mediated pathway) is characterized by the activations of cell- surface death receptors by the binding of extracellular ligands. Both extrinsic and intrinsic apoptotic pathways are highly dependent on the activation of caspases, which play a critical role in the proteolysis of specific targets [35]. Caspase-8 and caspase-9 are the initiator caspases for the extrinsic and intrinsic apoptotic pathways, respectively, while the effector caspase caspase-3 can be stimulated by activation of the initiator caspases in both pathways. To identify the mechanisms involved in (−)- asarinin (1)-induced apoptotic cell death, we investigated the activation of an effector caspase, caspase-3 and initiator caspases caspase-8 and caspase-9. (−)-Asarinin (1) markedly stimulated the activation of caspase-3, caspase-8, and caspase-9 in both A2780 and SKOV3 cells (Figure 3). To further confirm the involvement of caspases in (−)-asarinin (1)-induced apoptosis, the effect of caspase inhibitors on (−)-asarinin (1)-induced cell death was investigated. z-DEVD-fmk (a specific caspase-3 inhibitor), z-IETD-fmk (a specific caspase-8 inhibitor), and z-LEHD-fmk (a specific caspase-9 inhibitor) considerably attenuated (−)-asarinin (1)-induced cell death in both A2780 and SKOV3 cells (Figure 4). These results show that (−)-asarinin (1)-induced apoptosis is mediated by the caspase- dependent pathway in human ovarian cancer cells. The activation mechanisms that trigger apoptosis are often referred to as the extrinsic and intrinsic pathways [33,34]. The intrinsic pathway (also called the mitochondria-mediated pathway) is activated directly by a variety of intracellular death stimuli. On the other hand, the extrinsic pathway (also called the receptor-mediated pathway) is characterized by the activations of cell-surface death receptors by the binding of extracellular ligands. Both extrinsic and intrinsic apoptotic pathways are highly dependent on the activation of caspases, which play a critical role in the proteolysis of speciﬁc targets [35]. Caspase-8 and caspase-9 are the initiator caspases for the extrinsic and intrinsic apoptotic pathways, respectively, while the effector caspase caspase-3 can be stimulated by activation of the initiator caspases in both pathways. 2.4. (−)-Asarinin (1)-Induced Apoptotic Cell Death in Human Ovarian Cancer Cells was used as an assay positive control. (A and B) A2780 cells (A) and SKOV3 cells (B) were treated with (−)-asarinin (1) for the indicated concentrations for 48 h. The cells were stained with propidium iodide (PI) according to the protocol as described in method. The cell cycle distribution profiles were measured by flow cytometry. FACS images shown are representative of three independent experiments. Means ± SD from three independent experiments are graphed from the proportion of cells (%) in each phase of cell cycle (sub G1, G0/G1, S, and G2/M). Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. (C and D) A2780 cells (C) and SKOV3 cells (D) were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Apoptotic cell death analysis was performed using PI/Annexin V-fluorescein isothiocyanate (V-FITC) double staining assay. The data are representative of three independent experiments. Statistical significance was determined by a one- way ANOVA. * p < 0.05 as compared with the untreated group. Figure 2. The effect of (−)-asarinin (1) on the cell cycle and apoptosis of human ovarian cancer cells. (A and B) A2780 cells (A) and SKOV3 cells (B) were treated with (−)-asarinin (1) for the indicated concentrations for 48 h. The cells were stained with propidium iodide (PI) according to the protocol as described in method. The cell cycle distribution proﬁles were measured by ﬂow cytometry. FACS images shown are representative of three independent experiments. Means ± SD from three independent experiments are graphed from the proportion of cells (%) in each phase of cell cycle (sub G1, G0/G1, S, and G2/M). Statistical signiﬁcance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. (C and D) A2780 cells (C) and SKOV3 cells (D) were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Apoptotic cell death analysis was performed using PI/Annexin V-ﬂuorescein isothiocyanate (V-FITC) double staining assay. The data are representative of three independent experiments. Statistical signiﬁcance was determined by a one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 2. The effect of (−)-asarinin (1) on the cell cycle and apoptosis of human ovarian cancer cells. (A and B) A2780 cells (A) and SKOV3 cells (B) were treated with (−)-asarinin (1) for the indicated concentrations for 48 h. 2.4. (−)-Asarinin (1)-Induced Apoptotic Cell Death in Human Ovarian Cancer Cells was used as an assay positive control. We found that (−)-asarinin (1) increased the fractionation of the nuclei accumulated at Sub G1 in A2780 and SKOV3 cells, but it failed to induce cell cycle arrest, which is one mechanism of growth inhibition (Figure 2A,B). Therefore, we further explored the molecular mechanism of the cytotoxic activity of (−)-asarinin (1) in human ovarian cancer cells. Dysfunction of apoptosis signaling pathways is associated with cancer development [32]. Therefore, compounds that promote apoptosis in cancer cells are considered as good candidates for anti-cancer chemotherapeutics. To examine whether (−)-asarinin (1)-induced cell death is mediated by the induction of apoptosis, two ovarian cancer cells A2780 and SKOV3 cells were treated with (−)-asarinin (1) and subsequently co-stained with PI and Annexin V-FITC. Treatment with (−)-asarinin (1) increased the percentage of Annexin V-FITC positive/PI negative cells by up to 43% in A2780 cells and 48% in SKOV3 cells (Figure 2C,D). These results suggest that the (−)-asarinin (1)-induced cell death is associated with the induction of apoptosis in human ovarian cancer cells. 2.4. ( )-Asarinin (1)-Induced Apoptotic Cell Death in Human Ovarian Cancer Cells We found that (−)-asarinin (1) increased the fractionation of the nuclei accumulated at Sub G1 in A2780 and SKOV3 cells, but it failed to induce cell cycle arrest, which is one mechanism of growth inhibition (Figure 2A,B). Therefore, we further explored the molecular mechanism of the cytotoxic activity of (−)-asarinin (1) in human ovarian cancer cells. Dysfunction of apoptosis signaling pathways is associated with cancer development [32]. Therefore, compounds that promote apoptosis in cancer cells are considered as good candidates for anti-cancer chemotherapeutics. To examine whether (−)-asarinin (1)-induced cell death is mediated by the induction of apoptosis, two ovarian cancer cells A2780 and SKOV3 cells were treated with (−)-asarinin (1) and subsequently co-stained with PI and Annexin V-FITC. Treatment with (−)-asarinin (1) increased the percentage of Annexin V- FITC positive/PI negative cells by up to 43% in A2780 cells and 48% in SKOV3 cells (Figure 2C,D). These results suggest that the (−)-asarinin (1)-induced cell death is associated with the induction of apoptosis in human ovarian cancer cells. Figure 2. Cont. Cont Figure Figure 2. Cont. 5 of 11 5 of 11 Molecules 2018, 23, 1849 Molecules 2018 23 x Figure 2. The effect of (−)-asarinin (1) on the cell cycle and apoptosis of human ovarian cancer cells. 2.4. (−)-Asarinin (1)-Induced Apoptotic Cell Death in Human Ovarian Cancer Cells was used as an assay positive control. The cells were stained with propidium iodide (PI) according to the protocol as described in method. The cell cycle distribution profiles were measured by flow cytometry. FACS images shown are representative of three independent experiments. Means ± SD from three independent experiments are graphed from the proportion of cells (%) in each phase of cell cycle (sub G1, G0/G1, S, and G2/M). Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. (C and D) A2780 cells (C) and SKOV3 cells (D) were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Apoptotic cell death analysis was performed using PI/Annexin V-fluorescein isothiocyanate (V-FITC) double staining assay. The data are representative of three independent experiments. Statistical significance was determined by a one- way ANOVA. * p < 0.05 as compared with the untreated group. Figure 2. The effect of (−)-asarinin (1) on the cell cycle and apoptosis of human ovarian cancer cells. (A and B) A2780 cells (A) and SKOV3 cells (B) were treated with (−)-asarinin (1) for the indicated concentrations for 48 h. The cells were stained with propidium iodide (PI) according to the protocol as described in method. The cell cycle distribution proﬁles were measured by ﬂow cytometry. FACS images shown are representative of three independent experiments. Means ± SD from three independent experiments are graphed from the proportion of cells (%) in each phase of cell cycle (sub G1, G0/G1, S, and G2/M). Statistical signiﬁcance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. (C and D) A2780 cells (C) and SKOV3 cells (D) were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Apoptotic cell death analysis was performed using PI/Annexin V-ﬂuorescein isothiocyanate (V-FITC) double staining assay. The data are representative of three independent experiments. Statistical signiﬁcance was determined by a one-way ANOVA. * p < 0.05 as compared with the untreated group. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells * p < 0.05 as compared with the untreated group. (A) and SKOV3 (B) cells were treated with the indicated concentration of ( ) asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical signiﬁcance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. a d S O 3 ( ) ce s e e t eated it t e i dicated co ce t atio o ( ) asa i i ( ) o 8 Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells To identify the mechanisms involved in (−)-asarinin (1)-induced apoptotic cell death, we investigated the activation of an effector caspase, caspase-3 and initiator caspases caspase-8 and caspase-9. (−)-Asarinin (1) markedly stimulated the activation of caspase-3, caspase-8, and caspase-9 in both A2780 and SKOV3 cells (Figure 3). To further conﬁrm the involvement of caspases in (−)-asarinin (1)-induced apoptosis, the effect of caspase inhibitors on (−)-asarinin (1)-induced cell death was investigated. z-DEVD-fmk (a speciﬁc caspase-3 inhibitor), z-IETD-fmk (a speciﬁc caspase-8 inhibitor), and z-LEHD-fmk (a speciﬁc caspase-9 inhibitor) considerably attenuated (−)-asarinin (1)-induced cell death in both A2780 and SKOV3 cells (Figure 4). These results show that (−)-asarinin (1)-induced apoptosis is mediated by the caspase-dependent pathway in human ovarian cancer cells. 6 of 11 6 of 11 6 of 11 Molecules 2018, 23, 1849 M l l 2018 23 M l l 2018 23 Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical signiﬁcance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells * p < 0.05 as compared with the untreated group. Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical signiﬁcance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. Figure 3. The effect of (−)-asarinin (1) on caspase activation in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were treated with the indicated concentration of (−)-asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. * p < 0.05 as compared with the untreated group. and SKOV3 (B) cells were treated with the indicated concentration of ( ) asarinin (1) for 48 h. Procaspase-3, -8, and -9 levels were determined by Western blot assay. β-Actin was used as an internal control. A representative protein immunoblot of three independent experiments is shown. Data are presented as the means ± SD of three independent experiments. Statistical significance was determined by one-way ANOVA. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 µM), caspase-8 inhibitor z-IETD-fmk (50 µM), and caspase-9 inhibitor z-LEHD-fmk (50 µM) for 30 min, and then treated with the indicated concentration (A2780; 40 µM and SKOV3; 60 µM) of (−)-asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical signiﬁcance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 µM), caspase-8 inhibitor z-IETD-fmk (50 µM), and caspase-9 inhibitor z-LEHD-fmk (50 µM) for 30 min, and then treated with the indicated concentration (A2780; 40 µM and SKOV3; 60 µM) of (−)-asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical signiﬁcance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. Figure 4. The effect of caspase inhibitors on (−)-asarinin (1)-induced cell death in human ovarian cancer cells. A2780 (A) and SKOV3 (B) cells were pretreated with caspase-3 inhibitor z-DEVD-fmk (50 µM), caspase-8 inhibitor z-IETD-fmk (50 µM), and caspase-9 inhibitor z-LEHD-fmk (50 µM) for 30 min, and then treated with the indicated concentration (A2780; 40 µM and SKOV3; 60 µM) of (−)-asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. 3.3. Extraction and Isolation The dried roots of A. sieboldii (280.4 g) was extracted with 70% EtOH (3.65 L) three times at room temperature, each for 6 h. The extract (40.15 g) was suspended in H2O (350 mL) and successively extracted with EtOAc (350 mL × 3) to yield EtOAc- (5.73 g) and a water-soluble fraction (34.92 g), respectively. The EtOAc-soluble fraction (5.73 g) was chromatographed over silica gel CC (70–230 mesh, ø 4.7 × 38.5 cm) as stationary phase with n-hexane-EtOAc-MeOH (9:1:0–0:0:1, v/v) as a mobile phase to generate 14 fractions (E1~E14). Fraction E3 (260.0 mg) was subjected to silica gel CC (230–400 mesh, ø 3.7 × 22.5 cm) with an n-hexane-CH2Cl2-EtOAc mixture (7:3:0.1, v/v) to give compound 5 (1.3 mg). Compounds 1 (344.4 mg), 2 (20.0 mg), and 3 (45.2 mg) were obtained by recrystallization (in n-hexane) from fractions E6 (274.1 mg), E9 (161.3 mg), and E4 (847.3 mg), respectively. Fraction E8 (740 mg) was further subjected to Sephadex LH-20 CC (ø 3.6 × 73.5 cm) eluted with CH2Cl2-MeOH mixture (1:1, v/v), to produce six subfractions (E8-1~E8-6). Compounds 7 (4.0 mg) and 8 (100.0 mg) were isolated from subfraction E8-2 (530 mg) by reversed-phase CC (YMC gel 75 µm, ø 3.6 × 23.5 cm) with MeOH-H2O (4:1, v/v). Fraction E4 (847 mg) was separated using Sephadex LH-20 CC (ø 3.6 × 40 cm) eluted with CHCl3-MeOH (1:1, v/v), to give ﬁve subfractions (E4-1~E4-5). Compound 6 (487.5 mg) was obtained from subfraction E4-2 by a silica gel CC (230-400 mesh, ø 3.7 × 22.5 cm). Fraction E5 (271 mg) was fractionated using Sephadex LH-20 CC (ø 3.6 × 73.5 cm) with CH2Cl2-MeOH (1:1, v/v), to generate ﬁve subfractions (E5-1~E5-5). Compound 4 (29.8 mg) was puriﬁed from subfraction E5-4 (70 mg) by a ﬂash chromatographic system with a Redi Sep-C18 column (26 g, MeOH-H2O = 7:13 to 3:2, v/v). 3.2. Plant Meterial The roots of Asarum sieboldii Miq. (Aristolochiaceae) were purchased at Miryon Herbal Medicine Co. Gyeonggi-do, Korea, in March, 2013. Plant material was identiﬁed by Prof. Dae Sik Jang, one of the authors. A voucher specimen (No. 2013-ASSI01) has been deposited in the Lab of Natural Product Medicine, College of Pharmacy, Kyung Hee University, Republic of Korea. 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells 2.5. (−)-Asarinin (1) Induced Caspase-Dependent Cell Death in Human Ovarian Cancer Cells Statistical signiﬁcance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. (50 μM), caspase-8 inhibitor z-IETD-fmk (50 μM), and caspase-9 inhibitor z-LEHD-fmk (50 μM) for 30 min, and then treated with the indicated concentration (A2780; 40 μM and SKOV3; 60 μM) of (−)- asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical significance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. (50 µM), caspase-8 inhibitor z-IETD-fmk (50 µM), and caspase-9 inhibitor z-LEHD-fmk (50 µM) for 30 min, and then treated with the indicated concentration (A2780; 40 µM and SKOV3; 60 µM) of (−)-asarinin (1) for 48 h. A MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed to determine the cell viability after (−)-asarinin (1) treatment. Statistical signiﬁcance was determined by Student’s t-test. # p < 0.05 as compared with the untreated group and * p < 0.05 as compared with the (−)-asarinin (1) only-treated group. 7 of 11 7 of 11 Molecules 2018, 23, 1849 3.1. General Procedures Open column chromatography (CC) was performed with silica gel (70–230 or 230–400 mesh ASTM, Merck, Kenilworth, NJ, USA), Sephadex LH-20 (Amersham Pharmacia Biotech, Piscataway, NJ, USA), reversed-phase silica gel (ODS-A 12 nm S-75 µm, YMC Co., Tokyo, Japan), and Redi Sep-C18 (26 g, Teledyne Isco, Lincoln, NE, USA). Thin-layer chromatography (TLC) was performed on Silica gel 60 F254 (Merck) and RP-18 F254S (Merck) plates; compounds were visualized by ultraviolet (UV) light (254 and 365 nm) and 20% (v/v) H2SO4 reagent (Aldrich, St. Louis, MI, USA). NMR spectra were recorded on a Bruker (Boston, MA, USA) 400 MHz and Varian 500 MHz NMR spectrometer using TMS as an internal standard, and chemical shifts were expressed as δ values. All solvents used for the chromatographic separations were distilled before use. 3.5. MTT Assay A MTT assay was performed to evaluate the cell viability. The MTT was obtained from Molecular Probes Inc. (Eugene, OR, USA). Brieﬂy, the cells (1.0 × 105/well) were seeded in a 96-well plate and incubated for 24 h. The cells were treated with extracts (0.125–200 µg/mL) and compounds (0.125–200 µM) and incubated for 48 h. MTT solution was added into each well (ﬁnal concentration; 0.5 mg/mL) and the plates were incubated for an additional 4 h. The medium was removed and 50 µL of dimethyl sulfoxide (DMSO) was added. The optical density was measured at 540 nm using a microplate spectrophotometer (SpectraMax; Molecular Devices, Sunnyvale, CA, USA). To investigate the involvement of caspases in asarinin (1)-induced cell death, a MTT assay was also performed using caspase inhibitors. The cells were pretreated with caspase inhibitors (50 µM) for 30 min, and treated with (−)-asarinin (1) (A2780, 40 µM; and SKOV3, 60 µM) for 48 h. Caspase-3 inhibitor z-DEVD-fmk, caspase-8 inhibitor z-IETD-fmk, and caspase-9 inhibitor z-LEHD-fmk were purchased from Calbiochem (Bad Soden, Germany). 3.7. Annexin V-FITC/PI Double Staining Annexin V- FITC was obtained from BD Biosciences (San Jose, CA, USA). The cells were treated with (−)-asarinin (1) and incubated for 48 h. The cells were rinsed twice with ice-cold PBS and suspended with 100 µL of binding buffer (10 mM HEPES/NaOH, 140 mM Nacl, 2.5 mM CaCl2, PH 7.4). A total of 5 µL of FITC-conjugated annexin V and 5 µL of PI (50 mg/mL) were added into the cell suspension, and the mixture was incubated in a dark place at room temperature for 15 min. The cells were analyzed using FACS cater-plus ﬂow cytometry (guava easy cyteTM); at least 10,000 cells per each group were counted. 3.6. Cell Cycle Analysis The cells were treated with (−)-asarinin (1) and incubated for 48 h. At the time of collection, the cells were harvested and washed twice with ice-cold phosphate buffered saline (PBS). The cells were ﬁxed and permeabilized with 70% ice-cold ethanol at −20 ◦C for 4 h. The cells were washed once with PBS and resuspended in a staining solution containing propidium iodide (50 µg/mL) and RNase A (5 mg/mL). The cell suspensions were incubated for 30 min at room temperature in a dark place. After 30 min the suspensions were analyzed by ﬂuorescence-activated cell sorting (FACS) cater-plus ﬂow cytometry (guava easy cyteTM, Merk Millipore, Darmstadt, Germany) using 5000 cells per group. 3.4. Cell Culture Human ovarian cancer cell lines (A2780 and SKOV3 cells) were obtained from the American Type Culture Collection (ATCC), and immortalized ovarian surface epithelial cell lines (IOSE80PC) were provided by Dr. N. Auersperg (University of British Columbia, Vancouver, British Columbia, Canada) and Dr. A. Godwin (Fox Chase Cancer Center, Philadelphia, PA, USA). Cells were cultured in the Roswell Park Memorial Institute (RPMI) 1640, supplemented with penicillin (100 U/mL), streptomycin sulfate (100 µg/mL), and 5% fetal bovine serum (FBS) (Life Technologies, Inc., Grand Island, NY, USA) in a humidiﬁed atmosphere of 5% CO2–95% air at 37 ◦C. Molecules 2018, 23, 1849 8 of 11 3.9. Statistical Analysis One-way ANOVA and Student’s t-test were performed to determine statistically signiﬁcant differences. p-values of less than 0.05 were regarded as statistically signiﬁcant. 3.8. Western Blot Assay The cells were treated with (−)-asarinin (1) and incubated for 48 h. The cells were rinsed twice with ice-cold PBS and lysed with protein lysis buffer (Intron Biotechnology, Seoul, Korea) containing protease inhibitors (0.5 mM PMSF and 5 µg/mL aprotinin). The lysates were mixed with 5X sodium dodecyl sulfate (SDS) sample buffer and boiled for 5 min for denaturation. Total protein was run on 10–12% SDS-PAGE gels and electrotransferred onto a polyvinylidene diﬂuoride (PVDF) membrane. The membrane was immunoblotted using speciﬁc primary antibodies overnight at 4 ◦C following blocking with 5% non-fat dry milk for 30 min–1 h. After washing, the membrane was incubated with horseradish peroxidase-conjugated secondary antibody (1:1000–2000) at room temperature for 1−2 h. After washing, immunepositive bands were visualized using an ECL chemiluminescent system and analyzed by Image Quant LAS-4000 (Fujiﬁlm Life science, Tokyo, Japan). Anti-caspase-3 and b-actin antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Caspase-9 antibody was purchased from Cell Signaling (Beverly, MA, USA). Caspase-8 antibody was obtained from BD Biosciences (San Jose, CA, USA). 9 of 11 Molecules 2018, 23, 1849 4. Conclusions Fractionation of the active EtOAc-soluble fraction from the 70% EtOH extract of the roots of A. sieboldii resulted in the isolation and identiﬁcation of two tetrahydrofurofurano lignans (1 and 2), four phenylpropanoids (3–6), and two alkamides (7 and 8). Of the isolates, (−)-asarinin (1) exhibited the most potent cytotoxicity against two human ovarian cancer cells, A2780 and SKOV3, while it did not affect cell viability of normal human ovarian epithelial cells IOSE80PC. Treatment with (−)-asarinin (1) signiﬁcantly induced apoptotic cell death in both A2780 and SKOV3 cells. We further demonstrated that (−)-asarinin (1) stimulated the activation of caspase-3, caspase-8, and caspase-9, and caspase inhibitors signiﬁcantly reversed (−)-asarinin (1)-induced cell death in human ovarian cancer cells. Taken together, these results suggest that (−)-asarinin (1), a major component of the roots of A. sieboldii, induces apoptotic cell death speciﬁcally against cancer cells via the caspase-dependent pathway. Author Contributions: D.S.J. and J.-H.C. conceived and designed the experiments; M.J., H.M.K., and J.S.L. performed the experiments and analyzed the data; H.M.K. and M.J. wrote the paper. Funding: This research was funded by a grant from the Bio-Synergy Research Project (NRF-2015M3A9C4070483) of the Ministry of Science, ICT and Future Planning through the National Research Foundation of Korea (NRF), and by Basic Science Research Program through the NRF funded by the Ministry of Education (NRF-2016R1D1A1B03930222). Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Drew, A.K.; Whyte, I.M.; Bensoussan, A.; Dawson, A.H.; Zhu, X.; Myers, S.P. Chinese herbal medicine toxicology database: Monograph on Herba Asari, “Xi Xin”. J. Toxicol. Clin. Toxicol. 2002, 40, 169–172. [CrossRef] [PubMed] Ramalingam, M.; Kim, S.J. Phytochemical, toxicological and pharmacological studies of Asiasari Radix e Rhizoma: A Review. Trop. J. Pharm. Res. 2015, 14, 545–554. [CrossRef] 3. Kim, H.M.; Moon, Y.S. Asiasari radix inhibits immunoglobulin E production on experimental models in vitro and in vivo. Immunopharm. Immunot. 1999, 21, 469–481. [CrossRef] [PubMed] 4. Kosuge, T.; Yokota, M.; Nukaya, H.; Gotoh, Y.; Nagasawa, M. Studies on antitussive principles of Asiasari radix. Chem. Pharm. Bull. 1978, 26, 2284–2285. [CrossRef] 4. Kosuge, T.; Yokota, M.; Nukaya, H.; Gotoh, Y.; Nagasawa, M. Studies on antitussive principles of Asiasari radix. Chem. Pharm. Bull. 1978, 26, 2284–2285. [CrossRef] 5. Han, A.R.; Kim, H.J.; Shin, M.; Hong, M.; Kim, Y.S.; Bae, H. Constituents of Asarum sieboldii with inhibitory activity on lipopolysaccharide (LPS)-induced NO production in BV-2 microglial cells. Chem. Biodivers. 2008, 5, 346–351. [CrossRef] [PubMed] 5. Han, A.R.; Kim, H.J.; Shin, M.; Hong, M.; Kim, Y.S.; Bae, H. Constituents of Asarum sieboldii with inhibitory activity on lipopolysaccharide (LPS)-induced NO production in BV-2 microglial cells. Chem. Biodivers. 2008, 5, 346–351. [CrossRef] [PubMed] 6. Kim, S.J.; Zhang, C.G.; Lim, J.T. Mechanism of antinociceptive effects of Asarum sieboldii Miq. Radix: Potential role of bradykinin, histamine and opioid receptor-mediated pathways. J. Ethnopharmacol. 2003, 88, 5–9. [CrossRef] 6. Kim, S.J.; Zhang, C.G.; Lim, J.T. Mechanism of antinociceptive effects of Asarum sieboldii Miq. Radix: Potential role of bradykinin, histamine and opioid receptor-mediated pathways. J. Ethnopharmacol. 2003, 88, 5–9. [CrossRef] 7. Lee, J.Y.; Moon, S.S.; Hwang, B.K. Isolation and antifungal activity of kakuol, a propiophenone derivative from Asarum sieboldii rhizome. Pest. Manag. Sci. 2005, 61, 821–825. [CrossRef] [PubMed] 8. Yu, H.H.; Seo, S.J.; Hur, J.M.; Lee, H.S.; Lee, Y.E.; You, Y.O. Asarum sieboldii extracts attenuate growth, acid production, adhesion, and water-insoluble glucan synthesis of Streptococcus mutans. J. Med. Food 2006, 9, 505–509. [CrossRef] [PubMed] 9. Han, Y.; Kim, S.J. Memory enhancing actions of Asiasari radix extracts via activation of insulin receptor and extracellular signal regulated kinase (ERK) I/II in rat hippocampus. Brain Res. 2003, 974, 193–201. [CrossRef] 10. Oh, S.M.; Kim, J.; Lee, J.; Yi, J.M.; Oh, D.S.; Bang, O.S.; Kim, N.S. Anticancer potential of an ethanol extract of Asiasari radix against HCT-116 human colon cancer cells in vitro. Oncol. Lett. References 2013, 5, 305–310. [CrossRef] [PubMed] Molecules 2018, 23, 1849 10 of 11 10 of 11 11. Park, J.D.; Baek, N.I.; Lee, Y.H.; Kim, S.I. Isolation of a cytotoxic agent from Asiasari Radix. Arch. Pharm. Res. 1996, 19, 559–561. [CrossRef] 12. Hashimoto, K.; Katsuhara, T.; Itoh, H.; Ikeya, Y.; Okada, M.; Mitsuhashi, H. Monoterpenes from asiasari radix from Asiasarum sp. Phytochemistry 1990, 29, 3571–3574. [CrossRef] 13. Hashimoto, K.; Katsuhara, T.; Niitsu, K.; Ikeya, Y.; Hayashi, K.; Maruno, M.; Fujita, T. Enantioexcess monoterpenes from roots of Asiasarum sieboldi. Phytochemistry 1994, 35, 969–973. [CrossRef] 14. Wagner, H.; Bauer, R.; Melchart, D.; Xiao, P.G.; Staudinger, A. Radix et Rhizoma Asari—Xixin. In Chromatographic Fingerprint Analysis of Herbal Medicines; Springer: Vienna, Austria, 2011; pp. 45–57, ISBN 978-3-319-09298-0. 15. Lo, C.F.; Chen, C.M. Determination of higenamine in plasma and urine by high-performance liquid chromatography with electrochemical detection. J. Chromatogr. B 1994, 655, 33–39. [CrossRef] 16. Masaki, N.; Iizuka, H.; Yokota, M.; Ochiai, A. Crystal structure of higenamine [1, 2, 3, 4-tetrahydro-1-(4-hydroxybenzyl) isoquinoline-6, 7-diol] hydrobromide. J. Chem. Soc. Perkin 1 1977, 7, 717–719. [CrossRef] 17. Takara, K.; Horibe, S.; Obata, Y.; Yoshikawa, E.; Ohnishi, N.; Yokoyama, T. Effects of 19 herbal extracts on the sensitivity to paclitaxel or 5-ﬂuorouracil in HeLa cells. Biol. Pharm. Bull. 2005, 28, 138–142. [CrossRef] [PubMed] 18. Jeong, M.; Kim, H.M.; Kim, H.J.; Choi, J.H.; Jang, D.S. Kudsuphilactone B, a nortriterpenoid isolated from Schisandra chinensis fruit, induces caspase-dependent apoptosis in human ovarian cancer A2780 cells. Arch. Pharm. Res. 2017, 40, 500–508. [CrossRef] [PubMed] 19. Choi, E.H.; Choi, J.Y.; Lee, J.G.; Oh, J.S.; Kim, D.C.; Lee, H.S.; Son, J.K.; Ryang, S.A.; Kim, J.A.; Lee, S.H. Isolation of melanin biosynthesis inhibitory compounds from the roots of Asarum sieboldii. Korean J. Pharmacogn. 2007, 38, 394–399. 20. Roy, S.C.; Rana, K.K.; Guin, C. Short and stereoselective total synthesis of furano lignans (+/−)-dihydrosesamin, (+/−)-lariciresinol dimethyl ether, (+/−)-acuminatin methyl ether, (+/−)-sanshodiol methyl ether, (+/−)-lariciresinol, (+/−)-acuminatin, and (+/−)-lariciresinol monomethyl ether and furofuran lignans (+/−)-sesamin, (+/−)-eudesmin, (+/−)-piperitol methyl ether, (+/−)-pinoresinol, (+/−)-piperitol, and (+/−)-pinoresinol monomethyl ether by radical cyclization of epoxides using a transition-metal radical source. J. Org. Chem. 2002, 67, 3242–3248. [CrossRef] [PubMed] 21. Musso, L.; Dallavalle, S.; Merlini, L.; Farina, G. Synthesis and Antifungal Activity of 2-Hydroxy-4, 5-methylenedioxyaryl Ketones as Analogues of Kakuol. Chem. Biodivers. 2010, 7, 887–897. [CrossRef] [PubMed] 22. Hofer, O.; Greger, H.; Robien, W.; Werner, A. © 2018 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 13C NMR and 1H lanthanide induced shifts of naturally occurring alkamides with cyclic amide moieties-amides from achilleafalcata. Tetrahedron 1986, 42, 2707–2716. [CrossRef] 23. Bizzo, H.R.; Lopes, D.; Abdala, R.V.; Pimentel, F.A.; de Souza, J.A.; Pereira, M.V.; Bergter, L.; Guimarães, E.F. Sarisan from leaves of Piper afﬁnis hispidinervum C. DC (long pepper). Flavour Fragr. J. 2001, 16, 113–115. [CrossRef] 24. Lopes-Lutz, D.; Mudge, E.; Ippolito, R.; Brown, P.; Schieber, A. Puriﬁcation of alkylamides from Echinacea angustifolia (DC.) Hell. roots by high-speed countercurrent chromatography. J. Agric. Food Chem. 2010, 59, 491–494. [CrossRef] [PubMed] 25. Truan, J.S.; Chen, J.M.; Thompson, L.U. Comparative effects of sesame seed lignan and ﬂaxseed lignan in reducing the growth of human breast tumors (MCF-7) at high levels of circulating estrogen in athymic mice. Nutr. Cancer 2012, 64, 65–71. [CrossRef] [PubMed] 26. Hirose, N.; Doi, F.; Ueki, T.; Akazawa, K.; Chijiiwa, K.; Sugano, M.; Akimoto, K.; Shimizu, S.; Yamada, H. Suppressive effect of sesamin against 7,12-dimethylbenz[a]-anthracene induced rat mammary carcinogenesis. Anticancer Res. 1992, 12, 1259–1265. [PubMed] 27. Deng, P.; Wang, C.; Chen, L.; Wang, C.; Du, Y.; Yan, X.; Chen, M.; Yang, G.; He, G. Sesamin induces cell cycle arrest and apoptosis through the inhibition of signal transducer and activator of transcription 3 signalling in human hepatocellular carcinoma cell line HepG2. Biol. Pharm. Bull. 2013, 36, 1540–1548. [CrossRef] [PubMed] 11 of 11 Molecules 2018, 23, 1849 28. Hibasami, H.; Fujikawa, T.; Takeda, H.; Nishibe, S.; Satoh, T.; Fujisawa, T.; Nakashima, K. Induction of apoptosis by Acanthopanax senticosus HARMS and its component, sesamin in human stomach cancer KATO III cells. Oncol. Rep. 2000, 7, 1213–1219. [CrossRef] [PubMed] 29. Takasaki, M.; Konoshima, T.; Yasuda, I.; Hamano, T.; Tokuda, H. Inhibitory effects of shouseiryu-to on two-stage carcinogenesis. II. Anti-tumor-promoting activities of lignans from Asiasarum heterotropoides var mandshuricum. Biol. Pharm. Bull. 1997, 20, 776–780. [CrossRef] [PubMed] 30. Konoshima, T.; Takasaki, M. Anti-tumor-promoting activities (cancer chemopreventive activities) of natural products. Stud. Nat. Prod. Chem. 2000, 24, 215–267. [CrossRef] 31. Ju, Y.; Still, C.C.; Sacalis, J.N.; Li, J.; Ho, C.T. Cytotoxic coumarins and ligmans from extracts of the northern prickly ash (Zanthoxylum americanum). Phytother. Res. 2001, 15, 441–443. [CrossRef] [PubMed] 32. Elmore, S. Apoptosis: A Review of Programmed Cell Death. Toxicol Pathol. 2007, 35, 495–516. [CrossRef] [PubMed] 33. Wang, C.; Youle, R.J. The Role of Mitochondria in Apoptosis. Annu Rev. Genet. 2009, 42, 95–118. [CrossRef] [PubMed] 34. Kantari, C.; Walczak, H. References Caspase-8 and Bid: Caught in the act between death receptors and mitoch Biochim. Biophys. Acta 2011, 1813, 558–563. [CrossRef] [PubMed] 34. Kantari, C.; Walczak, H. Caspase-8 and Bid: Caught in the act between death receptors and mitochondria. Biochim. Biophys. Acta 2011, 1813, 558–563. [CrossRef] [PubMed] 35 C h E M C t lk I t ti f t ti d i t i i lli th 5. Creagh, E.M. Caspase crosstalk: Integration of apoptotic and innate immune signalling pathw Trends Immunol. 2014, 35, 631–640. [CrossRef] [PubMed] © 2018 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/W3042442795	https://thericejournal.springeropen.com/track/pdf/10.1186/s12284-020-00410-7	English	null	Characterization of Haplotype Diversity in the BADH2 Aroma Gene and Development of a KASP SNP Assay for Predicting Aroma in U.S. Rice	Rice	2,020	cc-by	8,333	© 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/. * Correspondence: afamoso@agcenter.lsu.edu 2H. Rouse Caffey Rice Research Station, Louisiana State University Agricultural Center, 1373 Caffey Rd, Rayne, LA 70578, USA Full list of author information is available at the end of the article ORIGINAL ARTICLE Open Access Characterization of Haplotype Diversity in the BADH2 Aroma Gene and Development of a KASP SNP Assay for Predicting Aroma in U.S. Rice Characterization of Haplotype Diversity in the BADH2 Aroma Gene and Development of a KASP SNP Assay for Predicting Aroma in U.S. Rice Christopher K. Addison1, Brijesh Angira2, Manoch Kongchum2, Dustin L. Harrell2, Niranjan Baisakh1, Steven D. Linscombe2 and Adam N. Famoso2* Abstract Background: Aroma is an important grain quality trait in rice, controlled by mutations within the BADH2 gene. The trait is simply inherited, and its importance in variety development makes it a practical target for marker-assisted selection in applied breeding programs. The predominant functional mutation within BADH2, an 8-bp indel, can be reliably detected using a PCR-based assay, but the available assays and associated genotyping platforms are insufficient for large-scale applied molecular breeding applications and are not compatible with outsourcing genotyping services. Results: We first characterized SNP diversity across the BADH2 gene in a collection of 2932 rice varieties to determine the number of gene haplotypes in O. sativa. Using 297 gene-based SNPs, 11 haplotype groups were detected, and subsequently identified a minimal set of nine informative SNPs that uniquely identified the BADH2 haplotypes. These nine SNPs were developed into KASP assays and used to examine a panel of 369 U.S. rice accessions. The panel represented modern breeding germplasm and included all known aroma pedigree sources in U.S. rice. Six haplotypes were detected within the U.S. panel, of which two were found in majority (85%) of varieties. A representative set of 39 varieties from all haplotype groups was evaluated phenotypically to distinguish aromatic from non-aromatic lines. Conclusion: One haplotype (Hap 6) was found to be perfectly associated with the aromatic phenotype. A single KASP SNP unique to Hap 6 was demonstrated to reliably differentiate aromatic from non-aromatic rice varieties across U.S. germplasm. Keywords: Rice, Aroma, Haplotype analysis, BADH2, SNP Background the aromatic rice global production (Annex, 2011; Giraud, 2013; Mahajan et al., 2018). The USDA forecasts imports of rice for the United States to be 27 million hundredweight (cwt) in 2018, with 23.5 million cwt being aromatic, long- grain varieties classified as either Jasmine or Basmati types (Baldwin and Childs, 2018). Rice is a staple crop that provides the largest amount of daily caloric intake for human populations globally. Most rice is consumed within the country where it was produced, with only 7% of rice production entering the global export market (Giraud, 2013; Muthayya et al., 2014). Aromatic var- ieties account for a significant portion of the export market and include widely recognized types, including Jasmine and Basmati rice. These two types account for the majority of The genetic relationship between fragrant rice varieties and other subpopulations of rice has become clearer with the use of molecular markers. Rice (Oryza sativa) was domesticated roughly 10,000 years ago from pre- differentiated gene pools of the wild grass species Oryza rufipogon (Kovach et al., 2007). These gene pools, Indica and Japonica, are hypothesized to have diverged between Addison et al. Rice (2020) 13:47 https://doi.org/10.1186/s12284-020-00410-7 Addison et al. Rice (2020) 13:47 https://doi.org/10.1186/s12284-020-00410-7 Addison et al. Rice (2020) 13:47 Addison et al. Rice (2020) 13:47 Page 2 of 9 Page 2 of 9 Addison et al. Rice was demonstrated to be an 8-base pair (bp) deletion lo- cated in the seventh exon of the gene (Bradbury et al., 2005a). The proposed pathway of 2AP production begins with proline being catabolized via putrescine into y- aminobutyraldehyde (AB-ald), which is a substrate of BADH2. A functioning BADH2 enzyme converts AB-ald into y-aminobutyric acid (GABA). Fragrant rice lacks a functioning BADH2 enzyme which causes an accumula- tion of AB-ald. Due to the inability of the enzyme to convert AB-ald into GABA, an increased synthesis of 2AP results from the accumulated AB-ald being acety- lated (Bradbury et al., 2008; Chen et al., 2008). Although 10 different alleles have been reported that appear to confer fragrance, the 8-bp deletion is the most predom- inant fragrance allele. The 8-bp insertion/deletion (indel) was present in 93/124 (80%) diverse fragrant varieties from around the world, including common fragrant var- ieties, such as KDML105, Basmati, Della, and Jasmine 85 (Kovach et al., 2009). The discovery of the 8-bp func- tional deletion facilitated the development of an indel DNA marker for use in research and breeding applica- tions (Bradbury et al., 2005a, b). 200,000–400,000 years prior to their domestication (Cai and Morishima, 2002; Ma and Bennetzen, 2004; Vitte et al., 2004). Several studies have identified five genetic- ally distinct subpopulations within these varietal groups. The Indica varietal group contains the indica and aus subpopulations, and the Japonica varietal group contains the temperate japonica, tropical japonica, and aromatic subpopulations (Glaszmann, 1987; Garris et al., 2005; Kovach et al., 2007; Civán et al., 2015; McCouch et al., 2016). The aromatic subpopulation was previously as- sumed to be most closely related to the indica types based on the morphological similarity of the grain, but research utilizing simple sequence repeats demonstrated that the aromatic subpopulation was genetically more closely related to the japonica than the indica subpopu- lation (Sweeney and McCouch, 2007). Today, based on re-sequencing studies, the aromatic subpopulation is interpreted to be an ancient admixture between temper- ate japonica and aus, with a minor dose of indica ances- try (Civán et al., 2015). In the United States, most rice germplasm belongs to the Japonica varietal group, with long-grain germplasm associated with the tropical japon- ica subpopulation and medium-grain germplasm con- taining both tropical and temperate japonica ancestry (Lu et al., 2005; Zhao et al., 2011). The fact that a single gene controls a large proportion of the phenotypic variation for aroma and that the cost and throughput of the phenotyping is a limitation in the breeding of new aromatic lines, makes this an ideal trait for marker-assisted selection. With an increasing interest in the development of fragrant varieties for U.S. rice pro- duction, the availability of a low-cost, high-throughput, single nucleotide polymorphism (SNP)-based DNA marker suitable for screening U.S. breeding material for fragrance and compatible with commonly used KASP genotyping platforms would be of immediate utility for applied breeding programs. The objectives of this re- search were to 1) characterize the haplotype diversity of the BADH2 gene across O. sativa, 2) determine which BADH2 haplotypes/alleles are present in U.S. rice germ- plasm, and 3) develop and validate a SNP-based kompe- titive allele specific polymerase chain reaction (KASP) assay (LGC Group, 2016) informative across U.S. breed- ing germplasm and useful for high-throughput genotypic selection for fragrance in U.S. breeding programs. Thirteen fragrant rice varieties have been developed in the United States, with the fragrance trait derived from six different aromatic varieties that were introduced into U.S. breeding germplasm. The introduced sources of aroma include Delitus, Basmati 370, Jasmine 85, 96A-8, and two additional Basmati types (Linscombe and Famoso, 2017; McClung, 2018; Marchetti et al., 1998; Sha et al., 2011; CCRRF, 2018). The method for detecting fragrance in rice was initially scent-based, involving KOH and ground leaf or stem tis- sue (Sood and Siddiq, 1978). The component, 2-acetyl- 1-pyrroline (2AP), was subsequently identified as the key factor of the aromatic qualities present in fragrant rice varieties (Buttery et al., 1982), and a method was devel- oped for quantifying 2AP with gas chromatography (GC) (Petrov et al., 1996). A strong correlation was observed between the GC method and the standard human scent technique, with the GC method demonstrating less sub- jectivity and increased accuracy (Lorieux et al., 1996). Although GC is a reliable method, the throughput is relatively low, and it is cost prohibitive in the context of an applied breeding program where large sample num- bers are typical. Allele 2: GAAGGTCGGAGTCAACGGATTGGTAG CAATCTAAATAGGCACGTAC, Common: TACTTG TATATACTTGCAGCCATGAATGTT. The functional 8 bp deletion was assayed using a gel based indel marker as described by Bradbury et al., 2005b. PCR was performed as described by Solis et al., 2018, with minor modifications. Briefly, 50 ng of gen- omic DNA was amplified in a total volume of 20 ul that contained 2 ul of 10x PCR buffer, 2.5 ul of 25 mM MgCl2, 2.5 ul of 2 mM dNTP mix, 0.5 ul of 10 uM primers, and 0.2 ul of Taq polymerase (Promega, Madi- son, WI). The thermal profile was initial denaturation for 5 min at 94 °C, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 58 °C for 30 s, elongation at 72 °C for 45 s, and a final extension at 72 °C for 5 min. The PCR products were resolved in 1.5% TAE-agarose gel and viewed/captured in GelLogic 200 imaging station (Kodak Inc., New Haven, CT). KASP Assay Development and Genotyping Kompetative allele specific polymerase chain reaction as- says were designed through LGC genomics (LGC Group, 2016) to characterize the nine SNPs used to differentiate BADH2 haplotype groups (described above). An add- itional KASP assay was designed to characterize the pre- viously reported functional 8-bp deletion in the BADH2 gene (Bradbury et al., 2005a, b). All KASP assay genotyp- ing was performed using the LGC SNPline system fol- lowing standard KASP protocols (LGC Group, 2016) at the H. Rouse Caffey Rice Research Station (HRCRRS). Leaf tissue was collected from the flag leaf of field- grown plants at harvest and dried at 38 °C for 3 days. DNA was extracted using a modified CTAB method for KASP genotyping (Khan et al., 2013). Minimum SNP Set To identify a minimal subset of SNPs that could unam- biguously differentiate the BADH2 haplotypes found in the 3 K genomes, the 255 SNPs found within the BADH2 gene were analyzed across the 11 haplotypes present in the 3 K accessions. The SNPs were sorted by minor al- lele frequency among the 11 haplotype groups. Each SNP was selected one at a time to distinguish the haplo- type groups. Upon each SNP selection, the haplotype groups were further distinguished, and the subsequent SNP selection was based on the SNPs capacity to distin- guish any haplotype groups that were yet to be differen- tiated by the minimum SNP set. A set of 8 SNPs (designated SNP2 – SNP9) was identified that could suc- cessfully distinguish all of the haplotypes, with the ex- ception of Hap 7 and Hap 9, rare aus haplotypes which could not be distinguished. We next sought to identify a single SNP that would be diagnostic of Haplotype 6 Materials and Methods Haplotype Variation in BADH2 in the 3 K Genomes Dataset Single nucleotide polymorphism variation across the BADH2 (LOC_Os08g32870) gene was obtained from the International Rice Informatics Consortium (IRIC) SNP- Seek database, from the “3kAll” SNP dataset (http://snp- seek.irri.org/) (Mansueto et al., 2017). Single nucleotide polymorphism data points with heterozygous allele calls were classified as missing data, and SNPs with > 100 total missing data points were excluded from haplotype A major gene (fgr) was detected in multiple studies for 2AP content. This gene was reported on Chromosome 8 and explained up to 69% of the variation for 2AP con- tent (Ahn et al., 1992; Lorieux et al., 1996; Chen et al., 2006). The underlying gene is a betaine aldehyde de- hydrogenase gene (BADH2), and the functional mutation Addison et al. Rice (2020) 13:47 Addison et al. Rice (2020) 13:47 Page 3 of 9 Page 3 of 9 (Hap 6), the haplotype carried by majority of lines known to be fragrant. None of the SNPs within the BADH2 gene met this criterion, so the 10-kb sequence on either side of the BADH2 gene were examined to identify a SNP (designated SNP 1) that would uniquely distinguish Hap 6. characterization. The graphical genotyping and haplotype visualization software Flapjack (Milne et al., 2010) was used for initial haplotype grouping and characterization. Haplotypes were defined based on a SNP similarity of > 0.99, with genotypes having three or less SNPs different being grouped as a haplotype class. Haplotypes that were present in less than 5% of the genotypes were defined as rare and subsequently omitted from haplotype analysis. Haplotype groups that contained genotypes that differed by three or less SNPs were further divided into subgroups. Subgroups were defined by all genotypes having identical SNP data and the haplotype being present at a minimal frequency of 1% of the total genotypes within the haplo- type group. These haplotype subgroups were designated with an A, B, or C. If a subgroup did not consist of 1% of the total number of haplotypes for the subpopulation, it was considered extremely rare and designated with an “X” (Supplemental Table 1). Genotypes that had missing data for the key SNPs and could not be accurately classified were designated as unclassified (U). U.S. Rice Germplasm Panel The trait marker, SNP1, was additionally designed into an assay by 3CRbio (www.3crbio.com) with the primer sequences: The U.S. rice germplasm panel consists of 369 acces- sions that represent all modern southern U.S. rice breed- ing germplasm, including modern varieties, parental lines, and advanced breeding materials. The panel in- cludes 27 aromatic lines and offspring derived from ped- igrees that include all known aromatic sources used in U.S. breeding. The panel consists of germplasm from Louisiana (131), Arkansas (93), Texas (74), Mississippi (44), and California (8). The international diversity is comprised of germplasm from Italy, Japan, Brazil, Taiwan, and Uruguay. The panel includes short- (7), medium- (45), and long-grain (289) rice accessions as well as a collection of herbicide resistance classes. Allele 1: GAAGGTGACCAAGTTCATGCTTGGTA GCAATCTAAATAGGCACGTAA, Allele 2: GAAGGTCGGAGTCAACGGATTGGTAG CAATCTAAATAGGCACGTAC, Common: TACTTG TATATACTTGCAGCCATGAATGTT. Allele 2: GAAGGTCGGAGTCAACGGATTGGTAG CAATCTAAATAGGCACGTAC, Common: TACTTG TATATACTTGCAGCCATGAATGTT. Allele 2: GAAGGTCGGAGTCAACGGATTGGTAG CAATCTAAATAGGCACGTAC, Common: TACTTG TATATACTTGCAGCCATGAATGTT. Analysis of Fragrance A subset of 39 varieties was selected from the U.S. germ- plasm panel and phenotyped for fragrance. These var- ieties represented all six of the haplotype classes observed in the U.S. germplasm, and 27 fragrant rice varieties that represented all known pedigree sources of aroma in U.S. germplasm. Seeds were obtained from panicle rows grown at the HRCRRS in 2017. Five grams Addison et al. Rice (2020) 13:47 Addison et al. Rice (2020) 13:47 Page 4 of 9 Page 4 of 9 of milled rice of each line were prepared by grinding them into a powder of less than 2.5 mm in diameter with a Cyclone Sample Mill (UDY Corporation). The 2AP concentration was determined using gas chromato- graphic separation on a Shimadzu GC-2010 Plus System (Shimadzu, Columbia, MD) with a flame thermionic de- tector (FTD) following a modified method of the proto- col described by Goufo et al., 2010. Pressuring time, pressure equilibrium time, and injection times were set at 1, 0.01, and 2 min, respectively. The headspace was transferred to the gas chromatograph with a heat trans- fer line for 0.5 min. Oven, sample line, and transfer line temperatures were set at 120, 150, and 160 °C, respect- ively. Gas chromatography and FTD were conducted with the temperatures of the detector set to 280 °C. He- lium was used as the carrier gas with a flow rate of 3.5 mL/min. Data were collected on three replications per sample using LabSolutions software (Shimadzu, Columbia, MD), and the average value was assigned for the 2AP concentration. in Table 1. These haplotypes were relatively minor (0.7– 7%) in the dataset as a whole; however, they are consist- ent with observations that there is greater genetic vari- ation in indica and aus than in the tropical japonica and temperate japonica subpopulations of O. sativa (Kovach et al., 2009; Zhao et al., 2011; Huang et al., 2012; McCouch et al., 2016). Haplotype 6 (Hap 6) was rare in the 3 K genomes (5.8%; n = 171) overall, but it was the most common haplotype within accessions belonging to the aromatic subpopula- tion, present in 51% (n = 38) of accessions in the subpopu- lation. Hap 6 was absent in aus and found at relatively low frequencies in the other subpopulations (1.3% in temper- ate japonica, 3.6% in indica, 11% in tropical japonica). Results BADH2 Sequence and Haplotype Diversity in O. sativa BADH2 Sequence and Haplotype Diversity in O. sativa To characterize the extent of sequence variation across the BADH2 gene, we examined the SNP-Seek database containing re-sequencing information for a collection of 2932 (hereafter referred to as 3 K) diverse varieties of O. sativa (http://snp-seek.irri.org). A total of 297 SNPs was identified in the BADH2 (LOC_Os08g32870) gene. Fil- tering for heterozygous calls and missing data eliminated 42 SNPs, leaving 255 SNPs that were utilized for con- struction of BADH2 gene haplotypes. The number of SNPs observed at the BADH2 gene in the SNP-Seek database was similar to previously reported genes (Wang et al., 2018). Eleven haplotype groups were identified and summarized in Supplemental Table 1. All but 143 accessions (4.8%) could be classified into one of the 11 haplotype groups. Accessions that could not be classified included 74 that had rare haplotypes (present at a fre- quency of 2.5%), and 69 that were unclassified due to missing data (> 100 data points missing). Known fragrant accessions, such as Domsiah, Basmati 1, Gerdeh, and the Thai Jasmine variety, Khao Dawk Mali 105, all carried Hap 6, and all had been docu- mented to contain the functional 8-bp deletion causing fragrance in rice (Mahatheeranont et al., 2001; Bradbury et al., 2005a; Kovach et al., 2009). Based on these obser- vations, it was hypothesized that Hap 6 carries the func- tional 8-bp deletion that is responsible for the fragrant phenotype, and that all Hap 6-containing accessions would be phenotypically fragrant. To confirm this hypothesis, we first interrogated the SNP- Seek database to find accessions carrying the 8-bp deletion. The deletion was observed in 128 of the 3 K accessions, all of which were classified as carrying Hap 6, but interestingly, 43 Hap 6-containing accessions in SNP-Seek did not appear to carry the deletion. Among these accessions were two well-known accessions, Khao Dawk Mali 105 and Nerica 1 which are known to be fragrant. Previous studies had clearly documented that these two fragrant varieties did, in fact, carry the 8-bp deletion (Kovach et al., 2009; Asante et al., 2010). This observation suggested that the indel data from the SNP-Seek database was not detecting the 8-bp deletion in some samples or that different seed sources of these ac- cessions were used across different studies. Results Haplotype 1 (Hap 1) was the most common haplotype in the dataset, present in 892 (30%) accessions. It was the most frequent haplotype in indica (42%; n = 749), and the second most frequent haplotype in the tropical japonica subpopulation (26%; n = 121) (Table 1). Haplo- type 2 (Hap 2) was the second most common haplotype in the 3 K genomes dataset, present in 755 (26%) acces- sions. It was the predominant haplotype within the Ja- ponica varietal group (70%; n = 586), present in 96% of temperate japonica and in 60% of tropical japonica ac- cessions (Table 1). Analysis of Fragrance The aromatic subpopulation only accounts for 2.6% (75 accessions) in the 3 K genomes dataset. Due to the rela- tively lower representation, there is a bias in the number of Hap 6 containing accessions across subpopulations in this study. For example, of the 171 accessions containing Hap 6, 22% were from the aromatic subpopulation, 38% were from the indica subpopulation, and 31% were from the tropical japonica subpopulation (Table 1). This obser- vation highlights the fact that some accessions do contain the aroma allele at BADH2, while belonging to another subpopulation on the whole genome level. Development of KASP Assays for Haplotype Characterization To attempt to resolve the discrepancy, attention was turned to the development of KASP assays that could Haplotypes 3, 4, 5, 7, 8, 9, 10, and 11 were predomin- antly carried by indica or aus accessions, as summarized Page 5 of 9 Addison et al. Rice (2020) 13:47 Addison et al. Rice Table 1 Haplotype characterization of the BADH2 gene across the global rice diversity panel Subpopulation Ancestryb 20374951a 20,381,787 20,382,161 20,384,350 20,382,286 20,380,447 20,381,308 20,380,886 20,382,480 SNP1c SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 Hap Group Aus Indica Admixd ARO. TEJ. TRJ. Japx Total 601e 548 536 603 531 849 852 855 N/A 1 1 749 12 0 1 121 8 892 G A G G A G G T C 2 2 148 19 26 215 281 64 755 G T A G A C G T C 3 0 201 1 0 1 0 0 203 G A G A A C C T C 4 9 183 5 0 0 0 0 197 G A G G A G C T C 5 1 186 5 2 0 1 0 195 G T G G A C C T C 6 (aromatic) 0 65 9 38 3 53 3 171 T T A G A C C T C 7 97 11 6 0 0 0 0 114 G A G G A C C T C 8 0 89 2 1 0 2 0 94 G A G G G G G T C 9 63 13 6 3 0 0 0 85 G A G G A C C T C 10 1 39 6 0 2 4 0 52 G A G G A G G T T 11 3 22 6 0 0 0 0 31 G A G G A G C C C R 22 42 6 2 0 2 0 74 – – – – – – – – – U 2 41 16 3 2 3 2 69 – – – – – – – – – Total 201 1789 99 75 224 467 77 2932 aSNP positions are based on IRGSPv.1 bSubpopulation assignments based on SNP-Seek assignments, with all Indica subgroups classified as Indica cRepresents the SNP ID used throughout manuscript dAdmix, admixture; ARO, aromatic; TEJ, temperate japonica; TRJ, tropical japonica; Japx, japonica admixture eRepresents the LSU KASP SNP ID. Development of KASP Assays for Haplotype Characterization Hap 10 was observed in only 52 genotypes in the 3 K genomes dataset, primarily of indica origin from South and Southeast Asia, and was considered unlikely to be present in the U.S. germplasm; thus, for the purposes of this study, U.S. genotypes characterized as having Hap 1 or Hap 10 were classified as Hap 1. In addition, a KASP assay was designed to directly characterize the 8-bp functional deletion in the BADH2 gene. This assay accurately classified all non-deletion, non-aromatic genotypes as such, but it was unreliable for classifying accessions carrying the deletion. Indeed, 50% of known fragrant U.S. varieties were misclassified into the non-fragrant group using this KASP marker, a problem that was similar to that observed in the SNP- Seek database. Given that the U.S. panel contained ex- perimentally confirmed fragrant and non-fragrant phe- notypes, had well-documented pedigrees, and that all donor varieties carried Hap 6 and the 8-bp deletion causing fragrance, it was concluded that the KASP assay was unreliable, and therefore, all subsequent analysis fo- cused on haplotype classification using the eight reliable SNPs (Table 2). To confirm the presence or absence of fragrance in the U.S. panel, a representative sample of each haplotype group was phenotyped for 2AP content. A set of 39 ac- cessions was phenotyped using gas chromatographic separation to quantify 2AP content in milled rice sam- ples (Table 2). Seventeen accessions that contained Hap 6 and had been previously characterized as fragrant var- ieties based on breeder descriptions were phenotyped and all were detected to contain 2AP. The average 2AP content for Hap 6-containing varieties was 1.30 ppm and ranged from 0.356 to 3.165 ppm. The range of measured 2AP content in fragrant genotypes demonstrates that add- itional, small-effect genes are involved in determining the level of 2AP production. A set of 22 varieties from the U.S. panel, representing all of the other haplotype classes, contained 2AP levels at or near zero, demonstrating that Hap 6 is the only aromatic haplotype in the U.S. rice panel. Based on these findings, it was concluded that the BADH2 allele referred to in this study as Hap 6, is the only allele conferring fragrance in U.S. germplasm. Of the 11 haplotype groups present in the 3 K ge- nomes dataset, six were identified in the U.S. Development of KASP Assays for Haplotype Characterization SNP9 was not developed into a KASP assay otype characterization of the BADH2 gene across the global rice diversity panel pulation Ancestryb 20374951a 20,381,787 20,382,161 20,384,350 20,382,2 emphasis on U.S. germplasm in our study, differentiating haplotypes 7 and 9 was not considered a priority. reliably identify the major BADH2 haplotypes, with par- ticular attention on a rapid and cost-effective assay that could diagnose Hap 6 and tag the functional 8-bp deletion causing fragrance. A minimal set of 8 SNPs (designated SNP2 – SNP9), selected from the 255 SNPs segregating in the BADH2 gene, was selected that could unequivocally identify all major haplotypes, with the exception of Hap 9 which could not be distinguished from Hap 7 (Fig. 1). These two, relatively rare haplotype groups were primarily found in the aus subpopulation (Table 1) and given the Next, a single SNP that would be uniquely diagnostic of Hap 6, the haplotype carried by majority of lines known to be fragrant, was sought to be identified. None of the SNPs within the BADH2 gene met this criterion, so the 10-kb sequence upstream and downstream of the BADH2 gene was examined. A Hap 6-diagnostic SNP, designated SNP 1, was identified 7908 bp from the BADH2 gene. This SNP is invariant in all the other haplotypes (Fig. 1). Fig. 1 Haplotype Characterization and Phenotypic Validation Workflow. Legend: Workflow utilized to leverage publicly available genomic information to develop a KASP SNP marker for targeted application in U.S. rice breeding germplasm Fig. 1 Haplotype Characterization and Phenotypic Validation Workflow. Legend: Workflow utilized to leverage publicly available genomic information to develop a KASP SNP marker for targeted application in U.S. rice breeding germplasm Fig. 1 Haplotype Characterization and Phenotypic Validation Workflow. Legend: Workflow utilized to leverage publicly available genomic information to develop a KASP SNP marker for targeted application in U.S. rice breeding germplasm Addison et al. Rice (2020) 13:47 Page 6 of 9 Addison et al. Rice Page 6 of 9 Hap 6 was observed in 27 accessions in the U.S. germ- plasm panel and 17 of these lines were selected as a sub- set for phenotyping for 2AP content. All 17 lines that contained Hap 6 were phenotypically characterized as fragrant. This is consistent with the hypothesis that Hap 6 carries the 8-bp deletion in the BADH2 gene, and that fragrance in southern U.S. germplasm is perfectly pre- dicted by the presence of Hap 6. Development of KASP Assays for Haplotype Characterization Thus it was concluded that fragrant U.S. germplasm contains the same BADH2 allele as identified in the Thai Jasmine line KDML105 (Kovach et al., 2009). BADH2 Diversity in U.S. Breeding Germplasm The 3 K genomes dataset included only 32 varieties from the United States and did not include any modern-day varieties or breeding germplasm. Therefore, to characterize the haplotype diversity present in modern U.S. breeding germplasm, a panel of 369 lines represent- ing modern varieties and elite breeding germplasm from the southern United States was assembled, including 27 accessions classified as aromatic based on variety/ breeder description (Supplemental Table 2). This panel was characterized for haplotype diversity using the SNP- based KASP assays described above. Three haplotypes (Hap 3, Hap 5, and Hap 7/9) strongly associated with the Indica varietal group in the 3 K genomes dataset were observed in a small set of U.S. varieties (Table 2). Hap 3 was identified in 11 breeding lines, 10 of which shared historical pedigree relationships that traced back to the Texas variety ‘Dawn’ (Supple- mental Table 2) (Bollich et al., 1968). The key intermedi- ates between Dawn and the 11 modern breeding lines include ‘Trenasse’ and ‘Catahoula’ (Linscombe et al., 2006; Blanche et al., 2009). It was concluded that, in these breeding lines, Hap 3 derives from Dawn, which inherited the BADH2 gene from an indica ancestor. The one additional U.S. line containing Hap 3 was ‘CL151’, which has the indica variety ‘Taducan’ in its pedigree and is the likely source of Hap 3 in that lineage (Blanche et al., 2011). Hap 5 was observed in five U.S. lines, two of which are recent introductions from outside the United States, and two are new Provisia herbicide- resistant breeding lines derived from indica breeding germplasm. Hap 7/9 is present in 11 U.S. varieties, eight of which were developed in Arkansas and share similar pedigree histories that trace back to the introduction of ‘Rexoro’ (Johnston, 1958). Eight of the nine SNPs selected for KASP assay devel- opment were successfully converted into KASP assays and used to screen the U.S. panel. SNP9, which was ini- tially included to differentiate Hap 1 from a rare haplo- type, Hap 10, in the 3 K genomes dataset, failed to convert and was omitted from further analysis. Discussion and Conclusions The objective of this research was to leverage publicly available, high density, SNP data from O. sativa to de- velop a rapid, cost-effective DNA marker assay for use in an applied U.S. rice breeding program. A common hurdle in the utilization of public genomic resources for applied breeding objectives is that elite germplasm of interest to plant breeders often is not well represented. A valuable feature of the 3 K genomes SNP-Seek data- base is that it represents the global diversity of domesti- cated Asian rice (O. sativa). It is reasonable to assume that haplotype alleles present in a specific rice breeding program are represented within the 3 K genomes, even if the specific accessions are not. In this study, the 3 K genotype data was utilized to characterize, identify, and validate that the BADH2 allele that confers fragrance in U.S. rice germplasm, without any fragrant U.S. varieties being present in the 3 K genomes dataset. By character- izing the haplotype diversity found in the 3 K genomes using 255 genic SNPs, a subset of nine SNPs was identi- fied that together could distinguish all 11 haplotypes, ef- fectively obtaining the same resolution as with the full 255-SNP set. Hap 6 was determined to be the aromatic haplotype due to 128 lines within the haplotype contain- ing the 8-bp deletion. The technical performance of the SNP1 KASP assay was evaluated in three separate genotyping jobs, run across 351 lines from the U.S. germplasm panel. The data return across the 1053 reactions was 99%., with 11 reactions resulting in missing data. Seven reactions failed to amplify and four reactions resulted in unclear cluster- ing, where the data point clustered between the hetero- zygous and homozygous G clusters. The reproducibility across the three genotyping jobs was 100%. A segregat- ing population of 96 F2:F3 plants from a bi-parental breeding population was evaluated to investigate the clustering of the assay and the potential for auto-scoring in the presence of the heterozygous class. The SNP1 KASP assay produced clearly defined clusters and was capable of autoscoring, thereby making this assay amen- able to high-throughput KASP genotyping for breeding applications (Supplemental Figure 1). Based on these validation procedure, we conclude that the KASP assay of SNP1 (Chr 8: 20,374,951 bp) is capable of differentiat- ing aromatic from non-aromatic lines across U.S. rice Then a panel of U.S. KASP Assay Validation for Use as BADH2 Trait Marker KASP Assay Validation for Use as BADH2 Trait Marker The haplotype SNP (SNP1) located at 20,374,951 MB is capable of differentiating the aromatic Hap 6 from all other non-aromatic haplotypes and hypothesized to be in linkage disequilibrium (LD) with the functional 8 bp deletion. The U.S. rice germplasm panel was screened with the previously reported gel-based marker that dir- ectly assays the 8 bp deletion in order to confirm the SNP is in LD with the indel (Bradbury et al., 2005b). The allele calls observed with the functional indel marker matched perfectly with the SNP1 KASP allele calls, demonstrating that the SNP1 KASP assay is in per- fect LD across the U.S. germplasm panel (Supplemental Table 2). Development of KASP Assays for Haplotype Characterization germplasm, with Hap 1 (49%, n = 181) and Hap 2 (36%, n = 134) cu- mulatively accounting for 85% of the panel (Table 2). Accessions carrying Hap 1 were predominantly (92%) long-grain varieties, but it was also found in 14 medium- grain and one short-grain variety. Hap 2 was most com- mon in medium-grain varieties, present in 67% of the 45 medium-grain accessions in the U.S. panel. Given that long-grain U.S. varieties are primarily classified as trop- ical japonicas, and medium-grain varieties as tropical and temperate japonica, the frequencies of Hap 1 and Hap 2 in the U.S. panel is consistent with observations of haplotype frequencies in the 3 K genomes dataset. Page 7 of 9 Addison et al. Rice (2020) 13:47 Table 2 Haplotype characterization and 2AP concentration across U.S. breeding germplasm panel 20374951a 20,381,787 20,382,161 20,384,350 20,382,28 Table 2 Haplotype characterization and 2AP concentration across U.S. breeding germplasm panel 20374951a 20,381,787 20,382,161 20,384,350 20,382,286 20,380,447 20,381,308 20,380,886 2AP Concentration Test (ppm)b SNP1c SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 Hap Group Number n Range Mean (SE)d 601e 548 536 603 531 849 852 855 1 181 5 0–0 0 G A G G A G G T 2 134 9 0–0.079 0.009 (0.005) G T A G A C C T 3 11 2 0–0 0 G A G A A C C T 5 5 3 0–0.001 0.0002 (0.0002) G T G G A C C T 6 (aromatic) 27 17 0.356– 3.165 1.30 (0.058) T T A G A C C T 7 or 9 11 3 0–0.009 0.003 (0.002) G A G G A C C T aSNP positions are based on IRGSPv.1 and are on Chromosome 8 bppm, parts per million cRepresents the SNP ID used throughout manuscript dValues in the parentheses represent the standard error of the 2AP Concentration eRepresents the LSU KASP SNP ID breeding germplasm and is suitable for high-throughput breeding applications. Funding The research project was supported in part by the Louisiana Rice Research Board, the USDA National Institute of Food and Agriculture, and the US National Science Foundation. Discussion and Conclusions germplasm was genotyped using the nine SNPs to characterize BADH2 gene diversity in this panel. This set of nine SNPs could be applied in a similar fashion to other breeding programs’ target germ- plasm, assuming the haplotypes present in the target germplasm were represented in the 3 K genomes dataset. Addison et al. Rice (2020) 13:47 Addison et al. Rice (2020) 13:47 Page 8 of 9 The panel of U.S. germplasm included all modern U.S. varieties and elite breeding lines, thereby representing the pool of potential parents of new breeding crosses. In addition, the panel included fragrant lines whose pedi- grees traced back to all original sources of fragrance used in U.S. rice breeding. Thus, the aroma trait SNP (SNP1) validated across this panel can be used with con- fidence in any breeding crosses that are derived from materials in the panel. classes showed clear clustering suitable for automated computer scoring with the exception of one line (indicated in grey). classes showed clear clustering suitable for automated computer scoring with the exception of one line (indicated in grey). Received: 7 April 2020 Accepted: 6 July 2020 Received: 7 April 2020 Accepted: 6 July 2020 Abbreviations Hap: Haplotype; KDML105: Khao Dawk Mali 105; BADH2: Betaine aldehyde dehydrogenase gene; KASP: Kompetitive allele specific polymerase chain reaction; SNP: Single nucleotide polymorphism; bp: Base pair; 2AP: 2-acetyl-1- pyrroline; SSRs: Simple sequence repeats; GC: Gas chromatography; AB-ald: y- aminobutyraldehyde; GABA: y-aminobutyric acid; indel: insertion/deletion; FTD: Flame thermionic detector Although this SNP has been validated across the target U.S. breeding germplasm, it should be noted that the na- ture of LD based markers is germplasm dependent. The aroma SNP (SNP1) identified in this work was in perfect LD across the target germplasm; however, it is very pos- sible that this SNP will not be in LD with the 8 bp dele- tion across all germplasm globally. Thus, as with any LD based marker, validation in the target breeding germ- plasm is necessary prior to deployment for breeding ap- plications. Similarly, attempts in this work to develop a KASP assay on the 8 bp deletion were not successful when test on U.S. japonica rice, thus it is possible that the KASP assay could be successfully utilized on indica germplasm or a different assay design could provide better results. Acknowledgements Susan McCouch provided valuable feedback through a comprehensive technical and editorial review. A special acknowledgement to Jennifer Dartez and Hanamareddy Biradar for the technical contributions in the SNP genotyping and gel assay, respectively and to Valerie Dartez for helpful contributions in manuscript preparation and proofreading. Authors’ Contributions CA, SL, and AF designed the experiments and analyzed and interpreted the data. CA and AF wrote the manuscript. CA, BA, NB, MK, and DH performed the experiments. All authors edited and approved the final manuscript. Consent for Publication Consent for Publication Yes, all authors have provided consent for publication. Author details 1 1School of Plant, Environmental and Soil Science, Louisiana State University, 104 Sturgis Hall, Baton Rouge, LA 70803, USA. 2H. Rouse Caffey Rice Research Station, Louisiana State University Agricultural Center, 1373 Caffey Rd, Rayne, LA 70578, USA. Availability of Data and Materials Datasets supporting the results are included as supplemental files. Datasets supporting the results are included as supplemental files. The ultimate objective of this research was to identify a single SNP that could be used for selection of fragrant varieties within U.S. breeding programs. For deploy- ment in an applied breeding program, a DNA marker must be accurate across the target germplasm, amen- able to high-throughput genotyping, and effective on low cost genotyping platforms/chemistries (Platten et al., 2019). Many breeding programs utilize KASP genotyping chemistry (LGC Group, 2016), which is ideal for SNP assays but can also detect small indels. In this study, the development of a KASP assay targeting the causal 8-bp deletion for fragrance was attempted but had poor and inconsistent genotyping results. It is common that a functional mutation would involve an indel, and in cases where developing a robust KASP assay on the indel itself is unsuccessful, the haplotype approach described in this study can be effectively uti- lized to identify a SNP in linkage disequilibrium with the functional polymorphism. Ethics Approval and Consent to Participate Not applicable. Ethics Approval and Consent to Participate Not applicable. Consent for Publication Yes, all authors have provided consent for publication. References Ahn SN, Bollich CN, Tanksley SD (1992) RFLP tagging of a gene for aroma in rice. Theor Appl Genet 84(7–8):825–828. https://doi.org/10.1007/BF00227391 Annex (2011) Business process analysis of the export of jasmine rice from h l d h U d S B P A l G d S l T d Ahn SN, Bollich CN, Tanksley SD (1992) RFLP tagging of a gene for aroma in rice. Theor Appl Genet 84(7–8):825–828. https://doi.org/10.1007/BF00227391 Annex (2011) Business process analysis of the export of jasmine rice from , , y ( ) gg g g Theor Appl Genet 84(7–8):825–828. https://doi.org/10.1007/BF00227391 Annex (2011) Business process analysis of the export of jasmine rice from thailand to the United States. Buiness Process Anal. Guid. to Simpl. Trade Proced., pp 1–68 http://www.unescap.org/tid/publication/tipub2558.asp (Accessed 2 Feb 2019) thailand to the United States. Buiness Process Anal. Guid. to Simpl. Trade Proced., pp 1–68 http://www.unescap.org/tid/publication/tipub2558.asp (Accessed 2 Feb 2019) Asante MD, Kovach MJ, Huang L, Harrington S, Dartey PK et al (2010) The genetic origin of fragrance in NERICA1. Mol Breed 26(3):419–424. https://doi.org/10. 1007/s11032-009-9382-8 Supplementary information Supplementary information accompanies this paper at https://doi.org/10. 1186/s12284-020-00410-7. Additional file 1: Supplemental Table 1. BADH2 SNP output and haplotype assignments from 3 K genomes data set. Supplemental Table 2. Haplotype characterization of the U.S. breeding germplasm panel using the minimal SNP set. Additional file 2: Supplemental Figure 1. Technical validation of Aroma SNP1 across segregating F2:F3 breeding population. All genotypic Supplementary information Supplementary information accompanies this paper at https://doi.org/10. 1186/s12284-020-00410-7. pp y Supplementary information accompanies this paper at https://doi.org/10. 1186/s12284-020-00410-7. Baldwin, K., and N. Childs. 2018. USDA Rice Outlook. USDA Rice Outlook. https:// www.ers.usda.gov/webdocs/publications/89598/rcs-18g.pdf?v=0 (Accessed 13 Dec 2018). Additional file 1: Supplemental Table 1. BADH2 SNP output and haplotype assignments from 3 K genomes data set. Supplemental Table 2. Haplotype characterization of the U.S. breeding germplasm panel using the minimal SNP set. Blanche SB, Linscombe SD, Sha X, Bearb KF, Groth DE et al (2009) Registration of ‘Catahoula’ Rice. J Plant Regist 3(2):146. https://doi.org/10.3198/jpr2008.11. 0677crc Supplemental Blanche SB, Sha X, Harrell DL, Groth DE, Bearb KF et al (2011) Registration of ‘CL151’ Rice. J Plant Regist 5(2):177. https://doi.org/10.3198/jpr2010.05.0245crc Bollich C, Atkins J, Scott J, Webb B (1968) Regristration of “Dawn” Rice. Crop Sci 8(3):1968. https://doi.org/10.2135/cropsci1968.0011183X000800030058x Additional file 2: Supplemental Figure 1. Technical validation of Aroma SNP1 across segregating F2:F3 breeding population. All genotyp Additional file 2: Supplemental Figure 1. References Technical validation of Aroma SNP1 across segregating F2:F3 breeding population. All genotypic Page 9 of 9 Page 9 of 9 Addison et al. Rice (2020) 13:47 Addison et al. Rice (2020) 13:47 Page 9 of 9 Addison et al. Rice (2 Bradbury LMT, Fitzgerald TL, Henry RJ, Jin Q, Waters DLE (2005a) The gene for fragrance in rice. Plant Biotechnol J 3(3):363–370. https://doi.org/10.1111/j. 1467-7652.2005.00131.x Mahatheeranont S, Keawsa-ard S, Dumri K (2001) Quantification of the rice aroma compound, 2-acetyl-1-pyrroline, in uncooked Khao Dawk Mali 105 brown rice. J Agric Food Chem 49(2):773–779. https://doi.org/10.1021/jf000885y Mansueto L, Fuentes RR, Borja FN, Detras J, Abrio-Santos JM et al (2017) Rice SNP-seek database update: New SNPs, indels, and queries. Nucleic Acids Res 45(D1):D1075–D1081. https://doi.org/10.1093/nar/gkw1135 Bradbury LMT, Gillies SA, Brushett DJ, Waters DLE, Henry RJ (2008) Inactivation of an aminoaldehyde dehydrogenase is responsible for fragrance in rice. Plant Mol Biol 68(4–5):439–449. https://doi.org/10.1007/s11103-008-9381-x Marchetti MA, Bollich CN, Webb BD, Jackson BR, McClung AM et al (1998) Registration of “Jasmine 85” Rice. Crop Sci (38):896 Marchetti MA, Bollich CN, Webb BD, Jackson BR, McClung Bradbury LMT, Henry RJ, Jin Q, Reinke RF, Waters DLE (2005b) A perfect marker for fragrance genotyping in rice. Mol Breed 16(4):279–283. https://doi.org/10. 1007/s11032-005-0776-y Registration of “Jasmine 85” Rice. Crop Sci (38):896 McClung, A.M. 2018. New Rice Varieties. United States Dep. Agric. - Agric. Res. Serv. https://www.ars.usda.gov/oc/br/rice/index/ (Accessed 29 Mar 2019). Buttery RG, Ling LC, Juliano BO (1982) 2-acetyl-1-pyrroline: an important aroma component of cooked rice. Chem Ind 23:958–959 McCouch SR, Wright MH, Tung CW, Maron LG, McNally KL et al (2016) Open access resources for genome-wide association mapping in rice. Nat Commun 7. https://doi.org/10.1038/ncomms10532 Cai HW, Morishima H (2002) QTL clusters reflect character associations in wild and cultivated rice. Theor Appl Genet 104(8):1217–1228. https://doi.org/10. 1007/s00122-001-0819-7 Milne I, Shaw P, Stephen G, Bayer M, Cardle L et al (2010) Flapjack-graphical genotype visualization. Bioinformatics 26(24):3133–3134. https://doi.org/10. 1093/bioinformatics/btq580 CCRRF. 2018. Varieties and Seed. Calif. Coop. Rice Res. Found. Exp. Stn. http:// www.crrf.org/ccrrf_res_019.htm (Accessed 29 Mar 2019). Muthayya, S., J.D. Sugimoto, S. Montgomery, and G.F. Maberly. 2014. An overview of global rice production, supply, trade, and consumption. (1324): 7–14. doi: https://doi.org/10.1111/nyas.12540. Chen S, Wu J, Yang Y, Shi W, Xu M (2006) The fgr gene responsible for rice fragrance was restricted within 69 kb. Plant Sci 171(4):505–514. https://doi. References https://doi.org/10.1093/aob/mcm128 Vitte C, Ishii T, Lamy F, Brar D, Panaud O (2004) Genomic paleontology provides evidence for two distinct origins of Asian rice (Oryza sativa L.). Mol Genet Genomics 272(5):504–511. https://doi.org/10.1007/s00438-004-1069-6 Wang W, Mauleon R, Hu Z, Chebotarov D, Tai S et al (2018) Genomic variation in 3010 diverse accessions of Asian cultivated rice. Nature 557(7703):43–49. https://doi.org/10.1038/s41586-018-0063-9 Goufo P, Wongpornchai S, Tang XR (2010) Decrease in rice aroma after application of growth regulators. Agron Sustain Dev 31(2):349–359. https:// doi.org/10.1051/agro/2010011 Sweeney M, McCouch S (2007) The complex history of the domestication of rice. Ann Bot 100(5):951–957. https://doi.org/10.1093/aob/mcm128 p g Vitte C, Ishii T, Lamy F, Brar D, Panaud O (2004) Genomic paleontology provides evidence for two distinct origins of Asian rice (Oryza sativa L.). Mol Genet Genomics 272(5):504–511. https://doi.org/10.1007/s00438-004-1069-6 Wang W, Mauleon R, Hu Z, Chebotarov D, Tai S et al (2018) Genomic variation in 3010 diverse accessions of Asian cultivated rice. Nature 557(7703):43–49. https://doi.org/10.1038/s41586-018-0063-9 p g Vitte C, Ishii T, Lamy F, Brar D, Panaud O (2004) Genomic paleontology provides evidence for two distinct origins of Asian rice (Oryza sativa L.). Mol Genet Genomics 272(5):504–511. https://doi.org/10.1007/s00438-004-1069-6 Wang W, Mauleon R, Hu Z, Chebotarov D, Tai S et al (2018) Genomic variation in 3010 diverse accessions of Asian cultivated rice. Nature 557(7703):43–49. https://doi.org/10.1038/s41586-018-0063-9 Huang X, Kurata N, Wei X, Wang Z, Wang A et al (2012) A map of rice genome variation reveals the origin of cultivated rice. Nature. https://doi.org/10.1038/ nature11532 hnston TH (1958) Registration of Rice Varieties’. Agron J 10:694–700 Khan NA, Bedre R, Parco A, Bernaola L, Hale A et al (2013) Identification of cold- responsive genes in energycane for their use in genetic diversity analysis and future functional marker development. Plant Sci 211:122–131. https://doi.org/ 10.1016/j.plantsci.2013.07.001 Zhao K, Tung CW, Eizenga GC, Wright MH, Ali ML et al (2011) Genome-wide association mapping reveals a rich genetic architecture of complex traits in Oryza sativa. Nat Commun 2(1):1–10. https://doi.org/10.1038/ncomms1467 Kovach MJ, Calingacion MN, Fitzgerald MA, McCouch SR (2009) The origin and evolution of fragrance in rice (Oryza sativa L.). Proc. Natl. Acad. Sci 106(34): 14444–14,449. https://doi.org/10.1073/pnas.0904077106 Publisher’s Note Kovach MJ, Sweeney MT, McCouch SR (2007) New insights into the history of rice domestication. Trends Genet 23(11):578–587. https://doi.org/10.1016/j.tig. 2007.08.012 Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. LGC Group. 2016. SNPline genotyping automation. https://www.lgcgroup.com/ products/genotyping-instruments/snpline/#.XFCv91xKhaQ (Accessed 29 Mar 2019). Linscombe, S.D., and A.N. Famoso. 2017. Rice Station Released Varieties 1917– 2015. LSU AgCenter. https://www.lsuagcenter.com/profiles/lblack/articles/ page1484946379503 (Accessed 29 Mar 2019). References org/10.1016/j.plantsci.2006.05.013 Petrov M, Danzart M, Giampaoli P, Faure J, Richard H (1996) Rice aroma analysis: discrimination between a scented and a non-scented rice. Sci Aliments (16): 347–360 g j Chen S, Yang Y, Shi W, Ji Q, He F et al (2008) Badh2, encoding betaine aldehyde dehydrogenase, inhibits the biosynthesis of 2-acetyl-1-pyrroline, a major component in rice fragrance. Plant Cell Online 20(7):1850–1861. https://doi. org/10.1105/tpc.108.058917 Platten JD, Cobb JN, Zantua RE (2019) Criteria for evaluating molecular markers: Comprehensive quality metrics to improve marker-assisted selection. PLoS One 14(1):1–20. https://doi.org/10.1371/journal.pone.0210529 Civán P, Craig H, Cox CJ, Brown TA (2015) Three geographically separate domestications of Asian rice. Nat Plants 1(November):1–5. https://doi.org/10. 1038/nplants.2015.164 One 14(1):1–20. https://doi.org/10.1371/journal.pone.0210529 Sha XY, Linscombe SD, Groth DE, Harrell DL, White LM et al (2011) Registration of ‘Jazzman’ Aromatic Long-Grain Rice. J Plant Regist 5(3):304–308. https://doi. org/10.3198/jpr2012.07.0012crc Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169(3):1631–1638. https://doi.org/10.1534/ genetics.104.035642 Solis J, Gutierrez A, Mangu V, Sánchez Timm E, Bedre R et al (2018) Genetic mapping of quantitative trait loci for grain yield under drought in rice under controlled greenhouse conditions. Front Chem 5(January):1–12. https://doi. org/10.3389/fchem.2017.00129 Solis J, Gutierrez A, Mangu V, Sánchez Timm E, Bedre R et al (2018) Genetic mapping of quantitative trait loci for grain yield under drought in rice under controlled greenhouse conditions. Front Chem 5(January):1–12. https://doi. org/10.3389/fchem.2017.00129 Giraud G (2013) The World Market of Fragrant Rice, Main Issues and Perspectives The Present Market of Fragrant Rice. Int Food Agribus Manag Rev 16(2):1–20. Glaszmann JC (1987) Isozymes and classification of Asian rice varieties Theor Giraud G (2013) The World Market of Fragrant Rice, Main Issues and Perspectives The Present Market of Fragrant Rice. Int Food Agribus Manag Rev 16(2):1–20. Glaszmann JC (1987) Isozymes and classification of Asian rice varieties. Theor Appl Genet 74(1):21–30. https://doi.org/10.1007/BF00290078 Sood BC, Siddiq EA (1978) A rapid technique for scent determination in rice. Indian J Genet Plant Breed 38(2):151–275 Sood BC, Siddiq EA (1978) A rapid technique for scent determination in rice. Indian J Genet Plant Breed 38(2):151–275 Sood BC, Siddiq EA (1978) A rapid technique for scent determination in rice. Indian J Genet Plant Breed 38(2):151–275 Sweeney M, McCouch S (2007) The complex history of the domestication of rice. Ann Bot 100(5):951–957. Linscombe SD, Sha X, Bond JA, Bearb K, Rush MC et al (2006) Registration of ‘Trenasse’ Rice. Crop Sci 46(5):2318. https://doi.org/10.2135/cropsci2006.03.0208 Lorieux M, Petrov M, Huang N, Guiderdoni E, Ghesquière A (1996) Aroma in rice: genetic analysis of a quantitative trait. Theor Appl Genet 93(7):1145–1151. https://doi.org/10.1007/BF00230138 Lu H, Redus MA, Coburn JR, Rutger JN, McCouch SR et al (2005) Population structure and breeding patterns of 145 U.S. rice cultivars based on SSR marker analysis. Crop Sci 45(1):66–76 Ma J, Bennetzen JL (2004) Rapid recent growth and divergence of rice nuclear genomes. Proc Natl Acad Sci 101(34):12404–12,410. https://doi.org/10.1073/ pnas.0403715101 Ma J, Bennetzen JL (2004) Rapid recent growth and divergence of rice nuclear genomes. Proc Natl Acad Sci 101(34):12404–12,410. https://doi.org/10.1073/ pnas.0403715101 Mahajan G, Matloob A, Singh R (2018) Basmati Rice in the Indian Subcontinent: Strategies to Boost Production and Quality Traits. Advances in Agronomy. Elsevier Ltd, pp 159–213. https://doi.org/10.1016/bs.agron.2018.04.002. Mahajan G, Matloob A, Singh R (2018) Basmati Rice in the Indian Subcontinent: Strategies to Boost Production and Quality Traits. Advances in Agronomy. Elsevier Ltd, pp 159–213. https://doi.org/10.1016/bs.agron.2018.04.002.
https://openalex.org/W4206833073	https://hal-insu.archives-ouvertes.fr/insu-03582742/file/EPSC2021-534-print.pdf	English	null	Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrometer	null	2,021	cc-by	1,558	To cite this version: Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, et al.. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrom- eter. Europlanet Science Congress 2021, Sep 2021, Virtual Meeting, Unknown Region. pp.EPSC2021- 534, 10.5194/epsc2021-534. insu-03582742 Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, et al.. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrom- eter. Europlanet Science Congress 2021, Sep 2021, Virtual Meeting, Unknown Region. pp.EPSC2021- 534, 10.5194/epsc2021-534. insu-03582742 Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrometer Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, Franck Montmessin, Olivier Forni, Jeffrey R. Johnson, Olivier Gasnault, Erwin Dehouck, et al. To cite this version: Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, et al.. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrom- eter. Europlanet Science Congress 2021, Sep 2021, Virtual Meeting, Unknown Region. pp.EPSC2021- 534, 10.5194/epsc2021-534. insu-03582742 Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, Franck Montmessin, Olivier Forni, Jeffrey R. Johnson, Olivier Gasnault, Erwin Dehouck, et al. To cite this version: Lucia Mandon, Cathy Quantin-Nataf, Pierre Beck, Thierry Fouchet, Clément Royer, et al.. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrom- eter. Europlanet Science Congress 2021, Sep 2021, Virtual Meeting, Unknown Region. pp.EPSC2021- 534, 10.5194/epsc2021-534. insu-03582742 HAL Id: insu-03582742 https://insu.hal.science/insu-03582742v1 Submitted on 21 Feb 2022 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. EPSC Abstracts Vol. 15, EPSC2021-534, 2021 https://doi.org/10.5194/epsc2021-534 Europlanet Science Congress 2021 © Author(s) 2022. This work is distributed under the Creative Commons Attribution 4.0 License. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrometer Lucia Mandon1, Cathy Quantin-Nataf2, Pierre Beck3, Thierry Fouchet1, Clément Royer1, Franck Montmessin4, Olivier Forni5, Jeffrey R. Johnson6, Olivier Gasnault5, Erwin Dehouck2, François Poulet7, Adrian Brown8, Jesse D. Tarnas9, Stéphane Le Mouélic10, Pernelle Bernardi1, Jean-Michel Reess1, Raymond T. Newell11, Sylvestre Maurice5, and Roger C. Wiens11 1LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Université de Paris, Meudon, France (lucia.mandon@obspm.fr) 2 2Université de Lyon, UCBL, ENSL, CNRS, LGL-TPE, Lyon, France 3Université Grenoble-Alpes, CNRS, IPAG, UMR 5274, Grenoble, France 4LATMOS, CNRS, Univ. Saint-Quentin-en-Yvelines, Sorbonne Univ., Guyancourt, France 6Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, USA 7Institut d’Astrophysique Spatiale, CNRS/Paris-Sud University, France 7Institut d’Astrophysique Spatiale, CNRS/Paris-Sud University, France 8Plancius Research, MD, USA 10LPG Nantes, UMR 6112, CNRS, OSUNA, Université de Nantes, Nantes, France 11Los Alamos National Laboratory, Los Alamos, NM, USA 10LPG Nantes, UMR 6112, CNRS, OSUNA, Université de Nantes, Nantes, France 11Los Alamos National Laboratory, Los Alamos, NM, USA The Perseverance rover (Mars 2020 mission, NASA) landed in the ancient lakebed of Jezero crater, Mars on February 18th, 2021. The main science objectives of the mission are the characterization of past habitable environments, the search for preserved biosignatures and the collection of samples to be returned to Earth by the next MSR (Mars Sample Return, NASA/ESA) mission [1]. The payload includes the SuperCam instrument (under the leadership of LANL, USA and IRAP, France), which plays a critical role in the exploration strategy, as it combines various remote-sensing techniques to investigate the elemental and mineralogical composition of rocks and soils: high-resolution color imaging, laser-induced breakdown spectroscopy, Raman spectroscopy, visible and near-infrared (VISIR) reflectance spectroscopy, and acoustic sensing [2, 3, 4, 5]. In particular, the near-infrared spectrometer is a novel instrument on the Martian surface. The spectra, together with data from two additional visible spectrometers, cover 0.39–0.85 µm (VIS) and 1.3–2.6 µm (IR), which allows the identification of a wide variety of minerals. Figure 1. Enhanced-color mosaic of images taken by the Mastcam-Z Left camera of the rover workspace on sol 78 [6]. NASA/Caltech-JPL/MSSS/ASU. Figure 1. Enhanced-color mosaic of images taken by the Mastcam-Z Left camera of the rover workspace on sol 78 [6]. NASA/Caltech-JPL/MSSS/ASU. Orbital data show that the landing site exhibits diverse mineralogical assemblages in various geological contexts, attesting in particular of an ancient aqueous activity during the Noachian and Hesperian eras (> 3 Ga), and including mafic minerals, carbonates, phyllosilicates and opaline silica [7, 8, 9]. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrometer Future in situ measurements on these potentially more hydrated and diverse rocks are expected to bring new insights into the past aqueous environment at Jezero crater. and diverse rocks are expected to bring new insights into the past aqueous environment at Jezero crater. Figure 2. Example of a SuperCam observation on a rock present in the workspace of the rover: 10x1 raster on target Bidziil. (a) RMI mosaic taken at 2.3 m from the rover mast on sol 56 with the IRS 68% field of view annotated in red. (b) Corresponding IRS spectra in color, compared to laboratory spectra of known minerals in black and grey. Figure 2. Example of a SuperCam observation on a rock present in the workspace of the rover: 10x1 raster on target Bidziil. (a) RMI mosaic taken at 2.3 m from the rover mast on sol 56 with the IRS 68% field of view annotated in red. (b) Corresponding IRS spectra in color, compared to laboratory spectra of known minerals in black and grey. Observing rocks in Jezero crater, Mars: results of the first months of operation of the SuperCam VISIR spectrometer We present the initial results of the SuperCam VISIR spectrometers, with calibrated [10] spectral data collected in the first months of operation of the instrument, with a focus on the rocks and soils present in the rover workspace and explored so far. Spectra have been obtained on nearby rocks so far within 50 m of the Octavia E. Butler landing site, at the boundaries of the Crater Floor Fractured Rough (CF-Fr) and Crater Floor Fractured (CF-F) units, which from orbit exhibit the spectral signatures of pyroxene and olivine, respectively [7]. A comparison with orbital data is presented in [11]. Most of the rocks analyzed in the close vicinity of the rover so far consists of light-toned ‘pavers’ and upstanding dark-toned rocks (Fig. 1). Some variability in reflectance levels and absorption bands is observed within and between the rocks (Fig. 2), but overall they share similar spectral characteristics in the VISIR. In particular, they all exhibit an absorption band near 1.9 µm (a spectral signature indicating the presence of the water molecule), a band that is usually not observed in the soils – at least not with the same depth. This absorption near 1.9 µm is also present in spectral observations of outcrops a few kilometers away from the rover [11]. In addition to this absorption at ~1.9 µm, a blue slope in the 2.1-2.5 µm region and subtle absorption bands near 1.4 µm and 2.28 µm show up locally, which suggests the presence of hydrated mineral species, such as an iron-rich phyllosilicate (like nontronite or hisingerite), potentially mixed with an oxy-hydroxide such as ferrihydrite (Fig. 2). The intensity of these absorption bands is consistent with at least a few weight percent of these hydrated species in the observed target, either in the bulk of the rock or in a varnish/coating. Finally, fewer spectra exhibit a weak absorption at 2.2 µm, suggesting the additional presence of an Al-OH or Si-OH-bearing phase. Overall, the detection of these widespread hydration features supports some pervasive water-rock interactions during the past, in the landing site of Perseverance. CRISM data suggest that we have landed in a relatively low hydration region, and that the rover will be driving through progressively increasing hydration throughout the mission, as it moves onto the delta, the olivine-carbonate and the region outside of Jezero crater. [1] Farley, K. A. et al., 2020. Space Sci. Rev.; [2] Wiens, R.C. et al., 2017. Spectroscopy; [3] Maurice, S. et al., 2021 Space Sci. Rev.; [4] Cousin, A. et al., this conference; [5] Chide, B. et al., this conference; [6] Bell, J. F. et al., 2021. Space Sci. Rev. [7] Goudge, T. A. et al., 2015. J. Geophys. Res. Planets; [8] Ehlmann, B. L. et al., 2008. Science; [9] Tarnas, J. D. et al., 2019. Geophys. Res. Lett.; [10] Royer, C. et al., this conference; [11] Quantin-Nataf et al., this conference. References [1] Farley, K. A. et al., 2020. Space Sci. Rev.; [2] Wiens, R.C. et al., 2017. Spectroscopy; [3] Maurice, S. et al., 2021 Space Sci. Rev.; [4] Cousin, A. et al., this conference; [5] Chide, B. et al., this conference; [6] Bell, J. F. et al., 2021. Space Sci. Rev. [7] Goudge, T. A. et al., 2015. J. Geophys. Res. Planets; [8] Ehlmann, B. L. et al., 2008. Science; [9] Tarnas, J. D. et al., 2019. Geophys. Res. Lett.; [10] Royer, C. et al., this conference; [11] Quantin-Nataf et al., this conference.
W2804296800.txt	https://www.biorxiv.org/content/biorxiv/early/2018/05/17/324269.full.pdf	en		Microscopy quantification of microbial birth and death dynamics	bioRxiv (Cold Spring Harbor Laboratory)	2,018	cc-by	10,235	bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Microscopy quantification of microbial birth and death dynamics Samuel F. M. Hart, David Skelding, Adam J. Waite, Justin Burton, Li Xie, and Wenying Shou# Fred Hutchinson Cancer Research Center, Seattle, WA, USA Equal contribution; #: Author of correspondence (wenying.shou@gmail.com) Abstract Microbes live in dynamic environments where nutrient concentrations fluctuate. Quantifying fitness (birth and death) in a wide range of environments is critical for understanding microbial evolution as well as ecological interactions where one species alters the fitness of another. Here, using high-throughput time-lapse microscopy, we have quantified how Saccharomyces cerevisiae mutants incapable of synthesizing an essential metabolite grow or die in various concentrations of the required metabolite. We establish that cells normally expressing fluorescent proteins lose fluorescence upon death and that the total fluorescence in an imaging frame is proportional to the number of live cells even when cells form multiple layers. We validate our microscopy approach of measuring birth and death rates using flow cytometry, cell counting, and chemostat culturing. For lysine-requiring cells, very low concentrations of lysine are not detectably consumed and do not support cell birth, but delay the onset of death phase and reduce the death rate. In contrast, in low hypoxanthine, hypoxanthine-requiring cells can produce new cells, yet also die faster than in the absence of hypoxanthine. For both strains, birth rates under various metabolite concentrations are better described by the sigmoidal-shaped Moser model than the well-known Monod model, while death rates depend on the metabolite concentration and can vary with time. Our work reveals how time-lapse microscopy can be used to discover non-intuitive microbial dynamics and to quantify growth rates in many environments. Introduction Understanding microbial evolution and ecology requires quantifying microbial fitness in diverse environments that the microbes typically encounter. Fitness is often measured as the net growth rate (“growth rate”) – the difference between birth and death rates. An easy and rapid method for measuring growth rate is to track optical density of a culture over time. This method is useful when death rate is low, since optical density cannot differentiate between live and dead cells. In contrast, flow cytometry can yield live and dead cell counts, but requires periodic manual sampling of the culture under observation. As an alternative method, high-throughput microscopy has been developed and applied to, for example, monitoring biofilm susceptibility to antibiotics 1, quantifying growth rate heterogeneity among microcolonies 2, and screening large collections of mutant strains 3. Here, we use microscopy to distinguish cell birth from cell death, especially at low metabolite concentrations where death rate is high. Distinguishing birth from death can be important. For bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. example, metabolite consumption is tied to birth and not to death. As another example, in the extreme case of cells not dividing or dying, then natural selection ceases. Based on the same reasoning, if two populations have the same net growth rate, then a population that divides and dies slowly should evolve slower per unit time compared to a population that divides and dies rapidly. Several mathematical models phenomenologically relate nutrient concentrations to population growth rates. The best known model is the Monod model 4, 𝑔 = 𝑔𝑚𝑎𝑥 𝑠⁄(𝐾𝑀 + 𝑠), where g is the net growth rate, s is the concentration of the limiting metabolite, gmax is the maximal growth rate, and KM is the concentration of s at which half gmax is achieved. Other growth models such as the Teisser and the Contois 5 models have also been proposed. However, like the Monod model, they do not consider cell death since they assume zero (instead of negative) growth rate at zero metabolite concentration. A different growth model by Kovárová-Kovar and Egli 6 incorporates a fixed death rate, although in reality, death rate could vary with metabolite concentrations. Previously, we constructed a two-strain synthetic yeast cooperative community as a model system to explore how cells in a cooperative community might evolve and how cooperation might shape species coexistence and spatial patterning 7–11. In this community, a red-fluorescent strain required lysine and released hypoxanthine (an adenine precursor) (BioRxiv), while a green-fluorescent strain required hypoxanthine and released lysine. A mathematical model for this community had model parameters including each strain’s birth and death rates at various concentrations of the required metabolite, metabolite release rate, and metabolite consumption per birth. Here, we describe a high-throughput microscopy assay that we developed and validated for quantifying a strain’s birth and death rates at various concentrations of the required metabolite. Our approach can be applied to quantifying the birth and death dynamics of other fluorescently labeled microbes. Results Using fluorescence to quantify cell birth and death Our inverted fluorescence microscope is equipped with motorized stage and filter wheel, and is enclosed in a temperature-controlled chamber (Figure 1A) to ensure a nearly constant temperature (Supp Fig 1; Supp Fig 10). To enable automated long-term imaging with minimal photo-damage, we wrote a LabView routine to perform autofocusing using the bright field, and then imaged in the fluorescence channel. However, despite controlling the temperature, condensation developed on the microplate lid over time, which sometimes interfered with autofocusing. To resolve this, we developed a “lid warmer” using transparent, conductive ITO glass (Figure 1B) to warm the microplate lid to ~0.7oC above the stage temperature (Supp Fig 1). This eliminated condensation (Figure 1C) and allowed reliable auto-focusing over tens of hours. For yeast cells expressing a fluorescent protein, total fluorescence intensity (after background subtraction) scaled linearly with live fluorescent cell density up to at least nine cell layers (Supp Fig 2). Occasionally, we also observed that cells lost fluorescence immediately upon losing cell bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. integrity (compare Supp Movie 3 vs Supp Movie 5). Thus, increases or decreases in fluorescence were proportional to cell division or cell death, respectively. To measure birth and death rates, we performed time-lapse imaging of fluorescent yeast cells at various concentrations of the required metabolite (e.g. Figure 1D). We started with a small number of cells to minimize metabolite depletion during growth rate measurements. Even in the absence of the required metabolite, total fluorescence intensity initially increased due to residual growth fueled by cellular storage of metabolites 12 (e.g. 0~3 hrs in Figure 1D). Thus, we only used images after the residual growth in our data analysis. Death rate is time-dependent Death rate is not a constant. For example, we measured the death rate of lys2- cells in zero lysine. Since new birth was negligible (no birth out of 603 cells over 30 hrs; Supp Movie 3), death rate could be estimated by quantifying the negative slope of ln(live population size) against time. We observed multi-phasic death kinetics, with a slow death rate followed by a faster death rate (Figure 2A, lightest grey; Figure 2B, [Lys]=0; Supp Fig 3). ade8- cells in the absence of hypoxanthine also displayed a time-dependent death rate (Supp Fig 4). For both strains, the death rate would eventually slow down, as shown in our previous work 7. As we will demonstrate below, death rate also depends on the metabolite concentration. Death rate based on total fluorescence is comparable to two other approaches where fluorescent cells were scored “live” and non-fluorescent cells were scored “dead”: flow cytometry and direct cell counting in microscopy images. As expected, subtle differences existed among the three methods (Supp Fig 3). For example, total fluorescence intensity (but not flow cytometry or cell counting) would be increased by cell swelling (Figure 3C). The microscopy cell counting method could detect the death of a bud, but in flow cytometry, since the attached mother cell was still fluorescent, the death event would not be recorded (unless a death dye was used). Overall, results from microscopy fluorescence and flow cytometry overlapped, while that from cell counting yielded a slightly lower death rate (Figure 2B, compare different symbols of the same color), possibly because sample handling in flow cytometry reduced the viability of starving cells. Diverse birth and death dynamics at low metabolite concentrations Low concentrations of metabolites may not be consumed. For example, low concentrations of lysine (e.g. 0.33~1 µM; Supp Movie 2) was barely consumed by lys2- cells, while concentrations >1.75 µM were depleted down to <1 µM (Fig 3B). Consistently, maximal fluorescence intensity only scaled linearly with lysine concentrations > 1.75 µM (Figure 3A), the level around which half maximal growth rate was achieved (Figure 4). For ade8- cells at low input concentrations (e.g. 0.1~0.2 µM) of hypoxanthine, supernatant hypoxanthine concentration was too low to be directly measured. However, we inferred that low concentrations of hypoxanthine were largely un-consumed based on the following inference. Since 1~3 fmole of hypoxanthine is consumed per ade8- cell (BioRxiv), the input medium (300 µl of 0.1 µM or 3x104 fmole hypoxanthine) bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. should support the birth of a total of 1x104~3x104 cells. Instead, starting with 3000 cells, we observed <50% increase in cell number (Figure 3D). Low concentrations of metabolites lead to diverse birth and death dynamics depending on the strain genotype. For lys2- cells in low lysine, although total fluorescence intensity increased for longer compared to zero lysine (Figure 1D), this increase generally corresponded to cell swelling rather than cell birth (Figure 3C). Birth events, if any, were restricted to the initial few hours and not sustained at later time points despite nearly-constant metabolite concentration (Figure 2A; Figure 3B; Supp Fig 5A). Thus, the initial birth events could be interpreted as low lysine prolonging the residual growth phase and delaying the onset of death phase. Low lysine also reduced early and late death rates (Figure 2B). In contrast, for ade8- cells, low input concentrations (e.g. 0.1~0.2 µM) of hypoxanthine led to increased birth and death rates compared to zero hypoxanthine (Figure 3D). ade8- cells were occasionally born in low hypoxanthine, even after the onset of death phase (30-66 hr, Supp Fig 5B). Moreover, a small number of ade8- cells transiently lost fluorescence, but then regained fluorescence and continued to divide (Supp Movie 4). In summary, for lys2- cells, low concentrations of lysine were not consumed, did not sustain birth beyond the initial stage, but delayed the onset of death phase and slowed the death rate once cells began to die (Figure 2). For ade8- cells, low concentrations of hypoxanthine were likely also largely unconsumed, and increased both birth and death rates. Overall, death rate was not only time-dependent, but also nutrient concentration-dependent. The Moser model is superior to the Monod model in describing lys2- and ade8- cell birth At relatively high lysine concentrations (≥1.75 µM, Supp Movie 1), we could not directly count birth and death events due to multiple cell layers. Since the death rate was small (<0.002/hr) compared to the net growth rate (  0.1/hr) in this range of lysine concentrations (BioRxiv), we approximated the net growth rate as the birth rate (Figure 4). That is, we measured the rate of ln(total fluorescence) increase across sliding time windows (generally four time points over a total of 1.7 hrs; Figure 1D), and used the steepest slope as birth rate. When lysine was not the limiting nutrient in the media (  5 µM), lys2- cells achieved a maximal net growth rate of 0.50 +/- 0.02/hr in microscopy assay (Figure 1D, magenta), consistent with the value measured from culture optical density over time (e.g. 0.49 +/- 0.03/hr). For lysine concentrations ≥ 2.5 µM, the maximal growth rate was maintained across at least two contiguous sliding windows (Supp Fig 6). However, at lower lysine concentrations (e.g. 1.75 µM), the growth rate continuously declined throughout the experiment, and thus it was unclear whether our measured maximal birth rate was truly maximal. To verify our observations, we employed an independent measurement method where we grew lys2- cells in lysine-limited chemostats 13 (Methods). In a steady-state chemostat, the population net growth rate is equal to the dilution rate. We set the dilution rate to various values, and measured the corresponding lysine concentrations (Supp Fig 7). Chemostat measurements were consistent with microscopy measurements (Figure 4, blue crosses). The birth rate of lys2- cell increased with lysine concentration in a sigmoidal fashion. The data were better characterized by the Moser model b(s)  bmax sn / (Kmn  sn ) (black line; Figure 4), bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. than the Monod model b(s)  bmax s / (Km  s) (grey dotted line; Figure 4). In both models, s is lysine concentration, bmax is the maximal birth rate, Km is the Monod constant (s required to achieve half maximal birth rate). The Moser model has an additional parameter, n, which is analogous to the cooperativity coefficient in the Hill equation. For ade8-, the Monod model worked relatively well (Supp Fig 8, grey), but the Moser model provided a more accurate estimate of the maximal birth rate (Supp Fig 8, black). Experimentally measured maximal birth rate was 0.44+/- 0.03/hr, comparable to the fit by the Moser model (0.44/hr) and significantly lower than the fit by the Monod model (0.49/hr). Increased accuracy in the Moser model is not surprising given the additional parameter. Nevertheless, an accurate fit to experimental data is useful when modeling population dynamics. Discussion We demonstrated that our microscopy setup allows quantitative measurements of cell growth in a wide range of nutrient environments. Using this approach, we tracked individual birth and death events in a monolayer at low metabolite concentrations. At higher metabolite concentrations, we quantified the net growth rate which was approximately the birth rate since death rate was relatively small. Microscopy quantification of growth rates requires careful cell preparation and image analysis. Growth rates can differ depending on the growth conditions before starvation (Supp Fig 9), how we define residual growth (Supp Fig 6), and whether birth and death rates are calculated from fluorescence or cell counts (Figure 3C). Since there is no single “correct” procedure, it is important to be aware of the limitations of each method. We recommend cross-checking one method against another, independent, method. For example, we cross-checked microscopy measurements against chemostat measurements when estimating growth-nutrient relationship (Figure 4), and cross-checked microscopy measurements against flow cytometry when estimating the death rate (Supp Fig 3). Population dynamics of the two strains shared certain features under nutrient limitation. For example, both lys2- and ade8- cells displayed time- and nutrient-dependent death rates (Figure 2; Supp Fig 3; Supp Fig 4). The two strains differed in other aspects. For lys2- cells, low concentrations (0.33-1 µM) of lysine were largely un-consumed (Figure 3A and B), and did not support birth beyond residual growth (Supp Fig 5A), but did delay and slow death when compared to full starvation (Figure 2). In contrast, for ade8- cells, low concentrations (0.1-0.2 µM) of hypoxanthine increased death rates (Figure 3D) compared to full starvation, and, at the same time, supported some birth (Supp Fig 5B). These seemingly counter-intuitive behaviors probably resulted from the fact that these mutations disrupted biosynthetic pathways that would normally produce the necessary metabolite. Since the cell does not “know” that it cannot make the metabolite, it continues to try to grow and divide in its absence, which results in an abnormally high death rate 14. The Monod model has been observed to fit, for example, the growth rates of yeast strains at various glucose concentrations 2. For lys2- cells, the Moser model was much better than the Monod model in describing birth rates as a function of lysine concentrations (Figure 4). The bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Monod model (Figure 4, grey) overestimated the birth rate at low lysine (<2 µM), underestimated the growth rate at medium lysine (2~6 µM), and overestimated the growth rate at high lysine (>6 µM). Experimentally, the birth rate of lys2- increased with lysine concentration in a sigmoidal fashion, corresponding to a lack of birth at low lysine (≤1 µM), a sharp increase in birth rate at medium lysine (1.75~2.5 µM), and leveling off at max birth rate at higher lysine (≥3.75 µM). The sigmoidal growth-nutrient relationship could be explained by, for example, cooperative binding by nutrient transporters, as seen in a variety of cells types, including yeast 15–17 . For ade8- cells in hypoxanthine, the Monod model closely fit experimental data. The Moser model was still more accurate than Monod model (Supp Fig 8), although a two-parameter model (Moser) is expected to improve the fit over a one-parameter model (Monod). Regardless, a growth model that faithfully captures experimental observations is useful. For example, when modeling a community of two cross-feeding strains, a Moser model of how fast each strain grows in various concentrations of partner-supplied metabolite can be incorporated into the community dynamics model. In summary, our work demonstrates the potential of microscopy assays in quantifying microbial birth and death dynamics. Methods Strains and growth medium We used strains from the RM11 background with the following genetic modifications introduced via transformation. Strain WY1335 (“lys2-”) has the genotype of ho::loxP AMN1-BY ste3::Hph fba1::FBA1-mCherry-loxP ade4::ADE4-PUR6 (o/e) lys2::loxP. Strain WY1340 (“ade8-)” has the genotype of ho::loxP AMN1-BY ste3::NAT fba1::FBA1-EGFP-loxP lys21::LYS21(o/e) ade8::loxP. For our bioassay of low metabolite concentrations, we used an evolved clone (WY2270) isolated after lys2- had grown for tens of generations under lysine limitation. This clone displayed an increased affinity for lysine. We stored these strains at -80oC in YPD+15% glycerol. We used rich medium YPD (10 g/L yeast extract, 20 g/L peptone, 20 g/L glucose) for streaking out single colonies and for growing saturated YPD overnight cultures which were then used as inoculum to grow exponential cultures. We found ade8- cells could grow to a higher density in YPD if supplemented with 100 µM hypoxanthine. We sterilized YPD media by autoclaving. YPD overnight cultures were stored at room temperature for no more than 4~5 days prior to experiments. We used defined minimal medium SD (6.7 g/L DifcoTM yeast nitrogen base w/o amino acids, 20 g/L glucose) for all experiments 18, with supplemental metabolites added as noted 19. To achieve higher reproducibility than autoclaving, we sterilized SD media by filtering through 0.22 µm filters. We performed all culturing at 29.5±1oC. lys2- cells were pre-grown to exponential phase in SD supplemented with excess (164 µM) lysine and washed 3~5 times with SD. Where noted, we starved lys2- cells for 3~6 hours to deplete intracellular lysine storage. Otherwise, we did not starve lys2- prior to starting an experiment. ade8- cells were pre-grown to exponential phase in SD supplemented with excess hypoxanthine (100 µM), washed 3~5 times with SD, and prestarved in SD for 24 hours to deplete cellular storage. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Microscope setup Imaging was performed using a Nikon Eclipse TE-2000U inverted fluorescence microscope. A temperature-controlled enclosure (In Vivo Scientific controller, model 300353) maintained the microscope at a temperature close to a set temperature (29.5˚C). We noticed that the air conditioning system could cause fluctuations in temperature, so we turned it off. The microscope had a motorized stage to allow z-autofocusing (Methods, “Autofocusing") and systematic xy-scanning of locations in microplate wells. Our LabVIEW (National Instruments) program controlled the microscope body, illumination, and stage position through serial port communication. The program moved the stage gently, so that cells were not disturbed and individual cells could be tracked from one time point to the next. The microscope was also equipped with motorized switchable filter cubes capable of detecting a variety of fluorophores. We used an ET DsRed filter cube (Exciter: ET545/30x, Emitter: ET620/60m, Dichroic: T570LP) for mCherry-expressing strains, and an ET GFP filter cube (Exciter: ET470/40x, Emitter: ET525/50m, Dichroic: T495LP) for GFP-expressing strains. Fluorescence and transmitted light images were taken with a Photometrics CoolSNAP HQ2 cooled CCD camera, interfaced with LabView through the Bruxton Inc. SIDX API. We used a 10x objective, with a numerical aperture of 0.30, because it provided a wide field of view while allowing easy observation of individual cells. Image acquisition was done with an in-house LabVIEW program, incorporating autofocusing in bright field followed by fluorescence imaging with automatically adjusted exposure time to avoid camera saturation. Optimal exposure times for fluorescence imaging may vary (~0.1-1 second). When we imaged ade8- cells using a particular light configuration, the very short exposure time (0.05 sec for initial images) created problems for image analysis. This could be due to the latency in shutter opening/closing becoming important in short exposure times. Alternatively, since high-intensity light was used (and hence the short exposure time), exposure time had to be reduced as cells grew to avoid camera saturation, and adjusting for variable exposure time during data analysis could introduce errors. When we added a neutral density filter and/or adjusted the size of light aperture so that the exposure time was ~0.3 sec, data analysis became normal. Four locations per well were imaged, with ~20-200 initial cells per image. Lid warmer During extended imaging, condensation could accumulate on the underside of microplate lid, even in the temperature-controlled chamber. We encountered failures in autofocusing due to heavy condensation, and condensation can degrade proper Kӧhlier illumination. In order to prevent condensation, we used an optically transparent heating plate to warm the lid (Figure 1B), which kept the lid temperature an average of 0.68°C with 2σ of 0.22°C higher than the stage temperature (Supp Fig 1). This eliminated condensation (Figure 1C). Our lid warmer used an ITO glass heating plate, with an integrated thermistor temperature sensor (Oven Industries TR91-170). The 1 mm-thick ITO glass had a 140nm ITO coating (from SPI Supplies), transmitted 88% of visible light, and had a sheet resistance between 30 – 60 Ohms/sq. We cut a 70mm x 120mm sheet with beveled corners to match the microplate lids. To apply bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. power for heating, we affixed strips of 5 mm-wide conductive copper foil tape to the ITO coating along two opposite ends of the heating plate, and applied silver conductive epoxy (MG Chemicals, 8332-13G) to the tape, extending 1-2 mm onto the ITO coating to ensure a reliable connection. Wires were connected through brass conductors, which were insulated with heat shrink tubing and epoxied to the edge of the heating plate. We measured the resistance of the plate to be 20 Ohms, so the application of 5V DC generated (5 V)2/26 Ohms = 1.0 W. This provided the needed heat when applied. The sensor readings were tested, and adjusted if necessary, using a Barnant 115 thermocouple thermometer with a T type probe. The temperature of the heating plate was controlled by a LabVIEW program, with the aid of a DAQ (National Instruments USB-6008) for reading temperature sensors. Based on the measured temperatures of the heating plate and the microscope stage, the LabVIEW program activated or deactivated the heating plate when the temperature difference was ≤0.5oC or ≥0.8oC, respectively. As discussed above, the heating plate was run at a fixed power of ~ 1W when active. Note: The 5V supply from the DAQ was used for the sensors, and a separate supply was used for heating the plate. Setting up samples in microtiter plates We used flat-bottom transparent microplates with wells joined together at the bottom by a continuous sheet of plastic, such as Costar 3370 96-well plates (rather than 96-well plates where the spaces between the wells was open to the surrounding environment and thus more susceptible to temperature fluctuations). In these plates, air warmed by the lid warmer thermally insulates the sides of individual wells, which improves temperature uniformity. We filled the outermost wells of a 96-well microtiter plate with water to reduce evaporation, leaving up to 60 wells for imaging. We diluted cells to low densities (1000~5000 cells inoculated in 300 µl medium) to maximize our growth window and to minimize metabolite depletion during measurements. When assaying death in the absence of supplements, we added 2~10 fold more cells since there was little cell birth. We spun the plates at 2000 rpm for 2 minutes to settle all cells to the bottom of the wells, set up the microscope as described above, and imaged the same four positions in each selected well periodically (every 0.5~2 hrs). For each position, a bright field and a fluorescent image was saved. For growth assays, we ran experiments until fluorescence leveled off (16+ hrs for lys2-, and 30+ hrs for ade8- cells). For death assays, we ran experiments for up to 66 hr. We found that growth rates for the same samples did not vary significantly across different well positions in a plate (Supp Fig 10). Autofocusing Automated imaging of the cells was performed using custom software written in LabVIEW. At each time point, it was necessary to auto-focus on the cells due to the small vertical drift caused by small changes in temperature or mechanical stress. At the beginning of an experiment, manual focusing was performed on four corner sample wells to ensure that the plate was level (otherwise, we need to adjust the screw positions of the plate holder). Then, coarse auto-focusing was performed for one position in each sample well at +/- 100 z-positions spaced at 3 µm apart, and the best focal plane was chosen to initiate an experiment. To identify the best focal plane for bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. imaging and to prevent a loss of focus during the experiment, fine auto-focusing (+/- 30 zpositions spaced at 2 µm apart) was performed for each time point. Each 16-bit image was imported directly into Labview, and converted into a twodimensional array of real numbers. The optimal z-position for focusing was chosen using a variant of the Brenner auto-focus algorithm 20,21. The quality of focus, 𝐴(𝑧), was measured by computing the averaged horizontal and vertical gradient in each image: 2 2 𝑀−2 𝐴(𝑧) = ∑𝑁−2 𝑖=1 (𝑝(𝑖 + 2, 𝑗) − 𝑝(𝑖, 𝑗)) + ∑𝑗=1 (𝑝(𝑖, 𝑗 + 2) − 𝑝(𝑖, 𝑗)) , where 𝑁 × 𝑀 were the dimensions of each image, and 𝑝(𝑖, 𝑗) was the pixel intensity at row 𝑖 and column 𝑗. Local maxima in 𝐴(𝑧) corresponded to sharply-focused images. Comparison between data two pixels apart rather than adjacent pixels reduced the effects of correlated noise and the natural illumination of nearby pixels. When observing yeast we often observed multiple local maxima in the range of z positions, which was due to the focusing of light within the cells. Assuming yeast cells behave like small spherical lenses of diameter 𝐷 ≈ 5 𝜇𝑚, then this phenomenon is likely a complex function of reflection, refraction, and diffraction 23. However, some qualitative features may be illustrated by simply considering refraction of a ball lens, where the effective focal length (𝐸𝐹𝐿) is 23: 𝑛𝑐 𝐷 𝐸𝐹𝐿 = , 4(𝑛𝑐 − 𝑛𝑤 ) where 𝑛𝑤 = 1.33 is the index of refraction of the surrounding water, and nc is the index of refraction of the cell. Supp Fig 12 shows that the first maxima is due to light that is focused approximately 14 µm from the yeast cells, which would correspond to 𝑛𝑐 = 1.46. This is consistent with previous measurements showing that 𝑛𝑐 ≈ 1.53 23, although we note that this value depends on the size and density of the cells. The second maxima is located at an image plane below the cells that contains light halos from the apparent source of the focused light. We found that the minimum in 𝐴(𝑧) between these two maxima conveniently corresponded to a focal plane adequate for identifying and imaging the cells (Supp Fig 12). Thus, we choose this local minimum for imaging in all of our experiments. The three points closest to the minimum are fit to a parabola, and the minimum of the parabola is chosen as the optimal focus position. The microscope stage is then moved to the optimal focal plane for imaging. A similar method for autofocusing, i.e. using the local minimum in the autofocus score, has been used before 24 . Image analysis We analyzed time-lapse images using Bioact2, an ImageJ plugin written by Adam Waite (available at https://github.com/nodice73/Java/tree/master/imagej_plugins/bioact). Bioact2 measured the background-subtracted total fluorescence intensity of all cells in an image. To distinguish fluorescent cells from background, each fluorescent image was blurred with a Gaussian filter using a standard deviation (σ) of 1 pixel. Low frequency noise was removed using the “rolling ball” background subtraction algorithm 25. The dynamic range of the image bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. was reduced from 16-bit to 8-bit, and each pixel was replaced by the maximum value in a 3 pixel radius. The resulting image was thresholded using the “maximum entropy” method for low density images or the “iso data” method for higher-density images and converted to a binary mask 26. After filling holes using binary closing, the mask was dilated by 2 pixels. From this mask, the percent pixels considered foreground was calculated. Each mask was then applied to the un-manipulated original image, and the foreground and background intensities were measured. At low cell densities, the background was calculated for each image and subtracted. We found that when the foreground made up more than a specified fraction of the total image area, the background estimate was no longer accurate, and the running average background value calculated before this threshold was met was used as the background value for subsequent images. This method occasionally failed if cells were nearly confluent during late stage growth. However, we were only interested in maximum growth rate, which occurred before confluency. We plotted background-subtracted fluorescence intensity over time for all four positions in each well to allow visual inspection. In rare occasions, all four positions were out-of-focus and none were used. In a small subset of experiments, a discontinuous jump in data appeared in all four positions for reasons we do not understand. We did not calculate rates across the jump. Occasionally, data from one or two positions deviated from the rest. This could have been due to a number of reasons, including spurious shifts in stage position, or black or bright dust particles in the field of view. In these cases, we inspected the images and outliers with obvious causes were excluded. If the fluorescence dynamics of four positions differed due to cell heterogeneity at low concentrations of metabolites, all positions were retained. We normalized intensity against that of time zero, and averaged across positions. We calculated growth rate over three to four consecutive time points, and plotted the maximal net growth rate against metabolite concentration. If maximal growth rate occurred at the end of an experiment, then the experimental duration was too short and data were not used. Individual cell tracking at low metabolite concentrations At very low concentrations of supplements, we manually counted birth and death events by scanning through fluorescent images in ImageJ (e.g. Supp Movie 2-3). After counting the initial number of cells in an image, we proceeded through each image and noted the appearance or disappearance of cells. We counted a birth event as the appearance of a new cell adjacent to a cell present in the previous image. We differentiated this from two vertically-stacked cells shifting orientation to create the appearance of a new cell birth, as the fluorescence intensity of stacked cells dropped noticeably during orientation shift. We counted a death event when a cell present in the previous image suddenly lost fluorescence. Occasionally cellular fluorescence slowly faded over time rather than disappearing suddenly, and in this case we counted the initial drop in intensity as the death of the cell. Bioassay quantification of metabolite concentrations We used a bioassay to quantify lysine concentrations from supernatants. In order to obtain supernatant, we filtered cell cultures through a 0.45 µm nitrocellulose filter and stored at -80oC until quantification. We mixed 150 µl sample with an equal volume of master mix containing 2x bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. SD and WY2270 lysine-requiring tester cells (~1x104 cells/ml) in a flat-bottom 96-well plate. We measured growth rates of WY2270 in unknown samples and compared them to growth rates of WY2270 run concurrently with known concentrations of lysine. The growth rate scaled linearly with lysine concentration up to 1 µM (i.e. 2 µM in undiluted sample, Supp Fig 11). From this standard curve, we inferred lysine concentrations of samples. Assay sensitivity was 0.1 µM. Quantifying population dynamics using flow cytometry We first prepared bead standards for quantifying cell density. Fluorescent beads (ThermoFisher Cat R0300, 3 μm red fluorescent beads) were autoclaved in a factory-clean glass tube, diluted into sterile 0.9% NaCl, and supplemented with sterile-filtered Triton X-100 to a final 0.05% (to prevent beads from clumping). We sonicated beads and kept them in constant rotation to prevent settling. We quantified bead concentrations by counting beads using a hemacytometer and a BD Coulter counter. The final bead stock was generally 4~8x106/ml. Culture samples were diluted to OD 0.01~0.1 (7x105~7x106/ml) in MilliQ H2O in unautoclaved 1.6ml Eppendorf tubes. In a 96-well plate, 90 µl sample were supplemented with 10 µl bead stock to calculate cell density from cell:bead ratio. We also added 2 µl of 1 µM nucleic acid dye ToPro3 (Molecular Probes T-3605) which stains dying/dead cells with compromised membrane. Flow cytometry was performed on Cytek DxP Cytometer equipped with four lasers, ten detectors, and an autosampler. Fluorescent tags GFP, mCherry, and ToPro were respectively detected by 50 mW 488 nm laser with 505/10 (i.e. 500~515nm) detector, 75 mW 561nm Laser with 615/25 detector, and 25mW 637nm laser with 660/20 detector. Each sample was run in triplicate and individually analyzed using FlowJo® software to identify numbers of events of beads, dead cells, and various live fluorescent cells. We calculated the mean cell density from triplicate measurements, with the coefficient of variation generally within 5%~10%. Chemostat culturing We constructed an eight-vessel chemostat with a design 13 modified from an existing multiplexed culturing device 27. A chemostat creates a nutrient-limited environment where the population is forced to grow at a constant, pre-determined rate slower than the maximal growth rate 28. Specifically, a medium containing a limiting metabolite is added to the culturing chamber at a constant flow rate f (ml/h). The culture effluent is removed from the chamber at the same rate f, thereby maintaining a constant culture volume V. Mathematically 28, live population density N satisfies dN dt   r  f V  N (1) where r (h-1) is the net growth rate (birth rate minus death rate), f is the flow rate, and f/V is the dilution rate (h-1). At steady state, the net growth rate and the dilution rate are equal: r = f/V (2) The limiting metabolite also reaches a steady state in the culturing vessel. Thus, the steady state concentration of the limiting metabolite supports a net growth rate equaling the dilution rate. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Due to rapid evolution, we designed experiments so that live and dead populations quickly reached steady state and the experiment lasted ≤26 hours (bioRxiv). We washed exponentiallygrowing cells to remove extracellular lysine and inoculated 1/4~1/2 of the volume at 1/3 of the expected steady-state density. We filled the rest of the 19ml chamber with reservoir media (resulting in less than the full 20 µM of reservoir lysine, but more than enough for maximal initial growth rate, ~10-15 µM). We sampled cultures periodically to track population dynamics using flow cytometry (Supp Fig 7A) and filtered supernatant through a 0.45 µm nitrocellulose filter. We froze the supernatants for metabolite quantification at the conclusion of an experiment (Supp Fig 7B). Acknowledgement We thank Jose Pineda for performing the experiment in Supp Fig 2 bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Figures Figure 1 Automated high-throughput microscopy (A) Microscope setup. An enclosure around the microscope provides a temperature-controlled environment. A motorized stage allows repeated bright field and fluorescent imaging of the same positions in specified wells of a microtiter plate (Supp Movies 1-3). Motorized filter cubes allows automated filter switching. (B) An ITO glass lid warmer prevents condensation. Sensors on the plate lid (“thermistor”) and the microscope stage provide temperature measurements to a LabVIEW program which turns the ITO lid warmer on or off to maintain the plate at ~ 0.7˚C warmer than the stage (Supp Fig 1). (C) The lid warmer eliminates condensation. The images were taken after a 24-hour imaging experiment at 30˚C. (D) Growth of lys2- (WY1335) cells at various lysine concentrations. Background-subtracted total fluorescence intensities from four picture frames were normalized against their respective initial values, averaged, and plotted. Dashed line marks the end of residual growth. Residual growth occurs even at zero lysine, and is presumably fueled by vacuolar lysine storage. When calculating growth rates, we only considered post residual growth data. Circles and triangles mark two independent experiments. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Figure 2 Death rate is nutrient- and time-dependent. Exponentially-growing lys2- (WY1335) cells were washed free of lysine at 0 hr, and imaged in minimal medium supplemented with 0 to 1 µM lysine. (A) Birth events were restricted to the first few hours, and may be regarded as part of residual growth. Early and late death rates were calculated using data before and after asterix (), respectively. For each experiment, we tracked birth and death events from an initial 200~300 cells. (B) Death rate started slow (cyan), and then increased (blue). For experiments in 0 µM lysine, data points are jittered along the x-axis to aid visualization. Death rates of lys2- cells at 0 µM lysine were tracked using microscopy cell counting (circles), microscopy fluorescence intensity (triangles), and flow cytometry (diamonds). The three methods resulted in compatible results, although death rate from cell counting seemed to be lower than that from microscopy fluorescence intensity or flow cytometry. Death rates at 0.33~1 µM lysine were tracked using microscopy cell counting only (circles). bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Figure 3 Very low concentrations of metabolites may not be consumed and can lead to diverse birth and death dynamics (A) Maximal fluorescence achieved by a lys2- (WY1335) culture scales with initial lysine concentration only for lysine concentrations above a threshold (1.75 µM). Maximal fluorescence was normalized against that at zero lysine. (B) Low concentrations of lysine (≤1 µM) remain largely unconsumed by lys2- over 40 hours. We measured lysine concentrations in supernatants using the rate bioassay (Methods). (C) An increase in fluorescence intensity may not correspond to cell birth. We imaged lys2- cells (161 total) in 0.33 µM lysine. Total fluorescence intensity (brown) and counts of fluorescent cells (blue) over time are plotted. We observed no birth events in this experiment, while fluorescence increased for over 15 hours after the start of the experiment. (D) A low level of hypoxanthine increases cell birth and death in ade8- cells. Exponential ade8- (WY1340) cells were washed and starved for 24 hours, and then imaged every 2 hours in 0 (light grey), 0.1 µM (dark grey) and 0.2 µM (black) hypoxanthine. A small amount of hypoxanthine increases both cell birth (0~20 hours; Supp Fig 5B) and cell death (2066 hours). bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Figure 4 The birth rate of lys2- as a function of lysine concentration is better described by the Moser model than the Monod model. For lysine concentrations ≥1.75 µM, we calculated net growth rates within 3~4 hr sliding windows. We used the maximal growth rates (before lysine depletion) to approximate birth rates, since death rates are at most 2% of growth rates and can thus be neglected (see Figure 2). At lower lysine concentrations (≤ 1µM), cells remain monolayer, which has enabled us to count birth (and death) events. Four microscopy assays are shown (green) with each assay designated by a different symbol. Despite increasing fluorescence at low lysine concentrations, we observed no births (e.g. Figure 3C). Growth rates and their approximate birth rates were also measured in chemostats (blue) where growth rates are controlled via dilution rates. The corresponding lysine concentrations in culturing vessels were measured using a bioassay (Methods). Results from different measurements are consistent. . The birth rate of lys2- as a function of lysine concentration can be described by the Moser equation (black) where b( L)  bmax L LnL / (K LnL  LnL ) with bmax L  0.50 (95% CI: 0.48 ~ 0.53), K L  1.9 (95% CI: 1.8 ~ 2.0), and n = 4.47 (95% CI: 3.2~ 5.7). In comparison, the best-fit of the Monod equation g (L)  bmax L L / (KL  L) (grey dotted, bmax L  0.85 and KL  4.6 ) fits the data poorly. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 1 Temperature-controlled stage and lid warmer. (A) Stage temperature was set to 29.5 oC and was maintained at within 0.5oC of the set temperature except for the initial period of time. (B) The lid warmer maintained the microplate lid 0.5~ 0.8oC above the stage temperature. The average temperature difference was 0.68oC. A LabVIEW program compared temperature readings from the stage and from the lid warmer, and turned on and off the heater to maintain the set range of temperature difference. The initial temperature overshoot may be due to non-equilibrium distribution of heat in the system. Three colors indicate three independent experiments. Growth rates at 5 µM (as seen in Figure 4) for all three experiments were similar (0.50 +/- 0.02/hr) despite the observed ~0.5oC differences in temperature. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 2 A linear relationship between total live cell number and total fluorescence intensity in a field of view Various numbers of live cells (counted by flow cytometry) were placed in minimal medium lacking glucose (to arrest cell growth) in microtiter wells and allowed to settle to the bottom of wells. The plate was then imaged using our microscope setup to measure the total fluorescence intensity. For both red (mCherry-expressing lys2-) and green (GFP-expressing ade8-) fluorescent strains, total cell numbers and total background-subtracted fluorescence intensities in a field of view display linear relationships. At the highest density, there were at least nine cell layers. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 3 Microscopy and flow cytometry measurements reveal multi-phasic death dynamics of lys2- cells. We grew lys2- (WY1335) in excess lysine. At time zero, we washed cells free of lysine, and prestarved cells for 4~5 hrs to deplete vacuolar lysine storage. We then diluted cells in SD (<20,000 cell/ml) and imaged them. (A) Microscopy total fluorescence intensity (circles) and flow cytometry (diamonds) revealed multi-phasic death kinetics (slow death rate followed by faster death rate). For a fraction of wells, we also periodically harvested and concentrated cells for flow cytometry measurements. Error bars are 2x standard deviations from multiple independent measurements (three technical replicates for FACS, and tens of images for fluorescence imaging). (B) Cell counts (diamonds) and total fluorescence intensity (circles) reveal multiphasic death kinetics. In cell counting, ~200 cells were followed. In each panel, different colors represent experiments done on different days. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 4 Time-dependent death rate of ade8- during starvation (A) Death dynamics of ade8- cells. Exponential ade8- (WY1340) cells were washed and starved. We tracked dead and live cell densities over time using flow cytometry. (B) Death rate is initially slow and then speeds up. Early death rates were generally calculated within 24-96 hr, while late death rates were generally calculated within 70-230 hr. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 5 Low concentrations of the required metabolite supported birth for ade8- but not lys2- cells. (A) lys2- cells grown in excess lysine were washed free of lysine at time zero, and the number of birth over a 1-hr interval was counted. Data source was identical to that in Figure 2A. (B) We grew ade8- in excess hypoxanthine to exponential phase, washed away hypoxanthine, and prestarved cells for 24 hrs. At time zero, we started imaging. The number of birth over a 2-hr interval was counted. Data source was identical to that in Figure 3D. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 6 Steady growth rate is maintained only for ≥2.5 µM lysine We calculated growth rates from fluorescence intensity over time. Each data point corresponds to the growth rate calculated over the four previous time points, a time window of approximately 1.7 hours. For ≥2.5 µM lysine, cells reached a maximal growth rate and maintained it across at least two sliding windows (e.g. hours ~5-8 for 3.75 and 5 µM). For 1.75 µM lysine, the growth rate declined throughout the duration of measurements. For ≤1 µM lysine, any positive net growth rate after residual growth was due to cell swelling not cell birth (Figure 3C). bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 7 Chemostat dynamics rapidly reach a steady state. lys2- cells growing exponentially in excess lysine were washed free of lysine, and inoculated in a chemostat vessel at 1/3 steady-state density in the presence of ~10-15 µM lysine. Minimal medium containing 20 µM lysine was dripped into the culturing vessel (19 ml) at a set rate to achieve desired growth rate (8-hr doubling time corresponding to a flow rate=19 mlln(2)/8 hr = 1.646 ml/hr; “Chemostat culturing” in Methods). We tracked live and dead cell densities (A) using flow cytometry (Methods, “Quantifying population dynamics using flow cytometry”) and tracked lysine concentrations (B) using a bioassay (Methods, “Bioassay quantification of metabolite concentrations”). bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 8 Birth rates of ade8- as a function of hypoxanthine concentrations ade8- (WY1340) cells were grown to exponential phase in minimal medium supplemented with excess hypoxanthine, washed free of hypoxanthine, and pre-starved for 24 hrs to deplete cellular storage. Cells were then incubated at various concentrations of hypoxanthine and imaged. The Monod model (grey dotted line) predicted a maximal birth rate of 0.49/hr (95% CI: 0.46~ 0.52/hr). In the Moser model (black), maximal birth rate bmax  0.44 /hr (95% CI: 0.43 ~ 0.45), hypoxanthine concentration for half maximal birth rate Km  1.3 µM (95% CI: 1.2 ~ 1.4), and n  1.5 (95% CI: 1.4~ 1.7). In comparison, experimentally-measured bmax is 0.44±0.03/hr. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 9 The duration of pre-starvation can affect growth rates Pre-starvation at zero lysine, which could be used to deplete cellular lysine storage, reduces net growth rate compared to no pre-starvation. lys2- cells were grown to exponential phase in excess lysine. At time zero, we washed cells free of lysine, and pre-starved them for 0, 3, or 6 hrs before imaging them in various concentrations of lysine. We analyzed data from hour 3 and on (postresidual growth). Using 3-hr sliding time windows, we calculated the net growth rate over time. We plotted maximal growth rate with error bars indicating 2x error of slope estimation. Note that we used fluorescence to quantify growth rate and did not correct for the absence of birth in 0~1µM lysine. This resulted in perceived positive net growth rates. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 10 Well location does not significantly impact growth rate. To examine potential well-position bias in growth rate, we quantified growth rates of lys2- cells across various wells of a 96-well plate at an identical initial lysine concentration. (A) We did not observe any correlation between well position and growth rate at 10 µM (non-limiting) lysine in either of the two experiments (left and right). (B) There was a slight correlation between well position and growth rate at 2.5 µM (limiting) lysine. Wells that supported faster growth in one experiment tended to support faster growth in the other (left vs right). This is likely the result of the order in which cells were imaged, as imaging could take over an hour per round, and at 2.5 µM lysine, growth rate continuously declined during measurement (Supp Fig 6). (C) Growth rates averaged across wells in a plate were similar between two independent experiments. Error bars represent 2*standard deviations. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 11 Using an evolved clone to measure low concentrations of lysine. WY2270, an evolved lys2- clone with significantly improved affinity for lysine compared to WY1335, can detect lysine as low as 0.1 µM. Dotted line marks detection limit. Circles and diamonds mark two independent replicates. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Fig 12 Autofocusing. Adequate fluorescent images can be obtained by choosing the local minimum of the autofocus score as our imaging plane. A) The autofocus score A(z) had two local maxima, corresponding to image artifacts created by lensing effects of yeast cells. Bright field (B-F) and fluorescent (C-G) images were taken near z = -24 µm, 0 µm, and 14 µm, corresponding to local maximal, minimal, and maximal A(z), respectively. The yeast cells were in focus at the local minimum. (H) A qualitative explanation of autofocusing. This figure shows the apparent origin of light rays reaching the camera. The density of light rays in a given focal plane corresponds to the apparent brightness. The optical effects are due to focusing of the light rays (blue) by the yeast cells (black circle). The image plane at z = 14 µm (dashed line) contained a sharp bright spot (which looks bigger than a point due to diffraction), and the image plane at z = -24 µm (dashed black line) contained bright halos due to the apparent source of the focused light rays (red lines). H bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. Supp Movie 1 Time-lapse fluorescence microscopy of lys2- at high lysine Supp Movie 2 Time-lapse fluorescence microscopy of lys2- at low (0.66 µM) lysine Supp Movie 3 Time-lapse fluorescence microscopy of lys2- at zero lysine Supp Movie 4 ade8- cell re-dividing after having transiently lost fluorescence. Supp Movie 5 Time-lapse bright field microscopy of lys2- at zero lysine References 1. Müsken M, Di Fiore S, Römling U, Häussler S. A 96-well-plate-based optical method for the quantitative and qualitative evaluation of Pseudomonas aeruginosa biofilm formation and its application to susceptibility testing. Nat Protoc. 2010 Aug;5(8):1460–1469. PMID: 20671729 2. Ziv N, Siegal ML, Gresham D. Genetic and nongenetic determinants of cell growth variation assessed by high-throughput microscopy. Mol Biol Evol. 2013 Dec;30(12):2568–2578. PMCID: PMC3840306 3. Shi H, Colavin A, Lee TK, Huang KC. Strain Library Imaging Protocol for high-throughput, automated single-cell microscopy of large bacterial collections arrayed on multiwell plates. Nat Protoc. 2017;12(2):429–438. PMCID: PMC5831406 4. Monod J. The growth of bacterial cultures. Annu Rev Microbiol. 1949 Oct;3(1):371–394. 5. Kargi F, Shuler ML. Generalized differential specific rate equation for microbial growth. Biotechnol Bioeng. 1979 Oct 1;21(10):1871–1875. 6. Kovárová-Kovar K, Egli T. Growth Kinetics of Suspended Microbial Cells: From Single-SubstrateControlled Growth to Mixed-Substrate Kinetics. Microbiol Mol Biol Rev. 1998 Sep 1;62(3):646–666. PMID: 9729604 7. Shou W, Ram S, Vilar JM. Synthetic cooperation in engineered yeast populations. Proc Natl Acad Sci USA. 2007;104:1877–1882. PMCID: 1794266 8. Momeni B, Chen C-C, Hillesland KL, Waite A, Shou W. Using artificial systems to explore the ecology and evolution of symbioses. Cell Mol Life Sci. 2011 Mar;68(8):1353–1368. PMID: 21424911 9. Waite AJ, Shou W. Adaptation to a new environment allows cooperators to purge cheaters stochastically. Proc Natl Acad Sci. 2012 Nov 20;109(47):19079–19086. PMID: 23091010 10. Momeni B, Brileya KA, Fields MW, Shou W. Strong inter-population cooperation leads to partner intermixing in microbial communities. eLife. 2013;2:e00230. PMCID: 3552619 11. Momeni B, Waite AJ, Shou W. Spatial self-organization favors heterotypic cooperation over cheating. eLife. 2013 Nov 12;2:e00960. PMID: 24220506 12. Wiemken A, Dürr M. Characterization of amino acid pools in the vacuolar compartment of Saccharomyces cerevisiae. Arch Microbiol. 1974 Jan 1;101(1):45–57. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. 13. Skelding DH, Hart SFM, Vidyasagar T, Pozhitkov AE, Shou W. Developing a low-cost milliliter-scale chemostat array for precise control of cellular growth. bioRxiv [Internet]. 2017 Jan 1; Available from: http://biorxiv.org/content/early/2017/11/22/223594.abstract 14. Boer VM, Amini S, Botstein D. Influence of genotype and nutrition on survival and metabolism of starving yeast. Proc Natl Acad Sci. 2008 May 13;105(19):6930–6935. 15. Puig S, Lee J, Lau M, Thiele DJ. Biochemical and Genetic Analyses of Yeast and Human High Affinity Copper Transporters Suggest a Conserved Mechanism for Copper Uptake. J Biol Chem. 2002 Jul 19;277(29):26021–26030. PMID: 11983704 16. Xiao C, Cant JP. Glucose transporter in bovine mammary epithelial cells is an asymmetric carrier that exhibits cooperativity and trans-stimulation. Am J Physiol - Cell Physiol. 2003 Nov 1;285(5):C1226–C1234. PMID: 12867359 17. Casillas T, Delicado EG, Garcia-Carmona F, Miras-Portugal MT. Kinetic and allosteric cooperativity in L-adenosine transport in chromaffin cells. A mnemonical transporter. Biochemistry (Mosc). 1993 Dec 1;32(51):14203–14209. 18. Guthrie C, Fink GR. Guide to yeast genetics and molecular biology. Academic Press; 1991. 19. Waite AJ, Shou W. Constructing synthetic microbial communities to explore the ecology and evolution of symbiosis. Methods Mol Biol Clifton NJ. 2014;1151:27–38. PMID: 24838876 20. Mir H, Xu P, Van Beek P. An extensive empirical evaluation of focus measures for digital photography. International Society for Optics and Photonics; 2014. p. 90230I. 21. Brenner JF, Dew BS, Horton JB, King T, Neurath PW, Selles WD. An automated microscope for cytologic research a preliminary evaluation. J Histochem Cytochem. 1976;24(1):100–111. 22. Smith AE, Zhang Z, Thomas C. Wall material properties of yeast cells: Part 1. Cell measurements and compression experiments. Chem Eng Sci. 2000;55(11):2031–2041. 23. Kim M-S, Scharf T, Mühlig S, Fruhnert M, Rockstuhl C, Bitterli R, Noell W, Voelkel R, Herzig HP. Refraction limit of miniaturized optical systems: a ball-lens example. Opt Express. 2016;24(7):6996–7005. 24. Gordon A, Colman-Lerner A, Chin TE, Benjamin KR, Yu RC, Brent R. Single-cell quantification of molecules and rates using open-source microscope-based cytometry. Nat Methods. 2007 Feb;4(2):175–181. 25. Hartig SM. Basic image analysis and manipulation in ImageJ. Curr Protoc Mol Biol. 2013;14–15. 26. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671. 27. Takahashi CN, Miller AW, Ekness F, Dunham MJ, Klavins E. A Low Cost, Customizable Turbidostat for Use in Synthetic Circuit Characterization. ACS Synth Biol. 2015 Jan 16;4(1):32–38. bioRxiv preprint doi: https://doi.org/10.1101/324269; this version posted May 17, 2018. The copyright holder for this preprint (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 made available under aCC-BY 4.0 International license. 28. Novick A, Szilard L. Experiments with the Chemostat on Spontaneous Mutations of Bacteria. Proc Natl Acad Sci U S A. 1950;36(12):708–719.
https://openalex.org/W2172937016	https://www.frontiersin.org/articles/10.3389/fncir.2015.00054/pdf	English	null	A computational relationship between thalamic sensory neural responses and contrast perception	Frontiers in neural circuits	2,015	cc-by	16,050	ORIGINAL RESEARCH published: 08 October 2015 doi: 10.3389/fncir.2015.00054 A computational relationship between thalamic sensory neural responses and contrast perception Yaoguang Jiang 1, Gopathy Purushothaman 2 and Vivien A. Casagrande 1,2,3* 1 Department of Psychology, Vanderbilt University, Nashville, TN, USA, 2 Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN, USA, 3 Department of Ophthalmology and Visual Sciences, Vanderbilt University, Nashville, TN, USA Uncovering the relationship between sensory neural responses and perceptual decisions remains a fundamental problem in neuroscience. Decades of experimental and modeling work in the sensory cortex have demonstrated that a perceptual decision pool is usually composed of tens to hundreds of neurons, the responses of which are signiﬁcantly correlated not only with each other, but also with the behavioral choices of an animal. Few studies, however, have measured neural activity in the sensory thalamus of awake, behaving animals. Therefore, it remains unclear how many thalamic neurons are recruited and how the information from these neurons is pooled at subsequent cortical stages to form a perceptual decision. In a previous study we measured neural activity in the macaque lateral geniculate nucleus (LGN) during a two alternative forced choice (2AFC) contrast detection task, and found that single LGN neurons were signiﬁcantly correlated with the monkeys’ behavioral choices, despite their relatively poor contrast sensitivity and a lack of overall interneuronal correlations. We have now computationally tested a number of speciﬁc hypotheses relating these measured LGN neural responses to the contrast detection behavior of the animals. We modeled the perceptual decisions with different numbers of neurons and using a variety of pooling/readout strategies, and found that the most successful model consisted of about 50–200 LGN neurons, with individual neurons weighted differentially according to their signal-to-noise ratios (quantiﬁed as d-primes). These results supported the hypothesis that in contrast detection the perceptual decision pool consists of multiple thalamic neurons, and that the response ﬂuctuations in these neurons can inﬂuence contrast perception, with the more sensitive thalamic neurons likely to exert a greater inﬂuence. Edited by: W. Martin Usrey, University of California, Davis, USA Reviewed by: Donald A. Wilson, New York University School of Medicine, USA Henry Joseph Alitto, University of California, Berkeley, USA Reviewed by: Donald A. Wilson, New York University School of Medicine, USA Henry Joseph Alitto, University of California, Berkeley, USA *Correspondence: Vivien A. Casagrande, Department of Cell and Developmental Biology, Vanderbilt University, PMB407935 RM U-3218 MRB 3, 465 21st Avenue South, Nashville, TN 37240-7935, USA vivien.casagrande@vanderbilt.edu Keywords: lateral geniculate nucleus (LGN), perception, contrast, neural model, choice probability Received: 28 July 2015 Accepted: 14 September 2015 Published: 08 October 2015 Citation: Jiang Y, Purushothaman G and Casagrande VA (2015) A computational relationship between thalamic sensory neural responses and contrast perception. Front. Neural Circuits 9:54. doi: 10.3389/fncir.2015.00054 Received: 28 July 2015 Accepted: 14 September 2015 Published: 08 October 2015 Received: 28 July 2015 Accepted: 14 September 2015 Published: 08 October 2015 Keywords: lateral geniculate nucleus (LGN), perception, contrast, neural model, choice probability Citation: Previous modeling work has revealed that interneuronal correlations can have a profound influence on the choice probability structure of the decision pool (Shadlen et al., 1996; Cohen and Newsome, 2009; Nienborg and Cumming, 2010; Haefner et al., 2013). In reality, however, a variety of factors such as the response variances of single neurons and the positive noise correlations between pairs of neurons constrain the pool size, requiring at least 10–1000 sensory neurons in an average sized decision pool (Shadlen et al., 1996; Cook and Maunsell, 2002; Purushothaman and Bradley, 2005; Cohen and Newsome, 2009; Liu et al., 2013). In sensory cortex, such perceptual decision pools have two prominent features. First, stimulus-independent, random fluctuations of sensory neural responses are known to covary with the perceptual decisions of the animal. The strength of this covariation is quantified as ‘‘choice probability’’ (Britten et al., 1996). Weak but significantly above chance choice probabilities have been observed in a number of sensory cortical areas (Britten et al., 1996; Dodd et al., 2001; Cook and Maunsell, 2002; Grunewald et al., 2002; Uka and Deangelis, 2004; Liu and Newsome, 2005; Purushothaman and Bradley, 2005; Uka et al., 2005; Nienborg and Cumming, 2006; Palmer et al., 2007). Second, cortical sensory neurons are also correlated with each other in their random response fluctuations (Averbeck et al., 2006; Cohen and Kohn, 2011). This correlation, known as the interneuronal noise correlation, is likely to reflect the shared feedforward, feedback, or lateral connections between neurons (Zohary et al., 1994; Shadlen and Newsome, 1998; Bair et al., 2001; Reich et al., 2001; Cohen and Maunsell, 2009). Previous modeling work has revealed that interneuronal correlations can have a profound influence on the choice probability structure of the decision pool (Shadlen et al., 1996; Cohen and Newsome, 2009; Nienborg and Cumming, 2010; Haefner et al., 2013). suggested a positive relationship between neural sensitivity, choice probability, and readout weight (for example see Britten et al., 1996; Shadlen et al., 1996; Purushothaman and Bradley, 2005; Haefner et al., 2013; Liu et al., 2013), we examined not only the standard uniform readout model but also several alternative weighted readout schemes in which individual neurons were assigned different weights based on their sensitivities. We accepted or rejected these models based on their ability to account for the behavioral performance of the monkeys as well as the measured choice probability values for LGN neurons (see Jiang et al., 2015). Subjects Two macaque monkeys (monkey 1: Macaca radiata, male, 7 kg, 10 years old; monkey 2: Macaca mulatta, male, 8 kg, 12 years old) served as subjects. The monkeys were treated and cared for in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the guidelines of Vanderbilt University Animal Care and Use Committee under an approved protocol. The monkeys underwent sterile procedures for the implantation of head posts and recording chambers. The chambers were centered over the right LGN of monkey 1 (AP = 7, ML = 12.5) and the left LGN of monkey 2 (AP = 7, ML = 12). g g Such interneuronal correlation or choice probability measurements, however, are rarely made in subcortical structures (but see Liu et al., 2013). In the mammalian visual system, the retina sends direct input to the lateral geniculate nucleus (LGN) of the thalamus which, in turn, relays this information to the visual cortex. Recently, we reported the first study in which LGN neural responses were examined in detail while the animals were required to make perceptual decisions using the information available within the receptive fields of those LGN neurons (Jiang et al., 2015). In a two alternative forced choice (2AFC) contrast detection task, we found that the majority of single LGN parvocellular (P) and magnocellular (M) neurons were not as sensitive as the monkeys. Importantly, the covariation between neural responses and perceptual decisions, measured as choice probability, was significant for both P and M neurons, even though the average interneuronal correlation between LGN neuron pairs was not different from zero. Additionally, both neural sensitivity and choice probability evolved throughout the stimulus presentation time, with M neurons exhibiting faster and more transient response profiles than P neurons (Jiang et al., 2015). Frontiers in Neural Circuits \| www.frontiersin.org Citation: Aspects of the modeling data presented here have been published in abstract form (Jiang et al., 2012, 2013). Materials and Methods All the experimental procedures regarding surgical preparation, animal training, stimulus presentation, and physiological recordings have been described in detail in previous publications (Jiang et al., 2013, 2015), and are therefore only briefly repeated here when relevant. Citation: From smelling a flower to recognizing the face of a loved one, every perceptual task we face, simple or complex, involves a number of neurons in a wide range of brain areas. Of essential interest to neuroscientists is the number of sensory neurons needed to sustain a perception, and the way these neurons are decoded at a later stage to form various decisions. Theoretically, every perceptual task can be accomplished by engaging only the few sensory neurons that are the most sensitive for that task (i.e., the lower envelope principle; Barlow, 1995; Parker and Newsome, 1998). Jiang Y, Purushothaman G and Casagrande VA (2015) A computational relationship between thalamic sensory neural responses and contrast perception. Front. Neural Circuits 9:54. doi: 10.3389/fncir.2015.00054 October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 1 Computation of contrast in thalamus Jiang et al. In reality, however, a variety of factors such as the response variances of single neurons and the positive noise correlations between pairs of neurons constrain the pool size, requiring at least 10–1000 sensory neurons in an average sized decision pool (Shadlen et al., 1996; Cook and Maunsell, 2002; Purushothaman and Bradley, 2005; Cohen and Newsome, 2009; Liu et al., 2013). In sensory cortex, such perceptual decision pools have two prominent features. First, stimulus-independent, random fluctuations of sensory neural responses are known to covary with the perceptual decisions of the animal. The strength of this covariation is quantified as ‘‘choice probability’’ (Britten et al., 1996). Weak but significantly above chance choice probabilities have been observed in a number of sensory cortical areas (Britten et al., 1996; Dodd et al., 2001; Cook and Maunsell, 2002; Grunewald et al., 2002; Uka and Deangelis, 2004; Liu and Newsome, 2005; Purushothaman and Bradley, 2005; Uka et al., 2005; Nienborg and Cumming, 2006; Palmer et al., 2007). Second, cortical sensory neurons are also correlated with each other in their random response fluctuations (Averbeck et al., 2006; Cohen and Kohn, 2011). This correlation, known as the interneuronal noise correlation, is likely to reflect the shared feedforward, feedback, or lateral connections between neurons (Zohary et al., 1994; Shadlen and Newsome, 1998; Bair et al., 2001; Reich et al., 2001; Cohen and Maunsell, 2009). Tasks The monkeys were first trained to fixate on a central fixation spot for an extended period of time. Next, the monkeys were trained to perform a two-alternative forced choice (2AFC) contrast detection task, in which a contrast stimulus was presented either at the receptive field location of the cell being recorded, or at a symmetrical location in the opposite visual hemi-field, for a fixed duration (200 ms). The monkeys saccaded to one of the two target locations to indicate the side on which the stimulus was presented. The stimulus diameter was always the sum of the classical receptive field diameter (center and surround) plus the fixation window diameter. During each recording session, stimuli of 5 or 9 different contrast levels (including 0% contrast, or blank trials, where no physical stimulus was presented) were presented at each location. Different contrast levels and presentation locations (i.e., left or right) were randomly mixed, with equal probabilities of left or right appearance and higher proportions of low to medium contrast trials to ensure accurate estimations of the psychophysical threshold. Taking advantage of this previously characterized dataset and using a computational approach, we investigated in this study how single LGN neurons contribute to our perception of contrast. We built a series of models to explore the interaction between the size of the decision pool, the duration of integration time, and the pooling/readout strategy of the neural system. Because previous experimental and computational work has October 2015 \| Volume 9 \| Article 54 2 Computation of contrast in thalamus Jiang et al. Cell Mapping and Classiﬁcation LGN cells were hand mapped using first a flashlight and then an elongated bar with a sharp contrast profile. Cells were classified as ON-center or OFF-center cells, and Parvocellular (P) or Magnocellular (M) cells, based on their visually driven responses (Norton and Casagrande, 1982; Norton et al., 1988; Xu et al., 2001; Royal et al., 2006; Jiang et al., 2015). Psychometric Functions plotting, as an ROC curve, the proportion of choice-inside- receptive-field trials (i.e., trials in which the monkey saccaded towards the receptive field location) against the proportion of choice-outside-receptive-field trials that exceeded the spike count criteria, and computing the area under that curve. The significance of individual or population choice probabilities was assessed using permutation tests (Britten et al., 1996; Jiang et al., 2015). To accurately estimate choice probability, only neural recordings that met the following criteria were included in this analysis: (1) Behavior ratio (choice-inside/choice-outside-receptive-field trials) > 0.25 and <4; and (2) For every contrast level that was included in the choice probability computation, at least 10 choice- inside and 10 choice-outside-receptive-field trials were recorded. Out of the 89 neurons in the above dataset, 75 (54 P neurons, 21 M neurons) were included in the choice probability analysis according to these criteria. We found in this dataset that, in the absence of any physical stimulus (i.e., 0% contrast, blank trials only), the average choice probability was 0.54 ± 0.01 for LGN P neurons and 0.54 ± 0.01 for LGN M neurons, both above chance according to permutation tests (P neuron: P = 0.015, M neuron: P = 0.033; Jiang et al., 2015). The proportion of correct responses from the monkeys was plotted for each contrast, and a Weibull function was fitted to the data: P(c) = 1 −0.5 ∗e−( c α )β (1) (1) Where P(c) is the probability of correct responses at contrast level c, α is the contrast level that supports threshold performance (82% correct), and β is the slope of the function. Pooling Model g The basic structure of our pooling model was similar to other bottom up pooling models previously proposed to account for the psychophysical threshold and choice probabilities measured during behavioral tasks (Shadlen et al., 1996; Purushothaman and Bradley, 2005; Cohen and Newsome, 2009; Haefner et al., 2013; Liu et al., 2013). Briefly, to simulate a perceptual decision pool of n units, n single neurons were randomly chosen, with replacement, from our entire dataset. To construct a single trial at a given contrast, we simulated each neuron’s response by randomly drawing a number from a Gaussian distribution; the mean and variance of this distribution were determined by that neuron’s measured response at that contrast level. In each trial, the model made a ‘‘choice’’ by comparing the summed activity of the neural pool at the test contrast level to the summed activity of the same neurons at the reference contrast (i.e., blank, 0% contrast trials). This procedure was repeated 50 times (i.e., to simulate 50 trials) for each of the 5 contrast levels, and the simulated ‘‘psychophysical’’ performance was recorded as the percentage of correct ‘‘choices’’ at each contrast. This performance was fitted with a cumulative Weibull function [1], and threshold and slope parameters were extracted as described above. The choice probability for each simulated neuron was quantified as the covariation between the simulated neural response and the simulated ‘‘psychophysical choice’’ at 0% contrast. The parameters for the model included the number of neurons in the pool (n = 1–512 neurons), the integration time window (t = 25–200 ms), the Fano factor (f = 0.25–3.0), the interneuronal noise correlation (r = 0–0.3), and the downstream pooling noise (p = 0–4.0). For each parameter combination, the set of simulations described above (50 trials ∗5 contrast levels) was repeated 200 times, each time with a new random Neurometric Functions Basic procedures in computing neurometric functions were similar to those described in previous studies (Barlow et al., 1971; Britten et al., 1992; Purushothaman and Bradley, 2005). For every contrast level, an ROC (Receiver Operating Characteristic) curve was computed (Green and Swets, 1966). Each ROC curve plotted, for all possible signal detection criteria (spikes), the proportion of stimulus-inside-receptive-field trials where the spike count exceeded a certain criterion, against the proportion of stimulus-outside-receptive-field trials that exceeded the same criterion. Next each area-under-ROC curve value was calculated and plotted against its corresponding contrast, a Weibull function ([1], as described above) was fitted, and the neurometric threshold and slope were obtained from the fitted curve. All the LGN cells that could be clearly mapped and maintained long enough to characterize both the psychophysical and neural responses (>150 trials, overall psychophysical performance >65% correct) in the detection task were included in this analysis (overall: n = 89 neurons; monkey 1: n = 61; monkey 2: n = 28). We identified in this dataset 41 ON-center P neurons, 27 OFF-center P neurons, 19 ON-center M neurons, and 2 OFF-center M neurons. We found that the average neurometric threshold (54.4 ± 4.78% contrast, n = 89 neurons, in 0–150 ms integration time windows) was significantly different from the simultaneously measured psychometric threshold (5.76 ± 0.72% contrast, P = 0.000, Wilcoxon signed rank test). The average ratio of neurometric to psychometric threshold was 40.74 ± 10.75, indicating that the average LGN neuron was much less sensitive than the monkey in contrast detection (Jiang et al., 2015). Frontiers in Neural Circuits \| www.frontiersin.org Model Parameters In our simulations we typically used fixed or experimentally measured values (Jiang et al., 2015) for variables such as the Fano factor, the interneural correlation, and the pooling noise (with the exception of Figure 1, where these parameters were systematically varied to probe the basic properties of the pooling model). We consider our choices for these parameter values to be neurobiologically realistic and meaningful for the following reasons: (1) The Fano factor: Recordings in anesthetized as well as alert animals have reported significant variabilities in the responses of single cortical neurons, with the Fano factor (response variance/mean) averaging 1.0–3.0 (Tolhurst et al., 1983; McAdams and Maunsell, 1999; Oram et al., 1999; Gu et al., 2007). The Fano factor of subcortical visual neurons, however, is relatively low (i.e., <1.0). This is true for retinal ganglion cells (Levine et al., 1992; Berry et al., 1997; Reich et al., 1997) as well as LGN cells (Kara et al., 2000). The Fano factors measured in our detection task (Jiang et al., 2015) and used in our models (0.8–1.4, depending on integration time) were in agreement these previously reported measurements (2) Interneuronal correlation: In sensory cortex, interneuronal correlations between pairs of nearby neurons are typically weak but positive (∼0.1–0.2; Averbeck et al., 2006). For the LGN P-P and M-M neuron pairs, because convergent feedforward, divergent feedforward, and lateral connections are sparser than those in the cortex (Casagrande and Xu, 2004; Nassi and Callaway, 2009), it is not surprising that we found an average interneuronal correlation (0.028) that was not significantly different from 0.0. The interneuronal correlation for a P-M neuron pair is very likely to be even smaller, as the P and M pathways receive different retinal inputs, remain segregated in different layers of the LGN (Casagrande and Norton, 1991; Nassi and Callaway, 2009), and retain separate feedback loops with V1 (Ichida and Casagrande, 2002; Briggs and Usrey, 2009, 2011; Ichida et al., 2014). In our model the P-M correlation was fixed at 0.01, but our simulations could always approach >99% GoF at some parameter combinations, given any P-M correlation values between 0.0–0.05 (data not shown). (3) Pooling noise: The downstream pooling noise can be thought of as the average Fano factor of the cortical neurons onto which LGN neurons converge (Shadlen et al., 1996). Choice Probability Basic procedures in computing choice probabilities also were similar to those described in previous studies (Britten et al., 1996; Purushothaman and Bradley, 2005). The choice probability for a certain contrast was measured by October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 3 Computation of contrast in thalamus Jiang et al. sample of n neurons (with replacement), thus giving reliable estimations of the model performance. The overall fitness of the model was evaluated by computing a Goodness-of-Fit (GoF) index: depend on this assumption. Similar model results could be obtained by simply assuming a true Poisson distribution for all downstream neurons (i.e., pooling noise = 1.0). Uniform Pooling and Alternative Pooling Strategies GoF = (1 −1/3 ∗(\|simulated threshold −measured threshold\|/measured threshold + \|simulated P choice probability −measured P choice probability\|/measured P choice probability + \|simulated M choice probability −measured M choice probability\|/measured M choice probability)) ∗100% (2) GoF = (1 −1/3 ∗(\|simulated threshold −measured threshold\|/measured threshold + \|simulated P choice In this paper, we examined uniform pooling models as well as several alternative weighted pooling models. All of these models shared the same overall structure: Xpooled = n X i = 1 wixi (3) (3) (2) Where Xpooled is the summed activity of the perceptual decision pool, xi is the response of a single neuron, and wi is the readout weight assigned to this neuron. Within this structure, the uniform pooling model simply assigned equal weights (i.e., wi = 1.0) to all neurons in the decision pool, regardless of their sensitivities (Shadlen et al., 1996). An alternative weighted pooling strategy, in contrast, calculated the sensitivity of each individual neuron and assigned weights accordingly. Depending on how this neural sensitivity was quantified, there were three main categories of weighted pooling schemes: (1) Amplitude- per-trial (amp/trial) weighted scheme, where every neuron was weighted according to its response amplitude in every trial (wi ∝xi), with the neuron with the highest spike rate carrying a weight of 1.0; (2) Mean amplitude (mean amp) weighted scheme, where every neuron was weighted according to its average response amplitude (wi ∝xi) at high contrast (80–99%), with the neuron with the highest average spike rate carrying a weight of 1.0; and (3) D-prime weighted scheme, where every neuron was weighted by its d-prime value (wi ∝d ′ i ) at high contrast (80–99%), with the neuron with the greatest d-prime carrying a weight of 1.0. Here d-prime was defined as: Frontiers in Neural Circuits \| www.frontiersin.org Model Parameters In our simulations this number was fixed at 2.0, which is the average estimation of the Fano factor in cortex (see above). The success of our simulations (i.e., approaching >99% GoF at some parameter combinations), however, did not . d ′ i = (xi −x0i) s s2 i + s02 i 2 (4) (4) Where xi is the neuron’s mean response amplitude at high contrast (80–99%), x0i is its mean response amplitude at reference contrast (i.e., blank, 0% contrast), and si and s0i represent the corresponding standard deviations. Because of their potential deviations from normality, for both weight and d-prime distributions we reported medians as well as means. Additionally, to characterize the spread of a distribution, we reported the interquartile range: Interquatile Range (IQR) = Q3 −Q1 (5) (5) Where Q1 is the 1st quartile (i.e., 25% percentile), and Q3 is the 3rd quartile (i.e., 75% percentile) in the range. To characterize the skewness of a distribution, we reported the skewness index: Skewness index (SI) = 1 n Pn i = 1 (xi −x)3 q 1 n Pn i = 1 (xi −x)23 (6) (6) Where n is the sample size, and x is the mean of the sample distribution. Where n is the sample size, and x is the mean of the sample distribution. Frontiers in Neural Circuits \| www.frontiersin.org October 2015 \| Volume 9 \| Article 54 4 Jiang et al. Computation of contrast in thalamus FIGURE 1 \| Parametric analysis of the pooling model. (A,B) Increasing the number of neurons in the pool decreased the simulated threshold and choice probability. Interneuronal correlation and pooling noise were assumed to be 0; integration time was 0–150 ms. Magenta: simulated P neuron; green: simulated M neuron; thick gray line: measured mean psychophysical threshold/choice probability; thin gray line/rectangle: mean ± SEM. (C,D) Increasing the Fano factor increased simulated threshold but maintained choice probability, for both P neurons (C) and M neurons (D). Interneuronal correlation and pooling noise were assumed to be 0; integration time was 0–150 ms. As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Gray line: measured mean psychophysical threshold/choice probability; rectangle: mean ± SEM. (E,F) Increasing interneuronal correlation increased simulated threshold and choice probability, for both P neurons (E) and M neurons (F). Model Parameters The Fano factor was assumed to be 1.03 (measured value); pooling noise was assumed to be 0; integration time was 0–150 ms. As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Legends as in (C,D). (G,H) Increasing pooling noise increased simulated threshold and decreased choice probability, for both P neurons (G) and M neurons (H). The Fano factor was assumed to be 1.03 (measured value); interneuronal correlation was assumed to be 0.028 (measured value); integration time was 0–150 ms. As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Legends as in (C,D). (A,B) and (E–H) were adapted from Jiang et al. (2015). FIGURE 1 \| Parametric analysis of the pooling model. (A,B) Increasing the number of neurons in the pool decreased the simulated threshold and choice probability. Interneuronal correlation and pooling noise were assumed to be 0; integration time was 0–150 ms. Magenta: simulated P neuron; green: simulated M neuron; thick gray line: measured mean psychophysical threshold/choice probability; thin gray line/rectangle: mean ± SEM. (C,D) Increasing the Fano factor increased simulated threshold but maintained choice probability, for both P neurons (C) and M neurons (D). Interneuronal correlation and pooling noise were assumed to be 0; integration time was 0–150 ms. As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Gray line: measured mean psychophysical threshold/choice probability; rectangle: mean ± SEM. (E,F) Increasing interneuronal correlation increased simulated threshold and choice probability, for both P neurons (E) and M neurons (F). The Fano factor was assumed to be 1.03 (measured value); pooling noise was assumed to be 0; integration time was 0–150 ms. As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Legends as in (C,D). (G,H) Increasing pooling noise increased simulated threshold and decreased choice probability, for both P neurons (G) and M neurons (H). The Fano factor was assumed to be 1.03 (measured value); interneuronal correlation was assumed to be 0.028 (measured value); integration time was 0–150 ms. Uniform Pooling Model: Parameters Uniform Pooling Model: Parameters The uniform pooling model we built was similar to a number of previous models used to account for psychophysical performance and choice probability measurements based on sensory neural responses (Shadlen et al., 1996; Purushothaman and Bradley, 2005; Cohen and Newsome, 2009; Haefner et al., 2013; Liu et al., 2013). Inputs to the model were single LGN P and M neural responses at different contrasts. Outputs were the simulated ‘‘psychophysical’’ threshold and the simulated choice probabilities for individual neurons. In agreement with Shadlen et al. (1996), this uniform pooling model behaved predictably when certain model parameters were changed. Specifically, increasing the number of neurons in the pool (n) decreased the psychophysical threshold and choice probability values (Figures 1A,B). Increasing the Fano factor increased the simulated threshold but maintained the same choice probability values (Figures 1C,D). Increasing the interneuronal correlation increased the threshold as well as choice probability values (Figures 1E,F). Increasing the downstream pooling noise increased the threshold and decreased choice probability values (Figures 1G,H). For the simulation results reported in the following sections, the Fano factor and interneuronal noise correlation were fixed at experimentally measured values and the pooling noise, which could be considered as the Fano factor of downstream neurons onto which LGN neurons converge, was assumed to be 2.0 (see ‘‘Material and Methods’’ Section). To evaluate the overall performance of this model, a Goodness-of-Fit (GoF) index (see equation [2]) was reported for each (n, t) parameter combination. A GoF (ranging from 0–100%) reflected three factors equally: (1) how close the simulated ‘‘psychophysical’’ threshold approached the measured psychophysical threshold; (2) how close the simulated P population choice probability approached the measured average choice probability for P neurons; and (3) how close the simulated M population choice probability approached the measured average choice probability for M neurons. A GoF of 100% indicated that our simulation perfectly matched the observed psychometric threshold and the choice probabilities for both types of neurons. Model Parameters As in (B), each simulation line moved from the upper right corner to the lower left corner as more neurons were added in the pool. Legends as in (C,D). (A,B) and (E–H) were adapted from Jiang et al. (2015). October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org Computation of contrast in thalamus Jiang et al. Uniform Pooling Model: Parameters By changing the duration of the integration window (25–200 ms), we found that: (1) at extremely short intervals (25 ms), even incorporating a large number of neurons from both groups (n = 512 P neurons, 512 M neurons) still failed to reproduce the observed threshold and choice probabilities (Figure 4A); (2) in 50 ms, preferably incorporating a large number of M neurons (n = 256–512) could explain the observed threshold and choice probabilities (Figure 4B); (3) in 75 ms, incorporating a large number of either P or M neurons (n = 128–512) could achieve good overall model performance (Figure 4C); and (4) at medium to long intervals (100–200 ms), a smaller number of P and M neurons were needed (n = 32–128) to achieve good model performance, but further increasing the number of neurons resulted in a decrease in model performance (Figures 4D–F). Uniform Pooling Model: Performance The first pooling scheme we investigated was the uniform pooling model, in which the responses of all neurons were weighted equally and summed up to form perceptual decisions. In this model, the simulated psychometric threshold consistently decreased as: (1) more neurons (n) were added into the pool; and (2) the integration time window (t) was extended. Specifically, we found that: (1) at extremely short intervals (25 ms), incorporating a large number of neurons from both the P and M populations (n = 512 P neurons, 512 M neurons) still failed to achieve great psychophysical sensitivities (i.e., threshold <10% contrast), but incorporating a large number of M neurons (n = 256–512) rather than P neurons was more beneficial to model performance (Figure 2A); (2) at relatively brief intervals (50 ms), preferably Results incorporating a large number of M neurons (n = 256–512) rather than P neurons produced good psychophysical performance (i.e., threshold <10% contrast; Figure 2B); (3) at medium to long intervals (75–200 ms), a wider range of M/P neuron combinations (n = 64–512) yielded good model performance (Figures 2C–F). As previously reported (Jiang et al., 2015), we found that single LGN P and M neurons, although not as sensitive as the monkeys in detecting contrast, were significantly correlated with the behavioral choices of the monkeys during a 2AFC contrast detection task. Based on these experimental data, we report in this paper a series of modeling results in the following order: first we describe the basic parameters of the uniform pooling model and its performance in different time frames, then we compare several alternative weighted pooling schemes to the uniform pooling model, and finally we zoom in on one of the best performing weighted schemes and examine its structure in detail. Next we analyzed the simulated choice probabilities for the P and M populations and compared them to the measured choice probability distributions (Jiang et al., 2015). Briefly, in the 2AFC contrast detection task, we found that in the absence of any physical stimulus (i.e., 0% contrast, blank trials only), the average choice probability was 0.54 ± 0.01 for LGN P neurons and 0.54 ± 0.01 for LGN M neurons, both above chance according to permutation tests (P neuron: P = 0.015, M neuron: P = 0.033). In the uniform pooling model, the simulated choice probability distributions for P and M neurons (n = 512 neurons, t = 0–150 ms, P choice probability = 0.54 ± 0.00, M choice probability = 0.54 ± 0.00) resembled their experimentally measured counterparts (P > 0.05, permutation tests; Figures 3A,B). These choice probability patterns remained unchanged throughout the 200 ms stimulus presentation time (n = 512 neurons, P > 0.05, 1-way ANOVAs; Figures 3C,D). Alternative Pooling Schemes In this section we examine several alternative pooling schemes where, instead of assigning the same weight to all neurons, each individual neuron was weighted differentially based on its response rate or sensitivity. We investigated three main categories of weighted pooling schemes, namely the Frontiers in Neural Circuits \| www.frontiersin.org October 2015 \| Volume 9 \| Article 54 6 Jiang et al. Computation of contrast in thalamus FIGURE 2 \| The simulated “psychophysical” thresholds for different pool sizes (n) and integration time windows (t). Red indicates high threshold (100% contrast), whereas blue indicates low threshold (0% contrast). (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. FIGURE 2 \| The simulated “psychophysical” thresholds for different pool sizes (n) and integration time windows (t). Red indicates high threshold (100% contrast), whereas blue indicates low threshold (0% contrast). (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. interaction effect; Figure 5A). Next we examined whether these alternative pooling strategies improved the sensitivity of the model when compared to the uniform pooling strategy. Here the mean amplitude weighted and the d-prime weighted models could be further divided into two subcategories, respectively, depending on whether P and M neurons were weighted separately or together in the model. Among all of these alternative pooling schemes, we found that only the d-prime weighted schemes consistently improved the psychophysical performance when compared with the uniform pooling scheme (mean difference, d-prime 1 = −4.88 ± 0.42% contrast; amplitude-per-trial (amp/trial) weighted, the mean amplitude (mean amp) weighted, and the d-prime weighted schemes (see ‘‘Material and Methods’’ Section). First we compared the simulated psychometric thresholds and found that for all pooling schemes the average psychometric threshold decreased with time (n = 1–512 neurons, F = 920.22, P = 0.00, 2- way ANOVA main effect for time). Furthermore, there was a significant difference in psychometric thresholds among different pooling schemes (F = 54.82, P = 0.00, 2-way ANOVA main effect for pooling strategy), and this difference changed across time (F = 8.43, P = 0.00, 2-way ANOVA October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 7 Jiang et al. Alternative Pooling Schemes Computation of contrast in thalamus FIGURE 3 \| The simulated choice probabilities of P and M neurons for different pool sizes (n) and integration time windows (t). (A) Cumulative choice probability distributions for P (magenta) and M (green) neurons in 200 simulations (n = 512 neurons, t = 0–150 ms). (B) Choice probability distributions for P (magenta) and M (green) neurons in 200 simulations (n = 512 neurons, t = 0–150 ms). Arrow: mean choice probability; solid line: choice probability = 0.5. (C) Cumulative choice probability distributions for P (magenta) neurons in different integration time windows (n = 512 neurons, 200 simulations for each integration time). (D) Cumulative choice probability distributions for M (green) neurons in different integration time windows (n = 512 neurons, 200 simulations for each integration time). FIGURE 3 \| The simulated choice probabilities of P and M neurons for different pool sizes (n) and integration time windows (t). (A) Cumulative choice probability distributions for P (magenta) and M (green) neurons in 200 simulations (n = 512 neurons, t = 0–150 ms). (B) Choice probability distributions for P (magenta) and M (green) neurons in 200 simulations (n = 512 neurons, t = 0–150 ms). Arrow: mean choice probability; solid line: choice probability = 0.5. (C) Cumulative choice probability distributions for P (magenta) neurons in different integration time windows (n = 512 neurons, 200 simulations for each integration time). (D) Cumulative choice probability distributions for M (green) neurons in different integration time windows (n = 512 neurons, 200 simulations for each integration time). mean difference, d-prime 2 = −4.67 ± 0.43% contrast; P < 0.05, Tukey’s HSD tests for multiple comparisons). The mean amplitude weighted schemes and the amplitude per trial weighted scheme all failed to perform as well as the uniform pooling scheme in terms of the threshold (mean difference, mean amp 1 = 2.59 ± 0.21% contrast; mean difference, mean amp 2 = 3.31 ± 0.23% contrast; mean difference, amp/trial = 7.51 ± 0.28% contrast; P < 0.05, Tukey’s HSD tests for multiple comparisons). Additionally, the two subtypes of mean amplitude weighted models did not differ from each other in terms of their simulated thresholds, and the two subtypes of d-prime weighted models did not differ from each other either (P > 0.05, Tukey’s HSD tests for multiple comparisons; Figure 5B). Frontiers in Neural Circuits \| www.frontiersin.org Alternative Pooling Schemes Finally, the minimal psychophysical threshold achieved by the model also decreased with time in all pooling schemes and plateaued at around 50–75 ms after stimulus onset (n = 1–512 neurons, minimal threshold = 2–3% contrast; Figure 5C). Taken together, Figures 5A–C demonstrated that the d-prime weighted pooling strategies were the most optimal in terms of the simulated psychophysical performance, and this advantage over other pooling strategies was the most apparent in short integration time windows (25–50 ms). Next, we compared the simulated choice probabilities in these different pooling schemes. First, in a fixed time window (t = 0–150 ms), the overall choice probability distributions for simulated P or M neurons did not differ significantly among the pooling schemes (n = 512 neurons, P > 0.05, 1-way ANOVAs; Figures 5D,E). Second, these choice probability distributions did not change with time in the case of the amplitude per trial and mean amplitude weighted models (n = 512 neurons, P > 0.05, 2-way ANOVAs main effect for time; Figures 5F,G). In the d-prime weighted pooling model, however, the choice probability distributions did shift significantly as the integration time window was extended (n = 512 neurons, F = 4.86, P = 0.00, 2-way ANOVA main effect for time; Figure 5H), corresponding well to the temporal dynamics of choice probability that were experimentally measured in LGN P and M neurons (see Figure 7 of Jiang et al., 2015). Finally, we investigated the overall performance of different pooling strategies by comparing their GoF indices in different time windows. We found that for all pooling schemes the overall October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 8 Jiang et al. Computation of contrast in thalamus FIGURE 4 \| The Goodness-of-Fit (GoF) indices for different pool sizes (n) and integration time windows (t). A GoF of 100% (white) indicates that the model perfectly matches the observed psychometric threshold as well as choice probabilities for both P and M neurons. (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. (A,B) and (D–F) were adapted from Jiang et al. (2015). FIGURE 4 \| The Goodness-of-Fit (GoF) indices for different pool sizes (n) and integration time windows (t). Frontiers in Neural Circuits \| www.frontiersin.org Alternative Pooling Schemes A GoF of 100% (white) indicates that the model perfectly matches the observed psychometric threshold as well as choice probabilities for both P and M neurons. (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. (A,B) and (D–F) were adapted from Jiang et al. (2015). GoF; mean difference, d-prime 2 = 0.95 ± 0.15% GoF; P < 0.05, Tukey’s HSD tests for multiple comparisons). The mean amplitude weighted schemes did not differ significantly from the uniform pooling scheme (mean difference, mean/amp 1 = −0.64 ± 0.25% GoF; mean difference, mean amp 2 = −0.79 ± 0.26% GoF; P > 0.05, Tukey’s HSD tests for multiple comparisons), whereas the amplitude per trial weighted scheme failed to perform as well as the uniform pooling scheme (mean difference, amp/trial = −2.49 ± 0.29% GoF; P < 0.05, Tukey’s HSD test for multiple comparisons). Additionally, the two subtypes of mean amplitude weighted models did not differ from each other in model performance improved with time (n = 1–512 neurons, F = 223.78, P = 0.00, 2-way ANOVA main effect for time). Furthermore, there was a significant difference in overall fitness among different pooling schemes (F = 19.06, P = 0.00, 2-way ANOVA main effect for pooling strategy), and this difference changed across time (F = 2.75, P = 0.00, 2-way ANOVA interaction effect; Figure 6A). Next, we examined whether the alternative pooling strategies improved upon the performance of the uniform pooling model. We found that, again, only the d-prime weighted schemes consistently improved the overall fitness of the model (mean difference, d-prime 1 = 1.10 ± 0.15% October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 9 Jiang et al. Computation of contrast in thalamus FIGURE 5 \| A comparison of threshold and choice probability values derived from different pooling schemes. (A) The average psychometric threshold decreased with time in all pooling schemes (n = 1–512 neurons). Black: uniform pooling (as above); gray: every neuron weighted by its response amplitude in every trial; blue: every neuron weighted by its average response amplitude at high contrast (80–99%); magenta: every neuron weighted by its d-prime value at high contrast (80–99%); error bar: mean ± SEM. Alternative Pooling Schemes (B) Different pooling schemes yielded signiﬁcantly different psychometric thresholds when compared with the uniform pooling model (n = 1–512 neurons). Gray: every neuron weighted by its response amplitude in every trial; blue with circle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal responses; blue with triangle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal response; magenta with circle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal d-primes; magenta with triangle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal d-prime; y axis: the difference in psychometric threshold (alternative pooling scheme—uniform, % contrast); error bar: mean ± SEM. (C) The minimal psychometric threshold achieved by the model decreased with time in all pooling schemes. (D,E) Cumulative choice probability distributions for P (D) and M (E) neurons (n = 512 neurons, t = 0–150 ms) in different pooling schemes. (F–H) Cumulative choice probability distributions for P (magenta) and M (green) neurons in different integration time windows (n = 512 neurons), with every neuron weighted by its response amplitude in every trial (F), by its average response amplitude (G), or by its d-prime (H). FIGURE 5 \| A comparison of threshold and choice probability values derived from different pooling schemes. (A) The FIGURE 5 \| A comparison of threshold and choice probability values derived from different pooling schemes. (A) The average psychometric threshold decreased with time in all pooling schemes (n = 1–512 neurons). Black: uniform pooling (as above); gray: every neuron weighted by its response amplitude in every trial; blue: every neuron weighted by its average response amplitude at high contrast (80–99%); magenta: every neuron weighted by its d-prime value at high contrast (80–99%); error bar: mean ± SEM. (B) Different pooling schemes yielded signiﬁcantly different psychometric thresholds when compared with the uniform pooling model (n = 1–512 neurons). d-prime Weighted Pooling Scheme: Structure d-prime Weighted Pooling Scheme: Structure Next, we examined the relationship between d-primes, weights, and choice probabilities for different cell types within the d-prime pooling model. First, intuitively, as the model was allowed to integrate firing rate information for longer durations, the overall d-prime distributions extended accordingly, for both P and M populations (F = 87.74, P = 0.00, 2-way ANOVA main effect for time; Figures 9A,B). Next, in a fixed time window of medium duration (n = 512 neurons, t = 0–150 ms), we compared the d-prime distributions for P and M neurons. In P neurons, the average d-prime was 1.70 ± 0.00 and the median was 1.40. In M neurons, the average d-prime was 1.69 ± 0.00 and the median was 1.69 as well. Even though the average d-primes were similar, the shapes of the distributions differed dramatically between the two cell types, with the P d-prime distribution much more widely spread (P interquartile range = 2.05, M interquartile range = 1.25) and positively skewed (P skewness index = 0.73, M skewness index = 0.12; Figures 9C,D). In the same time window (t = 0–150 ms), the pooling weight of each individual neuron was directly determined by its d-prime value, and the weight distributions for P and M neurons were therefore very reminiscent of the their corresponding d-prime distributions in terms of shape. As the pooling weight of a neuron could not exceed 1.0, however, the weight distributions were scaled-down versions of the corresponding d-prime distributions (P weight: mean = 0.29 ± 0.00, median = 0.23; M weight: mean = 0.44 ± 0.00, median = 0.43). As a result, the weight distributions for P and M neurons still differed from each other in terms of skewness (P skewness index = 0.74, M skewness index = 0.15), but they were no longer distinguishable in terms of spread (P interquartile range = 0.35, M interquartile range = 0.32; Figures 9E,F). In terms of the overall fitness quantified as GoF, at extremely short integration time windows (25 ms), the d-prime model failed to reproduce the observed threshold and choice probabilities even when it incorporated a large number of neurons from both the P and M populations (n = 512 P neurons, 512 M neurons; Figure 8A). Alternative Pooling Schemes Gray: every neuron weighted by its response amplitude in every trial; blue with circle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal responses; blue with triangle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal response; magenta with circle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal d-primes; magenta with triangle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal d-prime; y axis: the difference in psychometric threshold (alternative pooling scheme—uniform, % contrast); error bar: mean ± SEM. (C) The minimal psychometric threshold achieved by the model decreased with time in all pooling schemes. (D,E) Cumulative choice probability distributions for P (D) and M (E) neurons (n = 512 neurons, t = 0–150 ms) in different pooling schemes. (F–H) Cumulative choice probability distributions for P (magenta) and M (green) neurons in different integration time windows (n = 512 neurons), with every neuron weighted by its response amplitude in every trial (F), by its average response amplitude (G), or by its d-prime (H). October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org Frontiers in Neural Circuits \| www.frontiersin.org 10 Computation of contrast in thalamus Jiang et al. terms of their GoF indices, and neither did the two subtypes of d-prime weighted models (P > 0.05, Tukey’s HSD tests for multiple comparisons; Figure 6B). Additionally, the maximal GoF achieved by the model also increased with time in all pooling schemes and plateaued at around 50–75 ms after stimulus onset (n = 1–512 neurons, 98–99% GoF; Figure 6C). Finally, the number of neurons needed to achieve the maximal GoF decreased with time in all pooling schemes, but only the uniform and the d-prime weighted pooling schemes were able to achieve maximal fitness (98–99% GoF) with fewer than 100 neurons (Figure 6D). Taken together, Figures 6A–D demonstrated that the d-prime weighted pooling strategies were the most optimal in terms of the overall performance, and this advantage over other pooling strategies was the most apparent in short integration time windows (25–50 ms). Alternative Pooling Schemes number of neurons would result in a decrease in model performance (Figures 8C–F). Comparing Figure 8 (d-prime pooling) with Figure 4 (uniform pooling), it is clear that the temporal evolution of the GoF index for the d-prime model resembled that for the uniform model, but there were apparent differences between the two models in the 50 ms and 75 ms time windows. To be more precise, in the 50 ms window, the d-prime model demonstrated better overall performance than the uniform model (mean difference = 4.85 ± 0.51% GoF, P = 0.00, Wilcoxon signed rank test). In the 75 ms window, in contrast, the overall performance did not differ between the two types of pooling models (mean difference = 0.00 ± 0.48% GoF, P = 0.30, Wilcoxon signed rank test), but the number of neurons needed to achieve good model performance (>90% GoF) was significantly reduced in the d-prime model (d-prime model: n = 64–256 neurons, uniform model: n = 128–512 neurons). d-prime Weighted Pooling Scheme: Performance Of all the alternative pooling strategies described above, we were the most interested in the d-prime weighted strategy because of its superior performance. In the next few sections we discuss in detail the performance, structure and properties of the d-prime model. First, in terms of the simulated psychophysical threshold, at extremely short integration time windows (25 ms), the d-prime model failed to achieve good psychophysical performance (i.e., threshold < 10% contrast) even when it incorporated a large number of neurons from both the P and M populations (n = 512 P neurons, 512 M neurons), but incorporating a large number of M neurons (n = 128–512) rather than P neurons was more beneficial to model performance (Figure 7A). At relatively brief intervals (50 ms), incorporating a large number of either P or M neurons (n = 256–512) could achieve good psychophysical performance (i.e., threshold < 10% contrast; Figure 7B). At medium to long intervals (75–200 ms), a wider range of M/P neuron combinations (n = 64–512) yielded good model performance (Figures 7C–F). Comparing Figure 7 (d-prime pooling) to Figure 2 (uniform pooling), it is clear that the d-prime model behaved rather similarly to the uniform pooling model in terms of its psychophysical performance, but there were apparent differences between the two models in the 25 ms and 50 ms time windows. Frontiers in Neural Circuits \| www.frontiersin.org Alternative Pooling Schemes Specifically, the d-prime model achieved much lower psychophysical thresholds than the uniform model in both time windows (25 ms: mean difference = −25.00 ± 1.13% contrast, P = 0.00, Wilcoxon signed rank test; 50 ms: mean difference = −4.25 ± 0.28% contrast, P = 0.00, Wilcoxon signed rank test). d-prime Weighted Pooling Scheme: Structure In 50 ms windows, incorporating a large number of either P or M neurons (n = 256–512) could explain the observed threshold and choice probabilities (Figure 8B). Finally, at medium to long intervals (75–200 ms), a wider range of M/P neuron combinations (n = 32–256) yielded good model performance, but further increasing the We also analyzed the simulated choice probabilities for P and M neurons in the same time window (t = 0–150 ms), and found that the P and M choice probabilities in the d-prime model (P choice probability = 0.53 ± 0.00, M choice probability = 0.54 ± 0.00) resembled their experimentally measured counterparts (Jiang et al., 2015, also see above; Figure 9G). Furthermore, individual choice probability values were positively correlated October 2015 \| Volume 9 \| Article 54 11 Jiang et al. Computation of contrast in thalamus FIGURE 6 \| A comparison of the GoF indices derived from different pooling schemes. (A) The average GoF increased with time in all pooling schemes (n = 1–512 neurons). Black: uniform pooling (as above); gray: every neuron weighted by its response amplitude in every trial; blue: every neuron weighted by its average response amplitude at high contrast (80–99%); magenta: every neuron weighted by its d-prime value at high contrast (80–99%); error bar: mean ± SEM. (B) Different pooling schemes yielded signiﬁcantly different GoFs when compared with the uniform pooling model (n = 1–512 neurons). Gray: every neuron weighted by its response amplitude in every trial; blue with circle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal responses; blue with triangle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal response; magenta with circle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal d-primes; magenta with triangle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal d-prime; y axis: the difference in GoF (alternative pooling scheme—uniform, % GoF); error bar: mean ± SEM. (C) The maximal GoF achieved by the model increased with time in all pooling schemes. d-prime Weighted Pooling Scheme: Structure Black: uniform pooling (as above); gray: every neuron weighted by its response amplitude in every trial; blue: every neuron weighted by its average response amplitude at high contrast (80–99%); magenta: every neuron weighted by its d-prime value at high contrast (80–99%); error bar: mean ± SEM. (B) Different pooling schemes yielded signiﬁcantly different GoFs when compared with the uniform pooling model (n = 1–512 neurons). Gray: every neuron weighted by its response amplitude in every trial; blue with circle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal responses; blue with triangle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal response; magenta with circle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal d-primes; magenta with triangle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal d-prime; y axis: the difference in GoF (alternative pooling scheme—uniform, % GoF); error bar: mean ± SEM. (C) The maximal GoF achieved by the model increased with time in all pooling schemes. (D) The number of neurons needed to achieve the maximal GoF decreased with time in all pooling schemes. with d-prime values for both P neurons (r = 0.08, P = 0.00) and M neurons (r = 0.04, P = 0.00; Figure 9H), indicating that the more sensitive LGN neurons were also more correlated with the behavioral choices of the monkeys. making inferences as to which model was neurobiologically more meaningful. In the previous section we described the relationship between d-prime, weight, and choice probability in the scenario where P and M neurons were weighted separately according to their respective maximal d-primes, and in this section we perform similar analyses on the alternative d-prime model where P and M neurons were weighted together. d-prime Weighted Pooling Scheme: Structure (D) The number of neurons needed to achieve the maximal GoF decreased with time in all pooling schemes. FIGURE 6 \| A comparison of the GoF indices derived from different pooling schemes. (A) The average GoF increased with time in all pooling schemes (n = 1–512 neurons). Black: uniform pooling (as above); gray: every neuron weighted by its response amplitude in every trial; blue: every neuron weighted by its average response amplitude at high contrast (80–99%); magenta: every neuron weighted by its d-prime value at high contrast (80–99%); error bar: mean ± SEM. (B) Different pooling schemes yielded signiﬁcantly different GoFs when compared with the uniform pooling model (n = 1–512 neurons). Gray: every neuron weighted by its response amplitude in every trial; blue with circle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal responses; blue with triangle: every neuron weighted by its average response amplitude at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal response; magenta with circle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted separately in reference to their respective maximal d-primes; magenta with triangle: every neuron weighted by its d-prime value at high contrast (80–99%), with P and M neurons weighted together in reference to one maximal d-prime; y axis: the difference in GoF (alternative pooling scheme—uniform, % GoF); error bar: mean ± SEM. (C) The maximal GoF achieved by the model increased with time in all pooling schemes. (D) The number of neurons needed to achieve the maximal GoF decreased with time in all pooling schemes. GURE 6 \| A comparison of the GoF indices derived from different pooling schemes. (A) The average GoF increased with tim of the GoF indices derived from different pooling schemes. (A) The average GoF increased with time in all pooling schemes if li ( b ) i ht d b it lit d i t i l bl i ht d b it FIGURE 6 \| A comparison of the GoF indices derived from different pooling schemes. (A) The average GoF increased with time in all pooling schemes (n = 1–512 neurons). d-Prime Weighted Pooling Scheme: Which One to Choose? First, as the d-prime value is a direct reflection of the signal- to-noise ratio of single neural responses, it is not surprising that the d-prime distributions remained the same regardless of the pooling strategy (compare Figures 9C,D to 10A,B). Specifically, for the P population, the average d-prime here was 1.70 ± 0.00 and the median was 1.40. For the M population, the average As mentioned above, the d-prime weighted pooling scheme could be further divided into two subtypes depending on whether the P and M populations were weighted separately or together. These two types of d-prime models were indistinguishable from each other in terms of overall fitness, but we were interested in comparing their structures and detailed properties as well as October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 12 Jiang et al. Computation of contrast in thalamus g p FIGURE 7 \| The simulated “psychophysical” thresholds in the d-prime weighted scheme for different pool sizes (n) and integration time windows (t). Red indicates high threshold (100% contrast), whereas blue indicates low threshold (0% contrast). (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. FIGURE 7 \| The simulated “psychophysical” thresholds in the d-prime weighted scheme for different pool sizes (n) and integration time windows (t). Red indicates high threshold (100% contrast), whereas blue indicates low threshold (0% contrast). (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. was the case here, the weight distributions were still scaled-down versions of their corresponding d-prime distributions (P weight: mean = 0.29 ± 0.00, median = 0.24; M weight: mean = 0.29 ± 0.00, median = 0.29), but both distributions retained their shape and spread. In other words, the weight distributions for P and M neurons still differed from each other in terms of both spread (P interquartile range = 0.36, M interquartile range = 0.21) and skewness (P skewness index = 0.74, M skewness index = 0.14; Figures 10C,D). Two-way ANOVAs confirmed that while the d-prime distributions did not differ d-prime here was 1.70 ± 0.00 and the median was 1.70 as well. d-Prime Weighted Pooling Scheme: Which One to Choose? Additionally, the P and M d-prime distributions differed significantly in their shapes, with the P d-prime distribution much more widely spread (P interquartile range = 2.04, M interquartile range = 1.25) and positively skewed (P skewness index = 0.73, M skewness index = 0.11; Figures 10A,B). The weight distributions for P and M neurons in the same time window (t = 0–150 ms), however, were very different between the two types of d-prime models (compare Figures 9E,F to 10C,D). Specifically, when P and M neurons were pooled together, as October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 13 Jiang et al. Computation of contrast in thalamus FIGURE 8 \| The GoF indices in the d-prime weighted scheme for different pool sizes (n) and integration time windows (t). A GoF of 100% (white) indicates that the model perfectly matches the observed psychometric threshold as well as choice probabilities for both P and M neurons. (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. FIGURE 8 \| The GoF indices in the d-prime weighted scheme for different pool sizes (n) and integration time windows (t). A GoF of 100% (white) indicates that the model perfectly matches the observed psychometric threshold as well as choice probabilities for both P and M neurons. (A) t = 0–25 ms. (B) t = 0–50 ms. (C) t = 0–75 ms. (D) t = 0–100 ms. (E) t = 0–150 ms. (F) t = 0–200 ms. (F = 0.01, P = 0.91, 2-way ANOVA main effect for pooling strategy), the weight distributions differed dramatically between the two pooling schemes (F = 12073.18, P = 0.00, 2-way ANOVA main effect for pooling strategy). This difference in the weight distributions was presumably due to the fact that, compared with M d-prime distributions, P d-prime distributions were more widely spread with greater maximal values. Thus, when P and M populations were scaled together, as was the case here, both were most likely scaled in reference to the d-primes of a subset of P neurons, thus preserving the shapes as well as spreads of these distributions. Frontiers in Neural Circuits \| www.frontiersin.org d-Prime Weighted Pooling Scheme: Which One to Choose? When P and M populations were scaled separately, as was the case above, M neurons were scaled to a lesser degree when compared with P neurons, rendering the spreads of the two distributions indistinguishable. We also analyzed the simulated choice probabilities for P and M neurons in the same time window (t = 0–150 ms), and found that in this d-prime model the P and M choice probabilities (P choice probability = 0.53 ± 0.00, M choice probability = 0.54 ± 0.00) also resembled their experimentally measured counterparts (Jiang et al., 2015, also see above; Figure 10E). Furthermore, these P and M choice probability distributions developed throughout the 200 ms stimulus presentation time (n = 512 neurons, F = 3.78, October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 14 Jiang et al. Computation of contrast in thalamus FIGURE 9 \| Further analysis of the d-prime weighted pooling scheme: the relationship between d-prime, weight, and choice probability. (A,B) Cumulative d-prime distributions for P (magenta, A) or M (green, B) neurons in different integration time windows (n = 512 neurons). (C) Cumulative d-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). (D) D-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median d-prime; solid line: d-prime = 0. (E) Cumulative weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (F) Weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median weight; solid line: weight = 0. (G) Cumulative choice probability distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (H) Choice probability values were positively correlated with d-primes (n = 512 neurons, t = 0–150 ms, 200 simulations). Horizontal line: choice probability = 0.5; vertical line: d-prime = 0. FIGURE 9 \| Further analysis of the d-prime weighted pooling scheme: the relationship between d-prime, weight, and choice probability. (A,B) Cumulative d-prime distributions for P (magenta, A) or M (green, B) neurons in different integration time windows (n = 512 neurons). d-Prime Weighted Pooling Scheme: Which One to Choose? (C) Cumulative d-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). (D) D-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median d-prime; solid line: d-prime = 0. (E) Cumulative weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (F) Weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median weight; solid line: weight = 0. (G) Cumulative choice probability distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (H) Choice probability values were positively correlated with d-primes (n = 512 neurons, t = 0–150 ms, 200 simulations). Horizontal line: choice probability = 0.5; vertical line: d-prime = 0. October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 15 Jiang et al. Computation of contrast in thalamus FIGURE 10 \| Further analysis of the d-prime weighted pooling scheme: P and M neurons weighted together in reference to one maximal d-prime. (A) Cumulative d-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). (B) D-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median d-prime; solid line: d-prime = 0. (C) Cumulative weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (D) Weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median weight; solid line: weight = 0. (E) Cumulative choice probability distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (F) Cumulative choice probability distributions for P (magenta) and M (green) neurons in different integration time windows (n = 512 neurons). FIGURE 10 \| Further analysis of the d-prime weighted pooling scheme: P and M neurons weighted together in reference to one maximal d-prime. (A) Cumulative d-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). d-Prime Weighted Pooling Scheme: Which One to Choose? (B) D-prime distributions for P (magenta) and M (green) neurons in a 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median d-prime; solid line: d-prime = 0. (C) Cumulative weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (D) Weight distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). Arrow: median weight; solid line: weight = 0. (E) Cumulative choice probability distributions for P (magenta) and M (green) neurons in the same 0–150 ms window (n = 512 neurons, 200 simulations). (F) Cumulative choice probability distributions for P (magenta) and M (green) neurons in different integration time windows (n = 512 neurons). P = 0.00, 2-way ANOVA main effect for time), much like in the other d-prime model (compare Figures 5H to 10F), and confirming the temporal dynamics that we had observed in the LGN of awake monkeys (Jiang et al., 2015). effects of the neural pool size, the Fano factor, the interneuronal correlation and the downstream pooling noise on the simulated psychophysical performance and choice probability values. We investigated different pooling/readout schemes that ranged from basic, uniform pools to more optimal pools that preferably weighted the more sensitive single neurons. We compared these pooling strategies in integration time windows of different durations, and found that the most successful model consisted of a medium number of LGN neurons (n = ∼30–250) in medium to long integration time windows (75–200 ms), with individual neurons weighted differentially according to their d-prime Frontiers in Neural Circuits \| www.frontiersin.org Pooling/Readout Strategies The primate LGN provides major feedforward input to the visual cortex, and is an essential thalamic gateway to conscious vision (Sherman and Guillery, 2001; Jones, 2007; Schmid et al., 2010; Casagrande and Ichida, 2011; Saalmann and Kastner, 2011). The effort to understand how the information carried by LGN cells is utilized at later stages is therefore of great importance. Very generally, pooling /readout rules can be divided into two categories. In the first, perceptual decisions are based on signals provided by one or several of the most sensitive sensory neurons (i.e., lower envelope principle; Barlow, 1995). In the second category, perceptual decisions are based on some form of pooled responses from many sensory neurons. The uniform pooling as well as alternative weighted pooling schemes used in this paper all fall into the second category. p p p Compared with the simple uniform pooling scheme, the d-prime model was superior in several major ways: (1) The d-prime model achieved lower average and minimal psychophysical thresholds (Figures 5A–C), especially in shorter integration time windows; (2) The d-prime model more faithfully reflected the temporal developments of choice probabilities in LGN P and M neurons (Figure 5H); and (3) The d-prime model achieved greater average and maximal model fitness (Figures 6A–C) with fewer neurons (Figure 6D), especially in shorter time windows. Additionally, the d-prime weighted model also demonstrated a clear, direct relationship between choice probability and neural sensitivity (Figure 9H), indicating that neurons with higher signal-to-noise ratios were also more correlated with perceptual choices. This correlation was even more pronounced in shorter integration time windows, where fewer neurons demonstrated high signal-to-noise ratios (e.g., d-prime vs. choice probability, t = 0–25 ms: r for P neurons = 0.22, P = 0.00; r for M neurons = 0.28, P = 0.00. t = 0–50 ms: r for P neurons = 0.18, P = 0.00; r for M neurons = 0.18, P = 0.00). According to previous theoretical work (Haefner et al., 2013; Moreno-Bote et al., 2014), when choice probabilities and neural sensitivities (i.e., d-primes) exhibit such direct correlations, it is an indication that the pooling/readout strategy is optimal for the task. Last but not least, the uniform pooling scheme assumes that even after extensive practice of a perceptual task, the initial pattern of widespread and diffuse synaptic connections will remain unrefined. Pooling/Readout Strategies In reality, however, perceptual learning is known to dramatically alter the properties of single sensory neurons (Sasaki et al., 2010; Kumano and Uka, 2013; Watanabe and Sasaki, 2015). Therefore, neurobiologically speaking, the d-prime weighted model is also the more plausible solution in vivo. The lower envelope principle, however, always remains a theoretical possibility. This is because even in a detection task such as ours, where the sensitivities of most single neurons failed to match the psychophysical sensitivity of the subject, there were still a small but significant proportion (13.5%; Jiang et al., 2015) of single cells that matched or even outperformed the subject. That being said, if the lower envelop principle were true, we would expect a choice probability distribution that is qualitatively different from what was observed in physiological recordings. Briefly, if only a few neurons contribute to a perceptual decision, all of them should demonstrate very significant choice probabilities, with the rest of the entire neural population exhibiting chance choice probabilities (Nienborg et al., 2012; Haefner et al., 2013; but see below for the influence of interneuronal correlation on choice probability). In reality, most cortical recordings have reported a broad distribution of weakly significant choice probabilities (for example, see Britten et al., 1996; Uka and Deangelis, 2004; Liu and Newsome, 2005; Purushothaman and Bradley, 2005; Nienborg and Cumming, 2006; Palmer et al., 2007; Price and Born, 2010; Liu et al., 2013), a result that was confirmed in the LGN (Jiang et al., 2015). If the d-prime weighted pooling strategy is indeed utilized in the neural system, our simulations make several specific predictions that can be tested in future psychophysical and physiological recordings: (1) A single LGN neuron’s d-prime value should be directly correlated with its choice probability (Figure 9H), and this correlation should be stronger in shorter integration time windows (see above); (2) A single LGN neuron’s d-prime and choice probability values should both develop throughout the stimulus presentation time (see Figures 5H, 9A,B, 10F). This prediction was already confirmed in our previous publication (Figures 7C,D in Jiang et al., 2015); and Consequently, our current study as well as a number of other computational studies (Shadlen et al., 1996; Purushothaman and Bradley, 2005; Cohen and Newsome, 2009; Haefner et al., 2013) arrived at the conclusion that an ideal perceptual decision pool consists of not just a few, but rather tens to hundreds of single sensory neurons. Discussion We previously reported that, in a 2AFC contrast detection task, single LGN P and M neurons demonstrated significant choice probabilities despite their relatively poor neural sensitivities (Jiang et al., 2015). In this study, we examined quantitatively the October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 16 Computation of contrast in thalamus Jiang et al. The d-Prime Weighted Pooling Model According to signal detection theory (Green and Swets, 1966), d-prime is one of the most useful and widely used descriptors of signal-to-noise ratio. The d-prime model was one of several selective weighted pooling models that we examined in this paper. In this model the readout weight of each neuron was determined by its d-prime value at high contrast, with the neuron with the greatest d-prime value carrying a weight of 1.0. Our simulations showed that the d-prime weighted model provided a parsimonious and complete account of all of our experimental data including the monkeys’ psychophysical performance and the population distributions of LGN choice probabilities. values. These results indicated that both the psychophysical threshold and the LGN choice probabilities during contrast detection could be fully explained using simple, bottom-up pooling models without assuming significant interneuronal correlations, and that such modeling efforts helped elucidate the complicated relationship between neural sensitivity, readout weight, and choice probability. We now consider the significance of these results in light of previous experimental and theoretical findings. Frontiers in Neural Circuits \| www.frontiersin.org The Limitations of the Pooling Models g First, to appropriately interpret our modeling results, it is important to understand the role of interneuronal correlation in perceptual decision making. In medium-sized decision pools such as ours, interneuronal noise correlations can strongly influence not only the choice probability structure, but also the readout weight distribution (Chen et al., 2006; Haefner et al., 2013). In cortex, interneuronal correlations are considered to be mostly unavoidable (Averbeck et al., 2006; Cohen and Kohn, 2011) because of the extensively shared connections between neurons (Zohary et al., 1994; Shadlen and Newsome, 1998; Bair et al., 2001; Reich et al., 2001; Averbeck et al., 2006; Cohen and Maunsell, 2009). Recent studies, however, reported overall interneuronal correlations not different from chance in chronic recordings from large populations of V1 neurons in awake monkeys (Ecker et al., 2010, 2014). Compared to the visual cortex, neural circuitry in the LGN is simpler (Casagrande and Norton, 1991; Nassi and Callaway, 2009) and highly specific to cell types (Casagrande and Xu, 2004; Briggs and Usrey, 2011; Ichida et al., 2014). Furthermore, cognitive factors such as attention (Cohen and Maunsell, 2009; Mitchell et al., 2009) and perceptual learning (Gu et al., 2011) are known to decrease the existing interneuronal correlations in a perceptual decision task. It is therefore not entirely surprising that we found LGN interneuronal correlations to be not significantly different from 0.0 during a contrast detection task (Jiang et al., 2015). Additionally, our pooling models were abstract representations of the minimal computations required to account for our experimental data. These models did not specify and were not critically dependent on, for example, exactly when and where a perceptual ‘‘choice’’ is made in vivo. Furthermore, even though we were able to simulate neural pools of infinitely large sizes for a large number of trials, the accuracies of these simulations were constrained by the sample sizes in our original experimental data (Jiang et al., 2015). Finally, although we took into consideration the temporal evolution of a variety of critical factors such as the mean and variance of neural response, the d-prime, and the choice probability, we did not characterize how temporal changes in other parameters such as the interneuronal correlation and the downstream pooling noise might influence model performance. Pooling/Readout Strategies In this type of broad decision pool, the readout weight profile, or pooling strategy, of the neural system can be inferred from experimentally measurable quantities such as the behavioral threshold and the choice probability distributions (Haefner et al., 2013; Liu et al., 2013), as demonstrated in the current study. October 2015 \| Volume 9 \| Article 54 17 Computation of contrast in thalamus Jiang et al. interneuronal correlations, we could not rule out the possibility that, in reality, there exist some fine patterns within the LGN interneuronal correlation structure. In fact, recent modeling work has revealed that it is not the average interneuronal correlation level, but the structure of a specific type of differential correlation, that determines choice probability values in a perceptual decision pool (Haefner et al., 2013; Moreno-Bote et al., 2014). Briefly, in cortex, interneuronal correlations are known to be stronger for similarly tuned neurons rather than dissimilarly tuned ones (Zohary et al., 1994; Maynard et al., 1999; Averbeck and Lee, 2003; Gu et al., 2011; Adibi et al., 2013). In this scenario, the neurons at the center of the decision pool could have the largest choice probabilities simply because they are most correlated with all the other neurons in the same pool. In other words, choice probabilities could decrease in the direction of the pool boundaries solely because of the correlation structure, but not the readout weight structure, of the decision pool (Chen et al., 2006; Haefner et al., 2013). This is a possibility that we did not model, and therefore could not rule out for the LGN perceptual decision pool. Furthermore, thalamic interneuronal correlations may be qualitatively different from those measured in the cortex, as LGN neurons sharing the same retinal inputs are known to exhibit very strong temporal correlations in their firing patterns (Alonso et al., 1996; Dan et al., 1998). This is also a possible correlation structure that we did not explore in our models, and as a result we could not rule out its potential influence on the psychophysical sensitivity and choice probabilities of LGN neurons. (3) More importantly, if LGN responses are optimally pooled in subsequent stages, humans and monkeys should be able to maintain the same contrast detection performance with stimulus durations as short as 50–75 ms (see Figures 5A–C), even though LGN choice probabilities may decrease in such short integration time windows (Figures 5H, 10F). The Limitations of the Pooling Models Despite these limitations, we believe that our modeling results clearly and unarguably support the hypothesis that the neural pool consists of not just a few very sensitive neurons but many neurons, likely 100 or more, at the level of the visual thalamus, and that the response fluctuations in these thalamic neurons can influence perception, with the more sensitive neurons exerting a bigger influence on perception. Frontiers in Neural Circuits \| www.frontiersin.org Pooling/Readout Strategies As mentioned above, the d-prime weighted pooling model could be further divided into two types, with one weighing P and M neurons separately according to their respective maximal d-primes, and one weighing P and M neurons together according to one maximal d-prime. These two models were indistinguishable in terms of their overall performance, but they did differ from each other in their readout weight and choice probability distributions. When considered separately, M neurons were significantly more heavily weighted than P neurons (Figure 9F). In contrast, when weighted together, M neurons only had a very slight advantage over P neurons (Figure 10D). Computationally, we could not rule out one model in favor of the other. Neurobiologically, the former scenario is more likely to occur only in layer 4 of V1, where LGN P and M inputs remain segregated (Casagrande and Xu, 2004; Nassi and Callaway, 2009). The latter readout scheme, in contrast, is more likely to occur everywhere else in the cortex, where LGN P and M inputs are mixed and integrated. References Responses to single quanta of light in retinal ganglion cells of the cat. Vision Res. Suppl 3, 87–101. doi: 10. 1016/0042-6989(71)90033-2 Green, D. M., and Swets, J. A. (1966). Signal Detection Theory and Psychophysics. New York, NY: Wiley. Berry, M. J., Warland, D. K., and Meister, M. (1997). The structure and precision of retinal spike trains. Proc. Natl. Acad. Sci. U. S. A. 94, 5411–5416. doi: 10. 1073/pnas.94.10.5411 Grunewald, A., Bradley, D. C., and Andersen, R. A. (2002). Neural correlates of structure-from-motion perception in macaque V1 and MT. J. Neurosci. 22, 6195–6207. Briggs, F., and Usrey, W. M. (2009). Parallel processing in the corticogeniculate pathway of the macaque monkey. Neuron. 62, 135–146. doi: 10.1016/j.neuron. 2009.02.024 Gu, Y., Deangelis, G. C., and Angelaki, D. E. (2007). A functional link between area MSTd and heading perception based on vestibular signals. Nat. Neurosci. 10, 1038–1047. doi: 10.1038/nn1935 Briggs, F., and Usrey, W. M. (2011). Corticogeniculate feedback and visual processing in the primate. J. Physiol. 589, 33–40. doi: 10.1113/jphysiol.2010. 193599 Gu, Y., Liu, S., Fetsch, C. R., Yang, Y., Fok, S., Sunkara, A., et al. (2011). Perceptual learning reduces interneuronal correlations in macaque visual cortex. Neuron. 71, 750–761. doi: 10.1016/j.neuron.2011.06.015 Britten, K. H., Newsome, W. T., Shadlen, M. N., Celebrini, S., and Movshon, J. A. (1996). A relationship between behavioral choice and the visual responses of neurons in macaque MT. Vis. Neurosci. 13, 87–100. doi: 10. 1017/s095252380000715x Haefner, R. M., Gerwinn, S., Macke, J. H., and Bethge, M. (2013). Inferring decoding strategies from choice probabilities in the presence of correlated variability. Nat. Neurosci. 16, 235–242. doi: 10.1038/nn.3309 Britten, K. H., Shadlen, M. N., Newsome, W. T., and Movshon, J. A. (1992). The analysis of visual motion: a comparison of neuronal and psychophysical performance. J. Neurosci. 12, 4745–4765. Ichida, J. M., and Casagrande, V. A. (2002). Organization of the feedback pathway from striate cortex (V1) to the lateral geniculate nucleus (LGN) in the owl monkey ( Aotus trivirgatus). J. Comp. Neurol. 454, 272–283. doi: 10.1002/cne. 10441 Casagrande, V. A., and Ichida, J. M. (2011). ‘‘Processing in the Lateral Geniculate Nucleus (LGN),’’ in Adler’s Physiology of the Eye 11th edition, eds L. A. Levin, S. F. E. Nilsson, J. Ver Hoeve, S. Wu, P. L. Kaufman and A. Alm (Amsterdam: Elsevier), 574–585. Ichida, J. M., Mavity-Hudson, J. A., and Casagrande, V. A. (2014). Distinct patterns of corticogeniculate feedback to different layers of the lateral geniculate nucleus. References Cohen, M. R., and Maunsell, J. H. (2009). Attention improves performance primarily by reducing interneuronal correlations. Nat. Neurosci. 12, 1594–1600. doi: 10.1038/nn.2439 Adibi, M., Mcdonald, J. S., Clifford, C. W., and Arabzadeh, E. (2013). Adaptation improves neural coding efficiency despite increasing correlations in variability. J. Neurosci. 33, 2108–2120. doi: 10.1523/jneurosci.3449- 12.2013 Cohen, M. R., and Newsome, W. T. (2009). Estimates of the contribution of single neurons to perception depend on timescale and noise correlation. J. Neurosci. neurons to perception depend on timescale and noise c 29, 6635–6648. doi: 10.1523/jneurosci.5179-08.2009 neurons to perception depend on timescale and noise co 29, 6635–6648. doi: 10.1523/jneurosci.5179-08.2009 29, 6635–6648. doi: 10.1523/jneurosci.5179-08.2009 Cook, E. P., and Maunsell, J. H. (2002). Dynamics of neuronal responses in macaque MT and VIP during motion detection. Nat. Neurosci. 5, 985–994. doi: 10.1038/nn924 Alonso, J. M., Usrey, W. M., and Reid, R. C. (1996). Precisely correlated firing in cells of the lateral geniculate nucleus. Nature. 383, 815–819. doi: 10. 1038/383815a0 Dan, Y., Alonso, J. M., Usrey, W. M., and Reid, R. C. (1998). Coding of visual information by precisely correlated spikes in the lateral geniculate nucleus. Nat. Neurosci. 1, 501–507. doi: 10.1038/2217 Averbeck, B. B., Latham, P. E., and Pouget, A. (2006). Neural correlations, population coding and computation. Nat. Rev. Neurosci. 7, 358–366. doi: 10. 1038/nrn1888 Averbeck, B. B., and Lee, D. (2003). Neural noise and movement-related codes in the macaque supplementary motor area. J. Neurosci. 23, 7630–7641. Dodd, J. V., Krug, K., Cumming, B. G., and Parker, A. J. (2001). Perceptually bistable three-dimensional figures evoke high choice probabilities in cortical area MT. J. Neurosci. 21, 4809–4821. Bair, W., Zohary, E., and Newsome, W. T. (2001). Correlated firing in macaque visual area MT: time scales and relationship to behavior. J. Neurosci. 21, 1676–1697. Ecker, A. S., Berens, P., Cotton, R. J., Subramaniyan, M., Denfield, G. H., Cadwell, C. R., et al. (2014). State dependence of noise correlations in macaque primary visual cortex. Neuron. 82, 235–248. doi: 10.1016/j.neuron.2014. 02.006 Barlow, H. B. (1995). ‘‘The neuron doctrine in perception,’’ in The Cognitive Neurosciences, ed. M. S. Gazzaniga (Cambridge, MA: The MIT Press), 415–435. Ecker, A. S., Berens, P., Keliris, G. A., Bethge, M., Logothetis, N. K., and Tolias, A. S. (2010). Decorrelated neuronal firing in cortical microcircuits. Science. 327, 584–587. doi: 10.1126/science.1179867 Barlow, H. B., Levick, W. R., and Yoon, M. (1971). Funding This work was supported by National Institutes of Health from grants EY001778 (VAC), EY025422 (VAC), R21 EY019132 (VAC), core grants EY008126 and HD15052, and funds from the Department of Cell and Developmental Biology at Vanderbilt University. This work was supported by National Institutes of Health from grants EY001778 (VAC), EY025422 (VAC), R21 EY019132 (VAC), core grants EY008126 and HD15052, and funds from the Department of Cell and Developmental Biology at Vanderbilt University. Author Contributions YJ, GP and VC conceptualized and designed the study, YJ collected and analyzed the data, YJ and VC interpreted the results, YJ drafted the manuscript, YJ and VC revised the manuscript, YJ, GP and VC approved the final version of the Even though we were able to successfully model experimentally measured psychophysical performance and choice probabilities without assuming any significant October 2015 \| Volume 9 \| Article 54 18 Computation of contrast in thalamus Jiang et al. Acknowledgments manuscript and agreed to be accountable for all aspects of the work. We are grateful to our veterinarian Troy Apple, veterinary technician Mary Feurtado, and the staff of Division of Animal Care at Vanderbilt University for providing excellent support and care for our animals. We appreciate Dr. Jeffrey Schall’s advice on experimental design and his willingness to share lab equipment and programs. We thank Julia Mavity- Hudson for her continuous help with many aspects of this project. References Eye Brain 6(Suppl. 1), 57–73. doi: 10.2147/EB. S64281 Casagrande, V. A., and Norton, T. T. (1991). ‘‘The lateral geniculate nucleus: A review of its physiology and function,’’ in The Neural Basis of Visual Function, ed. A. G. Leventhal (London, UK: MacMillan Press), 41–84. doi: 10. 1007/springerreference_184100 Jiang, Y., Yampolsky, D., Purushothaman, G., and Casagrande, V. A. (2012). Neural response in the lateral geniculate nucleus (LGN) of awake, behaving monkeys during a contrast detection task: comparison of psychometric and neurometric functions. Abstract retrieved from Society for Neuroscience database (Accession No. 175.115/CC112). Casagrande, V. A., and Xu, X. (2004). ‘‘Parallel visual pathways: a comparative perspective,’’ in The Visual Neurosciences, eds L. Chalupa and J. S. Werner (Cambridge, MA: The MIT Press), 494–506. Jiang, Y., Yampolsky, D., Purushothaman, G., and Casagrande, V. A. (2013). Neural sensitivity in the lateral geniculate nucleus (LGN) of awake, behaving monkeys during a contrast detection task: comparison of neurometric and psychometric functions. J. Vis. 13, 1026. Abstract retrieved from Vision Sciences Society. doi: 10.1167/13.9.1026 Chen, Y., Geisler, W. S., and Seidemann, E. (2006). Optimal decoding of correlated neural population responses in the primate visual cortex. Nat. Neurosci. 9, 1412–1420. doi: 10.1038/nn1792 Cohen, M. R., and Kohn, A. (2011). Measuring and interpreting neuronal correlations. Nat. Neurosci. 14, 811–819. doi: 10.1038/nn.2842 October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 19 Computation of contrast in thalamus Jiang et al. Purushothaman, G., and Bradley, D. C. (2005). Neural population code for fine perceptual decisions in area MT. Nat. Neurosci. 8, 99–106. doi: 10.1038/ nn1373 Jiang, Y., Yampolsky, D., Purushothaman, G., and Casagrande, V. A. (2015). Perceptual Decision Related Activity in the Lateral Geniculate Nucleus (LGN). J. Neurophysiol. 114, 717–735. doi: 10.1152/jn.00068.2015 Reich, D. S., Mechler, F., and Victor, J. D. (2001). Independent and redundant information in nearby cortical neurons. Science. 294, 2566–2568. doi: 10. 1126/science.1065839 Jones, E. G. (2007). The Thalamus. Cambridge, UK: Cambridge University Press. Kara, P., Reinagel, P., and Reid, R. C. (2000). Low response variability in simultaneously recorded retinal, thalamic and cortical neurons. Neuron. 27, 635–646. doi: 10.1016/s0896-6273(00)00072-6 Reich, D. S., Victor, J. D., Knight, B. W., Ozaki, T., and Kaplan, E. (1997). Response variability and timing precision of neuronal spike trains in vivo. J. Neurophysiol. 77, 2836–2841. Kumano, H., and Uka, T. (2013). Neuronal mechanisms of visual perceptual learning. Behav. Brain Res. 249, 75–80. doi: 10.1016/j.bbr.2013.04.034 Levine, M. W., Cleland, B. References G., and Zimmerman, R. P. (1992). Variability of responses of cat retinal ganglion cells. Vis. Neurosci. 8, 277–279. doi: 10. 1017/s0952523800002923 Royal, D. W., Sary, G., Schall, J. D., and Casagrande, V. A. (2006). Correlates of motor planning and postsaccadic fixation in the macaque monkey lateral geniculate nucleus. Exp. Brain Res. 168, 62–75. doi: 10.1007/s00221-005- 0093-z Liu, J., and Newsome, W. T. (2005). Correlation between speed perception and neural activity in the middle temporal visual area. J. Neurosci. 25, 711–722. doi: 10.1523/jneurosci.4034-04.2005 Saalmann, Y. B., and Kastner, S. (2011). Cognitive and perceptual functions of the visual thalamus. Neuron. 71, 209–223. doi: 10.1016/j.neuron.2011. 06.027 Liu, S., Gu, Y., Deangelis, G. C., and Angelaki, D. E. (2013). Choice-related activity and correlated noise in subcortical vestibular neurons. Nat. Neurosci. 16, 89–97. doi: 10.1038/nn.3267 Sasaki, Y., Nanez, J. E., and Watanabe, T. (2010). Advances in visual perceptual learning and plasticity. Nat. Rev. Neurosci. 11, 53–60. doi: 10.1038/ nrn2737 Maynard, E., Hatsopoulos, N., Ojakangas, C., Acuna, B., Sanes, J., Normann, R., et al. (1999). Neuronal interactions improve cortical population coding of movement direction. J. Neurosci. 19, 8083–8093. Schmid, M. C., Mrowka, S. W., Turchi, J., Saunders, R. C., Wilke, M., Peters, A. J., et al. (2010). Blindsight depends on the lateral geniculate nucleus. Nature. 466, 373–377. doi: 10.3410/f.4829956.5455055 McAdams, C. J., and Maunsell, J. H. (1999). Effects of attention on the reliability of individual neurons in monkey visual cortex. Neuron. 23, 765–773. doi: 10. 1016/s0896-6273(01)80034-9 Shadlen, M. N., Britten, K. H., Newsome, W. T., and Movshon, J. A. (1996). A computational analysis of the relationship between neuronal and behavioral responses to visual motion. J. Neurosci. 16, 1486–1510. Mitchell, J. F., Sundberg, K. A., and Reynolds, J. H. (2009). Spatial attention decorrelates intrinsic activity fluctuations in macaque area V4. Neuron. 63, 879–888. doi: 10.1016/j.neuron.2009.09.013 Shadlen, M. N., and Newsome, W. T. (1998). The variable discharge of cortical neurons: implications for connectivity, computation and information coding. J. Neurosci. 18, 3870–3896. Moreno-Bote, R., Beck, J., Kanitscheider, I., Pitkow, X., Latham, P., and Pouget, A. (2014). Information-limiting correlations. Nat. Neurosci. 17, 1410–1417. doi: 10.1038/nn.3807 Sherman, S. M., and Guillery, R. W. (2001). Exploring the thalamus. San Diego, CA: Academic. Nassi, J. J., and Callaway, E. M. (2009). Parallel processing strategies of the primate visual system. Nat. Rev. Neurosci. 10, 360–372. doi: 10. 1038/nrn2619 Tolhurst, D. J., Movshon, J., and Dean, A. (1983). References The statistical reliability of signals in single neurons in cat and monkey visual cortex. Vision Res. 23, 775–785. doi: 10.1016/0042-6989(83)90200-6 Nienborg, H., Cohen, M. R., and Cumming, B. G. (2012). Decision-related activity in sensory neurons: correlations among neurons and with behavior. Annu. Rev. Neurosci. 35, 463–483. doi: 10.1146/annurev-neuro-062111-150403 Uka, T., and Deangelis, G. C. (2004). Contribution of area MT to stereoscopic depth perception: choice-related response modulations reflect task strategy. Neuron. 42, 297–310. doi: 10.1016/S0896-6273(04)00186-2 Nienborg, H., and Cumming, B. (2010). Correlations between the activity of sensory neurons and behavior: how much do they tell us about a neuron’s causality? Curr. Opin. Neurobiol. 20, 376–381. doi: 10.1016/j.conb.2010.05.002 Uka, T., Tanabe, S., Watanabe, M., and Fujita, I. (2005). Neural correlates of fine depth discrimination in monkey inferior temporal cortex. J. Neurosci. 25, 10796–10802. doi: 10.1523/jneurosci.1637-05.2005 Nienborg, H., and Cumming, B. G. (2006). Macaque V2 neurons, but not V1 neurons, show choice-related activity. J. Neurosci. 26, 9567–9578. doi: 10. 1523/jneurosci.2256-06.2006 Watanabe, T., and Sasaki, Y. (2015). Perceptual learning: toward a comprehensive theory. Annu. Rev. Psychol. 66, 197–221. doi: 10.1146/annurev-psych-010814- 015214 Norton, T. T., and Casagrande, V. A. (1982). Laminar organization of receptive-field properties in lateral geniculate nucleus of bush baby ( Galago crassicaudatus). J. Neurophysiol. 47, 715–741. Xu, X., Ichida, J. M., Allison, J. D., Boyd, J. D., Bonds, A. B., and Casagrande, V. A. (2001). A comparison of koniocellular, magnocellular and parvocellular receptive field properties in the lateral geniculate nucleus of the owl monkey ( Aotus trivirgatus). J. Physiol. 531, 203–218. doi: 10.1111/j.1469-7793.2001. 0203j.x Xu, X., Ichida, J. M., Allison, J. D., Boyd, J. D., Bonds, A. B., and Casagrande, V. A. (2001). A comparison of koniocellular, magnocellular and parvocellular receptive field properties in the lateral geniculate nucleus of the owl monkey ( A ) h l d / Norton, T. T., Casagrande, V. A., Irvin, G. E., Sesma, M. A., and Petry, H. M. (1988). Contrast-sensitivity functions of W-, X- and Y-like relay cells in the lateral geniculate nucleus of bush baby, Galago crassicaudatus. J. Neurophysiol. 59, 1639–1656. ( Aotus trivirgatus). J. Physiol. 531, 203–218. doi: 10.1111/j.1469-7793.2001. 0203j.x Zohary, E., Shadlen, M. N., and Newsome, W. T. (1994). Correlated neuronal discharge rate and its implications for psychophysical performance. Nature. 370, 140–143. doi: 10.1038/371358c0 Oram, M., Wiener, M., Lestienne, R., and Richmond, B. (1999). Stochastic nature of precisely timed spike patterns in visual system neuronal responses. J. Neurophysiol. 81, 3021–3033. References Conﬂict of Interest Statement: 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. Conﬂict of Interest Statement: 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. Palmer, C., Cheng, S. Y., and Seidemann, E. (2007). Linking neuronal and behavioral performance in a reaction-time visual detection task. J. Neurosci. 27, 8122–8137. doi: 10.1523/jneurosci.1940-07.2007 Parker, A. J., and Newsome, W. T. (1998). Sense and the single neuron: probing the physiology of perception. Annu. Rev. Neurosci. 21, 227–277. doi: 10. 1146/annurev.neuro.21.1.227 Copyright © 2015 Jiang, Purushothaman and Casagrande. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor 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. Copyright © 2015 Jiang, Purushothaman and Casagrande. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor 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. Price, N. S., and Born, R. T. (2010). Timescales of sensory-and decision-related activity in the middle temporal and medial superior temporal areas. J. Neurosci. 30, 14036–14045. doi: 10.1523/jneurosci.2336-10.2010 October 2015 \| Volume 9 \| Article 54 Frontiers in Neural Circuits \| www.frontiersin.org 20
https://openalex.org/W2991430030	https://hal-pasteur.archives-ouvertes.fr/pasteur-02387745/document	English	null	Structure and function of minor pilins of type IV pili	Medical microbiology and immunology	2,019	cc-by	6,452	To cite this version: Theis Jacobsen, Benjamin Bardiaux, Olivera Francetic, Nadia Izadi-Pruneyre, Michael Nilges. Struc- ture and function of minor pilins of type IV pili. Medical Microbiology and Immunology, 2019, 10.1007/s00430-019-00642-5. pasteur-02387745 Structure and function of minor pilins of type IV pili Theis Jacobsen, Benjamin Bardiaux, Olivera Francetic, Nadia Izadi-Pruneyre, Michael Nilges Structure and function of minor pilins of type IV pili Theis Jacobsen, Benjamin Bardiaux, Olivera Francetic, Nadia Izadi-Pruneyre, Michael Nilges Distributed under a Creative Commons Attribution 4.0 International License Abstract Type IV pili are versatile and highly flexible fibers formed on the surface of many Gram-negative and Gram-positive bacteria. Virulence and infection rate of several pathogenic bacteria, such as Neisseria meningitidis and Pseudomonas aeruginosa, are strongly dependent on the presence of pili as they facilitate the adhesion of the bacteria to the host cell. Disruption of the interactions between the pili and the host cells by targeting proteins involved in this interaction could, therefore, be a treat- ment strategy. A type IV pilus is primarily composed of multiple copies of protein subunits called major pilins. Additional proteins, called minor pilins, are present in lower abundance, but are essential for the assembly of the pilus or for its specific functions. One class of minor pilins is required to initiate the formation of pili, and may form a complex similar to that identified in the related type II secretion system. Other, species-specific minor pilins in the type IV pilus system have been shown to promote additional functions such as DNA binding, aggregation and adherence. Here, we will review the structure and the function of the minor pilins from type IV pili. Keywords Type IV pili · Minor pilins · Adhesion · Type II secretion system Abbreviations T2SS Type II secretion system T4P Type IV pilus T4aP Type IVa pilus T4bP Type IVb pilus Abbreviations T2SS Type II secretion system T4P Type IV pilus T4aP Type IVa pilus T4bP Type IVb pilus structure and assembly mechanism [3]. Both Gram-negative and Gram-positive bacteria can form pili, though the diver- sity of pili formed in Gram-negative bacteria is larger [3, 4]. In this review, we will focus on the type IV pili (T4P), which promote biological functions important for the patho- genicity of bacteria such as twitching motility, DNA uptake and adhesion to host cells [1]. The T4P family is generally subdivided into two groups, type IVa pili (T4aP) and type IVb pili (T4bP). This subdivision is based on the sequence of the pilins, the length of the leader peptide, and minor differences in their assembly [1, 5]. The organization of the genes required for pilus formation are different between T4aP and T4bP systems. T4aP genes are spread throughout the genome in several operons, whereas the T4bP genes are clustered in a single operon [2]. T4bP are involved in biofilm formation, bacterial colonization and cell adhesion [6]. Abstract T4aP have been identified as the molecular framework behind bac- terial twitching motility and other functions such as DNA uptake [7]. Their flexibility, elongation and retraction allow them to ensure these different functions. Theis Jacobsen1,2 · Benjamin Bardiaux1 · Olivera Francetic3 · Nadia Izadi‑Pruneyre1 · Michael Nilges1 Received: 1 July 2019 / Accepted: 14 November 2019 © The Author(s) 2019 * Michael Nilges michael.nilges@pasteur.fr Edited by Volkhard A.J. Kempf. HAL Id: pasteur-02387745 https://pasteur.hal.science/pasteur-02387745v1 Submitted on 30 Nov 2019 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. Distributed under a Creative Commons Attribution 4.0 International License Medical Microbiology and Immunology https://doi.org/10.1007/s00430-019-00642-5 REVIEW 3 Biochemistry of Macromolecular Interactions Unit, Department of Structural Biology and Chemistry, Institut Pasteur, CNRS UMR3528, Paris, France 1 Structural Bioinformatics Unit, Department of Structural Biology and Chemistry, C3BI, Institut Pasteur, CNRS UMR3528, CNRS USR3756, Paris, France REVIEW REVIEW Introduction Motility and pathogenicity of various bacteria are strongly dependent on proteinaceous appendages called pili, located at the cell surface [1, 2]. Pili are highly flexible, long fib- ers assembled in the envelope of bacteria. Several different types of pili have been identified and grouped based on their Edited by Volkhard A.J. Kempf. Edited by Volkhard A.J. Kempf. 1 Structural Bioinformatics Unit, Department of Structural Biology and Chemistry, C3BI, Institut Pasteur, CNRS UMR3528, CNRS USR3756, Paris, France f This review will highlight recent advances in the under- standing of the assembly of T4P as well as in the biological functions and the structure of their minor pilins. Although the T4P assembly system is conserved between bacterial species, the nomenclature of T4P proteins is very heteroge- neous. In this review, unless stated otherwise, we will use 2 Sorbonne Université, Complexité du Vivant, 75005 Paris, France 3 Biochemistry of Macromolecular Interactions Unit, Department of Structural Biology and Chemistry, Institut Pasteur, CNRS UMR3528, Paris, France :(0123 1 3456789) 3 Medical Microbiology and Immunology the nomenclature from Pseudomonas aeruginosa, one of the best-characterized T4P systems (Table 1). the minor pilins, have been shown to be essential for pilus formation and function [8]. Minor pilins can be further cat- egorized into core and non-core minor pilins [9]. Core minor pilins are important for the formation of pili, despite their low abundance. Building blocks of a pilus: the pilins Pilins, both major and minor, are produced as prepilins with a short N-terminal leader sequence (signal peptide), which serves as an anchor to ensure correct insertion of the pilins in the cytoplasmic membrane [10]. The SecYEG translocon promotes prepilin membrane insertion in such a way that the N-terminal peptide is cytosolically exposed and The major building blocks of a T4P are proteins called pilins, which are composed of a periplasmic domain and a transmembrane helix [3] (Fig. 1a). The bulk of a pilus is made up of thousands of copies of one subunit called the major pilin (Fig. 1b). Other pilins with a lower abundance, Table 1 P. aeruginosa T4aP minor pilins and their homologs in both T4aP and T2SS from different organisms T4a minor pilins T2SS minor pseudopilins Pseudomonas Escherichia Neisseria Pseudomonas Escherichia Vibrio FimU ppdA PilH GspH XcpH EspH PilV ppdC PilI GspI XcpI EspI PilW ppdB PilJ GspJ XcpJ EspJ PilX ygdB PilK GspK XcpK EspK PilE N/A PilX/L PilV N/A N/A N/A PilY1 N/A PilC2 N/A N/A N/A Fig. 1 A major pilin and the type IVa pilus. a Molecular represen- tation of the P. aeruginosa T4aP major pilin PilA (PDB:1OQW). The central α-helix is colored in red, the transmembrane N-termi- nal part (α1-N) and C-terminal (α1-C) part of the folded domain are indicated. The αβ-loop and D-region are highlighted in green and magenta, respectively. b The surface representation of the pilus assembled by subunits of PilA colored in gray and the αβ-loop and D-region are colored green and magenta, respectively (PDB: 5VXY). The zoom-in highlights the surface exposed of one PilA subunit in the pilus and the interactions between neighboring pilins and magenta, respectively. b The surface representation of the pilus assembled by subunits of PilA colored in gray and the αβ-loop and D-region are colored green and magenta, respectively (PDB: 5VXY). The zoom-in highlights the surface exposed of one PilA subunit in the pilus and the interactions between neighboring pilins Fig. 1 A major pilin and the type IVa pilus. a Molecular represen- tation of the P. aeruginosa T4aP major pilin PilA (PDB:1OQW). The central α-helix is colored in red, the transmembrane N-termi- nal part (α1-N) and C-terminal (α1-C) part of the folded domain are indicated. The αβ-loop and D-region are highlighted in green Fig. 1 A major pilin and the type IVa pilus. a Molecular represen- tation of the P. Building blocks of a pilus: the pilins aeruginosa T4aP major pilin PilA (PDB:1OQW). The central α-helix is colored in red, the transmembrane N-termi- nal part (α1-N) and C-terminal (α1-C) part of the folded domain are indicated. The αβ-loop and D-region are highlighted in green 1 3 3 3 3 Medical Microbiology and Immunology the C-terminal domain of the pilins is exported to the peri- plasm (Fig. 1a) [12, 13]. The leader peptide is cleaved off by a specific prepilin peptidase to obtain the mature pilins that are competent for assembly [11]. solvent as assessed by hydrogen–deuterium exchange in a T2SS pseudopilus [18] and a T4b pilus [19]. This conserved structural feature may enable the pilus to undergo spring- like extensions in response to shear forces [17]. Nothing is known about the structure of this region in minor pilins, since structural studies of individual minor pilins were lim- ited to the soluble globular domains, and no structure of a complete pilus with a minor pilin is known. The structural features of T4a pilins, both major and minor, are very similar (see Figs. 1a, 2a, b). They all adopt a lollipop-like shape with an N-terminal helix (α1), fol- lowed by a globular domain [9]. The α1 helix can further be separated into two segments, a hydrophobic N-terminal transmembrane (α1-N) and a C-terminal (α1-C) parts, which extends into the globular domain. Before assembly into the pilus, free pilins are anchored in the inner membrane by the transmembrane part (α1-N). At the C-terminal end of the transmembrane domain, the sequence contains a stretch of conserved residues usually limited by a glycine and a pro- line. In all known structures of T4P and the highly homolo- gous type II secretion system (T2SS) major pilin subunits in the context of the pilus, this stretch disrupts the helix and forms an extended structure when incorporated into the fiber [14–17]. In the pilus, this region is also accessible to p p p The central axis of the globular domain is formed by the α1-C helix, which is surrounded by one or two antiparallel β-sheets composed of 4–7 strands, slightly rotated compared to the α-helical axis (Figs. 1a, 2a, b). Major structural differences between different minor and major pilins are restricted to the characteristic hypervari- able regions: the αβ-loop connecting the α1-C helix to the β-sheet and the D-region enclosed by the C-terminal disulfide bridge found in some species. The αβ-loop is structurally very versatile and can contain various struc- tural motifs. The ɑβ-loop of FimU includes distorted β-strands, whereas the D-region contains two antiparallel Fig. 2 Comparison of minor pilins of T4aP from P. 3 aeruginosa (or equivalent); the α1-helix (red), αβ-loop (green) and D-region (magenta) are highlighted. Disulphide bonds are represented as sticks. a Structures of core minor pilins of P. aeruginosa or equivalent pilins minor from T. thermophilus. We include these structures although the assignment as core minor pilins is indirect, by homology to GspI and GspJ. b Structures of non-core minor pilins of P. aeruginosa or equivalent minor pilins from N. meningitis. c Complex of three core minor pseudopilins of enterotoxigenic E. coli, GspI-J-K. GspI is colored in blue, GspJ is colored in green, the pilin domain of GspK is colored in brown and the inserted domain of GspK is colored in gray. An area covered by this inserted domain, which would form a solvent accessible cavity otherwise, is highlighted. Figure was prepared with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC Fig. 2 Comparison of minor pilins of T4aP from P. aeruginosa (or equivalent); the α1-helix (red), αβ-loop (green) and D-region (magenta) are highlighted. Disulphide bonds are represented as sticks. a Structures of core minor pilins of P. aeruginosa or equivalent pilins minor from T. thermophilus. We include these structures although the assignment as core minor pilins is indirect, by homology to GspI and GspJ. b Structures of non-core minor pilins of P. aeruginosa or equivalent minor pilins from N. meningitis. c Complex of three core Fig. 2 Comparison of minor pilins of T4aP from P. aeruginosa (or equivalent); the α1-helix (red), αβ-loop (green) and D-region (magenta) are highlighted. Disulphide bonds are represented as sticks. a Structures of core minor pilins of P. aeruginosa or equivalent pilins minor from T. thermophilus. We include these structures although the assignment as core minor pilins is indirect, by homology to GspI and GspJ. b Structures of non-core minor pilins of P. aeruginosa or equivalent minor pilins from N. meningitis. c Complex of three core minor pseudopilins of enterotoxigenic E. coli, GspI-J-K. GspI is colored in blue, GspJ is colored in green, the pilin domain of GspK is colored in brown and the inserted domain of GspK is colored in gray. An area covered by this inserted domain, which would form a solvent accessible cavity otherwise, is highlighted. Figure was prepared with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC minor pseudopilins of enterotoxigenic E. coli, GspI-J-K. 3 GspI is colored in blue, GspJ is colored in green, the pilin domain of GspK is colored in brown and the inserted domain of GspK is colored in gray. An area covered by this inserted domain, which would form a solvent accessible cavity otherwise, is highlighted. Figure was prepared with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC 1 3 Medical Microbiology and Immunology Assembly, elongation and retraction of pili β-strands (Fig. 2a) [20]. Remarkably, the equivalents of PilV and PilW from Thermus thermophilus, Tt1218 and Tt1219, respectively, do not have a D-region, as the mostly conserved C-terminal disulfide bridge is absent in both proteins [21]. The C-terminal β-sheet has the same elongated shape in both pilins, the major difference is restricted to the αβ-loop. Tt1218 has a very short ɑβ-loop composed of only a short α-helix and a single β-strand. The ɑβ-loop of Tt1219 is much longer and contains a 5-stranded antiparallel β-sheet and two disulfide bridges. The T4P assembly system is responsible for the elongation and the retraction of pili. In Gram-negative bacteria, the T4P assembly system is composed of between 10 and 18 dif- ferent proteins forming complexes that are located in the inner membrane, in the periplasm and in the outer membrane [2]. Recent studies of the T4aP from Myxococcus xanthus [22] and T. thermophilus [23] by electron cryo-tomography have revealed the in situ organization of key elements in this assembly system that are summarized below (Fig. 3a) [22]. 1 3 p β g g y assembly system that are summarized below (Fig. 3a) [22]. Fig. 3 Type IVa pilus (T4aP) assembly. a The T4aP assembly machinery from M. xanthus (PDB:3JC8). On the left the outer membrane (OM), the periplasmic space (PS) and the inner membrane (IM) are marked. The individual proteins involved in the formation of T4P are shown and labeled. On the right the spanning of the different sub-complexes (pilus, OM pore, alignment complex, pilus assembly platform and motor) are indicated, see text for detailed description. b Schematic overview of the pilus initiation and elongation of tip complex. The orientations of the minor pilins are shown as hypothesized Fig. 3 Type IVa pilus (T4aP) assembly. a The T4aP assembly machinery from M. xanthus (PDB:3JC8). On the left the outer membrane (OM), the periplasmic space (PS) and the inner membrane (IM) are marked. The individual proteins involved in the formation of T4P are shown and labeled. On the right the spanning of the different sub-complexes (pilus, OM pore, alignment complex, pilus assembly platform and motor) are indicated, see text for detailed description. b Schematic overview of the pilus initiation and elongation of tip complex. The orientations of the minor pilins are shown as hypothesized 1 3 Medical Microbiology and Immunology Elongation and retraction of the pilus are made possi- ble by the addition or removal, respectively, of major pilin subunits (PilA) at the base of the pilus, and are facilitated by the central inner membrane platform protein, PilC and two antagonistic ATPase motors, PilB and PilT (Fig. 3) [24, 25]. PilC is in direct contact with the pilins and with the ATPases, which presumably rotate PilC to either elongate or retract the pilus [26, 27]. PilB and PilT occupy the center of the PilM cytoplasmic ring positioned near the inner mem- brane [28]. The PilM cytoplasmic ring interacts with the N-terminal transmembrane tail of PilN [28]. PilN forms a heterodimer with PilO, which assembles as a ring structure in the periplasm with the same stoichiometry as the cyto- plasmic ring. The inner membrane protein PilP binds to this complex and connects it to the periplasmic domain of PilQ. The large PilQ multimer forms the secretin pore in the outer membrane necessary for the passage of the pilus through the membrane [29, 30]. Thus, the PilMNOP subcomplex links the cytoplasmic ATPases to the secretin pore [22, 31–33] and ensures that the motors of the assembly platform, PilC and PilB/PilT, are aligned with the secretin [34–36]. The fibers built by inserting pilins into the pilus base extend through the secretin pore out of the cell to perform their biological functions. required for fully efficient piliation, by analogy to the role of minor pseudopilins in the T2SS, suggesting that they are core minor pilins. A detailed study of the role of individual minor pilins in piliation and twitching motility demonstrated the role of the core minor pilins in assembly initiation, and showed that in the pilT mutant background, the residual pili- ation is dependent on minor T2SS pseudopilins [8, 44]. This confirms that T4a and T2SS minor pilins are functionally interchangeable, as had been shown before [45]. The homology between the four core minor pilins in T4P and T2SS led to the suggestion that they interact and form a complex responsible for the initiation of the pilus assembly in similar ways in both systems (Fig. 2c). The X-ray crystal structure of a complex of three minor pseudopilins, GspI, GspJ and GspK, from the T2SS of enterotoxigenic E. coli (ETEC) [46] serves as a structural template. In this complex, the tip subunit GspK acts as a cap, strongly suggesting that the core minor pilins are assembled first to form the tip of the nascent pseudopilus [46]. The minor pseudopilins form a staggered complex in the membrane, promoting initiation and setting the register for helical pseudopilus assembly [39]. This minor pilin complex would stabilize the tip of the pilus and provide a template for assembly of major subunits (Fig. 3b). In analogy to the T2SS, some of the minor pilins should differ both in sequence and structure from the major pilin to form a stable complex, which can remain anchored at the pilus tip. The GspI-J-K complex is composed of pilins of different molecular weight and three-dimensional shape (Fig. 2c), the central GspI being the smallest and GspK the largest, more than twice the size of the major pilin GspG. GspK has a large insertion in the globular domain, which might ensure that the complex is only incorporated at the tip, and its increased size can be imagined to shield the cavity that is formed by pilins of similar size and structure at the tip of a pilus (Fig. 2c). Unfortunately, there is no structure available for the GspK equivalent PilX from P. aeruginosa, nor for an equivalent T4P pilin from other species. The T4P machinery is highly homologous to the type II secretion system (T2SS), in both assembly mechanism and pilus structure [37]. The sequence and structural organiza- tion of pilins of the two systems are conserved, and a T4P can be efficiently assembled by the T2SS [38]. The pilus, or the pseudopilus in T2SS, allows specific folded proteins in the periplasm to be secreted to the external medium. In T2SS, the minor pseudopilins promote assembly initiation [39] and have been implicated in substrate binding [40]. The absence of a retraction ATPase in the T2SS assembly plat- form is one of the major differences with the T4P system. The importance of calcium for the stability of the pseudopili from Klebsiella oxytoca suggests its implication in length control and passive disassembly of the pilus [14]. Although the sequence similarity makes it likely that a similar tip complex exists also in the T4P system, no struc- tural information of this type of complex has been obtained to this date in the T4aP system. Remarkably, the large inser- tion present in GspK is absent in the T4P equivalent PilX from P. aeruginosa, and one has to postulate a different mechanism to cap the pilus. For example, in the T4bP from ETEC a single minor subunit CofB promotes initiation of assembly. Structural studies revealed that this subunit con- tains four distinct domains corresponding those of T4aP core minor pilins [47]. An interesting model based on another crystal structure proposes that CofB forms a homo-trimer that caps and primes pilus assembly [48]. Its homologue TcpB, the unique minor pilin in the Vibrio cholerae T4bP has been implicated in initiation of pilus assembly and Concluding remarks With the emergence of antibiotic resistance, T4P are con- sidered as a target for innovative antibacterial therapeutics [61]. T4aP of EHEC, among others, have been shown to be involved in biofilm formation, twitching motility and adher- ence [60]. The core minor pilins that likely form the tip of the pilus might play multiple roles, being at the same time responsible for the initiation of pilus formation, capping the pilus, and mediate specific biological functions. However, it is at the moment not clear if only the tip complex or also the major pilins can interact with surfaces or host cells. Further work is thus required in order to resolve the specific struc- tural and biological mechanism of T4aP. Detailed structural knowledge of the minor pilins in the context of the entire fiber would be a major asset for the development of new vaccinal and therapeutic strategies. Non-core minor pilins are involved in the different bio- logical functions that are promoted by T4P, including aggre- gation, adhesion and natural competence—the acquisition of external DNA (either inter or intra species) [7]. In Neis- seria species, the minor pilin PilX promotes aggregation via pilus–pilus interactions [52], whereas PilV promotes adherence and signaling via receptors on eukaryotic cell surface [53]. In some species, additional proteins affect T4P- mediated adhesion. For example, PilY1 from P. aeruginosa is an adhesin which interacts with the complex formed by core minor pilins at the tip of the pilus [20]. PilY1 could also facilitate the adhesion to host epithelial cells. In addition, PilY1 and other minor pilins are involved in the regulation of P. aeruginosa virulence genes [54]. In EHEC no pilin- associated adhesins have been identified, but the expression of T4aP genes promotes twitching motility [55]. Acknowledgements We are grateful to the Institut Pasteur and the CNRS for their continued support for our research. This work was funded by the French Agence Nationale de la Recherche (ANR- 14-CE09-0004 to OF), the INCEPTION project (PIA/ANR-16- CONV-0005), the FRM (Equipe FRM 2017 M.DEQ 20170839114 to MN) and the European Union’s Horizon 2020 Research and Innova- tion Programme under the Marie Sklodowska-Curie Grant agreement 765042. The minor pilin of N. meningitidis ComP binds DNA, recognizing a specific DNA sequence called the uptake sig- nal [56] to promote natural transformation [57]. The DNA uptake mechanism mediated by T4P is not completely understood. Structure and function of core minor pilins In the T4aP system, there are typically four core minor pilin genes that are clustered in an operon, which are homolo- gous to the genes encoding four minor pseudopilins in the T2SS (Table 1). Deletion of these genes results in strong defects in pilus assembly and function, as demonstrated in P. aeruginosa [20] and enterohemorrhagic Escherichia coli (EHEC) [41]. In N. meninigitidis, deletion of each of the minor pilin genes—pilH, I, J and K—leads to a non-piliated phenotype [42]. Mutation in the ATPase pilT restores pilia- tion, which led to the suggestion [43] that the role of these minor pilins is “anti-retraction”. However, each of them is 1 3 Medical Microbiology and Immunology disassembly [49]. Furthermore, recent biochemical evidence suggest that it forms a trimer localized at the pilus tip [50]. not involved in pilus assembly, this minor pilin is likely located at the tip of the pilus, since the structural organi- zation and the large size of this additional domain cannot allow upward incorporation of major pilins. Interestingly, this domain of ComZ was shown to be involved in DNA binding. Structure and function of non‑core minor pilins The non-core minor pilins have a more conserved fold and the diversity in the αβ-loop, and the D-region is more lim- ited compared to the core minor pilins (see Fig. 2b). The αβ-loops of PilE [51] and the equivalents from N. menin- gitidis PilX [52] and PilV, are shorter and contain a small α-helical segment. The D-region in PilE and PilV from N. meningitidis both contain two β-strands, which are part of the conserved C-terminal β-sheet. Furthermore, PilE and PilX from N. meningitidis have a short helical segment in the D-region.f Concluding remarks However, it has been proposed that T4P play a role in the early step of DNA uptake [7]. A study of compe- tence T4P from Vibrio cholerae by live fluorescence micros- copy showed that DNA binds at the tip of the pilus [58]. This suggests that the binding of DNA is happening at the pilus tip principally via the minor pilins, despite their low abundance. ComP from the Neisseria species is also thought to be located at the pilus tip [59]. The retraction of the T4P brings the DNA to the cell surface where the DNA can be internalized via the secretin pore of the pilus or an alternate channel [7]. Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creativeco mmons.org/licenses/by/4.0/), which permits unrestricted use, distribu- tion, 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. 1 3 2. Pelicic V (2008) Type IV pili: e pluribus unum? Mol Microbiol 68:827–837. https://doi.org/10.1111/j.1365-2958.2008.06197.x 1. Craig L, Pique ME, Tainer JA (2004) Type IV pilus structure and bacterial pathogenicity. Nat Rev Microbiol 2:363–378 References Mancl JM, Black WP, Robinson H, Yang Z, Schubot FD (2016) Crystal structure of a type IV pilus assembly ATPase: insights into the molecular mechanism of PilB from Thermus thermophilus. Structure 24:1886–1897. https://doi.org/10.1016/j.str.2016.08.010 11. Strom MS, Nunn DN, Lory S (1993) A single bifunctional enzyme, PilD, catalyzes cleavage and N-methylation of proteins belonging to the type IV pilin family. Proc Natl Acad Sci USA 90:2404–2408 27. Georgiadou M, Castagnini M, Karimova G, Ladant D, Pelicic V (2012) Large-scale study of the interactions between pro- teins involved in type IV pilus biology in Neisseria menin- gitidis: characterization of a subcomplex involved in pilus assembly. Mol Microbiol 84:857–873. https://doi.org/10.111 1/j.1365-2958.2012.08062.x 12. Arts J, van Boxtel R, Filloux A, Tommassen J, Koster M (2007) Export of the pseudopilin XcpT of the Pseudomonas aeruginosa type II secretion system via the signal recognition particle-Sec pathway. J Bacteriol 189:2069–2076 28. Karuppiah V, Derrick JP (2011) Structure of the PilM-PilN inner membrane type IV pilus biogenesis complex from Thermus ther- mophilus. J Biol Chem 286:24434–24442. https://doi.org/10.1074/ jbc.M111.243535 13. Francetic O, Buddelmeijer N, Lewenza S, Kumamoto CA, Pug- sley AP (2007) Signal recognition particle-dependent inner membrane targeting of the PulG Pseudopilin component of a type II secretion system. J Bacteriol 189:1783–1793 29. Berry JL, Phelan MM, Collins RF, Adomavicius T, Tønjum T, Frye SA, Bird L, Owens R, Ford RC, Lian LY, Derrick JP (2012) Structure and assembly of a trans-periplasmic channel for type IV pili in Neisseria meningitidis. PLoS Pathog. https://doi. org/10.1371/journal.ppat.1002923 14. López-Castilla A, Thomassin JL, Bardiaux B, Zheng W, Nivas- kumar M, Yu X, Nilges M, Egelman EH, Izadi-Pruneyre N, Francetic O (2017) Structure of the calcium-dependent type 2 secretion pseudopilus. Nat Microbiol 2:1686–1695. https://doi. org/10.1038/s41564-017-0041-2 30. Wolfgang M, Putten JPM, Hayes S, Dorward D, Koomey M (2000) Components and dynamics of fiber formation define a ubiquitous biogenesis pathway for bacterial pili. EMBO J 19:6408–6418. https://doi.org/10.1093/emboj/19.23.6408 g 15. Bardiaux B, de Amorim GC, Luna Rico A, Zheng W, Guil- vout I, Jollivet C, Nilges M, Egelman EH, Izadi-Pruneyre N, Francetic O (2019) Structure and assembly of the enterohem- orrhagic Escherichia coli type 4 pilus. Structure. https://doi. org/10.1016/j.str.2019.03.021 31. Carter T, Buensuceso RNC, Tammam S, Lamers RP, Harvey H, Howell PL, Burrows LL (2017) The type IVa pilus machinery is recruited to sites of future cell division. MBio. https://doi. org/10.1128/mBio.02103-16 16. References 1. Craig L, Pique ME, Tainer JA (2004) Type IV pilus structure and bacterial pathogenicity. Nat Rev Microbiol 2:363–378 p g y 2. Pelicic V (2008) Type IV pili: e pluribus unum? Mol Microbiol 68:827–837. https://doi.org/10.1111/j.1365-2958.2008.06197.x 3. Hospenthal MK, Costa TRD, Waksman G (2017) A comprehen- sive guide to pilus biogenesis in Gram-negative bacteria. Nat Rev Microbiol 15:365–379l Some non-core minor pilins have large structural inser- tions. The recently determined structure of the minor T4a pilin ComZ from T. thermophilus [60] shows an unusual structure with two domains, a pilin-like domain and a large additional domain inserted into its β-sheet. Hence, although 4. Van Gerven N, Waksman G, Remaut H (2011) Pili and flagella: biology, structure, and biotechnological applications. Academic, Oxford, pp 21–72 1 3 Medical Microbiology and Immunology similarities with type II secretion system pseudopilins. J Struct Biol 196:375–384. https://doi.org/10.1016/j.jsb.2016.08.006 5. Ayers M, Howell PL, Burrows LL (2010) Architecture of the type II secretion and type IV pilus machineries. Future Microbiol 5:1203–1218 22. Chang Y-W, Rettberg LA, Treuner-Lange A, Iwasa J, Søgaard- Andersen L, Jensen GJ (2016) Architecture of the type IVa pilus machine. Science 351:aad2001. https://doi.org/10.1126/scien ce.aad2001f 6. Roux N, Spagnolo J, De Bentzmann S (2012) Neglected but amaz- ingly diverse type IVb pili. Res Microbiol 163:659–673. https:// doi.org/10.1016/j.resmic.2012.10.015 g j 7. Piepenbrink KH (2019) DNA uptake by type IV filaments. Front Mol Biosci 6:1–13. https://doi.org/10.3389/fmolb.2019.00001 23. Gold VA, Salzer R, Averhoff B, Kühlbrandt W (2015) Structure of a type IV pilus machinery in the open and closed state. Elife. 21:4. https://doi.org/10.7554/eLife.07380 8. Giltner CL, Habash M, Burrows LL (2010) Pseudomonas aer- uginosa minor pilins are incorporated into type IV Pili. J Mol Biol 398:444–461. https://doi.org/10.1016/j.jmb.2010.03.028 24. Takhar HK, Kemp K, Kim M, Howell PL, Burrows LL (2013) The platform protein is essential for type IV pilus biogenesis. J Biol Chem 288:9721–9728. https://doi.org/10.1074/jbc.M113.453506 9. Giltner CL, Nguyen Y, Burrows LL (2012) Type IV pilin pro- teins: versatile molecular modules. Microbiol Mol Biol Rev 76:740–772. https://doi.org/10.1128/MMBR.00035-12 25. Tsai CL, Tainer JA (2016) The ATPase motor turns for type IV pilus assembly. Structure 24:1857–1859 p g 10. Berry JL, Pelicic V (2015) Exceptionally widespread nanoma- chines composed of type IV pilins: the prokaryotic Swiss Army knives. FEMS Microbiol Rev 39:134–154 26. References Helaine S, Dyer DH, Nassif X, Pelicic V, Forest KT (2007) 3D structure/function analysis of PilX reveals how minor pilins can modulate the virulence properties of type IV pili. Proc Natl Acad Sci 104:15888–15893. https://doi.org/10.1073/pnas.0707581104 37. Thomassin JL, Santos Moreno J, Guilvout I, Tran Van Nhieu G, Francetic O (2017) The trans-envelope architecture and function of the type 2 secretion system: new insights raising new questions. Mol Microbiol 105:211–226. https://doi.org/10.1111/mmi.13704 p g 38. Sauvonnet N, Vignon G, Pugsley AP, Gounon P (2000) Pilus for- mation and protein secretion by the same machinery in Escheri- chia coli. EMBO J 19:2221–2228 52. Nguyen Y, Harvey H, Sugiman-Marangos S, Bell SD, Buensuceso RNC, Junop MS, Burrows LL (2015) Structural and functional studies of the Pseudomonas aeruginosa Minor Pilin, PilE. J Biol Chem 290:26856–26865. https://doi.org/10.1074/jbc.M115.68333 4 39. Cisneros DA, Bond PJ, Pugsley AP, Campos M, Francetic O (2012) Minor pseudopilin self-assembly primes type II secre- tion pseudopilus elongation. EMBO J 31:1041–1053. https://doi. org/10.1038/emboj.2011.454 53. Bernard SC, Simpson N, Join-Lambert O, Federici C, Laran- Chich MP, Maïssa N, Bouzinba-Ségard H, Morand PC, Chretien F, Taouji S, Chevet E, Janel S, Lafont F, Coureuil M, Segura A, Niedergang F, Marullo S, Couraud PO, Nassif X, Bourdoulous S (2014) Pathogenic Neisseria meningitidis utilizes CD147 for vas- cular colonization. Nat Med 20:725–731. https://doi.org/10.1038/ nm.3563 g j 40. Douzi B, Ball G, Cambillau C, Tegoni M, Voulhoux R (2011) Deciphering the Xcp Pseudomonas aeruginosa type II secretion machinery through multiple interactions with substrates. J Biol Chem 286:40792–40801. https://doi.org/10.1074/jbc.M111.29484 3 41. Luna Rico A, Zheng W, Petiot N, Egelman EH, Francetic O (2019) Functional reconstitution of the type IVa pilus assembly system from enterohaemorrhagic Escherichia coli. Mol Microbiol 111:732–749. https://doi.org/10.1111/mmi.14188 54. Marko VA, Kilmury SLN, MacNeil LT, Burrows LL (2018) Pseu- domonas aeruginosa type IV minor pilins and PilY1 regulate virulence by modulating FimS-AlgR activity. PLoS Pathog. https ://doi.org/10.1371/journal.ppat.1007074 42. Carbonnelle E, Hélaine S, Prouvensier L, Nassif X, Pelicic V (2005) Type IV pilus biogenesis in Neisseria meningitidis: PilW is involved in a step occurring after pilus assembly, essential for fibre stability and function. Mol Microbiol 55(1):54–64 55. Saldana Z, Giron JA, Monteiro-Neto V, Ledesma MA, Xicohten- catl-Cortes J, Puente JL, Monteiro-Neto V, Saldaña Z, Ledesma MA, Puente JL, Girón JA (2008) The type 4 pili of enterohem- orrhagic Escherichia coli O157:H7 are multipurpose structures with pathogenic attributes. J Bacteriol 91:411–421. https://doi. org/10.1128/jb.01306-08 i 43. References Wang F, Coureuil M, Osinski T, Orlova A, Altindal T, Ges- bert G, Nassif X, Egelman EH, Craig L (2017) Cryoelectron microscopy reconstructions of the Pseudomonas aeruginosa and Neisseria gonorrhoeae Type IV Pili at sub-nanometer res- olution. Structure 25:1423–1435.e4. https://doi.org/10.1016/j. str.2017.07.016 32. Siewering K, Jain S, Friedrich C, Webber-Birungi MT, Sem- chonok DA, Binzen I, Wagner A, Huntley S, Kahnt J, Klingl A, Boekema EJ, Søgaard-Andersen L, van der Does C (2014) Peptidoglycan-binding protein TsaP functions in surface assembly of type IV pili. Proc Natl Acad Sci 111:E953–E961. https://doi. org/10.1073/pnas.1322889111 17. Kolappan S, Coureuil M, Yu X, Nassif X, Egelman EH, Craig L (2016) Structure of the neisseria meningitidis type IV pilus. Nat Commun 7:1–12. https://doi.org/10.1038/ncomms13015 33. Burrows LL, Sundaram P, Koo J, Andrew Chong P, Howell PL, Sampaleanu LM, Tammam S, Ayers M, Forman-Kay JD (2011) Characterization of the PilN, PilO and PilP type IVa pilus sub- complex. Mol Microbiol 82:1496–1514. https://doi.org/10.111 1/j.1365-2958.2011.07903.x 18. Bardiaux B, Cordier F, Brier S, López-Castilla A, Izadi- Pruneyre N, Nilges M (2019) Dynamics of a type 2 secretion system pseudopilus unraveled by complementary approaches. J Biomol NMR. 73(6–7):293–303. https://doi.org/10.1007/s1085 8-019-00246-4 (Epub 2019 May 23) 34. Tammam S, Sampaleanu LM, Koo J, Manoharan K, Daubaras M, Burrows LL, Howell PL (2013) PilMNOPQ from the Pseu- domonas aeruginosa type IV pilus system form a transenvelope protein interaction network that interacts with PilA. J Bacteriol 195(10):2126–2135. https://doi.org/10.1128/jb.00032-13 (Epub 2013 Mar 1) 19. Li J, Lim MS, Li S, Brock M, Pique ME, Woods VL Jr, Craig L (2008) Vibrio cholerae toxin-coregulated pilus structure analyzed by hydrogen/deuterium exchange mass spectrom- etry. Structure. 16(1):137–148. https://doi.org/10.1016/j. str.2007.10.027 20. Sugiman-Marangos S, Junop MS, Nguyen Y, Bell SD, Burrows LL, Harvey H, Charlton CL (2014) Pseudomonas aeruginosa minor pilins prime Type IVa pilus assembly and promote surface display of the PilY1 adhesin. J Biol Chem 290:601–611. https:// doi.org/10.1074/jbc.m114.616904 35. Balasingham SV, Collins RF, Assalkhou R, Homberset H, Frye SA, Derrick JP, Tønjum T (2007) Interactions between the lipo- protein PilP and the secretin PilQ in Neisseria meningitidis. J Bacteriol. 189(15):5716–5727 (Epub 2007 May 25) 36. Drake SL, Sandstedt SA, Koomey M (1997) PilP, a pilus bio- genesis lipoprotein in Neisseria gonorrhoeae, affects expres- sion of PilQ as a high-molecular-mass multimer. Mol Microbiol 23(4):657–668 g j 21. Karuppiah V, Thistlethwaite A, Derrick JP (2016) Structures of type IV pilins from Thermus thermophilus demonstrate 1 3 Medical Microbiology and Immunology 51. References Carbonnelle E, Helaine S, Nassif X, Pelicic V (2006) A systematic genetic analysis in Neisseria meningitidis defines the Pil proteins required for assembly, functionality, stabilization and export of type IV pili. Mol Microbiol 61(6):1510–1522 g j 56. Cehovin A, Simpson PJ, McDowell MA, Brown DR, Noschese R, Pallett M, Brady J, Baldwin GS, Lea SM, Matthews SJ, Pelicic V (2013) Specific DNA recognition mediated by a type IV pilin. Proc Natl Acad Sci 110:3065–3070. https://doi.org/10.1073/ pnas.1218832110 44. Nguyen Y, Sugiman-Marangos S, Harvey H, Bell SD, Charlton CL, Junop MS, Burrows LL (2015) Pseudomonas aeruginosa minor pilins prime type IVa pilus assembly and promote surface display of the PilY1 adhesin. J Biol Chem. 290(1):601–611 57. Wolfgang M, Van Putten JPM, Hayes SF, Koomey M (1999) The comP locus of Neisseria gonorrhoeae encodes a type IV prepilin that is dispensable for pilus biogenesis but essential for natural transformation. Mol Microbiol 31:1345–1357. https://doi.org/10 .1046/j.1365-2958.1999.01269.x 45. Cisneros DA, Pehau-Arnaudet G, Francetic O (2012) The heter- ologous assembly of type IV pilin by a type II secretion system reveals the role of minor pilins in assembly initiation. Mol Micro- biol 86(4):805–818 46. Korotkov KV, Hol WGJ (2008) Structure of the GspK-GspI- GspJ complex from the enterotoxigenic Escherichia coli type 2 secretion system. Nat Struct Mol Biol 15:462–468. https://doi. org/10.1038/nsmb.1426 58. Ellison CK, Dalia TN, Vidal Ceballos A, Wang JCY, Biais N, Brun YV, Dalia AB (2018) Retraction of DNA-bound type IV competence pili initiates DNA uptake during natural transforma- tion in Vibrio cholerae. Nat Microbiol 3:773–780. https://doi. org/10.1038/s41564-018-0174-y 47. Kolappan S, Ng D, Yang G, Harn T, Craig L (2015) Crystal struc- ture of the minor pilin CofB, the initiator of CFA/III pilus assem- bly in enterotoxigenic Escherichia coli. J Biol Chem 290:25805– 25818. https://doi.org/10.1074/jbc.M115.676106 59. Aas FE, Løvold C, Koomey M (2002) An inhibitor of DNA bind- ing and uptake events dictates the proficiency of genetic transfor- mation in Neisseria gonorrhoeae: mechanism of action and links to Type IV pilus expression. Mol Microbiol 46:1441–1450. https ://doi.org/10.1046/j.1365-2958.2002.03265.x 48. Kawahara K, Oki H, Fukakusa S, Yoshida T, Imai T, Maruno T, Kobayashi Y, Motooka D, Iida T, Ohkubo T, Nakamura S (2016) Homo-trimeric structure of the type IVb minor pilin CofB sug- gests mechanism of CFA/III pilus assembly in human enterotoxi- genic Escherichia coli. J Mol Biol 428:1209–1226. https://doi. org/10.1016/j.jmb.2016.02.003 60. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Salleh MZ, Karuppiah V, Snee M, Thistlethwaite A, Levy CW, Knight D, Derrick JP (2019) Structure and properties of a natural competence-associated pilin suggest a unique pilus tip-associated DNA receptor. MBio 10:1–17. https://doi.org/10.1128/mBio 61. Duménil G (2019) Type IV pili as a therapeutic target. Trends Microbiol 27:658–661 49. Ng D, Harn T, Altindal T, Kolappan S, Marles JM, Lala R, Spiel- man I, Gao Y, Hauke CA, Kovacikova G, Verjee Z, Taylor RK, Biais N, Craig L (2016) The vibrio cholerae minor pilin TcpB initiates assembly and retraction of the toxin-coregulated pilus. PLoS Pathog 12(12):e1006109. https://doi.org/10.1371/journ al.ppat.1006109 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 50. Gutierrez-Rodarte M, Kolappan S, Burrell BA, Craig L (2019) The Vibrio cholerae minor pilin TcpB mediates uptake of the cholera toxin phage CTXφ. J Biol Chem 294:15698–15710 1 3
https://openalex.org/W4381308350	https://discovery.ucl.ac.uk/id/eprint/10172228/1/Semantic%20fluency%20in%20primary%20school%20age%20children%20with%20vision%20impairment.pdf	English	null	Semantic fluency in primary school-age children with vision impairment	International journal of speech-language pathology	2,023	cc-by	8,111	International Journal of Speech-Language Pathology ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/iasl20 Abstract Purpose: Semantic fluency is potentially a useful tool for vocabulary assessment in children with vision impairment because it contains no visual test stimuli. It is not known whether in the primary school years children with vision impair- ment perform more poorly on semantic fluency tasks compared to their sighted peers. Method: We compared semantic fluency performance of two groups of 5- to 11-year-old British English speaking chil- dren—one group with vision impairment and one without. We also investigated within-group differences in performance, based on severity of vision impairment. We administered one category (animals) to children with vision impairment (n ¼ 45) and sighted children (n ¼ 30). Participants had one minute to respond. Responses were coded for accuracy, error type, clusters, and switches. Result: Correct responses increased with age within each group. Groups did not differ significantly on any outcome meas- ure. Severity of vision impairment did not impact task performance. Conclusion: Results suggested that semantic fluency performance—at least for the category animals—is not different in chil- dren with vision impairment compared to sighted children. Findings also suggest that semantic fluency could be a suitable addition to the tools that speech-language pathologists use to assess language abilities in children with vision impairment. Keywords: vision impairment; blindness; semantic fluency; vocabulary; children; primary-school Jessica Hayton, Jessica Marshall, Helen Girvan & Chloe Marshall essica Hayton, Jessica Marshall, Helen Girvan & Chloe Marshall To cite this article: Jessica Hayton, Jessica Marshall, Helen Girvan & Chloe Marshall (2023): Semantic fluency in primary school-age children with vision impairment, International Journal of Speech-Language Pathology, DOI: 10.1080/17549507.2023.2214711 To link to this article: https://doi.org/10.1080/17549507.2023.2214711 © 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published online: 19 Jun 2023. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=iasl20 International Journal of Speech-Language Pathology, 2023; Early Online: 1–10 JESSICA HAYTON1 , JESSICA MARSHALL2, HELEN GIRVAN3 & CHLOE MARSHALL1 1Department of Psychology and Human Development, IOE, UCL’s Faculty of Education and Society, London, UK, 2Department of Psychology, University of York, UK, and 3IOE, UCL’s Faculty of Education and Society, London, UK Introduction et al., 2006). However, to the best of our knowledge there is only one, small-scale, study exploring how chil- dren with vision impairment perform on semantic flu- ency tasks. We therefore do not know whether they are equally productive as sighted children and whether they retrieve words in the same way. In this paper we present the first analysis of semantic fluency in primary school-age children with vision impairment (VI). The semantic fluency task requires participants to name as many members of a category (e.g. animals, foods, objects from around the house) as they can in a limited period of time (e.g. 30seconds or 1minute; see Chami et al., 2018; Mengisidou et al., 2020). Measures of interest include the number of cor- rect items produced, the number of errors, and the number and size of clusters (common clustered responses for the category animals, for example, include farm animals [e.g. horse, cow, sheep], pets [e.g. cat, dog], and zoo animals [e.g. lion, tiger, giraffe, elephant, zebra]). Semantic fluency is a widely used tool in language and cognitive assessments in children and adults (see Ardila The vast literature on sighted children’s word learn- ing demonstrates how the acquisition of words and their meanings is grounded in sensory experiences (Rose et al., 2022). In the case of spoken language acquisition, word learning frequently involves a cross-modal associ- ation between a spoken word and an object or action in view. The child is supported in making these associa- tions by caregivers who use visual cues such as eye gaze and a range of manual gestures such as pointing. Visual experiences help the child make and strengthen links between word and referent, supporting concept ISSN 1754-9507 print/ISSN 1754-9515 online 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published by Taylor & Francis DOI: 10.1080/17549507.2023.2214711 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Correspondence: Jessica Hayton, Department of Psychology and Human Development, IOE, UCL’s Faculty of Education and Society, London, UK. Email: jessica.hayton@ucl.ac.uk ISSN 1754-9507 print/ISSN 1754-9515 online 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published by Taylor & Francis DOI: 10.1080/17549507.2023.2214711 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. ISSN 1754-9507 print/ISSN 1754-9515 online 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published by Taylor & Francis ISSN 1754-9507 print/ISSN 1754-9515 online 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published by Taylor & Francis DOI: 10.1080/17549507.2023.2214711 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits sica Hayton, Department of Psychology and Human Development, IOE, UCL’s Faculty of Education and Society, London, UK. n@ucl.ac.uk Introduction The causes of blindness in the sample included Lebers congenital amaurosis (a congenital retinal dystrophy), retinopathy of prematurity (excessive development of retinal blood vessels in premature or low birthweight babies), retino- blastoma (a form of paediatric eye cancer), or an unknown birth/genetic defect as reported by parents. The semantic categories used were “all the things you might find around the house” and “all the things you might find in the supermarket”. Combining the scores for these two categories, the authors found that the chil- dren with VI produced an average of 23.34 (SD ¼ 7.10) correct responses in comparison to the sighted control group’s average of 27.50 (SD ¼ 5.40) responses. This group difference was not statistically significant (t ¼ 1.93, p ¼ .062), nor did the groups differ in cluster size. Age related differences were not investigated in this study. The authors point to Perez-Pereira and Conti- Ramsden (2013), who argued that the most prudent interpretation of the limited data on language develop- ment in children with VI is that there are large individual differences and that it is difficult to point to areas where differences are consistently found. Nevertheless, although the aforementioned group difference in correct responses did not emerge as statistically significant in the study by Wakefield et al. (2006), their data do suggest that productivity might be lower in children with VI and that further studies are warranted. With respect to language acquisition, areas of vulner- ability for children with VI include verbal concepts including words outside the child’s direct experience (Rose et al., 2022). Yet, seminal work by Landau and Gleitman (1985) reported that toddlers with blindness had just as complex semantic knowledge for some words compared with sighted peers. Landau and Gleitman’s findings are consistent with experimental studies of sighted toddlers by Wojcik and Saffran (2015) on word learning and syntactic sentence structure. Wojcik and Saffran (2015) demonstrated that the meaning of novel words can be learned effectively from speech input alone, without the presentation of visual information. Together, this body of work suggests that speech input alone can lead to effective semantic learning. Further studies of semantic fluency could also be valuable because in a review by Rose et al. (2022) of the language assessment tools that have been used with children with VI, semantic fluency does not appear (Rose et al., 2022). Introduction The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. J. Hayton et al. 2 formation with information about the what, who, how, and why. Children with VI are not likely to have the same quality of word learning experiences. also differs (zoos and savannahs for zebras, stables and pastures for horses), but again this information might not be available to a child with VI unless verbally pro- vided. These different experiences between sighted children and those with VI might give rise to different conceptual representations with different levels of rich- ness, and therefore potentially different patterns of retrieval during semantic fluency tests. Vision impairment is defined as any ocular or brain based condition that cannot be corrected with medi- cation, surgery, or prescription lenses, and the indi- vidual may have some degree of residual or functional vision (Solebo & Rahi, 2014; Wall, 2019). Blindness is defined under the same boundary, though the indi- vidual has no residual or functional vision, but may have light perception (Solebo & Rahi, 2014). Site (i.e. location) and severity of vision impairment have cascading effects on how children and young people explore and engage with their environment. The pres- ence of a VI can reduce or deny access to incidental learning (i.e. learning through observation) in add- ition to reducing motivation to move towards or away from visual stimuli (Wall, 2019). In addition, the age of onset of VI is relevant: Notably, the visual experi- ences of babies born with a congenital VI (i.e. blind- ness) will differ from infants and young children who have acquired a VI shortly after birth or in infancy. This is because the latter will have had some previous visual experiences, regardless of whether they were able to consciously process or understand such visual input (Wall, 2019). Further details regarding classifi- cation of VI relevant to the sample in the current study are presented in the Methods section. g y There is one small-scale study of semantic fluency in English speaking Australian children with blindness (n¼ 16) aged 11–18 (Wakefield et al., 2006). The sam- ple were either blind from birth (n¼ 13) or became blind before their first birthday (n¼ 3). The sample did not have additional cognitive disabilities, though two participants were medicated for epilepsy. Introduction The Royal National Institute for the Blind (RNIB) suggest that approximately 50% of children with VI have an add- itional educational need/disability (Emerson & Robertson, 2011; RNIB, 2022). Nevertheless, assess- ing the language of children with VI is important because the prevalence of language difficulties in these children may be underestimated (Rose et al., 2022). ( g p g ) Defining severity of vision impairment relative to visual acuity is offered by the World Health Organization (2022) via the International Classification of Diseases-11 (2022b). The classifica- tion uses both Logarithm of the Minimum Angle of Resolution (logMAR) and Snellen scores to explain and interpret severity (Solebo & Rahi, 2014). Both logMAR and Snellen scores are tests of visual acuity, characterised by a chart containing rows of capital let- ters that gradually decrease in size. Scores are based on the smallest discernible letter that the patient can read. The logMAR chart was introduced into clinical practice, resultant from its sensitivity threshold rela- tive to visual acuity despite no direct correlation with the Snellen chart (Lovie-Kitchin, 2015). In logMAR, lower scores correspond to better vision; as the deci- mal notation increases this represents worsening vis- ual acuity. The Snellen score offers a more relatable indication of a person’s visual field whereby the frac- tion relates to the distance (i.e. 6 metres [in the UK context] or 20 feet [in the USA context]) that a per- son with a vision impairment can see compared to a typically sighted person. To illustrate, a Snellen score of 6/6 metres, otherwise known as 20/20 vision, corre- sponds to a logMAR score of 0.0. This means that a person can see at 6 metres what a typically sighted person can see at 6 metres. If a person with VI has a logMAR score of 1.00, equivalent to a Snellen score of 6/60, this means that a person with VI can see at 6 metres what a typically sighted person can see at 60 metres. As Snellen is better recognised in some con- texts, conversions are offered where appropriate in this paper. Semantic fluency is a widely used task in language assessment and it might be particularly suitable for assessing the vocabulary skills of children with VI, because it requires no visual test stimuli. However, an open question concerns whether it is an appropriate measure for children with VI in the primary school years. Introduction Given the differences in language learning experiences for children with VI compared to sighted children, it is possible that semantic fluency scores will be lower in children with VI and that patterns of lexical retrieval will be different. It is this overarching research question that we set out to investigate in this study of 45 VI and 30 sighted children aged 5– 11years and resident in the United Kingdom. As far as we are aware, this is the first study of its kind. Specifically, we compare the performance of two groups of 5- to11-year-old British English speaking children—one group with VI and one group with- out—on a semantic fluency task using the following measures: number of correct responses, number of errors, number of clusters, cluster size, and number of switches. We investigate the extent to which the number of correct responses correlates with age within each group. We also investigate whether there are differen- ces in performance between children with more or less severe VI. The WHO (2022) taxonomy classifies vision impairment into four groups. These are: (a) mild or no vision impairment, defined as vision better than or equal to 0.48 logMAR (6/18 Snellen); (b) moderate vision impairment, whereby vision is worse than 0.48 logMAR but equal to 1.0 logMAR (6/60 Snellen); (c) severe sight impairment acuity is worse than 1.0 logMAR, but equal to or better than 1.3 logMAR (3/60 Snellen); and (d) blindness, whereby vision is Introduction Yet semantic fluency might have some advantages over other assessments for elicit- ing language, because it does not rely on picture stim- uli. On the other hand, given the importance of the sense of vision for word learning, we cannot assume that children with VI will give as many responses, and of the same type, as sighted children when responding to a semantic fluency task. It is this gap in the clinical and research literature on primary school-age children with VI that motivates our exploratory study. It is then an open question, and one we explore in this study, as to whether any differences in the word learning experience impact lexical retrieval measured by semantic performance in children with VI. Here we give an example relevant to the semantic category ani- mals that we use in the semantic fluency task in the current study. Zebra and horse are closely taxonomic- ally related, and are very similar to one another in shape. Yet there are very salient visual differences, not- ably the black and white stripes of the zebra that are lacking in the horse and the wider range of colours that horses can be. These visual characteristics that are obvious to a sighted child viewing toys, pictures, or real-life zebras and horses might not be accessible to a child with VI unless verbally described by a caregiver (or other interlocutor). The habitat of the two animals Vision impairment is a low incidence disability. For example, recent figures from the UK’s Department for Education (2021) suggest that 0.16% of school-age Semantic fluency in primary school-age children with vision impairment 3 3 children have a primary diagnosis of vision impairment, marginally lower than the two in every thousand (i.e. 0.2%) prevalence rate detailed by the World Health Organization (WHO; 2022). Keil (2019) recom- mended caution should be used with government sta- tistics, however, as most published materials focus only on primary diagnosis. UK statistics including VI as a primary and as a secondary diagnosis subsequently met the 0.2% threshold consistent with the WHO’s classifi- cation of childhood VI and blindness, with similar fig- ures in Australia (McLeod & Mckinnon, 2007). Higher figures have been reported in some other coun- tries (e.g. India [0.5%, see Kulkarni et al., 2022] and Malawi [1.1%, see Kalua et al., 2008]) with country income predicting prevalence, i.e. Introduction low income is associ- ated with a high prevalence of VI (WHO, 2022). Speech-language pathologists are therefore not likely to have much experience with this population. Likewise, research in language abilities of school-age children with VI without additional needs is sparse. The Royal National Institute for the Blind (RNIB) suggest that approximately 50% of children with VI have an add- itional educational need/disability (Emerson & Robertson, 2011; RNIB, 2022). Nevertheless, assess- ing the language of children with VI is important because the prevalence of language difficulties in these children may be underestimated (Rose et al., 2022). Vision impairment is defined as any ocular condi- tion that cannot be corrected using surgical interven- tion, prescription glasses/lenses, nor medication. The scope of vision impairment encompasses: (a) the globe of the eye(s); (b) genetic/hereditary conditions; and (c) brain based cerebral vision impairment, whereby the globe of the eye(s) remains structurally unaffected, though processing visual information is compromised (e.g. due to trauma, infection, child- hood stroke, or genetic mutations; see International Classification of Diseases-11 [ICD-11], 2022a). The site of vision impairment (i.e. globe of the eye[s], ret- ina, optic nerve, or cerebral vision impairment) can indicate underlying causes of vision impairment and the nature of the progression of the condition, i.e. if the condition is stable (e.g. Albinism) or degenerative (e.g. retinitis pigmentosa). children have a primary diagnosis of vision impairment, marginally lower than the two in every thousand (i.e. 0.2%) prevalence rate detailed by the World Health Organization (WHO; 2022). Keil (2019) recom- mended caution should be used with government sta- tistics, however, as most published materials focus only on primary diagnosis. UK statistics including VI as a primary and as a secondary diagnosis subsequently met the 0.2% threshold consistent with the WHO’s classifi- cation of childhood VI and blindness, with similar fig- ures in Australia (McLeod & Mckinnon, 2007). Higher figures have been reported in some other coun- tries (e.g. India [0.5%, see Kulkarni et al., 2022] and Malawi [1.1%, see Kalua et al., 2008]) with country income predicting prevalence, i.e. low income is associ- ated with a high prevalence of VI (WHO, 2022). Speech-language pathologists are therefore not likely to have much experience with this population. Likewise, research in language abilities of school-age children with VI without additional needs is sparse. Method In this section we first define VI in a way that is rele- vant to our study and subsequent interpretation of results. 4 J. Hayton et al. 4 with vision impairment did not produce any responses in the semantic fluency task. One child with vision impairment required verbal reassurance from the test administrator (e.g. a confirmatory response to the child’s question “is [sic] birds an animal?”), which voided the results. Seven sighted children dropped out due to time commitments as disclosed by parents/care- givers and one sighted child was identified as an outlier when examining data distribution. This participant named 32 animals, placing them 3 standard deviations above the sighted group mean and over 1 standard deviation above the next highest scores (25 animals) for two sighted children. Thus, data collected from 75 child participants are examined in this paper, whereby n¼ 45 children with vision impairment (female n¼ 21; age range 5.08–11.33 years; M¼ 8.29 years) and n¼ 30 sighted children (female n¼ 17; age range 5–11.75 years; M¼ 7.9years). worse than 1.3 logMAR. It is important to note that there is no current defined taxonomy for childhood vision impairment, though it is acknowledged that chronological and developmental ages may affect the accuracy of the categorisation (Solebo & Rahi, 2014). For this reason, severe sight impairment and blind- ness are often merged, despite distinct phenotypes and characteristics of their vision. For example, a child presenting with 1.1 logMAR may still be able to navigate elements of their environment using avail- able functional vision (a capacity that can be influ- enced by environmental factors, such as appropriate task lighting that may facilitate movement in the environment) but remains grouped with children in the category blindness who may have no light percep- tion. The WHO taxonomy classifies vision impair- ment into four groups, this has been adopted in the current study for understanding the visual acuity of recruited participants (WHO, 2022). Inclusion criteria included parental report con- firming no hearing difficulties (e.g. no hearing aid) of their child/ren (child participants had the ability to hear and respond to test instructions), a clinical diag- nosis of vision impairment (for the group with vision impairment), no diagnosed or suspected additional educational need/disability (both VI and sighted groups), and normal or corrected-normal vision (i.e. glasses/lenses for refractive error correction in the sighted group). Participants Data from 98 participants from a larger study (Hayton et al., 2021) were screened for eligibility in the current paper. The larger study explored sleeping profiles in children with vision impairment and sighted children (matched on chronological age) aged 4–11years. The number of responses to semantic and phonemic verbal fluency and digit span tasks were collected, but did not predict sleep quality nor quan- tity as measured by questionnaire, sleep diary, and actigraphy data (Hayton et al., 2021). That study did not focus on the semantic fluency data nor present them in any detail, which is our aim here. Table I shows the demographic characteristics of the sample, detailing timing of onset, severity, and diagnosis by site of vision impairment, where known. It is important to note the onset of vision impairment was based on time of diagnosis; all parents who pro- vided information reported their children having the visual condition from birth. To be eligible for inclusion in the analysis in the cur- rent paper, children needed to have produced at least one item in the semantic fluency task. In total, 23 par- ticipants were excluded from analysis. Specifically, five children with vision impairment and one sighted child had no or missing data. Two verbal children with vision impairment and one sighted child dropped out from the study due to illness. Three children with vision impairment did not participate as parental report cited that they were non-verbal. Two children Method Confirmation of visual and auditory status, no additional learning needs, and all demo- graphic information were based on parental/caregiver report. Child participants were born in the UK so had English as either a first or additional language. Before testing, a short conversation was held with each child participant confirming their understanding and appropriate response to spoken English. Parents had English as a first or additional language con- firmed via email correspondence and a telephone call. Parental first language was not an exclusion cri- terion for this study, as the data were based on child- ren’s language ability. Confirmation of proficiency in English was based on informed consent and child assent to participate in the parent study (Hayton et al., 2021), in addition to arranging a telephone call to inform parents/caregivers of the procedure and to speak with the child participants. Recruitment Ethics were approved by UCL’s Institute of Education research ethics committee. Parent/caregiver written informed consent was obtained, as was informed child consent and verbal assent during the testing phase. Participants were recruited from across the UK as part of a larger study, and parents/caregivers informed that depersonalised data would be processed. Sighted chil- dren were educated in mainstream settings. Children with vision impairment were educated in both special- ist and mainstream settings, arguably representative of the types of formal education a child with a vision impairment may experience (depending on severity of diagnosis and its implications for education). Semantic fluency in primary school-age children with vision impairment 5 5 Table I. Participant demographics for children with vision impairment and sighted children. Demographic characteristic Vision impairment (n ¼ 45) Sighted (n ¼ 30) Chronological age in years, mean (SD) 8.29 (1.84) 7.94 (2.04) Sex Female 21 (46.7%) 17 (56.7%) Male 24 (53.3%) 13 (43.3%) Severity of vision impairmenta Mild or no sight impairment 7 (15.6%) – Moderate 15 (33.3%) – Severe 16 (35.6%) – Blindness 7 (15.6%) Sighted 30 (100%) Diagnosis by site of vision impairment Ocular (including retina) 26 (57.8%) – Cerebral (including nystagmus and optic nerve) 19 (42.2%) – – Note. aSeverity of vision impairment using WHO classification (e.g. Solebo & Rahi, 2014); mild or no sight impair- ment ¼ vision acuity better than or equal to 0.48 logMAR; moderate vision impairment ¼ 0.6–1.0 logMAR; severe vision impairment ¼ 1.1–1.3 logMAR; blindness ¼ 1.4 logMAR. Table I. Participant demographics for children with vision impairment and sighted children. Note. aSeverity of vision impairment using WHO classification (e.g. Solebo & Rahi, 2014); mild or no sight impair- ment ¼ vision acuity better than or equal to 0.48 logMAR; moderate vision impairment ¼ 0.6–1.0 logMAR; severe vision impairment ¼ 1.1–1.3 logMAR; blindness ¼ 1.4 logMAR. benefit of the doubt regarding their use of clusters”. This post hoc, emergent approach to coding is used far more commonly in the research literature (e.g. Beal-Alvarez & Figueroa, 2017; Chami et al., 2018; Henry et al., 2015; Kosmidis et al., 2004; Marshall et al., 2018; Mengisidou et al., 2020; inter alia) than the imposition of a priori categories that arguably do not fully capture how children retrieve lexical items (e.g. Nash & Snowling, 2008). participation in the verbal fluency measures. It is important to note that many parents do not and will not know the cause, severity, or sometimes site of vision impairment. Semantic verbal fluency was measured using instructions provided by Strauss et al. (2006). First, a trial using the category “things that you find in the kitchen” was presented. Examples offered by the administrator were: knives, forks, spoons, and plates. Participants were then asked to continue and they included words such as microwave and fridge freezer. Upon completion of the trial, the target category of animals was introduced and the following instruc- tions were read aloud: “Now tell me the names of as many animals as you can. Name them as quickly as possible.” (Strauss et al., 2006). Semantic fluency in primary school-age children with vision impairment 5 Timing of 1 minute commenced at the end of the verbal instruction. If a participant paused for 15seconds, the instructions were repeated and the starting word dog was given. Clusters were defined as two or more adjacent responses that were closely related semantically, i.e. had a thematic association (e.g. the water animal clus- ter included shark, whale, fish) and/or a taxonomic association (e.g. the bear cluster included polar bear, grizzly bear, brown bear). We also identified phono- logical clusters, where the initial sound triggered fur- ther words beginning with the same sound but are semantically unrelated (e.g. /k/, koala, chameleon, cat, cow). It is unusual to code phonological clusters within the semantic fluency task and, where they have been coded, they have been reported to be much less fre- quent than semantic clusters (Koren et al., 2005). However, in keeping with our emergent approach to coding and mindful of advice from Troyer et al. (1997) that individuals vary considerably in how they tackle the task, we chose to include phonological clus- ters in our coding1. Some words fell into different cate- gories depending on the context of the responses e.g. butterfly was coded under “minibeast” in the sequence ladybird, caterpillar, butterfly or as part of a phonological cluster in the sequence bee, bear, butterfly, bird. Measures Medical history questionnaires were completed by parents/caregivers detailing background information such as diagnosis/es and additional needs. This infor- mation was important in understanding clinical con- dition and any impairment(s) that might affect Semantic fluency in primary school-age children with vision impairment 5 Semantic fluency in primary school-age children with vision impairment 5 Coding Responses were audio recorded for later transcrip- tion. Words were omitted from the correct semantic score if they violated the task instructions i.e. out of category words and repetitions (whereby repetitions were identified as either variations of the same word (e.g. dog and doggie) or exact repetition (e.g. dog and dog; Strauss et al., 2006). Words such as chocolate and cockadoodle were considered out of category errors. Although authors rarely describe how they treat mythical creatures such as unicorn, yeti, and dragon, we followed McGregor et al. (2018) in accepting them as correct responses. By way of examples to illustrate how cluster size was calculated, the cluster zebra lion was calculated as having a size of two; kangaroo, cheetah, lion was calcu- lated as three; and bee, bear, butterfly, bird was calcu- lated as four, etc. The number of switches was calculated as the number of consecutive items that were not related to one another either semantically or phonologically, e.g. the transition in snake bee, was classed as a switch, as was bird kangaroo. Responses (including repetitions, but not out of category responses) were coded for semantic and phonological clusters. Codes were allowed to emerge from the data. In doing so, we followed the recom- mendation of Troyer et al. (1997, p. 140) who wrote that the large number of subcategories thus gener- ated, “reflects the considerable individual variations in approach to this task and gives participants the All participant responses were independently coded by two authors. Inter-rater agreement of clus- ter categories was 86.6%. This level of inter-rater J. Hayton et al. 6 agreement is comparable to figures reported in other studies (83%, Beal-Alvarez & Figueroa, 2017; 87%, Chami et al., 2018; 89%, Marshall et al., 2018). All disagreements in coding were discussed to reach con- sensus, including with a third coder if necessary. minimum of five words and maximum 24 correct words. The number of errors was very low, with the majority of children not producing any errors. First, we investigated whether the number of correct responses correlated with age in each group and we found that it did so moderately. For the VI group, r ¼ .448, p ¼ .002. For the sighted group, r ¼ .493, p ¼ .006. Not surprisingly then, an increase in age was associated with an increase in productivity. The scatter- plot showing this association is presented in Figure 1. Result There was a great deal of variation in number of cor- rect responses produced by participants. The group with VI produced a minimum three and maximum of 25 correct words, and the sighted group produced a Determining the distribution of the correct seman- tic responses was important for determining the type Figure 1. Grouped scatterplot illustrating the association between chronological age and the number of correct responses for both the vision impaired and the sighted groups. Figure 1. Grouped scatterplot illustrating the association between chronological age and the number of correct responses for both the vision impaired and the sighted groups. Figure 2. Histogram illustrating the distribution of the total number of responses in the full sample2. Figure 2. Histogram illustrating the distribution of the total number of responses in the full sample2. 7 Semantic fluency in primary school-age children with vision impairment of statistical test used to compare group means. A Shapiro-Wilk test was performed (suitable for low sample sizes) and the distribution of number of semantic responses for both sighted and VI groups combined (Figure 2) was normally distributed (W[75] ¼ 9.68, p ¼ 0.051). A further research question was to investigate whether severity of VI affects semantic fluency per- formance. For this analysis, the participants with VI were divided into two groups according to the severity of VI and WHO classification (Solebo & Rahi, 2014). These results are presented in Table III. The results in Table III indicate that there is a trend for the children with a more severe VI to pro- duce fewer responses, in comparison to children with moderate sight impairment. However, the two groups were not well matched for age. Because the severe sight impairment group were younger than the mod- erate VI group, it might be the case that the near-sig- nificant difference in correct responses is driven by age rather than by severity of VI. We therefore reran the analysis on correct responses with age as a covari- ate. A univariate ANOVA exploring whether the type of vision impairment affected the number of correct responses confirmed a near-significant group differ- ence, F(1, 43) ¼ 3.85, p ¼ .056, gp 2 ¼ 0.82. The covariate, age, was significantly related to the number of correct responses, F(1, 42) ¼ 8.38, p ¼ .006, gp 2 ¼ 0.173. Note. at not calculated due to no out of category errors in the group with severe sight impairment and blindness. Result Rerunning the ANOVA with age as a covari- ate reduced the impact of severity of vision impair- ment on performance, F (1, 42) ¼ 1.30, p ¼ .286, gp 2 ¼ 0.089. We conclude, therefore, that the severity We therefore carried out a series of independent samples t-tests on the data2. In Table II we report the means and standard deviations for each outcome measure for the two groups. We also report the results of the t-test used to investigate whether the groups differed significantly for any of these measures. For the interpretation of effect sizes (i.e. d), we used the widely accepted interpretation of d values below 0.3 being judged small. The results of the t-tests presented in Table II indi- cate that the VI and sighted groups did not differ sig- nificantly on any of the measures (i.e. number of correct responses, number of clusters, mean cluster size, and number of switches). T-tests could not be calculated on out of category errors or repetitions, owing to sighted children not producing any errors. Furthermore, all differences between the groups had a very small effect size (0.17 and below). The two groups therefore perform similarly on the semantic fluency task. Table II. Vision impaired and sighted group means, standard deviations, independent samples t-test, and effect size (d) exploring differ- ences between vision impaired and sighted groups. red and sighted group means, standard deviations, independent samples t-test, and effect size (d) exploring differ- mpaired and sighted groups. ble II. Vision impaired and sighted group means, standard deviations, independent samples t-test, and effect size (d) exp ces between vision impaired and sighted groups. Table II. Vision impaired and sighted group means, standard deviations, independent samples t-test, and effect size (d) exploring differ- ences between vision impaired and sighted groups. Vision impairment (n ¼ 45) Sighted (n ¼ 30) t p d Number of correct responses 13.09 (5.65) 12.10 (5.42) 0.739 .462 0.17 Errors (total)a 0.24 (0.69) 0.00 (0.00) – – – Out of category errorsa 0.07 (0.33) 0.00 (0.00) – – – Repetitionsa 0.18 (.44) 0.00 (0.00) – – – Clusters Number of clusters 3.24 (1.71) 3.06 (1.69) 0.703 .484 0.17 Mean cluster size 3.83 (1.44) 4.03 (1.34) 0.358 .721 0.08 Number of switches 3.91 (2.49) 3.71 (2.52) 0.531 .597 0.13 Note. at not calculated due to no repetition/errors in sighted group. Table III. Discussion The aims of this exploratory study were to investigate the semantic fluency of sighted children and children with VI, and to investigate whether severity of VI affected task performance. The rationale behind this study was to offer insight into the potential suitability of semantic fluency—an auditory-only measure—for assessing children whose access to visual information is either compromised or absent. A further limitation of our study—which again comes as a result of drawing on data from a larger study with different aims—is that we used just one semantic category, albeit one of the most widely employed and most appropriate for this age group (animals). We therefore do not know whether the VI participants’ performance on this category is represen- tative of their performance on other categories. Nor do we know whether productivity for the animals cat- egory correlates with performance on other language assessments, including assessments of vocabulary. These are important avenues for future research. Our results suggest that the presence of a vision impairment does not necessarily impact semantic flu- ency. Patterns of retrieval were relatively consistent across both samples. Incorrect responses (out of cat- egory responses, repetitions) were minimal, indicat- ing that participants understood the task instructions. We found a wide variation in the number of correct responses produced by both the VI and sighted groups, which was not surprising given the wide age range. For both groups, the number of correct responses correlated moderately with age, indicating that the task is sensitive to development for children with VI just as it is for sighted children. Importantly, there were no differences between the groups on any of our outcome measures: The children with VI pro- duced similar numbers of clusters, clusters of similar size, and a similar number of switches compared to the sighted children, indicating similar patterns of lex- ical retrieval. These findings are consistent with the smaller scale study of semantic fluency in 16 older children with VI (11- to 18-year-olds) by Wakefield et al. (2006). Further, with respect to the severity of the vision impairment, our data showed that this fac- tor did not affect overall productivity or patterns of lexical retrieval either. This means that semantic flu- ency could be a suitable addition to the tools that speech-language pathologists use to assess vocabulary abilities in children with VI. Acknowledgements The authors thank the children and families who participated in this project. Result Split vision impairment group means, standard deviations, independent samples t-test, and effect size (d) exploring the differ- ences in severity of vision impairment. No sight impairment þ Severe sight Table III. Split vision impairment group means, standard deviations, independent samples t-test, and effect size (d) exploring the differ- ences in severity of vision impairment. Table III. Split vision impairment group means, standard deviations, independent samples t-test, and effect size (d) exploring the differ- ences in severity of vision impairment. No sight impairment þ moderate vision impairment (n ¼ 22) Severe sight impairment þ blindness (n ¼ 23) t p d Chronological age in years, mean (SD) 8.83 (1.75) 7.78 (1.83) 1.96 .056 0.59 Number of correct responses 14.73 (4.48) 11.52 (6.28) 1.98 .055 0.59 Errors (total) 0.36 (0.90) 0.13 (0.34) 1.14 .266 0.35 Out of category errorsa 0.14 (0.47) 0.00 (0.00) – – – Repetitions 0.23 (0.53) 0.13 (0.34) 0.73 .473 0.22 Clusters Number of clusters 3.64 (1.79) 2.87 (1.58) 1.53 .134 0.46 Mean cluster size 3.75 (1.46) 3.91 (1.47) 0.37 .711 0.11 Number of switches 4.32 (2.23) 3.52 (2.69) 1.08 .288 0.32 Note. at not calculated due to no out of category errors in the group with severe sight impairment and blindness. Table III. Split vision impairment group means, standard deviations, independent samples t-test, and effect size (d) ex ences in severity of vision impairment. sion impairment group means, standard deviations, independent samples t-test, and effect size (d) exploring the differ- f vision impairment. Note. at not calculated due to no out of category errors in the group with severe sight impairment and blindness. J. Hayton et al. 8 of the vision impairment does not impact semantic fluency performance. verbal had a likely language difficulty or language delay. Nor do we have information on whether partic- ipants used any languages at home other than English, and we only administered the task in English. Information about the home context of the participants may have enabled us to better under- stand media and cultural influences on semantic retrieval (e.g. McGregor et al. [2018] found retrieval of animals from the Chinese zodiac in their sample of children with a Taiwanese background). Notes 1. In total, 10 phonological clusters were coded. Within the VI group, six phonological clusters were coded, whereby four participants produced one phonological cluster and one participant produced two phonological clusters. Within the sighted group, four phonological clusters were coded, whereby two participants produced one phonological cluster and one participant produced two phonological clusters. Therefore, our findings are consistent with the literature demonstrating the rarity of phonological clusters in the semantic fluency task. However, it should be noted that our analysis was run on participants who produced at least one item in the semantic task. As mentioned in the method sec- tion, six children with VI were omitted from the ana- lysis as they were either non-verbal (three children), were not able to produce any responses to this task (two children), or required verbal reassurance from the test administrator (one child). This is arguably representative of the greater variability that is inherent in VI groups (Warren, 1994). However, a limitation of our study is that the data we drew on were part of a larger study (Hayton et al., 2021) examining sleep in children with vision impairment. We did not consider a post hoc power analysis appropriate, owing to the data having been collected via availability sampling. As language was not the focus of that larger study, we do not have any additional language measures that would help us to understand whether the participants who were unable to complete the task despite being 2. The Shapiro-Wilk test just crossed the threshold of normal distribution. To be conservative, we also ran non-parametric Mann-Whitney U tests and the pattern of results was identical. We therefore present just the results of the t-tests here. Mann Whitney U tests and the pattern of results was identical. We therefore present just the results of the t-tests here. Discussion Notwithstanding these limitations, the findings of our study offer an initial response to concerns raised by Rose et al. (2022), who called for suitable language assess- ment measures for children with VI but did not mention the semantic fluency task. Our results support the utility of this task for children with VI, despite their different sensory experiences compared to sighted children. Although this was only an exploratory study and our findings are preliminary (and need replicating in other languages and in larger and more diverse samples, employing a prospective power analysis), our findings do suggest that speech-language pathologists can be confi- dent in using semantic fluency tests with VI children. References Marshall, C. R., Jones, A., Fastelli, A., Atkinson, J., Botting, N., & Morgan, G. (2018). Semantic fluency in deaf children who use spoken and signed language, in comparison to hearing peers. International Journal of Language & Communication Disorders, 53(1), 157–170. https://doi.org/10.1111/1460-6984. 12333 Ardila, A., Ostrosky-Solıs, F., & Bernal, B. (2006). Cognitive testing toward the future: The example of Semantic Verbal Fluency (ANIMALS). International Journal of Psychology, 41(5), 324–332. https://doi.org/10.1080/00207590500345542 Beal-Alvarez, J. S., & Figueroa, D. M. (2017). Generation of signs within semantic and phonological categories: Data from deaf adults and children who use American Sign Language. Journal of Deaf Studies and Deaf Education, 22(2), 219–232. https://doi.org/10.1093/deafed/enw075 McGregor, K., Munro, N., Baker, E., Chen, S., & Oleson, J. (2018). Cultural influences on the developing semantic lexi- con. Journal of Child Language, 45(6), 1309–1336. https://doi. org/10.1017/S0305000918000211 Chami, S., Munro, N., Docking, K., McGregor, K., Arciuli, J., Baker, E., & Heard, R. (2018). Changes in semantic fluency across childhood: normative data from Australian-English speakers. International Journal of Speech-Language Pathology, 20(2), 262–273. https://doi.org/10.1080/17549507.2016. 1276214 Mcleod, S., & Mckinnon, D. H. (2007). Prevalence of commu- nication disorders compared with other learning needs in 14 500 primary and secondary school students. International Journal of Language & Communication Disorders, 42(S1), 37– 59. https://doi.org/10.1080/13682820601173262 Mengisidou, M., Marshall, C. R., & Stavrakaki, S. (2020). Semantic fluency difficulties in developmental dyslexia and developmental language disorder (DLD): poor semantic structure of the lexicon or slower retrieval processes? International Journal of Language & Communication Disorders, 55(2), 200–215. https://doi.org/10.1111/1460-6984.12512 Department for Education. (2021). Special educational needs in England. https://explore-education-statistics.service.gov.uk/ find-statistics/special-educational-needs-in-england/2020-21 Emerson, E., Robertson, J. (2011). The estimated prevalence of visual impairment among People with learning disabilities in the UK. Centre for Disability Research, School of Health & Medicine, Lancaster University report for RNIB and SeeAbility. http://www.rnib.org.uk/knowledge-and-research- hub/research-reports/prevention-sight-loss/prevalence-VI- learning-disabilities Nash, H. M., & Snowling, M. J. (2008). Semantic and phono- logical fluency in children with Down syndrome: Atypical organization of language or less efficient retrieval strategies? Cognitive Neuropsychology, 25(5), 690–703. https://doi.org/10. 1080/02643290802274064 Hayton, J., Marshall, J., & Dimitriou, D. (2021). Lights out: Examining sleep in children with vision impairment. Brain Sciences, 11(4), 421. https://doi.org/10.3390/brainsci11040421 Perez-Pereira, M., & Conti-Ramsden, G. (2013). Language devel- opment and social interaction in blind children. Psychology Press. Royal National Institute of Blind People (RNIB). (2022). Getting the right support – SEN and inclusion. https://www.rnib.org.uk/ living-with-sight-loss/education-and-learning/getting-the-right- support-sen-and-inclusion/ Henry, L. A., Messer, D. J., & Nash, G. (2015). Patient consent statement All participants and their parents gave written consent and ver- bal assent to anonymised/depersonalised data being presented in this manuscript. Keil, S. (2019). Summary of DfE official data on children and young people with vision impairment in England 2019. View briefing: Children and young people whose primary or sec- ondary SEN was vision impairment. https://viewweb.org.uk/ dfe-official-data-children-young-people-vision-impairment/ Ethical approval The project received ethical approval from UCL Institute of Education Research Ethics Committee. Semantic fluency in primary school-age children with vision impairment Semantic fluency in primary school-age children with vision impairment 9 9 Kalua, K., Patel, D., Muhit, M., & Courtright, P. (2008). Causes of blindness among children identified through village key informants in Malawi. Canadian Journal of Ophthalmology. Journal Canadien D’ophtalmologie, 43(4), 425– 427. https://doi.org/10.3129/i08-084 Data availability statement The data generated for this study will not be made publicly available. The ethics form states that only processed data will be presented in manuscripts. Data are not available to maintain anonymity and confidentiality of participants. Landau, B., & Gleitman, L. (1985). Language and experience. (Vol. 174). Harvard University Press. Lovie-Kitchin, J. E. (2015). Is it time to confine Snellen charts to the annals of history? Ophthalmic and Physiological Optics, 35(6), 631–636. https://doi.org/10.1111/opo.12252 Disclosure statement No potential conflict of interest was reported by the author(s). Koren, R., Kofman, O., & Berger, A. (2005). Analysis of word clustering in verbal fluency of school-aged children. Archives of Clinical Neuropsychology, 20(8), 1087–1104. https://doi.org/ 10.1016/j.acn.2005.06.012 Jessica Hayton http://orcid.org/0000-0001-8123-0681 Kulkarni, S., Gilbert, C., Giri, N., Hankare, P., Dole, K., & Deshpande, M. (2022). Visual impairment and blindness among children from schools for the blind in Maharashtra state, India: Changing trends over the last decade. Indian Journal of Ophthalmology, 70(2), 597–603. https://doi.org/10. 4103/ijo.IJO_1930_21 Funding The parent project (Hayton et al., 2021), under which these data were collected, was funded by the BA/Leverhulme Small Research Grant SRG18R1/181296. Kosmidis, M., Vlahou, C., Panagiotaki, P., & Kiosseoglou, G. (2004). The verbal fluency task in the Greek population: Normative data, and clustering and switching strategies. Journal of the International Neuropsychological Society : JINS, 10(2), 164–172. https://doi.org/10.1017/S1355617704102014 ORCID Jessica Hayton http://orcid.org/0000-0001-8123-0681 References Executive func- tion and verbal fluency in children with language difficulties. Learning and Instruction, 39, 137–147. https://doi.org/10. 1016/j.learninstruc.2015.06.001 Rose, S., Bates, K., & Greenaway, R. (2022). Chapter 10 Language and Communication Development. In N. Dale, A. Salt, J, Sargent & R. Greenaway (Eds), Children with Vision Impairment: Assessment, Development, and Management (pp.123–136). Mac Keith Press Practical Guides. International Classification of Diseases (ICD-11). (2022a). 09 Diseases of the visual system. https://icd.who.int/browse11/l-m/ en#/http%3a//id.who.int/icd/entity/868865918 International Classification of Diseases (ICD-11). (2022b). 9D90 Vision impairment including blindness. ICD-11 for Mortality and Morbidity Statistics. https:// icd.who.int/browse11/l-m/en#/http%3a//id.who.int/icd/ entity/1103667651 Solebo, A. L., & Rahi, J. (2014). Epidemiology, aetiology and management of visual impairment in children. Archives of Disease in Childhood, 99(4), 375–379. https://doi.org/10.1136/ archdischild-2012-303002 10 J. Hayton et al. 10 Wall, K. (2019). Habilitation and Rehabilitation. In J. Ravenscroft (Ed), The Routledge handbook of visual impairment. (pp.333–359). Routledge. Strauss, E., Sherman, E. M. S., & Spreen, O. (2006). A compen- dium of neuropsychological tests: Administration, norms, and com- mentary. Oxford University Press. Warren, D. H. (1994). Blindness and children: An individual differ- ences approach. Cambridge University Press. Troyer, A. K., Moscovitch, M., & Winocur, G. (1997). Clustering and switching as two components of verbal fluency: Evidence from younger and older healthy adults. Neuropsychology, 11(1), 138–146. https://doi.org/10.1037//0894-4105.11.1.138 World Health Organization (WHO). (2022). October 13. Blindness and vision impairment. https://www.who.int/news- room/fact-sheets/detail/blindness-and-visual-impairment Wakefield, C. E., Homewood, J., & Taylor, A. J. (2006). Early blindness may be associated with changes in performance on verbal fluency tasks. Journal of Visual Impairment & Blindness, 100(5), 306–310. https://doi.org/10.1177/0145482X0610000508 Wojcik, E. H., & Saffran, J. R. (2015). Toddlers encode similar- ities among novel words from meaningful sentences. Cognition, 138, 10–20. https://doi.org/10.1016/j.cognition.2015.01.015
https://openalex.org/W4394838563	http://fis.uni-bamberg.de/bitstreams/e6db2351-8814-463d-91e4-d9e02665b15b/download	English	null	Blaming the Mother : New Momism and Failed Matriarchy in Gillian Flynn’s Sharp Objects	Bamberger Studien zu Literatur, Kultur und Medien	2,024	cc-by	8,107	In: Kerstin-Anja Münderlein (Ed.), Crime Fiction, Femininities and Masculinities : Proceedings of the Eighth Captivating Criminality Conference, Bamberg : University of Bamberg Press, 16p.. 2024. DOI: 10.20378/irb-92502 Bookpart - Published Version DOI of the Article: Date of Publication: 15.04.2024 10.20378/irb-94615 1 University of Bamberg Press 1 University of Bamberg Press Blaming the Mother : New Momism and Failed Matriarchy in Gillian Flynn’s Sharp Objects In: Kerstin-Anja Münderlein (Ed.), Crime Fiction, Femininities and Masculinities : Proceedings of the Eighth Captivating Criminality Conference, Bamberg : University of Bamberg Press, 16p.. 2024. DOI: 10.20378/irb-92502 Legal Notice: This work is protected by copyright and/or the indication of a licence. You are free to use this work in any way permitted by the copyright and/or the licence that applies to your usage. For other uses, you must obtain permission from the rights-holder(s). This document is made available under the Creative Commons Lizense CC BY. This licencse information is available online: (D https://creativecommons.org/licenses/by/4.0/ Keywords Motherhood, Mother Blame, New Momism, patriarchal ideology, matriarchy, Mun chausen Syndrome by Proxy, domestic noir Abstract This chapter investigates the representation of motherhood in Gillian Flynn’s novel Sharp Objects (2006). The novel’s focus on the domestic and maternal illustrates its participation in contemporary thematic trends in two broader genres: crime fiction and domestic noir. Based on a close reading of the novel’s representation of moth erhood, this chapter argues that Sharp Objects reinforces patriarchal ideology on two levels. Firstly, the novel echoes the cultural myth of Mother Blame by creating a genealogy of perverted motherhood in the family of protagonist and first-person nar rator Camille, connecting them through Munchausen Syndrome by Proxy as a plot device. In so doing, it also reinforces the essentialist gender ideology of New Momism, coined by Douglas and Michaels, because motherhood is perverted either through neglect or ‘over’-nurture. Secondly, the novel implies that these perversions of motherhood are connected to a ‘lack’ of patriarchal order by representing Ca mille’s family as matriarchal, and therefore ‘toxic,’ because fathers are absent and passive. The chapter shows how the novel meaningfully contrasts Camille’s biolog ical, matriarchal family with an idealized example of a nuclear, patriarchal family particularly towards its end: because Camille experiences a significant improvement of her mental health while staying with the nuclear, patriarchal family, Sharp Objects thereby valorizes this family unit, suggesting that motherhood becomes perverted and harmful for children especially within familial structures that ‘lack’ a patriarchal order. Blaming the Mother: New Momism and Failed Matriarchy in Gillian Flynn’s Sharp Objects Janine Schwarz, Eberhard Karls Universität Tübingen y p j Janine Schwarz, Eberhard Karls Universität Tübingen Introduction Towards the end of Sharp Objects, one of the novel’s mother figures, named Adora, tells a story about her mother Joya to her adult daughter Camille: “When I was a little girl, my mother took me into the North Woods and left me,” Adora said. “She didn’t seem angry or upset. Indifferent. Almost bored. She didn’t explain why. She didn’t say a word to me, in fact. Just told me to get in the car. I was barefoot. When we got there, she took me by the hand and very efficiently pulled me along the trail, then off the trail, then dropped my hand and told me not to follow her. I was eight, just a small thing. My feet were ripped into strips by the time I got home, and she just looked up at me from the evening paper, and went to her room.” (304-05) Blaming the Mother When Camille asks Adora why she is telling her this story, Adora responds: “‘When a child knows that young that her mother doesn’t care for her, bad things happen’” (305). This is a key moment in the novel for two reasons. First, Adora’s story and evaluation of it clearly echo the concept of Mother Blame. According to this concept, a child’s development is dependent primarily on their mother and, thereby, any ‘de fects’ in children are the fault of mothers alone for failing to conform to societal ideals of mothering. Adora’s story does exactly that: It characterizes Joya and her performance of motherhood as not conforming to essentialist ideals of it as inher ently nurturing and loving. In the story, Joya is, after all, “[i]ndifferent” to Adora, “[a]lmost bored” (304). Adora’s suggestion that “bad things happen” when a child does not receive love from their mother then anticipates her own version of moth ering, although Adora herself does not evaluate it as bad (305): Because Joya did not provide loving care, Adora developed an obsessive need to care for her daughters. She even explains this to Camille once she is done with the story: “I wanted to love you, Camille. But you were so hard . . . . Let me take care of you, Camille. Just once, need me” (305). Second, Adora’s story also reproduces the concept of Mother Blame in the sense that she holds Joya responsible for her own non-conformity to essentialist ideals of moth erhood. 1 Although the term MSbP is no longer in usage in medical discourse, I am adopting it here because it is employed in the novel’s story world. The Diagnostic and Statistical Manual of Mental Disorders (2022) terms the disorder not MSbP but “Factitious Disorder Imposed on Another (Previously Factitious Disorder by Proxy)” and defines it as the “Falsification of physical or psychological signs or symptoms, or induction of injury or disease, in another, associated with identified perception” (367). The diagnosed individual thus “presents another individual (victim) to others as ill, impaired, or injured” (367). Importantly, then, “the perpetrator, not the victim, is given the diagnosis,” whereas “the victim may be given an abuse diagnosis” (368). Introduction A few pages before Adora tells the story, Camille is at the local hospital to investigate the death of her younger half-sister Marian. Before Marian’s death, Adora used to frequent that hospital with her daughter because Marian was often sick. There, however, Camille talks to a nurse who had filed a report years earlier in which she describes her suspicion that Adora suffers from Munchausen Syndrome by Proxy (MSbP) and that Marian’s symptoms of illness are induced by Adora. The nurse even provides a, if somewhat biased, definition of it: “The caregiver, usually the mother, almost always the mother, makes her child ill to get attention for herself” (293; italics in original). She continues: “You got MBP, you make your child sick to show what a kind, doting mommy you are” (293).1 When Camille learns about her mother’s illness, she also concludes that Adora must have poisoned Marian for years, killing her eventually when Camille was thirteen. Just like her mother Joya, then, Adora does not conform to essentialist ideals of motherhood and, in fact, per verts motherhood through ‘over’-nurture that results in murder. Whether Adora herself is aware that she suffers from MSbP is unclear, but it is also not relevant. 57 Janine Schwarz Janine Schwarz What is much more relevant is the causality Adora and her story imply: By juxtapos ing Joya’s and her own version of mothering, Adora implies her version of mother ing to be reactive to her mother’s and thereby reproduces the concept of Mother Blame. Such cautionary examples of ‘bad’ mothers who are allegedly responsible for defects and developmental issues in children are ubiquitous in Sharp Objects. And, as I will show, they are always placed within family structures that suggest a matriarchal hi erarchy. I argue here that, through this interplay of representing motherhood as perverted and as originating from matriarchal family structures, the novel eventually valorizes the patriarchal, nuclear family and thereby patriarchal ideology. To show this, I first demonstrate how the novel represents motherhood as perverted either through neglect or ‘over’-nurture in the form of MSbP. The novel thus perpetuates a narrow and essentialist ideal of motherhood that is similar to the one propagated by what Susan J. Douglas and Meredith W. Michaels have called “New Momism,” a gender ideology they claim emerged in the 1980s (4). Introduction In fact, with the onset of the twenty-first century, the crime novel has seen a growing interest in the domestic, not only as a geographical location but also as a spatial meaning-making entity regarding gender roles and motherhood. Nowhere does this become clearer than in the name ‘domestic noir,’ a currently immensely popular subgenre in crime fiction, whose beginnings are commonly traced back to Gillian Flynn’s 2012-bestseller Gone Girl (Joyce 4). Scholars of crime fiction locate one of the reasons for this contemporaneous interest in the domestic in the rise of neoliberalism, that is, the expansion of political and economic thinking into all public and private spheres governed by an emphasis on, i.a., privatization, individualism, and entrepreneurialism. Ruth Cain, for example, argues that the dominant neoliberal ethos of individualism and the growing privatization of family life lend themselves to a dramatization of “crucial conflicts of neoliberal maternal life” in the genre of domestic noir (290). While I concur with this contextual reading of domestic noir texts that focus on the maternal, neoliberalism as a concept does not explicitly factor into my interpretation here for reasons of brevity. A more holis tic interpretation of motherhood in Sharp Objects could include pondering the ques tion of how neoliberal contexts influence the understanding of this text’s represen tation of motherhood. Sharp Objects and New Momism: A Genealogy of Perverted Motherhood The genealogy of perverted motherhood, as mentioned in the introduction, reaches over three generations in the family of protagonist Camille. At the same time, it intensifies from generation to generation. This section unravels this genealogy and shows how the novel’s three central mother figures – Camille’s grandmother Joya, her mother Adora, and her half-sister Amma – pervert motherhood through either neglect or ‘over’-nurture. In the novel’s present, Camille travels from Chicago to her hometown Wind Gap as a journalist to investigate the murders of the local teenagers Ann and Natalie. Because Camille stays at her mother Adora’s mansion, she must confront memories of her childhood and her deceased half-sister Marian. As ex plained earlier, she learns towards the novel’s end that Adora, suffering from MSbP, killed Marian. While the novel conditions readers for a long while to assume that Adora has killed Ann and Natalie, too, they learn in a final plot twist that it was Camille’s thirteen-year-old half-sister Amma. Introduction By employing MSbP in the first place, the novel also echoes the concept of Mother Blame and creates a geneal ogy of perverted motherhood within Camille’s family that intensifies from genera tion to generation. In a second step, I show that Sharp Objects and particularly its ending further valorize the patriarchal, nuclear family by contrasting Camille’s bio logical family with her boss and his wife. While the former is represented as a ‘toxic’ and matriarchal family because fathers are absent either literally or figuratively by being passive and motherhood is perverted, the second family unit represents the ideal of the patriarchal, nuclear family. Importantly, Camille can improve her men tal health only within the latter. The novel thereby implies that motherhood becomes perverted and harmful for children especially within familial structures that ‘lack’ a patriarchal order. While I concur with interpretations of Sharp Objects that point out the novel’s sub versive potential regarding gender roles, I argue that this potential is not fully real ized. Alyson Miller, for example, claims that the novel negotiates cultural anxieties about the “monstrous mother” who poses a threat to the patriarchal system by not conforming to hegemonic ideals of motherhood (495). Through its insistence on the superiority of the nuclear, patriarchal family at the end, however, I argue that Sharp Objects not only dramatizes patriarchal anxieties, but also reinforces patriarchal ide ologies. Similarly, Malinda K. Hackett interprets the female characters in the novel as “creat[ing] a visual rhetoric that disrupts idealized versions of femaleness” (81). According to her, the women “partake in performative acts of pathology to reject notions of idealized femininity, beauty, and motherhood” (80). This is true: By per verting motherhood in the first place, the novel reconceptualizes and, to some ex tent, challenges essentialist conceptions of it as inherently nurturing. However, I contend that this critique has its limits because other essentialist and patriarchal 58 Blaming the Mother ideologies, most importantly those of New Momism and Mother Blame, are perpet uated at the same time. Such ideological tensions and contradictions regarding gender are ubiquitous in contemporary crime novels that focus on the domestic – and Sharp Objects clearly is part of this trend, too. Introduction Although Amma is not a mother bio logically, the way she kills Ann and Natalie echoes Adora’s perverted performance of motherhood and murder of Marian, and she thus appropriates motherhood 59 Janine Schwarz through these murders. As such, Amma stays true to her telling name, itself an an agram of ‘Mama.’ Through these perverted performances of motherhood, I argue, Sharp Objects per petuates a motherhood ideal as promulgated by the gender ideology of New Momism. Coined by Douglas and Michaels in 2004, New Momism prescribes heg emonic ideals of motherhood that are based on an essentialist perception of gender and rests on three core beliefs: It insists “that no woman is truly complete or fulfilled unless she has kids, that women remain the best primary caretakers of children, and that to be a remotely decent mother, a woman has to devote her entire physical, psychological, emotional, and intellectual being, 24/7, to her children” (Douglas and Michaels 4). In so doing, the ideology promotes a mothering ideal coined “intensive mothering” by Sharon Hays, requiring women to define themselves exclusively through their role as mother, to “expend a tremendous amount of time, energy, and money in raising their children,” and to exhibit “unselfish nurturing” while doing so (Hays x). These markers illustrate that New Momism is a distinctly narrow ideal: After all, it also “insist[s] that mothers perfectly regulate their behavior, that they never become over-invested in their kids but never seem underinvested, either” (Douglas and Michaels 152). Ultimately, women must walk a fine line to be deemed a good mother: They must be caring, but not too caring, and their mothering must be, above all, selfless. What New Momism’s narrow motherhood ideal reveals, then, is the ideology’s pro pensity to reinforce the concept of Mother Blame. The ideology suggests that a mother’s failure to conform to the ideal results in long-term damage in her children. It implies that “[i]f mothers screw up” by failing to perform motherhood according to the three core beliefs named earlier, the result would be “permanent psychological and/or physical damage” (6). This causality is also evidenced by a case study con ducted by Douglas and Michaels in which they analyze news stories from the late 1980s until the mid-90s which put forth the narrative that the largest threat for chil dren comes “from mom herself” (140). Introduction They show that, within that period, the me dia “built an interlocking, cumulative image of the dedicated, doting ‘mom’ versus the delinquent, bad ‘mother’” (7). Mother Blame thus is a central myth propagated by New Momism. As I have already detailed in the introduction, Joya deviates from the New Momism ideal through her neglectful and loveless mothering. In addition to how Adora’s story, quoted in the introduction, depicts her mother, Adora also tells Camille that Joya “never loved [her]” and characterizes her mother as “cold and distant” (190). Other characters in the novel describe Adora and Joya’s relationship as similarly lacking affection. For example, one of Adora’s childhood friends explains to Camille that Joya never “smile[d] at [Adora] or touch[ed] her in a loving way” (258). Instead of such “loving” touches, she recalls that Joya treated Adora like an object at her 60 Blaming the Mother convenience. When Adora needed some form of nurture, Joya seemed to hierarchize her own pleasure. For example, “[w]hen Adora peeled from a sunburn . . . Joya would sit next to [her], strip off her shirt, and peel the skin off in long strips” (258). And, most importantly: “Joya loved that,” meaning the pleasure Adora’s body could afford her but not Adora herself (258; emphasis added). Joya’s performance of motherhood as loveless thus perverts New Momism’s essentialist conception of it. Importantly, the same childhood friend who describes Joya and Adora’s relationship as lacking love also contends that “Adora was . . . overly mothered” by Joya (258). This suggests that Joya, too, may have suffered from MSbP. After all, the same child hood friend also remembers that “‘Adora was sick all the time. She was always hav ing tubes and needles and such stuck in her’” (259). The novel thereby draws a con nection between Adora’s illnesses as a child and those of her daughter Marian, and this connection is reinforced by using similar imagery. An image similar to the “tubes and needles” stuck in Adora also appears in one of Camille’s childhood mem ories of her sister: “Marian was a sweet series of diseases. She had trouble breathing from the start,” Camille narrates (96-97). “Later she had trouble digesting and sat murmuring to her dolls in a hospital bed set up in her room, while [Adora] poured sustenance into her through IVs and feeding tubes” (97; emphasis added). Introduction Particularly the image that Camille’s narration invokes in the reader’s mind of a young, drug abusing mother who forgot about her children for a quick drug fix establishes that Camille is speaking from an assumed moral high ground. The description of Tammy’s “pink rouge on her cheeks in two perfect circles the size of shot glasses” only broadens the notion of drug abuse through the reference to alcohol. At the same time, Tammy’s neglect of her own appearance – suggested by her ‘bad’ make-up application and because Camille describes her as “fat” – implies a similar neglect towards her children. Tammy’s ‘fatness’ furthermore characterizes her as selfish in Camille’s eyes: while her children are “flurrying like chickens” over the inadequate food Tammy left for them, her body bears signs of overeating. Camille’s judgment becomes more explicit later when she describes having had a dream about “the item on miserable Tammy Davis and her four locked-up children” (33; emphases added). That Camille judges Tammy is clear because she introduces the story to readers as a “limp sort of evil” in the first place. Although Camille concedes that Tammy had “just forgotten” her kids and “[s]ometimes that happens,” her initial description of the story as “evil” implies a degree of intentionality on Tammy’s part. Particularly the image that Camille’s narration invokes in the reader’s mind of a young, drug abusing mother who forgot about her children for a quick drug fix establishes that Camille is speaking from an assumed moral high ground. The description of Tammy’s “pink rouge on her cheeks in two perfect circles the size of shot glasses” only broadens the notion of drug abuse through the reference to alcohol. At the same time, Tammy’s neglect of her own appearance – suggested by her ‘bad’ make-up application and because Camille describes her as “fat” – implies a similar neglect towards her children. Tammy’s ‘fatness’ furthermore characterizes her as selfish in Camille’s eyes: while her children are “flurrying like chickens” over the inadequate food Tammy left for them, her body bears signs of overeating. Camille’s judgment becomes more explicit later when she describes having had a dream about “the item on miserable Tammy Davis and her four locked-up children” (33; emphases added). This introduction of ‘bad’ mothering via Tammy conditions readers to perceive ne glectful mothering as most harmful for children through Camille’s narration. Introduction Through this, the novel equates Adora’s and Marian’s experiences, implying that Adora may have been the victim of MSbP as well. In the end, whether Joya really had MSbP or not is not relevant. What is important is that the implication of her potentially having suffered from it and then having passed it on to her daughter Adora is there, and the novel thereby clearly echoes the concept of Mother Blame. Although my focus here is on the mother figures in Camille’s family and how they pervert motherhood ideals, various other ‘bad’ mothers appear in the novel, adding to its perpetuation of New Momism’s narrow motherhood ideal. As the first-person narrator in Sharp Objects, Camille functions to connect these examples of ‘bad’ motherhood through her judgmental narration. In fact, the first quarter of Sharp Objects functions as a general introduction to the theme of motherhood with a focus on ‘bad’ mothers who deviate from the ideal by being neglectful. One of these exam ples appears on the novel’s first page. Working as a journalist, and before traveling from Chicago to Wind Gap, Camille tells the story of Tammy Davis whose version of motherhood she clearly evaluates as ‘bad’: My story for the day was a limp sort of evil. Four kids, ages two through six, were found locked in a room on the South Side with a couple of tuna sandwiches and a quart of milk. They’d been left three days, flurrying like chickens over the food and feces on the carpet. Their mother had wandered off for a suck on the pipe and just forgotten. Sometimes that happens. No cigarette burns, no bone snaps. Just an irretrievable slipping. I’d seen the mother after the arrest: twenty-two-year-old Tammy Davis, blonde and fat, with pink rouge on her cheeks in two perfect circles the size of shot glasses. (1) 61 Janine Schwarz That Camille judges Tammy is clear because she introduces the story to readers as a “limp sort of evil” in the first place. Although Camille concedes that Tammy had “just forgotten” her kids and “[s]ometimes that happens,” her initial description of the story as “evil” implies a degree of intentionality on Tammy’s part. Introduction In this context, Adora’s performance of motherhood becomes significant: her version of mothering is introduced to readers as diametrically opposed to Tammy’s, the ‘bad’ mother, because she is hyper-focused on nurturing and her mother role, among other things. Adora thus fulfills, or at least seems to fulfill, the ideal of “intensive mothering” purported by New Momism (Hays x). Importantly, though, Camille’s narration still conditions readers to perceive even Adora’s version of mothering as ‘bad’ by presenting it as peculiar and inappropriate, on the one hand, and by estab lishing that it continues the genealogy of perverted motherhood in the family initi ated by Joya, on the other hand. The peculiarity and inappropriateness of Adora’s performance of motherhood is most obvious in her tendency to ‘babyfy’ her thirteen-year-old daughter Amma. This is exemplified best by the first representation of Adora in her role as mother that the reader encounters. A few days into Camille staying at her mother’s house, she nar rates the following episode: “Amma and my mother sat on the couch, my mother cradling Amma – in a woolen nightgown despite the heat – as she held an ice cube to her lips. My half-sister stared up at me with blank contentment, then went back to playing with a glowing mahogany table” for her dollhouse (Flynn 73). Although Amma is thirteen years old at the time, Adora still “cradl[es]” her daughter like a baby, something that seems even more peculiar when considering the size of Amma’s body at that age. That Amma is playing with furniture for a dollhouse only intensifies the peculiarity. Like Camille’s judgmental description of Tammy earlier, the way she narrates this scene between Adora and Amma as peculiar achieves a 62 Blaming the Mother similar effect in the reader, namely, to judge Adora’s performance of motherhood as undesirable and transgressive. This becomes clearest in the way she introduces the episode to readers: “The scene was startling, it was so much like the old days with Marian” (73). As such, the “scene” not only functions to condition readers to evaluate Adora’s version of mothering as “startling” and, therefore, peculiar, but also to foreshadow her suffering from MSbP by comparing the episode to Camille’s memories of her deceased half-sister Marian. The scene furthermore exemplifies how Adora perverts motherhood through condi tional motherly love. 2 It is meaningful that Adora refers to Marian as “a doll” when she praises her performance of daughterhood. In fact, Adora not only babyfies her daughters but also ‘dollifies’ them, thereby imposing specific gendered performances onto them and effecting, as Miller argues, “a static vision in which girls are little more than objects of beauty and play” (497). Although Adora’s habit Introduction In fact, as detailed in the introduction already, Adora herself suggests that because her own mother Joya neglected her, she developed a reaction ary version of mothering that is conditional on her daughters’ readiness to accept her love. Adora even admits this to Camille: “‘You remind me of my mother. Joya. . . . My mother never loved me, either. And if you girls won’t love me, I won’t love you’” (190). Adora thereby also stays true to her telling name: She wants to be loved and ‘adored’ by her daughters, and only then does she love them back. What is more, Adora’s conditional motherly love and her habit of babyfying Amma are linked, showcasing her insistence that her daughters perform their role in align ment with her performance of motherhood. The scene quoted earlier where Adora “cradl[es]” Amma exemplifies that Adora and Amma’s respective performances of mother- and daughterhood are reciprocal: Adora nurtures and Amma happily and passively receives her mother’s care, evidenced by her “star[ing] up at Camille with blank contentment” (73). Adora’s children thus only receive her love if they readily accept her mothering and thereby perform daughterhood in alignment with her role as mother. Camille, for example, remembers that she often refused her mother’s ‘care’ in the form of medicine, or rather poison; her half-sister Marian, by contrast, did not. In the novel’s present, Camille realizes that this made her lose Adora’s “full attention as a mother” (74). This is even corroborated by Adora herself. In a diary entry from 1982, six years before Marian’s death, she explains: “I’ve decided today to stop caring for Camille and focus on Marian. Camille has never become a good patient,” which is, of course, Adora’s way of expressing that Camille has never been a ‘good daughter’ (309). In fact, Adora’s identity as a mother is dependent on her daughters being both: daughters and patients. This becomes clear when Adora con tinues explaining that “‘Camille doesn’t like me to touch her. I’ve never heard of such a thing. . . . I hate her’” (309). Camille’s half-sister Marian, on the other hand, is a good patient/daughter: “Marian is such a doll when she’s ill, she dotes on me terribly and wants me with her all the time. I love wiping away her tears” (309).2 63 Janine Schwarz Adora’s and Camille’s relationship thus exemplifies Adora’s conditional motherly love. Introduction In the scene quoted earlier in which Adora “cradles” Amma, for example, Adora ignores Camille consciously precisely because Camille does not conform to her mother’s insistence on an idealized performance of daughterhood. Camille ex plains that she “was about to linger near [her] mother, waiting for her to put an arm around [her], too” (74). Adora, however, does not show any reaction: “My mother and Amma said nothing. My mother didn’t even look up at me, just nuzzled Amma in closer to her, and cooed into her ear” (74). While this description of Adora might be read as her conforming to motherhood ideals because she is seemingly so in vested in her role as mother that she forgets her surroundings, it must be read in stead as a conscious choice by Adora to ignore Camille because Camille has never performed femininity in alignment with her performance of motherhood. In fact, Camille’s stepfather Alan acknowledges and even speaks to Camille in the same scene, meaning that Adora not noticing Camille’s arrival is even more unlikely. On a broader level, Adora’s insistence that her daughters match her performance of motherhood relates to further, collectively perceived gendered ideals in Camille’s hometown, Wind Gap. According to Miller, because the novel is set in Missouri, this local ideal of femininity is highly influenced by “the social rules of Southern eti quette” which insist on “feminine behaviors that fulfil gendered stereotypes of pas sivity, silence, and compliance” (495). Adora herself regards these ideals highly in her daughters and attempts to enforce them through her mothering. Therefore, her loveless relationship with Camille must also be understood as Adora despising Ca mille’s infringement of these local ideals of femininity: Precisely because Camille does not perform her role as daughter by passively accepting her mother’s nurture, she fell out of favor with Adora. Adora’s conditional motherly love is thus dependent on her daughters not only matching her performance of motherhood but also con forming to the local ideal of femininity. The genealogy of perverted motherhood then finds its climax in Adora’s daughter and Camille’s half-sister Amma. Her murders of the local teenagers Ann and Natalie must be read as a climactic doubling of Adora’s actions and recall both Adora’s and Wind Gapians’ insistence on idealized femininity. The novel introduces Ann and Natalie to readers as violating these ideals, thereby equating them. of dollifying her daughters is another example of how she perverts motherhood, I am not further elaborating on this here for reasons of brevity. Introduction Ann’s father, for example, tells Camille a story of Ann rejecting markers of idealized femininity: “My wife wanted to put her hair in rollers the night before school photos. Ann chopped it off instead. She was a willful thing. A tomboy” (26). Similarly, Natalie’s brother describes Natalie as “kind of a tomboy,” too (160). The girls are also equated through their propensity for violence. According to stories by police chief Vickery and other 64 Blaming the Mother locals, “[t]he girls weren’t friends. Their only connection was their shared vicious ness” (90). Vickery, for example, tells Camille that Ann “had killed a neighbor’s pet bird with a stick” and Natalie and her family had “moved here [to Wind Gap] two years ago because she stabbed one of her classmates in the eye with a pair of scis sors” (57). He even agrees with Camille’s sarcastic comment that everyone in a small town like Wind Gap “knows who the bad seeds are” (57). Vickery’s agreement sug gests a communal perception of the girls as “bad” and highlights that these ideals of femininity are not just reinforced by Adora or a few locals but rather by the commu nity generally. Considering both Ann and Natalie’s obvious rejection of idealized femininity, then, Amma murdering them functions as a symbolic correction of the other, ‘not-ideal’ feminine. This is evidenced by Amma’s modus operandi both before and after the killings which imposes stereotypical feminine beautification onto them. According to the FBI agent supporting the police investigation, “Natalie’s fingernails were painted a bright pink. Ann’s legs were shaved. They both had lipstick applied at some point” (299). In so doing, Amma exaggerates Adora’s perverted performance of motherhood to an extreme: While Adora ‘nurtures’ only those daughters who conform to idealized femininity, Amma violently imposes those ideals onto Ann and Natalie. As Miller argues, Amma’s murders enforce “an ideological framework” of ideal femininity by “ironically expos[ing] an insistence on the articulation of femi nine behaviors that fulfil gendered stereotypes” (495). 3 That Amma pulls the girls’ teeth also recalls Adora’s habit of dollifying her daughters – which I mentioned in the previous footnote – albeit in an exaggerated way. Camille’s description of Natalie’s corpse corroborates this: With her teeth pulled, Natalie’s “lips [cave] in around her gums in a small circle. She look[s] like a plastic baby-doll, the kind with the built-in hole for bottle feedings” (35). Introduction What is more, Amma’s modus operandi after the murders recalls the babyfying she experiences at the hands of Adora in a much more perverted way: By pulling the girls’ teeth, Amma symbolically babyfies Ann and Natalie and violently enforces her self as their mother figure.3 When Camille finds Natalie’s corpse in the middle of town, the way Amma has placed her body functions as a symbolic reminder of how to perform femininity ‘correctly’: Her body is “aimed out at the sidewalk,” visible to passengers, and she is “[w]edged in the foot-wide space between the hardware store and the beauty parlor” (34), two places denoting stereotypical masculinity and fem ininity. This placement comments on Natalie’s liminal position as “tomboy,” on the one hand. On the other hand, it reflects Amma’s criticism of exactly that: Natalie’s gendered performance is liminal, a death sentence in a town that, according to Ca mille, “demands utmost femininity in its fairer sex” (17). As I have shown, Sharp Objects establishes a genealogy of perverted motherhood that intensifies from generation to generation. While the novel only implies that Joya 65 Janine Schwarz may have suffered from MSbP, a nurse diagnoses Adora within the story world. Moreover, it is never clear whether Joya had any other children apart from Adora, but she clearly never killed Adora – although Joya abandoning Adora in the forest could be interpreted as attempted murder. Adora, on the other hand, has killed Mar ian. Joya and Adora’s perversions of motherhood thus focus on their biological chil dren. Amma, by contrast, and of course this is also because she is only thirteen, kills children who are not her biological children, imposing her role as mother onto them violently. Through this genealogy of perverted motherhood within Joya, Adora, and Amma, the novel suggests that their versions of motherhood are causally connected. In so doing, Sharp Objects perpetuates the myth of Mother Blame, also one of the central beliefs of the New Momism ideology, and thereby also supports the patriar chal system within which this ideology and myth are fostered in the first place. Sharp Objects and Failed Matriarchy: Valorizing the Patriarchal Family Sharp Objects and Failed Matriarchy: Valorizing the Patriarchal Family Apart from its propagation of Mother Blame and New Momism, Sharp Objects fur thermore reinforces patriarchal ideology because it valorizes the patriarchal, nuclear family both throughout the novel and particularly through its ending. Tellingly, the novel’s tagline on its front cover already suggests with a pun that non-patriarchal family structures, such as matriarchally hierarchized ones, are a breeding ground for toxicity, to use the terminology in the tagline. Referring to Camille’s family, it declares that “[t]his family isn’t nuclear, it’s toxic.” Importantly, the previously ana lyzed perversions of motherhood in Camille’s family – the toxicity to which the ta gline refers – originate from family structures that ‘lack’ a patriarchal order. Sharp Objects thus implies that matriarchally structured family units are a ‘toxic,’ un healthy breeding ground for perverted motherhood because they cannot produce children without developmental issues. In so doing, the novel ultimately valorizes the patriarchal, nuclear family by representing it as the only family unit able to do so instead. The ‘lack’ of patriarchal order within Camille’s family, just like perverted mother hood, is multi-generational, and reaches across two generations. Camille’s family is matriarchally structured because, on the one hand, husbands and fathers are absent either literally – such as Camille’s biological father about whom she “knows almost nothing” (96) – or figuratively because they are quiet and passive. Camille’s grand father, for example, was so passive that a family friend “do[es]n’t even remember his name. Herbert? Herman? He was never around, and when he was, he was just quiet and… away” (259). The same family friend also equates both father figures in Ca mille’s family by describing them as similarly passive and by explaining that Ca mille’s grandfather was just “like Alan” (259), Marian and Amma’s biological father and Camille’s stepfather. On the other hand, the mother figures in Camille’s family, 66 Blaming the Mother particularly Adora, hold more power than the father figures. Early in the novel, Ca mille explains that Adora insisted that she “was to be considered Alan’s child” (96). Both Camille and Alan lived by that rule, negotiating Adora’s position of power in the family. As a father figure, however, Alan was only a façade, a means to portray a public image of an idyllic nuclear, patriarchal family. Sharp Objects and Failed Matriarchy: Valorizing the Patriarchal Family Camille remembers that she was “never really fathered by him, never encouraged to call him anything but his proper name” (96). Instead, Camille explains that “Adora prefers [them] to feel like strangers” and “wants all relationships in the house to run through her” (96), further cementing her position of power within the family. Adora’s powerful position as matriarch is substantiated by Camille’s characterization of Alan, which underlines not only his passivity but also his subordination to Adora: Alan is “a ribbon-winning equestrian who doesn’t ride anymore because it makes Adora nervous” and who “seems content to let my mother do most of the talking” (96). Evidently, then, the family structures within which perverted motherhood is fostered in Sharp Objects are distinctly non-patriarchal. In this context, Camille’s relationship to her boss at the newspaper in Chicago, Frank Curry, referred to as Curry, and his wife Eileen is meaningful. As a family unit, they function as a foil to Camille’s biological family and represent the ideal of a nuclear and patriarchally structured family. Although Curry and Eileen seemingly deviate from this ideal because they have no children, Camille’s relationship to them mirrors that of a child-parent relationship. Importantly, Camille assumes that Curry and Eileen are not without children by choice. Instead, she suspects that “they’d been unable to conceive” because they married late (100). Unlike her biological fam ily, they provide comfort to her, Camille phones them regularly for advice, and es pecially Eileen has a soothing effect on her. Eileen conforms to essentialist ideals of femininity and, by extension, motherhood, although she is not a mother per se. When Camille phones the Currys one time and Eileen picks up, for example, Ca mille describes her voice as “soft and steady as a hill” (295). Another time, “Eileen was warm when she answered the phone, which was what [Camille] needed” while staying at her mother’s house (101). Eileen’s warmth thus contrasts with Adora’s coldness, a characteristic Camille even points out to Adora: “I never felt anything but coldness from you” (190). Ultimately, Eileen is characterized to readers as par ticularly nurturing, and she thus also functions as a foil to Adora whose conditional motherly love deviates from idealized motherhood. Sharp Objects and Failed Matriarchy: Valorizing the Patriarchal Family What is more, Eileen and Curry represent the ideal of a nuclear, patriarchal family because they perform their gendered roles within this unit in a way that matches idealized and essentialist conceptions of gender. As I have shown, it is particularly Eileen’s warmth and soothing effect on Camille that mark her as conforming to ide alized woman- and motherhood. Curry is also nurturing towards Camille, not only 67 Janine Schwarz in his roles as her boss and professional mentor but also as a substitute father fig ure.4 Importantly, however, he retains a relative distance as well as a gender and age hierarchy and thereby performs his role in a way that matches essentialist and patri archal conceptions of it. This is exemplified by Curry’s nicknames for Camille. He refers to Camille either as “kiddo” (2, 6) or “my girl” (80, 101), two nicknames which invoke a stereotypical father-daughter relationship, or as “Cubby” (102, 215, 216, 296), an endearment of “cub reporter” (102). Curry’s double role of professional mentor and father figure is also implied in the nickname “Cubby”: it is both close to his own name, ‘Curry,’ and, at the same time, marks Camille as a “cub,” that is, ‘young reporter.’ What is more, by calling Camille “kiddo,” “girl,” and “cub,” Curry clearly establishes a hierarchy between the two based on their age difference, thereby retaining distance while expressing his fondness for Camille at the same time. Another instance towards the novel’s end furthermore exemplifies both Curry and Eileen’s adherence to traditional gender roles in a rescue scene Camille imagines. This imagined scene highlights not only that Curry and Eileen represent the ideal of a nuclear, patriarchal family but also that Camille obviously yearns to be part of such a family unit precisely because she imagines them as conforming to that ideal in the first place. By the time in the novel when Camille imagines the scene, she has already grown increasingly suspicious and afraid of Adora, suspecting that she has killed her half-sister Marian and possibly also the local teenagers Ann and Natalie. Paralyzed by fear, with her “hands . . . sweating” while her mother is outside her room, Camille turns her mind to the Currys for comfort and imagines them rescu ing her: “I had a flash of Curry, one of his crappy ties swinging wildly over his belly, bursting into the room to save me. 4 Although I stated in the introduction that neoliberalism does not factor into my interpretation, it is worth noting that Curry strikingly functions to connect neoliberalism and the domestic by being both Camille’s professional mentor and a father figure for her. Through this, Sharp Objects links neoliberalism and patriarchy and equates them as positive systems. Blaming the Mother Ann, and Natalie, assuming they are connected. Camille becomes Amma’s guard ian, and, together, they return to Chicago. There, Amma kills another teenager with the same modus operandi, thereby acquitting Adora of killing Ann and Natalie but not of killing Marian. Eventually, Amma is found guilty of murdering Ann, Natalie, and the teen in Chicago. Once both Adora and Amma are incarcerated, Curry and Eileen take Camille into their home. Now, their formerly childless family is no longer childless because Camille functions as their substitute child: Curry and Eileen packed my things and took me to their home, where I have a bed and some space in what was once a basement rec room. . . . I am learning to be cared for. I am learning to be parented. Eileen and Curry wake me in the morning and put me to bed with kisses (or in Curry’s case, a gentle chuck under the chin). . . . Eileen runs my bath and sometimes brushes my hair. (321) Although equality is suggested between the two, the way they care for Camille still recalls the separation of gendered roles in the earlier rescue scene. Eileen puts Ca mille “to bed with kisses,” emphasis on the plural. Curry, on the other hand, gives Camille “a,” meaning one, “gentle chuck under the chin.” Curry retains his relative distance, and Eileen continues her nurturing care. The Currys thereby confirm their role as a patriarchal, nuclear family. Within this family, Camille can now find comfort and improve her mental health. Recalling the novel’s tagline, Camille’s biological family is established as distinctly non-patriarchal and, therefore, ‘toxic’ and unhealthy. The Currys, by contrast, are the exact opposite, and only this family unit can allow Camille to better her mental health which has suffered greatly in her biological family. This is evidenced by the fact that neither her tendency to self-harm (which started in her youth and lasted into her adulthood), nor the genealogy of perverted motherhood continue now that she is living with Curry and Eileen, the ideal nuclear family. In fact, “[a]ll sharp ob jects have been locked up” by the two, but Camille explains that she “ha[s]n’t tried too hard to get at them” anyways (321). She does not need to, she can now be a child, or rather, child-like, because she is now “learning to be cared for . . . Sharp Objects and Failed Matriarchy: Valorizing the Patriarchal Family Carrying me off in his smoky Ford Taurus, Eileen stroking my hair on the way back to Chicago” (244-45). It is telling that, even though the scene only happens in Camille’s mind, the way that Curry and Eileen perform gender here conforms to essentialist conceptions of it and, by extension, to an ide alized image of the patriarchal, nuclear family: Frank is strong and active, perform ing a fatherly and heroic role by “bursting into the room to save” Camille (244-45); Eileen, at the same time, is calm and comforting, simply “stroking [Camille’s] hair” (245). Eventually, this imagined rescue scene becomes a reality, albeit not as Hollywood esque as in Camille’s imagination. Regardless, the scene recalls the same values that the imagined one did because Curry and Eileen still adhere to traditional gender roles. For my final line of arguments to make sense, I want to summarize the order of events towards the novel’s end: The police arrest Adora for the murders of Marian, 68 Blaming the Mother Blaming the Mother learning to be parented,” as she puts it (321). Ultimately, the novel’s last lines and particularly Camille’s thoughts effect a final valorization of the patriarchal, nuclear family. Sharp Objects ends with Camille pon dering the question whether she “was . . . good at caring for Amma because of kind ness” when she was her guardian, or whether she “like[d] caring for Amma because [she] ha[s] Adora’s sickness,” that is, MSbP (321). Camille’s answer to her question again suggests that the order the patriarchal, nuclear family provides and the devel opment of idealized mother- and womanhood are connected: “Lately,” Camille says, referring to her staying within the ‘healing’ patriarchal, nuclear family, “I’ve been leaning towards kindness” (321). By juxtaposing “kindness,” connoting idealized motherhood as promulgated by New Momism, and “sickness,” connoting perverted 69 Janine Schwarz motherhood, both Camille and the novel clearly hierarchize one over the other. At the same time, Camille’s self-assessment of having been “good at caring for Amma” and her final “leaning towards kindness” imply that a woman is inherently ‘good’ and can perform motherhood with “kindness” if – and only if – she is part of a pa triarchal family. The ending thus also implies that Camille can figuratively break the cycle of “sickness” within her family precisely because she is willing to join a patri archal family. Hays, Sharon. The Cultural Contradictions of Motherhood. Yale UP, 1996. Miller, Alyson. “‘Emissaries of Death and Destruction’: Reading the Child-as-Killer in We Need to Talk about Kevin and Sharp Objects.” Critique: Studies in Contem porary Fiction, vol. 60, no. 4, 2019, pp. 487-500. Conclusion In conclusion, I have shown that Sharp Objects reinforces patriarchal ideology by creating a genealogy of perverted motherhood in Camille’s family and thereby ech oing the cultural myth of Mother Blame. Because these instances of perverted moth erhood are placed within family units that ‘lack’ a patriarchal order, the novel also advocates for the superiority of the patriarchal, nuclear family. This is also evidenced by the ‘healing’ powers Camille experiences regarding her mental health while stay ing with Curry and Eileen, who represent an idealized version of the nuclear family and thereby function as a foil to Camille’s biological family. While I acknowledge that the novel does reconceptualize motherhood to some extent by subverting essen tialist conceptions of it as inherently nurturing, I have shown that this reconceptu alization is limited precisely because other essentialist and patriarchal ideologies are clearly perpetuated at the same time. Works Cited Cain, Ruth. “‘Just What Kind of Mother Are You?’ Neoliberal Guilt and Privatised Maternal Responsibility in Recent Domestic Crime Fiction.” We Need to Talk about Family: Essays on Neoliberalism, the Family and Popular Culture, edited by Roberta Garrett et al., Cambridge Scholars, 2016, pp. 289-311. Diagnostic and Statistical Manual of Mental Disorders: DSM-5-TR. 5th ed., text revi sion, American Psychiatric Association, 2022. Douglas, Susan J., and Meredith W. Michaels. The Mommy Myth: The Idealization of Motherhood and How It Has Undermined All Women. Free Press, 2004. Flynn, Gillian. Sharp Objects. 2006. UK paperback ed., Weidenfeld & Nicolson, 2007. Hackett, Malinda K. “Dirty Pretty Things: Female Trauma, Self-Mutilation, and Ran dom Acts of Violence in Gillian Flynn’s Sharp Objects.” Clues: A Journal of Detec tion, vol. 38, no. 2, 2020, pp. 80-89. Hays, Sharon. The Cultural Contradictions of Motherhood. Yale UP, 1996. 70 Blaming the Mother Blaming the Mother Joyce, Laura. “Introduction to Domestic Noir.” Domestic Noir: The New Face of 21st Century Crime Fiction, edited by Joyce and Henry Sutton, Palgrave Macmillan, 2018, pp. 1-7. Crime Files. Joyce, Laura. “Introduction to Domestic Noir.” Domestic Noir: The New Face of 21st Century Crime Fiction, edited by Joyce and Henry Sutton, Palgrave Macmillan, 2018, pp. 1-7. Crime Files. Miller, Alyson. “‘Emissaries of Death and Destruction’: Reading the Child-as-Killer in We Need to Talk about Kevin and Sharp Objects.” Critique: Studies in Contem porary Fiction, vol. 60, no. 4, 2019, pp. 487-500. 71
https://openalex.org/W3216493391	https://figshare.com/articles/thesis/Top_Marks_for_Quality_Parks/14657841/1/files/28139844.pdf	English	null	Top Marks for Quality Parks	null	2,021	cc-by	22,223	© Jennifer Roth, 2014 Master of Planning in Urban Development Ryerson University in Urban Development Ryerson University Key Words: Park Report Card; Parks Plan; Quality Parks; Toronto. Author’s Declaration I hereby declare that I am the sole author of this major research paper. This is a true copy of the major research paper, including any required final revisions, as accepted by my examiners. I authorize Ryerson University to lend this major research paper to other institutions or individuals for the purpose of scholarly research I further authorize Ryerson University to reproduce this major research paper by photocopying or by other means, in total or in part, at the request of other institutions or individuals for the purpose of scholarly research. I understand that my major research paper may be made electronically available to the public. ii ii ABSTRACT As the City of Toronto and the Greater Golden Horseshoe continue to become more densely populated resulting from the Provincial Growth Plan, increased stresses on existing parks spaces will occur due to overuse. The focus of this research is to evaluate Toronto’s effectiveness in maintaining quality park spaces according to the Park’s Plan (2013-2017) in the City through the trial of a Park Report Card that focuses on basic amenities in parks in the densest wards. The methodology was adapted from New York City’s Park Report Card which involved the definition of basic amenities and a scoring scale for their functionality. The observed parks did not meet quality park standards according to the Parks Plan and a series of recommendations to improve the level of quality were provided. Key Words: Park Report Card; Parks Plan; Quality Parks; Toronto. iii iii Acknowledgements I would like to express my very great appreciation to my supervisor, Dr. Ron Pushchak, for his ongoing support throughout the development and production of this research project. His advice was central to the clarity of the report and the completion of this work. I wish to acknowledge the help of Dr. Pamela Robinson, the Graduate Director and my second reader, for providing valuable feedback. iv iv Table of Contents Title Page ..................................................................................................................................... i Author’s Declaration .................................................................................................................... ii ABSTRACT ................................................................................................................................ iii Acknowledgements .................................................................................................................... iv List of Tables ............................................................................................................................ vii List of Appendices .................................................................................................................... viii 1.0 Introduction .......................................................................................................................... 1 2.0 Research Objective .............................................................................................................. 4 3.0 Toronto Park Planning: Recent History ............................................................................... 5 4.0 Legislative Guidelines .........................................................................................................10 5.0 Other Comparable Plans to the Parks Plan .........................................................................11 5.1 Ottawa Greenspace Master Plan .....................................................................................11 5.2 Chicago Plans .................................................................................................................11 6.0 Inspection Programs ...........................................................................................................13 6.1 New York City Park Inspection Program (PIP) .................................................................13 6.2 New York City Report Card .............................................................................................13 6.3 Toronto Park Inspection Program ....................................................................................14 7.0 Financing ............................................................................................................................15 8.0 Park Quality Assessment: Literature Review ......................................................................17 8.1 Overview of Park Planning ..............................................................................................17 Park Design Principles .......................................................................................................17 Park Carrying Capacity and Area Standards ......................................................................18 Park Valuing and Community Engagement ........................................................................19 Different Types and Designs of Parks ................................................................................20 Current Trends ...................................................................................................................21 8.2 Benefits of Urban Parks ...................................................................................................22 Physical Health Benefits ....................................................................................................22 Mental Health Benefits .......................................................................................................23 Economic Benefits .............................................................................................................23 Environmental Benefits ......................................................................................................25 Other Benefits ....................................................................................................................25 8.3 Equity Discussion ............................................................................................................26 v 9.0 Methodology .......................................................................................................................29 9.1 Overview .........................................................................................................................29 9.2 Choosing the Sample ......................................................................................................29 9.3 Choosing the Functional Criterion/Basic Amenities ..........................................................30 9.4 Summary of Steps: ..........................................................................................................33 9.5 Benefits ...........................................................................................................................33 9.6 Limitations .......................................................................................................................34 10.0 Findings and Analysis .......................................................................................................36 10.1 Overall Findings .............................................................................................................36 Pathways ...........................................................................................................................39 Benches .............................................................................................................................40 Passive Greenspace ..........................................................................................................41 Water Fountains.................................................................................................................42 Bathrooms .........................................................................................................................42 10.2 Application of Park Report Cards ...................................................................................43 10.3 Carrying Capacity ..........................................................................................................44 11.0 Recommendations ............................................................................................................46 12.0 Future Research ...............................................................................................................48 Works Cited ..............................................................................................................................56 Appendices ...............................................................................................................................51 Appendix A - Toronto Park Categorization .............................................................................51 Appendix B – Park Sample and Ward Densities ....................................................................52 Appendix C – Carrying Capacity ............................................................................................54 Appendix D – Findings………………………………………………………………………………54 vi List of Tables vii List of Appendices Appendix A – Toronto Park Categorization 1.0 Introduction Toronto’s parks in dense downtown neighbourhoods are experiencing strain due to increased density. St. Michael’s Urban HEART research report found that of Toronto’s 140 neighbourhoods, 28 of them have greenspace shortages (St. Michaels, 2014). Combined, a lack of greenspace and diminishing park quality has made it necessary to ensure that an appropriate method for monitoring park’s basic amenities is in place. This will ensure that the public is aware of the quality of their local parks. For clarification purposes, parks are “all publicly owned natural areas (forests, ravines, etc.) as well as playgrounds, playing fields, skateboarding parks, beaches, bike trails, river walks, cemeteries, hydro corridors, paved public squares, parkettes, and community gardens (Harvey, 2010, page 10).” According to the City of Toronto Parks Plan (2013-2017) (hereafter known as the Parks Plan) “quality means that a park is clean, safe, beautiful and well-maintained. A quality park also meets community needs and creates a sense of place (page 43).” This definition was developed based on public consultations that occurred prior to the development of the Parks Plan. This research focuses on using a Park Report Card to measure the level of cleanliness and maintenance in achieving two of the outlined components that makes up a quality park space. It is imperative in maintaining park quality that an effective and established monitoring system is in place as Toronto and the Greater Golden Horseshoe Region continue to grow to meet Ontario Government mandated growth targets. It has never been more important to ensure that Toronto provides its residents with quality park spaces, especially in the densest urban areas where private greenspace access is limited. The newly implemented Parks Plan goes beyond attempting to counteract the $230 million deficit in maintenance and repairs required by Parks, Forestry and Recreation (PFAR) and has a section entitled Quality of Parks that concentrates on making Toronto parks, quality parks. Since amalgamation, Toronto has struggled to update its park infrastructure, and as a result City parks have also languished. It is necessary to start moving park infrastructure forward to achieve quality parks as the population continues to rise and the capital deficit is expected to carry on increasing to $600 million if no action is taken (Harvey, 2010). Appendix B – Park Sample and Ward Density Appendix C – Carrying Capacity Calculations Appendix D – Findings Summary viii 1.0 Introduction To ensure that park spaces are being maintained and repaired appropriately, a new method of park monitoring needs to be implemented to ensure that the community knows how well their local park is meeting the City-wide park standards. Quality park spaces as outlined in Toronto’s Parks Plan must be maintained despite growing and dense populations. To achieve quality spaces maintenance standards must be implemented equally across the city, and should be monitored and recorded. The Parks Plan 1 firmly commits PFAR staff to producing and maintaining quality parks. The Parks Plan Mission aims to “enrich communities and lives by designing, building and operating quality parks (Parks Plan, 2013-2017, page ii).” PFAR must implement the Quality of Parks component of the Parks Plan as it is more difficult in dense urban environments to ensure that all residents have access to quality park spaces to gain the full variety of individual and community benefits. To ensure that parks are becoming and staying quality parks, it is necessary to monitor park spaces and make the findings public. Long term planning is required to achieve functional and well designed parks. There are numerous benefits that make this type of government investment worthwhile. Generally there are health, economic and social benefits for residents (Pack and Schanuel, 2005). Benefits from parks will be explored more thoroughly below. When developing strategic plans, maintenance standards and monitoring programs such as the ones used in New York City, must be implemented to ensure that the goals in the plan are truly being met. Park monitoring or inspection programs are techniques used to measure if maintenance and repair standards are being achieved. Further explanation of the New York Park Inspection Program (PIPs) follows on page 13. Park monitoring programs can also be used as a means to ‘benchmark’ a city’s park space compared to parks within the city and to parks in other cities (Pack and Schanuel, 2005). Benchmarking occurs when an evaluation occurs by comparison to a defined standard and measures the quality of the feature of concern (Merriam-Webster, 2014). Benchmarking is a worthwhile endeavor as it will encourage PFAR to guarantee that parks are quality public spaces. For improved comprehension by the public, monitoring programs can be translated into park report cards providing letter grades denoting the level of quality of the space experienced by users. 1.0 Introduction This has the potential to encourage more active participation by local community groups and more investment by the city if their local parks are receiving poor or failing grades. The research that has been conducted in Toronto takes the current parks inspection program further and develops a benchmarking system in the form of a park report card that is based on the New York City Park Report Card method that targets neighbourhoods under the duress of high density levels. Park planning relies on an interdisciplinary team approach which has become the leading responsibility of landscape architects and planners (Christiansen, 1977). When planning parks, challenges exist similar to other planning experiences such as the inability to meet everyone’s 2 preferences (Rutledge, 1971). The functionality cannot be bypassed for beauty’s sake as the “highest quality is a product which has no weaknesses” (Rutledge, 1971, page 53). This may seem like an unrealistic goal, but plausible plans with few weaknesses can be created for parks through interdisciplinary collaboration and public assistance. Striving not only for well designed and aesthetically pleasing park spaces, but for ones that are truly functional with high maintenance standards should be the main focus of park planners. It is important to understand that maintenance standards alter as the budget changes and to reflect changes in industry trends. These ongoing fluctuations must be taken into account when designing parks and maintenance standards to ensure that quality standards can always be achieved. To ensure that goals and maintenance standards are being met, park inspection programs and park report cards are required. As Alan Broadbent stated, “you can tell what a city thinks about itself from looking at the way it looks after its parks. Parks, principal among public spaces, are a telling face to the world” (Harvey, 2010). 3 3.0 Toronto Park Planning: Recent History 3.0 Toronto Park Planning: Recent History Before the Parks Plan was established, the Toronto Parks Renaissance Strategy was the existing park master plan. It was drafted in 2006 and was formulated as a “reinvestment program to align the City’s parks, trails and physical assets with the social, economic and cultural needs of residents (Harvey, 2010).” A series of public consultations were held by a consultant from Urban Strategies who was hired to produce a preliminary report. Key findings as indicated by public consultation participants were issues about maintenance and lack of repairs being made in existing parks and the ongoing need for new parks in new and redeveloping neighbourhoods. It was recognized that the park planner’s challenge is to find inventive ways to expand park space in already built-up areas. The report did not focus on a monitoring system or report card as a means of maintaining the quality of existing park spaces. Differences between user needs were identified with a central example of children and teenagers requiring different facilities (Harvey, 2010). Unfortunately, after the consultant filed their report, the Toronto Renaissance Strategy was not implemented. It was not until council approved the current Parks Plan that the information from the report, along with new information collected from public consultation, was used. The current Parks Plan was established after council approved the development of a multi-year Parks Plan in February 2010. The origin of the Parks Plan was based on a recommendation from the Division’s Our Common Ground report (PFAR, Staff report for action on the Parks Plan 2013-2017, 2013). Seven guiding principles were adopted in developing the Parks Plan (2013 – 2017) which included; parks and trails as city infrastructure, equitable access for all residents, nature in the city, placemaking, supporting a diversity of uses, community engagement and partnerships, and environmental goals and practices. Monitoring programs started to enter the language of the report; however, they were not made to be explicit or to be used as a form of community awareness and engagement. Following this, in June 2011, council approved a Stakeholder Consultation Strategy that produced four objectives; communicate and connect with users, preserve and promote nature, maintain quality parks and improve system planning (PFAR, 2013). Most consultation participants were generally satisfied with the state of Toronto parks; however, as previously stated there were some concerns expressed about existing maintenance levels. 2.0 Research Objective This research aims to investigate a park report card system that would measure the level of quality of park spaces in densely populated areas of the City. It will provide a benchmark for the parks sampled within the densest wards. It can be considered a small scale trial of the effectiveness of using a park report card method to compare parks and provide data for analysis. The research may be used as a confirmation of the current design and maintenance standards and may also provide information about areas that require improvements. The research will test the park report card method as a monitoring tool that can easily communicate and engage local communities while providing accurate information regarding the Quality of Parks as established by the Parks Plan. The research that was conducted involved an adaptation of the New York City Park Report Card applied to a sample of parks in Toronto’s densest wards to establish whether the parks are meeting the Quality of Park criteria as defined by the Parks Plan. The park report card method focused on basic amenities including; bathrooms, water fountains, benches, passive greenspace and pathways. The aim of using a park report card is to establish a monitoring tool to measure park quality that is public and easily comprehended as there currently is no established inspection program with public findings. 4 4 3.0 Toronto Park Planning: Recent History Other concerns included availability of washrooms and improved washroom maintenance and the limited number of benches and water fountains. Larger Toronto parks provide basic amenities such as washrooms, benches and water fountains along with shade from mature 5 trees which work together to allow parks to be the setting for social gatherings (Parks Plan, 2013-2017). Currently the City is developing a comprehensive approach to seating to encourage socialization in city parks which is centered on these basic amenities and permitting for social gatherings. Seating in parks predominately takes the form of benches but also includes picnic benches and objects that are not traditionally viewed as sitting areas such as a ledge around a flower planter. An important aspect of the public’s ability to use parks is the application of maintenance standards. City standards for maintenance aim to be equitable across the city and are influenced by factors such as user volume, intensity of use and neighbourhood density (Parks Plan, 2013-2017). For example, parks along the waterfront receive more tourists and therefore have a higher volume of use so they tend to be cleaned more frequently to ensure that the maintenance standards are met. By investing in and maintaining basic park amenities, the public will get higher quality park spaces as result. Despite not being made public at this time, there are monitoring systems in place that need to be made more explicit to increase public awareness of the quality of local parks. The Parks Plan not only included initial stages of community engagement, but is intended to include the public throughout the five years of implementation. Closer relations with ‘Friends of’ groups are expected to encourage public ownership of local parks and to streamline the process from idea development to implementation stage. According to the Parks Plan ‘Friends of’ groups are local groups of volunteers from neighbourhoods surrounding parks in Toronto that independently work to maintain parks and jointly with the City to raise money to improve the specified park. In addition to public open houses, a reporting system for the public to use to raise maintenance and repair concerns has been developed and will be implemented by 2017. The reporting method is, in essence an email and telephone hotline for community members to use if they see maintenance or repairs that need resolving in their parks. 3.0 Toronto Park Planning: Recent History This method of reporting problems with parks can work in tandem with established but internal monitoring programs or updated programs to ensure that the highest park quality can be achieved. The Parks Plan recognized that the City is growing and as a result city planners developed a park categorization system to provide a foundation for future service planning undertaken by the City. The park classification was developed as a means to help understand Toronto’s park and trail system. It works as an administrative tool that helps to categorize the over 1,500 parks and 90 square kilometers of park space. The Parks Plan provides five classes of parks primarily 6 based on the park’s size and are as follows; parkette, neighhourhood park, community park, district park and city park. Park categorization can aid in park monitoring as parks of varying sizes will have different features. The City has the potential to incorporate park categorization and monitoring to assist in streamlining the inspection process and to improve the communication of the maintenance results to the public. The total park space in Toronto compares favorably to other North American cities in acres per 1,000 residents. Table 1 demonstrates a comparison of the total park space of North American cities. Table 1: Total Park Space, North American Comparison (Harvey, 2010) City Total Park Space (acres/1000 residents) Toronto 7.88 Montreal 2.96 Chicago 4.2 New York City 4.6 San Francisco 7.0 Large concentrations of park space exist in the ravine systems in the east and west sections of the city. Despite high amounts of park space in Toronto, there is limited space in the more built up and dense sections of the city which is noted in Table 2 below. Table 2: Total Park Space, District of Toronto Comparison (2006) (Harvey, 2010) District of Toronto Total Park Space (acres/1000 residents) Scarborough 11.81 Etobicoke 10.13 East York 7.56 North York 7.43 York 4.67 Toronto 4.54 Currently there are neighbourhoods in Toronto that are underserved by parks, but the City is attempting to expand park space by capitalizing on underutilized areas (Harvey, 2010). Recent additions to downtown park spaces include inventive solutions such as Underpass Park in the West Don Lands, a park built underneath Eastern Avenue as it crosses the Don River. This is 7 7 an excellent example because it took formerly unused space beneath a bridge to create a park. 3.0 Toronto Park Planning: Recent History The use of park dedication requirements by the City is a successful method for expanding park space in areas that are experiencing low access to parks. Despite this achievement, there will be a continuous need to upgrade and maintain existing parks to ensure that they meet the standards mandated in the Parks Plan. Knowing that the City possesses large tracts of park space, monitoring all of it can prove to be a challenge. It is necessary to implement a tool that can be used by the City to monitor all the spaces for maintenance and repairs as well as notify residents of the quality of their local park spaces. Moving forward, PFAR will continue to implement the Parks Plan. The implementation of this strategic plan depends on both the operating and capital budget. PFAR will receive 4.3% of the approved operating budget for 2014 and 6% of the capital budget. The major source of funding is property tax and intergovernmental transfers (PFAR, 2014 Budget at a Glance, 2014). Despite some funding, many of the major initiatives in the Parks Plan have gone unfunded in 2014 including the area related to this research; of building and maintaining quality parks in dense urban neighbourhoods. As previously stated, there is an extensive maintenance and repairs deficit of $230 million in 2010 with an expected rise to $600 million by 2018. It is more imperative than ever that this become a political issue to ensure that parks receive the appropriate funding needed through both the capital and operating budgets. PFAR has developed their 2013-2021 capital budget to reflect their priority of providing park space in a state of good repair to keep pace with the growing and changing city. The approved 2014 capital budget will focus on reducing the State of Good Repair backlog by providing $491.8 million (PFAR, 2014 Operating Budget Highlights , 2014). As aforementioned, the Quality of Parks initiative outlined in the Parks Plan has not been provided with any funding in the 2014 City operating budget (PFAR, 2014 Operating Budget Highlights , 2014). It will be considered for funding in future budgets, but will require significant resources to fully implement. 2013 maintenance levels are being maintained at an incremental cost of $12.885 million to bring the total operating budget for PFAR to $409.740 million gross. 3.0 Toronto Park Planning: Recent History All operating budget changes will focus on improving service levels to meet demographic changes and growing population needs over the coming years. 8 Despite the lack of funding support, maintenance and repairs monitoring are ongoing but done with results only being known internally. A new method to work alongside or to replace the existing monitoring system needs to be implemented to support the increased concentrations of population in Toronto. A new method of monitoring needs to be implemented by PFAR as a means to raise public awareness, political concern and community engagement as the Parks Plan seeks future funding. 9 5.1 Ottawa Greenspace Master Plan The City of Ottawa developed its Greenspace Master Plan in 2006 which targets ways to preserve and expand the City’s parks and forests. This plan recognizes the efforts of past councilors who assembled land along waterways and in communities before any houses were built (Greenspace Master Plan, 2006). It aims to preserve this legacy of greenspace as the population continues to grow. Ottawa is well positioned to enforce the Greenspace Master Plan as a result of amalgamation of 11 local municipalities and a regional government into 1 local municipality with more authority to streamline this process. The major aspects of this plan are a greenspace inventory, the development of a greenspace network, expansions of greenspaces and increasing the quality of and access to greenspace. Quality greenspace is defined by the Ottawa Greenspace Master Plan (2006) as attractive, safe and well-designed, serving multiple needs of users while defining the unique identity of communities (page 27).” This definition is reflective of local values and aims to create spaces that residents need and enjoy. A complete inventory of City-owned parks was conducted which allowed for a more standard delivery and monitoring of parks in old and new neighbourhoods. The inventory also provides information on the contribution of parks to the larger greenspace network and areas where there is need for more parks. Updates to the parkland dedication by- law will establish a standardized way to calculate the land requirements and the criteria needed to take cash in lieu of land dedication. When developing new communities, the park space must be linked to the existing greenspace network. Currently there is no monitoring tool listed in the Greenspace Master Plan, such as an inspection program or a park report card, to ensure that maintenance and repairs are completed and that the results can be made public. 4.0 Legislative Guidelines The City of Toronto’s Official Plan and the Province’s Growth Plan have targeted the City to grow by half a million residents and a similar number of new jobs by 2031. This accounts for almost 14% of the Greater Golden Horseshoe’s growth (Harvey, 2010). As the City of Toronto continues to grow and become denser as a result of the Greenbelt preventing sprawl and density targets being achieved following the Growth Plan, it is becoming more vital to maintain quality park spaces and prevent overuse because parks provide the City and its residents with many benefits. Not only does the City preserve green and open space in parks in accordance with the Official Plan, but the Toronto Region Conservation Authority (TRCA) is in the process of finalizing its Living City Policies (LCP) which aim to combine the TRCA’s legislative power with new policies that are focusing on development initiatives (TRCA, 2013). It is reassuring to see the regional authority of green and open space keeping pace with the growth of the city and region. 10 5.2 Chicago Plans The City of Chicago has several plans that include aggressive provisions to improve the quality and amount of green and park space. The Chicago City Space Plan (1998) envisions parks of “unmatched quality and beauty” in the downtown core (page 85). Parks in the city core are meant to be places for relaxation for downtown workers and visitors alike. This plan was a response to the growing density in Chicago’s downtown which prompted a search for creative ways to improve and increase the amount of open space and parks. All district plans were made to reflect the initiatives outlined in the City Space Plan (1998). The central ideas of this plan focused on; expanding existing park and open space and creating new parks, creating a 11 Riverwalk to act as a linear park, and on improving the access points to the open space. As part of the plan to expand green and open space, an inventory of vacant lots was conducted to demonstrate existing opportunities. The implementation of the Chicago City Space Plan (1998) focused on developing public-private partnerships and creating relationships with community groups and organizations. To ensure that the City Space Plan (1998) was meeting its outlined goals and objectives, several review and monitoring systems were implemented. The Go To 2040 Plan is a comprehensive regional planning process focusing on creating a future for Chicago that is sustainable and aligns public policy and funding. The plan recognizes the system of parks that exists which adds to residents’ quality of life and that there is currently inadequate access to parks for nearly half of the region’s residents. As part of this plan, it is recommended to expand and improve the park system because of the personal benefits to the region’s residents and the benefits to the regional economy (CMAP, Go To 2040, 2010). Parks are valued as green infrastructure that is central to the prosperity and livability of the region. The plan focuses on obtaining more park space in areas that are being built up to preserve the heritage of green openspace that was central to the 1909 Plan of Chicago that Daniel Burnham envisioned. This plan has comparable objectives to the Parks Plan to improve and expand park space where possible. 6.0 Inspection Programs Park inspection programs are monitoring systems that are necessary to ensure that plans to improve and expand parks are being achieved. Several programs will be explored which were central to the development of the method for the conducted research. 5.2 Chicago Plans The Go to 2040 Plan states that the region should monitor any programs implemented that result from the policies within the plan, but does not directly stipulate a method or potential funding source. Both of these are necessary to ensure that a monitoring or inspection program can be implemented effectively. These comparisons provide an insight into how other park plans do not always include monitoring programs or having funding to implement them. Monitoring programs are essential to ensuring that maintenance standards are met and that quality park spaces are being sustained. None of the comparable park plans incorporated a means to use a monitoring program as a method to engage the public to improve awareness of park quality. 12 6.1 New York City Park Inspection Program (PIP) The aim of this program is to measure performance by providing detailed inspections of New York City’s parks and playgrounds. It is administered by the Operations and Management Planning division and provides information on the overall condition of New York City Parks as experienced by park users. Over 6,000 inspections are conducted each year by a group of trained inspectors who use hand-held electronic devices and digital cameras to track their observations (Parks N. , 2013). Park inspections examine cleanliness features, landscape features and structural features. Inspectors identify park elements that require maintenance and repairs which help to target limited funding to the areas of most need. Findings are made available online for the public and city officials. 6.2 New York City Report Card The NYC PIP was developed in 2002 and was central to the development of the NYC Report Card method in 2007 which was developed by New Yorkers for Parks (NY4P). There are several major steps: 1. Selection of survey population 2. Identification and weighting of major service areas 3. Feature forms: structures of the survey instrument 4. Assignment of numerical scores 4. Assignment of numerical scores 5. Conversion of numerical scores to letter grades 6. Conducting the survey 7. Comparison 8. Modifications The survey selected parks between 1 and 20 acres in area which are classified as neighbourhood parks (DiPalermo, 2007). There were 123 neighbourhood parks chosen to be in the sample; however, not all of them were included (ie. parks undergoing capital improvements). Throughout the identification and weighting process, a group of community leaders and elected officials determined what the basic amenities, such as benches and water fountains, would be 13 and their relative importance. A group of 30 park users were asked to rate the relative importance of each basic amenity which confirmed the weighting. Four categories for evaluation were determined by NY4P staff and Ernst and Young, a statistical consultant to be included in the park report card inspections. The categories were; maintenance, structural integrity, safety and cleanliness. Focus groups were involved to provide feedback and rank the importance of each category in park spaces. Field work was conducted by NY4P staff in June-August in 2006 after receiving uniform training on standards and handheld computers for recording findings. Once field work was completed, numerical scores were calculated for each major service area or basic amenity. To convert the numerical scores to letter grades, another focus group was held and its results are provided in the Table 9 in the Methodology, which was used to inform this research. 7.0 Financing As Olshansky (2003) states, “creating parks requires planning, citizen involvement and funding.” However, there are budget restraints that exist and at times an unwilling political climate to invest in parks emerges. There are a series of policies that enforce park development including; Section 42 of the Planning Act that requires a portion of the development area to be given as a Parkland Dedication or cash in lieu (Parks Levy Fee), and Section 37 of the Planning Act that provides for an exchange of increased height or density for a public benefit which includes park space. Transparency and accountability of all funds collected through Section 42 and Section 37 must be achieved to allow the public to benefit from the legislation that has been outlined on their behalf. Development charges are not a plausible way to expand land required for park space according to section 4 in the Development Charges Act, 1997. Ongoing funding needs to be secured to ensure that parks can be maintained and become quality spaces through the use of monitoring programs which are central to ensuring that this can occur. Funding for park inspection programs is invaluable and must expand as population density increases and the amount of park space in Toronto expands. Harvey (2010) says that more creative ways to gain funding for parks should and can be obtained. Corporate donations should be encouraged and can be done after a revision to the no-naming policy of the City. Selling naming rights for parks and park facilities can be used to increase the funding opportunities in developing new parks and improving existing parks. Improved relationships with ‘Friends of’ groups who have fundraising opportunities can enhance the community engagement and have funded parks that reflect what users really want. Park improvement areas can be developed to “facilitate individual and corporate funding for local parks (Harvey, 2010, page 7).” Private donations normally cover capital costs, but there is little improvement found in obtaining donations to cover operating costs of park maintenance which would include monitoring programs. The PFAR 2014 recommended operating budget is $409.740 million gross which is an increase in the operating budget of $12.888 million to continue the 2013 maintenance levels (PFAR, Toronto 2014 Budget, 2014). Of the operating budget, parks receive 34% of the total allotted amount. 6.3 Toronto Park Inspection Program Toronto’s Park Inspection Program was developed in 2010 “to monitor the condition of Toronto’s Parks (Parks Plan, 2013, page 46)” and was based on the New York Park Inspection Program. As part of the Parks Plan it was recommended that the current park inspection program in Toronto be updated to ensure that parks are safe, accessible and of high quality. Park employees in charge of operations and maintenance have and will continue to be central to the park inspection program as they continue to denote areas of concern regarding maintenance levels and repairs. This tool also provided information on types of successful service delivery and indicated quality. Safety concerns are followed up immediately and minor repairs are made within 48 hours (Parks Plan, 2013). As the Parks Inspection Program continues to develop, it will reflect the new park’s classification system as denoted in the Parks Plan. New inspection requirements will be added and the recording of observations will be made easier using mobile devices. The aim is to provide accurate and timely data reporting so that maintenance and repairs can occur quickly. This method is used internally to monitor parks and track data and the results are not made public. The park report card method being tested as part of this research aims to be a public resource denoting the quality of park spaces which is currently based on the functionality of basic amenities. 14 7.0 Financing There are a variety of government agencies such as the TRCA, Waterfront Toronto, Rouge National Park, who work alongside PFAR to create quality park spaces in Toronto. As stated, there are a number of ways that parks are funded from new development. An excellent example is Canoe Landing, which is a new 10 acre park and open space facility that was built by the developer as part of the City Park development in Toronto (Harvey, 2010). 15 For smaller sites, it is not as easy to obtain park space, so cash in lieu is obtained at a rate of 5% for residential development and 2% for commercial development. The funds from this parkland levy are placed in the City’s Parkland Acquisition Reserve Fund (PARF) and are spent equally in four ways: 25% for land acquisition in the local area, 25% for land acquisition across the city, 25% for development and upgrades to existing parks in the local area and 25% for development and upgrades to existing parks across the City (Harvey, 2010). PARF spending is determined at the same time as the budget is finalized by council. The section 37 public benefits from increased height and density allowances for developers are dependent on the local councilors’ negotiating abilities. Questions of accountability and transparency in these negotiations have been raised by the development community and the public, and the City developed guidelines to rectify this deficit in 2007. Many local park improvements and new parks have been obtained using Section 37 benefits in neighbourhoods like the Queen West Triangle and Fort York. Land Acquisition for Source Water Protection is a program that purchases land to protect the City’s drinking water supply. Funding for this program is derived from Toronto Water and has been successful in obtaining new park land. Through this acquisition program, 68 acres of land were added to Rouge Park (Harvey, 2010). This is an excellent method to expand green and open space within urban boundaries while promoting sustainable environmental treatment. There are a number of funding tools that are available to increase and improve existing parks in Toronto that should include the provision of funding for a monitoring system with publicly accessible results. The level of effectiveness of some of these funding tools needs to be enhanced with particular emphasis on gaining more private donations and building partnerships. 8.1 Overview of Park Planning Park planning has a long history and many different design progressions. Andrew Wright Crawford was an early park planner in Philadelphia who realized the need to bring parks to the people as “parks that are not accessible easily and cheaply are useless (page 289).” Frederick Olmstead agreed that all Americans should have access to parks and greenspace since prior in European Cities, parks had been a privilege for people who could afford to pay for them (Garvin, 2000). Knowing the benefits of parks and the positive experiences of the City Beautiful and City Healthful movements experienced in Bourneville, England, of lower death rates and their correlation with a higher number of parks, Wright sought to bring excellent park design standards to North America. He argued that parks must be oriented to the street and allow natural topography to influence the placement of park features (Crawford, 1910). 7.0 Financing Park levies must be used to gain more park space or improve existing parks and the transparency of this process must be increased; however this is beyond the scope of this research. Quality park spaces must be central to the development, redevelopment and ongoing maintenance of parks and their funding sources. Overall, park funding is central to supporting park monitoring programs that promote park quality and is unfortunately outside of this research’s scope. 16 8.0 Park Quality Assessment: Literature Review There is much research focusing on park planning and its many components as “parks are, in fact, a vital element of city building, one of the basic pieces of infrastructure needed to make cities fit for human habitation (Olshansky, 2003, page 318).” Within this literature review, broad industry movements and history have been explored. From this an overview of different principles, standards, valuing and trends of parks, the categorization of parks, the benefits of parks, and a discussion of equity have been included to inform the research that has been conducted. The literature has influenced the approach taken in developing the checklist criteria for the proposed monitoring system and the focus on basic amenities instead of aesthetic components of parks. Park Design Principles Garvin (2000) found that even in cities with a large amount of green space, parks must be located in areas where residents can access them. Mattocks (1937) outlined seven general guiding principles of an ideal park plan to aid in providing parks in built up areas and not in outlying areas so as to increase the ease of access for residents (Crawford, 1910; Moeller, 1965). The seven principles for designing park spaces are; use, planning, artistic unity, balance, scale, style, and the use of plantings. Planners must determine the use of the park and should be influenced by the surrounding environment’s zoning and population density. The plan should be reflective of the park’s purpose and must balance artistic and practical necessities (Mattocks, 1937). There needs to be harmony and artistic unity and a balance 17 between formal and natural areas. Scale, style of design and the use of plantings must all be reflective of the setting of the surrounding architecture. Mattocks (1937) viewed park spaces as a natural continuation of the city and argued that they must be planned with equal force. Ensuring that park planning principles are used effectively to bring park spaces to people so that they are easily accessible has been a common practice within this field of study. Park spaces must be viewed as an imperative part of the city and be planned using strong design principles to ensure that quality park spaces prevail. Knowing the importance of park spaces within cities it is important to ensure that their quality is maintained as cities grow and become denser through the application of monitoring and inspection programs. Park Carrying Capacity and Area Standards Overuse of park spaces can lead to low quality parks. To prevent overuse, the practice of calculating a carrying capacity or design load, when planning parks and developing additional park spaces, has been used. Carrying capacity or ‘design load’ is “the optimum number of people…that an area or facility has been designed to physically hold (Christiansen, 1977).” It is important to calculate the carrying capacity of a park that is in an area of heavy use so that it will not deteriorate and park planners will recognize when another park is required so the quality of a park space is maintained (Fogg, 1975). Therefore, a major component of determining the required park space is population density (Moeller, 1965). These calculations are integral to park planning in dense urban spaces. Park inspection programs can aid in carrying capacity calculations to ensure that park quality is maintained as cities continue to become denser. In 1914 Charles Downing, a landscape architect for the New York State Department of Parks calculated that a city of 100,000 residents would require 1,500 acres of park space. This was to be divided amongst parks of differing sizes, from the city park scale to local playgrounds. At the time about a quarter of American cities met this standard with some far exceeding it. Downing went on to argue that, for cities with populations of 500,000 or less, the standard of 1 acre per 200 residents be used and for cities with populations over 1 million residents, the standard of 2 acres per 300 residents should be used (Moeller, 1965). These area standards can aid in calculating the carrying capacity of parks and can assist cities in recognizing when more park land is required. A related concept is the social carrying capacity, which is the “degree of satisfaction of users” (Yapp and Barrow, 1979, page 199). This is an important factor that must be incorporated into 18 park planning procedures as park users must be satisfied with the provided park spaces; otherwise their quality may be considered lower. Public engagement is central to having a high degree of satisfaction among park users. The proposed use of a park report card to make the quality of park spaces known to the public can be viewed as a form of public engagement. Park Carrying Capacity and Area Standards The report card will notify residents of the quality of their park and will encourage the public to become active in improving their local parks. This can be done in a series of ways including lobbying political bodies, becoming a member of a ‘Friends of’ group and fundraising (Harvey, 2010). Park Valuing and Community Engagement As the recognition of the need for more park land grew, it corresponded to the increased value being placed on parks by the public. Francis (1987) found that there was a gap between city employees’ and users of parks’ valuations of park spaces; with the public valuing parks higher than City staff (Loukaitou-Sideris, 1995). Public support for urban park space has continued to grow as parks provide many different recreational activities and improve the quality of life of residents (Salazar and Menendez, 2005). Parks continue to gain value and recognition as a public resource and as an imperative component of the public realm, even by non-users. City employees have over time recognized the growing importance of parks to the public and as a result the designs have become more reflective of how people choose to spend their time in parks and not as focused on past design principles (Page et al., 1994). Park planners must assess community needs appropriately to give the public what they need and value in park spaces. In the past, park designs and plans would not have been as reflective of the park user groups and their ideals as they are currently. It is extremely challenging to design park spaces in multi-ethnic areas as park designs and plans should reflect the values and specific needs of a variety of ethnic groups and allow for the co-existence of all ethnic groups in a community (Loukaitou-Sideris, 1995). To successfully meet different user needs, park planners must seek community input. As a part of that process, park planners must learn to efficiently allocate and prioritize community needs as there are limited resources (Olshanksy, 2003; Page et al., 1994). As well, park planners will have to meet community needs within park budget constraints. Park master plans can be developed to reflect the community’s needs and goals as a basis for action plans that focus on implementation (Pack and Schanuel, 2005). Park master plans must also include monitoring and inspection programs to ensure that the quality of parks is 19 continuously achieved as cities continue to grow. As part of the implementation and monitoring processes, there is an ongoing struggle to find cost effective ways to deliver the park services required to meet the variety of public needs. Park Valuing and Community Engagement Despite this financial obstacle, parks need to remain accessible public goods that do not have user fees attached to them as they contribute to the wider public good (Page et al., 1994). Because user fees are a regressive tax, they would limit the equitable distribution of park space benefits. Different Types and Designs of Parks There are different types of parks that can be classified in a variety of ways and can be distinguished by their different features. Classification systems have been developed based on park size, park use, amenities present, location, and recreation uses. Moeller (1965) outlined a park classification method that has three categories: neighbourhood parks, community parks and major parks. Categorization of parks aids in the monitoring of park spaces. As there are varying sizes of park spaces, different degrees of inspection and time investments are required. Passive and active recreation opportunities for locals are a priority of neighbourhood parks of varying sizes. Community park spaces must be accessible by public transit and be large enough to serve several neighbourhoods. Major parks provide city residents with connections to nature and provide unique recreation options (Moeller, 1965). This classification varies based on size of the park space and reflects the types of features and uses contained within. Park features and activities are reflective of the type of users of the park space. Kemperman and Timmermans (2006) found a correlation between different age groups and the type of park space that is preferred. As a means to accommodate most user needs, a three division functional layout was designed. There are areas of active recreation, a naturalized buffer area and an area for passive recreation, all of which occur in varying ratios of size of each area that is reflective of local need (Mattocks, 1937). This park design layout allows for a variety of users to be content with the same park space despite using it differently. Applying a monitoring system that focuses on functionality of basic amenities in parks instead of one focusing on aesthetics will ensure that a greater number of users’ needs are met. There are various types of parks that have emerged in the field of park planning including but not limited to; the Pleasure Ground, the Reform Park, the Recreation Facility, the Open Space System and the Sustainable Park. Cranz and Boland’s (2004) research focused on the emergence of the Sustainable Park in 1990 which was reflective of growing ecological concerns (Jim and Chen, 2003). The three main attributes of the sustainable park were self-sufficiency of 20 material resources and maintenance, its ability to resolve larger urban problems (ie. storm water storage) and its creation of new aesthetic urban standards. Different Types and Designs of Parks Parks are now being designed with more sustainable features to help promote ecological awareness within urban centers and can reduce the level of maintenance that is required. As Rutledge (1971) stated, planners and landscape architects must design with purpose. Good quality park designs must be functional as well as aesthetically pleasing to meet community needs. For example, durable surface treatments are required along paths where there is constant foot traffic. This will ensure that paths remain intact despite heavy amounts of use, which contributes to quality park spaces. Challenges arise when a community has different opinions about how a park space should be designed. Some individuals would disagree that a durable material should be used for paths suggesting that a park should be naturalized. This is a major challenge for park planners since they will be faced with the task of meeting everyone’s preferences in designing new parks and redesigning existing parks. City-specific standards are used to aid park planners in trying to meet diverse park needs and wants but they should be viewed as guides and can be modified as required (Rutledge, 1971). Current Trends There are a variety of factors that influence park planning trends. There are current park trends that are reflective of community wants, needs and values. For example, mobile park furniture is a newer innovation that is being used when designing park spaces as it allows users to locate seating in areas that they prefer instead of in mandated locations (Loukaitou-Sideris, 1995). Toronto’s current Parks Plan reflects such new industry trends. Not only does PFAR’s planning require effective community engagement, it also requires a comprehensive analysis of citywide land uses (Loukaitou-Sideris, 1995). This allows planners to be strategic and allows for the most effective organization of land uses. Current trends in park planning reflect demographic changes. Newcomers and aging populations alter the types of activity that determine the use of parks. An example of changing demographics is seen in California where there are changes recognized in recreation patterns due to aging populations in parks which are reflected in changing trends in the use of parks (Parks C. S., 2005). Another example is that Caucasians are the most represented and African Americans are the most underrepresented ethnic group in outdoor recreation in parks and open space in California. As well, immigrants tend to recreate together as a way to preserve cultural heritage and tradition (Parks, 2005). Trends vary within ethnic groups as well. The California 21 Park and Recreation Trends (2005) analysed local trends of demographic changes and the corresponding changes in use of parks. For example, it was discovered that large immigrant groups use parks as a social gathering space (California State Parks, 2005). Park master plans ensure that trends are being implemented as well as ensuring there are appropriate funding tools in place to meet all user needs (Pack and Schanuel, 2005). There is limited literature regarding the use and application of park inspection programs as a tool to ensure the quality of park spaces is maintained or as a means of community engagement. Moving forward, the use of a park monitoring system such as the New York City Park Report Card will become a more prominent trend as a means of communicating how well parks are achieving city standards. 8.2 Benefits of Urban Parks There are extensive benefits to parks in urban settings which are well researched within the literature. Benefits vary widely and include but are not limited to; physical and mental health benefits, economic benefits, social benefits and environmental benefits. Monitoring programs can ensure that parks are maintained appropriately and all park benefits can be publicly received even in densely populated cities. As this research is so widely conducted, only some of the most current researchers have been included. Mental Health Benefits Not only do parks benefit physical health, but there are correlations between parks and mental health benefits (Nowak et al. 1998; Petros and Costa, 2011). There is a growing body of research that confirms contact with or awareness of nature improves the quality of life of residents in urban spaces (Kaplan, 1998; Hartig, 1991; Hartig 2004 and Ulrich, 1979). Parks can provide the necessary contact with nature to ensure that city residents can achieve a high quality of life and improved mental health. Links between low nature access and high levels of anxiety, attention deficit hyperactivity disorder (ADHD), depression and stress were observed in a study conducted by scientists in the Netherlands looking at the prevalence of anxiety disorders in 345,000 adults (NRPA, 2010). Findings demonstrated that there is a correlation between increased anxiety disorders in adults in residential areas that do not have access to greenspace in parks. Therefore, by improving access to parks in residential areas, a reduction of the levels of anxiety disorders and other mental health challenges may occur. Findings such as these indicate the potential for health industry savings. Parks provide greenspace in urban areas that act as nodes of stress relief and provide mental health respite. Not only are there stress relief benefits, but there are links between nature and improved cognitive functioning and impulse control (NRPA, 2010). The emotions and feelings of well-being that are evoked in park spaces are thought to be important contributors to mental health (Chiesura, 2004). Parks are an important component in improving and maintaining positive mental health outcomes, as depression was diagnosed in 33% more adults in neighbourhoods with low access to greenspace in the same controlled study from the Netherlands (NRPA, 2010). Therefore, those residents who have access to parks are at an advantage compared to residents with low access to parks because they have improved cognitive function and decreased levels of mental health diagnoses. Physical Health Benefits One of the most recognized benefits of parks is that they encourage recreational uses that increase levels of activity outdoors and have positive physical health implications that improve quality of life of residents (Godbey et al., 2005; Nowak et al., 1998; Petros and Costa, 2011; Salazar and Menendez, 2005). The National Recreation and Parks Association (NRPA) (2010) and Frumkin (2003) both found that parks provide a necessary venue to maintain good physical health. Parks are increasing in their importance as there is a growing movement to counteract obesity given that 1 in 3 American children are overweight and 1 in 6 are obese. Weight loss can result in a lower chance of getting diabetes which could have longterm healthcare cost implications (Harvey, 2010). Therefore, positive health outcomes can be the result of access to quality park space. Additionally, seniors benefit from greenspace in parks as it has been reported that they have improved functionality and live longer (NRPA, 2010). Frumkin (2003) found that basic amenities like good lighting, toilets, water fountains and well-maintained paths lead to higher levels of physical activity in both younger and more senior demographics. It is necessary to correlate the 22 features found in parks with the users of park spaces because parks are positioned as a health and wellness feature of cities (Godbey et al., 2005). Providing basic amenities that are well maintained is necessary in achieving positive physical health benefits. The application of a monitoring program to measure the quality of basic amenities can help achieve community physical health benefits by providing data on how well park standards are being met. Economic Benefits Potential savings in the health industry are not the only positive economic impacts that parks provide. There are also direct economic benefits to local governments as larger parks tend to 23 act as tourist attractions (Harvey, 2010; NRPA, 2010). Revenue from tourists increases total revenues and benefit local governments. Local governments can benefit indirectly from park investment as well through an increase in property tax revenue from surrounding properties which is a concept that has been extensively researched (Salazar and Menendez, 2005). Building from Olmstead’s findings of the increase in property values that are adjacent to parks, Kitchen and Hendon (1967) conducted a cost-benefit analysis of land values of properties adjacent to parks (Crompton, 2001). They discovered that that the further a property was from a park, the lower the value. Overall, they noted the market relationship of property value and the distance to a park. Pack and Schanuel (2005) confirmed this finding as they found that property values increased with increasing proximity to parks. However, they realized that the type of use within the park may affect property values. For example, active recreation opportunities like basketball courts may deter some investors due to associated noise levels. Pack and Schaunel (2005) suggest that park design can counteract this finding if passive recreation areas border surrounding properties and the active recreation areas are located in spaces that will have minimal effect. Therefore, proximity to parks increases property values and property tax revenues benefiting home and business owners as well as the local government. Parks can trigger revitalization and development of surrounding properties in urban areas (pack and Schanuel, 2005). An excellent recent example of this is the Highline in New York City. The former elevated freight line was converted into a linear trail and park system and, as a result, there has been growth in development surrounding it (NRPA, 2011). Parks spurring development in surrounding neighbourhoods will increase property values and the expected property tax which will benefit local governments. Cortes (2008) found that the $1.5 billion construction investment in the Highline is expected to increase residential and business property values by 10-15%. This will result in higher property tax revenues for New York City. Similar experiences are expected to occur in other cities that invest in their parks, even at a smaller scale. Economic Benefits Knowing the number of economic benefits that parks can provide, it is necessary that the quality of parks is maintained as increasing use and strain is placed on them resulting from increased population densities in cities. If parks fell to disrepair and were seen as blighted and unsafe areas, this would have the opposite effect to the above findings where parks increase properties 24 values. To ensure that parks are being maintained at a high level of quality, park inspection programs are necessary as they enforce city standards and ensure that parks can provide positive economic benefits. Environmental Benefits Parks provide cities with much needed green space and act as an escape for residents from concrete. They also provide dwelling space for wildlife living in the city as well as many climate benefits. For example, Rouge Park in Toronto is large enough to provide a lot of viable habitat for wildlife. Not only do parks provide wildlife with respite, but they also reduce the heat island effect in cities and provide benefits to the water system. A park’s greenspace can improve water quality if there are wetlands to assist in water purification (Harvey, 2010). As well, parks can act as areas of water storage which can reduce run-off and flooding in the event of storms or winter melts. Both have the potential to reduce stress on local treatment plants. Frederick Olmstead called New York City’s Central Park the “lungs of the city” which refers to the air quality benefits that a park provides for citizens. Greenspaces in urban centres provide invaluable benefits as they remove air pollution, store carbon and reduce temperature increases resulting from asphalt and concrete (Salazar and Menendez, 2005). Decreases in carbon dioxide and pollution may have a positive effect on respiratory illness levels. Overall, environmental benefits of park spaces can be viewed as part of the sustainable renovation of cities to improve their ability to adjust to climate change. Ensuring that parks are maintained to provide these environmental benefits is necessary especially during times of population growth and increased urban density. This can be accomplished through programs that guarantee the ongoing inspection of park spaces such as the NYC park report card. Maintaining passive greenspace and naturalized areas at their highest level of quality will ensure the maximum benefit can be provided for the public. Other Benefits There are other benefits that are less explored but are still positive outcomes resulting from parks in urban centres. Parks provide space for social gatherings and interactions with neighbours which decrease loneliness, crime and aggression (NRPA, 2010). Communities that invest in their parks for social purposes are investing in their youth as well. Parks provide safe, natural spaces for youth to develop into strong and contributing community members (NRPA, 2010). Parks provide spaces for urban agriculture which is an educational tool as well as a means to improve food security (Harvey, 2010). As parks provide so many benefits to 25 residents, they are a sought after feature when individuals are choosing where to live. Therefore, cities must consider this as they expand, grow and become denser and planners should be able to provide parks to attract new residents (Harvey, 2010). Pack and Schanuel (2005) note that when St. Louis was experiencing a housing boom its government created a comprehensive park plan to ensure that there was enough park space for all residents to receive equal benefit. Parks require strategic planning to capture their health, economic, environmental and other benefits. It is necessary to ensure that all the above benefits are being presented to city residents in their parks despite growing urban populations. To ensure that this is occurring, quality park spaces must be achieved through the use of park master plans such as the Parks Plan. Plans such as this must ensure that there is funding secured in the operating budget to implement all policies within the plan including monitoring or inspection programs. Monitoring programs can ensure that parks are maintained appropriately which will ensure that parks can provide the greatest community benefits. 8.3 Equity Discussion As outlined above there are numerous benefits to park spaces in urban areas. There are also researchers who explicitly state that access to parks must be equitable due to the need to share the benefits that result. City Council stated that one of its central principles in developing the Parks Plan was to ensure “equitable access for all residents (PFAR, 2013, page 4).” This was reiterated again in the 2014 operating budget (PFAR, Toronto 2014 Budget, 2014). Ongoing pressure to increase the equitable distribution of and ease of access to parks by the City’s residents must also be extended to equitable park quality. All residents should have access to Quality Parks and not just individuals who could afford them. Currently the Toronto maintenance standards are equal across the City and, despite some parks receiving more use than others, all parks are supposed to be maintained equally. To ensure that all parks are equally maintained despite growing strains caused by density, city- wide programs that provide ongoing monitoring must be implemented. A park’s ability to meet park standards as measured by inspections programs must be made public to allow the community to recognize discrepancies in park maintenance. The results must be conveyed in a way that is easily interpreted by the public so that there will be no barriers. The New York City Park Report Card provides an interesting opportunity to convey park standards in a grade. 26 Historically, park planners used to situate parks and recreation facilities on the edge of cities or in wealthy areas. Moving forward, park spaces are being more equitably distributed as parks are growing in importance, especially in dense urban areas as there is limited access to private greenspace areas (Harvey, 2010). In particular areas, of low income and high density have a higher need for parks as there is less of a financial opportunity to seek out greenspace and nature outside of local parks (Moeller, 1965). Those individuals who cannot afford private greenspace should not be penalized, but should be provided with access to quality parks. The benefits that individuals receive from parks, such as physical and mental health benefits must be equally distributed across the city. Deakin (1999) states that planners need to promote equity, fairness and justice within the city. Therefore park planning in the city must also promote equity, fairness and justice which encourages citizen equality as they access park spaces. 9.1 Overview The focus of this research is to evaluate the City’s effectiveness in creating and maintaining quality park spaces in Toronto through the trial of a Park Report Card that focuses on basic amenities in parks in the densest wards. The model that is being applied is derived from New York City’s Park Report Cards and their Parks Inspection Program (PIPs). A panel of park experts and community leaders from New York City defined the functional criteria that were used in NYC and those criteria will be used as the basis for this research design. A scoring system based on the relevant importance of each criterion or basic amenity was provided as well as features for evaluation (ie. cleanliness). 8.3 Equity Discussion Having equitable access to parks will aid a city in becoming more Just, as described by Fainstein (2000). Since parks provide a variety of benefits to the public that improve their quality of life, it is necessary to supply access to these benefits to ensure that cities are fair and just to their residents. The planner’s role in providing park spaces must be to ensure that all groups of residents are supported equally as they strive to improve access and the quality of their park spaces. It is the planner’s aim to redistribute benefits within the city (Davidoff, 1978). This suggests that ensuring the benefits of parks are distributed equitably is a necessary role for park planners to fill. Park planners must seek to equalize the benefits of greenspace so that minorities, low income groups and other marginalized groups have access to parks that is equal to those groups that are not marginalized. Park planners can utilize tools for monitoring parks not only for their physical condition, but for the type of park users. This will provide data that will ensure parks are targeting all user needs and if not, planners will know that they have a challenge to resolve. Throughout this review of park planning, it has become apparent that park spaces are essential components of city spaces that provide numerous benefits to the public, businesses and the local government. The history of park planning has led to the recognition of the need for equitable access to quality park spaces as a result of the wide variety of benefits that result from the presence of parks in cities. A number of concepts, designs, and trends have been touched on and provide context for the current research that is being conducted on monitoring quality park spaces in accordance with the Parks Plan in Toronto. It is important to note that park 27 quality can only be maintained with the use of inspection programs that will ensure that city standards are being maintained. 28 9.2 Choosing the Sample The sample selection process was adapted from the NYC Park Report Card model as it is recognized as an effective model within North America. The NYC Park Report Card provides the public with an easy method to interpret the quality of their parks compared to others in the City. Toronto’s PFAR department has used other initiatives that have been started in NYC, so it is a natural expansion to utilize this model. The sample was chosen from parks in Toronto that are between 1 and 20 acres in area and located in the densest wards. The wards that are the densest are based on the City’s ward profiles and the top 4 wards included were 14, 18, 20 and 27. These wards are located in the Old City of Toronto prior to amalgamation. The densest wards were chosen as the research is focused on parks in urban areas that are experiencing stresses caused by increased density. The parks within these wards were chosen based on the list and interactive map provided by PFAR on their website. The parks that had a size listed in the description that fell between 1 and 20 acres and located in the chosen wards were all included in the sample. Fourteen parks were chosen for site visits and analysis. The sample will represent the parks in wards in Toronto that are experiencing the stresses of increased density, and will not be representative of the total park system. However, the results will be indicative of the application of park maintenance standards across the City. This sample was chosen and can be replicated in other municipalities experiencing density stresses. See Appendix B for a list of the parks chosen and the ward densities. 29 29 9.3 Choosing the Functional Criterion/Basic Amenities The basic amenities were chosen due to the ease of observation of their functionality and that they are all regularly present in parks in a standardized format. Ultimately, the amenities chosen for observation can be agreed upon as being necessary for the ease of use of parks for all users. Benches are vital to parks as they ensure that all users can enjoy the space with ease and for longer periods of time than just walking through them on the way to work for example. Passive greenspace provides relief from the City, which is related to a number of benefits that were outlined in the literature review, and this is why it is necessary to be observed. Water fountains encourage longer use in parks during summer months and were a central concern during the Parks Plan’s (2013-2017) public consultations. Increasing the number of bathrooms and their overall cleanliness were also major concerns that were brought up during public consultations. Pathways are used daily by park users and may pose a risk for trips and falls if in disrepair. This made them an excellent basic amenity to observe for functionality to contribute to the overall park grades. All of the chosen basic amenities’ evaluation criteria were adapted from the NYC Park Report which had used these same basic amenities as a means to evaluate the quality of parks. Parks and the basic amenities were evaluated using a scoring method and checklist focusing on maintenance, cleanliness, safety and structural integrity that was developed based on NYC Report Cards and PIPs (Parks, 2013). For each criterion/basic amenity being evaluated, a total value was provided based on community input from NYC consultations that occurred during the development of the Report Card method. The scoring represents the value placed on that amenity in NYC, and it is assumed that it is equally applicable in a Toronto setting. The scoring can be viewed in Table 3 below. Table 3: Basic Amenity Scores Basic Amenity Score Passive greenspace 5 Bathroom 4 Drinking fountain 3 Sitting areas 5 Pathways 3 Total 20 30 Each park being evaluated was visited once during the day in October 2013 over the course of a week. The basic amenities were observed and a value was given based on the interpretation of each score. Tables 4-8 provide the descriptions of each amenity’s score. 9.3 Choosing the Functional Criterion/Basic Amenities The tables ensure that each park is scored the same way and aids in preventing bias during data collection. Table 4: Passive Greenspace Score 1 2 3 4 5 Interpretation Lawns are not maintained. A lot of dead branches and trees. Greenspace not useable due to amount of litter. Some maintenance of lawns and trees. A fair bit of garbage and litter. Lawns require mowing and additional maintenance. Trees need maintenance. Garbage, graffiti and weeds starting to impact. Lawns are slightly overgrown. No dead trees but some dead or low lying limbs. Minimal garbage, graffiti and weeds. Lawns are properly irrigated, moved, even, maintained. Trees do not have low hanging branches, no dead limbs and no dead trees. No litter, graffiti or weeds. Table 5: Bathroom Score 1 2 3 4 Interpretation Unusable due to litter, graffiti and lack of supplies Would rather not use due to a fair amount of litter and graffiti and minimal supplies Useable – clean enough, some litter or graffiti, stocked with supplies for the day. Useable – clean, no litter or graffiti and fully stocked Table 6: Water Fountain Score 1 2 3 Interpretation Doesn’t work and has too much litter and graffiti on it to use. Works but not the most efficient. Some litter and/or graffiti that would detract some users. Works well, no litter or graffiti impeding use. Table 4: Passive Greenspace Score 1 2 3 4 5 Interpretation Lawns are not maintained. A lot of dead branches and trees. Greenspace not useable due to amount of litter. Some maintenance of lawns and trees. A fair bit of garbage and litter. Lawns require mowing and additional maintenance. Trees need maintenance. Garbage, graffiti and weeds starting to impact. Lawns are slightly overgrown. No dead trees but some dead or low lying limbs. Minimal garbage, graffiti and weeds. Lawns are properly irrigated, moved, even, maintained. Trees do not have low hanging branches, no dead limbs and no dead trees. No litter, graffiti or weeds. Table 5: Bathroom Score 1 2 3 4 Interpretation Unusable due to litter, graffiti and lack of supplies Would rather not use due to a fair amount of litter and graffiti and minimal supplies Useable – clean enough, some litter or graffiti, stocked with supplies for the day. 9.3 Choosing the Functional Criterion/Basic Amenities Useable – clean, no litter or graffiti and fully stocked Table 6: Water Fountain Score 1 2 3 Interpretation Doesn’t work and has too much litter and graffiti on it to use. Works but not the most efficient. Some litter and/or graffiti that would detract some users. Works well, no litter or graffiti impeding use. Table 4: Passive Greenspace Score 1 2 3 4 5 Interpretation Lawns are not maintained. A lot of dead branches and trees. Greenspace not useable due to amount of litter. Some maintenance of lawns and trees. A fair bit of garbage and litter. Lawns require mowing and additional maintenance. Trees need maintenance. Garbage, graffiti and weeds starting to impact. Lawns are slightly overgrown. No dead trees but some dead or low lying limbs. Minimal garbage, graffiti and weeds. Lawns are properly irrigated, moved, even, maintained. Trees do not have low hanging branches, no dead limbs and no dead trees. No litter, graffiti or weeds. Table 5: Bathroom Score 1 2 3 4 Interpretation Unusable due to litter, graffiti and lack of supplies Would rather not use due to a fair amount of litter and graffiti and minimal supplies Useable – clean enough, some litter or graffiti, stocked with supplies for the day. Useable – clean, no litter or graffiti and fully stocked Table 6: Water Fountain Score 1 2 3 Interpretation Doesn’t work and has too much litter and graffiti on it to use. Works but not the most efficient. Some litter and/or graffiti that would detract some users. Works well, no litter or graffiti impeding use. 31 Table 7: Pathways Score 1 2 3 Interpretation Holes, cracks and protrusions in paved pathways making them unsafe. Some cracks in pavement but no major unevenness, so they can still be used. New pathway that is smooth and safe. Table 8: Sitting Areas Score 1 2 3 4 5 Interpretation User cannot sit due to amount of garbage/graffiti and/or level of damage (ie. missing slats) A fair amount of garbage/graffiti with some structural damage that reduces the use greatly. Some graffiti or garbage, but seating options still available. Some structural damage, but seating can still be used. Some graffiti or garbage, but seating options still available. No structural damage. Can still be used. No user impact. Clean, no graffiti. No structural damage. 12. Provide recommendations. A complete summary of the NYC Parks Inspection Program can be found in the Background section. A complete summary of the NYC Parks Inspection Program can be found in the Background section. 9.3 Choosing the Functional Criterion/Basic Amenities Table 8: Sitting Areas Score 1 2 3 4 5 Interpretation User cannot sit due to amount of garbage/graffiti and/or level of damage (ie. missing slats) A fair amount of garbage/graffiti with some structural damage that reduces the use greatly. Some graffiti or garbage, but seating options still available. Some structural damage, but seating can still be used. Some graffiti or garbage, but seating options still available. No structural damage. Can still be used. No user impact. Clean, no graffiti. No structural damage. After each park’s basic amenities were evaluated, a mark out of the total score will be calculated. From this, a numerical grade between 0-100 is determined for each park by multiplying the total by 5. The next step is a conversion from a numerical grade to a letter grade, which is based on the NYC conversion chart which is in Table 9. The conversion chart was determined through systematic consultations with the participating NYC community groups, local politicians and professionals within the field of park planning. Table 9: Conversion Chart (NYC Park Report Cards, 2007) Numerical Grade Letter Grade 97-100 A+ 93-96 A 90-92 A- 87-89 B+ 83-86 B 80-82 B- 77-79 C+ 73-76 C 70-72 C- 60-69 D 59 and below F Table 9: Conversion Chart (NYC Park Report Cards, 2007) Numerical Grade Letter Grade 97-100 A+ 93-96 A 90-92 A- 87-89 B+ 83-86 B 80-82 B- 77-79 C+ 73-76 C 70-72 C- 60-69 D 59 and below F Table 9: Conversion Chart (NYC Park Report Cards, 2007) Numerical Grade Letter Grade 97-100 A+ 93-96 A 90-92 A- 87-89 B+ 83-86 B 80-82 B- 77-79 C+ 73-76 C 70-72 C- 60-69 D 59 and below F Table 9: Conversion Chart (NYC Park Report Cards, 2007) Analysis of the research provides knowledge of the current state of quality of Toronto’s parks in the densest wards in the City. It can provide a baseline for the current status of parks and can 32 be used to measure the City’s progress in achieving quality park spaces as defined and outlined in the new Parks Plan. be used to measure the City’s progress in achieving quality park spaces as defined and outlined in the new Parks Plan. 9.4 Summary of Steps: 1. Background research is collected. 1. Background research is collected. 2. Identify functional park components/basic amenities to be evaluated. 3. Identify evaluation criteria. 4. Develop survey instrument based on PIPs. 4. Develop survey instrument based on PIPs. 5. Denote sample boundary. 5. Denote sample boundary. 6. Develop criterion for sample. 6. Develop criterion for sample. 7. Identify sample. 8. Data collection and evaluation of parks using survey instrument. 10. Calculate score out of 100 by multiplying totals by 5 and convert findings to letter grades. 10. Calculate score out of 100 by multiplying totals by 5 and convert findings to 11. Analyze data and state findings. 11. Analyze data and state findings. 12. Provide recommendations. 9.5 Benefits There are numerous benefits to this research from the local neighbourhood level to the broader regional level. As the Greater Golden Horseshoe continues to become denser as stipulated in the Provincial Growth Plan, cities of varying sizes can use the findings from this research as a means to enhance their park planning and maintenance standards. Providing park report cards allows for comparisons within the City of Toronto and between cities. This will encourage high standards of park design and maintenance to enhance the quality and resulting benefits that parks provide across the region. This research provides an easily understood translation of maintenance standards for the public. As a result, the public may be more encouraged to take a more active role in the maintenance and enhancement of their local parks. Higher degrees of ownership of local parks may lead to a more engaged public as they will be able to understand the quality standards attached to their park. Councilors in turn may recognize the growing importance of park spaces to their constituents and make park development, improvements and monitoring a higher 33 political priority. Once parks are of higher priority on the political agenda, increased funding to enforce plans and policies may be secured. Gaining more knowledge about the current state of parks has the potential to increase the awareness of the benefits that parks provide to the public, businesses and the local municipality. Gaining more awareness of the benefits derived from parks will encourage higher investment in them which in turn will improve the quality of life of residents. 9.6 Limitations There are a variety of limitations that are observed in this research. The time frame is a major restraint in this research. Ideally the fieldwork research would have occurred over a longer time period allowing for comparisons of the ongoing maintenance of parks. Instead, due to the time constraints on the research, a baseline was provided of the parks surveyed. The time limitation made this more challenging to fully explore the feasibility of using the park report card method on an ongoing basis in Toronto. Alternatively, it can be evaluated on its ability to effectively report on current maintenance levels. Weather was another limiting factor which resulted in a single field observation session for each park. There are limitations associated with not scoring all components in the parks that were observed. The unobserved factors in parks, such as active recreation fields and playground equipment, could have large influences on the quality of the park space. The data collected provided a score of the quality of each individual feature as well as that basic amenities influence of the overall quality of each park. There are also limitations attached to the sample size. A larger sample would have provided more data and possibly an improved understanding of the quality of parks that are in areas of high density. When conducting the field observations, personal bias may occur when ranking the indicators of quality. The scoring matrix is the strongest counteraction to bias which is found in Tables 4-8. As well, there was only one researcher conducting the site visits which ensures that the bias is uniform and the data collected do not vary dramatically from one observation to another. If there were multiple field researchers, a variety of personal biases would be at play which may lead to variability in the findings. A possible limitation is found in the difference between New York City and Toronto. Since the observation standards were developed in NYC, they may not be reflective of local values and opinions. Focus groups in Toronto would be required to develop City-specific scoring for each 34 criterion/basic amenity identified by the public to be included in the check list for monitoring to counteract this possible drawback. A regional park planning report card survey tool would be ideal to record the intensity changes in park use as the Greater Golden Horseshoe continues to increase in density. 9.6 Limitations criterion/basic amenity identified by the public to be included in the check list for monitoring to counteract this possible drawback. A regional park planning report card survey tool would be ideal to record the intensity changes in park use as the Greater Golden Horseshoe continues to increase in density. 35 10.1 Overall Findings Upon completion of the field research, the data were used to calculate the park report cards as outlined in the methodology. The results are summarized in Table 10 below. Appendix D provides the individual marks each basic amenity achieved, bench counts and bench to park size ratios. None of the surveyed parks achieved marks in the A range. There are a number of possible explanations for this occurrence which include, challenges with park maintenance and standards or funding is not available to meet the outlined standards in the Parks Plan. These possible explanations for the park findings will be explored in more depth. As well, the use of park report cards for benchmarking, evaluation and conveying information will be examined. Table 10: Park Grades Park Grade Alexandra Park C Bellevue Park B- Campbell Avenue Playground D Canoe Landing B- Dovercourt Park B+ Dufferin Grove Park C+ Grange Park C Jesse Ketchum Park D Little Norway Park B MacGregor Park B- Masaryk Park B- McCormick Park B- Moss Park C Vermont Square B- Table 10: Park Grades The grades are reflective of parks throughout Toronto that are experiencing levels of stress resulting from increases in density. As the overall median of surveyed parks is a B-, the quality standard defined by the Parks Plan was not met on the day surveyed using the adapted park report card method. None of the parks were in the A range, the majority were in the B range and several parks received C and D grades. This range of marks is indicative of different usage levels and different lengths of time from the last visit by maintenance crews. A combination of basic amenities contributed to each park’s scores and there was no specific feature that repeatedly had the lowest parks. However, passive greenspace never received the highest number of marks that could be allotted. This may not be reflective of different maintenance standards because there are City-wide standards to ensure all park users have access to the same level of quality of basic park amenities. 10.1 Overall Findings Park Grade Alexandra Park C Bellevue Park B- Campbell Avenue Playground D Canoe Landing B- Dovercourt Park B+ Dufferin Grove Park C+ Grange Park C Jesse Ketchum Park D Little Norway Park B MacGregor Park B- Masaryk Park B- McCormick Park B- Moss Park C Vermont Square B- Table 10: Park Grades The grades are reflective of parks throughout Toronto that are experiencing levels of stress resulting from increases in density. As the overall median of surveyed parks is a B-, the quality standard defined by the Parks Plan was not met on the day surveyed using the adapted park report card method. None of the parks were in the A range, the majority were in the B range and several parks received C and D grades. This range of marks is indicative of different usage levels and different lengths of time from the last visit by maintenance crews. A combination of basic amenities contributed to each park’s scores and there was no specific feature that repeatedly had the lowest parks. However, passive greenspace never received the highest number of marks that could be allotted. This may not be reflective of different maintenance standards because there are City-wide standards to ensure all park users have access to the same level of quality of basic park amenities. Park Grade Alexandra Park C Bellevue Park B- Campbell Avenue Playground D Canoe Landing B- Dovercourt Park B+ Dufferin Grove Park C+ Grange Park C Jesse Ketchum Park D Little Norway Park B MacGregor Park B- Masaryk Park B- McCormick Park B- Moss Park C Vermont Square B- Table 10: Park Grades The grades are reflective of parks throughout Toronto that are experiencing levels of stress resulting from increases in density. As the overall median of surveyed parks is a B-, the quality standard defined by the Parks Plan was not met on the day surveyed using the adapted park report card method. None of the parks were in the A range, the majority were in the B range and several parks received C and D grades. This range of marks is indicative of different usage levels and different lengths of time from the last visit by maintenance crews. A combination of basic amenities contributed to each park’s scores and there was no specific feature that repeatedly had the lowest parks. 10.1 Overall Findings However, passive greenspace never received the highest number of marks that could be allotted. This may not be reflective of different maintenance standards because there are City-wide standards to ensure all park users have access to the same level of quality of basic park amenities. Park Grade Alexandra Park C Bellevue Park B- Campbell Avenue Playground D Canoe Landing B- Dovercourt Park B+ Dufferin Grove Park C+ Grange Park C Jesse Ketchum Park D Little Norway Park B MacGregor Park B- Masaryk Park B- McCormick Park B- Moss Park C Vermont Square B- The grades are reflective of parks throughout Toronto that are experiencing levels of stress resulting from increases in density. As the overall median of surveyed parks is a B-, the quality standard defined by the Parks Plan was not met on the day surveyed using the adapted park report card method. None of the parks were in the A range, the majority were in the B range and several parks received C and D grades. This range of marks is indicative of different usage levels and different lengths of time from the last visit by maintenance crews. A combination of basic amenities contributed to each park’s scores and there was no specific feature that repeatedly had the lowest parks. However, passive greenspace never received the highest number of ot be reflective of different maintenance standards nsure all park users have access to the same level The median grade is a B- which may indicate that parks in Toronto are experiencing pressures from density. The median grade would be higher if the observed basic amenities received 36 higher individual marks which may reflect a higher quality of user experience in parks. The median grade should be higher to reflect the initiatives outlined in the Parks Plan. A mark of A+ would be reflective of parks that perfectly meet the outlined criteria of what defines a quality park. This should be the aim of park maintenance standards and should be reflected in the budget. There are a variety of challenges that may be associated with the current park maintenance standards and the ability for Toronto parks to truly achieve marks that are reflective of the aims in the Parks Plan. Unfortunately, the standards are not released for public or academic scrutiny so it is difficult to pinpoint if the maintenance standards are the cause of the low grades. 10.1 Overall Findings A possible reason for the lower marks is that the park maintenance standards may be too low across the City of Toronto. Park maintenance standards may be lower due to the transition between Summer and Fall standards which would explain the lower marks resulting from observations made in October. Alternatively, the standards may be consistently met; however, on the days that were observed, park maintenance crews may not have yet been present since their last visit. This would explain the lower grades received on the observed day, but it indicates that parks are maintained at too low a level to achieve grades that are reflective of the Parks Plan’s goals. The standards could be acceptable; howevers the frequency of use in dense areas of the City may be underestimated resulting in lower marks. The median grade of a B- may reflect one specific issue or a combination of challenges. The median grade and the lower grades overall may be reflective of the type of survey instrument that was used. The adaption of the NYC Park Report Card may be too strict resulting in the grades being lower. Because the tool and grading system was developed in NYC, the values and standards may not be reflective of Toronto’s park standards. The low median grade may not be reflective of the quality of the park spaces, rather that the survey tool was not appropriately adapted to a different setting outside of NYC. The adapted park report card may not be an effective inspection method for reporting the level of quality of Toronto parks which may explain the low marks the parks received. The adaption of the park report card method may not have been effective because there are different maintenance standards and public expectations in NYC. The adaption of the NYC park report may have been too strict and not provided enough flexibility to include all features functionality and just focused on basic amenities. Therefore, the focus on basic amenities may have been too restrictive and not contributed to an appropriate adaption of the NYC Park Report Card method. For example, this method did not account for local preferences and different budget challenges that are 37 experienced in Toronto. A more appropriate adaptation of the NYC Park Report Card may include more indicators and aesthetic features to help provide a more complete representation of Toronto’s park spaces that may alter grades. 10.1 Overall Findings The parks may have received their lower marks because this research has focused on basic amenities and did not include aesthetic features. The surveyed basic amenities may not be a complete representation of the intent of the definition of quality as outlined by the City in the Parks Plan which would explain the lower grades. During the adaption of the NYC Park Report Card, aesthetic and other features were not included in this analysis as they tend to be arbitrary, potentially controversial and may produce bias. Aesthetic features may have a more positive effect on the park grades if they were included in the observations and provide a more complete representation of quality parks. Other park features can be evaluated for their functionality, such as playgrounds to increase the number of indicators included in the analysis. This will ensure that no one particular feature will have a drastic effect on the park grades. The last reason that the grades may have been low may be reflective of the maintenance standards outlined in the Parks Plan not receiving enough funding. As the Quality of Parks initiative is unfunded according to the 2014 capital budget, this is a major indicator that park quality will not improve in 2014. The parks will continue to reflect the findings and grades from the observations in October 2013 since the Quality of Parks initiative is not being implemented at this time. 2013 service levels remain the same, and the baseline that the park grades provide will stay the same throughout 2014. As an unfunded mandate of the Parks Plan, the Quality of Parks initiative will have no effect in 2014 and will not change drastically until it is funded. The achieved grades represent pre-implementation of the Quality of Parks initiative of the Parks Plan and demonstrate that there is a need to improve the overall quality of basic amenities in parks in the densest wards of Toronto. Lack of funding for this initiative will see no alteration in the quality of Toronto parks despite increased number of users as Toronto’s density increases. This research has provided a benchmark or baseline mark for the parks surveyed representing the basic amenities of benches, passive greenspace, water fountains, bathrooms and pathways. 10.1 Overall Findings These data can be used to compare parks and as a basis for improving the maintenance conducted, increasing the maintenance standards or even altering park design to achieve quality parks. Baselines are beneficial for research and ongoing maintenance to ensure that standards are being met and that parks can truly achieve quality standards once funding is obtained. Baselines or benchmarking can provide a means for comparison within a city or 38 between cities (Daniels, 1996). The use of park report cards for benchmarking the City’s park standards is effective as it is easily understood by the public and professionals as well. The grades for each park are effective forms of communication for the public to understand the current status of their park’s quality. Lower grades may lead the public to lobby their councilor to support the Parks Plan through funding or become involved in a ‘Friends of’ group. Each basic amenity contributes to the overall grades and funding for these specific features must be obtained. To better understand the current status of each basic amenity, more in depth analysis is provided below. Benches Sitting areas achieved the most consistent and highest rankings with nine of the fourteen observed parks attaining the best marks. The remaining parks had minor issues with the quality of seating with only a few examples of broken benches and the main cause of lost marks was the result of graffiti and garbage. To resolve this, more frequent graffiti removal and maintenance and/or higher maintenance standards need to be encouraged in parks located in high density neighbourhoods. It was observed that there were too few benches in the parks that were visited during the research phase. Appendix C provides the number of benches for each park and the ratio of benches to park size. The number of benches in each park was divided by the size of each park to discover the number of benches per hectare which indicates that there are too few benches. This has the potential to reduce the use of park spaces and may negatively target demographic groups, such as seniors who may be more reliant on benches. Having an appropriate number of benches for park space that is reflective of surrounding environments is a necessary feature that must be explored by the City of Toronto as it will ensure inclusivity of all Toronto residents. There currently is no standard provided to ensure that an appropriate number of benches are in each park. From observations, it was found that parks that had 20 or more benchers per hectare appeared to have an appropriate number of seating options without looking too crowded. Therefore 20 benches per hectare is recommended as a minimum bench standard that PFAR should achieve when they gain funding to fully implement the Parks Plan. All but 2 of the parks sampled met this recommendation. Vermont Square far exceeded the recommended minimum and provided 57 benches per hectare. Concerns that this would appear crowded are not the case as the design of the seating is unique. Approximately 30 benches are provided in a barrier that separates and encloses the playground from the rest of the park. Through unique and unobtrusive park design a large number of seating options can successfully be incorporated into parks. Therefore no maximum number of benches is suggested. The City is currently working on a new plan to encourage social gathering places that has a central focus on providing an appropriate number of benches. Pathways The overall grades for pathways were fair, but can be improved as they are an integral part of making parks accessible. Pathways should always achieve the highest score as they ensure that there is equal access to park spaces for all users in dense areas, especially those who may have some mobility challenges, such as caretakers with baby strollers or seniors. Jesse Ketchum Park is an example of a poor score because the park did not have pathways. Only four of the fourteen parks achieved quality ranking on their pathways as they provided paths that were smooth and safe to navigate for all users. The remaining park spaces did not achieve quality rankings as they had cracks and were uneven, but not to the point of being unusable. Parks tended to receive midrange marks for their pathways. Park quality was determined based on the ease of mobility of pedestrians. Cracking resulting from winter and ground shifting and leaves covering the paths were the main reasons for lost marks. Other aspects of mobility should be included in the pathway evaluation in the future including the ease of access for individuals with wheelchairs or strollers. For example, Dufferin Grove Park had a number of staircases throughout the park which reduce accessibility. The level of accessibility should be included within this standard or be a separate evaluation as it is important for the quality of park spaces. The current Parks Plan requires that PFAR meet all accessibility requirements when developing new parks and redeveloping existing parks. Some parks may need redesigning to ensure that park users have safe pathways provided in maintained and manicured areas. Some pathways are not used as they do not follow the direction of user flow. This leads to degradation of passive greenspace unnecessarily. Park redesign can resolve this challenge and lead to quality pathways that follow the direction of user flow and improve the quality of passive greenspace. 39 Benches This plan works in tandem with other basic amenities such as access to bathrooms and water fountains. There are objectives to increase the number of benches, but this is dependent upon capital budget approvals. 40 Despite the Parks Plan aim to increase the number of benches to develop social gathering spaces, the 2014 budget has not provided any means for expanding on this initiative. Toronto has a bench donation system which is a method used to obtain additional seating in parks without being dependent upon yearly budget variations (PFAR, Commemorative Trees and Benches, 2014). This program, however, does not guarantee that the recommended minimum of 20 benches per hectare will be present in parks. There were benches of different quality which can partially be attributed to the variety of styles. As part of the social gathering plan, a uniform bench type is being distributed to all parks as benches require replacing and new benches are being installed. Durability of bench materials must be central to this plan to ensure trouble-free maintenance and longevity of the seating. Quality park benches and increases in the number of benches to the recommended minimum will contribute to the quality of parks in Toronto. The City should target the parks in dense areas to ensure than an appropriate number of benches are provided. Passive Greenspace Passive greenspace consistently required more maintenance than was being provided to achieve marks that indicate quality park space. There was not a single park that had high quality greenspace as indicated by the determinants in Tables 4-8. Each park’s greenspace received failing marks that resulted from poor maintenance of garbage, graffiti and/or dead tree branches. This is a challenging basic amenity to evaluate as it is the largest and varies the most in quality. There tend to be sections of quality greenspace and sections that are less well-maintained. The approach used during the evaluation was to give the mark reflective of the worst sections of passive greenspace as it lessens the overall quality of a park. The time of year of the observations highly impacts the quality of the passive greenspace. During the summer months, passive greenspace may achieve higher marks as a result of less moisture causing muddy grounds as people walk through. Higher maintenance standards in the summer would also result in lower levels of garbage being present. As well, there are strict grass lengths that the City maintains in an attempt to ensure that parks’ passive greenspaces are equally and highly maintained. Therefore it is concluded that passive greenspace marks will fluctuate according to seasonality. 41 Water Fountains Every park that was visited had a drinking fountain, but only two of the fourteen parks analysed received complete marks for having quality fountains. With three parks receiving the worst weighting, the remaining 10 parks achieved average levels. This indicates that the majority of parks have functioning water fountains that have some garbage or graffiti that may reduce the visual quality and use of this park amenity. Since the social gathering space plans that are currently being developed rely on water fountains being present in park spaces, it is necessary to ensure that they are useable and maintained to a high standard. Water fountains contribute to the overall quality of parks and can easily contribute to parks achieving a top grade in neighbourhoods that are becoming denser. Bathrooms Of the fourteen parks analysed, only four parks had bathrooms. This is a natural constraint on the length of time that park users can stay in parks. Parks without bathrooms also reduce the likelihood of being used as destinations for social gatherings. PFAR does not have requirements stating that parks of a certain class or size require on-site bathroom access. It may be assumed that parks in dense urban areas of Toronto will have access to bathrooms in surrounding private establishments; however, this can still impact the level of use and quality of parks. Parks that have community centers may provide access for public use but this was not evaluated as part of this study. If there was no access to separate bathrooms outside of a community center, no marks were awarded because it was not part of the park report card method as adapted from NYC. It was determined that parks without bathrooms make visits more challenging for individuals who may have mobility difficulties because they will always have to leave the park to use a bathroom. This has a negative impact on the use of parks and the benefits that can be drawn from them. Providing bathrooms that are available for public use and are accessible will ensure that parks can be used as social a gathering space. Resulting from the consultations that led to the development of the Parks Plan it was noted that the public wanted improved access to bathrooms within parks. Therefore parks need to be designed and redeveloped in a way that will encourage social gathering spaces and ensure that the public has access to bathrooms. The bathrooms that are built must be designed to be easy to clean and maintain to ensure that they contribute to the quality of the park space. Of the parks that did have bathrooms, they were mostly closed as the plumbing needed to be shut off to prevent pipes from freezing during the time of year that the field research occurred. 42 There were two bathrooms that were open to the public during the time that site visits were being conducted and they were well stocked and clean which contributed to those parks gaining higher marks. This may be reflective of the time of year and reduced use of outdoor space due to temperature changes and more frequent inclement weather. Bathrooms During busier times, the cleanliness of the bathrooms may be reduced due to the intensity of use. Ongoing maintenance observations need to occur to explore the continuing quality and upkeep of the bathrooms that contribute to the overall quality of parks in densely populated cities. 10.2 Application of Park Report Cards The use of park reports card in Toronto parks was easy to execute and can be conducted by maintenance staff or specially trained PFAR staff. This could occur in association with the park inspection program to create a legible and easily understood tool to measure the quality of parks based on the current park standards. The Parks Plan outlines plans to digitize the maintenance and repair component of the park inspection program to ensure that the process is easier and internally recorded. Combining the digitization of the parks inspection program and the introduction of a parks report card as the means of evaluation would be an aid to plan parks. During observations, the specific evaluation of each park was easy to conduct as it required an analysis of the entire park space. This involved walking throughout the park and observing the space for the indicators in Tables 4-8. Quality park spaces are truly reflective of their basic amenities which became apparent during a time of year (October) that did not provide aesthetically pleasing surroundings. Park reports that focused on basic amenity space, however, are not reflective of the complete picture of park spaces. Park report cards should be broader and encompass a variety of features including general care of aesthetic features but are not an evaluation of the artistic nature of those features. Aesthetic components of parks are important features and could be included in the application of a park report card, but a means to avoid bias would need to be implemented. Ultimately, the functionality and maintenance level of all park features should be evaluated and included in park report cards to provide a greater understanding of the quality of park spaces. Park report cards have the potential to increase the public awareness of the quality of parks within Toronto. The ease of legibility of a park grade may encourage locals to seek improvements to the quality of their parks if they repeatedly score poorly. The parks that achieved the lowest grades during this evaluation, such as Jesse Ketchum Park and Campbell Avenue Playground, could be improved by activism that may result from an understanding of 43 the reduced quality of parks. Grades should be made public so residents could monitor the quality of their parks. 10.2 Application of Park Report Cards Poor park quality could lead to resident’s lobbying their councilors to ensure that appropriate funding is secured to implement the Quality of Park initiative of the Parks Plan. Park report cards have the potential to be used as a tool to engage political bodies to ensure that appropriate budget allocations are made to support the Parks Plan. The application of park report cards in Toronto will be useful for knowing what parks need more frequent maintenance to ensure that quality parks are available for all residents. Park report cards can be incorporated into the parks inspection program that is being digitized which will increase the ease of access that the public has to this type of information. Public knowledge about the quality of local parks may encourage citizen participation in lobbying councilors to secure funding for the implementation of the Quality of Park initiative. 10.3 Carrying Capacity The current amount of park and green space in Toronto is approximately 19,760 acres and the required amount of space needed to meet the standard provided by Charles Downing of 2 acres per 300 residents in cities over 1 million residents is calculated to be 17,433 acres (See Appendix C). This is reflective of the 2011 census data for the City of Toronto’s population. Therefore the City has an appropriate amount of park and green space for its residents; however questions of quality and accessibility remain. The quality standards of basic amenities have been explored briefly in this research and it has found that there are basic amenities that require improvement since parks have not achieved high grades in Toronto’s densest wards. Accessibility has not been a focus of this research, but is imperative in ensuring a high quality of life for all residents is achieved and maintained. A large proportion of the green space included in the total acreage is the ravine system which may not be easily accessible to all residents; further research can explore the accessibility of the ravines. City of Toronto does not provide statistics that distinguish between the area of greenspace that includes the ravine system and the area of greenspace that is representative of parks. Residents in dense urban areas must have access to green and park space as there are many outlined benefits to it that must be equally distributed. The City of Toronto must work to maintain the ratio of park and green space to residents at an acceptable standard especially within dense areas of the City. City staff must ensure that there is enough park space for residents that do not have access to private green space and that the space is not being degraded by overuse. Utilizing a carrying capacity standard can resolve challenges of obtaining funding for parks as it is a quantitative 44 indicator that can trigger demand for new park space otherwise park quality will decrease as a result of overuse. This calculation should use the total area of park space excluding the ravine system to have a more representative indicator. Alternatively, this calculation can be completed on a ward level, taking into account all publicly accessible greenspace, which can include the ravine system and privately owned public open spaces. 11.0 Recommendations Upon analysis of the collected data a variety of recommendations are suggested for the City of Toronto’s Parks, Forestry and Recreation Department that should be made central to their implementation of the Parks Plan to ensure that their initiative of Quality of Parks is achieved. The park report card method that was tested encountered a number of challenges in parks under strain from density and below are a series of recommendations to resolve them. Overall, the park report card method was effective and should be implemented at the same time as the digitization of the new parks inspection program occurs. Only one method is needed to inspect parks; however, the findings should be made public through the park report card format. There are a number of recommendations for implementing a park report card method that build from the adapted trial of the NYC Park Report Card. It is important that the park report card be directly linked to Toronto’s park maintenance standards which can be used as the basis for evaluating the quality of parks. Therefore, the methodology from the NYC Park Report Card can be used, but new weighting scales for each amenity being evaluated must be developed based on consultations in Toronto that would be conducted by PFAR and Community Planning together. This monitoring method must be implemented across the City to ensure that the maintenance standards are achieving quality parks equitably. Moving forward, funding through the 2015 operating budget is necessary to implement the park report card method as soon as possible. In the future, it is recommended that Council should only approve plans that will receive funding so they can be implemented fully. From the field observations a number of recommendations for the basic amenities that were observed became apparent and needs to be implemented by PFAR. There is a need to include more criteria for each basic amenity being evaluated as well as new features being evaluated. The pathway indicator must not only focus on the maintenance and repairs of the physical structures, but also the modes of connecting the pathways such as ramps with handrails instead of stairs. Criteria focusing on accessibility must be included throughout the pathways evaluations. When evaluating benches and seating, there needs to be a criterion focusing on meeting a minimum of 20 benches per hectare. 10.3 Carrying Capacity A carrying capacity calculation should work in tandem with a park monitoring system such as a park report card as they both aim to maintain quality park space in cities experiencing density increases. This will aid City staff and Councilors in directing funding and lobbying for budget allocations to obtain new park land, to prevent degradation of existing park space and improve the quality of existing parks through design and maintenance standards. 45 11.0 Recommendations The city plans on implementing a uniform bench design; however, it was noted that creative designs, such as found in Vermont Square Park, can create lots of seating effectively. A balance must be achieved between restricting park design and ease of maintenance and repairs of benches. 46 Despite the City being aware that it must improve the features and amenities in parks, it must be stressed again since it is of great importance to achieve quality parks. The following are two of the most apparent areas of improvement based on the field research conducted. Despite the large capital investment, there needs to be a greater number of bathrooms in parks to ensure that the public can use parks for longer periods of time. An increase in bathrooms can be triggered by the size of parks. As well, there is a need to explore options to increase the longevity of turf which will improve the quality of passive greenspace. During months when it is extremely wet, techniques to prevent degradation of the passive greenspace must be explored, such as putting up temporary barriers to restrict the public from walking off pathways. PFAR must work with the community to seek appropriate solutions to these challenges. Since the City of Toronto has not released its maintenance standards in the past and their current standards are under review, it makes it impossible to truly evaluate them. Limited support from the City of Toronto has limited the scope of the analysis that has been completed regarding the current maintenance standards. Despite this, the method used to evaluate the parks clearly indicates that the parks are below the standard to meet the Quality of Parks initiative as defined in the Parks Plan. It is recommended that the maintenance standards need to be met to achieve quality park space and that PFAR should make the standards and results public to increase public awareness. The Quality of Parks initiative that is presently unfunded according to the 2014 budget must receive funding as the park report grades are low and reflective of poor quality parks (PFAR, 2014 Operating Budget Highlights , 2014). This lack of implementation may have occurred for a variety of reasons; however, the quality of Toronto’s parks in the densest wards is being negatively affected. 12.0 Future Research Park planning research can be expanded in a variety of formats to continue to explore the value of quality park space and the techniques required to secure and achieve quality parks. This research can be completed again with variations, such as a larger sample, case studies of specific parks or comparing urban and suburban park quality. Potential findings may confirm the viability of applying park report cards as a park inspection method in cities that are becoming denser. This research can be replicated using the same sample but the report card can be adapted to include more park features including aesthetic features which were originally excluded due to their arbitrariness. Aesthetic features can be included in future analysis as long as the functionality and maintenance of the feature is assessed. Precautions against bias must be taken to ensure that aesthetic features are not being evaluated based on personal preference. The park report card is a method that can be replicated easily outside of Toronto and be applied to the Greater Golden Horseshoe Region given the provincially legislated growth. Comparing park report cards between cities has the potential to encourage a positive form of competition among cities using a standardized evaluation and report method. Cities can become known for having quality parks and residents within the Greater Golden Horseshoe would understand the ranking as it would be a consistent benchmarking tool across the region. There are other benchmarking systems that can be tested in Toronto or the Greater Golden Horseshoe to guarantee comparability. PROGRAIS is an international benchmarking system from the National Recreation and Parks Association in the United States of America that ranks national parks. Applying a comparable approach can ensure comparisons of parks internationally. This would be beneficial to cities that are experiencing growth and density pressures as lessons can be learned from cities that consistently achieve high marks for quality park spaces. Utilizing a monitoring system such as the park report card method can lead to improved community awareness and engagement as the public would seek to improve the quality of their local parks. Confirming that the public can comprehend the park report grades across different demographics and ethnic groups is a necessary area to explore to ensure that the park report card method can be used effectively to engage communities. 11.0 Recommendations The Quality of Parks initiative must be implemented to ensure that Toronto residents, businesses and the City reap the benefits that parks can provide. This initiative can be established City-wide through phasing that targets the densest areas first and then gradually is provided to less dense areas. This may seem unfair initially; however those individuals living in the densest parts of the City do not have access to private greenspace as readily as those individuals who live in single detached homes in lower density areas of the City. To ensure that improvements are being achieved, a park report card method should be implemented by PFAR so that the public can be aware of the true quality of their parks according to the City’s standards. 47 12.0 Future Research Finding additional methods to engage the community to support park improvements can be the focus of future research. 48 Ensuring that parks reflect the public’s values and opinions is necessary so developing strategies to engage the public is necessary and would be a valuable form of future research. Ensuring that parks reflect the public’s values and opinions is necessary so developing strategies to engage the public is necessary and would be a valuable form of future research. Methods to improve the quality of park spaces outside of the municipal budget should be explored in the future. As there are limited municipal finances, partnerships with donors and ‘Friends of’ groups must be explored for their fundraising feasibility to improve local park spaces. Creative methods to secure funding that is directed at parks must be investigated and implemented to supplement the budget since there are a number of unfunded mandates in the Parks Plan. Ensuring that all mandates of the Parks Plan can be funded through a combination of funding tools should be the focus of future research. Future research can focus on whether Toronto parks are meeting the needs of minority groups. Minority groups can include newcomers to the City, ethnic minorities, seniors, teenagers and other groups that historically have not been included when planning and designing park spaces. Basic amenity features and additional uses that are specific to welcoming minority groups must be included in parks to ensure that parks spaces are meeting everyone’s needs. An accessibility study of parks in Toronto should be conducted as the pathways did not achieve the highest ranking. Through park observations there were a number of barriers to individuals requiring assistance. For example, stairs and lack of ramps with support rails have the potential to negatively impact some users who require assistance. A study that focuses on the accessibility or the overall ease of use of parks spaces must be conducted to ensure that all residents can use parks safely. Features that improve ease of use such as benches must be properly studied in the future. Ongoing efforts to increase the number of benches in Toronto must be studied for their effectiveness as the number of benches is brought up to achieve the recommended minimum of 20 benches per hectare. 12.0 Future Research Exploring a viable carrying capacity calculation for park spaces will be valuable to future park planning as Toronto and other cities in the Greater Golden Horseshoe continue to become denser. A carrying capacity indicator would ensure that when parks are being overused, cities would need to limit development in that area until park space is obtained or alternative solutions are reached. It would also indicate park spaces where quality would be decreased as a result of over use. Therefore a park carrying capacity indicator that is specific to Toronto is necessary to ensure that quality park spaces are maintained and that all City residents have access to quality park spaces. This will also ensure that tourists can enjoy access to a variety of quality park spaces even in the densest urban areas of Toronto. 49 As Toronto continues to experience density pressures, finding different forms of open space such as privately owned public spaces will become more prominent. San Francisco has completed an inventory of all privately owned public spaces and has placed plaques at each entrance. A comparable analysis of this work could be conducted in Toronto to ensure that Toronto residents and tourists would be able to utilize all available public spaces. Park planning research can explore a variety of avenues. The most pertinent areas to study are the ones pertaining to the park improvements within Toronto to advance the quality of park spaces. Developing strategies, partnerships and implementation methods in areas of Toronto that are experiencing density pressures must be central to moving forward with the Parks Plan’s (2013-2017) mandate to achieve quality parks. 50 Appendices Appendices Appendix A - Toronto Park Categorization Park Type Description Parkette Less than 0.5 hectares. Used by local residents for passive recreation. Supplement and not replace larger park spaces. Neighbourhood park Generally not less than 0.5 hectares. Used locally for both passive and some active recreation purposes. Community park Generally not less than 3 hectares. Serves several neighbourhoods. Provides passive, active and programmed recreation. District park Generally not less than 5 hectares. Serves several communities. Higher levels of use, has parking when other smaller ones do not. Passive and active recreation; a hub for special functions and programs. City park Serves users from across the city. Generally not less than 15 hectares. Destination for tourists. Can contain natural or cultural heritage features. Ie. Woodbine Beach. Note: Descriptions provided from the Parks Plan. Appendix A - Toronto Park Categorization 51 51 Appendix B – Park Sample and Ward Densities Park Name Ward Address Size(ha) Size(acres) Description Masaryk Park 14 212 Cowan Ave 0.5 1.235 Community garden, wading pool, playground. Campbell Avenue Playground 18 255 Campbell Ave 1.3 3.211 Three sports fields, wading pool, basketball court and playground. MacGregor Playground 18 346 Lansdowne Ave 1.4 3.458 Lighted ball diamond, a field house, a basketball court, a wading pool and a children's playground. McCormick Park 18 66 Sheridan Ave 1.5 3.705 A ball diamond, two basketball courts, a wading pool and a children’s playground. Dufferin Grove Park 18 875 Dufferin St 5.3 13.091 Large mature forest canopy. The park features a multi- purpose sports field, a basketball court, a picnic area, a wading pool and a children's playground and artificial ice rink. Dovercourt Park 18 155 Bartlett Ave 2.4 5.928 Lighted ball diamond, wading pool, two tennis courts and playground and has street parking around outskirts. Vermont Square 20 819 Palmerston Ave 1.5 3.705 No description. Bellevue Park 20 5 Bellevue Ave 0.49 1.2103 Wading pool, playground and water fountain. Alexandra Park 20 275 Bathurst 2.7 6.669 Featuring a children's playground, community garden, volleyball court. Located at Scadding Court. Grange Park 20 317 Dundas St W 1.8 4.446 Owned by AGO but run my PFAR. Canoe Landing 20 95 Fort York Blvd 3.1 7.657 Walking paths and multipurpose sports fields. Little Norway Park 20 659 Queens Quay 2.4 5.928 Ball diamond, wading pool, children's playground and walking paths. Appendices Moss Park 27 150 Sherbourne St 3.4 8.398 A lighted ball diamond, two tennis courts, a basket ball court, a wading pool and a children`s playground. Jesse Ketchum Park 27 1310 Bay St 1.2 2.964 Mature tree canopy. Note: Descriptions are based on descriptions from the PFAR website. Appendix B – Park Sample and Ward Densities 52 Ward Density (thousand people/km2) 14 9.93 18 9.49 20 10.27 27 9.49 53 53 Appendix C – Carrying Capacity Total Toronto Park Hectares 8000 Acres 19760 Total Toronto Population (2011 Census) 2615060 Carrying Capacity Standard for Cities over 1 million residents = 2 acres/300 residents 17433.7 Appendix D Appendix C – Carrying Capacity 54 Acres 19760 Total Toronto Population (2011 Census) 2615060 Carrying Capacity Standard for Cities over 1 million residents = 2 acres/300 residents 17433.7 Appendix D 54 55 55 Works Cited Association, N. R. (2010). Synopsis of 2010 Research Papers. United States: National Recreation and Parks Association. Barrow, G. Y. (1979). Zonation and Carrying Capactiy Estimates in Canadian Park Planning. England: Applied Science Publishers. Boland, G. C. (2004). Defining the Sustainable Park: A Fifth Model for Urban Parks. Landscape Journal , 102-120. Canada, S. (2011). Census Profile. Retrieved February 3, 2014, from Statistics Canada: http://www12.statcan.gc.ca/census-recensement/2011/dp- pd/prof/details/page.cfm?Lang=E&Geo1=CSD&Code1=3520005&Geo2=PR&Code2=35&Data= Count&SearchText=Toronto&SearchType=Begins&SearchPR=01&B1=All&GeoLevel=PR&Geo Code=3520005 Chen, C. J. (2003). Comprehensive greenspace planning based on landscape ecology principles in compact Nanjing city, China. Landscape and Urban Planning , 1-22. (1998). Chicago City Space Plan. Chicago: City of Chicago. (1998). Chicago City Space Plan. Chicago: City of Chicago. Christiansen, M. L. (1977). Park Planning Handbook. Canada: John Wiley & Sons Inc. CMAP. (n.d.). Retrieved March 3, 2013, from Chicago Metropolitan Agency for Planning: http://www.cmap.illinois.gov/ CMAP. (2010). Go To 2040. Chicago: Chicago Metropolitan Agency for Planning. Committee, E. D. (2004). Toronto Parks and Recreation Strategic Plan Our Common Ground. Toronto: City of Toronto. Costa, A. K. (2011). Planning of Recreation Parks in the University Campus. Journal of Environment Protection , 365-370. Crawford, A. W. (1910). City Planning and Philadelphia Parks. Annals of the American Academy of Political and Social Science , 71-80. Crompton, J. (2001). The Impact of Parks on Property Values: A Review of the Empircal Evidence. Journal of Leisure Research , 1-33. Daniels, S. (1996). Benchmarking. Emerald , 18-20. Daniels, S. (1996). Benchmarking. Emerald , 18-20. Davidoff, P. (1965). Advocacy and Pluralism in Planning. Journal of the American Institute of Planners , 331 - 338. Davidoff, P. (1978). The redistributive function of planning: Creating greater equity among citizens of communities. In R. B. Sternlieb, Planning Theory in the 1980s (pp. 69-72). New Jersey: Rutgers University. 56 Deakin, E. (1999). Social Equity in Planning. Berkeley Planning Journal , 1-5. (2010). Development of a City-wide Parks Plan. Toronto: City of Toronto. DiPalermo, C. (2007). The Report Card on Parks 2007. New York City: New Yorkers for Parks. DiPalermo, C. (2007). The Report Card on Parks 2007. New York City: New Yorkers for Parks. Fainstein, S. S. (2000). New Directions in Planning Theory. Urban Affairs Review , 451-478. Fainstein, S. S. (2000). New Directions in Planning Theory. Urban Affairs Review , 451-478. Francis, M. (1987). Some different meanings atached to a city park and community gardens. Landscape Journal , 101-112. Frumkin, H. (2003). Healthy Places: Exploring the Evidence. Works Cited American Jounal of Public Health , 1451-1456. Geoffrey C. Godbey, L. L. (2005). Contributions of Leisure Studies and Recreation and Park Management Research to the Active Living Agenda. American Journal of Preventive Medicine , 150-159. Gobster, P. (2001). Visions of nature: conflict and compatibility in urban park restoration. Landscape and Urban Planning , 35-51. (2006). Greenspace Master Plan. Ottawa: Ottawa. (2006). Greenspace Master Plan. Ottawa: Ottawa. Hartig, T. (2004). Restorative Environments. In C. Spielberger, Encyclopedia of Applied Psychology (pp. 273-279). San Diego: Academic Press. Harvey, D. (2010). Fertile Ground for New Thinking: Improving Toronto's Parks. Toronto: Metcalf Foundation. Hendon, J. K. (1967). Land Values Adjacent to an Urban Neighbourhood Park. Land Economics , 357-360. Highline, F. o. (2013). The Highline. Retrieved September 26, 2013, from The Highline: http://www.thehighline.org/ Infrastructure, M. o. (2013, February 13). Places to Grow. Retrieved November 4, 2013, from https://www.placestogrow.ca/index.php Kevin Thwaites, I. S. (2010). Towards socially restorative urbanism: exploring social and spatial implications for urban restorative experience. Landscape Review , 26-39. Loukaitou-Sideris, A. (1995). Urban Form and Social Context: Cultural Differentiation in the Uses of Urban Parks. Journal of Planning Education and Research , 89-102. Mattocks, R. H. (1937). Park Planning: Part II. The Town Review , 243-270. Menendez, S. d. (2005). Estimating the non-market benefits of an urban park: Does proximity matter? Land Use Policy , 1-10. 57 Merriam-Webster. (2014). Benchmark. Retrieved March 15, 2014, from Merriam-Webster: http://www.merriam-webster.com/dictionary/benchmark Michaels, S. (2014). Urban HEART @ Toronto. Toronto: Toronto Community Health Profiles Partnership. Moeller, J. (1965). Standards for Outdoor Recreational Areas. American Society of Planning Officials. Olshansky, R. (2003). Planning for Public Parks. Journal of the American Planning Association , 318. Ontario. (2006). Growth Plan for the Greater Golden Horseshoe. Ministry of Infrastructure. Ontario. (2006). Growth Plan for the Greater Golden Horseshoe. Ministry of Infrast Parks, C. S. (2005). Park and Recreation Trends in California. California: National Park Service. Parks, N. (2013). Parks Inspection Program. Retrieved January 14, 2013, from NYC Parks: http://www.nycgovparks.org/park-features/parks-inspection-program Parks, N. Y. (2013). Report Cards. Retrieved September 20, 2013, from New Yorkers for Parks: http://www.ny4p.org/research/report-cards Peterson, G. (1993). Parks and Recreation: Fundamental for this City. Parks and Recreation , 48-51. PFAR. (2014). 2014 Budget at a Glance. Toronto: City of Toronto. PFAR. (2014). 2014 Budget at a Glance. Toronto: City of Toronto. PFAR. (2014). 2014 Budget at a Glance. Toronto: City of Toronto. PFAR. (2014). 2014 Operating Budget Highlights . Works Cited Retrieved February 4, 2014, from Parks, Forestry & Recreation : PFAR. (2014). 2014 Operating Budget Highlights . Retrieved February 4, 2014, from Parks, Forestry & Recreation : y http://www1.toronto.ca/wps/portal/contentonly?vgnextoid=a2a68d5c19c52410VgnVCM1000007 1d60f89RCRD&vgnextfmt=default http://www1.toronto.ca/wps/portal/contentonly?vgnextoid=a2a68d5c19c52410VgnVCM1000007 1d60f89RCRD&vgnextfmt=default PFAR. (2014). Commemorative Trees and Benches. Retrieved January 31, 2014, from City of Toronto: PFAR. (2014). Commemorative Trees and Benches. Retrieved January 31, 2014, from City of Toronto: https://www1.toronto.ca/wps/portal/contentonly?vgnextoid=b7a1ede252801410VgnVCM100000 71d60f89RCRD&vgnextchannel=470bdada600f0410VgnVCM10000071d60f89RCRD https://www1.toronto.ca/wps/portal/contentonly?vgnextoid=b7a1ede252801410VgnVCM100000 71d60f89RCRD&vgnextchannel=470bdada600f0410VgnVCM10000071d60f89RCRD PFAR. (2013). Staff report for action on the Parks Plan. Toronto: Toronto. PFAR. (2013). Staff report for action on the Parks Plan. Toronto: Toronto. PFAR. (2014). Toronto 2014 Budget. Toronto: City of Toronto. PFAR. (2014). Toronto 2014 Budget. Toronto: City of Toronto. PFAR. (2014). Toronto 2014 Budget. Toronto: City of Toronto. PFAR. (2013). Toronto's geography. Retrieved February 3, 2014, from City of Toronto: http://www.toronto.ca/toronto_facts/geography.htm PFAR. (2013). Toronto's geography. Retrieved February 3, 2014, from City of Toronto: http://www.toronto.ca/toronto_facts/geography.htm Rutledge, A. J. (1971). Anatomy of a Park. New York City: McGraw-Hill, Inc. Rutledge, A. J. (1971). Anatomy of a Park. New York City: McGraw-Hill, Inc. Schanuel, A. P. (2005). The Economics of Urban Park Planning. Park & Recreation , 64-67. Schanuel, A. P. (2005). The Economics of Urban Park Planning. Park & Recreation , 64-67. 58 Stephen Page, K. N. (1994). Managing Urban Parks: User Perspectives and Local Leisure Needs in the 1990s. The Service Industries Journal , 216-238. T. Hartig, M. M. (1991). Restorative Effects of Natural Environment Experiences. Environment and Behaviour , 3-26. Timmermans, A. K. (2006). Heterogeneity in Urban Park Use of Aging Visitors: A Latent Class Analysis. Greenspace Project , 1-15. (2010). Toronto Official Plan. Toronto: City of Toronto. (2010). Toronto Official Plan. Toronto: City of Toronto. Toronto. (2013). Parks Plan Summary. Toronto: Toronto. Toronto. (2013). Parks Plan Summary. Toronto: Toronto. TRCA. (2013). The Living City Policies. Retrieved January 14, 2013, from Toronto and Region Conservation Authority: http://trca.on.ca/the-living-city/public-consultations/the-living-city- policies.dot Ulrich, R. (1979). Visual Landscapes and Psychological Well-being. Landscape Research , 17- 23. 59 59
https://openalex.org/W3037382805	https://europepmc.org/articles/pmc7344744?pdf=render	English	null	Spousal Concordance of Physical Frailty in Older Korean Couples	International journal of environmental research and public health/International journal of environmental research and public health	2,020	cc-by	6,639	Received: 23 May 2020; Accepted: 23 June 2020; Published: 25 June 2020 Abstract: Marital status is an important risk factor for physical frailty. However, there are limited data on spousal concordance of physical frailty among married couples. Here, we evaluate the spousal concordance of frailty as deﬁned by the Fried frailty phenotype and speciﬁc phenotype components that contribute to this association. Data on 315 married couples (630 individuals) aged between 70 and 84 years were obtained from the Korean Frailty and Aging Cohort Study (KFACS). Multivariate logistic regressions were used for the analysis. After adjusting for covariates (age, body mass index, education, house ownership, comorbidity, cognition, depressive symptoms, cohabitation with adult children for both partners), a husband’s frailty was positively associated with his wife’s frailty (odds ratio (OR) 3.34, 95% conﬁdence interval (CI) 1.04–10.73, p < 0.05), and a wife’s frailty was signiﬁcantly associated with her husband’s frailty (OR 4.62, 95% CI 1.31–16.33, p < 0.05), indicating a greater eﬀect of the frailty status of the spouse among women than among men. Among the ﬁve components of the Fried frailty phenotype, weight loss, slowness, and exhaustion were the main contributing factors to the spousal association for frailty. In conclusion, having a frail spouse is a strong and independent risk factor for frailty among community-living older adults. Keywords: frailty; spousal concordance; aging International Journal of Environmental Research and Public Health Received: 23 May 2020; Accepted: 23 June 2020; Published: 25 June 2020 Int. J. Environ. Res. Public Health 2020, 17, 4574; doi:10.3390/ijerph17124574 Article Spousal Concordance of Physical Frailty in Older Korean Couples Suah Kang 1 , Miji Kim 2,* and Chang Won Won 3,* 1 Department of Medicine, College of Medicine, Kyung Hee University, Seoul 02447, Korea; skang39@khu.ac.kr 2 Department of Biomedical Science and Technology, College of Medicine, East-West Medical Research Institute, Kyung Hee University, Seoul 02447, Korea Institute, Kyung Hee University, Seoul 02447, Korea 3 Elderly Frailty Research Center, Department of Family Medicine, College of Medicine, Kyung Hee University, Seoul 02447, Korea y * Correspondence: mijiak@khu.ac.kr (M.K.); chunwon62@naver.com (C.W.W.); Tel.: +82-2-958-2840 (M.K.); +82-2-958-8700 (C.W.W.); Fax: +82-2-958-2836 (M.K.); +82-2-958-8699 (C.W.W.) * Correspondence: mijiak@khu.ac.kr (M.K.); chunwon62@naver.com (C.W.W.); Tel.: +82-2-958-2840 (M.K.); +82-2-958-8700 (C.W.W.); Fax: +82-2-958-2836 (M.K.); +82-2-958-8699 (C.W.W.) www.mdpi.com/journal/ijerph 2.1. Participants Data were mainly derived from the Korean Frailty and Aging Cohort Study (KFACS) 2016 and 2017 baseline database. The KFACS is a 2-year, nationwide, multicenter, prospective cohort study to identify risk factors and outcomes of frailty and develop interventions for prevention and management. The KFACS recruited 3014 community-living elderly individuals aged between 70 and 84 years in 2016 and 2017. The participants were recruited from 10 study centers across diﬀerent regions and diﬀerent residential locations (urban, suburban, and rural) [20]. Each center recruited participants using quota sampling stratiﬁed by age (70–74, 75–79, and 80–84 years with a ratio of 6:5:4) and sex (male and female with same ratio) [20]. Participants were recruited from diverse settings (local senior welfare centers, community health centers, apartments, housing complexes, and outpatient clinics) to minimize selection bias [20]. The inclusion criteria for KFACS participants were as follows: aged 70–84 years, currently living in the community, having no plans to move out in the next 2 years, and having no problems with communication and no prior dementia diagnosis [20]. In this case, “move out” refers to relocating to areas outside the three neighboring towns [20]. Exclusion criteria included individuals with diﬃculty in giving opinion, those unable to comply with the study requirements, or those deemed inappropriate on the basis of the ﬁndings or evaluation by the researcher. The detailed study design has been published previously [20]. Spousal pairs were identiﬁed by marital status, cohabitation status, and home address. Subjects were considered spouses if both male and female adults reported being married, currently lived with their spouse, and had the same home address. A total of 333 samples of married couples were conﬁrmed to be eligible for the present study. The ﬁnal analysis included 315 married couples, after excluding 18 with missing data, for assessing the Fried physical frailty phenotype. The Clinical Research Ethics Committee of the Kyung Hee University Hospital approved the KFACS protocol (institutional review board (IRB) number 2015-12-103). The present study was exempt from the requirement for IRB approval by the Clinical Research Ethics Committee of the Kyung Hee University Medical Center (IRB number: 2020-05-066). 1. Introduction The aged population is growing rapidly. By 2050, the proportion of the population over 60 years old will almost double from 12% to 22%, reaching up to 434 million worldwide [1]. Frailty has become an area of interest for those concerned with healthy aging [2,3]. Frailty is a state of functional decline and increased vulnerability, commonly deﬁned by the Fried phenotype model and the frailty index [4–6]. The Fried phenotype model is a physical frailty criterion, and the frailty index is a cumulative health deﬁcits index including physical, psychological, mental, and social functions [4,6]. In community-dwelling adults aged ≥65 years, the estimated prevalence of frailty is around 10%, and the incidence may increase with age [7,8]. Frailty may have a high health care burden since it is associated with increased morbidity and mortality in the elderly [9,10]. Frailty is associated with sociodemographic, physical, biological, lifestyle, and psychological risk factors [11]. A recent study identiﬁed marital status as an important risk factor for physical frailty [12]. Most caregivers of frail older adults are their spouses [13–15]. Spousal caregivers experience reduced quality of life because of their restricted participation in daily activities [16]. There are negative impacts Int. J. Environ. Res. Public Health 2020, 17, 4574; doi:10.3390/ijerph17124574 www.mdpi.com/journal/ijerph 2 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 on the physical and mental health of spousal caregivers [17]. Furthermore, in a case–control study, frailty status worsened in one-third of the spousal caregivers of older patients with cognitive and functional impairment [18]. However, limited studies have examined the spousal correlation of frailty among community-living older adults. One study found a positive correlation between an individual’s frailty and the spouse’s frailty [19]. However, the study failed to control for the major risk factors for frailty, including comorbid conditions, cognitive impairment, and depression. It is unclear to what extent having a frail spouse contributes to one’s frailty status after adjusting for these risk factors. In this study, we evaluated the spousal concordance of frailty among older adults and determined the speciﬁc frailty phenotype components that contribute to this association. 3. Results The demographic and health characteristics of the study population (630 subjects, or 315 married couples) are listed in Table 1. The mean ages of husbands and wives were 77.6 ± 3.4 and 74.7 ± 3.6 years, respectively. Husbands had a lower mean BMI (23.7 ± 3.0 kg/m2) and higher education status (77.1%) than wives. Wives were more likely to have house ownership (31.1%). There was no signiﬁcant diﬀerence in the current employment status. Wives had higher prevalence of cognitive impairment (24.1%), comorbidities (61.2%), and depressive symptoms (23.5%) than husbands. In addition, the demographic and health characteristics of the frail and nonfrail groups of study participants are presented in Appendix A. p pp Table 2 shows the frailty status determined by the Fried frailty phenotype and its ﬁve components by sex. Subjects classiﬁed as robust or prefrail in the Fried frailty phenotype were considered nonfrail. Frailty was more prevalent among husbands than wives; 35 (11.1%) husbands and 28 (8.9%) wives were classiﬁed as frail. The percentages of all ﬁve components except exhaustion were higher in husbands. Tables 3 and 4 summarize the multivariate logistic regression models examining the spousal association of frailty and for the ﬁve components of the Fried frailty phenotype. The wife’s frailty was signiﬁcantly associated with the frailty of her husband in all models (OR 3.13–4.76, p < 0.05). Similarly, the husband’s frailty was positively associated with the frailty of his wife in all models (OR 2.91–4.76, p < 0.05) except model 3 (OR 3.19, 95% CI 1.00–10.22, p = 0.05). After adjusting for possible covariates, a frail husband had 4.62 odds of having a frail wife, and a frail wife had 3.34 odds of having a frail husband (model 4). While adjusting for other covariates reduced the odds ratio, adjusting for depressive symptoms (model 3) and cohabitation with adult children (model 4) increased the odds ratio. Table 2 shows the frailty status determined by the Fried frailty phenotype and its ﬁve components by sex. Subjects classiﬁed as robust or prefrail in the Fried frailty phenotype were considered nonfrail. Frailty was more prevalent among husbands than wives; 35 (11.1%) husbands and 28 (8.9%) wives were classiﬁed as frail. The percentages of all ﬁve components except exhaustion were higher in husbands. Tables 3 and 4 summarize the multivariate logistic regression models examining the spousal association of frailty and for the ﬁve components of the Fried frailty phenotype. 2.4. Statistical Analysis All analyses were conducted using SPSS (ver. 25.0; IBM Corp., Armonk, NY, USA). A comparison of characteristics by sex was performed using the chi-squared test for categorical variables and the independent t-test for continuous variables. The spousal association of frailty and the ﬁve components of the Fried frailty phenotype were analyzed using multivariate logistic regression models and described by odds ratios (ORs) with 95% conﬁdence intervals (CIs). Four models were considered in addition to the unadjusted model: (1) adjusted for age; (2) additionally adjusted for BMI, education, house ownership, comorbidity, and cognition; (3) additionally adjusted for depressive symptoms; and (4) additionally adjusted for cohabitation with adult children of both partners in all analyses. A two-sided p < 0.05 was considered statistically signiﬁcant. 2.2. Frailty Assessment The primary outcome of this study was the frailty status. Frailty was deﬁned by the Fried frailty phenotype. The Fried frailty phenotype comprises ﬁve criteria, namely weight loss, weakness, exhaustion, slowness, and low physical activity [4]. Those satisfying ≥3 components were considered frail. Weight loss was deﬁned as an unintentional weight loss of ≥10 pounds. Weakness was deﬁned as a baseline grip strength in the lower 20%, adjusted for sex and body mass index (BMI) in the KFACS population distribution. Exhaustion was assessed by the responses to two statements from the Center for Epidemiologic Studies Depression Scale: (1)” I felt that everything I did was an eﬀort”, and (2) “I could not get going”. Slowness was deﬁned as the slowest 20% at baseline on the basis of the 4-m usual gait speed, adjusted for sex and height in the KFACS population distribution. Low physical activity was deﬁned as the lower 20% of sex-speciﬁc kcals per week, calculated on the basis of the International Physical Activity Questionnaire in a general Korean population-based survey of older adults [21]. 3 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 2.3. Covariates Age, BMI, education, house ownership, comorbidity, cognition, depressive symptoms, and cohabitation with adult children of both husband and wife were used as covariates in this study. Data on demographics, including age, BMI, education, house ownership, and cohabitation with adult children were self-reported. Comorbidity was deﬁned as having ≥2 comorbid conditions. Cognitive impairment was assessed using the Mini-Mental State Examination (MMSE). Depressive symptoms were assessed using the short version of the Geriatric Depression Scale. Nutritional status was determined using the Korean version of the short form of the Mini Nutritional Assessment (MNA-SF). 3. Results The wife’s frailty was signiﬁcantly associated with the frailty of her husband in all models (OR 3.13–4.76, p < 0.05). Similarly, the husband’s frailty was positively associated with the frailty of his wife in all models (OR 2.91–4.76, p < 0.05) except model 3 (OR 3.19, 95% CI 1.00–10.22, p = 0.05). After adjusting for possible covariates, a frail husband had 4.62 odds of having a frail wife, and a frail wife had 3.34 odds of having a frail husband (model 4). While adjusting for other covariates reduced the odds ratio, adjusting for depressive symptoms (model 3) and cohabitation with adult children (model 4) increased the odds ratio. Among the ﬁve components of the Fried frailty phenotype, there was an association between the husband and wife in regard to weight loss, exhaustion, and slowness. Weight loss in one spouse was signiﬁcantly associated with the increased risk of weight loss in the partner in all models (husband, OR 6.89–8.56, p < 0.01; wife, OR 4.91–6.95, p < 0.05). After adjusting for covariates, wives with weight loss had 8.34 odds of having a husband with weight loss (p < 0.01). A husband with weight loss had 4.91 odds of having a wife with weight loss (p < 0.01). Exhaustion in one spouse was also associated with greater risk of exhaustion in the partner in all models (husband, OR 2.00–2.23, p < 0.05; wife, 4 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 OR 1.94–2.23, p < 0.05). Slowness in one spouse was associated with higher odds of slowness in the partner in all models (husband, OR 2.50–2.82, p < 0.05; wife, OR 2.52–2.82, p < 0.05) except model 2 (husband, OR 1.87, 95% CI 0.95–2.51; wife, OR 1.88, 95% CI 0.95–3.72). There was no signiﬁcant association between spouses for low activity and muscle weakness. Table 1. Demographic and health characteristics of study participants. 3. Results Variables Husband (n = 315) Wife (n = 315) p-Value Demographics Age, mean (SD), year 77.6 (3.4) 74.7 (3.6) <0.001 BMI, mean (SD), kg/m2 23.7 (3.0) 24.7 (2.8) <0.001 Education, n (%) Less than 7 years 72 (22.9) 168 (53.3) <0.001 Residence, n * (%) Urban 80 (25.6) 80 (25.6) 1.000 Suburban 148 (47.4) 148 (47.4) Rural 84 (26.9) 84 (26.9) House ownership, n (%) 30 (9.5) 98 (31.1) <0.001 Currently employed, n * (%) 81 (25.8) 68 (21.7) 0.223 Living with adult children, n * (%) 64 (20.6) 64 (20.6) Health Characteristics Cognitive Impairment MMSE score <24 46 (14.6) 76 (24.1) 0.002 Health Behavior Current smoker, n (%) 33 (10.5) 0 (0.0) <0.001 Heavy drinking, n (%) 106 (33.7) 10 (3.0) <0.001 Comorbid Conditions Number of comorbid conditions, mean (SD) 1.46 (1.24) 1.99 (1.36) <0.001 Comorbidity, n (%) 141 (42.7) 202 (61.2) <0.001 Hypertension, n (%) 174 (55.2) 174 (55.2) 1.000 Diabetes mellitus, n (%) 77 (24.4) 69 (21.9) 0.450 Mental Health Depressive symptom (GDS score ≥6), yes, n (%) 49 (15.6) 74 (23.5) 0.012 Physical Activity Moderate/vigorous activity, none, n (%) 107 (34.0) 116 (36.8) 0.453 Social Activity Religious meeting, none, n * (%) 155 (49.4) 93 (29.6) <0.001 Social meeting, none, n (%) 72 (22.9) 108 (34.3) 0.001 Possible Malnutrition MNA score ≤11, n * (%) 30 (9.6) 19 (6.1) 0.102 Self-perceived health, poor, n (%) 81 (25.7) 119 (37.8) 0.001 Notes: Comorbidity: ≥2 comorbid conditions; BMI = body mass index; GDS = Geriatric Depression Scale; MMSE Mi i M l S E i i MNA Mi i N i i l A Chi d f i l Table 1. Demographic and health characteristics of study participants. Notes: Comorbidity: ≥2 comorbid conditions; BMI = body mass index; GDS = Geriatric Depression Scale; MMSE = Mini-Mental State Examination; MNA = Mini Nutritional Assessment. Chi-squared test for categorical variables and independent t-test for continuous variables. n * indicates missing data. Table 2. Proportion of Fried frailty phenotype by sex. 3. Results Variable Husband (n = 315) Wife (n = 315) Both (n = 315 Pairs) p-Value Total, n (%) Total, n (%) Total, n (%) Fried Frailty Phenotype Frail 35 (11.1) 28 (8.9) 9 (2.9) 0.353 Five Components of Frailty Phenotype Weight loss 22 (7.0) 17 (5.4) 5 (1.6) 0.408 Exhaustion 77 (24.4) 133 (42.2) 44 (14.0) <0.001 Low activity 42 (13.3) 25 (7.9) 6 (1.9) 0.028 Weakness 74 (23.5) 56 (17.8) 17 (5.4) 0.076 Slowness 79 (25.0) 53 (16.8) 23 (7.3) 0.011 Notes: p-values were determined using the chi-squared test for categorical variables. 5 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 Table 3. Husband’s odds of frailty when the wife is frail in 315 married couples. Frailty of Wife Frailty of Husband (n = 315) Unadjusted Model 1 Model 2 Model 3 Model 4 OR (95% CI) Frailty of Wife Frail wife 4.76 (1.95–11.58) * 3.86 (1.50–9.92) 2.91 (1.04–8.16) * 3.19 (1.00–10.22) 3.34 (1.04–10.73) * Five Components of Frailty Phenotype among Wives Weight loss 6.89 (2.18–21.80) * 6.95 (2.10–23.02) * 6.87 (1.71–27.64) ** 5.03 (1.14–22.23) * 4.91 (1.10–21.97) * Exhaustion 2.23 (1.32–3.76) * 2.18 (1.28–3.70) * 1.94 (1.10–3.42) * 2.09 (1.09–3.98) * 2.14 (1.11–4.101) * Low activity 2.23 (0.83–5.95) 1.66 (0.60–4.62) 1.85 (0.59–5.78) 2.00 (0.60–6.63) 2.12 (0.63–7.08) Weakness 1.54 (0.81–2.93) 1.30 (0.67–2.52) 1.05 (0.51–2.15) 1.04 (0.50–2.18) 1.02 (0.49–2.13) Slowness 2.82 (1.52–5.23) *** 1.88 (0.95–3.72) 2.55 (1.17–5.59) * 2.58 (1.16–5.70) * 2.75 (1.23–6.14) * Notes: * p < 0.05, p < 0.01, * p < 0.005. BMI = body mass index; CI = conﬁdence interval; OR = odds ratio. Model 1 was adjusted for age; model 2 was adjusted for variables in model 1 plus BMI, education, house ownership, comorbidity, and cognition; model 3 was adjusted for variables in model 2 plus depressive symptoms; and model 4 was adjusted for cohabitation with adult children plus model 3. Table 3. Husband’s odds of frailty when the wife is frail in 315 married couples. CI = conﬁdence interval; OR = odds ratio. Model 1 was adjusted for age; model 2 was adjusted for variables in model 1 n; model 3 was adjusted for variables in model 2 plus depressive symptoms; and model 4 was adjusted for cohabitation Notes: * p < 0.05, p < 0.01, * p < 0.005. BMI = body mass index; CI = conﬁdence interval; OR = odds ratio. 4. Discussion In this study, we found that older adults with frail spouses had signiﬁcantly higher odds of being frail. Husbands and wives with a frail partner had 3.34 and 4.62 times higher odds of being frail, respectively, than if their partner was not frail. Thus, women were more aﬀected by spousal frailty status than men. Among the ﬁve components of the Fried frailty phenotype, there was a spousal association in regard to weight loss, slowness, and exhaustion. The odds ratio for a wife to experience weight loss if her husband had weight loss was twice the odds ratio for weight loss in a husband if his wife had weight loss. These results suggest the presence of sex diﬀerences in our ﬁndings. We found a spousal concordance for frailty among community-living older adults. This ﬁnding is consistent with the results from a previous study that reported that frailty in one spouse is related to greater subsequent frailty in the other [19]. It is also consistent with the previously reported concordance of physical, functional, and mental health and health behavior change in married couples [22–25]. Concordant health decline, especially in an elderly couple, is a signiﬁcant risk factor for diﬃculties in physical activities, functional disabilities, and depressive symptoms that have an increased caregiver burden [26]. Thus, spousal frailty can be used to detect, prevent, and manage a couple’s frailty among community-living older adults. Our study suggests an independent association between frailty status of marital partners in the older population. The logistic regression models were adjusted for recognized frailty-related factors, including age, education, cognition, comorbidity, and depression [27]. The signiﬁcant association for frailty within married couples remained after the adjustment for these factors. Additionally, the adjustment for the covariates not only decreased but also increased the size of the odds ratios. The change in the direction of the association may be due to the interplay of positive and negative confounding [28,29]. Further research is needed to examine the relationship between these factors, especially the cohabitation with adult children, as its impact on frailty has not been fully investigated. In previous studies, concordance of mood (depressive symptoms, neuroticism) [30,31], social activity [32], chronic conditions [23,33,34], and cognitive decline [35] between married couples have been observed. The multifactorial nature of spousal association and frailty may explain the varying strength of the spousal association. 3. Results Model 1 was adjusted for age; model 2 was adjusted for variables in model 1 plus BMI, education, house ownership, comorbidity, and cognition; model 3 was adjusted for variables in model 2 plus depressive symptoms; and model 4 was adjusted for cohabitation with adult children plus model 3. Table 4. Wife’s odds of frailty when the husband is frail in 315 married couples. Frailty of Husband Frailty of Wife (n = 315) Unadjusted Model 1 Model 2 Model 3 Model 4 OR (95% CI) Frailty of Husband Frail husband 4.76 (1.95–11.58) * 3.96 (1.52–10.33) * 3.13 (1.06–9.27) * 4.10 (1.16–14.56) * 4.62 (1.31–16.33) * Five Components of Frailty Phenotype among Husbands Weight loss 6.89 (2.18–21.80) * 7.07 (2.13–23.47) * 8.47 (2.02–35.52) * 8.56 (1.76–41.65) 8.34 (1.70–40.85) Exhaustion 2.23 (1.32–3.76) * 2.17 (1.27–3.68) *** 2.00 (1.13–3.57) * 2.11 (1.10–4.05) * 2.17 (1.12–4.22) * Low activity 2.23 (0.83–5.95) 1.75 (0.63–4.83) 2.02 (0.63–6.48) 2.46 (0.72–8.37) 2.49 (0.72–8.55) Weakness 1.54 (0.81–2.93) 1.29 (0.67–2.51) 1.04 (0.50–2.15) 1.03 (0.49–2.19) 1.03 (0.49–2.19) Slowness 2.82 (1.52–5.23) *** 1.87 (0.95–3.69) 2.50 (1.13–5.53) * 2.48 (1.12–5.50) * 2.77 (1.24–6.31) * Notes: * p < 0.05, p < 0.01, * p < 0.005. BMI = body mass index; CI = conﬁdence interval; OR = odds ratio. Model 1 was adjusted for age; model 2 was adjusted for variables in model 1 plus BMI, education, house ownership, comorbidity, and cognition; model 3 was adjusted for variables in model 2 plus depressive symptoms; and model 4 was adjusted for cohabitation with adult children plus model 3. Table 4. Wife’s odds of frailty when the husband is frail in 315 married couples. Frailty of Husband Notes: * p < 0.05, p < 0.01, * p < 0.005. BMI = body mass index; CI = conﬁdence interval; OR = odds ratio. Model 1 was adjusted for age; model 2 was adjusted for variables in model 1 plus BMI, education, house ownership, comorbidity, and cognition; model 3 was adjusted for variables in model 2 plus depressive symptoms; and model 4 was adjusted for cohabitation with adult children plus model 3. Int. J. Environ. Res. Public Health 2020, 17, 4574 6 of 10 4. Discussion We identiﬁed that three components of the Fried frailty phenotype, namely, weight loss, slowness, and exhaustion, mainly contribute to the spousal association for frailty. The strength of the spousal association was strongest for weight loss in both husbands and wives. There are various causes of weight loss among the elderly, and they often coexist. Common causes are classiﬁed as organic (neoplastic, nonneoplastic, and age-related physiological changes), psychological (depression, dementia, and anxiety disorders), nonmedical (socioeconomic conditions such as poverty), and unknown [36,37]. These conditions can be directly (such as environmental factors) and indirectly (such as depression [31]) shared by married couples. Furthermore, activities related to food preparation are mainly performed by women rather than men, especially among older adults in Korea [38]. The inﬂuence of gender role needs is discussed later. Exhaustion is associated with the psychosocial condition of older adults. Older adults can be aﬀected by the depressive symptoms and poor physical health of their spouses [39]. Those living with a depressed spouse are more likely to experience depressive mood or episodes [40,41], contributing to the increased association for exhaustion among married couples. A recent study suggested an association between slow gait speed and social networks among older adults [42]. Thus, the sharing of a couple’s social network may inﬂuence the spousal concordance in slowness. We found that women were more aﬀected by the frailty status and weight loss of their spouses than men. This ﬁnding may be due to gender role orientation and socialization [43]. These perspectives are deeply ingrained in the minds of the Korean elderly and are widely reﬂected in their daily lives. According to the 2017 national survey of living conditions and welfare needs of the Korean elderly, 93.7% of male adults aged >65 years received instrumental support from their spouse compared to only 54% of females aged >65 years [44]. Similarly, men were less likely to provide and more likely to 7 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 receive caregiving than women [44]. Poor spousal health, low education, and unemployment decreased the rate of receiving instrumental support and caregiving [44]. Thus, it can be deduced that elderly women do more housework and have a higher caregiving burden than men [45]. Furthermore, female caregivers are more likely to experience physical and psychosocial distress [46–48]. 4. Discussion Our ﬁndings are in agreement with women’s vulnerability to spousal inﬂuence and its negative health eﬀects as reported in previous studies on cognitive functioning [35], vision impairment [49], and depressive symptoms [50]. There are limitations to our study. First, due to the cross-sectional study design, it is diﬃcult to understand the causal relationship from our analysis. Further longitudinal studies are needed to evaluate the temporal changes in the frailty status in married couples. Second, as the KFACS was designed to target the elderly population regardless of marital status, the sample size of married couples was relatively small. However, our results suggest a similar prevalence of frailty and concordant frailty within couples as reported in previous studies [19]. In addition, the quality of the marital relationship was not accounted for in our study. 5. Conclusions In conclusion, having a frail spouse is a strong and independent risk factor for frailty. Among the ﬁve components of the Fried frailty phenotype, weight loss, exhaustion, and slowness were associated within married couples. Weight loss had the strongest association. Knowledge of the spousal association for frailty can help to prevent spousal frailty by managing risk factors and developing interventions designed for older married couples. Further prospective and longitudinal investigations will help corroborate our ﬁndings. Author Contributions: Conceptualization, C.W.W.; methodology, S.K., M.K., and C.W.W.; validation, C.W.W., M.K., and S.K.; formal analysis, S.K.; investigation, S.K.; data curation, S.K.; writing—original draft preparation, S.K.; writing—review and editing, S.K., M.K., and C.W.W.; supervision, M.K. and C.W.W.; project administration, S.K., M.K., and C.W.W.; and funding acquisition, C.W.W. All authors have read and agreed to the published version of the manuscript. Funding: This research was supported by a grant from the Korea Health Technology R&D Project through the Korean Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea (grant number: HI15C3153). Acknowledgments: We would like to thank the study participants and the staﬀof the Korean Frailty and Aging Cohort Study for their cooperation in this study. Acknowledgments: We would like to thank the study participants and the staﬀof the Korean Frailty and Aging Cohort Study for their cooperation in this study. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Appendix A Appendix A Table A1. Demographic and health characteristics of frail and nonfrail groups. Table A1. Demographic and health characteristics of frail and nonfrail groups. Variables Frail (n = 63) Nonfrail (n = 567) p-Value Demographics Sex Male 35 (5.6) 280 (44.4) 0.864 Female 28 (4.4) 287 (45.6) Age, mean (SD), year 78.2 (3.4) 75.9 (3.7) <0.001 BMI, mean (SD), kg/m2 24.0 (3.2) 24.2 (2.9) 0.551 Education, n (%) Less than 7 years 40 (63.5) 200 (35.3) <0.001 Residence, n * (%) Urban 4 (27.7) 4 (6.5) <0.001 Suburban 32 (51.6) 265 (47.0) Rural 26 (41.9) 143 (25.4) House ownership, n (%) 53 (84.1) 514 (90.7) <0.001 Currently employed, n * (%) 11 (17.5) 138 (24.4) 0.220 Int. J. Environ. Res. Public Health 2020, 17, 4574 8 of 10 Table A1. Cont. Table A1. Cont. Table A1. Cont. Variables Frail (n = 63) Nonfrail (n = 567) p-Value Living with adult children, n * (%) 10 (15.9) 121 (21.4) 0.026 Health Characteristics Cognitive Impairment MMSE score <24 27 (42.9) 95 (16.8) <0.001 Health Behavior Current smoker, n (%) 4 (6.3) 29 (5.1) 0.676 Heavy drinking, n (%) 12 (19.0) 104 (18.3) 0.891 Comorbid Conditions Number of comorbid conditions, mean (SD) 2.16 (1.36) 1.68 (1.32) 0.006 Comorbidity, n (%) 31 (49.2) 202 (35.6) 0.034 Hypertension, n (%) 41 (65.1) 307 (54.1) 0.098 Diabetes mellitus, n (%) 18 (28.6) 128 (22.6) 0.285 Mental Health Depressive symptom (GDS score ≥6), yes, n (%) 36 (57.1) 87 (15.3) <0.001 Physical Activity Moderate/vigorous activity, none, n (%) 38 (60.3) 185 (32.6) <0.001 Social Activity Religious meeting, none, n * (%) 31 (49.2) 217 (38.3) 0.094 Social meeting, none, n (%) 19 (30.2) 161 (28.4) 0.769 Possible Malnutrition MNA score ≤11, n * (%) 14 (22.6) 71.4 (6.2) <0.001 Self-perceived health, poor, n (%) 43 (68.3) 157 (27.7) <0.001 Notes: Comorbidity: ≥2 comorbid conditions; GDS = Geriatric Depression Scale; MNA = Mini Nutritional Assessment; MMSE = Mini-Mental State Examination. Chi-squared test for categorical variables and independent t-test for continuous variables. n * indicates missing data. Frail (n = 63) Nonfrail (n = 567) p-Value Notes: Comorbidity: ≥2 comorbid conditions; GDS = Geriatric Depression Scale; MNA = Mini Nutritional A t MMSE Mi i M t l St t E i ti Chi d t t f t i l i bl d i d d t References 1. World Health Organization: WHO. (2018b, February 5). Ageing and Health. Available online: https: //www.who.int/news-room/fact-sheets/detail/ageing-and-health (accessed on 20 April 2020). 1. World Health Organization: WHO. (2018b, February 5). Ageing and Health. Available online: https: //www.who.int/news-room/fact-sheets/detail/ageing-and-health (accessed on 20 April 2020). 2. Friedman, S.M.; Shah, K.; Hall, W.J. Failing to Focus on Healthy Aging: A Frailty of Our Discipline? J. Am. Geriatr. Soc. 2015, 63, 1459–1462. [CrossRef] 2. Friedman, S.M.; Shah, K.; Hall, W.J. Failing to Focus on Healthy Aging: A Frailty of Our Discipline? J. Am. Geriatr. Soc. 2015, 63, 1459–1462. [CrossRef] 3. Kojima, G.; Liljas, A.; Iliﬀe, S. Frailty Syndrome: Implications and Challenges for Health Care Policy. RMHP 2019, 12, 23–30. [CrossRef] 3. Kojima, G.; Liljas, A.; Iliﬀe, S. Frailty Syndrome: Implications and Challenges for Health Care Policy. RMHP 2019, 12, 23–30. [CrossRef] 4. Fried, L.P.; Tangen, C.M.; Walston, J.; Newman, A.B.; Hirsch, C.; Gottdiener, J.; Seeman, T.; Tracy, R.; Kop, W.J.; Burke, G.; et al. Frailty in Older Adults: Evidence for a Phenotype. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 2001, 56, M146–M157. [CrossRef] [PubMed] 4. Fried, L.P.; Tangen, C.M.; Walston, J.; Newman, A.B.; Hirsch, C.; Gottdiener, J.; Seeman, T.; Tracy, R.; Kop, W.J.; Burke, G.; et al. Frailty in Older Adults: Evidence for a Phenotype. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 2001, 56, M146–M157. [CrossRef] [PubMed] 5. Hoogendijk, E.O.; Aﬁlalo, J.; Ensrud, K.E.; Kowal, P.; Onder, G.; Fried, L.P. Frailty: Implications for Clinical Practice and Public Health. Lancet 2019, 394, 1365–1375. [CrossRef] 5. Hoogendijk, E.O.; Aﬁlalo, J.; Ensrud, K.E.; Kowal, P.; Onder, G.; Fried, L.P. Frailty: Implications for Clinical Practice and Public Health. Lancet 2019, 394, 1365–1375. [CrossRef] 6. Rockwood, K.; Mitnitski, A. Frailty in Relation to the Accumulation of Deﬁcits. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 2007, 62, 722–727. [CrossRef] [PubMed] 7. Collard, R.M.; Boter, H.; Schoevers, R.A.; Oude Voshaar, R.C. Prevalence of Frailty in Community-Dwelling Older Persons: A Systematic Review. J. Am. Geriatr. Soc. 2012, 60, 1487–1492. [CrossRef] 8. He, B.; Ma, Y.; Wang, C.; Jiang, M.; Geng, C.; Chang, X.; Ma, B.; Han, L. Prevalence and Risk Factors for Frailty Among Community-Dwelling Older People in China: A Systematic Review and Meta-Analysis. J. Nutr. Health Aging 2019, 23, 442–450. [CrossRef] 9. Buckinx, F.; Rolland, Y.; Reginster, J.-Y.; Ricour, C.; Petermans, J.; Bruyère, O. Burden of Frailty in the Elderly Population: Perspectives for a Public Health Challenge. Arch. Notes: Comorbidity: ≥2 comorbid conditions; GDS = Geriatric Depression Scale; MNA = Mini Nutritional Assessment; MMSE = Mini-Mental State Examination. Chi-squared test for categorical variables and independent t-test for continuous variables. n * indicates missing data. Assessment; MMSE Mini Mental State Examination. Chi squared test for categorical variables and in t-test for continuous variables. n * indicates missing data. References Jeon, S.Y.; Won, C.W.; Choi, H.R.; Kim, B.S.; Kim, S.Y.; Hur, J.H. Physical frailty predicts cognitive decline 21. Jeon, S.Y.; Won, C.W.; Choi, H.R.; Kim, B.S.; Kim, S.Y.; Hur, J.H. Physical frailty predicts cognitive decline in elderly people: Prospective ﬁndings from the living proﬁles of older people survey in Korea. Korean J. Fam. Pract. 2015, 5, 702–707. 22. Hoppmann, C.A.; Gerstorf, D.; Hibbert, A. Spousal Associations between Functional Limitation and Depressive Symptom Trajectories: Longitudinal Findings from the Study of Asset and Health Dynamics among the Oldest Old (AHEAD). Health Psychol. 2011, 30, 153–162. [CrossRef] [PubMed] 3. Kim, H.; Kang, D.; Choi, K.; Nam, C.; Thomas, G.; Suh, I. Spousal Concordance of Metabolic Syndrom 3141 Korean Couples: A Nationwide Survey. Ann. Epidemiol. 2006, 16, 292–298. [CrossRef] [PubMed] g p y 3141 Korean Couples: A Nationwide Survey. Ann. Epidemiol. 2006, 16, 292–298. [CrossRef] [PubMed] 24. Meyler, D.; Stimpson, J.P.; Peek, M.K. Health Concordance within Couples: A Systematic Review. Soc. Sci. Med. 2007, 64, 2297–2310. [CrossRef] 24. Meyler, D.; Stimpson, J.P.; Peek, M.K. Health Concordance within Couples: A Systematic Review. Soc. S 2007, 64, 2297–2310. [CrossRef] 25. Jackson, S.E.; Steptoe, A.; Wardle, J. The Inﬂuence of Partner’s Behavior on Health Behavior Change: The English Longitudinal Study of Ageing. JAMA Intern. Med. 2015, 175, 385. [CrossRef] [PubMed] 26. Lu, W.-H.; Chiou, S.-T.; Chen, L.-K.; Hsiao, F.-Y. Functional and Mental Health Outcomes of the Joint Eﬀects of Spousal Health: The Potential Threats of “Concordant Frailty”. J. Am. Med. Dir. Assoc. 2016, 17, 324–330. [CrossRef] 27. Clegg, A.; Young, J.; Iliﬀe, S.; Rikkert, M.O.; Rockwood, K. Frailty in Elderly People. Lancet 2013, 381, 752–762. [CrossRef] 28. Mehio-Sibai, A.; Feinleib, M.; Sibai, T.A.; Armenian, H.K. A Positive or a Negative Confounding Variable? A Simple Teaching Aid for Clinicians and Students. Ann. Epidemiol. 2005, 15, 421–423. [CrossRef] 29. Lachin, J.M. Biostatistical Methods: The Assessment of Relative Risks; John Wiley & Sons: Hoboken, NJ, US 30. Kreitman, N. The Patient’s Spouse. Br. J. Psychiatry 1964, 110, 159–173. [CrossRef] 31. Eagles, J.M.; Walker, L.G.; Blackwood, G.W.; Beattie, J.A.G.; Restall, D.B. The Mental Health of Elderly Couples II. Concordance for Psychiatric Morbidity in Spouses. Br. J. Psychiatry 1987, 150, 303–308. [CrossRef] 32. Hoppmann, C.A.; Gerstorf, D.; Luszcz, M. Spousal Social Activity Trajectories in the Australian Longitudinal Study of Ageing in the Context of Cognitive, Physical, and Aﬀective Resources. J. Gerontol. Ser. B 2008, 63, P41–P50. [CrossRef] 33. Hippisley-Cox, J. References Public Health 2015, 73, 19. [CrossRef] 10. Klein, B.E.K.; Klein, R.; Knudtson, M.D.; Lee, K.E. Frailty, Morbidity and Survival. Arch. Gerontol. Geriatr. 2005, 41, 141–149. [CrossRef] 11. Feng, Z.; Lugtenberg, M.; Franse, C.; Fang, X.; Hu, S.; Jin, C.; Raat, H. Risk Factors and Protective Factors Associated with Incident or Increase of Frailty among Community-Dwelling Older Adults: A Systematic Review of Longitudinal Studies. PLoS ONE 2017, 12, e0178383. [CrossRef] 9 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 12. Kojima, G.; Walters, K.; Iliﬀe, S.; Taniguchi, Y.; Tamiya, N. Marital Status and Risk of Physical Frailty: A Systematic Review and Meta-Analysis. J. Am. Med. Dir. Assoc. 2020, 21, 322–330. [CrossRef] [PubMed] 13. Aggar, C.; Ronaldson, S.; Cameron, I.D. Reactions to Caregiving during an Intervention Targeting Frailty in Community Living Older People. BMC Geriatr. 2012, 12, 66. [CrossRef] [PubMed] 14. Mui, A.C. Perceived Health and Functional Status among Spouse Caregivers of Frail Older Persons. J. Aging Health 1995, 7, 283–300. [CrossRef] [PubMed] 15. Wolﬀ, J.L.; Kasper, J.D. Caregivers of Frail Elders: Updating a National Proﬁle. Gerontology 2006, 46, 344–356. [CrossRef] [PubMed] 16. Comans, T.A.; Currin, M.L.; Brauer, S.G.; Haines, T.P. Factors Associated with Quality of Life and Caregiver Strain amongst Frail Older Adults Referred to a Community Rehabilitation Service: Implications for Service Delivery. Disabil. Rehabil. 2011, 33, 1215–1221. [CrossRef] 17. Pinquart, M.; Sörensen, S. Spouses, Adult Children, and Children-in-Law as Caregivers of Older Adults: A Meta-Analytic Comparison. Psychol. Aging 2011, 26, 1–14. [CrossRef] 18. Potier, F.; Degryse, J.-M.; Bihin, B.; Debacq-Chainiaux, F.; Charlet-Renard, C.; Martens, H.; de Saint-Hubert, M. Health and Frailty among Older Spousal Caregivers: An Observational Cohort Study in Belgium. BMC Geriatr. 2018, 18, 291. [CrossRef] 19. Monin, J.; Doyle, M.; Levy, B.; Schulz, R.; Fried, T.; Kershaw, T. Spousal Associations between Frailty and Depressive Symptoms: Longitudinal Findings from the Cardiovascular Health Study. J. Am. Geriatr. Soc. 2016, 64, 824–830. [CrossRef] 20. Won, C.W.; Lee, S.; Kim, J.; Chon, D.; Kim, S.; Kim, C.-O.; Kim, M.K.; Cho, B.; Choi, K.M.; Roh, E.; et al. Korean Frailty and Aging Cohort Study (KFACS): Cohort Proﬁle. BMJ Open 2020, 10, e035573. [CrossRef] 20. Won, C.W.; Lee, S.; Kim, J.; Chon, D.; Kim, S.; Kim, C.-O.; Kim, M.K.; Cho, B.; Choi, K.M.; Roh, E.; et al. Korean Frailty and Aging Cohort Study (KFACS): Cohort Proﬁle. BMJ Open 2020, 10, e035573. [CrossRef] 21. References Married Couples’ Risk of Same Disease: Cross Sectional Study. BMJ 2002, 325, 636. [CrossRef] [PubMed] 34. Stimpson, J.P.; Peek, M.K. Concordance of Chronic Conditions in Older Mexican American Couples. Prev. Chronic Dis. 2005, 2, 7. 10 of 10 Int. J. Environ. Res. Public Health 2020, 17, 4574 35. Gerstorf, D.; Hoppmann, C.A.; Anstey, K.J.; Luszcz, M.A. Dynamic Links of Cognitive Functioning among Married Couples: Longitudinal Evidence from the Australian Longitudinal Study of Ageing. Psychol. Aging 2009, 24, 296–309. [CrossRef] [PubMed] 6. Alibhai, S.M.H. An Approach to the Management of Unintentional Weight Loss in Elderly People. Can. M Assoc. J. 2005, 172, 773–780. [CrossRef] 37. McMinn, J.; Steel, C.; Bowman, A. Investigation and Management of Unintentional Weight Loss in Older Adults. BMJ 2011, 342, d1732. [CrossRef] 38. Kim, C. Having meals together state of couples of after retirement. Korea J. Popul. Stud. 2010, 33, 123–142. 38. Kim, C. Having meals together state of couples of after retirement. Korea J. Popul. Stud. 2010, 33, 123–142. 39. Siegel, M.J.; Bradley, E.H.; Gallo, W.T.; Kasl, S.V. The Eﬀect of Spousal Mental and Physical Health on Husbands’ and Wives’ Depressive Symptoms, Among Older Adults: Longitudinal Evidence from the Health and Retirement Survey. J. Aging Health 2004, 16, 398–425. [CrossRef] 40. Benazon, N.R.; Coyne, J.C. Living with a Depressed Spouse. J. Fam. Psychol. 2000, 14, 71–79. [CrossRef] 41. McLeod, J.D. Spouse Concordance for Depressive Disorders in a Community Sample. J. Aﬀect. Disord. 1993, 27, 43–52. [CrossRef] 42. Shaﬁe, S.; Shahwan, S.; Abdin, E.; Vaingankar, J.; Picco, L.; Sambasivam, R.; Zhang, Y.; Ng, L.L.; Chong, S.A.; Subramaniam, M. The Correlates of Slow Gait and Its Relation with Social Network among Older Adults in Singapore. Aging Ment. Health 2017, 21, 1171–1176. [CrossRef] 43. Miller, B. Gender Diﬀerences in Spouse Caregiver Strain: Socialization and Role Explanations. J. Marriage Fam. 1990, 52, 311. [CrossRef] 44. Chung, Kyunghee. Survey of Living Conditions and Welfare Needs of Korean Older Persons (Publication No. 11-1352000-000672-12). Korean Institute for Health and Social Aﬀair. 2017. Available online: https://data.kihasa.re.kr/micro/subject_view.jsp?WT.ac=favor_data&grp_seq=&project_seq=673 (accessed on 21 April 2020). 45. Oh, Y.; Han, E. Review of Studies on Spousal Caregivers of Frail Spouses in South Korea. Int. Soc. Work 2019, 62, 529–548. [CrossRef] 46. Barusch, A.S.; Spaid, W.M. Gender Diﬀerences in Caregiving: Why Do Wives Report Greater Burden? Gerontology 1989, 29, 667–676. [CrossRef] [PubMed] 7. Chiu, C.-J.; Lin, Y.-C. Spousal Health and Older Adults’ Biomarker Change over Six Years: Investigatio Gender Diﬀerences. © 2020 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 Arch. Gerontol. Geriatr. 2019, 83, 44–49. [CrossRef] 48. Pinquart, M.; Sorensen, S. Gender Diﬀerences in Caregiver Stressors, Social Resources, and Health: An Updated Meta-Analysis. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 2006, 61, P33–P45. [CrossRef] 49. Strawbridge, W.J.; Wallhagen, M.I.; Shema, S.J. Impact of Spouse Vision Impairment on Partner Health and Well-Being: A Longitudinal Analysis of Couples. J. Gerontol. Ser. B Psychol. Sci. Soc. Sci. 2007, 62, S315–S322. [CrossRef] 50. Hagedoorn, M.; Sanderman, R.; Ranchor, A.V.; Brilman, E.I.; Kempen, G.I.J.M.; Ormel, J. Chronic Disease in Elderly Couples. J. Psychosom. Res. 2001, 51, 693–696. [CrossRef] © 2020 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/W2194627710	https://comum.rcaap.pt/bitstream/10400.26/26625/1/1.pdf	English	null	Comprehensive measurements of<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mi>t</mml:mi></mml:math>-channel single top-quark production cross sections at<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mrow><mml:msqrt><mml:mrow><mml:mi>s</mml:mi></mml:mrow></mml:msqrt><mml:mo>=</mml:mo><mml:mn>7</mml:mn><mml:mtext> </mml:mtext><mml:mtext> </mml:mtext><mml:mi>TeV</mml:mi></mml:mrow></mml:math>with the ATLAS detector	Physical review. D. Particles, fields, gravitation, and cosmology/Physical review. D, Particles, fields, gravitation, and cosmology	2,014	cc-by	42,304	EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH (CERN) EUROPEAN ORGANISATION FOR NUCLEAR RESEARCH (CERN) Xiv:1406.7844v2 [hep-ex] 23 May 2016 arXiv:1406.7844v2 [hep-ex] 23 Ma Comprehensive measurements of t-channel single top-quark production cross sections at √s = 7 TeV with the ATLAS detector The ATLAS Collaboration ⃝ Reproduction of this article or parts of it is allowed as speciﬁed in the CC-BY-3.0 license. c⃝2016 CERN for the beneﬁt of the ATLAS Collaboration. c⃝2016 CERN for the beneﬁt of the ATLAS Collaboration. Reproduction of this article or parts of it is allowed as speciﬁed in the CC-BY-3.0 license. I. INTRODUCTION In general, measurements of single top-quark production provide insights into the properties of the Wtb vertex. The cross sections are proportional to the square of the coupling at the production vertex. In the Standard Model (SM), the cou- pling is given by the Cabibbo–Kobayashi–Maskawa (CKM) matrix element Vtb [1, 2] multiplied by the universal elec- troweak coupling constant. Angular distributions of top-quark decay products give access to the Lorentz structure of the Wtb vertex, which has a vector–axial vector structure in the SM. As illustrated in Fig. 1, the t-channel process features a b- quark in the initial state if described in LO Quantum Chro- modynamics (QCD), and therefore the cross section depends strongly on the b-quark parton distribution function (PDF), which is derived from the gluon PDF by means of the DGLAP evolution [3–5]. A measurement of the combined top-quark and top-antiquark cross section σ(tq+ ¯tq) = σ(tq)+σ(¯tq) is well suited to constrain Vtb or the b-quark PDF. In addition, the measurement of σ(tq + ¯tq) is sensitive to various models of new physics phenomena [6], such as extra heavy quarks, gauge bosons, or scalar bosons. In proton–proton (pp) collisions at the LHC, top quarks are produced at unprecedented rates, allowing studies that were intractable before. The production of single top quarks via weak charged-current interactions is among the top-quark phenomena becoming accessible to precise investigations. In leading-order (LO) perturbation theory, single top-quark pro- duction is described by three subprocesses that are distin- guished by the virtuality of the exchanged W boson. The dominant process is the t-channel exchange depicted in Fig. 1, which is the focus of the measurements presented in this arti- cle. A light quark from one of the colliding protons interacts with a b-quark from another proton by exchanging a virtual W boson (W ∗). Since the u-quark density of the proton is about twice as high as the d-quark density, the production cross sec- tion of single top quarks σ(tq), shown in Fig. 1(a), is expected to be about twice the cross section of top-antiquark production σ(¯tq), shown in Fig. 1(b). At LO, subleading single top-quark (b) (b) (a) (b) FIG. 1. Representative leading-order Feynman diagrams of (a) single top-quark production and (b) single top-antiquark production via the t-channel exchange of a virtual W ∗boson, including the decay of the top-quark and top-antiquark, respectively. PACS numbers: 14.65.Ha, 12.15.Hh, 13.85.Qk, 14.20.Dh Comprehensive measurements of t-channel single top-quark production cross sections at √s = 7 TeV with the ATLAS detector Th ATLAS ll b i The ATLAS collaboration (Dated: May 24, 2016) The ATLAS collaboration (Dated: May 24, 2016) This article presents measurements of the t-channel single top-quark (t) and top-antiquark (¯t) total production cross sections σ(tq) and σ(¯tq), their ratio Rt = σ(tq)/σ(¯tq), and a measurement of the inclusive production cross section σ(tq+ ¯tq) in proton–proton collisions at √s = 7 TeV at the LHC. Differential cross sections for the tq and ¯tq processes are measured as a function of the transverse momentum and the absolute value of the rapidity of t and ¯t, respectively. The analyzed data set was recorded with the ATLAS detector and corresponds to an integrated luminosity of 4.59 fb−1. Selected events contain one charged lepton, large missing transverse momentum, and two or three jets. The cross sections are measured by performing a binned maximum-likelihood ﬁt to the output distributions of neural networks. The resulting measurements are σ(tq) = 46 ± 1(stat.) ± 6(syst.)pb, σ(¯tq) = 23 ± 1(stat.) ± 3(syst.)pb, Rt = 2.04 ± 0.13(stat.) ± 0.12(syst.), and σ(tq + ¯tq) = 68 ± 2(stat.) ±8(syst.) pb, consistent with the Standard Model expectation. The uncertainty on the measured cross sections is dominated by systematic uncertainties, while the uncertainty on Rt is mainly statistical. Using the ratio of σ(tq+ ¯tq) to its theoretical prediction, and assuming that the top-quark-related CKM matrix elements obey the relation \|Vtb\| ≫\|Vts\|,\|Vtd\|, we determine \|Vtb\| = 1.02±0.07. PACS numbers: 14.65.Ha, 12.15.Hh, 13.85.Qk, 14.20.Dh Abstract This article presents measurements of the t-channel single top-quark (t) and top-antiquark (¯t) total production cross sections σ(tq) and σ(¯tq), their ratio Rt = σ(tq)/σ(¯tq), and a measurement of the inclusive production cross section σ(tq + ¯tq) in proton–proton collisions at √s = 7 TeV at the LHC. Differential cross sections for the tq and ¯tq processes are measured as a function of the transverse momentum and the absolute value of the rapidity of t and ¯t, respectively. The analyzed data set was recorded with the ATLAS detector and corresponds to an integrated luminosity of 4.59 fb−1. Selected events contain one charged lepton, large missing transverse momentum, and two or three jets. The cross sections are measured by performing a binned maximum-likelihood ﬁt to the output distributions of neural networks. The resulting measurements are σ(tq) = 46 ± 1(stat.) ± 6(syst.)pb, σ(¯tq) = 23 ± 1(stat.) ± 3(syst.)pb, Rt = 2.04 ± 0.13(stat.) ± 0.12(syst.), and σ(tq + ¯tq) = 68 ± 2(stat.) ± 8(syst.)pb, consistent with the Standard Model expectation. The uncertainty on the measured cross sections is dominated by systematic uncertainties, while the uncertainty on Rt is mainly statistical. Using the ratio of σ(tq + ¯tq) to its theoretical prediction, and assuming that the top-quark-related CKM matrix elements obey the relation \|Vtb\| ≫\|Vts\|,\|Vtd\|, we determine \|Vtb\| = 1.02±0.07. I. INTRODUCTION muon), large values of the magnitude of the missing transverse momentum Emiss T , and two or three hadronic jets with high transverse momentum. The acceptance for t-channel events is dominated by the 2-jet signature, where one jet is a b-quark jet, while the second jet is a light-quark jet. A signiﬁcant frac- tion of single top-quark events are also present in the 3-jet channel, whereas the t¯t background is dominant in the 4-jet channel. For this reason, the analysis is restricted to events with two or three jets. \| \| y for the ﬁrst time, differential cross sections are measured as a function of the transverse momentum of the top quark, pT(t), and the top antiquark, pT(¯t), and as a function of the absolute value of the rapidities \|y(t)\| and \|y(¯t)\|, respectively. Several other processes feature the same signature as single top-quark events, the main backgrounds being W+jets pro- duction and top-quark-antiquark (t¯t) pair production. Since a typical signature-based event selection yields only a rela- tively low signal purity, a dedicated analysis strategy is de- veloped to separate signal and background events. In both the 2-jet and 3-jet channels, several observables discriminat- ing between signal and background events are combined by a neural network (NN) to one discriminant (NN output). The cross-section measurements are based on a simultaneous ﬁt to these multivariate discriminants. In the 2-jet channel, a cut on the NN discriminant is applied to obtain a sample of events enriched in t-channel single top-quark events, facilitating the measurement of differential cross sections. In pp collisions at √s = 7 TeV, the total inclusive cross sections of top-quark and top-antiquark production in the t- channel are predicted to be σ(tq) = 41.9+1.8 −0.9 pb, σ(¯tq) = 22.7+0.9 −1.0 pb, and σ(tq + ¯tq) = 64.6+2.7 −2.0 pb, σ(tq) = 41.9+1.8 −0.9 pb, σ(¯tq) = 22.7+0.9 −1.0 pb, and σ(tq + ¯tq) = 64.6+2.7 −2.0 pb, with approximate next-to-next-to-leading-order (NNLO) pre- cision [9], assuming a top-quark mass of mt = 172.5 GeV and using the MSTW2008 NNLO [10] PDF set. The quoted uncertainty contains the scale uncertainty and the correlated PDF–αs uncertainty. The contributions due to the resumma- tion of soft-gluon bremsstrahlung included in the approximate NNLO result are relatively small and the cross-section pre- dictions are therefore very close to the plain next-to-leading- order (NLO) calculation [11]. I. INTRODUCTION (a) (a) Separate measurements of σ(tq) and σ(¯tq) extend the sen- sitivity to the PDFs of the u-quark and the d-quark, exploit- ing the different initial states of the two processes, shown in Fig. 1. At a center-of-mass energy of √s = 7 TeV, the typical momentum fraction x of the initial-state light quarks is in the range of 0.02 <∼x <∼0.5, with a median of 0.17 for u-quarks and a median of 0.13 for d-quarks. The additional measure- ment of the cross-section ratio Rt ≡σ(tq)/σ(¯tq) is sensitive to the ratio of the two PDFs in the x-range speciﬁed above and features smaller systematic uncertainties because of par- tial cancelations of common uncertainties. The measurements of σ(tq), σ(¯tq), and Rt provide complementary inputs in con- straining PDFs to data currently used in QCD ﬁts. Investigat- ing Rt also provides a way of searching for new-physics con- tributions in single top-quark (top-antiquark) production [7] and of elucidating the nature of physics beyond the SM if it were to be observed [8]. (b) (a) FIG. 1. Representative leading-order Feynman diagrams of (a) single top-quark production and (b) single top-antiquark production via the t-channel exchange of a virtual W ∗boson, including the decay of the top-quark and top-antiquark, respectively. FIG. 1. Representative leading-order Feynman diagrams of (a) single top-quark production and (b) single top-antiquark production via the t-channel exchange of a virtual W ∗boson, including the decay of the top-quark and top-antiquark, respectively. processes are the associated production of aW boson and a top quark (Wt) and the s-channel production of t ¯b, analogous to the Drell–Yan process. 2 In this article we present measurements of σ(tq + ¯tq), σ(tq), σ(¯tq), and the cross-section ratio Rt at a center-of-mass energy of √s = 7 TeV, using the full data set corresponding to an integrated luminosity of 4.59 fb−1. Final calibrations for the 7 TeV data set are used, resulting in reduced systematic uncertainties. The measurement of σ(tq+ ¯tq) is used to deter- mine the value of the CKM matrix element \|Vtb\|. Additionally, for the ﬁrst time, differential cross sections are measured as a function of the transverse momentum of the top quark, pT(t), and the top antiquark, pT(¯t), and as a function of the absolute value of the rapidities \|y(t)\| and \|y(¯t)\|, respectively. I. INTRODUCTION All predictions used in this arti- cle are based on the “ﬁve-ﬂavor scheme”, involving a b-quark in the initial state (see Fig. 1). An alternative approach is to consider the Born process qg →tqb, where the b-quark does not enter in the QCD evolution of the PDFs and the strong coupling constant, referred to as “four-ﬂavor scheme”. Re- cently, computations of differential cross sections have be- come available at approximate NNLO precision [12], com- plementing the predictions at NLO [11]. Measurements of these differential quantities will allow more stringent tests of the calculations. In addition, a thorough study of differential cross sections can give hints about the potential presence of ﬂavor-changing neutral currents or four-fermion operators in the single top-quark production process [13]. II. DATA SAMPLES AND SAMPLES OF SIMULATED EVENTS The analysis described in this article uses pp collision data collected at a center-of-mass energy of 7 TeV with the AT- LAS detector [20] at the LHC between March and November 2011. In this data-taking period, the average number of pp collisions per bunch crossing was nine. The selected events were recorded based on single-electron or single-muon trig- gers. Stringent detector and data quality requirements are ap- plied, resulting in a data set corresponding to an integrated luminosity of 4.59 ± 0.08 fb−1 [21]. A. Electrons Samples of simulated t-channel single top-quark events are produced with the NLO matrix-element generator POWHEG-BOX [24] interfaced to PYTHIA [25] (version 6.4.27) for showering and hadronization. In POWHEG- BOX the four-ﬂavor scheme calculation is used to simulate t-channel single top-quark production. The events are gen- erated using the ﬁxed four-ﬂavor NLO PDF set CT104f [26] and the renormalization and factorization scales are calculated event-by-event[27] with µR = µF = 4· q m2 b + p2 T,b, where mb and pT,b are the mass and pT of the b-quark from the initial gluon splitting. Electron candidates are selected from energy deposits (clus- ters) in the LAr electromagnetic calorimeter matched to tracks [37] and are required to have ET > 25 GeV and \|ηcl\| < 2.47, where ηcl denotes the pseudorapidity of the cluster. Clusters falling in the calorimeter barrel/endcap transition re- gion, corresponding to 1.37 < \|ηcl\| < 1.52, are ignored. The energy of an electron candidate is taken from the cluster, while its η and φ are taken from the track. The z-position of the track has to be compatible with the hard-scatter primary ver- tex. Electron candidates are further required to fulﬁl strin- gent criteria regarding calorimeter shower shape, track qual- ity, track–cluster matching, and fraction of high-threshold hits in the TRT to ensure high identiﬁcation quality. Samples of t¯t events, Wt events, and s-channel single top- quark events are generated with POWHEG-BOX interfaced to PYTHIA using the CT10 NLO PDF set [26]. All pro- cesses involving top quarks are produced assuming mt = 172.5 GeV, and the parameters of the PYTHIA generator controlling the modeling of the parton shower and the under- lying event are set to the values of the Perugia 2011 tune [28]. Vector-boson production in association with jets (W/Z+jets) is simulated using the multileg LO genera- tor ALPGEN [29] (version 2.13) using the CTEQ6L1 PDF set [30]. The partonic events are showered with HERWIG [31] (version 6.5.20), and the underlying event is simulated with the JIMMY [32] model (version 4.31) using values of the ATLAS Underlying Event Tune 2 [33]. W+jets and Z+jets events with up to ﬁve additional partons are generated. The MLM matching scheme [34] is used to remove overlap between partonic conﬁgurations generated Samples of t¯t events, Wt events, and s-channel single top- quark events are generated with POWHEG-BOX interfaced to PYTHIA using the CT10 NLO PDF set [26]. III. PHYSICS OBJECT DEFINITIONS In this section the deﬁnition of the physics objects is given, namely reconstructed electrons, muons, and jets, as well as Emiss T . The deﬁnition of these objects involves the recon- structed position of the hard interaction. Primary interaction vertices are computed from reconstructed tracks that are com- patible with coming from the luminous interaction region. The hard-scatter primary vertex is chosen as the vertex fea- turing the highest ∑p2 T, the sum running over all tracks with pT > 0.4 GeV associated with the vertex. The single muon trigger is based on muon candidates re- constructed in the muon spectrometer. At L1, a threshold of pT = 10 GeV is applied. At the HLT level, the requirement is tightened to pT > 18GeV. B. Trigger requirements ATLAS employs a three-level trigger system. The ﬁrst level (L1) is built from custom-made hardware, while the second and third levels are software based and collectively referred to as the High Level Trigger (HLT). The data sets used in this analysis are deﬁned by high-pT single electron or single muon triggers [23]. During the data-taking period slightly different trigger conditions were used to cope with the increasing num- ber of multiple pp collisions per bunch crossing (pile-up). After the event generation step, all samples are passed through the full simulation of the ATLAS detector [35] based on GEANT4 [36] and are then reconstructed using the same procedure as for collision data. The simulation includes the effect of multiple pp collisions per bunch crossing. The events are weighted such that the distribution of the number of colli- sions per bunch crossing is the same as in collision data. At L1, electron candidate events are required to have an electromagnetic energy deposit of ET > 14 GeV; in the sec- ond part of the data-taking period the requirement was ET > 16 GeV. At the HLT level, the full granularity of the calorime- ter and tracking information is available. The calorimeter cluster is matched to a track and the trigger electron object has to have ET > 20 GeV or ET > 22 GeV, exceeding the cor- responding L1 requirements by 6 GeV. A. The ATLAS detector Single top-quark production in the t-channel was ﬁrst es- tablished in p ¯p collisions at √s = 1.96 TeV at the Teva- tron [14]. Measurements of t-channel single top-quark and Wt production at the LHC at √s = 7 TeV were performed by the ATLAS collaboration [15, 16] and the CMS collabo- ration [17, 18]. The ATLAS measurements used only a frac- tion of the recorded data, corresponding to 1.04 fb−1 in the t-channel analysis. At √s = 8 TeV the CMS collaboration measured the t-channel cross sections and the cross-section ratio Rt [19]. The ATLAS detector [20] is built from a set of cylindrical subdetectors, which cover almost the full solid angle around the interaction point [22]. ATLAS is composed of an in- ner tracking detector (ID) close to the interaction point, sur- rounded by a superconducting solenoid providing a 2T ax- ial magnetic ﬁeld, electromagnetic and hadronic calorimeters, and a muon spectrometer (MS). The ID consists of a sili- con pixel detector, a silicon microstrip detector (SCT), and a straw-tube transition radiation tracker (TRT). The electro- magnetic calorimeter is a lead and liquid-argon (LAr) sam- pling calorimeter with high granularity. An iron/scintillator tile calorimeter provides hadronic energy measurements in the central pseudorapidity range. The endcap and forward regions are instrumented with LAr calorimeters for both the electro- magnetic and hadronic energy measurements. The MS con- sists of three large superconducting toroids with eight coils The measurements presented in this article are based on events in the lepton+jets channel, in which the lepton can be either an electron or a muon originating from a W-boson decay. The analysis has acceptance for signal events involv- ing W →τν decays if the τ lepton decays subsequently to either eνeντ or µνµντ. The experimental signature of can- didate events is thus given by one charged lepton (electron or 3 each, a system of trigger chambers, and precision tracking chambers. by the matrix element and by parton shower evolution. The double counting between the inclusive W + n-parton samples and samples with associated heavy-quark pair-production is removed utilizing an overlap removal based on a ∆R matching. The diboson processes WW, WZ and ZZ are generated using HERWIG and JIMMY. A. Electrons All pro- cesses involving top quarks are produced assuming mt = 172.5 GeV, and the parameters of the PYTHIA generator controlling the modeling of the parton shower and the under- lying event are set to the values of the Perugia 2011 tune [28]. Hadronic jets mimicking the signature of an electron, elec- trons from b-hadron or c-hadron decays, and photon conver- sions constitute the major backgrounds for high-pT electrons originating from the decay of a W boson. Since signal elec- trons from W-boson decay are typically isolated from jet ac- tivity, these backgrounds can be suppressed via isolation crite- ria that require minimal calorimeter activity (calorimeter iso- lation) and only few tracks (track isolation) in an (η,φ) re- gion around the electron. Electron candidates are isolated by imposing thresholds on the scalar sum of the transverse momenta of calorimeter energy deposits Σpcalo T within a sur- rounding cone of radius ∆R = 0.2, excluding the energy de- posit associated with the candidate, and on the scalar sum of the transverse momenta of tracks Σptrack T in a cone of radius ∆R = 0.3 around the candidate excluding the track associated Vector-boson production in association with jets (W/Z+jets) is simulated using the multileg LO genera- tor ALPGEN [29] (version 2.13) using the CTEQ6L1 PDF set [30]. The partonic events are showered with HERWIG [31] (version 6.5.20), and the underlying event is simulated with the JIMMY [32] model (version 4.31) using values of the ATLAS Underlying Event Tune 2 [33]. W+jets and Z+jets events with up to ﬁve additional partons are generated. The MLM matching scheme [34] is used to remove overlap between partonic conﬁgurations generated 4 tracks within the jet that originate from the hard-scatter pri- mary vertex to the ∑pT of all tracks matched to the jet. It is required that εjvf > 0.75 for those jets that have associated tracks. The εjvf criterion is omitted for jets without matched tracks. An overlap removal between jets and muons is applied, removing any muon with separation ∆R < 0.4 from a jet with pT > 25GeV and εjvf > 0.75. In the same way an overlap removal is applied between jets and electrons, removing any electron separated from a jet by 0.2 < ∆R < 0.4. with the electron candidate. The Σpcalo T variable is corrected for pile-up effects as a function of the number of reconstructed vertices. C. Jets and missing transverse momentum Jets are reconstructed using the anti-kt algorithm [39] with a radius parameter of 0.4, using topological clusters [40] iden- tiﬁed in the calorimeter as inputs to the jet clustering. The jet energy is corrected for the effect of multiple pp interactions, both in collision data and in simulated events. Further energy corrections apply factors depending on the jet energy and the jet η to achieve a calibration that matches the energy of sta- ble particle jets in simulated events [41]. Differences between data and Monte Carlo simulation are evaluated using in situ techniques and are corrected for in an additional step [42]. The in situ calibration exploits the pT balance in Z+jet, γ+jet, and dijet events. Z+jet and γ+jet data are used to set the jet energy scale (JES) in the central detector region, while pT bal- ancing in dijet events is used to achieve an η intercalibration of jets in the forward region with respect to central jets. A. Electrons The thresholds applied to Σpcalo T and Σptrack T vary as a function of the electron pT, the electron η, and the number of reconstructed primary vertices and are chosen such that the efﬁciency for electrons from W-boson or Z-boson decays to pass this isolation requirement is 90%. p j y Only jets having pT > 30 GeV and \|η\| < 4.5 are consid- ered. Jets in the endcap/forward-calorimetertransition region, corresponding to 2.75 < \|η\| < 3.5, must have pT > 35 GeV. The Emiss T is a measure of the momentum of the escaping neutrinos, but is also affected by energy losses due to detec- tor inefﬁciencies. The Emiss T is calculated based on the vector sum of energy deposits in the calorimeter projected onto the transverse plane and is corrected for the presence of electrons, muons, and jets [43]. Only jets having pT > 30 GeV and \|η\| < 4.5 are consid- ered. Jets in the endcap/forward-calorimetertransition region, corresponding to 2.75 < \|η\| < 3.5, must have pT > 35 GeV. B. Muons Muon candidates are reconstructed by combining track seg- ments found in the ID and in the MS [38]. The momentum as measured using the ID is required to agree with the momen- tum measured using the MS after correcting for the predicted muon energy loss in the calorimeter. Only candidates that have pT > 25 GeV and \|η\| < 2.5 are considered. Selected muons must additionally satisfy a series of requirements on the number of track hits present in the various tracking sub- detectors. Muon tracks are required to have at least two hits in the pixel detector, and six or more hits in the SCT. Tracks are rejected if they have more than two missing hits in the SCT and pixel detectors, or tracks with an excessive number of out- lier hits in the TRT. Isolated muon candidates are selected by requiring Σpcalo T < 4 GeV within a surrounding cone of radius ∆R = 0.2, and Σptrack T < 2.5 GeV within a surrounding cone of radius ∆R = 0.3. The efﬁciency of this combined isolation requirement varies between 95% and 97%, depending on the data-taking period. The Emiss T is a measure of the momentum of the escaping neutrinos, but is also affected by energy losses due to detec- tor inefﬁciencies. The Emiss T is calculated based on the vector sum of energy deposits in the calorimeter projected onto the transverse plane and is corrected for the presence of electrons, muons, and jets [43]. D. Identiﬁcation of b-quark jets The identiﬁcation of jets originating from the fragmenta- tion of b-quarks is one of the most important techniques for selecting top-quark events. Several properties can be used to distinguish b-quark jets from other jets: the long lifetime of b-hadrons, the large b-hadron mass, and the large branch- ing ratio to leptons. The relatively long lifetime of b-ﬂavored hadrons results in a signiﬁcant ﬂight path length, leading to re- constructable secondary vertices and tracks with large impact parameters relative to the primary vertex. The reconstruction, identiﬁcation and trigger efﬁciencies of electrons and muons are measured using tag-and-probe methods on samples enriched with Z →ℓℓ, J/ψ →ℓℓ, or W ± →ℓν (ℓ= e,µ) events [37, 38]. Jets containing b-hadrons are identiﬁed in the region \|η\| < 2.5 by reconstructing secondary and tertiary vertices from the tracks associated with each jet and combining lifetime-related information in a neural network [44]. Three different neu- ral networks are trained corresponding to an optimal separa- tion of b-quark jets, c-quark jets, and light-quark jets. The output of the networks is given in terms of probabilities pb, pc, and pl, which are then combined to form a ﬁnal discrim- inant. In order to achieve excellent rejection of c-quark jets the ratio pb/pc is calculated. The chosen working point cor- responds to a b-tagging efﬁciency of about 54% for b-quark jets in t¯t events. The misidentiﬁcation efﬁciency is 4.8% for c-quark jets and 0.48% for light-quark jets, as derived from simulated t¯t events. Jets passing the requirement on the iden- tiﬁcation discriminant are called b-tagged jets. Scale factors, determined from collision data, are applied to correct the b- tagging efﬁciency in simulated events to match the data. A. W/Z+jets background mT ℓEmiss T = q 2pT(ℓ)·Emiss T 1 −cos ∆φ ℓ,Emiss T , (1) The W+jets background is initially normalized to the theo- retical prediction and then subsequently determined simulta- neously both in the context of the multijet background estima- tion and as part of the extraction of the signal cross section. The estimated number of events of the much smaller Z+jets background is calculated using the theoretical prediction. to be larger than 30 GeV. Further reduction of the multijet background is achieved by placing an additional requirement on events with a charged lepton that is back-to-back with the leading jet in pT. This is realized by the following condition between the lepton pT and the ∆φ (j1,ℓ): The cross sections for inclusive W-boson production and Z-boson production are predicted with NNLO precision using the FEWZ program [45], resulting in a LO-to-NNLO scale factor of 1.2 and an uncertainty of 4%. The uncertainty in- cludes the uncertainty on the PDF and scale variations. The scale factor is applied to the prediction based on the LO ALP- GEN calculation for the W+b¯b, W+c¯c, and W+light-jets sam- ples. An uncertainty for associated jet production is estimated using variations of the factorization and renormalization scale and the ALPGEN matching parameter. These variations yield an uncertainty of 5% for the production of two additional light-quark jets and 15% for two additional heavy-quark jets. An additional relative uncertainty of 50% is assigned to the W+b¯b and W+c¯c production rates to take uncertainties on heavy-ﬂavor production into account. This uncertainty is esti- mated using a tag-counting method in control regions [15]. pT (ℓ) > 40 GeV· 1 −π −\|∆φ (j1,ℓ)\| π −1 (2) (2) where j1 denotes the leading jet. In the subsequent analysis, signal events are divided into different analysis channels according to the sign of the lep- ton charge and the number of jets. In the 2-jet channels, ex- actly one jet is required to be b-tagged. To further reduce the W+jets background in these channels, the absolute value of the difference in pseudorapidity \|∆η\| of the lepton and the b- tagged jet is required to be smaller than 2.4. In the 3-jet chan- nels, events with exactly one and exactly two b-tagged jets are considered and separated accordingly. A. W/Z+jets background In the 3-jet-2-tag cate- gory no distinction is made between events with positive and negative lepton charge, since this channel is dominated by t¯t background and can be used to further constrain the uncer- tainty on the b-tagging efﬁciency. Finally, the resulting chan- nels are referred to as: 2-jet-ℓ+, 2-jet-ℓ−, 3-jet-ℓ+-1-tag, 3-jet- ℓ−-1-tag, and 3-jet-2-tag. The ALPGEN prediction for the W+c process is scaled by a factor of 1.52 that is obtained from a study based on NLO calculations using MCFM [46]. Normalization uncertainties on the factorization and renormalization scale and PDF uncer- tainties are 24%. A control region is deﬁned to be orthogonal to the signal region in the same kinematic phase space to validate the mod- eling of the backgrounds by simulated events. Events in these control regions feature exactly one b-tagged jet, which was identiﬁed with a less stringent b-tagging algorithm than used to deﬁne the signal region. The signal region is excluded from the control region by applying a veto. The processes W+b¯b, W+c¯c, and W+light-jets, being asym- metric in lepton charge, are combined and are used as a single process in the binned maximum-likelihood ﬁt to determine the signal yield. B. Multijet background Multijet background events pass the signal selection if a jet is misidentiﬁed as an isolated lepton or if the event has a non- prompt lepton that appears to be isolated. Since it is neither possible to simulate a sufﬁcient number of those events nor possible to calculate the rate precisely, different techniques are developed to model multijet events and to estimate the pro- duction rate. These techniques employ both, collision data as well as simulated events. IV. EVENT SELECTION The event selection requires exactly one charged lepton, e or µ, exactly two or three jets, and Emiss T > 30 GeV. At least one of the jets must be b-tagged. A trigger matching require- ment is applied according to which the lepton must lie within ∆R = 0.15 cone around its trigger-level object. Candidate events are selected if they contain at least one good primary Jets with separation ∆R < 0.2 from selected electron can- didates are removed, as in these cases the jet and the electron are very likely to correspond to the same physics object. In order to reject jets from pile-up events, a quantity called the jet-vertex fraction εjvf is deﬁned as the ratio of ∑pT for all 5 vertex candidate with at least ﬁve associated tracks. Events containing jets with transverse momentum pT > 20 GeV fail- ing to satisfy quality criteria against misreconstruction [41] are rejected. a lepton, usually an electron, or a real high-pT lepton is pro- duced within a jet due to the semileptonic decay of a heavy- ﬂavor (b or c) hadron and satisﬁes the lepton isolation criteria. Other smaller backgrounds come from diboson (WW, WZ, and ZZ) and Z+jets production. Since the multijet background is difﬁcult to model pre- cisely, its contribution is reduced by requiring the transverse mass of the lepton-Emiss T system, V. BACKGROUND ESTIMATION (a) [GeV] miss T E GeV 1 miss T E d N d 0 50 100 150 [GeV] miss T E 0 50 100 150 Pred. Data-Pred. -0.2 0 0.2 (central) SR - e 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b, t ,Wt, tt W+jets Z+jets, diboson Multijet MC stat. (b) (b) (a) FIG. 2. Emiss T distributions in the signal region (SR) for the (a) 2-jet-e+ and (b) 2-jet-e−channels for central electrons. The distributions are normalized to the result of a binned maximum-likelihood ﬁt described in Sec. V B. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. tical to the signal selection. The number of multijet events Ntight fake passing the tight (signal) isolation requirements can be expressed as, converted photons. In addition, the jet must deposit 80–95% of its energy in the electromagnetic calorimeter. Events are selected using the same criteria as for the signal selection ex- cept for the selection of the electron. The event is accepted if exactly one such ‘jet lepton’ is found. The jet-lepton se- lection is applied to a PYTHIA dijet sample and the resulting set of events is used to model the multijet background in the electron channel. Ntight fake = εfake εreal −εfake ·(Nlooseεreal −Ntight), (3) (3) where εreal and εfake are the efﬁciencies for real and fake loose leptons being selected as tight leptons, Nloose is the number of selected events in the loose sample, and Ntight is the number of selected events in the signal sample. The fake efﬁciencies are determined from collision data in a sample of selected muon candidates with high impact parameter signiﬁcance which is deﬁned by the impact parameter divided by its uncertainty. The real efﬁciencies are also estimated from collision data us- ing a “tag-and-probe” method, which is based on the identiﬁ- cation of a tight lepton and a loose lepton in events originating from a leptonically decaying Z boson. To determine the normalization of the multijet background in the electron channel, a binned maximum-likelihood ﬁt to observed data in the Emiss T distribution is performed after ap- plying all selection criteria except for the Emiss T requirement. V. BACKGROUND ESTIMATION One of the largest backgrounds to single top-quark pro- cesses in the lepton+jets channel is W+jets production. If one of the jets contains b-hadrons or c-hadrons, these events have the same signature as signal events. Due to possible misiden- tiﬁcation of a light-quark jet as a b-quark jet, W+light-jets production also contributes to the background. An equally im- portant background comes from top-quark-antiquark (t¯t) pair production events, which are difﬁcult to separate from single top-quark events, since they contain top quarks as well. An- other background is due to multijet production via the strong interaction. In these events a hadronic jet is misidentiﬁed as In the electron channel, misidentiﬁed jets are the main source of multijet background events. This motivates the jet- lepton method in which an electron-like jet is selected with special requirements and redeﬁned as a lepton. This jet has to fulﬁl the same pT and η requirements as a signal electron, and contain at least four tracks to reduce the contribution from 6 [GeV] miss T E GeV 1 miss T E d N d 0 50 100 150 [GeV] miss T E 0 50 100 150 Pred. Data-Pred. -0.2 0 0.2 (central) SR + e 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data ,tq b ,Wt,t tt W+jets Z+jets, diboson Multijet MC stat. (a) [GeV] miss T E GeV 1 miss T E d N d 0 50 100 150 [GeV] miss T E 0 50 100 150 Pred. Data-Pred. -0.2 0 0.2 (central) SR - e 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b, t ,Wt, tt W+jets Z+jets, diboson Multijet MC stat. (b) FIG. 2. Emiss T distributions in the signal region (SR) for the (a) 2-jet-e+ and (b) 2-jet-e−channels for central electrons. The distributions are normalized to the result of a binned maximum-likelihood ﬁt described in Sec. V B. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. [GeV] miss T E GeV 1 miss T E d N d 0 50 100 150 [GeV] miss T E 0 50 100 150 Pred. Data-Pred. -0.2 0 0.2 (central) SR + e 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data ,tq b ,Wt,t tt W+jets Z+jets, diboson Multijet MC stat. V. BACKGROUND ESTIMATION In each channel two ﬁts are performed separately; one for electrons in the central (\|η\| < 1.5) region and one for the endcap (\|η\| > 1.5) region of the electromagnetic calorime- ter. The multijet template is ﬁtted together with templates de- rived from Monte Carlo simulation for all other background processes whose rate uncertainties are accounted for in the ﬁt- ting process in the form of additional constrained nuisance parameters. For the purpose of these ﬁts the contributions from W+light-jets and W+b¯b, W+c¯c, W+c, the contributions from t¯t and single top-quark production, and the contributions from Z+jets and diboson production, are each combined into one template. Distributions normalized to the ﬁt results in the 2-jet-e+ and 2-jet-e−signal regions for central electrons are shown in Fig. 2. An uncertainty of 50% is applied to the estimated yield of multijet background events based on comparisons of the rates obtained by using alternative methods, i.e. the matrix method in the electron channel and the jet-lepton method in the muon channel, and using an alternative variable, i.e. mT(ℓEmiss T ) in- stead of Emiss T for the binned maximum-likelihood ﬁt. VI. SIGNAL AND BACKGROUND DISCRIMINATION To separate t-channel single top-quark signal events from background events, several kinematic variables are combined to form powerful discriminants by employing neural net- works. A large number of potential input variables were stud- ied, including not only kinematic variables of the identiﬁed physics objects, but also variables obtained from the recon- struction of the W boson and the top quark. C. t¯t production and other backgrounds In the muon channel, the matrix method [47] is used to ob- tain both the normalization and shape of the multijet back- ground. The method estimates the number of multijet back- ground events in the signal region based on loose and tight lepton isolation deﬁnitions, the latter selection being a subset of the former. Hence, the loose selection is deﬁned to contain leptons of similar kinematics, but results in much higher event yields and is, except for the muon isolation requirement, iden- The t¯t cross section is calculated at NNLO in QCD in- cluding resummation of next-to-next-to-leading logarithmic (NNLL) soft gluon terms [48–52] with Top++2.0 [53]. The PDF and αs uncertainties are calculated using the PDF4LHC prescription [54] with the MSTW2008 NNLO [10, 55] at 68% conﬁdence level (CL), the CT10 NNLO [26, 56], and the 7 NNPDF2.3 [57] PDF sets, and are added in quadrature to the scale uncertainty, yielding a ﬁnal uncertainty of 6%. imaginary part vanishes and at the same time the transverse components of the neutrino momentum are kept as close as possible to the Emiss T . As a result of this algorithm, the four- momentum of the neutrino is reconstructed. Since Wt production is charge symmetric with respect to top-quark and top-antiquark production, the combined cross section of σ(Wt) = 15.7 ± 1.1 pb [58] is used in the analy- sis. The predicted cross sections for s-channel production are σ(t ¯b) = 3.1 ± 0.1 pb and σ(¯tb) = 1.4 ± 0.1 pb [59]. The pre- dictions of σ(Wt), σ(t ¯b), and σ(¯tb) are given at approximate NNLO precision, applying soft-gluon resummation. The- oretical uncertainties including PDF and scale uncertainties are 4.4% [59] for s-channel single top-quark production and 7.0% [58] for Wt production. The PDF uncertainties are eval- uated using the 40 associated eigenvector PDF sets of MSTW 2008 at 90% CL. The cross sections given above are used to compute the number of expected single top-quark events by normalizing the samples of simulated events. The top quark is reconstructed by adding the four-momenta of the reconstructed W boson and the b-tagged jet. Several angular variables, invariant masses and differences in pT are deﬁned using the reconstructed physics objects. B. Selection of discriminating variables The NeuroBayes [60] tool is used for preprocessing the in- put variables and for the training of the NNs. The ranking of the variables in terms of their discrimination power is auto- matically determined as part of the preprocessing step and is independent of the training procedure [15]. Only the highest- ranking variables are chosen for the training of the NNs. Sep- arate NNs are trained in the 2-jet channel and 3-jet channel. In the training, no separation is made according to lepton charge or lepton ﬂavour. Dedicated studies show that training in the channels separated by lepton charge does not lead to an im- provement in sensitivity. All top-quark background processes are shown combined in the ﬁgures and used as a single process in the analysis. The charge asymmetry in s-channel production is taken from the approximate NNLO prediction. Diboson events (WW, WZ and ZZ) are normalized to the NLO cross-section prediction calculated with MCFM [46]. The cross-section uncertainty for these processes is 5%. As a result of the optimization procedure in the 2-jet chan- nel, 13 kinematic variables are identiﬁed as inputs to the NN. In the 3-jet channel, 11 variables are used. It was found that reducing the number of variables further would result in a con- siderable loss of sensitivity. The input variables to the NNs are listed in Table II. The separation between signal and the two most important backgrounds, the top-quark background and the combined W+light-jets, W+c¯c, and W+b¯b background, is shown in Fig. 3 for the two most important discriminating variables in the 2-jet channel. C. Neural network training After choosing a set of variables based on the criteria out- lined above, the analysis proceeds with the training of the NNs using a three-layer feed-forward architecture. The number of hidden nodes was chosen to be 15 for both networks. Sam- ples of simulated events are used for the training process, the size of the signal samples in the 2-jet channel being about 37,000 events for top-quark and about 40,000 events for top- antiquark t-channel production. In the 3-jet channel the sizes of the training samples are 14,000 and 13,000 events, respec- tively. All background processes are used in the training, ex- cept for the multijet background whose modeling is associ- ated with large uncertainties. The total number of simulated background events used in the training is about 89,000 in the 2-jet channel and about 57,000 in the 3-jet channel. The ratio of signal events to background events in the training is cho- sen to be 1:1, while the different background processes are D. Event yields Table I provides the event yields after event selection. The yields are presented for the tagged channels, where exactly one b-tagged jet is required, separated according to the lepton charge and for the 3-jet-2-tag channel. Small contributions from the tq process in the ℓ−regions and the ¯tq process in the ℓ+ regions originate from lepton charge misidentiﬁcation. The modeling of the input variables is checked in a con- trol region (see Sec. IV for the deﬁnition) that is enriched in W+jets events. Figures 4 and 5 show the three most discrim- inating variables in the 2-jet-ℓ± and 3-jet-ℓ±-1-tag channels, respectively. Good modeling of the variables is observed. TABLE II. Input variables of the NNs in the 2-jet channels and in the 3-jet channels. The deﬁnitions of the variables use the term leading jet and 2nd leading jet, deﬁned as the jet with the highest or 2nd highest pT, respectively. In the 2-jet channels, exactly one jet is required to be b-tagged. The jet that is not b-tagged is denoted untagged jet. A. Top-quark reconstruction When reconstructing the W boson, the transverse momen- tum of the neutrino is given by the x- and y-components of the Emiss T , while the unmeasured z-component of the neutrino momentum pz(ν) is inferred by imposing a W-boson mass constraint on the lepton–neutrino system. Since the constraint leads to a quadratic equation for pz(ν), a two-fold ambigu- ity arises. In the case of two real solutions, the one with the lower \|pz(ν)\| is chosen. In case of complex solutions, which can occur due to the low Emiss T resolution, a kinematic ﬁt is performed that rescales the neutrino px and py such that the 8 TABLE I. Predicted and observed events yields for the 2-jet and 3-jet channels considered in this measurement. The multijet background is estimated using data-driven techniques (see Sec. V B); an uncertainty of 50% is applied. All the other expectations are derived using theoretical cross sections and their uncertainties (see Sec. V A and Sec. V C). ( ) 2-jet channels 3-jet channels ℓ+ ℓ− ℓ+ ℓ− 2-tag tq 2550± 220 3.6±0.3 845± 74 1.2±0.1 309± 26 ¯tq 1.5±0.1 1390± 120 0.52± 0.05 435±38 162± 14 t¯t,Wt,t ¯b, ¯tb 5250± 530 5130± 510 8200± 820 8180± 820 5850± 580 W ++b¯b,c¯c,light jets 5700± 2500 16.3± 8.2 2400± 1200 11.5±5.7 200± 100 W −+b¯b,c¯c,light jets 9.2±4.6 3400± 1700 4.1±2.0 1470± 740 137± 68 W+c 1460± 350 1620± 390 388± 93 430±100 6.5± 1.6 Z+jets, diboson 370±220 310± 180 190± 120 180±110 22± 13 Multijet 750±340 740± 370 320± 160 440±220 21± 11 Total expectation 16100± 2600 12600± 2000 12400± 1500 11100± 1100 6710± 610 Data 16198 12837 12460 10819 6403 TABLE II. Input variables of the NNs in the 2-jet channels and in the 3-jet channels. The deﬁnitions of the variables use the term leading jet and 2nd leading jet, deﬁned as the jet with the highest or 2nd highest pT, respectively. In the 2-jet channels, exactly one jet is required to be b-tagged. The jet that is not b-tagged is denoted untagged jet. TABLE II. Input variables of the NNs in the 2-jet channels and in the 3-jet channels. The deﬁnitions of the variables use the term leading jet and 2nd leading jet, deﬁned as the jet with the highest or 2nd highest pT, respectively. In the 2-jet channels, exactly one jet is required to be b-tagged. A. Top-quark reconstruction The jet that is not b-tagged is denoted untagged jet. Variables used in the 2-jet channels and the 3-jet channels m(ℓνb) The invariant mass of the reconstructed top quark. mT(ℓEmiss T ) The transverse mass of the lepton–Emiss T system, as deﬁned in Eq. (1). η(ℓν) The pseudorapidity of the system of the lepton and the reconstructed neutrino. m(ℓb) The invariant mass of the charged lepton and the b-tagged jet. HT The scalar sum of the transverse momenta of the jets, the charged lepton, and the Emiss T . Variables used in the 2-jet channels and the 3-jet channels m(ℓνb) The invariant mass of the reconstructed top quark. mT(ℓEmiss T ) The transverse mass of the lepton–Emiss T system, as deﬁned in Eq. (1). η(ℓν) The pseudorapidity of the system of the lepton and the reconstructed neutrino. m(ℓb) The invariant mass of the charged lepton and the b-tagged jet. HT The scalar sum of the transverse momenta of the jets, the charged lepton, and the Emiss T . Variables used in the 2-jet channels only m(jb) The invariant mass of the untagged jet and the b-tagged jet. \|η(j)\| The absolute value of the pseudorapidity of the untagged jet. ∆R(ℓ, j) ∆R between the charged lepton and the untagged jet. ∆R(ℓνb, j) ∆R between the reconstructed top quark and the untagged jet. \|η (b)\| The absolute value of the pseudorapidity of the b-tagged jet. \|∆pT (ℓ, j)\| The absolute value of the difference between the transverse momentum of the charged lepton and the untagged jet. \|∆pT (ℓνb, j)\| The absolute value of the difference between the transverse momentum of the reconstructed top quark and the untagged jet. Emiss T The missing transverse momentum. Variables used in the 3-jet channels only \|∆y(j1, j2)\| The absolute value of the rapidity difference of the leading and 2nd leading jets. m(j2 j3) The invariant mass of the 2nd leading jet and the 3rd leading jet. cosθ (ℓ, j)ℓνb r.f. The cosine of the angle θ between the charged lepton and the leading untagged jet in the rest frame of the reconstructed top quark. Ση (ji) The sum of the pseudorapidities of all jets in the event. m(j1 j2) The invariant mass of the two leading jets. pT (ℓνb) The transverse momentum of the reconstructed top quark. D. Extraction of the signal yield The cross sections σ(tq) and σ(¯tq) are extracted by per- forming a binned maximum-likelihood ﬁt to the NN discrimi- nant distributions in the 2-jet-ℓ+, 2-jet-ℓ−, 3-jet-ℓ+-1-tag, and 3-jet-ℓ−-1-tag channels and to the event yield in the 3-jet-2- tag channel, treating t-channel top-quark and t-channel top- antiquark production as independent processes. The signal rates, the rate of the combined top-quark background (t¯t, Wt, t ¯b, and ¯tb), the rate of the combined W+light-jets, W+c¯c, and W+b¯b background, and the b-tagging efﬁciency correc- tion factor (discussed in Sec. III D) are ﬁtted in all channels simultaneously. The event yields of the multijet background and the W+c background are not allowed to vary in the ﬁt, but instead are ﬁxed to the estimates given in Table I. The cross- section ratio is subsequently computed as Rt = σ(tq)/σ(¯tq). Regularization techniques are applied in the training pro- cess to dampen statistical ﬂuctuations in the training sample and to avoid overtraining. At the preprocessing stage men- tioned above (Sec. VI B), the input variables are transformed in several steps to deﬁne new input variables that are opti- mally prepared to be fed into an NN. First, the variables are transformed, such that they populate a ﬁnite interval and are distributed according to a uniform distribution. The inﬂuence of outliers is thereby strongly reduced. The distributions of the transformed variables are discretized using 100 bins, and the distributions for signal events are divided by the sum of signal and background events bin-by-bin, yielding the purity distributions in each variable. Next, these purity curves are ﬁtted with a regularized spline function, thereby yielding a continuous transformation from the original input variables to the purities. By means of the spline ﬁt statistical ﬂuctuations in the input variables are signiﬁcantly reduced. Applying the continuous purity functions to the input variables yields purity distributions that are further transformed, such that the distri- butions of the resulting variables are centered at zero and have an RMS of one. These variables are input to the NNs. In the training process, the network structure is pruned to arrive at a minimal topology, i.e. statistically insigniﬁcant network con- nections and nodes are removed. The maximum-likelihood function is given by the product of Poisson probability terms for the individual histogram bins (see Ref. [15]). A. Top-quark reconstruction 9 \| (j) η\| 0 1 2 3 4 5 Event fraction 0 0.05 0.1 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (a) [GeV] b) ν m(l 0 100 200 300 400 500 Event fraction 0 0.1 0.2 0.3 0.4 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (b) FIG. 3. Probability densities of the two most important discriminating variables in the 2-jet channels, shown in the 2-jet-ℓ+ channel in the signal region (SR). The distributions are normalized to unit area. The absolute value of the pseudorapidity of the untagged jet \|η(j)\| is shown in (a), and the invariant mass of the reconstructed top quark m(ℓνb) is shown in (b). \| (j) η\| 0 1 2 3 4 5 Event fraction 0 0.05 0.1 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (a) [GeV] b) ν m(l 0 100 200 300 400 500 Event fraction 0 0.1 0.2 0.3 0.4 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (b) (b) (a) FIG. 3. Probability densities of the two most important discriminating variables in the 2-jet channels, shown in the 2-jet-ℓ+ channel in the signal region (SR). The distributions are normalized to unit area. The absolute value of the pseudorapidity of the untagged jet \|η(j)\| is shown in (a), and the invariant mass of the reconstructed top quark m(ℓνb) is shown in (b). weighted relative to each other according to the number of expected events. D. Extraction of the signal yield \| (j) η\| \| (j) η d\| N d 0 10000 20000 30000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 10000 20000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 600 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) \| (j) η\| \| (j) η d\| N d 0 10000 20000 30000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 10000 20000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. D. Extraction of the signal yield Gaussian priors are added multiplicatively to the maximum-likelihood function to constrain the background rates subject to the ﬁt and the correction factor of the b-tagging efﬁciency to their predictions within the associated uncertain- ties. The sensitivity to the background rates is mostly given by the background-dominated region close to zero in the NN dis- criminant distributions, while the sensitivity to the b-tagging efﬁciency stems from the event yield in the 3-jet-2-tag channel with respect to the event yields in the 1-tag channels. In Fig. 7 the observed NN discriminant distributions are shown compared to the compound model of signal and back- ground normalized to the ﬁt results. Figures 8 and 9 show the three most discriminating variables normalized to the ﬁt re- sults in the 2-jet-ℓ± and 3-jet-ℓ±-1-tag channels, respectively. Differences between data and prediction are covered by the normalization uncertainty of the different processes after the ﬁt. In Fig. 6, the probability densities of the resulting NN discriminants are shown for the signal, the top-quark back- grounds, and the combined W+light-jets, W+c¯c, and W+b¯b background. The separation between signal and backgrounds is equally good for the positive and the negative charge chan- nels, which demonstrates that the choice of training the NNs with a charge-combined sample is appropriate. 10 \| (j) η\| \| (j) η d\| N d 0 10000 20000 30000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 10000 20000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 600 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 200 400 600 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 100 200 300 400 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) IG. 4. Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the control region (CR gures (a) and (b) display the absolute value of the pseudorapidity of the untagged jet \|η(j)\|. Figures (c) and (d) show the invariant mass e reconstructed top quark m(ℓνb), (e) and (f) the invariant mass of the untagged and the b-tagged jet m(jb). The last histogram bin includ verﬂows. The multijet and the W+jets event yields are determined by a ﬁt to the Emiss T distribution as described in Sec. V B. The uncertain and represents the normalization uncertainty due to the uncertainty on the jet energy scale and the Monte Carlo statistical uncertainty. T lative difference between the observed and expected number of events in each bin is shown in the lower panels. D. Extraction of the signal yield -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) (b) (a) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 600 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 200 400 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) GeV 1 b) ν m(l d N d (c) (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 200 400 600 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 100 200 300 400 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] m(jb) GeV 1 m(jb) d N d 0 100 200 300 400 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \|) 2,j 1 y(j ∆\| 0 5000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| 0 5000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 200 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) (a) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 200 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 200 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] m(jb) GeV 1 m(jb) d N d 0 200 400 600 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) (f) (e) FIG. 4. Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the control region (CR). Figures (a) and (b) display the absolute value of the pseudorapidity of the untagged jet \|η(j)\|. Figures (c) and (d) show the invariant mass of the reconstructed top quark m(ℓνb), (e) and (f) the invariant mass of the untagged and the b-tagged jet m(jb). The last histogram bin includes overﬂows. The multijet and the W+jets event yields are determined by a ﬁt to the Emiss T distribution as described in Sec. V B. The uncertainty band represents the normalization uncertainty due to the uncertainty on the jet energy scale and the Monte Carlo statistical uncertainty. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. 11 \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 200 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) IG. 5. D. Extraction of the signal yield Distributions of the three most important discriminating variables in the 3-jet-ℓ+ and 3-jet-ℓ−channels in the control region (CR gures (a) and (b) display the absolute value of the rapidity difference of the leading and 2nd leading jet \|∆y(j1, j2)\|, (c) and (d) the invaria ass of the 2nd leading jet and the 3rd jet m(j2 j3), and (e) and (f) show the invariant mass of the reconstructed top quark m(ℓνb). The l stogram bin includes overﬂows. The multijet and the W+jets event yields are determined by a ﬁt to the Emiss T distribution as described ec. V B. The uncertainty band represents the normalization uncertainty due to the uncertainty on the jet energy scale and the Monte Ca atistical uncertainty. The relative difference between the observed and expected number of events in each bin is shown in the lower panels \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) (c) (d) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 CR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. D. Extraction of the signal yield -0.2 0 0.2 CR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) (f) (e) FIG. 5. Distributions of the three most important discriminating variables in the 3-jet-ℓ+ and 3-jet-ℓ−channels in the control region (CR). Figures (a) and (b) display the absolute value of the rapidity difference of the leading and 2nd leading jet \|∆y(j1, j2)\|, (c) and (d) the invariant mass of the 2nd leading jet and the 3rd jet m(j2 j3), and (e) and (f) show the invariant mass of the reconstructed top quark m(ℓνb). The last histogram bin includes overﬂows. The multijet and the W+jets event yields are determined by a ﬁt to the Emiss T distribution as described in Sec. V B. The uncertainty band represents the normalization uncertainty due to the uncertainty on the jet energy scale and the Monte Carlo statistical uncertainty. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. 12 NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (a) NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 q t b t , Wt , tt ,light jets c ,c b b + - W SR -l 2-jet- Simulation ATLAS = 7 TeV s (b) NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 3-jet- Simulation ATLAS = 7 TeV s (c) NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 q t b t , Wt , tt ,light jets c ,c b b + - W SR -l 3-jet- Simulation ATLAS = 7 TeV s (d) FIG. 6. Probability densities of the NN discriminants in the 2-jet channels and 3-jet channels in the signal region (SR): (a) 2-jet-ℓ+ channel, (b) 2-jet-ℓ−channel, (c) 3-jet-ℓ+ channel, and (d) 3-jet-ℓ−channel. The distributions are normalized to unit area. D. Extraction of the signal yield NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 2-jet- Simulation ATLAS = 7 TeV s (a) Event fraction NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 q t b t , Wt , tt ,light jets c ,c b b + - W SR -l 2-jet- Simulation ATLAS = 7 TeV s (b) Event fraction (b) NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 tq b t , Wt , tt ,light jets c ,c b b + + W SR + l 3-jet- Simulation ATLAS = 7 TeV s (c) NN o 0 0.2 0.4 0.6 0.8 1 Event fraction 0 0.1 0.2 q t b t , Wt , tt ,light jets c ,c b b + - W SR -l 3-jet- Simulation ATLAS = 7 TeV s (d) Event fraction Event fraction (d) (c) FIG. 6. Probability densities of the NN discriminants in the 2-jet channels and 3-jet channels in the signal region (SR): (a) 2-jet-ℓ+ channel, (b) 2-jet-ℓ−channel, (c) 3-jet-ℓ+ channel, and (d) 3-jet-ℓ−channel. The distributions are normalized to unit area. E. High-purity region A high-purity region (HPR) is deﬁned to measure the dif- ferential cross sections in the 2-jet-ℓ+ and 2-jet-ℓ−channels, by requiring the NN discriminant to be larger than 0.8. In the 2-jet-ℓ+ HPR the signal contribution is twice as large as the background contribution. The signal and background contributions in the 2-jet-ℓ−HPR are of approximately the same size. The result of the ﬁt described above is used to normalize the background in the HPR. Figure 10 shows the three most discriminating variables in the 2-jet-ℓ+ and 2-jet- ℓ−high-purity channels, normalized to the ﬁt results. The data are well described by the predicted compound model. 13 13 NN o NN o d N d 0 10000 20000 30000 40000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) NN o NN o d N d 0 10000 20000 30000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) NN o NN o d N d 0 1000 2000 3000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) NN o NN o d N d 0 1000 2000 3000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) FIG. 7. E. High-purity region Neural network discriminant distributions normalized to the result of the binned maximum-likelihood ﬁt in (a) the 2-jet-ℓ+ chann (b) the 2-jet-ℓ−channel, (c) the 3-jet-ℓ+ channel, and (d) the 3-jet-ℓ−channel. The uncertainty band represents the normalization uncertai of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between the observed a expected number of events in each bin is shown in the lower panels. NN o NN o d N d 0 10000 20000 30000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band NN o NN o d N d 0 10000 20000 30000 40000 o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (b) (a) NN o NN o d N d 0 1000 2000 3000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) NN o NN o d N d 0 1000 2000 3000 NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) (d) (c) FIG. 7. Neural network discriminant distributions normalized to the result of the binned maximum-likelihood ﬁt in (a) the 2-jet-ℓ+ channel, (b) the 2-jet-ℓ−channel, (c) the 3-jet-ℓ+ channel, and (d) the 3-jet-ℓ−channel. The uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. E. High-purity region The relative difference between the observed and expected number of events in each bin is shown in the lower panels. \| (j) η\| \| (j) η d\| N d 0 5000 10000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 2000 4000 6000 8000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 50 100 150 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 50 100 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. E. High-purity region -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) (a) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) (c) (d) (c) (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 50 100 150 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 50 100 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] m(jb) GeV 1 m(jb) d N d 0 50 100 150 [GeV] m(jb) 0 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) GeV 1 m(jb) d N d 0 50 100 0 100 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS (f) (e) FIG. 8. Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the signal region normalized to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. Figures (a) and (b) display the absolute value of the pseudorapidity of the untagged jet \|η(j)\|. E. High-purity region -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) FIG. 8. Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the signal region normal to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. Figures (a) and (b) display the abso value of the pseudorapidity of the untagged jet \|η(j)\|. Figures (c) and (d) show the invariant mass of the reconstructed top quark m(ℓ (e) and (f) the invariant mass of the b-tagged and the untagged jet m(jb). The last histogram bin includes overﬂows. The uncertainty b represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. relative difference between the observed and expected number of events in each bin is shown in the lower panels. 14 \| (j) η\| \| (j) η d\| N d 0 5000 10000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 2000 4000 6000 8000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. E. High-purity region -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) \| (j) η\| \| (j) η d\| N d 0 2000 4000 6000 8000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \| (j) η\| \| (j) η d\| N d 0 5000 10000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| 0 1 2 3 4 D (a) \| (j) η\| 0 1 2 3 4 D (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 200 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 Pred. Data-Pred. E. High-purity region Figures (c) and (d) show the invariant mass of the reconstructed top quark m(ℓνb), (e) and (f) the invariant mass of the b-tagged and the untagged jet m(jb). The last histogram bin includes overﬂows. The uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. 15 \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 2000 4000 6000 8000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. E. High-purity region -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 2000 4000 6000 8000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) (b) (a) 0 [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. E. High-purity region -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) (c) (d) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) GeV 1 b) ν m(l d N d Pred Data-Pred. [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) (f) (e) FIG. 9. Distributions of the three most important discriminating variables in the 3-jet-ℓ+ and 3-jet-ℓ−channels in the signal region normalized to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. E. High-purity region -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 50 100 150 [GeV] b) ν m(l 100 200 300 400 500 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) FIG. 9. Distributions of the three most important discriminating variables in the 3-jet-ℓ+ and 3-jet-ℓ−channels in the signal region normali to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. Figures (a) and (b) display the abso value of the rapidity difference of the leading and 2nd leading jet \|∆y(j1, j2)\|, (c) and (d) the invariant mass of the 2nd leading jet and the jet m(j2 j3), and (e) and (f) show the invariant mass of the reconstructed top quark m(ℓνb). The last histogram bin includes overﬂows. uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. E. High-purity region -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 2000 4000 6000 8000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] ) 3j 2 m(j GeV 1 ) 3j 2 m(j d N d 0 50 100 150 [GeV] ) 3j 2 m(j 0 100 200 300 400 Pred. Data-Pred. -0.2 0 0.2 SR -l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 5000 10000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. -0.2 0 0.2 SR + l 3-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \|) 2,j 1 y(j ∆\| \|) 2,j 1 y(j ∆ d\| N d 0 2000 4000 6000 8000 \|) 2,j 1 y(j ∆\| 0 1 2 3 4 5 Pred. Data-Pred. E. High-purity region Figures (a) and (b) display the absolute value of the rapidity difference of the leading and 2nd leading jet \|∆y(j1, j2)\|, (c) and (d) the invariant mass of the 2nd leading jet and the 3rd jet m(j2 j3), and (e) and (f) show the invariant mass of the reconstructed top quark m(ℓνb). The last histogram bin includes overﬂows. The uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. 16 \| (j) η\| \| (j) η d\| N d 0 500 1000 1500 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| \| (j) η d\| N d 0 500 1000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \| (j) η\| (j) η d\| N d 0 500 1000 1500 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) \| (j) η\| (j) η d\| N d 0 500 1000 \| (j) η\| 0 1 2 3 4 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 20 40 [GeV] b) ν m(l 100 150 200 250 Pred. Data-Pred. E. High-purity region GeV 1 b) ν m(l d N d GeV 1 b) ν m(l d N d [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 20 40 [GeV] b) ν m(l 100 150 200 250 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 10 20 30 [GeV] b) ν m(l 100 150 200 250 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (c) (d) (c) (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 5 10 [GeV] m(jb) 0 200 400 600 800 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 2 4 [GeV] m(jb) 0 200 400 600 800 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) [GeV] m(jb) GeV 1 m(jb) d N d 0 5 10 [GeV] m(jb) 0 200 400 600 800 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) GeV 1 m(jb) d N d GeV 1 m(jb) d N d (f) (e) FIG. 10. E. High-purity region -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) [GeV] b) ν m(l GeV 1 b) ν m(l d N d 0 10 20 30 [GeV] b) ν m(l 100 150 200 250 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) [GeV] m(jb) GeV 1 m(jb) d N d 0 5 10 [GeV] m(jb) 0 200 400 600 800 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (e) [GeV] m(jb) GeV 1 m(jb) d N d 0 2 4 [GeV] m(jb) 0 200 400 600 800 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (f) FIG. 10. Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the high-purity re (HPR) normalized to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. Figures (a) and display the absolute value of the pseudorapidity of the untagged jet \|η(j)\|. Figures (c) and (d) show the invariant mass of the reconstru top quark m(ℓνb), (e) and (f) the invariant mass of the b-tagged and the untagged jet m(jb). The last histogram bin includes overﬂows. uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, adde quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. VII. SYSTEMATIC UNCERTAINTIES • Flavor response: In this category an uncertainty is con- sidered due to imperfect knowledge of the calorimeter response to light-quark jets and gluon jets. For both the physical object deﬁnitions and the background estimations, systematic uncertainties are assigned to account for detector calibration and resolution uncertainties, as well as the uncertainties of theoretical predictions. These variations affect both normalization and shape of distributions for sig- nal and backgrounds. The uncertainties can be split into the following categories: physics object modeling, Monte Carlo generators, PDFs, theoretical cross-section normalization, and luminosity. • b-JES: An additional JES uncertainty is evaluated for b-quark jets by varying the modeling of b-quark frag- mentation. The uncertainty due to the jet energy resolution is modeled by varying the pT of the jets according to the systematic un- certainties of the resolution measurement performed on data using the dijet-balance method [61]. The effect of uncertain- ties associated with the jet vertex fraction is also considered for each jet. A. Physics object modeling The tagging efﬁciencies of b-jets, c-jets, and light jets are derived from data [62–64] and parameterized as a function of pT and η of the jet. The corresponding efﬁciencies in sim- ulated events are corrected to be the same as those observed in data, and the uncertainties in the calibration method are propagated to the analysis. The difference in the b-tagging ef- ﬁciency between jets initiated by b-quark and b-antiquark is ∼1%, estimated from simulated tq and ¯tq events. To account for a possible uncertainty in the modeling of the detector re- sponse the full difference is taken as a systematic uncertainty. In Table III this uncertainty is called b/¯b acceptance. Systematic uncertainties on the reconstruction and en- ergy calibration of jets, electrons and muons are propagated through the entire analysis. The main source of object mod- eling uncertainty comes from the jet energy scale (JES). The JES uncertainty has been evaluated for the in situ jet calibra- tion [41, 42], which uses Z+jet, γ+jet, and dijet pT-balance measurements in data. The JES uncertainty is evaluated in several different categories: • Detector: The different pT-balance measurements have uncertainties due to the jet energy resolution, the elec- tron and photon energy scale and the photon purity. The uncertainties due to lepton reconstruction, identiﬁca- tion and trigger efﬁciencies are evaluated using tag-and-probe methods in Z →ℓℓevents. Uncertainties due to the energy scale and resolution are considered for electrons and muons. Additionally, the lepton charge misidentiﬁcation is taken into account and was evaluated to be about 0.1%. All lepton un- certainties are summarized in Table III in one item. • Physics modeling: The uncertainties in the in situ cal- ibration techniques due to the choice of Monte Carlo generator, radiation modeling, and the extrapolation of ∆φ between the jet and the Z boson. • Statistics: The uncertainty due to the limited size of the data sets of the in situ jet calibration measurements. Other minor uncertainties are assigned to the reconstruction of Emiss T and to account for the impact of pile-up collisions on the calculation of Emiss T . The uncertainties on Emiss T are summarized under Emiss T modeling in Table III. • Mixed detector and modeling: In this category the un- certainty due to the modeling of the underlying event and soft radiation as well as modeling of the jet frag- mentation are considered. E. High-purity region Distributions of the three most important discriminating variables in the 2-jet-ℓ+ and 2-jet-ℓ−channels in the high-purity region (HPR) normalized to the result of the binned maximum-likelihood ﬁt to the NN discriminant as described in Sec. VI D. Figures (a) and (b) display the absolute value of the pseudorapidity of the untagged jet \|η(j)\|. Figures (c) and (d) show the invariant mass of the reconstructed top quark m(ℓνb), (e) and (f) the invariant mass of the b-tagged and the untagged jet m(jb). The last histogram bin includes overﬂows. The uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. 17 B. Monte Carlo generators • η intercalibration modeling: The uncertainty in the dijet-pT-balance technique due to the modeling of ad- ditional parton radiation is estimated by comparing di- jet events simulated with PYTHIA and HERWIG. This JES category is the largest contribution from the jet en- ergy scale to the cross-section measurements. Systematic uncertainties arising from the modeling of the single top-quark signal, the t¯t background, and the W+jets background are taken into account. The uncertainty due to the choice of single top-quark t- channel generator and parton shower model is estimated by comparing events generated with POWHEG-BOX interfaced to PYTHIA and events generated with the NLO matrix- element generator MG5 aMC@NLO [65] and showered with HERWIG and JIMMY. Again the ﬁxed four-ﬂavor PDF set CT10f4 [26] is used, and the renormalization and factoriza- tion scales are set to µR = µF = 4 · q m2 b + p2 T,b, where mb = 4.75 GeV is the b-quark mass, and pT,b is the transverse mo- mentum of the b-quark. The uncertainty on the choice of µR and µF is estimated using events generated with POWHEG- BOX interfaced to PYTHIA. Factorization and renormaliza- tion scales are varied independently by factors of 0.5 and 2.0, while the scale of the parton shower is varied consistently with • Close-by jets: The jet calibration can be affected by the presence of close-by jets, located at radii ∆R < 1.0. • Pile-up: Uncertainties due to the modeling of the large pile-up effects in data are included as a function of jet pT and η. • Flavor composition: This uncertainty covers effects due to the difference in quark–gluon composition between the jets used in the calibration and the jets used in this analysis. Since the response to quark and gluon jets is different, the uncertainty on the quark–gluon composi- tion in a given data sample leads to an uncertainty in the jet calibration. 18 the renormalization scale. The uncertainty related to scale variations is then given by the envelope of all variations. Event Fraction 0 0.05 0.1 0.15 0.2 SR + l 2-jet- Simulation ATLAS = 7 TeV s tq intercalib. up η JES intercalib. down η JES NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.1 0 0.1 FIG. 11. C. Parton distribution function The systematic uncertainties related to the PDFs are taken into account for the acceptance of all single top-quark pro- cesses and t¯t production. The simulated events are reweighted according to each of the PDF uncertainty eigenvectors. The uncertainty is calculated following the recommendation of the respective PDF group. The ﬁnal PDF uncertainty is the enve- lope of the estimated uncertainties for the CT10 PDF set, the MSTW2008nlo [55] PDF set and the NNPDF2.3 [57] PDF set. For all PDFs the variable ﬂavor number scheme [67] is used. B. Monte Carlo generators The normalized shape variation of the NN discriminant for the JES variation due to the uncertainty on the η intercalibration in the 2-jet-ℓ+ channel, shown for the tq sample. The nominal shape is shown by the black points. Red denotes the JES shift-up and blue the NN response for JES shift-down. In the lower panel the relative difference between the number of expected events in the systematic variation and the nominal distribution is shown for each bin. The grey uncertainty band in the lower histogram represents the normal- ization uncertainty due to the Monte Carlo statistical uncertainty. Event Fraction 0 0.05 0.1 0.15 0.2 SR + l 2-jet- Simulation ATLAS = 7 TeV s tq intercalib. up η JES intercalib. down η JES NN o 0 0.2 0.4 0.6 0.8 1 Pred. Data-Pred. -0.1 0 0.1 The modeling uncertainty for the t¯t background is evalu- ated by comparing events simulated with the NLO genera- tor POWHEG-BOX interfaced to PYTHIA and the multi- leg LO generator ALPGEN interfaced to HERWIG. An ad- ditional uncertainty for the top-quark background processes comes from the amount of initial-state and ﬁnal-state radia- tion, estimated using dedicated ACERMC samples interfaced to PYTHIA where parameters controlling initial-state and ﬁnal-state radiation (ISR/FSR) emission are varied. The vari- ations of the parameters are constrained by a measurement of t¯t production with a veto on additional central jet activity [66]. A shape uncertainty is assigned to the W+jets background, based on variation of the choices of the matching scale and of the functional form of the factorization scale in ALPGEN. The impact of using simulation samples of limited size is also taken into account. FIG. 11. The normalized shape variation of the NN discriminant for the JES variation due to the uncertainty on the η intercalibration in the 2-jet-ℓ+ channel, shown for the tq sample. The nominal shape is shown by the black points. Red denotes the JES shift-up and blue the NN response for JES shift-down. In the lower panel the relative difference between the number of expected events in the systematic variation and the nominal distribution is shown for each bin. The grey uncertainty band in the lower histogram represents the normal- ization uncertainty due to the Monte Carlo statistical uncertainty. E. Luminosity The luminosity measurement is calibrated using dedicated beam-separation scans, referred to as van der Meer scans, where the absolute luminosity can be inferred from the mea- surement of the beam parameters [21]. The resulting uncer- tainty is 1.8%. B. Cross-section dependence on the top-quark mass B. Cross-section dependence on the top-quark mass sources of systematic uncertainty to the uncertainties on the measured values of σ(tq), σ(¯tq), Rt and σ(tq + ¯tq). The dominant systematic uncertainty on the cross sections is the JES η-intercalibration uncertainty since one of the prominent features of tq production is a jet in the forward region. The t-channel single top-quark cross sections are measured using a signal model with mt = 172.5 GeV. The dependence of the cross-section measurements on mt is mainly due to ac- ceptance effects and is expressed by the function: (4) σt = σt(172.5 GeV)+ p1 ·∆mt + p2 ·∆m2 t (4) A. Inclusive cross-section measurement A lower limit on \|Vtb\| is extracted in a Bayesian limit com- putation, assuming that the likelihood curve of \|Vtb\|2 has a Gaussian shape, centered at the measured value. A ﬂat prior in \|Vtb\|2 is applied, being one in the interval [0,1] and zero otherwise. The resulting lower limit is \|Vtb\| > 0.88 at the 95% CL. The inclusive t-channel cross section σ(tq+¯tq) is extracted by using only one scale factor β(tq + ¯tq) in the likelihood function, scaling the top-quark and the top-antiquark contri- butions simultaneously. The top-quark-to-antiquark ratio is taken from the approximate NNLO prediction [9] (see Sec. I). The systematic uncertainties on the measured value of inclu- sive cross-section are determined as described in Sec. VII. A detailed list of the uncertainties is given in Table III. C. Vtb extraction Since σ(tq + ¯tq) is proportional to \|Vtb\|2, \|Vtb\| can be ex- tracted from the measurement. The \|Vtb\| measurement is in- dependent of assumptions about the number of quark gener- ations and about the unitarity of the CKM matrix. The only assumptions required are that \|Vtb\| ≫\|Vtd\|,\|Vts\| and that the Wtb interaction is an SM-like left-handed weak coupling. The t¯t-background rate is unaffected by a variation of \|Vtb\|, since the decay to a quark of a potentially existing higher genera- tion are prohibited by kinematics, such that the branching ra- tio B(t →Wb) ∼1. On the other hand, the rates of single-top quark Wt and s-channel backgrounds also scale with \|Vtb\|2, but their contributions are small in the signal region. The re- sulting variation of the total top-quark background yield is less than its systematic uncertainty and thus considered negligible. assuming a top-quark mass of mt = 172.5 GeV. Figure 12 compares the measured values of σ(tq), σ(¯tq), and Rt to NLO predictions from MCFM [11] and Hathor [68] using differ- ent PDF sets. Uncertainties on the predicted values include the uncertainty on the renormalization and factorization scales and the combined PDF and αs uncertainty of the respective PDF set. All PDF predictions are in agreement with all measure- ments. For σ(¯tq), the predictions of all PDF sets agree well with each other and with the measured value. The predictions for σ(tq) and Rt with the ABM11 PDF set [69] show an offset compared to the other predictions. With increasing precision, the measurement of these observables could provide a way to further constrain the involved PDFs. y y g g The value of \|Vtb\|2 is extracted by dividing the measured value of σ(tq + ¯tq) by the prediction of the approximate NNLO calculation [9]. The experimental and theoretical un- certainties are added in quadrature. The result obtained is \|Vtb\| = 1.02 ± 0.01(stat.)± 0.06(syst.)± 0.02(theo.)+0.01 −0.00(mt) = 1.02 ± 0.07. \|Vtb\| = 1.02 ± 0.01(stat.)± 0.06(syst.)± 0.02(theo.)+0.01 −0.00(mt) = 1.02 ± 0.07. \|Vtb\| = 1.02 ± 0.01(stat.)± 0.06(syst.)± 0.02(theo.)+0.01 −0.00(mt) = 1.02 ± 0.07. VIII. TOTAL CROSS-SECTION MEASUREMENTS with ∆mt = mt −172.5 GeV. The parameters p1 and p2 are determined using dedicated signal samples with different mt and are given in Table IV for σ(tq), σ(¯tq) and σ(tq + ¯tq). The cross-section ratio Rt is largely independent of the top- quark mass. After performing the binned maximum-likelihood ﬁt and estimating the total uncertainty, the cross sections of top- quark and top-antiquark production in the t-channel and their cross-section ratio Rt are measured to be: σ(tq) = 46 ± 1(stat.)± 6(syst.)pb = 46 ± 6pb, σ(¯tq) = 23 ± 1(stat.)± 3(syst.)pb = 23 ± 4pb and Rt = 2.04 ± 0.13(stat.) ± 0.12(syst.) = 2.04 ± 0.18, IX. DIFFERENTIAL CROSS-SECTION MEASUREMENTS The binned maximum-likelihood ﬁt yields a cross section of Differential cross sections are measured as a function of the pT and \|y\| of t and ¯t in the 2-jet HPR channels, deﬁned in Sec. VI E. σt(tq + ¯tq) = 68 ± 2(stat.) ± 8(syst.) pb = 68 ± 8 pb, F. Uncertainties on the cross-section measurements In Sec. V the theoretical cross sections and their uncertain- ties are quoted for each background process. Since the t¯t, single top-quark Wt and s-channel processes are grouped to- gether in the statistical analysis, their uncertainties are added in proportion to their relative fractions, leading to a combined uncertainty of 6.7%. The uncertainty on the combined Z+jets and diboson background is 60% including a conservative es- timate of the uncertainty of the heavy-ﬂavor fraction of 50%, while the uncertainties of theW+jets backgrounds are 24% for W+c and 36% for the combined W+b¯b, c¯c and light jets in- cluding the same heavy-ﬂavor-fraction uncertainty on the b¯b and c¯c contributions. Additionally, an uncertainty on the rela- tive fraction of 2-jet to 3-jet events of 5% for events with light- ﬂavor jets and 7% for events with heavy-ﬂavor jets is applied for the W+jets estimation. This uncertainty was estimated by varying the following input parameters of the generation with ALPGEN by a factor of two: the hard scattering scale, the coupling of the hard interaction, and the minimum pT and ∆R separation of the partons. The systematic uncertainties on the individual top-quark and top-antiquark cross-section measurements and their ra- tio are determined using pseudo-experiments that account for variations of the signal acceptance, the background rates, and the shape of the NN discriminant due to all sources of un- certainty described above. As an example, Fig. 11 shows the shape variation of the NN discriminant for t-channel single top-quark signal events due to the variation of the JES because of the uncertainty on the η intercalibration. The correlations between the different channels and the physics processes are fully accounted for. The probability densities of all possible outcomes of the measurements of σ(tq), σ(¯tq) and Rt are ob- tained by performing the measurements on the pseudo-data. The values measured in data are used as central values when generating the pseudo-experiments. The root mean squares of the estimator distributions of the measured quantities are estimators of the measurement uncertainties. Table III summarizes the contributions of the various 19 B Cross section dependence on the top quark mass 19 A. Signal yield and reconstructed variables assuming mt = 172.5 GeV. Figure 12(d) compares the mea- sured value for σ(tq + ¯tq) to NLO predictions [11, 68] ob- tained with different PDF sets. All predictions are in agree- ment with the measurement. assuming mt = 172.5 GeV. Figure 12(d) compares the mea- sured value for σ(tq + ¯tq) to NLO predictions [11, 68] ob- tained with different PDF sets. All predictions are in agree- ment with the measurement. The signal and background composition in the 2-jet-ℓ+ and the 2-jet-ℓ−HPR channels can be found in Table V. Figure 13 20 TABLE III. Detailed list of the contribution of each source of uncertainty to the total uncertainty on the measured values of σ(tq), σ(¯tq), Rt, and σ(tq+ ¯tq). The evaluation of the systematic uncertainties has a statistical uncertainty of 0.3%. A. Signal yield and reconstructed variables Uncertainties contributing less than 1.0% k d ith “ 1” Source ∆σ(tq)/σ(tq) [%] ∆σ(¯tq)/σ(¯tq) [%] ∆Rt/Rt [%] ∆σ(tq+ ¯tq)/σ(tq+ ¯tq) [%] Data statistical ±3.1 ±5.4 ±6.2 ±2.7 Monte Carlo statistical ±1.9 ±3.2 ±3.6 ±1.9 Multijet normalization ±1.1 ±2.0 ±1.6 ±1.4 Other background normalization ±1.1 ±2.8 ±1.9 ±1.6 JES detector ±1.6 ±1.4 < 1 ±1.4 JES statistical < 1 < 1 < 1 < 1 JES physics modeling < 1 < 1 < 1 < 1 JES η intercalibration ±6.9 ±8.4 ±1.8 ±7.3 JES mixed detector and modeling < 1 < 1 < 1 < 1 JES close-by jets < 1 < 1 < 1 < 1 JES pile-up < 1 < 1 < 1 < 1 JES ﬂavor composition ±1.4 ±1.4 ±1.2 ±1.6 JES ﬂavor response < 1 < 1 ±1.0 < 1 b-JES < 1 < 1 < 1 < 1 Jet energy resolution ±2.1 ±1.6 ±1.0 ±1.9 Jet vertex fraction < 1 < 1 < 1 < 1 b-tagging efﬁciency ±3.8 ±4.1 < 1 ±3.9 c-tagging efﬁciency < 1 ±1.4 < 1 < 1 Mistag efﬁciency < 1 < 1 < 1 < 1 b/¯b acceptance ±1.0 < 1 < 1 −− Emiss T modeling ±2.3 ±3.4 ±1.6 ±2.6 Lepton uncertainties ±2.8 ±3.0 ±1.0 ±2.8 PDF ±3.2 ±5.8 ±2.5 ±3.2 W+jets shape variation < 1 < 1 < 1 < 1 tq generator + parton shower ±1.9 ±1.6 < 1 ±1.9 tq scale variations ±2.6 ±3.0 < 1 ±2.6 t¯t generator + parton shower < 1 ±2.1 ±1.6 < 1 t¯t ISR / FSR < 1 < 1 ±1.0 < 1 Luminosity ±1.8 ±1.8 ±0.5 ±1.8 Total systematic ±12.0 ±14.9 ±6.1 ±12.1 Total ±12.4 ±15.9 ±8.7 ±12.4 shows the measured distributions of the reconstructed top- quark pT and the reconstructed top-quark \|y\| normalized to the result of the binned maximum-likelihood ﬁt performed to measure σ(tq) and σ(¯tq). TABLE IV. Parameterization factors for the mt dependence (see Eq. (4)) of σ(tq), σ(¯tq) and σ(tq+ ¯tq). p1 [pb/GeV] p2 [pb/ GeV2] σ(tq+ ¯tq) −0.46 −0.06 σ(tq) −0.27 −0.04 σ(¯tq) −0.19 −0.02 TABLE IV. Parameterization factors for the mt dependence (see Eq. (4)) of σ(tq), σ(¯tq) and σ(tq+ ¯tq). The binning of the differential cross sections is chosen based on the experimental resolution of the pT and \|y\| distribu- tions as well as the data statistical uncertainty. A. Signal yield and reconstructed variables Typical values for the resolution of the top-quark pT are 10 GeV, increas- 21 [pb] (tq) σ 30 35 40 45 50 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (a) [pb] q) t( σ 12 14 16 18 20 22 24 26 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (b) t R 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (c) [pb] q) t (tq + σ 45 50 55 60 65 70 75 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (d) FIG. 12. Comparison between observed and predicted values of (a) σ(tq), (b) σ(¯tq), (c) Rt, and (d) σ(tq+ ¯tq). The predictions are calculated at NLO precision [11, 68] in the ﬁve-ﬂavor scheme and given for different NLO PDF sets [70–72] and the uncertainty includes the uncertainty on the renormalization and factorization scales, the combined internal PDF and αs uncertainty. The dotted black line indicates the central value of the measured value. The combined statistical and systematic uncertainty of the measurement is shown in green, while the statistical uncertainty is represented by the yellow error band. A. Signal yield and reconstructed variables The expectation for the signal and background yields corre- spond to the result of the binned maximum-likelihood ﬁt described in Sec. VI D. The uncertainty of the expectations is the normalization uncertainty of each processes after the ﬁt, as described in Sec. VII F. TABLE V. Event yields for the 2-jet-ℓ+ and 2-jet-ℓ−HPR chan- nels. The expectation for the signal and background yields corre- spond to the result of the binned maximum-likelihood ﬁt described in Sec. VI D. The uncertainty of the expectations is the normalization uncertainty of each processes after the ﬁt, as described in Sec. VII F. A. Signal yield and reconstructed variables [pb] (tq) σ 30 35 40 45 50 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (a) [pb] q) t( σ 12 14 16 18 20 22 24 26 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (b) ATLAS [pb] q) t (tq + σ 45 50 55 60 65 70 75 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS (d) t R 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) Measurement result NNPDF 2.3 MSTW2008 (68% CL) HERAPDF 1.5 GJR08 (VF) CT10 (+ D0 W asym.) CT10 ABM11 (5 flav.) syst. ⊕ stat. stat. =7 TeV s -1 dt = 4.59 fb L ∫ ATLAS ( ) (c) (d) FIG. 12. Comparison between observed and predicted values of (a) σ(tq), (b) σ(¯tq), (c) Rt, and (d) σ(tq+ ¯tq). The predictions are calculated at NLO precision [11, 68] in the ﬁve-ﬂavor scheme and given for different NLO PDF sets [70–72] and the uncertainty includes the uncertainty on the renormalization and factorization scales, the combined internal PDF and αs uncertainty. The dotted black line indicates the central value of the measured value. The combined statistical and systematic uncertainty of the measurement is shown in green, while the statistical uncertainty is represented by the yellow error band. ing to 25 GeV in the tail of the distribution. The resolution of the rapidity varies from 0.2 to 0.4 from central to forward rapidities. TABLE V. Event yields for the 2-jet-ℓ+ and 2-jet-ℓ−HPR chan- nels. B. Method 2-jet-ℓ+ HPR 2-jet-ℓ−HPR tq 1210± 150 1.3±0.2 ¯tq 0.29± 0.05 549± 87 t¯t,Wt,t ¯b,¯tb 161± 18 175± 19 W ++b¯b,c¯c,light jets 250± 48 0.35± 0.07 W −+b¯b,c¯c,light jets 0.7±0.2 166± 40 W+c 110± 26 125± 30 Z+jets, diboson 15± 10 11.4± 6.8 Multijet 59± 30 62± 31 Total expectation 1810± 160 1090± 110 Data 1813 1034 The measured distributions are distorted by detector effects and acceptance effects. The observed distributions are un- folded to the (parton level) four-momenta of the top quarks before the decay and after QCD radiation to correct for these distortions. In the following, each bin of the measured dis- tribution is referred to by the index i, while each bin of the parton-level distribution is referred to by the index j. The re- lation between the measured distribution and the differential cross section in each bin j of the parton-level distribution can be written as: dσ dXj = 1 ∆Xj · ∑ i M−1 ij ·(Ni −Bi) L ·εj ·B(t →ℓνb) (5) (5) where ∆Xj is the bin width of the parton-level distribution, Ni (Bi) are the data (expected background) yields in each bin of 22 [GeV] b) ν (l T p GeV 1 b) ν (l T p d N d -2 10 -1 10 1 10 2 10 3 10 [GeV] b) ν (l T p 0 100 200 300 400 500 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (a) [GeV] b) ν (l T p GeV 1 b) ν (l T p d N d -2 10 -1 10 1 10 2 10 [GeV] b) ν (l T p 0 100 200 300 400 500 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (b) \| b) ν y(l \| \| b) ν y(l d\| N d 0 1000 2000 \| b) ν y(l \| 0 1 2 3 Pred. Data-Pred. B. Method Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (b) (a) (b) \| b) ν y(l \| \| b) ν y(l d\| N d 0 500 1000 1500 \| b) ν y(l \| 0 1 2 3 Pred. Data-Pred. -0.2 0 0.2 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) \| \| \| \| \| b) ν y(l \| \| b) ν y(l d\| N d 0 1000 2000 \| b) ν y(l \| 0 1 2 3 Pred. Data-Pred. -0.2 0 0.2 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) (c) (d) FIG. 13. Measured distributions of (a) the top-quark pT, (b) top-antiquark pT, (c) top-quark \|y\|, and (d) top-antiquark \|y\| shown on recon- struction level in the HPR normalized to the result of the binned maximum-likelihood ﬁt. The uncertainty band represents the normalization uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between the observed and expected number of events in each bin is shown in the lower panels. bin j of each variable is deﬁned as the ratio of the parton-level yield before and after selection and is evaluated using simula- tion. The efﬁciencies are typically in the 0.5–2.2% range. the measured distribution, L is the integrated luminosity of the data sample, εj is the event selection efﬁciency and M−1 ij is the inverse of the migration matrix. The migration matrix accounts for the detector response and is deﬁned as the prob- ability to observe an event in bin i when it is generated in bin j. The migration matrix is built by relating the variables at the reconstruction and at the parton level using the signal simula- tion. B. Method -0.2 0 0.2 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band (c) \| b) ν y(l \| \| b) ν y(l d\| N d 0 500 1000 1500 \| b) ν y(l \| 0 1 2 3 Pred. Data-Pred. -0.2 0 0.2 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band (d) FIG. 13. Measured distributions of (a) the top-quark pT, (b) top-antiquark pT, (c) top-quark \|y\|, and (d) top-antiquark \|y\| shown on reco struction level in the HPR normalized to the result of the binned maximum-likelihood ﬁt. The uncertainty band represents the normalizati uncertainty of all processes after the ﬁt and the Monte Carlo statistical uncertainty, added in quadrature. The relative difference between t observed and expected number of events in each bin is shown in the lower panels. [GeV] b) ν (l T p GeV 1 b) ν (l T p d N d -2 10 -1 10 1 10 2 10 3 10 [GeV] b) ν (l T p 0 100 200 300 400 500 Pred. Data-Pred. -0.5 0 0.5 HPR + l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data tq b t , Wt , tt +c + W ,light jets c ,c b b + + W +jets, diboson Z Multijet Uncertainty band [GeV] b) ν (l T p GeV 1 b) ν (l T p d N d -2 10 -1 10 1 10 2 10 [GeV] b) ν (l T p 0 100 200 300 400 500 Pred. Data-Pred. -0.5 0 0.5 HPR -l 2-jet- ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ Data q t b t , Wt , tt +c - W ,light jets c ,c b b + - W +jets, diboson Z Multijet Uncertainty band [GeV] b) ν (l T p GeV 1 b) ν (l T p d N d -2 10 -1 10 1 10 2 10 3 10 [GeV] b) ν (l T p 0 100 200 300 400 500 Pred. B. Method Figure 14 shows the migration matrices for the pT and \|y\| distributions of the top quark and top antiquark. The in- verse of the matrix is determined by applying Bayes’ theorem iteratively [73] in order to perform the unfolding. The num- ber of iterations is chosen such that the absolute change in the unfolded distributions is on average smaller than 1% of the content in each bin. This procedure results in a total of ﬁve iterations for all distributions. The selection efﬁciency εj in The unfolding is applied to the reconstructed pT(ℓνb) and \|y(ℓνb)\| distributions after subtraction of the background con- tributions. When subtracting the background, all backgrounds are normalized according to Table V. Closure tests are performed in order to check the validity of the unfolding procedure. The shape of the parton-level dis- tributions in the Monte Carlo simulation are altered to verify that the simulation does not bias the results. It is checked that the altered parton-level distributions are recovered by unfold- ing the reconstructed distributions with the nominal migration matrix. 23 23 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 [GeV] b) ν (l T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 0.70 0.28 0.02 0.00 0.00 0.16 0.71 0.13 0.00 0.00 0.02 0.23 0.66 0.09 0.00 0.00 0.02 0.23 0.65 0.09 0.00 0.00 0.01 0.16 0.83 Simulation ATLAS = 7 TeV s (a) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 [GeV] b) ν (l T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 0.71 0.28 0.01 0.00 0.00 0.17 0.69 0.13 0.00 0.00 0.02 0.25 0.64 0.08 0.00 0.01 0.02 0.25 0.63 0.09 0.00 0.00 0.01 0.17 0.82 Simulation ATLAS = 7 TeV s (b) 0.1 0.2 0.3 0.4 0.5 0.6 \| b) ν y(l \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 0.69 0.30 0.01 0.00 0.23 0.67 0.10 0.00 0.06 0.26 0.62 0.05 0.01 0.05 0.28 0.66 Simulation ATLAS = 7 TeV s (c) 0.1 0.2 0.3 0.4 0.5 0.6 \| b) ν y(l \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 0.68 0.31 0.01 0.00 0.23 0.66 0.10 0.00 0.06 0.28 0.61 0.05 0.01 0.06 0.29 0.64 Simulation ATLAS = 7 TeV s (d) FIG. 14. B. Method Migration matrices relating the parton level shown on the y axis and reconstruction level shown on the x axis for the (a) top-quark pT, (b) top-antiquark pT, (c) top-quark \|y\|, and (d) top-antiquark \|y\| distribution. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 [GeV] b) ν (l T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 0.70 0.28 0.02 0.00 0.00 0.16 0.71 0.13 0.00 0.00 0.02 0.23 0.66 0.09 0.00 0.00 0.02 0.23 0.65 0.09 0.00 0.00 0.01 0.16 0.83 Simulation ATLAS = 7 TeV s 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 [GeV] b) ν (l T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 0.71 0.28 0.01 0.00 0.00 0.17 0.69 0.13 0.00 0.00 0.02 0.25 0.64 0.08 0.00 0.01 0.02 0.25 0.63 0.09 0.00 0.00 0.01 0.17 0.82 Simulation ATLAS = 7 TeV s (b) (b) (a) 0.1 0.2 0.3 0.4 0.5 0.6 \| b) ν y(l \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 0.69 0.30 0.01 0.00 0.23 0.67 0.10 0.00 0.06 0.26 0.62 0.05 0.01 0.05 0.28 0.66 Simulation ATLAS = 7 TeV s (c) 0.1 0.2 0.3 0.4 0.5 0.6 \| b) ν y(l \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 0.68 0.31 0.01 0.00 0.23 0.66 0.10 0.00 0.06 0.28 0.61 0.05 0.01 0.06 0.29 0.64 Simulation ATLAS = 7 TeV s (d) (d) (c) FIG. 14. Migration matrices relating the parton level shown on the y axis and reconstruction level shown on the x axis for the (a) top-quark pT, (b) top-antiquark pT, (c) top-quark \|y\|, and (d) top-antiquark \|y\| distribution. D. Results To reduce the impact of systematic uncertainties normal- ized differential cross sections 1/σ ·(dσ/dXj) are calculated by dividing the differential cross section by the total cross sec- tion evaluated by integrating over all bins. The absolute differential cross-section results are listed in Table VI and the normalized resutls in Table VII as a function of pT and \|y\| of the top quark. A graphical representation of the results is shown in Fig. 15 for the absolute cross sections and in Fig. 16 for the normalized case. They are compared to NLO predictions from MCFM [46] using the MSTW2008 PDF set for all variables. Uncertainties on the predicted val- ues include the uncertainty on the scale and the PDF. To com- pare the NLO prediction with the measurement, χ2 values are computed with HERAﬁtter [71, 72] taking into account the full correlation of the systematic and statistical uncertainties. The χ2 values for the differential cross sections are listed in Table VIII. The statistical uncertainty of the unfolded results is esti- mated using pseudo-experiments, propagating the uncertain- ties from the measured distribution and from the size of the Monte Carlo signal and background samples through the un- folding process. All sources of systematic uncertainty de- scribed in Sec. VII are included for the unfolded distributions. In the case of the background normalization, the uncertainties quoted in Table V are taken into account. The impact of the systematic uncertainties is evaluated by modifying the sub- tracted background before unfolding in the case of uncertain- ties on the backgrounds. To assign uncertainties on the signal modeling, systematic shifts are applied to the simulated signal sample. The shifted reconstructed distribution is unfolded and then compared to the nominal distribution at parton level. Systematic uncertainties dominate over the statistical un- certainty for the differential cross sections. X. CONCLUSIONS In summary, measurements of the single top-quark pro- duction cross sections, σ(tq), σ(¯tq), Rt, and σ(tq + ¯tq), with the ATLAS detector at the LHC are presented using an integrated luminosity of 4.59 fb−1 pp collision data at √s = 7 TeV. All measurements are based on neural net- work (NN) discriminants separating signal events from back- ground events. Binned maximum-likelihood ﬁts to the NN discriminants yield: σ(tq) = 46 ± 6pb, σ(¯tq) = 23 ± 4pb, and σ(tq + ¯tq) = 68 ± 8pb. The measured cross-section ra- tio is Rt = 2.04 ± 0.18. The corresponding coupling at the Wtb vertex is \|Vtb\| = 1.02±0.07, and the 95% CL lower limit on the CKM matrix element \|Vtb\| is 0.88. A high-purity region is deﬁned using the signal region of the NN discriminant for the differential cross-section measurements. Using an itera- tive Bayesian method, differential cross sections are extracted as a function of pT(t), pT(¯t), \|y(t)\|, and \|y(¯t)\|. Good agree- ment with the NLO QCD predictions is observed. D. Results ties originate from the background normalization, the tq gen- erator + parton shower uncertainty, the JES due to the un- certainty in the η intercalibration as well as the PDF uncer- tainties mainly in the top-antiquark distributions. A detailed list of the systematic contributions in each bin of each distri- butions is shown in Table IX for dσ/dpT(t), in Table X for dσ/dpT(¯t), in Table XI for dσ/d\|y(t)\|, and in Table XII for dσ/d\|y(¯t)\|. In the case of the normalized differential cross sections many systematic uncertainties cancel and thus the measurement is dominated by statistical uncertainties from the data distributions and the Monte Carlo sample size. The contribution of systematic uncertainties to the normalized dis- tribution is again dominated by the background normalization, tq generator + parton shower, and the JES η-intercalibration uncertainty. Details of the contribution of each systematic uncertainty in each bin of the normalized distributions are listed in Table XIII for 1/σ · dσ/dpT(t), in Table XIV for 1/σ · dσ/dpT(¯t), in Table XV for 1/σ · dσ/d\|y(t)\|, and in Table XVI for 1/σ · dσ/d\|y(¯t)\|. Bin-wise correlation matri- ces for the statistical uncertainty are given in Fig. 17 for the differential cross sections and in Fig. 18 for the normalized differential cross sections. The crucial computing support from all WLCG partners is acknowledged gratefully, in particular from CERN and the ATLAS Tier-1 facilities at TRIUMF (Canada), NDGF (Den- mark, Norway, Sweden), CC-IN2P3 (France), KIT/GridKA (Germany), INFN-CNAF (Italy), NL-T1 (Netherlands), PIC (Spain), ASGC (Taiwan), RAL (UK) and BNL (USA) and in the Tier-2 facilities worldwide. Overall, good agreement is observed between the NLO QCD predictions and the differential cross-section measure- ments. This is also supported by the χ2 values listed in Ta- ble VIII. The contents of Tables VI to XVI and the contents of Fig. 17 and Fig. 18 are provided in machine-readable format in the Supplemental Material [74]. D. Results Large uncertain- 24 We acknowledge the support of ANPCyT, Argentina; Yer- PhI, Armenia; ARC, Australia; BMWF and FWF, Austria; ANAS, Azerbaijan; SSTC, Belarus; CNPq and FAPESP, Brazil; NSERC, NRC and CFI, Canada; CERN; CONI- CYT, Chile; CAS, MOST and NSFC, China; COLCIEN- CIAS, Colombia; MSMT CR, MPO CR and VSC CR, Czech Republic; DNRF, DNSRC and Lundbeck Foundation, Den- mark; EPLANET, ERC and NSRF, European Union; IN2P3- CNRS, CEA-DSM/IRFU, France; GNSF, Georgia; BMBF, DFG, HGF, MPG and AvH Foundation, Germany; GSRT and NSRF, Greece; ISF, MINERVA, GIF, I-CORE and Benoziyo Center, Israel; INFN, Italy; MEXT and JSPS, Japan; CNRST, Morocco; FOM and NWO, Netherlands; BRF and RCN, Nor- way; MNiSW and NCN, Poland; GRICES and FCT, Portu- gal; MNE/IFA, Romania; MES of Russia and ROSATOM, Russian Federation; JINR; MSTD, Serbia; MSSR, Slo- vakia; ARRS and MIZˇS, Slovenia; DST/NRF, South Africa; MINECO, Spain; SRC and Wallenberg Foundation, Sweden; SER, SNSF and Cantons of Bern and Geneva, Switzerland; NSC, Taiwan; TAEK, Turkey; STFC, the Royal Society and Leverhulme Trust, United Kingdom; DOE and NSF, United States of America. ties originate from the background normalization, the tq gen- erator + parton shower uncertainty, the JES due to the un- certainty in the η intercalibration as well as the PDF uncer- tainties mainly in the top-antiquark distributions. A detailed list of the systematic contributions in each bin of each distri- butions is shown in Table IX for dσ/dpT(t), in Table X for dσ/dpT(¯t), in Table XI for dσ/d\|y(t)\|, and in Table XII for dσ/d\|y(¯t)\|. In the case of the normalized differential cross sections many systematic uncertainties cancel and thus the measurement is dominated by statistical uncertainties from the data distributions and the Monte Carlo sample size. The contribution of systematic uncertainties to the normalized dis- tribution is again dominated by the background normalization, tq generator + parton shower, and the JES η-intercalibration uncertainty. Details of the contribution of each systematic uncertainty in each bin of the normalized distributions are listed in Table XIII for 1/σ · dσ/dpT(t), in Table XIV for 1/σ · dσ/dpT(¯t), in Table XV for 1/σ · dσ/d\|y(t)\|, and in Table XVI for 1/σ · dσ/d\|y(¯t)\|. Bin-wise correlation matri- ces for the statistical uncertainty are given in Fig. 17 for the differential cross sections and in Fig. 18 for the normalized differential cross sections. ACKNOWLEDGEMENTS We thank CERN for the very successful operation of the LHC, as well as the support staff from our institutions without whom ATLAS could not be operated efﬁciently. 25 25 GeV pb )t( T p d σ d -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 (a) GeV pb )t( T p d σ d -3 10 -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 (b) [pb] )\| t( y d \| σ d 0 10 20 30 40 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (c) [pb] )\| t(\| y d σ d 0 10 20 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (d) G. 15. Differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The differential distributions are compa the QCD NLO calculation. The black vertical error bars on the data points denote the total combined uncertainty, the green error b note the statistical uncertainty, while the red band denotes the theory predictions calculated at NLO using MCFM [46]. Uncertainties on edicted values include the PDF and scale uncertainties. The horizontal error bars indicate the bin width. GeV pb )t( T p d σ d -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV pb )t( T p d σ d -3 10 -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. ACKNOWLEDGEMENTS NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV pb )t( T p d σ d -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV pb )t( T p d σ d -3 10 -2 10 -1 10 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 Data / NLO (a) [pb] )\| t( y d \| σ d 0 10 20 30 40 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (c) (b) [pb] )\| t(\| y d σ d 0 10 20 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (d) (d) (c) FIG. 15. Differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The differential distributions are compared to the QCD NLO calculation. The black vertical error bars on the data points denote the total combined uncertainty, the green error bars denote the statistical uncertainty, while the red band denotes the theory predictions calculated at NLO using MCFM [46]. Uncertainties on the predicted values include the PDF and scale uncertainties. The horizontal error bars indicate the bin width. 26 GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ACKNOWLEDGEMENTS ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 (a) GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 (b) )\| t( y d \| σ d ⋅ σ1 0 0.5 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (c) )\| t( y d \| σ d ⋅ σ1 0 0.5 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (d) G. 16. Normalized differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The normalized differen tributions are compared to the QCD NLO calculation. The black vertical error bars on the data points denote the total combined uncertai green error bars denote the statistical uncertainty, while the red band denotes the theory predictions calculated at NLO using MCFM [4 certainties on the predicted values include the PDF and scale uncertainties. The horizontal error bars indicate the bin width. GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ACKNOWLEDGEMENTS ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 GeV 1 )t( T p d σ d ⋅ σ1 -4 10 -3 10 -2 10 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ ) [GeV] t( T p 0 100 200 300 400 500 Data / NLO 0.5 1 1.5 Data / NLO (a) )\| t( y d \| σ d ⋅ σ1 0 0.5 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (c) (b) )\| t( y d \| σ d ⋅ σ1 0 0.5 1 Data sys unc. ⊕ stat stat unc. NLO (MSTW2008) scale unc. ⊕ PDF ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ )\| t( y\| 0 1 2 3 Data / NLO 0.5 1 1.5 (d) (d) (c) FIG. 16. Normalized differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The normalized differential distributions are compared to the QCD NLO calculation. The black vertical error bars on the data points denote the total combined uncertainty, the green error bars denote the statistical uncertainty, while the red band denotes the theory predictions calculated at NLO using MCFM [46]. Uncertainties on the predicted values include the PDF and scale uncertainties. The horizontal error bars indicate the bin width. ACKNOWLEDGEMENTS 27 27 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 [GeV] (t) T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 0.48 0.49 0.43 0.35 0.48 1.00 0.51 0.43 0.37 0.49 0.51 1.00 0.34 0.31 0.43 0.43 0.34 1.00 0.14 0.35 0.37 0.31 0.14 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (a) 0 0.2 0.4 0.6 0.8 1 [GeV] )t( T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 0.34 0.26 0.13 0.13 0.34 1.00 0.32 0.03 0.15 0.26 0.32 1.00 0.09 0.05 0.13 0.03 0.09 1.00 -0.06 0.13 0.15 0.05 -0.06 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (b) 0 0.2 0.4 0.6 0.8 1 \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.12 -0.14 -0.02 -0.12 1.00 -0.19 -0.09 -0.14 -0.19 1.00 -0.18 -0.02 -0.09 -0.18 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (c) -0.2 0 0.2 0.4 0.6 0.8 1 \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.05 -0.17 0.11 -0.05 1.00 -0.01 -0.06 -0.17 -0.01 1.00 -0.23 0.11 -0.06 -0.23 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (d) FIG. 17. Statistical correlation matrices for the differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The contents of this ﬁgure are provided in machine-readable format in the Supplemental Material [74]. ACKNOWLEDGEMENTS 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 [GeV] (t) T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 0.48 0.49 0.43 0.35 0.48 1.00 0.51 0.43 0.37 0.49 0.51 1.00 0.34 0.31 0.43 0.43 0.34 1.00 0.14 0.35 0.37 0.31 0.14 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ 0 0.2 0.4 0.6 0.8 1 [GeV] )t( T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 0.34 0.26 0.13 0.13 0.34 1.00 0.32 0.03 0.15 0.26 0.32 1.00 0.09 0.05 0.13 0.03 0.09 1.00 -0.06 0.13 0.15 0.05 -0.06 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ T (a) 0 0.2 0.4 0.6 0.8 1 \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.12 -0.14 -0.02 -0.12 1.00 -0.19 -0.09 -0.14 -0.19 1.00 -0.18 -0.02 -0.09 -0.18 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (c) (b) (b) -0.2 0 0.2 0.4 0.6 0.8 1 \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.05 -0.17 0.11 -0.05 1.00 -0.01 -0.06 -0.17 -0.01 1.00 -0.23 0.11 -0.06 -0.23 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (d) (d) (c) FIG. 17. Statistical correlation matrices for the differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The contents of this ﬁgure are provided in machine-readable format in the Supplemental Material [74]. ACKNOWLEDGEMENTS 28 28 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 [GeV] (t) T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 -0.32 -0.60 -0.49 -0.26 -0.32 1.00 0.21 -0.42 -0.04 -0.60 0.21 1.00 -0.17 0.16 -0.49 -0.42 -0.17 1.00 -0.06 -0.26 -0.04 0.16 -0.06 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (a) -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 [GeV] )t( T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 -0.42 -0.55 -0.24 -0.30 -0.42 1.00 -0.13 -0.22 -0.29 -0.55 -0.13 1.00 0.24 -0.18 -0.24 -0.22 0.24 1.00 -0.07 -0.30 -0.29 -0.18 -0.07 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (b) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.11 -0.14 -0.31 -0.11 1.00 -0.18 -0.51 -0.14 -0.18 1.00 -0.60 -0.31 -0.51 -0.60 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (c) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.08 -0.14 -0.28 -0.08 1.00 -0.03 -0.58 -0.14 -0.03 1.00 -0.68 -0.28 -0.58 -0.68 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (d) FIG. 18. Statistical correlation matrices for the normalized differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) y(¯t)\|. The contents of this ﬁgure are provided in machine-readable format in the Supplemental Material [74]. ACKNOWLEDGEMENTS -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 [GeV] (t) T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] (t) T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 -0.32 -0.60 -0.49 -0.26 -0.32 1.00 0.21 -0.42 -0.04 -0.60 0.21 1.00 -0.17 0.16 -0.49 -0.42 -0.17 1.00 -0.06 -0.26 -0.04 0.16 -0.06 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 [GeV] )t( T p [0,45] [45,75] [75,110] [110,150][150,500] [GeV] )t( T p [0,45] [45,75] [75,110] [110,150] [150,500] 1.00 -0.42 -0.55 -0.24 -0.30 -0.42 1.00 -0.13 -0.22 -0.29 -0.55 -0.13 1.00 0.24 -0.18 -0.24 -0.22 0.24 1.00 -0.07 -0.30 -0.29 -0.18 -0.07 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (b) (a) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \| y(t) \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.11 -0.14 -0.31 -0.11 1.00 -0.18 -0.51 -0.14 -0.18 1.00 -0.60 -0.31 -0.51 -0.60 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (c) (b) (b) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] \|)t y( \| [0.0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] 1.00 -0.08 -0.14 -0.28 -0.08 1.00 -0.03 -0.58 -0.14 -0.03 1.00 -0.68 -0.28 -0.58 -0.68 1.00 ATLAS =7 TeV s -1 dt = 4.59 fb L ∫ (d) (d) (c) FIG. 18. Statistical correlation matrices for the normalized differential cross section as a function of (a) pT(t), (b) pT(¯t), (c) \|y(t)\| and (d) \|y(¯t)\|. The contents of this ﬁgure are provided in machine-readable format in the Supplemental Material [74]. 29 TABLE VII. Normalized differential t-channel top-quark production cross section as a function of pT(t), pT(¯t), \|y(t)\| and \|y(¯t)\| with the uncertainties for each bin given in percent. The contents of this table are provided in machine-readable format in the Supplemental Mate- rial [74]. TABLE VI. Differential t-channel top-quark production cross section as a function of pT(t), pT(¯t), \|y(t)\| and \|y(¯t)\| with the uncertainties for each bin given in percent. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. pT(t) [GeV] dσ dpT(t) [ fb GeV] total [%] stat. [%] syst. ACKNOWLEDGEMENTS [%] [0,45] 440±70 ±15 ±7.4 ±13 [45,75] 370±60 ±16 ±6.5 ±14 [75,110] 250±40 ±15 ±7.7 ±13 [110,150] 133±27 ±20 ±12 ±16 [150,500] 7.8±1.9 ±24 ±16 ±19 pT(¯t) [GeV] dσ dpT(¯t) [ fb GeV] total [%] stat. [%] syst. [%] [0,45] 190±50 ±28 ±12 ±25 [45,75] 230±40 ±18 ±8.2 ±17 [75,110] 97±27 ±27 ±13 ±24 [110,150] 13.0±9.7 ±74 ±26 ±70 [150,500] 1.4±0.9 ±59 ±26 ±53 \|y(t)\| dσ d\|y(t)\|[pb] total [%] stat. [%] syst. [%] [0,0.2] 28±4 ±15 ±9.0 ±12 [0.2,0.6] 27.3±3.3 ±12 ±6.3 ±10 [0.6,1.1] 22.1±3.0 ±14 ±7.5 ±11 [1.1,3.0] 10.7±1.6 ±15 ±7.0 ±13 \|y(¯t)\| dσ d\|y(¯t)\| [pb] total [%] stat. [%] syst. [%] [0,0.2] 15.0±3.4 ±23 ±13 ±18 [0.2,0.6] 13.3±3.3 ±25 ±9.5 ±23 [0.6,1.1] 11.2±2.6 ±23 ±11 ±20 [1.1,3.0] 3.3±0.9 ±29 ±13 ±25 rial [74]. pT(t) [GeV] 1 σ dσ dpT(t)[ 10−3 GeV] total [%] stat. [%] syst. [%] [0,45] 9.2+0.8 −0.9 +8.4 −9.4 ±5.3 +6.5 −7.7 [45,75] 7.8±0.9 ±11 ±6.9 ±8.8 [75,110] 5.3±0.8 ±15 ±8.0 ±13 [110,150] 2.8±0.6 ±21 ±11 ±18 [150,500] 0.16±0.04 ±22 ±15 ±16 pT(¯t) [GeV] 1 σ dσ dpT(¯t)[ 10−3 GeV] total [%] stat. [%] syst. [%] [0,45] 9.6±1.6 ±17 ±8.2 ±15 [45,75] 11.6±1.8 ±15 ±8.8 ±12 [75,110] 4.9±1.2 ±25 ±13 ±21 [110,150] 0.7±0.4 +67 −61 ±25.8 +62 −56 [150,500] 0.07±0.04 ±51 ±26 ±45 \|y(t)\| 1 σ dσ d\|y(t)\| total [%] stat. [%] syst. [%] [0,0.2] 0.59±0.09 ±15 ±9.0 ±11 [0.2,0.6] 0.57±0.05 ±9.0 ±6.4 ±6.3 [0.6,1.1] 0.46±0.05 ±9.7 ±7.5 ±6.2 [1.1,3.0] 0.223±0.019 ±8.5 ±4.9 ±6.9 \|y(¯t)\| 1 σ dσ d\|y(¯t)\| total [%] stat. [%] syst. [%] [0,0.2] 0.75±0.14 ±19 ±13 ±13 [0.2,0.6] 0.66±0.11 ±17 ±9.1 ±14 [0.6,1.1] 0.555±0.095 ±17 ±11 ±13 [1.1,3.0] 0.163±0.030 ±18 ±11 ±15 TABLE VIII. Comparison between the measured differential cross sections and the predictions from the NLO calculation using the MSTW2008 PDF set. For each variable and prediction a χ2 value is calculated with HERAﬁtter using the covariance matrix of each measured spectrum. The theory uncertainties of the predictions are treated as uncorrelated. The number of degrees of freedom (NDF) is equal to the number of bins in the measured spectrum. The contents of this table are provided in machine-readable format in the Supple- mental Material [74]. TABLE VIII. Comparison between the measured differential cross sections and the predictions from the NLO calculation using the MSTW2008 PDF set. For each variable and prediction a χ2 value is calculated with HERAﬁtter using the covariance matrix of each measured spectrum. ACKNOWLEDGEMENTS The theory uncertainties of the predictions are treated as uncorrelated. The number of degrees of freedom (NDF) is equal to the number of bins in the measured spectrum. The contents of this table are provided in machine-readable format in the Supple- mental Material [74]. dσ dpT(t) dσ dpT(¯t) dσ d\|y(t)\| dσ d\|y(¯t)\| χ2/NDF 7.55/5 4.68/5 6.30/4 0.32/4 30 TABLE IX. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured dσ dpT(t) distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, jet vertex fraction, b/¯b acceptance, Emiss T modeling, W+jets shape variation, and t¯t generator. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. TABLE IX. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured dσ dpT(t) distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, jet vertex fraction, b/¯b acceptance, Emiss T modeling, W+jets shape variation, and t¯t generator. ACKNOWLEDGEMENTS The contents of this table are provided in machine-readable format in the Supplemental Material [74] dσ dpT(t) pT(t) bins [GeV] Source [0,45] [45,75] [75,110] [110,150] [150,500] Data statistical ±7.4 ±6.5 ±7.7 ±12 ±16 Monte Carlo statistical ±5.5 ±5.3 ±4.8 ±6.0 ±9.4 Background normalization ±6.1 ±7.5 ±5.2 ±3.0 ±5.2 JES η intercalibration < 1 +2.6/−1.3 +3.4/−1.9 < 1 +9.0/−4.2 b-JES < 1 +1.2/−2.3 < 1 ±1.6 < 1 Jet energy resolution ±1.0 ±2.4 ±2.3 ±3.0 < 1 b-tagging efﬁciency ±3.0 ±3.1 ±3.3 ±3.6 ±6.2 c-tagging efﬁciency ±1.3 ±1.5 < 1 < 1 < 1 Mistag efﬁciency ±2.0 ±1.9 < 1 < 1 ±1.2 Lepton uncertainties ±2.6 ±2.6 ±2.6 ±2.6 ±2.6 PDF ±3.0 ±1.8 ±2.3 ±2.8 ±2.4 tq generator + parton shower ±6.8 ±8.2 ∓7.9 ∓12 +9.2/−9.7 tq scale variation ±2.8 < 1 ±3.7 < 1 +6.0/−6.4 Unfolding ±1.3 ±1.4 < 1 < 1 < 1 Luminosity ±1.8 ±1.8 ±1.8 ±1.8 ±1.8 Total systematic ±13 ±14 ±13 ±16 ±19 Total ±15 ±16 ±15 ±20 ±25 31 31 TABLE X. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured dσ dpT(¯t) distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, mistag efﬁciency, b/¯b acceptance, Emiss T modeling, W+jets shape variation, and t¯t generator. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. ACKNOWLEDGEMENTS pT(¯t) bins [GeV] dσ dpT(¯t) pT(¯t) bins [GeV] Source [0,45] [45,75] [75,110] [110,150] [150,500] Data statistical ±12 ±8.2 ±13 ±26 ±26 Monte Carlo statistical ±12 ±9.1 ±14 ±28 ±28 Background normalization ±14 ±11 ±16 ±48 ±33 JES η intercalibration −9.0/+8.7 +1.9/−3.7 +4.9/−1.3 +15/−13 < 1 Jet energy resolution ±1.0 ±2.2 ±3.4 < 1 ±3.0 b-tagging efﬁciency ±3.0 ±3.1 ±3.2 ±3.6 ±5.9 c-tagging efﬁciency ±5.6 ±2.0 ±2.2 ±10 ±5.9 Lepton uncertainties ±2.6 ±2.6 ±2.6 ±2.6 ±2.7 PDF ±3.8 ±4.3 ±5.3 ±7.2 ±8.2 tq generator + parton shower ±12.2 < 1 ∓9.6 ±11 < 1 tq scale variation ±3.1 < 1 ±3.2 ±1.9 ±5.9 Unfolding < 1 < 1 < 1 ±6.9 ±2.6 Luminosity ±1.8 ±1.8 ±1.8 ±1.8 ±1.8 Total systematic ±25 ±17 ±24 ±70 ±53 Total ±27 ±18 ±27 ±74 ±59 32 32 TABLE XI. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured dσ d\|y(t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, jet vertex fraction, b/¯b acceptance, Emiss T modeling, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. dσ d\|y(t)\| \|y(t)\| bins Source [0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] Data statistical ±9.0 ±6.3 ±7.5 ±7.1 Monte Carlo statistical ±5.9 ±4.8 ±5.0 ±4.4 Background normalization ±5.3 ±6.5 ±6.7 ±4.7 JES η intercalibration +1.7/−0.6 < 1 +1.7/−0.4 < 1 b-JES +1.1/−1.7 < 1 +1.1/+0.2 < 1 Jet energy resolution ±3.2 ±1.7 < 1 ±3.1 b-tagging efﬁciency ±3.3 ±3.4 ±3.4 ±3.2 c-tagging efﬁciency ±1.3 ±1.2 ±1.2 ±1.0 Mistag efﬁciency < 1 ±1.3 ±2.0 ±1.4 Lepton uncertainties ±2.6 ±2.7 ±2.6 ±2.5 PDF ±3.6 ±3.6 ±2.8 ±2.8 tq generator + parton shower ∓5.7 ±0.8 ±4.0 ±8.7 tq scale variation ±3.5 < 1 ±2.6 ±4.7 Luminosity ±1.8 ±1.8 ±1.8 ±1.8 Total systematic ±12 ±10 ±11 ±14 Total ±15 ±12 ±14 ±15 33 33 TABLE XII. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured dσ d\|y(¯t)\| distribution given in percent for each bin. ACKNOWLEDGEMENTS The list includes only those uncertainties that contribute with more than 1%. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, b/¯b acceptance, mistag efﬁciency, Emiss T modeling, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. dσ d\|y(¯t)\| \|y(¯t)\| bins Source [0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] Data statistical ±13 ±9.5 ±11 ±13 Monte Carlo statistical ±11 ±12 ±11 ±17 Background normalization ±11 ±16 ±13 ±15 JES η intercalibration < 1 +1.0/−1.8 < 1 +2.3/−0.9 Jet energy resolution ±2.3 ±2.2 ±1.0 ±3.2 b-tagging efﬁciency ±3.4 ±3.3 ±3.2 ±3.2 c-tagging efﬁciency ±2.5 ±3.6 ±2.9 ±4.0 Lepton uncertainties ±2.7 ±2.7 ±2.6 ±2.4 PDF ±6.0 ±5.3 ±4.4 ±4.1 tq generator + parton shower ±1.0 ∓5.6 ±6.6 ±6.2 tq scale variation ±2.1 ±2.6 ±1.6 ±4.3 Luminosity ±1.8 ±1.8 ±1.8 ±1.8 Total systematic ±18 ±23 ±20 ±25 Total ±23 ±25 ±23 ±29 34 34 TABLE XIII. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ dpT(t) distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. The JES η intercalibration uncertainty has a sign switch from the ﬁrst to the second bin. For the tq generator + parton shower uncertainty a sign switch is denoted with ∓. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, b/¯b acceptance, c-tagging efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, and t¯t generator. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. ACKNOWLEDGEMENTS Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ d\|y(t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. Sign switches within one uncertainty are denoted with ∓and ±. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, b/¯b acceptance, b-tagging efﬁciency, c-tagging efﬁciency, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. TABLE XV. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ d\|y(t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. Sign switches within one uncertainty are denoted with ∓and ±. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, b/¯b acceptance, b-tagging efﬁciency, c-tagging efﬁciency, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. 1 σ dσ d\|y(t)\| \|y(t)\| bins Source [0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] Data statistical ±9.0 ±6.4 ±7.5 ±5.0 Monte Carlo statistical ±5.9 ±4.8 ±4.9 ±3.2 Background normalization < 1 < 1 ±1.1 ±1.0 JES η intercalibration +1.6/−1.5 −0.5/+2.3 +1.4/−1.5 < 1 Jet energy resolution ±1.2 < 1 ∓1.6 ±1.0 PDF ±1.7 ±1.8 < 1 ±2.3 tq generator + parton shower −9.0/+9.8 −2.8/+3.0 < 1 +4.8/−5.2 tq scale variation < 1 < 1 < 1 ±1.5 Total systematic ±11 ±6.3 ±6.2 ±6.9 Total ±15 ±9.0 ±9.7 ±8.5 \|y(t)\| bins TABLE XVI. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ d\|y(¯t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. ACKNOWLEDGEMENTS Sign switches within one uncertainty are denoted with ∓and ±. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet energy resolution, jet vertex fraction, b/¯b acceptance, b-tagging efﬁciency, c-tagging efﬁciency, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. 1 σ dσ d\|y(¯t)\| \|y(¯t)\| bins Source [0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] Data statistical ±13 ±9.1 ±11 ±11 Monte Carlo statistical ±12 ±11 ±12 ±14 Background normalization ±3.4 ±2.4 ±1.1 < 1 JES η intercalibration < 1 +0.5/−1.9 < 1 +1.5/−0.8 PDF ±1.6 ±1.0 < 1 ±1.8 tq generator + parton shower ∓1.4 −7.8/+8.2 +4.0/−4.3 +3.8/−3.9 tq scale variation ±1.9 < 1 < 1 < 1 Total systematic ±13 ±14 ±13 ±15 Total ±19 ±17 ±17 ±18 TABLE XVI. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ d\|y(¯t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. Sign switches within one uncertainty are denoted with ∓and ±. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet energy resolution, jet vertex fraction, b/¯b acceptance, b-tagging efﬁciency, c-tagging efﬁciency, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. TABLE XVI. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ d\|y(¯t)\| distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. Sign switches within one uncertainty are denoted with ∓and ±. ACKNOWLEDGEMENTS 1 σ dσ dpT(t) pT(t) bins [GeV] Source [0,45] [45,75] [75,110] [110,150] [150,500] Data statistical ±5.3 ±6.9 ±8.0 ±11 ±15 Monte Carlo statistical ±4.2 ±5.5 ±5.2 ±6.2 ±9.3 Background normalization < 1 ±1.7 < 1 ±3.0 < 1 JES η intercalibration −4.7/+1.5 +3.5/−2.3 +4.1/−0.8 < 1 +9.6/−3.1 Jet energy resolution < 1 < 1 < 1 ∓1.4 ±2.7 b-tagging efﬁciency < 1 < 1 < 1 < 1 ±2.8 Mistag efﬁciency < 1 < 1 < 1 ±1.0 < 1 tq generator + parton shower ±3.9 ±5.4 ∓11 ∓14 ±6.9 tq scale variation < 1 ∓1.8 ±1.3 ∓2.7 +4.4/−5.1 Unfolding < 1 ±1.7 < 1 < 1 ±1.1 Total systematic +6.5/-7.7 ±8.8 ±13 ±18 ±16 Total +8.4/-9.4 ±11 ±15 ±21 ±22 35 35 TABLE XIV. Detailed list of the contribution of each source of uncertainty to the total relative uncertainty on the measured 1 σ dσ dpT(¯t) distribution given in percent for each bin. The list includes only those uncertainties that contribute with more than 1%. Sign switches within one uncertainty are denoted with ∓and ±. The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet vertex fraction, b/¯b acceptance, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, and t¯t generator. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. 1 σ dσ dpT(¯t) pT(¯t) bins [GeV] Source [0,45] [45,75] [75,110] [110,150] [150,500] Data statistical ±8.2 ±8.8 ±13 ±26 ±26 Monte Carlo statistical ±8.7 ±9.6 ±14 ±28 ±27 Background normalization < 1 ±4.5 ±1.8 ±39 ±22 JES η intercalibration −7.5/+6.7 +3.8/−5.3 +6.9/−3.1 +17/−9.9 < 1 Jet energy resolution < 1 < 1 ∓1.6 ±1.8 ∓1.2 b-tagging efﬁciency < 1 < 1 < 1 < 1 +2.4/−2.8 c-tagging efﬁciency ∓1.8 ±2.0 ±1.7 −6.2/+5.9 ∓2.0 PDF < 1 < 1 < 1 ±2.5 ±3.6 tq generator + parton shower +7.7/−8.2 −3.6/+3.7 −13/+14 +6.4/−7.0 −4.2/+4.5 tq scale variation ± 1.3 ∓3.0 ± 1.4 ∓1.8 ± 5.1 Unfolding < 1 < 1 < 1 ±6.7 ±2.8 Total systematic ±15 ±13 ±21 +62/−56 ±45 Total ±17 ±15 ±25 +67/−61 ±52 36 TABLE XV. 04, 063 (2008). [6] T. M. Tait and C.-P. Yuan, Phys. Rev. D 63, 014018 (2000 [7] J. Aguilar-Saavedra, Nucl. Phys. B 804, 160 (2008). [40] ATLAS Collaboration, ATL-LARG-PUB-2008-002 (2008), https://cdsweb.cern.ch/record/1099735. g y [8] J. Gao, C. S. Li, L. L. Yang, and H. Zhang, Phys. Rev. Lett. ] , , Rev. D 69, 094008 (2004). [12] N. Kidonakis, Phys. Rev. D 88, 031504 (2013). [46] J. M. Campbell and R. Ellis, Nucl. Phys. Proc. Suppl. 205-206, 10 (2010). [13] R. Coimbra, A. Onofre, R. Santos, and M. W [13] R. Coimbra, A. Onofre, R. Santos, and M. Won, Eur. Phys. J. C 72, 2222 (2012). R. Frederix, 130 (2012). [62] ATLAS Collaboration, ATLAS-CONF-2012-043 (2012), https://cdsweb.cern.ch/record/1435197. [28] P. Z. Skands, Phys. Rev. D 82, 074018 (2010). [63] ATLAS Collaboration, ATLAS-CONF-2012-039 (2012), https://cdsweb.cern.ch/record/1435193. [29] M. L. Mangano, M. Moretti, F. Piccinini, R. Pittau, and A. D. Polosa, J. High Energy Phys. 07, 001 (2003). [64] ATLAS Collaboration, ATLAS-CONF-2012-040 (2012), https://cdsweb.cern.ch/record/1435194. [30] J. Pumplin, D. Stump, J. Huston, H.-L. Lai, P. M. Nadolsky, and W. Tung, J. High Energy Phys. 07, 012 (2002). [65] J. Alwall, R. Frederix, S. Frixione, V. Hirschi, F. Maltoni, O. Mattelaer, H.-S. Shao, T. Stelzer, P. Torrielli, and M. Zaro, (2014), arXiv:1405.0301 [hep-ph]. [31] G. Corcella, I. Knowles, G. Marchesini, S. Moretti, K. Odagiri, P. Richardson, and B. Seymour, M.H. Webber, J. High Energy Phys. 01, 010 (2001). g 250 (2003). [3] G. Altarelli and G. Parisi, Nucl. Phys. B 126, 298 (19 37] ATLAS Collaboration, (2014), arXiv:1404.2240 [hep- [4] Y. L. Dokshitzer, Sov. Phys. JETP 46, 641 (1977). [38] ATLAS Collaboration, (2014), arXiv:1404.4562 [hep-ex]. [5] V. Gribov and L. Lipatov, Sov. J. Nucl. Phys. 15, 438 (1972). [39] M. Cacciari, G. P. Salam, and G. Soyez, J. High Energy Phys. 04, 063 (2008). [39] M. Cacciari, G. P. Salam, and G. Soyez, J. High Energy Phys. 109, 132001 (2012). [17] CMS Collaboration, J. High Energy Phys. 12, 035 (2012). [18] CMS Collaboration, Phys. Rev. Lett. 110, 022003 (2013). [19] CMS Collaboration, (2014), arXiv:1403.7366 [hep-ex]. [17] CMS Collaboration, J. High Energy Phys. 12, 035 (2012). g gy y [18] CMS Collaboration, Phys. Rev. Lett. 110, 022003 (2013). 19 C S C ll b i (2014) i 1403 366 h [50] M. Czakon and A. Mitov, J. High Energy Phys. 12, 054 (201 [18] CMS Collaboration, Phys. Rev. Lett. 110, 022003 (2013). [50] M. Czakon and A. Mitov, J. High Energy Phys. 12, 054 (2012). [51] M. Czakon and A. Mitov, J. High Energy Phys. 01, 080 (2013). [19] CMS Collaboration, (2014), arXiv:1403.7366 [hep-ex]. [51] M. Czakon and A. Mitov, J. High Energy Phys. 01, 080 (2013). [52] M. Czakon, P. Fiedler, and A. Mitov, Phys. Rev. Lett. 110, 252004 (2013). [20] ATLAS Collaboration, JINST 3, S08003 (2008). , ( ) [9] N. Kidonakis, Phys. Rev. D 83, 091503 (2011). , ( ) [9] N. Kidonakis, Phys. Rev. D 83, 091503 (2011). [42] ATLAS Collaboration, (2014), arXiv:1406.0076 [hep-ex]. y [10] A. Martin, W. Stirling, R. Thorne, and G. Watt, Eu [43] ATLAS Collaboration, Eur. Phys. J. C 72, 1844 (2012 [10] A. Martin, W. Stirling, R. Thorne, and G. Watt, Eur. Phys. J. C 63, 189 (2009). T. Sjostrand, S. 05, 026 (2006). [59] N. Kidonakis, Phys. Rev. D 81, 054028 (2010). p Phys. Rev. Lett. 102, 182003 (2009). [45] C. Anastasiou, L. J. Dixon, K. Melnikov, and F. Petriello, Phys. Rev D 69 094008 (2004) 05, 026 (2006). y ( ) [60] M. Feindt and U. Kerzel, Nucl. Instrum. Meth. A 559, 190 (2006) [60] M. Feindt and U. Kerzel, Nucl. Instrum. Meth. A 559, 190 [26] H.-L. Lai, M. Guzzi, J. Huston, Z. Li, P. M. Nadolsky, J. Pumplin, and C.-P. Yuan, Phys. Rev. D 82, 074024 (2010). , C 72, 2222 (2012). [47] ATLAS Collaboration, Phys. Lett. B 711, 244 (2012). [14] V. Abazov et al. (D0 Collaboration), Phys. Lett. B 682, 363 (2010). [48] M. Cacciari, M. Czakon, M. Mangano, A. Mitov, and P. Nason, Phys. Lett. B 710, 612 (2012). [48] M. Cacciari, M. Czakon, M. Mangano, A. Mitov, and [15] ATLAS Collaboration, Phys. Lett. B 717, 330 (2012). [49] P. Baernreuther, M. Czakon, and A. Mitov, Phys. Rev. Lett. 109, 132001 (2012). [16] ATLAS Collaboration, Phys. Lett. B 716, 142 (2012). ACKNOWLEDGEMENTS The following uncertainties contribute to the total uncertainty with less than 1% to each bin content: JES detector, JES statistical, JES physics modeling, JES mixed detector and modeling, JES close-by jets, JES pile-up, JES ﬂavor composition, JES ﬂavor response, b-JES, jet energy resolution, jet vertex fraction, b/¯b acceptance, b-tagging efﬁciency, c-tagging efﬁciency, mistag efﬁciency, Emiss T modeling, lepton uncertainties, W+jets shape variation, t¯t generator, t¯t ISR/FSR, and unfolding. The contents of this table are provided in machine-readable format in the Supplemental Material [74]. 1 σ dσ d\|y(¯t)\| \|y(¯t)\| bins Source [0,0.2] [0.2,0.6] [0.6,1.1] [1.1,3.0] Data statistical ±13 ±9.1 ±11 ±11 Monte Carlo statistical ±12 ±11 ±12 ±14 Background normalization ±3.4 ±2.4 ±1.1 < 1 JES η intercalibration < 1 +0.5/−1.9 < 1 +1.5/−0.8 PDF ±1.6 ±1.0 < 1 ±1.8 tq generator + parton shower ∓1.4 −7.8/+8.2 +4.0/−4.3 +3.8/−3.9 tq scale variation ±1.9 < 1 < 1 < 1 Total systematic ±13 ±14 ±13 ±15 Total ±19 ±17 ±17 ±18 37 [1] N. Cabibbo, Phys. Rev. Lett. 10, 531 (1963). [35] ATLAS Collaboration, Eur. Phys. J. C 70, 823 (2010). [36] S. Agostinelli et al. (GEANT4), Nucl. Instrum. Meth. A 506, 250 (2003). [2] M. Kobayashi and T. Maskawa, Prog. Theor. Phys. 49, 652 (1973). [66] ATLAS Collaboration, Eur. Phys. J. C 72, 2043 (2012). [66] ATLAS Collaboration, Eur. Phys. J. C 72, 2043 (2012). [32] J. Butterworth, J. R. Forshaw, and M. Seymour, Z. Phys. C 72, 637 (1996). (2006). [61] ATLAS Collaboration, Eur. Phys. J. C 73, 2306 (2013). [27] R. Frederix, E. Re, and P. Torrielli, J. High Energy Phys. 09, 130 (2012). , 252004 (2013). [21] ATLAS Collaboration, Eur. Phys. J. C 73, 2518 (201 ( ) [53] M. Czakon and A. Mitov, (2011), arXiv:1112.5675 [hep-ph]. [22] ATLAS uses a right-handed coordinate system with its origin at the nominal interaction point in the center of the detector and the z-axis along the beam direction. The z-axis is parallel to the anti-clockwise beam viewed from above. The pseudorapidity η is deﬁned as η = −ln[tan(θ/2)], where the polar angle θ is measured with respect to the z-axis. The azimuthal angle φ is measured with respect to the x-axis, which points toward the center of the LHC ring. Transverse momentum and energy are deﬁned as pT = psinθ and ET = E sinθ, respectively. The ∆R distance in (η,φ) space is deﬁned as ∆R = p (∆η)2 +(∆φ)2. [54] M. Botje, J. Butterworth, A. Cooper-Sarkar, A. de Roeck, J. Feltesse, S. Forte, A. Glazov, J. Huston, R. McNulty, T. Sjos- trand, and R. Thorne, (2011), arXiv:1101.0538 [hep-ph]. [55] A. Martin, W. Stirling, R. Thorne, and G. Watt, Eur. Phys. J. C 64, 653 (2009). [56] J. Gao, M. Guzzi, J. Huston, H.-L. Lai, Z. Li, P. M. Nadolsky, J. Pumplin, D. Stump, and C.-P. Yuan, Phys. Rev. D 89, 033009 (2014). [57] R. D. Ball, V. Bertone, S. Carrazza, C. S. Deans, L. Del Debbio, S. Forte, A. Guffanti, N. P. Hartland, J. I. Latorre, J. Rojo, and M. Ubiali, Nucl. Phys. B 867, 244 (2013). [23] ATLAS Collaboration, JINST 7, C01092 (2012). ( ) [24] P. Nason, J. High Energy Phys. 11, 040 (2004). , y , ( ) [58] N. Kidonakis, Phys. Rev. D 82, 054018 (2010). [25] T. Sjostrand, S. Mrenna, and P. Z. Skands, J. High Energy Phys. 05, 026 (2006). 107, 092002 (2011). p [41] ATLAS Collaboration, Eur. Phys. J. C 73, 2304 (2013). 63, 189 (2009). [44] ATLAS Collaboration, ATLAS-CONF-2011-102 (2011), https://cdsweb.cern.ch/record/1369219. ( ) [11] J. M. Campbell, R. Frederix, F. Maltoni, and F. Tramontano, [11] J. M. Campbell, R. Frederix, F. Maltoni, and F. Tramontano, Phys. Rev. Lett. 102, 182003 (2009). [11] J. M. Campbell, R. Frederix, F. Maltoni, and F. Tra 637 (1996). [67] R. Thorne, PoS DIS2010, 053 (2010). [68] P. Kant, O. Kind, T. Kintscher, T. Lohse, T. Martini, S. Moelb- itz, P. Rieck, and P. Uwer, (2014), arXiv:1406.4403 [hep-ph]. [33] ATLAS Collaboration, ATL-PHYS-PUB-2011-008 (2011), https://cds.cern.ch/record/1345343. [69] S. Alekhin, J. Bluemlein, and S. Moch, Phys. Rev. D 86, 054009 (2012). [34] J. Alwall, S. Hoche, F. Krauss, N. Lavesson, L. Lonnblad, F. Maltoni, M. L. Mangano, M. Moretti, C. Papadopoulos, F. Piccinini, S. Schumann, M. Treccani, J. Winter, and M. Worek, Eur. Phys. J. C 53, 473 (2008). [70] M. Gluck, P. Jimenez-Delgado, E. Reya, and C. Schuck, Phys. Lett. B 664, 133 (2008). [70] M. Gluck, P. Jimenez-Delgado, E. Reya, and C. Schuck, Phys. Lett. B 664, 133 (2008). 38 38 [71] F. Aaron et al. (H1 and ZEUS Collaboration), J. High Energy Phys. 01, 109 (2010). [72] F. Aaron et al. (H1 Collaboration), Eur. Phys. J. C 64, 561 (2009). [73] G. D’Agostini, Nucl. Instrum. Meth. A 362, 487 (1995). [74] See Supplemental Material at http://link.aps.org/supplemental/10.1103/PhysRevD.90.112006 for machine readable tables of the results, uncertainties, and correlations of the differential cross-section measurements. 39 The ATLAS Collaboration C. Anastopoulos140, L.S. Ancu49, N. Andari30, T. Andeen35, C.F. Anders58b, G. Anders30, K.J. Anderson31, C. Anastopoulos140, L.S. Ancu49, N. Andari30, T. Andeen35, C.F. Anders58b, G. Anders30, K.J. Anderson31, g gy p M. Arratia28, O. Arslan21, A. Artamonov96, G. Artoni23, S. Asai156, N. Asbah42, A. Ashkenazi154, B. ˚Asman147a,147b, j V. Bansal170, H.S. Bansil18, L. Barak173, S.P. Baranov95, E.L. Barberio87, D. Barberis50a,50b, M. Barbero84, T. Barillari100, M. Beckingham171, C. Becot116, A.J. Beddall19c, A. Beddall19c, S. Bedikian177, V.A. Bednyakov64, C.P. Bee149, g g p g L. Bellagamba20a, A. Bellerive29, M. Bellomo85, K. Belotskiy97, O. Beltramello30, O. Benary154, D. Benchekroun136a, j M.F. Bessner42, N. Besson137, C. Betancourt48, S. Bethke100, W. Bhimji46, R.M. Bianchi124, L. Bianchini23, M. Bianco30, 99 54 15 135 49 4 , , , , y , , , g , A.G. Bogdanchikov108, A. Bogouch91,∗, C. Bohm147a, J. Bohm126, V. Boisvert76, T. Bold38a, V. Boldea26a, A.S. Boldyrev98, 40 M. Cavalli-Sforza12, V. Cavasinni123a,123b, F. Ceradini135a,135b, B. Cerio45, K. Cerny128, A.S. Cerqueira24b, A. Cerri150, L Cerrito75 F Cer tti15 M Cer 30 A Cer elli17 S A Cetin19b A Chafaq136a D Chakrabort 107 I Chal pko a128 g H. Chen25, K. Chen149, L. Chen33d,g, S. Chen33c, X. Chen146c, Y. Chen66, Y. Chen35, H.C. Cheng88, Y. Cheng31, , , y , , , , ers6, A. Chilingarov71, G. Chiodini72a, A.S. Chisholm18, R.T. Chislett77, A. Chitan26a, M.V. Chizhov64, g S. Chouridou9, B.K.B. Chow99, D. Chromek-Burckhart30, M.L. Chu152, J. Chudoba126, J.J. Chwastowski39, L. Chytka114, 133 133b 4 4 53 74 15 13b 173 p , , , , , , , , M. Citterio90a, M. Ciubancan26a, A. Clark49, P.J. Clark46, R.N. Clarke15, W. Cleland124, J.C. Clemens84, C. Clement147a,147b, , , , , , , , u84, M. Cobal165a,165c, A. Coccaro139, J. Cochran63, L. Coffey23, J.G. Cogan144, J. Coggeshall166, B. Cole3 bal , A. Coccaro , J. Cochran , L. Coffey , J.G. Cogan , J. Coggeshall , B. Cole , jn106, J. Collot55, T. Colombo58c, G. Colon85, G. Compostella100, P. Conde Mui˜no125a,125b, E. Coniavitis48 y g gg 7, A.P. Colijn106, J. Collot55, T. Colombo58c, G. Colon85, G. Compostella100, P. Conde Mui˜no125a,125b, E. C , j , , , , p , , M.C. Conidi12, S.H. Connell146b, I.A. Connelly76, S.M. Consonni90a,90b, V. Consorti48, S. Constantinescu26a, C. Conta120a,120b, j p M.C. Conidi12, S.H. Connell146b, I.A. Connelly76, S.M. Consonni90a,90b, V. Consorti48, S. Constantinescu26a, C. g , y , , j , , g , , S. Donati123a,123b, P. Dondero120a,120b, J. Donini34, J. Dopke130, A. Doria103a, M.T. Dova70, A.T. Doyle53, M. Dris10, J. Dubbert88, S. Dube15, E. Dubreuil34, E. Duchovni173, G. Duckeck99, O.A. Ducu26a, D. Duda176, A. Dudarev30, F. Dudziak63, L. Duﬂot116, L. Duguid76, M. D¨uhrssen30, M. Dunford58a, H. Duran Yildiz4a, M. D¨uren52, A. Durglishvili51b, M. Dwuznik38a, M. Dyndal38a, J. Ebke99, W. Edson2, N.C. Edwards46, W. Ehrenfeld21, T. Eifert144, G. Eigen14, K. Einsweiler15, T. Ekelof167, M. El Kacimi136c, M. Ellert167, S. Elles5, F. Ellinghaus82, N. Ellis30, J. Elmsheuser99, M. Elsing30, D. Emeliyanov130, Y. Enari156, O.C. Endner82, M. Endo117, R. Engelmann149, J. Erdmann177, A. Ereditato17, D. Eriksson147a, G. Ernis176, J. Ernst2, M. Ernst25, J. Ernwein137, D. Errede166, S. Errede166, E. Ertel82, M. Escalier116, H. Esch43, C. Escobar124, B. Esposito47, A.I. Etienvre137, E. Etzion154, H. Evans60, A. Ezhilov122, L. Fabbri20a,20b, G. Facini31, R.M. Fakhrutdinov129, S. Falciano133a, R.J. Falla77, J. Faltova128, Y. Fang33a, M. Fanti90a,90b, A. Farbin8, A. Farilla135a, T. Farooque12, S. Farrell15, S.M. Farrington171, P. Farthouat30, F. Fassi136e, P. Fassnacht30, D. Fassouliotis9, A. Favareto50a,50b, L. Fayard116, P. Federic145a, O.L. Fedin122,j, W. Fedorko169, M. Fehling-Kaschek48, S. Feigl30, L. Feligioni84, C. Feng33d, E.J. Feng6, H. Feng88, A.B. Fenyuk129, S. Fernandez Perez30, S. Ferrag53, J. Ferrando53, A. Ferrari167, P. Ferrari106, R. Ferrari120a, D.E. Ferreira de Lima53, A. Ferrer168, D. Ferrere49, C. Ferretti88, A. Ferretto Parodi50a,50b, M. Fiascaris31, F. Fiedler82, A. Filipˇciˇc74, M. Filipuzzi42, F. Filthaut105, M. Fincke-Keeler170, K.D. Finelli151, M.C.N. Fiolhais125a,125c, L. Fiorini168, A. Firan40, A. Fischer2, J. Fischer176, W.C. Fisher89, E.A. Fitzgerald23, M. Flechl48, I. Fleck142, P. Fleischmann88, S. Fleischmann176, G.T. Fletcher140, G. Fletcher75, T. Flick176, A. Floderus80, L.R. Flores Castillo174,k, A.C. Florez Bustos160b, M.J. Flowerdew100, A. Formica137, A. Forti83, D. Fortin160a, D. Fournier116, H. Fox71, S. Fracchia12, P. Francavilla79, M. Franchini20a,20b, S. Franchino30, D. Francis30, L. Franconi118, M. Franklin57, S. Franz61, M. Fraternali120a,120b, S.T. French28, C. Friedrich42, F. Friedrich44, D. Froidevaux30, J.A. Frost28, C. Fukunaga157, E. Fullana Torregrosa82, B.G. Fulsom144, J. Fuster168, C. Gabaldon55, O. Gabizon173, A. Gabrielli20a,20b, A. Gabrielli133a,133b, S. Gadatsch106, S. Gadomski49, G. Gagliardi50a,50b, P. Gagnon60, C. Galea105, B. Galhardo125a,125c, E.J. Gallas119, V. Gallo17, B.J. Gallop130, P. Gallus127, G. Galster36, K.K. Gan110, R.P. Gandrajula62, J. Gao33b,g, Y.S. Gao144,e, F.M. Garay Walls46, F. Garberson177, C. Garc´ıa168, J.E. Garc´ıa Navarro168, M. Garcia-Sciveres15, R.W. Gardner31, N. Garelli144, V. Garonne30, C. Gatti47, G. Gaudio120a, B. Gaur142, L. Gauthier94, P. Gauzzi133a,133b, I.L. Gavrilenko95, C. Gay169, G. Gaycken21, E.N. Gazis10, A.B. Fenyuk , S. Fernandez Perez , S. Ferrag , J. Ferrando , A. Ferrari , P. Ferrari , R. Ferrari , D.E. Ferreira de Lima53, A. Ferrer168, D. Ferrere49, C. Ferretti88, A. Ferretto Parodi50a,50b, M. Fiascaris31, F. Fiedler82, A. Filipˇciˇc74, M. Filipuzzi42, F. Filthaut105, M. Fincke-Keeler170, K.D. Finelli151, M.C.N. Fiolhais125a,125c, L. Fiorini168, A. Firan40, A. Fischer2, J. Fischer176, W.C. Fisher89, E.A. Fitzgerald23, M. Flechl48, I. Fleck142, P. Fleischmann88, S. Fleischmann176, G.T. Fletcher140, G. Fletcher75, T. Flick176, A. Floderus80, L.R. Flores Castillo174,k, A.C. Florez Bustos160b M.J. Flowerdew100, A. Formica137, A. Forti83, D. Fortin160a, D. Fournier116, H. Fox71, S. Fracchia12, P. Francavilla79, M. Franchini20a,20b, S. Franchino30, D. Francis30, L. Franconi118, M. Franklin57, S. Franz61, M. Fraternali120a,120b, S.T. French28, C. Friedrich42, F. Friedrich44, D. Froidevaux30, J.A. Frost28, C. Fukunaga157, E. Fullana Torregrosa82, B.G. Fulsom144, J. Fuster168, C. Gabaldon55, O. Gabizon173, A. Gabrielli20a,20b, A. Gabrielli133a,133b, S. Gadatsch106, S. Gadomski49, G. Gagliardi50a,50b, P. Gagnon60, C. Galea105, B. Galhardo125a,125c, E.J. Gallas119, V. Gallo17, B.J. Gallop130, P. Gallus127, G. Galster36, K.K. Gan110, R.P. Gandrajula62, J. Gao33b,g, Y.S. Gao144,e, F.M. Garay Walls46, F. Garberson177, C. Garc´ıa168, J.E. Garc´ıa Navarro168, M. Garcia-Sciveres15, R.W. Gardner31, N. Garelli144, V. Garonne30, C. Gatti47, G. Gaudio120a, B. Gaur142, L. Gauthier94, P. Gauzzi133a,133b, I.L. Gavrilenko95, C. Gay169, G. Gaycken21, E.N. Gazis10, , , , , , , a, B. Gaur142, L. Gauthier94, P. Gauzzi133a,133b, I.L. Gavrilenko95, C. Gay169, G. Gaycken21, E.N. Gazis10, The ATLAS Collaboration Conta120a,120b, G C ti57 F C ti103a h M C k 15 B D C 77 A M C S k 119 N J C S ith76 K C i 15 , , , p , p , p , p , T. Cornelissen176, M. Corradi20a, F. Corriveau86,i, A. Corso-Radu164, A. Cortes-Gonzalez12, G. Cortiana100, G. Costa90a, pe-Renaudin55, F. Crescioli79, W.A. Cribbs147a,147b, M. Crispin Ortuzar119, M. Cristinziani21, V. Croft105, osetti37a,37b, C.-M. Cuciuc26a, T. Cuhadar Donszelmann140, J. Cummings177, M. Curatolo47, C. Cuthbert15 P. Czodrowski , Z. Czyczula , S. D Auria , M. D Onofrio , M.J. Da Cunha Sargedas De Sousa , C. Da Via , W. Dabrowski38a, A. Daﬁnca119, T. Dai88, O. Dale14, F. Dallaire94, C. Dallapiccola85, M. Dam36, A.C. Daniells18, y g W. Dabrowski38a, A. Daﬁnca119, T. Dai88, O. Dale14, F. Dallaire94, C. Dallapiccola85, M. Dam36, A.C. Daniells18, M. Dano Hoffmann137, V. Dao48, G. Darbo50a, S. Darmora8, J.A. Dassoulas42, A. Dattagupta60, W. Davey21, C. David170, T. Davidek128, E. Davies119,c, M. Davies154, O. Davignon79, A.R. Davison77, P. Davison77, Y. Davygora58a, E. Dawe143, 140 85 8 103 20 20b 79 105 T. Davidek128, E. Davies119,c, M. Davies154, O. Davignon79, A.R. Davison77, P. Davison77, Y. Davygora58a, E. Dawe143, 140 85 8 103 20 20b 79 105 P. de Jong , H. De la Torre , F. De Lorenzi , L. De Nooij , D. De Pedis , A. De Salvo , U. De Sanctis , A. De Santo150, J.B. De Vivie De Regie116, W.J. Dearnaley71, R. Debbe25, C. Debenedetti138, B. Dechenaux55, A. De Santo150, J.B. De Vivie De Regie116, W.J. Dearnaley71, R. Debbe25, C. Debenedetti138, B. Dechenaux55, g y g y A. Dell’Acqua30, L. Dell’Asta22, M. Dell’Orso123a,123b, M. Della Pietra103a,h, D. della Volpe49, M. Delmastro5, P.A. Delsart55, L. Di Ciaccio5, A. Di Domenico133a,133b, C. Di Donato103a,103b, A. Di Girolamo30, B. Di Girolamo30, A. Di Mattia153, B. Di Micco135a,135b, R. Di Nardo47, A. Di Simone48, R. Di Sipio20a,20b, D. Di Valentino29, F.A. Dias46, M.A. Diaz32a, E.B. Diehl88, J. Dietrich42, T.A. Dietzsch58a, S. Diglio84, A. Dimitrievska13a, J. Dingfelder21, C. Dionisi133a,133b, P. Dita26a, S. Dita26a, F. Dittus30, F. Djama84, T. Djobava51b, M.A.B. do Vale24c, A. Do Valle Wemans125a,125g, T.K.O. Doan5, D. Dobos30, p E.B. Diehl88, J. Dietrich42, T.A. Dietzsch58a, S. Diglio84, A. Dimitrievska13a, J. Dingfelder21, C. Dionisi133a,133b, P. Dita26a, 26 30 84 51b 24 125 125 5 30 g , y , , j , , g , , S. Donati123a,123b, P. Dondero120a,120b, J. The ATLAS Collaboration Donini34, J. Dopke130, A. Doria103a, M.T. Dova70, A.T. Doyle53, M. Dris10, , , , , , , , , , L. Duﬂot116, L. Duguid76, M. D¨uhrssen30, M. Dunford58a, H. Duran Yildiz4a, M. D¨uren52, A. Durglishvili51b, M. Dwuznik38a, 6 5 6 , , , , , , , , , B. Esposito47, A.I. Etienvre137, E. Etzion154, H. Evans60, A. Ezhilov122, L. Fabbri20a,20b, G. Facini31, R.M. Fakhrutdinov129, 41 P. Ge33d, Z. Gecse169, C.N.P. Gee130, D.A.A. Geerts106, Ch. Geich-Gimbel21, K. Gellerstedt147a,147b, C. Gemme50a, 53 55 133 133b 54 76 164 154 A. Gemmell53, M.H. Genest55, S. Gentile133a,133b, M. George54, S. George76, D. Gerbaudo164, A. Gershon154, H. Ghazlane136 N. Ghodbane34, B. Giacobbe20a, S. Giagu133a,133b, V. Giangiobbe12, P. Giannetti123a,123b, F. Gianotti30, B. Gibbard25, A. Gemmell , M.H. Genest , S. Gentile , M. George , S. George , D. Gerbaudo , A. Gershon , H. Ghazlane N. Ghodbane34, B. Giacobbe20a, S. Giagu133a,133b, V. Giangiobbe12, P. Giannetti123a,123b, F. Gianotti30, B. Gibbard25, S.M. Gibson76, M. Gilchriese15, T.P.S. Gillam28, D. Gillberg30, G. Gilles34, D.M. Gingrich3,d, N. Giokaris9, M.P. Giordani165a,165c, R. Giordano103a,103b, F.M. Giorgi20a, F.M. Giorgi16, P.F. Giraud137, D. Giugni90a, C. Giuliani48, , , , g , , g , , M.P. Giordani165a,165c, R. Giordano103a,103b, F.M. Giorgi20a, F.M. Giorgi16, P.F. Giraud137, D. Giugni90a, C. Giuliani48, , , g , g , , g , Giulini58b, B.K. Gjelsten118, S. Gkaitatzis155, I. Gkialas155,l, L.K. Gladilin98, C. Glasman81, J. Glatzer30, j Glazov42, G.L. Glonti64, M. Goblirsch-Kolb100, J.R. Goddard75, J. Godfrey143, J. Godlewski30, C. Goerin y g S. Goldfarb88, T. Golling177, D. Golubkov129, A. Gomes125a,125b,125d, L.S. Gomez Fajardo42, R. Gonc¸alo125a, J. Goncalves Pinto Firmino Da Costa137, L. Gonella21, S. Gonz´alez de la Hoz168, G. Gonzalez Parra12, S. Gonzalez-Sevilla49, L. Goossens30, P.A. Gorbounov96, H.A. Gordon25, I. Gorelov104, B. Gorini30, E. Gorini72a,72b, A. Goriˇsek74, E. Gornicki39, J. Goncalves Pinto Firmino Da Costa , L. Gonella , S. Gonzalez de la Hoz , G. Gonzalez Parra , S. Gonzalez Sevilla , L. Goossens30, P.A. Gorbounov96, H.A. Gordon25, I. Gorelov104, B. Gorini30, E. Gorini72a,72b, A. Goriˇsek74, E. Gornicki39, Goshaw , C. Gossling , M.I. Gostkin , M. Gouighri , D. Goujdami , M.P. Goulette , A.G. Goussio zpinar23, H.M.X. Grabas137, L. Graber54, I. Grabowska-Bold38a, P. Grafstr¨om20a,20b, K-J. Grahn42, J. Gram p g E. Gramstad118, S. Grancagnolo16, V. Grassi149, V. Gratchev122, H.M. Gray30, E. Graziani135a, O.G. Grebenyuk122, Z.D. Greenwood , , K. Gregersen , I.M. Gregor , P. Grenier , J. Grifﬁths , A.A. Grillo , K. , g , , , y , , , G. Moser100, M. Mosidze51b, J. Moss110, K. Motohashi158, R. Mount144, E. Mountricha25, S.V. Mouraviev T.G. McCarthy29, N.A. McCubbin130, K.W. McFarlane56,∗, J.A. Mcfayden77, G. Mchedlidze54, S.J. McMahon130, R.A. McPherson170,i, A. Meade85, J. Mechnich106, M. Medinnis42, S. Meehan31, S. Mehlhase99, A. Mehta73, K. Meier58a, C. Meineck99, B. Meirose80, C. Melachrinos31, B.R. Mellado Garcia146c, F. Meloni17, A. Mengarelli20a,20b, S. Menke100, E. Meoni162, K.M. Mercurio57, S. Mergelmeyer21, N. Meric137, P. Mermod49, L. Merola103a,103b, C. Meroni90a, F.S. Merritt31, H. Merritt110, A. Messina30,z, J. Metcalfe25, A.S. Mete164, C. Meyer82, C. Meyer121, J-P. Meyer137, J. Meyer30, R.P. Middleton130, S. Migas73, L. Mijovi´c21, G. Mikenberg173, M. Mikestikova126, M. Mikuˇz74, A. Milic30, D.W. Miller31, C. Mills46, A. Milov173, D.A. Milstead147a,147b, D. Milstein173, A.A. Minaenko129, I.A. Minashvili64, A.I. Mincer109, B. Mindur38a, M. Mineev64, Y. Ming174, L.M. Mir12, G. Mirabelli133a, T. Mitani172, J. Mitrevski99, V.A. Mitsou168, S. Mitsui65 A. Miucci49, P.S. Miyagawa140, J.U. Mj¨ornmark80, T. Moa147a,147b, K. Mochizuki84, S. Mohapatra35, W. Mohr48, S. Molander147a,147b, R. Moles-Valls168, K. M¨onig42, C. Monini55, J. Monk36, E. Monnier84, J. Montejo Berlingen12, F. Monticelli70, S. Monzani133a,133b, R.W. Moore3, A. Moraes53, N. Morange62, D. Moreno82, M. Moreno Ll´acer54, P. Morettini50a, M. Morgenstern44, M. Morii57, S. Moritz82, A.K. Morley148, G. Mornacchi30, J.D. Morris75, L. Morvaj102, H.G. Moser100, M. Mosidze51b, J. Moss110, K. Motohashi158, R. Mount144, E. Mountricha25, S.V. Mouraviev95,∗, V. Radescu42, S.K. Radhakrishnan149, P. Radloff115, P. Rados87, F. Ragusa90a,90b, G. Rahal179, S. Rajagopalan25, M. Rammensee30, A.S. Randle-Conde40, C. Rangel-Smith167, K. Rao164, F. Rauscher99, T.C. Rave48, T. Ravenscroft53, M. Raymond30, A.L. Read118, N.P. Readioff73, D.M. Rebuzzi120a,120b, A. Redelbach175, G. Redlinger25, R. Reece138, K. Reeves41, L. Rehnisch16, H. Reisin27, M. Relich164, C. Rembser30, H. Ren33a, Z.L. Ren152, A. Renaud116, M. Rescigno133a, S. Resconi90a, O.L. Rezanova108,t, P. Reznicek128, R. Rezvani94, R. Richter100, M. Ridel79, P. Rieck16, J. Rieger54, M. Rijssenbeek149, A. Rimoldi120a,120b, L. Rinaldi20a, E. Ritsch61, I. Riu12, F. Rizatdinova113, E. Rizvi75, S.H. Robertson86,i, A. Robichaud-Veronneau86, D. Robinson28, J.E.M. Robinson83, A. Robson53, C. Roda123a,123b, L. Rodrigues30, S. Roe30, O. Røhne118, S. Rolli162, A. Romaniouk97, M. Romano20a,20b, E. Romero Adam168, N. Rompotis139, M. Ronzani48, L. Roos79, E. Ros168, S. Rosati133a, K. Rosbach49, M. Rose76, P. Rose138, P.L. Rosendahl14, O. Rosenthal142, V. Rossetti147a,147b, E. Rossi103a,103b, L.P. Rossi50a, R. Rosten139, M. Rotaru26a, I. Roth173, J. Rothberg139, D. Rousseau116, C.R. Royon137, A. Rozanov84, Y. Rozen153, X. Ruan146c, F. Rubbo12, I. Rubinskiy42, V.I. Rud98, C. Rudolph44, M.S. Rudolph159, F. R¨uhr48, A. Ruiz-Martinez30, Z. Rurikova48, N.A. Rusakovich64, A. Ruschke99, J.P. Rutherfoord7, N. Ruthmann48, Y.F. Ryabov122, M. Rybar128, G. Rybkin116, N.C. Ryder119, A.F. Saavedra151, S. Sacerdoti27, A. Saddique3, I. Sadeh154, H.F-W. Sadrozinski138, R. Sadykov64, F. Safai Tehrani133a, H. Sakamoto156, Y. Sakurai172, G. Salamanna135a,135b, A. Salamon134a, M. Saleem112, D. Salek106, P.H. Sales De Bruin139, D. Salihagic100, A. Salnikov144, J. Salt168, D. Salvatore37a,37b, F. Salvatore150, A. Salvucci105, A. Salzburger30, D. Sampsonidis155, A. Sanchez103a,103b, J. S´anchez168, V. Sanchez Martinez168, H. Sandaker14, R.L. Sandbach75, H.G. Sander82, M.P. Sanders99, M. Sandhoff176, T. Sandoval28, C. Sandoval163, R. Sandstroem100, D.P.C. Sankey130, A. Sansoni47, C. Santoni34, R. Santonico134a,134b, H. Santos125a, I. Santoyo Castillo150, K. Sapp124, A. Sapronov64, J.G. Saraiva125a,125d, B. Sarrazin21, G. Sartisohn176, O. Sasaki65, Y. Sasaki156, G. Sauvage5,∗, , g , p , , , , 4, R.L. Sandbach75, H.G. Sander82, M.P. Sanders99, M. Sandhoff176, T. Sandoval28, C. Sandoval163, , , , , , , , , mann49, Y. Munwes154, J.A. Murillo Quijada18, W.J. Murray171,130, H. Musheghyan54, E. Musto153, The ATLAS Collaboration Grimm , S. Grinstein , , Ph. Gris34, Y.V. Grishkevich98, J.-F. Grivaz116, J.P. Grohs44, A. Grohsjean42, E. Gross173, J. Grosse-Knetter54, , g , g , , , , , , Ph. Gris34, Y.V. Grishkevich98, J.-F. Grivaz116, J.P. Grohs44, A. Grohsjean42, E. Gross173, J. Grosse-Knetter54, G. Halladjian89, K. Hamacher176, P. Hamal114, K. Hamano170, M. Hamer54, A. Hamilton146a, S. Hamilton162, G.N. Hamity146c, g J.D. Hansen36, P.H. Hansen36, K. Hara161, A.S. Hard174, T. Harenberg176, F. Hariri116, S. Harkusha91, D. Harper88, 42 p y p pp y E.V. Korolkova140, V.A. Korotkov129, O. Kortner100, S. Kortner100, V.V. Kostyukhin21, V.M. Kotov64, A. Kotwal45, , g , , , g , , , y , A. Kruse174, M.C. Kruse45, M. Kruskal22, T. Kubota87, S. Kuday4a, S. Kuehn48, A. Kugel58c, A. Kuhl138, T. Kuhl42, , , g , , , g , , M. Lungwitz82, D. Lynn25, R. Lysak126, E. Lytken80, H. Ma25, L.L. Ma33d, G. Maccarrone47, A. Macchiolo100, , , g , , , , , , A.A. Maier100, A. Maio125a,125b,125d, S. Majewski115, Y. Makida65, N. Makovec116, P. Mal137,y, B. Malaescu79, Pa. Malecki3 , , , , , , y , y , J. Mamuzic13b, B. Mandelli30, L. Mandelli90a, I. Mandi´c74, R. Mandrysch62, J. Maneira125a,125b, A. Manfredini100, j g Mansoulie137, R. Mantifel86, L. Mapelli30, L. March168, J.F. Marchand29, G. Marchiori79, M. Marcisovsky 70, M. Marjanovic13a, C.N. Marques125a, F. Marroquim24a, S.P. Marsden83, Z. Marshall15, L.F. Marti17, j q q cia168, B. Martin30, B. Martin89, T.A. Martin171, V.J. Martin46, B. Martin dit Latour14, H. Martinez137, M. Martinez12,n, S. Martin-Haugh130, A.C. Martyniuk77, M. Marx139, F. Marzano133a, A. Marzin30, L. Masetti82, 43 E.J.W. Moyse85, S. Muanza84, R.D. Mudd18, F. Mueller58a, J. Mueller124, K. Mueller21, T. Mueller28, T. Mueller82, y , , , , , , , D. Muenstermann49, Y. Munwes154, J.A. Murillo Quijada18, W.J. Murray171,130, H. Musheghyan54, E. Mu Q j y g y A.G. Myagkov129,aa, M. Myska127, O. Nackenhorst54, J. Nadal54, K. Nagai61, R. Nagai158, Y. Nagai84, K. Nagano65, g g g H. Namasivayam41, G. Nanava21, R. Narayan58b, T. Nattermann21, T. Naumann42, G. Navarro163, R. Nayyar7, H.A. Neal8 y , , y , , , , y P.Yu. Nechaeva95, T.J. Neep83, P.D. Nef144, A. Negri120a,120b, G. Negri30, M. Negrini20a, S. Nektarijevic49 p g g g j T.K. Nelson144, S. Nemecek126, P. Nemethy109, A.A. Nepomuceno24a, M. Nessi30,ab, M.S. Neubauer166, y p R.M. Neves109, P. Nevski25, P.R. Newman18, D.H. Nguyen6, R.B. Nickerson119, R. Nicolaidou137, B. Nicquevert30, Nikiforou35, A. Nikiforov16, V. Nikolaenko129,aa, I. M. Teixeira Dias Castanheira75, P. Teixeira-Dias76, K.K. Temming48, H. Ten Kate30, P.K. Teng152, J.J. Teoh117, S. Terada65, K. Terashi156, J. Terron81, S. Terzo100, M. Testa47, R.J. Teuscher159,i, J. Therhaag21, T. Theveneaux-Pelzer34, J.P. Thomas18, J. Thomas-Wilsker76, E.N. Thompson35, P.D. Thompson18, P.D. Thompson159, A.S. Thompson53, L.A. Thomsen36, E. Thomson121, M. Thomson28, W.M. Thong87, R.P. Thun88,∗, F. Tian35, M.J. Tibbetts15, V.O. Tikhomirov95,ag, Yu.A. Tikhonov108,t, S. Timoshenko97, E. Tiouchichine84, P. Tipton177, S. Tisserant84, T. Todorov5, S. Todorova-Nova128, B. Toggerson7, J. Tojo69, S. Tok´ar145a, K. Tokushuku65, K. Tollefson89, L. Tomlinson83, M. Tomoto102, L. Tompkins31, K. Toms104, N.D. Topilin64, E. Torrence115, H. Torres143, E. Torr´o Pastor168, J. Toth84,ah, F. Touchard84, D.R. Tovey140, H.L. Tran116, T. Trefzger175, L. Tremblet30, A. Tricoli30, I.M. Trigger160a, S. Trincaz-Duvoid79, M.F. Tripiana12, W. Trischuk159, B. Trocm´e55, C. Troncon90a, M. Trottier-McDonald143, M. Trovatelli135a,135b, P. True89, M. Trzebinski39, A. Trzupek39, C. Tsarouchas30, J.C-L. Tseng119, P.V. Tsiareshka91, D. Tsionou137, G. Tsipolitis10, N. Tsirintanis9, S. Tsiskaridze12, V. Tsiskaridze48, E.G. Tskhadadze51a, I.I. Tsukerman96, V. Tsulaia15, S. Tsuno65, D. Tsybychev149, A. Tudorache26a, V. Tudorache26a, A.N. Tuna121, S.A. Tupputi20a,20b, S. Turchikhin98,a f , D. Turecek127, I. Turk Cakir4d, R. Turra90a,90b, P.M. Tuts35, A. Tykhonov49, M. Tylmad147a,147b, M. Tyndel130, K. Uchida21, I. Ueda156, R. Ueno29, M. Ughetto84, M. Ugland14, M. Uhlenbrock21, F. Ukegawa161, G. Unal30, A. Undrus25, G. Unel164, F.C. Ungaro48, Y. Unno65 C. Unverdorben99, D. Urbaniec35, P. Urquijo87, G. Usai8, A. Usanova61, L. Vacavant84, V. Vacek127, B. Vachon86, N. Valencic106, S. Valentinetti20a,20b, A. Valero168, L. Valery34, S. Valkar128, E. Valladolid Gallego168, S. Vallecorsa49, J.A. Valls Ferrer168, W. Van Den Wollenberg106, P.C. Van Der Deijl106, R. van der Geer106, H. van der Graaf106, R. Van Der Leeuw106, D. van der Ster30, N. van Eldik30, P. van Gemmeren6, J. Van Nieuwkoop143, I. van Vulpen106, M.C. van Woerden30, M. Vanadia133a,133b, W. Vandelli30, R. Vanguri121, A. Vaniachine6, P. Vankov42, F. Vannucci79, The ATLAS Collaboration N , , , , , , p , P. Nilsson8, Y. Ninomiya156, A. Nisati133a, R. Nisius100, T. Nobe158, L. Nodulman6, M. Nomachi117, I. Nomidis29, y S. Norberg112, M. Nordberg30, O. Novgorodova44, S. Nowak100, M. Nozaki65, L. Nozka114, K. Ntekas10, G. Nunes Hanninger87, T. Nunnemann99, E. Nurse77, F. Nuti87, B.J. O’Brien46, F. O’grady7, D.C. O’Neil g , , , , , g y , , , F.G. Oakham29,d, H. Oberlack100, T. Obermann21, J. Ocariz79, A. Ochi66, M.I. Ochoa77, S. Oda69, S. Odaka65, H. Ogren60, 83 45 15 167 117 25 31 156 26 , , , , , , , y , , A.G. Olchevski64, S.A. Olivares Pino46, D. Oliveira Damazio25, E. Oliver Garcia168, A. Olszewski39, J. Olszowska39, 5 5 5 5 5 , , , g , , g , g , E. Paganis140, C. Pahl100, F. Paige25, P. Pais85, K. Pajchel118, G. Palacino160b, S. Palestini30, M. Palka38b, D. Pallin34, S. Panitkin25, D. Pantea26a, L. Paolozzi134a,134b, Th.D. Papadopoulou10, K. Papageorgiou155,l, A. Paramonov6, T. Pauly30, J. Pearce170, M. Pedersen118, S. Pedraza Lopez168, R. Pedro125a,125b, S.V. Peleganchuk108, D. Pelikan167, H. Peng , B. Penning , J. Penwell , D.V. Perepelitsa , E. Perez Codina , M.T. Perez Garcıa Estan , V. Perez Reale L. Perini90a,90b, H. Pernegger30, R. Perrino72a, R. Peschke42, V.D. Peshekhonov64, K. Peters30, R.F.Y. Peters83, B.A. Petersen D.T. Pignotti110, J.E. Pilcher31, A.D. Pilkington77, J. Pina125a,125b,125d, M. Pinamonti165a,165c,ac, A. Pinder119, J.L. Pinfold3, 36 125 79 173 90 90b 145 25 128 64 g P. Plucinski147a,147b, S. Poddar58a, F. Podlyski34, R. Poettgen82, L. Poggioli116, D. Pohl21, M. Pohl49, G. Polesello120a, B.G. Pope89, G.A. Popeneciu26b, D.S. Popovic13a, A. Poppleton30, X. Portell Bueso12, S. Pospisil127, K. P 64 150 115 30 60 64 84 p p p pp p I.N. Potrap64, C.J. Potter150, C.T. Potter115, G. Poulard30, J. Poveda60, V. Pozdnyakov64, P. Pralavorio84, A. Pranko15, , , , , , , , , , K. Prokoﬁev47, F. Prokoshin32b, E. Protopapadaki137, S. Protopopescu25, J. Proudfoot6, M. Przybycien38a, H. Przysiezniak5, 5 5 5 5 5 5 Q Q A. Quadt54, D.R. Quarrie15, W.B. Quayle165a,165b, M. Queitsch-Maitland83, D. Quilty53, A. Qureshi160b, V g M. Raymond30, A.L. Read118, N.P. Readioff73, D.M. Rebuzzi120a,120b, A. Redelbach175, G. Redlinger25, R. Reece138, 44 E. Sauvan5, P. Savard159,d, D.O. Savu30, C. Sawyer119, L. Sawyer78,m, D.H. Saxon53, J. Saxon121, C. Sbarra20a, A. Sbrizzi3, R. Schaefer42, S. Schaepe21, S. Schaetzel58b, U. Sch¨afer82, A.C. Schaffer116, D. Schaile99, R.D. Schamberger149, V. Scharf58a, R. Schaefer42, S. P.M. Watkins18, A.T. Watson18, I.J. Watson151, M.F. Watson18, G. Watts139, S. Watts83, B.M. Waugh77, S. Webb83, M.S. Weber17, S.W. Weber175, J.S. Webster31, A.R. Weidberg119, P. Weigell100, B. Weinert60, J. Weingarten54, C. Weiser48, H. Weits106, P.S. Wells30, T. Wenaus25, D. Wendland16, Z. Weng152,ae, T. Wengler30, S. Wenig30, N. Wermes21, M. Werner48, P. Werner30, M. Wessels58a, J. Wetter162, K. Whalen29, A. White8, M.J. White1, R. White32b, S. White123a,123b, D. Whiteson164 D. Wicke176, F.J. Wickens130, W. Wiedenmann174, M. Wielers130, P. Wienemann21, C. Wiglesworth36, L.A.M. Wiik-Fuchs21, P.A. Wijeratne77, A. Wildauer100, M.A. Wildt42,ak, H.G. Wilkens30, J.Z. Will99, H.H. Williams121, S. Williams28, C. Willis89, S. Willocq85, A. Wilson88, J.A. Wilson18, I. Wingerter-Seez5, F. Winklmeier115, B.T. Winter21, M. Wittgen144, T. Wittig43, J. Wittkowski99, S.J. Wollstadt82, M.W. Wolter39, H. Wolters125a,125c, B.K. Wosiek39, J. Wotschack30, M.J. Woudstra83, K.W. Wozniak39, M. Wright53, M. Wu55, S.L. Wu174, X. Wu49, Y. Wu88, E. Wulf35, T.R. Wyatt83, B.M. Wynne46, S. Xella36, M. Xiao137, D. Xu33a, L. Xu33b,al, B. Yabsley151, S. Yacoob146b,am, R. Yakabe66, M. Yamada65, H. Yamaguchi156, Y. Yamaguchi117, A. Yamamoto65, K. Yamamoto63, S. Yamamoto156, T. Yamamura156, T. Yamanaka156, K. Yamauchi102, Y. Yamazaki66, Z. Yan22, H. Yang33e, H. Yang174, U.K. Yang83, Y. Yang110, S. Yanush92, L. Yao33a, W-M. Yao15, Y. Yasu65, E. Yatsenko42, K.H. Yau Wong21, J. Ye40, S. Ye25, A.L. Yen57, E. Yildirim42, M. Yilmaz4b, R. Yoosoofmiya124, K. Yorita172, R. Yoshida6, K. Yoshihara156, C. Young144, C.J.S. Young30, S. Youssef22, D.R. Yu15, J. Yu8, J.M. Yu88, J. Yu113, L. Yuan66, A. Yurkewicz107, I. Yusuff28,an, B. Zabinski39, R. Zaidan62, A.M. Zaitsev129,aa, A. Zaman149, S. Zambito23, L. Zanello133a,133b D. Zanzi100, C. Zeitnitz176, M. Zeman127, A. Zemla38a, K. Zengel23, O. Zenin129, T. ˇZeniˇs145a, D. Zerwas116, G. Zevi della Porta57, D. Zhang88, F. Zhang174, H. Zhang89, J. Zhang6, L. Zhang152, X. Zhang33d, Z. Zhang116, Z. Zhao33b, A. Zhemchugov64, J. Zhong119, B. Zhou88, L. Zhou35, N. Zhou164, C.G. Zhu33d, H. Zhu33a, J. Zhu88, Y. Zhu33b, X. Zhuang33a, K. Zhukov95, A. Zibell175, D. Zieminska60, N.I. Zimine64, C. Zimmermann82, R. Zimmermann21, S. Zimmermann21, S. Zimmermann48, Z. Zinonos54, M. Ziolkowski142, G. Zobernig174, A. Zoccoli20a,20b, M. zur Nedden16, G. Zurzolo103a,103b, V Zutshi107 L Zwalinski30 y , , , , , p , , quez Schroeder54, J. Veatch7, F. Veloso125a,125c, S. Veneziano133a, A. Ventura72a,72b, D. Ventura85, M. Ven nyan178, R. Vari133a, E.W. Varnes7, T. Varol85, D. Varouchas79, A. Vartapetian8, K.E. Varvell151, F. Vazeille The ATLAS Collaboration Schaepe21, S. Schaetzel58b, U. Sch¨afer82, A.C. Schaffer116, D. Schaile99, R.D. Schamberger149, V. Scharf58a VA Schegelsky122 D Scheirich128 M Schernau164 M I Scherzer35 C Schiavi50a,50b J Schieck99 C Schillo48 g y M. Schioppa37a,37b, S. Schlenker30, E. Schmidt48, K. Schmieden30, C. Schmitt82, S. Schmitt58b, B. Schneider17, y H.-C. Schultz-Coulon58a, H. Schulz16, M. Schumacher48, B.A. Schumm138, Ph. Schune137, C. Schwanenberger83, , g , g , , g , , M. Schwoerer5, F.G. Sciacca17, E. Scifo116, G. Sciolla23, W.G. Scott130, F. Scuri123a,123b, F. Scutti21, J. Searcy88, G. Sedov42, , , p , , , , , , H. Severini112, T. Sﬁligoj74, F. Sforza100, A. Sfyrla30, E. Shabalina54, M. Shamim115, L.Y. Shan33a, R. Shang166, J.T. Shank22, p , , , , , , , , M. Shimojima101, M. Shiyakova64, A. Shmeleva95, M.J. Shochet31, D. Short119, S. Shrestha63, E. Shulga97, M.A. Shupe7, , , p , p , p , , , , , A.M. Soukharev108, D. South42, S. Spagnolo72a,72b, F. Span`o76, W.R. Spearman57, F. Spettel100, R. Spighi20a, G. Spigo30, S. Stamm16, E. Stanecka39, R.W. Stanek6, C. Stanescu135a, M. Stanescu-Bellu42, M.M. Stanitzki42, S. Stapnes118, g G. Stoicea26a, P. Stolte54, S. Stonjek100, A.R. Stradling8, A. Straessner44, M.E. Stramaglia17, J. Strandberg148, g R. Stroynowski40, S.A. Stucci17, B. Stugu14, N.A. Styles42, D. Su144, J. Su124, R. Subramaniam78, A. Succurro12, Y. Sugaya117, 10 12 95 4 6 5 33 48 140 y y J. Tanaka156, R. Tanaka116, S. Tanaka132, S. Tanaka65, A.J. Tanasijczuk143, B.B. Tannenwald110, N. Tannoury21, 5 45 C. Wa otaya oj , . Wa bu to , C. . Wa d , . . Wa d ope , . Wa s s y , . Was b oo , C. Was c , P.M. Watkins18, A.T. Watson18, I.J. Watson151, M.F. Watson18, G. Watts139, S. Watts83, B.M. Waugh77, S. Webb83, M.S. Weber17, S.W. Weber175, J.S. Webster31, A.R. Weidberg119, P. Weigell100, B. Weinert60, J. Weingarten54, C. Weiser48, H. Weits106, P.S. Wells30, T. Wenaus25, D. Wendland16, Z. Weng152,ae, T. Wengler30, S. Wenig30, N. Wermes21, M. Werner48, P. Werner30, M. Wessels58a, J. Wetter162, K. Whalen29, A. White8, M.J. White1, R. White32b, S. White123a,123b, D. Whiteson164, D. Wicke176, F.J. Wickens130, W. Wiedenmann174, M. Wielers130, P. Wienemann21, C. Wiglesworth36, L.A.M. Wiik-Fuchs21, P.A. Wijeratne77, A. Wildauer100, M.A. Wildt42,ak, H.G. Wilkens30, J.Z. Will99, H.H. Williams121, S. Williams28, C. Willis89, S. Willocq85, A. Wilson88, J.A. Wilson18, I. Wingerter-Seez5, F. Winklmeier115, B.T. Winter21, M. Wittgen144, T. Wittig43, J. Wittkowski99, S.J. Wollstadt82, M.W. Wolter39, H. Wolters125a,125c, B.K. Wosiek39, J. The ATLAS Collaboration Wotschack30, M.J. Woudstra83, K.W. Wozniak39, M. Wright53, M. Wu55, S.L. Wu174, X. Wu49, Y. Wu88, E. Wulf35, T.R. Wyatt83, B.M. Wynne46, S. Xella36, M. Xiao137, D. Xu33a, L. Xu33b,al, B. Yabsley151, S. Yacoob146b,am, R. Yakabe66, M. Yamada65, H. Yamaguchi156, Y. Yamaguchi117, A. Yamamoto65, K. Yamamoto63, S. Yamamoto156, T. Yamamura156, T. Yamanaka156, K. Yamauchi102, Y. Yamazaki66, Z. Yan22, H. Yang33e, H. Yang174, U.K. Yang83, Y. Yang110, S. Yanush92, L. Yao33a, W-M. Yao15, Y. Yasu65, E. Yatsenko42, K.H. Yau Wong21, J. Ye40, S. Ye25, A.L. Yen57, E. Yildirim42, M. Yilmaz4b, R. Yoosoofmiya124, K. Yorita172, R. Yoshida6, K. Yoshihara156, C. Young144, C.J.S. Young30, S. Youssef22, D.R. Yu15, J. Yu8, J.M. Yu88, J. Yu113, L. Yuan66, A. Yurkewicz107, I. Yusuff28,an, B. Zabinski39, R. Zaidan62, A.M. Zaitsev129,aa, A. Zaman149, S. Zambito23, L. Zanello133a,133b, D. Zanzi100, C. Zeitnitz176, M. Zeman127, A. Zemla38a, K. Zengel23, O. Zenin129, T. ˇZeniˇs145a, D. Zerwas116, G. Zevi della Porta57, D. Zhang88, F. Zhang174, H. Zhang89, J. Zhang6, L. Zhang152, X. Zhang33d, Z. Zhang116, Z. Zhao33b, A. Zhemchugov64, J. Zhong119, B. Zhou88, L. Zhou35, N. Zhou164, C.G. Zhu33d, H. Zhu33a, J. Zhu88, Y. Zhu33b, X. Zhuang33a, K. Zhukov95, A. Zibell175, D. Zieminska60, N.I. Zimine64, C. Zimmermann82, R. Zimmermann21, S. Zimmermann21, S. Zimmermann48, Z. Zinonos54, M. Ziolkowski142, G. Zobernig174, A. Zoccoli20a,20b, M. zur Nedden16, G. Zurzolo103a,103b, V. Zutshi107, L. Zwalinski30. V. Zutshi107, L. Zwalinski30. The ATLAS Collaboration 1 Department of Physics, University of Adelaide, Adelaide, Australia 1 Department of Physics, University of Adelaide, Adelaide, Australia 2 Physics Department, SUNY Albany, Albany NY, United States of America 3 3 Department of Physics, University of Alberta, Edmonton AB, Canada 4 (a) Department of Physics, Ankara University, Ankara; (b) Department of Physics, Gazi University, Ankara; (c) Division of Physics, TOBB University of Economics and Technology, Ankara; (d) Turkish Atomic Energy Authority, Ankara, Turkey y y gy 5 LAPP, CNRS/IN2P3 and Universit´e de Savoie, Annecy-le-Vieux, France 6 6 High Energy Physics Division, Argonne National Laboratory, Argonne IL, United States of America 7 Department of Physics, University of Arizona, Tucson AZ, United States of America 8 epartment of Physics, University of Arizona, Tucson 8 Department of Physics, The University of Texas at Arlington, Arlington TX, United States of America 9 9 Physics Department, University of Athens, Athens, Greece 10 Physics Department, National Technical University of 11 Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan 12 Institut de F´ısica d’Altes Energies and Departament de F´ısica de la Univer 13 ( ) (b) 12 Institut de F´ısica d’Altes Energies and Departament de F´ısica de la Universitat Aut`onoma de Barcelona, Barcelona, Spain 13 (a) Institute of Physics, University of Belgrade, Belgrade; (b) Vinca Institute of Nuclear Sciences, University of Belgrade, Belgrade, Serbia 14 Department for Physics and Technology, University of Bergen, Bergen, Norway 15 15 Physics Division, Lawrence Berkeley National Laboratory and University of California, Berkeley CA, United States of America 16 16 Department of Physics, Humboldt University, Berlin, Germany 17 p y y y 17 Albert Einstein Center for Fundamental Physics and Laboratory for High Energy Physics, University of Bern, Bern, Switzerland stein Center for Fundamental Physics and Laboratory for High Energy Physics, University of Bern, Bern, 18 School of Physics and Astronomy, University of Birmingham, Birmi 19 ( ) (b) 18 School of Physics and Astronomy, University of Birmingham, Birmingham, United Kingdom ment of Physics, Bogazici University, Istanbul; (b) Department of Physics, Dogus University, Istanbul; (c) 19 (a) Department of Physics, Bogazici University, Istanbul; (b) Department of Physics, Dogus Univ tment of Physics Engineering, Gaziantep University, Gaziantep, Turkey Department of Physics Engineering, Gaziantep University, Gaziantep, Turkey 46 20 (a) INFN Sezione di Bologna; (b) Dipartimento di Fisica e Astronomia, Universit`a di Bologna, Bologna, Italy (a) INFN Sezione di Bologna; (b) Dipartimento di Fisica e Astronomia, Universit`a di Bologna, Bologna, It 20 (a) INFN Sezione di Bologna; (b) Dipartimento di Fisica e Astronomia, Universit`a di Bologna, Bologna 21 21 Physikalisches Institut, University of Bonn, Bonn, Germany y y y Department of Physics, Boston University, Boston MA, United States of America 22 Department of Physics, Boston University, Boston MA, United States of America p y y 24 (a) Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b) Federal University of Juiz de Fora (UFJF), 24 (a) Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b) Federal ( ) ( ) 24 (a) Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b) Federal University of Juiz de Fora (UFJF), ( ) (d) 25 Physics Department, Brookhaven National Laboratory, Upton NY, United States of America 26 (a) N i l I i f Ph i d N l E i i B h (b) N i l I i f R Physics Department, Brookhaven National Laboratory, Upton NY, United States of America ( ) (b) 25 Physics Department, Brookhaven National Laboratory, Upton NY, United States of America 26 (a) National Institute of Physics and Nuclear Engineering, Bucharest; (b) National Institute for Research and Development of ( ) (d) 25 Physics Department, Brookhaven National Laboratory, Upton NY, United States of America 26 (a) National Institute of Physics and Nuclear Engineering, Bucharest; (b) National Institute for Research and Development o Physics Department, Brookhaven National Laboratory, Upton NY, United States of America 26 (a) National Institute of Physics and Nuclear Engineering, Bucharest; (b) National Institute for Research and Development of Isotopic and Molecular Technologies, Physics Department, Cluj Napoca; (c) University Politehnica Bucharest, Bucharest; (d) y p , y, p , 26 (a) National Institute of Physics and Nuclear Engineering, Bucharest; (b) National Institute for Research and Development of Isotopic and Molecular Technologies, Physics Department, Cluj Napoca; (c) University Politehnica Bucharest, Bucharest; (d) Isotopic and Molecular Technologies, Physics Depar West University in Timisoara, Timisoara, Romania 27 Departamento de F´ısica, Universidad de Buenos Aires, Buenos Aires, Argentina 28 27 Departamento de F´ısica, Universidad de Buenos Aires, Buenos Aires, Argentina 27 Departamento de F´ısica, Universidad de Buenos A 28 Cavendish Laboratory, University of Cambridge, Cambridge, United Kingdom 29 28 Cavendish Laboratory, University of Cambridge, Cambridge, United Kingdom 29 29 Department of Physics, Carleton University, Ottawa ON, Canada 29 Department of Physics, Carleton University, Ottawa ON, Canada 30 CERN, Geneva, Switzerland 31 Enrico Fermi Institute, University of Chicago, Chicago IL, United States of America 32 ( ) (b) 31 Enrico Fermi Institute, University of Chicago, Chicago IL, United States of America 32 ( ) (b) 32 (a) Departamento de F´ısica, Pontiﬁcia Universidad Cat´olica de Chile, Santiago; (b) Departamento de F´ısica, Universidad T´ecnica Federico Santa Mar´ıa, Valpara´ıso, Chile 32 (a) Departamento de F´ısica, Pontiﬁcia Universidad Cat´olica de Chile, Santiago; (b) De 32 (a) Departamento de F´ısica, Pontiﬁcia Universi 32 (a) Departamento de F´ısica, Pontiﬁcia Universidad Cat´olica de Chile, Santiago; (b) Departamento d T´ecnica Federico Santa Mar´ıa, Valpara´ıso, Chile ( ) 33 (a) Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b) Department of Modern Physics, University of Science and Technology of China, Anhui; (c) Department of Physics, Nanjing University, Jiangsu; (d) School of Physics, 33 (a) Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b) Department of Modern of Science and Technology of China Anhui; (c) Department of Physics Nanjing University Jiangsu; (d) S 33 (a) Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b) Department of Modern Physics, University of Science and Technology of China, Anhui; (c) Department of Physics, Nanjing University, Jiangsu; (d) School of Physics, Shandong University, Shandong; (e) Physics Department, Shanghai Jiao Tong University, Shanghai, China 34 Laboratoire de Physique Corpusculaire, Clermont Universit´e and Universit´e Blaise Pascal and CNRS/IN2P3, Clermont-Ferrand, France Shandong University, Shandong; (e) Physics Department, Shanghai Jiao Tong University, Shanghai, China 34 Laboratoire de Physique Corpusculaire, Clermont Universit´e and Universit´e Blaise Pascal and CNRS/IN2P3, Clermont-Ferrand, France Shandong University, Shandong; Physics Department, Shanghai Jiao Tong University, Shanghai, Chin 34 Laboratoire de Physique Corpusculaire, Clermont Universit´e and Universit´e Blaise Pascal and CNRS/I Clermont-Ferrand, France 35 Nevis Laboratory, Columbia University, Irvington NY, United States of America 36 35 Nevis Laboratory, Columbia University, Irvington NY, United States of America 36 36 Niels Bohr Institute, University of Copenhagen, Kobenhavn, Denmark ( ) 36 Niels Bohr Institute, University of Copenhagen, Kobenhavn, Denmark 37 ( ) 37 (a) INFN Gruppo Collegato di Cosenza, Laboratori Nazionali di Frascati; (b) Dipartimento di Fisica, Universit`a della Calabria, Rende, Italy ( ) ( ) 37 (a) INFN Gruppo Collegato di Cosenza, Laboratori Nazionali di Frascati; (b) Dipartimento di Fisica, Universit`a della Calabria, Rende, Italy 38 ( ) (b) (a) AGH University of Science and Technology, Faculty of Physics and Applied Computer Science, Krako 38 (a) AGH University of Science and Technology, Faculty of Physics and Applied Computer Science, moluchowski Institute of Physics, Jagiellonian University, Krakow, Poland Smoluchowski Institute of Physics, Jagiellonian University, Krakow, Poland 39 The Henryk Niewodniczanski Institute of Nuclear Physics, Polish Academy of Sciences, Krak 39 The Henryk Niewodniczanski Institute of Nuclear Physics, Polish Academy of Sciences, Krakow, Poland 40 Physics Department, Southern Methodist University, Dallas TX, United States of America 41 40 Physics Department, Southern Methodist University, Dallas TX, United States of America 41 Ph i D t t U i it f T t D ll Ri h d TX U it d St t f A i y p y 41 Physics Department, University of Texas at Dallas, Richardson TX, United States of America 42 41 Physics Department, University of Texas at Dallas, Richardson TX, United States of America 42 rtment, University of Texas at Dallas, Richardson TX 42 DESY, Hamburg and Zeuthen, Germany Institut f¨ur Experimentelle Physik IV, Technische Universit¨at Dortmund, Dortmund, Germany 43 Institut f¨ur Experimentelle Physik IV, Technische Universit¨at Dortmund, Dortmund, Germany 43 Institut f¨ur Experimentelle Physik IV, Technisc p y 44 Institut f¨ur Kern- und Teilchenphysik, Technische Universit¨at Dresden, Dresden, Germany 44 Institut f¨ur Kern- und Teilchenphysik, Technische Universit¨at Dresden, Dresden, Germany 45 Department of Physics, Duke University, Durham NC, United States of America 45 Department of Physics, Duke University, Durham NC, United States of America p y , y, , 46 SUPA - School of Physics and Astronomy, University of Edinburgh, Edinburgh, United Kingdom 47 p y y 46 SUPA - School of Physics and Astronomy, University of Edinburgh, Edin 46 SUPA - School of Physics and Astronomy, University of Edinburgh, Edinburgh, United Kingdom 47 UPA - School of Physics and Astronomy, University of 47 INFN Laboratori Nazionali di Frascati, Frascati, Italy y 48 Fakult¨at f¨ur Mathematik und Physik, Albert-Ludwigs-Universit¨at, Freiburg, Germany 48 Fakult¨at f¨ur Mathematik und Physik, Albert-Ludwigs-Universit¨at, Freiburg, Germany Fakult¨at f¨ur Mathematik und Physik, Albert-Ludwig 49 Section de Physique, Universit´e de Gen`eve, Geneva, Switzerland ( ) ( ) 50 (a) INFN Sezione di Genova; (b) Dipartimento di Fisica, Universit`a di Genova, Genova, Italy 51 ( ) (b ; p , , , y 51 (a) E. The ATLAS Collaboration Andronikashvili Institute of Physics, Iv. Javakhishvili Tbilisi State University, Tbilisi; (b) High Energy Phys 51 (a) E. Andronikashvili Institute of Physics, Iv. Javakhishvili Tbilisi State University, Tbilisi; (b) High Energy Physics 51 (a) E. Andronikashvili Institute of Physics, Iv. The ATLAS Collaboration Stepanov Institute of Physics, National Academy of Sciences of Belarus, Minsk, Repub 91 B.I. Stepanov Institute of Physics, National Academy of Sciences of Belarus, Minsk, Republic of Bela National Scientiﬁc and Educational Centre for Particle and High Energy Physics, Minsk, Republic of Bela 92 National Scientiﬁc and Educational Centre for Particle and High Energy Physics, Minsk, Republic 93 Department of Physics, Massachusetts Institute of Technology, Cambridge MA, United S 93 Department of Physics, Massachusetts Institute of Technology, Cambridge MA, United States of America 95 P.N. Lebedev Institute of Physics, Academy of Sciences, Moscow, Russia 96 95 P.N. The ATLAS Collaboration Lebedev Institute of Physics, Academy of Sciences, Moscow, Russia 96 y y 96 Institute for Theoretical and Experimental Physics (ITEP), Moscow, Russia 96 Institute for Theoretical and Experimental Physics (ITEP), Moscow, Russia 97 Moscow Engineering and Physics Institute (MEPhI), Moscow, Russia oscow Engineering and Physics Institute (MEPhI), Mo 97 Moscow Engineering and Physics Institute (MEPhI), Moscow, Russia 98 D.V.Skobeltsyn Institute of Nuclear Physics, M.V.Lomonosov Moscow State University, Moscow, Rus 98 D.V.Skobeltsyn Institute of Nuclear Physics, M.V.Lomonosov Moscow State University 99 Fakult¨at f¨ur Physik, Ludwig-Maximilians-Universit¨at M¨unchen, M¨unchen, Germany 99 Fakult¨at f¨ur Physik, Ludwig-Maximilians-Universit¨at M¨unchen, M¨unchen, Germany y g y 100 Max-Planck-Institut f¨ur Physik (Werner-Heisenberg-Institut), M¨unchen, Germany 100 Max-Planck-Institut f¨ur Physik (Werner-Heisenberg-Institut), M¨unchen, Germany Max-Planck-Institut f¨ur Physik (Werner-Heisenberg-In 101 Nagasaki Institute of Applied Science, Nagasaki, Japan g pp g p 102 Graduate School of Science and Kobayashi-Maskawa Institute, Nagoya University, Nagoya, Japan 103 ( ) (b) 102 Graduate School of Science and Kobayashi-Maskawa Institute, Nagoya University, Nagoya ( ) ( ) duate School of Science and Kobayashi-Maskawa In 103 (a) INFN Sezione di Napoli; (b) Dipartimento di Fisica, Universit`a di Napoli, Napoli, Italy 103 (a) INFN Sezione di Napoli; (b) Dipartimento di Fisica, Universit`a di Napoli, Napoli, Italy INFN Sezione di Napoli; (b) Dipartimento di Fisica, U 104 Department of Physics and Astronomy, University of New Mexico, Albuquerque NM, Unit 104 Department of Physics and Astronomy, University of New Mexico, Albuquerque NM, United States o 105 Institute for Mathematics, Astrophysics and Particle Physics, Radboud University Nijmegen/Nikhef, Nijmeg 105 Institute for Mathematics, Astrophysics and Particle Physics, Radboud University Nijmegen/Nikhef, Nijmegen, Netherlan 107 Department of Physics, Northern Illinois University, DeKalb IL, United States of America 107 Department of Physics, Northern Illinois University, DeKalb IL, Unite 108 Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia 108 Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia 109 Department of Physics, New York University, New York NY, United States of America 109 Department of Physics, New York University, New York NY, United States of Ame 110 Ohio State University, Columbus OH, United States of America 110 Ohio State University, Columbus OH, United States of Americ ulty of Science, Okayama University, Okayama, Japa 112 Homer L. Dodge Department of Physics and Astronomy, University of Oklahoma, Norman OK, Unite 113 Department of Physics and Astronomy, University of Oklahoma, Norman OK, United States of America 112 Homer L. The ATLAS Collaboration Javakhishvili Tbilisi State University, Tbilisi; (b) High Energy Physics I tit t Tbili i St t U i it Tbili i G i Institute, Tbilisi State University, Tbilisi, Georgia y g 52 II Physikalisches Institut, Justus-Liebig-Universit¨at Giessen, Giessen, Germany 5 52 II Physikalisches Institut, Justus-Liebig-Universit¨at Giessen, Giessen, Germany 52 II Physikalisches Institut, Justus-Liebig-Univers 53 SUPA - School of Physics and Astronomy, University of Glasgow, Glasgow, United Kingdom 54 53 SUPA - School of Physics and Astronomy, University of Glasgow, Glasgow, U 54 II Physikalisches Institut, Georg-August-Universit¨at, G¨ottingen, Germany 54 II Physikalisches Institut, Georg-August-Universit¨at, G¨ottingen, Germany 55 Laboratoire de Physique Subatomique et de Cosmologie, Universit´e Grenoble-Alpes, CNRS/IN2P3, Grenoble, Fra 55 Laboratoire de Physique Subatomique et de Cosmologie, Universit´e Grenoble-Alpes, CNRS/IN2P3, Grenoble, France 55 Laboratoire de Physique Subatomique et de Cosmologie, Universit´e Grenoble-Alpes Department of Physics, Hampton University, Hampt 57 Laboratory for Particle Physics and Cosmology, Harvard University, Cambridge MA, United States of America 58 ( ) (b) 57 Laboratory for Particle Physics and Cosmology, Harvard University, Cambridge MA, U ( ) ( ( ) Kirchhoff-Institut fur Physik, Ruprecht-Karls-Universitat Heidelberg, Heidelberg; ( ) Physikalische Ruprecht-Karls-Universit¨at Heidelberg, Heidelberg; (c) ZITI Institut f¨ur technische Informatik, Ruprecht- Kirchhoff Institut fur Physik, Ruprecht Karls Universitat Heidelberg, Heidelberg; Physikalisches Institut, Ruprecht-Karls-Universit¨at Heidelberg, Heidelberg; (c) ZITI Institut f¨ur technische Informatik, Ruprecht-Karls-Universit¨at Heidelberg, Mannheim, Germany Heidelberg, Mannheim, Germany 60 Department of Physics, Indiana University, Bloomington IN, United States of America 61 60 Department of Physics, Indiana University, Bloomington IN, United States of America 61 Institut f¨ur Astro- und Teilchenphysik, Leopold-Franzens-Universit¨at, Innsbruck, Austria 61 Institut f¨ur Astro- und Teilchenphysik, Leopold-Franzens-Universit¨at, Innsbruck, Austri 62 University of Iowa, Iowa City IA, United States of America 62 University of Iowa, Iowa City IA, United States of America 63 Department of Physics and Astronomy, Iowa State University, Ames IA, United States of America 64 63 Department of Physics and Astronomy, Iowa State University, Ames IA, United States of Amer 64 63 Department of Physics and Astronomy, Iowa State University, Ames I 64 Joint Institute for Nuclear Research, JINR Dubna, Dubna, Russia 47 65 KEK, High Energy Accelerator Research Organization, Tsukuba, Japan 65 KEK, High Energy Accelerator Research Organization, Tsukuba, Japan g gy g 66 Graduate School of Science, Kobe University, Kobe, Japan g gy g 66 Graduate School of Science, Kobe University, Kobe, Japan 67 Faculty of Science, Kyoto University, Kyoto, Japan 68 Kyoto University of Education, Kyoto, Japan y y y p 69 Department of Physics, Kyushu University, Fukuoka, Japan y y y p 69 Department of Physics, Kyushu University, Fukuoka, Japan 69 Department of Physics, Kyushu University, Fukuoka, Japan 69 Department of Physics, Kyushu University, Fuk p y y y p 70 Instituto de F´ısica La Plata, Universidad Nacional de La Plata and CONICET, La Plata, Argentina p y y y p 70 Instituto de F´ısica La Plata, Universidad Nacional de La Plata and CONICET, La Plata, Argenti 70 Instituto de F´ısica La Plata, Universidad Nacional de La Plata and C 71 Physics Department, Lancaster University, Lancaster, United Kingdom 71 Physics Department, Lancaster University, Lancaster, United Kingdom ( ) ( ) ysics Department, Lancaster University, Lancaster, U (a) INFN Sezione di Lecce; (b) Dipartimento di Matematica e Fisica, Universit`a del Salento, Lecce, Italy 72 (a) INFN Sezione di Lecce; (b) Dipartimento di Matematica e Fisica, Universit`a del Salento, Le 73 Oliver Lodge Laboratory, University of Liverpool, Liverpool, United Kingdom 74 g y y p p g 74 Department of Physics, Joˇzef Stefan Institute and University of Ljubljana, Ljubljana, Slov 74 Department of Physics, Joˇzef Stefan Institute and University of Ljubljana, Ljubljana, Slovenia School of Physics and Astronomy, Queen Mary University of London, London, United Kingdom 75 School of Physics and Astronomy, Queen Mary University of London, London, United Kingdo 76 Department of Physics, Royal Holloway University of London, Surrey, United Kingdom 76 Department of Physics, Royal Holloway University of London, Surrey, United Kingdom 77 Department of Physics and Astronomy, University College London, London, United Kingdom 78 77 Department of Physics and Astronomy, University College London, London, United Kingd 78 78 Louisiana Tech University, Ruston LA, United States of America 78 Louisiana Tech University, Ruston LA, United States of America 79 Laboratoire de Physique Nucl´eaire et de Hautes Energies, UPMC and Universit´e Paris-Diderot and CNRS/IN2P3, Paris France 80 79 Laboratoire de Physique Nucl´eaire et de Hautes Energies, UPMC and Universit´e Paris-Diderot and CNRS/IN2P3, Paris, France 80 80 Fysiska institutionen, Lunds universitet, Lund, Sweden 80 Fysiska institutionen, Lunds universitet, Lund, Sweden 81 81 Departamento de Fisica Teorica C-15, Universidad Autonoma de Madrid, Madrid, Spain 82 81 Departamento de Fisica Teorica C-15, Universidad Autonoma de Madrid, Madr 82 Institut f¨ur Physik, Universit¨at Mainz, Mainz, Germany 82 Institut f¨ur Physik, Universit¨at Mainz, Mainz, Germany 83 School of Physics and Astronomy, University of Manchester, Manchester, United Kingdom 83 School of Physics and Astronomy, University of Manchester, M 85 Department of Physics, University of Massachusetts, Amherst MA, United States of America 85 Department of Physics, University of Massachusetts, Amherst MA, Un 86 Department of Physics, McGill University, Montreal QC, Canada 86 Department of Physics, McGill University, Montreal QC, Canada 87 School of Physics, University of Melbourne, Victoria, Australia 87 School of Physics, University of Melbourne, Victoria, Australia 88 Department of Physics, The University of Michigan, Ann Arbor MI, United States of America 88 Department of Physics, The University of Michigan, Ann Arbor MI, United States of Ameri p y y g y g 90 (a) INFN Sezione di Milano; (b) Dipartimento di Fisica, Universit`a di Milano, Milano, Italy p y y g y g 90 (a) INFN Sezione di Milano; (b) Dipartimento di Fisica, Universit`a di Milano, Milano, Italy 91 B.I. The ATLAS Collaboration Dodge Department of Physics and Astronomy, University of Oklahoma, N 112 Homer L. Dodge Department of Physics and Astronomy, University of Oklahoma, Norman OK, United States of America 113 D t t f Ph i Okl h St t U i it Still t OK U it d St t f A i 113 Department of Physics, Oklahoma State University, Stillwater OK, United States of America 113 Department of Physics, Oklahoma State University, Stillwater OK, United States of Am 114 Palack´y University, RCPTM, Olomouc, Czech Republic 114 Palack´y University, RCPTM, Olomouc, Czech Republic 115 Center for High Energy Physics, University of Oregon, Eugene OR, United States of America 6 115 Center for High Energy Physics, University of Oregon, Eugene OR, United Stat 117 Graduate School of Science, Osaka University, Osaka, Japan 117 Graduate School of Science, Osaka University, Osaka, Japan 118 117 Graduate School of Science, Osaka University, Osaka, Japan 118 Department of Physics, University of Oslo, Oslo, Norway 118 Department of Physics, University of Oslo, Oslo, Norway p y y y 119 Department of Physics, Oxford University, Oxford, United Kingdom ( ) ( ) epartment of Physics, Oxford University, Oxford, Un 119 Department of Physics, Oxford University, Oxford, United Kingdom p y y g 120 (a) INFN Sezione di Pavia; (b) Dipartimento di Fisica, Universit`a di Pavia, Pavia, Italy 120 (a) INFN Sezione di Pavia; (b) Dipartimento di Fisica, Universit`a di Pavia, Pavia, Italy 121 121 Department of Physics, University of Pennsylvania, Philadelphia PA, United States of America 48 122 Petersburg Nuclear Physics Institute, Gatchina, Russia 122 Petersburg Nuclear Physics Institute, Gatchina, Russia ( ) ( ) g y 123 (a) INFN Sezione di Pisa; (b) Dipartimento di Fisica E. Fermi, Universit`a di Pisa, Pisa, Italy 124 INFN Sezione di Pisa; (b) Dipartimento di Fisica E. The ATLAS Collaboration Fermi, Universit`a di Pisa, Pisa, Italy 124 Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh PA, United States of America 124 Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh PA, United States of America p Universidade de Lisboa, Lisboa; (c) Department of Physics, University of Coimbra, Coimbra; (d) Centro de F´ısica Nuclear da Universidade de Lisboa, Lisboa; (e) Departamento de Fisica, Universidade do Minho, Braga; (f) Depa eorica y del Cosmos and CAFPE, Universidad de Gra Teorica y del Cosmos and CAFPE, Universidad de Granada, Granada (Spain); (g) Dep Fisica and CEF Ciencias e Tecnologia, Universidade Nova de Lisboa, Caparica, Portugal 6 Institute of Physics, Academy of Sciences of the Cz Institute of Physics, Academy of Sciences of the Czech Republic, Praha, Czech Republic 127 Czech Technical University in Prague, Praha, Czech Republic y y p 127 Czech Technical University in Prague, Praha, Czech Republic y y p p 127 Czech Technical University in Prague, Praha, Czech Republic 127 Czech Technical University in Prague, Praha, Czech Republic y g p 128 Faculty of Mathematics and Physics, Charles University in Prague, Praha, Czech Republic 129 128 Faculty of Mathematics and Physics, Charles University in Pragu ate Research Center Institute for High Energy Physic 129 State Research Center Institute for High Energy Physics, Protvino, Russia 130 Particle Physics Department, Rutherford Appleton Laboratory, Didcot, United Kingdom 131 131 Physics Department, University of Regina, Regina SK, Canada y p y g g 132 Ritsumeikan University, Kusatsu, Shiga, Japan ( ) ( ) 132 Ritsumeikan University, Kusatsu, Shiga, Japan ( ) ( ) y g p 133 (a) INFN Sezione di Roma; (b) Dipartimento di Fisica, Sapienza Universit`a di Roma, Roma, Italy 134 ( ) (b) y g p INFN Sezione di Roma; (b) Dipartimento di Fisica, Sapienza Universit`a di Roma, Roma, Italy (b) 133 (a) INFN Sezione di Roma; (b) Dipartimento di 134 (a) INFN Sezione di Roma Tor Vergata; (b) Dipartimento di Fisica, Universit`a di Roma Tor Vergata ( ) ( ) INFN Sezione di Roma Tor Vergata; (b) Dipartimento di Fisica, Universit`a di Roma Tor Vergata, Roma, I (b) 134 (a) INFN Sezione di Roma Tor Vergata; (b) Dipartimento di Fisica, Universit`a di Roma Tor Vergata, Roma, Italy 135 (a) INFN S i di R T (b) Di i di M i Fi i U i i ` R T R I l FN Sezione di Roma Tre; (b) Dipartimento di Matematica e Fisica, Universit`a Roma Tre, Roma, Italy 135 (a) INFN Sezione di Roma Tre; (b) Dipartimento di Matematica e Fisica, Universit`a Roma Tre, Ro 136 ( ) ´e des Sciences Ain Chock, R´eseau Universitaire de Physique des Hautes Energies - Universit´e Hassan II, b) ( ) 136 (a) Facult´e des Sciences Ain Chock, R´eseau Universitaire de Physique des Hautes Energies - Univ ( ) ( ) Casablanca; (b) Centre National de l’Energie des Sciences Techniques Nucleaires, Rabat; (c) Facult´e des Sciences Semlalia ( ) ( Casablanca; (b) Centre National de l’Energie des Sciences Techniques Nucleaires, Rabat; (c) Facult´e des Sciences Semlalia, g q Universit´e Cadi Ayyad, LPHEA-Marrakech; (d) Facult´e des Sciences, Universit´e Mohamed Premi ersit´e Cadi Ayyad, LPHEA-Marrakech; (d) Facult´e de Universit´e Cadi Ayyad, LPHEA-Marrakech; (d) Facult´e des Sciences, Universit´e Mohamed Premier and LPTPM, Oujda; (e) Facult´e des sciences, Universit´e Mohammed V-Agdal, Rabat, Morocco Facult´e des sciences, Universit´e Mohammed V-Agdal, Rabat, Morocco g IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat `a l’En M/IRFU (Institut de Recherches sur les Lois Fondame 137 DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Sa Atomique et aux Energies Alternatives), Gif-sur-Yvette, France q g 138 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa Cruz CA, United States of America 138 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa 138 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa Cruz CA, United States of America epartment of Physics, University of Washington, Seat 140 Department of Physics and Astronomy, University of Shefﬁeld, Shefﬁeld, United Kingdom 140 Department of Physics and Astronomy, University of Shefﬁeld, Shefﬁeld, United Kingdom 141 Department of Physics, Shinshu University, Nagano, Japan 142 Fachbereich Physik, Universit¨at Siegen, Siegen, Germany 143 Department of Physics, Simon Fraser University, Burnaby BC, Canada SLAC National Accelerator Laboratory, Stanford CA ( ) 144 SLAC National Accelerator Laboratory, Stanford CA, United States of America 145 (a) l f h i h i & f i C i i i i 145 (a) Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b) Department of Subnuclear 145 (a) Faculty of Mathematics, Physics & Informatics, Comenius University, Bratisla 145 (a) Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b) Department of Subnuclear Physics, titute of Experimental Physics of the Slovak Academy of Sciences, Kosice, Slovak Republic ( ) ( ) Institute of Experimental Physics of the Slovak Academy of Sciences, Kosice, Slovak Republic ( ) ( ) (a) Department of Physics, University of Cape Town, Cape Town; (b) Department of Physics, University ( ) 146 (a) Department of Physics, University of Cape Town, Cape Town; (b) Department of Physics, U p y y p p p y y Johannesburg; (c) School of Physics, University of the Witwatersrand, Johannesburg, South Africa Johannesburg; (c) School of Physics, University of the Witwatersrand, Johannesburg, South Africa 147 (a) Department of Physics, Stockholm University; (b) The Oskar Klein Cen p y , y; , 148 Physics Department, Royal Institute of Technology, Stockholm, Sweden 149 p y y 148 Physics Department, Royal Institute of Technology, Stockholm, Sweden 149 148 Physics Department, Royal Institute of Technology, Stockholm, Sweden Department, Royal Institute of Technology, Stockholm 149 Departments of Physics & Astronomy and Chemistry, Stony Brook University, Stony Brook Departments of Physics & Astronomy and Chemistry, Stony Brook University, Stony Brook NY, United 150 Department of Physics and Astronomy, University of Sussex, Brighton, United Kingdom 5 Department of Physics and Astronomy, University of Sussex, Brighton, United Kingdom 151 School of Physics, University of Sydney, Sydney, Australia 152 Institute of Physics, Academia Sinica, Taipei, Taiwan 153 Department of Physics, Technion: Israel Institute of Technology, Haifa, Israel 5 3 Department of Physics, Technion: Israel Institute of 153 Department of Physics, Technion: Israel Institute of Technology, Haifa, Israel p y gy Raymond and Beverly Sackler School of Physics and Astronomy, Tel Aviv University, Tel Aviv, Israel p y gy 154 Raymond and Beverly Sackler School of Physics and Astronomy, Tel Aviv Univ 155 Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece 155 Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece 156 ternational Center for Elementary Particle Physics an International Center for Elementary Particle Physics and Department of Physics, The University of T 156 International Center for Elementary Particle Physics and Department of Physics, The Univer Graduate School of Science and Technology, Tokyo Metropolitan University, Tokyo, Japan 157 Graduate School of Science and Technology, Tokyo Metropolitan University, Tokyo, Japan 158 Department of Physics, Tokyo Institute of Technology, Tokyo, Japan 158 Department of Physics, Tokyo Institute of Technology, Tokyo, Japan ent of Physics, University of Toronto, Toronto ON, C ( ) p y y 160 (a) TRIUMF, Vancouver BC; (b) Department of Physics and Astronomy, York University, Toronto ON, Canada 161 160 (a) TRIUMF, Vancouver BC; (b) Department of Physics and Astronomy, York University, 161 Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Japan culty of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Japan Department of Physics and Astronomy, Tufts University, Medford MA, United States of America 162 Department of Physics and Astronomy, Tufts University, Medford MA, Unite 163 Centro de Investigaciones, Universidad Antonio Narino, Bogota, Colombia Centro de Investigaciones, Universidad Antonio Narino, Bogota, Colombia p y y, y , , 165 (a) INFN Gruppo Collegato di Udine, Sezione di Trieste, Udine; (b) ICTP, Trieste; (c) Dipartim p y y y 165 (a) INFN Gruppo Collegato di Udine, Sezione di Trieste, Udine; (b) ICTP, Trieste; (c) Dipartimento di Chimica, Fisica e 165 (a) INFN Gruppo Collegato di Udine, Sezione di Trieste, Udine; (b) ICTP, Trieste; (c) Dipartimento di Chimica, Fisica e 165 (a) INFN Gruppo Collegato di Udine, Sezione di Trieste, Udine; (b) ICTP, Trieste; (c) Dipartimento di Chimica, Fisica e Ambiente Universit`a di Udine Udine Italy Ambiente, Universit`a di Udine, Udine, Italy y Department of Physics, University of Illinois, Urbana IL, United States of America y 166 Department of Physics, University of Illinois, Urbana IL, United States of America 167 Department of Physics and Astronomy, University of Uppsala, Uppsala, Sweden 167 Department of Physics and Astronomy, University of Uppsala, Uppsala, Sweden 168 p y y y pp pp 168 Instituto de F´ısica Corpuscular (IFIC) and Departamento de F´ısica At´omica, Molecular y Nuclear and Departamento de 49 Ingenier´ıa Electr´onica and Instituto de Microelectr´onica de Barcelona (IMB-CNM), University of Valencia and CSIC, Valencia, Spain Department of Physics, University of British Columbia, Vancouver BC, Canada p 169 Department of Physics, University of British Columbia, Vancouver BC, Canada y y y 171 Department of Physics, University of Warwick, Coventry, United Kingdom Department of Physics, University of Warwick, Coventry, United Kingdom 172 Waseda University, Tokyo, Japan Department of Particle Physics, The Weizmann Institute of Science, Rehovot, Israel 3 Department of Particle Physics, The Weizmann Inst 173 Department of Particle Physics, The Weizmann Institute of Science, Rehovot, Israel Department of Physics, University of Wisconsin, Madison WI, United States of America 174 Department of Physics, University of Wisconsin, Madison WI, United States of America y 176 Fachbereich C Physik, Bergische Universit¨at Wuppertal, Wuppertal, Ger 177 Department of Physics, Yale University, New Haven CT, United States of America 178 Yerevan Physics Institute, Yerevan, Armenia y , , 179 Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules (IN2P3), Villeurbanne, France a Al f h i i ’ C ll d d i d i d y , , 179 Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules (IN2P3), Villeurbanne, France y 179 Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules (IN2P3), Villeurbanne, France a Also at Department of Physics King’s College London London United Kingdom 179 Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules (IN Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules (IN2P3), Ville a Also at Department of Physics, King’s College London, London, United Kingdom a Also at Department of Physics, King’s College London, London, United Kingdom p y g g g b Also at Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan titute of Physics, Azerbaijan Academy of Sciences, B b Also at Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan c Also at Particle Physics Department, Rutherford Appleton Laboratory, Didcot, Un c Also at Particle Physics Department, Rutherford Appleton Laboratory, Didcot, United Kingdom d c Also at Particle Physics Department, Ruthe d Also at TRIUMF, Vancouver BC, Canada e Also at Department of Physics, California State University, Fresno CA, United States of America f e Also at Department of Physics, California State University, Fresno CA, United e Also at Department of Physics, California Stat f Also at Tomsk State University, Tomsk, Russia f Also at Tomsk State University, Tomsk, Russia g Also at CPPM, Aix-Marseille Universit´e and CNRS/IN2P3, Mars g Also at CPPM, Aix-Marseille Universit´e and CNRS/IN2P3, Marseille, France h Also at Universit`a di Napoli Parthenope, Napoli, Ita i Also at Institute of Particle Physics (IPP), Canada i Also at Institute of Particle Physics (IPP), Canada y ( ), j Also at Department of Physics, St. The ATLAS Collaboration Petersburg State Polytechnical University, St. Petersburg, Russia k j Also at Department of Physics, St. The ATLAS Collaboration Petersburg Sta k Also at Chinese University of Hong Kong, China k Also at Chinese University of Hong Kong, China l Also at Department of Financial and Management Engineering, University of l Also at Department of Financial and Management Engineering, University l Also at Department of Financial and Management Engineering, University of the Aegean, Chios, Greece na Tech University, Ruston LA, United States of Ame n Also at Institucio Catalana de Recerca i Estudis Avancats, ICREA, Barcelona, Spain o Also at Department of Physics, The University of Texas at Austin, Austin TX, United S p Also at Institute of Theoretical Physics, Ilia State University, Tbilisi, Georgia q Also at CERN, Geneva, Switzerland q Also at CERN, Geneva, Switzerland r Also at Ochadai Academic Production, Ochanomizu University, Tokyo, Ja s Also at Manhattan College, New York NY, United States of Ame t hattan College, New York NY, United States of Amer t Also at Novosibirsk State University, Novosibirsk, Russia t Also at Novosibirsk State University, Novosibirsk, Russia u Also at Institute of Physics, Academia Sinica, Taipei, Taiwan u Also at Institute of Physics, Academia Sinica, Taipei, Taiwan v Also at LAL, Universit´e Paris-Sud and CNRS/IN2P3, Orsay, France y w Also at Academia Sinica Grid Computing, Institute of Physics, Academia Sinica, Taipei, Taiwan x Also at Laboratoire de Physique Nucl´eaire et de Hautes Energies, UPMC and Universit´e Paris-Diderot and CNRS/IN2P3, Paris, France y Also at School of Physical Sciences, National Institute of Science Education and Research, Bhubanesw z Also at Dipartimento di Fisica, Sapienza Universit`a di Roma, Roma, Italy aa Also at Moscow Institute of Physics and Technology State University, Dolgoprudny, Russia b de Physique, Universit´e de Gen`eve, Geneva, Switzerla ab Also at Section de Physique, Universit´e de Gen`eve, Geneva, Switzerland y q ac Also at International School for Advanced Studies (SISSA), Trieste, Italy onal School for Advanced Studies (SISSA), Trieste, It partment of Physics and Astronomy, University of South Carolina, Columbia SC, United States of Americ ad Also at Department of Physics and Astronomy, University of South Carolina, Columbia SC, U ae Also at School of Physics and Engineering, Sun Yat-sen University, Guangzhou, China a f Also at Faculty of Physics, M.V.Lomonosov Moscow State University, Moscow, Russia ag Also at Moscow Engineering and Physics Institute (MEPhI), Moscow, Russia igner Research Centre for Physics, Budapest, Hungar Particle and Nuclear Physics, Wigner Research Centr y g ai Also at Department of Physics, Oxford University, Oxford, United Kingdom a j Also at Department of Physics, Nanjing University, Jiangsu, China a j Also at Department of Physics, Nanjing University, Jiangsu, China ak Also at Institut f¨ur Experimentalphysik, Universit¨at Hamburg, Hamburg, Germany l al Also at Department of Physics, The University of Michigan, Ann Arbor MI, United States of America an Also at University of Malaya, Department of Physics, Kuala Lumpur, Malaysia ∗Deceased an Also at University of Malaya, Department of Physics, Kuala Lumpur, Malaysia ∗D d
https://openalex.org/W2523125414	https://eprints.leedsbeckett.ac.uk/id/eprint/3263/1/Potential%20tree%20species%20for%20use%20in%20urban%20areas%20in%20temperate%20and%20oceanic%20climates.pdf	English	null	Potential tree species for use in urban areas in temperate and oceanic climates	Heliyon	2,016	cc-by	12,583	Potential tree species for use in urban areas in temperate and oceanic climates Received: 30 May 2016 Revised: 23 July 2016 Accepted: 1 September 2016 Heliyon 2 (2016) e00154 Heliyon 2 (2016) e00154 Miklas Scholz a,b,, Vincent C. Uzomah b, Furat A.M. Al-Faraj b Miklas Scholz a,b,, Vincent C. Uzomah b, Furat A.M. Al-Faraj b Miklas Scholz a,b,, Vincent C. Uzomah b, Furat A.M. Al-Faraj b a Division of Water Resources Engineering, Faculty of Engineering, Lund University, P.O. Box 118, 22100 Lund, Sweden b Civil Engineering Research Group, School of Computing, Science and Engineering, The University of Salford, Newton Building, Salford M5 4WT, England, United Kingdom Corresponding author at: Division of Water Resources Engineering, Faculty of Engineering, Lund University, P.O. Box 118, 22100 Lund, Sweden. E-mail address: miklas.scholz@tvrl.lth.se (M. Scholz). a Division of Water Resources Engineering, Faculty of Engineering, Lund University, P.O. Box 118, 22100 Lund, Sweden b Civil Engineering Research Group, School of Computing, Science and Engineering, The University of Salford, Newton Building, Salford M5 4WT, England, United Kingdom b Civil Engineering Research Group, School of Computing, Science and Engineering, The University of Salford, Newton Building, Salford M5 4WT, England, United Kingdom * Corresponding author at: Division of Water Resources Engineering, Faculty of Engineering, Lund University, P.O. Box 118, 22100 Lund, Sweden. E-mail address: miklas.scholz@tvrl.lth.se (M. Scholz). http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Citation: Scholz, M and Uzomah, VC and Al-Faraj, FAM (2016) Potential tree species for use in ur- ban areas in temperate and oceanic climates. Heliyon, 2 (9). ISSN 2405-8440 DOI: https://doi.org/10.1016/j.heliyon.2016.e00154 Citation: Scholz, M and Uzomah, VC and Al-Faraj, FAM (2016) Potential tree species for use in ur- ban areas in temperate and oceanic climates. Heliyon, 2 (9). ISSN 2405-8440 DOI: https://doi.org/10.1016/j.heliyon.2016.e00154 Link to Leeds Beckett Repository record: https://eprints.leedsbeckett.ac.uk/id/eprint/3263/ Document Version: Article (Published Version) Creative Commons: Attribution 4.0 Creative Commons: Attribution 4.0 The aim of the Leeds Beckett Repository is to provide open access to our research, as required by funder policies and permitted by publishers and copyright law. The Leeds Beckett repository holds a wide range of publications, each of which has been checked for copyright and the relevant embargo period has been applied by the Research Services team. We operate on a standard take-down policy. If you are the author or publisher of an output and you would like it removed from the repository, please contact us and we will investigate on a case-by-case basis. Each thesis in the repository has been cleared where necessary by the author for third party copyright. If you would like a thesis to be removed from the repository or believe there is an issue with copyright, please contact us on openaccess@leedsbeckett.ac.uk and we will investigate on a case-by-case basis. 1.1. Rationale Trees play major roles in creating healthy urban ecosystems and sustainable environments. However, some trees may cause damage to urban structures such as permeable pavements, impermeable pavements, kerbs, roads, footpaths, buildings and retaining walls. For the purpose of this study, permeable pavements are defined as a sustainable system comprising a base and subbase allowing the movement of storm water through the surface, reducing runoff. Randrup et al. (2003) indicated that in some cities substantial amount of money has been allocated to address conflicts between the rooting system of trees and urban infrastructure. An assessment of this sort becomes important as part of a decision support tool for the fitting and retrofitting of sustainable drainage systems (SuDS), and in the planning of tree planting projects at urban development sites, regeneration projects, and sustainable drainage projects (Scholz and Uzomah, 2013). A SuDS is designed to lessen the potential impact of construction developments with concerning surface water drainage discharges (Scholz, 2010, 2015). Abstract This study aims to assess the potential of trees for integration in urban development by evaluating the damage caused by trees in relation to various tree characteristics. Tree damage to permeable pavement systems and other urban structures such as impermeable pavements, kerbs, roads, retaining walls, footpaths, walls and buildings were assessed to identify the most suitable trees for the urban environment. One hundred square sites of 100 m × 100 m were randomly selected in Greater Manchester for this representative example case study to demonstrate the assessment methodology. Among tree species in this study, Acer platanoides L. (Norway maple) occurred most frequently (17%); others were Tilia spp. L. (Lime; 16%), Fraxinus excelsior L. (common ash; 12%), Acer pseudoplatanus L. (sycamore; 10%) and Prunus avium L. (wild cherry; 8%). The study concludes that 44% of the damage was to impermeable pavements and 22% to permeable pavements. Other damage to structures included kerbs (19%), retaining walls (5%), footpaths (4%), roads (3%) and walls (3%). Concerning the severity of damage, 66% were moderate, 21% light and 19% severe. Aesculus hippocastanum L. (horse chestnut) caused the greatest damage (59%) expressed in percentage as a ratio of the tree number related to damage over the corresponding tree number that was found close to structures. Article No~e00154 Keywords: Engineering, Environmental science Keywords: Engineering, Environmental science Keywords: Engineering, Environmental science 2 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 1.2. Brief literature review Table 1 summarised tree development characteristics (Pliûra and Heuertz, 2003; Defra, 2007; Arbor Day Foundation, 2015; British Hardwood Tree Nursery, 2015; Garden Centre, 2015). The presence of urban trees often increases property prices (Sander et al., 2010; Scholz, 2010, 2015). Trees may be regarded as desirable by residents, because they increase the aesthetic value of a place, and provide ecosystem services including amenities (Scholz and Uzomah, 2013). The root system of vegetation such as most trees provides the essential functions of anchorage (known as structural stability), absorption of runoff water and nutrients as well as storage of vital food reserves (Scholz, 2010, 2015). Tree roots may cause damage to underground utility services by direct pressure on conduits as roots grow and expand in diameter, or by entry to hydraulic services such as sewers and storm water pipes occasionally causing destruction and more frequently blockage (Mather and Morton, 2008). Tree species that have large and vigorous root systems in terms of their growth rate may result in significant destruction of public infrastructure elements including roads, kerbs, footpaths, paved areas and underground utility services. Trees with these characteristics should preferably be avoided or at least controlled. Mechanical forces exerted by radial growth of tree roots can lift relatively light structures such as paths, curbs, paving slabs and n Centre (2015), Arbor Day Foundation (2015), British Hardwood Tree Nursery (2015), Pliûra and mum years) Early growth pattern Growth rate Root pattern Best soil condition Comments After 10 years of age After 20 years of age Height (m) Crown (m) Height (m) Crown (m) 250 8 4 13 7 Medium – Acidic, alkaline, loamy, moist, sandy, well-drained, wet and clay soils; some drought tolerance. Rapid growth rate till maturity; tolerates pollution and other urban conditions well. 35–60 cm/yr 0–250 10 5 15 8 Fast – All soils; tolerates salt-laden soils. Rapid growth rate till maturity. 35–70 cm/yr (Av.=50) ≤400 8 5 11 8 Medium – Prefers deep, moist and cool soil; tolerates pollution and exposed sites. – 35–60 cm/yr 20–90 8 5 14 7 Medium to fast Requires deep soil Prefers light and sandy soil, but grows also in moist and well-drained soil; Not drought-tolerant. – 35–60 cm/yr 500 8 3 12 8 Medium to fast Deep roots Any well-drained fertile soil; able to withstand shade and pollution. Article No~e00154 boundary walls as well as occasionally single story buildings including porches and garages (Biddle, 1998; Mather and Morton, 2008). However, poor construction of pavements can also cause structural failure (Sydnor et al., 2000). Some research studies have found strong correlations between the size of specific trees and serious conflicts with infrastructure demands (Mather and Morton, 2008). Large trees usually cause more conflicts than small trees. Damage to pavements correlate with nearby tree diameters. Most trees are linked to damage when they are between 11 and 20 cm in diameter (measured at breast height). However, most Quercus spp. (Oak) and A. hippocastanum do not cause harm, unless they are greater than 20 cm in tree diameter at breast height (DBH) according to Randrup et al. (2003). Randrup et al. (2003) pointed out that a concrete or asphalt pathway can act as a barrier preventing soil moisture loss by evaporation. This artificial evaporation barrier creates a more humid environment on the underside of the pavement surface, because of temperature differences between the soil and the above pavement. Tree roots are therefore naturally attracted to the condensation water at the soil and impermeable pavement interface (Randrup et al., 2003). This may eventually lead to pavement surface destruction through the radial forces generated during tree root growth. When tree roots encounter dense soil layers, they usually change direction, stop growing, or adapt by remaining unusually close to the surface. This superficial rooting makes urban trees more vulnerable to drought and can cause destructive pavement heaving (Randrup et al., 2003). The highly compacted soils commonly required for constructing pavements do not allow tree root penetration (Scholz, 2013). Viswanathan et al. (2011) undertook a research study concerned with the performance of Liquidambar styraciflua L. (American sweetgum) roots under permeable and impermeable pavements. Their results suggested that the standing live root lengths for the American sweetgum were longer in impermeable concrete than in permeable concrete for the first 0 to 20 cm of soil depth. Beyond this depth, the standing live roots were more abundant in permeable than in impermeable pavements. However, they came to the conclusion that pervious concrete does not give a quantifiable root production benefit in comparison to impervious concrete. Giuliani et al. (2015) used modelling tools to analyse tree growth in street pavements. The findings indicate the progressive reduction of deformations with the increase of the depth of root penetration. 5 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). However, these studies require a lot of data for individual trees and sites. 1.2. Brief literature review – 35–60 cm/yr (Continued) Article No~e00154 ximum (years) Early growth pattern Growth rate Root pattern Best soil condition After 10 years of age After 20 years of age Height (m) Crown (m) Height (m) Crown (m) 300 8 4 11 8 Medium – Acidic, loamy, moist, rich, sandy, silty loam well-drained and clay soils. 35–60 cm/yr 500 6 4 12 6 – Deep roots Good light loam 50–100 8 3 18 4 Fast – Rich humus and raw soil of mountainside. 35–70 cm/yr 100–150 4 3 8 5 Slow to medium – – 30–60 cm/yr (av.=40) 150–200 4 4 14 7 Slow to medium Does not need deep soil Acidic, loamy, moist, sandy, well-drained and clay soils; prefer moist and well-drained soi but has some drought toleranc 30–60 cm/yr Article No~e00154 distance from structures, DBH, tree height, crown spread (diameter), and tree hang- over characteristics. The outcomes should be used to focus on planting the most suitable tree species near specific urban structures in the future. The objectives are (1) to outline a method for random selection of representative sites in Greater Manchester to study the tree damage characteristics; (2) to identify the predominant trees causing damage to urban structures in Greater Manchester (example case study); and (3) to define a rapid methodology to assess the damage to structures such as permeable pavements, impermeable pavements, roads, kerbs, footpaths, and retention walls for which individual tree species are responsible for. This study provides valuable information for the retrofitting of structures such as permeable pavements in combination with existing trees and to developers in choosing the most suitable trees for the right urban environment minimising damage. The objectives are (1) to outline a method for random selection of representative sites in Greater Manchester to study the tree damage characteristics; (2) to identify the predominant trees causing damage to urban structures in Greater Manchester (example case study); and (3) to define a rapid methodology to assess the damage to structures such as permeable pavements, impermeable pavements, roads, kerbs, footpaths, and retention walls for which individual tree species are responsible for. 1.3. Aim, objectives and significance This study aims to provide a simple method to assess the damage caused by urban tree roots in relation to the corresponding tree characteristics such as species, 6 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2.2. Tree damage data collection Data related to tree damage assessments were collected to address objective (2). The data set included variables such as site number, tree number, tree species, common name and genus, tree DBH (1.5 m from ground level), estimated tree height, estimated tree crown diameter, structures near the tree, distance (no maximum threshold) of all aboveground nearby structures from the tree (defined at species level where possible), type of damage (if any) to structures and their severity as well as subjective aesthetic considerations. Site visits were carried out during 2013 and 2014. Predominantly summer periods were chosen because during these periods, trees have their full leaf canopies, which makes tree identifications and corresponding crown spread determinations easier. All trees within the marked 100 m × 100 m boundaries with a DBH of greater than 10 cm were assessed, except where a site was inaccessible for a valid reason; e.g., some relatively small areas within restricted (private) access areas such as gated private gardens were not assessed, and were subsequently marked as inaccessible sites. Other sites without any tree data entries were without any trees, had only trees where the DBH was less than 10 or were predominantly of a different land use category (without trees) such as water. The DBH was calculated by measuring the circumference at breast height using a tape measure, and dividing the value by π (approximately 3.14159). Trees less than 10 cm in diameter were not recorded as they were considered too young to cause any measureable damage. Tree heights were estimated using methods based on goniometry (Vernier, 2013), and also by comparing the tree height with nearby structures such as houses as well as electric and telephone poles. Goniometry involves walking away from the base of the trunk until the observer sees the top of the tree from an angle of 45° (which the observer can check using his or her arm). The height of the tree roughly equates to the distance from the tree to where the observer is standing plus his or her eye height from the ground. Article No~e00154 subsequently determined. The Greater Manchester area is located between 53°28′0′′N and 2°14′0′′W. The estimated population of Greater Manchester is around 2,680,000. 7 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2.1. Site selections In order to address objective (1), a total of 100 sites were randomly selected in the Greater Manchester area (North-west England) using the Google Earth map and tools, but restricted within the area of an ellipse covering the main urban areas around Manchester city centre for ease of assessment and to reduce transportation costs (Fig. 1). A square of 100 m × 100 m was drawn around the centre of each selected site to identify an outer boundary for the tree assessment studies. The coordinates, grid references, longitudes, latitudes and post codes of all sites were Fig. 1. Overview of the assessed case study sites located in the Greater Manchester area. Fig. 1. Overview of the assessed case study sites located in the Greater Manchester area. 2.3. Tree damage assessment method In order to address objective (3), the information and data collected and processed to show the methodology to assess damage by trees is summarised in Table 2, Table 3, Table 4, Table 5, Table 6. Table 2 shows a summary of the most frequently occurring trees and their corresponding damage to urban structures. The structures that were considered in this assessment were permeable pavements, impermeable pavements, kerbs, roads, retaining walls, buildings and footpaths. res locate ber of age on ning 0 0 4 0 1 3 0 2 1 0 11 exceed the d their corresponding damage recordings to key urban structures locat umber of es that used mage Number of damage on permeable pavement Number of damage on impermeable pavement Number of damage on kerb Number of damage on road Number of damage on retaining wall 38 10 29 12 2 0 21 9 19 3 1 0 20 4 11 8 4 4 6 4 0 1 0 0 20 7 13 9 0 1 22 6 21 6 0 3 10 8 2 2 0 0 8 1 2 2 0 2 2 0 2 0 0 1 4 2 2 2 0 0 151 51 101 45 7 11 e, and as such the addition of the number of damage to all the structures may exceed the gus vatica 0 2 0 0.1 68 28.3 0 – – – – 0 1 1 0 89 tinued) occurred ≥10 times and which were found in ≥5 different sites out of the 100 randomly selected sites in Acer pseudoplataneus L. Fraxinus excelsior L. Prunus avium L. Tilia platyphyllos L. Aesculus hippocastanum L. Tilia cordata L. Betula pendula L. Crataegus monogyna L. Fagus sylvatica L. 2.3. Tree damage assessment method 3 1 3 2 1 1 0 0 0 6 3 1 5 5 7 0 0 2 0 2 2 1 1 1 – – 0 0.5 1.8 0 0.2 0.8 0.8 – – 0.1 52 66 – 34 51 26 – – 68 20.5 10.7 – 14.2 18.7 8.6 – – 28.3 0 0 0 0 0 0 1 0 0 – – – – – – 0 – – – – – – – – 0 – – – – – – – – 20 – – – – – – – – 0 – – 3 4 0 3 3 0 2 0 0 13 7 0 10 15 0 0 2 1 1 1 – 0 1 – – 1 1 1.4 0.2 – 0 0.2 – – 0 0 57 25 – 52 70 – – 25 89 (Continued) a 0 1 1 oplataneus Fraxinus excelsior L. Prunus avium L. Tilia platyphyllos L. Aesculus hippocastanum L. Tilia cordata L. Betula pendula L. Crataegus monogyna L. Fagus sylvatica L. 22.7 8.5 – 8.1 20.5 – – 0 3 0 0 0 3 2 0 0 1 – – – 1 0 – – 1.4 – – – 0.3 0 – – 73 – – – 77 38 – – 24 – – – 12.1 0 – – 1 1 0 2 2 2 0 0 2 5 1 6 4 0 2 0 1 1 0 1 1 – 0 – 0 0.8 0 1 0.2 – 0 – 64 66 62 48 70 – 45 – 0 20.1 0 6.1 26.9 – 27.9 – 0 2 0 1 0 0 0 0 – 1 – 0 – – – – – 0.9 – 0 – – – – 0 – 36 – 43 – – – – – 3.7 – 0 – – – – 3 (Continued 0 – – – – Acer pseudoplataneus L. Fraxinus excelsior L. Prunus avium L. Tilia platyphyllos L. Aesculus hippocastanum L. Tilia cordata L. Betula pendula L. Crataegus monogyna L. Fagus sylvatica L. uctur the hors Ker (we RR 1 2.3. Tree damage assessment method 0 0 0 0 0 0 0 1 0 0 0 3 0 1 3 0 1 1 – – 0 – 0 0 – 0 0 – – 0.2 – 0 0.2 – 0 0 – – 58 – 46 61 – 125 20 – – 14.1 – 0 8.1 – 0 0 0 0 1 0 0 0 0 0 0 – – 0 – – – – – – – – 0 – – – – – – – – 38 – – – – – – – – 0 – – – – – – 2 0 0 0 0 0 0 1 0 1 0 1 2 2 0 0 0 0 3 – 1 3 0 – – – – 0 – 0 0 0 – – – – 3 – 53 49 45 – – – – 0 – 0 20.8 0 – – – – ing damage/sum of trees causing and not causing d Kerbs Roads Retaining walls Oct-16 Feb-31 0/0 02-Nov Jan-24 0/4 Jul-14 Apr-16 02-Apr Jan-14 0/13 0/2 Jul-13 0/22 01-Jun May-22 0/21 03-Apr 0/10 0/8 0/5 02-Jul 0/12 01-Feb 0/4 0/4 01-Mar 02-Sep 0/3 0/1 sing d using d s for damage he relative i ngs Over relat rank t = 10) (tota total WR 10 10 10 10 10 NA 10 10 10 10 uctural damage. The rankings indicate increases of ‘potential s the highest potential for damage. Weighting factors reflect t thors. Kerbs Roads Retaining walls Footpaths Buildin (weight = 4) (weight = 8) (weight = 4) (weight = 4) (weigh RR WR RR WR RR WR RR WR RR 10 40 3 24 NA NA 10 40 1 2 7 28 4 32 2 8 1 4 1 9 36 7 56 8 32 4 16 1 3 12 2 16 3 12 7 28 1 4 8 32 1 8 4 16 2 8 1 5 20 1 8 9 36 8 32 NA 9 1 4 5 40 1 4 9 36 1 6 6 24 2 16 7 28 3 12 1 A 2 8 6 48 5 20 6 24 1 4 16 6 48 6 24 5 20 1 ankings (RR g factors ref Overal rankin WR/tot WR – – – – – – – – – – on their growth and development characteristics. on t and ment Roa (we Article No~e00154 The footpath structure refers to a walkway though areas such as parks. The damage that was taken into account is lifting-up of structures, disjointing of structures by roots, sinking-in (depression) of structures and cracking-up of structures. The assessment was undertaken with care to distinguish between damage due to trees and/or poor construction (Sydnor et al., 2000). However, all assessments were based on civil engineering expert opinion considering that no disruptive and/or destructive tests could be undertaken on private and public land. Pictures of actual root damage were taken and analysed. The footpath structure refers to a walkway though areas such as parks. The damage that was taken into account is lifting-up of structures, disjointing of structures by roots, sinking-in (depression) of structures and cracking-up of structures. The assessment was undertaken with care to distinguish between damage due to trees and/or poor construction (Sydnor et al., 2000). However, all assessments were based on civil engineering expert opinion considering that no disruptive and/or destructive tests could be undertaken on private and public land. Pictures of actual root damage were taken and analysed. The severity of damage was determined by assigning numbers between 1 and 3, where 1 represents an emerging damage at an early stage (‘light damage’), 2 indicates a damage that is gradually advancing or already well-established (‘moderate damage’), and 3 equates to ‘severe damage’, which is an advanced damage (e.g., pavements completely separated or kerbs completely disjointed) or a well-advanced damage that has become a safety hazard to users requiring immediate attention (or a damage that has already been repaired). It follows that essentially a rather coarse three-category damage scale (Table 3) has been used to reflect the fact that damage to structures by trees is rare despite the large data set collected. Furthermore, the absolute majority of trees did not cause any damage, and could be seen as the control group, which was assigned 0 (no damage recorded). Table 4 and Table 5 indicate the proportion of tree species that caused structural damage and the relative tree rankings concerning the structural damage, respectively (see Section 3.4 for detailed descriptions and interpretations). Weighting factors reflect the relative importance of structures to the local infrastructure (key criterion of assessment), and have been determined by civil engineering expert judgements expressed by the authors (Table 5). For example, damage to a building receives a higher weighting than damage to pavements. However, pavements are more important than their corresponding kerbs. Finally, Table 6, which has also been informed by the literature review and Table 1, shows the predicted future damage potentials for tree species based on their growth and development characteristics. 2.3. Tree damage assessment method Relative ranking and 10 represents highest potential for damage. Weighting facto ent by the authors. Roads (weight = 8) Retaining walls (weight = 4) Footpaths (weight = 4) Buildings (weight = 10) O r W RR WR RR WR RR WR RR WR 3 24 – – 2 8 – – 1 8 – – – – – – 2 16 3 12 4 16 – – – – – – 1 4 – – – – 4 16 5 20 – – – – 2 8 – – – – – – – – – – – – – 1 4 – – – – – – 5 20 – – – – – – – – 3 12 – – 15 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.1. General overview A total of 536 mature trees were assessed in detail. After applying the criteria given above, the tree species percentage occurrence reduced accordingly. Table 2 shows a summary of the most frequently occurring trees and their corresponding damage recordings to key urban structures. The range of tree size has been limited by including only trees that have DBH entries of at least 10 cm to avoid skewing the data set towards small trees that might not survive. Furthermore, small and young trees have not yet developed sufficient size and strength to cause damage to the surrounding infrastructure. The fact that the standard deviations of DBH are high reflects the point that even mature trees are highly variable in size, which is natural. The application of the proposed methodology has been demonstrated in Table 2, Table 3, Table 4, Table 5, Table 6 as discussed in Sections 3.2–3.4 below. 16 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2.4. Statistics Microsoft Excel (www.microsoft.com) and IBM SPSS Statistics Version 20 (www.ibm.com) were used. All data collected have been quality-checked, and outliers have been identified and removed if there was a scientific reason to do so. Descriptive summary statistics, regression analysis and the non-parametric Mann- Whitney U-test have been performed for statistically valid data sets such as damage to structures, if data sets were sufficiently large. Significant (p < 0.05) findings have been highlighted, where appropriate. The ability to conduct further 15 6 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 statistical analyses of the data set was limited due to the small sample size for most species causing damage and the dynamic nature of the urban environment. Article No~e00154 follows: Aesculus hippocastanum, 59%; Tilia platyphyllos Scop. (large-leaved lime), 53%; F. excelsior, 45%; A. pseudoplatanus, 42%; Tilia cordata Mill. (small- leaved lime), 36%; Fagus sylvatica L. (common beech), 33%; Betula pendula L (silver birch), 32%; P. avium, 15%; and Crataegus monogyna L. (common hawthorn), 11%. The severity of corresponding damage was in the following order: moderate (66%); light (21%); and severe (13%). follows: Aesculus hippocastanum, 59%; Tilia platyphyllos Scop. (large-leaved lime), 53%; F. excelsior, 45%; A. pseudoplatanus, 42%; Tilia cordata Mill. (small- leaved lime), 36%; Fagus sylvatica L. (common beech), 33%; Betula pendula L (silver birch), 32%; P. avium, 15%; and Crataegus monogyna L. (common hawthorn), 11%. The severity of corresponding damage was in the following order: moderate (66%); light (21%); and severe (13%). follows: Aesculus hippocastanum, 59%; Tilia platyphyllos Scop. (large-leaved lime), 53%; F. excelsior, 45%; A. pseudoplatanus, 42%; Tilia cordata Mill. (small- leaved lime), 36%; Fagus sylvatica L. (common beech), 33%; Betula pendula L (silver birch), 32%; P. avium, 15%; and Crataegus monogyna L. (common hawthorn), 11%. The severity of corresponding damage was in the following order: moderate (66%); light (21%); and severe (13%). Findings indicate that there were no obvious patterns of damage to structures. This could be attributed to relatively small sample sizes and complex processes such as differences in soil moisture content, various levels of structural compactions, and average distance of trees from structures. For example, trees are normally planted closer to permeable pavements, impermeable pavements and kerbs compared to roads. In order to achieve maximum ecosystem service benefits, the most suitable trees that could be combined with SuDS are the one that (a) are as close to structure as possible; (b) have a large diameter; (c) cause the least or no damage; and (d) are readily available and desirable by residents. The closer trees are to the structures or residents, the more the effects are felt; e.g., reducing localised extreme temperatures. The greater a tree diameter, the more mature the tree is likely to be and, therefore, the more noticeable will be the tree benefits (Leuzinger et al., 2010). Trees linked to a low damage potential are usually preferred both for new construction or retrofitting of SuDS sites. Vegetation that is desirable by residents is usually associated with high aesthetic values such as mature and beautiful trees with a perceived rich character. 17 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.2. Structural damage The proportions of structures that were linked to damage from trees can be found in Table 2. Of the total 231 damaged structures observed, the following proportions expressed in percentages can be calculated: impermeable pavements (44%), permeable pavements (22%), kerbs (19%), retaining walls (5%), footpaths (4%), roads (3%) and walls (3%). No damage to buildings (0%) has been recorded. The patterns associated with damage linked to impermeable pavements compared to permeable pavements are in line with the findings by Randrup et al. (2003). This suggests the need for more retrofitting of robust SuDS techniques (Scholz and Uzomah, 2013). However, it is expected that the severity of damage will advance further with time. No damage to buildings was recorded. This is possibly due to the fact that the assessment was only based on an external visual observation. An internal structural assessment may reveal damage to buildings. Moreover, most buildings have formidable foundations and may not be easily damaged as compared to road structures and pathways. By dividing the number of a particular species causing damage by the number of the corresponding species occurrence, the proportion of damage caused by this species can be calculated from Table 2. In contrast to the control group (48% of A. platanoides that did not cause any damage), about 52% of all A. platanoides caused various kinds of damage (as detailed in Table 2) to urban structures determined by expert judgement. The proportions of the other species that caused damage were as 16 17 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 3.3.1. Overview Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from permeable pavements, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 2. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from permeable pavements, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. 18 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.3.1. Overview Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8 show the relationship of tree DBH, average distance of trees away from the structures, and the proportion of trees close to structures that caused moderate to severe damage. Note that only moderate and severe damage was considered, considering that the reason for light damage is often unclear. Apart from tree-related damage, other reasons for damage might be as important but only further destructive tests on site might reveal the main reason(s) for damage. For x(y/z), where x represents the DBH (cm), which is also signified by the relative size (diameter) of the circle. The diameters expressed by circles give a visual indication of the maturity of the average tree species, which makes visual comparisons between trees easier. The entry z indicates the number of the tree species within 10 m of the structure, out of which y trees caused moderate to severe damage. 17 [(Fig._2)TD$FIG] Norway maple 56(8/21) Sycamore 52(6/12) Common ash 66(3/9) Large-leaved lime 34(5/23) Horse chesnust 51(3/10) Small-Leaved lime 26(7/16) Silver birch 20(1/2) Common beech 68(2/4) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0% 20% 40% 60% 80% Distance (m) from permeable pavement Percentage of tree species causing damage Fig. 2. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from permeable pavements, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Article No~e00154 Article No~e00154 Norway maple 56(8/21) Sycamore 52(6/12) Common ash 66(3/9) Large-leaved lime 34(5/23) Horse chesnust 51(3/10) Small-Leaved lime 26(7/16) Silver birch 20(1/2) Common beech 68(2/4) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0% 20% 40% 60% 80% Distance (m) from permeable pavement Percentage of tree species causing damage Percentage of tree species causing damage Fig. 2. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from permeable pavements, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 2. 3.3.2. Permeable pavement For permeable pavements, most significant (p < 0.05) damage to permeable pavements was caused by trees located within 0–1.0 m away from a structure, except for those from F. excelsior. About 33% of F. excelsior located close to permeable pavements caused damage to these pavements if their average diameter was 66 cm and if their average distance was 2.3 m away from the permeable pavements (Fig. 2). The trees with the highest percentage of moderate and severe damage to permeable pavements (up to 50%) were F. sylvatica, A. pseudoplatanus and B. pendula. However, the corresponding sample sizes were rather small. The average distance of F. sylvatica and B. pendula to permeable pavements was 0 m, indicating that most of these two species were planted too close to the pavement. The average DBH of these trees was 68 cm and 20 cm, respectively (Fig. 3). [(Fig._3)TD$FIG] Norway maple 42(19/32) Sycamore 64(14/18) Common ash 30(7/10) Large -leaved lime 52(6/11) Horse chesnust 71(14/20) Small-leaved lime 38(2/5) Common hawthorn 25(2/4) Common beech 89(2/3) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Distance (m) from impermeable pavement Percentage of tree species close to impermeble pavements causing damage Fig. 3. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from impermeable pavements, and the proportion of trees within 10 m to this structure subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Norway maple 42(19/32) Sycamore 64(14/18) Common ash 30(7/10) Large -leaved lime 52(6/11) Horse chesnust 71(14/20) Small-leaved lime 38(2/5) Common hawthorn 25(2/4) Common beech 89(2/3) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Distance (m) from impermeable pavement Percentage of tree species close to impermeble pavements causing damage Horse chesnust 71(14/20) Percentage of tree species close to impermeble pavements causing damage Fig. 3. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from impermeable pavements, and the proportion of trees within 10 m to this structure subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. 19 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.3.2. Permeable pavement The average DBH of this tree was 52 cm and it was located about 0.4 m away from structures. Tilia platyphyllos caused five major damage occurrences to permeable Acer platanoides caused the most overall damage (2 light, 4 moderate and 4 severe damage) to permeable pavements (Table 3), which was statistically significant (p < 0.05). The average diameter of the tree was 56 cm and the mean distance from the permeable pavements was 0.8 m. However, Fig. 3 shows a comparison for only moderate and severe damage. Seven out of sixteen T. cordata trees located close to permeable pavements caused major damage. The corresponding average tree diameter was 26 cm and the mean distance from the structures was 0.8 m. Acer pseudoplatanus caused six major damage incidents to permeable pavements. The average DBH of this tree was 52 cm and it was located about 0.4 m away from structures. Tilia platyphyllos caused five major damage occurrences to permeable [(Fig._5)TD$FIG] Norway maple 50(2/31) Sycamore 73(1/24) Common ash 72(4/16) 0 1 2 3 4 5 0% 20% 40% 60% 80% 100% Distance (m) from road Percentage of tree species close to roads causing damage Fig. 5. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from roads, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. [(Fig._5)TD$FIG] Norway maple 50(2/31) Sycamore 73(1/24) Common ash 72(4/16) 0 1 2 3 4 5 0% 20% 40% 60% 80% 100% Distance (m) from road Percentage of tree species close to roads causing damage Common ash 72(4/16) Percentage of tree species close to roads causing damage Fig. 5. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from roads, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 5. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from roads, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. 3.3.2. Permeable pavement 18 [(Fig._4)TD$FIG] Norway maple 41(10/16) Sycamore 64(2/11) Common ash 62(7/14) Wild cherry 62(1/14) Large-leaved lime 48(7/13) Horse chesnust 71(5/22) Silver birch 45(2/7) Common beech 93(2/9) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0% 20% 40% 60% 80% 100% Distance (m) from kerb Percentage of tree species close to kerb causing damage Fig. 4. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from kerbs, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Article No~e00154 Article No~e00154 [(Fig._4)TD$FIG] Norway maple 41(10/16) Sycamore 64(2/11) Common ash 62(7/14) Wild cherry 62(1/14) Large-leaved lime 48(7/13) Horse chesnust 71(5/22) Silver birch 45(2/7) Common beech 93(2/9) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0% 20% 40% 60% 80% 100% Distance (m) from kerb Percentage of tree species close to kerb causing damage Percentage of tree species close to kerb causing damage Fig. 4. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from kerbs, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 4. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from kerbs, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Acer platanoides caused the most overall damage (2 light, 4 moderate and 4 severe damage) to permeable pavements (Table 3), which was statistically significant (p < 0.05). The average diameter of the tree was 56 cm and the mean distance from the permeable pavements was 0.8 m. However, Fig. 3 shows a comparison for only moderate and severe damage. Seven out of sixteen T. cordata trees located close to permeable pavements caused major damage. The corresponding average tree diameter was 26 cm and the mean distance from the structures was 0.8 m. Acer pseudoplatanus caused six major damage incidents to permeable pavements. 3.3.2. Permeable pavement Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. 19 [(Fig._6)TD$FIG] Common ash 53(2/4) Large-leaved lime 46(1/6) Horse chesnust 61(3/4) Silver birch 125(1/2) Common hawthorn 20(1/3) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0% 20% 40% 60% 80% 100% ll a w g n i n i a t e r m o rf ) m ( e c n a t s i D Percentage of tree species close to retaining walls causing damage Article No~e00 Article No~e00154 Percentage of tree species close to retaining walls causing damage Fig. 6. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from retaining walls, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 6. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from retaining walls, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. pavements; its average diameter was 34 cm and the corresponding mean distance from structures was 0.8 m. For F. excelsior, although three major damage incidents to permeable pavements were recorded, the average distance from the structure was 2.3 m and the mean diameter was 66.1 cm, indicating that these were mature trees located further away from the structure, but still causing damage. Therefore, F. excelsior is not best suited close to permeable pavements. pavements; its average diameter was 34 cm and the corresponding mean distance from structures was 0.8 m. For F. excelsior, although three major damage incidents to permeable pavements were recorded, the average distance from the structure was 2.3 m and the mean diameter was 66.1 cm, indicating that these were mature trees located further away from the structure, but still causing damage. Therefore, F. excelsior is not best suited close to permeable pavements. The F. sylvatica assessed were mature trees with an average diameter of 68.1 cm and a mean distance of 0 m (i.e. touching the structures) from the building elements. 20 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 0 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 3.3.2. Permeable pavement On the other hand, using the metrics detailed in Table 1, B. pendula included in the analysis had not yet reached maturity. Their average DBH was 20.1 [(Fig._7)TD$FIG] Norway maple 63(1/10) Common ash 53(1/6) Wild cherry 45(2/14) Large-leaved lime 45(2/11) Common beech 89(1/3) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0% 20% 40% 60% 80% 100% Distance (m) from footpath Percentage of tree species close to footpaths causing damage Percentage of tree species close to footpaths causing damage Fig. 7. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from footpaths, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 7. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from footpaths, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. Fig. 7. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from footpaths, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. 20 [(Fig._8)TD$FIG] Norway maple 51(3/14) Sycamore 73(1/12) Small-leaved Lime 45(1/4) Silver birch 73(1/6) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0% 20% 40% 60% 80% 100% ll a w m o rf ) m ( e c n a t s i D Percentage of tree species close to walls causing damage Article No~e0 Article No~e00154 Percentage of tree species close to walls causing damage Fig. 8. Relationships of tree diameters at breast height (DBH (cm); represented by circles), average distances of trees away from walls, and the proportion of trees within 10 m to these structures subjected to moderate to severe damage. Note: x(y/z), where x represents DBH and z indicates the number of the tree species out of which y trees caused damage. cm and they were planted too close to the structures. 3.3.2. Permeable pavement Betula pendula of this DBH was estimated to be about 20 years (see above). 21 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 21 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 Article No~e00154 This study revealed that the pavements of Greater Manchester roads consist of more impermeable pavements than permeable pavements. Considering the findings of Morgenroth (2011) and Viswanathan et al. (2011), the Greater Manchester case is more likely linked to the phenomenon of insufficient moisture in the compacted soil strata below the impermeable pavements, and the tendency of roots to remain close to the surface for oxygen and moisture availability. Hence, this is the reason for greater damage to impermeable pavements than permeable pavements. This phenomenon seems common where there are more impermeable pavements than porous surfaces. Acer pseudoplatanus caused the most damage to impermeable pavements (78%) from an average distance of 1.3 m and an average DBH of 64 cm. 3.3.3. Impermeable pavement Concerning impermeable pavements (Table 2), the majority of the damage occurred to these pavement structures (44%), which was statistically significant (p < 0.05). The reason for this is that impermeable pavements do not allow free circulation of moisture and air into and out of the pavement surface (Randrup et al., 2003; Day et al., 2010; Scholz, 2013). Because of this, pockets of moisture build- up below the surface of impermeable surfaces, causing the roots of trees below the impermeable surface to be attracted to these pockets of moisture, and thereby lifting-up of the corresponding pavement surface. This may have accounted for the relatively high number of damage to impermeable pavements. For an impermeable pavement, the further away the tree (up to a distance of 1.4 m), the higher is the percentage of this tree causing damage irrespective of the tree DBH (Fig. 3). This is rather unexpected, considering that a large tree DBH is usually linked to large roots, which would cause damage at close range. However, the tree DBH variability is relatively small and the sample sizes are rather small as well. Regression analysis did not reveal any significant findings. Wherever tree roots are deprived of air and moisture, they start to grow back towards the surface to obtain these resources. Morgenroth (2011) studied root distribution in relation to paved and normal surfaces in the top 30 cm of soil. He found that root abundance in the top 30 cm is greater in impermeable pavements than in normal soil. 21 3.3.4. Kerb Kerb damage comprised 19% of all recorded structural damage. Acer platanoides caused the most damage to kerbs (10 out of 16 trees were located close to kerbs) from an average distance of 0.6 m and with a mean DBH of 41 cm (Fig. 4). Similar to A. platanoides was the impact of T. platyphyllos (7 out of 13 nearby T. platyphyllos), A. pseudoplatanus (2 out of 11 surrounding trees) and F. excelsior (7 out of 14 surrounding trees) caused damage to kerbs from the furthest average distance of 1 m (Fig. 4). Other trees that caused damage were less than 1 m from the kerb as shown in Fig. 4. Prunus avium was the best tree suitable for kerbs: only 1 in 14 trees caused moderate to severe damage (Fig. 4). However, most P. avium were very closely located (0 m) to kerbs, and their average DBH was 62 cm. This was closely followed by F. sylvatica. Although for F. sylvatica of an average DBH of 93 cm (indicating trees well-advanced in age) and an average distance of 0.23 m from kerbs, only 2 out of 9 trees caused moderate to severe damage to kerbs (Fig. 4). The worst tree to be located close to kerbs is A. platanoides. For trees of this species with an average DBH of 41 cm (indicating middle age) and located about 0.6 m from the kerbs, about 10 out of 16 A. platanoides caused moderate to severe damage to kerbs (Fig. 4). 22 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.4. Trees Acer platanoides occurred the most frequently (17%) among other trees that were found in this survey (Fig. 1). Furthermore, A. platanoides caused the most severe damage to structures (Table 2 and Table 3). The damage done to structures by A. platanoides did not follow any particular pattern. In this survey, 38 out of 73 (52%) A. platanoides caused damage to various structures (Table 2). About 35% of all A. platanoides planted close to permeable pavements with an average DBH of 56 cm and an average distance of 0.75 m from the permeable pavements caused severe to moderate damage to the pavement structures. This average DBH represents maturing A. platanoides. On average, A. platanoides caused more damage (42%) to impermeable pavements than to permeable pavements. These 42% of A. platanoides had an average DBH of 42 cm with an average distance of 0.3 m from impermeable pavements. This DBH represents A. platanoides, which are still in their relatively fast growth phase. This indicates that A. platanoides has a greater potential to cause more damage to impermeable pavements than to permeable pavements. About 60% of A. platanoides with an average DBH of 41 cm caused severe to moderate damage to kerbs from an average distance of 0.6 m. Acer platanoides of this DBH are still in the growing stage, indicating a future potential to cause more damage to kerbs. It follows that A. platanoides should not be recommended for planting near kerbs, as it is ranked the least suitable tree for planting close to kerbs (Table 4). Only 7% of A. platanoides with an average DBH of 50 cm caused severe and moderate damage to roads. Their average distance from roads was 2.0 m. Roads are normally well-compacted during construction to bear heavy traffic and haulage loads, and will therefore resist most damage from tree roots. Moreover, trees are normally at least 2.0 m located from roads, because of spaces for permeable or impermeable pavements and kerbs. Therefore, roads were linked to rather few damage incidents by tree roots. There were no records of severe and moderate damage to retaining walls by A. platanoides. About 20% of A. platanoides planted close to walls of average DBH of 51 cm caused severe and moderate damage to these wall structures. Those that caused damage were placed at an average distance of 1.0 m from the walls. About 10% of A. Article No~e00154 caused the most damage to retaining walls. Three out of four A. hippocastanum with an average DBH of 61 cm and located at a mean distance of 0.25 m away caused moderate to severe damage to retaining walls (Fig. 6). 23 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 3.3.5. Other structures The percentages of damage to roads and retaining walls were 3% each. Only three trees (F. excelsior, A. platanoides and A. pseudoplatanus) caused moderate to severe damage to roads (Fig. 5). Trees that caused damage to roads were located within an average distance of 2–5 m away from roads, indicating that the majority of them were planted close to the pavements. Not many trees were found close to retaining walls. For B. pendula with a DBH of 125 cm and planted at an average distance to structures of close to 0 m, only 1 out of 2 trees caused moderate to severe damage (Fig. 6). Aesculus hippocastanum 22 23 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 Article No~e00154 A. platanoides even at a distance of 3 m are possible, because the underlying soils at footpaths are not as compacted as those associated with other road structures. Despite that A. platanoides caused the most damage, and was also ranked the lowest in the potential for retrofitting (Table 5). Concerning T. platyphyllos, most trees that caused damage to urban structures (for example, impermeable pavements, retaining walls and footpaths) were very closely located to these structures compared to other trees (Fig. 4, Fig. 6, Fig. 7). Based on the survey, there was no record of severe to moderate damage by T. platyphyllos to some structures such as roads and walls. About 20% of the T. platyphyllos planted close to permeable pavements caused severe and moderate damage to permeable pavements from an average distance of 0.7 m. The average DBH of T. platyphyllos that caused damage were 34 cm. Tilia platyphyllos of this diameter were considered as still being in their growing stage (Table 1). The older these trees become, the more severe the damage would be. About 55% of the T. platyphyllos planted close to impermeable pavements caused severe to moderate damage to these structures. These trees were very close located to impermeable pavements as their average distance to the structures was 0 m at a mean DBH of 52 cm. About 25% of T. platyphyllos planted close to kerbs with an average DBH of 48 cm caused severe to moderate damage to kerb structures. Their average distance to the kerbs was 0.5 m. About 17% of T. platyphyllos planted close to retaining walls having an average DBH of 46 cm caused severe to moderate damage to these wall structures. Their average distance to the retaining walls was 0.0 m, indicating that they were very close (virtually touching) to these structures. Similarly, about 18% of T. platyphyllos planted close to footpaths having an average DBH of 46 cm caused severe to moderate damage to footpaths. Their average distance to footpaths was also 0.0 m. When assessing the damage to structures caused by T. platyphyllos with the relative importance of these structures, T. platyphyllos came second in terms of choice (Table 5). Furthermore, T. platyphyllos did not rank high in terms of future potential for damage (Table 6). Fraxinus excelsior caused severe to moderate damage to permeable pavements, impermeable pavements, kerbs, roads and retaining walls, but none to walls. 3.4. Trees platanoides close to footpaths with an average DBH of 63 cm caused severe and moderate damage to footpaths. Those that caused damage were at an average distance of 3 m to the footpaths. Damage to footpaths by 23 24 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Article No~e00154 About 70% of F. excelsior that were close to impermeable pavements caused severe to moderate damage to these pavement structures. The trees were of an average DBH of 30 cm and were located at a mean distance of 0.7 m from the impermeable pavements. Fraxinus excelsior trees of such DBH are considered to be young and developing, and are likely to cause more damage to any urban structures in the future. About 50% of the F. excelsior trees that were located closely to kerbs (average distance of 0.9 m) caused severe to moderate damage. Their average DBH was 62 cm. Most of these trees could be considered as mature. About 25% of F. excelsior close to roads with an average DBH of 72 cm caused severe to moderate damage to these road structures. They were located at an average distance of 2.0 m to the roads. About 50% of the F. excelsior found close to retaining walls with an a mean DBH of 53 cm caused severe to moderate damage to the retaining walls. They were placed at an average distance of 0.3 m from the retaining walls. Fraxinus excelsior had the highest average distance from the retaining walls amongst other trees that caused damage to retaining walls. The percentage of F. excelsior that caused damage to footpaths was the least among damage to other structures. The percentage of the trees that caused damage to footpaths was about 18% with a mean DBH of 53 cm and located an average distance of 0.5 m from the footpaths. Fraxinus excelsior ranked very high (8/10) in terms of potential for damage (Table 5), but ranked lower (5/10) in terms of potential for future damage. Most F. excelsior trees recorded in this survey were already mature, but reached less than half of their life span when compared with data shown in Table 1. Secondly, none of the F. excelsior trees were located very close to any structure. Fraxinus excelsior received average scores (51%) in terms of aesthetics in spring and summer, but very low scores (24%) for aesthetics in autumn. Acer pseudoplatanus caused damage to structures, even if planted at distances that could be considered as far away from structures such as permeable pavements, impermeable pavements, kerbs and roads. However, there were no recorded damage by A. pseudoplatanus to footpaths and retaining walls. The average diameter of A. Based on the results of this study, it can be inferred that the roots of F. excelsior can spread well beyond 2.0 m on the ground surface. About 35% of F. excelsior planted close to permeable pavements with an average DBH of 66 cm caused severe to moderate damage to permeable pavements from an average distance of 2.3 m. A F. excelsior tree of this DBH is considered to be fully grown (Dobrowolska et al., 2011). Fraxinus excelsior was the tree furthest away that caused damage to permeable pavements. This may be due to its great size. 24 25 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Because of the potential to cause damage even from a relatively far distance, A. pseudoplatanus ranked very high (9/10) in the potential for damage (Table 5), and also ranked very high (10/10) in the potential for future damage (Table 6). Because of the potential to cause damage even from a relatively far distance, A. pseudoplatanus ranked very high (9/10) in the potential for damage (Table 5), and also ranked very high (10/10) in the potential for future damage (Table 6). Because of the potential to cause damage even from a relatively far distance, A. pseudoplatanus ranked very high (9/10) in the potential for damage (Table 5), and also ranked very high (10/10) in the potential for future damage (Table 6). Because of the potential to cause damage even from a relatively far distance, A. pseudoplatanus ranked very high (9/10) in the potential for damage (Table 5), and also ranked very high (10/10) in the potential for future damage (Table 6). Prunus avium caused moderate to severe damage only to kerbs and footpaths. The corresponding damage to kerbs was the lowest (1/14 trees) among other trees. Prunus avium had an average DBH of 62 cm and were located very close (touching distance) to kerbs (Fig. 4). The number of P. avium that caused damage to footpaths was also very small (2/14). The DBH was 45 cm and the average distance from the footpaths was 2.5 m. Concerning future damage, P. avium ranked third (Table 5), indicating that it is one of the preferred tree species when considering damage to structures. For predicted future damage potentials, it is ranking first (Table 6), highlighting that the damage from P. avium are unlikely to get worse compared to other trees. Prunus avium also scored very high (72%) concerning aesthetics in spring and summer, but low (36%) in autumn. Aesculus hippocastanum caused moderate to severe damage to permeable pavements, impermeable pavements, kerbs and retaining walls, but none to roads, footpaths and walls. About 32% of A. hippocastanum were responsible for moderate to severe damage to these structures. Most A. hippocastanum that caused damage were mature in size with a mean DBH ranging from 51 to 71 cm (Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8). pseudoplatanus that caused damage to structures ranged from 52 cm for permeable pavements to 73 cm for both roads and walls. Findings indicated that 6 out of 12 A. pseudoplatanus (50%) with a mean DBH of 52 cm caused damage to permeable pavements at an average distance of 0.4 m (Fig. 2). Acer pseudoplatanus was the only tree that consistently caused damage from the furthest distance concerning kerbs, impermeable pavements and roads (Fig. 3, Fig. 4, Fig. 5). Acer pseudoplatanus was responsible for the most damage to impermeable pavements from the furthest average distance of 1.2 m with a mean DBH of 64 cm (Fig. 3). 25 26 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Betula pendula caused moderate to severe damage to permeable pavements, kerbs, walls and retaining walls, but no harm to impermeable pavements, roads and footpaths. The DBH for B. pendula that caused damage varied widely: 20 cm for those trees near permeable pavements, 45 cm for those near kerbs, 73 cm for those near walls, and 125 cm for those near the retaining walls. Most B. pendula that caused harm were very close to the structures they damaged, except for those close to walls, which were located at an average of 0.9 m away from trees. Due to B. pendula being able to cause damage even at small DBH, it ranked very high in the potential for structural damage both at presence and in the future (Table 5 and Table 6). Crataegus monogyna caused moderate to severe destruction to only impermeable pavements and retaining walls at an average DBH of 25 cm and 20 cm, respectively (Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8), indicating that these were still relatively small trees. However, 2 out of 4 C. monogyna caused moderate to severe harm to impermeable pavements from an average distance of 1 m, while 1 out of 3 trees caused damage to retaining walls from a mean distance of 0 m (Fig. 3 and Fig. 6). Crataegus monogyna ranked fifth in terms of potential for damage. This tree was located close to most structures, but damaged only two (Table 5). However, it ranked second in terms of potential for future damage (Table 6). The overall size of this tree may not increase significantly in the future due to its natural size, which is rather small compared to other trees such as A. pseudoplatanus (Table 1). Fagus sylvatica was linked to moderate and severe destruction to permeable pavements, impermeable pavements, kerbs and footpaths, but no damage to roads, walls and retaining walls. The average DBH of most F. sylvatica trees causing damage was relatively large, ranging from 68 to 93 cm (Fig. 2, Fig. 3, Fig. 4 and Fig. 8), indicating that they are already large and mature (Table 1). In all cases of harm to structures, F. sylvatica trees appeared to be the largest trees in terms of DBH wherever they featured (Fig. 2, Fig. 3, Fig. 4 and Fig. 8). Aesculus hippocastanum was ranked as the second (2/10) best tree with regard to damage to structures, and ranked fourth best in the potential for future damage, because most of the assessed trees were already mature. However, A. hippocastanum leaves generally lead to considerable volumes of leaf litter on streets during autumn. Tilia cordata caused moderate to severe damage to permeable pavements, impermeable pavements and walls, and no damage to kerbs, roads, retaining walls and footpaths (Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8). Distances of T. cordata to structures were generally within a mean distance of 0 m (as for impermeable pavements) to 0.7 m (as for permeable pavements). Most T. cordata that caused damage could be classed as still being very young, since their average DBH were between 26 to 38 cm, compared with those of 146–200 cm for a mature T. cordata tree (Fig. 2, Fig. 3, Fig. 8 and Table 1). Tilia cordata was responsible for damage already at young age (indicated by a small DBH). Therefore, this tree was considered to have a high potential to cause damage both in the present but particularly in the future. 26 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 27 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 (2) Tree species that occurred in less than five different sites were also discarded. (2) Tree species that occurred in less than five different sites were also discarded. (2) Tree species that occurred in less than five different sites were also discarded. (2) Tree species that occurred in less than five different sites were also discarded. (3) All damage classed as ‘light’ was also not included in the detailed analysis to reduce the likelihood of making trees responsible for damage when in fact other causes of damage are potentially also likely. Alternative reasons for damage might be natural settling of structures, fatigue of old constructions and physical damage linked to road accidents. (4) For the analysis of structural damage, only the structure types with at least ten damage reports linked to a specific tree species were considered to decrease the likelihood of high variability linked to small data sets to lead to spurious findings. Furthermore, not all trees were located in areas where they had the opportunity to influence all categorised structures in the same manner; e.g., some of the trees were located, for example, in parks with no major structures (e.g., buildings and roads) around, and that may have reduced the proportion of trees that have caused damage to these types of structures. Therefore, it was important to base the study on a relatively large data set collected at random and to implement the above criteria limiting the risk of spurious findings. 3.5. Study limitations The study has limitations due to the complex nature of the dynamic urban environment. In order to make sure that tree species, which have a good spread in Greater Manchester and that are typical for urban areas are well-reflected in this analysis, and also that recorded damage were actually caused by trees and not by other causes such as soil settlements due to unforeseen heavy traffic, the following criteria were applied: (1) Tree species that had less than 10 occurrences in total were not included in the analysis; 27 28 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Article No~e00154 deeper into the soil causing little or no damage to these structures. It follows that trees to be planted along streets in the future require more space and less compacted soil to reduce the risk of damage to nearby structures. Considering that the proportion of trees causing damage to infrastructure is always relatively small, the corresponding sample size per species is also rather small. This makes a statistical analysis rather challenging. Therefore, the authors recommend to undertake further studies on a much larger scale, and to focus only on a specific group of trees. Such studies should also assess tree trunk and root flare developments to provide a better understanding of root growth and development, particularly under structures such as pavements as well as the interaction of roots and trunk flare with pavements. No additional information is available for this paper. No additional information is available for this paper. No additional information is available for this paper. Competing interest statement The authors declare no conflict of interest. The authors declare no conflict of interest. Funding statement Vincent Uzomah was supported by The University of Salford. Vincent Uzomah was supported by The University of Salford. Author contribution statement Miklas Scholz: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper. Vincent Uzomah: Performed the experiments; Analyzed and interpreted the data; Wrote the paper. Furat Al-Faraj: Contributed reagents, materials, analysis tools or data. 29 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 4. Conclusions Considering the damage to structures by trees, obvious patterns may not have been demonstrated due the small sample size for a given species combined with the variability of the growing conditions for each species and site. Nevertheless, based on the Greater Manchester case study, the ‘best trees’ (in terms of relatively low risk to infrastructure) to be recommended for temperate and oceanic climates are T. platyphyllos, P. avium, C. monogyna, B. pendula, F. sylvatica, F. excelsior, A. pseudoplatanus and T. cordata. The project also concludes that impermeable pavements were subject to the highest number of damage from trees (44%), followed by permeable pavements and kerbs (22% and 19%, respectively). Trees planted close to impermeable pavements will cause more damage to the structure compared to those planted close to permeable pavements under the same conditions, which should be considered by town planners in the future. Other structural damage to roads, retaining walls and houses ranged from 0 to 5%. These rather low figures can be explained by the high compaction of the underlying media during their construction. Planners should consider that the more compacted underlying materials are, the greater is the likelihood that tree roots will spread close to the surface, and thereby damaging roads and SuDS structures. Roots of trees planted in not compacted underlying soil media, for example, in parks, fields and footpaths, did not spread along the ground surface, but went 28 References Arbor Day Foundation. (2015). Tree guide. http://www.arborday.org/trees/ treeGuide/browsetrees.cfm. (Accessed 16.04.2016). Biddle, P.G., 1998. Tree Root Damage to Buildings: Causes, Diagnosis and Remedy, vol. 1, Willowmead Publishing, Wantage. Biddle, P.G., 1998. Tree Root Damage to Buildings: Causes, Diagnosis and Remedy, vol. 1, Willowmead Publishing, Wantage. British Hardwood Tree Nursery. (2015). Tree life expectance. http://www. britishhardwood.co.uk/tree-life-expectancy/67. (Accessed 16.04.2016). 29 30 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Day, S.D., Wiseman, P.E., Dickinson, S.B., Harris, J.R., 2010. Tree root ecology in the urban environment and implications for a sustainable rhizosphere. Arboricul- ture Urban For. 36, 193–205. Defra. (2007). Hedgerow Survey Handbook. A standard procedure for local surveys in the UK. 2nd ed. http://www.hedgelink.org.uk/files/hedgerow-survey- handbook.pdf. (Accessed 16.04.2016). Dobrowolska, D., Hein, S., Oosterbaan, A., Wagner, S., Clark, J., Skovsgaard, J.P., 2011. A review of European ash (Fraxinus excelsior L.): implications for silviculture. Forestry 84, 133–148. Garden Centre. (2015). A guide to buying and growing trees and shrubs. http:// www.garden-centre.org/trees.htm. (Accessed 16.04.2016). Giuliani, F., Autelitano, F., Degiovanni, E., Montepara, A., 2015. DEM modelling analysis of tree growth in street pavements. Int. J. Pavement Eng., 1–10. doi:http:// dx.doi.org/10.1080/10298436.2015.1019495. Leuzinger, S., Vogt, R., Korner, C., 2010. Tree surface temperature in an urban environment. Agric. For. Meteorol. 150, 56–62. Mather, I., Morton, A., 2008. Street Tree Design Guidelines. Landcom, Parramatta. Morgenroth, J., 2011. Root growth response of Platanus orientalis to porous pavements. Arboriculture Urban For. 37, 45–50. Pliûra, A., Heuertz, M., 2003. Euforgen Technical Guidelines for Genetic Conservation and Use for Common Ash (Fraxinus excelsior). International Plant Genetic Resources Institute, Rome. Randrup, T.B., Mcpherson, E.G., Costello, L.R., 2003. A review of tree root conflicts with sidewalks, curbs, and roads. Urban Ecosyst. 5, 209–225. Sander, H., Polasky, S., Haight, R.G., 2010. The value of urban tree cover: a hedonic property price model in Ramsey and Dakota Counties, Minnesota, USA. Ecol. Econ. 69, 1646–1656. Scholz, M., 2010. Wetland Systems—storm Water Management Control. Springer Verlag, Berlin. Scholz, M., 2013. Water quality improvement performance of geotextiles within permeable paving systems: a critical review. Water 5, 462–479. Scholz, M., 2015. Wetlands for Water Pollution Control, 2nd edn Elsevier, Amsterdam. Scholz, M., Uzomah, V., 2013. Rapid decision support tool based on novel ecosystem service variables for retrofitting of permeable pavement systems in the presence of trees. Sci. Tot. Environ. 458–460, 486–498. 30 31 http://dx.doi.org/10.1016/j.heliyon.2016.e00154 2405-8440/© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Article No~e00154 Sydnor, T.D., Gamstetter, D., Nichols, J., Bishop, B., Favorite, J., Blazer, C., Turpin, L., 2000. Trees are not the root of sidewalk problems. J. Arboriculture 26, 20–29. Vernier. (2013). Using the goniometer to measure the height of trees. http://www. vernier.com/innovate/using-the-goniometer-to-measure-the-height-of-trees. (Accessed 16.04.2016). Viswanathan, B., Volder, A., Watson, W.T., Aitkenhead-Peterson, J.A., 2011. Impervious and pervious pavements increase soil CO2 concentrations and reduce root production of American sweetgum (Liquidambar styraciflua). Urban For. Urban Green 10, 133–139. 31
https://openalex.org/W4238014622	https://www.comparativepopulationstudies.de/index.php/CPoS/article/download/196/211	English	null	Effects of Family Structure and the Experience of Parental Separation: A Study on Adolescents’ Well-Being	Comparative population studies	2,015	cc-by-sa	14,269	Comparative Population Studies Vol. 40, 3 (2015): 335-364 (Date of release: 27.11.2015) Comparative Population Studies Vol. 40, 3 (2015): 335-364 (Date of release: 27.11.2015) 1 The terms “youth” and “adolescent” are used interchangeably. © Federal Institute for Population Research 2015 DOI: 10.12765/CPoS-2015-12en URL: www.comparativepopulationstudies.de URN: urn:nbn:de:bib-cpos-2015-12en3 Keywords: Parental separation · Divorce · Stepfamily · Adolescent well-being · Economic deprivation © Federal Institute for Population Research 2015 URL: www.comparativepopulationstudies.de DOI: 10.12765/CPoS-2015-12en URN: urn:nbn:de:bib-cpos-2015-12en3 1 Introduction Only few recent changes in family life have received as much public and scientiﬁ c attention as the increasing instability of marriage. Ever since the “Golden Age of Marriage” of the 1960s started to wane, divorce became a prominent issue not only among demographers and family sociologists, but also among developmental and clinical psychologists seeking to determine the impact of parental divorce for the children involved. Meanwhile, marital instability has increased considerably across many countries – as reﬂ ected in the crude divorce rate for Europe (EU-27), which doubled from 1.0 divorces per year per 1 000 inhabitants in 1970 to 2.0 divorces by 2010 (Eurostat 2013). At the same time, marriage rates declined, cohabitation increased, and a rising share of children is born to unmarried parents. In the U.S., the percentage of births to unmarried women increased relatively steadily since the 1960s, peaking in 2009 and levelling off since, with 40.6 percent in 2013 (Curtin et al. 2014). Other countries show similar trends, albeit at different levels. Given that un- married childbearing increasingly occurs in cohabiting unions and that such unmar- ried unions have a higher risk of breakup (e.g. Liefbroer/Dourleijn 2006), even when joint children are involved (Bastin et al. 2012; Schnor 2012), it is no longer only the instability of parents’ marriages, but also the instability of unmarried unions which has become an issue for child development. Not surprisingly, concerns about the impact of these changes on the next gen- eration are high. However, the focus and theoretical framing have changed over the years. While early studies from the 1950s, ‘60s and ‘70s (Hetherington/Stanley- Hagan 1999) on the effects of parental divorce were largely guided by a common deﬁ cit model assuming that children from divorced homes are per se worse off than children from nuclear families, these generalising assumptions proved to be inadequate. The initial methodological approach, which relied on simple compari- sons of children’s development in nuclear and separated (single parent) families, also turned out to be insufﬁ cient. As pointed out by Amato (2010: 661), “focusing on the average effects of divorce masks the substantial degree of variability that ex- ists in people’s adjustment”. Sabine Walper, Carolin Thönnissen, Philipp Alt Sabine Walper, Carolin Thönnissen, Philipp Alt Abstract: Large numbers of studies, mostly from the U.S., have addressed the effects of parental separation and divorce, pointing to disadvantages of children and adolescents growing up in separated families. However, evidence on this topic varies across countries and is limited for Germany. Using longitudinal data from waves 1 and 3 of the German Family Panel pairfam, we investigated differences in adolescents’ well-being by comparing stable nuclear families (n = 1968), single mother families (n = 360), and stepfather families (n = 214), as well as an additional smaller group of adolescents whose parents separated between waves 1 and 3 (“prospective separators”; n = 76). Adolescents’ satisfaction with different domains of life (family, education/work, and their general life satisfaction) as well as their self-esteem were used as indicators of well-being. A series of multiple regression analyses tested the effects of family structure on well-being at T1 and changes in well-being over time, controlling for various background factors. Furthermore, likely mediation effects of infrequent contact to the non-resident father and economic strain were tested. The ﬁ ndings show (relatively minor) effects of parental separa- tion, namely lower well-being among youth1 in single mother families compared to nuclear families. Disadvantages of youth in single mother families could only be partly explained by the higher ﬁ nancial strain generally experienced in these fami- lies. Youth in stepfather families reported a similar overall well-being as adolescents in nuclear families, but indicated a greater decrease in family satisfaction over time. Pre-separation disadvantages among prospective separators were limited to great- er dissatisfaction with school. Infrequent contact with the non-resident father did not affect adolescents’ well-being. Effects of family structure did not differ between boys and girls, but maternal education moderated the effects of family structure on adolescents’ life satisfaction. Overall, the ﬁ ndings are in line with other evidence from Germany, which points towards only limited disadvantages of youth in sepa- rated or divorced families. Keywords: Parental separation · Divorce · Stepfamily · Adolescent well-being · Economic deprivation URL: www.comparativepopulationstudies.de URN: urn:nbn:de:bib-cpos-2015-12en3 © Federal Institute for Population Research 2015 DOI: 10.12765/CPoS-2015-12en 336 • Sabine Walper, Carolin Thönnissen, Philipp Alt 336 2 Perspectives on the Effects of Divorce Research on the well-being of children and adolescents in divorced or separated homes has examined a wide variety of consequences, including emotional well-be- ing and mental health (Chase-Lansdale et al. 1995; Strohschein 2005), behavioural problems and delinquency (Burt et al. 2008; Fergusson et al. 1992), cognitive com- petencies (Sanz-De-Galdeano/Vuri 2007), academic achievement and educational attainment (Francesconi et al. 2010; Hilmert 2002), as well as life-course trajectories with respect to home-leaving, employment and earnings, partnership stability, and early childbearing (Cherlin et al. 1995; Ross/Mirowsky 1999). Given the many empir- ical studies which address differences in children’s well-being by family structure, several meta-analyses have sought to integrate available empirical evidence. In an update of his earlier meta-analysis (Amato/Keith 1991), Amato reviewed empirical evidence from studies mainly conducted in the U.S. during the 1990s (Amato 2001) and found consistent evidence for overall disadvantages among children from di- vorced families when compared to children from nuclear families, even though the effect sizes were rather small. Elaborating on this approach and looking at evidence from Europe, a recent meta-analysis covered 17 studies from Europe (Amato 2014), encompassing inves- tigations on behavioural and emotional problems (from Bulgaria, the Netherlands, Greece, Germany, and Norway), research on educational achievement (from Italy and Sweden), studies that have looked at health problems, substance use (alcohol and tobacco), and risky sexual behaviour (from Germany, France, Greece, the Neth- erlands, and Slovakia), two studies on delinquency (from Denmark) and one study on attachment security (from Germany). Almost all of these studies reported poorer outcomes for children with divorced parents than for children with continuously married parents. The weighted effect (accounting for differences in sample size) revealed a mean value of -.17 across all studies – a value which proved small and very similar to the overall effect sizes from American studies. Accordingly, Amato (2014: 15) concluded that “irrespective of national and cultural characteristics, the gap between children with divorced and continuously married parents is about the same on both sides of the Atlantic.” At the same time, there is considerable variability in ﬁ ndings across studies and countries (Amato/James 2010). While some authors stress the dramatic consequenc- es which parental breakup may have for the offspring’s well-being (Wallerstein et al. 1988), others highlight the coping potential in divorced families (Hetherington/Kelly 2002). 1 Introduction Over the past decades, large numbers of studies have contributed to a reﬁ ned understanding of relevant conditions that may – or may not – occur in the context of divorce and may contribute to or ameliorate the many chal- lenges in children’s and adolescents’ coping with parental breakup (for an overview see Amato 2010). Accordingly, understanding the effects of parental breakup has become a complex enterprise. – occur in the context of divorce and may contribute to or ameliorate the many chal- lenges in children’s and adolescents’ coping with parental breakup (for an overview see Amato 2010). Accordingly, understanding the effects of parental breakup has become a complex enterprise. This paper starts by providing an overview of relevant perspectives, models, and ﬁ ndings focusing on several key issues: the role of economic deprivation in linking family structure to child development, the inﬂ uence of children’s contact with the non-resident parent on children’s well-being, effects of interparental conﬂ ict, the inﬂ uence of pre-separation stress in families which eventually separate, and costs and beneﬁ ts of stepfamily formation. The majority of such ﬁ ndings come from the U.S., but we will also look into evidence from Europe, particularly from Germany. The empirical analyses presented here investigate adolescents’ well-being in differ- ent family types and address the mediating role of ﬁ nancial hardship and reduced 337 Effects of Parental Separation • 337 contact with the non-resident father. We focus particularly on the experience of parental separation in mid- to late adolescence using prospective data from the German family panel pairfam, through which we can address possible strains prior to and following parental separation. 2 Perspectives on the Effects of Divorce Such divergent views also ﬁ nd support in evidence from Germany: several studies based on large samples did not ﬁ nd higher depressiveness, impaired self- esteem, more problems in peer relations, or increased behaviour problems among Sabine Walper, Carolin Thönnissen, Philipp Alt 338 children and adolescents from separated single parent families compared to nucle- ar families (Walper 2002; Walper/Wendt 2005; Wendt/Walper 2007). Furthermore, a recent study on the effects of parental separation on young children whose parents cohabited, but were not married when the children were born, found no evidence for increased emotional or behavioural problems among those who experienced parental separation (Walper/Langmeyer 2014). Similarly, the large-scale PISA as- sessment of academic competencies among 15-year-old adolescents did not sug- gest any differences between youth from single parent and two-parent families in Germany, whereas marked differences were found in other countries, such as the U.S. (Ehmke et al. 2004). However, other research conducted in Germany reported increased emotional and behavioural problems for up to two years after parental separation (Schmidt-Denter 2000), an increased likelihood of insecure attachment representation for boys in single parent families (Gloger-Tippelt/König 2007), and – in a clinical sample – more instances of conduct disorder among youth raised in separated single parent families (Steinhausen et al. 1987). In seeking to account for the variability of ﬁ ndings across as well as within sam- ples and to better understand the origin of differences between children from sepa- rated and nuclear families, researchers have long pointed out that divorce is not a uniform single event, but typically comes along with other stressors triggered by union dissolution (Amato 2000; Hetherington/Stanley-Hagan 1999). This divorce- stress-adjustment perspective views disadvantages among offspring from sepa- rated families as caused by divorce-related stressors and differences in coping re- sources. In contrast, the selection perspective cautions against overestimating the causal role of separation and disadvantages and points towards common causes for parental separation and offspring strain (Amato 2000). Both perspectives shall be shortly reviewed. 2.1 Parental Separation as Stressor The divorce-stress-adjustment perspective (Amato 2000) points out that parental breakup is neither a uniform nor a single event, but instead is likely to be embedded in a longer series of events and stressors which partly precede but mostly follow separation or divorce. Such stressors differ in prevalence and strength, and they occur in the context of varying coping resources. Both factors – divorce-triggered stressors and the available coping resources – jointly shape the individual experi- ence and outcomes of parental separation. Prominent divorce-related stressors are (1) economic problems which arise when available ﬁ nancial resources have to cover the expenses of two households, (2) reduced contact to the non-resident parent, and (3) conﬂ ict between parents (e.g. about child custody and child-support payments). Findings from many countries have shown that single parenthood is linked to an increased risk of income poverty and downward mobility (Amato 2010), even though effects of family structure on in- come vary by family policy and welfare programmes, with a weaker effect in Scan- dinavian countries than in the U.K. and the U.S. (Garriga/Härkönen 2009). Poverty, in turn, has been shown to increase the risk for compromised child development in 339 Effects of Parental Separation • 339 terms of health, education, emotional well-being, social relations, and behavioural problems (Conger et al. 2000; Duncan/Brooks-Gunn 1997; Walper 2009). In line with the economic strain hypothesis, several ﬁ ndings suggest that a large share of the disadvantages found for children in single parent families can be explained by the increased risk of inadequate ﬁ nancial resources (see also Amato 2010; McLanahan 1999), at least in countries where divorce typically leads to ﬁ nancial hardship (Gar- riga/Härkönen 2009). Support has also been provided for the interparental conﬂ ict hypothesis, as chil- dren’s increased emotional and behavioural problems in divorced families could be largely accounted for by the higher prevalence of intense conﬂ ict between separat- ed parents (e.g. Schick 2002). Similarly, looking at subgroups of children with differ- ent developmental trajectories after parental separation, more negative trajectories were linked to more severe problems in the post-divorce relationship (Schmidt- Denter 2001). 2.1 Parental Separation as Stressor involv- ing a large sample of adolescents from separated families even found that the link between fathers’ engagement and adolescent behaviour had to be interpreted in the opposite direction: Fathers’ engagement did not affect their child’s subsequent Sabine Walper, Carolin Thönnissen, Philipp Alt 340 well-being, but an increase in adolescents behavioural problems led to a decrease in paternal involvement in the subsequent year (Hawkins et al. 2007). While the ef- fects of contact frequency on children’s well-being seem to be less reliable, contact does affect the chances of establishing a close relationship with the non-resident parent (Walper/Krey 2009; Whiteside/Becker 2000). p p y Finally, union dissolution opens the options for parents to repartner and remarry, thus potentially triggering additional changes in children’s family composition and parenting experiences. In Germany, about eleven percent of all minors are estimat- ed to live with a stepparent in their primary household (Steinbach 2008). Further- more, in about 5 percent of all nuclear and 17 percent of stepfamilies, at least one parent has one or more children outside of the household, suggesting a secondary or “weekend” stepfamily constellation. Although parents usually beneﬁ t from the support of a new partner (Schmidt-Denter 2001), the advantages of such changes are less clear for children (Coleman et al. 2000; Sweeney 2010). Typically, they not only have to adapt to a new person in the household, who is likely to claim inﬂ u- ence on decisions and may be inexperienced in child rearing, but they also lose some of the time and attention of their biological parent. Adolescence has spe- ciﬁ cally been pointed out as a difﬁ cult phase in stepfamily life, and some studies suggest more negative ﬁ ndings for girl’s adjustment in stepfamilies (Coleman et al. 2000; Hetherington/Jodl 1994). Integrating ﬁ ndings from 61 studies, a meta-analysis found that stepchildren generally fare worse than children in nuclear families (with both biological parents) regarding academic achievement and psychological well- being (Jeynes 2006). When compared to children in single parent families, however, stepchildren did not show any advantages. They did not differ in psychological well- being and showed lower scholastic achievement than youth from divorced single parent families. 2.1 Parental Separation as Stressor In general, for nuclear as well as for separated families, interparental conﬂ ict has been pointed out as a major stressor for children, because it is likely to spill over into parenting (Buehler/Gerard 2002; Erel/Burman 1995), contributes to loyalty conﬂ icts, and undermines children’s emotional security in the family system (Davies et al. 2002; Walper/Beckh 2006). However, since post-divorce conﬂ ict may have its roots in pre-separation marital problems, it is less clear whether increased interparental tension and conﬂ ict in the post-separation period can be interpreted as resulting from parental breakup, or whether it actually preceded the parents’ decision to separate. In general, it is quite likely that conﬂ ict was already present in the pre-divorce period. Accordingly, interparental conﬂ ict is also a factor addressed by the selection perspective. In contrast, parental separation is very likely to affect a child’s relation to the non-resident parent, at least in terms of frequency of contact. Many studies ﬁ nd that contact with the non-resident parent – usually the father – diminishes as time goes by, especially if the father lacks the economic resources to provide for the children and if the interparental relation is strained (Amendt 2004). A decrease in contact with the non-resident parent can also be observed if the children were very young at the time of parental breakup and once they reach adolescence and invest more time in peer relations (see Walper/Krey 2009). In legal decision-making, counsel- ling, and public debate, considerable concern relates to such effects, because it is commonly held that children thrive on a continuous relation to both parents despite parental breakup. Interestingly, however, available evidence does not suggest that the frequency of contact is a crucial variable for well-being (Amato 1993, 2010). In a meta-analysis of links between child well-being and several aspect of non-resident fathers’ involvement, almost no effects of contact frequency were found, whereas the quality of fathers’ parenting had the strongest effect (Amato/Gilbreth 1999). Fur- thermore, a study from Germany on adolescents who never had contact to their separated father evidenced no differences in well-being compared to adolescents in nuclear families (Walper/Wendt 2011). A longitudinal study from the U.S. 2.2 The selection perspective While the divorce-stress-adjustment perspective highlights the causal role of di- vorce-related stressors in explaining disadvantages of children whose parents separated, the selection perspective addresses possible differences between sta- ble and disrupted families preceding parental breakup, which may affect not only the likelihood for separation or divorce but also for children’s compromised de- velopment. Likely factors are parents’ personality problems, low socio-economic resources, problems in the parents’ relationship, but also genetic factors that play a role in social behaviour (Amato 2000; Garriga/Härkönen 2009). Several approaches have been developed to address issues of selectivity (see Amato 2010; Garriga/Harkönen 2009). Taking into account pre-disruption differ- ences in family characteristics and child well-being is one of the main methods for estimating selection effects. In fact, much support for the selection perspec- tive came from prospective longitudinal studies which were able to trace family characteristics and child development across time prior to parental separation (e.g. Block et al. 1986; Cherlin et al. 1991). Many of these ﬁ ndings suggested that disad- vantages in child development can be observed well ahead of their parents’ separa- 341 Effects of Parental Separation • 341 tion. Interestingly, boys seemed to be more strongly affected in the pre-separation period than girls. In the study by Cherlin et al. (1991), effects of parental breakup were considerably reduced once such pre-separation differences were taken into account. In some cases, the effects of parental separation disappear when control- ling for children’s pre-divorce well-being and competencies. For example, a study based on data from the National Education Longitudinal Study observed teenagers from divorced families before and after the divorce and compared them to youth from intact homes (Sanz-De-Galdeano/Vuri 2007). The authors did not ﬁ nd that pa- rental divorce negatively affected adolescents’ cognitive skills when controlling for pre-divorce competencies. However, selection effects such as these are not consistently found in all stud- ies on the topic (e.g. Allison/Furstenberg 1989; Forehand et al. 1997). In a study on urban youth in Germany, only limited pre-separation differences were found when comparing adolescents from stable and instable families one year prior to parental separation, in the year of parental breakup, and one year later (Schwarz 1999). Pre- separation disadvantages were restricted to adolescents’ relationships to parents and peers, but were not found for their transgression proneness (positive attitudes to delinquency) or their self-derogation. 2.3 Differential effects across contexts and child characteristics Some attention has been paid to the question of whether the negative consequenc- es of divorce fade as divorce becomes more prevalent and a more common ex- perience in children’s lives. Such an increased “institutionalisation“ of divorce is likely to reduce the risk of stigmatisation and may facilitate successful coping with the demands of parental breakup given that many peers share the same experi- ence. However, comparing the disadvantages of children from divorced families across time, the meta-analysis by Amato (2001) does not provide evidence of a linear trend. Although effects in the 1980s were less marked than in older studies conducted up until the 1970s, more recent ﬁ ndings from the 1990s show an increase in disadvantages for children from divorced families. Since most of these studies were conducted in the U.S., where social beneﬁ ts for single parent families had been cut back during the 1990s, such changes in policy have been suggested as a likely explanation (see Amato 2001). Findings from other countries point towards similar evidence. A large multi-country study based on data from the Generations and Gender Survey provided no support for the „institutionalisation hypothesis“, which would have predicted less disadvantages for divorced families where divorce is highly prevalent. Instead, these data indicated larger effects of parental separa- tion on the offspring’s educational attainment in countries and at times with higher divorce rates (Bernardi/Radl 2014). Such ﬁ ndings also raise issues of self-selection into divorce. When divorces are rare, the threshold to separate is high. Hence, couples with highly dysfunctional relationships are more likely to select into the group of those who opt for divorce. Since many ﬁ ndings show that children’s well-being is undermined by frequent and intensive interparental conﬂ ict (Fincham 1998; Krishnakumar/Buehler 2000), one might expect more negative consequences of parental breakups for children in such highly stressed families. However, prospective studies suggest otherwise (see Amato 2010): when tracing the effects of divorce for families with more or less problems in the parents’ pre-divorce relationship, it rather seems that children from high-conﬂ ict families have less to loose and more to gain if their parents separate. In contrast, offspring from previously unstressed marriages showed considerable disadvantages in their adult life (e.g. Amato et al. 1995). Many studies have addressed the issue of children’s vulnerabilities, examining age and gender as likely moderators. 2.2 The selection perspective Other data suggest a combination of selec- tion effects as well as additional stresses caused by parental separation (Cherlin et al. 1998; Cherlin et al. 1995; Sigle-Rushton et al. 2005). For example, a 14-year longitudinal study which compared children of unmarried or divorced mothers to children with married mothers drew from information about the mothers’ devel- opment during their adolescent years, long before the children were born (Emery et al. 1999). As expected, children from single mother families showed increased problematic behaviour, which could partly be explained by the higher delinquency of these mothers during their adolescent years. It is quite likely that mothers‘ ear- lier problem behaviour prompts risky choices in mate selection and contributes to problems with the child’s father, thus triggering partnership instability at a later stage. Furthermore, problems in self-regulation or anti-social traits may undermine mothers’ parenting competencies, thus reducing the chances for positive child de- velopment. At the same time, parental separation seems to have caused additional problems in child development. Given such direct similarities in parents’ and children’s behaviour, addressing genetic issues seems to be a logical next step, even more so since some evidence points to genetic inﬂ uences on the likelihood to divorce (see Amato 2010; Garriga/ Harkonen 2009). However, when comparing the effects of parental divorce in bio- logical and adoptive families, child outcomes in terms of problem behaviour and substance use were similar for biological and adopted children (e.g. Amato/Cheadle 2008; O’Connor et al. 2000). Accordingly, it is unlikely that these effects of parental separation can simply be attributed to genetic selectivity. Overall, there is evidence for selection effects which often, but not consistently, lead to increased problems among children even prior to parental separation. In general, these ﬁ ndings are not at odds with the divorce-stress-adjustment perspec- tive, even more so since increased strain among youth shortly before their parents’ separation could still be seen as part of the separation process as a whole. This 342 Sabine Walper, Carolin Thönnissen, Philipp Alt phase of decision making is often particularly stressful for the adults involved. Thus, selection and socialisation may jointly be at work. 2.3 Differential effects across contexts and child characteristics Although some ﬁ ndings are in line with the wide-spread notion that younger children are more dependent on a well-function- ing family and thus more vulnerable to family disruption (Allison/Furstenberg 1989; Emery 1988), other researchers have suggested that it is adolescent children who are particularly vulnerable to parental breakups (Chase-Lansdale et al. 1995). The majority of ﬁ ndings do not suggest differential effects by children’s age (Amato 2001). Similarly, gender differences in children’s coping with parental breakup are not very clear. In line with evidence suggesting that boys are more vulnerable to 343 Effects of Parental Separation • 343 family stress than girls (e.g. Gloger-Tippelt/König 2007), pre-separation distress was found to be higher among boys whose parents break up during childhood (Block et al. 1986; Cherlin et al. 1991). In general, however, divorce does not seem to affect boys and girls differently (e.g. Fergusson et al. 1994). For the adolescent years, the evidence is even less consistent. Since divorce has been found to be more prevalent among families with low so- cio-economic resources, studies usually control for parental education, occupation, or a SES (socioeconomic status) composite when estimating the effects of divorce. However, little attention has been paid to differential effects of parental separation depending on parents’ SES resources. In a recent study using data from the Genera- tions and Gender Survey (covering 14 countries), Bernardi and Radl (2014) investi- gated the long-term consequences of parental separation for children’s educational attainment, focusing on the likelihood of achieving tertiary education. Comparing the effects of parental breakup for families with different levels of parental educa- tion, they found no evidence for the “social origin compensation hypothesis“, which claims that parents with higher SES resources are better able to cope with divorce and protect their children from negative consequences. Rather, their data provided clear support of the „ﬂ oor effect“, i.e. for weaker effects of parental breakup among families with low parental education whose children face substantially lower chanc- es for tertiary education anyway. Interestingly, this „ﬂ oor effect“ was not observed in countries with early tracking, i.e. early performance-based selection into differ- ent school tracks (like in Germany). In these countries, offspring from families with higher as well as lower parental education experienced markedly lower chances for tertiary education when their parents separated. 2.3 Differential effects across contexts and child characteristics In the following study, we sought to follow up on these issues by looking at various aspects of adolescents’ well- being, including their satisfaction with their domain of education and work. 2.4 Research Questions In the following analyses, we compare adolescents’ well-being in different family structures over a 2-year time period. We employ a broad concept of parental sepa- ration, including all cases with both parents alive but not living in a union, irrespec- tive of whether they have been married in the past or not or whether they are still formally married but separated and no longer live in an intact partnership. When testing possible disadvantages of adolescents in stable single parent households and stepfamilies as opposed to nuclear families, we control for important back- ground factors and explore whether both types of separated families are linked to similar disadvantages (separation disadvantage hypothesis) or whether either of these two types of separated families is less likely to promote adolescent well-being (single mother or stepfamily strain hypothesis). Although single parent families are more likely to suffer from ﬁ nancial hardship and may thus provide a more stressful environment for youth development, available evidence does not suggest advan- tages of stepfamilies compared to single parent families for children’s well-being (Jeynes 2006). Hence, we expect similar disadvantages for youth from both types of separated families when compared to adolescents from nuclear families. Further- • Sabine Walper, Carolin Thönnissen, Philipp Alt 4 344 more, we seek to explore effects of a recent parental separation employing a pro- spective design. Based on previous ﬁ ndings from prospective studies, we expect to ﬁ nd lower well-being among adolescents prior to parental separation than among adolescents from nuclear families (pre-separation strain hypotheses). Secondly, we investigate whether changes in outcomes differ across time by fam- ily type. In this respect, we expect a reduction in adolescents’ well-being after pa- rental separation among those whose parents separated between both waves, while controlling for pre-separation well-being and background factors (post-separation strain hypothesis). These analyses also explore whether other types of separated families show different trends in adolescents’ well-being than nuclear families do. Thirdly, we ask whether infrequent contact to the non-residential parent and/or economic deprivation explain any effects of family type on changes in well-being across time. Such evidence would suggest that contact and/or economic resources function as mediating links between family structure and adolescent outcomes. While public discourse gives major importance to the amount of contact children have to their non-residential parent, research on the topic has not found substantial effects of such contact on children’s and adolescents’ well-being. 2.4 Research Questions Hence, we in- clude this factor, but do not expect it to explain differences in adolescents’ well-be- ing when comparing youth from separated families to those from nuclear families. In contrast, economic deprivation has been pointed out as an important mediator for explaining disadvantages of youth in single mother families, who are most likely to be hit by ﬁ nancial strain. We expect to replicate this ﬁ nding (economic strain hypothesis). Finally, we test whether adolescents’ gender and parental education moderate the link between family structure and adolescent well-being. With respect to paren- tal education, we expect to ﬁ nd more negative effects of a recent parental separa- tion, single motherhood, and stepfamily life in the less educated group (resource hypothesis). Although previous ﬁ ndings have not pointed to the signiﬁ cance of a child’s gender as a factor that moderates the effects of parental separation, pro- spective studies have suggested that boys may be more strained prior to parental separation. Thus, we test such moderation effects for T1 differences between family types as well as for T3 changes in adolescents’ well-being in different family types. 3.1 Sample The paper’s sample was drawn from the German Family Panel pairfam (Huinink et al. 2011), a three-cohort longitudinal study on family development with annual as- sessments, which started in 2008/2009 for three birth cohorts. Participants were recruited through register data and personal visits of the interviewer, who conduct- ed the interviews in the participants’ home. For our purposes, the sample was re- stricted to the adolescent cohort (born 1991–1993). Adolescents were mostly 15 to 17 years old during wave 1, with only very few younger or older participants. Of the 345 Effects of Parental Separation • 345 initial adolescent cohort in wave 1 (n = 4334), 72.3 percent were re-interviewed two years later for wave 3 (n = 3132). Our analysis are based on the data from the ﬁ rst three waves of pairfam, release 4.0 (Nauck et al. 2013). Data on household composition in waves 1 and 3 was available for 3128 adoles- cents. The family type was identiﬁ ed by the adolescents’ report on whether both parents currently lived in a joint household and the adolescents’ household com- position at their main residence. In a ﬁ rst step, the sample was restricted to youth whose parents were both still alive (n = 3019) and who lived with at least one parent in waves 1 and 3 (n = 2760). This second criterion excluded youth who had left the parental home prior to wave 3 (n = 259). As might be expected, the likelihood of having left the parental household differed for youth from higher- and lower-track schools. Adolescents who had already left the parental home by wave 3 were less likely to be on the higher school track than those who still stayed with their parents (35.1 percent vs. 44.6 percent; Chi2 = 8.69; df = 2; p<.01). In a second step, we se- lected family types of sufﬁ cient group size. Adolescents who lived with a single father (n = 57) or with their father and his new partner/a stepmother (n = 27) in waves 1 and 3 could not be considered due to their low prevalence in the sample. For similar reasons, newly formed or separated stepfather families (n = 27 and 31) were not included in the analyses. In addition, we excluded all cases in which no information about either maternal or paternal education was provided (5.1 percent of all cases; n = 160). 3.1 Sample Based on data for waves 1 and 3, we categorised all remaining cases as stable nuclear families (with both biological parents in adolescents’ household, n = 1968), stable single mother families (biological mother, without a partner in the household; n = 360), stable stepfather families (biological mother and her partner in the house- hold; n = 214), or prospective separators, i.e. families who started as nuclear fami- lies in wave 1 but experienced parental separation between waves 1 and 3 (n = 76). In total, n = 2618 adolescents who participated in waves 1 and 3 were available, among these up to 2483 had complete data for our analyses. As analyses regarding selective drop-out across time revealed, longitudinal drop-out was not randomly distributed between nuclear and separated families (Chi2 = 13.44; df = 1; p < .01). Adolescents from separated families were more likely to drop out by wave 3 than adolescents from nuclear families were (32.5 percent vs. 26.8 percent). Table 1 provides an overview of the sample, subdivided by family type as well as providing totals. As shown in the last column, adolescents’ average age during wave 1 was 15.92 years (SD = .88), and about half of the sample was female. Almost half of all adolescents were currently enrolled or had already ﬁ nished grammar school (“Gymnasium”), the highest school track which – if successfully ﬁ nished – provides access to university studies. The share of families with high maternal education (university or advanced technical college entrance qualiﬁ cation) was about a third. One-ﬁ fth of the adolescents had a migration background, i.e. they and/or at least one of their parents were not born in Germany. In line with population statistics, only a minority (18.0 percent) of participants lived in eastern Germany. To test for signiﬁ cant differences by gender, maternal education, migration back- ground, and region of residence between the subgroups deﬁ ned by family type, • Sabine Walper, Carolin Thönnissen, Philipp Alt 346 Tab. 3.1 Sample 1: Descriptives of the Sample by Family Type Nuclear Families Prospective Separators* Single Mother Families Stepfather Families Total N 1968 76 360 214 2618 Age at T1 M 15.90 15.91 15.97 15.95 15.92 SD .87 .80 .89 .94 .88 Gender (% Female) 47.4 53.9 47.2 46.7 47.5 School Track (% Grammar School) 48.2 42.7 33.9 35.5 45.1 High Maternal Education (% yes) 32.9 27.6 27.2 26.6 32.5 Migration Background (% yes) 21.3 11.8 23.3 11.2 20.8 Region (% East) 16.9 22.4 18.3 26.2 18.0 Economic Deprivation T2/3 M 2.10 2.55 2.91 2.32 2.24 SD .87 .92 1.01 .90 .93 % With Contact to Non-Resident Father at T3 -- 97.3 77.7 76.7 79.7 * Adolescents whose parents separated between waves 1 and 3 Source: pairfam data release 4.0 Tab. 1: Descriptives of the Sample by Family Type Chi2-Tests were carried out. The share of adolescents in grammar schools proved signiﬁ cantly lower in single mother families and stepfather families than in nuclear families (Chi2 = 34.28; df = 3; p < .01). A similar pattern was found for maternal education. The share of high maternal education (graduation from grammar school) proved signiﬁ cantly lower in single mother families and stepfather families than in nuclear families (Chi2 = 8.03; df = 3; p < .05). Signiﬁ cant differences were also found by region (Chi2 = 12.39; df = 3; p < .01): Adolescents in stepfather families were more likely to live in eastern Germany, while regional distribution did not differ for the other family types. Furthermore, migration background differed signiﬁ cantly across family types (Chi2 = 15.42; df = 3; p < .01): Prospective separators and ado- lescents in stepfather families were less likely to have a migration background than youth in single mother families. In order to inspect the ﬁ nancial situation of the family, we decided not to rely on household income, which was only available as reported by adolescents. Instead, we used a presumably more valid indicator: a two-item-scale which was assessed in wave 3 and measured ﬁ nancial strain in the household (see indicators below). Analyses of variance comparing the four family types revealed highly signiﬁ cant effects of family type (F = 85.37; df = 3; p < .001): Economic deprivation was sig- niﬁ cantly lower in nuclear families than in all types of separated families. 3.1 Sample According to post-hoc tests, youth from single mother families reported signiﬁ cantly higher economic deprivation than youth from stepfather and prospective separation fami- lies. Stepfather families and prospective separators did not differ signiﬁ cantly. Effects of Parental Separation • 347 Effects of Parental Separation • 347 Finally, we compared the share of adolescents’ who still had contact to their non-residential father in wave 3 and found signiﬁ cant differences between the three types of separated families (Chi2 = 16.11; df = 2; p < .001). As might be expected, youth in recently separated families (prospective separators) were less likely to have lost contact to their biological fathers at T3 than youth in stable single mother and stable stepfather families. No signiﬁ cant differences regarding age or gender were found. 3.2 Indicators Economic Deprivation. To assess Economic Deprivation in the household, two items from a larger scale were used in wave 3 indicating the inadequacy of the family’s budget for ordinary living expenses (e.g. “My parents often have to forego some- thing because they have to watch their budget”). The indicator was computed as the mean of both items. More speciﬁ c information on this, as well as all other indicators used in our analyses is presented in table 2. Contact to Father. For separated families, information on the frequency of con- tact to the non-residential father was assessed by a single question: “How often are you in contact with your father, adding up all visits, letters, phone calls, etc.?” For the purpose of our analyses, we focused on the effects of infrequent contact and recoded the categories into a dichotomous variable combining the categories “con- Tab. 2: Description of Indicators Variable Number of items Source Response format Cron- bach’s Alpha Mean (standard deviation) Wave 1 Wave 3 Economic Deprivation 2 Schwarz et al. (1997) 1 = not at all correct to 5 = completely correct .89 (W3) 2.24 (.93) Contact to Father 1 pairfam 1 = daily to 7 = never 10 = contact never existed - 1.87 (1.75) Satisfaction with Family 1 pairfam 0 = very dissatisﬁ ed to 10 = very satisﬁ ed 8.83 (1.54) 8.55 (1.58) Satisfaction with School, Education, Career 1 pairfam 0 = very dissatisﬁ ed to 10 = very satisﬁ ed 7.36 (1.85) 7.42 (1.98) Overall Life Satisfaction 1 pairfam 0 = very dissatisﬁ ed to 10 = very satisﬁ ed 7.99 (1.47) 7.90 (1.45) Self-Esteem 3 Adapted from Rosenberg (1965) 1 = not at all to 5 = absolutely .78 (W3) 4.13 (.74) 3.98 (.82) Sabine Walper, Carolin Thönnissen, Philipp Alt 348 tact never existed”, ”never” and “less than several times per year” into the category infrequent (no or rare) contact (n = 186; coded as 1), while the other categories (from “several times per year” up to “daily”) served as reference category indicating more frequent contact (n = 2407; coded as 0). Although this coding implies a loss of information about variations in frequency of contact, it provided a variable which was less confounded with family structure than the original categories. Adolescents’ Well-Being. To assess adolescents’ well-being in waves 1 and 3, we relied on indicators which were similarly measured in both waves. 3.2 Indicators Three single- item indicators gave information about adolescents’ satisfaction with their family, school/work, and life in general. The items were introduced with the question “How satisﬁ ed are you with the following domains of your life?”. The domains were named “Family” and “School, Education, Career”. Overall life satisfaction was assessed with the question “All in all, how satisﬁ ed are you with your life at the moment?”. Furthermore, a 3-item-indicator of self-esteem was included in both waves. Since the assessment mode of self-esteem was changed between waves 1 and 2 from computer-assisted personal interviews (wave 1: CAPI) to computer-assisted self-completed interview (waves 2 and later waves: CASI), the scale had to be cor- rected for mean-level differences which emerged between wave 1 and later waves (with higher levels of self-esteem reported in the CAPI than in the CASI interviews). In order to correct for mean level differences without changing the rank order of participants at each wave, the scale was z-standardised for the entire sample of each wave. Due to the loss of sample size for the present analyses (with a higher loss in wave 1 than in wave 3), the distribution of the self-esteem scales does not completely match the original means and standard deviations. More detailed in- formation on all indicators used in the present analyses can be found in the scale manual for the pairfam project report, available online (Thönnissen et al. 2014). 3.3 Analyses Multiple regression analyses were used to test effects of family structure on ado- lescents’ well-being at waves 1 and 3 (T1 and T3), controlling for adolescents’ age, gender, maternal education, migration background, and region of residence. Fam- ily type was dummy-coded using nuclear families as the reference category for comparisons with recent separators, single mother families, and stepfather families (each coded as a dichotomous dummy variable). For each dependent variable, ﬁ ve analyses were conducted: (1) In model 1, data for T1 inform about mean differences between youth from dif- ferent family types, with a special focus on possible pre-separation strain among those whose parents later separated in the following two years (between T1 and T3). These ﬁrst analyses control for background variables only. (2) In model 2, we added interaction effects between family type and gender or ma- ternal education to test whether pre-separation well-being among prospective separators was more compromised for boys or youth from families with lower maternal education than their respective counterparts. In addition, these analy- ses examined similarities and differences in the effects of single motherhood 349 Effects of Parental Separation • 349 and stepfamily life between boys and girls and between families with higher or lower educational resources. (3) The third model used data for T3 as dependent variables to assess changes in well-being between both waves, controlling for the stability of the respective outcome variable across time. It tested effects of family type, controlling for the background variables as well as for effects of the outcome variable at T1. (3) The third model used data for T3 as dependent variables to assess changes in well-being between both waves, controlling for the stability of the respective outcome variable across time. It tested effects of family type, controlling for the background variables as well as for effects of the outcome variable at T1. (4) Model 4 added the effects of infrequent contact to the biological father at T3 and economic deprivation at T3 as likely mediators between family structure and adolescent outcomes. 3.3 Analyses In these latter analyses, we were interested in seeing (a) whether youth from separated family types evidenced less favourable changes in well-being across the two-year time period than youth from nuclear families, (b) whether infrequent contact to the father and/or economic deprivation affect- ed adolescent outcomes, and (c) whether and to what extent the effects of fam- ily structure were diminished when introducing infrequent contact to the father and economic deprivation as additional predictors. Given that infrequent con- tact to the father is only found in separated families and economic deprivation has been shown to be substantially higher in single mother families (see sample description), both are potential mediators which may explain disadvantages in adolescent well-being when comparing separated to nuclear families (for tests on mediator effects see Baron/Kenny 1986). Furthermore, particular attention was paid to effects of a prospective separation between T1 and T3. If a recent parental separation was more stressful than the pre-separation period, adoles- cents’ well-being should be signiﬁ cantly compromised over time for this group of prospective separators (controlling for the previous level of the respective outcome at T1). (5) In a ﬁ nal model, we tested interactions between family type and adolescent gen- der or maternal education, using multiplication terms between the dichotomous family type variables and either gender or education (each variable coded 0/1). All interaction terms were entered simultaneously (controlling for the predictors used in the model 4 for T3). If any interaction term proved signiﬁ cant, the mul- tiple regression analyses were repeated separately for the relevant subgroups (males vs. females or adolescents with higher- vs. lower-educated mother) to allow for more detailed comparisons. (5) In a ﬁ nal model, we tested interactions between family type and adolescent gen- der or maternal education, using multiplication terms between the dichotomous family type variables and either gender or education (each variable coded 0/1). All interaction terms were entered simultaneously (controlling for the predictors used in the model 4 for T3). If any interaction term proved signiﬁ cant, the mul- tiple regression analyses were repeated separately for the relevant subgroups (males vs. females or adolescents with higher- vs. lower-educated mother) to allow for more detailed comparisons. 4 4 Results Satisfaction With Family. The ﬁ ndings for multiple regressions on adolescents’ sat- isfaction with their family are depicted in table 3. As shown in the ﬁ rst column for family satisfaction at T1, youth from single mother families indicated signiﬁ cantly lower satisfaction with family life, whereas youth in stepfather families did not dif- fer from nuclear families. Furthermore, youth in nuclear families whose parents separated within the subsequent two years (prospective separators) also indicated signiﬁ cantly lower satisfaction with their family at T1, even though this effect was rather weak. There were neither age nor gender differences, nor did the region of Sabine Walper, Carolin Thönnissen, Philipp Alt 350 residence or a migration background have any effect. However, adolescents with highly-educated mothers reported signiﬁ cantly lower satisfaction with family life than youth with a less highly-educated mother. As can be seen in model 2, no inter- action effects were found, i.e. the effects of family type did not differ by adolescent gender or educational resources. At T3 (model 3), disadvantages of separated families had slightly increased, as evident in signiﬁ cant negative changes when controlling for family satisfaction at T1. This particularly holds for prospective separators, whose family satisfaction showed a signiﬁ cantly stronger decline than found among youth in nuclear families. Youth in single mother or stepfather families also indicated a small, but signiﬁ cantly stronger decrease in satisfaction with their family than youth in nuclear families, but these effects were weak overall. As can be seen in model 4, these effects were not diminished when controlling for economic deprivation and infrequent contact to the father even though economic deprivation had a signiﬁ cant negative effect on adolescents’ satisfaction with their family. As expected, infrequent contact to the T1 T3 (1) (2) (3) (4) (5) Gender (Male) -.006 -.005 .017 .016 .008 Age -.021 -.022 -.010 -.010 -.010 Migration Background .039 .038 .021 .035 .034 Maternal Education -.096 -.091* -.012 -.028 -.044* Region (East) .022 .021 -.014 -.001 -.003 Prospective Separators+ -.084 -.042 -.106* -.099* -.150* Single Mother Families -.139 -.152* -.081* -.059 -.080* Stepfather Families -.024 -.016 -.054* -.050 -.061* Family Satisfaction T1 .393* .387* .388* Economic Deprivation -.089* -.089*** Infrequent Contact to Father .004 .008 Prosp.Sep. X Gender -.053 .032 Single Fam. X Gender .029 .022 Stepfather Fam. X Gender -.010 -.006 Prosp.Sep. X Education -.012 .057 Single Fam. X Education -.013 .007 Stepfather Fam. 4 Results X Education .000 .027 R2 .031 .031 .186 .192 .193 N 2481 2481 2427 Tab. 3: Effects on Adolescents’ Family Satisfaction at T1 and T3: Standardised Regression Coefﬁ cients Note: Signiﬁ cance: * p < .05; p < .01; * p < .001 + Adolescents whose parents separated between waves 1 and 3 Source: pairfam data release 4.0 Tab. 3: Effects on Adolescents’ Family Satisfaction at T1 and T3: Standardised Regression Coefﬁ cients 351 Effects of Parental Separation Effects of Parental Separation father did not affect adolescents’ satisfaction with their family. Model 5 introduced the interaction terms, but none of these proved signiﬁ cant. Satisfaction With School, Education, and Occupation. Similar to the previous analyses, youth from single mother families as well as prospective separators re- ported lower satisfaction with their education or work situation at T1 when com- pared to adolescents from nuclear families (see Table 4). This indicates not only disadvantages of youth with single mothers, but also pre-separation strain in the scholastic or occupational domain among adolescents whose parents later separat- ed. However, both effects were rather small. Adolescents from stepfather families did not differ from youth who lived with both biological parents. Age, gender, ma- ternal education, and migration background showed no effects. There were small effects for region, as youth from eastern Germany tended to report less satisfaction with their education and work than youth from western Germany. In model 2, no interaction effect emerged. Regarding change across time (model 3), prospective separators as well as youth from single mother or stepfather families did not differ Tab. 4: Effects on Adolescents’ Satisfaction with School, Education, and Occupation at T1 and T3: Standardised Regression Coefﬁ cients T1 T3 (1) (2) (3) (4) (5) Gender (Male) -.008 -.012 .057 .055 .060** Age .038 -.038 -.018 -.017 -.018 Migration Background -.010 -.009 -.047* -.029 -.031 Maternal Education -.026 -.024 -.015 -.035 -.031 Region (East) -.052* -.051* -.043* -.028 -.030 Prospective Separators+ -.052** -.062* -.021 -.012 -.023 Single Mother Families -.090* -.102 -.032 .001 .016 Stepfather Families -.027 -.013 -.010 .000 .017 Satisfaction with School, Education, and Occupation T1 .208* .197* .197* Economic Deprivation -.106* -.106*** Infrequent Contact to Father -.014 -.016 Prosp.Sep. X Gender .028 -.009 Single Fam. X Gender .009 .006 Stepfather Fam. X Gender -.011 -.026 Prosp.Sep. X Education -.018 .033 Single Fam. X Education .011 -.035 Stepfather Fam. + Adolescents whose parents separated between waves 1 and 3 4 Results X Education -.010 .004 R2 .011 .010 .050 .059 .059 N 2481 2481 2426 Note: Signiﬁ cance: * p < .05; p < .01; * p < .001 + Adolescents whose parents separated between waves 1 and 3 Source: pairfam data release 4.0 Tab. 4: Effects on Adolescents’ Satisfaction with School, Education, and Occupation at T1 and T3: Standardised Regression Coefﬁ cients + Adolescents whose parents separated between waves 1 and 3 Sabine Walper, Carolin Thönnissen, Philipp Alt 352 from youth in nuclear families. A small gender effect was found, as male adoles- cents reported a slightly stronger increase in satisfaction than female adolescents. Economic deprivation proved to predict lower satisfaction with this domain but did not explain any effect of family structure (model 4), as there were no such effects in model 3. Interestingly, satisfaction with this domain was substantially less stable than family satisfaction. No interaction effects between family structure and gender or maternal education were found (model 5). Overall Life Satisfaction. Table 5 shows a summary of multiple regressions on adolescent’s life satisfaction. A signiﬁ cant effect of family type was found at T1 and T3 (model 1), with lower overall life satisfaction at T1 and more negative changes (T3) reported by youth from single mother families (see model 1 and 3). The lat- ter effect was reduced to insigniﬁ cance when controlling for economic deprivation (model 4), which proved to be a highly signiﬁ cant predictor of compromised life satisfaction. Infrequent contact to the father did not affect overall life satisfaction. As for the control variables, males and younger adolescents proved more satisﬁ ed Tab. 5: Effects on Adolescents’ Life Satisfaction at T1 and T3: Standardised Regression Coefﬁ cients T1 T3 (1) (2) (3) (4) (5) Gender (Male) .074* .063 .057 .055 .062 Age -.067* -.069** -.028 -.029 -.028 Migration Background .018 .017 -.045* -.013 -.014 Maternal Education -.024 -.006 .017 -.019 -.039 Region (East) -.028 -.028 -.070* -.042* -.043* Prospective Separators+ -.001 .014 -.030 -.014 -.012 Single Mother Families -.076* -.062* -.076* -.018 -.052 Stepfather Families -.013 -.040 -.015 .002 .021 Life Satisfaction T1 - .335* .308* .311* Economic Deprivation - -.188* -.190*** Infrequent Contact to Father - -.020 -.014 Prosp.Sep. X Gender - -.008 -.023 Single Fam. X Gender - .023 .011 Stepfather Fam. X Gender - .032 -.035 Prosp.Sep. X Education - -.019 .025 Single Fam. Note: Signiﬁ cance: * p < .05; p < .01; * p < .001 g p p p + Adolescents whose parents separated between waves 1 and 3 4 Results X Education - -.060* .048* Stepfather Fam. X Education - .010 .008 R2 .014 .015 .134 .163 .164 N 2483 2483 2426 Note: Signiﬁ cance: * p < .05; p < .01; * p < .001 + Adolescents whose parents separated between waves 1 and 3 Source: pairfam data release 4.0 5: Effects on Adolescents’ Life Satisfaction at T1 and T3: Standardised Regression Coefﬁ cients 353 Effects of Parental Separation • 353 Tab. 6: Effects on Adolescents’ Self-Esteem at T1 and T3: Standardised Regression Coefﬁ cients T1 T3 (1) (2) (3) (4) (5) Gender (Male) .213* .197* .113* .115* .107** Age -.009 -.010 .001 .000 .001 Migration Background .049* .049* .016 .026 .026 Maternal Education .016 .026 -.010 -.021 -.027 Region (East) -.028 -.028 -.025 -.016 -.017* Prospective Separators+ -.023 -.029 -.001 .002 -.031 Single Mother Families -.012 -.011 -.052** -.049* -.071* Stepfather Families -.016 -.050 -.017 .010 .013 Self-Esteem T1 - .380* .369* .369* Economic Deprivation - -.075* -.075*** Infrequent Contact to Father - .040 .043* Prosp.Sep. X Gender - -.020 .023 Single Fam. X Gender - .024 .023 Stepfather Fam. X Gender - -.036 .000 Prosp.Sep. X Education - -.014 .033 Single Fam. X Education - -.036 .009 Stepfather Fam. X Education - .016 -.007 R2 .047 .047 .131 .164 .165 N 2482 2416 Note: Signiﬁ cance: * p < .05; p < .01; * p < .001 + Adolescents whose parents separated between waves 1 and 3 Source: pairfam data release 4.0 Tab. 6: Effects on Adolescents’ Self-Esteem at T1 and T3: Standardised Regression Coefﬁ cients than females and older adolescents, while maternal education and migration back- ground did not matter. A small effect of region emerged at T3 (model 3), indicating a slightly stronger decrease in life satisfaction among youth in eastern than in west- ern Germany. In model 4, however, controlling for economic deprivation, this effect was no longer signiﬁ cant. Finally, both tests for interaction effects (model 2 and 5) evidenced differential effects of single motherhood among families with higher or lower educational resources. However, the effect differed for both models, i.e. changed over time. Separate regression analyses (as in model 1) were conducted for both educational groups. 5 Discussion The aim of the current study was to assess the effects of family structure on dif- ferent aspects of adolescent well-being across a period of two years. We not only wanted to explore possible disadvantages for adolescents in stable single mother and stepfather families compared to nuclear families, but also took the opportunity of following a group of adolescents whose parents separated within the two-year interval between the two longitudinal assessments. This group of so-called “pro- spective separators” allowed for addressing issues of selectivity and pre-separation strain that may compromise adolescents’ well-being prior to parental separation. We also examined whether changes in adolescents’ well-being differed by family type and expected to ﬁ nd more negative change among prospective separators. Secondly, we explored the mediating role of infrequent contact to the non-residen- tial parent as well as economic deprivation in linking family structure to adolescent outcomes in well-being. Finally, we tested whether gender and maternal education moderated the effects of family structure at T1 or across time. In the following, we summarise and discuss the overall pattern of results linking our ﬁ ndings to the hy- potheses previously speciﬁ ed. We end with a brief outline of strengths and limita- tions of our study and of challenges for future research. 4 Results Among families with higher maternal education, youth from single mother families reported lower life satisfaction than youth from nuclear families (beta = -.153, p < .001), while no such effect was evident among youth with less highly-educated mother (beta = -.046, n.s.). While this ﬁ nding does not support the resource hypothesis, changes over time (model 3) were more in line with the re- source notion. These changes were less favourable for youth from less resourceful families (beta = -.102, p < .001) than for adolescents with a higher-educated mother (beta = -.004, n.s.). • Sabine Walper, Carolin Thönnissen, Philipp Alt 354 Self-Esteem. Finally, turning to adolescents’ self-esteem, table 6 shows the ﬁ nd- ings from our regression analyses. As can be seen, no effects of family type were found at T1. Prospective separators did not differ from youth in nuclear families, nei- ther prior to parental separation (T1), nor shortly after (T3). Similarly, we did not ﬁ nd any disadvantage of youth in stepfather families at either measurement point. Only at T3 did adolescents growing up with a single mother indicate slightly lower self- esteem. Although signiﬁ cant, this effect was very weak and could not be accounted for by the negative effects of economic deprivation on adolescents’ self-esteem (model 4). There were no interaction effects (model 5). With respect to the control variables, males consistently reported higher self- esteem than females at T1 and T3, and youth with a migration background had slightly higher self-esteem than youth without a migration background. Age, ma- ternal education, and the region of residence did not have any signiﬁ cant effects at T1 (model 1) or over time (model 3). No interaction effects were found in model 5. 5.1 Effects of Family Type Effects of family structure on adolescents’ well-being were evident for all outcome variables, but differed with respect to the domain under consideration. Most con- sistent evidence was found for disadvantages of youth from single mother families, who reported lower satisfaction with family life and with the domain of education and occupation as well as a limited overall life satisfaction compared to their peers in nuclear families at T1. These youth with single mothers also evidenced a stronger decrease in family satisfaction, overall life satisfaction, and self-esteem than youth 355 Effects of Parental Separation • 355 in nuclear families. Although these ﬁ ndings are in line with the large body of re- search on parental separation (Amato 2010), it should be noted that the effects are mostly very small. As might be expected, the most pronounced effects were found for family satisfaction, indicating that the problems faced by these youth are indeed rooted in the family system and not in confounded external factors. In contrast, there was almost no evidence for disadvantages of adolescents raised in stepfather families. At T1, they did not differ in any of the outcome vari- ables from youth in nuclear families. When looking at changes over time, only fam- ily satisfaction was found to show a stronger decline among youth living with a stepfather, while the other three indicators did not reveal any less advantageous changes than found for youth in nuclear families. Overall, these ﬁ ndings suggest that the majority of stepfamilies manage to adjust successfully to the challenges of stepfamily life and provide a positive context for adolescents’ well-being. This ﬁ nd- ing is noteworthy, since other evidence has instead focused on the challenges and problems in stepfamily life (Amato 1994; Sweeney 2010). Contrary to the ﬁ ndings from the large body of data available from international studies (Jeynes 2006), our data do not suggest that adolescents raised in a stepfamily are similarly strained or even worse off than youth living in a single mother family. The slightly stronger de- crease in satisfaction with family life may indicate that stepfamily life becomes more stressful in late adolescence when autonomy is more fully established. However, this effect was quite small and should not be over-interpreted. With respect to prospective separators, the expected pre-separation strain at T1 was evidenced by lower family satisfaction than found for youth from stable nuclear families. 5.1 Effects of Family Type Furthermore, such pre-separation impairment of well-being was also evi- dent in the domain of education and work, which these adolescents were less satis- ﬁ ed with than those from stable nuclear families. It should be noted that – contrary to youth from single mother and stepfather families – prospective separators were not less likely than their peers from stable nuclear families to be enrolled in or have ﬁ nished the highest track of schooling. This may suggest that such problems with education/work were not yet evident at the end of elementary school but rather emerged more recently, reﬂ ecting distractions emanating from discord at home. While these effects are in line with other evidence suggesting higher levels of strain in the pre-separation phase (e.g. Cherlin et al. 1991), they prove to be rather weak. Moreover, neither overall life satisfaction nor adolescents’ self-esteem suggest any major general pre-separation strain extending into adolescents’ outlook on life. Overall, our ﬁ ndings at least partly resemble other evidence from Germany, which provided only limited evidence for pre-separation effects (Schwarz 1999). Similarly, more negative changes across time are restricted to family satisfaction, indicating that family problems increase when parents separate. This ﬁ nding is no surprise. Rather, it must be emphasized that this effect is quite limited, not only in content, but also in size. Neither did we ﬁ nd any signs for a spillover of post-separation problems into the domain of education or work, nor did the experience of parental separation seem to affect more general features of well-being, such as life satisfac- tion or self-esteem. 356 • Sabine Walper, Carolin Thönnissen, Philipp Alt In summary, while the expected effects of parental separation were evident for almost every outcome variable, these effects proved rather weak and fell below the average effect sizes reported in meta-analyses for data, most of which come from the US (Amato 2001). The most pronounced effects were found for adolescents’ family satisfaction, which – along with adolescents’ satisfaction in the domain of education and work – also provided support for the expected pre-separation strain. An additional decrease in well-being shortly after a recent parental separation was only evident for family satisfaction. Finally, stepfamilies seem to provide a rather unobtrusive context for adolescents’ well-being. 5.2 The Mediating Role of Infrequent Contact to the Father and Economic Deprivation Our analyses also tested for possible mediation effects of infrequent contact to the father and economic deprivation. Throughout the analysis, infrequent contact to father had no effect on adolescents’ well-being and thus could not explain any effects of the family type. Overall, these ﬁ ndings are well in line with other studies (Amato/Gilbreth 1999). Economic deprivation, however, proved to mediate at least some of the effects of family structure. Such mediation effects of economic deprivation were restricted to disadvantages of single mother families (who experienced the highest ﬁ nancial strain). The stronger decrease in overall life satisfaction which was found among adolescents in single mother families could be fully accounted for by the higher economic deprivation reported by these youth. With respect to family satisfaction, however, only partial mediation was found, and our results for self-esteem did not indicate any mediation by economic deprivation. Although self-esteem decreased more strongly in single mother families than in nuclear families, this effect remained unchanged when controlling for ﬁ nancial conditions. Overall, these ﬁ nding are at least partly in line with other evidence suggest- ing that a substantial share of disadvantages found for children raised by a single mother can be accounted for by the increased ﬁ nancial problems these families face (e.g. McLanahan 1999). At the same time, parental separation seems to also play a distinct role, as particularly evident in cases of recent separations. Elevated post-separation strain among prospective separators, as evident in reduced family satisfaction, could not be explained by economic deprivation. This suggests that a recent separation does provide a unique stressor which cannot be reduced to ﬁ nancial problems. 5.3 The Moderating Role of Gender and Maternal Education We also addressed possible moderating effects of gender and adolescents’ school type to test for differential effects of family type in the respective subgroups. In line with other ﬁ ndings, gender did not moderate any effects of growing up in a single mother family. While other evidence suggests that girls fare worse than boys in stepfather families (Hetherington/Jodl 1994), we did not ﬁ nd stronger disadvan- 357 Effects of Parental Separation • 3 tages in well-being among girls (or boys) growing up with a stepfather. Similarly, we could not replicate earlier ﬁ ndings indicating stronger pre-separation strain among boys. While the group of prospective separators may have been too small to detect such effects, the overall pattern suggests rather homogeneous effects of family type for males and females. Maternal education was found to moderate the effects of family type in one case only: for adolescents’ overall life satisfaction. At the same time, this effect was com- plex and not easy to interpret. Contrary to the resource hypothesis, but in line with the ﬂ oor effect found by Bernardi and Radl (2014), our data also indicate a “divorce penalty” among more privileged youth only: Negative effects of parental separa- tion at T1 were restricted to youth with a higher-educated mother. In this subgroup, youth from single mother families indicated lower life satisfaction than those from nuclear families, while no such difference was found for adolescents with a less highly educated mother. It may well be that adolescents from families with higher educational resources have more to lose when their parents separate, at least in terms of ﬁ nancial resources. Interestingly, however, support for the resource hy- potheses was found when looking at changes across time. Adolescents’ life satis- faction decreased more strongly in single mother than nuclear families with a lower education, while no such effect was found among adolescents with a higher-edu- cated mother. As expected, the decrease in life satisfaction among youth from less resourceful single mother families could be fully explained by ﬁ nancial strain. This suggests that ﬁ nancial resources linked to parental education play a major role in differential outcomes of single parenthood. 5.4 Strengths and Limitations of the Study This study used a large sample of adolescents in Germany for systematically testing several research questions regarding the effects of parental separation on adoles- cents’ well-being. Taking a broad view on separation and re-partnering, we did not only focus our analyses strictly on parental divorce or on remarried families, but also included youth whose parents were still or never married but separated, and those who lived with their mother’s new partner, no matter whether they were mar- ried or not. Overall, our data provide support for the detrimental effects of increased ﬁ nancial strain among single mother families, while the role of contact to the non- residential father for adolescents’ well-being seems to be much less marked. This latter ﬁ nding is well in line with broad evidence from international studies (Amato/ Gilbreth 1999) as well as studies conducted in Germany (Walper/Wendt 2011; Walp- er/Krey 2009), but not with current legal practice, which places much emphasis on preserving the continuity of adolescents’ relationship to their non-custodial par- ent, even against the children’s will. Furthermore, we could identify increased pre- separation strain in the subgroup of prospective separators, who also seem to face additional post-separation strain, not with a broad impact on general well-being, but more speciﬁ cally in the domain of family life. Finally, we found evidence for the moderating role of socio-economic resources as addressed by maternal education. 358 • Sabine Walper, Carolin Thönnissen, Philipp Alt 358 Nevertheless, these ﬁ ndings should be interpreted with caution. First, as pointed out above, the effects of family structure were mostly very weak. Many of them would have remained undetected using a smaller sample. While the large sample may be considered an advantage of this study, the subgroups in focus differ con- siderably in size. Particularly the group of prospective separators was very small in comparison to the large group of nuclear families. Restricting the number of nuclear families by selecting a random subsample or by matching them to the separated families might have solved this problem. However, such procedures carry other risks. We decided to capitalise on the full sample to provide a broader picture of adolescents growing up in different family types. Second, we restricted our analyses to adolescents who still lived with their mother in any case. Amato, Paul R. 1994: The implications of research ﬁ ndings on children in stepfamilies. In: Booth, Alan; Dunn, Judith (Eds.): Stepfamilies: Who beneﬁ ts? Who does not? Hill- sdale, NJ: Erlbaum: 81-87. Acknowledgement This research was supported by a grant from the German Research Foundation (DFG) to Sabine Walper (Wa 949/11-1 through 3) as principle co-investigator of the German Family Panel, pairfam, at Ludwig-Maximilian-University, Munich. We are grateful to all participants of the pairfam panel. 5.4 Strengths and Limitations of the Study Although the number of adolescents who had already moved out by age 17 to 19 (the majority age range at T3) was small, our selectivity analyses showed that offspring from separated families is overrepresented in this group. Hence, we may underestimate the effects of family structure if those who were most strained by their parents’ separation moved out earlier. Third, we could only rely on a limited range of outcome variables. Indicators of adolescents’ depression differed at both measurement points and thus could not be considered. Information about adolescents’ problem behaviour was not available at either measurement point. Hence, we may underestimate psychological and behav- ioural problems that have been pointed out to be more sensitive to negative effects of parental separation (Amato 2001). Finally, our analyses did not inform about some key variables previously dis- cussed, such as the role of interparental conﬂ ict and the number of family transi- tions adolescents went through prior to our study. Some of these data are available in the pairfam panel, and can be addressed in future analyses. Amato, Paul R. 1993: Children’s adjustment to divorce: Theories, hypotheses, and em- pirical support. In: Journal of Marriage and Family 55: 23-38 [doi: 10.2307/352954]. References Allison, Paul D.; Furstenberg, Frank F. Jr. 1989: How marital dissolution affects chil- dren: variations by age and sex. In: Developmental Psychology 25,4: 540-549 [doi: 10.1037/0012-1649.25.4.540]. Amato, Paul R. 1993: Children’s adjustment to divorce: Theories, hypotheses, and em- pirical support. In: Journal of Marriage and Family 55: 23-38 [doi: 10.2307/352954]. 359 Effects of Parental Separation • 359 Amato, Paul R. 2000: The consequences of divorce for adults and children. In: Journal of Marriage and Family 62: 1269-1287 [doi: 10.1111/j.1741-3737.2000.01269.x]. Amato, Paul R. 2001: Children of divorce in the 1990s: An update of the Amato and Keith (1991) meta-analysis. In: Journal of Family Psychology 15,3: 355-370 [doi: 10.1037/0893-3200.15.3.355]. Amato, Paul R. 2010: Research on divorce: Continuing trends and new developments. In: Journal of Marriage and Family 72,3: 650-666 [doi: 10.1111/j.1741-3737.2010.00723.x]. Amato, Paul R. 2014: The consequences of divorce for adults and children: an update. In: Drustvena istrazivanja: Journal for General Social Issues 23,1: 5-24 [doi: 10.5559/ di.5523.5551.5501]. Amato, Paul R.; Keith, Bruce 1991: Parental divorce and the well-being of children: a meta-analysis. In: Psychological Bulletin 110: 26-46 [doi: 10.1037/0012-1649.25.4.540]. Amato, Paul R.; Gilbreth, Joan G. 1999: Nonresident fathers and children’s well-being: A meta-analysis. In: Journal of Marriage and Family 61: 557-573 [doi: 10.2307/353560]. Amato, Paul R.; Loomis, Laura S.; Booth, Alan 1995: Parental divorce, marital conﬂ ict, and offspring well-being during early adulthood. In: Social Forces 73: 895-915 [doi: 10.2307/2580551]. Amato, Paul R; Cheadle, Jacob E. 2008: Parental divorce, marital conﬂ ict and children’s behavior problems: A comparison of adopted and biological children. In: Social Forc- es 86,3: 1139-1161 [doi: 10.1353/sof.0.0025]. Amato, Paul R.; James, Spencer 2010: Divorce in Europe and the United States: commonalities and differences across nations. In: Family Science 1,1: 2-13 [doi: 10.1080/19424620903381583]. Amendt, Gerhard 2004: Scheidungsväter. Schriftenreihe des Instituts für Geschlechter- und Generationenforschung Band 6. Bremen: Ikaru-Verlag. Baron, Reuben M.; Kenny, David A. 1986: The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. In: Journal of Personality and Social Psychology 51: 1173-1182 [doi: 10.1037/0022- 3514.51.6.1173]. Bastin, Sonja; Kreyenfeld, Michaela; Schnor, Christine 2012: Diversität von Familienfor- men in Ost- und Westdeutschland. In: Krüger, Dorothea; Herma, Holger; Schierbaum, Anja (Eds.): Familie(n) heute ശ Entwicklungen, Kontroversen, Prognosen. Wiesbaden: Juventa: 126-145. Bernardi, Fabrizio; Radl, Jonas 2014: The long-term consequences of parental divorce for children’s educational attainment. In: Demographic Research 30,61: 1653-1680 [doi: 10.4054/DemRes.2014.30.61]. Block, Jeanne H.; Block, Jack; Gjerde, Per F. References 1986: The personality of children prior to di- vorce: A prospective study. In: Child Development 57: 827-840 [doi: 10.2307/1130360]. Buehler, Cheryl; Gerard, Jean M. 2002: Marital conﬂ ict, ineffective parenting and chil- dren’s and adolescents’ maladjustment. In: Journal of Marriage and Family 64,1: 78-92 [doi: 10.1111/j.1741-3737.2002.00078.x]. Burt, S. Alexandra et al. 2008: Parental divorce and adolescent delinquency: ruling out the impact of common genes. In: Developmental Psychology 44,6: 1668-1677 [doi: 10.1037/a0013477]. Chase-Lansdale, P. Lindsay; Cherlin, Andrew J.; Kiernan, Kathleen E. 1995: The long- term effects of parental divorce on the mental health of young adults: A developmen- tal perspective. In: Child Development 66,6: 1614-1634 [doi: 10.1111/j.1467-8624.1995. tb00955.x]. • Sabine Walper, Carolin Thönnissen, Philipp Alt 360 360 Cherlin, Andrew J.; Chase-Landsdale, P. Lindsay; Mcrae, Christine 1998: Effects of pa- rental divorce on mental health throughout the life course. In: American Sociological Review 63,2: 239-249 [doi: 10.2307/2657325]. Cherlin, Andrew J. et al. 1991: Longitudinal studies of effects of divorce on children in Great Britain and the United States. In: Science 252: 1386-1389 [doi: 10.1126/sci- ence.2047851]. Cherlin, Andrew J.; Kiernan, Kathleen E.; Chase-Lansdale, P. Lindsay 1995: Parental di- vorce in childhood and demographic outcomes in young adulthood. In: Demography 32,3: 299-318 [doi: 10.2307/2061682]. Coleman, Marilyn; Ganong, Lawrence; Fine, Mark 2000: Reinvestigating remarriage: Another decade of progress. In: Journal of Marriage and Family 62,4: 1288-1307 [doi: 10.1111/j.1741-3737.2000.01288.x]. Conger, Katherine J.; Rueter, Martha A.; Conger, Rand D. 2000: The role of economic pressure in the lives of parents and their adolescents: The Family Stress Model. In: Crockett, Lisa J.; Silbereisen, Rainer K. (Eds.): Negotiating adolescence in times of social change. Cambridge: Cambridge University Press: 201-233. Curtin, Sally C.; Ventura, Stephanie J.; Martinez, Gladys M. 2014: Recent Declines in Nonmarital Childbearing in the United States. In: NCHS data brief 162: 1-8. Davies, Patrick T. et al. 2002: Child emotional security and interparental conﬂ ict. Boston, MA: Blackwell. Duncan, Greg J.; Brooks-Gunn, Jeanne (Eds.) 1997: The consequences of growing up poor. New York: Russell Sage Foundation. Ehmke, Timo et al. 2004: Familiäre Lebensverhältnisse, Bildungsbeteiligung und Kom- petenzerwerb. In: Deutschland, Pisa-Konsortium (Ed.): PISA 2003. Der Bildungsstand der Jugendlichen in Deutschland ശ Ergebnisse des zweiten internationalen Vergleichs. Münster: Waxmann. Emery, Robert E. 1988: Marriage, divorce, and children’s adjustment. Newbury Park, CA: Sage. Emery, Robert E. et al. 1999: Delinquent behavior, future divorce or nonmarital child- bearing, and externalizing behavior among offspring: A 14-year prospective study. References In: Journal of Family Psychology 13,4: 568-579 [doi: 10.1037/0893-3200.13.4.568]. Erel, Osnat; Burman, Bonnie 1995: Interrelatedness of marital relations and parent- child-relations: A meta-analytic review. In: Psychological Bulletin 118,1: 108-132 [doi: 10.1037/0033-2909.118.1.108]. Eurostat 2013: European social statistics. 2013 edition. Luxembourg: European Union. Fergusson, D. David M.; Horwood, L. John; Lynskey, Michael T. 1992: Family change, parental discord, and early offending. In: Journal of Child Psychology and Psychiatry Review 33: 1059-1075 [doi: 10.1111/j.1469-7610.1992.tb00925.x]. Fergusson, David M; Horwood, L. John; Lynskey, Michael T. 1994: Parental separation, adolescent psychopathology, and problem behaviors. In: Journal of the American Academy of Child & Adolescent Psychiatry 33,8: 1122-1133 [doi: 10.1097/00004583- 199410000-00008]. Fincham, Frank D. 1998: Child development and marital relations. In: Child Development 69,2: 543-574 [doi: 10.2307/1132183]. Forehand, Rex; Armistead, Lisa; David, Corinne 1997: Is adolescent adjustment follow- ing parental divorce a function of predivorce adjustment? In: Journal of Abnormal Child Psychology 25,2: 157-164 [doi: 10.1023/A:1025735629914]. 361 Effects of Parental Separation • 361 Francesconi, Marco; Jenkins, Stephen P.; Siedler, Thomas 2010: Childhood family struc- ture and schooling outcomes: Evidence for Germany. In: Journal of Population Eco- nomics 23,3: 1073-1103 [doi: 10.1007/s00148-009-0242-y]. Garriga, Anna; Härkönen, Juho 2009: The effects of marital instability on children’s well- being and intergenerational relations. In: EQUALSOC State-of-the-art report. Pompeu Fabra University and Stockholm University. Gloger-Tippelt, Gabriele; König, Lilith 2007: Attachment representations of 6-year-old children from one and two parent families in Germany. In: School Psychology Interna- tional 28,3: 313-330 [doi: 10.1177/0143034307078540]. Hawkins, Daniel N.; Amato, Paul R.; King, Valerie 2007: Nonresident father involvement and adolescent well-being: Father effects or child effects? In: American Sociological Review 72,6: 990-1010 [doi: 10.1177/000312240707200607]. Hetherington, Eileen Mavis; Jodl, Kathleen M. 1994: Stepfamilies as settings for child development. In: Booth, Alan; Dunn, Judy (Eds.): Stepfamilies. Who beneﬁ ts? Who does not? Hillsdale, N.J.: Lawrence Erlbaum: 55-79. Hetherington, Eileen Mavis; Stanley-Hagan, Margaret 1999: The adjustment of children with divorced parents: a risk and resiliency perspective. In: Journal of Child Psychol- ogy and Psychiatry 40,1: 129-140 [doi: 10.1177/000312240707200607]. Hetherington, E. Mavis; Kelly, John 2002: For better or for worse. Divorce reconsidered. New York: Norton. Hilmert, Steffen 2002: Familiale Resourcen und Bildungschancen: Konsequenzen eines frühzeitigen Elternverlustes (Family Resources and Educational Opportunities: Con- sequences of Early Loss of Parents). In: Zeitschrift für Familienforschung 14: 44-69. Huinink, Johannes et al. 2011: Panel Analysis of Intimate Relationships and Family Dy- namics (pairfam) ശ Conceptual Framework and Design. References In: Zeitschrift für Familien- forschung/Journal of Family Research 23,1: 77-100. Jeynes, William H. 2006: The impact of parental remarriage on children: A meta-analy- sis. In: Marriage & family review 40,4: 75-102 [doi: 10.1300/J002v40n04_05]. Krishnakumar, Ambika; Buehler, Cheryl 2000: Interparental conﬂ ict and parenting be- haviors. A meta-analytic review. In: Family Relations 49: 25-44 [doi: 10.1111/j.1741- 3729.2000.00025.x]. Liefbroer, Aart C.; Dourleijn, Edith 2006: Unmarried cohabitation and union stability: Testing the role of diffusion using data from 16 European countries. In: Demography 43,2: 203-221 [doi: 10.1353/dem.2006.0018]. McLanahan, Sara 1999: Father absence and the welfare of children. In: Hetherington, Eileen Mavis (Ed.): Coping with divorce, single parenting, and remarriage. A risk and resilience perspective. Mahwah: Lawrence Erlbaum Associates: 117-146. Nauck, Bernhard et al. 2013: The German Family Panel (pairfam). GESIS Data Archive, Cologne. ZA5678 Data ﬁ le Version 4.0.0, [doi:10.4232/pairfam.5678.4.0.0]. O’Connor, Thomas G. et al. 2000: Are associations between parental divorce and chil- dren’s adjustment genetically mediated? An adoption study. In: Developmental Psy- chology 36,4: 429-437 [doi: 10.1037/0012-1649.36.4.429]. Rosenberg, Morris 1965: Society and the adolescent self-image. Princeton: Princeton University Press. Ross, Catherine E.; Mirowsky, John 1999: Parental divorce, life-course disrup- tion, and adult depression. In: Journal of Marriage and Family 61: 1034-1045 [doi: 10.2307/354022]. • Sabine Walper, Carolin Thönnissen, Philipp Alt 362 Sanz-De-Galdeano, Anna; Vuri, Daniela 2007: Parental Divorce and Students’ Perfor- mance: Evidence from Longitudinal Data. In: Oxford Bulletin of Economics and Statis- tics 69,3: 321-338 [doi: 10.1111/j.1468-0084.2006.00199.x]. Schick, Andreas 2002: Behavioral and emotional differences between children of di- vorce and children from intact famlies: Clinical signiﬁ cance and mediating processes. In: Swiss Journal of Psychology 61,1: 5-14 [doi: 10.1024//1421-0185.61.1.5]. Schmidt-Denter, Ulrich 2000: Entwicklung von Trennungs- und Scheidungsfamilien: Die Kölner Längsschnittstudie. In: Schneewind, Klaus A. (Ed.): Familienpsychologie im Aufwind. Brückenschläge zwischen Forschung und Praxis. Göttingen: Hogrefe: 203- 221. Schmidt-Denter, Ulrich 2001: Differentielle Entwicklungsverläufe von Scheidungs- kindern. In: Walper, Sabine; Pekrun, Reinhard (Eds.): Familie und Entwicklung. Ak- tuelle Perspektiven der Familienpsychologie. Göttingen: Hogrefe: 292-313. Schnor, Christine 2012: Trennungsrisiko von Paaren mit Kindern: Der Einﬂ uss der Re- ligion in West- und Ostdeutschland. In: Huinink, Johannes; Kreyenfeld, Michaela; Trappe, Heike (Eds.): Familie und Partnerschaft in Ost- und Westdeutschland: Ähnlich und doch immer noch anders (Sonderheft 9 der Zeitschrift für Familienforschung). Leverkusen: Verlag Barbara Budrich: 229-255. Schwarz, Beate 1999: Die Entwicklung Jugendlicher in Scheidungsfamilien. Weinheim: Beltz PVU. Schwarz, Beate; Walper, Sabine; Gödde, Mechtild; Jurasic, Stephanie 1997: Dokumen- tation der Erhebungsinstrumente der 1. Haupterhebung (überarb. Version). References Berichte aus der Arbeitsgruppe „Familienentwicklung nach der Trennung“ # 14/1997. Sigle-Rushton, Wendy; Hobcraft, John; Kiernan, Kathleen 2005: Parental Divorce and Subsequent Disadvantage: A Cross-Cohort Comparison. In: Demography 42,3: 427- 446 [doi: 10.1353/dem.2005.0026]. Steinbach, Anja 2008: Stieffamilien in Deutschland. In: Zeitschrift für Bevölkerungswis- senschaft 33,2: 153-180 [doi: 10.1007/s12523-009-0009-2]. Steinhausen, Hans-Christoph; von Aster, Sigrid; Goebel, Dietmar 1987: Family compo- sition and child psychiatric disorders. In: Journal of the American Academy of Child & Adolescent Psychiatry 26,2: 242-247 [doi: 10.1097/00004583-198703000-00022]. Strohschein, Lisa 2005: Parental divorce and child mental health trajectories. In: Journal of Marriage and Family 67,5: 1286-1300 [doi: 10.1111/j.1741-3737.2005.00217.x]. Sweeney, Megan M. 2010: Remarriage and Stepfamilies: Strategic sites for scholarship in the 21st century. In: Journal of Marriage and Family 72,3: 667-684 [doi: 10.1111/j.1741- 3737.2010.00724.x]. Thönnissen, Carolin et al. 2014: Scales Manual of the German Family Panel. Waves 1 to 5. Release 5.0 (with contributions from: Petra Buhr, Veronika Salzburger, Nadia Lois and Claudia Schmiedeberg) [http://www.pairfam.de/ﬁ leadmin/user_upload/redak- teur/publis/Dokumentation/Manuals/Scales_Manual_pairfam_5.0.pdf]. Wallerstein, Judith S.; Corbin, Shauna B.; Lewis, Julia M. 1988: Children of divorce: A ten-year study. In: Hetherington, Eileen Mavis; Arasteh, J. D. (Eds.): Impact of divorce, single parenting, and stepparenting on children. Hilldale, N.J.: Erlbaum: 198-214. Walper, Sabine 2002: Einﬂ üsse von Trennung und neuer Partnerschaft der Eltern. Ein Vergleich von Jungen und Mädchen in Ost- und Westdeutschland. In: Zeitschrift für Soziologie der Erziehung und Sozialisation 22,1: 25-46. 363 Effects of Parental Separation Effects of Parental Separation Walper, Sabine 2009: Links of perceived economic deprivation to adolescents’ well- being six years later. In: Zeitschrift für Familienforschung/Journal of Family Research 21: 107-127. Walper, Sabine; Beckh, Katharina 2006: Adolescents’ development in high-conﬂ ict and separated families. Evidence from a German longitudinal study. In: Clarke-Stewart, Alison; Dunn, Judy (Eds.): Families count: Effects on child and adolescent develop- ment. Cambridge, MA: Cambridge University Press: 238-270. Walper, Sabine; Krey, Mari 2009: Familienbeziehungen nach Trennungen. In: Lenz, Karl; Nestmann, Frank (Eds.): Persönliche Beziehungen. Weinheim: Juventa. Walper, Sabine; Langmeyer, Alexandra 2014: Auswirkungen einer elterlichen Trennung auf die Entwicklung von Kindern in den ersten Lebensjahren. In: Kißgen, Rüdiger; Hei- nen, Norbert (Eds.): Trennung, Tod und Trauer in den ersten Lebensjahren: Begleitung und Beratung von Kindern und Eltern. Stuttgart: Klett-Cotta: 159-176. Walper, Sabine; Wendt, Eva-Verena 2005: Nicht mit beiden Eltern aufwachsen ശ ein Ri- siko? Kinder von Alleinerziehenden und Stieffamilien. In: Alt, Christian (Ed.): Kinderle- ben ശ Aufwachsen zwischen Familie, Freunden und Institutionen. Band 1: Aufwachsen in Familien. References Wiesbaden: VS Verlag für Sozialwissenschaften. DJI Deutsches Jugend- institut: 187-216. Walper, Sabine; Wendt, Eva-Verena 2011: Die Bedeutung der Abstammung für die Iden- titätsﬁ ndung und Persönlichkeitsentwicklung in der Adoleszenz: Adoption, Samen- spende und frühe Vaterabwesenheit nach Trennung der Eltern. In: Vaskovics, Laszlo (Ed.): Pluralisierung der Elternschaft und Kindschaft (Sonderheft der Zeitschrift für Familienforschung): 211-237. Wendt, Eva-Verena; Walper, Sabine 2007: Entwicklungsverläufe von Kindern in Ein- Eltern- und Stieffamilien. In: Alt, Christian (Ed.): Kinderleben ശ Start in die Grundschu- le Band 3: Ergebnisse aus der zweiten Welle. Wiesbaden: VS Verlag für Sozialwissen- schaften: 211-242. Whiteside, Mary F; Becker, Betsy Jane 2000: Parental factors and the young child’s post- divorce adjustment: a meta-analysis with implications for parenting arrangements. In: Journal of Family Psychology 14,1: 5-26 [doi: 10.1037/0893-3200.14.1.5]. Date of acceptance: 21.09.2015 Date of submission: 14.11.2014 Prof. Dr. Sabine Walper (). German Youth Institute. Munich, Germany. E-mail: walper@dji.de URL: http://www.dji.de/~walper Ludwig-Maximilians-University of Munich. Munich, Germany. URL: http://www.dji.de/~walperl Prof. Dr. Sabine Walper (). German Youth Institute. Munich, Germany. E-mail: walper@dji.de URL: http://www.dji.de/~walper Ludwig-Maximilians-University of Munich. Munich, Germany. URL: http://www.dji.de/~walperl Dr. Carolin Thönnissen, Philipp Alt. Ludwig-Maximilians-University of Munich. Munich, Germany. E-mail: c.thoennissen@edu.lmu.de, Philipp.Alt@edu.lmu.de URL: http://www.edu.lmu.de/apb/personen/wiss_ma/thoennissen/index.html http://www.edu.lmu.de/apb/personen/wiss_ma/alt/index.html URL: http://www.dji.de/~walperl Dr. Carolin Thönnissen, Philipp Alt. Ludwig-Maximilians-University of Munich. Munich, Germany. E-mail: c.thoennissen@edu.lmu.de, Philipp.Alt@edu.lmu.de URL: http://www.edu.lmu.de/apb/personen/wiss_ma/thoennissen/index.html http://www.edu.lmu.de/apb/personen/wiss_ma/alt/index.html Dr. Carolin Thönnissen, Philipp Alt. Ludwig-Maximilians-University of Munich. Munich, Dr. Carolin Thönnissen, Philipp Alt. Ludwig-Maximilians-University of Munich. Munich, Germany. E-mail: c.thoennissen@edu.lmu.de, Philipp.Alt@edu.lmu.de URL: http://www.edu.lmu.de/apb/personen/wiss_ma/thoennissen/index.html http://www.edu.lmu.de/apb/personen/wiss_ma/alt/index.html Comparative Population Studies www.comparativepopulationstudies.de ISSN: 1869-8980 (Print) – 1869-8999 (Internet) Published by / Herausgegeben von Prof. Dr. Norbert F. Schneider Federal Institute for Population Research D-65180 Wiesbaden / Germany Managing Editor / Verantwortlicher Redakteur Frank Swiaczny Assistant Managing Editor / Stellvertretende Redakteurin Katrin Schiefer Copy Editor (German) / Lektorat (deutsch) Dr. Evelyn Grünheid Layout / Satz Beatriz Feiler-Fuchs E-mail: cpos@bib.bund.de Scientiﬁ c Advisory Board / Wissenschaftlicher Beirat Paul Gans (Mannheim) Johannes Huinink (Bremen) Michaela Kreyenfeld (Rostock) Marc Luy (Wien) Clara H. Mulder (Groningen) Notburga Ott (Bochum) Peter Preisendörfer (Mainz) Zsolt Spéder (Budapest) © Federal Institute for Population Research Published by / Herausgegeben von Prof. Dr. Norbert F. Schneider Board of Reviewers / Gutachterbeirat Martin Abraham (Erlangen) Laura Bernardi (Lausanne) Hansjörg Bucher (Bonn) Claudia Diehl (Konstanz) Andreas Diekmann (Zürich) Gabriele Doblhammer-Reiter (Rostock) Jürgen Dorbritz (Wiesbaden) Anette Eva Fasang (Berlin) E.-Jürgen Flöthmann (Bielefeld) Alexia Fürnkranz-Prskawetz (Wien) Beat Fux (Salzburg) Joshua Goldstein (Berkeley) Karsten Hank (Köln) Sonja Haug (Regensburg) Hill Kulu (Liverpool) Aart C. Liefbroer (Den Haag) Kurt Lüscher (Konstanz) Emma Lundholm (Umeå) Nadja Milewski (Rostock) Dimiter Philipov (Wien) Roland Rau (Rostock) Tomáš Sobotka (Wien) Jeroen Spijker (Barcelona) Olivier Thévenon (Paris) Helga de Valk (Brussel) Heike Trappe (Rostock) Michael Wagner (Köln) © Federal Institute for Population Research 2015 – All rights reserved © Federal Institute for Population Research 2015 – All rights reserved
https://openalex.org/W2617034330	https://europepmc.org/articles/pmc5487087?pdf=render	English	null	Limited Service Availability, Readiness, and Use of Facility-Based Delivery Care in Haiti: A Study Linking Health Facility Data and Population Data	Global health science and practice	2,017	cc-by	11,611	ABSTRACT B k d Background: Understanding the barriers that women in Haiti face to giving birth at a health facility is important for improving coverage of facility delivery and reducing persistently high maternal mortality. We linked health facility survey data and population survey data to assess the role of the obstetric service environment in affecting women’s use of facility delivery care. y Methods: Data came from the 2012 Haiti Demographic and Health Survey (DHS) and the 2013 Haiti Service Provision Assessment (SPA) survey. DHS clusters and SPA facilities were linked with their geographic coordinate information. The final analysis sample from the DHS comprised 4,921 women who had a live birth in the 5 years preceding the survey. Service availability was measured with the number of facilities providing delivery services within a specified distance from the cluster (within 5 kilometers for urban areas and 10 kilometers for rural areas). We measured facility readiness to provide obstetric care using 37 indicators defined by the World Health Organization. Random-intercept logistic regres- sions were used to model the variation in individual use of facility-based delivery care and cluster-level service availability and readiness, adjusting for other factors. Results: Overall, 39% of women delivered their most recent birth at a health facility and 61% delivered at home, with disparities by residence (about 60% delivered at a health facility in urban areas vs. 24% in rural areas). About one-fifth (18%) of women in rural areas and one-tenth (12%) of women in nonmetropolitan urban areas lived in clusters where no facility offered delivery care within the specified distances, while nearly all women (99%) in the metropolitan area lived in clusters that had at least 2 such facilities. Urban clusters had better service readiness compared with rural clusters, with a wide range of variation in both areas. Regression models indicated that in both rural and nonmetropolitan urban areas availability of delivery services was significantly associated with women’s greater likelihood of using facility-based deliv- ery care after controlling for other covariates, while facilities’ readiness to provide delivery services was also important in nonmetropolitan urban areas. Conclusion: Increasing physical access to delivery care should become a high priority in rural Haiti. In urban areas, where delivery services are more available than in rural areas, improving quality of care at facilities could potentially lead to increased coverage of facility delivery. aThe Demographic and Health Surveys (DHS) Program, ICF, Rockville, MD, USA. Correspondence to Wenjuan Wang (wenjuan.wang@icf.com). Limited Service Availability, Readiness, and Use of Facility- Based Delivery Care in Haiti: A Study Linking Health Facility Data and Population Data Wenjuan Wang,a Michelle Winner,a Clara R Burgert-Bruckera Wenjuan Wang,a Michelle Winner,a Clara R Burgert-Bruckera Proximity to a health facility offering delivery services and readiness of the facilities to provide such services were poor in both rural and urban areas outside of Port-au-Prince. Availability of a proximate facility was significantly associated with women in rural and urban areas delivering at a facility, as was the quality of delivery care available at the facilities but only in urban areas. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Physical accessibility remains one of the biggest barriers to using health care in Haiti due to the mountainous terrain. INTRODUCTION A few studies have looked at social contextual factors such as community norms, media access, and the level of local devel- opment.12–15 In Haiti, a few studies have found that facility delivery was associated with mater- nal and birth characteristics, household poverty, use of antenatal care services, and community exposure to mass media.16,17 The effect of the supply side—delivery care offered in health facilities—on facility delivery has received only limited attention.8,11 One of the main reasons that research on the effects of service provision is limited is lack of suitable data. Supply-side data typically come from health facilities and need to be linked to data from population-based surveys in order to explore the relationship between the provision of services and women’s use of facility-based delivery care. Because of Haiti’s mountainous terrain, physi- cal accessibility remains one of the biggest barriers to using health care.22,23 Gage and Guirlène Calixte linked women’s reports on use of facility delivery care and community key informants’ reports on health services and found that in rural Haiti the physical accessibility of health facilities was strongly associated with women’s use of delivery care.23 Women’s odds of being attended by a skilled birth attendant were positively associ- ated with the presence of a health worker provid- ing antenatal care in the neighborhood but negatively associated with living in a mountain- ous terrain and with distance from the nearest hospital. Ruktanonchai and colleagues found that in the 5 East Africa countries studied, geographic inaccessibility was an important predictor of use of maternal health care services, including skilled birth attendance.24 This furthers the argument that straight-line distance linkage between a facil- ity and a DHS displaced cluster may cause misclas- sification in the Haiti context. Thus, the service environment approach is a better approximation of the likely access to health facilities for a woman. With the availability of geographic data from both household surveys and health facility sur- veys, it has become possible to link population data and health facility data within a geo- graphic information system (GIS). INTRODUCTION thousands of women in Haiti die from causes that could be prevented by access to comprehensive and skilled obstetric care during pregnancy, childbirth, and the postpartum period.2,3 Use of maternal health services, especially facility delivery, remains low in Haiti. Only 36% of births take place in health facilities, according to the 2012 Haiti Demographic and Health Survey (DHS).4 Unless a woman delivers at a health facility, she is H aiti has the highest maternal mortality ratio in Latin America and the Caribbean, at an estimated 359 deaths per 100,000 live births.1 Every year, 244 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org While these studies have contributed to estab- lishing a geospatial methodology to assess the relationship between service provision and use, linking DHS clusters and closest facilities is subject to misclassification errors. Because DHS cluster coordinates are displaced before release of the data (to ensure respondent confidentiality is maintained), the closest facility identified based on the released geographic data may not be the nearest facility in reality. Skiles and colleagues indicated that, due to displacement of cluster coor- dinates, the distance to the closest facility can be misclassified for 34% to 43% of clusters.21 The displacement is an important limitation of the data, so instead of only looking at the closest facil- ity our study measured the service environment (all available facilities) within a reasonable dis- tance of the displaced cluster, thus representing the likely service environment of the real cluster location. unlikely to have access to emergency obstetric care, which is considered the most important strat- egy for reducing maternal deaths.5,6 Increasing use of facility delivery is critical for reducing maternal mortality in Haiti. An extensive body of literature exists on fac- tors that influence facility delivery.7–11 The ma- jority of studies have focused on the demand side—for example, the characteristics of women and their families. Increasing use of facility delivery is critical for reducing maternal mortality in Haiti. Another key determinant of service utilization, beyond physical proximity, is the quality of care of services. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 We limited our analysis to women who had a live birth in the 5 years preceding the survey, with the final analysis sample comprising 4,921 women. INTRODUCTION In Nepal, when quality of care was meas- ured solely in structural terms (e.g., infrastruc- ture, availability of medicine, number of staff) a significant effect was also seen on the use of ante- natal care and immunization services.29 eligible for individual interview, which collected data on their socio-demographic characteristics and use of health care services. A total number of 14,287 women were interviewed. The Haiti DHS georeferenced the locations of the sampled clusters by using Global Positioning System (GPS) receivers to collect the coordinates of the center of the populated areas of the clusters. Prior to release of the geographic dataset, the clus- ter coordinates were verified and geographically displaced.30 Coordinates of urban clusters were displaced up to a maximum distance of 2 kilo- meters (km); average urban displacement was 0.8 km. In rural areas, the displacement distance was up to 5 km with a further, randomly selected 1% of rural clusters displaced up to 10 km; average rural displacement was 2.1 km. These displaced GPS locations were used in our analysis. Haiti has a hierarchical system of health care provision in which small facilities are located in villages or small communities and larger, better- equipped facilities are located in cities. Small or low-level facilities may not provide delivery serv- ices, however, and among those that provide these services there is little information on how prepared they are to provide good-quality delivery care and how their service preparedness affects utilization. The Haiti 2013 Service Provision Ass- essment (SPA) and 2012 DHS provide an opportu- nity to link data for health facilities and DHS clusters in order to explore the influence of service availability and readiness on use of delivery care in facilities. Given that the outcome of interest was facility delivery, we limited the analysis to women who had a live birth in the 5 years preceding the survey (N=5,218) with a focus on their most recent birth. We excluded 234 women interviewed in 45 camp clusters that housed the population displaced by the 2010 earthquake because they were likely to have lived in a different area when they had their most recent birth. Thus, the health care environ- ment where they had the birth was likely different from where they had been surveyed. Another 63 women were excluded from the analysis because they were from 8 clusters with missing georeferenced data. INTRODUCTION A few studies have linked DHS data and facility census data to assess how distance to the closest facility affects women’s use of reproductive health serv- ices.18–20 For example, a study in Malawi and Zambia linking DHS clusters and health facilities (from a facility census) found that in both coun- tries a longer straight-line distance from the DHS cluster to the closest facility offering emergency obstetric care significantly reduced the likelihood of facility delivery.19 Another study in Zambia with the same methodology but focusing on antenatal care found that distance to the closest facility had a significant effect on the content of care women received but had no effect on num- ber of antenatal care visits or timing of the first visit.18 In a rural setting in Ghana, Nesbitt and colleagues linked health facility census data and health and demographic surveillance data from about 600 villages and found a significant associ- ation between distance to the closest delivery fa- cility and women’s likelihood of delivering in a health facility.20 While physical access is important, another key determinant of service utilization is the qual- ity of care. Families may bypass the nearest health facility when quality is an issue.25–27 In examining the effect of the quality of care on use of services, some studies have looked at the quality of care from the user’s perspective.15,28 While this per- spective is indicative, it is subject to the respond- ents’ level of knowledge about the services provided at health facilities, which may be biased or misinformed. Among the limited number of 245 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org studies based on linked population data and facil- ity data, a few have looked at the quality of service provision in health facilities. The study in Zambia by Kyei and colleagues measured the level of care using an index that combined several process and structural aspects of antenatal care provided at facilities.18 The level of service provision at the closest health facility was found to be significantly associated with the content of antenatal care received. This study used geographic data collected in the Haiti Demographic and Health Survey (DHS) and the Service Provision Assessment (SPA) to link DHS clusters and SPA facilities. Data The Haiti DHS provides data on women’s socio- demographic characteristics and their use of maternal health care services, including facility- based delivery care. The Haiti SPA provides information on the availability of delivery care at health facilities and the readiness of facilities to provide good-quality services. This study used ge- ographic data collected in both surveys to link DHS clusters and SPA facilities. INTRODUCTION The final analysis sample con- sisted of 4,921 women from 392 clusters with GPS data available. Health Facility Data y The 2013 Haiti SPA is a health facility census of 907 public and private facilities, from hospitals at the highest level to dispensaries at the lowest level. The SPA provides data on availability of key health services and readiness to provide these services. Data were collected using 4 types of survey instruments: the inventory questionnaire, health provider interviews, observation of con- sultations, and client exit interviews. For our analysis, facility data primarily came from the in- ventory questionnaire, which was administered to the facility manager or the most knowledgeable person for specific service areas. The inventory questionnaire collects data on the facility’s infra- structure, supplies, medicines, staffing, trainings, and routine practices in providing general and specific services. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Data on 195 hospitals and health centers with a bed that offered delivery services were analyzed. Population Data The 2012 Haiti DHS is a population-based house- hold survey that provides representative esti- mates for both urban and rural areas and for the 10 administrative departments of Haiti. The sur- vey used a 2-stage cluster sampling design. At the first stage, 445 clusters were selected with probability proportional to their population size from a master national sample frame. At the sec- ond stage, a systematic sample of households was drawn in each of the selected clusters. All women ages 15–49 in the sampled households were The SPA georeferenced the locations of the health facilities by using GPS receivers. Unlike the 246 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org tance from an urban cluster and a 10-km buffer distance from a rural cluster. These buffer sizes around the displaced location were chosen to ensure that facilities within the same distance around the real (non-displaced) location of the cluster were included within the buffer, given the displacement radius used for urban (maximum 2 km) and rural locations (maximum 10 km). We also believe that facilities within the chosen buffer distance reasonably represent the service environ- ment where individuals seek health services, if not the exact facilities they visit. Lastly, data from all the facilities within the buffer distance were sum- marized for each cluster to measure the cluster’s service environment. DHS GPS data for each cluster, facilities’ GPS coor- dinates were not displaced. Data on 195 hospitals and health centers with a bed that offer delivery services were analyzed in this study. We excluded health centers without a bed and dispensaries from the analysis because they are not mandated to provide delivery care and are rarely used for delivery care.4,31 Figure 1 shows the location of the SPA facilities as well as the DHS (displaced) clusters that were included in the analysis. Linked Data Between DHS Clusters and SPA Facilities Using publicly available GPS data for DHS (dis- placed) clusters and SPA facilities, we linked facili- ties to clusters to measure the health service environment of the clusters. We first created a dis- tance matrix with the direct distance measure- ment from every DHS (displaced) cluster location to every health facility. The distances were then used to identify facilities within a 5-km buffer dis- Definitions of Key Measurements. We measured a facility’s readiness to provide good- quality delivery care by a readiness score based on 37 indicators defined by WHO. Statistical Analysis The analysis was stratified by urban and rural res- idence because of differences between urban and rural areas in the health service environment and women’s health care-seeking behaviors. We fur- ther separated the Port-au-Prince metropolitan area (comprising the capital city Port-au-Prince and the urban zones of the Ouest region) from other urban areas because of the substantial differ- ences in the density and types of health facilities. At the facility level, we measured availability using the SPA definition of facilities offering deliv- ery care, which the SPA obtained by asking the facility manager if the facility provides delivery services. We measured facility’s readiness to provide good-quality delivery care by a readiness score created with principal component analysis based on a set of service readiness indicators defined by the World Health Organization.32 For each indicator—according to whether the facility met the criteria for availability—facilities were assigned a binary variable: 1=available, 0=un- available. A total of 37 readiness indicators were constructed; their definitions are presented in Supplementary Table 1. The readiness score was computed based on the first component resulting from the principal component analysis, which explained the largest proportion of the total var- iance. The readiness score is a relative summary indicator of how ready a health facility is to pro- vide good-quality delivery services. A higher score represents better readiness and a lower score rep- resents poorer readiness compared with other facilities. Availability of delivery service and read- iness indicators were then summarized to the DHS cluster level, as described below. Random-intercept logistic regressions were used for the multivariable analysis. DHS data fol- low a hierarchical structure—that is, individuals are nested within households and households are nested within clusters. Respondents who live in the same household or cluster may not be inde- pendent of one another. Moreover, the outcome variable is at the individual level but the key ex- ploratory variables—level of service availability and readiness of facilities to provide delivery services—are at the cluster level. A multilevel analysis approach is more appropriate to allow for simultaneous investigation of the effect of the group-level and individual-level predictors on individual-level outcomes.33 Therefore, we applied multilevel (individual- and cluster-level) random intercept logistic models to investigate how the service environment measures affect women’s use of facility-based delivery care. Population Data The outcome variable was dichotomous, indicating whether a woman delivered her most recent birth at a health facility. We measured the service envi- ronment with 2 indicators: availability of facilities FIGURE 1. Geographic Distribution of 2012 DHS (Displaced) Clusters and 2013 SPA Facilities Included in the Analysis Abbreviations: DHS, Demographic and Health Survey; SPA, Service Provision Assessment. Abbreviations: DHS, Demographic and Health Survey; SPA, Service Provision Assessment. Abbreviations: DHS, Demographic and Health Survey; SPA, Service Provision Assessment. 247 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org the low-level readiness group; and the rest were put in the medium-level readiness group. offering delivery care and facilities’ readiness to provide good-quality delivery care. Both were measured at the DHS cluster level and derived from the facility-level data after linking facilities to DHS clusters. We categorized clusters into groups with 3 levels of availability of delivery services: low, medium, or high. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Statistical Analysis The household level was omitted since there were few women who had a child in the 5 years preceding the survey living in the same household. We categorized clusters into groups with 3 levels of availability of delivery services: low, medium, or high. To measure availability of delivery services, we counted the number of facilities offering delivery services within the specified buffer of each DHS cluster. We categorized clusters into groups with 3 levels of availability: We measured facilities’ readiness to provide delivery services by the median readiness score of all the facilities within the buffer and divided clusters into low-, medium-, or high- level readiness groups based on the score terciles. Other variables adjusted for included women’s age at birth, birth order, mother’s education, household wealth quintile, number of antenatal care visits, and region (department), all of which have been found to be associated with facility delivery.12,34 1. Low availability: no facility with delivery serv- ices within the buffer distance 2. Medium availability: 1 facility with delivery services within the buffer distance 3. High availability: 2 or more facilities with delivery services within the buffer distance Most facilities offering delivery services were hospitals in the metropolitan area and health centers in rural areas. Sample Characteristics Facilities’ readiness to provide delivery care was measured by the median readiness score of all the facilities within the buffer. Given that the readiness score is a relative measurement, we di- vided the clusters into low-, medium-, and high- level readiness groups based on the score terciles at the cluster level. Clusters with a median score (median score of all facilities within the buffer) falling in the top 33% were considered in the high-level readiness group; those with a median score in the bottom 33% were categorized into For health facilities, we analyzed data on hospitals and health centers with a bed that provide deliv- ery services. The distribution of these facilities in each residence location by facility background characteristics is presented in Supplementary Table 2. Overall, hospitals and health centers each accounted for half the facilities; but in the metropolitan area most facilities offering delivery services were hospitals, while in rural areas most were health centers. With regard to the managing Most facilities offering delivery services were hospitals in the metropolitan area and health centers in rural areas. 248 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org authority, more than half of the facilities in the metropolitan area were private for-profit, while government health facilities were more common in other urban areas and rural areas. Facility distri- bution also varied across regions in rural and other urban areas. In rural areas, some regions, such as the Ouest region, had many health facilities, while some, such as Sud-Est and Grand-Anse, had only a few. marily in the upper 3 wealth quintiles, while in ru- ral areas 71% of women were in the poorest 2 quintiles. As for women’s receipt of antenatal care during pregnancy for their most recent birth, more than three-quarters of women in the metro- politan and other urban areas had 4 or more antenatal care visits compared with less than two-thirds (60%) of rural women. For use of facility-based delivery services, data were based on 4,921 women who had a live birth in the 5 years preceding the DHS survey (includ- ing 2,878 women in rural areas, 1,214 in the met- ropolitan area, and 829 in other urban areas). Supplementary Table 3 provides the distribution of these women by background characteristics and receipt of antenatal care for the most recent birth. Sample Characteristics In all 3 location categories, a majority of the women reported having their most recent birth between ages 20–34, while 14% to 15% were under age 20 when they had their most recent birth. Most women in the metropolitan and other urban areas had 3 children or fewer, with about 40% having the most recent birth as their first birth. In contrast, in rural areas a higher percent- age of women (38%) had 4 children or more. In the metropolitan and other urban areas, most women reported having a primary education and close to 60% had secondary or higher educa- tion. In contrast, in rural areas only 25% of women had secondary or higher education, while 27% had no education. In the metropolitan and other urban areas, women in the study were pri- Availability of Delivery Services Availability of facilities offering delivery services varied by location; there was a greater number of facilities within the buffer of clusters in the metro- politan area than in other urban areas and rural areas. The number of facilities with delivery serv- ices within the buffer distances from the DHS clus- ters (i.e., within 5 km for the metropolitan and other urban areas and within 10 km for rural areas) ranged from 0–28 in the metropolitan area, from 0–6 in other urban areas, and from 0–7 in ru- ral areas. Figure 2 presents the percentage distri- bution of women by levels of availability to delivery services. Proximity to a facility with deliv- ery services was nearly universal in the metropol- itan area; 99% of women lived in an area with a high level of availability of delivery services (2 or more facilities within 5 km from their clusters). In contrast, 12% of women in other urban areas and 18% of rural women lived in areas with low avail- ability—that is, no facility offering delivery serv- ices within 5 km for women in other urban areas and 10 km for women in rural areas. Proximity to a facility with delivery services was nearly universal in the metropolitan area, whereas 12% of women in other urban areas and 18% of rural women lived in areas with no such facility. FIGURE 2. Percent Distribution of Women by Level of Availability of Facilities With Delivery Services Within the Buffer Distance,a Haiti, 2012–2013 18 12 1 27 44 55 44 99 0 20 40 60 80 100 Rural Other urban Metropolitan Percentage Low availability Medium availability High availability a Low availability=no facility with delivery services within the buffer distance; medium availability=1 facility with delivery services within the buffer; high availability=2 or more facilities with delivery services within the buffer. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. FIGURE 2. Percent Distribution of Women by Level of Availability of Facilities With Delivery Services Within the Buffer Distance,a Haiti, 2012–2013 18 12 1 27 44 55 44 99 0 20 40 60 80 100 Rural Other urban Metropolitan Percentage Low availability Medium availability High availability a Low availability=no facility with delivery services within the buffer distance; medium availability=1 facility with delivery services within the buffer; high availability=2 or more facilities with delivery services within the buffer. Availability of comprehensive obstetric care items was low across the board, but especially in facilities in rural areas. Facilities’ Readiness to Provide Delivery Services with 2 or more facilities within 5 km, the effect of availability of delivery services was assessed only for other urban and rural areas. In both types of area, use of facility delivery was significantly associated with levels of availability of facilities offering delivery services. In rural areas, facility delivery coverage increased incrementally with the level of availability. In rural clusters without a facility with delivery services within 10 km, only 8% of women delivered the most recent birth in a facility compared with 19% of women in clusters with access to 1 such facility, and 32% of those with 2 or more such facilities. Service availability was also associated with use in other urban areas but the difference was less remarkable between medium and high availability. Readiness to provide delivery services among facilities that offered delivery services was meas- ured by the availability of a number of basic and comprehensive obstetric care services and items available at the facility on the day of the survey. Table 1, which presents the availability of these 37 services and supplies, shows wide variation by the facility’s location. Facilities in the metropoli- tan and other urban areas had many of the items while those in rural areas had few. Some items were commonly available across locations, such as a delivery bed, gloves, injectable antibiotics, and parenteral administration of oxytocic drug, while other items were rare, such as manual vac- uum extractor and guidelines for integrated man- agement of pregnancy and childbirth. Overall, the availability of comprehensive obstetric care items was low, especially at facilities in rural areas. Figure 5 indicates that use of facility delivery services was also positively associated with the level of readiness at the facilities within the buffer distance in the metropolitan and other urban areas. In other urban clusters, 45% of women delivered in a facility with low level of readiness compared with 63% of women for facilities with a medium level of readiness and 71% of women for facilities with a high level of readiness. In rural areas, however, there was little association between facilities’ level of readiness and women’s use of facility delivery services. As discussed earlier, to assess the level of serv- ice readiness at facilities that the DHS cluster was linked to, an overall readiness score was generated for each cluster based on the median score of all facilities within the buffer. Availability of Delivery Services The buffer distance was 5 km for urban clusters and 10 km for rural clusters. 249 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org Results of Multivariable Analysis Using random-intercept logistic models, we ass- essed how women’s use of facility delivery was associated with availability of delivery services within the buffer and service readiness of the facilities. We first ran models for nonmetropolitan urban areas and rural areas to examine the effect of availability. This model was not run for the metropolitan area since almost all metropolitan women had a high level of availability to delivery services. We then ran models to examine the effect of service readiness for all 3 locations after controlling for availability and other covariates. Since readiness is only available for clusters with at least 1 facility within the buffer distance, women in clusters with no facility were dropped from this model. Facilities’ Readiness to Provide Delivery Services The clusters were di- vided into low-, medium-, and high-readiness groups based on the readiness score terciles at the cluster level. Figure 3 shows that the distribution of readiness scores differed by residence area. In the metropolitan area, the cluster readiness scores were more homogenous, while the interquartile range was wider in other urban and rural clusters. The readiness score distribution in other urban areas was highly skewed to high readiness scores, indicating that some facilities had much better readiness than others. A few outlier facilities with markedly different (better or worse) readiness compared with the rest of facilities were also observed in each residence area. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 In both urban and rural areas, availability of delivery services was significantly associated with women’s use of facility delivery care. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Overall, 39% of women delivered their most recent birth at a health facility and 61% delivered at home. Delivery at a health facility was far more common in the metropolitan and other urban areas (60% and 59%, respectively) compared with rural areas (24%). Table 2 presents odds ratios (ORs) and 95% confidence intervals (CIs) from the regression models of facility delivery on the availability indi- cator, as well as for the covariates adjusted for in the models. In both urban and rural areas, avail- ability of delivery services was significantly associ- ated with women’s use of facility delivery care after controlling for other covariates. In rural Figure 4 highlights the bivariate association between availability of delivery services and the percentage of women who delivered their most recent birth at a health facility. Since nearly all women in the metropolitan area lived in clusters 250 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org TABLE 1. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Continued Rural % Other Urban % Metropolitan % Total % Guidelines for CEmOC adapted for Haiti 6.8 15.9 7.7 10.8 Staff member providing delivery trained in CEmOC 44.6 45.1 38.5 43.6 Anesthesia equipment 10.8 15.9 25.6 15.9 Incubator 10.8 23.2 23.1 18.5 Blood typing 17.6 42.7 56.4 35.9 Cross matching test 4.1 3.7 10.3 5.1 Blood supply sufficiency 12.2 19.5 20.5 16.9 Blood supply safety 16.2 37.8 53.8 32.8 Total number of facilities 74 82 39 195 Abbreviations: CEmOC, comprehensive emergency obstetric care; D&C, dilation and curettage; IMPAC, integrated management of pregnancy and childbirth. a Assisted vaginal delivery was actually interpreted as "attended vaginal delivery" in the Haiti Service Provision Assessment. FIGURE 3. Distribution of Cluster-Level Median Readiness Scores Among Facilities Offering Delivery Services, by Residence, Haiti, 2012–2013 -2 -1 0 1 2 Median readiness score Rural Other urban Metropolitan Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 252 TABLE 1. Continued Rural % Other Urban % Metropolitan % Total % Guidelines for CEmOC adapted for Haiti 6.8 15.9 7.7 10.8 Staff member providing delivery trained in CEmOC 44.6 45.1 38.5 43.6 Anesthesia equipment 10.8 15.9 25.6 15.9 Incubator 10.8 23.2 23.1 18.5 Blood typing 17.6 42.7 56.4 35.9 Cross matching test 4.1 3.7 10.3 5.1 Blood supply sufficiency 12.2 19.5 20.5 16.9 Blood supply safety 16.2 37.8 53.8 32.8 Total number of facilities 74 82 39 195 Abbreviations: CEmOC, comprehensive emergency obstetric care; D&C, dilation and curettage; IMPAC, integrated management of pregnancy and childbirth. a Assisted vaginal delivery was actually interpreted as "attended vaginal delivery" in the Haiti Service Provision Assessment. FIGURE 3. Distribution of Cluster-Level Median Readiness Scores Among Facilities Offering Delivery Services, by Residence, Haiti, 2012–2013 -2 -1 0 1 2 Median readiness score Rural Other urban Metropolitan FIGURE 3. Distribution of Cluster-Level Median Readiness Scores Among Facilities Offering Delivery Services, by Residence, Haiti, 2012–2013 GURE 3. Distribution of Cluster-Level Median Readiness Scores Among Facilities Offering Deliver rvices, by Residence, Haiti, 2012–2013 -2 -1 0 1 2 Median readiness score Rural Other urban Metropolitan areas, women in clusters with medium-level availability to delivery services had 1.9 times higher odds of going to a facility for delivery (CI, 1.14 to 3.13; P<.05) and women in clus- ters with high availability had 3.6 times higher odds (CI, 2.23 to 5.66; P<.001) compared with women with low availability. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Although the odds ratio appears lower for high availability (OR, 2.64; 95% CI, 1.13 to 6.19; P<.05) than medium availability (OR, 3.83; 95% CI, 1.57 to 9.31; P<.01), the difference was not statistically signifi- cant (results not shown). As expected, maternal age at birth, the number of antenatal care visits, and household wealth were positively and significantly associated with TABLE 1. Continued Rural % Other Urban % Metropolitan % Total % Guidelines for CEmOC adapted for Haiti 6.8 15.9 7.7 10.8 Staff member providing delivery trained in CEmOC 44.6 45.1 38.5 43.6 Anesthesia equipment 10.8 15.9 25.6 15.9 Incubator 10.8 23.2 23.1 18.5 Blood typing 17.6 42.7 56.4 35.9 Cross matching test 4.1 3.7 10.3 5.1 Blood supply sufficiency 12.2 19.5 20.5 16.9 Blood supply safety 16.2 37.8 53.8 32.8 Total number of facilities 74 82 39 195 Abbreviations: CEmOC, comprehensive emergency obstetric care; D&C, dilation and curettage; IMPAC, integrated management of pregnancy and childbirth. a Assisted vaginal delivery was actually interpreted as "attended vaginal delivery" in the Haiti Service Provision Assessment. FIGURE 3. Distribution of Cluster-Level Median Readiness Scores Among Facilities Offering Delivery Services, by Residence, Haiti, 2012–2013 -2 -1 0 1 2 Median readiness score Rural Other urban Metropolitan Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 252 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org areas, women in clusters with medium-level availability to delivery services had 1.9 times higher odds of going to a facility for delivery (CI, 1.14 to 3.13; P<.05) and women in clus- ters with high availability had 3.6 times higher odds (CI, 2.23 to 5.66; P<.001) compared with women with low availability. Similarly, in other urban areas higher availability of delivery services was also significantly associated with a greater likelihood to deliver in a facility. Although the odds ratio appears lower for high availability (OR, 2.64; 95% CI, 1.13 to 6.19; P<.05) than medium availability (OR, 3.83; 95% CI, 1.57 to 9.31; P<.01), the difference was not statistically signifi- cant (results not shown). As expected, maternal age at birth, the number of antenatal care visits, and household wealth were positively and significantly associated with TABLE 1. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Percentage of Facilities Offering Delivery Services With Availability of Basic and Comprehensive Obstetric Care Services and Items, by Residence, Haiti, 2013 Rural % Other Urban % Metropolitan % Total % Basic obstetric care indicators Parenteral administration of antibiotics 68.9 78.0 94.9 77.9 Parenteral administration of oxytocic drug 86.5 95.1 94.9 91.8 Parenteral administration of anticonvulsants 40.5 67.1 48.7 53.3 Assisted vaginal deliverya 90.5 92.7 89.7 91.3 Manual removal of placenta 58.1 79.3 59.0 67.2 Manual removal of retained products 48.6 75.6 64.1 63.1 Neonatal resuscitation 51.4 72.0 53.8 60.5 Guidelines for IMPAC 21.6 31.7 12.8 24.1 Staff trained in IMPAC 48.6 53.7 46.2 50.3 Emergency transportaion 32.4 36.6 35.9 34.9 Sterilization equipment 75.7 84.1 87.2 81.5 Examination light 43.2 43.9 59.0 46.7 Delivery pack 87.8 92.7 87.2 89.7 Suction apparatus 93.2 98.8 79.5 92.8 Manual vacuum extractor 16.2 19.5 20.5 18.5 Vacuum aspirator or D&C kit 20.3 35.4 30.8 28.7 Newborn bag and mask 51.4 62.2 59.0 57.4 Delivery bed 97.3 100.0 94.9 97.9 Partograph 32.4 41.5 33.3 36.4 Gloves 93.2 93.9 89.7 92.8 Antibiotic eye ointment for newborns 66.2 84.1 66.7 73.8 Injectable uterotonic 62.2 74.4 64.1 67.7 Injectable antibiotics 89.2 92.7 94.9 91.8 Injectable magnesium sulphate 60.8 78.0 74.4 70.8 Skin disinfectant 60.8 81.7 61.5 69.7 Intravenous solution with infusion set 74.3 84.1 82.1 80.0 Regular reviews of maternal or newborn deaths 25.7 32.9 48.7 33.3 Comprehensive obstetric care indicators Cesarean delivery services 23.0 52.4 71.8 45.1 Blood transfusion 23.0 50.0 61.5 42.1 Continued 251 TABLE 1. Percentage of Facilities Offering Delivery Services With Availability of Basic and Comprehensive Obstetric Care Services and Items, by Residence, Haiti, 2013 R l % O h U b % M li % T l % 251 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 areas, women in clusters with medium-level availability to delivery services had 1.9 times higher odds of going to a facility for delivery (CI, 1.14 to 3.13; P<.05) and women in clus- ters with high availability had 3.6 times higher odds (CI, 2.23 to 5.66; P<.001) compared with women with low availability. Similarly, in other urban areas higher availability of delivery services was also significantly associated with a greater likelihood to deliver in a facility. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services b Women living in clusters with no health facility offering delivery services within the buffer distance were excluded from this figure since the readiness score was applied only to clusters with 1 or more health facilities. FIGURE 4. Percentage of Women Who Delivered at a Facility by Availability of Facilities With Delivery Services Within the Buffer Distance,a Haiti, 2012–2013 8 37 19 61 32 63 0 20 40 60 80 100 n a b r u r e h t O la r u R Percentage Low availability Medium availability High availability a Low availability=no facility with delivery services within the buffer distance; medium availability=1 facility with delivery services within the buffer; high availability=2 or more facilities with delivery services within the buffer. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. a Low availability=no facility with delivery services within the buffer distance; medium availability=1 facility with delivery services within the buffer; high availability=2 or more facilities with delivery services within the buffer. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. FIGURE 5. Percentage of Women Who Delivered at a Facility by Level of Readinessa at Facilities With Delivery Services Within the Buffer Distance,b Haiti, 2012–2013 Rural Other urba Metropolitan s 25 63 61 27 45 55 25 63 61 30 71 70 0 20 40 60 80 100 Rural Other urban Metropolitan Percentage Low readiness Medium readiness High readiness a Low-, medium-, and high-level readiness=median readiness score of the facilities within the buffer falling in the bottom 33.3%, middle 33.3%, and top 33.3%, respectively. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. b Women living in clusters with no health facility offering delivery services within the buffer distance were excluded from this figure since the readiness score was applied only to clusters with 1 or more health facilities. FIGURE 5. Percentage of Women Who Delivered at a Facility by Level of Readinessa at Facilities With Delivery Services Within the Buffer Distance,b Haiti, 2012–2013 a Low-, medium-, and high-level readiness=median readiness score of the facilities within the buffer falling in the bottom 33.3%, middle 33.3%, and top 33.3%, respectively. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. Service readiness was significantly associated with facility delivery only in urban areas. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Similarly, in other urban areas higher availability of delivery services was also significantly associated with a greater likelihood to deliver in a facility. Although the odds ratio appears lower for high availability (OR, 2.64; 95% CI, 1.13 to 6.19; P<.05) than medium availability (OR, 3.83; 95% CI, 1.57 to 9.31; P<.01), the difference was not statistically signifi- cant (results not shown). As expected, maternal age at birth, the number of antenatal care visits, and household wealth were positively and significantly associated with 252 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org women’s use of delivery care in a health facility, while the child’s birth order was negatively associated. Table 3 shows the results of the regression models of facility delivery on service readiness and availability after controlling for women’s characteristics and antenatal care variables. Service readiness was significantly associated with facility delivery only in urban areas (other than the metropolitan area). Women in clusters with a high level of service readiness had 2.7 times higher odds of delivery in a health facility FIGURE 4. Percentage of Women Who Delivered at a Facility by Availability of Facilities With Delivery Services Within the Buffer Distance,a Haiti, 2012–2013 8 37 19 61 32 63 0 20 40 60 80 100 n a b r u r e h t O la r u R Percentage Low availability Medium availability High availability a Low availability=no facility with delivery services within the buffer distance; medium availability=1 facility with delivery services within the buffer; high availability=2 or more facilities with delivery services within the buffer. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. FIGURE 5. Percentage of Women Who Delivered at a Facility by Level of Readinessa at Facilities With Delivery Services Within the Buffer Distance,b Haiti, 2012–2013 Rural Other urba Metropolitan s 25 63 61 27 45 55 25 63 61 30 71 70 0 20 40 60 80 100 Rural Other urban Metropolitan Percentage Low readiness Medium readiness High readiness a Low-, medium-, and high-level readiness=median readiness score of the facilities within the buffer falling in the bottom 33.3%, middle 33.3%, and top 33.3%, respectively. The buffer distance was 5 km for urban clusters and 10 km for rural clusters. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services b Women living in clusters with no health facility offering delivery services within the buffer distance were excluded from this figure since the readiness score was applied only to clusters with 1 or more health facilities. women’s use of delivery care in a health facility, while the child’s birth order was negatively associated. characteristics and antenatal care variables. Service readiness was significantly associated with facility delivery only in urban areas (other than the metropolitan area). Women in clusters with a high level of service readiness had 2.7 times higher odds of delivery in a health facility Table 3 shows the results of the regression models of facility delivery on service readiness and availability after controlling for women’s 253 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org essions of Facility Delivery on Availability of Facilities With Delivery ariables, Haiti, 2012–2013 Rural Other Urban Odds Ratio 95% CI Odds Ratio 95% CI 1.00 1.00 1.89* 1.14, 3.13 3.83 1.57, 9.31 3.56* 2.23, 5.66 2.64* 1.13, 6.19 1.06* 1.03, 1.08 1.08* 1.04, 1.13 1.00 1.00 0.30* 0.23, 0.39 0.30* 0.19, 0.48 0.15* 0.10, 0.22 0.23* 0.12, 0.46 0.10* 0.06, 0.17 0.14* 0.05, 0.37 1.00 1.00 1.37 0.99, 1.89 0.95 0.48, 1.89 1.85 1.27, 2.68 1.64 0.80, 3.35 1.00 NAa NAa 1.86* 1.38, 2.50 NAa NAa 2.99* 2.12, 4.20 1.00 5.72* 3.61, 9.07 1.43 0.91, 2.25 10.73* 4.83, 23.80 3.39* 1.91, 6.04 1.00 1.00 4.46* 2.05, 9.69 1.38 0.36, 5.29 3.90* 2.09, 7.28 1.40 0.51, 3.82 6.75*** 3.71, 12.27 2.85* 1.13, 7.17 1.00 1.00 1.17 0.61, 2.23 2.96 0.73, 12.08 0.92 0.53, 1.60 1.83 0.64, 5.28 1.62 0.79, 3.32 1.74 0.53, 5.78 Continued TABLE 2. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 254 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Multivariable Model Results for the Regressions of Facility Delivery on Availability of Facilities With Delivery Services Within the Buffer, Controlling for Other Variables, Haiti, 2012–2013 Variables Rural Other Urban Odds Ratio 95% CI Odds Ratio 95% CI Availability of facilities with delivery services Low availability 1.00 1.00 Medium availability 1.89* 1.14, 3.13 3.83 1.57, 9.31 High availability 3.56* 2.23, 5.66 2.64* 1.13, 6.19 Mother's age at birth 1.06* 1.03, 1.08 1.08* 1.04, 1.13 Birth order 1 1.00 1.00 2–3 0.30* 0.23, 0.39 0.30* 0.19, 0.48 4–5 0.15* 0.10, 0.22 0.23* 0.12, 0.46 6þ 0.10* 0.06, 0.17 0.14* 0.05, 0.37 Education None 1.00 1.00 Primary 1.37 0.99, 1.89 0.95 0.48, 1.89 Secondary or higher 1.85 1.27, 2.68 1.64 0.80, 3.35 Wealth quintile Lowest 1.00 NAa NAa Second 1.86* 1.38, 2.50 NAa NAa Middle 2.99* 2.12, 4.20 1.00 Fourth 5.72* 3.61, 9.07 1.43 0.91, 2.25 Highest 10.73* 4.83, 23.80 3.39* 1.91, 6.04 Antenatal care visits None 1.00 1.00 1 4.46* 2.05, 9.69 1.38 0.36, 5.29 2–3 3.90* 2.09, 7.28 1.40 0.51, 3.82 4þ 6.75*** 3.71, 12.27 2.85* 1.13, 7.17 Departement Ouest 1.00 1.00 Sud-est 1.17 0.61, 2.23 2.96 0.73, 12.08 Nord 0.92 0.53, 1.60 1.83 0.64, 5.28 Nord-est 1.62 0.79, 3.32 1.74 0.53, 5.78 Continued TABLE 2. Multivariable Model Results for the Regressions of Facility Delivery on Availability of Facilities With Delivery Services Within the Buffer, Controlling for Other Variables, Haiti, 2012–2013 254 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org TABLE 2. Continued Variables Rural Other Urban Odds Ratio 95% CI Odds Ratio 95% CI Artibonite 1.17 0.76, 1.78 1.37 0.49, 3.86 Centre 1.64 0.97, 2.78 1.42 0.39, 5.20 Sud 1.56 0.94, 2.59 1.90 0.52, 6.88 Grand'anse 1.21 0.57, 2.57 0.68 0.17, 2.73 Nord-ouest 1.04 0.56, 1.93 0.67 0.19, 2.39 Nippes 1.35 0.67, 2.70 1.71 0.26, 11.14 Number of clusters 241 85 Observations 2,878 829 a In other urban areas, no women were in the first wealth quintile and very few women were in the second wealth quintile; they were combined into the third quintile. *P<.001; P<.01; P<.05. q P<.001; P<.01; P<.05. compared with women in clusters with low readi- ness (95% CI, 1.34 to 5.60; P<.01). The difference in facility delivery between women in clusters with low readiness and those with medium readi- ness was not significant after controlling for other covariates. Use of Facility Delivery Services by Level of Availability and Readiness of Delivery Services Similar to the findings from the models on availability, in other urban areas there was no significant difference in coverage of facility deliv- ery between clusters with medium availability and clusters with high availability of delivery serv- ices. In rural areas, after controlling for service readiness and other variables, having high avail- ability of delivery services—that is, having 2 or more facilities within the buffer distance—was associated with significantly greater odds of deliv- ery in a health facility compared with medium availability—that is, having only 1 facility with delivery services within the buffer distance (OR, 1.94; 95% CI, 1.38 to 2.72; P<.001). 1 in 5 women in rural areas lives in a place where there is no facility that provides delivery services within a 10 km distance. It should be noted that the 10 km distance is a straight-line measurement. The actual travel distance could be longer because of travel paths, road networks, and the mountainous terrain. In rural areas, physical access to health care can be further constrained by poor road conditions and lack of transportation. The regression results highlight the impor- tance of service availability to the use of facility delivery in rural areas. Living reasonably near a fa- cility that provides delivery services is significantly associated with greater probability of women delivering at a health facility. Physical access as an important barrier to service use in rural Haiti was also demonstrated in previous research that measured access with the distance to the nearest facility reported by key community informants.23 Our finding also agrees with previous research that having physical access to services is a strong determinant of use of delivery care.28,34–36 Living reasonably near a facility that provides delivery services is significantly associated with greater probability of women delivering at a health facility. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Women in rural areas of Haiti have very limited physical access to obstetric care. Living reasonably near a facility that provides delivery services is significantly associated with greater probability of women delivering at a health facility. DISCUSSION The availability of the recent DHS and SPA surveys in Haiti, both with GPS data, enabled analysis link- ing women’s use of facility delivery care with their physical access to delivery care and facilities’ read- iness to provide the care. This analysis showed that women in rural areas of Haiti have very lim- ited physical access to obstetric care. Almost Quality of care—measured in our study by analyzing how prepared the facility is to provide the needed services—is important to the use of health services. Our results show that hospitals and health centers with a bed are not well pre- pared to provide delivery services. A large number Women in rural areas of Haiti have very limited physical access to obstetric care. 255 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org TABLE 3. DISCUSSION Multivariable Model Results for the Regressions of Facility Delivery on Service Readiness of Facilities With Delivery Services, Controlling for Availability of Services and Other Variables, Haiti, 2012–2013 Rural Other Urban Metropolitan Variables Odds Ratio 95% CI Odds Ratio 95% CI Odds Ratio 95% CI Facilities' readiness to provide delivery servicesa Low readiness 1.00 1.00 1.00 Medium readiness 1.43 0.93, 2.18 1.74 0.71, 4.26 1.02 0.74, 1.41 High readiness 1.33 0.95, 1.85 2.74* 1.34, 5.60 1.41 0.96, 2.05 Availability of facilities with delivery servicesb Medium availability 1.00 1.00 NA High availability 1.94* 1.38, 2.72 0.72 0.37, 1.41 NA Mother's age at birth 1.05* 1.03, 1.08 1.09* 1.05, 1.14 1.023 0.99, 1.05 Birth order 1 1.00 1.00 1.00 2–3 0.31* 0.23, 0.42 0.28* 0.17, 0.47 0.61 0.43, 0.88 4–5 0.15* 0.10, 0.24 0.22* 0.10, 0.48 0.50* 0.29, 0.86 6þ 0.12* 0.07, 0.20 0.17 0.06, 0.49 0.42* 0.20, 0.91 Education None 1.00 1.00 1.00 Primary 1.47* 1.04, 2.08 0.91 0.42, 1.96 3.23* 1.71, 6.12 Secondary or higher 2.06* 1.39, 3.06 1.91 0.86, 4.24 5.01* 2.63, 9.55 Wealth quintilec Lowest 1.00 NAc NAc NAc NAc Second 1.89* 1.37, 2.61 NAc NAc NAc NAc Middle 3.04* 2.11, 4.38 1.00 1.00 Fourth 5.72* 3.56, 9.21 1.47 0.87, 2.46 1.05 0.68, 1.63 Highest 11.14* 4.95, 25.07 2.82 1.49, 5.35 2.56* 1.55, 4.20 Antenatal care visits None 1.00 1.00 1.00 1 4.75* 2.05, 10.99 2.02 0.47, 8.73 3.60* 1.25, 10.32 2–3 4.17*** 2.13, 8.17 2.16 0.69, 6.76 2.20* 1.12, 4.32 4þ 7.24*** 3.81, 13.79 3.88* 1.36, 11.05 3.95* 2.17, 7.18 Continued TABLE 3. Multivariable Model Results for the Regressions of Facility Delivery on Service Readiness of Facilities With Delivery Services, Controlling for Availability of Services and Other Variables, Haiti, 2012–2013 256 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org TABLE 3. Hospitals and health centers with a bed in Haiti are not well prepared to provide delivery services. DISCUSSION Continued Rural Other Urban Metropolitan Variables Odds Ratio 95% CI Odds Ratio 95% CI Odds Ratio 95% CI Departement Ouest 1.00 1.00 Sud-est 0.98 0.47, 2.03 1.00 0.11, 9.03 Nord 0.89 0.50, 1.57 2.00 0.57, 6.98 Nord-est 1.63 0.79, 3.39 2.37 0.55, 10.15 Artibonite 1.19 0.75, 1.88 1.60 0.46, 5.62 Centre 1.64 0.92, 2.92 1.92 0.46, 7.94 Sud 1.62 0.95, 2.76 2.46 0.48, 12.64 Grand'anse 0.97 0.40, 2.34 0.58 0.12, 2.79 Nord-ouest 1.02 0.54, 1.92 0.65 0.16, 2.68 Nippes 1.36 0.66, 2.81 2.67 0.31, 22.83 Number of clusters 201 69 59 Observations 2,571 802 650 a Regressions excluded women living in clusters with no health facility offering delivery services within the buffer since the readiness score is applied only to clusters with 1 or more health facilities. b Since the analysis excluded women living in clusters with no health facility offering delivery services within the buffer (i.e., the low availability group), the medium availability group was set as the reference. c In the metropolitan and other urban areas, no women were in the first wealth quintile and very few women were in the second wealth quintile; they were combined into the third quintile. P<.001; P<.01; P<.05. of facilities lack essential equipment and supplies for routine delivery care, and most have limited capacity to provide emergency obstetric care. Only about one-third of facilities have functional emergency transportation, and the availability of Cesarean delivery services is limited, especially in rural areas. Delivery in a health facility itself can- not reduce maternal mortality unless women are assisted by a skilled birth attendant capable of managing common life-threatening obstetric complications.38 In Haiti, however, less than half of the facilities have staff who received training in comprehensive emergency obstetric care (CEmOC) during the last 2 years. Guidelines for CEmOC were not available in service areas at most facilities. All of these limitations increase the risk of death for mother and newborn when an obstetrical emergency occurs. to play an important role predicting facility deliv- ery in urban areas, where delivery services are more available, compared with rural areas. Urban areas have more facilities, more accessible trans- portation, and more financial resources, and thus women may be able to choose to deliver at facili- ties offering a better quality of care. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 DISCUSSION In rural areas, however, especially in mountainous areas, there is less availability of facilities and thus having access to at least 1 health facility seems more important to the use of delivery care than the facility’s level of readiness to provide the services. We did not find an association between service readiness and use of facility delivery in the metro- politan area. This could be because a 5 km buffer may not be appropriate for defining the service environment for clusters in the metropolitan area. Because of a high density of health facilities and DHS clusters in the metropolitan area, using a 5 km buffer may result in adjacent clusters Despite the poor readiness of health facilities to provide delivery services, service quality appears 257 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org linking to more or less the same group of facilities; therefore, there is limited variation across metro- politan clusters in terms of the service environ- ment. This is indicated by less variation in readiness among clusters in the metropolitan area compared with clusters in rural and other urban areas. Moreover, people in metropolitan areas are likely to have more transportation options avail- able and better accessibility to health services that are geographically further from their home. expect that there was a substantial change in the availability of services and facilities’ readiness to provide the service over the period. While linking women to all of the facilities that they likely used reduces misclassification errors from the DHS GPS displacement procedure, the straight-line buffer approach does not take into account the mountainous terrain or the impassi- bility of roads during the rainy season, which may limit access to a linked facility. However, Nesbitt and colleagues compared 6 different meas- ures of spatial access and found that the straight- line linkage yields results similar to other geospa- tial algorithms in a developing-country setting.20 Finally, the buffer linkage between DHS clusters and SPA facilities may not be appropriate in areas where there is a high density of both health facili- ties and population, such as Haiti’s metropolitan area. More precise measurements of the service environment are needed for such areas, as well as a better understanding of other drivers of use in areas where service availability may be less of an issue. Acknowledgments: The authors acknowledge the financial support from the United States Agency for International Development (USAID) through the DHS Program. Strengths and Limitations Linking women’s use of health services to the service environment within a reasonable distance from DHS clusters is less prone to misclas- sification errors (resulting from the displacement of DHS clusters) than linking to the closet facility. g As discussed, because of methodological con- straints in linking population data and health serv- ice data, most previous studies have been limited to measuring service provision from the client’s perspective. Several recent studies have taken advantage of geographic data to associate health facilities and DHS clusters. These studies focused primarily on distance to the closest facility or the service available in the closest facility; however, this approach can be problematic because DHS cluster locations have been displaced. Our study improved on this methodology; instead of looking at a single facility (the closest facility)—where esti- mates may be subject to misclassification errors— we measured the effect of the service environment within a reasonable distance. In addition to the methodological improvement of measuring the service environment, some of the other strengths of our study lie in the use of facility census data and nationally representative household data, measuring service readiness with a wide range of items that the World Health Organization has identified as essential for providing high-quality delivery services. Together, these improvements in methodology have led to more generalizable results. Additionally, the use of observed availabil- ity of equipment and items instead of self-reported data during facility data collection increases the robustness of the readiness indicators and thus the accuracy of the relationship between the pro- vision of delivery services and their use. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 CONCLUSION This study indicates the importance of improving physical access to delivery services in rural Haiti. Overall, health facilities in Haiti are poorly equipped and do not appear ready to provide high-quality delivery services. Improving the quality of care at health facilities could contribute to increased use of facility delivery particularly in nonmetropolitan urban areas, where 40% of women still deliver at home. Over the years, the global community has recognized the importance of providing women with quality maternal health services to reduce maternal mortality.38 After all, reducing maternal mortality by having women deliver in health facilities will only work if these facilities are ready to provide comprehensive obstetric care. This study also has some limitations. One limi- tation is the temporal gap between the outcome variables and the service variables. Facility data reflect the "current" service environment at the time of the Haiti 2013 SPA, while use of facility delivery was measured over a 5-year time period preceding the 2012 Haiti DHS. Associating service provision and facility delivery use data could be problematic if the service environment changed substantially over this time period. Given the na- ture of delivery services, however, we do not 1. World Health Organization (WHO). Trends in Maternal Mortality: 1990 to 2015. Estimates by WHO, UNICEF, UNFPA, the World Bank and the United Nations Population Division. Geneva: WHO; 2015. http://www.who.int/reproductivehealth/publications/ monitoring/maternal-mortality-2015/en/. Accessed May 4, 2017. Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti Access to institutional delivery care and reasons for home delivery in three districts of Tanzania. Int J Equity Health. 2014;13(1):48. CrossRef. Medline 8. Gabrysch S, Campbell OMR. Still too far to walk: literature review of the determinants of delivery service use. BMC Pregnancy Childbirth. 2009;9(1):34. CrossRef. Medline 26. Kinney MV, Kerber KJ, Black RE, et al; Science in Action: Saving the lives of Africa’s Mothers, Newborns, and Children working group. Sub-Saharan Africa’s mothers, newborns, and children: where and why do they die? PLoS Med. 2010;7(6):e1000294. CrossRef. Medline 9. Kitui J, Lewis S, Davey G. Factors influencing place of delivery for women in Kenya: an analysis of the Kenya Demographic and Health Survey, 2008/2009. BMC Pregnancy Childbirth. 2013;13(1):40. CrossRef. Medline 27. Yaffee AQ, Whiteside LK, Oteng RA, et al. Bypassing proximal health care facilities for acute care: a survey of patients in a Ghanaian Accident and Emergency Centre. Trop Med Int Health. 2012;17(6):775–781. CrossRef. Medline 10. Montagu D, Yamey G, Visconti A, Harding A, Yoong J. Where do poor women in developing countries give birth? A multi-country analysis of demographic and health survey data. PLoS One. 2011;6(2):e17155. CrossRef. Medline 28. Karkee R, Lee AH, Pokharel PK. Women’s perception of quality of maternity services: a longitudinal survey in Nepal. BMC Pregnancy Childbirth. 2014;14(1):45. CrossRef. Medline 11. Moyer CA, Dako-Gyeke P, Adanu RM. Facility-based delivery and maternal and early neonatal mortality in sub-Saharan Africa: a regional review of the literature. Afr J Reprod Health. 2013;17(3):30–43. Medline 29. Acharya LB, Cleland J. Maternal and child health services in rural Nepal: does access or quality matter more? Health Policy Plan. 2000;15(2):223–229. CrossRef. Medline 12. Diamond-Smith N, Sudhinaraset M. Drivers of facility deliveries in Africa and Asia: regional analyses using the Demographic and Health Surveys. Reprod Health. 2015;12:6. CrossRef. Medline 30. Burgert CR, Colston J, Roy T, Zachary B. Geographic displacement procedure and georeferenced data release policy for the Demographic and Health Surveys. DHS Statistical Analysis Reports No. 7. Calverton, Maryland: ICF International; 2013. https:// dhsprogram.com/pubs/pdf/SAR7/SAR7.pdf. Accessed May 4, 2017. 13. Moyer CA, Mustafa A. Drivers and deterrents of facility delivery in sub-Saharan Africa: a systematic review. Reprod Health. 2013;10(1):40. CrossRef. Medline 31. Ministère de la Santé Publique et de la Population. Plan Directeur de la Santé 2012-2022. Port-au-Prince, Haïti: MSPP; 2013. 14. Tey NP, Lai SL. Correlates of and barriers to the utilization of health services for delivery in South Asia and Sub-Saharan Africa. ScientificWorldJournal. Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org 19. Lohela TJ, Campbell OMR, Gabrysch S. Distance to care, facility delivery and early neonatal mortality in Malawi and Zambia. PLoS One. 2012;7(12):e52110. CrossRef. Medline 2. Anderson FWJ, Morton SU, Naik S, Gebrian B. Maternal mortality and the consequences on infant and child survival in rural Haiti. Matern Child Health J. 2007;11(4):395–401. CrossRef. Medline 20. Nesbitt RC, Gabrysch S, Laub A, et al. Methods to measure potential spatial access to delivery care in low- and middle-income countries: a case study in rural Ghana. Int J Health Geogr. 2014;13(1):25. CrossRef. Medline 3. Kirigia JM, Oluwole D, Mwabu GM, Gatwiri D, Kainyu LH. Effects of maternal mortality on gross domestic product (GDP) in the WHO African region. Afr J Health Sci. 2006;13(1-2):86–95. Medline 4. Cayemittes M, Busangu MF, Bizimana JdD, et al. Enquête Mortalité, Morbidité et Utilisation des Services, Haïti, 2012. Calverton, Maryland: MSPP, IHE, and ICF International; 2013. https:// dhsprogram.com/pubs/pdf/FR273/FR273.pdf. Accessed May 4, 2017. 21. Skiles MP, Burgert CR, Curtis SL, Spencer J. Geographically linking population and facility surveys: methodological considerations. Popul Health Metr. 2013;11(1):14. CrossRef. Medline 22. Alexandre PK, Saint-Jean G, Crandall L, Fevrin E. Prenatal care uti- lization in rural areas and urban areas of Haiti. Rev Panam Salud Publica. 2005;18(2):84–92. Medline 5. De Brouwere V, De Brouwere V, Tonglet R, Van Lerberghe W. Strategies for reducing maternal mortality in developing countries: what can we learn from the history of the industrialized West? Trop MedIntHealth.1998;3(10):771–782.CrossRef.Medline 23. Gage AJ, Guirlène Calixte M. Effects of the physical accessibility of maternal health services on their use in rural Haiti. Popul Stud (Camb). 2006;60(3):271–288. CrossRef. Medline 6. Graham WJ, Bell JS, Bullough CHW. Can skilled attendance at delivery reduce maternal mortality in developing countries? In: Brouwere VD, Lerberghe V, Antwerp BIP, editors. Studies in Health Service Organization and Policy. Antwerp, Belgium: TIG Press; 2001. 24. Ruktanonchai CW, Ruktanonchai NW, Nove A, et al. Equality in maternal and newborn health: modelling geographic disparities in utilisation of care in five East African countries. PloS One. 2016;11(8):e0162006. CrossRef. Medline 25. Choulagai B, Onta S, Subedi N, et al. Barriers to using skilled birth attendants services in mid- and far-western Nepal: a cross-sectional study. BMC Int Health Hum Rights. 2013;13(1):49. CrossRef. Medline 7. Exavery A, Kanté A, Njozi M, et al. Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 REFERENCES 1. World Health Organization (WHO). Trends in Maternal Mortality: 1990 to 2015. Estimates by WHO, UNICEF, UNFPA, the World Bank and the United Nations Population Division. Geneva: WHO; 2015. http://www.who.int/reproductivehealth/publications/ monitoring/maternal-mortality-2015/en/. Accessed May 4, 2017. 1. World Health Organization (WHO). Trends in Maternal Mortality: 1990 to 2015. Estimates by WHO, UNICEF, UNFPA, the World Bank and the United Nations Population Division. Geneva: WHO; 2015. http://www.who.int/reproductivehealth/publications/ monitoring/maternal-mortality-2015/en/. Accessed May 4, 2017. 258 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti 2013;2013:423403. CrossRef. Medline 32. World Health Organization (WHO). Service Availability and Readiness Assessment (SARA): An Annual Monitoring System for Service Delivery. Geneva: WHO; 2013. http://www.who.int/ healthinfo/systems/SARA_Reference_Manual_Full.pdf. Accessed May 4, 2017. 15. Thind A, Mohani A, Banerjee K, Hagigi F. Where to deliver? Analysis of choice of delivery location from a national survey in India. BMC Public Health. 2008;8(1):29. CrossRef. Medline 33. Raudenbush SW, Bryk AS. Hierarchical Linear Models: Applications and Data Analysis Methods. London: Sage Publications; 2002. 16. Babalola SO. Factors associated with use of maternal health services in Haiti: a multilevel analysis. Rev Panam Salud Publica. 2014;36(1):1–9. Medline 34. Moyer CA, Mustafa A. Drivers and deterrents of facility delivery in sub-Saharan Africa: a systematic review. Reprod Health. 2013;10(1):40. CrossRef. Medline 17. Séraphin MN, Ngnie-Teta I, Ayoya MA, et al. Determinants of institutional delivery among women of childbearing age in rural Haiti. Matern Child Health J. 2015;19(6):1400–1407. CrossRef. Medline 35. Gabrysch S, Cousens S, Cox J, Campbell OMR. The influence of dis- tance and level of care on delivery place in rural Zambia: a study of linked national data in a geographic information system. PLoS Med. 2011;8(1):e1000394. CrossRef. Medline 18. Kyei NNA, Campbell OMR, Gabrysch S. The influence of distance and level of service provision on antenatal care use in rural Zambia. PLoS One. 2012;7(10):e46475. CrossRef. Medline 36. Stekelenburg J, Kyanamina S, Mukelabai M, Wolffers I, Roosmalen J. Waiting too long: low use of maternal health services in Kalabo, 259 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Peer Reviewed Received: 2016 Oct 10; Accepted: 2017 Mar 13; First Published Online: 2017 May 25 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org Zambia. Trop Med Int Health. 2004;9(3):390–398. CrossRef. Medline Zambia. Trop Med Int Health. 2004;9(3):390–398. CrossRef. Medline results and a potential way forward. Bull World Health Organ. 2007;85(10):783–790. CrossRef. Medline results and a potential way forward. Bull World Health Organ. 2007;85(10):783–790. CrossRef. Medline 37. Harvey S, Blandón YC, McCaw-Binns A, et al; Nicaraguan Maternal and Neonatal Health Quality Improvement Group. Are skilled birth attendants really skilled? A measurement method, some disturbing 38. Koblinsky M, Moyer CA, Calvert C, et al. Quality maternity care for every woman, everywhere: a call to action. Lancet. 2016;388 (10057):2307–2320. CrossRef Medline Peer Reviewed results and a potential way forward. Bull World Health Organ. 2007;85(10):783–790. CrossRef. Medline 38. Koblinsky M, Moyer CA, Calvert C, et al. Quality maternity care for every woman, everywhere: a call to action. Lancet. 2016;388 (10057):2307–2320. CrossRef Medline Global Health: Science and Practice 2017 \| Volume 5 \| Number 2 Received: 2016 Oct 10; Accepted: 2017 Mar 13; First Published Online: 2017 May 25 Cite this article as: Wang W, Winner M, Burgert-Brucker CR. Limited service availability, readiness, and use of facility-based delivery care in Haiti: a study linking health facility data and population data. Glob Health Sci Pract. 2017;5(2):244-260. https://doi.org/10.9745/GHSP-D-16-00311 © Wang 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 properly cited. To view a copy of the license, visit http:// creativecommons.org/licenses/by/3.0/. When linking to this article, please use the following permanent link: https://doi.org/10.9745/GHSP-D-16- 00311 260 Limited Service Availability, Readiness, and Use of Facility Delivery in Haiti www.ghspjournal.org 261
https://openalex.org/W2927445962	https://research.wu.ac.at/ws/files/18576227/Jomrich2020_Article_SystemicImmune-InflammationInd.pdf	English	null	Systemic Immune-Inflammation Index (SII) Predicts Poor Survival in Pancreatic Cancer Patients Undergoing Resection	Journal of gastrointestinal surgery	2,020	cc-by	6,296	Citation for published version (APA): Jomrich, G., Gruber, E. S., Winkler, D., Hollenstein, M., Gnant, M., Sahora, K., & Schindl, M. (2019). Systemic Immune-Inflammation Index (SII) Predicts Poor Survival in Pancreatic Cancer Patients Undergoing Resection. Journal of Gastrointestinal Surgery. https://doi.org/10.1007/s11605-019-04187-z Link to publication Systemic Immune-Inflammation Index (SII) Predicts Poor Survival in Pancreatic Cancer Patients Undergoing Resection Jomrich, Gerd; Gruber, Elisabeth S.; Winkler, Daniel; Hollenstein, Marlene; Gnant, Michael; Sahora, Klaus; Schindl, Martin Published in: Journal of Gastrointestinal Surgery Published in: Journal of Gastrointestinal Surgery DOI: 10.1007/s11605-019-04187-z Document Version Publisher's PDF, also known as Version of record Link to publication Link to publication Download date: 24. Oct 2024 Journal of Gastrointestinal Surgery (2020) 24:610–618 https://doi.org/10.1007/s11605-019-04187-z ORIGINAL ARTICLE Abstract Background The systemic immune-inflammation index based on peripheral neutrophil, lymphocyte, and platelet counts has shown a prognostic impact in several malignancies. The aim of this study was to determine the prognostic role of systemic immune-inflammation index in patients with pancreatic ductal adenocarcinoma undergoing resection. Methods Consecutive patients who underwent surgical resection at the department of surgery at the Medical University of Vienna between 1995 and 2014 were included into this study. The systemic immune-inflammation index was calculated by the formula plateletneutrophil/lymphocyte. Optimal cutoffs were determined using Youden’s index. Uni- and multivariate analyses were calculated by the Cox proportional hazard regression model for overall survival. Results Three hundred twenty-one patients were included in this study. Clinical data was achieved from a prospective patient database. In univariate survival analysis, elevated systemic immune-inflammation index was found to be significantly associated with shortened patients’ overall survival (p = 0.007). In multivariate survival analysis, systemic immune-inflammation index remained an independent prognostic factor for overall survival (p = 0.004). No statistical significance could be found for platelet to lymphocyte ratio and neutrophil to lymphocyte ratio in multivariate analysis. Furthermore, area under the curve analysis showed a higher prognostic significance for systemic immune-inflammation index, compared to platelet to lymphocyte ratio and neutrophil to lymphocyte ratio. Conclusion A high systemic immune-inflammation index is an independent, preoperative available prognostic factor in patients with resectable pancreatic ductal adenocarcinoma and is superior to platelet to lymphocyte ratio and neutrophil to lymphocyte ratio for predicting overall survival in pancreatic ductal adenocarcinoma patients. Keywords Pancreatic ductal adenocarcinoma . Systemic immune-inflammation index . Inflammation Martin Schindl martin.schindl@meduniwien.ac.at Systemic Immune-Inflammation Index (SII) Predicts Poor Survival in Pancreatic Cancer Patients Undergoing Resection Gerd Jomrich1 & Elisabeth S. Gruber1 & Daniel Winkler2 & Marlene Hollenstein1 & Michael Gnant1 & Klaus Sahora1 & Martin Schindl1 Received: 21 November 2018 /Accepted: 26 February 2019 /Published online: 28 March 2019 # 2019 The Author(s) 2 Vienna University of Economics and Business, Welthandelsplatz 1, AD 1020 Vienna, Austria 1 Department of Surgery, Comprehensive Cancer Center (CCC), Medical University Vienna and Pancreatic Cancer Unit, Spitalgasse 23, 1090 Vienna, Austria Introduction survival rate for patients diagnosed with PADC remains poor with less than 5%.2 Pancreatic ductal adenocarcinoma (PDAC) is the ninth most common malignancy and ranks fifth place of cancer-related death in western countries with inclining incidence.1 Despite the development of multimodal approaches, combining surgi- cal resection with perioperative chemo-(radio)therapy, 5-year Inflammation, as one of the hallmarks of cancer, is an ac- knowledged factor in tumor biology.3,4 Inflammation-driven tumorigenesis and tumor progression plays a crucial role in malignant diseases.3,5 Systemic inflammatory response (SIR) in the context of tumor-associated inflammation has been demonstrated to diminish outcome and be of major prognostic importance in various cancers.6,7 A number of promising and potentially prognostic immunologic and histologic bio- markers have been investigated in PDAC.8,9 However, evalu- ation of these biomarkers is often expensive and time-consum- ing. Thus, investigation of tumor-driving inflammation-based components is of major importance and targeting pathways of inflammatory response might become a cornerstone of cancer treatment.10 In PDAC, outcome prediction mainly depends on * Martin Schindl martin.schindl@meduniwien.ac.at 611 J Gastrointest Surg (2020) 24:610–618 clinical and pathological factors, such as tumor size, lymph node involvement, and distant metastases as well as resection margin.11 Notably, these factors are obtained post-operatively through evaluation of the surgical specimen and current avail- able prognostic markers do not allow to preoperatively predict outcome. Therefore, identification of easy-available markers might help to determine individual treatment approaches. The utility of inflammation-based scores, such as neutrophil- lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR) and systemic immune-inflammation index (SII), are based on routinely obtained markers that are available before sur- gery. Elevated SII has been reported to be associated with clinico-pathological parameters and has been proven to be an independent prognostic factor in a number of malignancies, including PDAC.12–16 However, no data exists until now, de- scribing the prognostic value of the SII in PDAC after neoad- juvant treatment. The aim of the present study was to assess the prognostic value of SII in patients undergoing a potentially curative resection for PDAC with or without neoadjuvant ther- apy. Furthermore, the SII is compared with NLR and PLR in predicting survival in this cohort of patients. whereas gemcitabine-based regimens (gemcitabine monotherapy and combinations with oxaliplatin, erlotinib, and nab-Paclitaxel) or FOLFIRINOX were used from 1999 onwards. The present standard for neoadjuvant treatment since several years consists of either FOLFIRINOX or Gem/nab-Pac depending on patients’ condition. Statistical Analysis Statistical analysis was performed using the R statistical software (Version 3.4.3) with the BSurvival,^ BpROC,^ and BOptimal Cutpoints^ packages.18–21 To evaluate the discriminatory ability of the SII, NLR, and PLR, ROC curves were generated and the area under the ROC curves (AUROCs) was measured and compared. For bivariate analysis, to investigate relationships between SII, NLR, and PLR and clinico-pathological parameters, t test and the Wilcoxon test were used as appropriate. Univariate and multivariate analyses were conducted using the Cox proportional hazard model. In the multivariable model, SII, NLR, and PLR could not be included into together, due to multicollinearity. Therefore, the stepwise regres- sion analysis for multivariate Cox models, SII, NLR, and PLR could not be included into the model together due to multicollinearity. Discrimination ability was com- pared using the receiver operating curve. Optimal cutoff values for SII, NLR, and PLR were determined using Youden’s index, which maximizes the sum of sensitivity and specificity. Graphically, it is represented by the dis- tance between the 45-degree line and the ROC22,23. The graphical analysis was performed using the Kaplan-Meier survival curve estimator and analyzed by the log-rank test. Overall survival (OS) was defined as time between prima- ry surgery and the patients’ death. Death from other cause than PDAC or survival until the end of the observation period was considered as censored observations. Disease- free survival (DFS) was defined from the day of surgery until first evidence of disease progression. Categorical da- ta was analyzed using the chi-squared test. Continuous data was either analyzed using the t test form normally distributed values or the Wilcoxon rank-sum test. Introduction For adju- vant chemotherapy, 5-fluorouracil-based regimens were used between 1995 and 1998 and gemcitabine-based regi- mens were administered thereafter. All patients were regu- larly followed thereafter with physical examination, tumor marker, and computed tomographic scan every 3 months for the first 2 years and every 6 months until 5 years after surgery. Blood samples were obtained within 7 days prior to sur- gery. NLR, PLR, and SII were calculated as previously de- scribed: NLR = neutrophils / lymphocytes, PLR = platelet / lymphocytes, and SII = platelets × neutrophils / lymphocytes.16 Material and Methods Patients undergoing resection for PDAC between 1995 and 2014 at the Department of Surgery, Medical University of Vienna, were identified from a prospectively maintained PDAC database. Both patients who primarily underwent surgery and patients with borderline resectable disease ac- cording to NCCN guidelines who were treated by chemo- therapy or radio-chemotherapy before resection were included.17 The study was approved by the Ethics Committee of the Medical University Vienna, Austria, in accordance with the Helsinki declaration (EK 1166/2013). Clinico-pathological data were assembled from medical re- cords, including, gender, age, preoperative neutrophil, lymphocyte and platelet counts, tumor site, histopatholog- ical tumor grading, staging (TNM) according to the 8th edition of the Union for International Cancer Control (UICC)/American Joint Committee on Cancer (AJCC), neoadjuvant treatment and surgical resection technique. Tumor resection margin status (R) was classified as R0 or R1 (1-mm tumor-free margin). Patients with distant metas- tases at time of diagnosis and death from other cause within 30 days post-surgery as well as patients who had recently pyrexia (axillary ≥37.2 °C/99.0 °F), any form of active infection or chronic inflammatory disease were excluded from the study. Each patient was discussed in the multidis- ciplinary team meeting before surgery. Neo-/adjuvant che- motherapy was administered according to the standard reg- imens available at the respective period. ) g SD, standard deviation; UICC, Union for International Cancer Control; SII, systemic immune-inflammation index ( ) g SD, standard deviation; UICC, Union for International Cancer Control; SII, systemic immune-inflammation index Results for patients with high SII was 14.2 (range, 1.5–128.2) months and 20.5 (range, 1.6–200.8) months for patients with low SII respectively. In the entire cohort, using overall survival as an end-point, the area under the re- ceiver operator curve was 0.46 (CI, 0.37–0.55) for SII, 0.46 (CI, 0.36–0.56) for NLR, and 0.51 (CI, 0.42–0.61) for PLR with no significant difference in discrimination ability between SII, NLR, and PLR regarding OS was found (Fig. 1). for patients with high SII was 14.2 (range, 1.5–128.2) months and 20.5 (range, 1.6–200.8) months for patients with low SII respectively. In the entire cohort, using overall survival as an end-point, the area under the re- ceiver operator curve was 0.46 (CI, 0.37–0.55) for SII, 0.46 (CI, 0.36–0.56) for NLR, and 0.51 (CI, 0.42–0.61) for PLR with no significant difference in discrimination ability between SII, NLR, and PLR regarding OS was found (Fig. 1). In the present study, a total of 321 patients (169, 52.2% male) with a median age of 68.5 (range, 35.9–92.3) years were in- cluded. The majority of patients (201, 62.6%) presented with a moderately differentiated tumor (G2) and 43 (13.3%) patients received neoadjuvant treatment prior to resection. In 255 (78.7%) cases, the tumor was located in the head of the pan- creas and in 238 (73.5%) positive lymph node was found at pathological assessment. In accordance with the 8th edition UICC/AJCC classification, 265 (82.6%) patients showed stage II disease and 153 (47.2%) patients underwent biliary drainage preoperatively. Clinico-pathological data are present- ed in Table 1. Kaplan-Meier curve survival analysis for all patients re- vealed that low SII (p = 0.004) and PLR (p = 0.04) were sig- nificantly associated with longer OS, whereas no significance was found for NLR (Fig. 2). Univariate Cox proportional hazard regression revealed that SII, PLR, age, jaundice, resection margin, CA 19-9, lymph node-ratio, and tumor size were significantly associat- ed with OS. Similarly, age, jaundice, lymph node-ratio, and tumor size are significantly associated with DFS whereas SII, NLR, and PLR was not (Table 2). The median OS was 18.5 months (range, 1.5– 198 months) and the rate of 3- and 5-year OS was 25.63% and 8.44%, respectively. The optimal cutoff values for SII, PLR, and NLR were 873, 179, and 225, respectively. With the defined cutoffs, 119 patients had SII ≥873, 125 patients had PLR ≥179, and 225 patients NLR ≥2.15 before surgery. Material and Methods 1 Receiver operating characteristic (ROC) curves were generated to evaluate the discriminatory ability of the SII (a), NLR (b), and PLR (c). The area under the ROC curves (AUROCs) were measured and compared a Specificity Sensitivity 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 SII a Specificity Sensitivity 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.0 0.0 0.2 0.4 0.6 0.8 1.0 SII b Sp Sensitivity Fig. 1 Receiver operating characteristic (ROC) curves were generated to evaluate the discriminatory ability of the SII (a), NLR (b), and PLR (c). NLR 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 b Specificity Sensitivity PLR 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 c Specificity Sensitivity The area under the ROC curves (AUROCs) were measured and compared Fig. 1 Receiver operating characteristic (ROC) curves were generated to evaluate the discriminatory ability of the SII (a), NLR (b), and PLR (c). Material and Methods For neoadjuvant treatment, 5-fluorouracil was used from 1995 to 1998, J Gastrointest Surg (2020) 24:610–618 612 Table 1 Association of the SII with clinico-pathological parameters in pancreatic ductal adenocarcinoma Table 1 Association of the SII with clinico-pathological parameters in pancreatic ductal adenocarcinoma Factors All SII High (≥873) Low (< 873) p value n 324 120 204 Age, mean (SD) 68.25 (9.74) 68.81 (9.21) 67.91 (10.04) > 0.1(*) Sex > 0.1 Male 169 52.16% 65 104 Female 155 47.84% 55 100 (y)pT > 0.1 0 2 0.62% 1 1 1 24 7.41% 5 19 2 47 14.51% 18 29 3 234 72.22% 87 147 4 17 5.25% 9 8 (y)pN > 0.1 0 86 26.54% 27 59 1 238 73.46% 93 145 (y)G > 0.1 1 15 4.63% 4 11 2 204 62.96% 72 132 3 105 32.41% 44 61 R 0.027 0 254 78.40% 86 168 1 70 21.60% 34 36 UICC staging > 0.1 I 41 12.65% 15 26 II 265 81.79% 96 169 III 18 5.56% 10 8 Neoadjuvant treatment > 0.1 yes 43 13.27% 14 29 no 281 86.73% 106 175 Jaundice 0.041 No 163 50.31% 51 112 Yes 161 49.69% 69 92 CA 19-9 0.013 Unknown 24 7.41% 9 15 ≤114 kU/L 143 44.14% 42 101 > 114 kU/L 157 48.46% 69 88 Lymph node-ratio > 0.1 Unknown 23 7.10% 7 16 ≤0.2 196 60.49% 71 125 > 0.2 105 32.41% 42 63 Nicotine > 0.1 Unknown 3 0.93% 1 2 Yes 135 41.67% 55 80 No 186 57.41% 64 122 Pain > 0.1 Yes 130 40.12% 45 85 No 194 59.88% 75 119 Pancreatitis > 0.1 Yes 63 19.44% 20 43 No 261 80.56% 100 161 Diabetes > 0.1 Unknown 12 3.70% 3 9 Yes 73 22.53% 29 44 No 239 73.77% 88 151 Stent > 0.1 Yes 153 47.22% 60 93 No 171 52.78% 60 111 Surgical procedure 0.0345 PPPD 181 55.86% 65 116 Whipple 71 21.91% 35 36 Distal resection 60 18.52% 15 45 Total pancreatectomy 12 3.70% 5 7 Localization > 0.1 1 255 78.70% 100 155 2 24 7.41% 8 16 3 45 13.89% 12 33 J Gastrointest Surg (2020) 24:610–618 613 a Specificity Sensitivity 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 NLR 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 PLR 1.0 0.5 0.0 0.0 0.2 0.4 0.6 0.8 1.0 SII b c Specificity Specificity Sensitivity Sensitivity Fig. Results Using bivariate analysis, sig- nificant relationship between elevated SII and clinico- pathological parameters was found for resection margin (p = 0.03), jaundice (p = 0.04), CA 19–9 (p = 0.01), and surgical procedure (p = 0.04) (Table 1). The median OS Furthermore, no statistical significance was found for OS and DFS in univariate Cox proportional hazard regression for neoad- juvant treatment (p = 0.08; RR, 1.35; CI, 95 0.96–1.88; and p = 0.23; RR, 1.24; CI, 95 0.87–1.77; respectively; Table 2). Multivariate Cox-regression analysis using SII as bi- variate variable revealed that a high SII (p = 0.016; RR, 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.0037, log-rank test 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.068, log-rank test 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.039, log-rank test a b c Time since surgery (years) Cumulative overall suvival SII<873 NLR<2.15 PLR<179 Time since surgery (years) Cumulative overall suvival PLR≥179 NLR≥2.15 Time since surgery (years) Cumulative overall suvival SII≥873 Fig. 2 Kaplan-Meier curves for overall survival (OS) stratified by SII (a), NLR (b) and PLR (c) 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.039, log-rank test c PLR<179 PLR≥179 Time since surgery (years) Cumulative overall suvival 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.068, log-rank test b NLR<2.15 Time since surgery (years) Cumulative overall suvival NLR≥2.15 0 5 10 15 0.0 0.2 0.4 0.6 0.8 1.0 p=0.0037, log-rank test a Time since surgery (years) Cumulative overall suvival SII<873 SII≥873 Fig. Results 2 Kaplan-Meier curves for overall survival (OS) stratified by SII (a), NLR (b) and PLR (c) J Gastrointest Surg (2020) 24:610–618 614 Table 2 Univariate Cox regression analysis estimating the influence of the SII, NLR, PLR and clinico-pathological parameters on overall survival and disease-free survival in patients with ductal adenocarcinoma of the pancreas Table 2 Univariate Cox regression analysis estimating the influence of the SII, NLR, PLR and clinico-pathological p disease-free survival in patients with ductal adenocarcinoma of the pancreas gression analysis estimating the influence of the SII, NLR, PLR and clinico-pathological parameters on overall survival and nts with ductal adenocarcinoma of the pancreas 0 71; CI 95% 0 54 0 94) positive resection margin (p = Discussion Table 2 Univariate Cox regression analysis estimating the influence of the SII, NLR, PLR and clinico-pathological parameters on overall survival and disease-free survival in patients with ductal adenocarcinoma of the pancreas Overall survival Disease-free survival RR CI (95%) p value RR CI (95%) p value SII ≥873 vs. < 873 0.6979 0.5469–0.8907 0.0038 0.8147 0.6288–1.0556 0.1211 NLR ≥2.15 vs. < 2.15 0.7852 0.6054–1.0184 0.0683 0.8536 0.6455–1.1286 0.2666 PLR ≥179 vs. < 179 0.772 0.6031–0.9882 0.0399 0.808 0.6205–1.0522 0.1136 UICC stage II vs. I 1.0668 0.7438–1.5301 0.7252 1.1808 0.7898–1.7653 0.4179 III vs. I 1.4071 0.7658–2.5852 0.2712 1.2352 0.6366–2.3964 0.5323 Grading 2 vs. 1 0.7087 0.4174–1.2032 0.2023 0.6882 0.3973–1.1922 0.1826 3 vs. 1 1.0347 0.5996–1.7855 0.9026 1.0086 0.5719–1.7785 0.9765 Sex Male vs. female 1.0843 0.8562–1.373 0.5019 1.0475 0.8145–1.3472 0.7175 Age 0.9881 0.9769–0.9993 0.038 0.9838 0.9714–0.9964 0.0116 Jaundice Yes vs. no 1.4243 1.1246–1.8039 0.0033 1.4186 1.1013–1.8272 0.0068 R 1 vs. 0 1.4382 1.0866–1.9037 0.0111 1.3526 0.9989–1.8315 0.0508 CA 19-9 ≥114 vs. < 114 kU/L 1.3148 1.0281–1.6815 0.0292 1.2842 0.9899–1.6661 0.0596 Lymph node-ratio ≥0.2 vs. < 0.2 1.8094 1.4006–2.3375 < 0.0001 1.6615 1.264–2.184 0.0003 T-staging 1 vs. 0 0.2554 0.0592–1.1006 0.067 0.1249 0.0162–0.9622 0.0458 2 vs. 0 0.231 0.0553–0.9647 0.0445 0.1209 0.0162–0.9034 0.0395 3 vs. 0 0.2674 0.0657–1.0876 0.0654 0.1266 0.0173–0.9263 0.0418 4 vs. 0 0.3471 0.0786–1.5319 0.1624 0.1349 0.0172–1.0561 0.0564 N-staging 1 vs. 0 1.158 0.8836–1.5176 0.2876 1.0838 0.8162–1.4391 0.5781 Pain Yes vs. no 0.9929 0.7804–1.2632 0.9535 1.0445 0.8088–1.3488 0.7388 Pancreatitis Yes vs. no 0.9514 0.7064–1.2815 0.7431 0.8858 0.6426–1.2212 0.4592 Nicotine Yes vs. no 1.106 0.8705–1.406 0.409 1.0901 0.8457–1.4052 0.5053 Diabetes Yes vs. no 1.136 0.8545–1.51 0.38 1.1107 0.8214–1.502 0.4953 Neoadjuvant treatment Yes vs. no 1.3449 0.9616–1.881 0.0834 1.2416 0.8711–1.7696 0.2314 Stent Yes vs. Results no 1.1175 0.883–1.4142 0.3553 1.1664 0.9076–1.4989 0.2291 Surgical procedure PPPD Whipple 1.0712 0.7961–1.4414 0.6497 1.0691 0.7808–1.4638 0.6769 Distal resection 0.8498 0.6152–1.1739 0.3236 1.2939 0.6334–2.6432 0.4796 Total pancreatectomy 1.4092 0.7425–2.6747 0.294 0.9609 0.6824–1.353 0.8192 Localization 2 vs. 1 0.8828 0.5386–1.4469 0.621 1.0087 0.6055–1.6802 0.9736 3 vs. 1 0.9289 0.6605–1.3062 0.6714 1.0415 0.7199–1.5069 0.829 UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; Ref., reference; SII, systemic immune-inflammation index; PLR, platelet lymphocyte ratio; NLR, neutrophil lymphocyte ratio; PPPD, pylorus-preserving pancreaticoduodenectomy UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; Ref., reference; SII, systemic immune-inflammation index; PLR, platelet lymphocyte ratio; NLR, neutrophil lymphocyte ratio; PPPD, pylorus-preserving pancreaticoduodenectomy UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; Ref., reference; SII, system PLR, platelet lymphocyte ratio; NLR, neutrophil lymphocyte ratio; PPPD, pylorus-preserving pancreaticoduodenecto Discussion 0.71; CI, 95% 0.54–0.94), positive resection margin (p = 0.03; RR, 1.46; CI, 95% 1.05–2.03), and a high lymph node-ratio (p < 0.001; RR, 1.77; CI, 95% 1.32–2.36), but not NLR and PLR, are independent risk factors for OS (Table 3). No statistical significant association could be found for SII, NLR, and PLR in the multivariate Cox models for DSF (Table 4). 0.71; CI, 95% 0.54–0.94), positive resection margin (p = 0.03; RR, 1.46; CI, 95% 1.05–2.03), and a high lymph node-ratio (p < 0.001; RR, 1.77; CI, 95% 1.32–2.36), but not NLR and PLR, are independent risk factors for OS (Table 3). No statistical significant association could be found for SII, NLR, and PLR in the multivariate Cox models for DSF (Table 4). Inflammation plays a key role in tumor initiation, malignant conversion, and metastasis and influences the host anti-tumor immunity.3–5 The present study investigated the clinical and prognostic value of preoperative SII, NLR, and PLR in pa- tients with PDAC undergoing resection and competed their 615 J Gastrointest Surg (2020) 24:610–618 Table 3 Multivariate Cox regression analysis estimating the influence of the SII and clinico-pathological parameters on overall survival in pa- tients with ductal adenocarcinoma of the pancreas Overall survival SII RR CI (95%) p value SII ≥873 vs. < 873 0.7138 0.5427–0.9388 0.0159 UICC stage II vs. I 0.7796 0.493–1.2329 0.2871 III vs. I 0.9423 0.4591–1.9344 0.8714 Grading 2 vs. 1 0.5881 0.3181–1.0875 0.0906 3 vs. 1 0.8104 0.4285–1.5328 0.5179 Sex Male vs. female 0.9328 0.7155–1.2162 0.6074 Age 0.9874 0.9749–1.0001 0.0514 Jaundice Yes vs. no 1.1994 0.9094–1.5819 0.1979 R 1 vs. 0 1.4561 1.0457–2.0275 0.0261 CA 19-9 ≥114 vs. < 114 kU/L 1.1642 0.8795–1.541 0.288 Lymph node-ratio ≥0.2 vs. < 0.2 1.7687 1.3247–2.3615 < 0.001 UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; SII, systemic immune-inflammation index Table 4 Multivariate Cox regression analysis estimating the influence of the SII and clinico-pathological parameters on disease-free survival in patients with ductal adenocarcinoma of the pancreas of the SII and clinico pathological parameters on disease free survival in patients with ductal adenocarcinoma of the pancreas Disease-free survival SII RR CI (95%) p value SII ≥873 vs. < 873 0.787 0.5885–1.0524 0.1062 UICC stage II vs. I 1.0855 0.6388–1.8446 0.7617 III vs. I 1.0861 0.495–2.3831 0.8367 Grading 2 vs. 1 0.5279 0.2826–0.9861 0.0451 3 vs. 1 0.7407 0.3878–1.4148 0.3633 Sex Male vs. Discussion female 0.9176 0.6892–1.2217 0.5562 Age 0.9823 0.9685–0.9964 0.0138 Jaundice Yes vs. no 1.2138 0.9041–1.6295 0.1973 R 1 vs. 0 1.405 0.984–2.006 0.0613 CA 19-9 ≥114 vs. < 114 kU/L 1.1385 0.8459–1.5323 0.3921 Lymph node-ratio ≥0.2 vs. < 0.2 1.4942 1.0995–2.0306 0.0103 UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; SII, systemic immune-inflammation index UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; SII, systemic immune-inflammation index UICC, Union for International Cancer Control; CI, confidence interval; RR, relative risk; SII, systemic immune-inflammation index predictive accuracy. Overall, SII, but not NLR and PLR was an independent prognostic factor for OS in patients with PDAC undergoing resection. infiltrating immune cells that promote tumorigenesis and dissemination.5,32 Neutrophils activate endothelium and pa- renchymal cells via secretion of soluble factors that enhance tumor cell adhesion at distant sites.33–35 Increasing numbers of blood neutrophils and platelets have been associated with tu- mor progression and diminished clinical outcome in a number of solid tumors.36,37 Lymphocytes inhibit tumor cell prolifer- ation and migration through induction of cytotoxic cell death and thus play a key role in cancer immuno-surveillance.5 On the basis of these findings, several inflammation-based scores have emerged as prognostic indicators in cancer patients. PDAC undergoing resection. The relationship between tumor and inflammation was first described by Virchow in 1863, and later in 1986 by Dvorak as BTumors: Wounds that do not heal.^ Meanwhile, inflamma- tion is known as one of the hallmarks of cancer.4,24,25 Increasing data shows the close relationship between tumori- genesis, tumor progression, and metastasis.4–6 The major prognostic impact of inflammatory markers can be ascribed to a cytokine-driven immunogenic tumor microenvironment and a significant prognostic role of inflammation-based bio- markers and scores has recently been shown in a number of malignant diseases.5,7,26–28 One of the newly emerging prog- nostic scores is the SII, based on platelets, neutrophils, and lymphocytes. As a combination of both NLR and PLR, SII firstly has been confirmed as superior prognostic factor in hepatocellular carcinoma and then in small cell lung cancer reflecting patient’s inflammatory status.15,16 In a number of malignancies, including PDAC, an elevated preoperative SII plays a key role in prognosis estimation.12–16,29–31 This is the first study that has proven the prognostic value of the SII and is superior to PLR. Discussion The relationship between tumor and inflammation was first described by Virchow in 1863, and later in 1986 by Dvorak as BTumors: Wounds that do not heal.^ Meanwhile, inflamma- tion is known as one of the hallmarks of cancer.4,24,25 Recently published data is diverging regarding the prog- nostic value of NLR and PLR. The NLR, combining circulat- ing neutrophil and lymphocyte counts, and the PLR, combin- ing circulating platelets and lymphocyte counts, has been as- sociated with impaired survival in lung and ovarian cancers, while in PDAC results remain inconsistent.38–42 Whereas Mowbray et al. found preoperative NLR to be an independent prognostic predictor, Chawla et al. reports that neither NLR nor PLR predicts survival in patients who underwent pancre- atectomy for PDAC.43,44 A high SII, consisting of high neutrophil and platelet as well as low lymphocyte counts, indicates inflammation ac- tivity that may be associated with poor survival through enhanced tumor invasion and metastases. Investigating It has been proposed that SII is able to predict tumor recur- rence in a highly inflammatory tumor microenvironment with 616 J Gastrointest Surg (2020) 24:610–618 the prognostic capacity of SII, NLR, and PLR, our results were consistent with those of Chawla et al., revealing that preoperative SII, but not NLR and PLR, is an independent prognostic factor for OS in patients with resectable PDAC.43 Recently, Haldar and Ben-Eliyahu critically discussed the impact of perioperative β-adrenergic block- ade and COX2 inhibition on cancer outcomes.45 Thus, pa- tients with resectable PDAC who have elevated preopera- tive SII might benefit from anti-inflammatory and/or anti- immunotherapy prior and after surgery. mesopancreatic compartment both stimulates inflammatory activity and influences the likelihood of complete resection. Similarly, CA19-9, serves as a surrogate marker of tumor bur- den and thereby associated with inflammation. Conclusion In summary, the present study shows that preoperative SII is an independent predictor of OS in patients with PDAC under- going pancreatic resection that is superior to NLR and PLR. Measurement of SII is easily applicable and of low cost. Patients with preoperatively elevated SII might benefit from anti-inflammatory and/or anti-immunotherapy. Even though the results of the present study demonstrate that the SII is an independent prognostic factor in patients with PDAC undergoing resection, it has several limitations. Although the patients were prospectively entered into a database, a retrospective analysis was performed with a selection bias by the availability of complete blood cell count before surgery in daily practice. The cohort represents the experience of one center that needs to be validated by external cohort from another center. Acknowledgments Open access funding provided by Medical University of Vienna. The authors would like to thank Dr. Helmuth Haslacher for his assistance in the Material and Methods section (labora- tory methods). There are no consensual cutoff values for inflammation indices. The majority of studies determine individual cutoff levels by their relevance and significance, showing a signifi- cant prediction of survival when applied to the same patients’ cohort. As a result, there is a wide range of cutoff values for these indices. However, the present study demonstrated that the SII provided the strongest survival prediction compared to NLR and PLR in patients with PDAC undergoing surgery. Emphasis should be given to determine significant cutoff levels for inflammatory indices that stay valid when applied to independent cohorts of patients. Author Contribution All authors meet all the following criteria: substan- tial contributions to the conception or design of the work, or the acquisi- tion, analysis, or interpretation of data for the work; drafting the work or revising it critically for important intellectual content; final approval of the version to be published; and agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Conflict of Interest The authors declare no conflict of interest. The administration of different neo-/adjuvant chemothera- py regimens and changing policies of treating patients with borderline disease during the study period may have influ- enced the study result. However, it reflects the real-world sit- uation and patients included in the study were treated with standard regimens that were available at the respective period. We did not analyze the differences of prognostic strength of SII during different time intervals reflecting variations in neo-/ adjuvant chemotherapy regimens. However, we could proof the prognostic value of SII for the entire observation period, regardless neoadjuvant treatment was administered or not. We admit that in order to draw representative conclusions for the association between inflammatory activity and prognosis in patients undergoing neoadjuvant treatment, a cohort with little variations in neoadjuvant regimens should be analyzed. 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 appro- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. References 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin. 2011;61(2):69–90. doi:https:// doi.org/10.3322/caac.20107. 2. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010;46(4):765–81. doi:https://doi.org/10.1016/j.ejca.2009.12. 014. High and low SII was equally distributed in most of the items characterizing the study cohort. However, resection rates, CA19-9 and bilirubin blood concentrations were differ- ent between patients with high and low inflammatory activity. The difference of resection rates interesting observation that needs to be addressed by further studies as it may represent an important factor for treatment decision. There is evidence that tumor invasion initiates host inflammatory response and one could argue that the extent of tumor cell infiltration into the 3. Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell. 2010;140(6):883–99. doi:https://doi.org/10.1016/j. cell.2010.01.025. 4. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera- tion. Cell. 2011;144(5):646–74. doi:https://doi.org/10.1016/j.cell. 2011.02.013. 4. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera- tion. Cell. 2011;144(5):646–74. doi:https://doi.org/10.1016/j.cell. 2011.02.013. 5. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related in- flammation. Nature. 2008;454(7203):436–44. doi:https://doi.org/ 10.1038/nature07205. J Gastrointest Surg (2020) 24:610–618 617 Selecting Optimal Cutpoints in Diagnostic Tests. J Stat Soft. 2014;61(8). https://doi.org/10.18637/jss.v061.i08. 6. Proctor MJ, Morrison DS, Talwar D, Balmer SM, O'Reilly DS, Foulis AK et al. An inflammation-based prognostic score (mGPS) predicts cancer survival independent of tumour site: a Glasgow Inflammation Outcome Study. Br J Cancer. 2011;104(4):726–34. doi:https://doi.org/10.1038/sj.bjc.6606087. 22. Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3(1): 32–5. 23. Powers DMW. Evaluation: From precision, recall and f-measure to roc., informedness, markedness & correlation. Journal of Machine Learning Technologies. 2011;2(1):37–63. 7. Jomrich G, Hollenstein M, John M, Baierl A, Paireder M, Kristo I et al. The modified glasgow prognostic score is an independent prognostic indicator in neoadjuvantly treated adenocarcinoma of the esophagogastric junction. Oncotarget. 2018;9(6):6968–76. doi:https://doi.org/10.18632/oncotarget.24087. 24. Balkwill F, Mantovani A. Inflammation and cancer: back to Virchow? Lancet. 2001;357(9255):539–45. doi:https://doi.org/10. 1016/S0140-6736(00)04046-0. 8. Martinelli P, Carrillo-de Santa Pau E, Cox T, Sainz B, Jr, Dusetti N, Greenhalf W et al. GATA6 regulates EMT and tumour dissemina- tion, and is a marker of response to adjuvant chemotherapy in pan- creatic cancer. Gut. 2017;66(9):1665–76. doi:https://doi.org/10. 1136/gutjnl-2015-311256. 25. Dvorak HF. Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med. 1986;315(26):1650–9. doi:https://doi.org/10.1056/ NEJM198612253152606. 26. References Guthrie GJ, Roxburgh CS, Farhan-Alanie OM, Horgan PG, McMillan DC. Comparison of the prognostic value of longitudinal measurements of systemic inflammation in patients undergoing curative resection of colorectal cancer. Br J Cancer. 2013;109(1): 24–8. doi:https://doi.org/10.1038/bjc.2013.330. 9. Pelosi E, Castelli G, Testa U. Pancreatic Cancer: Molecular Characterization, Clonal Evolution and Cancer Stem Cells. Biomedicines. 2017;5(4). https://doi.org/10.3390/ biomedicines5040065. 10. Fidler IJ, Poste G. The Bseed and soil^ hypothesis revisited. Lancet Oncol. 2008;9(8):808. doi:https://doi.org/10.1016/S1470-2045(08) 70201-8. 27. Jomrich G, Paireder M, Gleiss A, Kristo I, Harpain L, Schoppmann SF. Comparison of Inflammation-Based Prognostic Scores in a Cohort of Patients with Resectable Esophageal Cancer. Gastroenterol Res Pract. 2017;2017:1678584. doi:https://doi.org/ 10.1155/2017/1678584. 11. Perini MV, Montagnini AL, Jukemura J, Penteado S, Abdo EE, Patzina R et al. Clinical and pathologic prognostic factors for cura- tive resection for pancreatic cancer. HPB (Oxford). 2008;10(5): 356–62. doi:https://doi.org/10.1080/13651820802140752. 28. Pinato DJ, Shiner RJ, Seckl MJ, Stebbing J, Sharma R, Mauri FA. Prognostic performance of inflammation-based prognostic indices in primary operable non-small cell lung cancer. Br J Cancer. 2014;110(8):1930–5. doi:https://doi.org/10.1038/bjc.2014.145. 12. Aziz MH, Sideras K, Aziz NA, Mauff K, Haen R, Roos D et al. The Systemic-Immune-Inflammation Index Independently Predicts Survival and Recurrence in Resectable Pancreatic Cancer and its Prognostic Value Depends on Bilirubin Levels: A Retrospective Multicenter Cohort Study. Ann Surg. 2018. doi:https://doi.org/10. 1097/SLA.0000000000002660. 29. Tong YS, Tan J, Zhou XL, Song YQ, Song YJ. Systemic immune- inflammation index predicting chemoradiation resistance and poor outcome in patients with stage III non-small cell lung cancer. J Transl Med. 2017;15(1):221. doi:https://doi.org/10.1186/s12967- 017-1326-1. 13. Feng JF, Chen S, Yang X. Systemic immune-inflammation index (SII) is a useful prognostic indicator for patients with squamous cell carcinoma of the esophagus. Medicine (Baltimore). 2017;96(4): e5886. doi:https://doi.org/10.1097/MD.0000000000005886. 30. Lolli C, Basso U, Derosa L, Scarpi E, Sava T, Santoni M et al. Systemic immune-inflammation index predicts the clinical outcome in patients with metastatic renal cell cancer treated with sunitinib. Oncotarget. 2016;7(34):54564–71. doi:https://doi.org/10.18632/ oncotarget.10515. 14. Geng Y, Shao Y, Zhu D, Zheng X, Zhou Q, Zhou Wet al. Systemic Immune-Inflammation Index Predicts Prognosis of Patients with Esophageal Squamous Cell Carcinoma: A Propensity Score- matched Analysis. Sci Rep. 2016;6:39482. doi:https://doi.org/10. 1038/srep39482. 31. Lolli C, Caffo O, Scarpi E, Aieta M, Conteduca V, Maines F et al. Systemic Immune-Inflammation Index Predicts the Clinical Outcome in Patients with mCRPC Treated with Abiraterone. Front Pharmacol. 2016;7:376. doi:https://doi.org/10.3389/fphar. 2016.00376. 15. Hong X, Cui B, Wang M, Yang Z, Wang L, Xu Q. References Systemic Immune-inflammation Index, Based on Platelet Counts and Neutrophil-Lymphocyte Ratio, Is Useful for Predicting Prognosis in Small Cell Lung Cancer. Tohoku J Exp Med. 2015;236(4):297– 304. doi:https://doi.org/10.1620/tjem.236.297. 32. McMillan DC. Cancer and systemic inflammation: stage the tu- mour and stage the host. Br J Cancer. 2013;109(3):529. doi: https://doi.org/10.1038/bjc.2013.418. 16. Hu B, Yang XR, Xu Y, Sun YF, Sun C, Guo W et al. Systemic immune-inflammation index predicts prognosis of patients after curative resection for hepatocellular carcinoma. Clin Cancer Res. 2014;20(23):6212–22. doi:https://doi.org/10.1158/1078-0432. CCR-14-0442. 33. Houghton AM, Rzymkiewicz DM, Ji H, Gregory AD, Egea EE, Metz HE et al. Neutrophil elastase-mediated degradation of IRS-1 accelerates lung tumor growth. Nat Med. 2010;16(2):219–23. doi: https://doi.org/10.1038/nm.2084. 34. De Larco JE, Wuertz BR, Furcht LT. The potential role of neutro- phils in promoting the metastatic phenotype of tumors releasing interleukin-8. Clin Cancer Res. 2004;10(15):4895–900. doi: https://doi.org/10.1158/1078-0432.CCR-03-0760. 17. Tempero MA, Malafa MP, Al-Hawary M, Asbun H, Bain A, Behrman SW et al. Pancreatic Adenocarcinoma, Version 2.2017, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2017;15(8):1028–61. doi:https://doi.org/10.6004/ jnccn.2017.0131. 35. Chen HC, Lin HC, Liu CY, Wang CH, Hwang T, Huang TT et al. Neutrophil elastase induces IL-8 synthesis by lung epithelial cells via the mitogen-activated protein kinase pathway. J Biomed Sci. 2004;11(1):49–58. doi:https://doi.org/10.1159/000075288. 18. R Development Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2018. 36. Teramukai S, Kitano T, Kishida Y, Kawahara M, Kubota K, Komuta K et al. Pretreatment neutrophil count as an independent prognostic factor in advanced non-small-cell lung cancer: an anal- ysis of Japan Multinational Trial Organisation LC00-03. Eur J Cancer. 2009;45(11):1950–8. doi:https://doi.org/10.1016/j.ejca. 2009.01.023. 19. Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77. doi: https://doi.org/10.1186/1471-2105-12-77. 20. Therneau T. A Package for Survival Analysis in S. 2.38 ed2015. Á 37. Labelle M, Begum S, Hynes RO. Direct signaling between platelets and cancer cells induces an epithelial-mesenchymal-like transition 37. Labelle M, Begum S, Hynes RO. Direct signaling between platelets and cancer cells induces an epithelial-mesenchymal-like transition 21. López-Ratón M, Rodríguez-Álvarez MX, Cadarso-Suárez C, Gude-Sampedro F. OptimalCutpoints: An R Package for 618 J Gastrointest Surg (2020) 24:610–618 and promotes metastasis. Cancer Cell. 2011;20(5):576–90. doi: https://doi.org/10.1016/j.ccr.2011.09.009. independent significant prognostic marker in resected pancreatic ductal adenocarcinoma. Am J Surg. 2009;197(4):466–72. doi: https://doi.org/10.1016/j.amjsurg.2007.12.057. References independent significant prognostic marker in resected pancreatic ductal adenocarcinoma. Am J Surg. 2009;197(4):466–72. doi: https://doi.org/10.1016/j.amjsurg.2007.12.057. 38. Walsh SR, Cook EJ, Goulder F, Justin TA, Keeling NJ. Neutrophil- lymphocyte ratio as a prognostic factor in colorectal cancer. J Surg Oncol. 2005;91(3):181–4. doi:https://doi.org/10.1002/jso.20329. 43. Chawla A, Huang TL, Ibrahim AM, Hardacre JM, Siegel C, Ammori JB. Pretherapy neutrophil to lymphocyte ratio and platelet to lymphocyte ratio do not predict survival in resectable pancreatic cancer. HPB (Oxford). 2018;20(5):398–404. doi:https://doi.org/10. 1016/j.hpb.2017.10.011. 39. Sarraf KM, Belcher E, Raevsky E, Nicholson AG, Goldstraw P, Lim E. Neutrophil/lymphocyte ratio and its association with surviv- al after complete resection in non-small cell lung cancer. J Thorac Cardiovasc Surg. 2009;137(2):425–8. doi:https://doi.org/10.1016/j. jtcvs.2008.05.046. 44. Mowbray NG, Griffith D, Hammoda M, Shingler G, Kambal A, Al- Sarireh B. A meta-analysis of the utility of the neutrophil-to- lymphocyte ratio in predicting survival after pancreatic cancer re- section. HPB (Oxford). 2018;20(5):379–84. doi:https://doi.org/10. 1016/j.hpb.2017.12.009. 40. Kao SC, Pavlakis N, Harvie R, Vardy JL, Boyer MJ, van Zandwijk N et al. High blood neutrophil-to-lymphocyte ratio is an indicator of poor prognosis in malignant mesothelioma patients undergoing sys- temic therapy. Clin Cancer Res. 2010;16(23):5805–13. doi:https:// doi.org/10.1158/1078-0432.CCR-10-2245. 45. Haldar R, Ben-Eliyahu S. Reducing the risk of post-surgical cancer recurrence: a perioperative anti-inflammatory anti-stress approach. Future Oncol. 2018;14(11):1017–21. doi:https://doi.org/10.2217/ fon-2017-0635. 41. Cho H, Hur HW, Kim SW, Kim SH, Kim JH, Kim YT et al. Pre- treatment neutrophil to lymphocyte ratio is elevated in epithelial ovarian cancer and predicts survival after treatment. Cancer Immunol Immunother. 2009;58(1):15–23. doi:https://doi.org/10. 1007/s00262-008-0516-3. Publisher’s Note Springer Nature remains neutral with regard to jurisdic- tional claims in published maps and institutional affiliations. 42. Smith RA, Bosonnet L, Raraty M, Sutton R, Neoptolemos JP, Campbell F et al. Preoperative platelet-lymphocyte ratio is an
https://openalex.org/W3158133555	https://art.torvergata.it/bitstream/2108/295211/2/nutrients-13-01389%20REPRINT.pdf	English	null	Dietary Strategies for Management of Metabolic Syndrome: Role of Gut Microbiota Metabolites	Nutrients	2,021	cc-by	14,738	Citation: Croci, S.; D’Apolito, L.I.; Gasperi, V.; Catani, M.V.; Savini, I. Dietary Strategies for Management of Metabolic Syndrome: Role of Gut Microbiota Metabolites. Nutrients 2021, 13, 1389. https://doi.org/ 10.3390/nu13051389 Keywords: metabolic syndrome; gut microbiota; dysbiosis; nutrition; inflammation; probiotics; prebiotics Academic Editor: Margarita Aguilera Received: 31 January 2021 Accepted: 16 April 2021 Published: 21 April 2021 nutrients nutrients nutrients nutrients nutrients Review Dietary Strategies for Management of Metabolic Syndrome: Role of Gut Microbiota Metabolites †, Lina Ilaras D’Apolito †, Valeria Gasperi * , Maria Valeria Catani ,‡ and Isabella Savini ‡ Sonia Croci †, Lina Ilaras D’Apolito †, Valeria Gasperi , Maria Valeria Catani ,‡ and Isab Department of Experimental Medicine, Tor Vergata University of Rome, 00133 Rome, Italy; Sonia.croci@gmail.com (S.C.); Dottoressadapolito@gmail.com (L.I.D.); savini@uniroma2.it (I.S.) Correspondence: gasperi@med.uniroma2.it (V.G.); catani@uniroma2.it (M.V.C.); Tel.: +39-06-72596465 (V.G. & M.V.C.) Department of Experimental Medicine, Tor Vergata University of Rome, 00133 Rome, Italy; Sonia.croci@gmail.com (S.C.); Dottoressadapolito@gmail.com (L.I.D.); savini@uniroma2.it (I.S.) * Correspondence: gasperi@med.uniroma2.it (V.G.); catani@uniroma2.it (M.V.C.); Tel.: +39-06-72596465 (V.G. & M.V.C.) † These authors contributed equally to this work. ‡ These authors contributed equally to this work. Department of Experimental Medicine, Tor Vergata University of Rome, 00133 Rome, Italy; Sonia.croci@gmail.com (S.C.); Dottoressadapolito@gmail.com (L.I.D.); savini@uniroma2.it (I.S.) * Correspondence: gasperi@med.uniroma2.it (V.G.); catani@uniroma2.it (M.V.C.); Tel.: +39-06-72596465 (V.G. & M.V.C.) † These authors contributed equally to this work. ‡ These authors contributed equally to this work. Abstract: Metabolic syndrome (MetS) is a complex pathophysiological state with incidence similar to that of a global epidemic and represents a risk factor for the onset of chronic non-communicable degenerative diseases (NCDDs), including cardiovascular disease (CVD), type 2 diabetes mellitus, chronic kidney disease, and some types of cancer. A plethora of literature data suggest the potential role of gut microbiota in interfering with the host metabolism, thus inﬂuencing several MetS risk factors. Perturbation of the gut microbiota’s composition and activity, a condition known as dysbiosis, is involved in the etiopathogenesis of multiple chronic diseases. Recent studies have shown that some micro-organism-derived metabolites (including trimethylamine N-oxide (TMAO), lipopolysaccharide (LPS) of Gram-negative bacteria, indoxyl sulfate and p-cresol sulfate) induce subclinical inﬂammatory processes involved in MetS. Gut microbiota’s taxonomic species or abundance are modiﬁed by many factors, including diet, lifestyle and medications. The main purpose of this review is to highlight the correlation between different dietary strategies and changes in gut microbiota metabolites. We mainly focus on the validity/inadequacy of speciﬁc dietary patterns to reduce inﬂammatory processes, including leaky gut and subsequent endotoxemia. We also describe the chance of probiotic supplementation to interact with the immune system and limit negative consequences associated with MetS. Citation: Croci, S.; D’Apolito, L.I.; Gasperi, V.; Catani, M.V.; Savini, I. Dietary Strategies for Management of Metabolic Syndrome: Role of Gut Microbiota Metabolites. Nutrients 2021, 13, 1389. https://doi.org/ 10.3390/nu13051389 1. Introduction Metabolic syndrome (MetS) is deﬁned by WHO as a pathological condition character- ized by obesity, insulin resistance, hypertension, hyperlipidemia and waist-to-hip ratio; MetS is present if three or more of the above-mentioned criteria are present [1]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Main MetS risk factors are a sedentary lifestyle and high sugar or fat diets [1], which over time lead to being overweight and obesity. An excess of adipose tissue, particularly in the visceral compartment, is associated with morbidity and complications of MetS, because of the increased production of inﬂammatory molecules (cytokines and adipokines) [2,3]. Adipose tissue is, indeed, an active and complex endocrine organ that secretes molecules, playing a key role in inﬂammation, immune response, appetite regulation, vascular events, reproductive functions, and insulin sensitivity [4]. Copyright: © 2021 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/). y Recent scientific data have linked MetS development also to intestinal dysbiosis [5,6]: because of increased intestinal permeability, the translocation of lipopolysaccharide (LPS), an outer membrane component of Gram-negative bacteria, occurs, thus determining metabolic endotoxemia, which can be viewed as a causal factor of chronic low-grade systemic inflam- mation [7,8]. The complex ecosystem of microorganisms (including bacteria, viruses, protozoa, and fungi) living in different districts of the human body (gastrointestinal tract, skin, mouth, https://www.mdpi.com/journal/nutrients Nutrients 2021, 13, 1389. https://doi.org/10.3390/nu13051389 Nutrients 2021, 13, 1389 2 of 19 respiratory and urogenital systems) is deﬁned as microbiota. Most microbiotas reside in the gastrointestinal tube [9]. The microbiota contains over 100 times more unique genes than those codiﬁed in the human genome [10]: it encompasses over 100 trillion microbes and 5000 different species, accounting for 5 million genes. respiratory and urogenital systems) is deﬁned as microbiota. Most microbiotas reside in the gastrointestinal tube [9]. The microbiota contains over 100 times more unique genes than those codiﬁed in the human genome [10]: it encompasses over 100 trillion microbes and 5000 different species, accounting for 5 million genes. Human microbiotas vary according to diet, lifestyle and the presence or absence of speciﬁc diseases. 1. Introduction Indeed, it has been shown that a high fat and sugar diet alters the intestinal ecosystem (at the phyla, genus, and/or species level), triggering changes in metabolic pathways and inducing pro-inﬂammatory cascade signals [11,12]. Some studies have shown that dysbiosis causes low-grade inﬂammation, obesity and consequently MetS [13], while others have shown that dysbiosis is the result of low-grade inﬂammation during obesity and MetS [2,14]. Further investigations are needed to understand if dysbiosis is a cause or consequence of MetS. Nonetheless, modulation of dysbiosis via dietary interventions and probiotic supplementation has provided evidence for supporting a possible beneﬁcial effect in treating MetS complications [15]. p g p There is growing evidence about the role of bacterial metabolites as pathophysi- ological biomarkers of cardio-metabolic diseases, with applications in diagnostics and prognosis [16]; for example, plasma levels of trimethylamine N-oxide (TMAO; derived from choline and carnitine) are associated with various diseases [17,18], while indoxyl sulfate (derived from tryptophan breakdown) and p-cresyl sulfate (derived from tyrosine and phenylalanine), as well as their metabolites [19], are involved in the progression of kidney and vascular diseases (Table 1) [20]. Table 1. Gut microbiota metabolites, diet, and host effects. Family Metabolites Diet Effects References Prevotellaceae SCFAs Fiber-rich diet • Energy source for colonocytes • Signaling molecules • Regulation of intestinal transit time • Modulation of host appetite and food intake [21,22] Enterobacteriaceae LPS Western diet • Metabolic endotoxemia • Pro-inﬂammatory signaling [11,23–25] Enterobacteriaceae, Bacteoridaceae, Ruminococcaceae, Lachnospiraceae TMAO Dietary source of choline • Risk factor for MetS, cardiovascular and cerebrovascular events [17,18,26–36] Clostridiaceae Indoxyl sulfate, p-cresyl sulfate Western diet; VLCKD • Chronic kidney disease • Cardiovascular, metabolic and brain disorders [20,37–43] LPS: lipopolysaccharide; MetS: metabolic syndrome; SCFA: short chain fatty acid; TMAO: trimethylamine N-oxide; VLCKD: very low carbohydrate ketogenic diet. Table 1. Gut microbiota metabolites, diet, and host effects. Table 1. Gut microbiota metabolites, diet, and host effects. haride; MetS: metabolic syndrome; SCFA: short chain fatty acid; TMAO: trimethylamine N-oxide; VLCKD: very low genic diet. The habitual diet plays an important role in deﬁning the composition of the intestinal microbiota and determining the microbial metabolites that can affect the host metabolism. In the literature, several studies have shown positive effects of some dietary models in MetS management. For example, adherence to the Mediterranean diet (MD) leads to signiﬁcantly higher levels of total short-chain fatty acids (SCFAs), important gut microbiota metabolites that modulate immune–endocrine processes [44]. 1. Introduction p The purpose of this review is to examine and evaluate potential microbial metabolites involved in MetS to deﬁne nutritional interventions aimed at modulating the intestinal microbiota and personalizing MetS treatment. 3 of 19 3 of 19 Nutrients 2021, 13, 1389 Nutrients 2021, 13, x FOR 3. LPS and Endotoxemia In healthy conditions, the integrity of the intestinal epithelial barrier, guaranteed by a tight junction network, blocks passage of antigens or microbe-derived endotoxins. Some pathological conditions give rise to gut microbiota perturbations (referred to as dysbiosis) and a subsequent impairment of intestinal barrier function (due to disorganized tight junction proteins, zonulin and occludin, in colonocytes); in these circumstances, microbial metabolites can cross the intestinal barrier and move to the bloodstream, triggering sys- temic pro-inﬂammatory signaling that, in turn, causes metabolic alterations (peripheral insulin resistance, hyperglycemia and non-alcoholic fatty liver disease (NAFLD)) in distant tissues [11,53]. Microbe-derived pro-inﬂammatory molecules include LPS, ﬂagellins and peptidoglycans, which can reach the bloodstream, leading to “metabolic endotoxemia” [54]; this condition causes or exacerbates human pathologies, from intestinal (inﬂammatory bowel diseases, colon cancer) to neurological ones (Parkinson’s disease, autism), including MetS, obesity, multiple organ failure, renal and traumatic brain injuries, autoimmunity, transplant rejection, depression, chronic fatigue, and HIV disease [23,55,56]. p j p g Obesity, insulin resistance, and NAFLD, linked to MetS, are usually associated with low diversity in the gut microbiome and chronically higher levels of pro-inﬂammatory and microbiota derived LPS in circulation [57]. A growing body of evidence suggests the potential role of LPS in obesity, insulin resistance, hepatic steatosis, and systemic and local inﬂammatory processes [11]. The ability of LPS to induce proliferation and adipogenesis is supported by in vitro and in vivo (mice models and human subjects) studies. LPS has been shown to promote differentiation of pre-adipocytes in culture, via Janus kinase/signal trans- ducers and activators of transcription (JAK/STAT) signaling and AMPK-regulated cytosolic phospholipases A2 (cPLA2) expression [58]. A low dose (300 µg/kg/day) LPS injection in lean mice grown with a normal chow diet leads to weight gain, adipose tissue stores, metabolic alterations, and increased expression of pro-inﬂammatory markers [54,59,60]. In human studies, E. coli LPS exposure (3 ng/kg) has been shown to promote systemic insulin resistance and adipose tissue-related inﬂammation [61]. Speciﬁcally, low dose (0.6 ng/kg) intravenous LPS administration induces a rapid, transient increase in plasma interleukin (IL)-6 (25-fold) and tumor necrosis factor α (TNFα; 100-fold), followed by modest induc- tion of pro-inﬂammatory (i.e., IL-6, TNFα, monocyte chemoattractant protein-1 (MCP-1), suppressor of cytokine signaling 1 and 3 (SOCS1 and SOCS3)) cytokine expression in adipose tissue [62]. In addition, whole body insulin sensitivity and homeostatic model assessment for insulin resistance (HOMA-IR) are signiﬁcantly impaired after 24 h of an LPS injection [61]. 2. SCFA Beneﬁcial Metabolic Effects 2. SCFA Beneficial Metabolic Effects Most bacteria found in the gastrointestinal tract belong to Firmicutes (Lactobacillus, Blautia, Clostridium and Enterococcus genera) and Bacteroidetes (Bacteroides, Parabac- teroides, Porphyromonas, Prevotellaceae (Prevotella) and Rikenellaceae (Alistipes) genera) phyla, followed by Actinobacteria (Biﬁdobacteria), Proteobacteria (Escherichia coli), Fu- sobacteria, Verrucomicrobia and Cyanobacteria phyla [45–47]. Most bacteria found in the gastrointestinal tract belong to Firmicutes (Lactobacillus, Blautia, Clostridium and Enterococcus genera) and Bacteroidetes (Bacteroides, Parabacteroides, Porphyromonas, Prevotellaceae (Prevotella) and Rikenellaceae (Alistipes) genera) phyla, fol- lowed by Actinobacteria (Bifidobacteria), Proteobacteria (Escherichia coli), Fusobacteria, Verrucomicrobia and Cyanobacteria phyla [45–47]. Intestinal bacteria play an important role in regulation of the host metabolism (in- ﬂuencing energy homeostasis, appetite and food eating behavior) and modulation of the immune system, through the production of SCFAs, vitamins, metabolites, and neuropep- tides (Figure 1) [48,49]. Intestinal bacteria play an important role in regulation of the host metabolism (influ- encing energy homeostasis, appetite and food eating behavior) and modulation of the im- mune system, through the production of SCFAs, vitamins, metabolites, and neuropep- tides (Figure 1) [48,49]. Figure 1. Gut microbiota metabolites in eubiosis and dysbiosis. GPR: G protein coupled receptor; GLP: glucagon-like peptide; PYY: peptide YY. Figure 1. Gut microbiota metabolites in eubiosis and dysbiosis. GPR: G protein coupled receptor; GLP: glucagon-like peptide; PYY: peptide YY. Figure 1. Gut microbiota metabolites in eubiosis and dysbiosis. GPR: G protein coupled receptor; GLP: glucagon-like peptide; PYY: peptide YY Figure 1. Gut microbiota metabolites in eubiosis and dysbiosis. GPR: G protein coupled receptor; GLP: glucagon-like peptide; PYY: peptide YY. Firmicutes bacteria produce SCFAs as products of the hydrolysis and fermentation of dietary polysaccharides; once taken up by enterocytes, butyrate is usually used as an energy source for intestinal epithelium, while acetate and propionate enter in systemic Firmicutes bacteria produce SCFAs as products of the hydrolysis and fermentation of dietary polysaccharides; once taken up by enterocytes, butyrate is usually used as an energy source for intestinal epithelium, while acetate and propionate enter in systemic circulation and reach peripheral tissues, where they are used for lipogenesis and gluconeogenesis. 2. SCFA Beneﬁcial Metabolic Effects 2. SCFA Beneficial Metabolic Effects Nutrients 2021, 13, 1389 4 of 19 SCFAs also act as signaling molecules, thereby exerting beneﬁcial metabolic effects, including the regulation of intestinal transit time [21]; modulation of food intake [50,51]; in- crease in intestinal absorption of monosaccharides (by inducing the expression of sodium/ glucose transporter-1) [50]; increase in lipogenesis by inhibiting the inhibitor of lipoprotein lipase in the small intestine (thus inhibiting fatty acid release from triglycerides) and pro- moting cellular uptake of triglycerides (thus stimulating their storage in depot organs) [52]. g p g y g g p g Speciﬁcally, SCFAs stimulate colonic L cells through activation of the G protein- coupled receptors GPR41 (Free Fatty Acid Receptor 3 or FFA3) and GPR43 (Free Fatty Acid Receptor 2 or FFA2); the ﬁrst receptor triggers the release of the anorexigenic peptide PYY, which increases the gut transit rate and satiety, while the second one reduces inﬂammation, stimulates release of glucagon-like peptide 1 and 2 (GLP1 and GLP2) and increases con- centrations of peripheral hormones, such as insulin, leptin and ghrelin, thus inﬂuencing appetite control [51]. 4. Nonalcoholic Fatty Liver Disease (NAFLD) and Microbiome NAFLD is the hepatic manifestation of cardiometabolic syndrome. Systemic LPS concentration is signiﬁcantly elevated in NAFLD compared to control groups, in both human and animal studies [68–71]. The gut microbiota contributes to liver fat deposition through modulation of the nuclear farnesoid X receptor (FXR), responsible for regulation of bile acid synthesis, and hepatic triglyceride accumulation [8,11,55]. After a meal, primary bile acids (chenodeoxycholic and cholic acids), stored in the gall bladder, are secreted in the duodenum, where they can be deconjugated by gut microbes, thus being metabolized into secondary bile acids in the colon [72]. Bacteria with the capability of producing secondary bile acids belong to Clostridium (clusters XIVa and XI) and Eubacterium [73,74]. g Gut microbiota involvement in NAFLD genesis has been demonstrated by micro- biota transplantation in recipient germ-free mice, as it generates fasting hyperglycemia, insulinemia and NAFLD; in particular, Lachnospiraceae bacterium 609 and Barnesiella intes- tinihominis species were speciﬁcally related to NAFLD [75]. The gut microbiota contributes to NAFLD development also by increasing the number of ethanol-producing bacteria (e.g., Escherichia coli, Klebsiella pneumoniae), thereby favoring gut permeability alteration, generation of reactive oxygen species (ROS), and liver inﬂammation [76]. However, it should be taken into account that the genetic background of the mice (wild-type, inbred strains and their transgenic derivatives) strongly affects gut microbiota composition [55] and, moreover, each mammalian species harbors a distinct microbial com- position [15], thus limiting the translation of results obtained in animal models to humans. 3. LPS and Endotoxemia However, a weakness of these studies is that the immune response can be stimulated by an administered circulating dose of LPS greater than that actually released by bacteria; there are no studies capable of quantifying bacterial LPS production. y p q y g p In vitro molecular studies show that LPS impairs insulin sensitivity through activation of Toll-like receptors (TLRs): in particular, LPS binds to LPS-binding protein, activates the CD14 receptor and transfers TLR4 in the plasma membrane of macrophages [61]. In Nutrients 2021, 13, 1389 5 of 19 macrophages and dendritic cells, LPS also regulates nucleotide oligomerization domain (NOD)-like receptors, thus triggering activation of pro-inﬂammatory transcription factors, including nuclear factor-κB (NF-κB), activator protein-1 (AP-1) and interferon regulatory factors (IRFs); among downstream targets, several components of the inﬂammasome (caspase-1 and apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC)) are transcribed [63–66]. Collectively, these events regulate glucose and lipid homeostasis; accordingly, Cani and colleagues [54] reported increased expression of pro-inﬂammatory (TNFα, IL6, IL-1), oxidative stress (NADPH oxidase, inducible nitric oxide synthase (iNOS)) and macrophage inﬁltration (CD86) markers in the liver of geneti- cally obese mice with metabolic endotoxemia [54,59,60]. Conversely, a systematic review, analyzing serum concentrations of LPS and/or LPS-binding protein in diabetic patients, highlighted that high LPS levels were observed both in diabetic subjects and healthy con- trols [67]. In order to study the actual involvement of LPS in glycemic regulation, these human studies require further conﬁrmation, conducted on larger and uniform samples (for age, diet, lifestyle, cigarette smoking, use of drugs, etc.). In addition, it would be interesting to investigate if, in diabetic subjects, inﬂammatory markers are really correlated with LPS serum levels. 4. Nonalcoholic Fatty Liver Disease (NAFLD) and Microbiome 5. Role of Bacterial TMAO and Development of Atherosclerosis TMAO is a biomarker of risk for major adverse cardiovascular and cerebrovascular events, such as myocardial infarction and stroke (Figure 2): increased plasma TMAO concentrations have indeed been correlated with the accumulation of fatty depots in blood vessels, fatty liver, visceral obesity, and atherosclerosis [29,30,77–83]. y y Lecithin (the phospholipid phosphatidylcholine, found in eggs, liver, red meat, and ﬁsh) is the major dietary source of choline, a semi-essential nutrient part of the B-complex vitamin family [84]. Choline takes part in lipid metabolism, as well as in the synthesis of acetylcholine, homocysteine, and methionine [31]. Choline can be metabolized by gut microbiotas (mainly Enterobacteriaceae), with the production of trimethylamine (TMA), dimethylamine (DMA) and monomethylamine (MMA), which reach the liver where they are converted into TMAO via hepatic ﬂavin-monooxygenase-3 [32]. Therefore, dietary con- sumption of choline may lead to the generation of nitrosamine precursors with carcinogenic Nutrients 2021, 13, 1389 6 of 19 activity; it is noteworthy that TMA-producing bacterial strains (such as Escherichia fergusonii, Clostridium sporogenes, and Proteus penneri) have been detected in the human gut [31,85]. 6 of 19 activity; it is noteworthy that TMA-producing bacterial strains (such as Escherichia fergusonii, Clostridium sporogenes, and Proteus penneri) have been detected in the human gut [31,85]. 6 of 19 activity; it is noteworthy that TMA-producing bacterial strains (such as Escherichia fergusonii, Clostridium sporogenes, and Proteus penneri) have been detected in the human gut [31,85]. 6 of 19 Figure 2. Metabolic syndrome and gut microbiota intestinal dysbiosis. Increased intestinal perme- ability causes translocation of lipopolysaccharide and tryptophan-derived metabolites, with subse- quent metabolic endotoxemia and chronic low-grade systemic inflammation. Figure 2. Metabolic syndrome and gut microbiota intestinal dysbiosis. Increased intestinal permeability causes translocation of lipopolysaccharide and tryptophan-derived metabolites, with subsequent metabolic endotoxemia and chronic low-grade systemic inﬂammation. Figure 2. Metabolic syndrome and gut microbiota intestinal dysbiosis. Increased intestinal perme- ability causes translocation of lipopolysaccharide and tryptophan-derived metabolites, with subse- quent metabolic endotoxemia and chronic low-grade systemic inflammation. Figure 2. Metabolic syndrome and gut microbiota intestinal dysbiosis. Increased intestinal permeability causes translocation of lipopolysaccharide and tryptophan-derived metabolites, with subsequent metabolic endotoxemia and chronic low-grade systemic inﬂammation. n (the phospholipid phosphatidylcholine, found in eggs, liver, red meat, and major dietary source of choline, a semi-essential nutrient part of the B-complex mily [84]. Choline takes part in lipid metabolism, as well as in the synthesis of ne, homocysteine, and methionine [31]. 5. Role of Bacterial TMAO and Development of Atherosclerosis Choline can be metabolized by gut mi- mainly Enterobacteriaceae), with the production of trimethylamine (TMA), di- ne (DMA) and monomethylamine (MMA), which reach the liver where they ed into TMAO via hepatic flavin-monooxygenase-3 [32]. Therefore, dietary n of choline may lead to the generation of nitrosamine precursors with car- ctivity; it is noteworthy that TMA-producing bacterial strains (such as Esche- onii, Clostridium sporogenes, and Proteus penneri) have been detected in the hu- ,85]. MAO serum levels are related to MetS and cardiovascular risk. In 4007 healthy , deuterium [D9]-labeled phosphatidylcholine and the ingestion of two hard- have been shown to increase plasma TMAO concentrations that are reduced levels by oral broad-spectrum antibiotics [33]. The same group demonstrated tionship between the incidence of major adverse cardiovascular events (death, infarction, or stroke) and TMAO plasma levels, in a long-term study (three ow up) [33]. Similarly, in a cohort of 339 patients, plasma TMAO concentra- higher in patients with diabetes compared to euglycemic subjects, as well as in th MetS compared to healthy individuals; also, a direct relationship between AO levels and cardiovascular adverse events has been reported [34]. In 220 ting serum TMAO levels positively correlate with age, body mass index (BMI), emia, blood lipids and carotid intima-media thickness [29]. A meta-analysis ella et al. demonstrated a positive dose-dependent association between circu- O levels and all causes of mortality for cardio/cerebrovascular events [35]. A l b ti l t d i d t 330 d lt bj t ith M tS d High TMAO serum levels are related to MetS and cardiovascular risk. In 4007 healthy participants, deuterium [D9]-labeled phosphatidylcholine and the ingestion of two hard- boiled eggs have been shown to increase plasma TMAO concentrations that are reduced to baseline levels by oral broad-spectrum antibiotics [33]. The same group demonstrated a direct relationship between the incidence of major adverse cardiovascular events (death, myocardial infarction, or stroke) and TMAO plasma levels, in a long-term study (three years of follow up) [33]. Similarly, in a cohort of 339 patients, plasma TMAO concentra- tions were higher in patients with diabetes compared to euglycemic subjects, as well as in patients with MetS compared to healthy individuals; also, a direct relationship between plasma TMAO levels and cardiovascular adverse events has been reported [34]. 5. Role of Bacterial TMAO and Development of Atherosclerosis In 220 sub- jects, fasting serum TMAO levels positively correlate with age, body mass index (BMI), fasting glycemia, blood lipids and carotid intima-media thickness [29]. A meta-analysis by Schiattarella et al. demonstrated a positive dose-dependent association between cir- culating TMAO levels and all causes of mortality for cardio/cerebrovascular events [35]. A cross-sectional observational study, carried out on 330 adult subjects with MetS, docu- mented that circulating TMAO levels positively increased with BMI, as well as with MetS indexes (such as the visceral adiposity index and the fatty liver index) [36]. Additionally, a gender-speciﬁc relationship between TMAO levels and MetS markers (obesity, blood pressure, serum lipids, serum glucose and insulin resistance-related index) was identiﬁed in 1081 subjects [26]. Finally, relationships between the gut microbiota composition and circulating metabolites, including fasting plasma TMAO levels, have been identiﬁed in 531 middle-aged Finnish men, enrolled in the METabolic Syndrome In Men (METSIM) study: speciﬁcally, (i) the most dominant species belong to Bacteoridaceae, Ruminococcaceae and Lachnospiraceae families; (ii) the abundance of Peptococcaceae and Prevotella appears to be strictly linked to TMAO levels; (iii) there is a negative association between TMAO concentrations and an abundance of Faecalibacterium prausnitzii [7]. nal observational study, carried out on 330 adult subjects with MetS, docu- t circulating TMAO levels positively increased with BMI, as well as with MetS ch as the visceral adiposity index and the fatty liver index) [36] Additionally What are the molecular pathways by which TMAO exerts its pathogenic effects? TMAO may promote dyslipidemia by regulating hepatic lipogenesis and gluconeogen- Nutrients 2021, 13, 1389 7 of 19 esis [10], macrophage scavenger receptors [27], while downregulating cholesterol and bile acid metabolism [28], as well as impairing macrophage reverse cholesterol trans- port [86], promoting movement of activated leukocytes to endothelial cells [18], activating NF-κB signaling [18], and enhancing platelet activation, thus promoting a pro-thrombotic phenotype [17] and inducing endothelial dysfunction through activation of the NLRP3 inﬂammasome [87]. In addition, TMAO also affects brain functions, as it induces neuronal senescence, increases oxidative stress, impairs mitochondrial function, inhibits mTOR signaling and upregulates expression of macrophage scavenger receptors and CD68, all phenomena that contribute to brain aging and cognitive impairment [18,83,88]. These studies suggest that analysis of serum or cerebrospinal ﬂuid TMAO levels could represent, in clinical practice, a novel prevention and treatment tool. 6. Gut Microbiota and Tryptophan Metabolism Tryptophan (Trp) is an essential aromatic amino acid that is present in oats, bananas, dried prunes, milk, tuna ﬁsh, cheese, bread, poultry, peanuts, and chocolate. Gut mi- crobiota can directly utilize Trp (approximately 4–6%), limiting its bioavailability [37]. Bacteria-derived indoles, produced from Trp metabolism, can modulate host physiological and pathological pathways, thus contributing to cardiovascular, metabolic, and brain disor- ders [38]. For example, Clostridium sporogenes produces indole propionic acid from dietary Trp that plays an important role in the integrity of the intestinal barrier. p p y p g y Trp metabolites are also involved in MetS. Indeed, in human patients with MetS, overactivation of indoleamine 2,3-dioxygenase and increased serum levels of kynurenine have been reported [89]. Indoxyl sulfate and p-cresyl sulfate are two other Trp metabo- lites that stimulate GLP-1 in L cells and subsequent insulin secretion from pancreatic β cells [39,90]. These two metabolites seem to be also related to chronic kidney disease, a MetS complication, and related risk factors (cardiovascular disease (CVD); hypertension, diabetes and hyperhomocysteinemia) [40,41]; however, studies are few and sometimes controversial [42,91], so more investigations will certainly be useful for identifying future MetS and chronic kidney disease diagnostic markers. 8. Dietary Strategies for MetS Management and Gut Microbiota Modulation In MetS patients, nutritional intervention should ﬁrstly aim at reducing CVD and type 2 diabetes risk and usually includes reduction of body weight by 7–10%, followed by weight maintenance and lifestyle changes (such as increases in physical activity and stopping cigarette smoking). Losing as little as 5% of initial weight results in insulin sensitivity improvement, serum triglycerides and LDL-cholesterol reduction and decrease in systolic and diastolic blood pressure [99,100]. y p Obesity is one of the ﬁve fundamental features of MetS; the relationship between obesity and microbes is clearly established. Colonization of germ-free mice with gut mi- crobiotas derived from obese subjects led to increases in total weight, compared with transplantation of lean gut microbiotas [101]. Similar results have been observed in hu- mans, where obesity developed after fecal microbiota transplantation from overweight donors [102], whereas transplantation of lean microbiotas into individuals with MetS improved insulin sensitivity [103]. In obese subjects, main microbiota changes concern the reduction of Bacteroidetes and a proportional increase in Firmicutes and Actinobacteria [104–107]; increase in Bac- teroidetes in overweight subjects has also been reported [108]. However, conﬂicting data still exist [105,109,110], especially considering differences in age, sex, physiologi- cal state, and ethnicity; therefore, we are far away from understanding the real Firmi- cutes/Bacteroidetes ratio in obesity, and further studies are recommended. The Western diet, rich in saturated and trans fatty acids as well as sucrose intake, and low in ﬁber from fruits and vegetables, is linked to insulin resistance, dysbiosis and chronic non-infectious degenerative diseases [111–114]. The Western diet decreases microbial richness and increases the Firmicutes/Bacteroidetes ratio, in addition to in- creasing pro-inﬂammatory bacteria, and triggering changes at family, genus and species levels [20,22,24,25]. Conversely, a diet rich in ω-3 polyunsaturated fatty acids (PUFAs) is generally associated with anti-inﬂammatory effects [115]. Human studies on ω-3 PU- FAs and gut microbiotas are based on PUFA supplementation or eicosapentaenoic (EPA) and docosahexaenoic (DHA) acids-rich drinks; nonetheless, these studies highlighted the PUFA ability to increase the abundance of SCFA-producing bacterial genera [116]. Among butyrate-producing bacteria, the following genera, belonging to the Lachnospiraceae fam- ily of the phylum Firmicutes, have been found: Eubacterium, Roseburia, Anaerostipes, and Coprococcus [117,118]. Mokkala et al. [119] showed that ω-3 PUFAs are associated with intestinal permeability in vivo and linked to concentrations of serum zonulin (a marker of intestinal permeability), but the study was performed only in overweight pregnant women. 7. Different Dietary Patterns and Microbiota Enterotypes Diet strongly affects gut microbiota composition. Comparison of the fecal microbiota of European children and of children from a rural African village of Burkina Faso (following a polysaccharide-rich diet), for example, revealed significant differences: in Burkina Faso chil- dren, the predominant bacteria were Bacteroidetes, mainly Prevotella and Xylanibacter genera, while European children had significantly less SCFAs and more Enterobacteriaceae (Shigella and Escherichia) [22]. Diet-based differences have been confirmed in other studies: animal- based diets decrease levels of Firmicutes that metabolize plant polysaccharides (Roseburia, Eubacterium rectale and Ruminococcus bromii), while increasing the abundance of bile-tolerant microorganisms (Alistipes, Bilophila and Bacteroides) [92]. Accordingly, Prevotellaceae are prevalent in vegetarians, whereas the microbiotas of omnivores includes taxa Clostridiaceae, Bacteroidales and Eubacterium [20,93–96]. The gut microbiome is able to respond to short-term macronutrient changes; volun- teers placed on either plant-(grains, legumes, fruits, and vegetables) or animal-(meats, eggs, and cheeses) based diets for ﬁve days showed signiﬁcant temporal alterations in their microbial communities [92]. In contrast, such modulation determined by short term dietary intervention has not been found in Wu’s investigation, highlighting that, for relevant composition changes, long-term dietary patterns are needed: indeed, in fecal samples from 98 individuals, changes in gut microbiota composition occur rapidly (within 24 h) when shifting to a new dietary regimen, but only after 10 days, samples could be speciﬁcally di- vided into three enterotypes (mainly Bacteroides, associated with animal proteins/saturated fats, and Prevotella, associated with carbohydrates) [20]. Nevertheless, large cohort studies demonstrated that enterotypes might not always be stable and that enterotype stratiﬁca- tion is not so robust among populations [97,98]. Further studies (especially concerning Nutrients 2021, 13, 1389 8 of 19 8 of 19 enterotype analysis in homogenous samples and long dietary interventions) are needed in order to deﬁne stable microbiota categorizations, although different endogenous and exogenous variables cannot always be controlled. 8. Dietary Strategies for MetS Management and Gut Microbiota Modulation Several in vivo studies have underlined the relevance of dietary fat quality on metabolic health and gut microbiota composition. Patterson and co-workers [120] demonstrated that mice fed with high-fat diets of different compositions displayed peculiar microbial ecosystems: dietary saturated fatty acids (palm oil) were associated with a low abundance of Bacteroidetes, obesity and MetS; mono-unsaturated fatty acids (MUFAs; olive oil) led to an increase in the Bacteroidaceae family; ω-3 PUFAs (ﬂaxseed/ﬁsh oil) increased EPA and DHA concentrations, as well as the intestinal abundance of Biﬁdobacterium genus. Accordingly, fatty acid proﬁles, by modulating the microbiota composition and integrity of the epithelial barrier, may contribute to low-intensity inﬂammation and, consequently, endotoxemia [24,121]. Similarly, in humans, ω-3 PUFAs have been shown to modulate the gut microbial community (increasing the Bacteroidetes/Firmicutes ratio) with posi- tive metabolic outcomes on BMI, weight gain, obesity, insulin resistance, type 2 diabetes mellitus, and inﬂammatory bowel diseases [118,122–127]. MUFAs were also proven to positively impact the Bacteroidetes/Firmicutes ratio in 25 volunteers with risk of MetS: three MUFA-rich diets, indeed, increased Parabacteroides, Prevotella, Turicibacter, and Enter- Nutrients 2021, 13, 1389 9 of 19 obacteriaceae’s populations, and changes in gut microbiota proﬁles were more evident in obese subjects [124]. obacteriaceae’s populations, and changes in gut microbiota proﬁles were more evident in obese subjects [124]. j MD adherence reduces chronic inﬂammation, improves lipid proﬁle, insulin sensitiv- ity and endothelial function, while decreasing CVD incidence and all causes of mortality (myocardial infarction, stroke, cancer) [128,129]. MD is characterized by high intakes of vegetables, legumes, fruits, nuts, grains, ﬁsh, seafood, and poultry as sources of proteins, and low intakes of dairy products, red and processed meat, cream, and sugar drinks; lipid sources mainly include foods rich in unsaturated fatty acids, as well as other beneﬁcial active compounds (e.g., olive oil, ﬁsh, and nuts) [130]. Recently, the focus has shifted to the anti-inﬂammatory properties of polyphenols, as the protective roles of coffee and tea con- sumption against MetS have been proven for a long time [131]. Polyphenolic anthocyanins (e.g., blueberry) can lower triglycerides and fasting plasma glucose, while ameliorating systolic blood pressure. Eight-week supplementation of 4 cups of a freeze-dried strawberry beverage in MetS patients improved atherosclerosis risk factors by decreasing total and LDL-cholesterol levels and vascular cell adhesion-1 (VCAM1) expression [132]. p High adherence to MD positively impacts the gut microbiota composition and micro- bial metabolomes. 8. Dietary Strategies for MetS Management and Gut Microbiota Modulation Individuals with low adherence show high urinary TMAO levels, while compliance to MD is linked to increased SCFA levels, and an abundance of Prevotella and ﬁber-degrading Firmicutes [133,134]. Likewise, an increase in Bacteroidetes and Clostridia, a decrease in species belonging to Protectobacteria and Bacillaceae, and a drop in TMAO levels, in following MD, have been reported [135]. Haro’s group [136] analyzed the ability of two dietary patterns (MD and low-fat diet) to resolve dysbiosis: 106 male subjects (33 obese patients with MetS, 32 obese individuals without metabolic complications and 41 non-obese controls), enrolled in the CORDIOPREV study, were subjected to the two dif- ferent dietary interventions for 2 years. MD provided a minimum of 35% fats (22% MUFAs, 6% PUFAs and <10% saturated fats), 15% proteins and no more than 50% carbohydrates; the low-fat diet provided <30% total fats (<10% saturated fats, 12–14% MUFAs and 6–8% PUFAs), 15% protein and a minimum of 55% carbohydrates. In both diets, the cholesterol content was adjusted to <300 mg/day. What emerged from this study was that dysbiosis, characterized by a reduction of bacteria with saccharolytic activity and a subsequent reduc- tion in SCFA production, was more severe in obese patients with MetS and both dietary patterns were able to reverse dysbiosis by inducing increases in Bacteroides, Prevotella and Faecalibacterium genera. In addition, MD also increased Roseburia and Ruminococcus genera, suggesting the greater potential of MD to restore gut microbiota functionality, as compared to a low-fat diet [136]. These results suggest that chronic dietary intervention may restore gut dysbiosis in obese patients with coronary artery disease, depending on the degree of metabolic dysfunction; it would be interesting to investigate gender differences since the study enrolled only men. y y Some ﬁndings indicated that nutritional ketosis, induced by very low carbohydrate ketogenic diets (VLCKD; <10% carbohydrates per day), allowed weight management and improved metabolic and inﬂammatory markers, including lipids, glycated hemoglobin, high-sensitivity C-reactive protein, fasting insulin and glucose levels [137,138]. A recent, small (only nine studies) meta-analysis underlined a contradictory role of VLCKD on the gut microbiome, as it decreased microbial α diversity and richness; in particular, this dietary pattern increased Enterobacteriaceae (potentially pathogenic and pro-inﬂammatory species) and decreased the abundance of beneﬁcial Biﬁdobacterium species [139]. However, the meta- analysis included experimental studies conducted both in humans and murine models, and the human samples were not homogeneous in terms of age or pathological conditions. 9. Prebiotics, Probiotics, and Next Generation Microbes Prebiotics are selectively fermented components leading to speciﬁc changes in the composition and/or activity of gut microorganisms that may be beneﬁcial for the host’s health and wellbeing [145]. Among them, oligosaccharides, such as fructo-oligosaccharides (FOS) and galacto-oligosaccharides (GOS), represent complex carbohydrates with the most inﬂuence in Biﬁdobacterium and Lactobacillus growth [146]: 30 g per day of dietary ﬁber (soluble fraction) resulted in a higher abundance of Biﬁdobacterium spp. and Lactobacillus spp., as well as elevated levels of fecal butyrate [147]. However, the beneﬁcial effects are attributed to the simultaneous use of probiotics and prebiotics, according to the “synbiotic” concept formulated in 1995 by Gibson and Roberfroid [148]. The widely used and recently revised deﬁnition of prebiotic is now a substrate selectively utilized by host microorganisms, conferring health beneﬁts [149]. Current probiotics include microorganisms proven to exert positive effects on human health, such as Lactobacillus and Biﬁdobacterium genera [150]. Most individuals use pro- biotics as preventive tools, and consumption of fermented foods (yogurt and keﬁr) may be inversely related to MetS [151,152], even if the available literature does not uniformly agree on effective gut mucosal colonization. Nonetheless, supplementation with speciﬁc strains of Lactobacillus and Biﬁdobacterium has been found to (i) improve epithelial and mucosal barrier functions, (ii) inhibit the growth of pathogenic enteric bacteria and reduce production of pathogenic toxins, (iii) mediate some of the negative consequences associated with the consumption of high-fat diets; (iv) modulate the immune system, blood glucose levels and lipid proﬁles [153]. Investigations on animal experimental models and humans suggested an important role of probiotics on MetS and its complications. In a mouse model of obesity, supplemen- tation with Lactobacillus rhamnosus reduced visceral adiposity and diet-induced obesity, while improving the integrity of the gastrointestinal microbial lining [154]. In experi- mental atherosclerotic models (apolipoprotein E knockout mice feeding a high-fat diet for 12 weeks), supplementation with Lactobacillus rhamnosus GG changed a proportion of gut microbiotas and signiﬁcantly reduced atherosclerotic plaque size [155]. Lactobacillus reuteri V340 added to a hypocaloric diet, together with regular physical activity, has shown promising results for MetS management in animals and adult patients: supplementation once daily for 12 weeks, indeed, reduced IL-6 and soluble VCAM-1 [156]. In the same way, although based on limited sampling (only 25 individuals), consumption of milk containing B. lactis HN019 reduced classical MetS parameters and related cardiovascular risk factors (BMI, total cholesterol and LDLs, pro-inﬂammatory cytokines) [157]. 8. Dietary Strategies for MetS Management and Gut Microbiota Modulation A diet rich in proteins and low in carbohydrates promotes dysbiosis, causing an increase in pro-inﬂammatory bacteria and a reduction in SCFA-producing bacteria, with subsequent changes in bacterial metabolites; indeed, increased branched-chain fatty acids, indoxyl sulfate, p-cresol sulfate and TMAO have been observed with this dietary regi- men [135,140,141]. 10 of 19 Nutrients 2021, 13, 1389 10 of 19 Epidemiological evidence identiﬁes the beneﬁts of vegetarian dietary patterns (rich in ﬁber, but low in EPA and DHA) in both the prevention and treatment of MetS, CVD mortality and risk of coronary heart disease [142,143]. In adults, plant-based diets have beneﬁcial effects on gut microbiota, increasing Biﬁdobacterium and Lactobacillus species, elevating SCFA-producing bacteria and promoting Ruminococcus enterotype. Accordingly, indoxyl sulfate and p-cresol sulfate mean excretion is approximately 60% lower in vege- tarians than in omnivores [43]. In omnivores, proteins are of plant and/or animal origin, each type with unique digestibility and degradation patterns, depending on the microbes involved: colon bacteria involved in protein metabolism primarily include potentially pathogenic Bacteroides, Coliform, and Clostridium. Inhibition of these potential pathogens is often associated with the restoration of micro-ecosystem homeostasis by reducing the release of enterotoxins and adverse microbial metabolites [144]. 9. Prebiotics, Probiotics, and Next Generation Microbes p y y Speciﬁc commensal species, as novel probiotics (the so-called next-generation pro- biotics), exert positive effects by reducing inﬂammation and strengthening the epithelial barrier [158]. Among them, Akkermansia municiphila is a major propionate-producing bacterium, while Eubacterium hallii, Eubacterium rectale, Roseburia inulinivorans, Faecalibac- terium prausnitzii, Clostridium lavalense, Bacteroides uniformis and Ruminococcus bromii are responsible for most butyrate production [159]. Akkermansia muciniphila (0.5–5% of total bac- Nutrients 2021, 13, 1389 11 of 19 11 of 19 teria) is inversely associated with obesity, diabetes, cardiometabolic diseases, inﬂammatory bowel disease, hypertension, and liver diseases [160–165]. Amuc_1100, an outer membrane protein of A. muciniphila, is involved in gut epithelium integrity and is negatively associated with metabolic endotoxemia [6,165]. In overweight and obese volunteers, daily Akkermansia muciniphila supplementation, for 3 months, resulted in improved insulin sensitivity and reduced insulinemia and plasma total cholesterol [166]; accordingly, dietary interventions for reducing obesity (calorie restriction, energy reduction, diet rich in prebiotic ﬁbers) are associated with increased levels of A. muciniphila [167]. These preliminary results suggest that A. muciniphila may be a promising new therapeutic agent for MetS, although safety and absence of side effects must be assessed. Consumption of yogurt and other fermented products is positively associated with gastrointestinal and cardiovascular health, cancer risk, weight management, diabetes, and bone density [168]. Several studies evaluated the effects of keﬁr, a fermented milk product, on MetS. Keﬁr grains have a speciﬁc combination of bacteria (Lactobacillus, Lactococcus, Streptococcus Leuconostoc, and acetic acid bacteria) and yeasts (Saccharomyces, Kluyveromyces, and Candida), which exert anticarcinogenic, antimicrobial and anti-inﬂammatory activi- ties [13,169]. In MetS patients, 12 week-consumption of keﬁr improved anthropometrical measurements, blood pressure, lipid proﬁle, glycemic homeostasis (insulin and HOMA–IR) and inﬂammation (TNF–α and IFN–γ) [151]. In a prospective cohort study, fermented milk was associated with a 4% reduction in risk of stroke, ischemic heart disease, and cardiovas- cular mortality [152]. However, all described studies should be interpreted with caution because of the high heterogeneity and different probiotic strains (or different microbial content of keﬁr grains) used in these investigations. 10. Conclusions A relationship between microbiota metabolites, endotoxiemia and MetS is clearly emerging. Recent ﬁndings have shown that some microorganism-derived metabolites (in- cluding TMAO, LPS, indoxyl sulfate and p-cresol sulfate) induce subclinical inﬂammatory processes involved in MetS and CVD [17–19,33,35,36]. Although many studies have been conducted in mice models, which display a different phylogenetic makeup of bacterial communities with respect to humans [170,171], nonetheless, some phenomena related to MetS are similar between animal and human studies, as is, for example, the link between endotoxemia and insulin resistance and diabetes [11,61,165,172]. Among factors (age, lifestyle, drugs, type of birth, cigarette smoking, etc.) inﬂuencing microbiota composition, dietary habits play a key role; much evidence underlines that some dietary patterns are more effective than others in modifying the gut community. In particular, MD, rich in beneﬁcial nutrients and bioactive compounds, is particularly effective in modulating gut microbiota composition and inﬂammatory processes (including leaky gut and subsequent endotoxemia); compliance with MD is indeed linked to resolved dysbiosis, increased SCFA levels, reduced TMAO levels, and an abundance of Prevotella and ﬁber-degrading Firmicutes [133,135]. Contrastingly, the Western diet is low in “carbohy- drates accessible by the microbiota” and leads to the extinction of speciﬁc bacterial strains, thus negatively affecting a series of metabolic functions. A vegetarian or high-ﬁber diet can reduce total choline intake, thereby modulating the composition of intestinal microbiotas and/or their metabolites, but these patterns do not provide an adequate intake of PUFAs, important for anti-inﬂammatory effects and integrity of intestinal epithelium. For these reasons, long term MD appears to be the best treatment for preserving the greatest richness in microbiota species. Functional foods enriched with probiotic or synbiotic bacteria may also be useful in targeting microbiota composition or speciﬁc biochemical pathways, as they have been shown to signiﬁcantly improve insulin sensitivity and the lipid proﬁle [166]. Nevertheless, we are still far from bacterial therapy: individual diversity is remarkable because everyone has their own distinctive microbial pattern and, even if you try to modify it, the trend Nutrients 2021, 13, 1389 12 of 19 12 of 19 always seems to be restoration of the initial proﬁle. References 1. Saklayen, M.G. The Global Epidemic of the Metabolic Syndrome. Curr. Hypertens. Rep. 2018, 20, 12. [C y p y yp p 2. Yu, R.; Kim, C.-S.; Kang, J.-H. Inﬂammatory Components of Adipose Tissue as Target for Treatment of Metabolic Syndrome. In Food Factors for Health Promotion; KARGER: Basel, Switzerland, 2009; pp. 95–103. y p y yp p 2. Yu, R.; Kim, C.-S.; Kang, J.-H. Inﬂammatory Components of Adipose Tissue as Target for Treatment of Metabolic Syndrome. In Food Factors for Health Promotion; KARGER: Basel, Switzerland, 2009; pp. 95–103. f pp 3. Grundy, S.M.; Brewer, H.B.; Cleeman, J.I.; Smith, S.C.; Lenfant, C. Deﬁnition of Metabolic Syndrome. Circulation 2004, 109, 433–438. [CrossRef] [PubMed] r, J.S. Adipose Tissue as an Endocrine Organ. J. Clin. Endocrinol. Metab. 2004, 89, 2548–2556. [CrossRef] 4. Kershaw, E.E.; Flier, J.S. Adipose Tissue as an Endocrine Organ. J. Clin. Endocrinol. Metab. 2004, 8 [PubMed] [ ] 5. Vallianou, N.; Stratigou, T.; Christodoulatos, G.S.; Dalamaga, M. Understanding the Role of the Gut Microbiome and Microbial Metabolites in Obesity and Obesity-Associated Metabolic Disorders: Current Evidence and Perspectives. Curr. Obes. Rep. 2019, 8, 317–332. [CrossRef] 6. Everard, A.; Cani, P.D. Diabetes, obesity and gut microbiota. Best Pract. Res. Clin. Gastroenterol. 2013, 2 i, P.D. Diabetes, obesity and gut microbiota. Best Pract. Res. Clin. Gastroenterol. 2013, 27, 73–83. [CrossRef] 7. Minihane, A.M.; Vinoy, S.; Russell, W.R.; Baka, A.; Roche, H.M.; Tuohy, K.M.; Teeling, J.L.; Blaak, E.E.; Fenech, M.; Vauzour, D.; et al. Low-grade inﬂammation, diet composition and health: Current research evidence and its translation. Br. J. Nutr. 2015, 114, 999–1012. [CrossRef] 8. Cani, P.D.; Delzenne, N.M. Interplay between obesity and associated metabolic disorders: New insights into the gut microbiota. Curr. Opin. Pharmacol. 2009, 9, 737–743. [CrossRef] 9. Pascale, A.; Marchesi, N.; Marelli, C.; Coppola, A.; Luzi, L.; Govoni, S.; Giustina, A.; Gazzaruso, C. Microbiota and metabolic diseases. Endocrine 2018, 61, 357–371. [CrossRef] 10. Canyelles, M.; Tondo, M.; Cedó, L.; Farràs, M.; Escolà-Gil, J.; Blanco-Vaca, F. Trimethylamine N-Oxide: A Link among Diet, Gut Microbiota, Gene Regulation of Liver and Intestine Cholesterol Homeostasis and HDL Function. Int. J. Mol. Sci. 2018, 19, 3228. [CrossRef] 11. Cani, P.D.; Delzenne, N.M. Involvement of the gut microbiota in the development of low grade inﬂammation associated with obesity: Focus on this neglected partner. Acta Gastroenterol. Belg. 2010, 73, 267–269. [PubMed] 12. 10. Conclusions For the same reason, data on fecal microbiota transplantation, clinically successful in the treatment of recurrent Clostridium difﬁcile infections [173,174], are not solid and conclusive; the currently available studies are few [107,175,176] and with several study shortcomings, among which the most relevant are the lack of long-term follow-ups and of homogenous fecal samples from donors. In conclusion, although promising, microbiotas and microbiota-derived metabolites may not actually represent diagnostic MetS markers and/or have real therapeutic power, since many limitations must be overcome. The future challenge is to fully understand how different dietary patterns may speciﬁcally modulate the composition of gut microbiota and whether these changes are long-lasting; only a strong and robust knowledge will allow stabilizing the microbial community toward a healthy phenotype and reducing the risk of MetS. Author Contributions: S.C. and L.I.D. wrote the manuscript; M.V.C., V.G. and I.S. conceived the project and reviewed/edited the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. f [ ] [ ] 17. Zhu, W.; Gregory, J.C.; Org, E.; Buffa, J.A.; Gupta, N.; Wang, Z.; Li, L.; Fu, X.; Wu, Y.; Mehrabian, M.; et al. Gut Microbial Metabolite TMAO Enhances Platelet Hyperreactivity and Thrombosis Risk. Cell 2016, 165, 111–124. [CrossRef] References p-Cre p y -W.; Hsu, K.-H.; Lee, C.-C.; Sun, C.-Y.; Hsu, H.-J.; Tsai, C.-J.; Tzen, C.-Y.; Wang, Y.-C.; Lin, C.-Y.; Wu, M.-S. p-C yp p y 20. Wu, I.-W.; Hsu, K.-H.; Lee, C.-C.; Sun, C.-Y.; Hsu, H.-J.; Tsai, C.-J.; Tzen, C.-Y.; Wang, Y.-C.; Lin, C.-Y.; Wu, M.-S. p-Cresyl sulphate and indoxyl sulphate predict progression of chronic kidney disease. Nephrol. Dial. Transplant. 2011, 26, 938–947. [CrossRef] 21. Samuel, B.S.; Shaito, A.; Motoike, T.; Rey, F.E.; Backhed, F.; Manchester, J.K.; Hammer, R.E.; Williams, S.C.; Crowley, J.; Yanagisawa, M.; et al. Effects of the gut microbiota on host adiposity are modulated by the short-chain fatty-acid binding G protein-coupled receptor, Gpr41. Proc. Natl. Acad. Sci. USA 2008, 105, 16767–16772. [CrossRef] p p 22. De Filippo, C.; Cavalieri, D.; Di Paola, M.; Ramazzotti, M.; Poullet, J.B.; Massart, S.; Collini, S.; Pieraccini, G.; Lionetti, P. Impact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa. Proc. Natl. Acad. Sci. USA 2010, 107, 14691–14696. [CrossRef] [PubMed] 23. Lepper, P.M.; Schumann, C.; Triantaﬁlou, K.; Rasche, F.M.; Schuster, T.; Frank, H.; Schneider, E.M.; Triantaﬁlou, M.; von Eynatten, M. Association of Lipopolysaccharide-Binding Protein and Coronary Artery Disease in Men. J. Am. Coll. Cardiol. 2007, 50, 25–31. [CrossRef] enas, R.D.C.G.; Bressan, J. Dysbiosis and metabolic endotoxemia induced by high-fat diet. Nutr. Hosp. 2018 sRef] 24. Candido, T.L.N.; Alfenas, R.D.C.G.; Bressan, J. Dysbiosis and metabolic endotoxemia induced by high- 35, 1432–1440. [CrossRef] 25. Bibbò, S.; Ianiro, G.; Giorgio, V.; Scaldaferri, F.; Masucci, L.; Gasbarrini, A.; Cammarota, G. The role of diet on gut microbiota composition. Eur. Rev. Med. Pharmacol. Sci. 2016, 20, 4742–4749. 26. Gao, X.; Tian, Y.; Randell, E.; Zhou, H.; Sun, G. Unfavorable Associations Between Serum Trimethylamine N-Oxide and L- Carnitine Levels With Components of Metabolic Syndrome in the Newfoundland Population. Front. Endocrinol. 2019, 10, 168. [CrossRef] [PubMed] 27. Falony, G.; Vieira-Silva, S.; Raes, J. Microbiology Meets Big Data: The Case of Gut Microbiota–Derived Trimethylamine. Annu. Rev. Microbiol. 2015, 69, 305–321. [CrossRef] 28. Janeiro, M.; Ramírez, M.; Milagro, F.; Martínez, J.; Solas, M. Implication of Trimethylamine N-Oxide (TMAO) in Disease: Potential Biomarker or New Therapeutic Target. Nutrients 2018, 10, 1398. [CrossRef] 29. Randrianarisoa, E.; Lehn-Stefan, A.; Wang, X.; Hoene, M.; Peter, A.; Heinzmann, S.S.; Zhao, X.; Königsrainer, I.; Königsrainer, A.; Balletshofer, B.; et al. Relationship of Serum Trimethylamine N-Oxide (TMAO) Levels with early Atherosclerosis in Humans. Sci. Rep. 2016, 6, 26745. References [CrossRef] p 30. Zhu, L.; Baker, S.S.; Gill, C.; Liu, W.; Alkhouri, R.; Baker, R.D.; Gill, S.R. Characterization of gut microbiomes in nonalcoholic steatohepatitis (NASH) patients: A connection between endogenous alcohol and NASH. Hepatology 2013, 57, 601–609. [CrossRef] [PubMed] 31. Zeisel, S.H.; Warrier, M. Trimethylamine N -Oxide, the Microbiome, and Heart and Kidney Diseas 157–181. [CrossRef] 32. Zeisel, S.H.; Wishnok, J.S.; Blusztajn, J.K. Formation of methylamines from ingested choline and lecithin. J. Pharmacol. Exp. Ther. 1983, 225, 320–324. [PubMed] 33. Tang, W.H.W.; Wang, Z.; Levison, B.S.; Koeth, R.A.; Britt, E.B.; Fu, X.; Wu, Y.; Hazen, S.L. Intestinal Microbial Metabolism of Phosphatidylcholine and Cardiovascular Risk. N. Engl. J. Med. 2013, 368, 1575–1584. [CrossRef] [PubMed] 34. Mueller, D.M.; Allenspach, M.; Othman, A.; Saely, C.H.; Muendlein, A.; Vonbank, A.; Drexel, H.; von Eckardstein, A. Plasma levels of trimethylamine-N-oxide are confounded by impaired kidney function and poor metabolic control. Atherosclerosis 2015, 243, 638–644. [CrossRef] [ ] 35. Schiattarella, G.G.; Sannino, A.; Toscano, E.; Giugliano, G.; Gargiulo, G.; Franzone, A.; Trimarco, B.; Esposito, G.; Perrino, C. Gut microbe-generated metabolite trimethylamine-N-oxide as cardiovascular risk biomarker: A systematic review and dose-response meta-analysis. Eur. Heart J. 2017, 38, 2948–2956. [CrossRef] y 36. Barrea, L.; Annunziata, G.; Muscogiuri, G.; Di Somma, C.; Laudisio, D.; Maisto, M.; de Alteriis, G.; Tenore, G.; Colao, A.; Savastano, S. Trimethylamine-N-oxide (TMAO) as Novel Potential Biomarker of Early Predictors of Metabolic Syndrome. Nutrients 2018, 10, 1971. [CrossRef] 37. Yokoyama, M.T.; Carlson, J.R. Microbial metabolites of tryptophan in the intestinal tract with special reference to skatole. Am. J. Clin. Nutr. 1979, 32, 173–178. [CrossRef] 37. Yokoyama, M.T.; Carlson, J.R. Microbial metabolites of tryptophan in the intestinal tract with special reference to skatole. Am. J. Clin. Nutr. 1979, 32, 173–178. [CrossRef] 38 Agus A ; Planchais J ; Sokol H Gut Microbiota Regulation of Tryptophan Metabolism in Health and Disease Cell Host Microbe ais, J.; Sokol, H. Gut Microbiota Regulation of Tryptophan Metabolism in Health and Disease. Cell Host Mic 4. [CrossRef] 38. Agus, A.; Planchais, J.; Sokol, H. Gut Microbiota Regulation of Tryptophan Metabolism in Health a 2018, 23, 716–724. [CrossRef] 39. Gao, H.; Liu, S. Role of uremic toxin indoxyl sulfate in the progression of cardiovascular disease. Life Sci. 2017, 185, 23–29. [CrossRef] [ ] 40. Weiner, D.E. Chronic Kidney Disease as a Risk Factor for Cardiovascular Disease and All-Cause Mortality: A Pooled Analysis of Community-Based Studies. J. Am. Soc. Nephrol. 2004, 15, 1307–1315. References Org, E.; Blum, Y.; Kasela, S.; Mehrabian, M.; Kuusisto, J.; Kangas, A.J.; Soininen, P.; Wang, Z.; Ala-Korpela, M.; Hazen, S.L.; et al. Relationships between gut microbiota, plasma metabolites, and metabolic syndrome traits in the METSIM cohort. Genome Biol. 2017, 18, 70. [CrossRef] 13. Wang, P.-X.; Deng, X.-R.; Zhang, C.-H.; Yuan, H.-J. Gut microbiota and metabolic syndrome. Chin. Med. J. 2020, 133, 808–816. [CrossRef] [PubMed] 14. Ley, R.E.; Backhed, F.; Turnbaugh, P.; Lozupone, C.A.; Knight, R.D.; Gordon, J.I. Obesity alters gut microbial ecology. Proc. Natl. Acad. Sci. USA 2005, 102, 11070–11075. [CrossRef] [PubMed] 15. Rastelli, M.; Knauf, C.; Cani, P.D. Gut Microbes and Health: A Focus on the Mechanisms Linking Microbes, Obesity, and Related Disorders. Obesity 2018, 26, 792–800. [CrossRef] 16. Brial, F.; Le Lay, A.; Dumas, M.-E.; Gauguier, D. Implication of gut microbiota metabolites in cardiovascular and metabolic diseases. Cell. Mol. Life Sci. 2018, 75, 3977–3990. [CrossRef] [PubMed] 17. Zhu, W.; Gregory, J.C.; Org, E.; Buffa, J.A.; Gupta, N.; Wang, Z.; Li, L.; Fu, X.; Wu, Y.; Mehrabian, M.; et al. Gut Microbial Metabolite TMAO Enhances Platelet Hyperreactivity and Thrombosis Risk. Cell 2016, 165, 111–124. [CrossRef] 13 of 19 13 of 19 Nutrients 2021, 13, 1389 18. Seldin, M.M.; Meng, Y.; Qi, H.; Zhu, W.; Wang, Z.; Hazen, S.L.; Lusis, A.J.; Shih, D.M. Trimethylamine N-Oxide Promotes Vascular Inﬂammation Through Signaling of Mitogen-Activated Protein Kinase and Nuclear Factor-κB. J. Am. Heart Assoc. 2016, 5, e002767. [CrossRef] [PubMed] ] [ ] Vanholder, R.; Vaneechoutte, M.; Glorieux, G. p-Cresyl Sulfate. Toxins 2017, 9, 52. [CrossRef] [PubMed] H K H L C C S C Y H H J T i C J T C Y W Y C Li C Y W M S C [ ] [ ] 19. Gryp, T.; Vanholder, R.; Vaneechoutte, M.; Glorieux, G. p-Cresyl Sulfate. Toxins 2017, 9, 52. [CrossR 19. Gryp, T.; Vanholder, R.; Vaneechoutte, M.; Glorieux, G. p-Cresyl Sulfate. Toxins 2017, 9, 52. [CrossRef] [PubMed] 20. Wu, I.-W.; Hsu, K.-H.; Lee, C.-C.; Sun, C.-Y.; Hsu, H.-J.; Tsai, C.-J.; Tzen, C.-Y.; Wang, Y.-C.; Lin, C.-Y.; Wu, M.-S. p-Cresyl sulphate and indoxyl sulphate predict progression of chronic kidney disease. Nephrol. Dial. Transplant. 2011, 26, 938–947. [CrossRef] Vanholder, R.; Vaneechoutte, M.; Glorieux, G. p-Cresyl Sulfate. Toxins 2017, 9, 52. [CrossRef] [PubMed] ; Hsu, K.-H.; Lee, C.-C.; Sun, C.-Y.; Hsu, H.-J.; Tsai, C.-J.; Tzen, C.-Y.; Wang, Y.-C.; Lin, C.-Y.; Wu, M.-S. References Diversity of the Human Intestinal Microbial Flora. Science 2005, 308, 1635–1638. [CrossR 46. Eckburg, P.B. Diversity of the Human Intestinal Microbial Flora. Science 2005, 308, 1635–1638. [CrossRef] [PubMed] 47. Arumugam, M.; Raes, J.; Pelletier, E.; Le Paslier, D.; Yamada, T.; Mende, D.R.; Fernandes, G.R.; Tap, J.; Bruls, T.; Batto, J.-M.; et al. Enterotypes of the human gut microbiome. Nature 2011, 473, 174–180. [CrossRef] 48. Hartstra, A.V.; Bouter, K.E.C.; Bäckhed, F.; Nieuwdorp, M. Insights Into the Role of the Microbiome in Obesity and Type 2 Diabetes. Diabetes Care 2015, 38, 159–165. [CrossRef] 49. Marchesi, J.R.; Adams, D.H.; Fava, F.; Hermes, G.D.A.; Hirschﬁeld, G.M.; Hold, G.; Quraishi, M.N.; K.M.; et al. The gut microbiota and host health: A new clinical frontier. Gut 2016, 65, 330–339. [Cro 50. Tolhurst, G.; Heffron, H.; Lam, Y.S.; Parker, H.E.; Habib, A.M.; Diakogiannaki, E.; Cameron, J.; Grosse, J.; Reimann, F.; Gribble, F.M. Short-Chain Fatty Acids Stimulate Glucagon-Like Peptide-1 Secretion via the G-Protein-Coupled Receptor FFAR2. Diabetes 2012, 61, 364–371. [CrossRef] 51. de La Serre, C.B.; Ellis, C.L.; Lee, J.; Hartman, A.L.; Rutledge, J.C.; Raybould, H.E. Propensity to high-fat diet-induced obesity in rats is associated with changes in the gut microbiota and gut inﬂammation. Am. J. Physiol. Liver Physiol. 2010, 299, G440–G448. [CrossRef] [ ] 52. den Besten, G.; Bleeker, A.; Gerding, A.; van Eunen, K.; Havinga, R.; van Dijk, T.H.; Oosterveer, M.H.; Jonker, J.W.; Groen, A.K.; Reijngoud, D.-J.; et al. Short-Chain Fatty Acids Protect Against High-Fat Diet–Induced Obesity via a PPARγ-Dependent Switch From Lipogenesis to Fat Oxidation. Diabetes 2015, 64, 2398–2408. [CrossRef] g L. Roles of the gut in the metabolic syndrome: An overview. J. Intern. Med. 2017, 281, 319–336. [CrossRef] 53. Fändriks, L. Roles of the gut in the metabolic syndrome: An overview. J. Intern. Med. 2017, 281, 319 g y 54. Cani, P.D.; Amar, J.; Iglesias, M.A.; Poggi, M.; Knauf, C.; Bastelica, D.; Neyrinck, A.M.; Fava, F.; Tuohy, K.M.; Chabo, C.; et al. Metabolic Endotoxemia Initiates Obesity and Insulin Resistance. Diabetes 2007, 56, 1761–1772. [CrossRef] 55. Cani, P.D.; Joly, E.; Horsmans, Y.; Delzenne, N.M. Oligofructose promotes satiety in healthy human: A pilot study. Eur. J. Clin. Nutr. 2006, 60, 567–572. [CrossRef] 56. Paolella, G. Gut-liver axis and probiotics: Their role in non-alcoholic fatty liver disease. World J. Gastroenterol. 2014, 20, 15518. [CrossRef] 57. Le Chatelier, E.; Nielsen, T.; Qin, J.; Prifti, E.; Hildebrand, F.; Falony, G.; Almeida, M.; Arumugam, M.; Batto, J.-M.; Kennedy, S.; et al. References Richness of human gut microbiome correlates with metabolic markers. Nature 2013, 500, 541–546. [CrossRef] 58. Chang, C.-C.; Sia, K.-C.; Chang, J.-F.; Lin, C.-M.; Yang, C.-M.; Huang, K.-Y.; Lin, W.-N. Lipopolysaccharide promoted proliferation and adipogenesis of preadipocytes through JAK/STAT and AMPK-regulated cPLA2 expression. Int. J. Med. Sci. 2019, 16, 167–179. [CrossRef] [ ] 59. Amar, J.; Burcelin, R.; Ruidavets, J.B.; Cani, P.D.; Fauvel, J.; Alessi, M.C.; Chamontin, B.; Ferriéres, J. Energy intake is associated with endotoxemia in apparently healthy men. Am. J. Clin. Nutr. 2008, 87, 1219–1223. [CrossRef] 60. Frisardi, V. Impact of Metabolic Syndrome on Cognitive Decline in Older Age: Protective or Harmful, Where is the Pitfall? J. Alzheimers Dis. 2014, 41, 163–167. [CrossRef] 61. Mehta, N.N.; McGillicuddy, F.C.; Anderson, P.D.; Hinkle, C.C.; Shah, R.; Pruscino, L.; Tabita-Martinez, J.; Sellers, K.F.; Rickels, M.R.; Reilly, M.P. Experimental Endotoxemia Induces Adipose Inﬂammation and Insulin Resistance in Humans. Diabetes 2010, 59, 172–181. [CrossRef] 62. Mehta, N.N.; Heffron, S.P.; Patel, P.N.; Ferguson, J.; Shah, R.D.; Hinkle, C.C.; Krishnamoorthy, P.; Shah, R.; Tabita-Martinez, J.; Terembula, K.; et al. A human model of inﬂammatory cardio-metabolic dysfunction; a double blind placebo-controlled crossover trial. J. Transl. Med. 2012, 10, 124. [CrossRef] 63. Hailman, E.; Lichenstein, H.S.; Wurfel, M.M.; Miller, D.S.; Johnson, D.A.; Kelley, M.; Busse, L.A.; Zukowski, M.M.; Wright, S.D. Lipopolysaccharide (LPS)-binding protein accelerates the binding of LPS to CD14. J. Exp. Med. 1994, 179, 269–277. [CrossRef] 64 Ki h R L Th PA M d l ff f CD14 d LBP LPS h ll i i J E d t i R 2005 11 63. Hailman, E.; Lichenstein, H.S.; Wurfel, M.M.; Miller, D.S.; Johnson, D.A.; Kelley, M.; Busse, L.A.; Zukowski, M.M.; Wright, S.D. Lipopolysaccharide (LPS)-binding protein accelerates the binding of LPS to CD14. J. Exp. Med. 1994, 179, 269–277. [CrossRef] 64. Kitchens, R.L.; Thompson, P.A. Modulatory effects of sCD14 and LBP on LPS-host cell interactions. J. Endotoxin Res. 2005, 11, 225–229. [CrossRef] p p y ( ) g p g J p 64. Kitchens, R.L.; Thompson, P.A. Modulatory effects of sCD14 and LBP on LPS-host cell interactions. J. Endotoxin Res. 2005, 11, 225–229. [CrossRef] 65. Ryan, K.A.; Smith, M.F.; Sanders, M.K.; Ernst, P.B. Reactive Oxygen and Nitrogen Species Differentially Regulate Toll-Like Receptor 4-Mediated Activation of NF-κB and Interleukin-8 Expression. Infect. Immun. 2004, 72, 2123–2130. [CrossRef] [PubMed] 66. Lu, Y.-C.; Yeh, W.-C.; Ohashi, P.S. LPS/TLR4 signal transduction pathway. Cytokine 2008, 42, 145–151. [CrossRef] 67. J [ ] 70. Harte, A.L.; da Silva, N.F.; Creely, S.J.; McGee, K.C.; Billyard, T.; Youssef-Elabd, E.M.; Tripathi, G.; Ashour, E.; Abdalla, M.S.; Sharada, H.M.; et al. Elevated endotoxin levels in non-alcoholic fatty liver disease. J. Inﬂamm. 2010, 7, 15. [CrossRef] References [CrossRef] einer, D.E. Chronic Kidney Disease as a Risk Factor for Cardiovascular Disease and All-Cause Mortality: A P ommunity-Based Studies. J. Am. Soc. Nephrol. 2004, 15, 1307–1315. [CrossRef] 41. Longenecker, J.C. Traditional Cardiovascular Disease Risk Factors in Dialysis Patients Compared with the General Population: The Choice Study. J. Am. Soc. Nephrol. 2002, 13, 1918–1927. [CrossRef] y p 42. Lin, C.-J.; Wu, V.; Wu, P.-C.; Wu, C.-J. Meta-Analysis of the Associations of p-Cresyl Sulfate (PCS) and Indoxyl Sulfate (IS) with Cardiovascular Events and All-Cause Mortality in Patients with Chronic Renal Failure. PLoS ONE 2015, 10, e0132589. [CrossRef] 14 of 19 14 of 19 Nutrients 2021, 13, 1389 43. Patel, K.P.; Luo, F.J.-G.; Plummer, N.S.; Hostetter, T.H.; Meyer, T.W. The Production of p-Cresol Sulfate and Indoxyl Sulfate in Vegetarians versus Omnivores. Clin. J. Am. Soc. Nephrol. 2012, 7, 982–988. [CrossRef] g p 44. Garcia-Mantrana, I.; Selma-Royo, M.; Alcantara, C.; Collado, M.C. Shifts on Gut Microbiota Associated to Mediterranean Diet Adherence and Speciﬁc Dietary Intakes on General Adult Population Front Microbiol 2018 9 890 [CrossRef] g p 44. Garcia-Mantrana, I.; Selma-Royo, M.; Alcantara, C.; Collado, M.C. Shifts on Gut Microbiota Associated to Mediterranean Diet Adherence and Speciﬁc Dietary Intakes on General Adult Population. Front. Microbiol. 2018, 9, 890. [CrossRef] h g 44. Garcia-Mantrana, I.; Selma-Royo, M.; Alcantara, C.; Collado, M.C. Shifts on Gut Microbiota Associated to Mediterranean Diet 44. Garcia-Mantrana, I.; Selma-Royo, M.; Alcantara, C.; Collado, M.C. Shifts on Gut Microbiota Associate Adherence and Speciﬁc Dietary Intakes on General Adult Population. Front. Microbiol. 2018, 9, 890. [C 45 Thursby E ; Juge N Introduction to the human gut microbiota Biochem J 2017 474 1823–1836 [Cros y rence and Speciﬁc Dietary Intakes on General Adult Population. Front. Microbiol. 2018, 9, 890. [CrossRef] Adherence and Speciﬁc Dietary Intakes on General Adult Population. Front. Microbiol. 2018, 9, 890. [C 45. Thursby, E.; Juge, N. Introduction to the human gut microbiota. Biochem. J. 2017, 474, 1823–1836. [Cros ce a d Spec c eta y takes o Ge e a du t opu at o ront Microbiol 0 8, 9, 890 [C oss e ] E.; Juge, N. Introduction to the human gut microbiota. Biochem. J. 2017, 474, 1823–1836. [CrossRef] y, ; J g , g J , , [ ] 46. Eckburg, P.B. Diversity of the Human Intestinal Microbial Flora. Science 2005, 308, 1635–1638. [CrossRef] [PubMed] y g g 46. Eckburg, P.B. References Gomes, J.M.G.; de Assis Costa, J.; Alfenas, R.D.C.G. Metabolic endotoxemia and diabetes mellitus: A systematic review. Metabolism 2017, 68, 133–144. [CrossRef] 68. Yang, S.Q.; Lin, H.Z.; Lane, M.D.; Clemens, M.; Diehl, A.M. Obesity increases sensitivity to endotoxin liver injury: Implications for the pathogenesis of steatohepatitis. Proc. Natl. Acad. Sci. USA 1997, 94, 2557–2562. [CrossRef] [PubMed] 69. Thuy, S.; Ladurner, R.; Volynets, V.; Wagner, S.; Strahl, S.; Königsrainer, A.; Maier, K.-P.; Bischoff, S.C.; Bergheim, I. Nonalco- holic Fatty Liver Disease in Humans Is Associated with Increased Plasma Endotoxin and Plasminogen Activator Inhibitor 1 Concentrations and with Fructose Intake. J. Nutr. 2008, 138, 1452–1455. [CrossRef] 70. Harte, A.L.; da Silva, N.F.; Creely, S.J.; McGee, K.C.; Billyard, T.; Youssef-Elabd, E.M.; Tripathi, G.; Ashour, E.; Abdalla, M.S.; Sharada, H.M.; et al. Elevated endotoxin levels in non-alcoholic fatty liver disease. J. Inﬂamm. 2010, 7, 15. [CrossRef] 15 of 19 15 of 19 Nutrients 2021, 13, 1389 71. Sharifnia, T.; Antoun, J.; Verriere, T.G.C.; Suarez, G.; Wattacheril, J.; Wilson, K.T.; Peek, R.M.; Abumrad, N.N.; Flynn, C.R. Hepatic TLR4 signaling in obese NAFLD. Am. J. Physiol. Liver Physiol. 2015, 309, G270–G278. [CrossRef] 72. de Aguiar Vallim, T.Q.; Tarling, E.J.; Edwards, P.A. Pleiotropic Roles of Bile Acids in Metabolism. Cell Metab. 2013, 17, 657–669. [CrossRef] [ ] 73. Kitahara, M.; Takamine, F.; Imamura, T.; Benno, Y. Clostridium hiranonis sp. nov., a human intestinal bacterium with bile acid 7alpha-dehydroxylating activity. Int. J. Syst. Evol. Microbiol. 2001, 51, 39–44. [CrossRef] [PubMed] 74. Hylemon, P.B.; Harris, S.C.; Ridlon, J.M. Metabolism of hydrogen gases and bile acids in the gut microbiome. FEBS Lett. 2018, 592, 2070–2082. [CrossRef] [PubMed] 75. Le Roy, T.; Llopis, M.; Lepage, P.; Bruneau, A.; Rabot, S.; Bevilacqua, C.; Martin, P.; Philippe, C.; Walker, F.; Bado, A.; et al. Intestinal microbiota determines development of non-alcoholic fatty liver disease in mice. Gut 2013, 62, 1787–1794. [CrossRef] [PubMed] 76. Yuan, J.; Chen, C.; Cui, J.; Lu, J.; Yan, C.; Wei, X.; Zhao, X.; Li, N.; Li, S.; Xue, G.; et al. Fatty Liver Disease Caused by High-Alcohol-Producing Klebsiella pneumoniae. Cell Metab. 2019, 30, 675–688.e7. [CrossRef] [PubMed] g g p 77. Stremmel, W.; Schmidt, K.V.; Schuhmann, V.; Kratzer, F.; Garbade, S.F.; Langhans, C.-D.; Fricker, G.; Okun, J.G. Blood Trimethylamine-N-Oxide Originates from Microbiota Mediated Breakdown of Phosphatidylcholine and Absorption from Small Intestine. PLoS ONE 2017, 12, e0170742. [CrossRef] 78. Velasquez, M.; Ramezani, A.; Manal, A.; Raj, D. References Trimethylamine N-Oxide: The Good, the Bad and the Unknown. Toxins 2016, 8, 326. [CrossRef] [PubMed] 79. Vogt, N.M.; Romano, K.A.; Darst, B.F.; Engelman, C.D.; Johnson, S.C.; Carlsson, C.M.; Asthana, S.; Blennow, K.; Zetterberg, H.; Bendlin, B.B.; et al. The gut microbiota-derived metabolite trimethylamine N-oxide is elevated in Alzheimer’s disease. Alzheimers Res. Ther. 2018, 10, 124. [CrossRef] 80. Kanitsoraphan, C.; Rattanawong, P.; Charoensri, S.; Senthong, V. Trimethylamine N-Oxide and Risk of Cardiovascular Disease and Mortality. Curr. Nutr. Rep. 2018, 7, 207–213. [CrossRef] [PubMed] 81. Schugar, R.C.; Shih, D.M.; Warrier, M.; Helsley, R.N.; Burrows, A.; Ferguson, D.; Brown, A.L.; Gromovsky, A.D.; Heine, M.; Chatterjee, A.; et al. The TMAO-Producing Enzyme Flavin-Containing Monooxygenase 3 Regulates Obesity and the Beiging of White Adipose Tissue. Cell Rep. 2017, 19, 2451–2461. [CrossRef] 82. Tang, W.H.W.; Wang, Z.; Kennedy, D.J.; Wu, Y.; Buffa, J.A.; Agatisa-Boyle, B.; Li, X.S.; Levison, B.S.; Hazen, S.L. Gut Microbiota- Dependent Trimethylamine N -Oxide (TMAO) Pathway Contributes to Both Development of Renal Insufﬁciency and Mortality Risk in Chronic Kidney Disease. Circ. Res. 2015, 116, 448–455. [CrossRef] y 83. Koeth, R.A.; Wang, Z.; Levison, B.S.; Buffa, J.A.; Org, E.; Sheehy, B.T.; Britt, E.B.; Fu, X.; Wu, Y.; Li, L.; et al. Intestinal microbiota metabolism of l-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat. Med. 2013, 19, 576–585. [CrossRef] [PubMed] 84. Zhang, A.; Mitchell, S.; Smith, R. Dietary Precursors of Trimethylamine in Man: A Pilot Study. Food Chem. Toxicol. 1999, 37, 515–520. [CrossRef] 85. Romano, K.A.; Vivas, E.I.; Amador-Noguez, D.; Rey, F.E. Intestinal Microbiota Composition Modulates Choline Bioavailability from Diet and Accumulation of the Proatherogenic Metabolite Trimethylamine-N-Oxide. mBio 2015, 6, e02481. [CrossRef] 86. Kzhyshkowska, J.; Neyen, C.; Gordon, S. Role of macrophage scavenger receptors in atherosclerosis. Immunobiology 2012, 217, 492–502. [CrossRef] [PubMed] 85. Romano, K.A.; Vivas, E.I.; Amador-Noguez, D.; Rey, F.E. Intestinal Microbiota Composition Modulates Choline Bioavailability from Diet and Accumulation of the Proatherogenic Metabolite Trimethylamine-N-Oxide. mBio 2015, 6, e02481. [CrossRef] 86 K h hk k J N C G d S R l f h t i th l i I bi l 2012 217 g y 86. Kzhyshkowska, J.; Neyen, C.; Gordon, S. Role of macrophage scavenger receptors in atherosclerosis. Immunobiology 2012, 217, 492–502. [CrossRef] [PubMed] 87. Chen, M.; Zhu, X.; Ran, L.; Lang, H.; Yi, L.; Mi, M. Trimethylamine-N-Oxide Induces Vascular Inﬂamm NLRP3 Inﬂammasome Through the SIRT3-SOD2-mtROS Signaling Pathway. J. Am. Heart Assoc. 95. Conlon, M.; Bird, A. The Impact of Diet and Lifestyle on Gut Microbiota and Human Health. Nutrients 2014, 7, 17–44. [CrossRef] [PubMed] References Rethink Microbe 2014, 16, 433–437. [CrossRef] [PubMed] g , ; , ; y, ; , ; , ; , ; g , g yp Microbe 2014, 16, 433–437. [CrossRef] [PubMed] 98. Cheng, M.; Ning, K. Stereotypes about Enterotype: The Old and New Ideas. Genomics. Proteom. Bioinform. 2019, 17, 4–12. [C f] [ b d] Microbe 2014, 16, 433–437. [CrossRef] [PubMed] 98. Cheng, M.; Ning, K. Stereotypes about Enterotype: The Old and New Ideas. Genomics. Proteom. Bioinform. 2019, 17, 4–12. [CrossRef] [PubMed] 98. Cheng, M.; Ning, K. Stereotypes about Enterotype: The Old and New Ideas. Genomics. Proteom [CrossRef] [PubMed] 99. Klein, S.; Burke, L.E.; Bray, G.A.; Blair, S.; Allison, D.B.; Pi-Sunyer, X.; Hong, Y.; Eckel, R.H. Clinical Implications of Obesity With Speciﬁc Focus on Cardiovascular Disease. Circulation 2004, 110, 2952–2967. [CrossRef] [PubMed] Grundy, S.M.; Cleeman, J.I.; Daniels, S.R.; Donato, K.A.; Eckel, R.H.; Franklin, B.A.; Gordon, D.J.; Krau 100. Grundy, S.M.; Cleeman, J.I.; Daniels, S.R.; Donato, K.A.; Eckel, R.H.; Franklin, B.A.; Gordon, D.J.; Krauss, R.M.; Savage, P.J.; Smith, S.C.; et al. Diagnosis and Management of the Metabolic Syndrome. Circulation 2005, 112, 2735–2752. [CrossRef] [PubMed] 100. Grundy, S.M.; Cleeman, J.I.; Daniels, S.R.; Donato, K.A.; Eckel, R.H.; Franklin, B.A.; Gordon, D.J.; Krauss, R.M.; Savage, P.J.; Smith S C ; et al Diagnosis and Management of the Metabolic Syndrome Circulation 2005 112 2735 2752 [CrossRef] [PubMed] 101. Ridaura, V.K.; Faith, J.J.; Rey, F.E.; Cheng, J.; Duncan, A.E.; Kau, A.L.; Grifﬁn, N.W.; Lombard, V.; Henrissat, B.; Bain, J.R.; et al. Gut Microbiota from Twins Discordant for Obesity Modulate Metabolism in Mice. Science 2013, 341, 1241214. [CrossRef] y g Gut Microbiota from Twins Discordant for Obesity Modulate Metabolism in Mice. Science 2013, 341, 1241214. [CrossRef] 103. Vrieze, A.; Van Nood, E.; Holleman, F.; Salojärvi, J.; Kootte, R.S.; Bartelsman, J.F.W.M.; Dallinga–Thie, G.M.; Ackermans, M.T.; Serlie, M.J.; Oozeer, R.; et al. Transfer of Intestinal Microbiota from Lean Donors Increases Insulin Sensitivity in Individuals With Metabolic Syndrome. Gastroenterology 2012, 143, 913–916.e7. [CrossRef] y gy 104. Turnbaugh, P.J.; Ley, R.E.; Mahowald, M.A.; Magrini, V.; Mardis, E.R.; Gordon, J.I. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 2006, 444, 1027–1031. [CrossRef] 105. Turnbaugh, P.J.; Hamady, M.; Yatsunenko, T.; Cantarel, B.L.; Duncan, A.; Ley, R.E.; Sogin, M.L.; Jones, W.J.; Roe, B.A.; Affourtit, J.P.; et al. A core gut microbiome in obese and lean twins. Nature 2009, 457, 480–484. [CrossRef] [PubMed] 106. References Santacruz, A.; Collado, M.C.; García-Valdés, L.; Segura, M.T.; Martín-Lagos, J.A.; Anjos, T.; Martí-Romero, M.; Lopez, R.M.; Florido, J.; Campoy, C.; et al. Gut microbiota composition is associated with body weight, weight gain and biochemical parameters in pregnant women. Br. J. Nutr. 2010, 104, 83–92. [CrossRef] [PubMed] p g 107. Ley, R.E.; Turnbaugh, P.J.; Klein, S.; Gordon, J.I. Human gut microbes associated with obesity. Nature 2006, 444, 1022–1023. [CrossRef] 108. Schwiertz, A.; Taras, D.; Schäfer, K.; Beijer, S.; Bos, N.A.; Donus, C.; Hardt, P.D. Microbiota and SCFA in Lean and Overweight Healthy Subjects. Obesity 2010, 18, 190–195. [CrossRef] [PubMed] y j y 109. Collado, M.C.; Isolauri, E.; Laitinen, K.; Salminen, S. Distinct composition of gut microbiota during pregnancy in overweight and normal-weight women. Am. J. Clin. Nutr. 2008, 88, 894–899. [CrossRef] [PubMed] 110. Mai, V.; McCrary, Q.M.; Sinha, R.; Glei, M. Associations between dietary habits and body mass index with gut microbiota composition and fecal water genotoxicity: An observational study in African American and Caucasian American volunteers. Nutr. J. 2009, 8, 49. [CrossRef] 111. Kopp, W. How Western Diet and Lifestyle Drive The Pandemic Of Obesity And Civilization Diseases. Diabetes Metab. Syndr. Obes. Targets Ther. 2019, 12, 2221–2236. [CrossRef] g 112. Cordain, L.; Eaton, S.B.; Sebastian, A.; Mann, N.; Lindeberg, S.; Watkins, B.A.; O’Keefe, J.H.; Brand-Miller, J. Origins and evolution of the Western diet: Health implications for the 21st century. Am. J. Clin. Nutr. 2005, 81, 341–354. [CrossRef] p y ontes-Villalba, M.; O’Keefe, J.H.; Lindeberg, S.; Cordain, L. The western diet and lifestyle and diseases of orts Clin. Cardiol. 2011, 15. [CrossRef] 113. Carrera-Bastos, P.; Fontes-Villalba, M.; O’Keefe, J.H.; Lindeberg, S.; Cordain, L. The western diet civilization. Res. Reports Clin. Cardiol. 2011, 15. [CrossRef] 114. Moszak, M.; Szuli´nska, M.; Bogda´nski, P. You Are What You Eat—The Relationship between Diet, Microbiota, and Metabolic Disorders—A Review. Nutrients 2020, 12, 1096. [CrossRef] [ ] 115. Bagga, D.; Wang, L.; Farias-Eisner, R.; Glaspy, J.A.; Reddy, S.T. Differential effects of prostaglandin derived from -6 and -3 polyunsaturated fatty acids on COX-2 expression and IL-6 secretion. Proc. Natl. Acad. Sci. USA 2003, 100, 1751–1756. [CrossRef] 115. Bagga, D.; Wang, L.; Farias-Eisner, R.; Glaspy, J.A.; Reddy, S.T. Differential effects of prostaglandin derived from -6 and -3 polyunsaturated fatty acids on COX-2 expression and IL-6 secretion. Proc. Natl. Acad. Sci. USA 2003, 100, 1751–1756. [CrossRef] 116. Wang, L.; Christophersen, C.T.; Sorich, M.J.; Gerber, J.P.; Angley, M.T.; Conlon, M.A. 121. Parolini, C. Effects of Fish n-3 PUFAs on Intestinal Microbiota and Immune System. Mar. Drugs 2019, 17, 374. [CrossRef] [PubMed] References 2017, Ran, L.; Lang, H.; Yi, L.; Mi, M. Trimethylamine-N-Oxide Induces Vascular Inﬂammation by Activating the me Through the SIRT3-SOD2-mtROS Signaling Pathway. J. Am. Heart Assoc. 2017, 6. [CrossRef] M.; Zhu, X.; Ran, L.; Lang, H.; Yi, L.; Mi, M. Trimethylamine-N-Oxide Induces Vascular Inﬂammation by 3 Inﬂammasome Through the SIRT3-SOD2-mtROS Signaling Pathway. J. Am. Heart Assoc. 2017, 6. [CrossR 88. Minett, T.; Classey, J.; Matthews, F.E.; Fahrenhold, M.; Taga, M.; Brayne, C.; Ince, P.G.; Nicoll, J.A.R.; Boche, D. Microglial immunophenotype in dementia with Alzheimer’s pathology. J. Neuroinﬂamm. 2016, 13, 135. [CrossRef] p yp p gy ﬂ 89. Mallmann, N.H.; Lima, E.S.; Lalwani, P. Dysregulation of Tryptophan Catabolism in Metabolic Syndrome. Metab. Syndr. Relat. Disord. 2018, 16, 135–142. [CrossRef] 90. Chimerel, C.; Emery, E.; Summers, D.K.; Keyser, U.; Gribble, F.M.; Reimann, F. Bacterial Metabolite Indole Modulates Incretin Secretion from Intestinal Enteroendocrine L Cells. Cell Rep. 2014, 9, 1202–1208. [CrossRef] 91. Liabeuf, S.; Barreto, D.V.; Barreto, F.C.; Meert, N.; Glorieux, G.; Schepers, E.; Temmar, M.; Choukroun, G.; Vanholder, R.; Massy, Z.A. Free p-cresylsulphate is a predictor of mortality in patients at different stages of chronic kidney disease. Nephrol. Dial. Transplant. 2010, 25, 1183–1191. [CrossRef] [PubMed] p 92. David, L.A.; Maurice, C.F.; Carmody, R.N.; Gootenberg, D.B.; Button, J.E.; Wolfe, B.E.; Ling, A.V.; Devlin, A.S.; Varma, Y.; Fischbach, M.A.; et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature 2014, 505, 559–563. [CrossRef] 93. do Rosario, V.A.; Fernandes, R.; Trindade, E.B.S. de M. Vegetarian diets and gut microbiota: Important shifts in markers of metabolism and cardiovascular disease. Nutr. Rev. 2016, 74, 444–454. [CrossRef] 94. Wu, G.D.; Compher, C.; Chen, E.Z.; Smith, S.A.; Shah, R.D.; Bittinger, K.; Chehoud, C.; Albenberg, L.G.; Nessel, L.; Gilroy, E.; et al. Comparative metabolomics in vegans and omnivores reveal constraints on diet-dependent gut microbiota metabolite production. Gut 2016, 65, 63–72. [CrossRef] [PubMed] 95. Conlon, M.; Bird, A. The Impact of Diet and Lifestyle on Gut Microbiota and Human Health. Nutrients 2014, 7, 17–44. [CrossRef] [PubMed] 16 of 19 16 of 19 Nutrients 2021, 13, 1389 96. Fava, F.; Gitau, R.; Grifﬁn, B.A.; Gibson, G.R.; Tuohy, K.M.; Lovegrove, J.A. The type and quantity of dietary fat and carbohydrate alter faecal microbiome and short-chain fatty acid excretion in a metabolic syndrome ‘at-risk’ population. Int. J. Obes. 2013, 37, 216–223. [CrossRef] [PubMed] 97. Knights, D.; Ward, T.L.; McKinlay, C.E.; Miller, H.; Gonzalez, A.; McDonald, D.; Knight, R. References Increased abundance of Sutterella spp. and l 116. Wang, L.; Christophersen, C.T.; Sorich, M.J.; Gerber, J.P.; Angley, M.T.; Conlon, M.A. Increased abundance of Sutterella spp. and Ruminococcus torques in feces of children with autism spectrum disorder. Mol. Autism 2013, 4, 42. [CrossRef] [PubMed] ues in feces of children with autism spectrum disorder. Mol. Autism 2013, 4, 42. [CrossRef] [PubMed] q p 117. Rajkumar, H.; Mahmood, N.; Kumar, M.; Varikuti, S.R.; Challa, H.R.; Myakala, S.P. Effect of Probiotic (VSL#3) and Omega-3 on Lipid Proﬁle, Insulin Sensitivity, Inﬂammatory Markers, and Gut Colonization in Overweight Adults: A Randomized, Controlled Trial. Mediat. Inﬂamm. 2014, 2014, 1–8. [CrossRef] ﬂ 118. Watson, H.; Mitra, S.; Croden, F.C.; Taylor, M.; Wood, H.M.; Perry, S.L.; Spencer, J.A.; Quirke, P.; Toogood, G.J.; Lawton, C.L.; et al. A randomised trial of the effect of omega-3 polyunsaturated fatty acid supplements on the human intestinal microbiota. Gut 2018, 67, 1974–1983. [CrossRef] [PubMed] 119. Mokkala, K.; Röytiö, H.; Munukka, E.; Pietilä, S.; Ekblad, U.; Rönnemaa, T.; Eerola, E.; Laiho, A.; Laitinen, K. Gut Microbiota Richness and Composition and Dietary Intake of Overweight Pregnant Women Are Related to Serum Zonulin Concentration, a Marker for Intestinal Permeability. J. Nutr. 2016, 146, 1694–1700. [CrossRef] 120. Patterson, E.; O’Doherty, R.M.; Murphy, E.F.; Wall, R.; O’Sullivan, O.; Nilaweera, K.; Fitzgerald, G.F.; Cotter, P.D.; Ross, R.P.; Stanton, C. Impact of dietary fatty acids on metabolic activity and host intestinal microbiota composition in C57BL/6J mice. Br. J. Nutr. 2014, 111, 1905–1917. [CrossRef] 121. Parolini, C. Effects of Fish n-3 PUFAs on Intestinal Microbiota and Immune System. Mar. Drugs 2019, 17, 374. [CrossRef] [PubMed] 17 of 19 17 of 19 Nutrients 2021, 13, 1389 122. Wu, G.D.; Chen, J.; Hoffmann, C.; Bittinger, K.; Chen, Y.-Y.; Keilbaugh, S.A.; Bewtra, M.; Knights, D.; Walters, W.A.; Knight, R.; et al. Linking Long-Term Dietary Patterns with Gut Microbial Enterotypes. Science 2011, 334, 105–108. [CrossRef] g g y yp 123. Noriega, B.S.; Sanchez-Gonzalez, M.A.; Salyakina, D.; Coffman, J. Understanding the Impact of Omega-3 Rich Diet on the Gut Microbiota. Case Rep. Med. 2016, 2016, 1–6. [CrossRef] 124. Pu, S.; Khazanehei, H.; Jones, P.J.; Khaﬁpour, E. Interactions between Obesity Status and Dietary Intake of Monounsaturated and Polyunsaturated Oils on Human Gut Microbiome Proﬁles in the Canola Oil Multicenter Intervention Trial (COMIT). Front. Microbiol. 2016, 7. [CrossRef] 125. Menni, C.; Zierer, J.; Pallister, T.; Jackson, M.A.; Long, T.; Mohney, R.P.; Steves, C.J.; Spector, T.D.; Valdes, A.M. References Omega-3 fatty acids correlate with gut microbiome diversity and production of N-carbamylglutamate in middle aged and elderly women. Sci. Rep. 2017, 7, 11079. [CrossRef] unsaturated fatty acids (n-3 PUFAs) and inﬂammatory bowel disease (IBD): Pathogenesis and treatment. acol. Sci. 2004, 8, 225–229. 126. Belluzzi, A. Polyunsaturated fatty acids (n-3 PUFAs) and inﬂammatory bowel disease (IBD): Path Rev. Med. Pharmacol. Sci. 2004, 8, 225–229. 127. Santoru, M.L.; Piras, C.; Murgia, A.; Palmas, V.; Camboni, T.; Liggi, S.; Ibba, I.; Lai, M.A.; Orrù, S.; Blois, S.; et al. Cross sectional evaluation of the gut-microbiome metabolome axis in an Italian cohort of IBD patients. Sci. Rep. 2017, 7, 9523. [CrossRef] 128. Perez-Martinez, P.; Garcia-Rios, A.; Delgado-Lista, J.; Perez-Jimenez, F.; Lopez-Miranda, J. Mediterranean Diet Rich in Olive Oil and Obesity, Metabolic Syndrome and Diabetes Mellitus. Curr. Pharm. Des. 2011, 17, 769–777. [CrossRef] 129. Grosso, G.; Mistretta, A.; Frigiola, A.; Gruttadauria, S.; Biondi, A.; Basile, F.; Vitaglione, P.; D’Orazio, N.; Galvano, F. Mediterranean Diet and Cardiovascular Risk Factors: A Systematic Review. Crit. Rev. Food Sci. Nutr. 2014, 54, 593–610. [CrossRef] 130. Abete, I.; Goyenechea, E.; Zulet, M.A.; Martínez, J.A. Obesity and metabolic syndrome: Potential beneﬁt from speciﬁc nutritional components. Nutr. Metab. Cardiovasc. Dis. 2011, 21, B1–B15. [CrossRef] p 131. Grosso, G.; Marventano, S.; Galvano, F.; Pajak, A.; Mistretta, A. Factors Associated With Metabolic Synd Population: Role of Caffeinated Beverages. J. Epidemiol. 2014, 24, 327–333. [CrossRef] p g p 132. Basu, A.; Fu, D.X.; Wilkinson, M.; Simmons, B.; Wu, M.; Betts, N.M.; Du, M.; Lyons, T.J. Strawberries decrease atherosclerotic markers in subjects with metabolic syndrome. Nutr. Res. 2010, 30, 462–469. [CrossRef] j y 133. De Filippis, F.; Pellegrini, N.; Vannini, L.; Jeffery, I.B.; La Storia, A.; Laghi, L.; Serrazanetti, D.I.; Di Cagno, R.; Ferrocino, I.; Lazzi, C.; et al. High-level adherence to a Mediterranean diet beneﬁcially impacts the gut microbiota and associated metabolome. Gut 2016, 65, 1812–1821. [CrossRef] [PubMed] Marrone, G.; Cintoni, M.; Tarsitano, M.G.; Capacci, A.; De Lorenzo, A. Inﬂuence of Mediterranean Diet on iota. Nutrients 2020, 13, 7. [CrossRef] [PubMed] 134. Merra, G.; Noce, A.; Marrone, G.; Cintoni, M.; Tarsitano, M.G.; Capacci, A.; De Lorenzo, A. Inﬂuen Human Gut Microbiota. Nutrients 2020, 13, 7. [CrossRef] [PubMed] 135. De Angelis, M.; Garruti, G.; Minervini, F.; Bonfrate, L.; Portincasa, P.; Gobbetti, M. The Food-gut Human Axis: The Effects of Diet on Gut Microbiota and Metabolome. Curr. Med. Chem. 2019, 26, 3567–3583. [CrossRef] 136. References Haro, C.; García-Carpintero, S.; Rangel-Zúñiga, O.A.; Alcalá-Díaz, J.F.; Landa, B.B.; Clemente, J.C.; Pérez-Martínez, P.; López- Miranda, J.; Pérez-Jiménez, F.; Camargo, A. Consumption of Two Healthy Dietary Patterns Restored Microbiota Dysbiosis in Obese Patients with Metabolic Dysfunction. Mol. Nutr. Food Res. 2017, 61, 1700300. [CrossRef] [PubMed] y 137. Gershuni, V.M.; Yan, S.L.; Medici, V. Nutritional Ketosis for Weight Management and Reversal of Metabolic Syndrome. Curr. Nutr. Rep. 2018, 7, 97–106. [CrossRef] 138. Kosinski, C.; Jornayvaz, F. Effects of Ketogenic Diets on Cardiovascular Risk Factors: Evidence from Animal and Human Studies. Nutrients 2017, 9, 517. [CrossRef] 139. Paoli, A.; Mancin, L.; Bianco, A.; Thomas, E.; Mota, J.F.; Piccini, F. Ketogenic Diet and Microbiota: Friends or Enemies? Genes 2019, 10, 534. [CrossRef] 140. Gentile, C.L.; Weir, T.L. The gut microbiota at the intersection of diet and human health. Science 2018, 362, 776–780. [CrossRef] 141. Altuntas, Y. Microbiota and Metabolic Syndrome. Turk Kardiyol. Dern. Ars. 2017, 45, 286–296. [CrossRef] 140. Gentile, C.L.; Weir, T.L. The gut microbiota at the intersection of diet and human health. Science 2018, 362, 776 780. [CrossRef] 141. Altuntas, Y. Microbiota and Metabolic Syndrome. Turk Kardiyol. Dern. Ars. 2017, 45, 286–296. [CrossRef] Microbiota and Metabolic Syndrome. Turk Kardiyol. D 142. Rizzo, N.S.; Sabate, J.; Jaceldo-Siegl, K.; Fraser, G.E. Vegetarian Dietary Patterns Are Associated Wi Syndrome: The Adventist Health Study 2. Diabetes Care 2011, 34, 1225–1227. [CrossRef] 142. Rizzo, N.S.; Sabate, J.; Jaceldo-Siegl, K.; Fraser, G.E. Vegetarian Dietary Patterns Are Associated With a Lower Risk of Metabolic Syndrome: The Adventist Health Study 2. Diabetes Care 2011, 34, 1225–1227. [CrossRef] 143 K hl H L i S B d N D V t i Di t P tt d C di l Di P C di Di 2018 61 142. Rizzo, N.S.; Sabate, J.; Jaceldo Siegl, K.; Fraser, G.E. Vegetarian Dietary Patterns Are Associated With a Lower Risk of Metabolic Syndrome: The Adventist Health Study 2. Diabetes Care 2011, 34, 1225–1227. [CrossRef] 143. Kahleova, H.; Levin, S.; Barnard, N.D. Vegetarian Dietary Patterns and Cardiovascular Disease. Prog. Cardiovasc. Dis. 2018, 61, 54 61 [C R f] [P bM d] ova, H.; Levin, S.; Barnard, N.D. Vegetarian Dietary Patterns and Cardiovascular Disease. Prog. Cardiovasc . [CrossRef] [PubMed] 144. Zhao, J.; Zhang, X.; Liu, H.; Brown, M.A.; Qiao, S. Dietary Protein and Gut Microbiota Composition and Function. Curr. Protein Pept. Sci. 2018, 20, 145–154. [CrossRef] p 145. 148. Gibson, G.R.; Roberfroid, M.B. Dietary Modulation of the Human Colonic Microbiota: Introducing the Concept of Prebiotics. J. Nutr. 1995, 125, 1401–1412. [CrossRef] References Gibson, G.R.; Probert, H.M.; Van Loo, J.; Rastall, R.A.; Roberfroid, M.B. Dietary modulation of the human colonic microbiota: Updating the concept of prebiotics. Nutr. Res. Rev. 2004, 17, 259–275. [CrossRef] [PubMed] 146. Nicolucci, A.C.; Hume, M.P.; Martínez, I.; Mayengbam, S.; Walter, J.; Reimer, R.A. Prebiotics Reduce Body Fat and Alter Intestinal Microbiota in Children Who Are Overweight or With Obesity. Gastroenterology 2017, 153, 711–722. [CrossRef] [PubMed] g y gy 147. So, D.; Whelan, K.; Rossi, M.; Morrison, M.; Holtmann, G.; Kelly, J.T.; Shanahan, E.R.; Staudacher, H.M.; Campbell, K.L. Dietary ﬁber intervention on gut microbiota composition in healthy adults: A systematic review and meta-analysis. Am. J. Clin. Nutr. 2018, 107, 965–983. [CrossRef] 148. Gibson, G.R.; Roberfroid, M.B. Dietary Modulation of the Human Colonic Microbiota: Introducing the Concept of Prebiotics. J. Nutr. 1995, 125, 1401–1412. [CrossRef] 18 of 19 18 of 19 Nutrients 2021, 13, 1389 149. Gibson, G.R.; Hutkins, R.; Sanders, M.E.; Prescott, S.L.; Reimer, R.A.; Salminen, S.J.; Scott, K.; Stanton, C.; Swanson, K.S.; Cani, P.D.; et al. Expert consensus document: The International Scientiﬁc Association for Probiotics and Prebiotics (ISAPP) consensus statement on the deﬁnition and scope of prebiotics. Nat. Rev. Gastroenterol. Hepatol. 2017, 14, 491–502. [CrossRef] p p p 150. Fijan, S. Microorganisms with Claimed Probiotic Properties: An Overview of Recent Literature. Int. 2014, 11, 4745–4767. [CrossRef] [ ] 151. Bellikci-Koyu, E.; Sarer-Yurekli, B.P.; Akyon, Y.; Aydin-Kose, F.; Karagozlu, C.; Ozgen, A.G.; Brinkmann, A.; Nitsche, A.; Ergunay, K.; Yilmaz, E.; et al. Effects of Regular Keﬁr Consumption on Gut Microbiota in Patients with Metabolic Syndrome: A Parallel-Group, Randomized, Controlled Study. Nutrients 2019, 11, 2089. [CrossRef] 152. Companys, J.; Pedret, A.; Valls, R.M.; Solà, R.; Pascual, V. Fermented dairy foods rich in probiotics and cardiometabolic risk factors: A narrative review from prospective cohort studies. Crit. Rev. Food Sci. Nutr. 2020, 1–10. [CrossRef] [PubMed] ´ 153. Markowiak, P.; ´Slizewska, K. Effects of probiotics, prebiotics, and synbiotics on human health. Nutri É 153. Markowiak, P.; Slizewska, K. Effects of probiotics, prebiotics, and synbiotics on human health. Nutrients 2017, 9, 1021. [CrossRef] 154. Le Barz, M.; Daniel, N.; Varin, T.V.; Naimi, S.; Demers-Mathieu, V.; Pilon, G.; Audy, J.; Laurin, É.; Roy, D.; Urdaci, M.C.; et al. In vivo screening of multiple bacterial strains identiﬁes Lactobacillus rhamnosus Lb102 and Biﬁdobacterium animalis ssp. lactis Bf141 as probiotics that improve metabolic disorders in a mouse model of obesity. FASEB J. 2019, 33, 4921–4935. [CrossRef] 155. References Chan, Y.K.; Brar, M.S.; Kirjavainen, P.V.; Chen, Y.; Peng, J.; Li, D.; Leung, F.C.-C.; El-Nezami, H. High fat diet induced atherosclerosis is accompanied with low colonic bacterial diversity and altered abundances that correlates with plaque size, plasma A-FABP and cholesterol: A pilot study of high fat diet and its intervention with Lactobacillus rhamno. BMC Microbiol. 2016, 16, 264. [CrossRef] 156. Tenorio-Jiménez, C.; Martínez-Ramírez, M.J.; Del Castillo-Codes, I.; Arraiza-Irigoyen, C.; Tercero-Lozano, M.; Camacho, J.; Chueca, N.; García, F.; Olza, J.; Plaza-Díaz, J.; et al. Lactobacillus reuteri V3401 Reduces Inﬂammatory Biomarkers and Modiﬁes the Gastrointestinal Microbiome in Adults with Metabolic Syndrome: The PROSIR Study. Nutrients 2019, 11, 1761. [CrossRef] [PubMed] 157. Bernini, L.J.; Simão, A.N.C.; Alﬁeri, D.F.; Lozovoy, M.A.B.; Mari, N.L.; de Souza, C.H.B.; Dichi, I.; Costa, G.N. Beneﬁcial effects of Biﬁdobacterium lactis on lipid proﬁle and cytokines in patients with metabolic syndrome: A randomized trial. Effects of probiotics on metabolic syndrome. Nutrition 2016, 32, 716–719. [CrossRef] [PubMed] 158. Hiippala, K.; Jouhten, H.; Ronkainen, A.; Hartikainen, A.; Kainulainen, V.; Jalanka, J.; Satokari, R. The Potential of Gut Commensals in Reinforcing Intestinal Barrier Function and Alleviating Inﬂammation. Nutrients 2018, 10, 988. [CrossRef] [PubMed] 159. Morrison, D.J.; Preston, T. Formation of short chain fatty acids by the gut microbiota and their impact on human metabolism. Gut Microbes 2016, 7, 189–200. [CrossRef] 160. Zhang, X.; Shen, D.; Fang, Z.; Jie, Z.; Qiu, X.; Zhang, C.; Chen, Y.; Ji, L. Human Gut Microbiota Changes Reveal the Progression of Glucose Intolerance. PLoS ONE 2013, 8, e71108. [CrossRef] [PubMed] 161. Dao, M.C.; Everard, A.; Aron-Wisnewsky, J.; Sokolovska, N.; Prifti, E.; Verger, E.O.; Kayser, B.D.; Levenez, F.; Chilloux, J.; Hoyles, L.; et al. Akkermansia muciniphila and improved metabolic health during a dietary intervention in obesity: Relationship with gut microbiome richness and ecology. Gut 2016, 65, 426–436. [CrossRef] g gy 162. Yassour, M.; Lim, M.Y.; Yun, H.S.; Tickle, T.L.; Sung, J.; Song, Y.-M.; Lee, K.; Franzosa, E.A.; Morgan, X.C.; Gevers, D.; et al. Sub-clinical detection of gut microbial biomarkers of obesity and type 2 diabetes. Genome Med. 2016, 8, 17. [CrossRef] 162. Yassour, M.; Lim, M.Y.; Yun, H.S.; Tickle, T.L.; Sung, J.; Song, Y.-M.; Lee, K.; Franzosa, E.A.; Morgan, X.C.; Gevers, D.; et al. Sub-clinical detection of gut microbial biomarkers of obesity and type 2 diabetes. Genome Med. 2016, 8, 17. [CrossRef] 163. 174. Mullish, B.H.; Quraishi, M.N.; Segal, J.P.; McCune, V.L.; Baxter, M.; Marsden, G.L.; Moore, D.J.; Colville, A.; Bhala, N.; Iqbal, T.H.; et al. The use of faecal microbiota transplant as treatment for recurrent or refractory Clostridium difﬁcile infection and other potential indications: Joint British Society of Gastroenterology (BSG) and Healthcare Infection Society (HIS) guidelines. Gut 2018, 67, 1920–1941. [CrossRef] [PubMed] 173. Kassam, Z.; Lee, C.H.; Yuan, Y.; Hunt, R.H. Fecal Microbiota Transplantation for Clostridium difﬁcile Infection: Systematic Review and Meta-Analysis. Am. J. Gastroenterol. 2013, 108, 500–508. [CrossRef] [PubMed] p [ ] 176. Smits, L.P.; Kootte, R.S.; Levin, E.; Prodan, A.; Fuentes, S.; Zoetendal, E.G.; Wang, Z.; Levison, B.S.; Cleophas, M.C.P.; Kemper, E.M.; et al. Effect of Vegan Fecal Microbiota Transplantation on Carnitine- and Choline-Derived Trimethylamine- N-Oxide Production and Vascular Inﬂammation in Patients With Metabolic Syndrome. J. Am. Heart Assoc. 2018, 7. [CrossRef] 175. Kootte, R.S.; Levin, E.; Salojärvi, J.; Smits, L.P.; Hartstra, A.V.; Udayappan, S.D.; Hermes, G.; Bouter, K.E.; Koopen, A.M.; Holst, J.J.; et al. Improvement of Insulin Sensitivity after Lean Donor Feces in Metabolic Syndrome Is Driven by Baseline Intestinal Microbiota Composition. Cell Metab. 2017, 26, 611.e6–619.e6. [CrossRef] References Grander, C.; Adolph, T.E.; Wieser, V.; Lowe, P.; Wrzosek, L.; Gyongyosi, B.; Ward, D.V.; Grabherr, F.; Gerner, R.R.; Pﬁster, A.; et al. Recovery of ethanol-induced Akkermansia muciniphila depletion ameliorates alcoholic liver disease. Gut 2018, 67, 891–901. [CrossRef] 164. Li, J.; Zhao, F.; Wang, Y.; Chen, J.; Tao, J.; Tian, G.; Wu, S.; Liu, W.; Cui, Q.; Geng, B.; et al. Gut microbiota dysbiosis contributes to the development of hypertension. Microbiome 2017, 5, 14. [CrossRef] 165. Cani, P.D.; de Vos, W.M. Next-Generation Beneﬁcial Microbes: The Case of Akkermansia muciniphila [CrossRef] [PubMed] 166. Depommier, C.; Everard, A.; Druart, C.; Plovier, H.; Van Hul, M.; Vieira-Silva, S.; Falony, G.; Raes, J.; Maiter, D.; Delzenne, N.M.; et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: A proof-of-concept exploratory study. Nat. Med. 2019, 25, 1096–1103. [CrossRef] p y y 167. Verhoog, S.; Taneri, P.E.; Roa Díaz, Z.M.; Marques-Vidal, P.; Troup, J.P.; Bally, L.; Franco, O.H.; Glisic, M.; Muka, T. Dietary Factors and Modulation of Bacteria Strains of Akkermansia muciniphila and Faecalibacterium prausnitzii: A Systematic Review. Nutrients 2019, 11, 1565. [CrossRef] 168. Savaiano, D.A.; Hutkins, R.W. Yogurt, cultured fermented milk, and health: A systematic review. N 168. Savaiano, D.A.; Hutkins, R.W. Yogurt, cultured fermented milk, and health: A systematic review. Nutr. Rev. 2020. [CrossRef] 169. Ahmed, Z.; Wang, Y.; Ahmad, A.; Khan, S.T.; Nisa, M.; Ahmad, H.; Afreen, A. Keﬁr and Health: A Contemporary Perspective. Crit. Rev. Food Sci. Nutr. 2013, 53, 422–434. [CrossRef] [PubMed] 169. Ahmed, Z.; Wang, Y.; Ahmad, A.; Khan, S.T.; Nisa, M.; Ahmad, H.; Afreen, A. Keﬁr and Health: A Contemporary Perspective. Crit. Rev. Food Sci. Nutr. 2013, 53, 422–434. [CrossRef] [PubMed] 170. Nguyen, T.L.A.; Vieira-Silva, S.; Liston, A.; Raes, J. How informative is the mouse for human gut microbiota research? Dis. Model. Mech. 2015, 8, 1–16. [CrossRef] 171. Turnbaugh, P.J.; Ridaura, V.K.; Faith, J.J.; Rey, F.E.; Knight, R.; Gordon, J.I. The Effect of Diet on the Human Gut Microbiome: A Metagenomic Analysis in Humanized Gnotobiotic Mice. Sci. Transl. Med. 2009, 1, 6ra14. [CrossRef] g y 172. Pussinen, P.J.; Havulinna, A.S.; Lehto, M.; Sundvall, J.; Salomaa, V. Endotoxemia Is Associated With an Increased Risk of Incident Diabetes. Diabetes Care 2011, 34, 392–397. [CrossRef] [PubMed] 172. Pussinen, P.J.; Havulinna, A.S.; Lehto, M.; Sundvall, J.; Salomaa, V Diabetes. Diabetes Care 2011, 34, 392–397. [CrossRef] [PubMed] 19 of 19 19 of 19 Nutrients 2021, 13, 1389 173. References Kassam, Z.; Lee, C.H.; Yuan, Y.; Hunt, R.H. Fecal Microbiota Transplantation for Clostridium difﬁcile Infection: Systematic Review and Meta-Analysis. Am. J. Gastroenterol. 2013, 108, 500–508. [CrossRef] [PubMed] [ ] [ ] 175. Kootte, R.S.; Levin, E.; Salojärvi, J.; Smits, L.P.; Hartstra, A.V.; Udayappan, S.D.; Hermes, G.; Bouter, K.E.; Koopen, A.M.; Holst, J.J.; et al. Improvement of Insulin Sensitivity after Lean Donor Feces in Metabolic Syndrome Is Driven by Baseline Intestinal Microbiota Composition. Cell Metab. 2017, 26, 611.e6–619.e6. [CrossRef] 176. Smits, L.P.; Kootte, R.S.; Levin, E.; Prodan, A.; Fuentes, S.; Zoetendal, E.G.; Wang, Z.; Levison, B.S.; Cleophas, M.C.P.; Kemper, E.M.; et al. Effect of Vegan Fecal Microbiota Transplantation on Carnitine- and Choline-Derived Trimethylamine- N-Oxide Production and Vascular Inﬂammation in Patients With Metabolic Syndrome. J. Am. Heart Assoc. 2018, 7. [CrossRef]
https://openalex.org/W4205221548	https://tos.org/oceanography/assets/docs/ocean-observing-2021-wriston.pdf	English	null	PacIOOS Water Quality Sensor Partnership Program	Oceanography	2,021	cc-by	1,575	PacIOOS Water Quality Sensor Partnership Program By Shaun Wriston, Gordon Walker, Margaret Anne McManus, Simon Ellis, Fiona Langenberger, Ocean observing systems provide scientists, resource and conservation managers, industry, recreationists, and the general public with ocean data and information to improve decision-making (Iwamoto et al., 2016). While the need for improved open ocean data is global, data and information to address the needs of local, coastal commu- nities is increasingly important and must be identified to effectively aid decision-making by local governing bodies, stakeholders, and those mentioned above. Partnership Program (WQSPP) supports scientists, nat- ural resource managers, and citizens to collect data for research, conservation, planning, and resource manage- ment projects. Accurate and reliable oceanographic param- eters are often difficult to obtain due to a lack of resources and/or technical expertise. The WQSPP aims to fill this gap by partnering with local project coordinators to increase the understanding of coastal marine ecosystems. State and government resource agencies, colleges, nongovern- mental organizations, private businesses, and citizens, as well as independent researchers within the PacIOOS region can apply to use a Sea-Bird 16plus V2 SeaCAT water quality sensor for a period ranging from six months to two years. Sensor packages autonomously measure conductivity, temperature, and pressure with high precision, and up to seven auxiliary sensors can be simultaneously deployed to measure additional parameters such as chlorophyll and turbidity. In addition to the sensor suite, participants are provided with data management and technical capacity building assistance to allow for robust data collection. Initiated as a pilot project in Honolulu in 2007, the Pacific Islands Ocean Observing System (PacIOOS; https:// www.pacioos.hawaii.edu/) was first certified in 2015 by the US Integrated Ocean Observing System (IOOS) as a Regional Information Coordination Entity. Its area of interest includes the State of Hawai‘i; the territories of Guam, American Samoa, and the Commonwealth of Northern Mariana Islands; the Freely Associated States of the Federated States of Micronesia (FSM); the Republic of the Marshall Islands; the Republic of Palau; and the Minor Outlying Islands of Howland, Baker, Johnston, Jarvis, Kingman, Midway, Palmyra, and Wake. PacIOOS collects ocean data from partners across the region, develops and maintains numerical models and forecasts, and integrates this information into freely accessible data services and user-friendly web interfaces (Iwamoto et al., 2016). PacIOOS Water Quality Sensor Partnership Program An application for the WQSPP is initiated with a con- cept paper outlining the participant’s location, eligibility, purpose, and impact; project duration; permitting require- ments; management practices; and a map of the proposed field site. Successful applicants then work with PacIOOS staff to complete a more comprehensive application, and if selected, a project agreement plan. While PacIOOS sup- plies all the equipment, training, and data management, applicants are expected to pay for all costs associated with transporting and maintaining the sensor on site. PacIOOS also collaborates with various organizations and individuals who need water quality data on shorter timeframes to inform their work. The Water Quality Sensor FIGURE 1. Map of current (orange) and previous (yellow) Pacific Islands Ocean Observing System (PacIOOS) Water Quality Sensor Partnership Program (WQSPP) projects. To date, partnership deployments extend throughout the PacIOOS region, including the Hawaiian Islands (Maui, O‘ahu, and Kaua‘i); Palmyra Atoll; Pohnpei, Federated States of Micronesia; and Palau. Palau Federated States of Micronesia Palmyra Atoll Hawai‘i Islands Current WQSSP Projects Previous WQSSP Projects Presently, four sensors are being deployed in the WQSPP program (Figure 1). Data from these sensor packages are used to characterize temporal variability in the water column properties at each site. After data are collected, partners receive a summary document providing specific FIGURE 1. Map of current (orange) and previous (yellow) Pacific Islands Ocean Observing System (PacIOOS) Water Quality Sensor Partnership Program (WQSPP) projects. To date, partnership deployments extend throughout the PacIOOS region, including the Hawaiian Islands (Maui, O‘ahu, and Kaua‘i); Palmyra Atoll; Pohnpei, Federated States of Micronesia; and Palau. FIGURE 1. Map of current (orange) and previous (yellow) Pacific Islands Ocean Observing System (PacIOOS) Water Quality Sensor Partnership Program (WQSPP) projects. To date, partnership deployments extend throughout the PacIOOS region, including the Hawaiian Islands (Maui, O‘ahu, and Kaua‘i); Palmyra Atoll; Pohnpei, Federated States of Micronesia; and Palau. FIGURE 2. A diver maintains the water quality sensor at Kewalo Basin, O‘ahu, which provided valuable data to the nonprofit organization Friends of Kewalos. Photo credit: PacIOOS 66 FIGURE 3. Data produced by PacIOOS WQSPP sensors for Friends of Kewalos. The figure focuses on in-water construction periods (red box) during the Kewalo Basin Harbor construction project and a rain event (blue box). PacIOOS Water Quality Sensor Partnership Program Temperature Pressure Chlorophyll Turbidity Salinity Stream Gauge Height, Manoa-Palolo Drainage Canal Wave Height, Pearl Harbor Temperature (°C) Salinity (PSU) Chlorophyll (µg/L) Wave Height (m) Pressure (m) Stream Gauge Height (ft) Turbidity (NTU) 28.5 27.0 25.5 24.0 2.0 1.5 1.0 0.5 1.5 1.0 0.5 0.0 129630 34 32 30 28 6 4 2 0 1.5 1.0 0.5 0.0 Pressure Chlorophyll Stream Gauge Height, Manoa-Palolo Drainage Canal Stream Gauge Height, Manoa-Palolo Drainage Canal FIGURE 3. Data produced by PacIOOS WQSPP sensors for Friends of Kewalos. The figure focuses on in-water construction periods (red box) during the Kewalo Basin Harbor construction project and a rain event (blue box). FIGURE 3. Data produced by PacIOOS WQSPP sensors for Friends of Kewalos. The figure focuses on in-w during the Kewalo Basin Harbor construction project and a rain event (blue box). details of the project, visualizations, and descriptions of the data gathered by the sensor along with a case study addressing specific objectives of the deployment. Thus far, sensors at individual sites have provided insights into runoff patterns caused by heavy rainfall and progress in pollution cleanup, and they have clarified the impact from periods of in-water construction on local water quality. environment and those who enjoy it. Measurements taken by the nearshore sensor did not show any significant changes in the parameters measured over the course of each in-water construction period. A second result from this WQSPP deployment focused on a rain event that occurred on December 26, 2017. As the neighboring stream gauge height (US Geological Survey) rose to 6.21 ft, salinity levels dropped down to 28.81 prac- tical salinity units, chlorophyll and turbidity levels rose, and temperature dropped to its lowest reading during deploy- ment, 24.27°C, showing how a single environmental event can affect many aspects of nearshore waters (Figure 3). One of the WQSPP projects was with Friends of Kewalos, based in Honolulu, Hawai‘i. Friends of Kewalos worked with the PacIOOS WQSPP to deploy and manage a sensor located on the south shore of O‘ahu in Māmala Bay, at Kewalo Basin (Figure 2). Friends of Kewalos is comprised of recreational users committed to protecting and preserving the Kewalo Basin Park and the surrounding shoreline and ocean. Their intent is to mālama (care for) Kewalo Basin, to ensure that its users will continue to have access to this site and the ability to enjoy it for generations to come. Iwamoto, M.M., F. Langenberger, and C.E. Ostrander. 2016. Ocean observ- ing: Serving stakeholders in the Pacific Islands. Marine Technology Society Journal 50(3):47–54, https://doi.org/10.4031/MTSJ.50.3.2. PacIOOS Water Quality Sensor Partnership Program Through the collection of water quality information, partnerships like the PacIOOS WQSPP and Friends of Kewalos help natural resource managers and researchers to better evaluate their projects and make more informed decisions about them. PacIOOS aims to grow the number of sensor packages available and projects supported as interest and resources continue to grow. Deployment for this partnership extended from September 2017 to May 2019, including a period of in- water construction at Kewalo Basin Harbor in October and November 2017. In September 2017, Kewalo Basin Harbor initiated an improvements project aimed at increasing harbor berth count, rehabilitating piers, and replacing a condemned loading dock and fueling system. This proj- ect included both out of water and in-water construction, and standard erosion barriers were put in place to pre- vent sediment from leaving the construction site. Friends of Kewalos and other community members were con- cerned that this construction (particularly in-water) may have adverse effects for both the surrounding marine PacIOOS would like to acknowledge and thank our part- ners at Maui Nui Marine Resource Council, Ebiil Society, Conservation Society of Pohnpei, Kaua‘i Sea Farms, The Mariana Islands Nature Alliance, US Fish and Wildlife Service, Friends of Kewalos, and Mālama Maunalua. ARTICLE CITATION Wriston, S., G. Walker, M.A. McManus, S. Ellis, F. Langenberger, and M. Iwamoto. 2021. PacIOOS Water Quality Sensor Partnership Program. Pp. 66–67 in Frontiers in Ocean Observing: Documenting Ecosystems, Understanding Environmental Changes, Forecasting Hazards. E.S. Kappel, S.K. Juniper, S. Seeyave, E. Smith, and M. Visbeck, eds, A Supplement to Oceanography 34(4), https://doi.org/10.5670/oceanog.2021. supplement.02-25. REFERENCE ARTICLE DOI: https://doi.org/10.5670/oceanog.2021.supplement.02-25 67 Shaun Wriston (swriston@hawaii.edu), Department of Oceanography, University of Hawai‘i at Mānoa, USA. Gordon Walker, Pacific Islands Ocean Observing System, and Department of Oceanography, University of Hawai‘i at Mānoa, USA. Margaret Anne McManus, Department of Oceanography, University of Hawai‘i at Mānoa, USA. Simon Ellis, Pacific Islands Ocean Observing System, and University of Hawai‘i Sea Grant, USA. Fiona Langenberger and Melissa Iwamoto, Pacific Islands Ocean Observing System, USA. COPYRIGHT & USAGE This is an open access article made available under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/ licenses/by/4.0/), which permits use, sharing, adaptation, distribution, and repro- duction in any medium or format as long as users cite the materials appropriately, provide a link to the Creative Commons license, and indicate the changes that were made to the original content.
https://openalex.org/W2143942987	https://europepmc.org/articles/pmc3575313?pdf=render	English	null	Screening and managing cannabis use: comparing GP’s and nurses’ knowledge, beliefs, and behavior	Substance abuse treatment, prevention, and policy	2,012	cc-by	8,138	RESEARCH Open Access © 2012 Norberg et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Screening and managing cannabis use: comparing GP’s and nurses’ knowledge, beliefs, and behavior Norberg1, Peter Gates1, Paul Dillon1, David J Kavanagh2, Ramesh Manocha3 and Jan Copeland1 Melissa M Norberg1, Peter Gates1, Paul Dillon1, David J Kavanagh2, Ramesh Manocha3 and Jan Copeland1 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 * Correspondence: mnorberg@unsw.edu.au 1National Cannabis Prevention and Information Centre, UNSW, PO Box 684, Randwick, NSW 2031, Australia Full list of author information is available at the end of the article Abstract Background: General practitioners (GPs) and nurses are ideally placed to address the significant unmet demand for the treatment of cannabis-related problems given the numbers of people who regularly seek their care. The aim of this study was to evaluate differences between GPs and nurses’ perceived knowledge, beliefs, and behaviors toward cannabis use and its screening and management. Methods: This study involved 161 nurses and 503 GPs who completed a survey distributed via conference satchels to delegates of Healthed seminars focused on topics relevant to women and children’s health. Differences between GPs and nurses were analyzed using χ2- tests and two-sample t-tests, while logistic regression examined predictors of service provision. Results: GPs were more likely than nurses to have engaged in cannabis-related service provision, but also more frequently reported barriers related to time, interest, and having more important issues to address. Nurses reported less knowledge, skills, and role legitimacy. Perceived screening skills predicted screening and referral to alcohol and other drug (AOD) services, while knowing a regular user increased the likelihood of referrals only. Conclusions: Approaches to increase cannabis-related screening and intervention may be improved by involving nurses, and by leveraging the relationship between nurses and doctors, in primary care. Keywords: Cannabis, Marijuana abuse, Diagnosis, Therapeutics, Primary health care, Attitude Participants If more cannabis screening and intervention were to occur within primary care, a key question is whether nurses or GPs could more feasibly undertake this work. While nurses may be less subject to the time constraints for opportunistic intervention that doctors experience [23,24], findings from the alcohol literature indicate that nurses compared to GPs have less knowledge and skills about screening and intervention [16,25], have less fa- vorable attitudes towards discussing substance use with their patients [16], and are less likely to believe that this work should be a part of their role [16]. In addition, qualitative research found that nurses can be hesitant to provide alcohol interventions because of their own use and enjoyment of alcohol and because of their beliefs that alcohol use may serve beneficial social and coping functions for some patients [21]. If these alcohol-related findings are transferrable to cannabis use, they suggest that nurses may be less inclined than GPs to screen for cannabis use and provide early intervention. The study sample was comprised of delegates attend- ing a Healthed seminar in one of five Australian cities (Sydney, Melbourne, Adelaide, Brisbane, and Perth) in 2011. Each Healthed seminar featured approximately sixteen lectures on women and children’s health and represented a rare opportunity to meet with several hun- dred GPs and nurses in one location. GPs and nurses received professional education points for attending the lectures. Of the 1,925 delegates who attended the Healthed seminars, 664 (35%) completed the study sur- vey; 503 surveys were completed by GPs (76%) and 162 were completed by nurses (24%). This distribution was representative of the Healthed population, in which 71% were GPs and 26% were nurses. The study sample was predominately female (n = 527, 79%), which again reflects the Healthed population, as only 41% of Australian GPs are female [26]. The sample had a mean age of 48 years (SD = 9.88, range 21-81) and most practiced in a metro- politan area (n = 463, 71%). GPs were significantly more likely to be male (27% vs. Participants <1%; χ2(1) = 51.23, p < 0.001), and younger (GPs Mage = 47.61 years, SD = 10.25; nurse’s Mage = 49.93 years, SD = 8.04; t(343.77) = -2.95, p = 0.003), and reported spending significantly more hours in the clinic per week than did nurses (GP Mhours = 30.45, SD = 12.65; nurse’s Mhours = 26.56, SD = 12.04; t(646) = 3.35, p = 0.001. While a modest amount of research has examined GP’s and nurses’ attitudes towards alcohol and substance use in general, relatively little has focused specifically on cannabis use. In a UK survey of 97 GPs, most believed that cannabis use posed a health risk and may lead to mental health problems, but less than half believed that they had adequate knowledge about cannabis use, and only a third felt confident in their ability to advise patients about their cannabis use [22]. In addition, around half believed that GPs should refer patients with cannabis-related problems to specialist drug services, with only a third believing it was appropriate to treat cannabis dependence in primary care. GPs who believed it was appropriate to treat cannabis dependence in pri- mary care felt more confident in their ability to advise cannabis-using patients, whereas GPs who believed cannabis-using patients should be referred to specialist drug services reported a greater need to improve their knowledge of cannabis-related risks. Data collection D h During the opening of each Healthed seminar, we informed delegates that the survey and its information and consent form were located in their conference satchel. We advised delegates that if they were interested in participating that they should complete the entire self-report survey, as item nonresponse would render their survey invalid given the brevity of the survey. We instructed delegates to turn in their surveys, before after- noon tea, into dropboxes at the conference. Lastly, we instructed delegates that participation would result in entry into a lottery held during afternoon tea to win one of five gift certificates ($75 AUD). These procedures re- flect a cross-sectional design. The aim of this study was to assess potential differ- ences between GPs and nurses. Based on findings reviewed above, we hypothesized that while few GPs and nurses would be confident in their knowledge and skills about cannabis use and its treatment, nurses would re- port having less training (Hypothesis 1) and less role le- gitimacy for screening and treating cannabis use (Hypothesis 2), and perceive their knowledge and skills to be poorer than GPs (Hypothesis 3). In addition, we hypothesized that GP’s and nurses’ training, personal Background provides substantial opportunity for cannabis use screening and intervention. Doctors and nurses in primary care may be able to influence their patients’ cannabis use through a variety of health strategies [13], but their attitudes, knowledge, and skills may limit the provision of such care [14,15]. g Cannabis remains the most commonly used illicit drug, with around 200 million current users world- wide [1]. Cannabis use increases the risk of chronic respiratory and cardiovascular problems [2-5], and around one in nine users are at risk of developing de- pendence [6,7]. In Australia, these health risks are of significant concern given that cannabis use contri- butes to 10% of the burden relating to illicit drug use [8]. While few cannabis users seek specialist drug treatment [9,10], general practitioners (GPs) are the most frequently sought resource for treatment of can- nabis use [11]. As approximately 80% of Australians visit a GP at least once a year [12], primary care Although primary care practitioners are encouraged to, and typically believe it is appropriate for them to screen and provide early interventions for substance use [16], many do not feel comfortable diagnosing or treat- ing substance use problems [17-19]. Further, the content of a typical substance use brief intervention is not well known [16] and commonly disregarded as ineffective [18,20]. Many doctors and nurses have reported avoiding substance use discussions due to anticipated negative reactions from patients, believing that patients will not be honest about their substance use, not having enough training and resources, and time constraints [18,20-22]. * Correspondence: mnorberg@unsw.edu.au 1National Cannabis Prevention and Information Centre, UNSW, PO Box 684, Randwick, NSW 2031, Australia Full list of author information is available at the end of the article Page 2 of 10 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 experience, knowledge, attitudes, and beliefs would be related to whether or not they engaged in screening or intervention for cannabis use (Hypothesis 4). Conversely, doctors and nurses report that having readily available support services, quick and easy screening instruments and counseling material, more training, and evidence of the successful impact for early intervention would lead to their greater involvement in screening and intervention [16,20,22]. Survey h The survey was devised primarily for use in this study, and thus, had not been psychometrically validated. The survey had 31 items; of which 25 were relevant to the current study. Five items addressed basic demographic Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 3 of 10 significance level for all statistical tests was set at p < 0.01 to reduce Type I errors associated with multiple testing. information (career, gender, age, physical location of practice, and weekly clinical hours), and two items cov- ered exposure to cannabis issues (relevant training and personal experience with a regular cannabis user). Five items assessed knowledge regarding cannabis use (e.g., Cannabis users are more likely to have a mental health problem than those who do not use cannabis); four of these items were rated on a Likert scale from 0 (disagree completely) to 4 (agree completely). The fifth knowledge item was answered with a percentage (What percentage of people who try cannabis will someday develop canna- bis dependence?). Knowledge about Cannabis use and its treatment Knowledge about Cannabis use and its treatment As set out in Table 2, GPs were more likely than nurses to have agreed that effective psychological treatments exist for reducing cannabis use (p = 0.001) and that can- nabis users are more likely to have a mental health prob- lem than non-cannabis users (p < 0.001). Further, 34% (n = 221) of health practitioners somewhat or completely agreed that effective pharmacological treatments exist for cannabis use. On average, GPs believed that 21% (SD = 21.58) of people who try cannabis will someday develop cannabis dependence, which was significantly less than nurses believed (M = 25%, SD = 21.58), t(131.51) = -1.14, p = 0.001. Of the 516 participants who responded to this item, only 24% (n = 126) responded with an estimate close to 9% (6-12%) [6,7]. Twenty-one percent of partici- pants did not answer the question. Table 1 Exposure to Cannabis Use Item n GPs Nurses χ2(1) n (%) n (%) Cannabis Related Training 660 None 132 (27%) 66 (41%) 12.26** Small Amount 300 (60%) 77 (48%) 7.51* Moderate Amount 60 (12%) 17 (11%) 0.25 Substantial Amount 7 (1%) 1 (<1%) 0.62 Great Deal 0 (0%) 0 (0%) – Cannabis users known to participantsa 657 No one 331 (67%) 68 (43%) 28.48 Relative (other than child) 43 (9%) 35 (22%) 20.23 Child 13 (3%) 7 (4%) 1.27 Friend 129 (26%) 65 (41%) 12.52** Self 1 (<1%) 0 (0%) 0.32 Note. aPercentages do not add up to 100% as some participants choose more than one response option. p < 0.01, p ≤0.001. Overview of the study sample Table 1 contains summary data for cannabis-related training and exposure to cannabis users. GPs were sig- nificantly more likely to not have any personal experience with cannabis use compared to nurses (p < 0.001), while nurses were more likely to have a relative (p < 0.001) or know a friend (p < 0.001) who regularly uses cannabis. Further, GPs were significantly more likely to have received a small amount of training in cannabis-related issues (p = 0.006), while nurses were more likely to have not received any cannabis-related training (p < 0.001). The remaining 13 items used in this study were based on previous surveys assessing health practitioners’ beliefs, attitudes, and behavior regarding substance use and its treatment. Three items required GPs and nurses to assess their own knowledge about cannabis use and their skills related to screening and managing cannabis use [16–18,27]. These three items were rated on a Likert scale from 0 (very poor) to 4 (very strong). In addition, three items assessed beliefs regarding role legitimacy— factors related to the appropriateness of a GP or nurse to intervene with someone’s cannabis use [21,25,27,28] and were rated on a Likert scale from 0 (disagree com- pletely) to 4 (agree completely). One item assessed atti- tudes regarding cannabis use policy (Cannabis use should be illegal / decriminalized / available for med- ical use) [22,29]. Four open-ended items assessed behav- ior related to screening, intervention, and referral [22,28,30]. Participants responded to these items with numbers (e.g., I have screened ____ patients for cannabis use in the last month). The final two items assessed bar- riers and facilitators to screening and treatment provision [18,20,25,27-29]. These items were answered by checking all applicable barriers or facilitators. Options referred to attitudes, motivation, confidence, and support. Statistical analyses Differences between GPs and nurses on categorical vari- ables were analyzed using 2 Χ 2 χ2- tests (e.g., Hypothesis 1), while two-sample t-tests were used for continuous variables (e.g., Hypothesis 2, Hypothesis 3). When Levene’s test for equality of variances indicated that the variances between GPs and nurses responses were signifi- cantly different, a two-sample t-test was performed that did not assume equal variances. Logistic regressions using a generalized linear function examined the prediction of service provision (Hypothesis 4; 0 = none, 1 = at least once; reference category= 0). Before conducting the logistic regressions, zero-order correlations were conducted to determine which variables were appropriate for model inclusion. Only items statistically significant at the uni- variate level were included at the multivariate level. The Norberg et al. Statistical analyses Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 4 of 10 Table 2 Knowledge and Beliefs about Cannabis Use and its Treatment Item Disagree completely Somewhat disagree Neutral Somewhat agree Agree completely t df n (%) n (%) n (%) n (%) n (%) Knowledge Cannabis users are more likely to have a mental health problem than those who do not use cannabis GPs (N = 502) 4 (<1%) 11 (2%) 19 (4%) 159 (32%) 309 (62%) Nurses (N = 160) 4 (3%) 6 (4%) 20 (13%) 64 (40%) 66 (41%) 4.54* 224.80 Effective psychological treatments exist for helping people to reduce their cannabis use GPs (N = 501) 9 (2%) 31 (6%) 118 (23%) 196 (39%) 147 (29%) Nurses (N = 159) 7 (4%) 14 (9%) 42 (26%) 73 (46%) 23 (14%) 3.49 658 Effective pharmacological treatments exist for assisting with cannabis withdrawal GPs (N = 499) 43 (9%) 93 (19%) 195 (39%) 116 (23%) 52 (10%) Nurses (N = 159) 11 (7%) 20 (13%) 75 (47%) 41 (26%) 12 (8%) -0.65 656 Withdrawal can be a barrier to quitting cannabis GPs (N = 500) 10 (2%) 23 (5%) 67 (13%) 220 (44%) 180 (36%) Nurses (N = 158) 2 (<1%) 3 (2%) 32 (20%) 71 (45%) 50 (32%) 0.44 656 Role Legitimacy Conducting a 10 minute brief assessment of someone’s cannabis use can lead to reductions in their use GPs (N = 501) 22 (4%) 24 (5%) 105 (21%) 210 (42%) 140 (28%) Nurses (N = 159) 31 (20%) 26 (!6%) 37 (23%) 47 (30%) 18 (11%) 7.72 223.44 People in my position are effective in treating patients with cannabis use problems GPs (N = 499) 22 (4%) 60 (12%) 103 (21%) 218 (44%) 96 (19%) Nurses (N = 160) 23 (14%) 36 (23%) 52 (33%) 38 (24%) 11 (7%) 7.64** 657 People in my position should receive education about cannabis GPs (N = 501) 1 (<1%) 1 (<1%) 14 (3%) 116 (23%) 369 (74%) Nurses (N = 160) 0 (0%) 1 (<1%) 6 (4%) 43 (27%) 110 (69%) 1.20 659 Note. p < 0.01, p ≤0.001. Table 2 Knowledge and Beliefs about Cannabis Use and its Treatment Note. p < 0.01, p ≤0.001. than GPs reported that cannabis should be available for medicinal purposes, χ2(1) = 30.11, p < 0.001, as shown in Figure 1. Role legitimacy Nurses reported significantly less role legitimacy about providing cannabis-related services. Compared to nurses, GPs more strongly agreed that conducting a 10-minute brief assessment of someone’s cannabis use can lead to reductions in cannabis use (p < 0.001) and that people in their position are effective in treating patients with can- nabis use problems (p < 0.001). Almost all GPs and nurses agreed that people in their position should receive education regarding cannabis (See Table 2). Self-assessment of knowledge and skills Few GPs and nurses felt confident in their knowledge and skills about cannabis use and its treatment. As shown in Table 3, GPs rated their knowledge about can- nabis (p < 0.001), and their skills in screening (p < 0.001) and managing cannabis use (p < 0.001), higher than nurses rated their own knowledge and skills. Screening, intervention, and referral g In the previous month, GPs reported screening an aver- age of 3.84 (SD = 7.82, n = 498) patients for cannabis use and treating an average of 1.71 (SD = 4.17, n = 500) patients for cannabis use; although, approximately half (46%, n = 230) had not engaged in any screening or intervention provision (51%, n = 254). Nurses reported screening an average of 2.22 (SD = 10.00, n = 158) patients for cannabis use and treating an average of 1.33 (SD = 8.40, n = 159) patients for cannabis use in the last month. Most nurses, however, had not screened (85%, n = 134) or provided an intervention to any patients for cannabis use (87%, n = 139). In addition, GPs and nurses had referred less than one patient in the last month to an alcohol and other drug (AOD) facility (GPs: M = 0.48, SD = 1.42, n = 500, 374 (75%) had not engaged in referral; Nurses: M = 0.84, SD = 4.62, n = 160, 139 (87%) had not engaged in referral) or to a mental health service (GPs: M = 0.76, SD = 1.78, n = 500, 318 (64%) had not engaged Policy attitudes A significantly greater percentage of GPs than nurses reported that cannabis use should be illegal, χ2(1) = 14.59, p < 0.001, and a significantly greater percentage of nurses Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 5 of 10 Table 3 Self-Assessment of Knowledge about Cannabis Use and Skills Related to Screening and Managing Cannabis Use Item Very poor Poor Acceptable Strong Very strong t df n (%) n (%) n (%) n (%) n (%) Knowledge about Cannabis GPs (N = 501) 35 (7%) 241 (48%) 209 (42%) 14 (3%) 2 (<1%) 3.85 658 Nurses (N = 159) 30 (19%) 76 (48%) 49 (31%) 4 (3%) 0 (0%) Skills in Screening for Cannabis Use GPs (N = 499) 51 (10%) 241 (48%) 187 (37%) 18 (4%) 2 (<1%) 7.32 654 Nurses (N = 157) 57 (36%) 68 (43%) 29 (18%) 3 (2%) 0 (0%) Skills in Managing Cannabis Use GPs (N = 495) 57 (12%) 300 (61%) 127 (26%) 9 (2%) 2 (<1%) Nurses (N = 156) 55 (35%) 85 (54%) 15 (10%) 1 (<1%) 0 (0%) 7.13 649 Note. p < 0.01, p ≤0.001. Table 3 Self-Assessment of Knowledge about Cannabis Use and Skills Related to Screening an Use of Knowledge about Cannabis Use and Skills Related to Screening and Managing Cannabis Note. p < 0.01, p ≤0.001. in referral; Nurses: M = 0.44, SD = 2.39, n = 160, 143 (89%) had not engaged in referral). GPs were statistically signifi- cantly more likely to have engaged in screening (χ2(1) = 72.45, p < 0.001), treatment (χ2(1) = 64.85, p < 0.001), and referral (AOD: χ2(1) = 10.21, p = 0.001; Mental Health: χ2(1) = 38.24, p < 0.001) at least once (versus not at all) compared to nurses. 4.73 times more likely to have engaged in screening provision than nurses, were 4.22 times more likely to have provided an intervention, and were 3.47 times more likely to have referred someone to a mental health service, after controlling for other predictors. Confidence in the provision of screening was import- ant, with every unit increase in perceived screening skills associated with a 3.28 times greater chance of having engaged in screening provision in the previous month and an 8.69 times greater chance of having referred a patient to an AOD service. Discussion likely to screen for cannabis use than those who believed it should be available. While that attitude was a signifi- cant univariate correlate of intervention and referral to a mental health service, it did not provide a unique predic- tion of those actions. This study evaluated GP’s and nurses’ perceived know- ledge, beliefs, and behaviors toward cannabis use and its screening and management. As a whole, the findings suggest that despite GPs and nurses having little to no training in cannabis-related issues, most are aware of basic treatment-related issues and believe that it is part of their role to be educated about cannabis use. Yet, many GPs and nurses believed their knowledge about cannabis use and their skills in screening and managing cannabis use to be poor. The findings also suggest that nurses may be less knowledgeable about cannabis use and feel less role legitimacy for its screening and treat- ment. Thus, it is not surprising that being a GP was con- sistently associated with cannabis service provision compared with being a nurse. When controlling for other relevant variables, GPs were three to five times more likely than nurses to have engaged in screening, brief intervention, and referral to mental health services. Study findings also indicate that policy attitudes, per- sonal experience, and perceived screening skills may influence service provision. Policy attitudes Table 5 Barriers and Facilitators for Cannabis Use Screening and Intervention (N = 664) Statement GPs agreement n (%) Nurses agreement n (%) χ2(1) Table 5 Barriers and Facilitators for Cannabis Use Screening and Intervention (N = 664) Statement GPs agreement n (%) N GPs agreement n (%) Nurses agreement n (%) χ2(1) Policy attitudes Neither the degree of train- ing, knowledge, nor perceived skills in cannabis inter- vention, however, offered a significant unique prediction of screening, intervention, or referral. In order to identify if any variables had a unique pre- dictive variance regarding increased provision of screen- ing or intervention for cannabis use, only variables correlated with screening, intervention, and referral to AOD or mental health services at a significance level less than 0.01 were entered into the logistic regression mod- els (Table 4). Interaction terms between health profes- sion and other predictor variables were examined; however, as none were significant they were not included in the reported models. When the significant univariate correlates were included in the regressions, profession remained a significant predictor of screening, interven- tion and referral to mental health services. GPs were Health professionals who had a friend who regularly used cannabis were 1.80 to 1.83 times more likely to have referred someone to an AOD or mental health ser- vice than those without such a friend. In regards to atti- tudes, participants who believed cannabis should not be available for medicinal purposes were 2.06 times more ** 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% Illegal Decriminalised Available for Medicinal Purposes GPs Nurses Figure 1 This figure illustrates GP’s and nurses’ percentage agreement with three different cannabis use policy options. This figure illustrates GP’s and nurses’ percentage agreement with three different cannabis use policy options. d Referral Provision in the Previous Month tion Referral to AOD Services Referral to Menta 99% CI Wald Χ2 OR 99% CI Wald Χ2 OR 2.32, 7.67 4.64 1.88 1.06, 3.34 15.52* 3.4 – – – – 5.61 0.5 – – – – – – 0.88, 2.02 – – – 0.22 1.1 0.44, 1.04 7.45* 0.55 0.35, 0.84 7.19* 0.5 1.01, 1.98 1.25 1.22 0.86, 1.72 0.39 1.1 0.77, 1.71 2.37 0.72 0.47, 1.09 0.54 1.1 1.02, 2.09 8.69* 1.79 1.22, 2.65 4.54 1.4 0.96, 2.26 0.70 1.21 0.78, 1.88 1.14 1.2 0.97, 1.42 – – – – – 0.85, 1.28 2.06 1.17 0.94, 1.45 – – 0.89, 1.31 – – – 0.42 1.0 ch independent variable: a GPs, b Should be. . . (illegal, decriminalized, or available for med 3.4 0.5 – 1.1 0.5 1.1 1.1 1.4 1.2 – – 1.0 or med Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 Page 7 of 10 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Policy attitudes Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 7 of 10 Page 7 of 10 Table 5 Barriers and Facilitators for Cannabis Use Screening and Intervention (N = 664) Statement GPs agreement n (%) Nurses agreement n (%) χ2(1) Barriers Do not have the skills to screen or provide an intervention 336 (67%) 124 (77%) 5.94 Cannabis use often is not the most important issue that needs to be addressed during a patient’s visit 286 (57%) 56 (35%) 23.80 Patients may not be receptive to screening or intervention 279 (55%) 98 (61%) 1.45 Do not have time to screen or provide an intervention 242 (48%) 41 (25%) 25.56 Do not have support from colleagues/organization to screen or provide an intervention 138 (27%) 55 (34%) 2.68 The effort required to motivate a cannabis user to quit is not justified by the likelihood of a positive outcome 101 (20%) 15 (9%) 9.80* Do not have a personal interest in providing screening or treatment for cannabis use 95 (19%) 12 (7%) 11.78 Do not have professional interest in providing screening or treatment for cannabis use 77 (15%) 13 (8%) 5.45 I do not want to attract more cannabis users to my clinic 76 (15%) 5 (3%) 16.41 Most people who use cannabis do not need screening or treatment 30 (6%) 7 (4%) 0.61 Cannabis users are unpleasant to work with 24 (5%) 3 (2%) 2.64 Cannabis users should only be treated by specialists in the field 15 (3%) 18 (11%) 17.36 Facilitators More training 408 (81%) 144 (89%) 6.03 Access to up-to-date management guidelines and recommendations 396 (79%) 130 (81%) 0.30 Having more options for referrals 391 (78%) 88 (55%) 32.31 Having more resources to assist me (e.g., web interventions) 315 (63%) 103 (81%) 0.10 Believing that screening and intervention would lead to positive outcomes 232 (46%) 67 (42%) 1.00 Believing that screening and intervention are clinically important 134 (27%) 51 (32%) 1.54 Nothing would facilitate the screening or intervention of cannabis use 6 (1%) 2 (1%) <0.01 Note. p < 0.01, *p ≤0.001. Implications for policymakers and educators General practitioners and nurses represent the over- whelming majority of primary care providers in Australia. They have enormous potential to reach a broad range of people for early intervention, where the focus is on pre- vention and health promotion. Although Australian GPs are being encouraged via government reimbursement schemes to provide brief advice in many prevention areas, including drug and alcohol use [13], efforts to in- crease service provision for cannabis users may need to be directed at both nurses and GPs, in order to utilize the opportunity for collaboration between GPs and nurses in the primary care setting. There are several rea- sons why nurses should be involved in this process. First, nurses were consistently less likely to have engaged in service provision than GPs. Second, GPs in this study were substantially more likely to identify bar- riers to service provision that are difficult to resolve, such as not having enough time and cannabis use not being the most important issue. Third, in 2010 the Aus- tralian Government announced a $523 million invest- ment in Australia’s nurses in order reduce GP’s clinical burden [32]. Such an investment increases the need to improve nurses’ skills to ensure that they are fully able to undertake effective preventive health activities. Al- though the aim of the initiative is to allow GPs to focus on diagnosis, while nurses focus on tasks such as care co-ordination, health assessments, and health education [32], the most suitable model of practice for nurses has not been established. Thus, when developing training models, researchers will need to keep in mind that nurses may be operating from a substitution model (only doing those tasks delegated to them) or from a collaborative model (working autonomously within a practice). In addition, researchers will need to find an effective method for resolving the low legitimacy that nurses feel for managing cannabis use. Consistent with previous research [18,20-22], GPs and nurses identified not having enough skills, patients not being receptive, and not having enough time as common reasons for not engaging in screening and intervention for cannabis use. Not believing that cannabis use is the most important issue was a new common barrier identi- fied in this study. Barriers and facilitators As shown in Table 5, GPs more often than nurses reported not having enough time for screening and intervention provision (p < 0.001), not being personally interested in screening and treatment provision (p = 0.001), not believing that cannabis use is the most important clinical issue (p < 0.001), not wanting to at- tract more cannabis users to their clinic (p < 0.001), and believing that the effort required to motivate a cannabis user to quit is not justified by the likelihood of a positive outcome (p < 0.002). Nurses more often reported that cannabis users should only be treated by specialists in the field (p < 0.001). In respect to facilitators, GPs more so than nurses reported that having more referral options would facilitate their provision of screening and intervention for cannabis use (p < 0.001). Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 8 of 10 The finding that nurses perceive themselves to be less knowledgeable and to have less role legitimacy than GPs is consistent with alcohol use research [16,25]. While GPs rated their knowledge and skills better than did nurses, it is important to note that less than half of GPs considered their knowledge and screening skills to be ac- ceptable to very strong and less than one-third consid- ered their skills in managing cannabis use to be acceptable to very strong. In addition, only a quarter of GPs and nurses reported an accurate estimate of the chance of developing dependence [6,7]. As 21% of parti- cipants did not answer the question, 24% may overesti- mate the actual rate of accurate responses, as those who did not know the accurate figure may have chosen to skip this item. Both GPs and nurses also overestimated the effectiveness of pharmacological treatments [31]. These findings indicate that even though nurses may benefit from training in cannabis-related issues more so than GPs, training also has the potential to improve GP’s service provision to cannabis users. services, these findings suggest that training in cannabis- related issues should incorporate discussions on the ef- fectiveness of cannabis for medicinal purposes and how personal experience can shape their professional behav- ior. Doing so may help to motivate all health practi- tioners to engage in evidenced-based service provision. Implications for policymakers and educators An additional limitation of this study is the high amount of missing data regarding the percentage of patients who will develop dependence (21% missing data). Research suggests that the high nonresponse on this item may have been due to the cognitively demanding nature of the question [36]. In order to reduce the non-response rate for this item, future researchers may want to adapt the open-ended item to include response options. Lastly, respondents were not provided with operational defini- tions for screening, brief interventions, or referrals; there- fore, these items may have been interpreted differently. In light of these limitations, results may only generalize to GPs and nurses interested in women’s and children’s health who also have an interest in contributing to research. Future research with professionals interested in men’s health and that uses operational definitions and a validated survey may obtain different results. in their role as health educators. This would allow GPs to remain in the referral process, but minimize their clinical burden, while appropriately managing patients concerns. In those circumstances where nurses and nurse practitioners are working in a relatively autono- mous context, nurses could be trained in both screening and intervention, referring to the GP only when the ex- tremity of clinical features might warrant it. Regardless of whom efforts are focused on, combining motivational interviewing with screening and interven- tion training for cannabis use may lead to greater provision of services to individuals who use cannabis. Research has found that motivating GPs to provide an intervention does not need to involve long professional development sessions. For example, a 20-minute motiv- ational discussion with GPs has substantially increased their likelihood of discussing methods for managing dependent cannabis users [30]. Such motivational train- ing may circumvent the negative effect that health prac- titioners’ pre-existing low motivation has on service provision. Providing GPs with training and support has been found to only increase provision of alcohol screen- ing and brief advice for those GPs who were already motivated to work with patients with alcohol problems prior to receiving training [28]. Almost all GPs and nurses in the current study reported it was part of their role to receive education about cannabis use, and over three-fourths reported that receiving more training would increased their service provision. Authors’ information MMN is a clinical psychologist and senior lecturer. PG is a doctoral student in Community Medicine and Public Health. PD is a drug and alcohol educator. DJK is a clinical psychologist and professor. RM is a general practitioner and senior lecturer. JC is a professor who has a PhD in Community Medicine and Public Health. This work was carried out at the National Cannabis Prevention and Information Centre at the University of New South Wales. All surveys were distributed during HealthEd seminars conducted in major cities across Australia. Abbreviation l GPs: General Practitioners. When evaluating the effects of training on GPs’ and nurses’ service provision, researchers should carefully monitor increases in knowledge and attitudes toward ser- vice provision. Previous research found that training in a brief intervention led to increases in the understanding of a brief intervention for nurses, but not GPs [33]. Fur- thermore, while nurses’ knowledge about alcohol increased as a result of training, their positive attitudes towards discussing alcohol with patients decreased. Thus, training needs to be carefully developed and admi- nistered to avoid potential negative consequences. Authors’ contributions MMN, PG, PD, and RM contributed to the development of the study survey. MN wrote the first draft of the article and conducted all statistical analyses. PG revised the first draft and provided statistical advice. PD, RM, DJK, and JC contributed to other manuscript revisions. All authors read and approved the final manuscript. Competing interests Dr Ramesh Manocha is the CEO of Healthed pty ltd, a company that conducts educational seminars for GPs, nurses and other health professionals. This study was conducted at Healthed seminars. The other authors do not have any conflicts of interest. Implications for policymakers and educators As such, provid- ing training in screening (that includes information about screening resources, practice guidelines, and refer- ral options) and in brief motivational interventions (that includes information about treatment resources, and dis- cussion of personal experience and medicinal cannabis use) during professional development seminars may in- crease service provision for cannabis users. Conclusions This study found that as a group, GPs and nurses perceive their knowledge and skills pertaining to cannabis-related issues to be poor. In addition, this study identified that health professionals’ occupation, attitudes, personal ex- perience, and skills may affect their provision of cannabis- related care. Education and training, therefore, may be particularly important to improving service provision for individuals who use cannabis, especially training that uti- lizes the complementary roles that each professional plays in general practice, with a special emphasis on in- creasing nurses’ knowledge, skills, and role legitimacy. Implications for policymakers and educators Given that more GPs than nurses endorsed this barrier, it is not surprising that GPs were more likely than nurses to report not having enough time and that having more referral options would facili- tate their service provision. These findings suggest that GPs may see more legitimacy for their role in screening than they do intervention provision. As this study did not examine health practitioners’ beliefs about their role legitimacy for screening, future research would benefit from examining role legitimacy beliefs for both screen- ing and intervention provision. Prior to this study, scant research had statistically examined whether personal views and experience with cannabis use influenced service provision. An exception is Johnson and colleagues’ study [18], which found that primary care physicians who had family and friends with substance use problems had less difficulty discussing substance use with patients. The current study found that having a friend who regularly uses cannabis increased referral provision, but not screening and inter- vention provision. This study also identified that com- pared to health practitioners who believed that cannabis should be available for medicinal purposes, those who believed cannabis should not be available for medicinal purposes were twice as likely to have screened a patient for cannabis use. Along with Lock and colleagues’ [21] finding that nurses’ beliefs about alcohol’s perceived ben- efits affect their willingness to provide alcohol-related Hence, part of a training strategy may require educa- tion and empowerment of nurses in terms of their clin- ical potential to assist in managing cannabis use problems. In a substitution model, it may be advanta- geous to educate both GPs and nurses in how best to utilize their respective clinical strengths. This might in- volve nurses performing the screening and assessment of patients’ readiness for change and then referring moti- vated patients on to GPs for a brief intervention. GPs may develop clinical plans that involve utilizing nurses Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Page 9 of 10 Page 9 of 10 Notably, the obtained survey response rate was similar to other surveys conducted with Australian GPs [34,35], and thus, may not be indicative of non-respondents being less interested in cannabis use than responders. References Sherman MP, Campbell LA, Gong HJ, Roth MD, Tashkin DP: Antimicrobial and respiratory burst characteristics of pulmonary alveolar macrophaes recovered from smokers of marijuana alone, smokers of tobacco alone, smokers of marijuana and tobacco, and nonsmokers. Am Rev Respir Dis 1991, 144:1351–1356. 27. Silins E, Conigrave KM, Rakvin C, Dobbins T, Curry K: The influence of structured education and clinical experience on the attitudes of medical students towards substance misuers. Drug Alcohol Rev 2007, 26:191–200. 6. Degenhardt L, Hall W, Lynskey M: The relationship between cannabis use and other substance use in the general population. Drug Alcohol Depend 2001, 64:319–327. 28. Anderson P, Kaner E, Wutzke S, Funk M, Heather N, Wensing M, Grol R, Gual A, Pas L: Attitudes and managing alcohol problems in general practice: An interaction analysis based on findings from a WHO Collaborative Study. Alcohol Alcohol 2004, 39:351–356. 7. Anthony JC, Warner LA, Kessler RC: Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalants: Basic findings from the National Comorbidity Survey. Exp Clin Psychopharmacol 1994, 2:244–268. 29. Agrawal S, Everett WW, Sharma S: Medical students views of substance abuse treatment, policy and training. Drugs Educ Prev Pol 2010, 17:587–602. 30. McCambridge J, Strang J, Platts S, Witton J: Cannabis use and the GP: Brief motivational intervention increases clinical enquiry by GPs in a pilot study. Br J Gen Pract 2003, 53:637–639. 8. Begg S, Vos T, Barker B, Stevenson C, Stanley L, Lopez A: The burden of disease and injury in Australia 2003. Canberra: AIHW; 2007 http://www.aihw. gov.au/publication-detail/?id=6442467990&tab=2. 8. Begg S, Vos T, Barker B, Stevenson C, Stanley L, Lopez A: The burden of disease and injury in Australia 2003. Canberra: AIHW; 2007 http://www.aihw. gov.au/publication-detail/?id=6442467990&tab=2. g p 9. Mariani JJ, Cheng WY, Bisaga A, Sullivan M, Carpenter K, Nunes EV, Levin FR: Comparison of clinical trial recruitment populations: Treatment-seeking characteristics of opioid-, cocaine-, and cannabis-using participants. J Subst Abuse Treat 2011, 40(4):426–430. 31. van den Brink W: Evidence-based pharmacological treatment of subtance use disorders and pathological gambling. Curr Drug Abuse Rev 2012, 5:3–31. 32. Roxon N: More doctors and more nurses in National Health and Hospitals Network [media release].; 2010. 11 May 2010. http://www.health.gov.au/ internet/ministers/publishing.nsf/Content/mr-yr10-nr-nr092.htm. 10. Gates P, Taplin S, Copeland J, Swift W, Martin G: Barriers and facilitators to cannabis treatment. Drug Alcohol Rev 2011, 3:311–319. 11. Received: 5 June 2012 Accepted: 10 July 2012 Published: 24 July 2012 23. Venning P, Durie A, Roland M, Roberts C, Leese B: Randomise controlled trial comparing the cost-effectiveness of general practitioners and nurse practitioners in primary care. Br Med J 2000, 320:1048–1053. Author details 1 1National Cannabis Prevention and Information Centre, UNSW, PO Box 684, Randwick, NSW 2031, Australia. 2School of Psychology and Counselling, Institute of Health and Biomedical Innovation, Queensland University of Technology, Kelvin Grove, QLD 4059, Australia. 3Department of Psychiatry, Sydney Medical School, Level 5 Building 36, Royal North Shore Hospital, St Leonards, NSW 2065, Australia. 20. Geirsson M, Bendtsen P, Spak F: Attitudes of Swedish general practitioners and nurses to working with lifestyle change, with special reference to alcohol consumption. Alcohol Alcohol 2005, 40:388–393. 21. Lock CA, Kaner E, Lamont S, Bond S: A qualitative study of nurses' attitudes and practices regarding brief alcohol intervention in primary health care. J Adv Nurs 2002, 39:333–342. 22. Wilson I, Whiting M, Scammell A: Addressing cannabis use in primary care: GPs knowledge of cannabis-related harm and current practice. Prim Health Care Res Dev 2007, 8:216–225. Strengths and limitations MMN is a clinical psychologist and senior lecturer. PG is a doctoral student in Community Medicine and Public Health. PD is a drug and alcohol educator. DJK is a clinical psychologist and professor. RM is a general practitioner and senior lecturer. JC is a professor who has a PhD in Community Medicine and Public Health. The strength of the study is its large sample size. Limita- tions include the sole recruitment of health practitioners from Healthed seminars, reliance on self-reports of ser- vice provision, using a survey without established psy- chometric properties, and the low response rate (35%). This work was carried out at the National Cannabis Prevention and Information Centre at the University of New South Wales. All surveys were distributed during HealthEd seminars conducted in major cities across Australia. Page 10 of 10 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Norberg et al. Substance Abuse Treatment, Prevention, and Policy 2012, 7:31 http://www.substanceabusepolicy.com/content/7/1/31 Acknowledgements f 16. Aalto M, Pekuri P, Seppa K: Primary health care nurses' and physicians' attitudes, knowledge and beliefs regarding brief intervention for heavy drinkers. Addiction 2001, 96:305–311. The authors gratefully acknowledge the Department of Health and Aging (DoHA) for providing funding for the National Cannabis Prevention and Information Centre, and subsequently, the present study. DoHA had no role in study design or in the decision to submit the study for publication. DoHA also had no role in the collection, analysis and interpretation of data in the writing of the manuscript. The authors also acknowledge Ms Morag Millington, Julia Tropiano, and Lucy Albertella for their assistance with data collection and data entry. Early findings from this study were presented at the 2011 Australasian Professional Society on Alcohol and other Drugs Conference. 17. Jacka D, Clode D, Patterson S, Wyman K: Attitudes and practices of general practitioners training to work with drug-using patients. Drug Alcohol Rev 1999, 18:287–291. 18. Johnson TP, Booth AL, Johnson P: Physician beliefs about substance misuse and its treatment: Findings from a U.S. survey of primary care practitioners. Subst Use Misuse 2005, 40:1071–1084. 19. Amaral-Sabadini MB, Saitz R, Souza-Formigoni MLO: Do attitudes about unhealthy alcohol and other drug (AOD) use impact primary care professionals' readiness to implement AOD-related preventive care? Drug Alcohol Rev 2010, 29:655–661. References 1. United Nations Office on Drugs and Crime: World Drug Report 2011. Vienna; 2011 http://www.unodc.org/unodc/en/data-and-analysis/WDR-2011.html. 1. United Nations Office on Drugs and Crime: World Drug Report 2011. Vienna; 2011 http://www.unodc.org/unodc/en/data-and-analysis/WDR-2011.html. 24. Kinnersley P, Anderson E, Parry K, Clement J, Archard L, Turton P, Stainthorpe A, Fraser A, Butler CC, Rogers C: Randomised controlled trial of nurse practitioner versus general practitioner care for patients requesting "same day" consultations in primary care. Br Med J 2000, 320:1043–1048. p g y 2. Aldington S, Williams M, Nowitz M, Weatherall M, Pritchard A, McNaughton 2. Aldington S, Williams M, Nowitz M, Weatherall M, Pritchard A, McNaughton A, Robinson G, Beasley R: Effects of cannabis on pulmonary structure, function and symptoms. Thorax 2007, 62:1058–1063. 2. Aldington S, Williams M, Nowitz M, Weatherall M, Pritchard A, McNaughton A, Robinson G, Beasley R: Effects of cannabis on pulmonary structure, function and symptoms. Thorax 2007, 62:1058–1063. 2. Aldington S, Williams M, Nowitz M, Weatherall M, Pritchard A, McNaughton A, Robinson G, Beasley R: Effects of cannabis on pulmonary structure, A, Robinson G, Beasley R: Effects of cannabis on pulmonary structure, function and symptoms. Thorax 2007, 62:1058–1063. y p 3. Mittleman MA, Lewis RA, Maclure M, Sherwood JB, Muller JE: Triggering myocardial infarction by marijuana. Circulation 2001, 103:2805–2809. 3. Mittleman MA, Lewis RA, Maclure M, Sherwood JB, Muller JE: Triggering myocardial infarction by marijuana. Circulation 2001, 103:2805–2809. 25. Johansson K, Bendtsen P, Akerlind I: Early intervention for problem drinkers: Readiness to participate among general practitioners and nurses in Swedish primary health care. Alcohol Alcohol 2002, 37:38–42. 4. Taylor DR, Poulton R, Moffitt TE, Ramankutty P, Sears MR: The respiratory effects of cannabis dependence in young adults. Addiction 2000, 95:1669–1677. 4. Taylor DR, Poulton R, Moffitt TE, Ramankutty P, Sears MR: The respiratory effects of cannabis dependence in young adults. Addiction 2000, 95:1669–1677. 26. Australian Government Department of Health and Ageing: General Practice Statistics.; 2012. Archived at: http://www.webcitation.org/68wkrqPfI; 2012 [cited 2011 6-Jul-2012]. 5. Sherman MP, Campbell LA, Gong HJ, Roth MD, Tashkin DP: Antimicrobial and respiratory burst characteristics of pulmonary alveolar macrophaes recovered from smokers of marijuana alone, smokers of tobacco alone, smokers of marijuana and tobacco, and nonsmokers. Am Rev Respir Dis 1991, 144:1351–1356. 5. References Stinson FS, Ruan WJ, Pickering R, Grant B: Cannabis use disorders in the USA: prevalence, correlates and co-morbidity. Psychol Med 2006, 36:1447–1460. 33. Aalto M, Pekuri P, Seppa K: Implementation of brief alcohol intervention in primary health care: do nurses' and general practitioners' attitudes, skills and knowledge change? Drug Alcohol Rev 2005, 24:555–558. 34. Britt H, Miller GC, Charles J, Henderson J, Bayram C, Pan Y, Valenti L, Harrison C, Fahridin S, O'Halloran J: General practice activity in Australia 2008- 09. Canberra: AIHW; 2009. 12. Australian Bureau of Statistics: Health Services: Patient Experiences in Australia. Canberra: ABS; 2009. http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/ 4839.0.55.001Main+Features12009?OpenDocument. 35. Magin P, Horton G, Foster M, Girgis A: Response rates in GP surveys. Aust Fam Physician 2011, 40:427–430. 13. The Royal Australian College of General Practitioners: Putting Prevention into Practice: Guidelines for the implementation of prevention in the general practice setting. South Melbourne, Victoria: The Royal Australian College of General Practitioners; 2006. http://www.racgp.org.au/Content/ NavigationMenu/ClinicalResources/RACGPGuidelines/TheGreenBook/ RACGPgreenbook2nd.pdf. y 6. Shoemaker PJ, Eichholz M, Skewes EA: Item nonresponse: Distin b d ' k d f bl O 36. Shoemaker PJ, Eichholz M, Skewes EA: Item nonresponse: Distinguishing between don't know and refuse. Int J Publ Opin Res 2002, 14:193–201. doi:10.1186/1747-597X-7-31 Cite this article as: Norberg et al.: Screening and managing cannabis use: comparing GP’s and nurses’ knowledge, beliefs, and behavior. Substance Abuse Treatment, Prevention, and Policy 2012 7:31. 14. Crothers CE, Dorrian J: Determinants of nurses' attitudes toward the care of patients with alcohol problems. Nursing: ISRN; 2011. 14. Crothers CE, Dorrian J: Determinants of nurses' attitudes toward the care of patients with alcohol problems. Nursing: ISRN; 2011. 15. Kelleher S: Health care professionals' knowledge and attitudes regarding substance use and substance users. Accid Emerg Nurs 2007, 15:161–165. 15. Kelleher S: Health care professionals' knowledge and attitudes regarding substance use and substance users. Accid Emerg Nurs 2007, 15:161–165.
https://openalex.org/W4307816895	https://umu.diva-portal.org/smash/get/diva2:1792153/FULLTEXT01	English	null	Role of surface morphology on bed material activation during indirect gasification of wood	Fuel	2,023	cc-by	10,659	Role of surface morphology on bed material activation during indirect gasification of wood Robin Faust a,, Ali Valizadeh b, Ren Qiu c, Alyona Tormachen a, Jelena Maric d, Teresa Berdugo Vilches d, Nils Skoglund e, Martin Seemann d, Mats Halvarsson c, Marcus ¨Ohman b, Pavleta Knutsson a a Department of Chemistry and Chemical Engineering, Chalmers University of Technology, Kemigården 4, 412 96 Gothenburg, Sweden b Energy Engineering, Division of Energy Science, Luleå University of Technology, Luleå SE-971 87, Sweden c Chalmers University of Technology, Department of Physics, Kemigården 1, Gothenburg SE-41296, Sweden d Division of Energy Technology, Department of Space, Earth, and Environment (SEE), Chalmers University of Technology, Gothenburg 41296, Sweden e Thermochemical Energy Conversion Laboratory, Department of Applied Physics and Electronics, Umeå University, SE-901 87 Umeå, Sweden A R T I C L E I N F O Keywords: Fluidized bed Bed material Layer formation Olivine Feldspar Material characterization Olivine and alkali-feldspar were utilized in separate campaigns in an indirect dual fluidized bed gasification campaign with woody biomass as fuel. After three days, both bed materials were reported to be active towards tar removal and exhibited oxygen-carrying abilities and had formed an ash layer consisting of an outer ash deposition layer and an inner interaction layer. X-ray microtomography analysis concluded that a preferred deposition of ash happens onto convex regions of the bed particles, which results in an increase in thickness of the ash layer over convex regions. This effect is most pronounced for the outer layer which is a product of ash deposition. The inner layer exhibits a homogeneous thickness and is probably formed by interaction of Ca from the outer layer with the particles. Transmission electron microscopy revealed the presence of Fe and Mn on the surface of the particles in a solid solution with Mg. The oxygen-carrying effect which is found for aged particles is therefore attributed to the presence of Fe and Mn on the surface of aged particles. Alkali were found on the surface of both particles which are likely contributing to the catalytic activity of the material towards tar removal. the heat demand of the conversion by combustion of a part of the feedstock. This results in a product gas containing H2 and CO which can be applied for further chemical processes. To avoid dilution with com bustion products, the energy required for the endothermic gasification reaction can be provided in a dual fluidized bed (DFB) gasification reactor system. DFB gasification is a process consisting of two inter connected reactors which share the bed material while keeping the gases separated. In one reactor, fluidized bed combustion of biomass is con ducted where an oxygen-containing gas (e.g., air) is employed as the fluidizing gas. Hence, the bed material will be heated by the exothermic combustion reaction and conveyed by the gas stream into the second reactor where steam or CO2 is used for fluidization. In this reactor, the hot bed material supplies thermal energy which is required for the endothermic gasification reaction which converts the biomass into the product gas. i Corresponding author. E-mail address: faust@chalmers.se (R. Faust). Fuel 333 (2023) 126387 Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Fuel 333 (2023) 126387 Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Fuel 333 (2023) 126387 Contents lists available at ScienceDirect Available online 24 October 2022 0016-2361/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1016/j.fuel.2022.126387 Received 22 August 2022; Received in revised form 10 October 2022; Accepted 13 October 2022 Available online 24 October 2022 0016-2361/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction This study concluded that the composition of the ash layer varies, where a Ca-rich inner layer and a Mg-rich outer layer was found for the utilized bed materials. The outer layer had a thickness of 1–2 µm, which is close to the resolution limit of SEM which is why the resolution of SEM is therefore insufficient for detailed analysis of the outer layer. As the outer layer is in contact with the gas in the gasifier it is decisive for both catalytic tar conversion by the aged bed material as well as the oxygen-carrying ability the material exhibits after prolonged exposures. Another characterization technique which is mostly employed to analyze the particle layers is X-ray diffraction (XRD) which supplements the data from SEM with infor mation on the formation and rearrangement of crystallographic struc tures of the consisting phases. While XRD is useful to derive general information of the sample, this technique is not surface sensitive and therefore limited in its applicability to draw conclusions about the species which are in contact with the gas phase. In general, when woody biomass is utilized as the fuel, layer for mation proceeds by deposition of ash compounds which interact with the bed particles. For the case of quartz, the interaction between ash and bed particles could lead to the formation of alkali-silicates which have a low-melting point and therefore cause agglomeration of the particles [16]. On the other hand, olivine and feldspar are more resistant to agglomeration. For these materials, the interaction of deposited ash particles results in the formation of an interaction layer underneath the deposited ash, commonly denoted as an inner layer. The inner layer has a homogeneous appearance with a composition resembling both bed material and fuel ash. On the surface of the particles, the outer layer is heterogeneous consisting mainly of elements found in the fuel ash. However, ash layer characteristics are dependent on the bed material, fuel ash and the selected conversion process [17–18]. For feldspar, the developed inner layer comprises both Ca-Al-silicates and Ca-silicate while in the outer layer Ca3Mg(SiO4)2 and MgSiO3 could be detected [19–20]. In the inner layer formed around olivine bed particles, the formation of CaMgSiO4 was found [17,21]. In both cases, incorporation of higher amounts of Ca into the inner layer structure was associated with cracks in the inner layer perpendicular to the bed particle surface [17,19]. 1. Introduction [24] compared two different quartz beds in a floatation sep aration process. The authors observed that bed material with higher surface area has considerably larger capacity to adsorb sodium oleate due to higher wettability as well as Ca2+ and due to a larger number of adsorption sites. Similarly, the importance of surface morphology of solid particles was demonstrated, for example on the adsorption ca pacity of different silicates [25–26] or on the wettability of calcite particles [27]. Other studies with similar results have also portrayed the importance of surface roughness on the adsorption behavior of different solid surfaces [28–29]. Even though the cited studies have pointed on the influence of the surface morphology on the material behavior, there is still lack of understanding of how surface irregularities influence the interaction of the bed material with the formed fuel ash and further the development of surface activation (catalytic and/or oxygen carrying). Thus, the current study aims to contribute to filling this knowledge gap. In addition to the inherent properties of the minerals used as bed material, interactions between the bed material and the fuel ash influ ence their catalytic properties as well [11]. The ash forms a layer on the bed particles which decreases the concentration of tar in the produced gas. This effect has been extensively reported in the literature and was the focus of a recent review by Kuba et al. [12] For the case of woody biomass, the layer contains elements commonly found in woody biomass ash, which are mostly Ca, K, and Mg. Besides the impact of ash elements added through the fuel, olivine has shown an increase in activity due to different additives, such as salts of potassium [6,13] or calcium [14–15]. , yi g g g p The commonly employed characterization method for assessment of the surface layer is scanning electron microscopy (SEM). With SEM, it is possible to acquire information about composition and thickness of the layer of several particles, and conclusions about the bed inventory can be drawn by analyzing representative particles. In an earlier study, the formation of an outer layer was found for the used particles, which was only observable with SEM, when cross-sections of the particles were prepared with a more sophisticated sample preparation method, i.e., broad ion-beam milling (BIB) [18]. 1. Introduction Thermal conversion of biogenic fuels can play a major role in reducing fossil CO2-emissions and the dependence on fossil fuels. Emissions from thermal conversion of biomass can be considered CO2- neutral, as the CO2 is reabsorbed from the atmosphere into the biomass during the growth. Biomass is a complex feedstock and a suitable technique for its thermal conversion is based on the utilization of a fluidized bed reactor. Fluidized bed reactors have been assessed to be advantageous compared to other methods in thermal conversion of biomass as they offer higher fuel flexibility [1]. The technology is based on the utilization of fine sand-like particles which are fluidized by a gas stream. This provides a more uniform heat distribution throughout the reactor and improved conversion efficiency. li i In fluidized bed gasification of biomass, the oxygen supply to the fuel during the thermal conversion is limited to the level sufficient to cover One major challenge in biomass gasification is the presence of major challenge in biomass gasification is the presenc Fuel 333 (2023) 126387 R. Faust et al. polyaromatic hydrocarbons (tar) as side products of the gasification reaction which can cause problems by clogging of the down-stream equipment. Earlier attempts to reduce the concentration of the pro duced tar in the effluent gas, utilizing a bed material which exhibits catalytic properties has been proven successful [2]. Fluidized beds are most commonly operated with quartz sand (SiO2); however, quartz does not show catalytic impact towards tar reduction. This is why alternative bed materials, such as dolomite [CaMg(CO3)2] [2–4], olivine [(Mg, Fe)2SiO4] [5–7] and feldspar [(K, Na)AlSi3O8] [8–10] have been examined in fluidized bed gasification. Of those, dolomite has the highest impact in tar reduction. However, as it is comparably softer, the attrition rate will be higher, accordingly [3–4]. On the other hand, feldspar and olivine have shown beneficial behavior towards tar removal without causing problems associated with attrition [7–8]. chemical interactions between fuel ash and bed material. Nevertheless, physical properties of the bed particles (e.g., surface morphology) in fluence the ash layer formation mechanism and consequently alter performance of the bed material in the process. For instance, crack layers in quartz bed particles have been aligned with irregularities over the bed particle surface [23]. Furthermore, the surface morphology has been shown to have a significant impact on surface adsorption processes. Zhu et al. Table 1 2.1. Experimental campaigns 1. Introduction Table 1 Chemical composition of the investigated bed materials given in mol per centage on an oxygen free basis, according to the suppliers. mol % Olivine Feldspar Na 7.56 Mg 60.48 0.05 Al 0.44 20.10 Si 34.11 61.25 P 0.06 K 9.72 Ca 1.17 Ti 0.01 Cr 0.20 Fe 4.55 0.08 Ni 0.21 1. Introduction Apart from the catalytic activity, studies investigating the properties of used bed material have shown the development of oxygen-carrying abilities [7,22]. For instance, in DFB gasification, the particles trans port oxygen from the combustion side to the gasification side. Although this will decrease the overall calorific value of the product gas, as H2 and CO could be oxidized to H2O and CO2, respectively, the development of oxygen-carrying ability in DFB gasification has been observed to be connected to tar reduction in the product gas [7]. However, it has been shown in a recent study by Pissot et al. [22], that the two effects (oxygen carrying and catalytic activity) are independent, as the utilization of oxygen-carrying bed material (e.g., ilmenite), did not result in lower of the tar yield compared to used feldspar, despite ilmenite’s higher oxygen-carrying ability. Regardless of whether the tar decrease is due to catalytic or oxygen-carrying activity or a combination thereof, in either case, surface accumulated species have been considered to influence the properties of the bed material. Thus, it is essential for practical purposes to examine the outermost surface of the particles in more details in order to comprehend the properties of the bed material and how they could be affected by prolonged exposure times. In this study, the discussed shortcomings of SEM and XRD are resolved by the application of transmission electron microscopy (TEM). TEM is utilized to acquire higher resolution images along with the elemental and phase distribution information from the outermost layer around olivine and feldspar bed particles. This is decisive to list out which elements are in contact with the gas phase in the reactors and therefore participate in the previously listed characteristics of aged bed material, i.e., catalytic activity and oxygen-carrying ability. Addition ally, TEM diffraction offers the possibility to obtain surface sensitive crystallographic information. Furthermore, X-ray microtomography (XMT) was employed to investigate the effect of surface morphology on the bed particle layer formation and to obtain a 3D understanding on how the layer is distributed over the bed particle surface. Apart from novel knowledge, the present study demonstrates the importance to use the different methods complementary to each other and derive conclu sions from similarities that are observed. Most of the aforementioned studies that investigated the activity of the bed materials as a results of layer formation, have focused on the 2 Fuel 333 (2023) 126387 R. Faust et al. 2.1. Experimental campaigns This study investigates olivine and feldspar which were utilized as bed materials. Their elemental composition is shown in Table 1 and the particles size distribution in Fig. 2. The fresh bed material is mined as rocks and crushed to the desired particle size. Olivine is a solid solution of Mg2SiO4 (forsterite) and Fe2SiO4 (fayalite), corresponding to Mg1.86Fe0.14SiO4. The feldspar used for this study is an alkali-feldspar and consists of 48 % K-feldspar (KAlSi3O8), 40 % Na-feldspar (NaAl Si3O8), 6 % Ca-feldspar (CaAl2Si2O8), and 6 % quartz (SiO2). Each bed material was utilized in a separate experimental campaign conducted in the Chalmers dual fluidized bed gasifier (see Fig. 1), which consists of a 2–4 MWth bubbling fluidized bed gasifier (6) and a 12 MWth circulating fluidized bed combustor (1). The temperature ranges between 800 and 850 ◦C. Woody biomass was utilized as fuel; 2 t/h of wood chips were fed to the combustor and 300 kg/h of wood pellets were fed to the gasifier. Fig. 1. Schematic representation of the Chalmers DFB gasifier [1]. Fig. 1. Schematic representation of the Chalmers DFB gasifier [1]. Fuel 333 (2023) 126387 R. Faust et al. Fig. 2. Particle size distribution of the fresh materials according to the suppliers. a small fraction of the layer on one particle can be analyzed. The different techniques are therefore required complementary to each other. The statistics in analysis as limitation in sample characterization is exemplified in Fig. 3. Table 2 Ash composition of the feedstock (dry fuel basis) used during the experimental campaigns from which the bed material samples were extracted. The wood chips (0.5 wt% ash) were fed to the furnace, the wood pellets (0.4 wt% ash) were fed to the gasifier. i Wood Chips [mmol/kg] Wood Pellets [mmol/kg] Na 1.3 2.2 Mg 7.0 9.1 Al 1.5 0.7 Si 5.7 2.8 P 2.3 3.2 K 10.2 21.0 Ca 25.0 32.4 Ti <0.2 <0.2 Mn 2.2 1.1 Fe 0.7 0.4 Ba 0.1 0.1 2.3. X-ray microtomography Analysis. X-ray Microtomography (XMT) utilizes X-rays to produce 2D image cross sections of an object, which can be combined to recreate a 3D image of the probed material. Images produced through XMT portray optical density of the scanned object meaning that the image is con structed from relatively darker and brighter areas which resemble lower The ash composition of the fuel is shown in Table 2. The bed samples were extracted after 3 days residence time from the loop seal located after the gasifier (7). No addition of fresh bed material was performed during the experimental campaign in order to ensure the specified age of the bed particles. Note that in the case of the experiment with olivine, a solution of ammonium sulfate was added to the cyclone to control the emissions of CO. A summary of the exposure conditions is given in Table 3. Fig. 3. Estimated fraction of sample (log-scale) usually analyzed with the different characterization techniques. In the Chalmers semi-industrial DFB system, about 4 tons of bed material are circulating. Of the samples taken from the bed inventory, only about 1 – 10 g of sample is usually used for XRD analysis. Using the commonly employed epoxy embedding method for the preparation of a sample investigated with SEM, around 100 – 1000 particles can be analyzed (~1 – 10 mg); even fewer particles (around 10) can be pre pared and analyzed when the previously mentioned BIB method is employed. The number of particles analyzed with XMT is comparable with SEM, where XMT offers the advantage of 3D imaging. TEM offers the possibility of higher resolution, however, with the trade-off that only Fig. 3. Estimated fraction of sample (log-scale) usually analyzed with the different characterization techniques. 2.2. Electron microscopy characterization To reveal cross-sections of the bed particles, broad-ion beam (BIB) milling was employed where around 10 particles for each bed material were fixated separately between two silicon wafers and subsequently polished with a Leica EM TIC 3X. The particles were subsequently analyzed by scanning electron microscopy (SEM) with the back- scattered electron (BSE) signal. The BSE signal strength depends on the average atomic number of the investigated location, where heavier elements result in a stronger signal (i.e., brighter contrast). A FEI Quanta 200 FEG SEM was used which was equipped with an energy dispersive X-ray spectroscopy (EDS) system. 15 kV acceleration voltage was applied for imaging and EDS analysis and the low vacuum mode was utilized with 20 Pa water as charge neutralizing gas. Fig. 2. Particle size distribution of the fresh materials according to the suppliers. More detailed microstructures were studied by (scanning) trans mission electron microscopy (STEM/TEM) on the thin foil samples, using an FEI Titan 80–300 TEM/STEM instrument. The thin foil samples were prepared using an FEI Versa3D focused ion beam – scanning electron microscopy (FIB – SEM) instrument equipped with an Omni- probe micromanipulator. In the TEM mode, bright field (BF) imaging and selected area electron diffraction (SAED) were performed with a Gatan US1000 charge-coupled device (CCD) camera. STEM imaging was performed with a BF detector and a high angle annular dark field (HAADF) detector. EDS data was collected in STEM mode using an Oxford X-sight detector. Table 3 Summary of the exposure conditions of the two materials investigated in this study. The materials were exposed to the same environment except for the addition of ammonium sulfate to control CO emissions in the case of olivine. Bed Material Formula Combustor Gasifier Interval Additives Fuel Rate [t/h] Fuel Rate [kg/h] Olivine Mg1.86Fe0.14SiO4 Wood Chips 2 Wood Pellets 300 3 days Ammonium Sulfate Feldspar (K, Na)AlSi3O8 Wood Chips 2 Wood Pellets 300 3 days None Fig. 4. Terms used in this study to describe the different layers. R. Faust et al. Fig. 4. Terms used in this study to describe the different layers. R. Faust et al. R. Faust et al. Fuel 333 (2023) 126387 Fig. 4. Terms used in this study to describe the different layers. Raw data from XMT characterization were exported to DragonFly software (version 2020.2 build 941), where X-ray scans from several planes of the sample could be combined to create a 3D image of the entire sampling tube. Contrast difference (Fig. 9 left) and pre-installed color filters (Fig. 9 right) were used to distinguish difference in den sity phases where each color denotes a certain optical intensity in the studied object. To gauge the bed layer thickness around over the bed particle surface, a pre-installed thickness measurement feature in the software was utilized. and higher densities, respectively. Moreover, other physical properties of the sample including surface area, volume and volume fraction of different phases could also be measured. A plastic tube with 1 mm diameter was used as the sample holder. 2 mm of the tube’s height was filled with the bed particles and left for 16 h in order to neutralize the sample from any possible electric charges. Feldspar and olivine samples were separately inserted into the XMT machine (Xradia 620 Versa) with the optical resolution of 0.8 μm which provided a field of view covering around 15 bed particles. The scanning was performed with an X-ray tube voltage of 40 kV and 3 W tube power with a LE3 X-ray filter from Edmund industrial optics. During each scan, 2001 projection radiograph images were acquired over 360◦sample rotation with 10- and 9-seconds exposure time for olivine and feldspar, respectively. 2.4. Batch reactors The oxygen-carrying ability of the fresh and used bed material was Fig. 5. Back-scattered electron micrographs of the particles before and after exposure. The fresh particles exhibit rugged features and sharp edges, which disappear after 3 days residence time in the gasifier. A and B indicate locations with difference in morphology that have been formed as a result of introduction of additives to the process. Fig. 5. Back-scattered electron micrographs of the particles before and after exposure. The fresh particles exhibit rugged features and sharp edges, which disappear after 3 days residence time in the gasifier. A and B indicate locations with difference in morphology that have been formed as a result of introduction of additives to Fig. 5. Back-scattered electron micrographs of the particles before and after exposure. The fresh particles exhibit rugged features and sharp edges, which disappear after 3 days residence time in the gasifier. A and B indicate locations with difference in morphology that have been formed as a result of introduction of additives to the process. 5 5 R. Faust et al. Fuel 333 (2023) 126387 Fig. 6. BSE overview and elemental intensity maps acquired by SEM-EDS on the surface of olivine particles which have remained in the gasifier for 3 days. A phase rich in K and S (probably K2SO4) is deposited onto the edges of the particles. This phase has most likely formed to due to the addition of ammonium sulfate in the case of olivine. Fig. 6. BSE overview and elemental intensity maps acquired by SEM-EDS on the surface of olivine particles which have remained in the gasifier for 3 days. A phase rich in K and S (probably K2SO4) is deposited onto the edges of the particles. This phase has most likely formed to due to the addition of ammonium sulfate in the case of olivine. ntensity maps acquired by SEM-EDS on the surface of olivine particles which have remained in the gasifier for 3 days. A phase posited onto the edges of the particles. This phase has most likely formed to due to the addition of ammonium sulfate in the case investigated in a laboratory-scale batch fluidized bed reactor using 15 g of the bed particles in each experiment. The gas stream applied for fluidization was alternated between oxidizing, inert, and reducing. The reactor was electrically heated and maintained at 850 ◦C during the test. 2.4. Batch reactors More details about the laboratory set-up can be found elsewhere [30]. Each bed sample was exposed first to an inert atmosphere (N2) for 300 s and subsequently oxidized with 5 % O2 in N2 for 600 s to fully oxidize the elements which can contribute to the oxygen-carrying ability of the material. The remaining O2 was purged with another inert (N2) in the next step for 300 s after which the reduction step was conducted. The reduction was done with syngas (50 % CO and 50 % H2) which is representative of the atmosphere during biomass gasification. The dif ference in O2 entering and leaving the reactor during the oxidizing step (after the material had been exposed to the reducing atmosphere) was Fig. 7. 3D XMT image of the overall bed layer thickness of a bed particle that after 3 days residence time in the gasifier. The layer thickness measurement is shown (right) and compared to the XMT image from the same bed particle (left) for (a) feldspar and (b) olivine. Fig. 7. 3D XMT image of the overall bed layer thickness of a bed particle that after 3 days residence time in the gasifier. The layer thickness measurement is shown (right) and compared to the XMT image from the same bed particle (left) for (a) feldspar and (b) olivine. Fig. 7. 3D XMT image of the overall bed layer thickness of a bed particle that after 3 days residence time in the gasifier. The layer thickness measurement is shown (right) and compared to the XMT image from the same bed particle (left) for (a) feldspar and (b) olivine. 6 R. Faust et al. Fuel 333 (2023) 126387 Fig. 8. Schematic representation of the possible layer formation locations. Fig. 8. Schematic representation of the possible layer formation locations. groups, the terms used in the current study are defined in Fig. 4. measured to obtain comparable results on the oxygen-carrying ability of the material. measured to obtain comparable results on the oxygen-carrying ability of the material. i The term “ash layer” is used here as the sum of all layers around the bed particles formed due to interaction and deposition of ash compounds. 3.1. Morphology Both bed materials are mined as rocks and crushed to the desired particles size prior to utilization in the gasifier. Therefore, they exhibit sharp edges in their fresh state (see Fig. 5, left). After three days of exposure time, the particles appear to be much smoother, which was attributed to either attrition [6,15,31] and ash layer formation [8,14,32] or both [19–21] in previous studies. While the aged feldspar exhibits quite uniform BSE contrast, aged olivine exhibits two features of different contrast (indicated with A and B in Fig. 5). Region A is pref erentially located on the edges of the particles, whereas B is located mostly on the particles’ sides. An EDS intensity mapping (shown in Fig. 6) reveals that region A contains significantly higher concentrations of K and S compared to region B. As this behavior was found only in olivine but not for feldspar, the difference in exposure history (i.e., the addition of ammonium sulfate) could be attributed to the formation of K- and S-rich regions. Therefore, addition of ammonium sulfate can be attributed to cause deposition of a K- and S-rich phase. EDS point The surface roughness was quantified with XMT by measuring the total bed volume and the total bed surface. Thereby the specific surface area for both bed samples through the entire sampling tube was calcu lated and represented in Table 4. Comparing the specific surface areas of both bed materials reveals that the same volume of feldspar exposes ~50 % more surface area than olivine which means that feldspar Fig. 10. Back-scattered electron micrographs of typical ash layers formed on feldspar (a) and olivine (d). EDS line-scans (shown in atomic %) were recorded on a convex and a concave location in the ash layer. The normalized elemental compositions are shown in (b) and (c) for feldspar and (e) and (f) for olivine. The two dashed grey lines indicate the location of the “inner” layer. The cyan line in (e) indicates the location of the K- and S-rich layer. The inset shows the concentration of minor elements on the topmost 1 µm; in figure (e) underneath the K- and S-rich layer (i.e., between 1 and 2 µm from the surface). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 10. bed particle. bed particle. The “bulk” describes the bed particle with a composition similar to the fresh bed material. 3. Results and discussion Table 4 Table 4 Physical properties of the bed materials measured with XMT. Physical properties of the bed materials measured with XMT. Bed material total bed sample volume (mm3) total bed sample surface (mm2) specific surface (mm2/mm3) Olivine 0.48 12.40 25.83 Feldspar 0.46 18.74 40.74 Bed material total bed sample volume (mm3) total bed sample surface (mm2) specific surface (mm2/mm3) Olivine 0.48 12.40 25.83 Feldspar 0.46 18.74 40.74 XMT allowed to confirm the smoothening which both samples exhibited, and attribute it to layer formation. Further, XMT analysis allowed to appreciate the layer thickness formed around the individual bed material particles. Fig. 7 shows the layer thickness measurement result for typical feldspar and olivine bed particles, respectively. A color graphed image for the layer thickness is represented side by side to the original bed particle in the figure so the layer thickness over different surface morphologies of the bed particle could be observed. Comparing the two figures reveals that the average layer thickness is higher on olivine than on feldspar. Yet in both cases, there is a noticeable differ ence in the layer thickness upon the convex and concave areas over the bed particle surface. Based on the differences in the surface morphology, three different possibilities for layer build-up have been identified. The three possibilities of layer formation depending on surface morphology are schematically shown in Fig. 8. From the XMT analysis, it can be seen that the convex shaped regions possess a thicker layer meaning that there is a higher tendency towards layer formation in these areas (rep resented by mechanism “A” in Fig. 8). This effect can be observed in more detail on the cross-sectional views, shown in Fig. 9. Furthermore, it could also be observed that the frequency of the convex and concave areas is higher in the case of feldspar. This was attributed to the higher surface roughness which leads to less uniformity in the layer thickness around the feldspar bed particles compared to olivine. i 2.5. Nomenclature The “outer layer” is used to refer to the layer commonly described as inhomogeneous and formed by ash deposition. A multitude of studies exist which investigate and describe the layers formed by the interaction of different bed materials with different fuels. To facilitate a comparison between publications from different research The “inner layer” is used to refer to the layer commonly described as homogeneous and formed by the interaction of the deposited ash and the Fig. 9. XMT measured density distribution in the bed particle layer (right) compared to the 2D cross-sectional view of the exact same plane (left) for a typical (a) feldspar and (b) olivine bed particle. The measured relative optical densities were normalized where 0 represents the lowest detected density and 1 represents regions with the relatively highest density in the analyzed sample. Fig. 9. XMT measured density distribution in the bed particle layer (right) compared to the 2D cross-sectional view of the exact same plane (left) for a typical (a) feldspar and (b) olivine bed particle. The measured relative optical densities were normalized where 0 represents the lowest detected density and 1 represents regions with the relatively highest density in the analyzed sample. 7 Fuel 333 (2023) 126387 R. Faust et al. analysis within this location showed a molar ratio of K:S of about 2 which agrees with the formation of K2SO4, which was also found with XRD in the previous study of the material [18]. i Table 4 Physical properties of the bed materials measured with XMT. 3.1. Morphology Back-scattered electron micrographs of typical ash layers formed on feldspar (a) and olivine (d). EDS line-scans (shown in atomic %) were recorded on a convex and a concave location in the ash layer. The normalized elemental compositions are shown in (b) and (c) for feldspar and (e) and (f) for olivine. The two dashed grey lines indicate the location of the “inner” layer. The cyan line in (e) indicates the location of the K- and S-rich layer. The inset shows the concentration of minor elements on the topmost 1 µm; in figure (e) underneath the K- and S-rich layer (i.e., between 1 and 2 µm from the surface). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 8 R. Faust et al. Fuel 333 (2023) 126387 Fig. 11. Overview TEM micrographs of the two samples investigated. (a) olivine, (b) feldspar. The location of the EDS-mapping (as shown in Fig. 12) and the EDS line-scan (as shown in Fig. 13) are indicated in red. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 11. Overview TEM micrographs of the two samples investigated. (a) olivine, (b) feldspar. The location of the EDS-mapping (as shown in Fig. 12) and the EDS line-scan (as shown in Fig. 13) are indicated in red. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 11. Overview TEM micrographs of the two samples investigated. (a) olivine, (b) feldspar. The location of the EDS-mapp ine-scan (as shown in Fig. 13) are indicated in red. (For interpretation of the references to color in this figure legend, the re his article.) this article.) 1–2 µm [12,17–21,32–35]. This layer can be found on both convex and concave locations and is indicated as the location between the two dashed grey lines in the EDS line-scans shown in Fig. 10. The constant thickness and composition suggest that this layer is growing inwards into the original bed particle [34]. However, the materials differ regarding the formation of the outer layer. This layer is usually described as inhomogeneous in thickness and associated with the deposition of ash particles and therefore often resembles the fuel ash composition [23,32–33,36–38]. 3.1. Morphology It can be speculated that the ash- derived Ca migrates from the deposited outer layer into the original bed particle, thereby forming the homogeneous inner layer. For the case of the convex layer on olivine, the previously discussed deposition of a K- and S-rich phase (most likely K2SO4) is indicated with the cyan line. The difference in curvature of the layer appears to be related to the thickness of the outer layer: while the inner layer is almost homogeneous in thickness (following mechanism “B” in Fig. 8), the outer layer is exhibits relatively higher surface roughness. However, the volume fraction of the total layer volume around the bed particles was measured to be 0.078 for olivine and 0.049 for feldspar showing ~50 % thicker bed layer formation for the olivine. This can be decisive for the catalytic activity of the bed material as the surface area of aged bed material can have a stronger impact on their catalytic activity towards the water–gas shift reaction than the concentration of ash elements [9]. 1–2 µm [12,17–21,32–35]. This layer can be found on both convex and concave locations and is indicated as the location between the two dashed grey lines in the EDS line-scans shown in Fig. 10. The constant thickness and composition suggest that this layer is growing inwards into the original bed particle [34]. However, the materials differ regarding the formation of the outer layer. This layer is usually described as inhomogeneous in thickness and associated with the deposition of ash particles and therefore often resembles the fuel ash composition [23,32–33,36–38]. It can be speculated that the ash- derived Ca migrates from the deposited outer layer into the original bed particle, thereby forming the homogeneous inner layer. For the case of the convex layer on olivine, the previously discussed deposition of a K- and S-rich phase (most likely K2SO4) is indicated with the cyan line. The difference in curvature of the layer appears to be related to the thickness of the outer layer: while the inner layer is almost homogeneous in thickness (following mechanism “B” in Fig. 8), the outer layer is Fig. 12. TEM dark-field micrographs of (a) olivine and (b) feldspar, recorded at the locations indicated with the red box in Fig. 11. An EDS elemental intensity map is shown below which was recorded at the locations indicated inside the red box. The numbers illustrate the similar ities in layer composition. layer 1 is rich in alkali or alkaline earth metals; layer 2 contains transition metals (Fe and Mn) and Mg; layer 3 is rich in Si and P. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 3.2. Elemental composition The analysis with XMT has shown the importance of the morphology of the layer, where thicker ash layers are formed on convex locations. To compare the elemental composition on convex and concave locations, EDS line-scans were recorded on both materials and the results are shown in Fig. 10. Both materials exhibit the typical formation of a ho mogeneous inner Ca-rich layer which has a constant thickness of around 9 Fig. 12. TEM dark-field micrographs (a) olivine and (b) feldspar, recorded the locations indicated with the red b in Fig. 11. An EDS elemental intens map is shown below which was record at the locations indicated inside the r box. The numbers illustrate the simil ities in layer composition. layer 1 is ri in alkali or alkaline earth metals; laye contains transition metals (Fe and M and Mg; layer 3 is rich in Si and P. (F interpretation of the references to co in this figure legend, the reader referred to the web version of th article.) 9 Fig. 12. TEM dark-field micrographs of (a) olivine and (b) feldspar, recorded at the locations indicated with the red box in Fig. 11. An EDS elemental intensity map is shown below which was recorded at the locations indicated inside the red box. The numbers illustrate the similar ities in layer composition. layer 1 is rich in alkali or alkaline earth metals; layer 2 contains transition metals (Fe and Mn) and Mg; layer 3 is rich in Si and P. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 12. TEM dark-field micrographs of (a) olivine and (b) feldspar, recorded at the locations indicated with the red box in Fig. 11. An EDS elemental intensity map is shown below which was recorded at the locations indicated inside the red box. The numbers illustrate the similar ities in layer composition. layer 1 is rich in alkali or alkaline earth metals; layer 2 contains transition metals (Fe and Mn) and Mg; layer 3 is rich in Si and P. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 12. TEM dark-field micrographs of (a) olivine and (b) feldspar, recorded at the locations indicated with the red box in Fig. 11. 3.2. Elemental composition An EDS elemental intensity map is shown below which was recorded at the locations indicated inside the red box. The numbers illustrate the similar ities in layer composition. layer 1 is rich in alkali or alkaline earth metals; layer 2 contains transition metals (Fe and Mn) and Mg; layer 3 is rich in Si and P. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fuel 333 (2023) 126387 line-scans at the outermost part of the layer at convex locations indicated as red squares in Fig. 11. (a). olivine, (b). feldspar. The data is normalized and a C- and O-free basis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) R. Faust et al. Fuel 333 (2023) 126387 Fig. 13. EDS line-scans at the outermost part of the layer at convex locations indicated as red squares in Fig. 11. (a). olivine, (b). feldspar. The data is normalized and presented on a C- and O-free basis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) more details. For the case of olivine, a location free of the K- and S-rich layer was selected, to investigate similarities between the two materials independent of their exposure history. thicker on the convex areas (mechanism “A”). The deposition of K2SO4 occurred preferentially on the convex location which agrees with the previously discussed stronger tendency of layer formation on convex locations. The morphology is therefore decisive for the formation of the outer layer, which is associated with deposited ash-derived elements. Overviews of the olivine and the feldspar samples are shown in Fig. 11 (a) and (b), respectively. As it was presented in our previous study, [18] olivine exhibits an ash layer of around 4 – 7 µm thickness after 3 days of interaction with woody biomass ash. The lift-out was done on the ash layer, where only a small part of the bulk olivine can be observed in the bottom left corner of the micrograph [Fig. 11 (a)]. The homogeneous inner layer of around 1–2 µm thickness can be seen adjacent to the bulk olivine (as indicated with the dashed blue line). Above the inner layer, a heterogeneous outer layer was observed. 3.2. Elemental composition Again, this agrees with the previously listed studies on the interactions between olivine and woody biomass ash. The higher resolution achieved by the current analysis enables the differentiation of several features within the inhomogeneous outer layer. An outermost layer, brighter in contrast, which was found previously [18] is visible and covers the entire sample. However, contrary to our previous results, another darker layer can be detected on the sample. The feldspar sample exhibits an inner layer with around 2 – 3 µm thickness after 3 days of exposure, which is also similar to our previous study [18]. Again, an outer layer could be observed as a line brighter in contrast. Both samples were further investigated with Generally, around 50–70 at. % Mg can be found in the outer layer as well as 10 at. % Ca and 5 at. % Mn. Al is almost absent in both the fuel ash and the outer layer, which supports the suggested formation of the outer layer by ash deposition. Unlike olivine, around 10 at. % Na can be found on the surface of feldspar which suggests a driving force of diffusion of Na from the bulk feldspar towards the surface. Similarly, a small increase in Fe appears to be present on the surface of convex olivine. Together with the Mn found on both samples, this could contribute to the oxygen-carrying ability of the bed material. TEM-EDS which is shown in Fig. 12 and Fig. 13. The diffrac tion pattern recorded on the location of layer 3 (which contains high concentrations of Si and P) revealed the presence of an amorphous phase. This could influence the capacity of the layer to retain small ash particles, as an amorphous phase is likely to be sticky. The presence of a sticky phase on convex layers could have aided deposition of ash- derived particles and therefore amplified the layer formation on these locations. Its presence on the surface of feldspar suggests that the Mg originates from the biomass ash and not from the bed material itself. The diffrac tion pattern recorded on the location of layer 3 (which contains high concentrations of Si and P) revealed the presence of an amorphous phase. This could influence the capacity of the layer to retain small ash particles, as an amorphous phase is likely to be sticky. The presence of a sticky phase on convex layers could have aided deposition of ash- derived particles and therefore amplified the layer formation on these locations. The presence of a layer consisting of Mg, Mn, and Fe can be observed for both materials underneath layers 1a and 1b. This layer is denoted as 2 in Fig. 12 and Fig. 13. The migration of Fe to the surface of olivine particles is a known phenomenon in the literature [2,15,44] and the formation of a Mg-rich surface layer has also been discovered previously [18,44]. According to Berguerand et al. [8] where the process parame ters of the feldspar used for this study are investigated, the bed material develops an oxygen-carrying ability at later stages of the experiment. As a consequence, the concentration of H2 in the syngas and the heating value of the product gas will be decreased which is not desirable in the gasification process. The coincidence of Fe together with CaO or MgO was found to have synergistic effects regarding both oxygen-carrying ability of the bed material and its catalytic activity [42,45–46]. 3.4. Diffraction To identify the crystallographic phases which are present at the surfaces of the particles, an attempt was made to record and assign the electron diffraction patterns obtained with TEM. Due to the complexity of the elemental composition of the surface layer, this is a challenging endeavor. The layers indicated with 1a were too thin for both samples to record any diffraction patterns. The phases that were identified in the respective layers are shown in Table 5. This list is not exhaustive and should be treated as a result of some of the phases which have formed on the surface. The diffraction patterns and the matched phases can be found in the supplementary material. CaCO3 was found on the surface of olivine which agrees with the elemental composition shown in Fig. 13, where the data is presented on a C- and O-free basis. The presence of CaCO3 can be beneficial for the catalytic activity of the material and was found by Kuba et al. [44] with Raman spectroscopy during surface analysis of ash-layered olivine. On the oxygen-carrying layer (denoted 2), Mn2O3 was a possible match with the recorded pattern. Mn2O3 can form a solid solution with Fe2O3 and is therefore a possible phase corresponding to the elemental composition measured in layer 2. TEM-EDS which is shown in Fig. 12 and Fig. 13. TEM-EDS which is shown in Fig. 12 and Fig. 13. Fig. 12 (a) and (b) portray EDS elemental intensity maps for olivine and feldspar, respectively. The location selected for olivine was at the previously mentioned darker phase, which is denoted as 1b in Fig. 12. As can be seen from the intensity maps, this layer mainly contains Ca and O. This was confirmed with the EDS line-scan (Fig. 13), where the results are normalized to a C- and O-free basis. The presence of Ca on the sur face is commonly found to improve the catalytic activity of bed mate rials [8,14–15,39–40]. However, both map and line-scan reveal the presence of yet another layer above this Ca-rich layer which exhibits a high concentration of K. In Fig. 12 (a), this layer is denoted as 1a. The presence of K on the surface of olivine particles was found previously by Knutsson et al. [6] and the interplay of Ca and K was suggested to be responsible for the observed catalytic activity towards reduction of tars. A similar layer consisting of alkali was found for the case of feldspar, where both K and Na were found to be present. Alkali are known to decrease the tar concentration in the product gas [8,41–43] which is why their presence at the surface of the particles could be a possible explanation of the observed catalytic activity of the investigated particles. Fig. 14. Oxygen-carrying ability of the materials measured from the difference in O2 entering and leaving the reactor during the oxidizing step. The materials were investigated as obtained (“fresh”) and after 3 days residence time in the gasifier (“aged”). was matched with both periclase (MgO) and hematite (Fe2O3). Hematite is a known catalyst for the water–gas shift reaction [42] and was also found previously on used olivine. [44] Periclase is also commonly found with XRD on used olivine, [14,18,21] but it was not found on feldspar. was matched with both periclase (MgO) and hematite (Fe2O3). Hematite is a known catalyst for the water–gas shift reaction [42] and was also found previously on used olivine. [44] Periclase is also commonly found with XRD on used olivine, [14,18,21] but it was not found on feldspar. Its presence on the surface of feldspar suggests that the Mg originates from the biomass ash and not from the bed material itself. 3.5. Oxygen-carrying ability The oxygen-carrying (OC) ability of the materials in their fresh and active state was measured in the batch reactor and the results are shown in Fig. 14. The materials are considered active after 3 days of residence time in the system, as the tar concentration was significantly decreased at that stage [7–8,31]. Comparing the two materials in their fresh state reveals that olivine exhibits an inherent OC ability which is higher than fresh feldspar. This is expected due to the Fe-content in olivine which is able to change its oxidation state from Fe3+ at high oxygen partial pressure to Fe2+ at lower oxygen partial pressure and thereby transport oxygen. However, in its active state, feldspar exhibits more than twice as much OC as fresh olivine, which can be attributed to the transition metals (i.e., Fe and Mn) found on the surface of aged feldspar. The re sults of the OC ability of the two materials agrees with the previously discussed Fe- and Mn-rich layer (layer 2). The oxygen-carrying ability of aged olivine is higher than for aged feldspar. The migration of Fe from the core of the olivine is a possible explanation for the significantly higher OC ability of olivine to feldspar due to the driving force for Fe to diffuse to the surface. While the impact of S cannot be disregarded regarding the oxygen-carrying effect, feldspar exhibits a significant OC ability although no S was found on the surface. This supports the sug gested role of transition metals (Fe and Mn). Table 5 Fig. 14. Oxygen-carrying ability of the materials measured from the difference in O2 entering and leaving the reactor during the oxidizing step. The materials were investigated as obtained (“fresh”) and after 3 days residence time in the gasifier (“aged”). 3.3. Surface analysis To investigate the surface of the particles in more details, a TEM lamella was lifted out from one of the particles. The choice of location for the lift-out was made based on the previous results acquired with SEM and XMT. A convex location was chosen for both samples, to investigate the underlying reason for the amplified layer formation in 10 Table 5 Phases identified with TEM diffraction on the layers denoted according to Fig. 13. Sample Layer Phase Olivine 1b CaCO3 Olivine 2 Mn2O3 Feldspar 2 MgO and Fe2O3 Feldspar 3 Amorphous R. Faust et al. Fuel 333 (2023) 126387 R. Faust et al. R. Faust et al. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Thicker ash layers were formed on convex areas, which impacts especially the inhomogeneous outer layer which is formed due to deposition of ash particles. Data will be made available on request. Acknowledgements The deposition of K2SO4 which formed due to the addition of ammonium sulphate appeared preferentially on convex areas. We would like to thank the Swedish Energy Agency project nr 50450- 1. A.V., M.¨O., and N.S. gratefully acknowledge financial support from the Swedish Energy Agency project no. P46533-1. The inner Ca-rich layer has a constant thickness on convex and concave areas and is therefore likely formed due to the interaction of the deposited outer layer with the bed material particle. Both materials exhibit a surface layer containing alkali and alkaline earth metal elements which is in line with the observation of their cat alytic activity towards limiting tar formation. happens, which is influenced by their chemical affinity: li Ash particles preferentially deposit on convex areas of the bed particles. Robin Faust: Conceptualization, Methodology, Investigation, Writing – original draft, Visualization. Ali Valizadeh: Methodology, Investigation, Writing – original draft. Ren Qiu: Methodology, Investi gation. Alyona Tormachen: Investigation. Jelena Maric: Resources, Writing – review & editing. Teresa Berdugo Vilches: Resources, Writing – review & editing. Nils Skoglund: Methodology, Writing – review & editing. Martin Seemann: Resources, Writing – review & editing. Mats Halvarsson: Writing – review & editing. Marcus ¨Ohman: Methodology, Writing – review & editing. Pavleta Knutsson: Concep tualization, Writing – review & editing, Funding acquisition. The Ca from the ash can interact with the bed material which leads to the formation of the homogeneous inner layer. Oxygen-carrying species such as Fe and Mn are prone to migrate to higher oxygen partial pressures, i.e., the surface of the particles. Mg can form a solid solution with Fe and Mn which is why the three elements (Fe, Mn, Mg) coincide on the surface. Alkali which can be present as gaseous hydroxides can condensate on the surface of the bed material where they most likely influence the catalytic activity of the particles. The remaining ash elements (Si and P) are possibly present as a sticky melt which can facilitate deposition of additional ash onto the ash layers. 4. Conclusion Two different materials (olivine and feldspar) were investigated regarding ash layer formation and the importance of particle morphology was emphasized. The following conclusions were drawn: Supplementary data to this article can be found online at https://doi. org/10.1016/j.fuel.2022.126387. Supplementary data to this article can be found online at https://doi. org/10.1016/j.fuel.2022.126387. A layer containing transition metal elements (Fe and Mn) was found which most likely contributes to the oxygen-carrying ability of both bed materials. 3.6. Mechanism The general mechanism for layer formation on feldspar and olivine is summarized in Fig. 15. Ash is deposited onto the bed particles which occurs preferentially on convex areas. A segregation of ash elements For the case of feldspar, the diffraction pattern recorded on layer 2 11 Fuel 333 (2023) 126387 R. Faust et al. Fig. 15. Schematic depiction of the general mechanism of layer formation found for both olivine and feldspar. Fig. 15. Schematic depiction of the general mechanism of layer formation found for both olivine and feldspar. References The accumulation of Fe and Mn on the surface and the associated development of an oxygen-carrying effect suggests that for longer exposure times, fuels containing low amounts of Fe and Mn should be utilized. l [1] Larsson A, Seemann M, Neves D, Thunman H. Evaluation of performance of industrial-scale dual fluidized bed gasifiers using the chalmers 2–4-MWth gasifier. Energy Fuels 2013;27:6665–80. [2] Devi L, et al. Catalytic decomposition of biomass tars: use of dolomite and untreated olivine. Renewable Energy 2005;30:565–87. [2] Devi L, et al. Catalytic decomposition of biomass tars: use of dolomite and untreated olivine. Renewable Energy 2005;30:565–87. The influence of the particle morphology on the tendency to form ash layers could aid the selection of starting material. If the formation of a layer is beneficial for the process because the catalytic activity is required to decrease the tar concentration, particles exhibiting convex surface should be utilized. This is likely to be the case for bed material acquired from a sand deposit which usually exhibit a more spherical morphology. Conversely, if the layer formation is detrimental because the development of the oxygen-carrying ability should be avoided, the more rugged features of particles which were acquired from crushed rocks could be used to delay the layer formation. [3] Ferreira de Almeida V, G´omez-Barea A, Nilsson S, Tuomi S. Distribution of inorganics and trace elements during waste gasification in a bench-scale fluidized bed. Energy Fuels 2021;35:15802–16. i [3] Ferreira de Almeida V, G´omez-Barea A, Nilsson S, Tuomi S. Distribution of inorganics and trace elements during waste gasification in a bench-scale fluidized bed. Energy Fuels 2021;35:15802–16. i [4] Rapagna S, Jand N, Kiennemann A, Foscolo PU. Steam-gasification of biomass in a fluidised-bed of olivine particles. Biomass Bioenergy 2000;11. [4] Rapagna S, Jand N, Kiennemann A, Foscolo PU. Steam-gasification of biomass in a fluidised-bed of olivine particles. Biomass Bioenergy 2000;11. l [5] Knutsson P, et al. Potassium speciation and distribution for the K2CO3 additive- induced activation/deactivation of olivine during gasification of woody biomass. Appl Energy 2019;248:538–44. l [5] Knutsson P, et al. Potassium speciation and distribution for the K2CO3 additive- induced activation/deactivation of olivine during gasification of woody biomass. Appl Energy 2019;248:538–44. [6] Knutsson P, Cantatore V, Seemann M, Tam PL, Panas I. Role of potassium in the enhancement of the catalytic activity of calcium oxide towards tar reduction. Appl Catal B 2018;229:88–95. CRediT authorship contribution statement happens, which is influenced by their chemical affinity: References Layer formation on feldspar bed particles during indirect gasification of wood. 2. Na-Feldspar. Energy Fuels 2019;33:7333–46. i [21] Faust R, Sattari M, Maric J, Seemann M, Knutsson P. Microscopic investigation of layer growth during olivine bed material aging during indirect gasification of biomass. Fuel 2020;266:117076. [40] Fürsatz K, Fuchs J, Benedikt F, Kuba M, Hofbauer H. Effect of biomass fuel ash and bed material on the product gas composition in DFB steam gasification. Energy 2021;219:119650. [22] Pissot S, et al. Development of oxygen transport properties by olivine and feldspar in industrial-scale dual fluidized bed gasification of woody biomass. Energy Fuels 2021;35:9424–36. ¨ [41] Anis S, Zainal ZA. Tar reduction in biomass producer gas via mechanical, catalytic and thermal methods: A review. Renew Sustain Energy Rev 2011;15:2355–77. l gy [42] Xie YR, Shen LH, Xiao J, Xie DX, Zhu J. Influences of additives on steam gasification of biomass. 1. Pyrolysis procedure. Energy Fuels 2009;23:5199–205. [43] Lee D, et al. Recent progress in the catalytic thermochemical conversion process of [23] He H, Skoglund N, ¨Ohman M. Time-dependent crack layer formation in quartz bed particles during fluidized bed combustion of woody biomass. Energy Fuels 2017; 31:1672–7. l [42] Xie YR, Shen LH, Xiao J, Xie DX, Zhu J. Influences of additives on steam gasification of biomass. 1. Pyrolysis procedure. Energy Fuels 2009;23:5199–205. [42] Xie YR, Shen LH, Xiao J, Xie DX, Zhu J. Influences of additives on steam gasification of biomass. 1. Pyrolysis procedure. Energy Fuels 2009;23:5199–205. [43] Lee D, et al. Recent progress in the catalytic thermochemical conversion process of biomass for biofuels. Chem Eng J 2022;447:137501. l gi y y p gy 009; 3 5 99 05 [43] Lee D, et al. Recent progress in the catalytic thermochemical conversion process of biomass for biofuels. Chem Eng J 2022;447:137501. l [24] Zhu Z, et al. The role of surface roughness in the wettability and floatability of quartz particles. Appl Surf Sci 2020;527:146799. [44] Kuba M, et al. Surface characterization of ash-layered olivine from fluidized bed biomass gasification. Biomass Conv Bioref 2021;11:29–38. biomass gasification. Biomass Conv Bioref 2021;11:29–38. [25] Wang M, Wen B, Fan B, Zhang H. Study on adsorption mechanism of silicate adsorbents with different morphologies and pore structures towards formaldehyde in water. Colloids Surf, A 2020;599:124887. i [45] Niu X, Shen L. Ca- and Mg-rich waste as high active carrier for chemical looping gasification of biomass. Chin J Chem Eng 2020. References Bed material as a catalyst for char gasification: The case of ash-coated olivine activated by K and S addition. Fuel 2018;224:85–93. b l b ld d h b h d l fl d b d [33] Wagner K, et al. Layer formation mechanism of K-feldspar in bubbling fluidized bed combustion of phosphorus-lean and phosphorus-rich residual biomass. Appl Energy 2019;248:545–54. l [14] Kuba M, et al. Deposit build-up and ash behavior in dual fluid bed steam gasification of logging residues in an industrial power plant. Fuel Process Technol 2015;139:33–41. l ; [15] Siriwongrungson V, et al. Influence of bed materials on the performance of the Nong Bua dual fluidized bed gasification power plant in Thailand. Biomass Conv Bioref 2020. https://doi.org/10.1007/s13399-020-00908-6. gy [34] Kirnbauer F, Hofbauer H. Investigations on bed material changes in a dual fluidized bed steam gasification plant in Güssing, Austria. Energy Fuels 2011;25:3793–8. gi p g gy [35] Faust R, et al. Interactions between automotive shredder residue and olivine bed material during indirect fluidized bed gasification. Energy Fuels 2021;35: 15935–49. [16] Morris JD, Daood SS, Chilton S, Nimmo W. Mechanisms and mitigation of agglomeration during fluidized bed combustion of biomass: A review. Fuel 2018; 230:452–73. [36] Grimm A, Skoglund N, Bostr¨om D, ¨Ohman M. Bed agglomeration characteristics in fluidized quartz bed combustion of phosphorus-rich biomass fuels. Energy Fuels 2011;25:937–47. l [17] Kuba M, et al. Mechanism of layer formation on olivine bed particles in industrial- scale dual fluid bed gasification of wood. Energy Fuels 2016;30:7410–8. [37] Brus E, ¨Ohman M, Nordin A. Mechanisms of bed agglomeration during fluidized- bed combustion of biomass fuels. Energy Fuels 2005;19:825–32. ¨ li [18] Faust R, Berdugo Vilches T, Malmberg P, Seemann M, Knutsson P. Comparison of ash layer formation mechanisms on si-containing bed material during dual fluidized bed gasification of woody biomass. Energy Fuels 2020;34:8340–52. i [38] De Geyter S, ¨Ohman M, Bostr¨om D, Eriksson M, Nordin A. Effects of non-quartz minerals in natural bed sand on agglomeration characteristics during fluidized bed combustion of biomass fuels. Energy Fuels 2007;21:2663–8. li [19] Faust R, et al. Layer formation on feldspar bed particles during indirect gasification of wood. 1. K-Feldspar. Energy Fuels 2019;33:7321–32. [39] Kuba M, et al. Thermal stability of bed particle layers on naturally occurring minerals from dual fluid bed gasification of woody biomass. Energy Fuels 2016;30: 8277–85. [20] Hannl TK, et al. References [6] Knutsson P, Cantatore V, Seemann M, Tam PL, Panas I. Role of potassium in the enhancement of the catalytic activity of calcium oxide towards tar reduction. Appl Catal B 2018;229:88–95. [7] Berdugo Vilches T, Marinkovic J, Seemann M, Thunman H. Comparing active bed materials in a dual fluidized bed biomass gasifier: olivine, bauxite, quartz-sand, and ilmenite. Energy Fuels 2016;30:4848–57. [7] Berdugo Vilches T, Marinkovic J, Seemann M, Thunman H. Comparing active bed materials in a dual fluidized bed biomass gasifier: olivine, bauxite, quartz-sand, and ilmenite. Energy Fuels 2016;30:4848–57. 12 R. Faust et al. Fuel 333 (2023) 126387 [8] Berguerand N, Berdugo Vilches T. Alkali-feldspar as a catalyst for biomass gasification in a 2-MW indirect gasifier. Energy Fuels 2017;31:1583–92. [27] Wang X, Zhang Q. Role of surface roughness in the wettability, surface energy and flotation kinetics of calcite. Powder Technol 2020;371:55–63. ii [9] Fürsatz K, et al. Impact of residual fuel ash layers on the catalytic activation of K- feldspar regarding the water–gas shift reaction. Biomass Conv Bioref 2021;11: 3–14. i l [28] Darbha GK, Sch¨afer T, Heberling F, Lüttge A, Fischer C. Retention of latex colloids on calcite as a function of surface roughness and topography. Langmuir 2010;26: 4743–52. l [10] Khan MM, Xu S, Wang C. Catalytic biomass gasification in decoupled dual loop gasification system over alkali-feldspar for hydrogen rich-gas production. Biomass Bioenergy 2022;161:106472. i [29] Wang X, Zhang Q. Insight into the influence of surface roughness on the wettability of apatite and dolomite. Minerals 2020;10:114. l [30] Leion H, Frick V, Hildor F. Experimental method and setup for laboratory fluidized bed reactor testing. Energies 2018;11:2505. [11] Trubetskaya A. Reactivity effects of inorganic content in biomass gasification: a review. Energies 2022;15:3137. ¨ [31] Berguerand N, Marinkovic J, Berdugo Vilches T, Thunman H. Use of alkali-feldspar as bed material for upgrading a biomass-derived producer gas from a gasifier. Chem Eng J 2016;295:80–91. l [12] Kuba M, Skoglund N, ¨Ohman M, Hofbauer H. A review on bed material particle layer formation and its positive influence on the performance of thermo-chemical biomass conversion in fluidized beds. Fuel 2021;291:120214. i [32] Kirnbauer F, Hofbauer H. The mechanism of bed material coating in dual fluidized bed biomass steam gasification plants and its impact on plant optimization. Powder Technol 2013;245:94–104. l l [13] Berdugo Vilches T, et al. References https://doi.org/10.1016/j. cjche.2020.09.024. S1004954120305280. [26] Prouzet E, Boissi`ere C, Kim SS, Pinnavaia TJ. Roughness of mesoporous silica surfaces deduced from adsorption measurements. Microporous Mesoporous Mater 2009;119:9–17. [46] Hu Q, Wang C-H. Insight into the Fe2O3/CaO-based chemical looping process for biomass conversion. Bioresour Technol 2020;310:123384. 13
https://openalex.org/W1982398105	https://acuresearchbank.acu.edu.au/download/581fe831a931b977e99e5edbf571309d0a6b16afa2e5e470be649c6abd0b59b5/582955/Stadelmann_2015_Impaired_postural_control_in_healthy_men.pdf	English	null	Impaired Postural Control in Healthy Men at Moderate Altitude (1630 M and 2590 M): Data from a Randomized Trial	PloS one	2,015	cc-by	7,330	RESEARCH ARTICLE Methods Received: August 13, 2014 Accepted: December 10, 2014 Published: February 27, 2015 In 51 healthy men, median age 24 y (quartiles 20;28), static control was evaluated on a bal- ance platform in Zurich, 490 m, and during a 4-day sojourn in Swiss mountain villages at 1630 m and 2590 m, 2 days each. The order of altitude exposure was randomized. Total center of pressure path length (COPL) and sway amplitude measured in two directions by a balance platform, and pulse oximetry were recorded. Data were compared between altitudes. Copyright: © 2015 Stadelmann 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. Katrin Stadelmann1,2, Tsogyal D. Latshang3, Christian M. Lo Cascio3, Ross A. Clark5, Reto Huber2,4, Malcolm Kohler2,3, Peter Achermann1,2☯, Konrad E. Bloch2,3☯* 1 Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland, 2 Zurich Center for Integrative Human Physiology, University of Zurich, Zurich, Switzerland, 3 Pulmonary Division, University Hospital Zurich, Zurich, Switzerland, 4 University Children’s Hospital Zurich, Zurich, Switzerland, 5 Department of Physiotherapy, Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne, Melbourne, Australia a1111 ☯These authors contributed equally to this work. * konrad.bloch@usz.ch Objectives Intact postural control is essential for safe performance of mountain sports, operation of ma- chinery at altitude, and for piloting airplanes. We tested whether exposure to hypobaric hyp- oxia at moderate altitude impairs the static postural control of healthy subjects. Academic Editor: Larisa G. Tereshchenko, Johns Hopkins University SOM, UNITED STATES OPEN ACCESS Citation: Stadelmann K, Latshang TD, Lo Cascio CM, Clark RA, Huber R, Kohler M, et al. (2015) Impaired Postural Control in Healthy Men at Moderate Altitude (1630 M and 2590 M): Data from a Randomized Trial. PLoS ONE 10(2): e0116695. doi:10.1371/journal.pone.0116695 * konrad.bloch@usz.ch ☯These authors contributed equally to this work. Impaired Postural Control in Healthy Men at Moderate Altitude (1630 M and 2590 M): Data from a Randomized Trial Katrin Stadelmann1,2, Tsogyal D. Latshang3, Christian M. Lo Cascio3, Ross A. Clark5, Reto Huber2,4, Malcolm Kohler2,3, Peter Achermann1,2☯, Konrad E. Bloch2,3☯* Katrin Stadelmann1,2, Tsogyal D. Latshang3, Christian M. Lo Cascio3, Ross A. Clark5, Reto Huber2,4, Malcolm Kohler2,3, Peter Achermann1,2☯, Konrad E. Bloch2,3☯* Postural Control at Altitude Trial Registration ClinicalTrials.gov NCT01130948. Introduction Control of posture is essential for the safe performance of many activities of daily life, and in particular for sports, operating cars, machinery and for piloting airplanes. Furthermore, poor postural control is a major contributing factor to an increased risk for falls in the elderly [1]. Altitude travel for professional and leisure activities is increasingly common among millions of lowlanders worldwide [2,3]. As the central nervous system is sensitive to hypoxia, ascent to alti- tude may lead to impaired cognitive performance and motor control [4,5]. Dizziness, mental and muscle fatigue, and decrements in alertness and psychomotor performance have been shown to occur in hypoxic environments at high altitude (> 4000 m) [6]. Furthermore, impair- ments in postural stability during short exposures of a few hours to simulated altitudes between 1500 m and 5500 m have been reported [7–9]. However, it remains unclear whether such im- pairments persist or even aggravate during a prolonged stay at altitude, and to which extent they are altitude-dependent [7,9,10]. Baumgartner et al. [11] did not observe an improvement of posturographic performance in healthy mountaineers evaluated over the course of 3 days at the Capanna Regina Margherita research station at 4559 m. Whether similar impairments in postural control occur and persist during a prolonged stay even at moderate altitudes of 1600 to 2600 m remains unknown. Since most mountain resorts with frequent tourism are located at such moderate elevations more knowledge on this issue is highly desirable and relevant. Therefore, the purpose of the current study was to investigate static postural control in healthy subjects at 490 m and during a four-day sojourn at an alpine resort at moderate altitude. Ap- plying a randomized cross-over study design we tested the hypothesis that static control was impaired in an altitude dependent manner at Davos Jakobshorn (2590 m) compared to Davos Wolfgang (1630 m) and Zurich (490 m), respectively. Conclusions Competing Interests: The authors have declared that no competing interests exist. Exposure to 1630 and 2590m was associated with impaired static postural control even when visual references were available. Materials and Methods The protocol for this trial and supporting CONSORT checklist are available as supporting in- formation; see S1 CONSORT Checklist and S1 Protocol. Results Data Availability Statement: All relevant data are included in the paper and its Supporting Information files. Median (quartiles) COPL during standing on both legs with eyes open at 490 m and in the evenings on the first and second days at 1630 and 2590 m, respectively were: 50 (45;57), 55 (48;62), 56 (49;61), 53 (47;59), 54 (48;60) cm, P<0.001 ANOVA. Corresponding arterial oxygen saturation was 97% (96;97), 95% (94;96), 95%(94;96), 92%(90;93), 93%(91;93), P<0.001. Anterior-posterior sway amplitudes were larger at 1630 and 2590 m compared to 490 m, P<0.001. Multiple logistic regression analysis confirmed that higher altitudes (1630 and 2590m) were independently associated with increased COPL when controlled for the order of altitude exposure and age (P=0.001). Funding: This study was supported by unconditional grants from the Zurich Center for Integrative Human Physiology, University of Zurich; the Lunge Zurich, Lung Ligue of Zurich, Zurich; and the Schweizerische Unfallversicherungsgesellschaft, SUVA, Luzern. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. 1 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 Standing balance Standing balance was measured with a rectangular balance board 30x50cm in size, containing four transducers to assess movements of the center of pressure (COP; Wii Balance Board, Red- mond, WA, USA). The balance board was interfaced with a computer using specialized software as described previously [15]. Repeated calibrations with known weights were performed. Mea- surements of the standing balance by the balance board have been extensively validated [15]. Standing balance was assessed in the following three conditions: 1) standing on both legs, 2) on the right leg, and 3) on the left leg. Each condition was repeated 3 times with eyes open and then with eyes closed resulting in a total of 18 tests per session (see Table 1). Measurements on both legs lasted 30 s, and measurements on one leg 10 s, with breaks of at least 20 s between tests. The balance board was positioned 1.5 m in front of a wall. During eyes open tests subjects were instructed to focus on a black dot, 2 cm in diameter, fixed at the wall at the height of the subject’s eyes, and to keep their arms parallel to the body during all measurements. Standing balance on both legs was performed with a 30° angle between feet and with heels 2 cm apart [16]. For assessment of standing balance on one leg with the foot positioned in the center of the balance board, subjects had to rise the other leg to a 90° angle in the hip and bend the knee to a 90° angle so that the lower leg was in vertical position [17]. Standing balance was measured with a rectangular balance board 30x50cm in size, containing four transducers to assess movements of the center of pressure (COP; Wii Balance Board, Red- mond, WA, USA). The balance board was interfaced with a computer using specialized software as described previously [15]. Repeated calibrations with known weights were performed. Mea- surements of the standing balance by the balance board have been extensively validated [15]. Standing balance was assessed in the following three conditions: 1) standing on both legs, 2) on the right leg, and 3) on the left leg. Each condition was repeated 3 times with eyes open and then with eyes closed resulting in a total of 18 tests per session (see Table 1). Subjects Fifty-one healthy male volunteers, mean age ± SD: 26.9 ± 9.3 years (range: 20–67), were re- cruited. Subjects were accepted only if they were in good health, taking no medications regular- ly, had no history of altitude related illness during previous stays at < 2500 m, had not travelled to altitudes > 1500 m in the two weeks prior to the study. The study protocol was ap- proved by the ethical committee of the Canton of Zurich (Switzerland), and participants gave their written informed consent. 2 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 Postural Control at Altitude Standing balance Measurements on both legs lasted 30 s, and measurements on one leg 10 s, with breaks of at least 20 s between tests. The balance board was positioned 1.5 m in front of a wall. During eyes open tests subjects were instructed to focus on a black dot, 2 cm in diameter, fixed at the wall at the height of the subject’s eyes, and to keep their arms parallel to the body during all measurements. Standing balance on both legs was performed with a 30° angle between feet and with heels 2 cm apart [16]. For assessment of standing balance on one leg with the foot positioned in the center of the balance board, subjects had to rise the other leg to a 90° angle in the hip and bend the knee to a 90° angle so that the lower leg was in vertical position [17]. Oxygen saturation Oxygen saturation (SpO2) was measured by pulsoximetry every morning and evening while subjects were resting quietly in supine position. Protocol and interventions This study was part of a randomized cross-over trial (www.ClinicalTrials.gov, NCT01130948) evaluating the effects of altitude exposure on various physiologic outcomes. Data on the effects of altitude on sleep and cardiovascular function are described elsewhere [12–14], including the consort flow chart (Fig. 1). The data on static postural control, the topic of the current paper, have not been published. Measurements were performed in the period from July to October 2010 during one day at the University Hospital of Zurich (baseline, 490 m, 1608 ft, barometric pressure [PB] 719 Torr), and for 4 days at two study locations in the Swiss Alps, i.e. for 2 days at Davos Wolfang (1630 m, 5348 ft, PB 630 Torr), and 2 days at Davos Jakobshorn (2590 m, 8497 ft, PB 562 Torr). The order of altitude exposure was randomized with a balanced design by letting partici- pants select a study time slot according to their preference and availability without being aware of the corresponding altitude exposure sequence. Each time slot corresponded to one of the fol- lowing 4 sequences of altitude exposure (Fig. 1): A) 490 m—1630 m—1630 m—2590 m—2590 m; B) 490 m—2590 m—2590 m-1630 m—1630 m; C) 1630 m—1630 m—2590 m—2590 m— 490m; D) 2590 m—2590 m—1630 m—1630 m—490 m. A wash-out phase of 2–4 weeks was interposed between studies at altitude and at 490 m and vice versa. Subjects traveled from Zu- rich to the altitude locations by train and cable car. On study days at 490 m and during the stay at altitude subjects had to avoid strenuous exercise and to stay at the same altitude and within 200 m of the study location. The nocturnal rest period was from 23:00 to 06:00. Standing bal- ance was measured between 5 and 6 pm and between 10 and 11 am. Data analysis Three outcome measures were derived from each balance test: the pathway length of the COP movement; the movement amplitude in anterior-posterior (AP) direction; and the movement 3 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 Postural Control at Altitude Fig 1. Consort flow chart. 51 participants were randomized to 4 differend schedules of altitude exposure. The flow chart is the same as reported in [12]. doi:10.1371/journal.pone.0116695.g001 Fig 1. Consort flow chart. 51 participants were randomized to 4 differend schedules of altitude exposure. The flow chart is the same as that for the study reported in [12]. doi:10.1371/journal.pone.0116695.g001 doi:10.1371/journal.pone.0116695.g001 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 4 / 12 Postural Control at Altitude Table 1. Test protocol. Test number Condition Duration [s] Eyes Both legs Right leg Left leg 1 X 30 open 2 X 10 open 3 X 10 open 4 X 30 open 5 X 10 open 6 X 10 open 7 X 30 open 8 X 10 open 9 X 10 open 10 X 30 closed 11 X 10 closed 12 X 10 closed 13 X 30 closed 14 X 10 closed 15 X 10 closed 16 X 30 closed 17 X 10 closed 18 X 10 closed doi:10.1371/journal.pone.0116695.t001 Table 1. Test protocol. Eyes doi:10.1371/journal.pone.0116695.t001 amplitude in medial-lateral (ML) direction (Fig. 2). Means of the three repeated measurements in each condition and with eyes open and eyes closed, respectively, were computed and the re- sults used for further analysis. Data were summarized by means (SD) and medians (quartiles) for normal and non-normal distribution, respectively. For statistical analysis, data were transformed to obtain a normal dis- tribution where appropriate. Path length was transformed to 1/square root (path length). The anterior-posterior and medial-lateral amplitude were log transformed. A mixed model ANOVA was performed to investigate differences between the five test days at the different altitudes (i.e., one at 490 m, two at 1630 m and two at 2590 m). Post-hoc Wilcoxon signed ranks tests with Bonferroni correction were performed comparing results from different altitudes whenever the overall ANOVA was significant (p<0.05). In order to evaluate the effect of altitude on static control while controlling for potential confounders, univariate and multivariate ordinal logistic regression analysis was employed with COP path length, antero-posterior and medial-lateral sway, respectively, as dependent variables after transformation to quintiles. Inde- pendent variables in this analysis were altitude (490, 1630, 2590 m), oxygen saturation (%), alti- tude exposure sequence (coded 1 to 4), consecutive number of test days (1 to 5), evening and morning tests (coded 1, 2), eyes open and closed (coded 1 and 2), and age (years). Altitude and age, and all other independent variables with p<0.1 in univariate analysis were entered into the multivariate analysis. Sample size estimations indicating a minimal required number of 50 partic- ipants were based on the outcomes apnea/hypopnea index and psychomotor vigilance response reported previously [12]. A probability of P<0.05 was considered statistically significant. doi:10.1371/journal.pone.0116695.t001 Postural Control at Altitude Fig 2. Center of pressure recorded in an individual during a 30 s balance board test on both legs with eyes open at 490 m and 2590 m. Center of pressure path length (COPL) = 45 and 56 cm; anterior-posterior (AP) sway amplitude = 1.3 and 2.8 cm; medial-lateral (ML) sway amplitude = 1.4 and 1.9 cm. d i 10 1371/j l 0116695 002 Fig 2. Center of pressure recorded in an individual during a 30 s balance board test on both legs with eyes open at 490 m and 2590 m. Center of pressure path length (COPL) = 45 and 56 cm; anterior-posterior (AP) sway amplitude = 1.3 and 2.8 cm; medial-lateral (ML) sway amplitude = 1.4 and 1.9 cm. Fig 2. Center of pressure recorded in an individual during a 30 s balance board test on both legs with eyes open at 490 m and 2590 m. Center of pressure path length (COPL) = 45 and 56 cm; anterior-posterior (AP) sway amplitude = 1.3 and 2.8 cm; medial-lateral (ML) sway amplitude = 1.4 and 1.9 cm. doi:10.1371/journal.pone.0116695.g002 doi:10.1371/journal.pone.0116695.g002 summarize the COPL and the anterior-posterior sway amplitude at the different altitudes along with oxygen saturation for tests on both legs. S1 and S2 Tables show corresponding val- ues for single leg tests. The COPL during balance tests with eyes open on both legs, right leg, and left leg was longer at the higher altitudes compared to 490 m (Table 2, S1 Table) and this was observed in both morning and evening measurements. The findings from evening tests on both legs are summarized in Fig. 3. In contrast, the COPL of measurements with eyes closed was not consistently changed with increasing altitude although the path lengths were longer than corresponding values with eyes open. Consistent with the longer COPL, the anterior-posterior sway amplitude was also increased at 1630 m during measurements in the eyes open condition with the exception of standing bal- ance on both legs in the morning (Table 3). Furthermore, standing balance on the right and left leg with eyes open also showed significant changes on the medial-lateral sway amplitude on morning and evening measurements (S3 Table). Individual changes in path length at altitude are illustrated in S1 Fig. Results Fig. 1 shows the participant flow. An example of a center of pressure path length (COPL) re- cording at 490 m and at 2590 m with a large path length is illustrated in Fig. 2. Tables 2 and 3 5 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 * p<0.0125 compared to 490 m (Bonferroni correction), post-hoc Wilcoxon signed ranks test. ¶ p<0.0125 compared to 1650 m, day 1 and 2 (Bonferroni correction), post-hoc Wilcoxon signed ranks test. doi:10.1371/journal.pone.0116695.t002 Data are presented as median path length in cm [25 , 75 percentile]. P ANOVA overall: Mixed model ANOVA with factor condition (490 m, 1630 m day 1, 1630 m day 2, 2590 m day 1, 2590 m day 2). * p<0.0125 compared to 490 m (Bonferroni correction), post-hoc Wilcoxon signed ranks test. ¶ p<0.0125 compared to 1650 m, day 1 and 2 (Bonferroni correction), post-hoc Wilcoxon signed ranks test. doi:10.1371/journal.pone.0116695.t002 Table 3. Anterior-posterior sway amplitude at different altitudes. Measurement Time Eyes 490 m 1630 m Day 1 1630 m Day 2 2590 m Day 1 2590 m Day 2 P ANOVA Overall Both legs evening open 1.9 [1.7, 2.4] 2.2 [2.0, 2.6] * 2.3 [2.1, 2.8] * 2.1 [1.7, 2.4] 2.1 [1.7, 2.7] <0.001 Both legs evening closed 2.8 [2.3, 3.3] 2.9 [2.5, 3.4] 2.7 [2.3, 3.2] 2.7 [2.1, 3.4] 2.8 [2.3, 3.4] 0.480 Both legs morning open 2.3 [1.9, 2.8] 2.5 [2.0, 3.0] 2.4 [2.1, 3.0] 2.3 [1.8, 3.0] 2.3 [1.9, 2.8] 0.099 Both legs morning closed 2.8 [2.4, 3.4] 3.0 [2.5, 3.5] 2.8 [2.4, 3.5] 3.2 [2.4, 3.5] 3.0 [2.5, 3.5] 0.366 Data are presented as median sway ampltitude in cm [25th, 75th percentile]. Oxygen saturation values are the same as in Table 2. P ANOVA overall: Mixed model ANOVA with factor condition (490 m, 1630 m day 1, 1630 m day 2, 2590 m day 1, 2590 m day 2). * p<0.0125 vs. to 490 m (Bonferroni correction), post-hoc Wilcoxon signed ranks. Table 3. Anterior-posterior sway amplitude at different altitudes. Data are presented as median sway ampltitude in cm [25th, 75th percentile]. Oxygen saturation values are the same as in Table 2. P ANOVA overall: Mixed model ANOVA with factor condition (490 m, 1630 m day 1, 1630 m day 2, 2590 m day 1, 2590 m day 2). * p<0.0125 vs. to 490 m (Bonferroni correction), post-hoc Wilcoxon signed ranks. separately were observed. The odds ratio for an increase in COPL by one quintile with ascent from 490 m to altitude (1650 and 2590 m) was 1.70 (95% confidence limits 1.29 to 2.53, P <0.001); the corresponding odds ratio for an increase in antero-posterior sway amplitude was 1.31 (95% CI 0.97 to 1.79, P = 0.074). There was no association of altitude with medial-lateral sway amplitude (S4 Table). Seventy-seven and 69% of the subjects showed an increase in path length at 1630 m and 2590 m, respectively for measurements on both legs with eyes open. Seventy-eight and 82% showed a reduction in bal- ance performance for the measurements on the left and right leg respectively at 1630 m. Changes in path length on one leg at 2590 m were not significant. Anterior-posterior and medi- al-lateral sway amplitudes were both increased at 1630 m, but remained unchanged at 2590 m (Table 3 and S3 Table). Multivariate ordinal logistic regression analysis confirmed an independent effect of altitude on the COPL (Table 4) even when controlled for potential confounders. Thus, the odds-ratio that the COPL increased by one quintile with ascent from 490 to 1630 m, and with ascent from 490 to 2590 m, respectively, was 1.7 (Table 4). Performing tests with eyes closed was a strong predictor of a larger path length (odds-ratio 7.1, Table 4). Multivariate logistic regression anal- ysis further revealed an independent effect of altitude on the anterior-posterior sway amplitude at 1630 m (Table 5). When ordinal regression analysis was performed with data from both alti- tudes (1650 and 2590 m) vs. 490 m baseline, similar results as those with all 3 altitudes entered 6 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 Postural Control at Altitude Table 2. Center of pressure path length at different altitudes. Measurement Time Eyes 490 m 1630 m Day 1 1630 m Day 2 2590 m Day 1 2590 mDay 2 P ANOVA overall Both legs evening open 50 [45, 57] 55 [48, 62] * 56 [49, 61] * 52 [47, 59] 54 [48, 60] * <0.001 Both legs evening closed 62 [55, 71] 65 [57, 76] 64 [57, 76] 64 [56, 76] 63 [58, 75] 0.150 Both legs morning open 52 [47, 58] 54 [49, 63] * 56 [49, 60] * 55 [47, 61] 55 [48, 60] * 0.003 Both legs morning closed 61 [57, 69] 65 [60, 74] 63 [57, 69] 65 [58, 78] 68 [59, 73] 0.042 Pulse oximetry, % evening open 97 [96, 97] 95 [94, 96] * 95 [94, 96] * 92 [90, 93] * 93 [91, 93] * <0.001 Pulse oximetry, % morning open 97 [97, 98] 96 [95, 96] * 96 [95, 96] * 93 [92, 94] ¶ 94 [93, 94] ¶ <0.001 Table 2. Center of pressure path length at different altitudes. Table 2. Center of pressure path length at different altitudes. Measurement Time Eyes 490 m 1630 m Day 1 1630 m Day 2 2590 m Day 1 2590 mDay 2 P ANOVA overall Both legs evening open 50 [45, 57] 55 [48, 62] * 56 [49, 61] * 52 [47, 59] 54 [48, 60] * <0.001 Both legs evening closed 62 [55, 71] 65 [57, 76] 64 [57, 76] 64 [56, 76] 63 [58, 75] 0.150 Both legs morning open 52 [47, 58] 54 [49, 63] * 56 [49, 60] * 55 [47, 61] 55 [48, 60] * 0.003 Both legs morning closed 61 [57, 69] 65 [60, 74] 63 [57, 69] 65 [58, 78] 68 [59, 73] 0.042 Pulse oximetry, % evening open 97 [96, 97] 95 [94, 96] * 95 [94, 96] * 92 [90, 93] * 93 [91, 93] * <0.001 Pulse oximetry, % morning open 97 [97, 98] 96 [95, 96] * 96 [95, 96] * 93 [92, 94] ¶ 94 [93, 94] ¶ <0.001 Data are presented as median path length in cm [25th, 75th percentile]. P ANOVA overall: Mixed model ANOVA with factor condition (490 m, 1630 m day 1, 1630 m day 2, 2590 m day 1, 2590 m day 2). PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 median path length in cm [25th, 75th percentile]. Discussion We studied the effect of a stay at moderate altitude for four days on static control in a large co- hort of healthy men. The results of our randomized cross-over study demonstrate that the center of pressure (COPL) and the anterior-posterior sway amplitude were increased at moderate alti- tude which is consistent with an impaired static control at 1630 m and 2590 m compared to 490 m. The findings may be relevant for a large number of persons traveling to moderate altitude. This is the first field study to examine changes in balance performance at two moderate alti- tudes (1630 m and 2590 m) compared to low altitude (490 m). Compared to investigations at simulated altitudes, the results from the current four-day study performed at two real moderate altitudes represent the environmental conditions of many mountain tourists worldwide. They are also relevant for persons operating machinery at altitude, as well as for aircrew members, as a cabin pressure equivalent to up to 2430 m (8000 ft) is reached in commercial airplanes [18]. Our results show that the COPL and the anterior-posterior sway amplitude are affected already at 1630 m. Thus, static postural control was impaired at a relatively low altitude. These 7 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 Postural Control at Altitude Fi 3 R l f i f i l i b l i Fig 3. Results from evening measurements of arterial oxygen saturation by pulse oximetry (SpO2, medians and quartiles, top panel), and center of pressure path length by the balance board (middle panel) in the 51 study participants standing on both legs with eyes open along with altitude profile. Box plots show medians, quartiles, and whiskers representing the 10th and 90th percentiles, and dots the 5th and 95th percentiles. * p<0.05 vs. 490 m. D1 and D2 = 1st and 2nd day at corresponding altitude. doi:10.1371/journal.pone.0116695.g003 Fig 3. Results from evening measurements of arterial oxygen saturation by pulse oximetry (SpO2, medians and quartiles, top panel), and center of pressure path length by the balance board (middle panel) in the 51 study participants standing on both legs with eyes open along with altitude profile. Box plots show medians, quartiles, and whiskers representing the 10th and 90th percentiles, and dots the 5th and 95th percentiles. * p<0.05 vs. 490 m. D1 and D2 = 1st and 2nd day at corresponding altitude. doi:10.1371/journal.pone.0116695.g003 Fig 3. Postural Control at Altitude Table 4. Effect of altitude on center of pressure path length: ordinal logistic regression analysis. Univariate Multivariate Dependent variable: quintiles of center of pressure path length Odds ratio 95% CI P Odds ratio 95% CI P Altitude 1630 vs. 490 m 1.55 1.22 to 1.97 <0.001 1.72 1.27 to 2.34 <0.001 2590 vs. 490 m 1.53 1.19 to 1.96 0.001 1.73 1.25 to 2.39 0.001 SpO2, % 0.95 0.89 to 1.02 0.164 Altitude exposure sequence 2 vs. 1 0.51 0.22 to 1.21 0.129 0.43 0.15 to 1.22 0.113 3 vs. 1 0.34 0.12 to 0.95 0.040 0.23 0.06 to 0.86 0.029 4 vs. 1 0.45 0.21 to 0.10 0.049 0.40 0.15 to 1.03 0.057 Consecutive number of test days 1.02 0.95 to 1.10 0.596 Eyes closed vs. eyes open 6.27 4.43 to 8.87 <0.001 7.12 4.82 to 10.50 <0.001 Morning vs. evening test 1.09 0.96 to 1.23 0.176 Age, y 1.00 0.96 to 1.03 0.710 0.98 0.95 to 1.02 0.394 Univariate and multivariate logistic regressions were performed on quintiles of COP path lengths with an odds ratio of 1 corresponding to the lowest quintile. Altitude, age, and all variables with P<0.1 in univariate analysis were entered into the multivariate model. Altitude exposure sequence was: 1 = 2590-1630-490 m; 2 = 1630-2590-490m; 3 = 1630-2590-490 m; 4 = 2590-1630-490 m; consecutive number of test days was 1 to 5; eyes open and closed were coded as 1 and 2, respectively; morning and evening tests were coded as 1 and 2, respectively. CI = conﬁdence interval. Table 4. Effect of altitude on center of pressure path length: ordinal logistic regression analysis. Univariate and multivariate logistic regressions were performed on quintiles of COP path lengths with an odds ratio of 1 corresponding to the lowest quintile. Altitude, age, and all variables with P<0.1 in univariate analysis were entered into the multivariate model. Altitude exposure sequence was: 1 = 2590-1630-490 m; 2 = 1630-2590-490m; 3 = 1630-2590-490 m; 4 = 2590-1630-490 m; consecutive number of test days was 1 to 5; eyes open and closed were coded as 1 and 2, respectively; morning and evening tests were coded as 1 and 2, respectively. CI = conﬁdence interval. Discussion Results from evening measurements of arterial oxygen saturation by pulse oximetry (SpO2, medians and quartiles, top panel), and center of pressure path length by the balance board (middle panel) in the 51 study participants standing on both legs with eyes open along with altitude profile. Box plots show medians, quartiles, and whiskers representing the 10th and 90th percentiles, and dots the 5th and 95th percentiles. * p<0.05 vs. 490 m. D1 and D2 = 1st and 2nd day at corresponding altitude. doi:10 1371/journal pone 0116695 g003 doi:10.1371/journal.pone.0116695.g003 doi:10.1371/journal.pone.0116695.g003 observations corroborate and extend data from previous studies demonstrating impaired static postural control at simulated higher altitudes and with shorter exposures of a few minutes to 24 h [8,9,19]. In line with these findings Holness et al. [8] observed an increase in anterior- posterior sway in seven healthy men if oxygen saturation was reduced to 70% by letting them breathe a normobaric hypoxic gas mixture. Moreover, Cymerman et al. [19] reported worsen- ing of postural stability in 19 subjects during measurements on a dynamic platform balance system at an altitude equivalent to 4300 m simulated in a decompression chamber, and Nor- dahl et al. [9] observed an increase in body sway in 16 military aircrew members standing on a balance platform at simulated altitudes of 5500 m, 4300 m and 2400 m. 8 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 doi:10.1371/journal.pone.0116695.t004 Consistent with previous studies that reported alterations mainly or exclusively in the anterior-posterior direction during altitude exposure [8,9,11] we observed an increase predomi- nantly in the anterior-posterior sway amplitude (Table 3) and, to a lesser extent, in medial-lateral sway amplitude (S3 Table, online supplement). However, the latter was only affected when Table 5. Effect of altitude on anterior-posterior sway: ordinal logistic regression analysis. Table 5. Effect of altitude on anterior-posterior sway: ordinal logistic regression analysis. Univariate Multivariate Dependent variable: quintiles of anterior-posterior sway Odds ratio 95% CI P Odds 95% CI P Altitude 1630 vs. 490 m 1.35 1.04 to 1.75 0.023 1.46 1.09 to 1.96 0.012 2590 vs. 490 m 1.11 0.80 to 1.53 0.543 1.15 0.83 to 1.64 0.449 SpO2, % 1.01 0.94 to 1.09 0.838 Altitude exposure sequence 2 vs. 1 0.59 0.26 to 1.33 0.204 3 vs. 1 0.81 0.36 to 1.80 0.605 4 vs. 1 0.58 0.27 to 1.24 0.162 Consecutive number of test days 0.92 0.85 to 1.00 0.058 0.91 0.83 to 0.98 0.045 Eyes closed vs. eyes open 4.49 3.24 to 6.23 <0.001 4.62 3.28 to 6.52 <0.001 Morning vs. evening test 1.63 1.41 to 1.88 <0.001 1.68 1.43 to 1.98 <0.001 Age, y 1.00 0.96 to 1.04 0.922 1.00 0.95 to 1.04 0.883 Univariate and multivariate logistic regressions were performed on quintiles of values of the anterior-posterior sway amplitude with an odds ratio of 1 corresponding to the lowest quintile. Altitude, age, and all variables with P<0.100 in univariate analysis were entered into the multivariate model. Altitude exposure sequence was: 1 = 2590-1630-490 m; 2 = 1630-2590-490m; 3 = 1630-2590-490 m; 4 = 2590-1630-490 m; consecutive number of test days was 1 to 5; eyes open and closed were coded as 1 and 2, respectively; morning and evening tests were coded as 1 and 2, respectively. CI = conﬁdence interval. doi:10.1371/journal.pone.0116695.t005 Table 5. Effect of altitude on anterior-posterior sway: ordinal logistic regression analysis. Univariate and multivariate logistic regressions were performed on quintiles of values of the anterior-posterior sway amplitude with an odds ratio of 1 Univariate and multivariate logistic regressions were performed on quintiles of values of the anterior-posterior sway amplitude with an odds ratio of 1 corresponding to the lowest quintile. Altitude, age, and all variables with P<0.100 in univariate analysis were entered into the multivariate model. regressions were performed on quintiles of values of the anterior-posterior sway amplitude with an odds ratio of 1 Univariate and multivariate logistic regressions were performed on quintiles of values of the anterior-posterior sway amplitude with an odds ratio of 1 corresponding to the lowest quintile. Altitude, age, and all variables with P<0.100 in univariate analysis were entered into the multivariate model. Altitude exposure sequence was: 1 = 2590-1630-490 m; 2 = 1630-2590-490m; 3 = 1630-2590-490 m; 4 = 2590-1630-490 m; consecutive number of test days was 1 to 5; eyes open and closed were coded as 1 and 2, respectively; morning and evening tests were coded as 1 and 2, respectively. CI = conﬁdence interval. Altitude exposure sequence was: 1 = 2590-1630-490 m; 2 = 1630-2590-490m; 3 = 1630-2590-490 m; 4 = 2590-1630-490 m; consecutive number of test days was 1 to 5; eyes open and closed were coded as 1 and 2, respectively; morning and evening tests were coded as 1 and 2, respectively. CI = conﬁdence interval. PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 9 / 12 Postural Control at Altitude standing on one leg, and not when standing on both legs, most likely because postural control gets more challenging when holding balance on one leg. In addition, our larger sample size pro- vided a greater power to detect minor changes. In contrast to the medial-lateral sway amplitude, the sway amplitude in the anterior-posterior plane was increased in all three conditions (right leg, left leg and both legs) at moderate altitude. These findings therefore support prior assump- tions, that control over movements in the anterior-posterior plane is most sensitive to hypobaric hypoxia at altitude [8,9,11]. We found larger COPL and sway amplitudes with eyes closed than with eyes open (Tables 2 and 3). Vision is an important sensory input for postural control along with the proprioceptive and vestibular feedback [20]. Related to that and in agreement with Hol- ness et al. [8], we observed that, independent of altitude, performing tests with eyes closed was a strong predictor of a large path length. The dominant effect of withdrawing visual references by closing the eyes (and the large individual variability of this effect) has concealed the more mod- erate effect of altitude in univariate analyses (Tables 2 and 3). When the effects of closing the eyes and other confounders were controlled for by multivariate analyses, significant, indepen- dent effects of altitude on COPL and sway amplitude were confirmed (Tables 4 and 5). In the current altitude field study the variables representing postural control indicated im- pairment at 1630 and 2590 m compared to 490 m, respectively, but the differences between 1630 and 2590 m were not statistically significant (Tables 2 and 3). It remains elusive whether the lack of further impairment with ascent from 1630 to 2590 m is due to compensatory mecha- nisms such as an increase in cerebral perfusion or other unknown mechanisms. Although the multiple regression analysis revealed a significant effect of altitude on postural control, SpO2 was not independently associated with measures of postural control (Table 4). PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 S3 Table. Medial-lateral sway amplitude at different altitudes. Summary statistics for the re- sults obtained during measurements on both legs, the right and left leg, with eyes open and closed, and in the evening and morning, respectively. (DOCX) S3 Table. Medial-lateral sway amplitude at different altitudes. Summary statistics for the re- sults obtained during measurements on both legs, the right and left leg, with eyes open and closed, and in the evening and morning, respectively. (DOCX) S4 Table. Effect of altitude on medial-lateral sway: ordinal logistic regression analysis. Re- sults of univariate and multivariate logistic regression analysis performed on medial-lateral sway amplitude with altitude, pulse oximetry, altitude exposure sequence, consecutive number of tests, eyes open/closed, evening/morning, and age as independent variables. (DOCX) Author Contributions Conceived and designed the experiments: KS TDL CML RC RH MK PA KEB. Performed the experiments: KS TDL CML KEB. Analyzed the data: KS TDL CML RC RH MK PA KEB. Con- tributed reagents/materials/analysis tools: KS TDL CML RC RH MK PA KEB. Wrote the paper: KS TDL CML RC RH MK PA KEB. Conceived and designed the experiments: KS TDL CML RC RH MK PA KEB. Performed the experiments: KS TDL CML KEB. Analyzed the data: KS TDL CML RC RH MK PA KEB. Con- tributed reagents/materials/analysis tools: KS TDL CML RC RH MK PA KEB. Wrote the paper: KS TDL CML RC RH MK PA KEB. This might be re- lated to the flat shape of the oxygen-hemoglobin dissociation curve as the arterial oxygen partial pressures corresponding to 1630 and 2590 m resulting in poor sensitivity of pulse oximetry to subtle changes in arterial PO2. Consistent with the study of Baumgartner et al. [12], who re- ported that posturographic parameters remained impaired over the course of three consecutive days at 4559 m, we did not find an effect of the number of consecutive tests at altitude on COPL (Table 4). We therefore conclude that in regard to COPL, no acclimatization or learning effect was detectable during the four days at altitude. Correspondingly, as reported in our previous paper [13], the participants of the current study did not reveal significant changes in the mean nocturnal oxygen saturation and end-tidal PCO2 between night 1 and 2 at 1630 m, and an in- crease in mean nocturnal oxygen saturation of only 1% from night 1 to night 2 at 2590 m. Although we found that participants in the current study experienced impairments in postur- al control during their stay at altitude we failed to demonstrate a cognitive impairment (data re- ported in [12]). Studies have shown that during exposure to hypoxia oxygen delivery to the brainstem and cerebellum, areas involved in static control, is better defended than oxygen deliv- ery to cortical regions involved in cognitive processing [21]. Our findings are therefore more likely explaind by insufficient sensitivity of the cognitive tests than by a greater vulnerability to hypoxia of brain regions involved static control compared to regions involved in cognition. Since our study included predominantly young, male subjects the conclusions may not necessarily apply to older or female individuals or to patients with cardiorespiratory or neurologic diseases. In conclusion, this is the first field study investigating postural control over the course of several days at moderate altitudes corresponding to an elevation of many tourist destinations worldwide. Postural control was impaired already at 1630 m while several cognitive tests per- formed in the same subjects at 1630 m and 2590 m were not affected [12]. Since optimal pos- tural control is essential for many activities of daily living, and poor postural control contributes to the risk of falls in the elderly [1], the findings of the current study may have major implications for many persons exposed to mild hypobaric hypoxia at moderate altitude. S1 Data. (ZIP) S1 Fig. Individual changes in Center of Pressure Path Length (COPL) with ascent to alti- tude. The difference in COPL [cm] recorded at 1630 m and 2590 m, respectively, minus the corresponding value at 490 m is plotted for each subject (n = 51) at 1630 m and 2590 m (mean values of day 1 and 2 at altitude and of morning and evening measurements are illustrated). The bars representing individual values of the difference in COPL are sorted by size. Negative values indicate an improvement and positive values a decrease of postural control at altitude. The vertical line indicates the transition from negative to positive values. P values are indicated for comparisons (Wilcoxon signed ranks test) of median values at moderate altitude (morning, evening and day1, day2) with corresponding medians at 490 m. ns = not significant. (TIF) 10 / 12 Postural Control at Altitude S1 Protocol. Trial Protocol. (PDF) S1 Table. Center of pressure path length at different altitudes, single leg tests. Summary sta- tistics for the results obtained during measurements on the right and left leg, with eyes open and closed, and in the evening and morning, respectively. (DOCX) S2 Table. Anterior-posterior sway amplitude at different altitudes, single leg tests. Summa- ry statistics for the results obtained during measurements on the right and left leg, with eyes open and closed, and in the evening and morning, respectively. (DOCX) S2 Table. Anterior-posterior sway amplitude at different altitudes, single leg tests. Summa- ry statistics for the results obtained during measurements on the right and left leg, with eyes open and closed, and in the evening and morning, respectively. (DOCX) References 1. Muir SW, Berg K, Chesworth B, Klar N, Speechley M (2010) Quantifying the magnitude of risk for bal- ance impairment on falls in community-dwelling older adults: a systematic review and meta-analysis. PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 11 / 12 Postural Control at Altitude J Clin Epidemiol 63: 389–406. S0895-4356(09)00186-3 [pii]; doi: 10.1016/j.jclinepi.2009.06.010 PMID: 19744824 2. World Tourism Organization UNTWO (2014) 8th world congress on snow and mountain tourisms. Con- clusions. Available: http://dtxtq4w60xqpw.cloudfront.net/sites/all/files/pdf/conclusions_english_0 pdf. Accessed 2014 Oct 29. 3. United Nations Environmental Programme (2002) Mountains. Available: http://www.unep.org/ resourceefficiency/Home/Business/SectoralActivities/Tourism/WorkThematicAreas/ EcosystemManagement/Mountains/tabid/78801/Default aspx. Accessed 2014 Oct 29. 4. Browne SM, Halligan PW, Wade DT, Taggart DP (2003) Postoperative hypoxia is a contributory factor to cognitive impairment after cardiac surgery. J Thorac Cardiovasc Surg 126: 1061–1064. doi: 10. 1016/S0022 S0022522303006160 [pii]. PMID: 14566247 5. Regard M, Landis T, Casey J, Maggiorini M, Bartsch P et al. (1991) Cognitive changes at high altitude in healthy climbers and in climbers developing acute mountain sickness. Aviat Space Environ Med 62: 291–295. PMID: 2031628 6. West JB, Schoene RB, Luks AM, Milledge JS (2013) High altitude medicine and physiology. Boca Raton: CRC Press, Taylor & Francis Group. 584 p. 7. Fraser WD, Eastman DE, Paul MA, Porlier JA (1987) Decrement in postural control during mild hypoba- ric hypoxia. Aviat Space Environ Med 58: 768–772. PMID: 3632536 8. Holness DE, Fraser WD, Eastman DE, Porlier JA, Paul MA (1982) Postural stability during slow-onset and rapid-onset hypoxia. Aviat Space Environ Med 53: 647–651. PMID: 7115252 9. Nordahl SH, Aasen T, Owe JO, Molvaer OI (1998) Effects of hypobaric hypoxia on postural control. Aviat Space Environ Med 69: 590–595. PMID: 9641406 10. Wagner LS, Oakley SR, Vang P, Noble BN, Cevette MJ, et al. (2011) Hypoxia-induced changes in standing balance. Aviat Space Environ Med 82: 518–522. PMID: 21614865 11. Baumgartner RW, Bartsch P (2002) Ataxia in acute mountain sickness does not improve with short- term oxygen inhalation. High Alt Med Biol 3: 283–287. doi: 10.1089/152702902320604269 PMID: 12396882 12. Latshang TD, Lo Cascio CM, Stowhas AC, Grimm M, Stadelmann K, et al. (2013) Are Nocturnal Breath- ing, Sleep, and Cognitive Performance Impaired at Moderate Altitude (1,630–2,590 m)? Sleep 36: 1969–1976. doi: 10.5665/sleep.3242 PMID: 24293773 13. Stadelmann K, Latshang TD, Lo Cascio CM, Tesler N, Stoewhas AC, et al. (2013) Quantitative Changes in the Sleep EEG at Moderate Altitude (1630 m and 2590 m). PLoS One 8: e76945. doi: 10. PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015 References 1371/journal.pone.0076945 PONE-D-13-23155 [pii]. PMID: 24167552 14. Stoewhas AC, Latshang TD, Lo Cascio CM, Lautwein S, Tesler N, et al. (2013) Effects of acute expo- sure to moderate altitude on vascular function, metabolism and systemic inflammation. PLoS One 8: e70081. doi: 10.1371/journal.pone.0070081 PMID: 23936377 15. Clark RA, Bryant AL, Pua Y, McCrory P, Bennell K, et al. (2010) Validity and reliability of the Nintendo Wii Balance Board for assessment of standing balance. Gait Posture 31: 307–310. S0966-6362(09) 00664-X [pii]; doi: 10.1016/j.gaitpost.2009.11.012 PMID: 20005112 16. Bauer C, Groger I, Rupprecht R, Gassmann KG (2008) Intrasession reliability of force platform parame- ters in community-dwelling older adults. Arch Phys Med Rehabil 89: 1977–1982. S0003-9993(08) 00437-1 [pii]; doi: 10.1016/j.apmr.2008.02.033 PMID: 18929026 17. Ageberg E, Roberts D, Holmstrom E, Friden T (2003) Balance in single-limb stance in healthy subjects —reliability of testing procedure and the effect of short-duration sub-maximal cycling. BMC Musculos- kelet Disord 4: 14. doi: 10.1186/1471-2474-4-14 PMID: 12831402 18. Aerospace Medical Association ASCCAS (2008) Cabin cruising altitudes for regular transport aircraft. Aviat Space Environ Med 79: 433–439. PMID: 18457303 19. Cymerman A, Muza SR, Beidleman BA, Ditzler DT, Fulco CS (2001) Postural instability and acute mountain sickness during exposure to 24 hours of simulated altitude (4300 m). High Alt Med Biol 2: 509–514. doi: 10.1089/152702901753397072 PMID: 11809091 20. Redfern MS, Yardley L, Bronstein AM (2004) Visual influences on balance. J Anxiety Disord 15: 81– 94. 21. Binks AP, Cunningham VJ, Adams L, Banzett RB (2008) Gray matter blood flow change is unevenly distributed during moderate isocapnic hypoxia in humans. J Appl Physiol (1985) 104: 212–217. 00069.2007 [pii]; doi: 10.1152/japplphysiol.00069.2007 PMID: 17991793 12 / 12 PLOS ONE \| DOI:10.1371/journal.pone.0116695 February 27, 2015
https://openalex.org/W1584369701	https://europepmc.org/articles/pmc4490214?pdf=render	English	null	Theory of Mind and Context Processing in Schizophrenia: The Role of Social Knowledge	Frontiers in psychiatry	2,015	cc-by	7,451	Theory of mind and context processing in schizophrenia: the role of social knowledge Maud Champagne-Lavau1,2* and Anick Charest 3 1 Aix-Marseille Université, CNRS, LPL UMR7309, Aix-en-Provence, France, 2 Department of Psychiatry, University of Montreal, QC, Canada, 3 Hôpital du Sacré-Cœur de Montréal, Pavillon Albert-Prévost, Montreal, QC, Canada The present study sought to determine whether social knowledge such as speaker occu- pation stereotypes may impact theory of mind (ToM) ability in patients with schizophrenia (SZ). Thirty individuals with SZ and 30 matched healthy control (HC) participants were tested individually on their ToM ability using a paradigm showing that stereotypes such as speaker occupation influences the extent to which speaker ironic intent is understood. ToM ability was assessed with open questions on the speaker ironic intent, irony rating, and mockery rating. Social perception was also assessed through politeness rating. The main results showed that SZ participants, like HC participants, were sensitive to the social stereotypes. They used these stereotypes adequately to attribute mental states such as speaker ironic intent to a protagonist while they found it difficult to explicitly judge and attribute negative attitude and emotion, as evidenced by mockery rating. No difference was found between the two groups regarding social perception ability. These performances were not associated with clinical symptoms. The integration of contextual information seems to be a good target for cognitive remediation aiming to increase social cognition ability. Reviewed by: Reviewed by: Bernhard J. Mitterauer, Volitronics-Institute for Basic Research Psychopathology and Brain Philosophy, Austria Mihaly Hajos, Yale University School of Medicine, USA Eszter Varga, University of Pécs, Hungary *Correspondence: Maud Champagne-Lavau, LPL, CNRS UMR 7309, Aix-Marseille Université, 5 Av. Pasteur, B.P. 80975, Aix-en-Provence 13604, France maud.champagne-lavau@univ-amu.fr Introduction A deficit in social cognition – including theory of mind (ToM), emotion processing, social perception, and attributional bias – is one of the most disabling clinical characteristics of schizophrenia (SZ) (1). For this reason, social cognition has become a high priority domain of research for the study of SZ (2). It has been included as one of the seven domains represented in the MATRICS Consensus Cognitive Battery for clinical trials in SZ (2). In addition, results from recent meta-analysis and literature review have pointed out that there are strong correlations between social cognition and outcome, underly- ing that these correlations were strongest with ToM and, to a lesser degree, with social perception (1, 3). Thus, given the prominence of the impairments in social functioning in SZ, it is important to understand the cognitive processes underlying impaired social cognition in this disorder. This paper focuses on two aspects of social cognition, namely ToM and social perception, and particularly on the impact of social contextual information on ToM abilities in SZ.h Specialty section: This article was submitted to Schizophrenia, a section of the journal Frontiers in Psychiatry Received: 28 January 2015 Accepted: 19 June 2015 Published: 03 July 2015 Original Research published: 03 July 2015 doi: 10.3389/fpsyt.2015.00098 Keywords: theory of mind, ironic intent, social knowledge, context, stereotypes, social perception, schizophrenia Edited by: Mihaly Hajos, Yale University School of Medicine, USA Citation: Few studies (33) have dealt with the role of social knowledge in ToM ability. However, social knowledge is also a kind of contextual information that contributes to the attribution of mental states to others (34). Social knowledge refers “to awareness of the roles, rules, and goals that characterize social situations and guide social interactions” [(35): p. 1212, Ref. (1, 33, 36)]. These findings related to social knowledge and social percep- tion were used in the present study to determine whether social knowledge influences the attribution of intention by patients with SZ. As far as we know, no research investigating ToM ability via irony comprehension in SZ has assessed the impact of social knowledge on such ability. Accordingly, the present study aimed at determining whether social knowledge such as stereotypes (type of speaker’s occupation) – which has been demonstrated to be social factors that cue speaker ironic intent in healthy individuals – also cues comprehension of ironic intent in SZ. It also aimed at assess- ing social perception in SZ participants. To these aims, we used a task of irony understanding manipulating contextual information according to the presence of a speaker occupation cueing or not ironic intent. Based on previous results, we hypothesized that difficulty in using social contextual data that cue speaker ironic intent will have an impact on the ToM ability of SZ participants. To our knowledge, only the work by Corrigan and Penn and their colleagues (see below) directly investigated the processing of social knowledge and social perception in SZ. Social perception refers to the ability to identify social cues from the behavior of others in a given social context. Tasks used to investigate social perception abilities generally assess the ability of a person to identify social roles, social context, and rules of a society (37, 38). They require to process social cues which can be concrete (i.e., actions, roles, and dialogs explicitly displayed by the protagonists) or abstract (i.e., knowledge inferred from action and dialog of the protagonists such as affects, rules, and goals) [(39, 40), see Ref. (41) for examples]. The authors found that, compared to healthy participants, individuals with SZ were impaired in social perception (42). These disorders have been related to a difficulty in processing the social context (38). Citation: Champagne-Lavau M and Charest A (2015) Theory of mind and context processing in schizophrenia: the role of social knowledge. Front. Psychiatry 6:98. doi: 10.3389/fpsyt.2015.00098 Theory of mind is the ability to form representations of other people’s mental states (e.g., intention, thought, belief) and to use these representations to understand, predict, and judge their statements and behaviors (4, 5). A number of studies have reported impaired ToM ability in patients with SZ [see Ref. (6–9) for a review] assessed in natural communication situations (10, 11) and also Front. Psychiatry 6:98. doi: 10.3389/fpsyt.2015.00098 July 2015 \| Volume 6 \| Article 98 Frontiers in Psychiatry \| www.frontiersin.org 1 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest in tasks assessing first- and second-order false beliefs (12–15), irony comprehension (16–21), hinting comprehension (22–24), and picture-sequencing tasks (25, 26). This ToM impairment in SZ patients has been supposed to result from their inability to process contextual information (27–31). But research investigating this hypothesis in SZ has only used non-social contextual information using different kinds of visual and verbal tasks (27, 32). In line with the findings of Uhlhaas et al. (32) and Green et al. (29, 30), Champagne-Lavau et al. (21) showed that participants with SZ cor- rectly perceived contextual information, but they showed difficulty in correctly integrating this information, performing significantly worse than healthy participants when attributing mental states. people perceive members of certain occupations as likely to use irony while members of other occupations are perceived as unlikely to use irony. For example, one is more likely to interpret a statement as ironic and mocking if it is said by an actor rather than a clergyman (44, 50). According to Pexman and Olineck (44), “the occupation stereotype contributes to the ironic environment by indicating that the speaker is likely to have a negative attitude (tendency to be critical) and that such an attitude is likely to be indirectly expressed (through humor and insincerity)” (page 268). These authors also reported that saying something negative in an ironic way (i.e., “you are a wonderful singer” in a context where you sang and people started to throw things at you) is perceived as more polite than saying something in a direct way (i.e., “you are a horrible singer”). Thus, assessing the degree of perceived politeness is a way to assess social perception. Citation: Thus, because of their difficul- ties in identifying social information, individuals with SZ would have difficulties in recognizing the rules and social conventions associated with a given situation (41, 43). These difficulties would rather concern the abstract social cues than the concrete ones (39, 43). Moreover, these difficulties would be relatively independent of the symptomatology (38).h Frontiers in Psychiatry \| www.frontiersin.org Measures The data were analyzed using 2 groups (SZ, HC) × 2 statements (ironic, literal) × 2 speaker occupations (sarcastic, non-sarcastic) repeated-measures ANOVA. A Spearman correlation analysis was conducted on the performances (i.e., percentage of correct responses to the question on the speaker iconic intent, irony rating scale, mocking rating scale, politeness rating scale) of the SZ group for each condition to examine the relationship between ToM and symptoms. The alpha level was set at p < 0.05 for all the analyses. A task of irony understanding was used to test the attribution of intention. The stimuli were composed of 48 stories from Champagne-Lavau et al. (21) controlled for familiarity and plausibility. To assess how social knowledge, such as occupation stereotypes, influenced participants’ ToM ability, the context was manipulated according to the presence of an occupation cueing or not ironic intent (cf. Supplementary Material, for example). Following Pexman and Olineck (44), a no occupation condition was also devised to create filler stimuli. This condition, in which the speakers were identified by their first names, was included to prevent the participants from developing response strategy that would help them foresee job information. The speaker occupations were chosen following a pilot study conducted with the procedure used by Pexman and Olineck (44). Forty undergraduate students from the University of Montreal were recruited for this pilot study. For each occupation, they were asked to imagine that a member of that occupation got a flat tire on the way to work. They were then asked to rate, on a 7-point scale (1 = low probability, 7 = high prob- ability), the likelihood that the member of this occupation would make an ironic remark about the situation. Forty five occupations were tested in that pilot study. The following eight occupations were judged as having the highest probability (p > 4.50) of ironic remarks (“sarcastic occupations”): comedian, talk show host, actress, artist, movie critic, mechanic, plumber, and insurance agent (M = 4.91, SD = 0.43). The following eight occupations were judged as having the lowest probability (p < 3.50) of ironic remarks (“non-sarcastic occupations”): accountant, clergyman, scientist, librarian, soldier, waiter, bank teller, and veterinarian (M = 2.78, SD = 0.30). These two occupation conditions were significantly different [t(14) = 11.49; p < 0.0001]. Participantsh p The sample consisted of 60 participants: 30 individuals with SZ and 30 healthy control (HC) participants with no history of psychiatric disorders. SZ patients were diagnosed by clinicians according to DSM-IV diagnostic criteria. All SZ participants were outpatients who had been recruited from the pavillon Albert prévost of the Hôpital du Sacré-Coeur of Montréal. They were stable and on antipsychotic medication with a normal recommended range of dosage (the average chlorpromazine equivalent was 642 ± 651 mg/ day). The severity of symptoms was measured using the Positive and Negative Symptom Scale [PANSS; (51)]. HC participants had been recruited in the local community. They were matched with the SZ patients for age and educational level (cf. Table 1 for the demographic data). The two groups did not significantly differ with regard to age [t(58) = −0.94, p > 0.05], educational level [t(58) = −0.65, p > 0.05], and IQ estimated using the NART (52) [t(58) = −0.74, p > 0.05]. All participants were native French- speakers with no previous neurological history. The present study investigated whether social knowledge, which is a kind of social contextual information, impacts the attribu- tion of mental states to others. Irony understanding is a relevant paradigm to assess this issue in SZ, since social knowledge such as speaker occupation stereotypes is contextual information that has been found to influence the attribution of speaker intent (ToM) in healthy individuals (44). Stereotypes are knowledge people have on other people as members of a group (34, 45). They are stored in memory, processed and activated automatically and unconsciously (45, 46). They were shown to be present even in people with limited social experience and in people with impaired social behavior such as children with autism (47). Regarding irony processing, it has been demonstrated that, among several factors (e.g., level of incongruity between context and speaker’s utterance, prosody), speaker features (e.g., his/ her social status or his/her occupation) influence the extent to which ironic intent is perceived among healthy subjects (44, 48–50). More precisely, Katz and Pexman (50) pointed out that Written consent forms were obtained from all participants, according to ethics guidelines set out by the University of Montreal and the Hôpital du Sacré-Coeur de Montréal. July 2015 \| Volume 6 \| Article 98 Frontiers in Psychiatry \| www.frontiersin.org 2 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest Table 1 \| Demographic and clinical data. Measures The stories ended either with a literal statement (e.g., Marie has a poor memory) or with an ironic statement (e.g., Marie has a phenomenal memory) to prevent judgments at chance level. Thus, the materials for the experiment were 48 stories involving a 2 statements (ironic or literal) by 3 occupations (occupation that cues ironic intent, occupation that does not cue ironic intent, and no occupation) combination of conditions (cf. Supplementary Material, for example). The stories were presented in random order. To control for prosody, stimuli were presented on a sheet of paper. Each stimulus was placed in front of the participant and remained there throughout the reading and questioning so that participants did not have to remember it. This was done in order to minimize memory and attention requirements. Participantsh Schizophrenia Healthy control p-value Mean SD Mean SD Age 43.9 8.3 41.7 9.6 0.94 Educational level 12.8 1.9 13.2 2.4 0.80 Gender (male/ female) (17/13) (10/20) Duration of illness 17.3 8.9 PANSS (positive) 16.3 5.8 PANSS (negative) 15.8 6.3 PANSS (general) 33.4 10.2 NART 37.7 7 35.9 5.5 0.54 Table 1 \| Demographic and clinical data. In line with the study by Pexman and Olineck (44), participants were asked to read each of these 48 stimuli and then answer the following question: “What does X (the speaker) really mean?” Then, they had to judge on a 7-point scale if the speaker was being ironic (1 = not at all ironic, 7 = extremely ironic), if he/ she was mocking someone (1 = not at all mocking, 7 = extremely mocking), and if he/she was saying something polite (1 = not at all polite, 7 = extremely polite). Rating scales on irony and mockery were used to assess attribution of speaker intent while the question on politeness was used to assess social perception. All participants were tested individually over one session in a quiet room. Group Comparison on Theory of Mind Percentage of Correct Responses to the Open Questionh The 2 × 2 × 2 ANOVA on the percentage of correct responses to the questions on the speaker intent revealed a significant main effect of group [F(1, 58) = 6.52, p < 0.01], showing that SZ patients (71.9%) made more errors than HC participants (81.94%). There was a significant main effect of speaker occupation [F(1, 58) = 21.81, p < 0.0001]. The percentage of correct responses was higher when the speaker had a sarcastic occupation than when he/she had a non-sarcastic occupation. There was also a significant main effect of statement [F(1,58) = 58.58, p < 0.0001] showing a higher percentage of correct responses for the literal statements than for the ironic ones. The speaker occupation × statement interaction was significant [F(1,58) = 17.12, p < 0.001] reveal- ing that, for ironic statements, irony ratings were higher when the speaker had a sarcastic occupation than when he/she had a non-sarcastic occupation (p < 0.0001) whatever the group, while for literal statements such difference did not exist (p > 0.05). The group × speaker occupation × statement [F(1,58) = 1.59, p > 0.05], the group × statement [F(1,58) = 1.54, p > 0.05], and the speaker occupation × statement interactions [F(1,58) = 0.87, p > 0.05] were not significant (cf. Figure 1). Frontiers in Psychiatry \| www.frontiersin.org Mocking ratingh effect of statement [F(1,58) = 196.46, p < 0.0001] showing that irony ratings were higher when the statement was ironic than when it was literal. There was no main effect of group [F(1, 58) = 1.52, p > 0.05]. The group × speaker occupation × statement interaction was not significant [F(1, 58) = 0.02, p > 0.05]. The speaker occu- pation × statement interaction was significant [F(1, 58) = 14.71, p < 0.0001], showing that, for ironic statements, irony ratings were higher when the speaker had a sarcastic occupation than when he/she had a non-sarcastic occupation (p < 0.0003) whatever the group, while for literal statements such difference did not exist (p > 0.05). The speaker occupation × group interaction was also significant [F(1,58) = 6.62, p < 0.01], revealing that, whatever the type of statement, there was no difference between the sarcastic and non-sarcastic occupation conditions (p > 0.05) in the SZ group, while such difference was present in the HC group (p < 0.001). And finally, the group × statement interaction was significant [F(1,58) = 12.91, p < 0.001]. In both groups, irony ratings were higher when the statement was ironic than when it was literal (SZ group: p < 0.001; HC group: p < 0.001) (cf. Figure 2). g g The 2 × 2 × 2 ANOVA on the ratings of the extent to which the speaker was mocking someone revealed a main effect of the state- ment [F(1,58) = 30.70, p < 0.001] showing that mocking ratings were higher for the ironic statements than for the literal ones. There was a marginal effect of group [F(1,58) = 3.43, p = 0.06] and no main effect of speaker occupation [F(1,58) = 2.22, p > 0.05]. The speaker occupation × statement interaction was significant [respectively F(1, 58) = 12.10, p < 0.001] revealing that, for ironic statements, irony ratings were higher when the speaker had a sarcastic occupation than when he/she had a non-sarcastic occupa- tion (p < 0.0001) whatever the group, while for literal statements such difference did not exist (p > 0.052). The group × statement interaction was significant [F(1,58) = 20.10, p < 0.0001]. Rating Scales Irony ratingh The 2 × 2 × 2 ANOVA on the ratings of the extent to which the speaker was ironic revealed a significant main effect of speaker occupation [F(1, 58) = 9.20, p < 0.01]. Irony ratings were higher when the speaker had a sarcastic occupation than when he/she had a non-sarcastic occupation. There was also a significant main July 2015 \| Volume 6 \| Article 98 3 Champagne-Lavau and Charest Theory of mind and context processing in schizophrenia Figure 2 \| Level of irony of statement for each type of occupation according to the type of participant. Figure 1 \| Percentage of correct responses to the open question on ironic speaker intent in schizophrenia and healthy control participants. Figure 1 \| Percentage of correct responses to the open question on ironic speaker intent in schizophrenia and healthy control participants. gure 1 \| Percentage of correct responses to the open question on ironic speaker intent in schizophrenia and healthy e of correct responses to the open question on ironic speaker intent in schizophrenia and healthy control participants Figure 1 \| Percentage of correct responses to the open question on ironic speaker intent in schizophrenia and healthy control participants. Figure 2 \| Level of irony of statement for each type of occupation according to the type of participant. \| Figure 2 \| Level of irony of statement for each type of occupation according to the type of participant. Mocking ratingh In the SZ group, the mocking ratings were the same for the literal statements and the ironic ones (p > 0.05), while in the HC group, the mocking ratings were higher for the ironic statements than for the literal ones (p < 0.001), whatever the speaker occupation. The group × speaker occupation × statement interaction [F(1, 58) = 0.37, p > 0.05] July 2015 \| Volume 6 \| Article 98 Frontiers in Psychiatry \| www.frontiersin.org 4 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest and the group × speaker occupation interaction [F(1, 58) = 0.60, p > 0.05] were not significant (cf. Figure 3). politeness ratings were higher when the speaker had a non-sarcastic occupation than when he/she had a sarcastic occupation. There was also a significant main effect of statement [F(1,58) = 98.66, p < 0.0001] showing that politeness ratings were higher when the statements were ironic than when they were literal. There was no main effect of group [F(1,58) = 0.03, p > 0.05]. The group × speaker occupation interaction was significant [F(1,58) = 5.77, p < 0.05], showing that in the SZ group, the politeness ratings were higher when the speaker had a non-sarcastic occupation than when he/ she had a sarcastic occupation (p < 0.0002), while such difference did not exist in the HC group (p > 0.05), whatever the type of statement. The group × occupation × statement [F(1,58) = 0.16, p > 0.05], the group × statement [F(1,58) = 0.03, p > 0.05] and the occupation × statement interactions [F(1,58) = 0.32, p > 0.05] were not significant (cf. Figure 4). To sum up, patients with SZ had lower performances than HC participants on the open question on the speaker intent in all the conditions. This does not lead to conclude that they have any impairment in attributing ironic intent. Using the irony rating scales, they showed performances similar to those of HC control participants, except they rated the statements in the same way for both speaker occupation conditions, whatever the type of statement. Surprisingly, SZ patients also judged mockery in the same way in all the conditions, meaning they rated both ironic and literal statements as much as mocking (around 4), by contrast to HC participants who rated ironic statements as more mocking than the literal ones. Discussion participants were also able to use social contextual information to attribute ironic intent to the speakers in the stories. Thus, social knowledge, such as speaker occupation stereotypes, seemed to help them in understanding speaker ironic intent, probably because ste- reotypes are social knowledge that is automatically activated. They do not require inferences. Sarcastic occupations are associated to specific characteristics (e.g., speaker’s tendencies to be humorous, critical, insincere, and to have lower levels of education), which are activated automatically without any inference when the participant thinks of the occupation. Such occupation stereotypes contribute to the ironic environment (44).h The present study sought to determine whether social knowledge, such as speaker occupation stereotypes, may impact ToM ability in patients with SZ. We used a psycholinguistic paradigm showing that stereotypes, such as speaker occupation, influence the extent to which speaker ironic intent is understood. ToM ability was assessed with open questions on the speaker ironic intent, irony rating, and mockery rating while social perception was assessed with politeness rating.h g The main results revealed that, like HC participants, SZ participants were sensitive to the social stereotypes when they judged the statements and when they answered the open question on the speaker ironic intent. Indeed both groups found that the speaker meant something ironic more often when the speaker had a sarcastic occupation than when he had a non-sarcastic occupa- tion. Interestingly, SZ participants seemed to adequately use these stereotypes to cue ironic intent since they did not show specific difficulties in attributing ironic intent to the speaker of the stories. Indeed, although they performed worse than HC participants, committing more errors than HC participants in answering the open questions, there was no difference across the different condi- tions. While such result disagrees with that of previous studies showing ToM difficulty with paradigms involving irony or sarcasm understanding (17–21), it is consistent with the study of Varga et al. (53) and with the results of Rapp et al. (54) finding relatively good performances (85% of correct responses) in SZ. This is also probably the reason why our result disagrees with those of Corrigan and Penn and colleagues showing that SZ participants had specific impairment in recognizing and in using abstract social knowledge requiring inferences from the social situation (41). Contrary to the abstract social knowledge tested by these authors, stereotypes are social knowledge that does not require inference. The Case of Mockery Rating y g Surprisingly here, by contrast with irony rating, SZ patients rated the extent to which the speaker was mocking someone or not in the same way (around 4), the statement being ironic or literal and the speaker occupation being sarcastic or not. To our knowledge, no research has investigated how SZ patients understand or feel mockery. Our result is probably to be linked with research on emotion recognition and emotion judgment by SZ patients since irony is used to express negative emotion (55). There is a general consensus in the literature on SZ processing of emotion, report- ing that SZ participants have more difficulties in recognizing and judging the negative emotions than the positive ones (56). In these studies, SZ participants had to attribute an emotional state to a face, for example, while in our study, participants had to attribute negative attitude and emotion to the speaker. The attribution of a negative attitude to the speaker is probably more explicit in the case of mockery rating than in the case of irony rating, as it is the case when SZ participants are asked to judge emotions of a face. This hypothesis would account for the bad performance of SZ participants when they had to judge mockery while they succeeded in judging irony. Surprisingly, while we expected the same performances for irony rating and mockery rating, since both scales are supposed to measure attribution of mental states, the results showed that, in contrast to HC participants, SZ patients judged the extent to which the speaker was mocking in the same way whatever the four conditions. Finally, SZ participants did not show any difficulty regarding social perception according to politeness rating in the present study. Group Comparison on Social Perception Politeness Ratingh Association of ToM performances with clinical symptoms No correlation was found between ToM performances and PANSS in the SZ group. The 2 × 2 × 2 ANOVA on the ratings of the extent to which a speaker said something polite revealed a significant main effect of speaker occupation [F(1, 58) = 19.54, p < 0.0001] showing that Figure 3 \| Level of mockery of statement for each type of occupation according to the type of participant. Figure 4 \| Level of politeness of statement for each type of occupation according to the type of participant. p g p gif of speaker occupation [F(1, 58) = 19.54, p < 0.0001] showing that p in the SZ group. Figure 3 \| Level of mockery of statement for each type of occupation according to the type of participant. Figure 3 \| Level of mockery of statement for each type of occupation according to the type of participant. Figure 4 \| Level of politeness of statement for each type of occupation according to the type of participant. Figure 4 \| Level of politeness of statement for each type of occupation according to the type of participant. Figure 4 \| Level of politeness of statement for each type of occupation according to the type of participant. Frontiers in Psychiatry \| www.frontiersin.org Frontiers in Psychiatry \| www.frontiersin.org July 2015 \| Volume 6 \| Article 98 5 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest Discussion In addition, following the results of Hirschfeld et al. (47), such social knowledge would be acquired and used outside any social engagement, in contrast to the abstract knowledge tested by Corrigan and Nelson (41). For these reasons (automaticity and apart from social engagement), SZ participants would be able to attribute ironic intent depending on the speaker occupation stereotypes. Frontiers in Psychiatry \| www.frontiersin.org Influence of the Knowledge of Social Stereotypes on ToM Abilityh The main results of the present study concur with previous findings in healthy individuals, showing that ironic statements uttered by a speaker with a sarcastic occupation are better understood as ironic and judged as more mocking than ironic statements uttered by a speaker with a non-sarcastic occupation (44). In other words, this means that the knowledge of social stereotypes, such as the speaker’s occupation, influences the extent to which we attribute ironic intent. In the present study, like HC participants, SZ patients seemed also sensitive to such social knowledge, showing different performances according to the speaker occupation. This result is in line with those of previous studies showing such sensitivity to contextual information (21, 30, 32, 36, 53). More specifically, Varga et al. (53) found that a linguistic help inserted into the context improves the SZ patients’ ability to understand irony. Indeed, SZ patients performed as well as HC participants when a word cueing irony (e.g., Sarah furiously remarks: “You did a pretty haircut” meaning “you did an ugly haircut”) was added in the context, while they performed worse than the HC participants when the stimuli did not contain such linguistic help. In the present research, SZ In addition, research has also shown that SZ patients with positive symptoms exhibited this specific impairment regarding negative emotion (57). However, our data do not support this spe- cific profile since we did not find any correlation between mockery rating and the positive symptoms measured with the PANSS. Social Perception using Politeness Rating Social Perception using Politeness Rating Social perception was assessed through the recognition of polite- ness. Participants were asking to rate the extent to which the speaker of the story was saying something polite or not. To our knowledge, none of the tests previously used to investigate social July 2015 \| Volume 6 \| Article 98 6 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest perception in SZ assessed politeness perception. The main results showed that SZ patients and HC participants rated politeness in the same way. Indeed, they rated the ironic statements as more polite than the literal ones. And statements uttered by a speaker with a non-sarcastic occupation were rated as more polite than statements uttered by a speaker with a sarcastic occupation. These results confirm those of Pexman and Olineck (44) and suggest that SZ patients did not have difficulties with social perception. occupation concerns “the others” rather than “the other” as an individual. Thus, during an in-live communication, if a SZ patient knows that that the interlocutor has this or that occupation it may cue the attribution of ironic intent to his/her interlocutor. But other kind of contextual information, such as information on his/her interlocutor as a specific individual, will not be integrated leading to disturbance in the interaction. Results of the present study have important consequences in terms of cognitive remediation since ToM ability could have an impact on social functioning in SZ (1, 3). The integration of contextual information seems to be a good target for cognitive remediation aiming to increase the social cognition ability. fi Although politeness is part of the rules that govern social situa- tions referring to abstract social cues as defined by Corrigan et al. (40), it was correctly identified by the SZ patients in our study. Such finding diverges from the results of some of the studies by Corrigan and colleagues (39, 43). However, it is consistent with the result obtained by Corrigan et al. (40) showing that, in contrast to SZ inpatients, SZ outpatients are able to correctly identify both abstract and concrete cues. Indeed, our study only included outpa- tients. Such absence of failure may be due to symptom remission, as suggested by Corrigan et al. (40). Social Perception using Politeness Rating Given that an effect of the speaker occupation was also found in both SZ and HC group, it is also possible that speaker occupation stereotypes help SZ patients to rate politeness. Author Contributions All authors significantly contributed to the conception of the study, the analysis, and the interpretation of the data. All authors contributed to the final version of the manuscript. Acknowledgments This study was supported by the Fonds de la Recherche en Santé du Québec (New investigator and operating grant #5793) and the Natural Sciences and Engineering Research Council of Canada (Operating grant #327313-06) to MCL. We would like to thank Dr. Guy Blouin and Dr. Jean-Pierre Rodriguez for their help in patient recruitment. We also thank Prof. Eric Fakra for his helpful comments on the manuscript. In conclusion, the present study showed that individuals with SZ may be able to attribute mental states, such as speaker ironic intent, to a protagonist while they find it difficult to explicitly judge and attribute negative attitude, as evidenced by mockery rating. These findings provide a meaningful argument that the ability to attribute intention to others varies as a function of contextual information processing in SZ. While SZ patients fail to integrate non-social contextual information to attribute ironic intent (21), this attribution may be facilitated by social contextual information, such as stereotypes, whose distinctive feature is that they are automatically activated. We have to notice that social context information and particularly stereotypes on the speaker’s Supplementary Material The Supplementary Material for this article can be found online at http://journal.frontiersin.org/article/10.3389/fpsyt.2015.00098 10. Champagne-Lavau M, Fossard M, Martel G, Chapdelaine C, Blouin G, Rodriguez JP, et al. Do patients with schizophrenia attribute mental states in a referential communication task? Cogn Neuropsychiatry (2009) 14:217–39. doi:10.1080/13546800903004114 10. Champagne-Lavau M, Fossard M, Martel G, Chapdelaine C, Blouin G, Rodriguez JP, et al. Do patients with schizophrenia attribute mental states in a referential communication task? Cogn Neuropsychiatry (2009) 14:217–39. doi:10.1080/13546800903004114 References 1. Couture SM, Penn DL, Roberts DL. The functional significance of social cognition in schizophrenia: a review. Schizophr Bull (2006) 32(Suppl 1):S44–63. doi:10.1093/ schbul/sbl029 11. McCabe R, Leudar I, Antaki C. Do people with schizophrenia display theory of mind deficits in clinical interactions? Psychol Med (2004) 34:401–12. doi:10.1017/ S0033291703001338 11. McCabe R, Leudar I, Antaki C. Do people with schizophrenia display theory of mind deficits in clinical interactions? Psychol Med (2004) 34:401–12. doi:10.1017/ S0033291703001338 2. Green MF, Penn DL, Bentall R, Carpenter WT, Gaebel W, Gur RC, et al. Social cognition in schizophrenia: an NIMH workshop on definitions, assessment, and research opportunities. Schizophr Bull (2008) 34:1211–20. doi:10.1093/schbul/ sbm145 12. Mazza M, De Risio A, Surian L, Roncone R, Casacchia M. Selective impairments of theory of mind in people with schizophrenia. Schizophr Res (2001) 47:299–308. doi:10.1016/S0920-9964(00)00157-2 3. Fett AKJ, Viechtbauer W, Dominguez MD, Penn DL, van Os J, Krabbendam L. The relationship between neurocognition and social cognition with functional outcomes in schizophrenia: a meta-analysis. Neurosci Biobehav Rev (2011) 35:573–88. doi:10.1016/j.neubiorev.2010.07.001 3. Fett AKJ, Viechtbauer W, Dominguez MD, Penn DL, van Os J, Krabbendam L. The relationship between neurocognition and social cognition with functional outcomes in schizophrenia: a meta-analysis. Neurosci Biobehav Rev (2011) 35:573–88. doi:10.1016/j.neubiorev.2010.07.001 13. Brüne M. Emotion recognition, ‘theory of mind,’ and social behavior in schizo- phrenia. Psychiatry Res (2005) 133:135–47. doi:10.1016/j.psychres.2004.10.007 h 14. Bora E, Sehitoglu G, Aslier M, Atabay I, Veznedaroglu B. Theory of mind and unawareness of illness in schizophrenia: is poor insight a mentalizing deficit? Eur Arch Psychiatry Clin Neurosci (2007) 257:104–11. doi:10.1007/s00406-006- 0681-3 j 4. Premack D, Woodruff G. Does the chimpanzee have a theory of mind? Behav Brain Sci (1978) 4:515–26. doi:10.1017/S0140525X00076512 5. Baron-Cohen S, Leslie AM, Frith U. Does the autistic child have a “theory of mind”? Cognition (1985) 21:37–46. doi:10.1016/0010-0277(85)90022-8 5. Baron-Cohen S, Leslie AM, Frith U. Does the autistic child have a “theory of mind”? Cognition (1985) 21:37–46. doi:10.1016/0010-0277(85)90022-8 15. Champagne-Lavau M, Stip E. Pragmatic and executive dysfunction in schizophre- nia. J Neurolinguistics (2010) 23:285–96. doi:10.1016/j.jneuroling.2009.08.009 6. Frith CD. Schizophrenia and theory of mind. Psychol Med (2004) 34(3):385–9. doi:10.1017/S0033291703001326 h 6. Frith CD. Schizophrenia and theory of mind. Psychol Med (2004) 34(3):385–9. doi:10.1017/S0033291703001326 16. Langdon R, Coltheart M, Ward PB, Catts SV. Disturbed communication in schizophrenia: the role of poor pragmatics and poor mind-reading. Psychol Med (2002) 32:1273–84. doi:10.1017/S0033291702006396 7. Brüne M. References Theory of mind in schizophrenia: a review of the literature. Schizophr Bull (2005) 31:21–42. doi:10.1093/schbul/sbi002 7. Brüne M. Theory of mind in schizophrenia: a review of the literature. Schizophr Bull (2005) 31:21–42. doi:10.1093/schbul/sbi002 8. Harrington L, Siegert RJ, McClure J. Theory of mind in schizophrenia: a critical review. Cogn Neuropsychiatry (2005) 10:249–86. doi:10.1080/13546800444000056 h 17. Leitman DI, Ziwich R, Pasternak R, Javitt DC. Theory of mind (ToM) and counterfactuality deficits in schizophrenia: misperception or misinterpretation? Psychol Med (2006) 36:1075–83. doi:10.1017/S0033291706007653 8. Harrington L, Siegert RJ, McClure J. Theory of mind in schizophrenia: a critical review. Cogn Neuropsychiatry (2005) 10:249–86. doi:10.1080/13546800444000056 9. Sprong M, Schothorst P, Vos E, Hox J, van Engeland H. Theory of mind in schizophrenia: meta-analysis. Br J Psychiatry (2007) 191:5–13. doi:10.1192/bjp. bp.107.035899 y 18. Kosmidis MH, Aretouli E, Bozikas VP, Giannakou M, Ioannidis P. Studying social cognition in patients with schizophrenia and patients with frontotemporal 18. Kosmidis MH, Aretouli E, Bozikas VP, Giannakou M, Ioannidis P. Studying social cognition in patients with schizophrenia and patients with frontotemporal July 2015 \| Volume 6 \| Article 98 Frontiers in Psychiatry \| www.frontiersin.org 7 Theory of mind and context processing in schizophrenia Champagne-Lavau and Charest dementia: theory of mind and the perception of sarcasm. Behav Neurol (2008) 19:65–9. doi:10.1155/2008/157356 dementia: theory of mind and the perception of sarcasm. Behav Neurol (2008) 19:65–9. doi:10.1155/2008/157356 38. Penn DL, Ritchie M, Francis J, Combs D, Martin J. Social perception in schizo- phrenia: the role of context. Psychiatry Res (2002) 109:149–59. doi:10.1016/ S0165-1781(02)00004-5 19. Kern RS, Green MF, Fiske AP, Kee KS, Lee J, Sergi MJ, et al. Theory of mind deficits for processing counterfactual information in persons with chronic schizophrenia. Psychol Med (2009) 39:645–54. doi:10.1017/S0033291708003966 39. Corrigan PW, Green MF. Schizophrenic-patients sensitivity to social cues – the role of abstraction. Am J Psychiatry (1993) 150:589–94. doi:10.1176/ajp.150.4.589 39. Corrigan PW, Green MF. Schizophrenic-patients sensitivity to social cues – the role of abstraction. Am J Psychiatry (1993) 150:589–94. doi:10.1176/ajp.150.4.589 20. Sparks A, McDonald S, Lino B, O’Donnelle M, Green MJ. Social cognition, empa- thy and functional outcome in schizophrenia. Schizophr Res (2010) 122:172–8. doi:10.1016/j.schres.2010.06.011 40. Corrigan PW, Buican B, Toomey R. Construct validity of two tests of social cognition in schizophrenia. Psychiatry Res (1996) 63:77–82. doi:10.1016/0165-1781(96)02897-1 41. Corrigan PW, Nelson DR. Factors that affect social cue recognition in schizo- phrenia. Psychiatry Res (1998) 78:189–96. doi:10.1016/S0165-1781(98)00013-4 21. References Interpreting figurative statements: speaker occupation can change metaphor to irony. Metaphor Symb (1997) 12:19–41. doi:10.1207/ s15327868ms1201_3 h 50. Katz AN, Pexman PM. Interpreting figurative statements: speaker occupation can change metaphor to irony. Metaphor Symb (1997) 12:19–41. doi:10.1207/ s15327868ms1201_3 h 28. Hardy-Bayle MC, Sarfati Y, Passerieux C. The cognitive basis of disorganization symptomatology in schizophrenia and its clinical correlates: toward a pathogenetic approach to disorganization. Schizophr Bull (2003) 29:459–71. doi:10.1093/ oxfordjournals.schbul.a007019 51. Kay SR. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull (1987) 13:261–78. doi:10.1093/schbul/13.2.261 51. Kay SR. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull (1987) 13:261–78. doi:10.1093/schbul/13.2.261 52. Nelson HE. National Adult Reading Test (NART). Test Manuel. Windsor: NFER- Nelson (1982). 52. Nelson HE. National Adult Reading Test (NART). Test Manuel. Windsor: NFER- Nelson (1982). 29. Green MJ, Uhlhaas PJ, Coltheart M. Context processing and social cognition in schizophrenia. Curr Psychiatry Rev (2005) 1:11–22. doi:10.2174/1573400052953529 53. Varga E, Simon M, Tényi T, Schnell Z, Hajnal A, Orsi G, et al. Irony comprehension and context processing in schizophrenia during remission – a functional MRI study. Brain Lang (2013) 126:231–42. doi:10.1016/j.bandl.2013.05.017 53. Varga E, Simon M, Tényi T, Schnell Z, Hajnal A, Orsi G, et al. Irony comprehension and context processing in schizophrenia during remission – a functional MRI study. Brain Lang (2013) 126:231–42. doi:10.1016/j.bandl.2013.05.017 y y 30. Green MJ, Waldron JH, Coltheart M. Emotional context processing is impaired in schizophrenia. Cogn Neuropsychiatry (2007) 12:259–80. doi:10.1080/13546800601051847 hff 54. Rapp AM, Langohr K, Mutschler DE, Wild B. Irony and proverb comprehension in schizophrenia: do female patients “dislike” ironic remarks? Schizophr Res Treatment (2014) 2014:841086. doi:10.1155/2014/841086 i 31. Chung YS, Mathews JR, Barch DM. The effect of context processing on different aspects of social cognition in schizophrenia. Schizophr Bull (2011) 37:1048–56. doi:10.1093/schbul/sbq012 55. Roberts RM, Kreuz RJ. Why do people use figurative language? Psychol Sci (1994) 5:159–63. doi:10.1111/j.1467-9280.1994.tb00653.x q 32. Uhlhaas PJ, Phillips WA, Schenkel LS, Silverstein SM. Theory of mind and perceptual context-processing in schizophrenia. Cogn Neuropsychiatry (2006) 11:416–36. doi:10.1080/13546800444000272 56. Edwards J, Jackson HJ, Pattison PE. Emotion recognition via facial expression and affective prosody in schizophrenia: a methodological review. Clin Psychol Rev (2002) 22:789–832. doi:10.1016/S0272-7358(02)00162-9 33. Billeke P, Aboitiz F. Social cognition in schizophrenia: from social stimuli processing to social engagement. Front Psychiatry (2013) 4:4. doi:10.3389/ fpsyt.2013.00004 57. Mandal MK, Jain A, Haque-Nizamie S, Weiss U, Schneider F. References Champagne-Lavau M, Charest A, Anselmo K, Rodriguez JP, Blouin G. Theory of mind and context processing in schizophrenia: the role of cognitive flexibility. Psychiatry Res (2012) 200:184–92. doi:10.1016/j.psychres.2012.06.011 42. Corrigan PW, Penn DL. Social Cognition and Schizophrenia. 1st ed. Washington, DC: American Psychological Association (2001). 22. Corcoran R, Mercer G, Frith CD. Schizophrenia, symptomatology and social inference – investigating theory of mind in people with schizophrenia. Schizophr Res (1995) 17:5–13. doi:10.1016/0920-9964(95)00024-G y g 43. Corrigan PW, Garman A, Nelson D. Situational feature recognition in schizophrenic outpatients. Psychiatry Res (1996) 62:251–7. doi:10.1016/0165-1781(96)02798-9 23. Janssen I, Krabbendam L, Jolles J, van Os J. Alterations in theory of mind in patients with schizophrenia and non-psychotic relatives. Acta Psychiatr Scand (2003) 108:110–7. doi:10.1034/j.1600-0447.2003.00092.x 44. Pexman PM, Olineck KM. Understanding irony. How do stereotypes cue speaker intent? J Lang Soc Psychol (2002) 21:245–74. doi:10.1177/0261927X02021003003 45. Fiske ST. Stereotyping, prejudice, and discrimination at the seam between the centuries: evolution, culture, mind, and brain. Eur J Soc Psychol (2000) 30:299–322. doi:10.1002/(SICI)1099-0992(200005/06)30:3<299::AID-EJSP2>3.0.CO;2-F 24. Greig TC, Bryson GJ, Bell MD. Theory of mind performance in schizophrenia: diagnostic, symptom, and neuropsychological correlates. J Nerv Ment Dis (2004) 192:12–8. doi:10.1097/01.nmd.0000105995.67947.fc 46. Macrae CN, Bodenhausen GV. Social cognition: thinking categorically about others. Annu Rev Psychol (2000) 51:93–120. doi:10.1146/annurev.psych.51.1.93 25. Langdon R, Corner T, McLaren J, Ward PB, Coltheart M. Externalizing and per- sonalizing biases in persecutory delusions: the relationship with poor insight and theory-of-mind. Behav Res Ther (2006) 44:699–713. doi:10.1016/j.brat.2005.03.012 47. Hirschfeld L, Bartmess E, White S, Frith U. Can autistic children predict behavior by social stereotypes? Curr Biol (2007) 17:R451–2. doi:10.1016/j.cub.2007.04.051 h 26. Brune M, Abdel-Hamid M, Lehmkamper C, Sonntag C. Mental state attribution, neurocognitive functioning, and psychopathology: what predicts poor social competence in schizophrenia best? Schizophr Res (2007) 92:151–9. doi:10.1016/j. schres.2007.01.006 48. Ivanko SL, Pexman PM. Context incongruity and irony processing. Discourse Process (2003) 35:241–79. doi:10.1002/ijop.12003 48. Ivanko SL, Pexman PM. Context incongruity and irony processing. Discourse Process (2003) 35:241–79. doi:10.1002/ijop.12003 j 49. Bryant GA, Fox Tree JE. Recognizing verbal irony in spontaneous speech. Metaphor Symb (2002) 17:99–117. doi:10.1207/S15327868MS1702_2 i 49. Bryant GA, Fox Tree JE. Recognizing verbal irony in spontaneous speech. Metaphor Symb (2002) 17:99–117. doi:10.1207/S15327868MS1702_2 i 27. Schenkel LS, Spaulding WD, Silverstein SM. Poor premorbid social functioning and theory of mind deficit in schizophrenia: evidence of reduced context pro- cessing? J Psychiatr Res (2005) 39:499–508. doi:10.1016/j.jpsychires.2005.01.001 h 50. Katz AN, Pexman PM. References Generality and specificity of emotion-recognition deficit in schizophrenic patients with positive and negative symptoms. Psychiatry Res (1999) 87:39–46. doi:10.1016/ S0165-1781(99)00047-5 57. Mandal MK, Jain A, Haque-Nizamie S, Weiss U, Schneider F. Generality and specificity of emotion-recognition deficit in schizophrenic patients with positive and negative symptoms. Psychiatry Res (1999) 87:39–46. doi:10.1016/ S0165-1781(99)00047-5 y 34. Frith CD, Frith U. How we predict what other people are going to do. Brain Res (2006) 1079:36–46. doi:10.1016/j.brainres.2005.12.126 35. Green MF, Penn DL, Bentall R, Carpenter WT, Gaebel W, Gur RC, et al. Social cognition in schizophrenia: an NIMH workshop on definitions, assessment, and research opportunities. [Consensus Development Conference, NIH Research Support, N.I.H., Extramural]. Schizophr Bull (2008) 34:1211–20. doi:10.1093/ schbul/sbm145 Conflict of Interest Statement: 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. 36. Green MJ, Waldron JH, Simpson I, Coltheart M. Visual processing of social context during mental state perception in schizophrenia. J Psychiatry Neurosci (2008) 33:34–42. Copyright © 2015 Champagne-Lavau and Charest. 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) or licensor 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. 37. Toomey R, Schuldberg D, Corrigan P, Green MF. Nonverbal social perception and symptomatology in schizophrenia. Schizophr Res (2002) 53:83–91. doi:10.1016/ S0920-9964(01)00177-3 July 2015 \| Volume 6 \| Article 98 Frontiers in Psychiatry \| www.frontiersin.org
https://openalex.org/W4391900747	https://journal.unibos.ac.id/ijlf/article/download/3928/1925	Indonesian	null	EFEKTIVITAS PENERAPAN ASAS PRADUGA TAK BERSALAH PADA PROSES PERADILAN PIDANA DI PENGADILAN NEGERI WATAMPONE (PUTUSAN NOMOR: 114/Pid.B/2023/PN Wtp)	Indonesian journal of legality of law	2,023	cc-by	5,818	Effectiveness of the Application of the Principle of Presumption of Innocence in the Criminal Justice Process at the Watampone District Court (Decision Number: 114/Pid.B/2023/PN Wtp) Effectiveness of the Application of the Principle of Presumption of Innocence in the Criminal Justice Process at the Watampone District Court (Decision Number: 114/Pid.B/2023/PN Wtp) Baharuddin Badaru1, Ibnu Hibban Sabil2 1Program Studi Ilmu Hukum Fakultas Hukum Universitas Muslim Indonesia 2Program Studi Ilmu Hukum Program Pascasarjana Universitas Muslim Maros Email: baharuddin.badaru.fh@umi.ac.id Diterima: 10 Agustus 2023/Disetujui: 05 Desember 2023 ABSTRAK Penelitian bertujuan untuk mengkaji faktor-faktor yang mempengaruhi efektivitas penerapan asas praduga tak bersalah dalam proses peradilan pidana sebagai wujud penghormatan hak asasi manusia seperti perkara Putusan Nomor 114/Pid.B/2023/Pn Wtp yang mana perkara tersebut terjadi tindak pidana pembunuhan yang dilakukkan inisial IR terhadap korban yang tak lain adalah mertuanya inisial MK meninggal dunia. Hak-hak yang dimiliki tersangka sangat dijunjung tinggi, dan sedapat mungkin harus tetap diberikan oleh aparat penegak hukum. Tipe penelitian yang digunakan yaitu penelitian hukum empiris terhadap penerapan asas praduga tidak bersalah dalam proses penegakan hukum. Hasil penelitian menunjukan bahwa Pertimbangan hakim dalam memutuskan perkara tindak pidana pembunuhan nomor 114/Pid.B/2023/PN Wtp berdasarkan asas praduga tidak bersalah dengan memanfatkan Knowledge, Comprehensif, Aplikatif, Analysis, syntesis , evaluatif telah efektif namun saran penulis pada penelitian ini, perlu ditingkatkan sinergitas antara Para Penegak Hukum mulai dari Kepolisian, Kejaksaan sampai kepada Pengadilan dalam proses peradilan pidana tindak pidana pembunuhan berdasarkan asas praduga tidak bersalah dalam rangka mencegah tindak pidana baru seperti tindak pidana main hakim sendiri oleh Masyarakat. Kata Kunci: Asas Praduga Tak Bersalah, Peradilan Pidana, Perkara Pembunuhan ABSTRACT This study aims to examine the factors that influence the effectiveness of the application of the presumption of innocence in the criminal justice process as a form of respect for human rights such as Case No. 114/Pid.B/2023/Pn Wtp in which the case involved the crime of murder committed by the initials IR against the victim who was none other than his father-in-law MK died. The rights of the suspect are highly upheld, and as far as possible must still be given by law enforcement officials. The type of research used is empirical legal research on the application of the presumption of innocence in the law enforcement process. The results showed that the judge's consideration in deciding the criminal murder case number 114/Pid.B/2023/PN Wtp based on the presumption of innocence by utilizing Knowledge, Comprehensive, Applicative, Analysis, syntesis, evaluative has been effective but the author's suggestion in this study, it is necessary to increase the synergy between Law Enforcement Officers ranging from the Police, Prosecutors' Office to the Court in the criminal justice process of murder based on the presumption of innocence in order to prevent new crimes such as vigilantism by the community. Keywords: Presumption Of Innocence, Criminal Justice, Murder Case This work is licensed under Creative Commons Attribution License 4.0 CC-BY International license I.dn.J.of Law 6(1) : 185-191, Desember 2023 DOI: 10.35965/ijlf.v6i1.3928 I.dn.J.of Law 6(1) : 185-191, Desember 2023 DOI: 10.35965/ijlf.v6i1.3928 1. PENDAHULUAN dalam mewujudkan cita-cita ketenteraman dan ketertiban masyarakat. Dinamika kehidupan masyarakat yang menunjukkan adanya perilaku pelanggaran norma-norma atau kejahatan merupakan gejala kontra produktif dalam masyarakat. Kepentingan hukum telah memberikan hak dan kewajiban kepada setiap orang dan tentunya untuk memperoleh atau mewujudkan haknya, diperlukan rambu- rambu agar tidak bertentangan satu sama lainnya. Pada konteks tersebutlah hukum pidana hadir mengatur tata perilaku pribadi seseorang agar tetap serasi dan seimbang Dalam konteks terjadinya suatu perbuatan pidana, maka untuk menentukan salah tidaknya seseorang, setiap penegak 185 Efektivitas Penerapan Asas Praduga Tak Bersalah pada Proses Peradilan Pidana di (Baharuddin Badaru, Ibnu Hibban Sabil) hukum akan berpedoman pada ketentuan Undang-Undang Republik Indonesia Nomor 8 Tahun 1981 Tentang Hukum Acara Pidana atau biasa disingkat KUHAP. Asas Praduga Tak Bersalah merupakan salah satu asas yang melatar belakangi di rumuskannya pasal-pasal KUHAP. memiliki harkat dan martabat yang wajib dijunjung dalam proses penegakan hukum sebagaimana mestinya. Penerapan asas tersebut dalam proses peradilan pidana sangat penting sebagai wujud penghormatan hak asasi manusia, hal ini tentunya tergantung pula pada pemahaman penegakan hukum terhadap asas praduga tak bersalah, apabila asas tersebut tidak di terapkan maka akan membawa dampak berkurangnya kepercaya terhadap masyarakat terhadap pelaksanaan proses peradilan pidana yang seharusnya bertujuan untuk tegaknya hukum dan keadilan. Kehadiran asas praduga tak bersalah diprediksi muncul pada abad ke-XI. Asas ini lahir karena adanya ideologi individualistic-liberalistik. Dalam sistem Common Law, asas praduga tak bersalah merupakan syarat utama untuk menetapkan bahwa suatu proses telah dilaksanakan secara jujur, adil, dan tidak memihak. Asas praduga tak bersalah adalah bagian yang tidak bisa dipisahkan dari prinsip due process of law. Prinsip ini telah melembaga dalam proses peradilan sejak dua ratus tahun yang lalu dan telah melembaga pula dalam kehidupan sosial saat ini. Salah satu komponen yang ada dalam sistem peradilan pidana (Criminal Justice System) adalah Lembaga Kejaksaan. Lembaga Kejaksaan Republik Indonesia merupakan lembaga pemerintahan yang melaksanakan kekuasaan negara secara merdeka terutama dalam pelaksanaan tugas dan kewenangan di bidang penuntutan, melaksanakan tugas dan kewenangan dibidang penyidikan dan penuntutan perkara tindak pidana khusus seperti halnya korupsi dan pelanggaran Hak Asasi Manusia (HAM) berat serta kewenangan lain berdasarkan undang-undang, Dalam sistem peradilan pidana khususnya dalam hal penuntutan, seorang Jaksa harus memperhatikan mengenai hak-hak seorang terdakwa maupun korban. Penerapan asas praduga tak bersalah di Indonesia tercermin dalam Undang-Undang Republik Indonesia Nomor 8 Tahun 1981 Tentang HukumAcara Pidana (KUHAP) dan Undang-Undang Nomor 48 Tahun 2009 tentang Kekuasaan Kehakiman. 1. PENDAHULUAN Penjelasan Umum KUHAP butir ketiga huruf c menerangkan bahwa setiap orang yang disangka, ditangkap, ditahan, dituntut, dan atau dihadapkan di muka sidang pengadilan, wajib dianggap tidak bersalah sampai adanya putusan pengadilan yang menyatakan kesalahannya dan memperoleh kekuatan hukum tetap. g g p Hakim dalam sistem peradilan pidana di Indonesia, di wewenang oleh undang-undang untuk memutuskan seseorang bersalah atau tidak bersalah melalui proses penegakan hukum dan keadilan di pengadilan. Hal ini berarti bahwa kesalahan seseorang dapat di tentukan dalam suatu putusan hakim yang sudah mempunyai kekuatan tetap, oleh karena itu seseorang dianggap tidak bersalah sebelum sebelum ada putusan hakim yang mempunyai kekuatan tetap. Seperti dalam tindak pidana pembunuhan penerapan asas praduga tidak bersalah harus di junjung tinggi karena sebelum menyatakan motif yang sebenarnya terjadi maka belumlah bisa menyatakan pelakunya bersalah karena penyebab seseorang melakukan pembunuhan di latar belakangi oleh beberapa faktor seperti faktor stress, gangguan jiwa, balas dendam dan dalam keadaan mabuk Tindak pidana pembunuhan secara terminologi adalah perkara membunuh, perbuatan membunuh.Sedangkan dalam istilah KUHP pembunuhan adalah kesengajaan menghilangkan nyawa orang lain. Dari definisi tersebut, maka tindak pidana pembunuhan dianggap sebagai delik material bila delik tersebut selesai dilakukan oleh pelakunya dengan timbulnya akibat yang dilarang atau yang tidak dikehendaki oleh Undang-undang. Asas praduga tak bersalah bukan mengklasifikasikan tersangka/terdakwa sebagai orang yang tidak bersalah, melainkan memberikan jaminan HAM dalam proses peradilan. Sementara itu, Pasal 8 ayat (1) Undang-Undang Nomor 48 Tahun 2009 Tentang Kekuasaan Kehakiman menerangkan bahwa setiap orang yang disangka, ditangkap, ditahan, dituntut, atau dihadapkan di depan pengadilan wajib dianggap tidak bersalah sebelum ada putusan pengadilan yang menyatakan kesalahannya dan telah memperoleh kekuatan hukum tetap. Dari ketentuan-ketentuan tersebut, dapat disimpulkan bahwa secara sederhana pengertian asas praduga tak bersalah adalah asas yang menginginkan agar setiap orang menjalani proses perkara secara tidak bersalah hingga ada putusan pengadilan berkekuatan hukum tetap yang menyatakan kesalahan orang tersebut. Asas praduga tak bersalah tidak dapat dipisahkan dari proses peradilan pidana karena keduanya berkaitan erat. Proses peradilan pidana merupakan proses di mana sejak seseorang menjadi tersangka atau terdakwa dengan penangkapannyahingga adanya putusan hakim yang menyatakan kesalahannya. Dalam proses inilah asas praduga tak bersalah diterapkan. Sehubungan dengan kepentingan terdakwa karena dianggap “tidak bersalah”, hak-hak terdakwa atau tersangka harus dihormati sebagaimana mereka yang tidak bersalah. Dalam KUHP, ketentuan tindak pidana pembunuhan diatur dalam buku II bab XIX, yang terdiri dari 13 Pasal, yakni Pasal 338 sampai Pasal 350. 3. HASIL DAN PEMBAHASAN Perkara Pidana Nomor 114/Pid.B/2023/PN Wtp di Pengadilan Negeri watampone merupakan perkara tindak pidana pembunuhan dengan terdakwa berinisial IR yang terbukti secara sah dan meyakinkan bersalah melakukan tindak pidana pembunuhan sebagaimana dalam dakwaan primair melanggar pasal 338 KUHPidana dan mendapatkan sanksi pidana penjara kepada MK selama 12 tahun, dikurangkan seluruhnya selama terdakwa berada dalam tahanan sementara dengan perintah agar terdakwa tetap berada dalam tahanan. Penerapan asas praduga tidak bersalah menempatkan negara dalam hal ini diwakili oleh jaksa penuntut umum yang dibebani untuk membuktikan unsur-unsur pidana yang didakwakan, dan bukan sebaliknya terdakwa yang harus membuktikan ketidak- bersalahannya. Selanjutnya hakim yang akan memutus untuk menjatuhkan pidana kepada terdakwa apabila alat bukti tersebut secara limitatif ditentukan undang-undang dan didukung pula adanya keyakinan hakim terhadap eksistensinya alat-alat bukti tersebut. Tindak pidana pembunuhan terjadi pada hari jumat tangga10 Februari 2023 sekira pukul 13.30 WIB bertempat di Dusun Tirong Desa Tirong Kecematan Palakka Kabupaten Bone. Berawal ketika sakit hati IR membuat kandang ayam potong dibelakang rumah korban MK yang tak lain adalah mertuanya sendiri, semenjak terdakwa membuat kandang ayam potong terdakwa sering dikatai bodoh mengurus ayam oleh korban MK. Akibat dari perbuatan terdakwa, korban mendapat luka dan meninggal dunia berdasarkan Visum Et Repertum dari Rumah Sakit Umum Daerah Tenriawaru Kabupaten Bone Nomor; 350/28/III/RSU Tanggal 2 Maret 2023 yang ditanda tangani oleh dr.Yuarni Mellu, yang menerangkan bahwa pada hari jumat tanggal 10 Febuari 2023 jam 14.30 Wita telah melakukan pemeriksaan dan tindakan terhadap seseorang mayat dengan inisial MK dengan hasil pemeriksaan tampak luka robek dikepala bagian kanan tembus ke batok kepala, dan juga luka robek di leher bagian kanan mulai dari dagu belakang telinga kanan sampai leher. Seperti perkara Nomor 114/Pid.B/2023/Pn Wtp yang mana perkara tersebut sebagai berikut : telah terjadi tindak pidana pembunuhan yang di lakukan oleh inisial pelaku IR terhadap korban yang tak lain adalah mertuanya inisial MK meninggal dunia, tersangka melakukan pembunuhan dengan cara mendatangi rumah korban dengan membawa sebilah parang kemudian membunuh korban tepat di dalam kamarnya dengan luka robek di bagian kepala dan leher sehingga mengakibatkan korban mati ditempat, karna perbuatan tersangka di kenakan sanksi pidana yang di kenakan pasal 338 KUHP. 1. PENDAHULUAN Bentuk kesalahan tindak pidana menghilangkan nyawa orang lain ini dapat berupa sengaja (dolus) dan tidak sengaja (alpa). Kesengajaan (dolus) adalah suatu perbuatan yang dapat terjadi dengan direncanakan terlebih dahulu atau tidak direncanakan. Tetapi yang penting dari suatu peristiwa itu adalah adanya ”niat” yang diwujudkan melalui perbuatan yang dilakukan sampai selesai. Penerapan asas praduga tak bersalah bukan menitikberatkan pada salah atau tidaknya tersangka atau terdakwa. Pasalnya, sebagaimana yang diungkapkan Marc Weber Tobias, jika semua orang dianggap tidak bersalah, maka tidak seorang pun dapat disangka atau didakwa sampai penyelidikannya menghasilkan bukti bahwa orang tersebut telah melakukantindak pidana. Berdasarkan unsur kesalahan, tindak pidana pembunuhan dapat dibedakan menjadi: tindak pidana pembunuhan biasa tindak pidana yang diatur dalam Pasal 338 KUHP merupakan tindak pidana dalam bentuk pokok (Doodslag In Zijn Grondvorm), yaitu delik yang telah dirumuskan secara lengkap dengan semua unsur- unsurnya. Adapun rumusan Pasal 338 KUHP adalah: “Barangsiapa sengaja merampas nyawa orang lain, diancam, karena pembunuhan, dengan pidana penjara paling lama lima belas tahun”. Meskipun dengan adanya asas praduga tak bersalah ini, bukan berarti asas tersebut mengklasifikasikan tersangka/terdakwa sebagai orang yang tidak bersalah, melainkan memberikan jaminan hak asasi manusia (HAM) dalam proses peradilan. Selain itu, asas praduga tak bersalah juga merupakan dasar pemenuhan jaminan dan perlindungan hak asasi tersangka atau terdakwa sebagai makhluk yang 186 Idn. J. of Legality of law 6(1): 185-191,Desember 2023 Sedangkan Pasal 340 KUHP menyatakan: “Barang siapa sengaja dan dengan rencana lebih dahulu merampas nyawa orang lain diancam, karena pembunuhan dengan rencana (moord), dengan pidana mati atau pidana penjara seumur hidup atau selama waktu tertentu, paling lama dua puluh tahun. peradilan pidana karena seseorang tidak boleh dianggap bersalah sebelum adanya putusan pengadilan yang menyatakan bersalah kepadanya dan mempunyai kekuatan hukum. Hak-hak yang dimiliki tersangka sangat dijunjung tinggi, dan sedapat mungkin harus tetap diberikan oleh aparat penegak hukum. Pada pembunuhan biasa ini, Pasal 338 KUHP menyatakan bahwa pemberian sanksi atau hukuman pidananya adalah pidana penjara paling lama lima belas tahun. Di sini disebutkan “paling lama” jadi tidak menutup kemungkinan hakim akan memberikan sanksi pidana kurang dari lima belas tahun penjara. Penelitian bertujuan untuk mengkaji faktor-faktor yang mempengaruhi efektivitas penerapan asas praduga tak bersalah dalam proses peradilan pidana sebagai wujud penghormatan hak asasi manusia seperti perkara Putusan Nomor 114/Pid.B/2023/Pn Wtp. 2. METODE Penelitian ini menggunakan tipe penelitian hukum empiris, adapun Teknik pengumpulan data melalui wawancara dan literatur-literatur serta peraturan perundang-undangan yang berkaitan dalam penelitian ini, penulis melakukan penelitian di Pengadilan Negeri Watampone yang beralamatkan Jln.MTHaryonno, Macanang, Kec.Tanete Riattang Barat Kab.Bone, Sulawaesi Selatan 1. PENDAHULUAN p j Tindak pidana pembunuhan merupakan kejahatan yang menyebabkan hilangnya nyawa orang lain sehingga tidak mudah untuk mengungkapnya, dengan demikian asas praduga tak bersalah tetap di terapkan dalam proses penyelesaian perkara perkara tindak pidana pembunuhan di Pengadilan seperti di Pengadilan Negeri Bone, hakim Pengadilan Negeri Bone harus memahami terhadap konsep asas praduga tak bersalah karena selama proses peradilan masih berjalan seperti di Pengadilan Negeri, Pengadilan Tinggi dan Mahkamah Agung dan belum memperoleh kekuatan hukum tetap maka terdakwa belum dapat di katagorikan bersalah sebagai pelaku dari tindak pidana, sehingga selama proses peradilan pidana tersebut harus mendapatkan hak-haknya sebagaimana di atur undang-undang, selanjutnya asas praduga tidak bersalah yang dianut KUHAP, memberikan pedoman kepada Hakim untuk mempergunakan prinsip asas yang berkenaan dengan proses pemeriksaan terdakwa di Pengadilan (akusator). hakim menjauhkan diri dari cara-cara pemeriksaan yang “inkuisatur‟ atau inquirisatorial system‟ yang menempatkan terdakwa dalam pemeriksaan sebagai obyek yang dapat diperlakukan dengan sewenang-wenang. 3. HASIL DAN PEMBAHASAN Pengetahuan Walaupun sama-sama berpendidikan sarjana hukum sebagai syarat pekerjaan, masing-masing profesi mengikuti pendidikan yang berbeda berupa pendidikan calon hakim dan calon jaksa 2. Pemahaman Sistem rekrutmen jaksa dan hakim berbeda sehingga kualifikasi yang dibutuhkan juga berbeda. Akibatnya tingkat pemahaman antara jaksa dan hakim atas perundang-undangan berbeda 3. Penerapan Masing-masing institusi memiliki aturan teknis yang menerapkan pertauran perundang-undangan. Aturan teknis disusun instansi masing-masing yang cenderung mengarah pada instansi sentris, merasa instansinya lah yang paling betul dalam menerpakan ketentuan perundang- undangan. Jaksa dan hakim di lapangan tunduk pada aturan teknis dari pimpinan di tingkat pusat 4. Sintetis Satu ketentuan yang diterapkan berulang- ulang akan menjadi kebiasaan sehingga walaupun terdapat kesalahan dalam penerapan prosedur dianggap benar karena dianggap biasa 5. Evaluatif Praktek penerapan hukum akan dievaluasi menurut versi sendiri, sehingga mengarahkan pada pembenaran pendapat atas apa yang dilakukan walaupun salah 6. Analitis Kedua instansi tidak melakukan analisis secara komperhensif, sehingga kesalahan dalam memberi persepsi terus terjadi Tabel 1. Faktor Yang Mempengaruhi Persepsi No Faktor Yang Mempengaruhi Persepsi Analisis 1. Pengetahuan Walaupun sama-sama berpendidikan sarjana hukum sebagai syarat pekerjaan, masing-masing profesi mengikuti pendidikan yang berbeda berupa pendidikan calon hakim dan calon jaksa 2. Pemahaman Sistem rekrutmen jaksa dan hakim berbeda sehingga kualifikasi yang dibutuhkan juga berbeda. Akibatnya tingkat pemahaman antara jaksa dan hakim atas perundang-undangan berbeda 3. Penerapan Masing-masing institusi memiliki aturan teknis yang menerapkan pertauran perundang-undangan. Aturan teknis disusun instansi masing-masing yang cenderung mengarah pada instansi sentris, merasa instansinya lah yang paling betul dalam menerpakan ketentuan perundang- undangan. Jaksa dan hakim di lapangan tunduk pada aturan teknis dari pimpinan di tingkat pusat 4. Sintetis Satu ketentuan yang diterapkan berulang- ulang akan menjadi kebiasaan sehingga walaupun terdapat kesalahan dalam penerapan prosedur dianggap benar karena dianggap biasa 5. Evaluatif Praktek penerapan hukum akan dievaluasi menurut versi sendiri, sehingga mengarahkan pada pembenaran pendapat atas apa yang dilakukan walaupun salah 6. Analitis Kedua instansi tidak melakukan analisis secara komperhensif, sehingga kesalahan dalam memberi persepsi terus terjadi Tabel 1. Faktor Yang Mempengaruhi Persepsi j Penuntut umum harus menuntut 2/3 dari ancaman pidana pasal yang didakwakan dan ketentuan bahwa Majelis Hakim harus memidana minimal 2/3 dari tuntutan pidana penuntut umum merupakan kebijakan pimpinan Kejaksaan yang harus dilaksanakan oleh penuntut umum. 3. HASIL DAN PEMBAHASAN Berdasarkan kronologi perkara penerapan asas paraduga tidak bersalah terhadap perkara ini banyak timbul perbedaan persepsi di kalangan penegak hukum terutama di Pihak Jaksa dan Hakim, dimana ”kebijakan” kejaksaan yang harus menuntut pidana atau mengajukan upaya hukum apabila putusan pengadilan menyatakan pidana penjara tidak 2/3 dari tuntutan penuntut umum dengan pertimbangan terdakwa sangat Atas perbuatan tersangka Jaksa penuntut umum, menuntut tersangka dengan dakwaan primer pasal 338 KUHP dan menjatuhkan pidana penjara selama 12 tahun terhadap tersangka Dan Majelis Hakim Pengadilan Negeri Bone memutuskan dengan putusan selama 11 tahun penjara Berdasarkan uraian diatas bahwa asas praduga tak bersalah merupakan salah satu asas penting dalam proses 187 Efektivitas Penerapan Asas Praduga Tak Bersalah pada Proses Peradilan Pidana di (Baharuddin Badaru, Ibnu Hibban Sabil) Efektivitas Penerapan Asas Praduga Tak Bersalah pada Proses Peradilan Pidana di (Baharuddin Badaru, Ibnu Hibban Sabil) Efektivitas Penerapan Asas Praduga Tak Bersalah pada Proses Peradilan Pidana di (Baharuddin Badaru, Ibnu Hibban Sabil) menyesali perbuatannya dan kondisi sering sakit-sakitan setelah kejadian, karena dengan adanya kebijakan demikian, maka proses pemeriksaan di muka sidang pengadilan yang dilaksanakan oleh penuntut umum dengan ”kaca mata kuda”, artinya penuntut umum kurang memperdulikan terhadap fakta- fakta yang ada di persidangan dan lebih hanya berdasarkan Berita Acara Pemeriksaan (BAP) penyidik yang sudah ada. Pada gilirannya setelah pemeriksaan di persidangan selesai, penuntut umum tetap menuntut terdakwa bersalah atau dengan tuntutan pidana penjara yang tinggi, walaupun dalam persidangan terungkap fakta- fakta bahwa terdakwa pelakunya, tetapi terdapat hal-hal yang meringankan sehingga putusan pengadilan membebaskan terdakwa atau memidana dengan pidana penjara yang ringan sesuai dengan kesalahan dari terdakwa (Wawancara dengan Muswandar, S.H.,M.H, Hakim Pengadilan Negeri Watampone tanggal 25 September 2023). dihukum atas keyakinan berdasarkan berita acara pemeriksaan 4) Analysis, Pengalaman, penerapan tugas, dan pengetahuan yang berbeda dianalisis secara instansi sentris dengan kacamata masing-masing, sehingga persepsi praduga tak bersalah menjadi berbeda 4) Analysis, Pengalaman, penerapan tugas, dan pengetahuan yang berbeda dianalisis secara instansi sentris dengan kacamata masing-masing, sehingga persepsi praduga tak bersalah menjadi berbeda 5) Syntesis, Perbedaan persepsi secara berulang-ulang menyebabkan secara otomatis terus terjadi kesalahan dalam mempersepsikan asas praduga tak bersalah. Hal ini menyebabkan penerapan hukum dilakukan semaunya dan seenaknya 6) Evaluatif, Atas pengalaman yang sering terjadi para pihak menyimpulkan setelah melakukan evaluasi yang mengarah pada pembenaran persepsi masing- masing atas tugas yang dilakukan Berdasarkan kelima faktor di atas, maka masing-masing instansi melakukan evaluasi atas tugas masing-masing yang justru makin menyebabkan perbedaan persepsi makin melebar. 3. HASIL DAN PEMBAHASAN Ringkasnya, persepsi seseorang dipengaruhi oleh pengetahuannya (knowlwdge) dan pemahaman yang luas (comprehensif) serta praktek-praktek nyata di lapangan (aplicatif). Kemudian dengan analisis, sintesis dan evaluasi, seseorang mempersiapkan sesuatu menurut persepsi orang tersebut (Wawancara dengan Sulfiandi, S.H, Advokat, tanggal 25 September 2023). Penuntut umum membuat surat dakwaan dan melimpahkan perkara ke pengadilan, penuntut umum mendapat ”perintah” dari instansinya (kejaksaan) harus memidana terdakwa dengan tuntutan pidana yang juga diperintahkan oleh atasannya (melalui lembaga RENTUT). Tetapi apabila ternyata bahwa fakta-fakta yang terungkap dalam persidangan meringankan perbuatan terdakwa, maka hakim sebagai ”penegak hukum dan keadilan” tidak dapat memenuhi permintaan penuntut umum, sehingga terjadi perbedaan persepsi antara penuntut umum dengan hakim yang keduanya sebenarnya hanya melaksanakan tugas dan fungsinya saja. Untuk menggambarkan perbedaan persepsi antara jaksa dan hakim atas praduga tak bersalah jika dilihat dari teori persepsi dapat dilihat dalam bentuk Tabel 1. Berikut. y y y g g y saja. Penuntut umum harus menuntut 2/3 dari ancaman pidana pasal yang didakwakan dan ketentuan bahwa Majelis Hakim harus memidana minimal 2/3 dari tuntutan pidana penuntut umum merupakan kebijakan pimpinan Kejaksaan yang harus dilaksanakan oleh penuntut umum. Apabila penuntut umum gagal melaksanakan kebijakan tersebut, maka seorang penuntut umum akan dieksaminasi (diperiksa oleh Jaksa Pengawas) dan apabila hal tersebut karena kesalahan penuntut umum, maka yang bersangkutan dikenakan sanksi dari teguran sampai dengan pemecatan dengan menggunakan sanksi disiplin pegawai negeri sipil (Wawancara dengan Sulfiandi, S.H, Advokat, tanggal 25 September 2023). Untuk mengkaji adanya perbedaan persepsi antara hakim dengan penuntut umum tentang asas praduga tak bersalah dalam proses peradilan pidana yang menyatakan faktor-faktor yang mempengaruhi persepsi seseorang adalah: 1) Knowledge, Jaksa dan hakim walaupun sama-sama sarjana hukum, dalam perkembangannya mendapatkan pengetahuan yang berbeda karena mendapatkan pelatihan yang berbeda. Asas praduga tak bersalah dipersepsikan secara berbeda oleh kedua penegak hukum berdasarkan pengetahuan yang dimiliki berdasarkan pembekalan pengetahuan yang dimiliki dari instansi masing-masing. 2) Comprehensif, Pemahaman secara komprehensif juga mempengaruhi perbedaan persepsi. Padahal, semestinya pemahaman atas asas praduga tak bersalah atas tersangka/terdakwa harus dipahami secara komprehensif oleh petugas sebagai keseluruhan sistem. 3) Aplikatif, Penerapan praktek berdasarkan tugas yang berbeda menyebabkan pengalaman yang berbeda dalam memahami tugas masing-masing. Dalam persepsi hakim, penjatuhan hukuman didasarkan pada fakta-fakta di persidangan, sedangkan menurut persepsi jaksa penuntut umum terdakwa mesti p p p Tabel 1. Faktor Yang Mempengaruhi Persepsi No Faktor Yang Mempengaruhi Persepsi Analisis 1. 3. HASIL DAN PEMBAHASAN Tidak semua perkara harus divonis dengan hukuman (pidana) karena lebih baik melepaskan seribu orang bersalah daripada menghukum satu orang yang tidak bersalah Hakim lebih sering bernegosiasi dengan terdakwa sehingga orang yang seharsnya dihukum bisa dibebaskan, atau hukuman yang dijatuhkan jauh lebih ringan daripada tuntutan. 3. Masyarakat lah yang sering tidak mematuhi hukum dengan main hakim sendiri Masyarakat tidak percaya dengan putusan hakim, karenanya main hakim sendiri dirasa lebih adil. No Persepsi Hakim Persepsi Masyarakat 1. Masyarakat tidak memahami sepenuhnya kasus yang terjadi, masyarakat terlalu percaya pada pemberitaan pers Hakim tidak objektif dalam menangani perkara, sehingga terjadi disparitas pidana yang terlalu jauh antara perkara yang satu dengan yang lain 2. Masyarakat tidak sepenuhnya memahami kesulitan hakim dalam mengungkap perkara di persidangan. Tidak semua perkara harus divonis dengan hukuman (pidana) karena lebih baik melepaskan seribu orang bersalah daripada menghukum satu orang yang tidak bersalah Kasus yang umum adalah dalam hal pra penuntutan. Polisi seringkali merasa sudah benar dalam melakukan penyidikan, sedangkan jaksa seringkali merasa polisi belum maksimal sehingga perlu melakukan pra penuntutan (P19). Dalam hal penerapan asas praduga tak bersalah, yang muncul adalah persepsi jaksa yang selalu hamper selalu menganggap hakim salah dalam memutus suatu perkara sehingga dirasa perlu untuk melakukan upaya hukum berupa banding dan kasasi bahkan peninjauan kembali. Masyarakat tidak percaya dengan putusan hakim, karenanya main hakim sendiri dirasa lebih adil. Masyarakat tidak percaya dengan putusan hakim, karenanya main hakim sendiri dirasa lebih adil. 3. Masyarakat lah yang sering tidak mematuhi hukum dengan main hakim sendiri Menurut Minoru Shikita, perbedaan persepsi antar penegak hukum yang berakibat pada perbedaan peranan yang diterapkan dalam penegakan hukum akan menimbulkan setidaknya 3 (tiga) kerugian yaitu (Minoru Shikita dalam Indriyanto Seno Adji, 1998:21): Selanjutnya sehubungan dengan perkara nomor 114/Pid.B/2023/PN Wtp pedoman internal hakim dalam membuat putusan Pengadilan harus memperhatikan tujuan bahwa pidana dijatuhkan untuk (Wawancara dengan Muswandar, S.H.,M.H, Hakim Pengadilan Negeri Watampone tanggal 25 September 2023): 1) Kesukaran dalam menilai diri sendiri keberhasilan atau kegagalan masing- masing instansi, sehubungan dengan tugas mereka bersama g 2) Kesulitan dalam memecahkan sendiri masalah (masalah) pokok masing- masing instansi (sebagai sub system dari system peradilan pidana) ; gg 1) Mencegah dilakukannya tindak pidana dengan menegakan norma hukum demi masyarakatnya 2) Memasyarakatkan terpidana dengan mengadakan pembinaan sehingga menjadikannya orang yang baik dan berguna. 3. HASIL DAN PEMBAHASAN Apabila penuntut umum gagal melaksanakan kebijakan tersebut, maka seorang penuntut umum akan dieksaminasi (diperiksa oleh Jaksa Pengawas) dan apabila hal tersebut karena kesalahan penuntut umum, maka yang bersangkutan dikenakan sanksi dari teguran sampai dengan pemecatan dengan menggunakan sanksi disiplin pegawai negeri sipil (Wawancara dengan Sulfiandi, S.H, Advokat, tanggal 25 September 2023). Faktor Yang Mempengaruhi Persepsi Untuk mengkaji adanya perbedaan persepsi antara hakim dengan penuntut umum tentang asas praduga tak bersalah dalam proses peradilan pidana yang menyatakan faktor-faktor yang mempengaruhi persepsi seseorang adalah: 1) Knowledge, Jaksa dan hakim walaupun sama-sama sarjana hukum, dalam perkembangannya mendapatkan pengetahuan yang berbeda karena mendapatkan pelatihan yang berbeda. Asas praduga tak bersalah dipersepsikan secara berbeda oleh kedua penegak hukum berdasarkan pengetahuan yang dimiliki berdasarkan pembekalan pengetahuan yang dimiliki dari instansi masing-masing. 2) Comprehensif, Pemahaman secara komprehensif juga mempengaruhi perbedaan persepsi. Padahal, semestinya pemahaman atas asas praduga tak bersalah atas tersangka/terdakwa harus dipahami secara komprehensif oleh petugas sebagai keseluruhan sistem. 3) Aplikatif, Penerapan praktek berdasarkan tugas yang berbeda menyebabkan pengalaman yang berbeda dalam memahami tugas masing-masing. Dalam persepsi hakim, penjatuhan hukuman didasarkan pada fakta-fakta di persidangan, sedangkan menurut persepsi jaksa penuntut umum, terdakwa mesti 188 Idn. J. of Legality of law 6(1): 185-191,Desember 2023 masyarakat dengan hakim dapat digambarkan dalam tabel berikut ini: Perbedaan persepsi tersebut menyebabkan para penegak hukum memainkan peranannya secara berbeda. Undang- undang dalam hal ini KUHAP mengharuskan semua pihak khususnya jaksa dan hakim mepersepsikan hal yang sama atas prinsip praduga tak bersalah. Nyatanya, kedua instrumen penegak hukum ini memainkan peranannya sendiri-sendiri berdasarkan persepsi yang dibangun sendiri sebagai akibat proses di atas. Dari empat peranan yang digambarkan dalam teori peranan, maka hakim dan jaksa lebih sering menggunakan peranan yang dianggap oleh diri sendiri (perceived role) dan peranan yang sebenarnya dilakukan (actual role). Tabel 2. Perbedaan Persepsi Antara Hakim dan Masyarakat Atas Praduga Tak Bersalah Perbedaan Persepsi Antara Hakim dan Masyarakat Atas Praduga Tak Bersalah No Persepsi Hakim Persepsi Masyarakat 1. Masyarakat tidak memahami sepenuhnya kasus yang terjadi, masyarakat terlalu percaya pada pemberitaan pers Hakim tidak objektif dalam menangani perkara, sehingga terjadi disparitas pidana yang terlalu jauh antara perkara yang satu dengan yang lain 2. Masyarakat tidak sepenuhnya memahami kesulitan hakim dalam mengungkap perkara di persidangan. 3. HASIL DAN PEMBAHASAN Berarti dalam keadaan seperti ini hak penuh dari hakim sangat berperan besar dalam memberikan suatu putusan mengenai berat ringannya pidana. akan membantu memberikan arti terhadap objek psikologi tersbut. Apa yang ingin dilakukan atau dikehendaki oleh hakim tidak dapat ditentukan secara penuh, melainkan sangat tergantung pula pada nilai-nilai dan susunan masyarakatnya. Secara konkrit Julius Stone menyebut ada suatu ’expected reaction” yang datangnya dari masyarakat. seorang hakim tidak dapat bertindak mengikuti kebutuhan sendiri, melainkan tindakan-tindakannya itu didisiplinkan oleh sistem dari ”expected reaction” tersebut. Tugas dan pekerjaan sehari-harinya harus dijalankan oleh hakim, didukung oleh harapan atau tuntutan-tuntutan anggota masyarakat yang diorganisasi disekitar tugas-tugas hakim tersebut (Muladi dan Barda Nawawi Arief, 1986). Pada hakikatnya, kebebasan hakim adalah gebinden vrijheid, kebebasan terikat / terbatas karena diberi batas oleh undang-undang yang berlaku dalam batas tertentu. Hakim memiliki kebebasan dalam menetapkan, menentukan jenis pidana (stfaafsorrt), ukuran pidana atau berat ringan pidana (strafmaat), cara pelaksanaan pidana ( straf modus) dan kebebasan untuk menemukan hukum (rechtsvinding). Penjatuhan sanksi pidana oleh hakim mempunyai konsekwensi yang luas, baik yang menyangkut langsung pelaku tindak pidana maupun masyarakat secara luas karena keputusan pidana yang dianggap tidak tepat akan menimbulkan reaksi kontroversial, dalam hal ini disebut disparitas pidana, yaitu penerapan pidana yang tidak sama terhadap tidak pidana yang sama (same offence) atau terhadap tidank pidana yang sifatnya berbahaya dapat diperbandingkan. Dampak disparitas pidana adalah terpidana yang telah memperbandingkan pidana kemudian merasa menjadi korban ’the judicial caprice” akan menjadi terpidana yang tidak menghargai hukum, padahal pengahargaan terhadap hukum tersebut merupakan salah satu target di dalam tujuan pemidanaan. Kebebasan hakim mutlak dibutuhkan terutama untuk menjamin keobjektifan hakim dalam mengambil keputusan. Hakim memberikan keputusannya mengenai hal-hal sebagai berikut (Wawancara dengan Muswandar, S.H.,M.H, Hakim P dil N i W t t l 25 S t b 2023) Pengadilan Negeri Watampone tanggal 25 September 2023): Pengadilan Negeri Watampone tanggal 25 September 2023): 1) Keputusan mengenai peristiwanya, ialah apakah terdakwa melakukan perbuatan yang telah dituduhkan kepadanya, dan kemudian. 2) Keputusan mengenai hukumnya ialah apakah yang dilakukan terdakwa itu merupakan suatu tindakan pidana dan apakah terdakwa bersalah dan dapat dipidana, dan akhirnya. 3) Keputusan mengenai pidananya, apabila terdakwa memang dapat dipidana 3) Keputusan mengenai pidananya, apabila terdakwa memang dapat dipidana Pendekatan untuk memperkecil disparitas pidana dan pengaruhnya diperlukan suatu pedoman pemidanaan (statutory guidelines for sentencing) untuk memudahkan hakim dalam menetapkan pemidanaannya, setelah terbukti bahwa terdakwa telah melakukan perbuatan yang dituduhkan kepadanya. 3. HASIL DAN PEMBAHASAN 3) Karena tanggung jawab masing- masing instansi sering kurang jelas terbagi, maka setiap instansi tidak terlalu memperhatikan efektivitas menyeluruh dari system peradilan pidana 3) Menyelesaikan konflik yang ditimbulkan oleh tindak pidana, memulihkan keseimbangan dan mendatangkan rasa damai dalam masyarakat. 4) Membebaskan rasa bersalah pada terpidana 4) Membebaskan rasa bersalah pada terpidana Oleh karena itu, dapat dikatakan pada dasarnya putusan pengadilan merupakan suatu upaya harmonisasi konflik- konflik yang terjadi dalam masyarakat, dimana dengan adanya konflik menimbulkan disharmonisasi yang kemudian persoalannya diserahkan kepada Pengadilan dan pengadilan harus mengharmoniskan kembali masyarakat. Namun demikian, hampir saja selalu terjadi perbedaan persepsi tentang kesalahan seseorang dalam adanya statu putusan pengadilan Kekecewaan dan ketidakpuasan masyarakat juga terhadap putusan-putusan yang berbeda-beda dalamperkara- perkara yang sama (disparitas pidana). Adanya disparitas pidana putusan pengadilan dapat saja terjadi, karena keyakinan hakim bahwa putusan yang diberikan dirasakan sudah sesuai dengan rasa keadilan atas pertimbangan hukum dan non hukum. Pada sisi lain penetapan ancaman pidana ini sesuai dengan sistem yang ada dalam KUHP, khususnya pada pada Pasal 12 ayat (1) dan (2) menyatakan bahwa pidana penjara selama waktu tertentu paling pendek satu hari dan paling lama lima belas tahun berturut-turut. p Perbedaan dan polemik atau konflik peranan antar sub sistem peradilan pidana jelaslah tidak sesuai dengan prinsip yang ingin dibangun dengan adanya Integrated Criminal Justice System (Sistem Peradilan Pidana Terpadu) yang dapat diartikan sebagai keterpaduan yang sama dalam memahami satu tugas bersama yaitu mengungkapkan kebenaran materil dari suatu tindak pidana. Karena itulah, hubungan koordinasi yang ada daam sistem KUHAP adalah koordinasi horizontal, bukan vertikal. Semua instansi dalam sistem peradilan pidana memiliki tugas yang sama, tetapi dalam lingkungan yang teknis dan lebih kecil menjalankan tugas dan peranan sesuai tujuan kecil yang tidak dapat dipisahkan dari tujuan bersama berdasarkan kewenangan masing-masing pihak yang ditentukan undang-undang. Persepsi masyarakat bahwa pengadilan yang berhak menyatakan terdakwa bersalah atau tidak sebagai wujud dari asas praduga tak bersalah mempunyai perbedaan berdasarkan pengetahuan, pemahaman dan sikapnya. Perbedaan persepsi atas asas praduga tak bersalah antara Akibatnya dari adanya ketentuan pasal 12 ayat 1 dan 2 189 Efektivitas Penerapan Asas Praduga Tak Bersalah pada Proses Peradilan Pidana di (Baharuddin Badaru, Ibnu Hibban Sabil) KUHP ini, maka tidak menutup kemungkinan dapat terjadi disparitas pidana, yaitu penerapan pidana yang tidak sama terhadap tindak pidana yang sama (same offence) atau terhadap tindak pidana yang sifatnya bahaya dapat dibandingkan (offences of comparable seriousness) tanpa dasar pembenaran yang jelas. 3. HASIL DAN PEMBAHASAN Munculnya perbedaan persepsi di kalangan masyarakat dengan pengadilan karena adanya perkembangan ilmu pengetahuan dan tekhnologi serta komunikasi dan informasi yang mengakibatkan terjadinya terpaan informasi yang begitu deras di kalangan masyarakat, padahal perkembangan dan perubahan hukum berjalan sangat lambat. Para Hakim juga mempertimbangkan hal-hal yang memperingan pidana antara lain: Para Hakim juga mempertimbangkan hal-hal yang memperingan pidana antara lain: 1) Perbuatan yang merupakan percobaan atau pembantuan tindak pidana. 2) Pelaku tindak pidana dengan sukarela menyerahkan diri kepad yang berwajib; p y g j 3) Pelaku tindak pidana seorang wanita yang sedang hamil muda; 4) Pelaku tindak pidana dengan sukarela memberi ganti rugi yang layak atau memperbaiki kerusakan akibat perbuatannya; Masyarakat yang menghendaki agar peradilan membuat putusan yang seadil-adilnya dalam rangka menegakan supremasi hukum, tidak dapat berharap banyak sepanjang peraturan perundangundangan, aparat penegak hukum dan budaya hukum yang ada di pengadilan masih bersifat konvensional 5) Pelaku melakukan tindak pidana karena kegoncangan jiwa yang sangat hebat sebagai akibat yang sangat berat dari keadaan pribadi atau keluarganya; g y 6) Pelaku tindak pidana tidak mampu bertanggungjawab 5. DAFTAR PUSTAKA Achmad Ali. 2017. Menguak Teori Hukum (Legal Theory) dan Teori Peradilan (Judicialprudence) Termasuk Interpretasi Undang- Undang (Legisprudence). Volume 1 Pemahaman Awal. Kencana: Jakarta. Muladi dan Barda Nawawi Arief. ”Ruang Lingkup Penegakan Hukum Pidana dalam Konteks Politik Kriminal”. Makalah Seminar Kriminologi V. Fakultas Hukum Undip Semarang, 1986 Adami Chazawi. 2002. Pelajaran Hukum Pidana Bagian I, Raja Grafindo Persada. Ahmad Zaenal Fanani. 2011. Berpikir Falsafati Dalam Putusan Hakim, Artikel ini pernah dimuatdi Varia Peradilan No. 304. Nur Agus Susanto. 2014. Dimensi Aksiologis Dari Putusan Kasus “ST” Kajian Putusan Peninjauan Kembali Nomor 97 PK/Pid.Sus/2012, Jurnal Yudisial Vol. 7 No. 3. Andi Hamzah. 2011. Hukum Acara Pidana Indonesia, Sinar Grafika, Jakarta Romli Atmasasmita. 1996. Perbandingan Hukum Pidana, CV. Mandar Maju. Azhari M. Tahir. 2005. Negara Hukum: Suatu Studi tentang Prinsip-prinsip Dilihat Dari Segi Hukum Islam, Implementasinya Pada Periode Negara Madinah dan Masa Kini, Bulan Bintang, Jakarta. Satjipto Rahardjo. 2012. Ilmu Hukum, Editor Awaludin Marwan, PT Citra Aditya Bakti, Bandung. Sidharta Arief. 2007. Meuwissen Tentang Pengembanan Hukum, Ilmu Hukum, Teori Hukum dan Filsafat Hukum, PT Refika Aditama, Bandung. Bernard L Tanya dkk. 2013. Teori Hukum: Strategi Tertib Manusia Lintas Ruang dan Generasi, Genta Publishing, Yogyakarta. g Soerjono Soekanto. 2005. Faktor-Faktor yang Mempengaruhi Penegakan Hukum, PT RajaGrafindo Persada, Jakarta Bodenheimer. Dalam Satjipto Rahardjo. 2006. Ilmu Hukum, Bandung: PT. Citra Aditya Bakti. Soerjono Soekanto. 2007. Pokok-Pokok Sosiologi Hukum, PT Raja Grafindo Persada, Jakarta. Dominikus Rato. 2010. Filsafat Hukum Mencari: Memahami dan Memahami Hukum, LaksbangPressindo, Yogyakarta. Tongat. 2011. Dasar-Dasar Hukum Pidana Indonesia dalam Perspektif Pembaharuan, UMMPress, Malang. Utrecht, 1962. Pengantar Hukum Administrasi Negara Indonesia, Ichtiar, Jakarta. H.R Otje Salman. 2010. Filsafat Hukum (Perkembangan & Dinamika Masalah), PT. Refika Aditama, Bandung. Jimly Asshiddiqie. 2009. e-book Konstitusi & Konstitusionalisme Indonesia. John Rawls. 2006. A Theory of Justice, London: Oxford University press, yang sudah diterjemahkan dalam bahasa indonesia oleh Uzair Fauzan dan Heru Prasetyo, Teori Keadilan, Yogyakarta: Pustaka Pelajar. Kartanegara Satochid.1989. Hukum Pidana Kumpulan Kuliah dan Pendapat Pendapat Para Ahli Hukum Terkemuka, Balai Lektur Mahasiswa. Keputusan Menteri Kehakiman Republik Indonesia Nomor M.01.PW.07.03 Tahun 1982 tentang Pedoman Pelaksana Lili Rasjidi dan I.B Wyasa Putra. 1993. Hukum sebagai Suatu Sistem, Bandung: Remaja Rosdakarya. M. Yahya Harahap. 2008. Pembahasan Permasalahan Dan Penerapan KUHAP Penyidikan dan Penuntutan, Sinar Grafika, Jakarta. Mardjono Reksodiputro. 1993. Mardjono Reksodiputro, Sistem Peradilan Pidana Indonesia (Melihat Kepada Kejahatan dan Penegakkan Hukum Dalam Batas-batas Toleransi), Pidato Pengukuhan Penerimaan Jabatan Guru Besar Tetap Dalam Ilmu Hukum Pada Fakultas Hukum Universitas Indonesia, Jakarta. 4. KESIMPULAN DAN SARAN Keputusan yang proporsional tersebut dapat dicapai dengan memperlihatkan tujuan pemidanaan yang hendak dicapai, pedoman pemidaan (statutory guidelines for sentencing) aturan pemidanaan yang berlaku serta keyakinan hakim tersebut dalam menjatuhkan sanksi, sehingga terlihat faktor-faktor yang menjadi dasar pertimbangan hakim dalam penjatuhan sanksi pidana dalam rangka mengurangi disparitas pidana. Hasil penelitian dan pembahasan dapat disimpulkan bahwa penerapan asas praduga tidak bersalah menurut Undang-undang Hukum Acara Pidana (KUHAP) yang mensyaratkan harus adanya dugaan kuat bahwa orang tersebut bersalah melakukan suatu tindak pidana dan Hak tersangka atau terdakwa untuk mendapatkan bantuan hukum merupakan salah satu hak dalam mewujudkan asas praduga tak bersalah. Dalam pertimbangannya hakim memutuskan perkara tindak pidana pembunuhan nomor 114/Pid.B/2023/PN Wtp berdasarkan asas praduga tak bersalah dengan memanfaatkan Knowladge, Comprehensif, Aplikasif, Analisis, Syntesis, Evaluatif yang bertujuan memulihkan keseimbangan dan mendatangkan rasa damai dalam Masyarakat berdasarkan hasil penelitian telah efektif, namun dalam pelaksnaan proses peradilan terjadi perbedaan pendapat antara penegak hukum Kepolisian, Kejaksaan, dan Pengadilan dalam penerapan asas praduga tidak bersalah. Adanya penjatuhan pidana oleh hakim secara langsung mengkongkritkan tugas hakim tersebut, yaitu sebagai alat pemaksa agar norma dapat juga berfungsi sebagai alat prefentif dan sekaligus sebagai alat refresif sehingga tujuan dari hukum pidan tersebut dapat dicapai secara efektif. Adapun faktor-faktor yang mempengaruhi kemampuan mempersepsi suatu objek adalah: pengalaman hidup, proses belajar (sosialisasi), cakrawala dan pengetahuan. Faktor pengalaman hidup dan proses belajar yang dimiliki seseorang akan membantu memberi bentuk dan struktur terhadap objek persepsi. Adapun faktor cakrawala dan kemampuan seseorang 190 Idn. J. of Legality of law 6(1): 185-191,Desember 2023 Minoru Shikita dalam Indriyanto Seno Adji. 1998. Penyiksaan dan HAM dalam Perspektif KUHAP, Jakarta. Sinar HarapanMoh. Kusnardi dan Harmaily Ibrahim. 1983. Pengantar Hukum Tata Negara Indonesia, Pusat Studi Hukum Tata Negara Fakultas Hukum Universitas Indonesia dan Sinar Bakti. Perlunya ditingkatkan sinergitas antara pihak penegak hukum mulai dari kepolisian, kejaksaan sampai kepada pengadilan dalam proses peradilan pidana tindak pidana pembunuhan berdasarkan asas praduga tidak bersalah dalam rangka mencegah tindak pidana baru seperti tindak pidana main hakim sendiri oleh Masyarakat. Moh. Mahfud MD. 2009. Penegakan Hukum DanTata Kelola Pemerintahan Yang Baik, Bahan pada Acara Seminar Nasional “Saatnya Hati Nurani Bicara” yang diselenggarakan oleh DPP Partai HANURA. Mahkamah Konstitusi Jakarta. 5. DAFTAR PUSTAKA Masrudi Muchtar, Sistem Peradilan Pidana Di Bidang Perlindungan dan Pengelolaan Lingkungan Hidup, Jakarta, Prestasi Pustakaraya. 191
https://openalex.org/W2940451327	https://repository.publisso.de/resource/frl:6415583/data	English	null	The Interplay Between Stress, Inflammation, and Emotional Attention: Relevance for Depression	Frontiers in neuroscience	2,019	cc-by	7,415	MINI REVIEW published: 24 April 2019 doi: 10.3389/fnins.2019.00384 MINI REVIEW published: 24 April 2019 doi: 10.3389/fnins.2019.00384 The Interplay Between Stress, Inﬂammation, and Emotional Attention: Relevance for Depression Viktoriya Maydych* Department Psychology and Neurosciences, Leibniz Research Centre for Working Environment and Human Factors, TU Dortmund (IfADo), Dortmund, Germany Depression is among the most signiﬁcant public mental health issues. A growing body of research implicates inﬂammation in the etiology and pathophysiology of depression. Yet, the results are somewhat inconsistent, leading to burgeoning attempts to identify associations between components of innate immune system involved in inﬂammation and speciﬁc symptoms of depression, including attention to emotional information. Negative attentional bias, deﬁned as a tendency to direct attention toward negatively valenced information, is one of the core cognitive features of depression and is reliably demonstrated in depressed and vulnerable individuals. Altered attentional processing of emotional information and immunological changes are often precipitated by stressful events. Psychological stress triggers inﬂammatory activity and affective-cognitive changes that play a critical role in the onset, maintenance, and recurrence of depression. Using various designs, recent studies have reported a positive relationship between markers of inﬂammation and negative attentional bias on behavioral and neural levels, suggesting that the association between inﬂammation and emotional attention might represent a neurobiological pathway linking stress and depression. This mini-review summarizes current research on the reciprocal relationships between different types of stressors, emotional attention, inﬂammation, and depression, and discusses potential neurobiological mechanisms underlying these interactions. The integration provided aims to contribute toward understanding how biological and psychological processes interact to inﬂuence depression outcomes. Edited by: Deborah Suchecki, Federal University of São Paulo, Brazil Edited by: Deborah Suchecki, Federal University of São Paulo, Brazil Deborah Suchecki, Federal University of São Paulo, Brazil Reviewed by: Charles Barnet Nemeroff, The University of Texas at Austin, United States Marisa Toups, University of Texas Dell Medical School, United States, in collaboration with reviewer CBN Bruno Bonaz, Centre Hospitalier Universitaire de Grenoble, France *Correspondence: Viktoriya Maydych maydych@ifado.de Specialty section: This article was submitted to Neuroendocrine Science, a section of the journal Frontiers in Neuroscience Received: 10 January 2019 Accepted: 02 April 2019 Published: 24 April 2019 Specialty section: This article was submitted to Neuroendocrine Science, a section of the journal Frontiers in Neuroscience Received: 10 January 2019 Accepted: 02 April 2019 Published: 24 April 2019 Keywords: inﬂammation, cytokines, psychological stress, emotional attention, attentional bias, negative bias, depression, depressive disorders INFLAMMATION, STRESS, AND DEPRESSION A considerable body of evidence suggests that individuals with diagnosed depression exhibit signiﬁcantly higher levels of IL-1, IL-6, TNF-α, and C-reactive protein (CRP) compared to non-depressed counterparts (Howren et al., 2009; Dowlati et al., 2010). Concurrently, depressive symptoms are more frequent in patients with conditions involving inﬂammation (e.g., autoimmune diseases) and can be reversed through the use of anti-inﬂammatory drugs (Kojima et al., 2009; Köhler et al., 2014). Notably, increased inﬂammatory activity has been documented only in some patients with depression. This indicates that the depression-inﬂammation link may be modulated by further vulnerability factors, such as genes or cognitive vulnerability. Alternatively, since depression is a heterogeneous disorder, it is also plausible that the association between cytokine-mediated inﬂammatory processes and depression is more nuanced in terms of the groups of depressive symptoms (somatic vs. aﬀective-cognitive). In support of this notion, a large data set has documented mechanistic links between somatic symptoms of depression and increased inﬂammation in animals and humans (DellaGioia and Hannestad, 2010); however, studies are lacking in aﬀective-cognitive changes related to inﬂammatory activity. There is some evidence that inﬂammatory processes may have diﬀerential eﬀects on somatic and aﬀective- cognitive depressive symptoms that are based on distinct neurophysiological mechanisms. For example, studies examining the development of depressive symptoms during the course of IFN-α therapy found that all patients developed somatic symptoms, including fatigue, altered sleep and appetite, motor slowing, during the ﬁrst weeks of therapy (Capuron and Miller, 2004). In contrast, only 30–50% of patients developed aﬀective-cognitive symptoms such as negative mood, anhedonia, or cognitive impairment during the later stages of therapy. Strikingly, the development of this group of symptoms could be prevented by prophylactic antidepressant administration (Musselman et al., 2001). According to cognitive models of depression, one of the key features of depressed and vulnerable individuals is biased cognitive processing of emotional and social information. Cognitive biases manifest themselves in a consistent shift toward (self-referential) negative or threatening information in all aspects of cognition, including perception, attention, interpretation, memory, or sensitivity to feedback (Miskowiak and Carvalho, 2014). Negatively biased processing of emotional information is usually regarded both as a neuropsychiatric symptom and as a lingering trait factor that confers cognitive vulnerability to depression and may, when triggered by adverse environmental factors (e.g., stress), initiate the development or reoccurrence of depression (Ingram et al., 1998). The following mini-review focuses primarily on one cognitive domain: attention. Citation: Diﬀerent studies using a variety of study designs and populations have found positive associations between increased levels of proinﬂammatory cytokines and symptoms of depression (Dowlati et al., 2010; Valkanova et al., 2013). However, ﬁndings have not been entirely consistent for all types of depression (Rothermundt et al., 2001), raising the need for identifying more speciﬁc links between inﬂammation and diﬀerent somatic and aﬀective-cognitive symptoms, rather than merely testing associations between increased inﬂammation and categorically deﬁned depression. and (2) emotional attention as well as the potential link between stress-inﬂammation and stress-cognition pathways (3), and depression. Frontiers in Neuroscience \| www.frontiersin.org Citation: Maydych V (2019) The Interplay Between Stress, Inﬂammation, and Emotional Attention: Relevance for Depression. Depression is a highly prevalent mood disorder in modern society and is associated with signiﬁcant impairments in the patients’ quality of life. A multitude of basic research and clinical studies have been performed, with the aim of understanding the interaction between biological, psychological, and environmental factors involved in the etiology of depression. There is growing evidence Front. Neurosci. 13:384. doi: 10.3389/fnins.2019.00384 April 2019 \| Volume 13 \| Article 384 Frontiers in Neuroscience \| www.frontiersin.org Stress, Inﬂammation and Emotional Attention Maydych implicating increased levels of markers of inﬂammation in the pathogenesis of depressive disorders (Raison et al., 2006). Inﬂammation is a part of the innate immune system’s response to infection or injury. The main mediators of the inﬂammatory response, proinﬂammatory cytokines, such as interleukin (IL)- 1β, interleukin (IL)-1 receptor antagonist (RA), interleukin (IL)- 6, tumor necrosis factor (TNF)-α, and interferon (IFN)- γ, have been recently shown to communicate with the brain and aﬀect neurotransmission, neuroendocrine activity, and brain structure and functions, thereby inducing emotional, cognitive, and behavioral changes (Haroon et al., 2012). If the inﬂammatory response remains unresolved, the chronic release of proinﬂammatory cytokines can promote pathology, including depression. Diﬀerent studies using a variety of study designs and populations have found positive associations between increased levels of proinﬂammatory cytokines and symptoms of depression (Dowlati et al., 2010; Valkanova et al., 2013). However, ﬁndings have not been entirely consistent for all types of depression (Rothermundt et al., 2001), raising the need for identifying more speciﬁc links between inﬂammation and diﬀerent somatic and aﬀective-cognitive symptoms, rather than merely testing associations between increased inﬂammation and categorically deﬁned depression. implicating increased levels of markers of inﬂammation in the pathogenesis of depressive disorders (Raison et al., 2006). Inﬂammation is a part of the innate immune system’s response to infection or injury. The main mediators of the inﬂammatory response, proinﬂammatory cytokines, such as interleukin (IL)- 1β, interleukin (IL)-1 receptor antagonist (RA), interleukin (IL)- 6, tumor necrosis factor (TNF)-α, and interferon (IFN)- γ, have been recently shown to communicate with the brain and aﬀect neurotransmission, neuroendocrine activity, and brain structure and functions, thereby inducing emotional, cognitive, and behavioral changes (Haroon et al., 2012). If the inﬂammatory response remains unresolved, the chronic release of proinﬂammatory cytokines can promote pathology, including depression. EMOTIONAL ATTENTION, STRESS, AND DEPRESSION Cognitive symptoms of depression include attentional biases (AB) toward negative information (Mathews and MacLeod, 2005). A number of studies using diﬀerent attention allocation tasks (MacLeod et al., 1986) demonstrated that compared to non-depressed counterparts, depressed individuals exhibit increased diﬃculty in disengaging their attention from negative stimuli than from positive or neutral stimuli, especially when negative material is related to depression (e.g., feelings of worthlessness, guilt) (Gotlib et al., 2004a,b; Koster et al., 2005; Caseras et al., 2007). Negative AB has been also documented in patients with remitted depression and in individuals exposed to ELA (Luecken and Appelhans, 2005; Joormann and Gotlib, 2007; Raymond et al., 2018). , ) The causal role of stress in inﬂammatory activity was also examined in laboratory settings that enable the assessment of temporal patterns of cytokine responses and use of standardized stress induction procedures such as the Trier Social Stress Test (TSST) (Kirschbaum et al., 1993; Slavich and Irwin, 2014). Laboratory studies showed that acute stress was associated with signiﬁcant increases in IL-1β (Yamakawa et al., 2009), IL-1RA and IL-6 (Goebel et al., 2000; O’Donnell et al., 2008; Hackett et al., 2012), and TNF-α (O’Donnell et al., 2008), with IL-β, IL-6, and TNF-α demonstrating the most robust increases (Marsland et al., 2017). At the same time, higher cytokine levels were reported to be associated with increases in negative mood and anxiety in some studies (Yamakawa et al., 2009; Moons et al., 2010; Carroll et al., 2011). The notion that increases in inﬂammatory activity can lead to negative emotional states was also supported by studies that induced low-grade inﬂammation through the injection of bacterial endotoxin (i.e., lipopolysaccharide, LPS) or vaccines (i.e., ﬂu, typhoid). Stimulated increases in proinﬂammatory cytokines were associated with symptoms such as fatigue, negative mood, anhedonia, cognitive impairment, social withdrawal, motor slowing – a variety of symptoms collectively referred to as sickness behavior and resembling those of aﬀective-cognitive and somatic symptoms of depression (Dantzer et al., 2008; Eisenberger et al., 2010). Moreover, the associations between inﬂammatory activity and sickness behavior were not restricted to the laboratory, but also predicted depressive symptoms and cognitive impairment 1 week later (Kuhlman et al., 2018). Similarly, increases in IL-1β in response to TSST predicted the increase of depressive symptoms 1 year later (Aschbacher et al., 2012). INFLAMMATION, STRESS, AND DEPRESSION Heightened inﬂammation and negative attentional bias (AB) are often the results of psychological stress. Acute stressful challenges lead to increases in inﬂammatory activity and other neurophysiological changes that modulate aﬀective, cognitive, and behavioral processes (Allen et al., 2014; Slavich and Irwin, 2014). Chronic exposure to stressors causes endocrine and immune system dysfunction that contributes to sustained low-grade inﬂammation, which is involved in the pathogenesis of depressive symptoms (Rohleder, 2014). In parallel, acute stress has been shown to trigger aﬀective and cognitive changes similar to biased information processing characteristic to depression (Gotlib and Joormann, 2010). This evidence has recently led to promising attempts to investigate interactions between emotional attention and inﬂammation in the context of stress, thereby identifying speciﬁc neurocognitive pathways that may be relevant for the etiology of depression and development of novel treatments. The aim of this mini-review is to summarize independent lines of inquiry focusing on the eﬀects of stress (1) on inﬂammation There is considerable evidence that psychological stress can activate the inﬂammatory response. Diﬀerent types of stressors are capable of eliciting increases in inﬂammatory activity in a manner that may promote depressive symptoms (Slavich and Irwin, 2014). Moreover, the link between stressor-evoked increases in CRP and proinﬂammatory cytokines and depression appears to be bidirectional, as chronic stressors and current depressive symptoms, both associated with neurophysiological changes (e.g., glucocorticoid resistance), were found to increase stress reactivity, including cytokine changes in response to stressful challenges. One of the most robust predictors of increased levels of proinﬂammatory cytokines is early life adversity (ELA). Usually April 2019 \| Volume 13 \| Article 384 2 Stress, Inﬂammation and Emotional Attention Maydych indicated by parental maltreatment and low socioeconomic status during childhood, ELA is considered as a chronic and severe stressor causing long-lasting psychological and biological abnormalities that considerably increase the risk of depression (Hostinar et al., 2018). Psychological alterations are manifested in exaggerated reactivity to negative information and stress; biological abnormalities include HPA axis activity dysregulation (in most cases hyperactivity leading to glucocorticoid resistance), low parasympathetic activity, and frontolimbic circuit alterations that promote reactivity to threatening stimuli (Callaghan and Tottenham, 2016). For example, individuals exposed to ELA showed stronger increases in proinﬂammatory cytokines in response to laboratory stress than those who were not (Pace et al., 2006). INFLAMMATION, STRESS, AND DEPRESSION Moreover, exposure to ELA was prospectively and retrospectively associated with an increased inﬂammation in later life (Danese et al., 2008; Kiecolt-Glaser et al., 2011; Coelho et al., 2014). proinﬂammatory cytokines are prospectively associated with depressive symptoms. Frontiers in Neuroscience \| www.frontiersin.org EMOTIONAL ATTENTION, STRESS, AND DEPRESSION April 2019 \| Volume 13 \| Article 384 3 Stress, Inﬂammation and Emotional Attention Maydych STRESS, INFLAMMATION, AND EMOTIONAL ATTENTION stem from methodological and design issues, especially from diﬀerences in stress/mood induction procedures, the timing of cytokine assessments, and types of attention tasks. It is also plausible that endogenous concentrations of cytokines, in particular in healthy samples, even after a stressful challenge, may be too low to map on behavioral attention measures. Alternatively, attention tasks may not be sensitive enough to the cognitive changes produced by cytokines. As outlined in previous sections, distinct lines of research show that diﬀerent stressors can trigger inﬂammatory activity and increase the attentional processing of negative information. Both inﬂammatory processes and cognitive stressor-evoked changes were associated with mood lowering and an increase in anxiety and depressive symptoms. Stressor-evoked elevations in proinﬂammatory cytokines and attention shift toward negative information can represent stress responses at multiple levels that independently contribute to depressive symptoms. Alternatively, inﬂammatory and cognitive stress responses may act together, potentiating one another’s impact on promoting depression. The following section provides a summary of studies that examined the relationship between AB and markers of inﬂammation. Increased inﬂammatory activity appears not only to be associated with AB toward negative information but has also been suggested to increase stress reactivity (Dooley et al., 2018). As outlined earlier, depressed individuals and those exposed to ELA exhibit higher increases in proinﬂammatory cytokines in response to acute stress. Thus, it is possible that exogenously induced inﬂammation prior to stress manipulation would increase stress reactivity and drive even stronger changes in emotional processing than individual treatments. Increases in IL-6 levels have been demonstrated to be positively associated with negative AB only in response to typhoid vaccine in women with partially remitted depression, but not in response to laboratory stress or both treatments (Niemegeers et al., 2019). In another study, slower processing of negative information was observed in response to LPS treatment and at a trend level in combined LPS and negative mood induction condition in healthy males (Benson et al., 2017). EMOTIONAL ATTENTION, STRESS, AND DEPRESSION Individuals with diagnosed depression have demonstrated stronger increases in proinﬂammatory cytokines in response to laboratory stress than non-depressed individuals (Weinstein et al., 2010; Fagundes et al., 2013), indicating an increased inﬂammatory stress responsiveness in depression. Although the aforementioned studies provide interesting ﬁndings, the ecological validity of most results is limited due to laboratory settings and mainly samples of healthy young adults. Future research could seek to examine whether naturalistic stressor-induced increases in The common assumption of cognitive stress-diathesis models is that depression is a result of the interaction between cognitive vulnerability and stressful life events (Ingram et al., 1998). Therefore, given cognitive vulnerability, experiencing stressful events can initiate a depressive episode. Although the causal role of stress and cognitive biases that jointly increase the subsequent depression risk is theoretically now well-established, surprisingly few studies have examined this etiological pathway with assessments of stressful events and attention measures that do not rely on the participants’ self-report. A number of laboratory-based studies have examined whether laboratory stress would increase negative AB and whether attention shift would be associated with mood change. Indeed, AB toward negative vs. neutral material has been shown to be increased after a stressful challenge (Ellenbogen et al., 2002; Tsumura and Shimada, 2012). Moreover, attention shift toward negative information was associated with mood lowering (Ellenbogen et al., 2002, 2006) cortisol responses (Ellenbogen et al., 2010; Roelofs et al., 2007) in healthy participants and slower stress recovery in a depressed cohort (Sanchez et al., 2017). Although these results suggest a causal link between stressor-evoked AB and negative mood, the main limitation of this work is that stress eﬀects on AB and mood change reﬂect short-term prime eﬀects rather than providing ecologically valid evidence of the stress-diathesis hypothesis. To examine the long-term eﬀects, several studies examined whether baseline or stressor-related negative AB shifts would prospectively predict depressive symptoms. These studies reported that AB shift following induction of negative mood interacted with subsequent stressful events in predicting increases in dysphoria 7 weeks later in dysphoric students (Beevers and Carver, 2003). Similarly, negative AB was predictive of the exacerbation of depressive symptoms in adults with subclinical depression after 5 weeks (Disner et al., 2017). Finally, a signiﬁcant interaction eﬀect of stressful life events and dysfunctional attitudes on clinical depression incidence after 12 months has been reported (Lewinsohn et al., 2001). BEHAVIORAL STUDIES To analyze an association between negative AB and increased inﬂammatory activity, several cross-sectional and clinical studies examined performance in attention tasks using emotional material and tested for relationships with inﬂammatory markers. Levels of CRP were reported to positively correlate with increased AB toward sad vs. happy and angry faces in breast cancer survivors (Boyle et al., 2017). In addition, hepatitis C patients showed an AB away from positive vs. neutral and fearful faces and an increase in symptoms of depression, anxiety, and fatigue 6–7 weeks after commencing IFN-α therapy (Cooper et al., 2018). Concurrently, a greater increase in AB toward self-referential positive vs. negative words and improvement of aﬀective- cognitive and somatic depressive symptoms was observed after completion of anti-TNF-α therapy in patients with inﬂammatory bowel disease (Gray et al., 2018). This preliminary evidence suggests that aﬀective processing and depressive symptoms may, at least, be partially driven by inﬂammatory activity. The pattern of AB toward negative and away from positive information is consistent with AB usually observed in depression. However, the inﬂuence of disease or environmental confounding factors cannot be ruled out in these studies. In summary, the ﬁndings from various behavioral studies indicate that stressor-related or endotoxin-induced increases in inﬂammatory activity may aﬀect emotional attention similar to AB in depression. Yet, there were some inconsistent results and null results, which can reﬂect methodological diﬀerences between studies. In addition, the results obtained in laboratory studies do not allow for conclusions on long-term causal relationships between immune and cognitive processes. Future studies should determine whether increased inﬂammation can prospectively predict alterations in emotional attention. Frontiers in Neuroscience \| www.frontiersin.org FUNCTIONAL NEUROIMAGING STUDIES the synthesis of corticotrophin-releasing hormone (CRH), adrenocorticotropic hormone (ACTH), and cortisol (Raison and Miller, 2003). The extent to which cytokines induce the release of ACTH and cortisol is predictive of the development of aﬀective-cognitive but not of somatic symptoms of depression (Capuron et al., 2003). This implies that HPA axis sensitivity to inﬂammatory stimulation is particularly relevant for the development of aﬀective-cognitive symptoms of depression. Finally, the parasympathetic nervous system has been suggested to modulate aﬀective-cognitive and immune processes involved in stress cascade and depression (Thayer and Sternberg, 2006; Ondicova et al., 2010). Lower activity of the vagus nerve is predictive of higher levels of cortisol and cytokine acute stress response (Hamer and Steptoe, 2007; Smeets, 2010; Woody et al., 2017) as well as slower stress recovery (Weber et al., 2010). In addition, the vagal tone has been implicated in the detection and (down-) regulation of inﬂammatory processes. The anti-inﬂammatory eﬀects are mediated by the vagal release of acetylcholine, which activates a7 nicotinic Ach receptors in macrophages, thereby inhibiting the release of proinﬂammatory cytokines from these lymphocytes (the mechanism referred to as “cholinergic anti-inﬂammatory reﬂex”) (Rosas-Ballina and Tracey, 2009). It has been also suggested that reduced activity of the vagus nerve is associated with disturbed emotion regulation and further aﬀective-cognitive symptoms that stem from the impaired inhibitory control of the prefrontal cortex over the limbic system (Thayer and Sternberg, 2006; Thayer et al., 2012) as well as deﬁciency in monoamines (Dorr and Debonnel, 2006). p g p p y Although the data is not yet suﬃcient to draw generalized conclusions, the majority of studies have documented the increased activity of amygdala and dACC in response to increases in inﬂammatory activity and negative social stimuli/feedback. This is consistent with the literature on AB in depressive or at-risk individuals (e.g., those exposed to ELA) that found enhanced and long-lasting activity of amygdala in response to negative material (Disner et al., 2011). A simultaneous activation increase in of PFC and ACC was attributed to cortical insuﬃciency and abnormal frontolimbic circuit function (Wagner et al., 2006; Matsuo et al., 2007). Along with amygdala and AI, dACC was suggested to constitute a so-called “neural alarm system,” which is responsible for the detection of environmental threats and the regulation of responses to danger including SNS system and HPA axis response (Muscatell and Eisenberger, 2012). FUNCTIONAL NEUROIMAGING STUDIES Although the literature is rather sparse at present, the eﬀects of inﬂammation on neural activity and functional connectivity during the processing of emotional stimuli have also been the subject of investigation. The experimental designs of these studies induced increased inﬂammation through either LPS, vaccines, or laboratory stress and measured neural activity and connectivity during exposure to emotional stimuli or receiving social feedback. LPS-induced inﬂammation was shown to increase amygdala activity while viewing negative facial expression images (Inagaki et al., 2012). Furthermore, peripheral levels of IL-6 were associated with increased activation of the amygdala and increased functional connectivity between the amygdala and dorsomedial prefrontal cortex (dmPFC) in response to negative social feedback (Muscatell et al., 2015). Laboratory stressor-evoked increases in the soluble TNF-α receptor (sTNFαRII) have been shown to be positively correlated with increased activation in dorsal anterior cingulate cortex (dACC) and anterior insula (AI) in response to social rejection To examine whether a causal relationship exists between stressor-evoked inﬂammatory response and AB, a number of experimental studies investigated the eﬀects of acute stress or mood induction on cytokine levels and emotional attention in healthy and depressed individuals; however, the ﬁndings have been mixed. Signiﬁcant increases in both pro- and anti-inﬂammatory cytokines following laboratory stress were reported by some studies (Boyle, 2018; Maydych et al., 2018). Elevations of cytokine levels were, in turn, positively associated with increased AB toward negative and decreased AB toward positive information. While these ﬁndings provide support for the notion that stressor-evoked cytokine increases may drive, at least, short-term changes in attention processing with these eﬀects depending on the valence of emotional material, other studies could not conﬁrm this hypothesis (Benson et al., 2017; Niemegeers et al., 2019). The inconsistency of the ﬁndings may April 2019 \| Volume 13 \| Article 384 4 Stress, Inﬂammation and Emotional Attention Maydych (Slavich et al., 2010; Muscatell et al., 2016). Comparatively, the LPS-stimulated increase in IL-6 was positively associated with increased activation of the bilateral amygdala, dACC, and dorsal prefrontal cortex (dPFC) in response to negative feedback from the confederate based on a 10-min interview previously provided by participants (Muscatell et al., 2016). Other studies reported on the enhanced activity of right inferior orbitofrontal cortex (iOFC) (Kullmann et al., 2013) and subgenual ACC (Harrison et al., 2009) in response to viewing emotional pictures and faces, respectively. CONCLUSION In summary, preliminary evidence suggests that acute and chronic stress is associated with increased inﬂammatory activity and enhanced attentional processing of negative information. Both are predictive of negative mood and depression symptoms that, in turn, increase inﬂammatory and cognitive stress reactivity. Increased inﬂammation was associated with a pattern of attentional changes characteristic to depression, whereas aﬀective-cognitive states were predictive of inﬂammatory stress responses. These ﬁndings indicate that immune and aﬀective- cognitive processes are interconnected and may potentiate one another’s impact on depression onset, maintenance or recurrence. An improved understanding of the interplay between inﬂammatory activity and emotional cognition in the context of stress may help to optimize treatment strategies for depression. One of the most important molecular mechanisms linking inﬂammation to emotional cognition is the cytokine eﬀect on the serotonergic system (Capuron and Castanon, 2016). Proinﬂammatory cytokines activate indoleamine-2,3- dioxygenase (IDO), an enzyme involved in the synthesis of kynurenine from dietary tryptophan. Central and peripheral activation of IDO causes increased catabolism of tryptophan, an important precursor of serotonin, leading to serotonin deﬁciency (O’connor J. et al., 2009; O’Connor J.C. et al., 2009). Furthermore, the products of kynurenine metabolism, such as quinolinic acid, stimulate the N-methyl-D-aspartate (NMDA) receptor, thereby unfolding neurotoxic eﬀects leading to neuronal damage (Campbell et al., 2014). AUTHOR CONTRIBUTIONS VM has conceptualized and written the manuscript. MECHANISMS LINKING INFLAMMATION TO EMOTIONAL ATTENTION Peripherally released cytokines communicate with the brain and are capable of eliciting changes in emotional processing that mimic aﬀective-cognitive symptoms of depression. Research has identiﬁed several pathways by which cytokine signals can access the brain (Haroon et al., 2012). Brieﬂy, cytokines can enter the brain through leaky regions in the blood-brain barrier (e.g., circumventricular organs) or activated monocytes/macrophages recruited to the brain. In addition, cytokine release can be stimulated through brain blood vessel cells (e.g., endothelial cells). Furthermore, aﬀerent vagus nerve ﬁbers can be stimulated to transduce cytokine signals from the periphery into the brain, where the cytokine signals activate cytokine-producing glia cells. Frontiers in Neuroscience \| www.frontiersin.org REFERENCES Stress, Inﬂammation and Reward Processing: Implications for Major Depressive Disorder. Los Angeles, US: UCLA. Ellenbogen, M. A., Carson, R. J., and Pishva, R. (2010). Automatic emotional information processing and the cortisol response to acute psychosocial stress. Cognit. Aﬀect. Behav. Neurosci. 10, 71–82. doi: 10.3758/CABN.10.1.71 Boyle, C. C., Ganz, P. A., Van Dyk, K. M., and Bower, J. E. (2017). Inﬂammation and attentional bias in breast cancer survivors. Brain Behav. Immun. 66, 85–88. doi: 10.1016/j.bbi.2017.05.016 Callaghan, B. L., and Tottenham, N. (2016). The stress acceleration hypothesis: eﬀects of early-life adversity on emotion circuits and behavior. Curr. Opin. Behav. Sci. 7, 76–81. doi: 10.1016/j.cobeha.2015.11.018 Ellenbogen, M. A., Schwartzman, A. E., Stewart, J., and Walker, C.-D. (2002). Stress and selective attention: the interplay of mood, cortisol levels, and emotional information processing. Psychophysiology 39, 723–732. doi: 10.1111/1469-8986. 3960723 Campbell, B. M., Charych, E., Lee, A. W., and Möller, T. (2014). Kynurenines in CNS disease: regulation by inﬂammatory cytokines. Front. Neurosci. 8:12. doi: 10.3389/fnins.2014.00012 Ellenbogen, M. A., Schwartzman, A. E., Stewart, J., and Walker, C.-D. (2006). Automatic and eﬀortful emotional information processing regulates diﬀerent aspects of the stress response. Psychoneuroendocrinology 31, 373–387. doi: 10. 1016/j.psyneuen.2005.09.001 Capuron, L., and Castanon, N. (2016). “Role of inﬂammation in the development of neuropsychiatric symptom domains: evidence and mechanisms,” in Inﬂammation-Associated Depression: Evidence, Mechanisms and Implications, eds R. Dantzer and L. Capuron (Berlin: Springer), 31–44. doi: 10.1007/7854_ 2016_14 Fagundes, C. P., Glaser, R., Hwang, B. S., Malarkey, W. B., and Kiecolt-Glaser, J. K. (2013). Depressive symptoms enhance stress-induced inﬂammatory responses. Brain Behav. Immun. 31, 172–176. doi: 10.1016/j.bbi.2012.05.006 eds R. Dantzer and L. Capuron (Berlin: Springer), 31–44. doi: 10.1007/7854_ 2016_14 Capuron, L., and Miller, A. H. (2004). Cytokines and psychopathology: lessons from interferon-α. Biol. Psychiatry 56, 819–824. doi: 10.1016/j.biopsych.2004. 02.009 Goebel, M. U., Mills, P. J., Irwin, M. R., and Ziegler, M. G. (2000). Interleukin-6 and tumor necrosis factor-α production after acute psychological stress, exercise, and infused isoproterenol: diﬀerential eﬀects and pathways. Psychos. Med. 62, 591–598. doi: 10.1097/00006842-200007000-00019 Capuron, L., Raison, C. L., Musselman, D. L., Lawson, D. H., Nemeroﬀ, C. B., and Miller, A. H. (2003). Association of exaggerated HPA axis response to the initial injection of interferon-alpha with development of depression during interferon-alpha therapy. Am. J. Psychiatry 160, 1342–1345. doi: 10.1176/appi. ajp.160.7.1342 Gotlib, I. H., and Joormann, J. (2010). Cognition and depression: current status and future directions. Annu. Rev. Clin. Psychol. 6, 285–312. doi: 10.1146/annurev. clinpsy.121208.131305 Gotlib, I. REFERENCES H., Kasch, K. L., Traill, S., Joormann, J., Arnow, B. A., and Johnson, S. L. (2004a). Coherence and speciﬁcity of information-processing biases in depression and social phobia. J. Abnorm. Psychol. 113, 386. doi: 10.1037/0021- 843X.113.3.386 Carroll, J. E., Low, C. A., Prather, A. A., Cohen, S., Fury, J. M., Ross, D. C., et al. (2011). Negative aﬀective responses to a speech task predict changes in interleukin (IL)-6. Brain Behav. Immun. 25, 232–238. doi: 10.1016/j.bbi.2010. 09.024 Gotlib, I. H., Krasnoperova, E., Yue, D. N., and Joormann, J. (2004b). Attentional biases for negative interpersonal stimuli in clinical depression. J. Abnorm. Psychol. 113, 127. doi: 10.1037/0021-843X.113.1.121 Caseras, X., Garner, M., Bradley, B. P., and Mogg, K. (2007). Biases in visual orienting to negative and positive scenes in dysphoria: an eye movement study. J. Abnorm. Psychol. 116, 491. doi: 10.1037/0021-843X.116.3.491 Gray, M. A., Chao, C. Y., Staudacher, H. M., Kolosky, N. A., Talley, N. J., and Holtmann, G. (2018). Anti-TNFα therapy in IBD alters brain activity reﬂecting visceral sensory function and cognitive-aﬀective biases. PLoS One 13:e0193542. doi: 10.1371/journal.pone.0193542 Coelho, R., Viola, T., Walss-Bass, C., Brietzke, E., and Grassi-Oliveira, R. (2014). Childhood maltreatment and inﬂammatory markers: a systematic review. Acta Psychiatr. Scand. 129, 180–192. doi: 10.1111/acps.12217 Cooper, C. M., Godlewska, B., Sharpley, A. L., Barnes, E., Cowen, P. J., and Harmer, C. J. (2018). Interferon-α induces negative biases in emotional processing in patients with hepatitis C virus infection: a preliminary study. Psychol. Med. 48, 998–1007. doi: 10.1017/S0033291717002379 Hackett, R. A., Hamer, M., Endrighi, R., Brydon, L., and Steptoe, A. (2012). Loneliness and stress-related inﬂammatory and neuroendocrine responses in older men and women. Psychoneuroendocrinology 37, 1801–1809. doi: 10.1016/ j.psyneuen.2012.03.016 Danese, A., Moﬃtt, T. E., Pariante, C. M., Ambler, A., Poulton, R., and Caspi, A. (2008). Elevated inﬂammation levels in depressed adults with a history of childhood maltreatment. Arch. Gen. Psychiatry 65, 409–415. doi: 10.1038/ nrn2297 Hamer, M., and Steptoe, A. (2007). Association between physical ﬁtness, parasympathetic control, and proinﬂammatory responses to mental stress. Psychos. Med. 69, 660–666. doi: 10.1097/PSY.0b013e318148c4c0 Haroon, E., Raison, C. L., and Miller, A. H. (2012). Psychoneuroimmunology meets neuropsychopharmacology: translational implications of the impact of inﬂammation on behavior. Neuropsychopharmacology 37:137. doi: 10.1038/ npp.2011.205 Dantzer, R., O’Connor, J. C., Freund, G. G., Johnson, R. W., and Kelley, K. W. (2008). From inﬂammation to sickness and depression: when the immune system subjugates the brain. Nat. Rev. Neurosci. 9:46. doi: 10.1001/archpsyc.65. 4.409 Harrison, N. REFERENCES Disner, S. G., Beevers, C. G., Haigh, E. A., and Beck, A. T. (2011). Neural mechanisms of the cognitive model of depression. Nat. Rev. Neurosci. 12:467. doi: 10.1038/nrn3027 Allen, A. P., Kennedy, P. J., Cryan, J. F., Dinan, T. G., and Clarke, G. (2014). Allen, A. P., Kennedy, P. J., Cryan, J. F., Dinan, T. G., and Clarke, G. (2014). Biological and psychological markers of stress in humans: focus on the Trier Social Stress Test. Neurosci. Biobehav. Rev. 38, 94–124. doi: 10.1016/j.neubiorev. 2013.11.005 Disner, S. G., Shumake, J. D., and Beevers, C. G. (2017). Self-referential schemas and attentional bias predict severity and naturalistic course of depression Disner, S. G., Shumake, J. D., and Beevers, C. G. (2017). Self-referential schemas and attentional bias predict severity and naturalistic course of depression symptoms. Cognit. Emot. 31, 632–644. doi: 10.1080/02699931.2016.1146123 Aschbacher, K., Epel, E., Wolkowitz, O., Prather, A., Puterman, E., and Dhabhar, F. (2012). Maintenance of a positive outlook during acute stress protects against pro-inﬂammatory reactivity and future depressive symptoms. Brain Behav. Immun. 26, 346–352. doi: 10.1016/j.bbi.2011. 10.010 Dooley, L. N., Kuhlman, K. R., Robles, T. F., Eisenberger, N. I., Craske, M. G., and Bower, J. E. (2018). The role of inﬂammation in core features of depression: Insights from paradigms using exogenously-induced inﬂammation. Neurosci. Biobehav. Rev. 94, 219–237. doi: 10.1016/j.neubiorev.2018.09.006 Biobehav. Rev. 94, 219–237. doi: 10.1016/j.neubiorev.2018.09.006 Dorr, A. E., and Debonnel, G. (2006). Eﬀect of vagus nerve stimulation on serotonergic and noradrenergic transmission. J. Pharmacol. Exp. Therap. 318, 890–898. doi: 10.1124/jpet.106.104166 Beevers, C. G., and Carver, C. S. (2003). Attentional bias and mood persistence as prospective predictors of dysphoria. Cognit. Ther. Res. 27, 619–637. doi: 10.1023/A:1026347610928 Dowlati, Y., Herrmann, N., Swardfager, W., Liu, H., Sham, L., Reim, E. K., et al. (2010). A meta-analysis of cytokines in major depression. Biol. Psychiatry 67, 446–457. doi: 10.1016/j.biopsych.2009.09.033 Benson, S., Brinkhoﬀ, A., Lueg, L., Roderigo, T., Kribben, A., Wilde, B., et al. (2017). Eﬀects of acute systemic inﬂammation on the interplay between sad mood and aﬀective cognition. Transl. Psychiatry 7, 1281. doi: 10.1038/s41398-017- 0043-0 (2010). A meta-analysis of cytokines in major depression. Biol. Psychiatry 67, 446–457. doi: 10.1016/j.biopsych.2009.09.033 Eisenberger, N. I., Berkman, E. T., Inagaki, T. K., Rameson, L. T., Mashal, N. M., and Irwin, M. R. (2010). Inﬂammation-induced anhedonia: endotoxin reduces ventral striatum responses to reward. Biol. Psychiatry 68, 748–754. doi: 10.1016/ j.biopsych.2010.06.010 Boyle, C. C. (2018). ACKNOWLEDGMENTS The author thanks Dr. Thomas Kleinsorge for proofreading of the ﬁrst version of the manuscript. The publication of this article was funded by the Open Access Fund of the Leibniz Association. Another mechanism linking inﬂammation with cognition is the eﬀect of cytokines on the HPA axis. Cytokines can act on glucocorticoid receptors and indirectly upregulate April 2019 \| Volume 13 \| Article 384 Frontiers in Neuroscience \| www.frontiersin.org 5 Stress, Inﬂammation and Emotional Attention Maydych REFERENCES Paroxetine for the prevention of depression induced by high-dose interferon alfa. N. Engl. J. Med. 344, 961–966. doi: 10. 1056/NEJM200103293441303 Kiecolt-Glaser, J. K., Gouin, J. P., Weng, N. P., Malarkey, W. B., Beversdorf, D. Q., and Glaser, R. (2011). Childhood adversity heightens the impact of later-life caregiving stress on telomere length and inﬂammation. Psychos. Med. 73:16. doi: 10.1097/PSY.0b013e31820573b6 Niemegeers, P., de Boer, P., Schuermans, J., Dumont, G. J., Coppens, V., Spittaels, K., et al. (2019). Digging deeper in the diﬀerential eﬀects of inﬂammatory and psychosocial stressors in remitted depression: eﬀects on cognitive functioning. J. Aﬀect. Disord. 245, 356–363. doi: 10.1016/j.jad.2018. 11.020 Kirschbaum, C., Pirke, K.-M., and Hellhammer, D. H. (1993). The ‘Trier Social Stress Test’–a tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology 28, 76–81. doi: 10.1159/000119004 O’connor, J., Lawson, M., Andre, C., Moreau, M., Lestage, J., Castanon, N., et al. (2009). Lipopolysaccharide-induced depressive-like behavior is mediated by indoleamine 2, 3-dioxygenase activation in mice. Mol. Psychiatry 14:511. doi: 10.1038/sj.mp.4002148 Köhler, O., Benros, M. E., Nordentoft, M., Farkouh, M. E., Iyengar, R. L., Mors, O., et al. (2014). Eﬀect of anti-inﬂammatory treatment on depression, depressive symptoms, and adverse eﬀects: a systematic review and meta- analysis of randomized clinical trials. JAMA Psychiatry 71, 1381–1391. doi: 10.1001/jamapsychiatry.2014.1611 O’Connor, J. C., Lawson, M. A., André, C., Briley, E. M., Szegedi, S. S., Lestage, J., et al. (2009). Induction of IDO by bacille calmette-guerin is responsible for development of murine depressive-like behavior. J. Immunol. 182, 3202–3212. doi: 10.4049/jimmunol.0802722 Kojima, M., Kojima, T., Suzuki, S., Oguchi, T., Oba, M., Tsuchiya, H., et al. (2009). Depression, inﬂammation, and pain in patients with rheumatoid arthritis. Arthritis Care Res. 61, 1018–1024. doi: 10.1002/art.24647 O’Donnell, K., Brydon, L., Wright, C. E., and Steptoe, A. (2008). Self-esteem levels and cardiovascular and inﬂammatory responses to acute stress. Brain Behav. Immun. 22, 1241–1247. doi: 10.1016/j.bbi.2008.06.012 Koster, E. H., De Raedt, R., Goeleven, E., Franck, E., and Crombez, G. (2005). Mood-congruent attentional bias in dysphoria: maintained attention to and impaired disengagement from negative information. Emotion 5:446. doi: 10. 1037/1528-3542.5.4.446 Ondicova, K., Pecenak, J., and Mravec, B. (2010). The role of the vagus nerve in depression. Neuroendocrinol. Lett. 31:602. Kuhlman, K. R., Robles, T. F., Dooley, L. N., Boyle, C. C., Haydon, M. D., and Bower, J. E. (2018). Within-subject associations between inﬂammation and features of depression: Using the ﬂu vaccine as a mild inﬂammatory stimulus. Brain Behav. Immun. 69, 540–547. REFERENCES A., Brydon, L., Walker, C., Gray, M. A., Steptoe, A., and Critchley, H. D. (2009). Inﬂammation causes mood changes through alterations in subgenual cingulate activity and mesolimbic connectivity. Biol. Psychiatry 66, 407–414. doi: 10.1016/j.biopsych.2009.03.015 DellaGioia, N., and Hannestad, J. (2010). A critical review of human endotoxin administration as an experimental paradigm of depression. Neurosci. Biobehav. Rev. 34, 130–143. doi: 10.1016/j.neubiorev.2009.07.014 April 2019 \| Volume 13 \| Article 384 Frontiers in Neuroscience \| www.frontiersin.org 6 Stress, Inﬂammation and Emotional Attention Maydych Moons, W. G., Eisenberger, N. I., and Taylor, S. E. (2010). Anger and fear responses to stress have diﬀerent biological proﬁles. Brain Behav. Immun. 24, 215–219. doi: 10.1016/j.bbi.2009.08.009 Hostinar, C. E., Nusslock, R., and Miller, G. E. (2018). Future directions in the study of early-life stress and physical and emotional health: implications of the neuroimmune network hypothesis. J. Clin. Child Adolesc. Psychol. 47, 142–156. doi: 10.1080/15374416.2016.1266647 Muscatell, K. A., Dedovic, K., Slavich, G. M., Jarcho, M. R., Breen, E. C., Bower, J. E., et al. (2015). Greater amygdala activity and dorsomedial prefrontal-amygdala coupling are associated with enhanced inﬂammatory responses to stress. Brain Behav. Immun. 43, 46–53. doi: 10.1016/j.bbi.2014.06.201 Howren, M. B., Lamkin, D. M., and Suls, J. (2009). Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychos3. Med. 71, 171–186. doi: 10.1097/PSY.0b013e3181907c1b Muscatell, K. A., and Eisenberger, N. I. (2012). A social neuroscience perspective on stress and health. Soc. Pers. Psychol. Compass 6, 890–904. doi: 10.1111/j.1751- 9004.2012.00467.x Inagaki, T. K., Muscatell, K. A., Irwin, M. R., Cole, S. W., and Eisenberger, N. I. (2012). Inﬂammation selectively enhances amygdala activity to socially threatening images. Neuroimage 59, 3222–3226. doi: 10.1016/j.neuroimage. 2011.10.090 Muscatell, K. A., Moieni, M., Inagaki, T. K., Dutcher, J. M., Jevtic, I., Breen, E. C., et al. (2016). Exposure to an inﬂammatory challenge enhances neural sensitivity to negative and positive social feedback. Brain Behav. Immun. 57, 21–29. doi: 10.1016/j.bbi.2016.03.022 Ingram, R. E., Miranda, J., and Segal, Z. V. (1998). Cognitive Vulnerability to Depression. New York, NY: Guilford Press. Joormann, J., and Gotlib, I. H. (2007). Selective attention to emotional faces Joormann, J., and Gotlib, I. H. (2007). Selective attention to emotional faces following recovery from depression. J. Abnorm. Psychol. 116:80. doi: 10.1037/ 0021-843X.116.1.80 following recovery from depression. J. Abnorm. Psychol. 116:80. doi: 10.1037/ 0021-843X.116.1.80 Musselman, D. L., Lawson, D. H., Gumnick, J. F., Manatunga, A. K., Penna, S., Goodkin, R. S., et al. (2001). REFERENCES doi: 10.1016/j.bbi.2018.02.001 Pace, T. W., Mletzko, T. C., Alagbe, O., Musselman, D. L., Nemeroﬀ, C. B., Miller, A. H., et al. (2006). Increased stress-induced inﬂammatory responses in male patients with major depression and increased early life stress. Am. J. Psychiatry 163, 1630–1633. doi: 10.1176/ajp.2006.163.9.1630 Kullmann, J. S., Grigoleit, J. S., Lichte, P., Kobbe, P., Rosenberger, C., Banner, C., et al. (2013). Neural response to emotional stimuli during experimental human endotoxemia. Hum. Brain Mapp. 34, 2217–2227. doi: 10.1002/hbm.22063 Raison, C. L., Capuron, L., and Miller, A. H. (2006). Cytokines sing the blues: inﬂammation and the pathogenesis of depression. Trends Immunol. 27, 24–31. doi: 10.1016/j.it.2005.11.006 Lewinsohn, P. M., Joiner, T. E., and Rohde, P. (2001). Evaluation of cognitive diathesis-stress models in predicting major depressive disorder in adolescents. J. Abnorm. Psychol. 110:203. doi: 10.1037/0021-843X.110.2.203 Raison, C. L., and Miller, A. D. (2003). When not enough is too much: the role of insuﬃcient glucocorticoid signaling in the pathophysiology of stress- related disorders. Am. J. Psychiatry 160, 1554–1565. doi: 10.1176/appi.ajp.160.9. 1554 Luecken, L. J., and Appelhans, B. (2005). Information-processing biases in young adults from bereaved and divorced families. J. Abnorm. Psychol. 114:309. doi: 10.1037/0021-843X.114.2.309 Raymond, C., Marin, M.-F., Majeur, D., and Lupien, S. (2018). Early child adversity and psychopathology in adulthood: HPA axis and cognitive dysregulations as potential mechanisms. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 85, 152–160. doi: 10.1016/j.pnpbp.2017.07.015 MacLeod, C., Mathews, A., and Tata, P. (1986). Attentional bias in emotional disorders. J. Abnorm. Psychol. 95:15. doi: 10.1037/0021-843X.95.1.15 Marsland, A. L., Walsh, C., Lockwood, K., and John-Henderson, N. A. (2017). The eﬀects of acute psychological stress on circulating and stimulated inﬂammatory markers: a systematic review and meta-analysis. Brain Behav. Immun. 64, 208–219. doi: 10.1016/j.bbi.2017.01.011 Roelofs, K., Bakvisa, P., Hermans, E. J., van Pelt, J., and van Honk, J. (2007). The eﬀects of social stress and cortisol responses on the preconscious selective attention to social threat. Biol. Psychol. 75, 1–7. doi: 10.1016/j.biopsycho.2006. 09.002 Mathews, A., and MacLeod, C. (2005). Cognitive vulnerability to emotional disorders. Annu. Rev. Clin. Psychol. 1, 167–195. doi: 10.1146/annurev.clinpsy. 1.102803.143916 Rohleder, N. (2014). Stimulation of systemic low-grade inﬂammation by psychosocial stress. Psychos. Med. 76, 181–189. doi: 10.1097/PSY. 0000000000000049 Matsuo, K., Glahn, D., Peluso, M., Hatch, J., Monkul, E., Najt, P., et al. (2007). Prefrontal hyperactivation during working memory task in untreated individuals with major depressive disorder. Mol. Psychiatry 12:158. doi: 10. 1038/sj.mp.4001894 Rosas-Ballina, M., and Tracey, K. J. (2009). Cholinergic control of inﬂammation. J. Int. REFERENCES Med. 265, 663–679. doi: 10.1111/j.1365-2796.2009.02098.x Rothermundt, M., Arolt, V., Peters, M., Gutbrodt, H., Fenker, J., Kersting, A., et al. (2001). Inﬂammatory markers in major depression and melancholia. J. Aﬀect. Disord. 63, 93–102. doi: 10.1016/S0165-0327(00)00157-9 Maydych, V., Claus, M., Watzl, C., and Kleinsorge, T. (2018). Attention to emotional information is associated with cytokine responses to psychological stress. Front. Behav. Neurosci. 12:687. doi: 10.3389/fnins.2018.00687 Sanchez, A., Romero, N., and De Raedt, R. (2017). Depression-related diﬃculties disengaging from negative faces are associated with sustained attention to negative feedback during social evaluation and predict stress recovery. PLoS One 12:e0175040. doi: 10.1371/journal.pone.0175040 Miskowiak, K. W., and Carvalho, A. F. (2014). Hot cognition in major depressive disorder: a systematic review. CNS Neurol. Disord. Drug Targets 13, 1787–1803. doi: 10.2174/1871527313666141130205713 April 2019 \| Volume 13 \| Article 384 Frontiers in Neuroscience \| www.frontiersin.org Stress, Inﬂammation and Emotional Attention Maydych event-related FMRI study with the Stroop task. Biol. Psychiatry 59, 958–965. doi: 10.1016/j.biopsych.2005.10.025 Slavich, G. M., and Irwin, M. R. (2014). From stress to inﬂammation and major depressive disorder: a social signal transduction theory of depression. Psychol. Bull. 140:774. doi: 10.1037/a0035302 Weber, C. S., Thayer, J. F., Rudat, M., Wirtz, P. H., Zimmermann-Viehoﬀ, F., Thomas, A., et al. (2010). Low vagal tone is associated with impaired post stress recovery of cardiovascular, endocrine, and immune markers. Eur. J. Appl. Physiol. 109, 201–211. doi: 10.1007/s00421-009-1341-x Slavich, G. M., Way, B. M., Eisenberger, N. I., and Taylor, S. E. (2010). Neural sensitivity to social rejection is associated with inﬂammatory responses to social stress. Proc. Natl. Acad. Sci. U.S.A. 107:201009164. doi: 10.1073/pnas. 1009164107 Physiol. 109, 201–211. doi: 10.1007/s00421-009-1341-x Weinstein, A. A., Deuster, P. A., Francis, J. L., Bonsall, R. W., Tracy, R. P., and Kop, W. J. (2010). Neurohormonal and inﬂammatory hyper-responsiveness to acute mental stress in depression. Biol. Psychol. 84, 228–234. doi: 10.1016/j.biopsycho. 2010.01.016 Smeets, T. (2010). Autonomic and hypothalamic-pituitary-adrenal stress resilience: impact of cardiac vagal tone. Biol. Psychol. 84, 290–295. doi: 10.1016/j.biopsycho.2010.02.015 Thayer, J. F., Åhs, F., Fredrikson, M., Sollers, J. J. Jr., and Wager, T. D. (2012). A meta-analysis of heart rate variability and neuroimaging studies: implications for heart rate variability as a marker of stress and health. Neurosci. Biobehav. Rev. 36, 747–756. doi: 10.1016/j.neubiorev.2011. 11.009 Woody, A., Figueroa, W. S., Benencia, F., and Zoccola, P. M. (2017). Stress-induced parasympathetic control and its association with inﬂammatory reactivity. Psychos. Med. 79, 306–310. Frontiers in Neuroscience \| www.frontiersin.org April 2019 \| Volume 13 \| Article 384 REFERENCES doi: 10.1097/PSY.0000000000000426 Yamakawa, K., Matsunaga, M., Isowa, T., Kimura, K., Kasugai, K., Yoneda, M., et al. (2009). Transient responses of inﬂammatory cytokines in acute stress. Biol. Psychol. 82, 25–32. doi: 10.1016/j.biopsycho.2009.05.001 Thayer, J. F., and Sternberg, E. (2006). Beyond heart rate variability: vagal regulation of allostatic systems. Ann. N. Y. Acad. Sci. 1088, 361–372. doi: 10. 1196/annals.1366.014 Conﬂict of Interest Statement: The author declares that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Tsumura, H., and Shimada, H. (2012). Acutely elevated cortisol in response to stressor is associated with attentional bias toward depression-related stimuli but is not associated with attentional function. Appl. Psychophysiol. Biofeedback 37, 19–29. doi: 10.1007/s10484-011-9172-z Copyright © 2019 Maydych. 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. Copyright © 2019 Maydych. 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. Valkanova, V., Ebmeier, K. P., and Allan, C. L. (2013). CRP, IL-6 and depression: a systematic review and meta-analysis of longitudinal studies. J. Aﬀect. Disord. 150, 736–744. doi: 10.1016/j.jad.2013.06.004 Wagner, G., Sinsel, E., Sobanski, T., Köhler, S., Marinou, V., Mentzel, H.-J., et al. (2006). Cortical ineﬃciency in patients with unipolar depression: an April 2019 \| Volume 13 \| Article 384 Frontiers in Neuroscience \| www.frontiersin.org 8
https://openalex.org/W2007009846	https://europepmc.org/articles/pmc4145629?pdf=render	English	null	Statistical evaluation of the medium components for the production of high biomass, α-amylase and protease enzymes by Piriformospora indica using Plackett–Burman experimental design	3 biotech	2,013	cc-by	4,102	Statistical evaluation of the medium components for the production of high biomass, a-amylase and protease enzymes by Piriformospora indica using Plackett–Burman experimental design S. Swetha • Ajit Varma • T. Padmavathi Received: 8 April 2013 / Accepted: 26 August 2013 / Published online: 6 September 2013 The Author(s) 2013. This article is published with open access at Springerlink.com Abstract Piriformospora indica, a member of basidio- mycota is an axenically cultivable endophytic fungus which exerts plant growth promoting effects on its host plant. P. indica is known to produce a-amylase and pro- tease. Since the organism exhibits beneﬁcial role in plant growth promotion, achieving high biomass is immensely essential. Hence to enable the commercial production, screening of medium components is a necessary step. The present paper investigates the screening of medium com- ponents using Plackett–Burman experimental design wherein the parameters such as a-amylase, protease and biomass have been examined. The parameters a-amylase, protease and biomass was found to vary from 0.25 to 0.45 mg-1 ml-1 min-1, 0.1 to 0.15 mg-1 ml-1 h-1 and 0.8 to 22.6 g l-1, respectively, in 16 runs which demon- strates the strong inﬂuence of the medium components. straight parenthosomes. It was isolated from the rhizo- sphere of the woody shrubs Prosopsis juliﬂora and Zizy- phus nummularia growing in Indian Thar desert (Varma et al. 1998). P. indica enhances the growth and yield of plants, protects against biotic (resistance against patho- gens) and abiotic stress (salt stress) (Rai et al. 2001, Waller et al. 2008) by colonizing a broad spectrum of mono and dicotyledenous plants. It also produces a large number of thick walled, pear shaped chlamydospores having longer shelf life. The sporulating nature of this fungus has made it competent enough for exploiting it for commercial appli- cation (Kumar et al. 2011). Microbial enzymes ﬁnd increasing industrial applica- tions; among them amylases and proteases occupy large share of the total enzyme market. a-Amylases and prote- ases have wide applications in starch processing, brewing, alcohol production, dairy industries, and textile industries, pharmaceutical and detergent industries, respectively (Sumrin et al. 2011; Saxena and Singh 2010). Keywords Piriformospora indica Submerged fermentation Plackett–Burman experimental design a-Amylase Protease Keywords Piriformospora indica Submerged fermentation Plackett–Burman experimental design a-Amylase Protease Keywords Piriformospora indica Submerged fermentation Plackett–Burman experimental design a-Amylase Protease The cell growth, spore production and accumulation of metabolic products are strongly inﬂuenced by the medium components such as carbon sources, nitrogen sources, various inorganic salts and trace elements. 3 Biotech (2014) 4:439–445 DOI 10.1007/s13205-013-0168-7 3 Biotech (2014) 4:439–445 DOI 10.1007/s13205-013-0168-7 SHORT REPORT A. Varma Amity Institute of Herbal and Microbial Studies, Sector 125, Noida 201303, UP, India S. Swetha T. Padmavathi (&) Department of Microbiology, Centre of PG Studies, Jain University, 9th Main, Jayanagar 3rd Block, Bangalore, India e-mail: vam2010tpraviju@gmail.com Statistical evaluation of the medium components for the production of high biomass, a-amylase and protease enzymes by Piriformospora indica using Plackett–Burman experimental design Screening of medium components and their optimization is therefore an important criterion for large-scale production. Classical method, factorial combination of medium optimization involving one variable at a time by keeping others at ﬁxed level fails as it is laborious, time consuming; moreover, it does not guarantee the optimal conditions. Hence statistical approach such as Plackett–Burman design (provide statis- tical model) is a useful tool for the screening of nutrients as it helps in determining the signiﬁcant impact on growth rate, which in turn aids in understanding the interactions among the process parameters at different levels (Rajen- dran et al. 2007). The use of statistical experimental design 123 Protease activity Protease activity using casein as substrate was performed according to Rodarte et al. (2011) with some modiﬁcations. The reaction mixture containing cell-free extract, 0.5 % casein along with citrate buffer of pH 5 was incubated for 4 h at room temperature followed by the addition of 5 % trichloroacetic acid and centrifuged at 8,000 rpm for 10 min. The supernatant was further analyzed by Lowry’s method (Lowry et al 1951). One unit of the protease activity was deﬁned as the amount of enzyme required to liberate 1 lg of tyrosine per hour under the experimental conditions. Plackett–Burman design A total of 12 variables including different carbon, nitrogen sources, trace element solution [zinc sulphate (ZnSO47H2O), boric acid (H3BO3), manganous chloride (MnCl2), cobaltous chloride (CoCl2), copper sulphate (CuSO4) and ammonium molybdate ((NH4)6Mo7O244H2O) and 209 salt solution [sodium nitrate (NaNO3), potassium chloride, magnesium sulphate (MgSO47H2O) and potas- sium di hydrogen phosphate (KH2PO4)] at different levels as per the statistical model were used along with potato dex- trose media for screening (Table 1). The Plackett–Burman design based on the ﬁrst order model was used to screen and evaluate the important medium components that inﬂuence the production of a-amylase and protease. All the experiments were carried out according to designed matrix using the equation Y ¼ bo þ P biXi i ¼ ð l. . .kÞ, where Y is the estimated target function (yield of a- amylase/protease/biomass), bo is a model intercept, bi is the regression coefﬁcient. X is the independent variable and k is the number of variables (Rajendran et al. 2007; Seraman et al. 2010). The Student’s t test was performed to determine the signiﬁcance of each variable employed. The regression coefﬁcient was determined by least square method. a-Amylase activity a-Amylase activity a-Amylase activity a-Amylase activity a-Amylase activity was performed using 1 % soluble starch as substrate (optimum condition) followed by the estimation of reducing sugars using dinitrosalicylic acid (Miller 1959). One unit of amylase activity was deﬁned as the amount of enzyme that liberates reducing sugar equivalent to 1.0 mg glucose under speciﬁc assay conditions. Submerged fermentation 1 9 10-6 ml-1 of the culture was inoculated to all the ﬂasks (16 runs) designed as per the statistical experimental design and these ﬂasks were incubated at 30 C under continuous shaking conditions (120 rpm) and the response was measured in terms of biomass, amylase and protease production. Microorganism and culture maintenance The culture of P. indica with an accession number of AF014929 USA was obtained from Prof. Ajit Varma (Amity Institute of Herbal and Microbial Studies, Noida, India). The stock culture was maintained on potato dextrose agar at optimum conditions and stored at 4 C for further studies. Materials and methods Microorganism and culture maintenance Introduction Piriformospora indica is an axenically cultivable, faculta- tive root endophytic fungus belonging to basidiomycota possessing simple dolipore septum with continuous and 123 123 3 Biotech (2014) 4:439–445 440 48–72 h and the growth of P. indica was expressed in terms of dry cell weight per liter of the culture broth. in medium optimization has gained considerable attention in recent years and also number of publications describing the application of these methods for the production of various enzymes and biomolecules has appeared in the literature (Seraman et al. 2010). However, there seems to be no reports available for the production of a-amylases and proteases by P. indica. The present paper demonstrates the screening of the important medium components affecting the large-scale cultivation of P. indica and also the production of a-amylase and protease enzymes by P. indica, which has tremendous applications in the ﬁeld of plant biotechnology as a biological hardening tool, bio- fertilizer and biocontrol agent. Results higher level (0.1 g), - lower level (0.05 g) C1 þ C2 soluble starch þ C3 sodium nitrate þ C4 maltose þ C5 ammonium chloride þ C6 malt extract þ C7 peptone þ C8 sucrose þ C9 TE solution þ C10 glycerol þ C11 yeast extract þ C12 CMC þ C13 salt solution: C1 þ C2 soluble starch þ C3 sodium nitrate þ C4 maltose þ C5 ammonium chloride þ C6 malt extract þ C7 peptone þ C8 sucrose þ C9 TE solution þ C10 glycerol þ C11 yeast extract þ C12 CMC þ C13 salt solution: negative variables based on their main effect which in turn highlights the direction of optimization. More than half of the considered parameters were signiﬁcantly acting on the system. The components such as starch, glycerol, maltose, TE solution and salt solution have positive effect, whereas sodium nitrate and ammonium chloride has a negative effect on the above-mentioned activities. Pareto plots give a clear view about the role of each var- iable (Fig. 1a–c). Soluble starch, maltose, sucrose, TE solution and glycerol had a positive effect on a-amylase activity where as sodium nitrate, ammonium chloride, yeast extract, malt extract, carboxymethyl cellulose and salt solution had negative effect as shown in Fig. 1a. Figure 1b depicts the pareto plot of protease activity which reveals that maltose, malt extract, TE solution,glycerol, yeast extract and salt solution had a positive effect, whereas the soluble starch, sodium nitrate, ammonium chloride, peptone, sucrose, car- boxymethyl cellulose had a negative effect. Soluble starch, maltose, ammonium chloride, malt extract, peptone, TE solution, yeast extract and salt solution supported in enhancing the biomass but sodium nitrate, sucrose, glycerol and carboxymethyl cellulose depicted negative role in the enhancement of growth of P. indica (Fig. 1c). Results The coefﬁcient and the p values of the factors were studied using the statistical software MATLAB. a-Amylase activ- ity, protease activity and biomass were found to vary from 0.2555 to 0.4554 mg-1 ml-1 min-1, 0.1 to 0.158 mg-1 ml-1 h-1 and 0.8 to 22.6 g l-1, respectively, in 16 runs which shows the strong inﬂuence of the medium compo- nents. The variables were classiﬁed into positive and Measurement of cell growth The culture broth was ﬁltered through pre-weighed What- man No. 1 ﬁlter paper, dried in a hot air oven at 60 C for 123 441 3 Biotech (2014) 4:439–445 Table 1 Plackett–Burman experimental design for the evaluation of 12 variables for a-amylase, protease and biomass Run Soluble starch Maltose Sucrose Glycerol Carboxy methyl cellulose Sodium nitrate Ammonium chloride Malt extract Peptone Yeast extract TE solution 209 salt solution 1 ? ? - ? - ? ? - ? - ? ? 2 ? ? ? - ? ? - ? - - ? - 3 ? - ? - - ? ? - ? ? - - 4 ? ? - ? - - - ? ? - - - 5 ? - - - - ? ? ? - - ? ? 6 ? ? ? - ? - ? - - - - ? 7 - ? - - ? ? - - ? ? - ? 8 ? - - ? ? ? - ? - ? - ? 9 ? - - ? ? - ? - - ? ? - 10 - ? ? ? - - - - - ? ? ? 11 - - ? ? ? ? - - ? - ? - 12 ? - ? - - - - ? ? ? ? ? 13 - - ? ? ? - ? ? ? - - ? 14 - ? - - ? - ? ? ? ? ? - 15 - ? ? ? - ? ? ? - ? - - 16 - - - - - - - - - - - - ? higher level (0.1 g), - lower level (0.05 g) tt–Burman experimental design for the evaluation of 12 variables for a-amylase, protease and biomass ? a-Amylase activity The linear effect of soluble starch, ammonium chloride and TE solution were found to be more signiﬁcant than other variables for the production of a-amylase enzyme by P. indica as given in Table 2. Figure 2 represents the response surface plots of the effect of various combinations of independent variables such as (1) ammonium chloride and soluble starch, (2) glycerol and TE solution for the production of a-amylase. The interactions of ammonium chloride and soluble starch on a-amylase production was signiﬁcant as shown in Fig. 2a. Activity increased with increase in concentration of soluble starch, whereas the effect was opposite for ammonium chloride as depicted in Table 2. The maximum value observed was 0.43 mg-1 ml-1 min-1 at 0.05 % (w/v) of ammonium chloride and 0.1 % (w/v) of soluble starch. Similarly in case of glycerol and TE solution, the maximum value observed was 0.42 mg-1 ml-1 min-1 at 0.1 % (w/v) of glycerol and TE solution as the effect of both the variables were positive and dominating (Fig. 2b). Total terms in the quadratic equation were 91 including 1 constant, 66 interactions, 12 linear and 12 pure quadratic terms. To obtain the relevant parameter, the constant term and the linear effect alone were taken into consideration, and the following formula was used. 123 3 Biotech (2014) 4:439–445 442 -4 -2 0 2 4 6 8 Medium Components Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt Amm Chl ME -4 -2 0 2 4 T value Effect Medium Components -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 Medium Components (a) (b) (c) Suc Glycerol TE solution Constant Sol Starch Sod nit Maltose YE CMC Salt Pep T value Effect T value Effect Fig. a-Amylase activity 1 Pareto plots depicting the main effect of the variables on a a-amylase activity, b protease activity and c biomass -4 -2 0 2 4 6 8 Medium Components Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt -4 -2 0 2 4 T value Effect Medium Components (a) (b) T value Effect Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt -4 -2 0 2 4 T value Effect Medium Components (a) -4 -2 0 2 4 6 8 Medium Components Constant Sol Starch Sod nit Maltose Amm Chl ME Pep Suc TE solution Glycerol YE CMC Salt (b) T value Effect (b) (a) T value Effect T value Effect Medium Components Amm Chl ME -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 Medium Components (c) Suc Glycerol TE solution Constant Sol Starch Sod nit Maltose YE CMC Salt Pep T value Effect Fig. 1 Pareto plots depicting the main effect of the variables on a a-amylase activity, b protease activity and c biomass Medium Components Pareto plots depicting the main effect of the variables on a a-amylase activity, b protease activity and c biomass icting the main effect of the variables on a a-amylase activity, b protease activity and c biomass Table 2 Regression coefﬁcient results from the data of central composite designed experiments for a-amylase activity and protease activity p value \0.05 is signiﬁcant Sl. no. a-Amylase activity Components a-Amylase activity Protease activity t stat p value t stat p value 1 Constant 4.1280 0.0258 8.3769 0.0036 2 Soluble starch 4.9329 0.0160 -1.6904 0.1895 3 Sodium nitrate -1.7097 0.1859 -3.4698 0.0404 4 Maltose 1.8143 0.1673 4.0036 0.0279 5 Ammonium chloride -4.1014 0.0262 -3.2919 0.0460 6 Malt extract -1.8143 0.1673 2.2242 0.1126 7 Peptone -0.7736 0.4956 -2.2242 0.1126 8 Sucrose 0.6745 0.5483 -0.9787 0.3999 9 TE solution 4.3088 0.0230 0.8007 0.4818 10 Glycerol 3.1596 0.0509 2.4022 0.0957 11 Yeast extract -0.2635 0.8093 1.3345 0.2743 12 CMC -0.1532 0.8880 -0.9787 0.3999 13 Salt solution -2.2273 0.1123 2.7581 0.0703 123 3 Biotech (2014) 4:439–445 443 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.32 0.34 0.36 0.38 0.4 0.42 Soluble Starch Ammonium Chloride Amylase Activity (mg/ml/min) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.32 0.34 0.36 0.38 0.4 TE solution Glycerol Amylase Activity (mg/ml/min) (a) (b) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Maltose Salt Solution Acid Protease (mg/ml/hr) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Sodium Nitrate Maltose Acid Protease (mg/ml/hr) (c) (d) Fig. 2 3D response plots showing the effects of independent variables on a-amylase and protease activity 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.32 0.34 0.36 0.38 0.4 TE solution Glycerol Amylase Activity (mg/ml/min) (b) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.32 0.34 0.36 0.38 0.4 0.42 Soluble Starch Ammonium Chloride Amylase Activity (mg/ml/min) (a) (b) (a) Ammonium Chloride 0.05 TE solution Glycerol 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Maltose Salt Solution Acid Protease (mg/ml/hr) 0.05 0.06 0.07 0.08 0.09 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Sodium Nitrate Maltose Acid Protease (mg/ml/hr) (c) (d) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Maltose Salt Solution Acid Protease (mg/ml/hr) (c) 0.05 0.06 0.07 0.08 0.09 0.1 0.06 0.08 0.1 0.12 0.125 0.13 0.135 Sodium Nitrate Maltose Acid Protease (mg/ml/hr) (d) (d) (c) Fig. 2 3D response plots showing the effects of independent variables on a-amylase and protease activity of 0.135 mg-1 ml-1 h-1 at 0.1 % (w/v) of salt solution and 0.1 % (w/v) of maltose (Fig. 2d). of 0.135 mg-1 ml-1 h-1 at 0.1 % (w/v) of salt solution and 0.1 % (w/v) of maltose (Fig. 2d). Protease activity The effect of sodium nitrate, maltose and 209 salt solution were found to be more signiﬁcant than those of other variables for the production of protease enzyme is as evi- dent from the Table 2. Figure 2c, d represents the response surface plots of the effect of combinations of independent variables such as (1) maltose and sodium nitrate (2) salt solution and maltose for the production of protease enzyme. The interactions of maltose and sodium nitrate on protease production were signiﬁcant as shown in Fig. 2c. With the increase in concentration of maltose, the activity increased where as the activity decreased with increase in concentration of sodium nitrate. The maximum value observed was 0.136 mg-1 ml-1 h-1 at 0.1 % (w/v) of maltose and 0.05 % (w/v) of sodium nitrate. The interac- tions of salt solution and maltose gave the highest activity Biomass The linear effects of peptone, glycerol and yeast extract for biomass were found to be moderately signiﬁcant. The biomass varied signiﬁcantly with the concentration of glycerol, peptone and achieved the maximum biomass of 12 g l-1 at 0.1 % (w/v) of peptone and 0.05 % (w/v) of glycerol. References Babu KR, Satyanarayana T (1993) Parametric optimization of extracellular a-amylase production by thermophilic Bacillus coagulans. Folia Microbiol 38:77–80 Carbon sources such as glucose, sucrose, maltose, etc., have a positive inﬂuence on the production of protease and also salts such as KH2PO4, MgSO47H2O and KCl have positive impact on protease production as reported by Kezia et al. 2011. Present study is in concordance with Luciana and Sato (2009) who reported that concentration of KH2PO4 plays an important role in production of pro- teolytic enzymes. Gangadharan D, Sivaramakrishnan S, Namboothiri KM, Pandey A (2006) Solid culturing of Bacillus amyloliquefaciens for a- amylase Production. Food Technol Biotechnol 44(2):269–274 Kezia D, Swarnalath G, Dinesh Yadav V, Naidu SV, Narasimha Rao M (2011) Screening of nutritional components for alkaline protease production in submerged fermentation by Bacillus subtilis DKMNR using Plackett–Burman design. Res J Pharm Biol Chem Sci 2(3):598–606 Kumar V, Sahai V, Bisaria VS (2011) High-density spore production of Piriformospora indica, a plant growth-promoting endophyte, by optimization of nutritional and cultural parameters. Bioresour Technol 102:3169–3175 Carbon sources play a major role in growth of the microorganisms and also in obtaining high biomass. Pres- ent ﬁndings reveal that glycerol as a carbon source (eco- nomical source) and peptone as a nitrogen source promotes the cell growth and also yield, higher biomass which are in accordance with the earlier reports by Meinicke et al. (2012); Sreekumar and Soundarajan (2010). Kumar V, Sahai V, Bisaria VS (2012) Production of amylase and Chlamydospores by Piriformospora indica, a root endophytic fungus. Biocatal Agric Biotechnol 1:124–128 Lowry OH, Rosebrough MJ, Farr AL, Randall RJ (1951) Protein measurement by Folin phenol reagent. J Biol Chem 193:265–275 Luciana FS, Sato HH (2009) Production of alkaline protease from Cellulosimicrobium cellulans. Braz J Microbiol 40(1):54–60 The novelty of present paper lies in the screening of medium components to obtain higher enzyme activity and biomass of P. indica, which has tremendous applications in the ﬁeld of plant biotechnology as a biological hardening tool, biofertilizer and biocontrol agent. Meinicke RM, Vendruscolo F, Moritz DE, de Oliveira D, Schmidell W, Samohy RW, Ninow JL (2012) Potential use of glycerol as substrate for the production of red pigments by Monascus ruber in submerged fermentation. Biocatal Agric Biotechnol 1:238–242 Miller GL (1959) Use of dinitrosalicylic acid reagent for determina- tion of reducing sugar. Discussion Kumar et al. (2012) have reported that the increase in concentration of soluble starch increased the production of 123 444 3 Biotech (2014) 4:439–445 a-amylase by P. indica. Starch is a preferable carbon source over glucose and its concentration has a direct effect on a-amylase production as reported by Gangadharan et al. (2006) and Tanyildizi et al. (2006). Babu and Satyanara- yana (1993) and Narang and Satyanarayana (2001) repor- ted that the organic nitrogen sources produce high a- amylase yield compared to inorganic nitrogen sources. This is in accordance with our results that describe positive impact of soluble starch and negative impact of inorganic nitrogen sources on a-amylase activity. Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, dis- tribution, and reproduction in any medium, provided the original author(s) and the source are credited. References Anal Chem 31:426–428 Narang S, Satyanarayana T (2001) Thermostable a-amylase produc- tion by an extreme thermophile Bacillus thermooleovorans. Lett Appl Microbiol 32:31–35 123 Waller F, Mukherjee K, Deshmukh SD, Achatz B, Sharma M, Scha¨fer P, Kogel KH (2008) Systemic and local modulation of plant responses by Piriformospora indica and related Sebacinales species. J Plant Physiol 165:60–70 Tanyildizi MS, Elibol M, Ozer D (2006) Optimization of growth medium for the production of a´-amylase from Bacillus amylo- liquefaciens using response surface methodology. J Chem Technol Biotechnol 81:618–622 Varma S, Varma A, Rexer KH, Hassel A, Kost G, Sarabhoy A, Bisen P, Beutehorn B, Franken P (1998) Piriformospora indica, gen. et sp. nov., a new root-colonizing fungus. Mycologia 90:896–903 123 Conclusion Rai M, Acharya D, Singh A, Varma A (2001) Positive growth responses of the medicinal plants Spilanthes calva and Withania somnifera to inoculation by Piriformospora indica in a ﬁeld trial. Mycorrhiza 11:123–128 The Plackett–Burman design based on the ﬁrst order model to screen and evaluate the important medium components greatly inﬂuenced the production of a-amylase, protease and also biomass. a-Amylase, protease and biomass were found to vary from 0.25 to 0.45 mg-1 ml-1 min-1, 0.1 to 0.15 mg-1 ml-1 h-1 and 0.8 to 22.6 g l-1, respectively, in 16 runs which showed the strong inﬂuence of the medium components. Soluble starch, ammonium chloride, TE solution and glycerol were found to be signiﬁcant for a- amylase activity. Sodium nitrate, ammonium chloride, maltose and salt solution were found to be signiﬁcant for protease activity. Peptone and glycerol supported well the production of high biomass. Rajendran A, Thirugnanam M, Thangavelu V (2007) Statistical evaluation of medium components by Plackett–Burman exper- imental design and kinetic modeling of lipase production by Pseudomonas ﬂuorescens. Indian J Biotechnol 6:469–478 ﬂ Rodarte MP, Dias DR, Vilela DM, Schwan RF (2011) Proteolytic activities of bacteria, yeasts and ﬁlamentous fungi isolated from coffee fruit (Coffea arabica L). Acta Sci Agron 33(3):457–464 Saxena R, Singh R (2010) Statistical optimization of conditions for protease production from Bacillus sp. Acta Biol Szeged 54(2):135–141 Rodarte MP, Dias DR, Vilela DM, Schwan RF (2011) Proteolytic activities of bacteria, yeasts and ﬁlamentous fungi isolated from coffee fruit (Coffea arabica L). Acta Sci Agron 33(3):457–464 Saxena R, Singh R (2010) Statistical optimization of conditions for protease production from Bacillus sp. Acta Biol Szeged 54(2):135–141 Seraman S, Rajendran A, Thangavelu V (2010) Statistical optimiza- tion of anticholesterolemic drug lovastatin production by the red mold Monascus purpureus. Food Bioprod Process 88:266–276 Sreekumar G, Soundarajan K (2010) Isolation and characterization of probiotic Bacillus subtilis SK09 from dairy efﬂuent. Indian J Sci Technol 3:863–866 Acknowledgments The authors wish to acknowledge Dean of Research and Jain University, Bangalore, for carrying out this work successfully. Sumrin A, Ahmad W, Ijaz B, Sarwar MT, Gull S, Kausar H, Shahid I, Jahan S, Asad S, Hussain M, Riazuddin S (2011) Puriﬁcation and medium optimization of a-amylase from Bacillus subtilis 168. Afr J Biotechnol 10(11):2119–2129 Conﬂict of interest The authors declare that they have no conﬂict of interest. 123 123 3 Biotech (2014) 4:439–445 445
https://openalex.org/W2786667712	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0205608&type=printable	English	null	A novel core promoter element induces bidirectional transcription in CpG island	bioRxiv (Cold Spring Harbor Laboratory)	2,017	public-domain	14,651	RESEARCH ARTICLE OPEN ACCESS Citation: Mahpour A, Scruggs BS, Smiraglia D, Ouchi T, Gelman IH (2018) A methyl-sensitive element induces bidirectional transcription in TATA-less CpG island-associated promoters. PLoS ONE 13(10): e0205608. https://doi.org/10.1371/ journal.pone.0205608 Editor: Tamar Juven-Gershon, Bar-Ilan University, ISRAEL Copyright: This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication. A methyl-sensitive element induces bidirectional transcription in TATA-less CpG island-associated promoters Amin MahpourID1¤, Benjamin S. Scruggs2, Dominic Smiraglia1, Toru Ouchi1, Irwin H. GelmanID1 min MahpourID1¤, Benjamin S. Scruggs2, Dominic Smiraglia1, Toru Ouchi1, Irwin G l 1 1 Department of Cancer Genetics, Roswell Park Comprehensive Cancer Center, Buffalo, NY, 14263, United States of America, 2 Epigenetics and Stem Cell Biology Laboratory, National Institute of Environmental Health Sciences, Research Triangle Park, NC, 27709, United States of America a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 ¤ Current address: Gastrointestinal Unit, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, 02114, United States of America * amahpour@mgh.harvard.edu (AM); irwin.gelman@roswellpark.org (IHG) Abstract How TATA-less promoters such as those within CpG islands (CGI) control gene expression is still a subject of active research. Here, we have identified the “CGCG element”, a ten- base pair motif with a consensus sequence of TCTCGCGAGA present in a group of pro- moter-associated CGI-enriched in ribosomal protein and housekeeping genes. This element is evolutionarily conserved in vertebrates, found in DNase-accessible regions and employs RNA Pol II to activate gene expression. Through analysis of capped-nascent transcripts and supporting evidence from reporter assays, we demonstrate that this element activates bidi- rectional transcription through divergent start sites. Methylation of this element abrogates the associated promoter activity. When coincident with a TATA-box, directional transcription remains CGCG-dependent. Because the CGCG element is sufficient to drive transcription, we propose that its unmethylated form functions as a heretofore undescribed promoter ele- ment of a group of TATA-less CGI-associated promoters. The CGCG element induces bidirectional transcription The best-known DNA element that mediates the initiation of transcription of protein-cod- ing genes is the TATA box with the consensus sequence TATAA [7]. This element is usually located 25 to 34 base pairs upstream of transcription start sites (TSS). However, many human promoters, including those regulating housekeeping genes lack this DNA element [8], suggest- ing that TATA-less promoters are controlled by different yet poorly understood mechanisms. A few novel elements have been described that presumably function as core promoter ele- ments in TATA-less promoters [9–12]. Yet, most of these promoter elements (e.g. GC-box or Inr motif) require additional transcriptional activator binding sites in order to drive efficient directional transcription. Vertebrate genomes contain short GC-rich sequences that are typically less than 1 kb long, termed CpG islands (CGIs) [13, 14]. These regions are considered to be critical for transcriptional regulation of a large group of genes that include housekeeping genes [15]. Most CGI-associated promoters lack a TATA-box yet contain “GC-box” binding sites for the general transcription factor SP1 although these GC boxes are not sufficient to induce tran- scription on their own [15–18]. CGI-associated promoters typically induce bidirectional transcription that produces coding and non-coding transcripts [19, 20]. Thus, depending on the stability of the non-coding RNA, CGI-associated promoters can generate more stable long non-coding RNAs (lncRNA) or short-lived transcripts [21]. To date, no specific inde- pendently-acting promoter element governing these CGI-associated bidirectional promoters has been described. In this study, we analyzed DNase-accessible CGIs in the K562 cell line and found an enriched motif with the consensus sequence of TCTCGCGAGA, which we termed the “CGCG element” due to the characteristic invariable core sequence. This element confers transcriptional activity independent of other transcriptional activator sequences. Promoter sequences related to the CGCG element have been reported previously for several individual genes, but their functional significance was never explored [22–25]. A genome-wide computa- tional study identified a similar motif among those most enriched in human promoters, sug- gesting a possible functional role [26]. Our data indicate that the CGCG element is enriched in TATA-less CGI-associated promoters and evolutionarily conserved among vertebrates. Importantly, it is associated with bidirectional transcription only in the context of CGI-associ- ated promoters as assessed by analysis of GRO-Cap and Start-seq datasets that identify sense versus anti-sense TSS-associated nascent transcripts. Using novel reporter constructs, we dem- onstrate that the CGCG element suffices as a promoter element to drive bidirectional tran- scription. Gene Ontology analysis indicates that this element is enriched in the promoters of housekeeping genes, most notably those controlling RNA metabolism and translation, and of long non-coding RNAs. Together, our results indicate that the CGCG element functions as a previously unknown driver of CGI-associated TATA-less promoters. Introduction Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Gene expression is one of the most critical, yet enigmatic, biological processes that defines cel- lular and organismal identity, and that mediates cellular response to internal and external sti- muli [1]. Importantly, dysregulation of this process is known to contribute to various human diseases such as cancer [2]. With the discovery of RNA polymerases, the mechanisms of how transcription occurs have been extensively studied in many organisms [3]. In contrast to the relatively simple prokaryotic transcriptional system, metazoan transcription is considerably more elaborate and involves complicated promoter structures, multiple functional DNA ele- ments and a repertoire of specific general transcription factors. These factors and DNA ele- ments are required to facilitate accurate transcriptional initiation, elongation, and termination [4–6]. Funding: This work was supported by Roswell Park Comprehensive Cancer Center and National Cancer Institute (NCI) grant P30CA016056 involving the use of the Genomics Shared Resource. The work was also funded by institutional funds provided to TO. Competing interests: The authors have declared that no competing interests exist. 1 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 Reporter constructions and assays One to three copies of the CGCG elements from the DENR promoter were synthesized as dou- ble stranded oligonucleotides (IDT DNA) and cloned into the BglII and MluI restriction sites of a luciferase reporter construct that lacks promoter sequences (pGL2-basic, Promega). 1 μg of cloned reporter DNA along with 100 ng of a Renilla luciferase reporter construct (pRL-TK) as transfection control were transfected into HEK293T using X-tremeGENE 9 (Roche) reagent according to manufacturer’s protocol. The luciferase activities were measured 24h after trans- fection according to the Dual Luciferase assay protocol (Promega). The bidirectional luciferase (Empty-LuBiDi) reporter was constructed by PCR amplifica- tion and subsequent cloning of the firefly luciferase gene from pGL2-Basic into the BglII site of the promoterless Renilla luciferase cassette from the pRL-Null plasmid followed by site- directed mutagenesis to remove secondary the BglII recognition site downstream of firefly luciferase poly-A site. The primer sequences used are described in S1 File. Bioluminescence assays were performed as described above except that transfection was normalized using a co-transfection vector that expresses secretory alkaline phosphatase (pSELECT-zeo-SEAP, Invivogene). For the construction of the bidirectional fluorescence reporter, pmCGFP, we first PCR amplified and cloned the h2b-mCherry fused gene (plasmid Addgene id #20972) head-to- head into a promoterless eGFP containing construct. The resulting construct (eGFP + h2b- mCherry) was then digested with AgeI to release h2b-coding fragment and self-ligated to generate the pmCGFP (eGFP + mCherry). Double stranded oligonucleotides encoding one or three copies of TCTCGCGAGA were spliced into the AgeI restriction site of this reporter. For CpG-free reporter and methylation experiments, an oligonucleotide encoding a single copy of TCTCGCGAGA was inserted into the HindIII restriction site of pCpGfree-basic- Lucia (Invivogen). 10 μg of purified plasmid was incubated at 37˚C for 8h with 10 enzymatic units (U) of M.SssI methyltransferase (NEB) supplemented with fresh 100 μM S-adenosyl methionine (SAM) as the methyl donor. DNA was extracted using phenol-chloroform fol- lowed by ethanol precipitation. The DNA was subjected to another an additional 8h incuba- tion with M.SssI was followed by DNA extraction. As a control, a mock reaction was also carried out lacking M.SssI enzyme. To test the methylation efficiency, 300ng of DNA was digested with 10 U of NheI and BstUI for 30 min at 37˚C. Because CGCG methylation blocks BstUI cleavage, empty vector or and methylated constructs digested only by NheI enzyme pro- ducing two indistinguishable bands at 2.4 kb. The CGCG element induces bidirectional transcription Reporter constructions and assays However, unmethylated TCTCGCGAGA which is cut by BstUI enzyme as well as NheI produced three smaller bands. The sequences of inserts for each promoter fragment and related mutations are provided in the S1 File. Cell culture and treatments Human embryonic kidney 293T and mouse NMuMG cells were cultured in Dulbecco’s Modi- fied Eagle Medium (DMEM) media supplemented with 10% fetal bovine serum, penicillin and streptomycin at 37˚C and 5% CO2. For the α-amanitin treatment experiment, HEK293T cells were transfected with SV40 promoter-driven firefly luciferase reporter (pGL2-pro), or a con- struct containing a copy of TCTCGCGAGA. 24h post-transfection, cells were treated with 5 μg/ml α-amanitin (Santa Cruz Biotechnologies) as described [27] or with PBS (control), and firefly and Renilla luciferase bioluminescence activities were measured 24h after treatment. PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 2 / 25 Chromatin immunoprecipitation (ChIP) ChIP was performed according to a protocol described in Lee et al. [28]. In brief, 10 million HEK293T cells were cultured in 15 cm dishes and transfected with 10 μg reporter DNA using X-tremeGENE 9 transfection reagent. 48h post-transfection, cells were treated with the cross- linking reagent formaldehyde (1% in PBS, Sigma) for 5min. Cross-linking was stopped by add- ing glycine solution (0.125 M) for 10min, followed by 2 washes with ice-cold PBS. The cells were lysed in buffer A (50 mM HEPES-KOH, pH 7.5, 140 mM NaCl, 1 mM EDTA, 10% glyc- erol, 0.5% NP-40, 0.25% Triton X-100, 1X protease inhibitors [Roche]), and their nuclei were isolated by centrifugation (5 min, 1000 RPM) and lysed by sonication in buffer B (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% Na-deoxycholate, 0.5% N-lauroylsarcosine, 1X protease inhibitors) using a Biorupter(Diagenode), specifically, five rounds of 10sec sonication to obtain 300-600bp range chromatin fragments. The resulting sheared chromatin was immunoprecipitated (IP) using 20 μg of antibody against Pol II (Santa Cruz, N-10) or a non-specific Isotype Mouse IgG as a mock control (Santa Cruz). The IP com- plexes were then bound to Protein A/G Dynabeads and washed five times using RIPA washing buffer (50 mM HEPES-KOH, pKa 7.55, 500 mM LiCl, 1 mM EDTA, 1.0% NP-40, 0.7% Na- deoxycholate). The DNA was extracted from the beads using buffer C (50 mM Tris-HCl, pH 8.0, 10 mM EDTA, 1.0% SDS) and quantified by qPCR using primers designed to amplify the promoter region of the reporter construct. qRT-PCR HEK 293T cells were transfected with 1 μg of LuBiDi reporter DNAs containing 0, 1, 2, 4 cop- ies of TCTCGCGATA. Cells were lysed after 24h using TRIzol (Life Technologies), RNA was extracted using chloroform-isopropanol, and after resuspension, incubated with RNase-free DNase I (Thermo Fischer Scientific) followed by extraction by acidic-phenol chloroform pro- tocol, precipitation using ethanol and resuspension in RNase-free water. 1 μg of the resulting purified RNA was used to prepare cDNA using M-MLV reverse transcriptase according to manufacturer’s recommended protocol (Life Technologies). Transcript levels were measured using iTaq Universal SYBR Green Supermix (Bio-Rad) on an ABI-7900 RT-PCR instrument. Transcript levels were normalized using primers for HPRT1. Primers designed to amplify PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 3 / 25 The CGCG element induces bidirectional transcription the bacterially-expressed ampicillin resistance gene in the LuBiDi construct were used as a negative control to rule out plasmid contamination. Melting curve analyses for all PCR experi- ments were performed to validate faithful amplification of PCR products. Information on primer sequences is described in S1 File. Fluorescent microscopy and live imaging 1 μg of pmCGFP reporter constructs containing 0, 1, 2 copies of TCTCGCGAGA along with a CMV promoter-driven Blue Fluorescent Protein expression plasmid (CMV-BFP) were trans- fected into HEK293T cells. Images were taken 24h post-transfection using a Nikon Eclipse TE2000-E fluorescence microscope. For live imaging, images were taken every 30min with an exposure time of 1sec immediately after reporter transfection for 24h in an incubating cham- ber supplied with humidity and 5% CO2. 16-bit Tiff images from individual channels were used to generate MOV files using Videomach software (http://gromada.com/videomach/). The final video was produced using Adobe Premiere CC 2017. Oligonucleotide pull-down assay Kaiso binding to CGCG elements was assessed using biotin-tagged DNA duplexes of unmodi- fied TCTCGCGAGA, TCTCTCGAGA or completely methylated (TCTmeCGmeCGAGA). 10 μM aliquots of each duplex was bound to 100 μL streptavidin Dynabeads as recommended by the manufacturer (Invitrogen). HEK293T cells were lysed using NET-N buffer containing pro- tease inhibitors cocktail (Sigma) and incubated on ice for 30 min. Lysates were centrifuged at 12000 RPM for 10 min to pellet cellular debris, and supernatant representing 500 μg protein was mixed with duplex-charged beads and incubated at 4˚C overnight. The beads were washed five times with NET-N buffer, incubated with 50 μL Laemmli loading buffer (1X: 0.02% w/v bromophenol blue, 4% SDS, 20% glycerol, 120 mM Tris-HCl, pH 6.8) and boiled for 5 min to elute bound proteins. The proteins were analyzed by immunoblotting for Kaiso (Santa Cruz #D-10). Double nickase Cas9-mediated genome editing of DENR promoter Short guide RNAs (sgRNAs) to target DENR promoter were designed using the MIT CRISPR sgRNA design tool (http://crispr.mit.edu/). Double-stranded oligonucleotides representing sgRNAs (S1 File) were then cloned into pSpCas9 (BB)-2A-GFP (PX458) and pSpCas9 (BB)- 2A-Puro (PX459) V2.0 (Addgene plasmids 48138 and 62988). Constructs were then co-trans- fected into HEK293T cells and 24h later selected for puromycin resistance (3 μg/mL) for another 72h. GFP-expressing single cells were sorted using an Aria II FACS and incubated in 96 well dishes for two weeks to form visible cellular clones. DNA was extracted from the clones using QuickExtract solution (Epibio), and successful deletions were confirmed by Sanger sequencing of PCR products. Ribbon sequences were produced using the pyRibbon software which we deposited in https://github.com/AminMahpour/pyRibbon/. 4 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription Lentivirus mediated knockdown of ZBED1 and immunoblotting For ZBED1 knockdown, lentiviruses targeting ZBED1 or non-targeting (pLKO.1) were pur- chased from Sigma (MISSION shRNA). Cells were transduced with the virus according to the manufacturer’s protocol (Sigma) and transduced cells were selected for puromycin at 3μg/mL for four days. For immunoblots, cells were lysed in NET-N buffer (100 mM NaCl, 20 mM Tris-HCl pH 8.0, 0.5 mM EDTA, 0.5% NP-40) supplemented with protease inhibitors cocktail at 4˚C. In all experiments, 20 μg of total proteins/lane were analyzed by SDS-PAGE followed by blotting as described Previs et al. [29]. Antibodies included those specific for DENR (Santa Cruz #22), ZBED1 (Santa Cruz H-9) GFP (Santa Cruz B-2), mCherry (Abcam 1C51) or alpha-tubulin (Santa Cruz A-6) as a loading control. Rapid amplification of cDNA ends (50-RACE) To determine divergent TSSs, we transfected near confluent HEK293T cells in 10 cm dishes with 5 μg LuBiDi construct along with 0.5 μg pEGFP-C1 as a transfection control. RNA was extracted 72h after transfection, and the quality and purity evaluated using an Agilent 2100 Bioanalyzer, with samples with RNA integrity number (RIN) values >= 8.0 selected for further analysis. The SMARTer 50-end RACE (Clontech) protocol was used to determine divergent TSSs from 10 μg of total RNA. Briefly, the RNA was first reverse-transcribed at 42 C for 90 min using poly-dT primers and extended beyond TSS using RT-mediated template switching that employs the SMARTer IIA oligonucleotide only when the 50-cap is encountered. The resulting cDNA products were amplified using specific internal primers for either firefly or Renilla luciferases plus the Clontech Universal Primer Mix (UPM). A GFP primer set was used as an internal control (S1 File). The PCR products containing TSS were directionally cloned into the linearized pRACE vector using the In-fusion HD system, and individual bacterial clones were obtained following transformation of the ligated products into Stellar competent cells. Sanger sequencing of the resulting plasmid clones (using M13 primer) was used to iden- tify TSSs. ENCODE conservation, DNase-seq, GRO-Cap, WGBS data analysis Processed data points for hg38 were extracted and processed using Wigman software for 50 bp upstream and downstream windows for each motif occurrence. For ENCODE WGBS (acces- sion number ENCFF867JRG). The PhyloP and PhastCons conservation scores for hg38 assem- bly were downloaded from the UCSC genome browser (http://hgdownload.cse.ucsc.edu/ downloads.html). ENCODE accession number ENCFF867JRG was used for K562 DNase-seq data. The GRO-Cap dataset for K562 and GM12878 cell lines with GEO accession number of GSM1480321 was used to analyze nascent transcripts in promoters. Pol II ChIP-seq from K562 cell line with the accession number of ENCFF000YWS was used to determine Pol II occupancy state on CGCG elements. Heatmap plots were generated using the in-house written Wigman software (https://github.com/AminMahpour/Wigman). Start-seq analysis Start-seq datasets from mouse bone-marrow derived macrophages were published previously and is available for download from the GEO website (GSE62151, https://www.ncbi.nlm.nih. gov/geo/). Data were analyzed as follows: briefly, reads were aligned uniquely to the mm9 genome allowing a maximum of two mismatches with Bowtie version 0.12.8 (-m1 -v2). Sense and divergent TSS were assigned as defined above. Start-seq heat maps depict Start-RNA reads in 10 bp bins at the indicated distances with respect to the TSS. Heatmap plots were generated using Partek Genomics Suite version 6.12.1012. Individual CGCG element occurrences were identified with FIMO [33]. A ±1 kbp window around TSSs was scanned with a position weight matrix for the CGCG motif with a p-value threshold of 0.001. Motif occurrences were mapped with respect to TSS locations using custom scripts and counted in 10-mer bins. Composite Metagene distributions were generated by summing motifs at each indicated position with respect to the TSS and dividing by the number of TSSs included within each group. Genomic annotation and metagene analysis The scanMotifgenomewide script from the Homer program version 4.8 was used to locate all instances of motif 7 and 10 in human (hg38) and mouse (mm9) genomes. The annotatePeaks script (Homer) was used to identify motif co-occurrence, genomic annotations, metagene, and enrichment analysis. Motif discovery The CpG island annotation track in the human genome (hg38) was downloaded from the UCSC genome browser (https://genome.ucsc.edu/), and sequences that overlap with K562 DNase-seq peak track extracted using Bedtools [30]. The resulting sequences were used for 5 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription motif discovery using the findMotifgenomewide script in the Homer bioinformatics software suite using default command line arguments for the human genome [31]. CAGE-seq analysis of cells treated with DAC/DMSO CAGE-seq data derived from DAC/DMSO treated cells was published by brocks et al. [32]. The raw sequencing data (GSE81322) were aligned using the Bowtie2 software on the hg38 reference genome. Average mapped CAGE-seq reads were plotted ± 1kb relative to CGCG elements. Statistical analysis All plots were generated and analyzed using GraphPad Prism version 7. Unless noted other- wise, all statistical analyses were performed using Student t-test. The following p-values are presented as : p < 0.05, : p < 0.01, : p < 0.001, : p < 0.0001. Error bars represent standard deviation (S.D) from the mean. Gene ontology and gene network analysis Bedtools Closest feature was used to compile a list of genes with annotated TSS less than 500 bp from CGCG elements on both plus and minus strands from the latest hg38 GTF annotation file (http://www.ensembl.org/info/data/ftp/index.html). A custom script was written and used to determine the number of CGCG elements in annotated coding, non-coding, uni- and bi- 6 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription directional CGI promoters. Gene Ontology (GO) analysis performed using the GOrilla gene enrichment analysis platform. A list of CpG islands-associated genes was used as the back- ground genes for enrichment analysis [34]. GO enrichment score is defined as (b/n)/(B/N) where N is the total number of background CpG island-associated genes that have a GO term, B is the number of genes associated with a specified GO term, n is the number of genes whose promoter contain CGCG element and b is the number of genes in the intersection. Gene set interaction networks were generated and analyzed using REACTOME package v53 (http:// www.reactome.org/). Networks were visualized graphically using Cytoscape software version 3.5 (http://www.cytoscape.org/). Motif discovery in DNase-sensitive CpG islands Previous analysis suggested that roughly 50 percent of human promoters are associated with a CGI [35]. To identify novel CGI-associated, independently-functioning promoter elements that potentially drive transcription in human CGIs (30k), we extracted CGI sequences that overlapped with DNase-accessible regions (192k DNase-seq peaks) in the K562 cell line. We then performed an unbiased motif discovery to identify top enriched motifs in transcription- ally active CGI-associated promoters (Fig 1A and 1B). As expected, the SP1 binding site (GC box) had the highest enrichment score consistent with its purported role in driving TATA-less promoters. Binding sites for NRF and ETS were also identified, consistent with roles for these transcription factors in the regulation of CGI-associated housekeeping genes [36]. We also identified two novel sequence motifs (numbered 7 and 10) that were highly conserved within vertebrates (Fig 1C). There were more than 400 incidences of motif #10 that coincided with DNase-seq footprints in multiple cell lines (K562 is shown), suggesting that this motif repre- sents a shared regulatory element (Fig 1C and A in S1 Fig). Although most CGI-associated promoters contain one copy of the motifs shown in Fig 1E, motifs 7 and 10 tend to occur in multiple copies in a given promoter (Fig 1D and 1E). Genome ontology and metagene profile analyses showed that motifs 7 and 10 are enriched significantly in annotated human CGI-containing promoters, with motif 10 being far more enriched in promoters of annotated coding and non-coding genes despite being less frequent (Fig 1B; motif 7 = 1408 copies vs. motif 10 = 413 copies) (Fig 1F and 1G). CGCG elements recruit transcriptional machinery and activate gene expression To determine whether motif 7 and 10 could confer transcriptional activity independently, we cloned the sequence of the most common variant of each motif (ACTACAATTCCC and TCTCGCGAGA, respectively) into a promoterless construct that encodes a firefly luciferase reporter gene (Empty pGL2-basic). The resulting constructs were then separately cotransfected along with a control reporter for Renilla luciferase driven by the HSV-1 thymidine kinase pro- moter (pRL-TK) into human embryonic kidney (HEK293T) cells. Motif 10, but not Motif 7, significantly activated firefly reporter gene expression (Fig 2A). This result encouraged us to 7 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription Fig 1. Identification of enriched motifs in human CGIs. A) The computational pipeline used to identify the most enriched motifs in the DNase-accessible CGIs from K562 cells. Bedtools was used to identify CGIs that overlap with ENCODE derived DNase-sensitive peaks. Homer was then used to identify the most enriched DNA sequence motifs in the DNase-accessible CGI B) K i i f bi di i i h i h d if li C) H h i b i Fig 1. Identification of enriched motifs in human CGIs. A) The computational pipeline used to identify the most enriched motifs in the DNase-accessible CGIs from K562 cells. Bedtools was used to identify CGIs that overlap with ENCODE derived DNase-sensitive peaks. Homer was then used to identify the most enriched DNA sequence motifs in the DNase-accessible CGIs. B) Known transcription factor binding sites in the enriched motif list. C) Heatmaps showing vertebrate conservation and DNase-seq profiles for motifs #7 and #10 including their flanking 50 bp. PhyloP scores represent conservation of individual base-pairs throughout vertebrate genomes. Positive scores in PhyloP heatmaps, shown by red colors, indicate high sequence conservation whereas negative scores (blue colors) indicate acceleration of base pairs. PhastCons heatmaps show probability scores (ranging from 0 to 1) of conserved DNA elements. DNase-seq heatmaps of K562 and HeLa cells indicate DNase accessibility. Low accessibility, as indicated in dark colors, identifies a central DNase-seq footprint associated with motif #10 in both cell lines. Both motifs occur in DNase-sensitive CGIs of many different cell lines, but only motif #10 was consistently associated with a DNase-seq footprint (Fig A in S1 Fig). D) Motif co-occurrence odds-ratio matrix in DNase- sensitive CGIs. The odds-ratio is the value of observed-to-expected coincidence if motifs were distributed by chance. PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 https://doi.org/10.1371/journal.pone.0205608.g001 focus on motif 10, which we named the “CGCG element” based on its central motif. A genome-wide analysis found that this element maps within 50bp of annotated TSSs in human and mouse genomes (Fig B in S1 Fig) suggesting that this element could potentially function to regulate gene expression [37]. To address the function of a specific naturally-occurring CGCG element, we analyzed the CGI-containing promoter of the human Density Regulated gene (DENR). The DENR promoter contains three tandem CGCG elements separated by 21 and 11 nucleotides (Fig 2B). To determine the role of each CGCG element in this promoter, we inserted promoter fragments containing CGCG #1, CGCG #1,2 and CGCG #1,2,3 into the promoter-less pGL2-basic construct. Although a single copy of the CGCG element signifi- cantly increased reporter activity, there was a 7- and 17-fold increase in reporter activity with the addition of the second and third CGCG elements, respectively. Introducing G to T muta- tions in all CGCG elements (CTCG #1,2,3) dramatically decreased promoter activity, suggest- ing that the CGCG element is necessary and sufficient to drive reporter expression and that there is a cooperativity between multiple CGCG elements (Fig 2C). To determine if CGCG element-driven gene expression is dependent on RNA polymerase II (Pol II), we transfected HEK293T cells with reporter constructs that contain either the con- sensus motif (TCTCGCGAGA) or a G!T transversion mutation (TCTCTCGAGA) and per- formed chromatin immunoprecipitation (ChIP) for RNA Pol II [38]. As shown in Fig 2D, Pol II bound the wild-type (WT) CGCG but not to the mutant CTCG site. Analysis of the Pol II ChIP-seq ENCODE dataset in HEK293T cells identified binding peaks coincident with the DENR promoter containing CGCG elements (Fig 2B). α-amanitin, a Pol II inhibitor [39], decreased CGCG element-driven reporter expression (Fig 2E), suggesting that Pol II is indis- pensable for CGCG dependent gene expression. To assess the effect of removing CGCG elements on the endogenous DENR promoter activ- ity, we employed a CRISPR/Cas9 double-nickase strategy [40] to delete a small CGCG-con- taining DENR region in the HEK293T cell line. One cell clone, containing a deletion of approximately 200 base pairs (bp) removed all three CGCG elements in one allele, and a sepa- rate 100bp deletion removed one of the CGCG elements in the other allele without affecting the remaining CGI in the promoter (Fig 2F). https://doi.org/10.1371/journal.pone.0205608.g001 Removal of these CGCG-containing regions resulted in a significant decrease in DENR transcript and protein levels compared to those in WT controls (Fig 2G). Together with the reporter analyses, these findings suggest that CGCG elements actively recruit transcriptional machinery and promote gene expression in the CGI- associated promoter of DENR gene. CGCG elements recruit transcriptional machinery and activate gene expression Higher values indicate a higher likelihood of co-occurrence of indicated motifs in the matrix. E) Pie charts indicate frequency of the motif copy number in human CGIs. F) Metagene profiles for all human CGIs, motifs #7 and #10 were generated by the Homer annotated script, showing that both motifs are associated with TSSs. G) Homer annotated script was used to generate Fig 1. Identification of enriched motifs in human CGIs. A) The computational pipeline used to identify the most enriched motifs in the DNase-accessible CGIs from K562 cells. Bedtools was used to identify CGIs that overlap with ENCODE derived DNase-sensitive peaks. Homer was then used to identify the most enriched DNA sequence motifs in the DNase-accessible CGIs. B) Known transcription factor binding sites in the enriched motif list. C) Heatmaps showing vertebrate conservation and DNase-seq profiles for motifs #7 and #10 including their flanking 50 bp. PhyloP scores represent conservation of individual base-pairs throughout vertebrate genomes. Positive scores in PhyloP heatmaps, shown by red colors, indicate high sequence conservation whereas negative scores (blue colors) indicate acceleration of base pairs. PhastCons heatmaps show probability scores (ranging from 0 to 1) of conserved DNA elements. DNase-seq heatmaps of K562 and HeLa cells indicate DNase accessibility. Low accessibility, as indicated in dark colors, identifies a central DNase-seq footprint associated with motif #10 in both cell lines. Both motifs occur in DNase-sensitive CGIs of many different cell lines, but only motif #10 was consistently associated with a DNase-seq footprint (Fig A in S1 Fig). D) Motif co-occurrence odds-ratio matrix in DNase- sensitive CGIs. The odds-ratio is the value of observed-to-expected coincidence if motifs were distributed by chance. Higher values indicate a higher likelihood of co-occurrence of indicated motifs in the matrix. E) Pie charts indicate frequency of the motif copy number in human CGIs. F) Metagene profiles for all human CGIs, motifs #7 and #10 were generated by the Homer annotated script, showing that both motifs are associated with TSSs. G) Homer annotated script was used to generate PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 8 / 25 The CGCG element induces bidirectional transcription genomic annotation enrichment scores to identify genomic regions where these motifs are enriched. The data indicate that motif #7 and #10 are significantly enriched in promoters and CGIs. The enrichment values were calculated using the cumulative hypergeometric distribution method. CGCG elements recruit transcriptional machinery and activate gene expression genomic annotation enrichment scores to identify genomic regions where these motifs are enriched. The data indicate that motif #7 and #10 are significantly enriched in promoters and CGIs. The enrichment values were calculated using the cumulative hypergeometric distribution method. https://doi.org/10.1371/journal.pone.0205608.g001 CGCG element confers bidirectional transcription activity in bidirectional reporter constructs Due to the palindromic nature of the TCTCGCGAGA motif, we wondered whether the CGCG elements could also activate bidirectional transcription. To test this, we developed a novel bidirectional reporter construct (LuBiDi) to measure promoter activity using firefly and Renilla luciferase genes as reporters of directional transcription from a central control motif (Fig 3A). We inserted one or two copies of the TCTCGCGAGA motif into the LuBiDi plasmid and measured both reporter activities. A single CGCG element was sufficient to induce both firefly and Renilla reporters whereas two CGCG elements induced an additional 4-fold increase 9 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 Fig 2. CGCG elements recruit RNA polymerase 2 and activate reporter expression. A) Firefly reporter activity driven by motif 7 and 10. B) The structure and sequence of the human DENR promoter and promoter fragments used for reporter studies. Note that the sequence selected from this promoter do not contain a known core promoter element. The DENR promoter contains three highly conserved copies of the CGCG element. ENCODE Pol II ChIP-seq data from HEK293T cells (bottom) demarcates Pol II occupancy in the promoter region. C) Reporter activity of the corresponding DENR fragments as described in section b. D) Pol II-ChIP on chromatin from HEK293T cells transfected with reporter constructs driven by the wild-type (TCTCGCGAGA) or mutant (TCTCTCGAGA) CGCG motifs, using the human GAPDH promoter as a positive control. E) The effect of α-amanitin on TCTCGCGAGA-driven firefly reporter expression. α-amanitin significantly reduced both SV40 and CGCG-driven reporter expression. F) CRISPR/Cas9 double- nickase strategy was employed to target CGCG elements in the endogenous DENR promoter. The agarose gel image on the left side of this panel shows genomic PCR amplification products of the DENR promoter in parental and CRISPR-edited cells. Two shorter PCR products associated with CRISPR-edited clone indicate deletions in different alleles. Ribbon plots reveal Sanger sequences of parental and edited alleles shown in the gel image. G) qRT-PCR (“Transcript”) and immunoblot (“Protein”) analyses of DENR expression from the parental and edited cell lines, with the data from three replicate qRT-PCR experiments quantified at right (+/- SD). https://doi.org/10.1371/journal.pone.0205608.g002 The CGCG element induces bidirectional transcription The CGCG element induces bidirectional transcription Fig 2. CGCG elements recruit RNA polymerase 2 and activate reporter expression. A) Firefly reporter activity driven by motif 7 and 10. https://doi.org/10.1371/journal.pone.0205608.g002 The CGCG element induces bidirectional transcription Fig 3. CGCG elements promote bidirectional gene expression in the LuBiDi reporter system. A) The structure of the LuBiDi reporter construct. B) One copy of the TCTCGCGAGA motif inserted in the LuBiDi construct was sufficient to activate the expression of both firefly and Renilla reporters. Flank-exchanged (AGACGCGTCT), G!T transversion mutation (TCTCTCGAGA), or an insertion mutation in the middle of CGCG (TCTCGACGAGA) abolished the dual activation. C) The frequency of common CGCG element sequence variants in the human genome. D) The bidirectional promoter activity of selected naturally- occurring CGCG element variants. E) The effect of CGCG element copy number (TATCGCGAGA motif) in LuBiDi on reporter activity. F) Corresponding transcript levels from reporters in panel E. https://doi.org/10.1371/journal.pone.0205608.g003 Fig 3. CGCG elements promote bidirectional gene expression in the LuBiDi reporter system. A) The structure of the LuBiDi reporter construct. B) One copy of the TCTCGCGAGA motif inserted in the LuBiDi construct was sufficient to activate the expression of both firefly and Renilla reporters. Flank-exchanged (AGACGCGTCT), G!T transversion mutation (TCTCTCGAGA), or an insertion mutation in the middle of CGCG (TCTCGACGAGA) abolished the dual activation. C) The frequency of common CGCG element sequence variants in the human genome. D) The bidirectional promoter activity of selected naturally- occurring CGCG element variants. E) The effect of CGCG element copy number (TATCGCGAGA motif) in LuBiDi on reporter activity. F) Corresponding transcript levels from reporters in panel E. https://doi.org/10.1371/journal.pone.0205608.g003 (Fig 3B). To study the motif sequence requirement for this activation, we introduced muta- tions in the motif that disrupted the wild-type sequence in various locations. First, to deter- mine whether the palindromic structure was more important than the sequence content in conferring the bidirectional transcriptional activity, we exchanged the flanking sequences to form AGACGCGTCT, which maintains both symmetry and CpG content. This mutation abrogated the dual activation of reporters (Fig 3B), suggesting that the promoter function of the CGCG element relies on its sequence polarity. A CGCG ! CTCG (underline represents changes) transition mutation (TCTCTCGAGA, reduced CpG content) and an “A” insertion into CGCG (TCTCGACGAGA, unchanged CpG content) abrogated dual reporter activity (Fig 3B). The inclusion of two copies of the A insertion mutant failed to induce transcription. Altogether, these results indicate that the WT element, CGCG core plus the flanking (Fig 3B). CGCG element confers bidirectional transcription activity in bidirectional reporter constructs B) The structure and sequence of the human DENR promoter and promoter fragments used for reporter studies. Note that the sequence selected from this promoter do not contain a known core promoter element. The DENR promoter contains three highly conserved copies of the CGCG element. ENCODE Pol II ChIP-seq data from HEK293T cells (bottom) demarcates Pol II occupancy in the promoter region. C) Reporter activity of the corresponding DENR fragments as described in section b. D) Pol II-ChIP on chromatin from HEK293T cells transfected with reporter constructs driven by the wild-type (TCTCGCGAGA) or mutant (TCTCTCGAGA) CGCG motifs, using the human GAPDH promoter as a positive control. E) The effect of α-amanitin on TCTCGCGAGA-driven firefly reporter expression. α-amanitin significantly reduced both SV40 and CGCG-driven reporter expression. F) CRISPR/Cas9 double- nickase strategy was employed to target CGCG elements in the endogenous DENR promoter. The agarose gel image on the left side of this panel shows genomic PCR amplification products of the DENR promoter in parental and CRISPR-edited cells. Two shorter PCR products associated with CRISPR-edited clone indicate deletions in different alleles. Ribbon plots reveal Sanger sequences of parental and edited alleles shown in the gel image. G) qRT-PCR (“Transcript”) and immunoblot (“Protein”) analyses of DENR expression from the parental and edited cell lines, with the data from three replicate qRT-PCR experiments quantified at right (+/- SD). https://doi.org/10.1371/journal.pone.0205608.g002 10 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription palindromic sequences found in motif 10, are required for promotion of bidirectional tran- scriptional activity. To analyze the expression dynamics associated with CGCG elements in single cells, we developed another promoter-less bidirectional reporter (pmCGFP) that codes for enhanced Green Fluorescent Protein (eGFP) and mCherry reporters in opposite directions (Fig A in S2 Fig). One or three copies of TCTCGCGAGA motifs were inserted into this reporter construct, which were then cotransfected into HEK293T cells along with a CMV promoter construct driving the Blue Fluorescent Protein (BFP) as a transfection control. Cells simultaneously expressed both eGFP and mCherry reporter genes starting 12 hours after transfection only for constructs containing the TCTCGCGAGA element (B in S2 Fig). Immunoblot analysis indi- cated that eGFP and mCherry protein levels were correlated to the number of inserted TCTCGCGAGA motifs (C in S2 Fig). We also tracked individual cells using live imaging microscopy and observed that the two reporter genes are expressed simultaneously after transfection (D in S2 Fig; S1 Video). We also performed a similar imaging experiment using pmCGFP-H2b. This construct is similar to the pmCGFP construct except that the mCherry reporter is fused to the histone H2b protein. Introduction of this reporter construct to the HEK293T and NMuMG mouse mammary cell lines resulted in simultaneous expression of both reporters. Because H2b is a nuclear protein, mCherry signals were also restricted to the nuclei of cells (E and F in S2 Fig). Collectively, these results suggest that this element is a potent bidirectional transcription activator in multiple species. An analysis of human CGI-associated promoters indicated that CGCG elements could also contain less frequent, single nucleotide variations in TCT or AGA flanking sequences (Fig 3C). To determine the impact of these minor variations on bidirectional transcription activity, we compared LuBiDi constructs with one TCTCGCGAGA motif to those containing naturally variant sequences, using the AGA $ TCT flank-exchanged mutant as a negative control (Fig 3D, the variation in a specific nucleotide is underlined). CCT, AGG or ATG flanking sequences decreased relative dual reporter activity whereas variants that contain ATA or TAT showed similar activity to that of the TCTCGCGAGA motif (Fig 3D). The data suggest that some, but not all, variability in the flanking sequences confer promoter activity, albeit at lower efficiencies compared to the TCTCGCGAGA motif. The data also showed that imperfect pal- indrome elements can still drive bidirectional transcription. To study the role of copy number variation on bidirectional transcription activity in more detail, we generated LuBiDi reporters that contain one, two or four copies of TATCGCGAGA, a common variant of the CGCG element with an imperfect palindrome. Reporter activity increased proportionally with the number of motifs as measured by luciferase activity or lucif- erase transcript levels (Fig 3E and 3F). To study the motif sequence requirement for this activation, we introduced muta- tions in the motif that disrupted the wild-type sequence in various locations. First, to deter- mine whether the palindromic structure was more important than the sequence content in conferring the bidirectional transcriptional activity, we exchanged the flanking sequences to form AGACGCGTCT, which maintains both symmetry and CpG content. This mutation abrogated the dual activation of reporters (Fig 3B), suggesting that the promoter function of the CGCG element relies on its sequence polarity. A CGCG ! CTCG (underline represents changes) transition mutation (TCTCTCGAGA, reduced CpG content) and an “A” insertion into CGCG (TCTCGACGAGA, unchanged CpG content) abrogated dual reporter activity (Fig 3B). The inclusion of two copies of the A insertion mutant failed to induce transcription. Altogether, these results indicate that the WT element, CGCG core plus the flanking PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 11 / 25 The CGCG element induces bidirectional transcription Fig 4. CGCG elements are transcriptionally active in CpG islands and methylation abolishes its activity. A) CGCG elements that occur in CGIs in K562 cells coincide with DNase-seq footprints and associate with divergent plus and minus GRO-Cap transcripts and RNA Pol II occupancy. ENCODE WGBS methylation data show increased methylation of CGCG elements in non-CGIs compared to those in CGIs. B) Methylation of CGCG elements suppresses their promoter activity. Methylation of a TCTCGCGAGA insert in a CpG-free Lucia reporter using M.SssI CpG methyltransferase and SAM. C) Methylation of TCTCGCGAGA in the construct assessed by agarose gel analyses after digestion with NheI plus BstUI (only cuts non-methyl CGCG), which produces a doublet band of 2.3Kb. D) Lucia vs SEAP (transfection control) activity driven by WT, mutant or methylated CGCG inserts in transfected HEK293T cells. Data represent the mean of three replicates +/- SD. E) Strand-specific CAGE-seq signals generated from raw CAGE-seq sequence reads described by Brocks et al. [32] from NCI-H1299 cells treated with either the DNMT inhibitor, DAC, or DMSO control. https://doi.org/10.1371/journal.pone.0205608.g004 Fig 4. CGCG elements are transcriptionally active in CpG islands and methylation abolishes its activity. A) CGCG elements that occur in CGIs in K562 cells coincide with DNase-seq footprints and associate with divergent plus and minus GRO-Cap transcripts and RNA Pol II occupancy. ENCODE WGBS methylation dat show increased methylation of CGCG elements in non-CGIs compared to those in CGIs. B) Methylation of CGCG elements suppresses their promoter activity. Methylation of a TCTCGCGAGA insert in a CpG-free Lucia reporter using M.SssI CpG methyltransferase and SAM. C) Methylation of TCTCGCGAGA in the construct assessed by agarose gel analyses after digestion with NheI plus BstUI (only cuts non-methyl CGCG), which produces a doublet band of 2.3Kb. D) Lucia vs SEAP (transfection control) activity driven by WT, mutant or methylated CGCG inserts in transfected HEK293T cells. Data represent the mean of three replicates +/- SD. E) Strand-specific CAGE-seq signals generated from raw CAGE-seq sequence reads described by Brocks et al. [32] from NCI-H1299 cells treated with either the DNMT inhibitor, DAC, or DMSO control. https://doi.org/10.1371/journal.pone.0205608.g004 Fig 4. CGCG elements are transcriptionally active in CpG islands and methylation abolishes its activity. A) CGCG elements that occur in CGIs in K562 cells coincide with DNase-seq footprints and associate with divergent plus and minus GRO-Cap transcripts and RNA Pol II occupancy. Endogenous CGCG elements confer bidirectional transcriptional activity in CGI-associated promoters and methylation abrogates its promoter activity To determine if CGCG elements are associated with the in vivo bidirectional transcription from endogenous promoters, we analyzed a previously published GRO-cap (global run-on sequencing followed by enrichment for 50-cap structure) dataset performed on K562 cells [41]. GRO-cap allows for the detection of nascent, often unstable endogenous strand-specific RNA transcripts that are usually undetectable by common RNA-seq methods, likely because of the greatly increased sequencing depth near the transcriptionaly active regions. We found that the bidirectional transcription is associated in the immediate vicinity of the CGCG elements found in CGI-enriched promoters (Fig 4A). Further analysis showed that the endogenous PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 12 / 25 The CGCG element induces bidirectional transcription Table 1. Gene ontology analysis of genes whose promoter contain at least one CGCG element. Term p-value FDR Enrichment nuclear-transcribed mRNA catabolic process, nonsense-mediated decay 1.21E-08 1.71E-04 5.45 Co-translational protein targeting to membrane 2.57E-08 1.82E-04 5.96 mRNA metabolic process 1.97E-07 9.26E-04 2.43 amide biosynthetic process 2.77E-07 9.80E-04 3.31 protein targeting to ER 3.26E-07 9.21E-04 5.32 establishment of protein localization to endoplasmic reticulum 4.89E-07 1.15E-03 5.15 protein targeting to membrane 5.69E-07 1.15E-03 4.21 SRP-dependent co-translational protein targeting to membrane 6.11E-07 1.08E-03 5.47 protein localization to endoplasmic reticulum 1.34E-06 2.10E-03 4.75 nuclear-transcribed mRNA catabolic process 3.01E-06 4.25E-03 3.57 https://doi.org/10.1371/journal.pone.0205608.t001 ene ontology analysis of genes whose promoter contain at least one CGCG element. GRO-cap TSS are within 50bp distance of CGCG elements and that CGCG elements are largely positioned in nucleosome depleted regions as determined by the MNase-seq [42] (S3 Fig). Gene Ontology (GO) analysis showed that genes containing CGCG promoter elements produce protein-coding transcripts whose products form discernible protein-protein interact- ing networks (S4 Fig). Specifically, these genes encode core components of RNA metabolism and the translational apparatus (Table 1). Because CpG dinucleotides in CGI-associated promoters are invariably unmethylated [13], we asked if the methylation state of the CGCG elements might explain the observation that only the elements within CGIs are transcriptionally active. Analysis of ENCODE Whole Genome Bisulfate Sequencing (WGBS) from K562 cells indicated that in contrast to CpG-poor regions of the genome, CGCG elements in CGIs are largely unmethylated (Fig 4A). This observation prompted us to determine experimentally whether CpG methylation could alter the promoter activity of the CGCG element. To address this question, we inserted a single copy of TCTCGCGAGA into a secretory luciferase reporter construct that is devoid of CpG sequences (CpG-free Lucia). In this construct, the only CpG sequences are the ones contributed by the CGCG element (Fig 4B). The CpG sequences in this construct were then fully methylated using a CpG methyltransferase that specifically methylates CpG dinucleotides and the results were confirmed by saturated methyl-sensitive enzymatic digestion (Fig 4C). In comparison to the high reporter activity induced by the unmethylated TCTCGCGAGA-containing construct, methylation abrogated the promoter activity (Fig 4D), strongly suggesting that the CGCG methylation antagonizes its promoter function. PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 ENCODE WGBS methylation data show increased methylation of CGCG elements in non-CGIs compared to those in CGIs. B) Methylation of CGCG elements suppresses their promoter activity. Methylation of a TCTCGCGAGA insert in a CpG-free Lucia reporter using M.SssI CpG methyltransferase and SAM. C) Methylation of TCTCGCGAGA in the construct assessed by agarose gel analyses after digestion with NheI plus BstUI (only cuts non-methyl CGCG), which produces a doublet band of 2.3Kb. D) Lucia vs SEAP (transfection control) activity driven by WT, mutant or methylated CGCG inserts in transfected HEK293T cells. Data represent the mean of three replicates +/- SD. E) Strand-specific CAGE-seq signals generated from raw CAGE-seq sequence reads described by Brocks et al. [32] from NCI-H1299 cells treated with either the DNMT inhibitor, DAC, or DMSO control. htt //d i /10 1371/j l 0205608 004 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 13 / 25 The CGCG element induces bidirectional transcription significantly alter endogenous DENR protein levels (B in S5 Fig). These results indicate that Kaiso does not bind to the unmodified (i.e. not methylated) CGCG element. Indeed, ectopic expression of Kaiso does not suppress the DENR expression likely because the DENR promoter is not methylated in vivo. Thus, Kaiso along with other zBTB family members likely only sup- press the CGCG element-driven gene expression when this element is methylated. PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element activates gene expression in different human promoter configurations Given that the CGCG element drives bidirectional transcription, we were interested to deter- mine the frequency of this element in annotated uni- vs. bidirectional promoters. The vast majority of CGCG elements (93%) occur in annotated unidirectional promoters that drive coding or long non-coding RNAs, while 7% occur in an annotated bidirectional promoter (Table 2). However, recent studies suggest that the majority of what were classically defined as unidi- rectional promoters produce unstable “promoter upstream transcripts” (PROMPTS) [46]. Based on this, we investigated the role of CGCG elements in three different endogenous pro- moters that differ in their annotated directionality and whether they combine CGCG element with TATA-boxes. In order to determine the role of endogenous CGCG elements, we simulta- neously disrupted CGCG elements but maintained CG content by exchanging the flanking sequences (i.e. TCTCGCGAGA ! AGACGCGTCT). We first focused on the POLR1C/YIPF3 bidirectional promoter region, which has divergent TSS separated by 30 nucleotides that flank a single CGCG element. We inserted a promoter fragment (30bp) containing the wild-type CGCG element into the LuBiDi construct, and as a comparison, constructs were generated in which the flanking sequences (AGA and TCT) were exchanged. The WT fragment from POLR1C/YIPF3 promoter induced bidirectional expression irrespective of its orientation (Fig 5A). In contrast, the flank-exchanged mutants, regardless of insert orientation, did not show any discernible reporter activity. Next, we analyzed the ZZZ3 promoter which is similar to the DENR promoter in that it contains three CGCG elements (Fig 5B). Although the promoter is annotated as directional, PROMPTs on the opposite strand in both the K562 and GM12878 GRO-Cap datasets were found (Fig 5B, UCSC genome browser plot). To determine whether these elements are respon- sible for the divergent ZZZ3 transcripts, we inserted CGCG elements or flank-exchanged ele- ments from the ZZZ3 promoter into a LuBiDi construct. As shown in Fig 5B, WT sequences but not flank-exchanged could induce bidirectional reporter expression. An analysis of the DENR promoter also showed that their three CGCG elements drive bidirectional transcription in the LuBiDi assay, and that disruption of CGCG core sequences with A insertions abrogated the bidirectional promoter activity (S6 Fig). We also studied the PRDX1 promoter, a rare example in which both a single CGCG ele- ment plus a TATA-box map within the CpG-enriched promoter [47]. An analysis of GRO- Cap datasets indicated a predominant TSS approximately 25 nucleotides downstream of the Table 2. To extend these results to endogenous CGCG elements that occur in non-CGI context, we compared CAGE-seq reads of a published dataset from cells treated with either the demethyl- ating agent, deoxyazacitidine (DAC) or vehicle (DMSO) [32]. DAC treated cells showed signif- icant transcriptional activation in regions surrounding CGCG elements located in non-CGI regions on both strands. This observation supports the result of our CpG-free reporter experi- ment in that CpG methylation of CGCG elements inhibits transcriptional activation, and that demethylation of this element promotes bidirectional transcription in vivo(Fig 4E). A transcription factor zBTB33, also known as Kaiso, was shown previously to be enriched on methylated “CGCG” nucleotides [43]. Kaiso has been shown to interact with the repressive complex SMRT, leading to suppression of gene expression [44]. As illustrated in A of S5 Fig, this transcription factor interacts only with the methylated CGCG element confirming previ- ous observations [45]. The transient overexpression of Kaiso in HEK293T cells did not PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 14 / 25 Interestingly, the reporter containing a flank-exchanged CGCG element did not show any reporter activity even in the presence of a WT TATA-box, suggesting that the CGCG element not only promotes divergent transcription but also acts as a required activator for the TATA-box in this promoter. To further study the role of CGCG elements in the context of bidirectional promoters, we analyzed a set of mouse bidirectional promoters previously defined using Start-seq [48]. We assessed the presence of CGCG elements throughout the intervening regions in such bidirec- tional promoters. The coupled sense/anti-sense TSS form boundaries that flank a nucleosome- depleted region, characterized by an open chromatin structure that permits high accessibility for transcriptional machinery (Fig 5D). This analysis indicated that although CGCG elements do not show a fixed distance to sense or anti-sense Start-seq TSSs, they mostly positioned within 50 nucleotides of the predominant TSS in mouse bidirectional promoters. Associated TSS occur in context of PyPu motifs where Py and Pu are -1 and +1 nucleotides, respectively. (Fig 5E). The CGCG element induces bidirectional transcription analysis indicated associated divergent transcripts on the opposite strand. C) The promoter of PRDX1 gene contains both a TATA-box (⊳) and a CGCG element. D) Start-seq data analysis of CGCG elements in the mouse genome. The results suggest that CGCG elements are positioned in the nucleosome depleted regions between sense and anti-sense TSS. E) CGCG elements occur mostly within 50bp of sense and anti-sense Start-seq TSSs. Sense and anti-sense TSS occur in Py/Pu dinucleotide motif where Pu is the putative +1 nucleotide. Data are represented as the mean of three replicates +/- SD. https://doi.org/10.1371/journal.pone.0205608.g005 https://doi.org/10.1371/journal.pone.0205608.g005 TATA-box (Fig 5C), yet divergent transcripts were found starting roughly 50-70 bp upstream of the coding region in both K562 and GM12878 cells. To investigate the role of the TATA- box in this configuration, we inserted a fragment containing the TATA-box and CGCG ele- ment from this promoter into LuBiDi. We also produced mutants including one that disrupted the first TA in the TATA-box with CC sequences and another in which the TATA-box orien- tation was reversed relative to the CGCG element. The WT PRDX1 promoter fragment mainly drove unidirectional downstream transcription (Fig 5C) although some opposite direction reporter activity was noted. Mutation of the TATA-box severely attenuated downstream direc- tional promoter activity (Fig 5C). Interestingly, the reporter containing a flank-exchanged CGCG element did not show any reporter activity even in the presence of a WT TATA-box, suggesting that the CGCG element not only promotes divergent transcription but also acts as a required activator for the TATA-box in this promoter. TATA-box (Fig 5C), yet divergent transcripts were found starting roughly 50-70 bp upstream of the coding region in both K562 and GM12878 cells. To investigate the role of the TATA- box in this configuration, we inserted a fragment containing the TATA-box and CGCG ele- ment from this promoter into LuBiDi. We also produced mutants including one that disrupted the first TA in the TATA-box with CC sequences and another in which the TATA-box orien- tation was reversed relative to the CGCG element. The WT PRDX1 promoter fragment mainly drove unidirectional downstream transcription (Fig 5C) although some opposite direction reporter activity was noted. Mutation of the TATA-box severely attenuated downstream direc- tional promoter activity (Fig 5C). The CGCG element activates gene expression in different human promoter configurations Occurrence of CGCG elements in the annotated human promoters. Annotated configuration CGCG elements Percent Unidirectional coding 364 80 Bidirectional coding pair 22 5 Unidirectional non-coding 58 13 Non-coding and coding pair 9 2 https://doi.org/10.1371/journal.pone.0205608.t002 Table 2. Occurrence of CGCG elements in the annotated human promoters. Annotated configuration CGCG elements Percent Unidirectional coding 364 80 Bidirectional coding pair 22 5 Unidirectional non-coding 58 13 Non-coding and coding pair 9 2 https://doi.org/10.1371/journal.pone.0205608.t002 Table 2. Occurrence of CGCG elements in the annotated human promoters. Table 2. Occurrence of CGCG elements in the annotated human promoters. 15 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 15 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription The CGCG element induces bidirectional transcriptio Fig 5. CGCG elements in CGI promoters drive gene expression. A) The bidirectional promoter driving the POLR1C/YIPF3 genes (left panel) contains a conserved CGCG element between annotated TSSs (middle), which was cloned and tested for bidirectional promoter activity in LuBiDi. Right panel- the promoter insert orientation is defined by a large grey arrow, with the WT CGCG element flank orientation represented by ^, and the flank-exchanged variants represented by ⊳⊲. B) The ZZZ3 promoter contains three CGCG elements. Although this promoter is annotated as unidirectional, the GRO-Cap Fig 5. CGCG elements in CGI promoters drive gene expression. A) The bidirectional promoter driving the POLR1C/YIPF3 genes (left panel) contains a conserved CGCG element between annotated TSSs (middle), which was cloned and tested for bidirectional promoter activity in LuBiDi. Right panel- the promoter insert orientation is defined by a large grey arrow, with the WT CGCG element flank orientation represented by ^, and the flank-exchanged variants represented by ⊳⊲. B) The ZZZ3 promoter contains three CGCG elements. Although this promoter is annotated as unidirectional, the GRO-Cap PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 16 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 CGCG elements promote transcription through divergent TSS D) Bidirectional TSS associated with the LuBiDi construct containing four tandem copies of TCTCGCGAGA. E) Divergent TSS for the LuBiDi construct that contained a TATA-box and a CGCG element from the PRDX1 promoter. TSS positions are indicated in nucleotide relative to the nearest feature (the CGCG element or the TATA-box). The number of sequenced clones for each reporter constructs are indicated above coding regions. https://doi.org/10.1371/journal.pone.0205608.g006 Fig 6. CGCG element is associated with divergent transcription start sites. A) The location of gene-specific primers d i 0 RACE i id if bidi i l TSS i h L BiDi b d Fi fl d R ill Fig 6. CGCG element is associated with divergent transcription start sites. A) The location of gene-specific primers used in 50-RACE experiments to identify bidirectional TSSs in the LuBiDi based reporter constructs. Firefly and Renilla luciferase primers were designed 199 and 259 bp away from the BglII cloning site, respectively. B) Agarose gel image of firefly and Renilla luciferase RACE PCR products for the LuBiDi constructs containing none or one copy of TCTCGCGAGA. GFP transcripts were used as an internal control for RACE. C) TSS were determined for the LuBiDi construct containing a single copy of the TCTCGCGAGA motif. The sequence of +1 nucleotide and flanking five nucleotides are also shown on major TSSs. D) Bidirectional TSS associated with the LuBiDi construct containing four tandem copies of TCTCGCGAGA. E) Divergent TSS for the LuBiDi construct that contained a TATA-box and a CGCG element from the PRDX1 promoter. TSS positions are indicated in nucleotide relative to the nearest feature (the CGCG element or the TATA-box). The number of sequenced clones for each reporter constructs are indicated above coding regions. Fig 6. CGCG element is associated with divergent transcription start sites. A) The location of gene-specific primers used in 50-RACE experiments to identify bidirectional TSSs in the LuBiDi based reporter constructs. Firefly and Renilla luciferase primers were designed 199 and 259 bp away from the BglII cloning site, respectively. B) Agarose gel image of firefly and Renilla luciferase RACE PCR products for the LuBiDi constructs containing none or one copy of TCTCGCGAGA. GFP transcripts were used as an internal control for RACE. C) TSS were determined for the LuBiDi construct containing a single copy of the TCTCGCGAGA motif. The sequence of +1 nucleotide and flanking five nucleotides are also shown on major TSSs. CGCG elements promote transcription through divergent TSS Previously identified promoter elements such as the TATA box and the TCT motif promote transcription through a focused putative TSS that occurs either at a fixed distance downstream (in the case of TATA box) or on a specific nucleotide within the element in the case of the TCT motif [49]. To map the bidirectional TSSs associated with the CGCG element in both directions, we employed 50-end RACE (Rapid Amplification of cDNA Ends) strategy using RNA extracted from HEK293T cells transfected with LuBiDi reporter constructs along with pEGFP as a transfection control (Fig 6A). This robust method has been successfully used to determine the TSS of many genes in human and other organisms previously [50, 51]. As shown in Fig 6B, 50-end RACE analysis of RNA from cells transfected with the promo- terless LuBiDi construct (A in S7 Fig) did not produce transcripts in either direction, suggest- ing that the reporter construct does not exhibit intrinsic transcriptional activity. In contrast, 50-end RACE produced major products for firefly and Renilla transcripts from a LuBiDi con- struct containing one copy of the TCTCGCGAGA motif (B in S7 Fig). Sequencing of the RACE products from both firefly and Renilla transcripts showed a strong preference for purines as the +1 nucleotide, and pyrimidines as the -1 nucleotide, conforming to previous genome-wide observations [47]. Although multiple TSSs were found in the sense or anti-sense directions, there was a predominant firefly luciferase TSS (7 of 25 clones) 28 nucleotides and a predominant Renilla luciferase TSS (9 of 21 clones) 51 nucleotides from the TCTCGCGAGA element (Fig 6C). However, we observed that a majority of preferred Renilla TSS were PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 17 / 25 The CGCG element induces bidirectional transcription Fig 6. CGCG element is associated with divergent transcription start sites. A) The location of gene-specific primers used in 50-RACE experiments to identify bidirectional TSSs in the LuBiDi based reporter constructs. Firefly and Renilla luciferase primers were designed 199 and 259 bp away from the BglII cloning site, respectively. B) Agarose gel image of firefly and Renilla luciferase RACE PCR products for the LuBiDi constructs containing none or one copy of TCTCGCGAGA. GFP transcripts were used as an internal control for RACE. C) TSS were determined for the LuBiDi construct containing a single copy of the TCTCGCGAGA motif. The sequence of +1 nucleotide and flanking five nucleotides are also shown on major TSSs. https://doi.org/10.1371/journal.pone.0205608.g006 The CGCG element induces bidirectional transcription downstream of the initiation codon (ATG), and thus, unlikely to produce active Renilla lucif- erase product. This likely explains why the relative Renilla luciferase activity, but not the tran- script level, is always lower than that of the firefly luciferase, as was previously observed in Fig 3E and 3F. We also used 50-RACE to determine TSS in the reporter construct containing four tandem copies of TCTCGCGAGA motif (B in S7 Fig). As shown in Fig 6D, no TSS was observed in TCTCGCGAGA or in the 6 nucleotide spacer (BglII site) regions between the four elements. In contrast to the results observed with the construct containing one CGCG element, TSS were relocated closer to CGCG elements likely as a result of a potential cooperative Pol II recruiting mechanism when CGCG elements occur in tandem copies. Next, we determined how the presence of a TATA-box affects CGCG element-driven TSS choice using the LuBiDi reporter containing these elements from the PRDX1 promoter (B in S7 Fig). In this construct, the TATA-box was arranged between the Renilla reporter and the CGCG element. Sequencing of the Renilla RACE products showed a predominant TSS (7 of 13 clones) 26 nucleotides downstream of the TATA box on the Renilla-coding strand (Fig 6E). In contrast, on the firefly reporter coding strand, there was a concentration of multiple TSSs 40-43 nucleotides downstream of CGCG element. This TSS pattern differs from those induced from the construct containing just a single copy of the CGCG element (Fig 6C). Together with reporter data presented in Fig 5C, these results suggest that the CGCG element and TATA box cooperate to induce transcription in the PRDX1 promoter. CGCG elements promote transcription through divergent TSS D) Bidirectional TSS associated with the LuBiDi construct containing four tandem copies of TCTCGCGAGA. E) Divergent TSS for the LuBiDi construct that contained a TATA-box and a CGCG element from the PRDX1 promoter. TSS positions are indicated in nucleotide relative to the nearest feature (the CGCG element or the TATA-box). The number of sequenced clones for each reporter constructs are indicated above coding regions. Fig 6. CGCG element is associated with divergent transcription start sites. A) The location of gene-specific primers used in 50-RACE experiments to identify bidirectional TSSs in the LuBiDi based reporter constructs. Firefly and Renilla luciferase primers were designed 199 and 259 bp away from the BglII cloning site, respectively. B) Agarose gel image of firefly and Renilla luciferase RACE PCR products for the LuBiDi constructs containing none or one copy of TCTCGCGAGA. GFP transcripts were used as an internal control for RACE. C) TSS were determined for the LuBiDi construct containing a single copy of the TCTCGCGAGA motif. The sequence of +1 nucleotide and flanking five nucleotides are also shown on major TSSs. D) Bidirectional TSS associated with the LuBiDi construct containing four tandem copies of TCTCGCGAGA. E) Divergent TSS for the LuBiDi construct that contained a TATA-box and a CGCG element from the PRDX1 promoter. TSS positions are indicated in nucleotide relative to the nearest feature (the CGCG element or the TATA-box). The number of sequenced clones for each reporter constructs are indicated above coding regions. https://doi.org/10.1371/journal.pone.0205608.g006 18 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription sequences that account for the majority of CGCG elements do not overlap with the hDRE (a list of human genes that are potentially regulated by the CGCG element is provided in the S2 File). As outlined in the S5C and S5D Fig, successful knockdown of ZBED1 using specific len- tivirus-induced shRNAs do not change the endogenous protein or transcript levels of the DENR gene whose promoter contains three copies of the CGCG element. This result suggests that ZBED1 may not regulate the CGCG element-driven transcription of genes from the TATA-less CpG islands-associated promoters. Therefore, we speculate that the CGCG element is a different promoter element from the hDRE motif that function through transcription fac- tor ZBED1 and is regulated by different mediators. How housekeeping genes whose products are core components of cellular processes are transcriptionally regulated is poorly understood. In this study, we found that genes whose products play a central role in translation and transcription are enriched for CGCG elements in their CGI-associated promoters. This analysis led us to identify a group of ribosomal genes whose CpG rich promoters contain one or multiple CGCG elements (A in S8 Fig). As shown in the panel B of S8 Fig, a group of ribosomal protein promoters containing CGCG elements do not coincide with the previously described TCT motif that is thought to regulate the tran- scription of this gene class in humans [49]. These results suggest that TCT and CGCG ele- ments either cooperate or act independently to regulate the expression of specific sets of ribosomal protein genes in the human genome. In addition to genes encoding ribosomal pro- teins, promoters of key translation initiation factor genes encoding EIF5, EIF3H, and DENR, as well as the essential translation termination factor ETF1, contain CGCG elements in their promoters. This is consistent with the current perspective that different classes of promoter elements regulate functionally distinct protein coding genes [1]. Additionally, we demonstrated that methylation of CpGs in the CGCG element could sup- press promoter activity. Indeed, roughly 80 percent of CpG sites in the genome, particularly CpGs that occur outside of CGIs, are methylated [56]. We speculate a switch-like mechanism that could activate or repress gene expression based on the methylation status of CGCG ele- ments. Accordingly, we propose a model where CGCG elements, when occurring in CGIs, are protected from methylation thereby maintaining promoter activity in housekeeping genes. In contrast, CGCG elements in other regions of the genome would be more subject to methyla- tion, resulting in transcriptional silencing. In theory, DNA methylation of CGCG elements could protect the genome from spurious transcription, as reviewed elsewhere [57]. A similar switch-like mechanism for a group of transcription factors that contain CpG motif has been described in the past in which CpG methylation would affect the binding affinity of transcrip- tion factors [58]. Although the nature of the factor, or factors, that bind to non-methyl CGCG element has yet to be clarified, our results suggest that ChIP-seq studies should be interpreted with greater consideration to account for the differential binding of proteins to methyl or non- methyl CpG-containing motif sequences. Discussion In this study, we identify a novel promoter element that drives bidirectional transcription mainly in the context of TATA-less promoters. Although in previous studies, sequences simi- lar to this element were found in the promoter of individual genes, the functional role of the CGCG element in CGIs and TATA-less promoters in the human genome was never explored [22–25]. Whereas other promoter elements (e.g. TATA and GC boxes) require an activator binding site to initiate directional transcription [6], a single instance of the CGCG element is both necessary and sufficient to promote bidirectional transcription. However, in comparison to other known promoter elements that induce transcription, which typically occur once in most promoters, CGCG elements occur in multiple copies in a small percentage of CGI-con- taining promoters, a phenomenon that likely influences RNA polymerase recruitment and subsequent transcriptional rates. An interesting yet poorly studied feature of vertebrate genomes is the presence of CpG-rich regions known as CGIs [14]. Although CGIs mark transcriptionally active regions of the genome, the mechanism of RNA polymerase recruitment in these regions has been elusive [13]. Through enrichment analysis, we found that CGCG elements are enriched in a group of CGI-containing promoters and that they can recruit transcriptional machinery to promote bidirectional transcription, a feature common among transcriptionally active CpG islands [19]. Additionally, we provide evidence that in some rare cases, the CGCG element could interact functionally with an adjacent TATA-box to activate directional gene expression. Simi- lar synergistic activities have been described previously [52, 53] suggesting that the CGCG ele- ment also shares this attribute with other known promoter elements. In previous studies, a motif called hDRE (TGTCGYGAYA, where Y is a pyridimide nucleo- tide) was shown to positively activate metabolism-related and ribosomal protein genes through binding of ZBED1(also known as DREF) protein [54, 55]. The sequence of hDRE is very similar to that of CGCG element. However, as we showed in Fig 3C, TCTCGCGAGA and TATCGCGAGA (underlined nucleotides denote deviations from the hDRE sequence) PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 19 / 25 Conclusion In this study, we provide strong evidence that CGCG elements are evolutionarily conserved in vertebrates, functioning as an active component of CGI-associated promoters. The unmethy- lated form of the element may be sufficient to drive bidirectional transcription of TATA-less promoters. An interesting, yet very important question to address in future studies is whether the CGCG element functions as a core promoter element or as a sequence-specific transcrip- tion factor binding site (SSTFBS). An argument for the core promoter element characteristic of the CGCG element includes its ability to initiate local de novo bidirectional transcription in the absence of a core promoter element from nearby Py/Pu(+1) sites. This notion is supported PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 20 / 25 The CGCG element induces bidirectional transcription by studies that show that SSTFBSs cannot initiate transcription in the absence of a core pro- moter element. However, similar to SSTFBSs, CGCG elements can occur in tandem copies and the copy number modulates transcription intensity. Unlike traditional core promoter ele- ments, the CGCG element is not positioned at a fixed distance to TSSs, but it is found within a range of 20 to 70 nucleotides (with a peak of apporoximatly 50 nucleotides) upstream of TSS on either strands as was determined by the analysis of Start-seq and GRO-Cap datasets. It is also conceivable that CGCG elements modulate the induction of transcription by recruiting factors that could affect nucleosomal positioning in CGIs. Identification of the CGCG element interacting factor will likely clarify whether the CGCG element functions as a core promoter element and its role in driving transcription of housekeeping genes from CpG-rich and TATA-less promoters. Supporting information S1 Fig. The CGCG element (motif 10) is associated with DNase-seq footprint in different cell lines. A) ENCODE DNase-seq footprints of motif 7 and 10 for available cell lines. B) TCTCGCGAGA motif occurs within 50bp of annotated TSSs in the human and mouse genomes. S2 Fig. The CGCG elements promote simultaneous expression of GFP and mCherry genes in the pmCGFP reporter construct. A) The pmCGFP bidirectional reporter structure. B) Fluorescence image 24h after transfection of HEK293T cells with pmCGFP constructs contain- ing 0, 1 or 3 copies of the TCTCGCGAGA motif. CMV-driven BFP expression was used as an internal control. C) Immunoblots showing levels of GFP and mCherry expression 24 and 48h post transfection. D) Time-lapse imaging of HEK293T cells transfected with pmCGFP con- taining three copies of TCTCGCGAGA for 24h shows that both reporters are simultaneously expressed. Scale bar is 100 μm. E) CGCG element confers bidirectional expression of GFP and H2b-mCherry reporter genes in HEK293T. Time-lapse images of HEK293T cells transfected with pmCGFP-H2b (h2b-mCherry fused gene). Please note delayed H2b-mCherry signals as the fused mCherry protein is being trafficked into the nucleus. F) Images of NMuMG mouse cells transfected with a pmCGFP-H2b construct containing either 3 copies of the wild-type TCTCGCGAGA motif or 3 copies of the TCTCTCGAGA mutant motif. (EPS) S3 Fig. GRO-cap profiles were generated for +/- 1000bp from TCTCGCGAGA and TATA- box sites in K562 cells. The GRO-cap signals, which map TSS in vivo, clearly show TSS peaks are located 50bp from the TCTCGCGAGA in both directions. The figure also includes MNase-seq plots for the two promoter element +/-1000bp. The data indicates that genomic regions containing TCTCGCGAGA and TATA-boxes are nucleosome depleted. (EPS) S4 Fig. REACTOME gene network analysis of CGCG containing promoters. An analysis of genes that contain CGCG elements in their promoters found that most of these genes can be clustered into distinct functional groups as indicated in the figure. (EPS) S4 Fig. REACTOME gene network analysis of CGCG containing promoters. An analysis of genes that contain CGCG elements in their promoters found that most of these genes can be clustered into distinct functional groups as indicated in the figure. (EPS) S5 Fig. Kaiso binds specifically to the methylated CGCG element. A) Pulldown experiment using biotin-tagged CGCG oligonucleotides followed by immunoblotting for Kaiso or tubulin. B) Immunoblot analysis of transient expression of Kaiso in HEK293T cells. Acknowledgments Authors would like to thank William Burhans for his valuable input and critical review of the manuscript. First author (AM) would like to dedicate this manuscript to his parents. S1 Video. CGCG elements activate dual expression of GFP and mCherry reporter genes. S1 Video. CGCG elements activate dual expression of GFP and mCherry reporter genes. Live cells were transfected with pmCGFP reporter constructs containing none, one or three tandem copies of TCTCGCGAGA. Cells were incubated in a chamber supporting normal cel- lular growth and images were captured to record mCherry and GFP signals every 30 minutes. The top panel for each reporter assay shows overlapping signals. Lower panels show GFP or mCherry fluorescence channels. (MP4) The CGCG element induces bidirectional transcription does not change the DENR protein product. D) RT-PCR analysis of ZBED1 knockdown cells does not attenuate the transcript levels of the DENR gene. (EPS) S6 Fig. CGCG elements in the DENR promoter promote divergent transcription. A) GRO- cap signal tracks of the CpG-rich DENR promoter. B) The CGCG elements in the DENR pro- moter, regardless of the insert direction, activated bidirectional reporter luciferase activities. Insertion of an “A” in the center of CGCG elements eliminated the promoter activity. Data are represented as the mean of three replicates ± SD. (EPS) S6 Fig. CGCG elements in the DENR promoter promote divergent transcription. A) GRO- cap signal tracks of the CpG-rich DENR promoter. B) The CGCG elements in the DENR pro- moter, regardless of the insert direction, activated bidirectional reporter luciferase activities. Insertion of an “A” in the center of CGCG elements eliminated the promoter activity. Data are represented as the mean of three replicates ± SD. (EPS) S7 Fig. Sequence of LuBiDi constructs used for RACE experiments. A) The sequence flank- ing the BglII restriction site where DNA duplexes were inserted into the promoterless LuBiDi construct. The reporter genes in either direction are shown in bold. B) The sequences of DNA fragments inserted into LuBiDi, containing flanking BglII restriction sites (boxed), used for RACE analyses. (EPS) S7 Fig. Sequence of LuBiDi constructs used for RACE experiments. A) The sequence flank- ing the BglII restriction site where DNA duplexes were inserted into the promoterless LuBiDi construct. The reporter genes in either direction are shown in bold. B) The sequences of DNA fragments inserted into LuBiDi, containing flanking BglII restriction sites (boxed), used for RACE analyses. (EPS) S7 Fig. Sequence of LuBiDi constructs used for RACE experiments. A) The sequence flank- ing the BglII restriction site where DNA duplexes were inserted into the promoterless LuBiDi construct. The reporter genes in either direction are shown in bold. B) The sequences of DNA fragments inserted into LuBiDi, containing flanking BglII restriction sites (boxed), used for RACE analyses. (EPS) S8 Fig. CGCG elements are enriched in ribosomal protein promoters. A) Aligned sequences of CGCG elements and flanking regions in the promoters of ribosomal proteins genes, with the resulting consensus sequence of TCTCGCGAGA shown below. The list also indicates whether the TCT or hDRE motifs are present in a given promoter. B) Venn diagram showing the distribution of TCT and CGCG elements in human ribosomal proteins promoters. (TIF) S8 Fig. CGCG elements are enriched in ribosomal protein promoters. A) Aligned sequences of CGCG elements and flanking regions in the promoters of ribosomal proteins genes, with the resulting consensus sequence of TCTCGCGAGA shown below. The list also indicates whether the TCT or hDRE motifs are present in a given promoter. B) Venn diagram showing the distribution of TCT and CGCG elements in human ribosomal proteins promoters. (TIF) Author Contributions Conceptualization: Amin Mahpour. Supporting information Two independent replicates (Rep1 and Rep2) were used in this experiment. C) Successful knockdown of ZBED1 S5 Fig. Kaiso binds specifically to the methylated CGCG element. A) Pulldown experiment using biotin-tagged CGCG oligonucleotides followed by immunoblotting for Kaiso or tubulin. B) Immunoblot analysis of transient expression of Kaiso in HEK293T cells. Two independent replicates (Rep1 and Rep2) were used in this experiment. C) Successful knockdown of ZBED1 21 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 References 1. Levine M, Cattoglio C, Tjian R. Looping back to leap forward: transcription enters a new era. Cell. 2014; 157(1):13–25. https://doi.org/10.1016/j.cell.2014.02.009 PMID: 24679523 2. Lee TI, Young RA. Transcriptional regulation and its misregulation in disease. Cell. 2013; 152(6):1237– 51. https://doi.org/10.1016/j.cell.2013.02.014 PMID: 23498934 3. Roeder RG, Rutter WJ. Multiple forms of DNA-dependent RNA polymerase in eukaryotic organisms. Nature. 1969; 224(5216):234–7. https://doi.org/10.1038/224234a0 PMID: 5344598 4. Roeder RG. The complexities of eukaryotic transcription initiation: regulation of preinitiation complex assembly. Trends in Biochemical Sciences. 1991; 16(11):402–8. https://doi.org/10.1016/0968-0004 (91)90164-Q PMID: 1776168 5. Juven-Gershon T, Hsu JY, Theisen JW, Kadonaga JT. The RNA polymerase II core promoter—the gateway to transcription. Current Opinion in Cell Biology. 2008; 20(3):253–9. https://doi.org/10.1016/j. ceb.2008.03.003 PMID: 18436437 6. Smale ST, Kadonaga JT. The RNA polymerase II core promoter. Annual Review of Biochemistry. 2003; 72:449–79. https://doi.org/10.1146/annurev.biochem.72.121801.161520 PMID: 12651739 7. Sandelin A, Carninci P, Lenhard B, Ponjavic J, Hayashizaki Y, Hume DA. Mammalian RNA polymerase II core promoters: insights from genome-wide studies. Nature Reviews Genetics. 2007; 8(6):424–36. https://doi.org/10.1038/nrg2026 PMID: 17486122 8. Yang C, Bolotin E, Jiang T, Sladek FM, Martinez E. Prevalence of the initiator over the TATA box in human and yeast genes and identification of DNA motifs enriched in human TATA-less core promoters. Gene. 2007; 389(1):52–65. https://doi.org/10.1016/j.gene.2006.09.029 PMID: 17123746 9. Deng W, Roberts SG. A core promoter element downstream of the TATA box that is recognized by TFIIB. Genes & Development. 2005; 19(20):2418–23. https://doi.org/10.1101/gad.342405 10. Lim CY, Santoso B, Boulay T, Dong E, Ohler U, Kadonaga JT. The MTE, a new core promoter element for transcription by RNA polymerase II. Genes & Development. 2004; 18(13):1606–17. https://doi.org/ 10.1101/gad.1193404 11. Anish R, Hossain MB, Jacobson RH, Takada S. Characterization of transcription from TATA-less pro- moters: identification of a new core promoter element XCPE2 and analysis of factor requirements. PLOS ONE. 2009; 4(4):e5103. https://doi.org/10.1371/journal.pone.0005103 PMID: 19337366 12. Burke TW, Kadonaga JT. The downstream core promoter element, DPE, is conserved from Drosophila to humans and is recognized by TAFII60 of Drosophila. Genes & Development. 1997; 11(22):3020–31. https://doi.org/10.1101/gad.11.22.3020 13. Deaton AM, Bird A. CpG islands and the regulation of transcription. Genes & Development. 2011; 25 (10):1010–22. https://doi.org/10.1101/gad.2037511 14. Gardiner-Garden M, Frommer M. CpG islands in vertebrate genomes. Journal of Molecular Biology. 1987; 196(2):261–82. https://doi.org/10.1016/0022-2836(87)90689-9 PMID: 3656447 15. Zhu J, He F, Hu S, Yu J. On the nature of human housekeeping genes. Trends in Genetics. 2008; 24 (10):481–4. https://doi.org/10.1016/j.tig.2008.08.004 PMID: 18786740 16. Wierstra I. Conceptualization: Amin Mahpour. Data curation: Amin Mahpour, Benjamin S. Scruggs. Data curation: Amin Mahpour, Benjamin S. Scruggs. Formal analysis: Amin Mahpour. Funding acquisition: Toru Ouchi. Funding acquisition: Toru Ouchi. Methodology: Amin Mahpour, Dominic Smiraglia, Irwin H. Gelman. 22 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 The CGCG element induces bidirectional transcription Software: Amin Mahpour, Dominic Smiraglia, Irwin H. Gelman. Software: Amin Mahpour, Dominic Smiraglia, Irwin H. Gelman. Supervision: Irwin H. Gelman. Validation: Amin Mahpour. Writing – original draft: Amin Mahpour. Writing – review & editing: Irwin H. Gelman. Validation: Amin Mahpour. Writing – original draft: Amin Mahpour. Writing – review & editing: Irwin H. Gelman. References Sp1: emerging roles–beyond constitutive activation of TATA-less housekeeping genes. Bio- chemical and Biophysical Research Communications. 2008; 372(1):1–13. https://doi.org/10.1016/j. bbrc.2008.03.074 PMID: 18364237 17. Hargreaves DC, Horng T, Medzhitov R. Control of inducible gene expression by signal-dependent tran- scriptional elongation. Cell. 2009; 138(1):129–45. https://doi.org/10.1016/j.cell.2009.05.047 PMID: 19596240 18. Yang MQ, Laflamme K, Gotea V, Joiner CH, Seidel NE, Wong C, et al. Genome-wide detection of a TFIID localization element from an initial human disease mutation. Nucleic Acids Research. 2011; 39 (6):2175–87. https://doi.org/10.1093/nar/gkq1035 PMID: 21071415 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 23 / 25 The CGCG element induces bidirectional transcription 19. Core LJ, Waterfall JJ, Lis JT. Nascent RNA Sequencing Reveals Widespread Pausing and Divergent Initiation at Human Promoters. Science. 2008; 322(5909):1845–1848. https://doi.org/10.1126/science. 1162228 PMID: 19056941 20. Seila AC, Calabrese JM, Levine SS, Yeo GW, Rahl PB, Flynn RA, et al. Divergent transcription from active promoters. Science. 2008; 322(5909):1849–51. https://doi.org/10.1126/science.1162253 PMID: 19056940 21. Preker P, Nielsen J, Kammler S, Lykke-Andersen S, Christensen MS, Mapendano CK, et al. RNA exo- some depletion reveals transcription upstream of active human promoters. Science. 2008; 322 (5909):1851–4. https://doi.org/10.1126/science.1164096 PMID: 19056938 22. Haun RS, Moss J, Vaughan M. Characterization of the human ADP-ribosylation factor 3 promoter. Transcriptional regulation of a TATA-less promoter. Journal of Biological Chemistry. 1993; 268 (12):8793–800. PMID: 8473323 23. Wyrwicz LS, Gaj P, Hoffmann M, Rychlewski L, Ostrowski J. A common cis-element in promoters of protein synthesis and cell cycle genes. Acta Biochimica Polonica. 2007; 54(1):89–98. PMID: 17351670 24. Mikula M, Gaj P, Dzwonek K, Rubel T, Karczmarski J, Paziewska A, et al. Comprehensive analysis of the palindromic motif TCTCGCGAGA: a regulatory element of the HNRNPK promoter. DNA Research. 2010; 17(4):245–60. https://doi.org/10.1093/dnares/dsq016 PMID: 20587588 25. Guo G, Rodelsperger C, Digweed M, Robinson PN. Regulation of fibrillin-1 gene expression by Sp1. Gene. 2013; 527(2):448–55. https://doi.org/10.1016/j.gene.2013.07.006 PMID: 23860323 26. Xie X, Lu J, Kulbokas EJ, Golub TR, Mootha V, Lindblad-Toh K, et al. Systematic discovery of regula- tory motifs in human promoters and 3’ UTRs by comparison of several mammals. Nature. 2005; 434 (7031):338–45. https://doi.org/10.1038/nature03441 PMID: 15735639 27. Nguyen VT, Giannoni F, Dubois MF, Seo SJ, Vigneron M, Kedinger C, et al. In vivo degradation of RNA polymerase II largest subunit triggered by alpha-amanitin. Nucleic Acids Research. 1996; 24(15):2924– 9. https://doi.org/10.1093/nar/24.15.2924 PMID: 8760875 28. Lee TI, Johnstone SE, Young RA. Chromatin immunoprecipitation and microarray-based analysis of protein location. Nature Protocols. 2006; 1(2):729–48. https://doi.org/10.1038/nprot.2006.98 PMID: 17406303 29. References Analysis of nascent RNA identifies a uni- fied architecture of initiation regions at mammalian promoters and enhancers. Nature Genetics. 2014; 46(12):1311–20. https://doi.org/10.1038/ng.3142 PMID: 25383968 42. Gaffney DJ, McVicker G, Pai AA, Fondufe-Mittendorf YN, Lewellen N, Michelini K, et al. Controls of Nucleosome Positioning in the Human Genome. PLoS Genetics. 2012; 8(11):e1003036. https://doi.org/ 10.1371/journal.pgen.1003036 PMID: 23166509 43. Prokhortchouk A, Hendrich B, Jorgensen H, Ruzov A, Wilm M, Georgiev G, et al. The p120 catenin part- ner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes & Development. 2001; 15 (13):1613–8. https://doi.org/10.1101/gad.198501 44. Perissi V, Jepsen K, Glass CK, Rosenfeld MG. Deconstructing repression: evolving models of co- repressor action. Nature Reviews Genetics. 2010; 11(2):109–23. https://doi.org/10.1038/nrg2736 PMID: 20084085 45. Raghav SK, Waszak SM, Krier I, Gubelmann C, Isakova A, Mikkelsen TS, et al. Integrative genomics identifies the corepressor SMRT as a gatekeeper of adipogenesis through the transcription factors C/ EBPbeta and KAISO. Molecular Cell. 2012; 46(3):335–50. https://doi.org/10.1016/j.molcel.2012.03.017 PMID: 22521691 46. Chen Y, Pai AA, Herudek J, Lubas M, Meola N, Jarvelin AI, et al. Principles for RNA metabolism and alternative transcription initiation within closely spaced promoters. Nature Genetics. 2016; 48(9):984– 94. https://doi.org/10.1038/ng.3616 PMID: 27455346 47. Carninci P, Sandelin A, Lenhard B, Katayama S, Shimokawa K, Ponjavic J, et al. Genome-wide analy- sis of mammalian promoter architecture and evolution. Nature Genetics. 2006; 38(6):626–35. https:// doi.org/10.1038/ng1789 PMID: 16645617 48. Scruggs BS, Gilchrist DA, Nechaev S, Muse GW, Burkholder A, Fargo DC, et al. Bidirectional Tran- scription Arises from Two Distinct Hubs of Transcription Factor Binding and Active Chromatin. Molecu- lar Cell. 2015; 58(6):1101–12. https://doi.org/10.1016/j.molcel.2015.04.006 PMID: 26028540 49. Parry TJ, Theisen JW, Hsu JY, Wang YL, Corcoran DL, Eustice M, et al. The TCT motif, a key compo- nent of an RNA polymerase II transcription system for the translational machinery. Genes & Develop- ment. 2010; 24(18):2013–8. https://doi.org/10.1101/gad.1951110 50. Kovalskaya E, Buzdin A, Gogvadze E, Vinogradova T, Sverdlov E. Functional human endogenous ret- roviral LTR transcription start sites are located between the R and U5 regions. Virology. 2006; 346 (2):373–8. https://doi.org/10.1016/j.virol.2005.11.007 PMID: 16337666 51. Denoeud F, Kapranov P, Ucla C, Frankish A, Castelo R, Drenkow J, et al. Prominent use of distal 5- transcription start sites and discovery of a large number of additional exons in ENCODE regions. Genome Research. 2007; 17(6):746–759. https://doi.org/10.1101/gr.5660607 PMID: 17567994 52. Burke TW, Willy PJ, Kutach AK, Butler JE, Kadonaga JT. References Previs MJ, Beck Previs S, Gulick J, Robbins J, Warshaw DM. Molecular mechanics of cardiac myosin- binding protein C in native thick filaments. Science. 2012; 337(6099):1215–8. https://doi.org/10.1126/ science.1223602 PMID: 22923435 30. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformat- ics. 2010; 26(6):841–2. https://doi.org/10.1093/bioinformatics/btq033 PMID: 20110278 31. Heinz S, Benner C, Spann N, Bertolino E, Lin YC, Laslo P, et al. Simple combinations of lineage-deter- mining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Molecular Cell. 2010; 38(4):576–89. https://doi.org/10.1016/j.molcel.2010.05.004 PMID: 20513432 32. Brocks D, Schmidt CR, Daskalakis M, Jang HS, Shah NM, Li D, et al. DNMT and HDAC inhibitors induce cryptic transcription start sites encoded in long terminal repeats. Nature Genetics. 2017; 49 (7):1052–1060. https://doi.org/10.1038/ng.3889 PMID: 28604729 33. Grant CE, Bailey TL, Noble WS. FIMO: scanning for occurrences of a given motif. Bioinformatics. 2011; 27(7):1017–8. https://doi.org/10.1093/bioinformatics/btr064 PMID: 21330290 34. Eden E, Navon R, Steinfeld I, Lipson D, Yakhini Z. GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics. 2009; 10:48. https://doi.org/10.1186/ 1471-2105-10-48 PMID: 19192299 35. Ioshikhes IP, Zhang MQ. Large-scale human promoter mapping using CpG islands. Nature Genetics. 2000; 26(1):61–3. https://doi.org/10.1038/79189 PMID: 10973249 36. Rozenberg JM, Shlyakhtenko A, Glass K, Rishi V, Myakishev MV, FitzGerald PC, et al. All and only CpG containing sequences are enriched in promoters abundantly bound by RNA polymerase II in multi- ple tissues. BMC Genomics. 2008; 9:67. https://doi.org/10.1186/1471-2164-9-67 PMID: 18252004 37. Butler JE, Kadonaga JT. The RNA polymerase II core promoter: a key component in the regulation of gene expression. Genes & Development. 2002; 16(20):2583–92. https://doi.org/10.1101/gad.1026202 38. Lavrrar JL, Farnham PJ. The Use of Transient Chromatin Immunoprecipitation Assays to Test Models for E2F1-specific Transcriptional Activation. Journal of Biological Chemistry. 2004; 279(44):46343– 46349. https://doi.org/10.1074/jbc.M402692200 PMID: 15328355 39. Kedinger C, Gniazdowski M, Mandel J J L, Gissinger F, Chambon P. Alpha-amanitin: a specific inhibitor of one of two DNA-pendent RNA polymerase activities from calf thymus. Biochemical and Biophysical Research Communications. 1970; 38(1):165–71. https://doi.org/10.1016/0006-291X(70)91099-5 PMID: 4907405 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 24 / 25 The CGCG element induces bidirectional transcription 40. Ran FA, Hsu PD, Lin CY, Gootenberg JS, Konermann S, Trevino AE, et al. Double nicking by RNA- guided CRISPR Cas9 for enhanced genome editing specificity. Cell. 2013; 154(6):1380–9. https://doi. org/10.1016/j.cell.2013.08.021 PMID: 23992846 41. Core LJ, Martins AL, Danko CG, Waters CT, Siepel A, Lis JT. PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018 References The DPE, a conserved downstream core pro- moter element that is functionally analogous to the TATA box. Cold Spring Harb Symp Quant Biol. 1998; 63:75–82. https://doi.org/10.1101/sqb.1998.63.75 PMID: 10384272 53. Emami KH, Jain A, Smale ST. Mechanism of synergy between TATA and initiator: synergistic binding of TFIID following a putative TFIIA-induced isomerization. Genes & Development. 1997; 11(22):3007– 3019. https://doi.org/10.1101/gad.11.22.3007 54. Ohshima N, Takahashi M, Hirose F. Identification of a human homologue of the DREF transcription fac- tor with a potential role in regulation of the histone H1 gene. The Journal of Biological Chemistry. 2003; 278(25):22928–22938 https://doi.org/10.1074/jbc.M303109200 PMID: 12663651 55. Yamashita D, Sano Y, Adachi Y, Okamoto Y, Osada H, Takahashi T, Yamaguchi T, Osumi T, Hirose F. hDREF Regulates Cell Proliferation and Expression of Ribosomal Protein Genes. Molecular and Cellu- lar Biology. 2007; 27:2003–2013 https://doi.org/10.1128/MCB.01462-06 PMID: 17220279 56. Lister R, O’Malley RC, Tonti-Filippini J, Gregory BD, Berry CC, Millar AH, et al. Highly integrated single- base resolution maps of the epigenome in Arabidopsis. Cell. 2008; 133(3):523–36. https://doi.org/10. 1016/j.cell.2008.03.029 PMID: 18423832 57. Ndlovu MN, Denis H, Fuks F. Exposing the DNA methylome iceberg. Trends in Biochemical Sciences. 2011; 36(7):381–7. https://doi.org/10.1016/j.tibs.2011.03.002 PMID: 21497094 58. Domcke S, Bardet AF, Adrian Ginno P, Hartl D, Burger L, Schu¨beler D. Competition between DNA methylation and transcription factors determines binding of NRF1. Nature. 2015; 528(7583):575–579. http://www.nature.com/nature/journal/v528/n7583/abs/nature16462.html. PMID: 26675734 25 / 25 PLOS ONE \| https://doi.org/10.1371/journal.pone.0205608 October 17, 2018
https://openalex.org/W1965381766	https://www.scielo.br/j/rlae/a/vDBpwytKNhBsLbzyYkPygFq/?lang=en&format=pdf	English	null	The Bristol Stool Form Scale: its translation to Portuguese, cultural adaptation and validation	Revista latino-americana de enfermagem	2,012	cc-by	4,374	Original Article Original Article Rev. Latino-Am. Enfermagem 2012 May.-June;20(3):583-89 www.eerp.usp.br/rlae Corresponding Author: Anna Paula Martinez Rua Portugal, 63 Bairro: Jardim Europa CEP: 18045-280, Sorocaba, SP, Brasil 1 Master's student, Faculdade de Ciências Médicas e da Saúde, Pontifícia Universidade Católica de São Paulo, Brazil. Scholarship holder of the Scientific Initiation Program at the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). 2 h f ld d d ê éd d úd fí d d ól d ã l l The Bristol Stool Form Scale: its translation to Portuguese, cultural adaptation and validation Anna Paula Martinez1 Gisele Regina de Azevedo2 The Bristol Stool Form Scale is used for describing feces. The objective of this research was its translation, cultural adaptation and validation for Brazil. The methodology was translation, back-translation and discussion. Validation involved 85 nurses, 80 doctors, and 80 patients, who correlated images of seven types of feces with the descriptions. Results: there was a difference in sex distribution, with males predominating among the doctors and females among nurses and patients. In relation to concordance between definitions and pictures, the highest percentage was in type 5 in all three groups and the lowest was in types 6 and 7 for the doctors, in type 3 for the nurses, and type 6 for the patients. The general Kappa index was 0.826. Conclusion: the scale demonstrated high reliability for all the groups studied. Descriptors: Translation; Validation Studies; Stools. 1 Master's student, Faculdade de Ciências Médicas e da Saúde, Pontifícia Universidade Católica de São Paulo, Brazil. Scholarship holder of the Scientific Initiation Program at the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). 2 PhD, Assistant Professor, Faculdade de Ciências Médicas e da Saúde, Pontifícia Universidade Católica de São Paulo, Brazil. Assistant Professor, Faculdade de Ciências Médicas e da Saúde, Pontifícia Universidade Católica de São Paulo, Brazil. Corresponding Author: Anna Paula Martinez Rua Portugal, 63 Bairro: Jardim Europa CEP: 18045-280, Sorocaba, SP, Brasil 584 Descriptores: Traducción; Estudios de Validación; Hece. Descriptores: Traducción; Estudios de Validación; Hece. Tradução, adaptação cultural e validação da Bristol Stool Form Scale para a população brasileira A Escala de Bristol para Consistência de Fezes é usada na descrição de fezes. O objetivo deste estudo foi realizar a tradução, adaptação cultural e validação para o Brasil, dessa escala. Como metodologia realizou-se a tradução, tradução reversa e discussão. Para essa validação, incluíram-se 85 enfermeiros e 80 médicos e pacientes que correlacionaram imagens de sete tipos de fezes com descrições. Os resultados mostraram que houve diferença de distribuição do sexo, com predomínio do sexo masculino para médicos e feminino para enfermeiros e pacientes. Em relação à concordância entre definições e imagens, o maior percentual, no tipo 5, e os menores percentuais, nos tipos 6 e 7, relacionaram-se aos médicos; no tipo 3 referiram-se aos enfermeiros e no tipo 6 aos pacientes. O índice Kappa geral foi de 0,826. Conclui-se que a escala demonstrou alta confiabilidade em todos os grupos estudados. Descritores: Tradução; Estudos de Validação; Fezes. Traducción, adaptación cultural y validación de la "Bristol Stool Form Scale” La “Bristol Stool Form Scale” es usada para describir las heces. Objetivo: traducción, adaptación cultural y la validación para ser utilizada en Brasil. Metodología: Fue realizada la traducción, la traducción inversa y la discusión final. Para validar, se incluyeron 85 enfermeros y 80 médicos y pacientes que correlacionaron diseños de siete tipos de heces con descripciones. Resultados — Hubo diferencia en cuanto a la distribución del sexo con predominio masculino entre los médicos y femenino para los enfermeros y pacientes. Con respecto a la concordancia entre los conceptos y las imágenes, la mayor concordancia fue del tipo 5 en cuanto que el de menor correspondencia para los médicos fueron los tipos 6 y 7, para los enfermeros el 3 y el 6 para los pacientes. El índice de Kappa general fue de 0,826. Conclusión: Los valores obtenidos demuestran la alta confiabilidad de este cuestionario con respecto a los grupos estudiados. www.eerp.usp.br/rlae Introduction Anal incontinence can be idiopathic, congenital, of neurological origin or secondary to trauma(9). According to the Rome II Criteria, AI can be classified as light, characterized by loss of gases or soiling; moderate, with the loss of mushy or liquid feces up to three times per week, and severe, when there is total incontinence with loss three or more times per week(10-14). Analysis of bowel habits and feces type has always been part of evaluations carried out by healthcare professionals for characterizing physiological aspects of patients, as well as in the diagnosis or follow-up of diseases which involve alterations in intestinal transit. In physiological situations, the characterization of feces can be quite useful. For example, the relationship with the type of diet and bowel habits and the resulting form of the feces has been explored in the literature(1-3). Another condition which is characterized by alterations in the form of the feces is Irritable Bowel Syndrome(15), which is functional by nature and which has as its principal symptoms abdominal pain and discomfort, related to alteration in bowel habits(15-17) Another area where analysis of feces can be useful is in the post-operative stage of treatment for gastro-intestinal illnesses(18). In relation to the diagnosis and medical supervision, on the other hand, one can cite its importance in anal incontinence (AI), irritable bowel syndrome, and in the post-operative stage after abdominal surgery, among others(4-7). Anal function declines with age and women experience a greater functional loss than men(8). Anal incontinence is frequent during the menopause and lesions sustained during childbirth increase the risk of compromise of anal function(8). A significant portion of this compromise is purely the result of aging(8). www.eerp.usp.br/rlae Some instruments are utilized to help in classification and treatment(19-20); one which stands out is the Bristol Stool Form Scale (BSFS), which was developed and validated by Kenneth W. Heaton and S. J. Lewis. Its objective is to descriptively evaluate the form of fecal content, using graphic methods which represent seven types of feces according to 585 Martinez AP, Azevedo GR. Following this, back translation was carried out by a different pair of bilingual specialists. their shape and consistency. The innovation of this scale is that it presents images which illustrate the feces, along with precise descriptions as to the shape and consistency, using easily recognizable examples(21). Method Permission for the translation, validation and adaptation of the Bristol Stool Form Scale was obtained from the company Norgine Ltd which holds copyright and use of the same, conceded by the authors. The study was approved by the Research Ethics Committee of the Medical Sciences and Health Faculty of the Pontifical Catholic University of São Paulo. Statistical analysis 1 – Comparison of the make-up of the groups, in relation to sex and level of schooling (descriptive analysis with n and percentages). The scale is descriptive and visual, and consists of seven types of feces, being made up of images and their respective definitions: 2 – Description of the level of success and error in correlating images and their descriptions, using the absolute number of choices. 2 – Description of the level of success and error in correlating images and their descriptions, using the absolute number of choices. -Type 1: Separate hard lumps, like nuts (hard to pass). 3 – Description of the total of the subjects evaluated, and of each group, in relation to their success in matching the images and the descriptions for each of the types of feces. -Type 2: Sausage-shaped, but lumpy. -Type 3: Like a sausage but with cracks on its surface. -Type 4: Like an Italian sausage or snake, smooth and soft. 4 – Analysis of agreements between the groups (internal reliability and comparison between the groups). -Type 5: Soft blobs with clear cut edges (passed easily). -Type 6: Fluffy pieces with ragged edges, a mushy stool. -Type 7: Watery, no solid pieces. Entirely liquid. 5 – Analysis of the test-retest. Introduction All the patient has to do is select the type of feces which most resembles his or her own, according to their shape and consistency(21-22). At this point, the Portuguese version was applied in a pilot test by nine people (three doctors, three nurses and three patients) to detect possible comprehension problems, and, after some final adjustments, the definitive version of the scale was obtained. The BSFS was validated in its original version in English and has been translated, adapted and validated in Spanish. Due to its simplicity and efficacy, it has been incorporated into clinical practice for the evaluation of patients with irritable bowel syndrome, HIV-diarrhea related, and anal incontinence, besides being a much-used instrument in scientific studies related to these(21,23). To be used in Brazil, its translation, cultural adaptation and, consequently, validation are needed. Validation The definitive version (Figure 1) was applied to a sample of 165 healthcare professionals (80 doctors and 85 nurses) and 80 patients over 18 years of age, chosen at random from outpatient clinics at the Sorocaba Hospital Complex. All the evaluations were carried out in public health service facilities. The subjects were presented with a chart with the pictures representing the seven types of feces. Next, they were asked to match a text selected at random from the translated scale with one of the pictures. Ten per cent of the interviewees from each group (doctors, nurses and patients), chosen at random, were instructed to repeat the evaluation fifteen days later (re-test), to test its reproducibility. The objectives of the present study were the Bristol Stool Form Scale´s translation into Portuguese, cultural adaptation, and validation, so that it might be used in Brazil. www.eerp.usp.br/rlae 1 To a Brazilian, the word ‘nuts’ suggests walnuts. Translator’s note. 2 Small nuts from a type of coconut tree. Translator’s note. www.eerp.usp.br/rlae Figure 1 – The Bristol Stool Form Scale adapted to Portuguese Figure 1 – The Bristol Stool Form Scale adapted to Portuguese Table 1 – Characteristics referent to the sex of the respondents who made up the sample. Sorocaba, São Paulo, Brazil, 2011 Translation and cultural adaptation The size of the sample is calculated to estimate the concordance of 95%, with 5% precision and 5% significance. The quantitative variables were described with averages and standard deviation and, in the case of the qualitative variables, with absolute numbers and percentages. The reliability was analyzed by means of percentage and through a Kappa index. The comparative analyses were carried out through Kendall’s test of concordance for the percentage of agreement, and the Fleiss method was used for the Kappa values. The reproducibility after the period of 15 days was evaluated by means of the percentage of concordance and the Kappa index between the answers and a value below 0.05 was considered statistically significant. The translation into Portuguese was undertaken by two specialists in digestive function disorders who spoke both languages well. Once the scale had been translated, some descriptions were modified, so as to adapt it to the language and culture of Brazil, such as in the cases of the words “nuts” - which to an English-speaker suggests hazelnuts1 and was substituted with the term coquinhos2, which supplies an analogous image - and “sausage” which, for different types, was translated with the words “lingüiça” (German-style sausage) and “salsicha” (wiener), for better comprehension across Brazil. www.eerp.usp.br/rlae 586 Rev. Latino-Am. Enfermagem 2012 May.-June;20(3):583-89. Discussion Table 3 shows the concordance (as a percentage) between definitions and images in relation to the type of feces and the persons studied. The highest percentage for concordance was type 5 (100% for doctors and nurses, and 96. 3% for patients), and the lowest percentage was for types 6 and 7 for doctors (91.3%), type 3 for nurses (83.5%) and type 6 for patients (82.5%). The form of feces is modified in various intestinal illnesses such as, for example, infectious diarrheas, colitides, constipation, anal incontinence and irritable bowel syndrome(14-15). Its description in the medical history can be determinant in the diagnosis and medical supervision of these illnesses associated with other characteristics of bowel movements(14). Its use as a parameter for improvement or worsening can facilitate the healthcare professional’s dealing with these pathologies. This description varies according to the patient’s language and its interpretation also depends on the healthcare professional’s experience. Table 3 – Concordance between texts and images relating to feces type and persons studied Feces type Doctors Nurses Patients Total n= 80 n = 85 n = 80 n = 245 % % % % 1 98.8 98.8 95.0 97.6 2 92.5 87.1 87.5 89.0 3 92.5 83.5 85.0 86.9 4 97.5 100.0 95.0 97.6 5 100.0 100.0 96.3 98.8 6 91.3 87.1 82.5 86.9 7 91.3 85.9 83.8 86.9 Table 3 – Concordance between texts and images relating to feces type and persons studied It is, however, very difficult to obtain a precise description of the appearance and consistency of feces, which is why Heaton and Thompson created a visual and descriptive scale called “The Bristol Stool Form Scale”, which has been widely used, so as to guarantee that the patients describe their own fecal form precisely(15). The translation, validation and cultural adaptation of health evaluation instruments, as with this scale, is important for making it possible to carry out multinational studies or for comparing data from different Brazilian studies(17). The general Kappa index was 0.826, and the concordance summary for this index in each of the groups is presented in Table 4. The Bristol Scale has been recognized in the literature as a valuable instrument in the evaluation of intestinal illnesses, leading to a need for its translation and validation. Results The study population was divided in groups according to their profession or whether they were patients. The first group was made up of doctors, who were mainly male. The second group was made up of nurses, who were mainly female. The majority of patients were also female. As for level of education, none of the patients had been educated at university level. The patients were chosen at random too, irrespective of their illness or which specialist they were being treated by. None of the patients had any cognitive deficiency which could influence the results. Table 1 shows the numerical distribution of the sample, related to sex. Variables Doctors Nurses Patients n=80 % n=85 % n=80 % Sex (%) Male 47 58.75 5 5.88 31 38.75 Female 33 41.25 80 94.12 49 61.25 Table 2 shows the results for concordance or harmony (definition of text and image) for the validation text in all groups of individuals www.eerp.usp.br/rlae 587 Martinez AP, Azevedo GR. e of responses between texts and images in all the series. Sorocaba, São Paulo, Brasil, 2011 Texts Images Total 1 2 3 4 5 6 7 1 239 1 2 0 0 3 0 245 2 1 218 11 6 0 7 2 245 3 1 13 213 2 3 8 5 245 4 0 2 1 239 0 1 2 245 5 0 1 1 0 242 1 0 245 6 1 3 5 0 0 213 23 245 7 3 6 5 1 0 17 213 245 Total 245 244 238 248 245 250 245 1715 Table 2- Correspondence of responses between texts and images in all the series. Sorocaba, São Paulo, Brasil, 2011 www.eerp.usp.br/rlae References 1. Saad RJ, Rao SS, Koch KL, Kuo B, Parkman HP, McCallum RW, Sitrin MD, Wilding GE, Semler JR, Chey WD. Do stool form and frequency correlate with whole-gut and colonic transit? Am J Gastroenterol. 2010 Feb;105(2):403-11 2. Heaton KW, Radvan J, Cripps H, Mountford RA, Braddon FE, Hughes AO. Defecation frequency and timing, and stool form in the general population: a prospective study. GUT. 1992;33:818-24. 3. Degen LP, Phillps SF. How well does stool form reflect colonic transit? GUT. 1996;39:109-13. 4. López G, Antonio J. Incontinência anal/ Anal incontinence. Iatreia. 2002;15(3):190-9. 5. Engsbro AL, Simren M, Bytzer P. Short-term stability of subtypes in the irritable bowel syndrome: prospective evaluation using the Rome III classification. Aliment Pharmacol Ther. 2012 Feb;35(3):350-9. 6. Palsson OS, Baggish JS, Turner MJ, Whitehead WE. IBS Patients Show Frequent Fluctuations Between Loose/Watery and Hard/Lumpy Stools: Implications for Treatment. Am J Gastroenterol. 2012 Feb;107(2):286-95. 7. Domansky RC. Avaliação do hábito intestinal e fatores de risco para incontinência anal na população geral [tese de doutorado]. São Paulo (SP). São Paulo: Escola de Enfermagem da Universidade de São Paulo; 2009. 125 p. 8. Martínez ECJ, Flores CO, García JE, Velázquez PI, González MI, Márquez RA. Prevalence of urinary and anal incontinente in women from metropolitan area of Guadalajara. Ginecol Obstet Mex. 2006;74(6):300-5. 9. Bemonte-Monte C, Cervera-Servín JA, García-Vázquez JL. Fecal incontinence, occult lesions to the anal sphincter and related factors in primiparous patients without median routine episiotomy. Cir Cir. 2006;74(6):449-55. 10. Martínez P, Godínez MA, Rivas E, Herrera D, Barrón R. Anal incontinence caused by an obstetric trauma. Experience with the technique of overlapping sphincteroplasty. Ginecol Obstet Mes. 2006;74(8):418-23. 11. Azevedo GR, Negri AF. Dispositivo para Incontinência anal. Estima. (Sociedade Brasileira de Estomaterapia). 2007;5:40-2. 12. Alós R, Solana A, Ruiz MD, Moro D, García-Armengol J, Roig-Vila JV. Novel techniques in treatment of anal incontinence. Cir Española. 2005;78(3):41-9. 1. Saad RJ, Rao SS, Koch KL, Kuo B, Parkman HP, McCallum RW, Sitrin MD, Wilding GE, Semler JR, Chey WD. Do stool form and frequency correlate with whole-gut and colonic transit? Am J Gastroenterol. 2010 Feb;105(2):403-11 2. Heaton KW, Radvan J, Cripps H, Mountford RA, Braddon FE, Hughes AO. Defecation frequency and timing, and stool form in the general population: a prospective study. GUT. 1992;33:818-24. References Martínez P, Godínez MA, Rivas E, Herrera D, Barrón R. Anal incontinence caused by an obstetric trauma. Experience with the technique of overlapping sphincteroplasty. Ginecol Obstet Mes. 2006;74(8):418-23. 10. Martínez P, Godínez MA, Rivas E, Herrera D, Barrón R. Anal incontinence caused by an obstetric trauma. Experience with the technique of overlapping sphincteroplasty. Ginecol Obstet Mes. 2006;74(8):418-23. With regard to reliability, measured by the percentage of concordance between image and definition and the Kappa index, this study showed higher values than obtained by the Spanish group. 11. Azevedo GR, Negri AF. Dispositivo para Incontinência anal. Estima. (Sociedade Brasileira de Estomaterapia). 2007;5:40-2. 11. Azevedo GR, Negri AF. Dispositivo para Incontinência anal. Estima. (Sociedade Brasileira de Estomaterapia). 2007;5:40-2. Acknowledgments The sample made up of health care professionals was considered the most accurate in relation to the instrument’s being understood, as it was observed that there was a high correlation among the results, compared to a sample made up of non-health care respondents. As the instrument was devised to be used irrespective of educational level, it may be inferred that this difference between the groups is related to their knowledge about the area of health. The absence of any patients educated at university level reflected the reality of that hospital unit, which acts as a tertiary referral center for a low-income population of this region of the state of São Paulo. The rate of correlation obtained was higher than what the researchers had expected, which shows the care taken in the phases of translation and cultural adaptation. We thanks Jessica Clementino for their assistance on the data collection. Discussion This process has already been carried out for English and Spanish, allowing the opportunity for its use on an international level and for the comparison of data from various research centers in countries with the latter language(23). Table 4 - Concordance results for the global study. Sorocaba, Brazil, 2011 Variables Concordance p* Kappa p† Total group 89.5% <0.0001 0.826 <0.0001 Doctors 93.0% <0.0001 0.884 <0.0001 Nurses 91.1% <0.0001 0.825 <0.0001 Patients 84.6% <0.0001 0.773 <0.0001 Table 4 - Concordance results for the global study. Sorocaba, Brazil, 2011 ⃰ Kendall’s concordance test; †Fleiss method. This study was based on validation processes accepted worldwide and the methodology was divided into translation, cultural adaptation and validation. In this process, it was judged that cultural adaptation is the most important stage, as – apart from the language – the local culture can influence the form of expression. After translation, practically all of the items were modified to some degree so as to facilitate understanding. In the retesting phase there were no errors and, therefore, there was perfect concordance irrespective of the group. www.eerp.usp.br/rlae Rev. Latino-Am. Enfermagem 2012 May.-June;20(3):583-89. Rev. Latino-Am. Enfermagem 2012 May.-June;20(3):583-89. 588 References The comparison between the results obtained in this study and in that undertaken for the validation of the Spanish-language version differs significantly, despite similarity in relation to the number of persons studied(23). 3. Degen LP, Phillps SF. How well does stool form reflect colonic transit? GUT. 1996;39:109-13. 4. López G, Antonio J. Incontinência anal/ Anal incontinence. Iatreia. 2002;15(3):190-9. The samples are similar in relation to demographic data (being aged over 18, predominance of females). In relation to the type of feces most accurately identified, differences may be observed as much in the total of evaluations as in each group, although a high percentage of correct identification was observed in both studies. The types of feces with the greatest degree of correct or incorrect identification among the three groups were different in the two studies. In the Brazilian study, the type of feces most correctly identified was type 5, and the least correctly identified was type 6; while in the Spanish study, it may be observed that the most correctly identified type was type 4 and the least correctly identified was type 5. This fact is attributed to the difference in language and expressions used for the description of feces in professional and lay circles. 5. Engsbro AL, Simren M, Bytzer P. Short-term stability of subtypes in the irritable bowel syndrome: prospective evaluation using the Rome III classification. Aliment Pharmacol Ther. 2012 Feb;35(3):350-9. 5. Engsbro AL, Simren M, Bytzer P. Short-term stability of subtypes in the irritable bowel syndrome: prospective evaluation using the Rome III classification. Aliment Pharmacol Ther. 2012 Feb;35(3):350-9. In relation to the three groups studied, the Brazilian doctors obtained a higher rate of correct identification for type 5, whereas the Spanish doctors were more likely to correctly identify types 2 and 7. The Brazilian nurses were most likely to identify types 4 and 5, while the Spanish nurses tended to correctly identify type 6. Lastly, the Brazilian patients correctly identified type 6 most, whereas the Spanish patients correctly identified type 4 most. 9. Bemonte-Monte C, Cervera-Servín JA, García-Vázquez JL. Fecal incontinence, occult lesions to the anal sphincter and related factors in primiparous patients without median routine episiotomy. Cir Cir. 2006;74(6):449-55. 9. Bemonte-Monte C, Cervera-Servín JA, García-Vázquez JL. Fecal incontinence, occult lesions to the anal sphincter and related factors in primiparous patients without median routine episiotomy. Cir Cir. 2006;74(6):449-55. 10. Received: Ago. 30th 2011 Accepted: Apr. 3rd 2012 Conclusion 12. Alós R, Solana A, Ruiz MD, Moro D, García-Armengol J, Roig-Vila JV. Novel techniques in treatment of anal incontinence. Cir Española. 2005;78(3):41-9. 12. Alós R, Solana A, Ruiz MD, Moro D, García-Armengol J, Roig-Vila JV. Novel techniques in treatment of anal incontinence. Cir Española. 2005;78(3):41-9. The Bristol Stool Scale for the Consistency of Feces, translated and validated for Portuguese for use in Brazil, showed high reliability, indicating its usefulness in clinical practice for the purpose for which it was devised. 13. Lumi CM, Munoz JP, La Rosa L. Neuromodulación sacra para el tratamiento de la incontinencia anal. Experiencia piloto prospectiva en Argentina/ Sacral neuromodulation www.eerp.usp.br/rlae 589 Martinez AP, Azevedo GR. in fecal incontinente. The pilot prospective experience in Argentina. Acta Gastroenterol Latinoam. 2007;37(1): 29-36. in fecal incontinente. The pilot prospective experience in Argentina. Acta Gastroenterol Latinoam. 2007;37(1): 29-36. 14. Whitehead WE, Wald A, Norton NJ. Treatment options for fecal incontinence. Dis Colon & Rectum. 2001;44(1):131-42. 15. Yilmaz S, Dursun M, Ertem M, Canoruc F, Turhanoglu A. The epidemiological aspects of irritable bowel syndrome in Southeastern Anatolia: a stratified randomised community-based study. Int J Clin Pract. 2005;59:361-9. 16. Longstreth GF, Thompson WG, Chey WD, Houghton LA, Mearin F, Spiller RC et al. Functional Bowel Disorders. Gastroenterology. 2006;130:1480-91. 17. Heaton KW, Ghosh S, Braddon. How bad are the symptoms and bowel dysfunction of patients with the irritable bowel syndrome? A prospective, controlled study with emphasis on stool form. GUT. 1991;32:73-9. 17. Heaton KW, Ghosh S, Braddon. How bad are the symptoms and bowel dysfunction of patients with the irritable bowel syndrome? A prospective, controlled study with emphasis on stool form. GUT. 1991;32:73-9. 17. Heaton KW, Ghosh S, Braddon. How bad are the symptoms and bowel dysfunction of patients with the irritable bowel syndrome? A prospective, controlled study with emphasis on stool form. GUT. 1991;32:73-9. 18. Hearing SD, Thomas L, Heaton K, Hunt L. Effect of cholecystectomy on bowel function: a prospective, controlled study. GUT. 1999;45:889-94. 19. Macmillan AK, Merrie AE, Marshall RJ, Parry BR. Design and validation of a comprehensive fecal incontinence questionnaire. Dis Colon Rectum. 2008;51(10):1502-22. 20. Yusuf SA, Jorge JM, Habr-Gama A, Kiss DR, Gama J. Evaluation of quality of life in anal incontinence: validation of the questionnaire FIDL (Fecal Incontinence Quality of Life). Arq Gastroenterol. 2004;41(3):202-8. 19. Macmillan AK, Merrie AE, Marshall RJ, Parry BR. Design and validation of a comprehensive fecal incontinence questionnaire. Dis Colon Rectum. 2008;51(10):1502-22. www.eerp.usp.br/rlae Conclusion 20. Yusuf SA, Jorge JM, Habr-Gama A, Kiss DR, Gama J. Evaluation of quality of life in anal incontinence: validation of the questionnaire FIDL (Fecal Incontinence Quality of Life). Arq Gastroenterol. 2004;41(3):202-8. 20. Yusuf SA, Jorge JM, Habr-Gama A, Kiss DR, Gama J. Evaluation of quality of life in anal incontinence: validation of the questionnaire FIDL (Fecal Incontinence Quality of Life). Arq Gastroenterol. 2004;41(3):202-8. 21. Peréz MM, Martínez AB. The Bristol scale - a useful system to assess stool form? Rev Esp Enferm Dig. 2009;101(5):305-11. 22. Riegler G, Esposito I. Bristol scale stool form. A still valid help in medical practice and clinical research. Tech Coloproctol. 2001;5(3):163-4. 22. Riegler G, Esposito I. Bristol scale stool form. A still valid help in medical practice and clinical research. Tech Coloproctol. 2001;5(3):163-4. 23. Parés D, Comas M, Dorcaratto D, Araujo MI, Vial M, Bohle B, et al. Adaptation and validation of the Bristol scale stool form translated into the Spanish language among health professionals and patients. Rev Esp Enferm Dig. 2009;101(5):312-6. 23. Parés D, Comas M, Dorcaratto D, Araujo MI, Vial M, Bohle B, et al. Adaptation and validation of the Bristol scale stool form translated into the Spanish language among health professionals and patients. Rev Esp Enferm Dig. 2009;101(5):312-6. Received: Ago. 30th 2011 Accepted: Apr. 3rd 2012 www.eerp.usp.br/rlae www.eerp.usp.br/rlae
https://openalex.org/W4235163771	https://zenodo.org/records/2440717/files/article.pdf	English	null	RICHARD GARNETT	Library/The library	1,906	public-domain	3,345	1 Eariv Spanish-American Printing,' voL i, pp. 139-1+6; cOn the De Missione Legatorum Japonensium' (Macao, 1590), voL ii, pp. 172-182; 'Some Notes on Ancient Writing and Writing Materials,' vol. iv, pp. 225-235. RICHARD GARNETT. RICHARD GARNETT. 246 of 1903, references, evidently made in all good faith, to the ' Ingoldsby Legends' and the ' Bab Ballads." Neither Barham nor Mr. Gilbert, two writers of equally intoxicating brilliance of non- sense, ever wasted thought on an idea deeper than a pin would scratch. Dr. Garnett's book would have been publicly burned by any government in any age [which had really taken seriously the be- liefs which it professed in theory. It is a text-book of intellectual anarchy; it is loaded with symbols of revolution; but the air of our century is proof against it, it will never go off with the least damage to our idols. ARTHUR SYMONS. BEFORE an attempt is made to supple- \ ment in any other way what has already .been written about Dr. Garnett from various points of view, a special word J of gratitude has to be said for the help and encouragement which he constantly gave to this magazine. Not only did he lend it countenance and authority by acting as one of its consulting editors, but he found time to write for it three very character- istic articles; he was quick to express his pleasure at any contribution of unusual interest which ap- peared in its pages, and he frequently offered sug- gestions of subjects which might be written on, RICHARD GARNETT. 247 though the suggestions too often required a learn- ing akin to his own to carry them out. Two quite recent instances of his kindness may be specially mentioned. In the conviction, in which he was probably right, that it is only by an abundance of pictures that a bibliographical magazine can attain a satisfactory circulation, he offered himself to sub- scribe to provide more illustrations, an offer which could not be entertained, but is very gratefully re- membered. Again, only just before last Christmas, when asked for his opinion on the principles by which municipal librarians should be guided in their book-purchases, he wrote for our January number no mere hasty expression of his ideas, but a considered and carerally thought out view of the whole matter, which in its mellow reasonableness seemed to sum up almost all that could be said on the subject. RICHARD GARNETT. To lose a friend such as this is a griev- ous loss indeed to those who are carrying on a maga- zine to work a little closer towards ideals for which there are few enthusiasts, and while the editors of * The Library' have individually many other reasons for lamenting Dr. Garnett's death, the loss of his help and sympathy in their difficult task comes speci- ally home to them. though the suggestions too often required a learn- ing akin to his own to carry them out. Two quite recent instances of his kindness may be specially mentioned. In the conviction, in which he was probably right, that it is only by an abundance of pictures that a bibliographical magazine can attain a satisfactory circulation, he offered himself to sub- scribe to provide more illustrations, an offer which could not be entertained, but is very gratefully re- membered. Again, only just before last Christmas, when asked for his opinion on the principles by which municipal librarians should be guided in their book-purchases, he wrote for our January number no mere hasty expression of his ideas, but a considered and carerally thought out view of the whole matter, which in its mellow reasonableness seemed to sum up almost all that could be said on the subject. To lose a friend such as this is a griev- ous loss indeed to those who are carrying on a maga- zine to work a little closer towards ideals for which there are few enthusiasts, and while the editors of * The Library' have individually many other reasons for lamenting Dr. Garnett's death, the loss of his help and sympathy in their difficult task comes speci- ally home to them. Besides the notes here printed, and the one or two obituaries in the daily papers which appeared to be written from personal knowledge and with personal feeling, some very interesting tributes by Mr. F, M. Hueffer, and by three of Dr. Garnett's women friends—Miss Beatrice Harraden, Miss Agnes Adams, and Miss Alice Zimmern—have 7 been published in the June number of' The Book- i7 RICHARD GARNETT. RICHARD GARNETT. 248 man.' Necessarily slight as are all these contribu- tions, they yet point to the possibility that by co-operation some sketch of this unique personality might be evolved which should be different from the ordinary biography, so unflinching in its tedious detail, so swollen with letters no longer interesting, which passes through the circulating libraries, and is dead within the year. ' Every night of his life he went to the pillar-box at the top of the street,' Miss Adam writes in c The Bookman.' * He had an enormous correspondence, and insisted on posting his own letters. His friends who lived near used to say they knew it was ten minutes to twelve when a slow, hesitating step passed their windows. Latterly he leaned heavily on his stick—the stick that used to be Ford Madox Brown's.' It may seem capricious to pick out the record of so small a characteristic as this, and make much of it, but to at least one reader these few sentences brought back the living personality, and with it the sense of individual loss by his death, more than all the columns of formal obituaries. Dr. Garnett was proud of being a fellow-towns- man of Dr. Johnson, and it seems probable that, if his memory endures, it will be, as in the case of Johnson, less for what he wrote than for what he was. There is, indeed, a rather tempting parallel- ism in the a<5tual literary output of the two men. Against Johnson's Dictionary we may set Dr. Garnett's share in the British Museum Catalogue. The biographies of Milton, Carlyle and Emerson need not fear comparison with the once famous ' Lives of the Poets.' In poetry, * The Vanity of RICHARD GARNETT. 249 Human Wishes' and' London,' adaptations though they be, have more individuality and a stronger ring than Dr. Garnctt's graceful verses; but if the scale turns against him at this weighing, the author of' The Twilight of the Gods' and « De Flagello Myrteo' might well hope to make a corresponding recovery when these are contrasted with ' Rasselas' and ' Irene.' The epigrams in these two books of Dr. Garnett's are indeed as quotable as the best things in Boswell, and are only too likely to prove the chief memorials of his powers as a talker. RICHARD GARNETT. The revelation of the authorship of ' De Flagello Myrteo' within a few days of the writer's death may suggest a possible need of waiting a little before the chorus of admiration with which it has been greeted can be accepted as a final verdict. On the other hand, 'The Twilight of the Gods,' which has been enthusiastically praised in so many notices, has won its way to this favour after a most chilling first reception. Amid the gentle regrets now uttered that Dr. Garnett wasted on mere literary and official taskwork time which might have been devoted to producing more such books as this, it is interesting to remember that these intensely characteristic stories found their way to the remainder market with disconcerting rapidity. It would certainly have pleased their author had he ever known that it was the loyal appreciation of his staff at the British Museum that helped them to emerge very quickly from this undignified position. When the remainder-man's catalogue was received in the Printed Book Department, it was promptly taken the round of the different rooms, and the resultant RICHARD GARNETT. »$o order for twenty-five copies so surprised the vendor that he refused to execute it except at the very ad- vance of price it was partly intended to bring about While Dr. Garnett took the cold reception of his stories with cheerful philosophy, he never pre- tended to be indifferent to criticism. One of his most characteristic utterances, both for its feline reference and for the genial assurance with which it was spoken, was provoked by overhearing a doubt expressed as to whether authors who knew their own worth really care for the praise of critics. • Do'poets like praise'? * Do cats like cream'? was his comment, and as he made it there was a delight- ful beam of amusement in his face, which is good to remember. Before very long the cream for his own ' Twilight of the Gods' came to him in a form which he greatly appreciated—that of a warm letter of thanks and praise from the late Lord Lytton, with whom he had, I believe, no personal acquaintance. RICHARD GARNETT. It was typical of his modesty that this touch of sympathy from a single distinguished reader gave him as much pleasure as if he himself had been a raw beginner, and that when he had his own copy of the book bound a pocket was made in one of the covers, and Lord Lytton's letter placed in it. It is pleasant to know that the success of' The Twilight of the Gods' was not wholly posthumous, and that Dr. Garnett lived to see it pass into a second edition. That it was not more quickly appreciated is perhaps no real matter for regret. Not to succeed too rapidly is the surest of all safe- guards for artistic integrity. No man of letters of RICHARD GARNETT. RICHARD GARNETT. J51 Dr. Garnett's generous nature and limited official income can ever be quite indifferent to the tempta- tions offered by publishers, and had he been bom- barded with applications for more such stories, the pure gold which he extracted from this vein of fancy might have been alloyed with metal less truly characteristic. As it was, he found later on a new literary diversion in the aphorisms of the * De Flagello Myrteo,' and not many authors have more than two absolutely original books to their credit. One of the many reasons for hoping that some authentic memoir of Dr. Garnett may be written is that there are already signs that without some really discriminating record his reputation may at- tract to itself many of the Joe-Millerisms of librarian- ship. There is much to be grateful for in Mr. Hueffer's article in ' The Bookman,' but it may be wished that he had abstained from the obviously imaginative story of the engraving of a Mero- vingian buckle, for which Dr. Garnett directed him to about page 274 of the tenth volume of a work to be found on the fourth row of the fifth shelf (sic)> on the right from the entrance to the Reading Room. Dr. Garnett's feats of memory were too really extraordinary to need embellish- ment of this kind, and they become much more human and interesting when they are traced to their source instead of being treated as semi-magi- cal prodigies. Even his acquaintance with the names of the Derby winners, by which he amused so many of his friends, was not quite fortuitous, for by a whimsical survival from the manners of an . older generation, he was for many years an amused RICHARD GARNETT. 252 subscriber to a half-crown sweepstake on the Derby. There is even a tradition that he was once the winner of it, and used his gains to present all his innocent fellow gamblers with flowers. g One may be more more grateful to Mr. Hueffer when he speaks of Dr. RICHARD GARNETT. Garnett's ' enigmatic and very wonderful presence.' Gentle, easy of approach, and entirely unassuming as he was, it may be doubted whether any man every ventured to take a liberty with him; and for myself, to the end of his days, I paid him the unpleasant compliment of stammering more consistently when talking with him than with any other person in the world with whom I was on the same terms. But however long the query took to explain it was always heard with the same benign smile from the spedacled eyes, and when the end came there was usually some modest disclaimer and then a stream of suggestions, not always precisely to the point, but almost always opening up new vistas and pointing out connec- tions I had never suspected. When time served it was worth while to venture on a story for the sake of the better one with which he was sure to cap it, and which would be made more humorous if it hap- pened to bring in his accustomed pronunciation of the vowel u as short as possible. An anecdote of a biitcher, who exclaimed when he had slain a re- fractory sheep, c I've conciliated that one, anyway,' was the only story I remember to have heard him tell twice, and for the sake of the * but' in * btit- cher' I would gladly have heard it often. A h i h I fi k D G h At the time when I first knew Dr. Garnett he had already been a third of a century in the British RICHARD GARNETT. RICHARD GARNETT. 253 Museum. Mr. Fortescue's recollections go back thirteen years earlier than this. Those who remem- ber him in his freshman days are now sadly few. He was a * tall, lanky youth,' one of them tells me, always reading, and reputed to possess the gift of eating his lunch, going on with his work, and skimming the 'Athenaeum' all at the same time. The newspaper he seems to have read only in the street as he came down to the Museum of a morn- ing, holding it up before him with one hand, while he held bag and umbrella in the other. Perhaps it was this early habit of reading as he walked which accounted for his rather peculiar gait. When he entered the Museum he probably felt himself a Croesus, for the pay of assistants had just been altered from a daily or weekly wage to an annual salary of £130; and for a lad of sixteen to be able to start on a salary of £ 130 a year was no more common in those days than now. While his appointment illustrates the occasional advantages of the old system of patronage, that of his immedi- ate senior can hardly be quoted on the same side, for the legend runs that he had obtained his post as the only way of acknowledging his services in bringing over to the Queen some Barbary horses as a present from the Sultan of Morocco, and his stay at the Museum was neither very long nor very successful. Two places higher up, and in receipt of about £45 a year more salary, was Coventry Pat- more, with whom the young Garnett, already thinking of poetry, associated more than with any- one else. Two places below him was E. A. Roy, who had entered the Museum some ten years earlier RICHARD GARNETT. 254 in an inferior grade, and whose merits, as recounted by Panizzi before the Commission of 1850 formed one of the levers by which the position of the staflF had been recently improved. It seems that he knew French and Italian fluently, had a catalogu- ing acquaintance with German and Spanish, and could even transcribe Arabic. RICHARD GARNETT. Yet ' this young man,' complained Panizzi, ' receives twenty-five shillings a week, and if he catches a cold and is absent he gets nothing during his absence.' Panizzi was a good friend to Garnett, but his admiration for Mr. Roy's merits caused him in 1856 to pro- mote the latter, as the older man, over Garnett's head. Later on he was passed over again, much less justifiably, in favour of Ralston, the well-known Russian scholar, and on this occasion resented the slight so much that until an explanation was offered him he wished to be allowed to resign. But despite this share of official troubles it cannot be doubted that his life at the British Museum, from first to last, was a very happy one. He loved it so much that he very seldom took his full allowance of holi- days, and he knew nothing of Museum headaches. The atmosphere produced in winter by its hot- water pipes he used to compare for its warmth and dryness to the air of Egypt, and he seems to have found it sufficiently bracing to keep him in con- stant health. Would that more of his old colleagues were alive to tell us what he was like in these early days, when he sat first in the King's Library, and afterwards in the Arched Room, and catalogued the old books in their order shelf after shelf! I l i f d b id In conclusion, a few words may be said as to RICHARD GARNETT. 255 Dr. Gamett's connection with the Bibliographical Society of which, as Mr. Tedder has already noted, he was one of the founders. Despite the obstacles which his duties as one of the Resident Officers at the British Museum threw in his way, he was a frequent attendant at its meetings, and with the courtesy which never failed him in his intercourse with his colleagues, made a special point of being present when anyone from the Museum was reading a paper. It is no exaggeration to say that his pre- sence by itself sufficed to make a meeting a success, for he chatted delightfully with everyone who went up to him, and could always be relied on for an interesting speech. He had no oratorical gift, and when not entirely at his ease was far from an effective speaker. RICHARD GARNETT. But here he was among friends; the small audience and the room both suited him, and after some deprecatory remarks as to his own ignorance, he would settle down to play round any subject on which he was asked to speak with a wealth of learning and fancy which was quite de- lightful. It was the charm of his suggestions that they were almost always far fetched and yet tri- umphantly relevant. I remember that in speaking of English books printed abroad he pointed out as an example of the haps by which the spread of printing was ruled, that if England had been abso- lutely supreme at sea, when, under Elizabeth, she was supporting the claim of Don Antonio to the throne of Portugal, the Portuguese islands off the coast of Africa would no doubt have been captured, and proclamations have been issued there in his interest, and thus the appearance of printing in RICHARD GARNETT. 256 Africa would have been accelerated by at least fifty years. The speculation was quite to die point, but it may be doubted whether any other human being than Dr. Garnett, who was always fascinated by the might-have-beens of history, would have thought of it. During the two years that he held the office of President he worked really hard for the Society, and was always ready to promote its interests. He frequently also acted as deputy to his immediate suc- cessor, the Earl of Crawford, during the latter's absence from England, and to do this on one occa- sion came straight to the Society's rooms after com- pleting his last day's work at the British Museum. With some misgivings, lest he might already be a little overwrought, it was hastily resolved that something should be said as to the affection and esteem which he carried with him in his retire- ment. Dr. Garnett was taken by surprise, but the mingled dignity and cheerfulness of his brief reply were characteristic of his whole attitude to life. ALFRED W. POLLARD.
https://openalex.org/W4384432248	https://journal.unindra.ac.id/index.php/teraputik/article/download/1561/pdf	Indonesian	null	Konsep Bimbingan Dan Konseling Islam Di Dalam Al-Qur’an Surah Luqman Ayat 17-19	Teraputik : jurnal bimbingan dan konseling	2,023	cc-by	3,618	How to cite: Andri Oktaviandi1, & Irman2. (2023). Konsep Bimbingan Dan Konseling Islam Di Dalam Al-Qur’an Surah Luqman Ayat 17-19. Teraputik: Jurnal Bimbingan dan Konseling, 7(1), pp. 40–45. DOI: 10.26539/ teraputik.631561 This is an open access article distributed under the Creative Commons 4.0 Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2023, Andri Oktaviandi, & Irman (s). Original Article Original Article Konsep Bimbingan Dan Konseling Islam Di Dalam Al-Qur’an Surah Luqman Ayat 17-19 Andri Oktaviandi1), Irman2 SDIT An Nahl Kab. Lima Puluh Kota, UIN Batusangkar )Alamat korespondensi: Jl. Jenderal Sudirman No.137, Lima Kaum, Kab. Tanah Datar, Sumatera Barat, 27217, Indonesia; E-mail: andrioktaviandi90@gmail.com Andri Oktaviandi1), Irman2 SDIT An Nahl Kab. Lima Puluh Kota, UIN Batusangkar )Alamat korespondensi: Jl. Jenderal Sudirman No.137, Lima Kaum, Kab. Tanah Datar, Sumatera Barat, 27217, Indonesia; E-mail: andrioktaviandi90@gmail.com Article History: Received: 15/04/2023; Revised: 24/05/2023; Accepted: 12/06/2023; Published: 30/06/2023 Article History: Received: 15/04/2023; Revised: 24/05/2023; Accepted: 12/06/2023; Published: 30/06/2023 Abstract: This study aims to describe how the concept of Islamic guidance and counseling is contained in surah Luqman in verses 17-19, namely the approach, principles, and fields of service. This research is a type of library research (Library Research), using content analysis method. The results of this study indicate that in Surah Luqman verses 17-19. the keywords bil-ma'ruufi do good or in a good way and approach al-hikmah in a wise way. There are several principles, including worship guidance (developing a relationship with Allah through acts of worship) found in verse 17 "aqimish-sholaat" (carry out prayers) and carry out "amar ma'ruf and nahi munkar". Then in the 18th verse there is the word fil ardimaraaa (don't walk on earth arrogantly or arrogantly) and in the 19th verse there is the word "waksidek fii maisika wakdud minsautiha" (simple in walking and soften your voice). Keywords: The Concept of Islamic Guidance and Counseling and Surah Luqman Verses 17-19 Abstrak: Penelitian ini bertujuan untuk mendeskripsikan bagaimana konsep bimbingan dan konseling Islam yang terdapat di dalam surah Luqman pada ayat 17-19 yaitu adanya pendekatan, asas, dan bidang layanan. Penelitian ini merupakan jenis penelitian pustaka (Library Reseach), dengan menggunakan metode analisis isi. Hasil penelitian ini menunjukan bahwa dalam Surah Luqman ayat 17-19. kata kunci bil-ma‟ruufi mengerjakan yang baik atau dengan cara yang baik dan pendekatan al-hikmah dengan cara yang bijaksan. ada beberapa asas, diantaranya ada bimbingan ibadah (mengembangkan hubungan kepada Allah melalui amal ibadah) terdapat pada ayat ke 17 “aqimish-sholaat” (laksanakanlah shalat) serta melaksanakan “amar ma‟ruf dan nahi mungkar”. Kemudian pada ayat ke 18 ada kata fil ardi marahaa (jangan berjalan dimuka bumi dengan angkuh atau sombong) dan pada ayat ke 19 ada kata “waksidek fii maisika wakdud minsautiha” (sederhana dalam berjalan dan lunakkan suaramu). Kata Kunci: Konsep Bimbingan dan Konseling Islam dan Surah Luqman Ayat 17-19 Volume 7, Number 1, June, (2023), pp. 40-45 ISSN 2580-2046 (Print) \| ISSN 2580-2054 (Electronic) Pusat Kajian Penelitian dan Pengembangan Bimbingan dan Konseling DOI: 10.26539/ teraputik.631561 Open Access \| Url: https://journal.unindra.ac.id/index.php/teraputik/index Pendahuluan Cakupan bimbingan dan konseling itu sangat luas, salah satunya ada bimbingan dan konseling berbasis islami, yang tidak sepenuhnya dikenal oleh masyarakat luas, mereka hanya mengenali bimbingan dan konseling biasa yang nantinya akan memasuki dunia pendidikan, semata-mata hanya menjadi guru yang ditakuti oleh siswa, sering kali disebut sebagai polisi sekolah yang sangat menakutkan, seperti pada saat dipanggil untuk memasuki ruangan BK. Apabila pikiran seseorang itu bisa dikontrol dengan baik terhadap adanya bimbingan dan konseling, dan memiliki pandangan yang positif, tentunya akan menghasilkan kedamaian serta kebahagiaan antara satu dengan yang lainnya. Begitu juga dengan bimbingan dan konseling Islam, walaupun nama ini tidak terlalu dikenal secara menyeluruh, namun apabila mereka menyadari, ternyata konseling islami merupakan hal yang sangat dibutuhkan oleh semua orang dimulai dari anak-anak sampai lansia. Bimbingan islami yang dilakukan oleh orang yang sudah profesional akan sangat bermanfaat terhadap orang lain, apabila semua layanan itu berdasarkan ketentuan Allah SWT dan mengikuti petunjuk Al-Qur’an. g p j Namun seringkali bimbingan dan konseling berbasis Islam ini dianggap sebelah mata oleh sebagian manusia. Apabila pada saat menyebutkan nama bimbingan dan konseling Islam, 40 41 Hubungan Pola Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA Cibinong 41 Hubungan Pola Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademi Cibinong a Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA seringkali masyarakat tidak bisa membedakan bimbingan konseling Islam dengan bimbingan konseling biasa atau secara umum. Melihat dari sudut pandang yang ada, penulis sekilas bisa menyimpulkan bahwa prodi bimbingan dan konseling Islam itu belum diangap penting dan belum terlalu kenal. Alasan lain, karena sudah ada penelitian sebelumnya, melakukan bertemu secara langsung kepada manusia. Sedangkan diangkatan saya hampir semuanya menggunakan penelitian berbasis lapangan. Jadi untuk bisa mencari referensi yang lebih mendalami lagi, saya memutuskan untuk melakukan library riseach. y Sehingga dengan adanya argumen ini, bisa mewakili harapan-harapan baik, serta menjadi ladang pahala amal jariyah untuk bisa mengenalkan nilai- nilai Islam dan bisa diberikan kepada semua orang, dengan adanya permasalahan ini sehingga dilakukan adanya penelitian lebih lanjut. Salah satunya dengan melalui Al-Qur’an Surah Luqman Ayat 12-19 ini sangat istimewa, bisa menjadi contoh teladan yang baik terhadap konselor ataupun kebaikan klien. Pendahuluan Maka dari itu bimbingan dan konseling Islam tidak bisa lepas dari Al- Qur’an sebagai petunjuk bagi umat Islam yang bisa dijadikan sumber hukum dalam pelaksanaan layanan bimbingan yang dapat dilakukan dengan cara yang baik kepada manusia, sebagaimana salah satu contoh khususnya pada Al-Qur’an Surah Yunus ayat 57. y p y Artinya: “Hai manusia, sesungguhnya telah datang kepadamu pelajaran dari Tuhanmu dan penyembuh bagi penyakit-penyakit (yang berada) dalam dada dan petunjuk serta rahmat bagi orang-orang yang beriman”. (Yunus:57).1 g g g y g ( ) Bahwa Al-Qur’an adalah salah satu rujukan yang dibutuhkan oleh manusia pada saat mengalami masalah. Al-Qur’an ini adalah sebagai jawaban dari pertanyaan-pertanyaan dari berbagai macam permasalahan yang dihadapi oleh manusia dan dapat dijadikan landasan dalam memberikan bimbingan kepada manusia. Al-Qur’an sebagai dasar dari pelaksanaan layanan bimbingan dan konseling Islam juga sudah menyediakan garis-garis besar dalam praktik layanan, seperti pendekatan, bidang bimbingan, dan asas-asas yang terdapat dalam Al-Qur’an Surah Luqman ayat 12-19. TERAPUTIK \| Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 ISSN 2580-2046 (Print) \| ISSN 2580-2054 (Electronic) \| Url: https://journal.unindra.ac.id/index.php/teraputik/index Tafsir Surah Luqman Ayat 12-19 Tafsir Surah Luqman Ayat 12-19 Artinya: “dan Sesungguhnya telah Kami berikan hikmat kepada Luqman, Yaitu: "Bersyukurlah kepada Allah. dan Barangsiapa yang bersyukur (kepada Allah), Maka Sesungguhnya ia bersyukur untuk dirinya sendiri; dan Barangsiapa yang tidak bersyukur, Maka Sesungguhnya Allah Maha Kaya lagi Maha Terpuji". (QS. Luqman:12) Luqman ialah seorang yang arif bijaksana. Luqman telah diberi hikmah oleh Allah, yakni perintah bersyukur kepada Allah atas nikmat yang tidak terhingga banyaknya. Seorang dikatakan arif bijaksana apabila ia berlaku adil yang mutlak, meskipun terhadap keluarga bahkan diri sendiri. Ayat ini menerangkan bahwa Allah SWT menganugerahkan kepada Luqman hikmah: yaitu ilmu agama, akal pikiran, benar dan bijak dalam berucap, yang benar sehingga menyampaikannya kepada kebahagiaan abadi, sambil menjelaskan beberapa butir hikmah yang pernah beliau sampaikan kepada anakanya. Ayat di atas menyatakan: dan sesungguhnya Kami Yang Maha Perkasa dan Bijaksana telah menganugerahkan Luqman, yaitu: “bersyukur kepada Allah. Kata hikmah telah disinggung makna dasarnya ketika menafsirkan ayat dua di atas, Qurais Shihab bahwa para ulama mengajukan aneka keterangan tentang makna hikmah. Antara lain bahwa hikmah berarti “mengetahui yang paling utama dari segala sesuatu, baik pengetahuan, maupun perbuatan. Ia adalah ilmu amaliah, ia adalah ilmu yang didukung oleh amal yang tepat dan didukung oleh ilmu. Artinya: ”dan (ingatlah) ketika Luqman berkata kepada anaknya, di waktu ia memberi pelajaran kepadanya: "Hai anakku, janganlah kamu mempersekutukan Allah, Sesungguhnya mempersekutukan (Allah) adalah benar-benar kezaliman yang besar".(QS.Luqman:13). Kata ya‟izhuhu dengan arti memberi pelajaran kepadanya, yang terdapat pada bagian ayat 13. Ya‟izhuhu terambil dari kata wa‟zh yaitu nasihat yang menyangkut berbagai kebajikan dengan cara menyentuh hati. Penyebutan kata ini sesudah kata dia berkata untuk memberi gambaran tentang bagaimana pendekatan itu beliau sampaikan, yakni dengan cara tidak membentak, tetapi dengan penuh kasih sayang, sebagaimana dipahami dari panggilan mesrahnya kepada seorang anak. Ayat di atas dilukiskannya pengalaman hikmah itu oleh Luqman yaitu salah satu bentuk pelestariannya terhadap anaknya, ayat ini berbunyi: “dan ingatlah ketika Luqman berkata kepada anaknya dalam keadaan dia dari saat kesaat menasihatinya bahwa wahai anakku sayang!. Setelah ayat yang lalu menguraikan hikmah yang dianugerahi kepada Luqman yang intinya adalah kesyukuran kepada Allah, dan yang tercermin pada pengenalan hikmah itu oleh Luqman. Luqman menjelaskan kepada anaknya, bahwa perbuatan syirik itu merupakan kedzaliman yang benar. Syirik dinamakan perbuatan zalim, karena perbuatan syirik berarti meletakan sesuatu tidak pada tempatnya, dan dikatakan sebagai dosa besar, karena perbuatan itu menyamakan kedudukan Tuhan. TERAPUTIK Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 Pusat Kajian Penelitian dan Pengembangan Bimbingan dan Konseling, Unindra \| DOI: 10.26539/ teraputik.631561 TERAPUTIK Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 Metode Penelitian yang penulis lakukan ini merupakan jenis penelitian kepustakaan (Library Reseach). Studi literatur atau kepustakaan adalah kegiatan untuk menghimpun informasi yang relevan dengan topik atau masalah yang menjadi obyek penelitian. Informasi tersebut dapat diperoleh dari buku-buku, karya ilmiah, tesis, disertasi internet dan sumber-sumber lain.1 Jenis penelitian yang berusaha menghimpun data penelitian dari berbagai sumber informasi seperti dari buku-buku, jurnal yang bersifat tidak melalui wawancara dan sebagainnya dengan fokus pada penelaahan, yakni penelitian yang dilakukan dengan cara mengumpulkan data yang ada dipustaka yang ada kaitannya dengan fokus masalah penelitian dan pokok- pokok pertanyaan penelitian. Penelitian kepustakaan merupakan suatu jenis penelitian yang digunakan dalam pengumpulan informasi dan data secara mendalam melalui literatur, buku, catatan, majalah dan referensi lainnya, untuk mendapatkan jawaban dan landasan teori mengenai masalah yang akan diteliti.2 Menurut ahli lain studi kepustakaan merupakan kajian teoritis, referensi serta literatur ilmiah lainnya yang berkaitan dengan budaya, nilai dan norma yang berkembang pada situasi sosial yang diteliti. Penelitian ini merupakan penelitian pustaka dimana data-data yang diperoleh bersumber dari buku-buku, jurnal ilmiah, koran, majalah dan dokumen lainnya. Sedangkan metode yang penulis gunakan adalah metode analisis isi yang bertujuan untuk menganalisis yang digunakan untuk mendapatkan inferensi yang valid dan dapat diteliti ulang berdasarkan konteksnya. Penelitian yang dilakukan ini untuk mengetahui konsep-konsep bimbingan dan konseling Islam yang terdapat dalam Surah Luqman ayat 12-19. Penjelasan di atas maka dapat disimpulkan bahwa penelitian studi pustaka ini sangat berbeda dengan metode lainnya. Sedangkan kepustakaan ini lebih bersifat menganalisis isi dari buku- buku yang sumber referansinya tidak melalui lapangan secara langsung. Dalam hal ini penulis menggunakan pendekatan analisis isi untuk memahami dan menafsirkan. Hasil dan Diskusi Chandra Dewi, Fitniwilis, Tuti Lestari, Khalifah Pauziyah, & Elda Widayani 42 Tafsir Surah Luqman Ayat 12-19 Maka seorang anak diwajibkan senantiasa berterimakasih kepada Allah SWT yang telah memberikan segala kenikmatan kepada hamba-Nya dan berterimakasihlah kepada ayah dan ibu karena dari keduanya kamu lahir ke dunia.7 Artinya: “dan jika keduanya memaksamu untuk mempersekutukan dengan aku sesuatu yang tidak ada pengetahuanmu tentang itu, Maka janganlah kamu mengikuti keduanya, dan pergaulilah keduanya di dunia dengan baik, dan ikutilah jalan orang yang kembali kepada-Ku, kemudian hanya kepada-Kulah kembalimu, Maka Kuberitakan kepadamu apa yang telah kamu kerjakan.(QS. Luqman:15). Setelah ayat yang lalu menekankan pentingnya berbakti kepada ibu bapak, kini diuraikan kasus yang merupakan pengecualian mentaati perintah kedua orang tua, sekaligus menggaris bawahi wasiat Luqman kepada anaknya tentang keharusan meninggalkan kemusyrikan dalam bentuk serta kapan dan dimanapun.8 Pada ayat ke-15 ini Allah memberikan pengecualian, taat yang Allah maksud hanyalah pada sesuatu hal yang baik. Ketika mereka memerintahkan untuk mempersekutukan Allah, maka seorang anak diwajibkan untuk tidak menaatinya. Namun kita tetap diperintahkan untuk berinteraksi dengan mereka menggunakan cara yang baik dan sopan. “Dan ikutilah jalan orang yang kembali kedapa-Ku.” Yaitu jalan yang ditempuh oleh orang- orang yang beriman. Karena itulah jalan yang selamat, yang tidak berbahaya. “Kemudian kepada- Kulah kamu sekalian kembali. Karena datangnya kita ini adalah dari Allah, perjalanan hidup di dunia dalam jaminan Allah dan kelak akan kembali kepada Allah. “Maka akan Aku berikan kepadamu apa yang telah kamu kerjakan.” Allah kelak yang akan menilai baik buruknya apa yang kamu amalkan selama dalam dunia ini. Sebab itulah dari sekarang pula bimbingan Tuhan wajib diterima, dengan menempuh jalan orang-orang yang beriman.9 Artinya: “(Luqman berkata): "Hai anakku, Sesungguhnya jika ada (sesuatu perbuatan) seberat biji sawi, dan berada dalam batu atau di langit atau di dalam bumi, niscaya Allah akan mendatangkannya (membalasinya). Sesungguhnya Allah Maha Halus lagi Maha mengetahui” (QS. Luqman: 16). Dasar ayat 16 surah Luqman, tokoh yang dianugerahi hikmah ini kembali kepada akidah dengan memperkenalkan sifat Tuhan, khususnya yang berkaitan dengan sifat Maha Mengetahui, Allah mampu mengungkapkan segala sesuatu, betapapun kecilnya.10 Luqman melanjutkan kembali wasiatnya. Pada ayat ini, ia berwasiat kepada anaknya untuk berbuat ikhlas, yaitu mengerjakan seluruh amalan atas dasar niat karena Allah SWT. Tafsir Surah Luqman Ayat 12-19 Kesyirikan itu amat buruk dan kezaliman yang nyata, kesyirikan adalah meletakan sesuatu bukan pada tempatnya. Siapa yang menyamakan antara pencipta dengan yang diciptakan mahluk antara patung dengan Tuhan. Antara patung dengan Tuhan tidak diragukan lagi.4 Artinya: “dan Kami perintahkan kepada manusia (berbuat baik) kepada dua orang ibu- bapanya; ibunya telah mengandungnya dalam keadaan lemah yang bertambah- tambah, dan menyapihnya dalam dua tahun. bersyukurlah kepadaku dan kepada dua orang ibu bapakmu, hanya kepada- Kulah kembalimu.” (QS. Luqman: 14). Ayat di atas dinilai oleh banyak ulama bukan bagian dari pengajaran Luqman kepada anaknya. Ia disisipkan dalam pengajaran Luqman di dalam Al-Qur’an untuk menunjukan betapa penghormatan dan kebaktian kepada kedua orang tua tempat kedua setelah pengagungan kepada Allah SWT. Memang Al-Qur’an sering kali mengandengkan perintah menyembah Allah SWT dan perintah berbakti kepada kedua orang tua, tetapi kendati nasihat itu bukanlah nasihat Luqman, namun itu berarti bahwa beliau menasihati anaknya dengan nasihat serupa. Luqman menyatakan hal itu kepada anaknya dengan nasihat itu menyangkut hak kami. Tetapi lanjut Biqa’i redaksinya diubah agar mencakup semua manusia. Sesudah Allah menuturkan apa yang telah diwasiatkan oleh Luqman terhadap anaknya yaitu supaya ia bersyukur kepada Tuhan yang telah memberikan semua nikmat, yang tidak seorangpun bersekutu dengan-Nya di dalam penciptaan sesuatu. Kemudian Luqman menegaskan bahwa syirik itu adalah perbuatan yang zalim. Dengan wasiatnya kepada semua anak supaya mereka berbuat baik terhadap kedua orang tuanya. Qurais Shihab menjelaskan kata wahnan berarti kelemahan atau kerapuhan, yang dimaksud berarti kurangnya memikul 43 Hubungan Pola Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA Cibinong uh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA beban kehamilan, penyusuan dan pemeliharaan anak. Patron kata yang digunakan ayat inilah yang mengisyaratkan betapa lemahnya sang ibu sampai-sampai ia dilukiskan bagaikan kelemahan itu sendiri yaitu segala sesuatu yang berkaitan. Berbakti kepada orang tua merupakan wasiat Luqman selanjutnya, setelah sebelumnya ia menerangkan akan kewajiban kita untuk meng-Esakan Allah SWT. Karena berbakti kepada kedua orang tua menepati tempat kedua setelah pengagungan kepada Allah. Hal ini meyiratkan akan penting dan wajibnya kita untuk berbakti kepada orang tua. “Dan Kami perintahkan kepada manusia (berbuat baik) kepada dua orang ibu-bapaknya”. Penulis mengutip M. Qurais Shihab dalam (tafsir Al-Mishbah) beliau menyatakan, bahwa ayat diatas tentang bakti seorang anak kepada ayah dan ibunya yang menempati posisi kedua setelah Allah SWT, terutama berbakti kepada ibu dikarenakan telah mengandungnya selama sembilan bulan berturut-turut dengan kondisi tidak berdaya dan bersusah payah. TERAPUTIK \| Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 ISSN 2580-2046 (Print) \| ISSN 2580-2054 (Electronic) \| Url: https://journal.unindra.ac.id/index.php/teraputik/index Tafsir Surah Luqman Ayat 12-19 Meskipun amalan tersebut hanya seberat biji sawi dan bahkan bisa jadi terlihat remeh di hadapan manusia, lalu ia berada di tempat yang paling tersembunyi dan paling tidak kelihatan, seperti di dalam batu besar atau di tempat yang paling tinggi seperti di langit, atau tempat yang paling bawah seperti di dalam bumi, niscaya hal itu akan dikemukakan oleh Allah SWT kelak di hari kiamat. Yaitu pada hari ketika Allah meletakkan timbangan amal perbuatan yang tepat, lalu pelakunya akan menerima pembalasan amal perbuatannya, apabila amalnya itu baik, maka balasannyapun baik pula, dan apabila amalnya buruk, maka balasannyapun buruk pula.11 Dalam nasihat yang singkat ini, terkandung beberapa makna. Pertama, bahwa seberapa kecilnya setiap perbuatan pasti akan mendapat balasan dari Allah. Oleh karena itu, jangan pernah menganggap remeh amal baik yang kecil, karena hal itu akan tetap diperhitungkan oleh Allah. Demikian juga jangan pernah menganggap remeh perbuatan dosa seberapa kecilpun, karena Chandra Dewi, Fitniwilis, Tuti Lestari, Khalifah Pauziyah, & Elda Widayani 44 Allah pasti akan memberikan balasannya juga. Kedua, bahwa Allah mengetahui segala sesuatu sekecil-kecilnya perbuatan tersebut, tidak ada satupun yang luput dari pengetahun Allah. Allah mengetahui apa yang tampak dan apa yang tesembunyi. Allah Allah pasti akan memberikan balasannya juga. Kedua, bahwa Allah mengetahui segala sesuatu sekecil-kecilnya perbuatan tersebut, tidak ada satupun yang luput dari pengetahun Allah. Allah mengetahui apa yang tampak dan apa yang tesembunyi. Allah mengetahui setiap niat yang terlintas dalam pikiran manusia. Oleh karena itu, jangan pernah mengira seseorang bisa lolos dari pengamatan Allah. Allah memberikan balasan kepada setiap hambanya yang melakukan perbuatan baik atau buruk seadil-adilnya tanpa menzalimi siapapun karena Allah adalah Tuhan yang maha adil. Sekecil dan sebesar apapun perbuatan itu Allah akan memberinya balasan, sekarang saat di dunia atau nanti saat di akhirat semua perbuatan akan diperhitungkan dan mendapat balasan yang sesuai dengan tingkat perbuatan tersebut, karena Allah Maha Mengetahui dan Maha Bijaksana. Artinya: “Hai anakku, dirikanlah shalat dan suruhlah (manusia) mengerjakan yang baik dan cegahlah (mereka) dari perbuatan yang mungkar dan bersabarlah terhadap apa yang menimpa kamu. Sesungguhnya yang demikian itu Termasuk hal-hal yang diwajibkan (oleh Allah)”. (QS. Luqman: 17) Luqman melanjutkan nasehatnya kepada anak, nasehat yang dapat menjamin kesinambungan Tauhid serta kehadiran Ilahi dalam qalbu sang anak. Beliau berkata sambil tetap memanggilnya dengan pangilan mesra: “Wahai anakku sayang, laksanakanlah Sholat dengan sempurna sesuai syarat, rukun dan sunah-sunahnya. Dan disamping engkau memperhatikan dirimu dan membentenginya dari kekejian dan kemungkaran, anjurkan pula orang lain berlaku serupa. Simpulan Berdasarkan uraian yang telah penulis paparkan dari bab I sampai bab IV, maka kesimpulan dari konsep bimbingan dan konseling Islam dalam Al- Qur’an Surah Luqman Ayat 13-17 ada beberapa hal, yaitu: p y Pendekatan bimbingan dan konseling Islam yaitu pendekatan al-mauidzah al-hasanah (memberi nashat kepada orang lain dengan bahasa yang baik yang dapat menggugah hatinya sehingga pendengar mau menerima nasihat tersebut), yaitu pada ayat 13 dengan kata kunci ya‟izhuhu (memberi pelajaran kepadanya dengan bijaksana), ayat 16 kata lathif (halus yang artinya lemahlembut) dan ayat 17 kata bil-ma‟ruufi (mengerjakan yang baik atau dengan yang baik). Pendekatan bil-hikmah (suatu layanan yang dilakukan dengan penuh kebijaksanaan) terdapat pada ayat 12. Tafsir Surah Luqman Ayat 12-19 Karena itu, perintahkanlah secara baik-baik siapapun yang mampu engkau ajak dalam melaksanakan aneka tugasmu. Sesungguhnya yang demikian itu yang sangat tinggi kedudukannya dan jauh tingkatnya dalam kebaikan yakni Sholat, amr ma‟ruf nahi mungkar dan kesabaran termasuk hal-hal yang diperintahkan Allah agar diutamakan, sehingga tidak ada alasan untuk mengabaikannya.”12 Nasihat Luqman di atas menyangkut hal-hal yang berkaitan dengan amal-amal shaleh yang puncaknya adalah sholat, serta amal-amal perbuatan yang tercermin amr ma‟ruf nahi mungkar, juga nasihat berupa perisai yang membentengi seseorang dari kegagalan yaitu sabar dan tabah. Menyuruh mengerjakan yang baik, karena tidak wajar jika menyuruh orang lain sebelum diri sendiri yang mengerjakannya. Demikian pula melarang kemungkaran, menuntut agar yang melarang terlebih dahulu mencegah kemungkaran dari dirinya. Itulah sebab Luqman tidak memerintahkan anaknya melaksanakan ma‟ruf dan menjauhi mungkar, tetapi memerintah, menyuruh, dan mencegah. Disisi lain membiasakan anak melaksanakan tuntunan ini agar dapat timbul dalam dirinya jiwa kepemimpinan dan kepedulian sosial. TERAPUTIK Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 TERAPUTIK Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40 45 Pusat Kajian Penelitian dan Pengembangan Bimbingan dan Konseling, Unindra \| DOI: 10.26539/ teraputik.631561 Ucapan Terima Kasih Ucapan terimakasih yang paling istimewa kepada istri tercinta yang telah membantu, menemani dan menyemangati dalam penulisan artikel ini. Kemudian kepada teman-teman pasca sarjana BKPI yang selalu sama-sama menyemangati penulis serta kepada bapak Irman sebagai dosen pengampu mata kuliah Tafsir Ayat-ayat konseling 45 Hubungan Pola Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA Cibinong TERAPUTIK \| Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 ISSN 2580-2046 (Print) \| ISSN 2580-2054 (Electronic) \| Url: https://journal.unindra.ac.id/index.php/teraputik/index Competing interests: The authors declare that they have no significant competing financial, professional or personal interests that might have influenced the performance or presentation of the work described in this manuscript. 45 Hubungan Pola Asuh Orangtua Dengan Kecerdasan Emosional, Kemandirian Belajar, Dan Prokrastinasi Akademik Siswa SMA Cibinong Daftar Rujukan Aisyah. “Analisis Kemampuan Penalaran Logis Mahasiswa Program Studi Pendidikan Matematika Pada Mata Kuliah Pengantar Dasar Matematika” Jurnal Ilmiah Dikdayah (Juni, 2016), hal. 1-20. ( ) Al Arif. N.F.A. (2017). “Analisis Terhadap Pembatalan Penetapan Imbalan Jasa Kurator Dalam Kepailitan PT. Telkomsel Studi Putusan Peninjauan Kembali No. 48PK/PDT. SUS. PAILIT/2013,” Skripsi Sarjana, Fakultas Syariah dan Hukum. Al-Qaradhawi, Anwar, F. ( 2019). Landasan Bimbingan dan Konseling Islam. Yogyakarta: Penerbi Defepublish Grup Penerbit CV Budi Utama. p p Ardi, Z. “Konsep Dasar Self Disclosure dan Pentingnya Bagi Mahasiswa Bimbingan dan Konseling.,” Jurnal Ilmu Pendidikan, 13 (Januari, 2013), hal. 1-25. Assyaukanie, L. (2009). Metodologi Studi Al-Quran. Jakarta: PT Gramedia Pustaka Utama Anggota IKAPI. Aziz H. S. (2018). Qashashul Quran Bekal Utama Juru Kisah. Yogyakarta: Katalog Dalam Terbitan KDT. Basit, A. (2017). Konseling islam. Cimangis Depok: Kencana. g g p Bukhori, B. “Dakwah Melalui Bimbingan dan Konseling Islam,”Konseling Religi: Jurnal Bimbingan Konseling islam, 1 (Juni 2014), hal. 1-15. g g ( ) Bungin, B. (2010). Analisis Data Penelitian Kualitatif. Jakarta: Raja Grafindo Persada. Chusna, P.A. “Konsep Dasar Pendidikan Anak Selang Pandang Lukman Al_Hakim,” Jurnal Al-Makrifat (April, 2018), hal.1-30. Dahyani, A. N. (2020). Pengantar Bimbingan dan Konseling Islam. Bengkulu: CV Brimedia Global. yanto. (2011). Analisis Isi Pengantar Metodelogi Untuk Penelitian Ilmu Komunikasi da Ilmu-Ilmu Sosial Lainnya. Jakarta: Kencana. y Kementrian Agama RI. (2002). Al Quran dan Terjemahan. Alfatih. mentrian Agama RI. (2002). Al Quran dan Terjemahan. Alfatih. TERAPUTIK \| Jurnal Bimbingan dan Konseling \| Vol. 7, No. 1, June, (2023), pp. 40-45 ISSN 2580-2046 (Print) \| ISSN 2580-2054 (Electronic) \| Url: https://journal.unindra.ac.id/index.php/teraputik/index
https://openalex.org/W4221097114	https://www.frontiersin.org/articles/10.3389/fpsyg.2022.813620/pdf	English	null	Intergenerational Educational Inequality and Its Transmission in China’s Elite Universities	Frontiers in psychology	2,022	cc-by	8,725	Intergenerational Educational Inequality and Its Transmission in China’s Elite Universities China is experiencing high social inequality accompanying inﬂuential education reforms. The Independent Freshmen Admission (IFA) policy was one of the multiple strategies in higher education reforms in China against the social context of high social inequality and the expansion of higher education. By comparing students admitted through IFA with those admitted by the National College Entrance Examination (NCEE), we examined how family advantages contributed to higher education inequality in terms of educational opportunity, process, and results. Using data from an elite university in Beijing, we found that: (1) Family advantages improved a student’s likelihood of being admitted through IFA, exhibiting opportunity inequality. (2) No signiﬁcant difference in academic grades existed between the students admitted through IFA and NCEE. In comprehensive quality, however, those recruited through IFA performed signiﬁcantly better than those admitted through NCEE. (3) Family social capital not only increased the likelihood of students being admitted through IFA but also, through direct and indirect effects, increased their comprehensive quality performance in terms of receiving student association and social practice awards. Edited by: Leandro S. Almeida, University of Minho, Portugal Reviewed by: Antonio Cervero, University of Oviedo, Spain Joana R. Casanova, University of Minho, Portugal Claisy Marinho-Araujo, University of Brasilia, Brazil Correspondence: Shuanglong Li shuanglongli@gzhu.edu.cn †These authors have contributed equally to this work and share ﬁrst authorship Edited by: Leandro S. Almeida, University of Minho, Portugal Reviewed by: Antonio Cervero, University of Oviedo, Spain Joana R. Casanova, University of Minho, Portugal Claisy Marinho-Araujo, University of Brasilia, Brazil Correspondence: Shuanglong Li shuanglongli@gzhu.edu.cn †These authors have contributed equally to this work and share ﬁrst authorship ORIGINAL RESEARCH published: 07 March 2022 doi: 10.3389/fpsyg.2022.813620 INTRODUCTION Education plays an essential role in modern society as it is a channel to achieve social mobility and socioeconomic status. However, it is also a tool for reproducing social inequality as family resources are important potential advantages for children’s educational opportunities and achievements (Blau and Otis, 1967; Wu, 2010). With the expansion of education, some scholars argued that family advantage contributed less to education inequality (Treiman, 1970), while other studies found that family background remained a signiﬁcant eﬀect on educational opportunities, despite the expansion of education (Shavit and Blossfeld, 1993; Buchmann and Hannum, 2001). Specialty section: This article was submitted to Educational Psychology, a section of the journal Frontiers in Psychology Received: 12 November 2021 Accepted: 16 February 2022 Published: 07 March 2022 The eﬀect of family background on higher education inequality should be discussed in the context of speciﬁc systems and backgrounds. China is a unique setting to study how social inequality contributed to education reproduction as it is experiencing a high social inequality accompanying signiﬁcant reforms in the higher education system. On the one hand, China is experiencing high-income inequality measured by the Gini coeﬃcient. A Gini coeﬃcient is ranged from 0 to 1, with a higher score meaning higher income inequality. The Gini coeﬃcient in China increased from 0.30 approximately in 1980 to 0.467 in 2017, indicating that a huge gap existed Keywords: family background, admission methods, academic performance, China, elite universities Family Background and Educational Opportunity Family background has a vital function in educational attainment, and a privileged family social status is advantageous for acquiring educational opportunities. From the microscopic perspective of Blau and Otis (1967) status attainment model, family social resources are explained as potent factors in children’s educational attainment. Studies have uncovered, despite the expansion of education, family background has continued to exert a signiﬁcant inﬂuence on educational opportunities, regardless of whether a child lives in a developed or developing country (Treiman and Yip, 1989; Shavit and Blossfeld, 1993; Buchmann and Hannum, 2001; Wu, 2011). Research in China has similarly shown that educational inequality has persisted alongside educational expansion, suggesting that the eﬀect of the family background remains signiﬁcant (Li, 2010; Ye and Ding, 2015). Independent Freshmen Admission is a comprehensive admission process that has challenged traditional admissions that are solely based on test scores, as it admits students through multifaceted assessment approaches such as individual applications, qualiﬁcation evaluations, written examinations, and group interviews (Liu et al., 2014). The IFA of colleges and universities has become an important measure to change the disadvantages of the system of “one exam determines one’s life” and reﬂected the innovation and exploration of education and the requirements of quality education (Hu, 2020). IFA is only limited to a small number of elite universities, combining the independent admission program in selected universities with the NCEE (Wu et al., 2019). Students who have passed the IFA examination (including written test and interview) can enjoy special preferential treatment such as lower scores of the uniﬁed NCEE or preferential choice of major in enrollment (Wu et al., 2019). In addition to providing an alternative path toward higher education nationwide, the purpose of IFA is to recruit outstanding students with academic specialization and innovative potential who could be missed in the NCEE (Ministry of Education, 2012; Wu et al., 2019). The problem of inequality is equally present in the IFA process (Bao, 2012; Ma and Bu, 2019). The new educational opportunities oﬀered by the IFA system may beneﬁt only the students from higher socioeconomic status (SES) families, and the most privileged educational resources would be continuously used by advantaged groups (DiPrete and Gregory, 2006; Liu et al., 2014; Li, 2016; Qian and Yang, 2019). Citation: Wei J, Li S, Han Y and Fu W (2022) Intergenerational Educational Inequality and Its Transmission in China’s Elite Universities. Front. Psychol. 13:813620. doi: 10.3389/fpsyg.2022.813620 March 2022 \| Volume 13 \| Article 813620 1 Frontiers in Psychology \| www.frontiersin.org Intergenerational Educational Inequality in China Wei et al. between the rich and the poor in the society (Xie and Xiang, 2014; Household Survey Oﬃce of National Bureau of Statistics, 2017); on the other hand, China has experienced huge changes in the education system since 1949—from the radical egalitarianism of the Cultural Revolution to the resumption of the National College Entrance Examination (NCEE). The inﬂuence of family background on educational inequality has been signiﬁcantly diﬀerent across diﬀerent periods (Zhou et al., 1998; Li, 2010; Wu, 2010, 2011; Wu, 2013a,b; Yeung, 2013; Liu, 2014). between the rich and the poor in the society (Xie and Xiang, 2014; Household Survey Oﬃce of National Bureau of Statistics, 2017); on the other hand, China has experienced huge changes in the education system since 1949—from the radical egalitarianism of the Cultural Revolution to the resumption of the National College Entrance Examination (NCEE). The inﬂuence of family background on educational inequality has been signiﬁcantly diﬀerent across diﬀerent periods (Zhou et al., 1998; Li, 2010; Wu, 2010, 2011; Wu, 2013a,b; Yeung, 2013; Liu, 2014). Although previous studies have examined the eﬀect of higher education expansion on educational inequality, few have studied how the IFA, the new admission method among several strategies in the education reform and expansion, aﬀects higher education inequality in China. Thus, this study ﬁlled the gap by answering the research questions: (1) whether family background aﬀects IFA and the subsequent academic performance of students, and (2) whether it contributes to retaining the advantages of students from privileged backgrounds in achieving educational success. g ) At the beginning of reform (i.e., the founding of the PRC in 1949), higher education was characterized by intense class struggle in terms of “equality within the class,” and political upbringing within the family was accordingly the primary cause for diﬀerentiated educational opportunities (Ying and Liu, 2015). Inﬂuenced by the dramatic increase in social stratiﬁcation and diﬀerentiation since 1992, the educational system was stricken by marketization, revealing the eﬀect of family resources, and the generational mechanism of educational inequality was further transformed into dual modes of resource conversion and cultural reproduction (Li, 2006). Citation: In the twenty-ﬁrst century, higher education in China has undergone several fundamental changes. The expansion of higher education has signiﬁcantly aﬀected inequality in educational opportunities (Li, 2010; Ye and Ding, 2015). Aiming to solve the disadvantages of the traditional college enrollment system based on test scores, especially the questioning and criticism of the exam-oriented college entrance (i.e., NCEE), the Ministry of Education has also begun a large-scale reform of the college enrollment and examination system. One of the crucial strategies is Independent Freshmen Admission (IFA). LITERATURE REVIEW AND RESEARCH HYPOTHESES Education is crucial in attaining higher social status and achieving upward social mobility in modern societies. Existing literature has demonstrated that people in prominent positions are those who have been educated (Treiman and Yip, 1989; Muller and Shavit, 1998); accordingly, education is positioned at the core of studies on social stratiﬁcation, and educational attainment has become a vital subﬁeld in the study of intergenerational mobility (Shavit and Blossfeld, 1993; Hannum and Yu, 1994; Deng and Treiman, 1997; Li, 2006; Wu, 2011). Frontiers in Psychology \| www.frontiersin.org Family Background and Educational Opportunity Previous sociological studies have focused on several institutional contexts when discussing the family background and educational inequality: ﬁrst, institutional transformation (i.e., the marketization revolution) (Zhou et al., 1998; Li, 2006; Wu, 2010; Li, 2014); second, educational expansion since the expansion of higher education recruitment in 1998 (Li, 2010; Wu, 2013a); and third, the establishment of the key secondary school and tracking systems (Wu, 2013b; Tam and Jin, 2015). Since 2003, implementing an independent recruitment system (i.e., IFA) has provided an alternative institutional context for examining the relationship between family background and educational inequality. In the 21st century, reform in the Chinese higher education system has given birth to the IFA. If the expansion of higher education recruitment in 1998 can represent an educational expansion in quantity, then the IFA can be regarded as Frontiers in Psychology \| www.frontiersin.org March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 2 Intergenerational Educational Inequality in China Wei et al. increasing the pathways for educational attainment (Liu et al., 2014). From a theoretical point of view, “maximally maintained inequality (MMI, hereafter)” (Raftery and Michael, 1993) provides a noteworthy explanation for the relationship between educational expansion and inequality. According to MMI, when educational opportunities increase, families of privileged social status continue to control a large share of the educational resources, and only when the educational attainment of these privileged individuals is maximized can the beneﬁt of these increasing educational opportunities reach individuals of lower social status. In addition, family cultural capital exerts a signiﬁcant eﬀect on the educational attainment of a child (Bourdieu, 1974). Families with higher cultural capital place greater emphasis on education and are thus willing to pay higher costs; parents in such families may also have higher capabilities to assist in their children’s learning. Inﬂuenced by their family background, these children have higher cultural capital, place greater emphasis on education, and accordingly achieve higher educational performance (De Graaf et al., 2000; Sullivan, 2001; Li, 2006). The social capital of a family is another crucial factor in educational opportunities. Social capital is the sum of actual or potential resources attainable from institutionalized social networks; the unequal distribution of social capital thereby results in disparate educational attainment among children of diﬀering family backgrounds (Bourdieu, 1986). Background on Academic Performance The literature review revealed that family background signiﬁcantly aﬀects university admissions, and disparities may similarly exist between individuals who entered universities through diﬀerent admission methods: students admitted through IFA achieve higher academic performance than students who took the standardized NCEE. Furthermore, the eﬀects of family background on academic performance can be divided into direct and indirect inﬂuences. Two types of admission methods mediate the eﬀect of family background on academic performance. Through IFA, the privilege of family background is transformed into the advantage of academic achievement, further widening disparities in the social hierarchy. The social capital of family background may aﬀect academic performance among university students primarily because the networking resources generated by family social capital provide these university students with greater educational opportunities that foster their comprehensive quality in various aspects. y g Through more subjective and ﬂexible methods such as initial qualiﬁcation evaluations, paper examinations, and groups interviews, the range of standardized assessments in the IFA is far greater than the knowledge that students acquire in school, rendering it less objective than standardized examinations and consequently beneﬁcial to students with privileged family backgrounds. Students with higher social and cultural capital are more likely to be admitted through IFA than those with lower social and cultural capital, meaning that this system contributes to educational inequality. The following research hypothesis is thus proposed: The third hypothesis of this study is therefore proposed as follows: Hypothesis 3a: Holding other factors constant, family background displays a signiﬁcant and direct inﬂuence on academic performance. Hypothesis 3b: Family background aﬀects academic achievement indirectly by inﬂuencing the admission method. Hypothesis 3a: Holding other factors constant, family background displays a signiﬁcant and direct inﬂuence on academic performance. Hypothesis 3b: Family background aﬀects academic achievement indirectly by inﬂuencing the admission method. Hypothesis 3a: Holding other factors constant, family background displays a signiﬁcant and direct inﬂuence on academic performance. Hypothesis 3b: Family background aﬀects academic achievement indirectly by inﬂuencing the admission method. Hypothesis 1: University students from more privileged family backgrounds are more likely to be admitted into higher education through IFA than those who are less privileged. Family Background and Educational Opportunity assessments to identify the capabilities of an applicant among the massive number of applications; under such institutional requirements, students who can stand out from the rest are therefore more competent. This study thus proposes that disparities exist between students admitted through the two diﬀerent systems. The second research hypothesis is, accordingly: Hypothesis 2: Academic achievement among university students admitted through IFA is higher than university students admitted through NCEE. Hypothesis 2: Academic achievement among university students admitted through IFA is higher than university students admitted through NCEE. Data The purpose of the IFA system is to “select innovative talents, cultivate specializations, and actively explore a new system that, with standardized examination as its base, integrates diversiﬁed examinations, versatile admission selections, and independent recruitment to eﬀectively select outstanding and innovative talents through academic autonomy, comprehensive government instruction, and service” (Ministry of Education, 2003). In the recent decade, the Chinese Ministry of Education has further deﬁned IFA as intended for students with “academic specialization” and “innovative potential” (Ministry of Education, 2011, 2013). Through step-by-step assessment procedures (i.e., qualiﬁcation evaluations, paper examinations, and interviews), IFA features more comprehensive and practical Data were collected from 2011 to 2014 among undergraduates in an elite university in Beijing. According to Population Census data in 2010 (National Bureau of Statistics of China, 2012), which is one year before our ﬁrst wave survey, around 118 million people in China had completed tertiary education (including college, undergraduate, and postgraduate), accounting for 8.64% of the total population in 2010 in China. The proportion of people getting a bachelor’s degree in elite universities is even much lower than the 8.64%, which means the opportunity of getting an education in our selected university is rare. A random sampling method was adopted to select undergraduates of all years and majors. First, a list of all undergraduates was obtained from the student aﬀairs oﬃce, and the research participants March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 3 Intergenerational Educational Inequality in China Wei et al. summed it. The sum of family social capital was coded as follows: the ﬁrst 25% of the lowest power index was coded as 1 = “low family social capital”; a power index ranging between 25 and 75% was denoted as 2 = “intermediate family social capital”; and power index higher than 75% was coded as 3 = “high family social capital.” were retrieved through a random sampling procedure using SPSS software. Subsequently, the student aﬀairs oﬃce invited the selected students through the school counselors in each school/department and asked them to complete a questionnaire in diﬀerent classrooms. Supervisor were employed to examine the quality of the answers and collect the questionnaires on site. A total of 5422 valid questionnaires were collected over the 4- year survey. The numbers of valid questionnaires acquired each year were as follows: 2,014 in 2011; 1,298 in 2012; 1,102 in 2013; and 1,008 in 2014. Variables Key variables in this study are shown in the analysis framework (Figure 1). The ﬁrst step of the study involved examining the inﬂuence of family background on the admission method. The admission method was adopted as a dependent variable and subsequently processed into a dummy variable: admission through IFA was coded as 1, and admission through the standardized NCEE was coded as 0. The third step of the analysis was investigating the direct inﬂuence of family background on academic performance and the indirect inﬂuence generated through the admission method, with variables measured the same as in Step 1 and 2. The key independent variables were family background, measured by father’s work unit, father’s membership in the Chinese Communist Party (CCP), family cultural capital, and family social capital. Father’s work unit was categorized into three categories: 1 = private organizations; 2 = state-owned enterprises; 3 = Party (i.e., CCP) or government organizations. Father’s Party membership is a dummy variable, with non-party members as the reference group. The control variables included in this study were as follows: gender (1 = male; 0 = female), enrollment age (continuous variable), area of study in high school (1 = humanities; 0 = sciences), registered household residence (hukou) before admission (1 = urban; 0 = rural), family residence (1 = city; 2 = county; 3 = town or village), ethnicity (1 = Han; 0 = minority ethnic groups), and rank/type of secondary school (1 = national key level; 2 = provincial level; 3 = city level and below). In addition, the frequency of self-study was included in the analysis on academic achievement in Step 2 to control for the level of individual eﬀort (1 = never; 2 = rarely; 3 = occasionally; 4 = sometimes; 5 = often). Family cultural capital was evaluated using average parental educational attainment, with a maximum measure of 12 years: average parental educational attainment below 12 years was labeled “1 = low family cultural capital”; average parental educational attainment above 12 years but below 16 years was labeled “2 = intermediate family cultural capital”; average parental educational attainment greater than or equal to 16 years was labeled as “3 = high family cultural capital.” Data The second step of our analysis is to examine the eﬀects of family background and the admission method using both of these as the core independent variables. Academic performance was employed as the dependent variable and was classiﬁed primarily into academic grades and comprehensive quality. Additionally, academic grades were measured using the overall class ranking from the previous semester (1 = low; 2 = low-intermediate; 3 = intermediate; 4 = upper-intermediate; 5 = superior); comprehensive quality included whether the participants received awards in student associations, social practices, and essay competitions (1 = awarded; 0 = not awarded). METHODS The position generator technique in the “Chinese Lunar New Year greeting network” (Bian and Li, 2002) was adopted to measure family social capital. We assigned a power index to each occupation within one’s network (Wei and Zhao, 2011) and then The statistical analyses proceed in three steps. In the initial stage, a binary logistic regression model was employed, with the dependent variables as a binary variable. In the second step, FIGURE 1 \| Analysis framework. FIGURE 1 \| Analysis framework. Frontiers in Psychology \| www.frontiersin.org Frontiers in Psychology \| www.frontiersin.org March 2022 \| Volume 13 \| Article 813620 Intergenerational Educational Inequality in China Wei et al. ordinal logistic regression and binary logistic regression models were applied given that academic grade (i.e., overall class ranking) was an ordinal variable, and awards in student associations, social practices, and essay competitions were binary variables. Finally, the decomposition method developed by Buis (2010) was used to analyze the direct and indirect eﬀect of family background on academic performance since the decomposition was based on logistic regression models. TABLE 1 \| Statistical descriptions of the key variables used. Frontiers in Psychology \| www.frontiersin.org Inﬂuence of Family Background on the Admission Method Table 2 displays the results of the logistic regression model. Model 1 is the baseline model, which only included the controlled variables that possibly inﬂuenced the admission method. The variable of family social status was added in Model 2. According to Model 2, father’s work unit types and Party membership signiﬁcantly inﬂuenced the admission method (p < 0.001). When other variables were controlled for, the odds of IFA among participants whose fathers worked in Party or government organizations increased by 69% (e0.526–1) compared to those whose fathers worked in private organizations; students whose fathers were party members had a 74% (e0.553–1) higher odds of being admitted through IFA compared to those whose fathers were non-party members. Inﬂuence of Family Background and the Admission Method on Academic Performance Model 5 in Table 3 revealed that when other variables were controlled for, no signiﬁcant diﬀerence was found in the relationship between family background or the admission method and overall class ranking (p > 0.05). Models 6, 7, and 8 were the results of binary logistic regression models. According to the results of Model 6, when other variables were controlled for, students with high family social capital achieved a 59% (e0.461– 1) higher odds of achievement in student associations than did students with low family social capital (p < 0.01), and students who were admitted through IFA attained a 38% (e0.319–1) higher odds of obtaining student association award than did those who took the NCEE (p < 0.05). p y In Model 3, family cultural capital was added. The results revealed that holding other factors constant, participants from family backgrounds with intermediate cultural capital displayed TABLE 2 \| The effects of family background on admission method. Model 1 Model 2 Model 3 Model 4 Independent variables Father’s Work Unit Types (Ref.: Private organizations) State-owned enterprises 0.182 0.0920 0.0753 (0.116) (0.119) (0.119) Party or government organizations 0.526* 0.409 0.376 (0.125) (0.128) (0.129) Father’s Party Membership (Ref.: No) 0.553* 0.355* 0.334 (0.0988) (0.103) (0.103) Cultural Capital (Ref.: Low) Intermediate 0.544* 0.491* (0.148) (0.149) High 1.085* 1.018* (0.150) (0.151) Social Capital (Ref.: low) Intermediate 0.302* (0.123) High 0.511*** (0.135) N 5099 4571 4454 4454 pseudo R2 0.069 0.085 0.101 0.105 Standard errors in parentheses; p < 0.05, p < 0.01, *p < 0.001. We controlled gender, area of study, hukou before admission, family residence before admission, ethnicity, the rank of secondary school, and survey year in Model 1 to Model 4 (see Supplementary Table 2 for details). TABLE 2 \| The effects of family background on admission method. The results of Model 7 displayed similarities to those of Model 6: when other variables were controlled for, students with high family social capital displayed a 64% (e0.497–1) higher odds of obtaining social practice award than did students with low family social capital (p < 0.001); students with intermediate family social capital displayed a 36% (e0.306–1) higher odds of than did those of low family social capital (p < 0.01); and students admitted through IFA manifested a 39% (e0.329–1) higher odds of obtaining social practice awards (p < 0.01). Model 8 displays the eﬀects of family background and admission method on essay competition awards. Sample Characteristics Table 1 shows the sample characteristics. In our survey, 63.04% of respondents were male. The mean enrollment age was 18.28 (SD = 0.793). Among the respondents, 69.76% studied science in high school, 82.48% held an urban hukou before being admitted to the university, and 60.48% came from cities. In Model 4, family social capital was employed based on Model 3. According to the results, family social capital had signiﬁcant eﬀects on the admission method (p < 0.05). When other variables were controlled for, participants with intermediate family social capital demonstrated a 35% (e0.302–1) (p < 0.05) higher odds of being admitted via IFA than did those with low family social capital; furthermore, participants with high family social capital manifested a 67% (e0.511–1) (p < 0.001) higher odds of being admitted via IFA than did those with low family social capital. In both models, the eﬀects of the father’s work unit, Party membership, and family cultural capital remained positive and signiﬁcant (p < 0.05). METHODS Whole Sample IFA NCEE N Controls Gender (%) Male 36.96 41.98 35.98 5336 Female 63.04 58.02 64.02 Enrollment age [Mean (SD)] 18.28 (0.793) 18.20 (0.742) 18.29 (0.802) 5214 Area (%) Science 69.76 60.05 71.63 5159 Humanities 30.24 39.95 28.37 Hukou (%) Rural 17.52 7.84 19.40 5178 Urban 82.48 92.16 80.60 Family Residence (%) City 60.48 78.82 56.92 County 23.30 15.98 24.71 5207 Town or village 16.23 5.21 18.36 Ethnicity (%) Minority 13.66 9.10 14.55 5211 Han 86.34 90.90 85.45 Rank/Type of Secondary School (%) National Key Level 11.00 12.69 10.68 Provincial level 47.31 54.21 45.97 5198 City level and below 41.69 33.10 43.35 Family background Father’s Work Unit Types (%) Private organizations 31.26 22.43 33.07 4702 State-owned enterprises 44.13 42.73 44.42 Party or government organizations 24.61 34.84 22.52 Father’s Party Membership [Mean (SD)] 0.50 (0.070) 0.72 (0.016) 0.49 (0.008) 4925 Family Cultural Capital (%) Low 33.24 13.91 37.03 4828 Intermediate 34.86 33.50 35.13 High 31.9 52.59 27.84 Family Social Capital (%) Low 24.88 15.25 26.74 5214 Intermediate 49.60 49.17 49.68 High 25.53 35.58 23.58 Academic performance Overall Class Ranking (%) Low 4.82 5.10 4.77 3567 Low-intermediate 18.73 19.28 18.63 Intermediate 29.55 33.08 28.93 Upper-intermediate 31.37 26.84 32.16 Superior 15.53 15.69 15.50 Student association awards [Mean (SD)] 0.18 (0.007) 0.26 (0.019) 0.17 (0.007) 3411 Social practice awards [Mean (SD)] 0.28 (0.008) 0.35 (0.021) 0.27 (0.008) 3427 Essay competition awards [Mean (SD)] 0.07 (0.004) 0.11 (0.015) 0.06 (0.005) 3194 Standard deviations in parentheses. March 2022 \| Volume 13 \| Article 813620 Intergenerational Educational Inequality in China Wei et al. Inﬂuence of Family Background and the Admission Method on Academic Performance The results showed no signiﬁcant eﬀect of family background and the admission method on essay competition award (p > 0.05). RESULTS a 72% (e0.544–1) (p < 0.001) higher odds of being admitted via IFA compared to those from families with low cultural capital; moreover, participants from families with high cultural capital were 2.96 (e1.085) times more likely (p < 0.001) to be admitted through IFA than participants from families with low cultural capital. In both conditions, the eﬀects of father’s work unit and Party membership remained signiﬁcant (p < 0.001). Independent variables Standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001. We controlled gender, area of study, hukou before admission, family residence before admission, ethnicity, the rank of secondary school, grade, frequency of self-study, and survey year in Model 5 to Model 8 (see Supplementary Table 2 for details). on these above-mentioned ﬁndings, this section shows the extent to which these inﬂuences were directly generated from family social capital, and to what extent they were indirectly inﬂuenced by the eﬀect of family social capital on academic achievement through admission method. In Table 4, the direct and indirect eﬀects of family background on receiving awards were estimated using the decomposition method developed by Buis (2010). As shown in Model 9, when the distribution of admission methods is the same between students from family backgrounds with low and high social capital, the achievement odds of obtaining student association award among students with low family social capital can be enhanced by 5% (e0.047–1) as an indirect eﬀect through admission method (p < 0.01); when admission method was controlled for, students with high family social capital attained a 73% (e0.551–1) higher odds of obtaining student association award than did those with low family social capital as a direct eﬀect of family social capital (p < 0.001). Similar diﬀerences were observed between students from families with intermediate and low social capital and between students from families with high and intermediate social capital. TABLE 4 \| The direct and indirect effects of family background on receiving awards. Mode 9 Mode 10 Student association award Social practice award High/Low Total effects 0.598 (0.123) 0.603* (0.132) Indirect effects 0.047 (0.018) 0.046 (0.018) Direct effects 0.551* (0.121) 0.557* (0.137) Intermediate/Low Total effects 0.312* (0.132) 0.352** (0.118) Indirect effects 0.027* (0.017) 0.028* (0.012) Direct effects 0.285* (0.132) 0.323** (0.122) High/intermediate Total effects 0.286* (0.114) 0.252** (0.095) Indirect effects 0.019* (0.010) 0.018* (0.008) Direct effects 0.266* (0.113) 0.234* (0.095) N 2960 2974 Bootstrap standard errors in parentheses; p < 0.05, p < 0.01, p < 0.001. TABLE 4 \| The direct and indirect effects of family background on receiving awards. Model 10 showed the eﬀect of family social capital on the social practice awards, displaying similar results to Model 9. Direct and Indirect Effect of Family Background Table 3 displays the ﬁndings of Models 6 and 7 on the signiﬁcant inﬂuences of family social capital and the admission method on student association awards and social practice awards (p < 0.05); similarly, the signiﬁcant inﬂuence of family social capital on admission method is demonstrated by Model 4 in Table 2. Based March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 6 Intergenerational Educational Inequality in China Wei et al. TABLE 3 \| The effects of family background and admission method on academic performance. Mode 5 Mode 6 Mode 7 Mode 8 Overall class ranking Student association awards Social practice awards Essay competition awards Independent variables Father’s Work Unit Types (Ref.: Private organizations) State-owned enterprises –0.0101 0.198 0.145 –0.118 (0.0908) (0.132) (0.113) (0.156) Party or government institutions –0.0765 0.187 0.00187 –0.140 (0.105) (0.151) (0.132) (0.178) Father’s Party Membership 0.119 –0.0220 0.0459 0.187 (Ref.: No) (0.0807) (0.116) (0.100) (0.137) Cultural Capital (Ref.: Low) Intermediate 0.0191 –0.123 0.0545 –0.0539 (0.0990) (0.147) (0.124) (0.172) High –0.0578 0.0347 –0.0782 –0.0258 (0.108) (0.154) (0.136) (0.183) Social Capital (Ref.: Low) Intermediate 0.0413 0.267 0.306 0.108 (0.0861) (0.136) (0.114) (0.154) High –0.0418 0.461 0.497* 0.152 (0.103) (0.153) (0.132) (0.178) Admission Method –0.0782 0.319* 0.329** 0.279 (Ref.: NCEE) (0.0986) (0.130) (0.118) (0.157) N 3011 2960 2974 2796 pseudo R2 0.058 0.037 0.037 0.052 Standard errors in parentheses; p < 0.05, p < 0.01, **p < 0.001. We controlled gender, area of study, hukou before admission, family residence before admission, ethnicity, the rank of secondary school, grade, frequency of self-study, and survey year in Model 5 to Model 8 (see Supplementary Table 2 for details). TABLE 3 \| The effects of family background and admission method on academic performance. Frontiers in Psychology \| www.frontiersin.org Independent variables Each column indicates the indirect eﬀect of family social capital on obtaining awards through the admission method; in other words, the eﬀect of family social capital on obtaining awards may vary when the admission method distribution changes. For instance, 13.5% of the students with low family social capital earned student association awards. If the distribution of admission methods of the students with low family social capital did not change, and their family social capital was the same as that of the students with high family social capital, the proportion of obtaining student association awards among the students with low family social capital would increase from 13.5 to 21.4%. If the distribution of admission methods was the same between families with low and high social capital, the proportion of obtaining student association awards among the students with low family social capital would increase to 14.1%. Similar results were obtained regarding social practice awards. Since 1949, China’s elite higher education has undergone many changes, including the “silent revolution” (Liang et al., 2012), which has attracted much controversy and attention. In this study, an elite university in Beijing was used as a case to compare diﬀerences between the family backgrounds and academic performances of students admitted through IFA and NCEE. The inﬂuence of IFA on educational inequality was examined and controversies surrounding the silent revolution were discussed. The primary ﬁndings are as follows: In terms of admission methods, university students with more privileged family backgrounds had a higher likelihood of being admitted into higher education through IFA, revealing signiﬁcant inequality in educational opportunity. It is consistent with the existing studies (Qian and Yang, 2019; Wang, 2019; Wu et al., 2019). As a system design for selecting elite talents, IFA naturally tends to favor urban families with a considerable accumulation of cultural capital in the evaluation methods, assessment contents, and implementation procedures (Qian and Yang, 2019). This situation is evident when the following occurs: advantages in family social status, father’s party membership status, and the father’s employment in a party or government organization signiﬁcantly increased a child’s likelihood of being admitted through IFA. It is worth noting that the fathers’ party status’ inﬂuence in China may be related to China’s political capital, providing high social status and income (Wang and Li, 2020). The high social status contributes to the family’s use of social networks to get information and connect with relevant people (Liu, 2018). Independent variables In addition, advantages in family cultural and social capitals substantially increased a child’s likelihood of being admitted through IFA. Independent variables If the distribution of admission methods is the same between families with low and high social capital, the odds of obtaining social practice awards among students from families with low social capital increased by 5% (e0.046–1) as an indirect eﬀect through admission method (p < 0.01). When the admission method was March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 7 Intergenerational Educational Inequality in China Wei et al. TABLE 5 \| Predicted proportions of receiving awards and counterfactual proportions (%). Student association awards Social practice awards Association Association Distribution Low Intermediate High Low Intermediate High Low 13.5 17.2 21.4 21.5 27.5 32.4 Intermediate 13.9 17.6 21.8 22.0 28.0 33.0 High 14.1 17.9 22.2 22.3 28.4 33.4 TABLE 5 \| Predicted proportions of receiving awards and counterfactual proportions (%). and racial/ethnic groups (Hurwitz et al., 2017). However, there are still substantial class diﬀerences in test prep for college entrance exams and private tutoring, which can signiﬁcantly increase test scores and college admissions chances (Moore et al., 2018; Kim et al., 2021). Multi-dimensional and comprehensive admission indicators were initially used to ensure that students from the bottom of society have equal opportunities to enter college. Nonetheless, in practice, the ambiguity of admission criteria has led to most opportunities occupied by students from advantageous families (Rosinger et al., 2021). In addition, exam retakes could increase students’ chances of entering a college, but students from poor and minority backgrounds have lower retake rates than students from wealthier families (Goodman et al., 2020). controlled for, students from families with high social capital had a 75% (e0.557–1) higher odds of obtaining social practice award than did students from families with low social capital (p < 0.001). Similar results were identiﬁed between students from families with intermediate and low social capital and students from families with high and intermediate social capital. Table 5 also displays the direct eﬀect of family social capital and the indirect eﬀect of family social capital through admission methods on student association and social practice awards. The main diagonal indicates the factual odds of obtaining each award for each group. g p Each row indicates the direct eﬀect of family social capital on obtaining awards when the admission method was controlled for. Frontiers in Psychology \| www.frontiersin.org DISCUSSION AND CONCLUSION Family social capital signiﬁcantly aﬀected both the student association awards and social practice awards of students admitted through independent recruitment. High family social capital increased the likelihood that a student can receive both types of awards. The eﬀect mechanism of family social capital had a direct inﬂuence; the indirect beneﬁts created through the admission method were non-signiﬁcant. FUNDING This study was ﬁnancially supported by a grant from the National Social Science Fund of China (Grant No. 20CSH019) awarded to SL. This study was ﬁnancially supported by a grant from the National Social Science Fund of China (Grant No. 20CSH019) awarded to SL. The focus of the controversy over the IFA contradiction between diﬀerence and equality is the balance of talent selection and educational equity. It is also the key needed to be considered in the reform and optimization of college entrance examination policy (Qian and Yang, 2019). In order to make up for the unequal right to education in the independent enrollment SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg. 2022.813620/full#supplementary-material ETHICS STATEMENT The studies involving human participants were reviewed and approved by the Central University of Finance and Economics, China. The patients/participants provided their written informed consent to participate in this study. DISCUSSION AND CONCLUSION Unequal access to higher education between students from diﬀerent class backgrounds is a global issue. Many countries, including the United States and Brazil, have adopted aﬃrmative action to improve educational equity (French, 2021). In 2012, the Brazilian government initiated the quota law requesting all public universities to allocate 50% of the admission vacancies to public high school graduates based on the criteria such as race and income. Although time has not quelled controversy over this policy (Turgeon and Habel, 2021), extant research suggests no signiﬁcant diﬀerence in academic performance between quota and non-quota students, indicating that this policy is worthy of recognition (Vidigal, 2018; Pelegrini et al., 2022). In the United States, both SAT and ACT examination has undergone signiﬁcant changes to boost enrollment of lower socioeconomic Students admitted through IFA and NCEE showed signiﬁcant diﬀerences in comprehensive quality. No signiﬁcant diﬀerence existed between students admitted through March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 8 Intergenerational Educational Inequality in China Wei et al. policy, the comprehensive evaluation enrollment system and the special national plan were introduced accordingly in China. Safeguarding education equity requires diversiﬁed selection approaches rather than a simple “one examination system” model. Additionally, regional diﬀerences should be balanced. More attention and educational resources should be given to rural and educationally backward areas so that more students in disadvantaged areas can get more opportunities to attend higher education institutions. Since 2020, to improve the fairness of admission to Chinese elite universities, elite universities must allocate a certain percentage of vacancies to students from remote, poor, populated, and minority-inhabited areas. We expect more follow-up studies that provide timely analysis of the substantive consequences of the new policy. IFA and NCEE regarding academic grades measured by overall class ranking in our study. In terms of comprehensive quality, however, students admitted through IFA performed signiﬁcantly better in award obtaining than those admitted through NCEE. That means students admitted through IFA comprehensively perform better when they enter universities (Guo, 2020). It may be because the students with advantaged family backgrounds have better ability in the independent recruitment process, such as speech, self-exploration, artistic talent, etc. (Liu, 2018). Also, the students selected by universities through IFA are not necessarily the most competent but the most suitable for the school’s talent cultivation model (Ma and Bu, 2019). The inﬂuence of family social capital was signiﬁcant in conveying educational inequality. AUTHOR CONTRIBUTIONS JW and SL designed the research. YH and WF analyzed the data. All authors wrote the manuscript, contributed to the article, and approved the submitted version. DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Some limitations remain in this study. Firstly, these analyses were restricted to students of one elite university admitted in 2011 and 2014. Thus, it may not be suitable to generalize to the whole country. Secondly, the samples analyzed here were restricted to students who attend universities in Beijing only. Further work with a nationally representative sample is expected to reveal the whole picture. However, we focus on the mechanisms of educational inequality, that is, the underlying mechanisms by which diﬀerent admissions approaches play a role. Given that the speciﬁc implementation policy of IFA has not changed in the past few years, the conclusion of our research will not be substantially aﬀected by the timeliness of the data. Also, as the speciﬁc implementation policy of IFA is similar among each university, we speculate that the mechanisms could also be applicable to other universities in China. Thirdly, there is a lack of uniﬁed deﬁnition and evaluation of comprehensive quality among domestic universities in China, so our measurement of comprehensive quality may be biased. We hope future research can conduct a more comprehensive measurement and comparison of their performance and a more detailed tracking survey comparing their employment rate and postgraduate entrance examination rate after graduation. Bourdieu, P. (1974). “The school as a conservative force: Scholastic and cultural inequalities,” in Contemporary Research in the Sociology of Education, ed. J. Eggleston (London: Methuen), 32–46. Blau, P. M., and Otis, D. D. (1967). The American Occupational Structure. New York, NY: John Wiley & Sons, Inc. Bourdieu, P. (1974). “The school as a conservative force: Scholastic and cultural inequalities,” in Contemporary Research in the Sociology of Education, ed. J. Eggleston (London: Methuen), 32–46. Bourdieu, P. (1986). “The forms of capital,” in Handbook of Theory and Research for the Sociology of Education, ed. J. Richardson (New York, NY: Greenwood Press), 241–258. Blau, P. M., and Otis, D. D. (1967). The American Occupational Structure. New York, NY: John Wiley & Sons, Inc. Blau, P. M., and Otis, D. D. (1967). The American Occupational Structure. New York, NY: John Wiley & Sons, Inc. REFERENCES The role of selective college admissions criteria in interrupting or reproducing racial and economic Hurwitz, M., Mbekeani, P. P., Nipson, M. M., and Page, L. C. (2017). Surprising ripple eﬀects: how changing the SAT score-sending policy for low-income students impacts college access and success. Educ. Eval. Policy Anal. 39, 77–103. doi: 10.3102/0162373716665198 inequities. J. High. Educ. 92, 31–55. doi: 10.1080/00221546.2020.1795504 Shavit, Y., and Blossfeld, H. P. (1993). Persistent Inequality: Changing Educational Attainment in Thirteen Countries. Boulder: Westview Press. Kim, E., Goodman, J., and West, M. R. (2021). Kumon in: The Recent, Rapid Rise of Private Tutoring Centers. Rhode Island: Annenberg Institute at Brown University EdWorkingPaper, 21–367. doi: 10.26300/z79x-mr65 Sullivan, A. (2001). Cultural capital and educational attainment. Sociology 35, 893–912. doi: 10.1017/S0038038501008938 Tam, T., and Jin, J. (2015). Divergent Urban-rural Trends in College Attendance: State Policy Bias and Structural Exclusion in China. Soc. Educ. 88, 160–180. doi: 10.1177/0038040715574779 Li, C. (2010). Expansion of higher education and inequality in opportunity of education: a study on eﬀect of “kuozhao” policy on equalization of educational attainment. Soc. Stud. 31, 82–113. Treiman, D. J. (1970). Industrialization and social stratiﬁcation. Soc. Inq. 40, 207–234. doi: 10.7129/jject.3.1 Li, C. (2014). Education experience and inequality of opportunity among the post- 80s generation—with comments on “the silent revolution”. Soc. Sci. China 35, 66–77. Treiman, D. J., and Yip, K. B. (1989). “Educational and occupational attainment in 21 countries,” in Cross-National Research in Sociology, ed. M. Kohn (Newbury Park: Sage), 373–394. Li, J. (2016). Urban Origin and cumulative advantages: a longitudinal study of workers with higher education. Soc. Stud. 36, 57–81. doi: 10.19934/j.cnki.shxyj. 2016.02.004 Turgeon, M., and Habel, P. (2021). Prejudice, Political Ideology, and Interest: understanding Attitudes Toward Aﬃrmative Action in Brazil. Polit. Psychol. doi: 10.1111/pops.12777 Li, Y. (2006). The institutional change and educational inequality: mechanisms in educational stratiﬁcation in urban China (1966–2003). Soc. Sci. China 26, 97–109. Vidigal, C. B. R. (2018). Racial and low-income quotas in Brazilian universities: impact on academic performance. J. Econ. Stud. 45, 156–176. doi: 10.1108/jes- 10-2016-0200 Liang, C., Lee, J., Zhang, H., Li, L., Ruan, D., Campbell, C., et al. (2012). A silent revolution: research on family backgrounds of students of Peking University and Soochow University (1952–2002). Soc. Sci. China 193, 98–118. Wang, Q., and Li, Z. (2020). Party membership, self-identity, and women’s family status. Econ. Prob. China. 62, 105–120. doi: 10.19365/j.issn1000-4181.2020. 02.08 Liu, J. (2014). REFERENCES Blau, P. M., and Otis, D. D. (1967). The American Occupational Structure. New York, NY: John Wiley & Sons, Inc. Bao, W. (2012). Analysis on the eﬃciency of the autonomy admission of higher education: a perspective of fairness and eﬃciency. Res. Educ. Dev. 33, 1–7. doi: 10.14121/j.cnki.1008-3855.2012.19. 004 Bourdieu, P. (1974). “The school as a conservative force: Scholastic and cultural inequalities,” in Contemporary Research in the Sociology of Education, ed. J. Eggleston (London: Methuen), 32–46. Bourdieu, P. (1986). “The forms of capital,” in Handbook of Theory and Research for the Sociology of Education, ed. J. Richardson (New York, NY: Greenwood Press), 241–258. Bian, Y., and Li, Y. (2002). Social network capital of the Chinese family. Tsinghua Sociol. Rev. 2, 1–18. Frontiers in Psychology \| www.frontiersin.org March 2022 \| Volume 13 \| Article 813620 9 Wei et al. Intergenerational Educational Inequality in China Buchmann, C., and Hannum, E. (2001). Education and stratiﬁcation in developing countries: a review of theories and research. Annu. Rev. Soc. 27, 77–102. doi: 10.1146/annurev.soc.27.1.77 Ministry of Education (2011). Notice of the General Oﬃce of the Ministry of Education on the Eﬀective Implementation of the 2012 Pilot Program of Independent Recruitment for Higher Education Admission Reform. New Delhi: Ministry of Education Buis, M. L. (2010). Direct and indirect eﬀects in a logit model. Stata. J. 10, 11–29. doi: 10.1177/1536867X1001000104 Ministry of Education (2012). Instruction Opinions of the Ministry of Education on Consolidating the Pilot Program of Independent Recruitment for the Higher Education Student Admission Reform. New Delhi: Ministry of Education De Graaf, N. D., De Graaf, P. M., and Kraaykamp, G. (2000). Parental cultural capital and educational attainment in the Netherlands: a reﬁnement of the cultural capital perspective. Soc. Educ. 73, 92–111. doi: 10.2307/2673239 Ministry of Education (2013). Notice of the General Oﬃce of the Ministry of Education on Strengthening the Administration of the Pilot Program for Higher Ministry of Education (2013). Notice of the General Oﬃce of the Ministry of Education on Strengthening the Administration of the Pilot Program for Higher Education Independent Recruitment Reform. New Delhi: Ministry of Education Deng, Z., and Treiman, D. J. (1997). The impact of the cultural revolution on trends in educational attainment in the People’s Republic of China. Am. J. Soc. 103, 391–428. doi: 10.1086/231212 Education Independent Recruitment Reform. New Delhi: Ministry of Education p f y Moore, R., Sanchez, E., and San Pedro, M. O. (2018). REFERENCES Investigating Test Prep Impact Moore, R., Sanchez, E., and San Pedro, M. O. (2018). Investigating Test Prep Impact on Score Gains Using Quasi-Experimental Propensity Score Matching.ACT Working Paper 2018-6.Iowa: ACT, Inc DiPrete, T., and Gregory, E. (2006). Cumulative advantage as a mechanism for inequality: a review of theoretical and empirical developments. Annu. Rev. Soc. 32, 271–297. doi: 10.1146/annurev.soc.32.061604.123127 on Score Gains Using Quasi-Experimental Propensity Score Matching.ACT Working Paper 2018-6.Iowa: ACT, Inc Muller, W., and Shavit, Y. (1998). “The institutional embeddedness of the stratiﬁcation process: a comparative study of qualiﬁcations and occupations in 13 countries,” in From School to Work: A Comparative Study of Educational Qualiﬁcations and Occupational Destinations, eds Y. Shavit and W. Müller. (New York, NY: Clarendon Press), 1–48. doi: 10.1007/978-1-349-16431-8_1 French, J. H. (2021). Race, Racism, and Aﬃrmative Action in Brazil and the United States. Latin Am. Res. Rev. 56, 988–997. Goodman, Joshua, Oded Gurantz, and Jonathan Smith. (2020). Take Two! SAT Retaking and College Enrollment Gaps. Am. Econ. J. 12, 115–158. doi: 10.1257/ pol.20170503 National Bureau of Statistics of China (2012). Tabulation on the 2010 Population Census of the People’s Republic of China. Beijing: China Statistics Press. Guo, J. (2020). An empirical study on equity and eﬃciency of independent enrollment in colleges and universities – based on the perspective of students’ academic development. Jiangsu High. Educ. 36, 37–43. doi: 10.13236/j.cnki.jshe. 2020.03.006 Pelegrini, T., Braga, P. L. S., Frio, G. S., and França, M. T. A. (2022). Are There Performance Diﬀerentials Between Quota and Non-quota Brazilian Students? J. Econ. Race Policy 5, 41–53. doi: 10.1007/s41996-021-00080-7 Hannum, E., and Yu, X. (1994). Trends in educational gender inequality in China:1949–1985. Res. Soc. Strat. Mob. 13, 73–98. Qian, X., and Yang, C. (2019). Optimization and reform of independent enrollment system in colleges and universities from the perspective of educational equity. Educ. Sci. 35, 55–60. doi: 10.3969/j.issn.1002-8064.2019.05.014 Household Survey Oﬃce of National Bureau of Statistics. (2017). China Yearbook of Household Survey. Beijing: China Statistics Press. Raftery, A. E., and Michael, H. (1993). Maximally maintained inequality: expansion, reform, and opportunity in Irish education, 1921-75. Soc. Educ. 66, 41–62. doi: 10.2307/2112784 Hu, H. (2020). Analysis of regional diﬀerences in independent enrollment of “985 Universities” – based on an empirical study of candidates enrolled in 2018. Res. Educ. Dev. 41, 8–14. doi: 10.14121/j.cnki.1008-3855.2020.03.004 Rosinger, K. O., Sarita Ford, K., and Choi, J. (2021). REFERENCES Ability vs background: an analysis of the distribution mechanism of higher education enrolment opportunities. Soc. Sci. China. 35, 109–128. doi: 10.1080/02529203.2015.1029665 Wang, X. (2019). The impact of diversiﬁed enrollment methods on rural students’ enrollment opportunities in key universities: an analysis of Peking University. Educ. Acad. Month. 36, 107–111. doi: 10.16477/j.cnki.issn1674-2311.2019.04. 013 Liu, L., Wolfgang, W., Bettina, S., Wu, X., and Ulrich, T. (2014). Independent freshman admission and educational inequality in the access to elite higher education. Chinese Soc. Rev. 46, 41–67. doi: 10.2753/CSA2162-0555460403 Wei, J., and Zhao, Y. (2011). Power or prestige? The debates and veriﬁcations in measurement of social capital. Sociol. Stud. 3, 64–83. Liu, Y. (2018). Research on Equity of Independent Enrollment Class in First-Class Universities. Master thesis. Shandong: Ludong University. Wu, X. (2010). Economic transition, school expansion and educational inequality in China, 1990–2000. Res. Soc. Stratiﬁcat. Mob. 28, 91–108. doi: 10.1016/j.rssm. 2009.12.003 Ma, L., and Bu, S. (2019). An analysis of the selection eﬀect of independent enrollment policy in key universities. Peking University Educ. Rev. 17, 109–126. doi: 10.12088/pku1671-9468.201902006 Wu, X. (2011). The household registration system and rural-urban educational inequality in contemporary China. Chinese Soc. Rev. 44, 31–51. doi: 10.2753/ CSA2162-0555440202 Ministry of Education (2003). Notice of the General Oﬃce of the Ministry of Education on Eﬀective. Implementation of the Pilot Program of Independent Recruitment for Higher Education Admission Reform. New Delhi: Ministry of Education Wu, X., Li, Z., and Wang, N. (2019). Independent freshman admission program (IFAP) in China’s higher education: evidence from three national elite March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 10 Wei et al. Intergenerational Educational Inequality in China Zhou, X., Moen, P., and Tuma, N. B. (1998). Educational stratiﬁcation in urban China: 1949-94. Soc. Educ. 71, 199–222. doi: 10.2307/2673202 universities in Beijing. Chinese Soc. Rev. 51, 1–28. doi: 10.1080/21620555.2018. 1452608 universities in Beijing. Chinese Soc. Rev. 51, 1–28. doi: 10.1080/21620555.2018. 1452608 Wu, Y. (2013a). Educational opportunities for rural and urban residents in China, 1978-2008: inequality and evolution. Soc. Sci. China. 34, 58–75. doi: 10.1080/ 02529203.2013.820555 Wu, Y. (2013a). Educational opportunities for rural and urban residents in China, 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. 1978-2008: inequality and evolution. Soc. Sci. China. 34, 58–75. doi: 10.1080/ 02529203.2013.820555 Wu, Y. (2013b). Frontiers in Psychology \| www.frontiersin.org March 2022 \| Volume 13 \| Article 813620 REFERENCES The key-point school system, tracking, and educational stratiﬁcation in China, 1978–2008. Soc. Stud. 4, 179–202. doi: 10.19934/j.cnki. shxyj.2013.04.008 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. Xie, Y., and Xiang, Z. (2014). Income inequality in today’s China. Proc. Natl. Acad. Sci. 111, 6928–6933. doi: 10.1073/pnas.1403158111 Ye, X., and Ding, Y. (2015). Expanding Chinese higher education: quality and social stratiﬁcation. Chinese J. Soc. 35, 193–220. doi: 10.15992/j.cnki.31-1123/c.2015. 03.008 Yeung, W. J. (2013). Higher education expansion and social stratiﬁcation in China. Chinese Soc. Rev. 45, 54–80. doi: 10.2753/CSA2162-055545 0403 Copyright © 2022 Wei, Li, Han and Fu. 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. Ying, X., and Liu, Y. (2015). “Silent revolution” is exaggerated rhetoric: some idea exchange with Liang Chen, Li Zhongqing, et al. Chinese. J. Soc. 35, 81–93. doi: 10.15992/j.cnki.31-1123/c.2015.02.004 March 2022 \| Volume 13 \| Article 813620 Frontiers in Psychology \| www.frontiersin.org 11
https://openalex.org/W2155751959	https://ascpjournal.biomedcentral.com/track/pdf/10.1186/1940-0640-7-S1-A68	English	null	Setting-specific factors influencing trajectories of alcohol consumption in untreated control groups in early intervention studies for problematic drinking	Addiction science & clinical practice	2,012	cc-by	667	Bischof et al. Addiction Science & Clinical Practice 2012, 7(Suppl 1):A68 http://www.ascpjournal.org/content/7/S1/A68 Bischof et al. Addiction Science & Clinical Practice 2012, 7(Suppl 1):A68 http://www.ascpjournal.org/content/7/S1/A68 Bischof et al. Addiction Science & Clinical Practice 2012, 7(Suppl 1):A68 http://www.ascpjournal.org/content/7/S1/A68 Gallus Bischof1*, Jennis Freyer-Adam2, Christian Meyer2, John Ulrich2, Hans-Juergen Rumpf1 From International Network on Brief Interventions for Alcohol Problems (INEBRIA) Meeting 2011 Boston, MA, USA. 21-23 September 2011 While the efficacy of screening and brief intervention (SBI) for problematic alcohol use has been proven in various outpatient settings, findings from inpatient studies remain inconclusive. Some research has shown that inpatient treatment leads to elevated motivation to change problem behaviors. This study analyzed changes in drinking beha- vior among untreated patients with problematic alcohol use recruited proactively in general practice (GP) settings and in two general hospitals. We compared problem drin- kers randomly assigned to an untreated control group from a study of GP patients (n = 99) with inpatients recruited at the surgical and internal wards of the two gen- eral hospitals (n = 173). In both studies, all incoming patients aged 18-64 years were screened for alcohol use. Patients meeting criteria for at-risk drinking according to British Medical Association guidelines (>20-30 g alcohol per day) or meeting criteria for alcohol abuse or depen- dence according to DSM-IV criteria were included. Both samples received a non-alcohol–specific brochure on healthy living and were re-assessed 12 months later (response rate, 90.4%). The hospitalized patients were older, more often male, had less schooling, and showed elevated readiness to change at baseline compared with GP patients. The groups did not differ concerning alcohol- related diagnoses. At 12-month follow-up, significantly more GH patients than GP patients reported abstinence or low-risk drinking (50.0% versus 26.1%, p < 0.001). Recruitment setting (GH versus GP) remained a signifi- cant predictor for nonproblematic drinking or abstinence even after controlling for baseline differences between groups. Findings suggest that natural processes of change from problematic alcohol use are elevated after non- alcohol–related inpatient treatment. This might, in part, explain the lack of findings supporting BI for problematic alcohol use in inpatient settings. 1Department Of Psychiatry and Psychotherapy, University of Luebeck, Luebeck, Germany Full list of author information is available at the end of the article © 2012 Bischof et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. MEETING ABSTRACT Open Access Author details 1D Of Author details 1Department Of Psychiatry and Psychotherapy, University of Luebeck, Luebeck, Germany. 2Institute of Epidemiology and Social Medicine, Ernst- Moritz-Arndt-University of Greifswald, Greifswald, Germany. Author details 1Department Of Psychiatry and Psychotherapy, University of Luebeck, Luebeck, Germany. 2Institute of Epidemiology and Social Medicine, Ernst- Moritz-Arndt-University of Greifswald, Greifswald, Germany. Published: 9 October 2012 doi:10.1186/1940-0640-7-S1-A68 Cite this article as: Bischof et al.: Setting-specific factors influencing trajectories of alcohol consumption in untreated control groups in early intervention studies for problematic drinking. Addiction Science & Clinical Practice 2012 7(Suppl 1):A68. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission 1Department Of Psychiatry and Psychotherapy, University of Luebeck Luebeck, Germany Full list of author information is available at the end of the article 1Department Of Psychiatry and Psychotherapy, University of Luebeck Luebeck, Germany Full list of author information is available at the end of the article © 2012 Bischof et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
https://openalex.org/W4250245393	https://hal.science/hal-01007152/file/HAL-On%20the%20relevance.pdf	English	null	On the relevance of Continuum Damage Mechanics as applied to the Mullins effect in elastomers	Journal of the mechanics and physics of solids/Journal of the Mechanics and Physics of Solids	2,004	public-domain	10,394	To cite this version: Grégory Chagnon, Erwan Verron, Laurent Gornet, Gilles Marckmann, Pierre Charrier. On the rel- evance of Continuum Damage Mechanics as applied to the Mullins effect in elastomers. Journal of the Mechanics and Physics of Solids, 2004, 52 (7), pp.1627-1650. 10.1016/j.jmps.2003.12.006. hal- 01007152 On the relevance of Continuum Damage Mechani applied to the Mullins effect in elastomers Grégory Chagnon, Erwan Verron, Laurent Gornet, Gilles Marckmann, Pierre Charrier Grégory Chagnon, Erwan Verron, Laurent Gornet, Gilles Marckmann, Pierre Charrier Corresponding author. Tel.: +33-2-40-37-25-14; fax: +33-2-40-37-25-73. E-mail address: erwan.verron@ec-nantes.fr(E. Verron). HAL Id: hal-01007152 https://hal.science/hal-01007152v1 Submitted on 7 Oct 2016 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. Distributed under a Creative Commons Public Domain Mark 4.0 International License Distributed under a Creative Commons Public Domain Mark 4.0 International License G. Chagnona, E. Verron a; , L. Gorneta, G. Marckmanna, P. Charrierb aLaboratoire de M ecanique et Materiaux, Ecole Centrale de Nantes, BP 92101, 44321 Nantes cedex 3, France bGroupe Trelleborg, Societe Modyn, Service Recherche et Innovations, Zone Industrielle Nantes Carquefou, BP 419, 44474 Carquefou cedex, France aLaboratoire de M ecanique et Materiaux, Ecole Centrale de Nantes, BP 92101, 44321 Nantes cedex 3, France bGroupe Trelleborg, Societe Modyn, Service Recherche et Innovations, Zone Industrielle Nantes Carquefou, BP 419, 44474 Carquefou cedex, France Keywords: Rubber material; Stress-softening; Finite strain; Constitutive equation; Continuum damage mechanics On the relevance of Continuum Damage in elastomers G. Chagnona, E. Verron a; , L. Gorneta, G. Marckmanna, P. Charrierb Abstract The present paper reports and rationalizes the use of Continuum Damage Mechanics (CDM) to with isotropic damage is considered. Since it is demonstrated that stress-softening is not strictly speaking a damage phenomenon, the limitations of the CDM approach are highlighted. Moreover, connections with two-network-based constitutive models proposed by other authors are exhibited through the choice of both the damage criterion and the measure of deformation. Experimental data are used to establish the evolution equation of the stress-softening variable, and the choice of the maximum deformation endured previously by the material as the damage criterion is revealed as questionable. Nevertheless, the present model agrees qualitatively well with experiments except to reproduce the strain-hardening phenomenon that takes place as reloading paths rejoin the paths on material response and thereby demonstrates the importance of considering the Mullins Keywords: Rubber material; Stress-softening; Finite strain; Constitutive equation; Continuum damage mechanics Rubber material; Stress-softening; Finite strain; Constitutive equation; Continuum damage 1. Introduction Both natural rubber and synthetic elastomers are widely used in engineering applica- tions, such as tyres, engine mounts or bump stoppers. Due to the growing importance of numerical simulation in the design process of elastomeric parts, a well-established modelling of the material behaviour is an essential prerequisite for the development of new products. It is well known that rubber-like materials exhibit a strongly non-linear behaviour characterized by large strain and a non-linear stress–strain response under static conditions. Moreover, this behaviour is time-dependent as demonstrated by relax- ation and creep experiments. Finally, under cyclic loading conditions, both hysteresis and stress-softening phenomena are observed. The hysteretic behaviour may be related to viscoelasticity ( ) or viscoplasticity (Lion, 1996), and is Drozdov and Dorfmann, 2001). There is no unanimous microscopic explanation for the stress-softening of elastomers (Marckmann et al., 2002). Pioneering works focused on the experimental description of the phenomenon, and on the development of physically motivated one-dimensional constitutive models (seeMullins (1969) and the references therein). For the last decade, the interest of researchers in the development of constitutive equations involving the of oading and that it only depends on the maximum strain endured previously by the material. These prerequisites are quite acceptable with regard to experimental ob- servations. As the physical explanations of the phenomenon are not well established, initial concepts are derived. First, many authors use the thermodynamical framework Kachanov (1958) , and the general theory was derived later byLemaitre and Chaboche (1985) for metallic materials. This theory is a particular application of the most general frame- work of thermodynamics with internal variables (Maugin, 1999). It was successfully for example Voyiadjis et al., 1998). Recently, CDM was extended to the case of elas- to Gurtin and Francis (1981) who proposed a one-dimensional hyperelastic damage model for solid propellants. Later, the general three-dimensional case was studied by Simo (1987) who derived a large strain viscoelastic constitutive equation with dam- age for rubber-like materials. This model was improved by introducing microscopic concepts (Govindjee and Simo, 1991, 1992). More recently, several authors developed engineering applications (De Souza Neto et al., 1994; Miehe, 1995; Miehe and Keck, 2000). The second approach used to develop constitutive equations for the Mullins Green and Tobolsky (1946) and on the two-phase model ofMullins and Tobin (1957). 2. Thermodynamics of rubber-like materials with damage 2.1. Preliminary remarks on the use of damage mechanics to describe the Mullins 2.1. Preliminary remarks on the use of damage mechanics to describe the Mullins using the CDM. Before examining in detail the isotropic CDM as applied to non-linear theory considers that, under loading, the material surface on which internal forces apply is decreasing because of the emergence of micro-defects and micro-voids. Taking into account these physical considerations, three consequences are induced for the damage evolution: 1. the damage cannot decrease during material life because micro-defects and micro- voids cannot disappear, 1. the damage cannot decrease during material life because micro-defects and micro- voids cannot disappear, micro-defects and voids appear, and under compression only micro-defects are acti- vated because micro-voids are closed, cause the active surface tends to a non-zero threshold value determined experimen- tally (or to zero). be refuted: be refuted: recovery is highly accelerated by annealing (Mullins, 1969). So, the damage may decrease if the whole life of the material is considered, authors observed the occurrence of stress-softening under compressive conditions ( ), 3. it is mainly recognized that rupture in rubber parts is not directly related to stress- softening. Therefore, high values of the damage parameter should not be considered as a criterion for rupture or occurrence of cracks. 3. it is mainly recognized that rupture in rubber parts is not directly related to stress- softening. Therefore, high values of the damage parameter should not be considered as a criterion for rupture or occurrence of cracks. It is due to the rearrangement of the polymer network under deformation when some links between chains (entanglements), or between chains and reinforcement particles (for example carbon black) are broken down (Bueche, 1960, 1961). Considering these elastomers with some restrictions and care. 1. Introduction Both concepts postulate that the poly- mer network evolves under deformation; they consider that parts of the network are broken and others are reformed during loading. So,Wineman and Rajagopal (1990), and Rajagopal and Wineman (1992) proposed a general theory for materials under strain induced micro-structural changes. Their approach is successfully applied to the Wineman and Huntley (1994), and Huntley et al. (1996, 1997). Au- thors considered that the stress should be corrected by a scalar reformation function that depends on a given measure of deformation expressed as a scalar function of the principal invariants of the left Cauchy–Green tensor. Similarly,Beatty and Krish- naswamy (2000) derived a constitutive equation that generalized the previous works of Johnson and Beatty (1993a, b, 1995)on particular deformation states. In this work, the two-phase theory ofMullins and Tobin (1957) is considered, and the transformation of hard regions into soft regions is entirely controlled by a stress-softening function that depends on the maximum strain previously endured by the material. This function posed byEl u˜niga and Beatty (2002). Last, let us mention the recent work ofOgden and Roxburgh (1999) the stress-softening function (that depends on the maximum strain energy endured by of network changes during loading. Finally, it should be noticed that the constitutive appears that the constitutive equations obtained by the two approaches, i.e. CDM and network evolution, are close, as shown in the next section. The aim of the present paper is to demonstrate that Continuum Damage Mechan- are based on this theory, it is of great importance to rationalize its use and to exhibit its advantages and limitations in this context. Our work is restricted to hyperelastic- ity with isotropic damage. In Section2, the thermodynamical framework is derived. regard to physical phenomena and on the choice of the damage criterion. The equiva- lence of CDM and two-network or two-phase approaches is demonstrated. Section 3 is devoted to the experimental part of the work. Experimental data are used to construct the evolution equation of the stress-softening variable and to exhibit some limitations agreement for both uniaxial tensile and simple shear loading conditions. The numerical part of the work is presented in Section4. Both plane stress and three-dimensional implementations are considered, and some examples are presented in order to highlight are given in Section5. 2.2. Derivation of the constitutive equation Consider an isotropic, homogeneous and incompressible rubber-like material. This material is considered hyperelastic and subject to isotropic damage in order to describe function, that depends on the deformation gradient F and on a scalar damage variable d. This variable characterizes the elastic stress-softening of the material. Taking into account the objectivity requirement, the isotropy and the incompressibility, the strain energy function can be written as function, that depends on the deformation gradient F and on a scalar damage variable d. This variable characterizes the elastic stress-softening of the material. Taking into account the objectivity requirement, the isotropy and the incompressibility, the strain energy function can be written as W = W(I1; I 2; d); (1) W = W(I1; I 2; d); (1) where I1 and I2 B = F TF where I1 and I2 B = F TF I1 = tr B and I2 = 1 2(I 2 1 −tr(B2)): (2 (2) The third invariant I3 = det B is equal to 1 due to the incompressibility assump The third invariant I3 = det B is equal to 1 due to the incompressibility assumption. Considering now that the e)ective Cauchy stress 0, which acts on the damaged material, is related to the applied Cauchy stress by Lemaitre and Chaboche (1985) Considering now that the e)ective Cauchy stress 0, which acts on the damaged material, is related to the applied Cauchy stress by Lemaitre and Chaboche (1985) 0 = 1 −d 0 = 1 −d (3 (3) then, the strain energy function Eq. (1) can be considered as the product of the surface reducing parameter 1 −d and the strain energy function of the virgin undamaged material, denoted W0: then, the strain energy function Eq. (1) can be considered as the product of the surface reducing parameter 1 −d and the strain energy function of the virgin undamaged material, denoted W0: W(I1; I2; d) = (1 −d)W0(I1; I2): (4 (4) In order to establish the laws of state, the Clausius–Duhem inequality has to be con- sidered. It yields (Miehe, 1995) In order to establish the laws of state, the Clausius–Duhem inequality has to be con- sidered. 2.2. Derivation of the constitutive equation It yields (Miehe, 1995) D def= : D −˙W ¿ 0 (5 D def= : D −˙W ¿ 0 (5) in which D stands for the internal dissipation and D is the rate of deformation tensor that satisFes tr D=0 due to incompressibility. After some algebraic manipulations, Eq. (5) can be cast into the following form: in which D stands for the internal dissipation and D is the rate of deformation tensor that satisFes tr D=0 due to incompressibility. After some algebraic manipulations, Eq. (5) can be cast into the following form: D = −2B @W @B : D −@W @d ˙d ¿ 0 (6) (6) with tr D = 0. Thus, examining the two terms of this inequality, both the stress–strain relationship and the dissipation can be established: with tr D = 0. Thus, examining the two terms of this inequality, both the stress–strain relationship and the dissipation can be established: = −pI + (1 −d)2B @W0 @B and D = W0 ˙d ¿ 0; (7 (7) where pI is an arbitrary spherical tensor which expresses that only stress di)erences are deFned for incompressible materials. Indeed, the hydrostatic pressure p cannot be ob- tained using the stress–strain relationships but is deduced from the governing equations of the problem. Eq. (7)1 is the constitutive equation that relates the current damaged stress to the damage parameter, the strain energy function of the virgin material and the current strain state. Using that W0 only depends on the two Frst principal invariants of B, it becomes = −pI + (1 −d) 2 @W0 @I1 + I1 @W0 @I2 B −2 @W0 @I2 B2 (8) (8) in which the functions @W0=@I1 and @W0=@I2 are the material parameters. Eq. (7)2 shows that the damage process is dissipative and that the thermodynamic force associated with the damage variable d is in which the functions @W0=@I1 and @W0=@I2 are the material parameters. Eq. (7)2 shows that the damage process is dissipative and that the thermodynamic force associated with the damage variable d is Y = @W @d = −W0: (9) Y = @W @d = −W0: (9) Next, in order to completely deFne the damaged hyperelastic constitutive equation, the evolution equation for the damage variable d has to be considered. 2.2. Derivation of the constitutive equation As mentioned in the introduction, it is mainly recognized that the stress-softening in the material exclusively depends on the maximum deformation endured by the material during its history (Mullins, 1969). Therefore, a measure of the deformation state and its maximum should be deFned. This measure depends on the deformation history and is denoted . The choice of will be discussed later. Its maximum is R (t) = max ∈]−∞;t] (B()): (10) R (t) = max ∈]−∞;t] (B()): (10) In the following, notations proposed by Simo and Hughes (1998) are employed. Thus, considering previous deFnitions, the damage criterion provides a sub-region of the deformation space given by (Miehe, 1995; Simo and Hughes, 1998): In the following, notations proposed by Simo and Hughes (1998) are employed. Thus, considering previous deFnitions, the damage criterion provides a sub-region of the deformation space given by (Miehe, 1995; Simo and Hughes, 1998): ( (B(t)); R (t)) def= (B(t)) −R (t) 6 0: (11) (11) ( (B(t)); (t)) (B(t)) (t) 6 0: (11) he corresponding damage surface and its normal are, respectively The corresponding damage surface and its normal are, respectively The corresponding damage surface and its normal are, respectively = 0 and n def= @ @B: (12) (12) So, for a particular deformation state, four physical situations and their mathematical counterparts are possible: So, for a particular deformation state, four physical situations and their mathematical counterparts are possible: (i) the current deformation state is not maximum ( (t) ̸= R (t)), then the inequality ¡ 0 is satisFed (no damage evolution), (i) the current deformation state is not maximum ( (t) ̸= R (t)), then the inequality ¡ 0 is satisFed (no damage evolution), (ii) the current deformation state is maximum ( =0) and is in an unloading direction from a damage state (n : ˙B ¡ 0), (iii) the current deformation state is maximum ( = 0) and is in a neutral direction from a damage state (n : ˙B = 0), (iv) the current deformation state is maximum ( = 0) and is in a loading direction from a damage state (n : ˙B ¿ 0). (iv) the current deformation state is maximum ( = 0) and is in a loading direction from a damage state (n : ˙B ¿ 0). 2.2. Derivation of the constitutive equation The three Frst cases (i)–(iii) correspond with physical situations in which the damage should not evolve, and the fourth one (iv) represents the case in which the maximum of the deformation in the material life changes, so that the damage parameter increases. Consequently, an irreversible evolution equation for the damage can be proposed as follows: ˙d = h( ) ˙ if = 0 and n : ˙B ¿ 0; 0 otherwise; ˙d = h( ) ˙ if = 0 and n : ˙B ¿ 0; (13) otherwise; where the functionh characterizes the damage evolution. In the case of the Mullins h on the current value of the damage parameter d is classically 13) reduces to where the functionh characterizes the damage evolution. In the case of the Mullins h on the current value of the damage parameter d is classically 13) reduces to ˙d = h( ) ˙ with and ˙ 0: (14) ˙d = h( ) ˙ with and ˙ 0: (14) As the damage is an irreversible non-decreasing variable, i.e.˙d ¿ 0, the functionh is a positive function on ]0; + ∞). Moreover, considering thath is ensured by the physics of stress-softening) and that the damage in the natural state is equal to zero, Eq. (14) can be integrated to give d = ˜d ) = 0 h( ) d d = ˜d ) = 0 h( ) d (15) (15) Consequently, the stress–strain relationship Eq. (8) simply becomes Consequently, the stress–strain relationship Eq. (8) simply becomes = −pI + (1 −˜d )) 2 @W0 @I1 + I1 @W0 @I2 B −2 @W0 @I2 B2 : (16) (16) It may be noted that this equation is similar to constitutive equations proposed by others for the Mullins e)ect (Gurtin and Francis, 1981; De Souza Neto et al., 1994; Miehe, 1995; Beatty and Krishnaswamy, 2000; ElJKas-ZJu˜niga and Beatty, 2002), even if some of these theories were not established using the CDM theory. 2.3. Discussion on the scalar measure of deformation state The last problem to solve is the choice of the measure of deformation state. In the general theory of thermodynamics of irreversible processes, the evolution equation of an internal variable has to be written in terms of the thermodynamical force associated with this variable, because it avoids considering constitutive equations that do not sat- isfy the second law of thermodynamics. In the present case, it means that the measure should be chosen as a function of Y, the thermodynamical force associated with d, that is equal to −W0. In fact, W0 can be seen as a measure of the deformation state and this choice is consistent with data provided in experiments. The Frst authors who proposed to apply CDM to the modelling of the Mullins e)ect adopted this ap- proach, because they decided to apply the CDM theory without deviation (Simo, 1987; De Souza Neto et al., 1994; Miehe, 1995). This choice exhibits one major problem: the form of the strain energy function should be postulated before derivation of the whole model. Moreover, due to physical considerations, it seems to be more appropriate to formulate explicitly the damage evolution equation in terms of the deformation state. That was proposed, in the context of CDM, by Govindjee and Simo (1991, 1992), and later by Miehe and Keck (2000). Note that this approach can be related to the use of the two-network theory ofGreen and Tobolsky (1946) or the two-phase theory ofMullins and Tobin (1957) Wineman and Rajagopal (1990), Johnson and Beatty (1993a), and Beatty and Krishnaswamy (2000) The evolution of the network is described by a reformation function that depends on the maximum strain state endured by the material during its life. In the aforementioned papers, the deformation state is measured through a scalar function which depends on I1 and I2. As proposed by Beatty and Krishnaswamy, this function can be seen as a damage parameter (see footnote 5 inBeatty and Krishnaswamy, 2000). First the basic application of the CDM theory is examined. Let (B)= f (W0(B)) be the measure of the deformation state in terms of the thermo-dynamical force associated with d f (W0) is adopted. Thus, even if f has not to be a norm, it can be reasonably stated that: • f ; + ∞), • f (0) = 0, i.e. • g I1 and I2 on [3; + ∞) × [3; + ∞), • g ; + ∞) × [3; + ∞), 2.3. Discussion on the scalar measure of deformation state • the choiceg(I1 = 3 ; I 2 = 3) = 0, that ensures that the measure is null for the virgin unloaded state, is recommended. 2.3. Discussion on the scalar measure of deformation state the measure of the undeformed state is null, • f is an increasing function on [0; + ∞). Consequently, f is a positive increasing function on ]0; + ∞). Using this measure of 14) becomes ˙d = h(f (W0)) f; W0 (W0;I1 ˙I1 + W0;I2 ˙I2): (17) ˙d = h(f (W0)) f; W0 (W0;I1 ˙I1 + W0;I2 ˙I2): Taking into account the E-inequalities proposed byTruesdell and Noll (1965, p. 158): W0;I1 ¿ 0 and W0;I2 ¿ 0 W0;I1 ¿ 0 and W0;I2 ¿ 0 (18) the previous Eq. (17) can be written as the previous Eq. (17) can be written as the previous Eq. (17) can be written as ˙d = f 1 ˙I1 + f 2 ˙I2; ˙d = f 1 ˙I1 + f 2 ˙I2; where ( f i ) i=1 ;2 are functions ofI1 and I2, and satisfy where ( f i ) i=1 ;2 are functions ofI1 and I2, and satisfy f 1 ¿ 0 and f 2 ¿ 0 on ]0; + ∞): f 1 ¿ 0 and f 2 ¿ 0 on ]0; + ∞): (20) The classical CDM approach being derived, the second method that consists in directly expressing the scalar measure of the deformation state in terms of the left Cauchy–Green tensor is examined. For isotropic and incompressible materials, the mea- sure (B (B) = g(I1; I 2): (21) Following exactly the same method as above, Eq. (14) gives ˙d = h(g(I1; I 2))( g;I ˙I1 + g;I ˙I2): (22) (B) = g(I1; I 2): (B) = g(I1; I 2): (21) Following exactly the same method as above, Eq. (14) gives ˙d = h(g(I1; I 2))( g;I1 ˙I1 + g;I2 ˙I2): (22) Comparing this equation with Eqs. (19) and (20), and recalling the properties ofh, the properties thatg should satisfy can be easily established: • g ; + ∞) × [3; + ∞), • g ; + ∞) × [3; + ∞), • g I1 and I2 on [3; + ∞) × [3; + ∞), • g;I1 ¿ 0 and g;I2 ¿ 0, • g;I1 ¿ 0 and g;I2 ¿ 0, 1 2 • the choiceg(I1 = 3 ; I 2 = 3) = 0, that ensures that the measure is null for the virgin unloaded state, is recommended. 3.1. Experimental results by the Trelleborg group and all experiments were conducted in its French Research Department. In order to determine material parameters, that are the virgin strain energy function W0 and the evolution equation of the damage variable ˜d ), cyclic uniaxial tensile and simple shear tests were performed. The uniaxial tensile experiments were conducted experiments were performed under enforced displacement conditions. For several strain levels from 25% up to 500% for uniaxial tensile and from 20% up to 480% for simple 4:167 s−1. An example of experimental data is depicted in Figs.1(a) for uniaxial tensile the maximum stress measured in uniaxial tension. In the case of uniaxial tension, is considered. It can be assimilated to the term “discontinuous damage” previously introduced byMiehe (1995). It is to be noted that reloading paths do not rejoin the primary loading curve at the maximum strain. In fact, as samples are unloaded to coupon specimens that buckle during unloading. Consequently the next reloading stretch Johnson and Beatty, 1993a). should be separated from other inelastic phenomena such as hysteresis and relaxation. Concerning the hysteresis phenomenon, unloading paths of cycles are not considered and it is assumed that the equilibrium paths are the loading paths. In fact, static paths that correspond to the thermodynamic equilibrium of the material are situated within hysteresis cycles, i.e. between loading and unloading paths, but their exact positions are not well-established ( ). Nevertheless, the stress-softening model can be develop for loading paths, then in further works the results obtained here phenomenon can be eliminated by correcting the reference dimensions of samples in Tensile stretch 1 2 3 4 5 6 Adimensional nominal stress 0.0 0.2 0.4 0.6 0.8 1.0 Simple shear stretch 0 1 2 3 4 Adimensional nominal shear stress 0.0 0.2 0.4 0.6 0.8 1.0 5 (a) (b) Fig. 1. Experimental results: (a) uniaxial tensile and (b) simple shear. First principal stretch 1 2 5 Adimensional nominal tensile stress 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 First principal stretch Adimensional nominal pure shear stress 4 6 3 1 2 5 4 3 (a) (b) Fig. 2. Corrected data: (a) uniaxial tensile and (b) pure shear. Tensile stretch 1 2 3 4 5 6 Adimensional nominal stress 0.0 0.2 0.4 0.6 0.8 1.0 Simple shear stretch 0 1 2 3 4 Adimensional nominal shear stress 0.0 0.2 0.4 0.6 0.8 1.0 5 (a) (b) Fig. 3.1. Experimental results 1. Experimental results: (a) uniaxial tensile and (b) simple shear. Tensile stretch 1 2 3 4 5 6 Adimensional nominal stress 0.0 0.2 0.4 0.6 0.8 1.0 (a) (b) Fig. 1. Experimental results: (a) uniaxial tensile and (b) simple shear. First principal stretch 1 2 5 Adimensional nominal tensile stress 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 First principal stretch Adimensional nominal pure shear stress 4 6 3 1 2 5 4 3 (a) (b) Fig. 2. Corrected data: (a) uniaxial tensile and (b) pure shear. Fig. 2. Corrected data: (a) uniaxial tensile and (b) pure shear. the unloaded conFguration for the next loadings. Practically, experimental results are corrected in three steps: (i) only loading paths are considered, (ii) reloading paths are shifted to zero to eliminate the permanent set and (iii) they are extended to rejoin the Frst loading path. Finally, note that the analysis of simple shear experiments leads to some diNculties for the determination of material parameters, because principal directions of strain do not remain constant in time. Here, in order to simplify the analysis, simple shear ex- perimental data are transformed into pure shear data, as proposed by Charlton et al. (1994). Note that pure shear data could have been obtained with an appropriate ex- perimental apparatus (Rivlin, 1948) but it is practically easier to consider simple shear experiments. The corrected experimental data are presented in Figs. 2(a) and (b) for uniaxial tensile and pure shear results, respectively. 3.2. Determination of material parameters 3.2. Determination of material parameters As shown in Section 2, the general framework of hyperelasticity with damage ne- cessitates the determination of two material functions: the virgin strain energy function W0 and the evolution equation of the stress-softening variable (Eqs. ( 15) and (21)). Forms of these two functions are established in the next paragraphs. 3.2.1. Strain energy function 3.2.1. Strain energy function ( 8) and (24), the stress–strain relationship reduces to y Using Eqs. ( 8) and (24), the stress–strain relationship reduces to = −pI + 2(1 −˜d )) 3 i=1 iC i0(I1 −3) i−1B: (25) (25) Hence, the analytical stress–strain equations corresponding to our experiments can be easily derived (seeOgden (1984) for details). First, in the case of uniaxial tensile tests, both the stretch and the nominal stress in the tensile direction are measured. The corresponding equation is given by = 2(1 −˜d )) −1 2 3 i=1 iCi0 2 + 2 −3 i−1 : (26) (26) Second, after transformation of simple shear results into pure shear data, the correspond- ing relation between the nominal stress and the stretch in the Frst principal direction, i.e. the one that corresponds to the maximum stretch, reduces to Second, after transformation of simple shear results into pure shear data, the correspond- ing relation between the nominal stress and the stretch in the Frst principal direction, i.e. the one that corresponds to the maximum stretch, reduces to = 2(1 −˜d( R )) −1 3 3 i=1 iCi0 2 + 1 2 −1 i−1 : (27) (27) 3.2.1. Strain energy function Several forms for the strain energy function of rubber-like materials have been pro- based on experimental observations and phenomenological considerations. The corre- Mooney, 1940; Rivlin and Saunders, 1951; Gent and Thomas, 1958; Hart-Smith, 1966; Ogden, 1972; Yeoh, 1990). The second group of models were developed by considering the physics of the elastomer network. First, conformations of a single polymer chain are determined; then, the behaviour of the whole network is derived using statistical developments. For small and moderate strains, the neo-hookean ( Treloar, 1944) and the phantom (James and Guth, 1947) models can be mentioned. For large strains, the stretching limit of chains is taken into account and non-Gaussian statistics are employed ( Kuhn and Grun, 1942; James and Guth, 1943; Treloar and Riding, 1979; Arruda and Boyce, 1993; Wu and van der Giessen, 1993). The use of these models reduces the number of rele- vant material parameters because of their physical foundations, but their mathematical derivations are more complicated than those of phenomenological constitutive equa- tions. For a complete review of the constitutive models for rubber elasticity, the reader can refer toBoyce and Arruda (2000). It has to be mentioned that the present approach can be easily applied to every form of the strain energy function. In the present study, the emphasis is laid on the stress-softening phenomenon, so that a simple phenomenological strain energy function is chosen. Then, recalling that in the case of ncompressible materialsW0 is a func- I1 and I2, the general expansion ofRivlin (1948) is considered: W0 = i;j Cij (I1 −3) i(I2 −3) j (23) W0 = i;j Cij (I1 −3) i(I2 −3) j (23) in which Cij are the material parameters. Several phenomenological models were de- rived by truncating the series Eq. (23). In order to satisfactorily reproduce the large strain response of the material, the Yeoh model is adopted ( Yeoh, 1990, 1993). The corresponding strain energy function is given by W0 = 3 i=1 Ci0(I1 −3) i; W0 = 3 i=1 Ci0(I1 −3) i; (24) (24) where C10, C20 and C30 are the three material parameters. As this model only depends not improve the Gaussian theory. Nevertheless, its mathematical simplicity and ability where C10, C20 and C30 are the three material parameters. As this model only depends not improve the Gaussian theory. Nevertheless, its mathematical simplicity and ability study. study. y Using Eqs. 3.2.2. Evolution equation of the damage variable Once the virgin strain energy function has been chosen, the form of the damage evolution equation (Eqs. (14) and (15)) has to be established. As detailed in the theoretical part of the paper (Section 2), the present approach focuses on the use of a measure of deformation expressed as a function of the two Frst invariants of B (Eq. (21)). As a Frst approach, we adopt a very simple measure of the deformation state that only depends on I1: = I1=3 −1 (28) = I1=3 −1 (28) in which I1=3 is the stretch of the diagonal of a material cube. This measure is rele- vant to the properties exhibited in the theoretical section: it is deFned and di)erentiable on [3; +∞)×[3; +∞), its partial derivative with respect to I1 is positive and its partial derivative with respect to I2 is equal to zero, and Fnally the measure is null as I1 = 3 (whatever the value of I2). Note that the change of deformation measure is obvious, the only problem being the availability of suNcient experimental results to identify the corresponding function. ˜ The damage function ˜d( ) will be now determined using experimental results. The general method followed here is described without reference to the kind of experiments, but results for both uniaxial tensile and pure shear data are presented. First, new notations are deFned. Consider experimental results as sketched in Fig. 3. Using reduced experimental data, the stress–strain relationship reduces to a sequence of loading curves at di)erent maximum strains. The corresponding curves can represent either uniaxial tensile and pure shear data. In this Fgure, the nominal stress is given as a function of the Frst invariant I1. Two types of stress–strain curves can be identiFed: the primary Ι1 Nominal stress 1 1I 2 1I iI1 p π 1 s π 2 s π i s π Fig. 3. Notations used for the construction of the evolution equation of the damage variable: (- -) primary loading curve and (—) secondary loading curves. Ι1 Nominal stress 1 1I 2 1I iI1 p π 1 s π 2 s π i s π Fig. 3. Notations used for the construction of the evolution equation of the damage variable: (- -) primary loading curve and (—) secondary loading curves. curve represented by a dashed line and secondary curves represented by solid lines. 3.2.2. Evolution equation of the damage variable Consequently, for two di)erent secondary curves numbered i and j (i ¡ j), the corre- sponding nominal stresses are simply related by a multiplicative factor: Consequently, for two di)erent secondary curves numbered i and j (i ¡ j), the corre- sponding nominal stresses are simply related by a multiplicative factor: j s is = k j s kis on [1; I i 1]: j s is = k j s kis on [1; I i 1]: (33) j s is = k j s kis on [1; I i 1]: (33) (33) I1 3.0 4.0 5.0 6.0 0.65 0.70 0.75 0.85 0.80 0.90 0.95 1.00 0.75 0.85 0.80 0.90 0.95 1.00 I1 4 10 12 3.5 4.5 5.5 Non-proportional zone B Non-proportional zone B Proportional zone A Proportional zone A 8 6 s 10/s 9 s 7/s 6 (a) (b) Fig. 4. Examples of ratio i+1 s =i s: (a) uniaxial tensile and (b) pure shear. (b) Fig. 4. Examples of ratio i+1 s =i s: (a) uniaxial tensile and (b) pure shear. This property is now thoroughly used to construct the evolution equation of d. The method employed here is similar to the one recently proposed by Beatty and Krishnaswamy (2000). Before determining the evolution equation, its validity should be examined. For each secondary curve numbered i, the ratio of the corresponding stress to the stress measured on the next loading curve can be drawn as a function of the deformation state (using I1 for example). An example of such a graph is depicted in Figs. 4(a) and (b) for uniaxial tensile and pure shear data, respectively. The shapes of all graphs are qualitatively similar. At small and moderate strain, the ratio is entirely constant; the corresponding region of the graph is denoted A in Fig. 4 and is termed “proportional”. At large strain, the ratio decreases as the deformation increases. It cor- responds with region B in the Fgure and this zone can be termed “non-proportional”. Depending on the maximum deformation measure that corresponds with the considered secondary curve, the range of stretch in which the relation is proportional changes. In other words, as I i 1 increases, the width of part A in Fig. 4 increases. So, the present theory is proved to be valid in the deformation range of the proportional part of the graph. 3.2.2. Evolution equation of the damage variable Depending on the type of curve, the stress–strain relationship di)ers: curve represented by a dashed line and secondary curves represented by solid lines. Depending on the type of curve, the stress–strain relationship di)ers: • for the primary curve, the current stress p is given by p = kp0; (29) p = kp0; where 0 is the nominal stress of the virgin material and kp depends on I1: where 0 is the nominal stress of the virgin material and kp depends on I1: kp = 1 −˜d( R ) with R = I1=3 −1: (30 (30) So, on the primary curve kp is decreasing as I1 increases, i So, on the primary curve kp is decreasing as I1 increases, • on the ith secondary curve delimited by its intersection at I i 1 with the primary curve, the current stress i s can be written as • on the ith secondary curve delimited by its intersection at I i 1 with the primary curve, the current stress i s can be written as i s = ki s0; (31 i s = ki s0; (31) i s = ki s0; (31) where ki s is a constant parameter given by where ki s is a constant parameter given by where ki s is a constant parameter given by ki s = 1 −˜d( R ) with R = I i 1=3 −1: (32 (32) Indeed, on this curve, the current value of is lower than R previously reached on the primary curve for I1 = I i 1. Indeed, on this curve, the current value of is lower than R previously reached on the primary curve for I1 = I i 1. k k1s = a + b exp −c I1=3 −1 −1 ; 3.2.2. Evolution equation of the damage variable The non-proportional part of the graph corresponds with strain-hardening of the material exhibited during the secondary loading i. This strain-hardening phenomenon joins the secondary loading curves with the primary master curve (see Fig. 3). In- applicability of the present model in this part of the graph is discussed in Section 3.2.3. The applicability range of our theory being determined, it is now possible to con- struct the evolution equation of the stress-softening variable in this range. In that way, experimental curves ki s=k1 s vs I i 1=3−1 are drawn in Figs. 5(a) for uniaxial tensile and (b) for pure shear data. Both curves have a similar exponential-like decreasing shape that can be Ftted by a relation of the following form: k k1s = a + b exp −c I1=3 −1 −1 ; (34) 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.2 0.4 0.6 0.8 1.0 ks i /ks 1 ks i /ks 1 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1 3 / 1 − iI 1 3 / 1 − iI (a) (b) Fig. 5. Evolution of the non-dimensional damage parameter ki s=k1 s as a function of the measure of deformation Ii 1=3 −1: (a) uniaxial tensile and (b) pure shear. 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.2 0.4 0.6 0.8 1.0 ks i /ks 1 ks i /ks 1 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1 3 / 1 − iI 1 3 / 1 − iI (a) (b) Fig. 5. Evolution of the non-dimensional damage parameter ki s=k1 s as a function of the measure of deformatio Ii 1=3 −1: (a) uniaxial tensile and (b) pure shear (b) Fig. 5. Evolution of the non-dimensional damage parameter ki s=k1 s as a function of the measure of deformation Ii 1=3 −1: (a) uniaxial tensile and (b) pure shear. where a, b and c are material parameters. Moreover, considering that in the initial unloaded state the material is undamaged, i.e. d = 0 and k = 1 for I1 = 3, then the evolution equation of k reduces to where a, b and c are material parameters. Moreover, considering that in the initial unloaded state the material is undamaged, i.e. 3.2.3. Model limitations and =tting results 3.2.3.1. Limitations As mentioned above, the continuum damage mechanics as applied to the modelling of the Mullins e)ect is only e)ective in the proportional part of the experimental curves (see Fig. 4). This limitation of the present theory is 3.2.3. Model limitations and =tting results 3.2.2. Evolution equation of the damage variable d = 0 and k = 1 for I1 = 3, then the evolution equation of k reduces to k = 1 + b′ exp −c I1=3 −1 −1 : (35 (35) Recalling that k = 1 −d, the evolution equation of d can be cast into the following form: Recalling that k = 1 −d, the evolution equation of d can be cast into the following form: ˜d( R ) = d∞ 1 −exp −R ; (36) (36) where the two new material parameters are the maximum damage d∞and a scale factor . They are related to the previous parameters by where the two new material parameters are the maximum damage d∞and a scale factor . They are related to the previous parameters by d∞= bk1 s and = 1=c: d∞= bk1 s and = 1=c: (37) d∞= bk1 s and = 1=c: (37) The form of the evolution equation Eq. (36) is similar to the one proposed phenomenologically by Miehe (1995), except that his model is expressed in terms of the strain energy function: the measure in Eq. (36) is replaced by W0. Finally, note that experimental damage curves for uniaxial tensile and pure shear data are qual- itatively similar but their corresponding parameters in Eq. (36) are di)erent (see Figs. 5(a) and (b)). This di)erence is also explained in the next discussion. The form of the evolution equation Eq. (36) is similar to the one proposed phenomenologically by Miehe (1995), except that his model is expressed in terms of the strain energy function: the measure in Eq. (36) is replaced by W0. Finally, note that experimental damage curves for uniaxial tensile and pure shear data are qual- itatively similar but their corresponding parameters in Eq. (36) are di)erent (see Figs. 5(a) and (b)). This di)erence is also explained in the next discussion. 3.2.3.1. Limitations As mentioned above, the continuum damage mechanics as applied to the modelling of the Mullins e)ect is only e)ective in the proportional part of the experimental curves (see Fig. 4). 3.2.2. Evolution equation of the damage variable This limitation of the present theory is First principal stretch Adimensional nominal shear stress 60 % 75 % 100 % 140 % 175 % 210 % 250 % 280 % 360 % 400 % First principal stretch 1 2 Adimensional nominal tensile stress 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 50% 75% 100% 150% 200% 250% 300% 350% 400% 500% 6 5 4 3 1 2 5 4 3 (a) (b) Fig. 6. Superimposition of the secondary loading curves: (a) uniaxial tensile and (b) pure shear. Note that strain-hardening is not satisfactorily reproduced. (b) Fig. 6. Superimposition of the secondary loading curves: (a) uniaxial tensile and (b) pure shear. Note that strain-hardening is not satisfactorily reproduced. highlighted in Fig. 6, in which experimental secondary loading curves were superim- posed using the scale factors ki s. It shows that the superimposition is almost perfect until the phenomenon of strain-hardening takes place. This phenomenon occurs as sec- ondary curves rejoin the primary curve and corresponds to the level of stress-softening previously endured by the material. In fact, this limitation is inherent in the choice of the damage mechanism: as the evolution equation of the damage variable is written in terms of a maximum deformation state (using the strain energy or a measure of defor- mation), theoretical secondary curves are proportional. Physically, this discrepancy can be explained by considering the evolution of the network and the breaking of di)erent links in it, as recently proposed by Marckmann et al. (2002). The second diNculty evoked in the previous paragraph is the quantitative di)erence between evolution equations of the stress-softening variable obtained with uniaxial ten- sile and pure shear experimental data. In our opinion, this di)erence is a consequence of the choice of the deformation measure. Here, for the sake of simplicity, the measure only depends on I1. In order to reproduce several loading conditions, the in;uence of I2 should be also retained. Nevertheless, it requires a large number of experimental data corresponding with di)erent loading conditions. Finally, the range of deformation in which the model is applicable should be esti- mated. Theoretically, the damage d should never be greater than 1. If the long-term damage parameter d∞in Eq. (36) is lower than 1, then d 6 1 is satisFed for all deformation states. The constraint d∞6 1 could be imposed during the identiFcation process. 3.2.2. Evolution equation of the damage variable Here, this constraint is not imposed and d∞is allowed to be greater than 1. In that case, the model remains valid only if d 6 1, and this condition can be written as I1 6 3 1 + ln d∞ d∞−1 : I1 6 3 1 + ln d∞ d∞−1 : (38) Table 1 Values of material parameters Parameters Values C10 3:99e−2 MPa C20 −4:05e−4 MPa C30 1:31e−4 MPa d∞ 1.744 3.85 Thus, for Fnite element applications, the previous inequality should be satisFed in all points of the structure in order to ensure the validity of the results. This is especially the case in parts of the mesh where important local deformation is predicted. 3.2.3.2. Fitting results Five material parameters have to be determined: three for the virgin strain energy, i.e. C10, C20 and C30, and two for the evolution equation of the damage, i.e. d∞and . Due the use of a measure of deformation instead of the strain energy function as the damage criterion, the evolution equation could be Ftted Frstly using data reported in Fig. 5; and then the determination of W0 reduces to the classical problem of Ftting a hyperelastic model using the primary loading curve. Nevertheless, taking into consideration limitations discussed above, it is more eNcient to perform a global identiFcation using the stress–strain data for both uniaxial tensile and pure shear tests. For the identiFcation process, the error between the model and experimental re- sults is calculated simultaneously on all curves. Computations are performed using two di)erent algorithms. First, an optimization program based on a genetic algorithm es- timates approximately the material parameters. Second, these parameters are precisely determined with the help of a classical steepest descent algorithm using the genetic algorithm results as initial guess solutions. The values of material parameters obtained are presented in Table 1 and the identiF- cation results are compared with experiments in Figs. 7(a) and (b) for uniaxial tensile and pure shear deformation state, respectively. Theoretical results are globally in good agreement with experiments; so, the model describes successfully the transition between di)erent loading curves in cyclic experiments and exhibits the in;uence of the max- imum strain on the softened behaviour. As anticipated, the present model is not able to satisfactorily predict the behaviour of the material at the intersection of secondary loading curves with the primary curve. 4. Numerical examples In order to demonstrate the eNciency of the present approach for engineering applications, the constitutive equation was implemented in the commercial software Abaqus (Hibbitt and Karlsson and Sorensen Inc., S., 1999) with the help of the UMAT facility. Both the Cauchy stress tensor and the fourth-order elasticity tensor are needed. An accurate derivation of this elasticity tensor is of major importance to ensure and accelerate convergence. Both plane stress and three-dimensional formulations were implemented. The plane stress implementation is quite simple, because the numerical diNculty due to the in- compressibility assumption is automatically satisFed through the change of thickness in the normal direction. Two examples are described in the next paragraphs. 3.2.2. Evolution equation of the damage variable Indeed, the high curvature of secondary load- ing curves due to strain-hardening is not well reproduced. Nevertheless, such results with only Fve material parameters are good, and the use of the present model in Fnite element simulations will provide quality results. Finally, considering Eq. (38) and the values of material parameters given in Table 1, the present approach is applicable until 630% of deformation in uniaxial tension and up to 400% in equibiaxial tension. In this way, simulation of large strain problems can be considered. First principal stretch Adimensional nominal pure shear stress First principal stretch 1 Adimensional nominal tensile stress 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 6 5 4 3 2 1 5 4 3 2 (a) (b) Fig. 7. IdentiFcation results, (a) uniaxial tensile and (b) pure shear: (· · · ◦· · ·) experiments and (—) our hyperelastic model with damage. (b) Fig. 7. IdentiFcation results, (a) uniaxial tensile and (b) pure shear: (· · · ◦· · ·) experiments and (—) our hyperelastic model with damage. (ii) the sheet is stretched to 185% in direction 2, unloaded, then stretched to 185% in direction 1 and Fnally unloaded (Fig. 8(c)), max d = 0:33, (i) the sheet is stretched to 185% in direction 2 then unloaded (Fig. 8(b)), max d = 0:33, 4.1. Square perforated sheet The Frst example is a well-known plane stress hyperelasticity problem where an initial 20 × 20 mm2 strip with a circular hole of diameter 10 mm is stretched in both horizontal and vertical directions. The corresponding initial mesh is shown in Fig. 8(a). This example will exhibit the strong dependence of stress-softening on the deformation history. Three di)erent deformation paths are investigated (amplitudes of cycles are identical) and their corresponding damage distributions are presented in Fig. 8: Fig. 8. Finite element simulation of a square perforated sheet: (a) undeformed mesh, (b) damage distribution after a cycle in dir. 2, (c) damage distribution after a cycle in dir. 2 followed by a cycle in dir. 1 and (d) damage distribution after a cycle in the diagonal direction. Fig. 8. Finite element simulation of a square perforated sheet: (a) undeformed mesh, (b) damage distribution after a cycle in dir. 2, (c) damage distribution after a cycle in dir. 2 followed by a cycle in dir. 1 and (d) damage distribution after a cycle in the diagonal direction. (iii) the sheet is stretched to 185 √ 2% in one of the diagonal directions then unloaded (Fig. 8(d)), max d = 0:35. The in;uence of the loading history on the damage distribution can be highlighted by comparing results (ii) and (iii). The two samples were subjected to the same deforma- tion state in both directions 1 and 2. Only loading paths di)er. In case (ii), the sample is Frst stretched then relaxed in direction 2. This Frst cycle weakens the structure in the neighbourhood of the hole, in direction 1 (see Fig. 8(b)). Then, the second cycle applied in direction 1 modiFes a new unloaded non-homogeneous damage distribution. So, the Fnal damage distribution is non-symmetric as shown in Fig. 8(c). In the load- ing case (iii), both cycles (in directions 1 and 2) are applied simultaneously and the Fnal damage distribution respects the structural symmetry as displayed in Fig. 8(d). In order to exhibit more precisely this phenomenon, consider Fig. 9 which presents the evolution of the damage state in elements denoted A and B in Fig. 8(a) for the two loading history (ii) and (iii). In this Fgure, it is clearly shown that case (ii) leads to a non-symmetric damage distribution. Numerical time increment 0 Damage 0.0 0.1 0.2 0.3 3 2 1 Fig. 9. 4.1. Square perforated sheet Evolution of the damage in elements lying on the edge of the hole: (- -▽- -) loading (ii) at point A, (- - - -) loading (ii) at point B, (—◦—) loading (iii) at points A and B. Numerical time increment 0 Damage 0.0 0.1 0.2 0.3 3 2 1 Damage Numerical time increment Numerical time increment Fig. 9. Evolution of the damage in elements lying on the edge of the hole: (- -▽- -) loading (ii) at point A, (- - - -) loading (ii) at point B, (—◦—) loading (iii) at points A and B. 4.2. Simultaneous uniaxial tension and torsion of a diabolo sample The second example is a three-dimensional problem that highlights the capability of our approach to predict the behaviour of complex parts. The structure is a rubber diabolo sample. Its length and width are, respectively, 30 and 20 mm. The sample is stretched in its length direction (uniaxial tension) and twisted by rotating one of its ends. Fig. 10(a) presents an example of damage distribution in the part. The maximum of damage takes place in the central region of the sample because of its geometry. Indeed, this sample geometry is used to maximize strain and stress in the minimum section. In order to investigate the in;uence of loading conditions on damage, the sample is twisted for di)erent levels of uniaxial tension. This tensile loading is performed by controlling the relative displacement of diabolo ends. The corresponding results are presented in Fig. 10(b), in which damage in the central region is drawn as a function of the torsional angle for di)erent values of the relative displacement of ends. As exhibited above, under uniaxial tensile conditions, i.e. when the torsional angle is null, damage is an increasing function of the tension. Moreover, for a given value of pre-tension, damage increases as the torsional angle increases, because the deformation measure also increases. Nevertheless, for a torsional angle equal to 2, the smaller is the pre-tension, the larger is the damage. This can be explained by the deformation measure adopted to drive the evolution of damage. In fact, this measure being a function of the Frst principal invariants of B, it takes into account the whole deformation state that may Torsional angle Damage 0.0 0.1 0.2 0.3 0.4 0 mm 5 mm 7.5 mm 10 mm 12.5 mm 15 mm 17.5 mm 20 mm π π 2 2 / π 2 / 3π 0 (a) (b) Fig. 10. (a) Damage distribution in the diabolo sample under simultaneous uniaxial tensile and torsional loading conditions. (b) Maximum damage vs torsional angle, in;uence of the tensile pre-loading. (a) Torsional angle Damage 0.0 0.1 0.2 0.3 0.4 0 mm 5 mm 7.5 mm 10 mm 12.5 mm 15 mm 17.5 mm 20 mm π π 2 2 / π 2 / 3π 0 (b) Damage Damage (b) (a) Fig. 10. (a) Damage distribution in the diabolo sample under simultaneous uniaxial tensile and torsional loading conditions. (b) Maximum damage vs torsional angle, in;uence of the tensile pre-loading. 4.2. Simultaneous uniaxial tension and torsion of a diabolo sample be larger without pre-tension. Indeed, for large torsional angles, local shear stresses dominate and increase as pre-tension decreases, this phenomenon being rendered by the measure of deformation. Finally, note that these samples will be used in further works to investigate crack initiation in elastomers under fatigue loading, more precisely to examine the in;u- ence of the deformation state on crack occurrence. As shown with these prelimi- nary simulations, the Mullins e)ect should be considered in fatigue problems because stress-softening highly reduces local sti)ness and will certainly in;uence the fatigue life of elastomers. 5. Conclusion The present work demonstrates the ability and limitations of Continuum Damage Mechanics to describe the Mullins e)ect in elastomers. The material considered here is natural rubber; it is assumed non-linear elastic, isotropic and incompressible, viscous e)ects not being taken into account. The general framework of hyperelasticity with damage is derived. Restrictions of this theory are exhibited and it is shown that CDM should be employed with care to model stress-softening in elastomers. Moreover, the connection with two-network (Green and Tobolsky, 1946) or two-phase (Mullins and Tobin, 1957) theories is demonstrated; the properties of the deformation measure used in the evolution equation of the damage variable are established. Then, this evolu- tion equation is constructed using both uniaxial tensile and simple shear experiments. The inapplicability of models of this kind in strain-hardening parts of the stress–strain curve is demonstrated. Nevertheless, the identiFcation of material parameters leads to good qualitative results. From a numerical point of view, the model is implemented in Abaqus for both plane stress and three-dimensional cases. Convergence of the model paths on the response of the material, and demonstrate the importance of considering stress-softening for the development of ndustrial parts. Finally, one last remark has to be made. Connections between the two phenomeno- is the determination of the evolution equation that drives the parameter used to cor- rect the strain-energy function, whatever the terminology employed: damage parameter, network reformation function or transformation function between hard and soft regions of rubber. Here, we showed that the use of the maximum deformation endured previ- strain-hardening. Thus, the evolution equation of the stress-softening variable should be also written in terms of the current deformation measure and not only in terms of the maximum deformation previously endured by the material. Such approaches were recently proposed byMiehe and Keck (2000) and El u˜niga and Beatty (2002). References Arruda, E.M., Boyce, M.C., 1993. A three dimensional constitutive model for the large stretch behavior of rubber elastic materials. J. Mech. Phys. Solids 41, 389–412. y Beatty, M.F., Krishnaswamy, S., 2000. A theory of stress-softening in incompressible isotropic materials. J. Mech. Phys. Solids 48, 1931–1965. elastomers. J. Mech. Phys. Solids 46, 931–954. Boyce, M.C., Arruda, E.M., 2000. Constitutive models of rubber elasticity: a review. Rubber Chem. Technol. 73, 505–523. Charlton, D.J., Yang, J., Teh, K.K., 1994. A review of methods to characterize rubber elastic beh De Souza Neto, E.A., Peric, D., Owen, D.R.J., 1994. A phenomenological three-dimensional rate-independent Phys. Solids 42, 1533–1550. El u˜niga, A., Beatty, M.F., 2002. A new phenomenological model for stress-softening in e Z. Angew. Math. Phys. 53, 794–814. Gent, A.N., Thomas, A.G., 1958. Forms of the stored (strain) energy function for vulcanized rubber. J. Polym. Sci. 28, 625–637. J. Polym. Sci. 28, 625–637. Int. J. Solids Struct. 29, 1737–1751. Int. J. Solids Struct. 29, 1737–1751. Green, M.S., Tobolsky, A.V., 1946. A new approach for the theory of relaxing polymeric media. J. Chem. Phys. 14, 87–112. Green, M.S., Tobolsky, A.V., 1946. A new approach for the theory of relaxing polymeric media. J. Chem. Phys. 14, 87–112. Gurtin, M.E., Francis, E.C., 1981. Simple rate-independent model for damage. J. Spacecraft 18, 285–286. Phys. 17, 608–626. Hibbitt and Karlsson and Sorensen Inc., S., 1999. ABAQUS/Standard User’s Manual V5. 8. Hibbitt and Karlsson and Sorensen Inc., S., 1999. ABAQUS/Standard User’s Manual V5. 8. Huntley, H.E., Wineman, A.S., Rajagopal, K.R., 1996. Chemorheological relaxation, residual stress and permanent set arising in radial deformation of an elastomeric hollow sphere. Math. Mech. Solids 1 267–299. Huntley, H.E., Wineman, A.S., Rajagopal, K.R., 1996. Chemorheological relaxation, residual stress and permanent set arising in radial deformation of an elastomeric hollow sphere. Math. Mech. Solids 1, 267–299. Huntley, H.E., Wineman, A.S., Rajagopal, K.R., 1997. Stress softening, strain localization and permanent set in the circumferential shear of an incompressible elastomeric cylinder. IMA J. Appl. Math. 59, 309–338. James, H.M., Guth, E., 1943. Theory of the elastic properties of rubber. J. Chem. Phys. 11, 455–481. James, H.M., Guth, E., 1947. Theory of the increase in rigidity of rubber during cure. J. Chem. Phys. 15, Huntley, H.E., Wineman, A.S., Rajagopal, K.R., 1997. Stress softening, strain localization and permanent set in the circumferential shear of an incompressible elastomeric cylinder. IMA J. Appl. Math. 59, 309–338. James, H.M., Guth, E., 1943. References Theory of the elastic properties of rubber. J. Chem. Phys. 11, 455–481. James, H.M., Guth, E., 1947. Theory of the increase in rigidity of rubber during cure. J. Chem. Phys. 15, 669–683. extension experiments. Continuum Mech. Thermodyn. 5, 301–318. vibration of rubber string. Continuum Mech. Thermodyn. 5, 83–115. Kachanov, L.M., 1958. Time of the rupture process under creep conditions. Izv. Akad. Nauk. S.S.R. Otd. Tech. Nauk. 8, 26–31. Kuhn, W., Gr un, F., 1942. Beziehungen zwichen elastischen Konstanten und Dehnungsdoppelbrechung mathematical representation. Continuum Mech. Thermodyn. 8, 153–169. Marckmann, G., Verron, E., Gornet, L., Chagnon, G., Charrier, P., Fort, P., 2002. A theory of network Maugin, G.A., 1999. The Thermodynamics of Nonlinear Irreversible Behaviors: an Introductio Maugin, G.A., 1999. The Thermodynamics of Nonlinear Irreversible Behaviors: an Introduction. In: World Miehe, C., 1995. Discontinuous and continuous damage evolution in Ogden-type large-strain elastic materials Eur. J. Mech. A/Solids 14, 697–720. 48, 323–365. Mooney, M., 1940. A theory of arge elastic deformation. J. Appl. Phys. 11, 582–592. Mullins, L., 1969. Softening of rubber by deformation. Rubber Chem. Technol. 42, 339–362. rubbers. Rubber Chem. Technol. 30, 551–571. Ogden, R.W., 1972. Large deformation isotropic elasticity—on the correlation of theory and experiment for incompressible rubberlike solids. Proc. Roy. Soc. London A 326, 565–584. Ogden, R.W., 1984. Non-Linear Elastic Deformations. Ellis Horwood, Chichester. Roy. Soc. London A 455, 2861–2877. Rajagopal, K.R., Wineman, A.S., 1992. A constitutive equation for nonlinear solids which undergo deformation induced by microstructural changes. Int. J. Plast. 8, 385–395. y g Rivlin, R.S., 1948. Large elastic deformation of sotropic materials—IV. Further developments of the general theory. Philos. Trans. Roy. Soc. A 241, 379–397. y y Rivlin, R.S., Saunders, D.W., 1951. Large elastic deformations of sotropic materials—vii. experiments on the deformation of rubber. Philos. Trans. Roy. Soc. A 243, 251–288. y computational aspects. Comput. Methods Appl. Mech. Eng. 60, 153–173. Simo, J.C., Hughes, T.J.R., 1998. Computational Inelasticity. Springer, New York. computational aspects. Comput. Methods Appl. Mech. Eng. 60, 153–173. Simo, J.C., Hughes, T.J.R., 1998. Computational Inelasticity. Springer, New York. Treloar, L.R.G., 1944. Stress–strain data for vulcanised rubber under various types of deformation. Trans. Faraday Soc. 40, 59–70. Treloar, L.R.G., Riding, G., 1979. A Non-Gaussian theory for rubber in biaxial strain. I. Mechanical properties. Proc. Roy. Soc. London A 369, 261–280. Truesdell, C., Noll, W., 1965. Handbuch der Physik Bd, III/3, The Nonlinear Field Theories of Mechanics. Springer, Berlin. Voyiadjis, G.Z., Ju, J.W., Chaboche, J.L. (Eds.), 1998. Yeoh, O.H., 1993. Some forms of the strain energy function for rubber. Rubber Chem. Technol. 66, 754–771. References Damage Mechanics in Engineering Materials, Studies in Applied Mechanics, Vol. 46. Elsevier, Amsterdam. pp Wineman, A.S., Rajagopal, K.R., 1990. On a constitutive theory for materials undergoing microstructural changes. Arch. Mech. 42, 53–75. Wu, P.D., van der Giessen, E., 1993. On improved network models for rubber elasticity and their applications to orientation hardening in glassy polymers. J. Mech. Phys. Solids 41, 427–456. Chem. Technol. 63, 792–805. Yeoh, O.H., 1993. Some forms of the strain energy function for rubber. Rubber Chem. Technol. 66, 754–771.
https://openalex.org/W2013651230	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0050018&type=printable	English	null	Re-Introduction of Transmembrane Serine Residues Reduce the Minimum Pore Diameter of Channelrhodopsin-2	PloS one	2,012	cc-by	10,369	Introduction quantitative description of the biphasic decay after the light is turned off [13]. In contrast, the four-state model can quantitatively describe the experimental values obtained from photocurrent decay with both fast and slow time constants (Figure 1A) [12–14]. Channelrhodopsin-2 (ChR2) is a light-activated cation channel found in the eyespot region of the green algae Chlamydomonas reinhardtii [1]. Together with channelrhodopsin-1, ChR2 helps mediate light-sensing and phototaxis [1–4] in C. reinhardtii. ChR2 is activated upon photoisomerization of the chromophore all-trans retinal with blue light [5,6]. Like other microbial-type rhodopsins, ChR2 is comprised of seven transmembrane domains [4,7,8] with a retinal moiety bound to a single lysine residue [9]. However, unlike other rhodopsins, ChR2 is able to conduct a large array of mono- and divalent cations with a high selectivity for protons [4]. Large cations such as guanidinium (ionic radius ,2.4 A˚ ) [10] and dimethyl-amine (ionic radius ,2.6 A˚ ) [10] can pass through the channel which suggests a large pore diameter [4,11]. Under continuous illumination, ChR2 current decays from an initial peak current to stationary current. The stationary current then decays biexponentially to a desensitized closed state once the light is turned off [4,12,13]. ChR2 shares a high sequence homology with other microbial type rhodopsins, but in contrast with these proteins, ChR2 functions mainly as a cation channel and not an ion pump. However, recent studies have shown that when ChR2 is reconstituted in planar lipid membranes in the absence of an electrochemical gradient, it functions as an outwardly rectified proton pump [17]. Thus, it has been proposed that ChR2 is a proton pump with an inherent leak current that exhibits ion channel properties [17]. ChR2 also undergoes large conformational changes when compared to other opsin proteins [18]. These structural changes have been attributed to the backbone movement of the channel through FTIR and Raman difference spectroscopy [18]. These changes, which occur before and after the conductive state, suggest that pore formation precedes gating upon activation of the channel. Furthermore, an analysis of the sequence of ChR2 reveals that this protein lacks specific residues which contribute to non-covalent forces in bacteriorhodopsin (bR). These residues have been shown to play a role in the hydrogen-bonding network and overall stability of bR [8,19,20]. Ryan Richards, Robert E. Dempski* y , p Department of Chemistry and Biochemistry, Worcester Polytechnic Institute, Worcester, Massachusetts, United States of America Department of Chemistry and Biochemistry, Worcester Polytechnic Institute, Worcester, Massachusetts, United States of America Abstract The funders had no role in st analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: rdempski@wpi.edu Competing Interests: The authors have declared that no competing interests exist. * E-mail: rdempski@wpi.edu * E-mail: rdempski@wpi.edu Abstract Channelrhodopsin-2 (ChR2) is a microbial-type rhodopsin found in the green algae Chlamydomonas reinhardtii. Under physiological conditions, ChR2 is an inwardly rectifying cation channel that permeates a wide range of mono- and divalent cations. Although this protein shares a high sequence homology with other microbial-type rhodopsins, which are ion pumps, ChR2 is an ion channel. A sequence alignment of ChR2 with bacteriorhodopsin, a proton pump, reveals that ChR2 lacks specific motifs and residues, such as serine and threonine, known to contribute to non-covalent interactions within transmembrane domains. We hypothesized that reintroduction of the eight transmembrane serine residues present in bacteriorhodopsin, but not in ChR2, will restrict the conformational flexibility and reduce the pore diameter of ChR2. In this work, eight single serine mutations were created at homologous positions in ChR2. Additionally, an endogenous transmembrane serine was replaced with alanine. We measured kinetics, changes in reversal potential, and permeability ratios in different alkali metal solutions using two-electrode voltage clamp. Applying excluded volume theory, we calculated the minimum pore diameter of ChR2 constructs. An analysis of the results from our experiments show that reintroducing serine residues into the transmembrane domain of ChR2 can restrict the minimum pore diameter through inter- and intrahelical hydrogen bonds while the removal of a transmembrane serine results in a larger pore diameter. Therefore, multiple positions along the intracellular side of the transmembrane domains contribute to the cation permeability of ChR2. Citation: Richards R, Dempski RE (2012) Re-Introduction of Transmembrane Serine Residues Reduce the Minimum Pore Diameter of Channelrhodopsin-2. PLoS ONE 7(11): e50018. doi:10.1371/journal.pone.0050018 Received July 31, 2012; Accepted October 17, 2012; Published November 19, 2012 Received July 31, 2012; Accepted October 17, 2012; Published November 19, 2012 ards, Dempski. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits tion, and reproduction in any medium, provided the original author and source are credited. Copyright: 2012 Richards, Dempski. This is an open-access article distributed under the terms of the Creative Commons Att unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. was funded by the Worcester Polytechnic Institute Research Foundation. The funders had no role in study design, data collection and ish, or preparation of the manuscript. Funding: The research was funded by the Worcester Polytechnic Institute Research Foundation. Re-Introduction of Transmembrane Serine Residues Reduce the Minimum Pore Diameter of Channelrhodopsin-2 Ryan Richards, Robert E. Dempski* Department of Chemistry and Biochemistry, Worcester Polytechnic Institute, Worcester, Massachusetts, United States of America Ryan Richards, Robert E. Dempski* Introduction The photocycle of ChR2 has been described by both a three-state [4,14,15] and four-state model [4,7,12], each of which contains several spectral intermediates as identified by time-resolved UV-Vis spectroscopy [16]. The three-state model can accurately predict the peak to stationary current in addition to the fast kinetic values associated with the rise and decay rates; however it fails to provide a November 2012 \| Volume 7 \| Issue 11 \| e50018 1 November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org Channelrhodopsin-2 Pore Diameter Non-covalent forces, such as inter- and intrahelical hydro- percentage of serine and threonine residues found wit Figure 1. Simplified photocycle and ChR2 single serine mutation model. (A) Simple four-state model of the channelrhodopsin-2 pho The two nonconducting states consist of a dark-adapted (C1) and desensitized (C2) state. The conducting states represent an early conduct (O1) and a late conducting state (O2). Blue arrows represent illumination with blue light. (B) Structural model of ChR2 based on the channelrh chimera C1C2 (PDB entry: 3UG9) highlighting the locations of serine mutations. Three residues (black) had highly reduced photocurrents mutations showed changes in pore size, permeability, and/or kinetics (orange). Residue S136 is shown in purple. The retinal chromophore, covalently bound to K257, is shown in yellow. This figure was prepared using Visual Molecular Dynamics [37]. (C) Sequence compa transmembrane domains between bacteriorhodopsin and channelrhodopsin-2. Residues that correspond to serine mutations are highli orange. Adapted from [9]. doi:10.1371/journal.pone.0050018.g001 Figure 1. Simplified photocycle and ChR2 single serine mutation model. (A) Simple four-state model of the channelrhodopsin-2 photocycle. The two nonconducting states consist of a dark-adapted (C1) and desensitized (C2) state. The conducting states represent an early conducting state (O1) and a late conducting state (O2). Blue arrows represent illumination with blue light. (B) Structural model of ChR2 based on the channelrhodopsin chimera C1C2 (PDB entry: 3UG9) highlighting the locations of serine mutations. Three residues (black) had highly reduced photocurrents, while 5 mutations showed changes in pore size, permeability, and/or kinetics (orange). Residue S136 is shown in purple. The retinal chromophore, which is covalently bound to K257, is shown in yellow. This figure was prepared using Visual Molecular Dynamics [37]. (C) Sequence comparison of transmembrane domains between bacteriorhodopsin and channelrhodopsin-2. Residues that correspond to serine mutations are highlighted in orange. Adapted from [9]. Reagents The mMessage mMachine SP6 kit was purchased from Ambion, Inc. (Austin, TX). Restriction enzymes were acquired from New England Biolabs, Inc. (Ipswich, MA). The Quikchange site-directed mutagenesis kit was bought from Stratagene. (La Jolla, CA). All other reagents were purchased from Sigma-Aldrich Corp. (St. Louis, MO) unless otherwise noted. The importance of hydrogen-bonding between specific residues has also been observed in ChR2 [7,16]. Mutations located at C128 have been shown to increase the conduction lifetime of ChR2 [26]. C128, located on helix 3, has been proposed to form a hydrogen bond with D156 on helix 4 and is important for proper channel function. This hydrogen bond pair is consistent through- out most microbial-type rhodopsins [27]. In bR, the pair corresponds to T90 and D115 where its contribution to protein stability has been studied previously [20,28]. Neither residue is directly involved in the proton conduction pathway, but disruption of this hydrogen-bond alters the kinetics and proton pumping ability. However, the recent crystal structure of the channelrho- dopsin chimera C1C2 shows that the distance between the thiol group of C167 and carboxyl oxygen of D196, which correspond to C128 and D156 in ChR2, is 4.4 A˚ , too large for hydrogen- bonding [29]. Instead, it was suggested that the thiol group interacts with the conjugated region of the retinal moiety [29]. Introduction doi:10.1371/journal.pone.0050018.g001 Non-covalent forces, such as inter- and intrahelical hydro- gen bonding [21–23] and helix-helix interactions [24,25] are critical for proper folding [25], insertion [23], and function [8,19] of membrane proteins. Specifically, serine and threo- nine residues have been shown to be involved in helical packing interactions and motifs [23]. Furthermore, a high percentage of serine and threonine residues found within the transmembrane domains of membrane proteins form hydrogen bonds with carbonyl oxygen atoms in the backbone structure [21] or other polar amino acids [19]. These hydrogen bonds stabilize protein structures and can be involved in ion transport across the cell membrane. November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 2 Channelrhodopsin-2 Pore Diameter The surgical protocol was approved by the WPI Institutional Animal Care and Use Committee (protocol number- A3277–01). Surgery was performed using the anesthetic tricaine to minimize animal suffering. A sequence alignment of bR and ChR2 reveals that there are eight serine residues within the transmembrane domains of bR that are not present in ChR2. Previously [20], these residues in bR were individually mutated to Ala in order to disrupt the hydrogen- bonding and proton transfer network. It was shown that some, but not all, serine mutations had altered proton pumping ability, kinetics, and chromophore regeneration. Additionally, all eight mutants were functional pumps. These experiments elucidated specific residues that contribute to the stability and structural changes of bR through hydrogen-bonding, but are not essential for proton transport. Oocyte Membrane Preparation and Western Blotting Oocyte Membrane Preparation and Western Blotting Xenopus oocyte total membrane fractions were prepared as described previously [34,35]. Briefly, oocytes were homogenized in 20 mL/oocyte homogenization buffer A (20 mM Tris, 5 mM MgCl2, 5 mM NaH2PO4, 1 mM EDTA, and 80 mM sucrose; pH 7.4) supplemented with 1 mM PMSF and 5 mg/ml leupeptin and pepstatin A, Homogenized oocytes were spun at 200 g for 5 minutes at 4uC. After removal of supernatant, centrifugation was repeated. The supernatant was removed and spun at 14,000 g for 20 minutes at 4uC to pellet the membrane. Protein samples were solubilized in 4 ml/oocyte Laemmli buffer and incubated at 37uC for 30 minutes for denaturing. Protein samples were separated on a 12% gel using SDS-PAGE and transferred onto PVDF membranes. The membranes were blocked in 3% BSA overnight at 4uC. Blocked membrane were incubated with anti-HA polyclonal antibody (1:1000; Thermo Fisher Scientific, Waltham, MA) for 1 hour at room temperature. Membranes were washed 3 times for 10 minutes in TBST (0.1%) and incubated with an AP conjugated anti-rabbit secondary antibody (1:5000; Rockland Inc., Gilbertsville, PA). The ChR2 band was visualized using developing solution (Bio-Rad Labora- tories Inc., Waltham, MA). Molecular Biology A truncated wild type channelrhodopsin-2 (ChR2, amino acids 1–309) with the hemagluttanin (HA) epitope sequence (YPYDVP- DYA) was unidirectionally cloned into the vector pTLN [32] using EcoRV and XbaI restriction sites. Serine mutations were made using the Quikchange site-directed mutagenesis kit. All mutations were verified by full gene sequencing. Materials and Methods Microelectrodes were fabricated by pulling borosilicate glass capillaries (World Precision Instruments, Sarasota, Florida) using a PC-10 pipette puller (Narishige, Japan). The microelectrodes had resistances between 0.5–2.0 MV. Oocytes were placed in an RC- 10 oocyte chamber (Warner Instruments, LLC; Hamden, Connecticut). Solutions were introduced to the chamber using a Oocyte Preparation and mRNA Synthesis Oocytes were isolated according to previously published protocols [33]. In brief, oocytes were obtained via collagenase digestion (3 mg/mL; 3 hours, 17uC, shaking) in ORI+ solution (90 mM NaCl, 2 mM KCl, 2 mM CaCl2, 5 mM MOPS; pH 7.4) after partial ovarectomy of Xenopus laevis frogs. ChR2 mRNA was prepared using the SP6 mMessage and mMachine kit. A volume of 50 nL of 1 mg/mL mRNA was injected into each oocyte and incubated in ORI+ containing 1 mg/ml gentamycin and 1 mM all- trans-retinal (1 mM stock in DMSO) in the dark at 17uC for 3–4 days. Why ChR2 undergoes much larger structural changes com- pared to other opsin proteins is still unknown. However, ChR2 lacks specific residues and motifs that contribute to inter- and intrahelical interactions that are present in other microbial-type rhodopsins. We hypothesize that reintroduction of the eight transmembrane serine residues present in bR, but not in ChR2, will restrict the conformational flexibility and reduce the pore diameter of ChR2. In our present work, we created single serine ChR2 constructs that correspond to homologous residues in bR. We used two-electrode voltage clamp techniques on Xenopus laevis oocytes to study channel permeability, kinetics, and pore size in varying alkali metal solutions. ChR2 has an estimated pore size [4] larger than voltage gated sodium channels (,4.7 A˚ ) [30] but smaller than the nicotinic acetylcholine receptor (,8.4 A˚ ) [10]. Minimum pore sizes can be determined using excluded volume theory [10,31] and can help elucidate specific residues involved in forming the narrowest region of the pore. Three substitutions, I197S, P234S, and M255S, resulted in largely reduced photocur- rents compared to wild-type ChR2. Furthermore, G224S showed changes in permeability and kinetics while C87S, V269S, G181S and Q210S showed changes in permeability and pore size with no effect on kinetics. Lastly, S136A resulted in a larger minimum pore diameter. These results indicate the importance of hydrogen- bonding on the ChR2 pore size, conformational flexibility, and ion selectivity. An analysis of our experiments demonstrates that the size selectivity filter for ion conductance is not located at a single region within the transmembrane domains. Instead determinants that contribute to ion conductance are distributed along the intracellular side of channelrhodopsin-2. Ethical Animal Treatment Ethical Animal Treatment Isolation of Xenopus laevis oocytes were performed in strict accordance to the guidelines given in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. PLOS ONE \| www.plosone.org November 2012 \| Volume 7 \| Issue 11 \| e50018 3 Channelrhodopsin-2 Pore Diameter P234S, and M255S was determined by Western blotting (Figure 2C). Both I197S and M255S have similar band intensity compared to WT ChR2, suggesting that these mutations have reduced functionality. Conversely, P234S showed very low expression compared to the wild-type. However, because of the low photocurrents I197S, P234S, and M255S were not analyzed further. The five mutants that were measured retained typical inwardly rectified behavior (Figure 2D). VC-6 6 channel perfusion valve control system (Warner Instru- ments LLC; Hamden, Connecticut). Activation of ChR2 was achieved using 75 W xenon arc lamp (Specialty Optical Systems, Inc., Dallas, Texas) and 2 mm light guide (,461021 photons s21 m2). Currents were measured using a Turbo-Tec 03X amplifier (npi electronic GmbH, Germany) set to voltage clamp mode. The membrane potential was varied from 2120 mV to +60 mV in 20 mV steps. Data was recorded using pClamp10 software (Axon Instruments, Inc., Burlingame, California) and further analyzed using Clampfit 10.2 and SigmaPlot. Apparent kinetic values were determined by fitting current traces to a biexponential equation at 2120 mV in Na+ solution. To measure permeability ratios of mutant ChR2 constructs, we measured shifts in reversal potential using Li+, K+, Rb+, and Cs+ solutions (pH = 9) (Figure 3) relative to Na+ for individual constructs. Typical electrophysiological recordings for each mutant are shown in Figure 2D. Mutants retained the inwardly rectified current of WT ChR2 in each alkali metal solution. Photocurrents were normalized to the largest signal in Na+ solution on the same day for Erev analysis. Large negative changes in Erev were observed for C87S (DErev = 24761 mV; Rb+) and V269S (DErev = 25261 mV; Rb+) in the presence of larger cations when compared to WT DErev values (DE- rev = 229.260.4 mV; Rb+). A negative shift in reversal potential indicates that the channel becomes less permeable to cation X+. In contrast, G224S had much smaller shifts in Erev (DE- rev = 214.260.1 mV; K+) which indicated alkali metals became more permeable relative to other constructs and the phenotype (DErev = 227.960.5 mV; K+). Structural Modeling of ChR2 Structural models were created using Visual Molecular Dynamics software [37] with PDB entry 3UG9 [29]. Serine residues were inserted into the structure by creating an automatic protein structure file (psf) from 3UG9. Next, the psf was modified using the mutator plugin for VMD to insert serine residues at the homologous positions in the model. We measured kinetics and steady-state to peak current ratios (Iss/Ip) to determine the effect of single serine mutations on the ChR2 photocycle. Current traces were fit in Na+ solution from the dark to peak region for on rates (ton), the peak to steady-state region for decay rates (tdecay), and the steady state to dark region for closing rates (toff) (Figure 4A). Additionally, opening and decay rates were fit at 2120 mV to ensure ChR2 was fully recovered. WT ChR2 had an off rate of 12.160.3 ms (260 mV), on rate of 4.660.2 ms, and decay rate of 12265 ms (n = 16) (Figure 4B and 4C). Each mutant ChR2 had similar results for channel closing except for G224S which showed a two-fold increase in channel closing (27.060.8 ms; n = 14; 260 mV) (Figure 4B). Decay rates for ChR2 mutants demonstrated a similar trend with the off rates (Figure 4C). Only G224S was slower with a decay rate of 17568 ms (n = 14). Single serine ChR2 on rates were statistically significant when compared to wild-type, except for C87S which had no change. G224S had the largest change in on rate with ton = 10.360.8 ms (n = 14). Q210S also showed a slight slowdown with ton = 5.560.1 ms (n = 8) (Figure 4C). G181S and V269S had ton values of 5.460.2 ms and 5.660.3 ms, respectively. Our kinetic data show that only G224S exhibited a global effect on the Minimum Pore Estimation The minimum pore diameter of ChR2 was calculated using excluded volume theory [10,36]. In brief, the permeability of an ion moving through a channel is proportional to the region left unoccupied by the ion. Plotting the square-root of permeability ratios versus ionic radius, a linear regression is observed according to the equation: ﬃﬃﬃﬃﬃﬃﬃﬃ Px PNa r ~a{bRx ð1aÞ ð1aÞ a~ Rc Rc{RNa ð1bÞ b~ 1 Rc{RNa ð1cÞ a~ Rc Rc{RNa ð1bÞ ð1bÞ b~ 1 Rc{RNa ð1cÞ Next, we used DErev to calculate relative permeability ratios, Px/PNa, of each ion for ChR2 mutants. Changes in Erev between X+ and Na+ for each mutant were used to exclude differences in the cytoplasmic ionic composition of oocytes. Ratios were calculated relative to 115 mM Na+ solution for each ChR2 mutant (Figure 3). Values.1 indicated a higher permeability for cation X+ over Na+, where values,1 indicated reduced perme- ability for cation X+. In general, the largest changes in permeability ratios were seen for the larger cations (K+, Rb+, and Cs+). V269S showed highly reduced permeability ratios when compared to WT in all solutions. In contrast, G224S had higher permeability ratios when compared to WT except for PCs/PNa, where it was comparable to WT. ð1cÞ where Rx is the radius of ion X+, Rc is the radius of the channel, and RNa is the radius of the sodium ion. Calculating 2a/b gives the minimum pore diameter of the channel. Ethical Animal Treatment Both G181S and Q210S had mixed results, with both larger and smaller shifts in Erev compared to WT ChR2 (Figure 3). Interestingly, three mutants (G87S, G181S, and V269S) had highly positive sodium Erev indicating an increased selectivity for this ion over the other alkali metals. Other mutants had Na+ Erev similar to wild type. Results Eight individual serine mutations were made at residues that corresponded to transmembrane serine residues in bR. Mutations were located in various areas of the transmembrane domain; C87S, G181S, G224S, and M255S were located near the retinal binding pocket, I197S, Q210S, and V269S were located at the cytoplasmic end of the transmembrane domain, and P234S was located at the extracellular end of the transmembrane domain (Figure 1B). ChR2 constructs were expressed in Xenopus laevis oocytes and photocurrents were recorded using two-electrode voltage clamp techniques. Photocurrent measurements were made in the presence of 115 mM alkali metal solution (115 mM XCl, 2 mM BaCl2, 1 mM MgCl2, Tris; pH 9). Function of mutant constructs was determined by measuring photocurrents at 2120 mV in 115 mM Na+ solution and compared to WT ChR2 photocurrents on the same day (Figure 2A). The mutants I197S, P234S, and M255S were functional, but had highly reduced photocurrents (Figure 2B). Surface expression of I197S, November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 4 Channelrhodopsin-2 Pore Diameter Figure 2. Representative ChR2 photocurrent traces recorded at 2120 mV. Black bars indicate illumination with 473 nm light. (A) Normalized photocurrent traces. Mutant photocurrents were normalized to WT ChR2 on the same day in 115 mM Na+ solution at pH 9. Values are reported as mean6S.E.M (n = 3–9). Statistically significant values are denoted by a (P,0.001; *P,0.04). (B) Reduced photocurrent for I197S compared to WT ChR2 on the same day. M255S and P234S had similar results. (C) Western blot of I197S, P234S, and M255S ChR2 mutants. ChR2-HA tagged band appears at ,35 kDa. (D) Typical electrophysiological recordings for single serine mutants. Mutants retained inward rectification of cations, but with altered permeability, kinetics, and/or inactivation ratios. doi:10.1371/journal.pone.0050018.g002 Figure 2. Representative ChR2 photocurrent traces recorded at 2120 mV. Black bars indicate illumination with 473 nm light. (A) Normalized photocurrent traces. Mutant photocurrents were normalized to WT ChR2 on the same day in 115 mM Na+ solution at pH 9. Values are reported as mean6S.E.M (n = 3–9). Statistically significant values are denoted by a (P,0.001; P,0.04). (B) Reduced photocurrent for I197S compared to WT ChR2 on the same day. M255S and P234S had similar results. (C) Western blot of I197S, P234S, and M255S ChR2 mutants. ChR2-HA tagged band appears at ,35 kDa. (D) Typical electrophysiological recordings for single serine mutants. Results Changes in reversal potential were calculated by subtracting Erev in cation X+ from Erev in Na+ (mean±S.E.M; n = 7–25). Permeability ratios were calculated using changes in reversal potential. Values are reported as mean6S.E.M (n = 7–25). For significance testing, mutant ChR2 values were compared to WT in the same solution. Statistically significant values are denoted by a (P,0.001). doi:10.1371/journal.pone.0050018.g003 Figure 3. Shifts in reversal potential and permeability ratio comparison of single serine ChR2 constructs. Changes in reversal potential were calculated by subtracting Erev in cation X+ from Erev in Na+ (mean±S.E.M; n = 7–25). Permeability ratios were calculated using changes in reversal potential. Values are reported as mean6S.E.M (n = 7–25). For significance testing, mutant ChR2 values were compared to WT in the same solution. Statistically significant values are denoted by a (P,0.001). doi:10.1371/journal.pone.0050018.g003 Figure 3. Shifts in reversal potential and permeability ratio comparison of single serine ChR2 constructs. Changes in reversal potential were calculated by subtracting Erev in cation X+ from Erev in Na+ (mean±S.E.M; n = 7–25). Permeability ratios were calculated using changes in reversal potential. Values are reported as mean6S.E.M (n = 7–25). For significance testing, mutant ChR2 values were compared to WT in the same solution. Statistically significant values are denoted by a (P,0.001). doi:10.1371/journal.pone.0050018.g003 Figure 3. Shifts in reversal potential and permeability ratio comparison of single serine ChR2 constructs. Changes in reversal potential were calculated by subtracting Erev in cation X+ from Erev in Na+ (mean±S.E.M; n = 7–25). Permeability ratios were calculated using changes in reversal potential. Values are reported as mean6S.E.M (n = 7–25). For significance testing, mutant ChR2 values were compared to WT in the same solution. Statistically significant values are denoted by a (P,0.001). doi:10.1371/journal.pone.0050018.g003 6.02 A˚ . The largest change in pore diameter was seen for V269S which had a pore size of 4.62 A˚ . The change in pore size was consistent with our permeability calculations for this mutant ChR2, where all cations had decreased Px/PNa. Calculating (1– a)/b (Equations 1b and 1c) gives the radius of a sodium ion permeating the channel. The value obtained for WT was 0.84 A˚ (Figure 6B). Calculated sodium radii were found to have ,15% error compared to the dehydrated radius of sodium (0.98 A˚ ) and within reason compared to previously published reports [10,39,40]. Results Mutants retained inward rectification of cations, but with altered permeability, kinetics, and/or inactivation ratios. doi:10.1371/journal.pone.0050018.g002 ChR2 photocycle and that several ChR2 mutants selectively increased the opening channel kinetics. diameter of each construct. We applied excluded volume theory using our calculated permeability ratios to estimate the narrowest region of the channel. The permeability of an ion permeating the channel is proportional to the region that is left unoccupied by the ion in the pore [10,36,38]. Thus, the diameter of the narrowest region of the pore can be determined by plotting the square root of permeability ratios Px/PNa versus ionic radii. Values of the y-intercept, a, and slope, b, were obtained by fitting the data with a linear regression (Figure 6). Fitting the square root of WT permeability ratios to a linear regression afforded values of 1.37 and 20.44 for a and b, respectively with an R2 value of 0.73. R2 values were consistent with previously published pore size calculations [36]. According to equation 1a, the minimum pore diameter for WT ChR2 was 6.22 A˚ , in good agreement with the previously described estimates [4]. Single serine mutations had calculated minimum pore sizes smaller than that of WT ChR2, ranging from 4.62 A˚ to 5.74 A˚ (Figure 6B) excluding G224S, which had a pore size of Steady-state to peak current ratios were also calculated for each mutant ChR2. Ratios were calculated in Li+, Na+, and K+ solutions by normalizing current traces at 2120 mV to 21 mA. Iss/Ip ratios are used as a measure of inactivation of the channel during prolonged light exposure. Additionally, because monova- lent cation selectivity decreases in tandem with an increase in proton selectivity during the steady state, these ratios can be used to infer increased cation selectivity in the absence of protons (pH = 9) [14]. G181S had larger ratios when compared to WT in the same solution, which indicated less inactivation as ChR2 transitions to the O2 state (Figures 1A and 5). Other ChR2 mutants had no significant difference or a change of Iss/Ip in only one solution (Figure 4C). To determine whether introducing serine residues into the transmembrane domain of ChR2 facilitates inter- and intrahe- lical hydrogen-bonding, we estimated the minimum pore November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 5 Channelrhodopsin-2 Pore Diameter Figure 3. Shifts in reversal potential and permeability ratio comparison of single serine ChR2 constructs. Discussion 6.02 A˚ . The largest change in pore diameter was seen for V269S which had a pore size of 4.62 A˚ . The change in pore size was consistent with our permeability calculations for this mutant ChR2, where all cations had decreased Px/PNa. Calculating (1– a)/b (Equations 1b and 1c) gives the radius of a sodium ion permeating the channel. The value obtained for WT was 0.84 A˚ (Figure 6B). Calculated sodium radii were found to have ,15% error compared to the dehydrated radius of sodium (0.98 A˚ ) and within reason compared to previously published reports [10,39,40]. ChR2 has become a popular tool in the neurosciences for the temporal and spatial activation of excitable cells with light [41– 44]. However, the full mechanism for ion conductance is not well understood. Furthermore, why ChR2 is an ion channel when it shares a high sequence homology with other microbial-type rhodopsins, which are ion pumps, continues to be an unanswered question. Previous experiments utilizing FT-IR difference spectra show that ChR2 undergoes large conformational changes in its backbone structure compared to bR [18]. A sequence analysis of these two proteins indicates ChR2 also lacks eight transmembrane serine residues present in bR. Considering that serine residues within transmembrane domains contribute to inter- and intrahe- lical hydrogen-bonding, we hypothesized that the lack of these serine residues were an important contributor to the molecular determinants that define ChR2 function. By reintroducing the eight transmembrane serine residues present in bR at homologous positions in ChR2, we have shown that the pore size is decreased. Also, replacement of S136 with alanine increased the pore Lastly, we replaced an endogenous serine residue located at position 136 with alanine in order to remove an inter- or intrahelical hydrogen bond. We hypothesized that this would cause an opposite effect on ChR2 by increasing the pore diameter. Using permeability ratios (Figure 3) and applying excluded volume theory, we determined the minimum pore diameter to be 7.04 A˚ with an estimated sodium radius of 0.86 A˚ (Figure 6). Removal of S136 provides further evidence that introduction of transmem- brane serine residues constrict the pore diameter through inter- or intrahelical hydrogen bonds. November 2012 \| Volume 7 \| Issue 11 \| e50018 November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 6 Channelrhodopsin-2 Pore Diameter Figure 4. Kinetic parameters for single serine ChR2 constructs. Discussion Kinetic values were obtained by fitting ChR2 photocurrents to a standard biexponential equation. All parameters were calculated in 115 mM Na+ solution at pH 9. For on and decay times, traces were fit at 2120 mV to ensure ChR2 was in the dark adapted state. Values reported as mean6S.E.M (n = 8–23). Significant differences are marked with a (P,0.05). (A) Summary of photocurrent exponential fits. On time was calculated from baseline current to peak current, decay time was calculated from peak to stationary current, and off time was calculated from stationary to baseline current when the light is switched off. (B) Off rate comparison of the slow G224S mutant versus WT at different membrane potentials. The off rate of ChR2 had a strong dependence on the holding potential of the cell. Diamond, WT; crossed X, G224S. All values were statistically significant (as described in C) (C) Summary of kinetic parameters and inactivation ratios for single serine ChR2 mutants. Values are reported as mean6S.E.M (n = 8–23). Statistically significant values are denoted by a (P,0.05). doi:10.1371/journal.pone.0050018.g004 Figure 4. Kinetic parameters for single serine ChR2 constructs. Kinetic values were obtained by fitting ChR2 photocurrents to a standard biexponential equation. All parameters were calculated in 115 mM Na+ solution at pH 9. For on and decay times, traces were fit at 2120 mV to ensure ChR2 was in the dark adapted state. Values reported as mean6S.E.M (n = 8–23). Significant differences are marked with a (P,0.05). (A) Summary of photocurrent exponential fits. On time was calculated from baseline current to peak current, decay time was calculated from peak to stationary current, and off time was calculated from stationary to baseline current when the light is switched off. (B) Off rate comparison of the slow G224S mutant versus WT at different membrane potentials. The off rate of ChR2 had a strong dependence on the holding potential of the cell. Diamond, WT; crossed X, G224S. All values were statistically significant (as described in C) (C) Summary of kinetic parameters and inactivation ratios for single serine ChR2 mutants. Values are reported as mean6S.E.M (n = 8–23). Statistically significant values are denoted by a (P,0.05). doi:10.1371/journal.pone.0050018.g004 Figure 4. Kinetic parameters for single serine ChR2 constructs. Kinetic values were obtained by fitting ChR2 photocurrents to a standard biexponential equation. All parameters were calculated in 115 mM Na+ solution at pH 9. Discussion This region has also been shown to form a hydrophobic pocket that contains a water molecule, which could form a hydrogen-bonded network between V269S and T139 [46,47]. We postulate that locking this helix into place through a water-facilitated hydrogen-bonding network causes the decrease in pore size for this ChR2 mutant. Although there are no explicit hydrogen-bonding partners closer than 3.78 A˚ , the exposure of this residue to the cytosolic fluid may facilitate a water-mediated hydrogen bonding network that holds TM1 in place during the photocycle. It has been suggested that TM1 may form the pore exit during the photocycle [29]. Additionally, it has been previously shown for the M intermediate in bR that S226 (V269S in ChR2) moves ,2 A˚ closer to L100 (T139 in ChR2) [45]. This region has also been shown to form a hydrophobic pocket that contains a water molecule, which could form a hydrogen-bonded network between V269S and T139 [46,47]. We postulate that locking this helix into place through a water-facilitated hydrogen-bonding network causes the decrease in pore size for this ChR2 mutant. diameter. An analysis of the crystal structure of the channelrho- dopsin chimera C1C2 (PDB entry: 3UG9) indicates that some serine mutations have the potential to form hydrogen-bonds with residues on opposite helices or with the carbonyl backbone. Serine residues were inserted into the crystal structure using the mutator plugin for VMD. Three ChR2 mutants, I197S, P234S, and M255S, showed little ion conductance upon activation with light. The expression of both I197S and M255S were similar to WT, suggesting that these two mutations cause decreased function. In contrast, the P234S had highly reduced expression compared to WT. The reduced photocurrent is therefore a consequence of expression and not an alteration to the channel. The C87S and G181S mutants also had large reductions in pore size (5.10 A˚ and 5.07 A˚ , respectively). G181S is located next to the b-ionone ring of the retinal moiety (,2.36 A˚ ), but has no effect on the kinetics of the photocycle. However, G181S is 2.08 A˚ from the backbone carbonyl of T159 and 3.61 A˚ from the hydroxyl group of T159, which provides two possible hydrogen-bonding sites. Discussion For on and decay times, traces were fit at 2120 mV to ensure ChR2 was in the dark adapted state. Values reported as mean6S.E.M (n = 8–23). Significant differences are marked with a (P,0.05). (A) Summary of photocurrent exponential fits. On time was calculated from baseline current to peak current, decay time was calculated from peak to stationary current, and off time was calculated from stationary to baseline current when the light is switched off. (B) Off rate comparison of the slow G224S mutant versus WT at different membrane potentials. The off rate of ChR2 had a strong dependence on the holding potential of the cell. Diamond, WT; crossed X, G224S. All values were statistically significant (as described in C) (C) Summary of kinetic parameters and inactivation ratios for single serine ChR2 mutants. Values are reported as mean6S.E.M (n = 8–23). Statistically significant values are denoted by a (P,0.05). doi:10.1371/journal.pone.0050018.g004 Figure 5. Steady-state to peak current ratios for WT and G181S ChR2. Ratios were calculated after normalization to 21 mA at 2120 mV in cationic solutions (115 mM XCl, 2 mM BaCl2, 1 mM MgCl2, Tris; pH 9; n = 14–18). Significant differences are marked with a (P,0.05). (A) Comparison of WT and G181S photocurrents in Na+ solution. (B) Iss/Ip comparison for WT and G181S in different cationic solutions. Wild-type ChR2 is colored white while G181S is colored black. doi:10.1371/journal.pone.0050018.g005 Figure 5. Steady-state to peak current ratios for WT and G181S ChR2. Ratios were calculated after normalization to 21 mA at 2120 mV in cationic solutions (115 mM XCl, 2 mM BaCl2, 1 mM MgCl2, Tris; pH 9; n = 14–18). Significant differences are marked with a *(P,0.05). (A) Comparison of WT and G181S photocurrents in Na+ solution. (B) Iss/Ip comparison for WT and G181S in different cationic solutions. Wild-type ChR2 is colored white while G181S is colored black. doi:10.1371/journal.pone.0050018.g005 November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 7 Channelrhodopsin-2 Pore Diameter Figure 6. Relationship of relative permeability and alkali ionic radii for ChR2 serine mutations. Relative permeability ratios were calculated using reversal potentials for each mutant. (A) Plotting the square root of relative permeability ratios vs. ionic radius yielded a linear fit for each data set. Solid line/diamond, WT; dashed line/square, V269S; broken line/X S136A. Discussion For all mutants, the larger alkali metal ions (K+, Rb+, and Cs+) were less permeable than the smaller alkali metal ions (Li+, Na+). These results were consistent with excluded volume theory [36]. (B) Minimum pore fitting parameters and pore size calcualtions for single serine ChR2 mutants and S136A. The coefficients a and b correspond to the y-intercept and slope of the fitted data, respectively. doi:10.1371/journal.pone.0050018.g006 Figure 6. Relationship of relative permeability and alkali ionic radii for ChR2 serine mutations. Relative permeability ratios were calculated using reversal potentials for each mutant. (A) Plotting the square root of relative permeability ratios vs. ionic radius yielded a linear fit for each data set. Solid line/diamond, WT; dashed line/square, V269S; broken line/X S136A. For all mutants, the larger alkali metal ions (K+, Rb+, and Cs+) were less permeable than the smaller alkali metal ions (Li+, Na+). These results were consistent with excluded volume theory [36]. (B) Minimum pore fitting parameters and pore size calcualtions for single serine ChR2 mutants and S136A. The coefficients a and b correspond to the y-intercept and slope of the fitted data, respectively. doi:10.1371/journal.pone.0050018.g006 Figure 6. Relationship of relative permeability and alkali ionic radii for ChR2 serine mutations. Relative permeability ratios were calculated using reversal potentials for each mutant. (A) Plotting the square root of relative permeability ratios vs. ionic radius yielded a linear fit for each data set. Solid line/diamond, WT; dashed line/square, V269S; broken line/X S136A. For all mutants, the larger alkali metal ions (K+, Rb+, and Cs+) were less permeable than the smaller alkali metal ions (Li+, Na+). These results were consistent with excluded volume theory [36]. (B) Minimum pore fitting parameters and pore size calcualtions for single serine ChR2 mutants and S136A. The coefficients a and b correspond to the y-intercept and slope of the fitted data, respectively. doi:10.1371/journal.pone.0050018.g006 Although there are no explicit hydrogen-bonding partners closer than 3.78 A˚ , the exposure of this residue to the cytosolic fluid may facilitate a water-mediated hydrogen bonding network that holds TM1 in place during the photocycle. It has been suggested that TM1 may form the pore exit during the photocycle [29]. Additionally, it has been previously shown for the M intermediate in bR that S226 (V269S in ChR2) moves ,2 A˚ closer to L100 (T139 in ChR2) [45]. November 2012 \| Volume 7 \| Issue 11 \| e50018 Channelrhodopsin-2 Pore Diameter Channelrhodopsin-2 Pore Diameter activation [48]. Furthermore, we observed an increase in steady- state to peak current ratios for G181S under continuous light conditions, which suggests that this residue has an indirect effect on ChR2 inactivation. This is most likely a result of an interaction with T159. The C87S mutant has one possible interaction with the backbone carbonyl of S63 at a distance of ,2.32 A˚ . This region of the pore is constricted through hydrogen-bonding of S63, E123 and N258. The introduction of a hydrogen-bond with C87S would further constrict this region by coupling TM1 and TM2. Interestingly, cysteine residues also have the propensity toward forming hydrogen-bonds similar to serine in membrane protein structures [49]. end or near the retinal binding pocket. Furthermore, the G224S mutant was the only construct that had a change in the apparent kinetic parameters of the channel. The slowdown in kinetics we observed was a global effect which persisted throughout the entire conducting state. Excluding G224S, serine mutants had little effect on apparent channel kinetics. This suggests that pore constriction is not a result of a global change in the protein, but rather a local effect arising from the introduction of a strong hydrogen bond donor. Calculation of the pore size can be determined both empirically and computationally [10,50]. Our approach used excluded volume theory, which has general applicability to the translocation of hydrated and dehydrated cations [31,34,36]. This method has been well established for accurate estimations of the minimum pore size for ion permeation [10,36]. It is important to note that pore calculations are determined using relative permeability ratios. Therefore, it is possible to see a net decrease in the permeability of all cations with no change in pore diameter. Likewise, a proportional increase in cation permeability would result in no change in pore size. An example of this is the G224S mutant. Both Q210S and G224S had small changes in the pore diameter. Analysis of the channelrhodopsin chimera crystal structure suggests that Q210S is oriented toward the lipid bilayer where it has no potential for hydrogen bonding to the channel backbone. This explains why only a small change in pore diameter was observed for this mutation. Lastly, G224S had nearly no change in pore diameter compared to WT ChR2. The side chain, similar to Q210S, is oriented into the bilayer. Channelrhodopsin-2 Pore Diameter There is potential for a hydrogen-bond to be formed to the carbonyl of F220, although the distance is 3.91 A˚ ; above the average for this type of interaction. The changes in kinetics may be explained by having a glycine at this position in the wild-type. G224 could facilitate a favorable helix-helix interaction during the photocycle. Addition of a hydrogen-bond would remove this interaction but the added rigidity of the protein would cause slower kinetics and a pore diameter consistent with the phenotype. Recently, Nack et al. reported that 0.3 H+/ChR2 are pumped during the photocycle, a contribution that we have not taken into account for pore size estimations [51]. However, the amount of protons pumped during the photocycle is small in comparison to the amount of ions flowing through the channel. Therefore, we believe that the proton pumping ability of ChR2 has a negligible effect on the pore size calculations. In summary, the recently elucidated crystal structure of the chimeric channelrhodopsin protein (C1C2) exhibited a large pore lining surface on the extracellular side of C1C2 [29]. The putative conducting pathway forms between transmembranes 1, 2, 3, and 7 and is lined with negatively charged residues. Our results suggest that the size selectivity filter is located further along the channel and at multiple points along the conductance pathway (Figure 1B). Furthermore, evidence has been presented that residues located adjacent to the retinal binding pocket, along the intracellular side of the protein, and an interaction between TM1 and TM2 all affect the permeability of ChR2. Thus, residues that are located on the cytoplasmic side of the channel are determinants of ion conductance. We also suggest that the high affinity for serine residues to form inter- and intrahelical hydrogen bonds is the cause for reduction in pore size, decrease in permeability ratios, and slowdown of kinetics. The channel activity of ChR2 appears to be a direct consequence of its ability to undergo large conformational changes in the backbone structure for pore formation. To further test our hypothesis that serine residues play an important role in the conformational flexibility of ChR2, we replaced a serine at position 136 with alanine. The results from experiments with S136A indicated a pore size larger than that of WT ChR2, which was the opposite effect that we observed when introducing serine residues into the transmembrane domain. 1. Sineshchekov OA, Kwang-Hwan Jung, Spudich JL (2002) Two rhodopsins mediate phototaxis to low- and high-inten‘sity light in chlamydomonas reinhardtii. Proceedings of the National Academy of Sciences of the United States of America 99: 8689–8694. Available: http://www.jstor.org/stable/3059077; http:// AU4SB9AX7M.search.serialssolutions.com?sid= CentralSearch:null&genre = article &atitle = Two+Rhodopsins+Mediate+Phototaxis+to+Low-+and+High- Intensity+Light+in+Chlamydomonas+reinhardtii&volume = 99&issue = 13&title = Proceedings+of+the+National+Academy+of+Sciences+of+the+United+States+of+ America&issn=0027-8424&date=2002-06-25&spage=8689&aulast=Sineshchekov &aufirst=Oleg via the Internet. 2. Braun FJ, Hegemann P (1999) Two light-activated conductances in the eye of the green alga volvox carteri. Biophysical Journal 76: 1668–1678. Available: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db = pubmed&cmd= Retrieve& dopt = AbstractPlus&query_hl = &itool = pubmed_docsum&list_uids = 10049347; http://AU4SB9AX7M.search.serialssolutions.com?sid = CentralSearch:null&genre =article&atitle=Two+light-activated+conductances+in+the+eye+of+the+green+alga +Volvox+carteri.&volume= 76&issue = 3&title = Biophysical+journal&issn = 0006- 3495&date=1999-03-01&spage=1668&aulast=Braun&aufirst=F via the Internet. 3. Ehlenbeck S, Gradmann D, Braun FJ, Hegemann P (2002) Evidence for a light- induced H(+) conductance in the eye of the green alga chlamydomonas Acknowledgments We thank Olga Gaiko, Sagar Antala, and Tuong-Vi Nguyen for their help in the laboratory and for invaluable discussions. Channelrhodopsin-2 Pore Diameter S136 resides on TM4, approximately 3.2 A˚ away from T149 on TM5, which provides a potential hydrogen bonding site. Mutation of S136 to alanine would remove this interaction and thus allow for pore expansion. These results suggest that the reduced pore diameter of single serine ChR2 mutants is caused by the introduction of a serine mediated hydrogen-bond. y y g We calculated permeability ratios for the five single serine insertions and a single serine removal as a measure of channel selectivity to further support the notion that introduction of hydrogen-bonds would restrict the conformational flexibility of the channel. In general, ChR2 mutants that decreased the size of the pore had reduced relative permeability ratios for alkali metals. The V269S mutant had the largest effect on permeability, which was expected because of the large change in pore diameter. C87S, G181S, and Q210S had permeability ratios smaller than wild type for most ions, but larger than V269S. G181S also was the only mutant ChR2 that had a change in steady-state to peak current ratios compared to WT ChR2. An analysis of our experimental results suggest that the size selectivity filter is not restricted to any specific location within the transmembrane domain, as these residues are located on different helices either at the intracellular Author Contributions Conceived and designed the experiments: RR RED. Performed the experiments: RR. Analyzed the data: RR RED. Contributed reagents/ materials/analysis tools: RR RED. Wrote the paper: RR RED. Conceived and designed the experiments: RR RED. Performed the experiments: RR. Analyzed the data: RR RED. Contributed reagents/ materials/analysis tools: RR RED. Wrote the paper: RR RED. Discussion This threonine residue has been shown to form part of the retinal binding pocket where it is important for photocycle kinetics, magnitude of photocurrent, and the light intensity required for The most drastic change in pore diameter was observed for the V269S mutant, which had permeability ratios smaller than WT ChR2. V269S resides on the intracellular end of TM 7 near the transmembrane and intracellular interface. This helix extends past the membrane ,18 A˚ into the cytosol [29]. The closest hydrogen- bonding acceptor is E273, where the carboxylate ion is ,3.78 A˚ away from the hydroxyl proton on V269S. V269S is also in close proximity to the intracellular ends of TM1 and TM2, but the structure is unresolved at the residues of interest (residues 80–88). November 2012 \| Volume 7 \| Issue 11 \| e50018 November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 8 November 2012 \| Volume 7 \| Issue 11 \| e50018 Channelrhodopsin-2 Pore Diameter 2+Photocycle&volume=98&issue=5&title=Biophysical+Journal&date=2010-01-01 &spage=753&aulast=Berndt&aufirst=Andre´ via the Internet. reinhardtii. Biophysical Journal 82: 740–751. Available: http://www.ncbi.nlm. nih.gov/entrez/query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus& query_hl = &itool = pubmed_docsum&list_uids = 11806916; http:// AU4SB9AX7M.search.serialssolutions.com?sid=CentralSearch:null&genre=article &atitle=Evidence+for+a+light-induced+H%28%2B%29+conductance+in+the+eye +of+the+green+alga+Chlamydomonas+reinhardtii.&volume=82&issue=2&title= Biophysical+journal&issn = 0006-3495&date = 2002-02-01&spage = 740&aulast = Ehlenbeck&aufirst=S via the Internet. 13. Nikolic K, Grossman N, Grubb MS, Burrone J, Toumazou C, et al. (2009) Photocycles of channelrhodopsin-2. Photochemistry and Photobiology 85: 400– 411. Available: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl = &itool = pubmed_docsum&list_uids = 19161406; http://AU4SB9AX7M. search.serialssolutions.com?sid=CentralSearch:null&genre = article&atitle = Photocycles +of+channelrhodopsin-2.&volume = 85&issue = 1&title = Photochemistry+and+ photobiology&issn = 0031-8655&date = 2009-01-01&spage = 400&aulast = Nikolic&aufirst = K via the Internet. 4. Nagel G, Szellas T, Huhn W, Kateriya S, Adeishvili N, et al. (2003) Channelrhodopsin- 2, a directly light-gated cation-selective membrane channel. Proceedings of the National Academy of Sciences of the United States of America 100: 13940–13945. Available: http://www.jstor.org/stable/3148891; http://AU4SB9AX7M.search.serialssolutions. com?sid = CentralSearch:null&genre = article&atitle = Channelrhodopsin- 2%2C+a+Directly+Light-Gated+Cation-Selective+Membrane+Channel&volume=100 &issue=24&title=Proceedings+of+the+National+Academy+of+Sciences+of+the+ United+States+of+America&issn = 0027-8424&date= 2003-11-25&spage = 13940 &aulast = Nagel&aufirst = Georg via the Internet. 14. Hegemann P, Ehlenbeck S, Gradmann D (2005) Multiple photocycles of channelrhodopsin. Biophysical Journal 89: 3911–3918. Available: http://www. ncbi.nlm.nih.gov/entrez/query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_hl = &itool = pubmed_docsum&list_uids = 16169986; http://AU4SB9AX7M.search.serialssolutions.com?sid = CentralSearch:null& genre = article&atitle = Multiple+photocycles+of+channelrhodopsin. &volume = 89&issue = 6&title = Biophysical+journal&issn = 0006- 3495&date = 2005-12-01&spage = 3911&aulast = Hegemann&aufirst = P via the Internet. g g 5. Lawson MA, Zacks DN, Derguini F, Nakanishi K, Spudich JL (1991) Retinal analog restoration of photophobic responses in a blind chlamydomonas reinhardtii mutant. evidence for an archaebacterial like chromophore in a eukaryotic rhodopsin. Biophysical Journal 60: 1490–1498. Available: http://www.ncbi.nlm.nih.gov/entrez/ query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl=&itool=pubmed_docsum&list_uids=1777569; http://AU4SB9AX7M.search. serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Retinal+analog +restoration+of+photophobic+responses+in+a+blind+Chlamydomonas+reinhardtii+ mutant.+Evidence+for+an+archaebacterial+like+chromophore+in+a+eukaryotic+ rhodopsin.&volume = 60&issue = 6&title = Biophysical+journal&issn = 0006- 3495&date=1991-12-01&spage=1490&aulast=Lawson&aufirst=M via the Internet. 15. Nikolic K, Degenaar P, Toumazou C (2006) Modeling and engineering aspects of Channelrhodopsin2 system for neural photostimulation. Engineering in Medicine and Biology Society, 2006 EMBS ‘06 28th Annual International Conference of the IEEE : 1626–1629. 16. Stehfest K, Ritter E, Berndt A, Bartl F, Hegemann P (2010) The branched photocycle of the slow-cycling channelrhodopsin-2 mutant C128T. Journal of Molecular Biology 398: 690–702. Available: http://www.ncbi.nlm.nih.gov/ entrez/query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl = &itool = pubmed_docsum&list_uids = 20346954; http://AU4SB9AX7M. search.serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = The +branched+photocycle+of+the+slow-cycling+channelrhodopsin-2+mutant+C128T. Channelrhodopsin-2 Pore Diameter &volume = 398&issue = 5&title = Journal+of+molecular+biology&issn = 0022- 2836&date=2010-05-21&spage=690&aulast=Stehfest&aufirst=K via the Internet. 6. Hegemann P, Ga¨rtner W,Uhl R (1991) All-trans retinal constitutes the functional chromophore in chlamydomonas rhodopsin. Biophysical Journal 60: 1477–1489. Available: http://www.sciencedirect.com/science?_ ob = GatewayURL&_method = citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B94RW-4V8X9MX- J&md5 = f6f0c7d2d07f09c8586cc8a1a73c73bc; http://AU4SB9AX7M.search. serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = All- trans+retinal+constitutes+the+functional+chromophore+in+Chlamydomonas+rhodopsin &volume = 60&issue=6&title=Biophysical+Journal&date=1991-01-01&spage=1477& aulast=Hegemann&aufirst=P via the Internet. 17. Feldbauer K, Zimmermann D, Pintschovius V, Spitz J, Bamann C, et al. (2009) Channelrhodopsin-2 is a leaky proton pump. Proceedings of the National Academy of Sciences of the United States of America 106: 12317–12322. Available: http://www.pnas.org/cgi/content/abstract/106/30/12317; http:// AU4SB9AX7M.search.serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Channelrhodopsin-2+is+a+leaky+proton+pump&volume = 106& issue = 30&title = PNAS&date= 2009-07-28&spage = 12317&aulast = Feldbauer& aufirst= Katrin via the Internet. g 7. Bamann C, Gueta R, Kleinlogel S, Nagel G, Bamberg E (2010) Structural guidance of the photocycle of channelrhodopsin-2 by an interhelical hydrogen bond. Biochemistry 49: 267–278. Available: http://find.galegroup.com/ openurl/openurl?url_ver = Z39.88-2004&url_ctx_fmt = info:ofi/ fmt:kev:mtx:ctx&req_dat = info:sid/gale:ugnid:mlin_c_worpoly&res_ id = info:sid/gale:AONE&ctx_enc = info:ofi:enc:UTF-8&rft_val_fmt = info:ofi/ fmt:kev:mtx:unknown&rft.artnum = A217560374; http://AU4SB9AX7M. search.serialssolutions.com?sid = CentralSearch:IAO&genre = article&atitle = Structural+guidance+of+the+photocycle+of+channelrhodopsin- 2+by+an+interhelical+hydrogen+bond.&volume=49&issue=2&title=Biochemistry &issn= 0006-2960&date = 2010-01-19&spage = 267&aulast = Bamann&aufirst= Christian via the Internet. 18. Radu I, Bamann C, Nack M, Nagel G, Bamberg E&H J (2009) Conformational changes of channelrhodopsin-2. Journal of the American Chemical Society 131: 7313–7319. Available: http://find.galegroup.com/openurl/openurl?url_ ver = Z39.88-2004&url_ctx_fmt = info:ofi/fmt:kev:mtx:ctx&req_dat = info:sid/ gale:ugnid:mlin_c_worpoly&res_id = info:sid/gale:AONE&ctx_ enc = info:ofi:enc:UTF-8&rft_val_fmt = info:ofi/fmt:kev:mtx:unknown&rft. artnum = A202727189; http://AU4SB9AX7M.search.serialssolutions. com?sid = CentralSearch:IAO&genre = article&atitle = Conformational+changes +of+channelrhodopsin-2.&volume=131&issue=21&title=Journal+of+the+American +Chemical+Society&issn=0002-7863&date=2009-06-03&spage =7313&aulast=Radu &aufirst=Ionela via the Internet. 8. Lanyi JK (2004) Bacteriorhodopsin. Annual Review of Physiology 66: 665–688. Available: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_hl = &itool = pubmed_docsum&list_ uids = 14977418 via the Internet. 19. Luecke H, Richter HT, Lanyi JK (1998) Proton transfer pathways in bacteriorho- dopsin at 2.3 angstrom resolution: 1. Science (New York, N.Y.) 280: 1934–1937. 20. Marti T, Otto H, Mogi T, Rosselet SJ, Heyn MP, et al. (1991) Bacteriorhodopsin mutants containing single substitutions of serine or threonine residues are all active in proton translocation. Journal of Biological Chemistry 266: 6919–6927. Available: http://xml.engineeringvillage2.org/controller/servlet/Controller?CID=expertSearch DetailedFormat&EISESSION=1_29c58e13084759c4941e4ses4&SYSTEM_USE_ SESSION_PARAM = true&SEARCHID = 1f99eea13077b5788b5744prod1data1& DOCINDEX = 1&PAGEINDEX = 1&RESULTSCOUNT = 1&database = 1& format = expertSearchDetailedFormat; http://AU4SB9AX7M.search. serialssolutions.com?sid=CentralSearch:BJJ&genre=article&atitle=Bacteriorhodopsin +mutants+containing+single+substitutions+of+serine+or+threonine+residues+are+all+ active+in+proton+translocation&volume=266&issue=11&title=Journal+of+Biological +Chemistry&issn = 0021-9258&date = 1991-01-01&spage= 6919&aulast =Marti& aufirst =Thomas via the Internet. 9. Nagel G, Szellas T, Kateriya S, Adeishvili N, Hegemann P, et al. References PLOS ONE \| www.plosone.org November 2012 \| Volume 7 \| Issue 11 \| e50018 9 Channelrhodopsin-2 Pore Diameter Ihara K, Umemura T, Katagiri I, Kitajima-Ihara T, Sugiyama Y, et al. (1999) Evolution of the archaeal rhodopsins: Evolution rate changes by gene duplication and functional differentiation. Journal of Molecular Biology 285: 163–174. 10.1006/jmbi.1998.2286. 42. 42. Bruegmann T, Malan D, Hesse M, Beiert T, Fuegemann CJ, et al. (2010) Optogenetic control of heart muscle in vitro and in vivo. Nature Methods 7: 897–900. Available: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl = &itool = pubmed_docsum&list_uids = 20881965; http://AU4SB9AX7M. search.serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Optogenetic+control+of+heart+muscle+in+vitro+and+in+vivo.&volume=7&issue =11&title=Nature+methods&issn=1548-7091&date=2010-11-01&spage=897& aulast=Bruegmann&aufirst=T via the Internet. 28. Pera´lvarez-Marı´n A, Ma´rquez M, Bourdelande J, Querol E, Padro´s E (2004) Thr-90 plays a vital role in the structure and function of bacteriorhodopsin. The Journal of Biological Chemistry 279: 16403–16409. J g y 29. Kato HE, Zhang F, Yizhar O, Ramakrishnan C, Nishizawa T, et al. (2012) Crystal structure of the channelrhodopsin light-gated cation channel. Nature 482: 369–374. 10.1038/nature10870. 30. Payandeh J, Scheuer T, Zheng N, Catterall WA. (2011) The crystal structure of a voltage-gated sodium channel. Nature 475: 353–358. 43. Lin JY, Lin MZ, Steinbach P, Tsien RY (2009) Characterization of engineered channelrhodopsin variants with improved properties and kinetics. Biophysical Journal 96: 1803–1814. Available: http://www.sciencedirect.com/science?_ ob = GatewayURL&_method = citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B94RW-4VRPN44- D&md5 = 436eef729f13d9e019c9994117fb63d1; http://AU4SB9AX7M. search.serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Characterization+of+Engineered+Channelrhodopsin+Variants+with+Improved +Properties+and+Kinetics&volume=96&issue=5&title=Biophysical+Journal&date= 2009-01-01&spage=1803&aulast=Lin&aufirst=John via the Internet. 31. Hille B (1973) Potassium channels in myelinated nerve. selective permeability to small cations: Selective permeability to small cations. The Journal of General Physiology 61: 669–686. 32. Lorenz L, Pusch M, Jentsch TJ (1996) Heteromultimeric CLC chloride channels with novel properties. Proceedings of the National Academy of Sciences of the United States of America 93: 13362–13366. Available: http://gateway.isiknowledge.com/ gateway/Gateway.cgi?&GWVersion=2&SrcAuth=SerialsSolutions&SrcApp=360& DestLinkType = FullRecord&DestApp = WOS&KeyUT = A1996VT05400123; http://AU4SB9AX7M.search.serialssolutions.com?sid=CentralSearch:EGQ&genre= article&atitle=Heteromultimeric+CLC+chloride+channels+with+novel+properties& volume = 93&issue = 23&title = PROCEEDINGS+OF+THE+NATIONAL +ACADEMY+OF+SCIENCES+OF+THE+UNITED+STATES+OF+AMERICA &date=1996-11-12&spage=13362&aulast=Lorenz&aufirst=L via the Internet. 2009-01-01&spage=1803&aulast=Lin&aufirst=John via the Internet p g J 44. Kleinlogel S, Feldbauer K, Dempski RE, Fotis H, Wood PG, et al. (2011) Ultra light-sensitive and fast neuronal activation with the Ca,SUP.2+,/SUP.- permeable channelrhodopsin CatCh. Nature Neuroscience 14: 513–518. Available: http://find.galegroup.com/openurl/openurl?url_ver = Z39.88- 2004&url_ctx_fmt = info:ofi/fmt:kev:mtx:ctx&req_dat = info:sid/ gale:ugnid:mlin_c_worpoly&res_id = info:sid/gale:AONE&ctx_ enc = info:ofi:enc:UTF-8&rft_val_fmt = info:ofi/fmt:kev:mtx:unknown&rft. artnum = A253536112; http://AU4SB9AX7M.search.serialssolutions. com?sid = CentralSearch:IAO&genre = article&atitle = Ultra+light- sensitive+and+fast+neuronal+activation+with+the+%5BCa.sup.2%2B%5D- permeable+channelrhodopsin+CatCh.&volume = 14&issue = 4&title = Nature+ Neuroscience&issn = 1097-6256&date = 2011-04-01&spage = 513&aulast = Kleinlogel&aufirst = Sonja via the Internet. 33. Channelrhodopsin-2 Pore Diameter search.serialssolutions.com?sid = CentralSearch:null&genre=article&atitle=Serine+ and+Threonine+Residues+Bend+%CE%B1-Helices+in+the+%CF%871%C2%A0% 3D%C2%A0g%E2%88%92+Conformation&volume=79&issue=5&title=Biophysical +Journal&date=2000-01-01&spage=2754&aulast=Ballesteros&aufirst=Juan via the Internet. +and+Biophysical+Research+Communications&date=2001-01-01&spage=683&aulast =Kamsteeg&aufirst=Erik-Jan via the Internet. 36. McKinnon NK, Reeves DC, Akabas MH (2011) 5-HT3 receptor ion size selectivity is a property of the transmembrane channel, not the cytoplasmic vestibule portals. The Journal of General Physiology 138: 453–466. 10.1085/ jgp.201110686. 23. Dawson JP, Weinger JS, Engelman DM (2002) Motifs of serine and threonine can drive association of transmembrane helices. Journal of Molecular Biology 316: 799–805. Available: http://www.ncbi.nlm.nih.gov/entrez/query. fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl = &itool = pubmed_docsum&list_uids = 11866532; http://AU4SB9AX7M. search.serialssolutions.com?sid= CentralSearch:null&genre= article&atitle= Motifs +of+serine+and+threonine+can+drive+association+of+transmembrane+helices. &volume = 316&issue = 3&title = Journal+of+molecular+biology&issn = 0022- 2836&date=2002-02-22&spage=799&aulast=Dawson&aufirst=J via the Internet. 37. Humphrey W, Dalke A, Schulten K (1996) VMD: Visual molecular dynamics. Journal of Molecular Graphics 14: 33–38. 10.1016/0263–7855(96)00018–5. 38. Hille B (1972) The permeability of the sodium channel to metal cations in myelinated nerve. The Journal of General Physiology 59: 637–658. 10.1085/ jgp.59.6.637. 39. Eickhorst AN, Berson A, Cockayne D, Lester HA, Khakh BS (2002) Control of P2X(2) channel permeability by the cytosolic domain. The Journal of General Physiology 120: 119–131. 10.1085/jgp.20028535. 40. McKinnon NK, Reeves DC, Akabas MH (2011) 5-HT3 receptor ion size selectivity is a property of the transmembrane channel, not the cytoplasmic vestibule portals. The Journal of General Physiology 138: 453–466. 10.1085/ jgp.201110686. 24. Adamian L, Jackups Jr R, Binkowski TA, Liang J (2003) Higher-order interhelical spatial interactions in membrane proteins. Journal of Molecular Biology 327: 251–272. 10.1016/S0022–2836(03)00041-X. 25. Popot JL, Engelman DM (1990) Membrane protein folding and oligomerization: The two-stage model. Biochemistry 29: 4031–4037. 41. Boyden E, Zhang F, Bamberg E, Nagel GD K (2005) Millisecond-timescale, genetically targeted optical control of neural activity. Nature Neuroscience 8: 1263–1268. Available: http://find.galegroup.com/openurl/openurl?url_ ver = Z39.88-2004&url_ctx_fmt = info:ofi/fmt:kev:mtx:ctx&req_dat = info:sid/ gale:ugnid:mlin_c_worpoly&res_id = info:sid/gale:AONE&ctx_ enc = info:ofi:enc:UTF-8&rft_val_fmt = info:ofi/fmt:kev:mtx:unknown&rft. artnum = A185538797; http://AU4SB9AX7M.search.serialssolutions. com?sid = CentralSearch:IAO&genre = article&atitle = Millisecond- timescale%2C+genetically+targeted+optical+control+of+neural+activity. &volume = 8&issue = 9&title = Nature+Neuroscience&issn = 1471- 003X&date = 2005-09-01&spage = 1263&aulast = Boyden&aufirst = Edward via the Internet. 26. Berndt A, Yizhar O, Gunaydin LA, Hegemann P, Deisseroth K (2009) Bi-stable neural state switches. Nature Neuroscience 12: 229–234. Available: http://www. ncbi.nlm.nih.gov/entrez/query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_hl = &itool = pubmed_docsum&list_uids = 19079251; http://AU4SB9AX7M.search.serialssolutions.com?sid = CentralSearch: null&genre = article&atitle = Bi-stable+neural+state+switches. &volume = 12&issue = 2&title = Nature+neuroscience&issn = 1097- 6256&date = 2009-02-01&spage = 229&aulast = Berndt&aufirst = A via the Internet. 27. Channelrhodopsin-2 Pore Diameter Richards R, Dempski RE (2011) Examining the conformational dynamics of membrane proteinsin situwith site-directed fluorescence labeling. Journal of Visualized Experiments. 10.3791/2627. Available: http://www.ncbi.nlm.nih. gov/entrez/query.fcgi?db = pubmed&cmd= Retrieve&dopt = AbstractPlus&query _hl = &itool= pubmed_docsum&list_uids = 21673634 via the Internet. 34. Du¨rr KL, Tavraz NN, Zimmermann D, Bamberg E, Friedrich T (2008) Characterization of na,K-ATPASE and H,K-ATPase enzymes with glycosylation- deficient b-subunit variants by voltage-clamp fluorometry in xenopus oocytes. Biochemistry 47: 4288–4297. Available: http://xml.engineeringvillage2.org/ controller/servlet/Controller?CID = expertSearchDetailedFormat&EISESSION= 1 _ c 2 e e 1 5 1 3 0 8 4 7 7 4 6 4 e 3 f d 3 s e s 1 & S Y S T E M _ U S E _ S E S S I O N _ PARAM=true&SEARCHID=b957ea13077dbb6d85599prod1data2&DOCINDEX = 1&PAGEINDEX = 1&RESULTSCOUNT = 1&database = 1&format = expertSearchDetailedFormat; http://AU4SB9AX7M.search. serialssolutions . com?sid = CentralSearch : BJJ & genre = article & atitle = C h a r a c t e r i z a t i o n + o f + N a % 2 C K - A T P A S E + a n d + H % 2 C K - ATPase+enzymes+with+glycosylation-deficient+%26beta%3B- subunit+variants+by+voltage-clamp+fluorometry+in+Xenopus+oocytes& v o l u m e = 4 7 & i s s u e = 1 4 & t i t l e = B i o c h e m i s t r y & i s s n = 0 0 0 6 - 2960&date = 2008-04-08&spage = 4288 via the Internet. 34. Du¨rr KL, Tavraz NN, Zimmermann D, Bamberg E, Friedrich T (2008) Characterization of na,K-ATPASE and H,K-ATPase enzymes with glycosylation- deficient b-subunit variants by voltage-clamp fluorometry in xenopus oocytes. Biochemistry 47: 4288–4297. Available: http://xml.engineeringvillage2.org/ controller/servlet/Controller?CID = expertSearchDetailedFormat&EISESSION= 1 2 1 5 1 3 0 8 4 7 7 4 6 4 3 f d 3 1 & S Y S T E M U S E S E S S I O N 45. Nicole Radzwill, Klaus Gerwert, Heinz-Jurgen Steinhoff (2001) Time-resolved detection of transient movement of helices F and G in double spin-labeled bacteriorhodopsin. Biophysical Journal 80: 2856–2866. 46. Sass HJ, Bu¨ldt G, Gessenich R, Hehn D, Neff D, et al. (2000) Structural alterations for proton translocation in the M state of wild-type bacteriorhodop- sin. Nature 406: 649–653. 10.1038/35020607. 47. Henderson R (2000) Molecular mechanism of vectorial proton translocation by bacteriorhodopsin. Nature 406: 653–657. 10.1038/35020614. 48. Channelrhodopsin-2 Pore Diameter (2005) Channelrhodopsins: Directly light-gated cation channels. Biochemical Society Transactions 33: 863–866. Available: http://www.ncbi.nlm.nih.gov/entrez/ query.fcgi?db = pubmed&cmd = Retrieve&dopt = AbstractPlus&query_ hl = &itool = pubmed_docsum&list_uids = 16042615; http://AU4SB9AX7M. search.serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Channelrhodopsins%3A+directly+light-gated+cation+channels.&volume = 33 &issue = Pt+4&title = Biochemical+Society+transactions&issn = 0300- 5127&date = 2005-08-01&spage = 863&aulast = Nagel&aufirst = G via the Internet. 10. Cohen BN, Labarca C, Davidson N, Lester HA (1992) Mutations in M2 alter the selectivity of the mouse nicotinic acetylcholine receptor for organic and alkali metal cations. The Journal of General Physiology 100: 373–400. 21. Gray TM, Matthews BW (1984) Intrahelical hydrogen bonding of serine, threonine and cysteine residues within a-helices and its relevance to membrane-bound proteins. Journal of Molecular Biology 175: 75–81. Available: http://www.sciencedirect.com/ science?_ob = GatewayURL&_method = citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B6WK7-4DN8W90- GS&md5=d4e9737882860a4b6a08d41cbe56a454; http://AU4SB9AX7M.search. serialssolutions.com?sid=CentralSearch:null&genre=article&atitle=Intrahelical+hydrogen+ bonding+of+serine%2C+threonine+and+cysteine+residues+within+%CE%B1- helices+and+its+relevance+to+membrane-bound+proteins&volume=175&issue=1&title= Journal+of+Molecular+Biology&date=1984-01-01&spage=75&aulast=Gray&aufirst=T via the Internet. 11. Tsunoda SP, Hegemann P (2009) Glu 87 of channelrhodopsin-1 causes pH- dependent color tuning and fast photocurrent inactivation. Photochemistry and Photobiology 85: 564–569. Available: http://AU4SB9AX7M.search. serialssolutions.com?sid = CentralSearch:AKQ&genre= article&atitle = Glu+87+ of+Channelrhodopsin-1+Causes+pH-dependent+Color+Tuning+and+Fast+Photocurrent +Inactivation&volume=85&issue=2&title=Photochemistry+and+Photobiology&issn = 0031-8655&date=2009-03-01&spage=564&aulast=Tsunoda&aufirst=Satoshi via the Internet. 12. Berndt A, Prigge M, Gradmann D, Hegemann P (2010) Two open states with progressive proton selectivities in the branched channelrhodopsin-2 photocycle. Biophysical Journal 98: 753–761. Available: http://www.sciencedirect.com/ science?_ob = GatewayURL&_method = citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B94RW-4YGHRMD- 2&md5 = 0d5fe3e1a9f27ab628a00b2dae529262; http://AU4SB9AX7M.search. serialssolutions.com?sid = CentralSearch:null&genre = article&atitle = Two+Open +States+with+Progressive+Proton+Selectivities+in+the+Branched+Channelrhodopsin- 22. Ballesteros JA, Deupi X, Olivella M, Eric E.J Haaksma, Pardo L (2000) Serine and threonine residues bend a-helices in the x1 = g2 conformation. Biophysical Journal 79: 2754–2760. Available: http://www.sciencedirect.com/science?_ ob = GatewayURL&_method = citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B94RW-4TYB1G0- 1M&md5 = 94097dc101264836aabfc6dc32f3ffc2; http://AU4SB9AX7M. PLOS ONE \| www.plosone.org November 2012 \| Volume 7 \| Issue 11 \| e50018 10 Channelrhodopsin-2 Pore Diameter Berndt A, Schoenenberger P, Mattis J, Tye KM, Deisseroth K, et al. (2011) High-efficiency channelrhodopsins for fast neuronal stimulation at low light levels. Proceedings of the National Academy of Sciences of the United States of America 108: 7595–7600. 10.1073/pnas.1017210108. 35. Erik-Jan Kamsteeg, Peter MTD (2001) Detection of aquaporin-2 in the plasma membranes of oocytes: A novel isolation method with improved yield and purity. Biochemical and Biophysical Research Communications 282: 683–690. Available: http://www.sciencedirect. com/science?_ob = GatewayURL&_method= citationSearch&_urlVersion = 4&_ origin = SERIALSCENT&_version = 1&_uoikey = B6WBK-45S4JS4- 12S&md5 = 28962839b74ea2164a7d77c3d127d24a; http://AU4SB9AX7M. search.serialssolutions.com?sid=CentralSearch:null&genre=article&atitle=Detection +of+Aquaporin-2+in+the+Plasma+Membranes+of+Oocytes%3A+A+Novel+Isolation+ Method+with+Improved+Yield+and+Purity&volume=282&issue=3&title=Biochemical 49. Gregoret LM, Rader SD, Fletterick RJ, Cohen FE (1991) Hydrogen-bonds involving sulfur-atoms in proteins,br/.. Proteins-Structure Function and Genetics 9: 99–107. 50. Smart OS, Breed J, Smith GR, Sansom MS (1997) A novel method for structure-based prediction of ion channel conductance properties. Biophysical Journal 72: 1109–1126. 10.1016/S0006–3495(97)78760–5. 51. Nack M, Radu I, Schultz B, Resler T, Schlesinger R, et al. (2012) Kinetics of proton release and uptake by channelrhodopsin-2. FEBS Letters 586: 1344– 1348. 10.1016/j.febslet.2012.03.047. November 2012 \| Volume 7 \| Issue 11 \| e50018 PLOS ONE \| www.plosone.org 11
W2997070819.txt	https://hal.umontpellier.fr/hal-02402300/document	en		Main rivers of watershed Volta, Africa	HAL (Le Centre pour la Communication Scientifique Directe)	2,018	public-domain	571	5°O 0° ² Algérie Mauritanie Mali Mauritanie Niger Burkina Faso Benin Côte d'Ivoire Ghana Togo Nigeria Burkina Faso Libéria 15°N 15° Mali NAKONG g pon Bek Kule Kutong Disa ndj ari Oti Pe SABARI Kou rtiag ou ALEHERIDE BONGOULOU SOKODE Benin Togo Nigeria osu Ob yim Afra On i m br i Su bon g Asu bon o r Wo m D DZOGBEGAN i ay Fo HO METEO ed e m ru Ak paw m D 5°N DJOUGOU KPESSIDE LAMA KARA 1 KPETCHU Apap a Ku YENDI METEO Dak poin Paw m MISIKROM SENCHI FERRY LOME Vo lta ACCRA Cours principal Tributaires 10° N'NABOUPI ADA METEO Reseau_hydro_Modele_SIEREM Tributaires NATITINGOU KETE-KRACHI Ghana e Sis m ra Af Bassin Versant de la Volta de l'origine des postes de mesures à 2016 TIELE YEJI e fi in Tan Sa Bakes i ASUBENDE Tan fi KUMASSI Ya pit i NABOULGOU Oti Pe ndjari Juni Datib o ri nja Bo Ny om on Fak o Aje rantie Gb on bo n Chuko Pa sso KALBRUIPE Jo lo ulo BORGOU SANSANNE MANGO u angop Koum SABOBA Ko lus i jar nd e iP Ot PORGA ARLY Pod iéga Kara Boinya élé BENJA DABOYA TAMALE METEO Sene Bresu Sa re i se Y bo EJURA ng a f im r i Ch bon Dra Ong wam Ghana w Kula Tain Nasia Jeba TAINSO AMONT BOUAKE NAWUNI hi nc Bo BAMBOI NASIA re nf Bo BONDOUKOU Côte d'Ivoire Niam w Pa Laboni TAGADI n ou uh Mo Zola BOLE METEO Nakambé do Ku o gb on Gb Koh odio Ku ld Na a lor u Denbu ini ny na Bo VONKORO le Du o Yaég Guiméb é NAKPANDURI DAPAONG Ku lda lo Sa u so as b n né uè NOUMBIEL Ba Po i ur Koulbi nt Ko Koulda Bin éda NANGODI KOMPIENGA Nagla Aruba Dagere WA METEO BATIE Pale Kolodio ila BITTOU mb Le YARUGU Kumo Boin DAPOLA BAGRE Pa nt ya ni ri Bo NAVRONGO TENKODOGO i Sis sili Do PONT D'OUESSA ga n DIEBOUGOU ba m Ka GAOUA 10°N KAMPALA AMONT Do ub od o Bo ko un go u Do ub ég ué NEBBOU FADA N GOURMA No uh ao Ka dé ga Koulip élé a ss Nia o ok NOBERE lou u Ko atue Agrum BANFORA a NIAOGHO a Asibelik ib ur go u Bo DAKAYE Kabo uti igui ir u g Kaba rvaro Bou DAN POUYTENGA Yamoko ourou Sanbay BOBO DIOULASSO GUENA Bo lo BOROMO a dag Pèn Da ka é Ko ray Ka ro uk o Gra nd Ba lé Tu i Ko do no u Dougbéra Bao ou ulé Kou Maran kandé Koutoum TENADO OUAGADOUGOU WAYEN SAKOINSE Na zin on ar ndj Pe Oti a SAMANDENI BADARA POUN KOUDOUGOU BISSIGA LOUMBILA Barrage BARRAGE de ZIGA lé ua O Léy ess Massili é mb ka Na o or ni Sa é ék Ti Sio u Voun Hou TOUROUBA ré Wéna NWOKUY Burkina Faso ala téb é i K Kalu raku n TOUMANI KOURI Voun Hou KAYA YILOU MANE Ku r Barrage du sourou a YARAN KONGOUSSI Singou Barrage de Dourou (Kanazoe) DI Bo m bo Do ré ug M o ou on ul Le o a m Sourou RAMBO M o OUAHIGOUYA Mo rag o BAI Bourzanga Bouko Niger Océan Atlantique ABIDJAN Poste de mesure Bassin versant VOLTA Limites politiques 0 50 100 200 Kilometres 5°N © SIEREM Déc. 2018 N. Rouché
https://openalex.org/W4313544866	https://discovery.dundee.ac.uk/ws/files/92955390/sciadv.abn5429.pdf	English	null	Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo	Science advances	2,023	cc-by	12,768	Citation for published version (APA): Chuai, M., Serrano Najera, G., Serra, M., Mahadevan, L., & Weijer, C. J. (2023). Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo. Science Advances, 9(1), Article eabn5429. https://doi.org/10.1126/sciadv.abn5429 INTRODUCTION as part of robust developmental control mechanisms remains to be resolved (11–14). During gastrulation, embryos transform from a simple epithelial layer into a three-layer structure. In vertebrate embryos, this process requires large morphogenetic movements that locate the three primary germ layers—the ectoderm, mesoderm, and endo- derm—in their topologically correct positions (1, 2). However, the detailed tissue movements and morphology of the embryo during gastrulation vary greatly among vertebrate animals (3). For example, Xenopus forms a blastopore through which the mesoderm and endoderm precursors internalize as epithelial sheets, while, at the other end of the spectrum, amniote embryos such as chick, mouse, and human embryos internalize these precursors as mesen- chymal cells through a structure known as the primitive streak (4, 5). The morphogenetic tissue flows in all organisms result from the integration of a few critical cell behaviors, including cell division, differentiation, shape changes, and movement. These cell behaviors are patterned and coordinated through a combination of short- and long-range chemical and mechanical signals (6–8). Different spatio- temporal organizations of these cell behaviors, in conjunction with geometrical and mechanical constraints imposed by the size and shape of the embryo and extraembryonic tissues, are thought to de- termine the diverse morphogenetic programs in vertebrate gastru- lation (9, 10). However, how the large-scale cell movements are coordinated and coupled via feedback to the signaling processes During gastrulation, embryos transform from a simple epithelial layer into a three-layer structure. In vertebrate embryos, this process requires large morphogenetic movements that locate the three primary germ layers—the ectoderm, mesoderm, and endo- derm—in their topologically correct positions (1, 2). However, the detailed tissue movements and morphology of the embryo during gastrulation vary greatly among vertebrate animals (3). For example, Xenopus forms a blastopore through which the mesoderm and endoderm precursors internalize as epithelial sheets, while, at the other end of the spectrum, amniote embryos such as chick, mouse, and human embryos internalize these precursors as mesen- chymal cells through a structure known as the primitive streak (4, 5). The formation of the primitive streak in the chick embryo in- volves intercalation of cells in the sickle-shaped mesendoderm pre- cursor domain initially located in the posterior epiblast of the embryo followed by ingression of these cells in the primitive streak (15, 16). University of Dundee Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo Chuai, Manli; Serrano Najera, Guillermo; Serra, Mattia; Mahadevan, Lakshminarayanan; Weijer, Cornelis J. Published in: Science Advances DOI: 10.1126/sciadv.abn5429 Publication date: 2023 Licence: CC BY Document Version Publisher's PDF, also known as Version of record Link to publication in Discovery Research Portal Citation for published version (APA): Chuai, M., Serrano Najera, G., Serra, M., Mahadevan, L., & Weijer, C. J. (2023). Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo. Science Advances, 9( Article eabn5429. https://doi.org/10.1126/sciadv.abn5429 University of Dundee Document Version Publisher's PDF, also known as Version of record Link to publication in Discovery Research Portal Citation for published version (APA): Chuai, M., Serrano Najera, G., Serra, M., Mahadevan, L., & Weijer, C. J. (2023). Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo. Science Advances, 9(1), Article eabn5429. https://doi.org/10.1126/sciadv.abn5429 General rights Copyright and moral rights for the publications made accessible in Discovery Research Portal 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. 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. Download date: 24. Oct. 2024 Download date: 24. Oct. 2024 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E EVO LU TIO NA RY B I OLO GY Copyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY). RESULTS h Using the chick embryo as a model organism, we explore how the patterning of cell behaviors determines primitive streak morpho- genesis during gastrulation (Fig. 1, A to C and G). To study the tissue-wide coordination of the cell behaviors in the epiblast, we image, as before, gastrulation in a chick line expressing a mem- brane-targeted green fluorescent protein (GFP) using a custom- built light sheet microscope (15) and image analysis (see Materials and Methods and movie S1). INTRODUCTION Previously, we established quantitative light sheet microscopy and computational data analysis methods to character- ize both the embryo-wide tissue flows and critical cell behaviors driving them during streak formation (15). This showed quantita- tively how the tissue flows underlying streak formation are driven by directional cell intercalations and ingression of the mesendo- derm precursor cells. Furthermore, the directional cell intercala- tions were shown to be associated with and dependent on myosin cables in cellular junctions aligned in the direction of intercalation (15). We here confirm and expand these findings and undertake a study to investigate how the modulation of a limited number of crit- ical cell behaviors during gastrulation in the chick embryo alters the spatiotemporal dynamics and morphology of gastrulation. org at University of Dundee on January 06, 2023 y g p The morphogenetic tissue flows in all organisms result from the integration of a few critical cell behaviors, including cell division, differentiation, shape changes, and movement. These cell behaviors are patterned and coordinated through a combination of short- and long-range chemical and mechanical signals (6–8). Different spatio- temporal organizations of these cell behaviors, in conjunction with geometrical and mechanical constraints imposed by the size and shape of the embryo and extraembryonic tissues, are thought to de- termine the diverse morphogenetic programs in vertebrate gastru- lation (9, 10). However, how the large-scale cell movements are coordinated and coupled via feedback to the signaling processes ty of Dundee on January 06, 2023 1Division of Cell and Developmental Biology, School of Life Sciences, University of Dundee, Dundee DD1 5EH, UK. 2Department of Physics, University of California San Diego, La Jolla, CA 92093, USA. 3School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02134, USA. 4Departments of Physics and Or- ganismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA. Corresponding author. Email: gs714@cam.ac.uk (G.S.N.); c.j.weijer@dundee.ac.uk (C.J.W.) †These authors contributed equally to this work. ‡Present address: Department of Genetics, University of Cambridge, Cambridge CB2 3EH, UK. 1Division of Cell and Developmental Biology, School of Life Sciences, University of Dundee, Dundee DD1 5EH, UK. 2Department of Physics, University of California San Diego, La Jolla, CA 92093, USA. 3School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02134, USA. 4Departments of Physics and Or- ganismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138, USA. Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo Manli Chuai1†, Guillermo Serrano Nájera1†‡, Mattia Serra2, Lakshminarayanan Mahadevan3,4, Cornelis J. Weijer1* 1†, Guillermo Serrano Nájera1†‡, Mattia Serra2, Lakshminarayanan Mahadevan3,4, 1 The morphology of gastrulation driving the internalization of the mesoderm and endoderm differs markedly among vertebrate species. It ranges from involution of epithelial sheets of cells through a circular blastopore in amphibians to ingression of mesenchymal cells through a primitive streak in amniotes. By targeting signaling pathways controlling critical cell behaviors in the chick embryo, we generated crescent- and ring-shaped mes- endoderm territories in which cells can or cannot ingress. These alterations subvert the formation of the chick primitive streak into the gastrulation modes seen in amphibians, reptiles, and teleost fish. Our experimental manipulations are supported by a theoretical framework linking cellular behaviors to self-organized multicel- lular flows outlined in detail in the accompanying paper. Together, this suggests that the evolution of gastru- lation movements is largely determined by changes in a few critical cell behaviors in the mesendoderm territory across different species and controlled by a relatively small number of signaling pathways. Corresponding author. Email: gs714@cam.ac.uk (G.S.N.); c.j.weijer@dundee.ac.uk (C.J.W.) Quantitation of cell flows and behaviors (C) Two sections taken at the white lines in (A) showing SNAI (green), FN1 (red), and actin (blue). Note that in the streak, the FN1 layer is fragmented as indicated by the white arrow. (D) Overview images of embryo from light sheet microscope at the end (HH3+) of the experiment overlaid with a deformation mesh. (E) Strain rate tensor for stages shown in (D). Isotropic strain rate component shown as color blue for contraction and red for expansion. The anisotropic part of the strain rate tensor is shown as black and white bars indicating the magnitude and direction of contraction. (F) Cells that will ingress through the streak [domain of attraction (DOA)] during the experiment calculated from the dynamic morphoskeleton (DM). (G) Schematic summary of development (stages HH1, HH2, and HH3), extraembryonic area in blue, mesendoderm precursors in yellow/red, and embryonic area in purple; black bars show the direction of cell intercalation, section taken at the broken white line in HH3 embryo schematic in (E). Scale bars, 500 μm except in (C), which is 250 μm. For (E) and (L), the anisotropic scale bar length represents 0.5/hour for the anisotropic component of the strain rate. (H to N) Same layout as (A) to (F) for FGF2-treated embryo for the same time as the control embryo. FGF (50 μg/ml on the hypoblast; see Materials and Methods)–treated embryos generate a circular primitive streak surrounding the epiblast (eight of eight treated embryos in the light sheet microscope). Downloaded from https://www.science.org at University of Dundee on January 06, 2023 he ie ur ve d o- o- ar in ed E) te A)] in in ic ee ). Downloaded from https://www.science.org at University of Dundee on January 06, 2023 n d ) e ] n n c e Downloaded from https://www.science.org at University of Dundee on January 06, 2023 Fig. 1. Characterization of gastrulation for control embryo and after FGF2 treatment. (A) Control, bright-field image of embryo at HH3+. (B) SNAI expression in embryo shown in (A). (C) Two sections taken at the white lines in (A) showing SNAI (green), FN1 (red), and actin (blue). Note that in the streak, the FN1 layer is fragmented as indicated by the white arrow. (D) Overview images of embryo from light sheet microscope at the end (HH3+) of the experiment overlaid with a deformation mesh. (E) Strain rate tensor for stages shown in (D). Quantitation of cell flows and behaviors The dynamics of tissue deformation (Fig. 1D) can be directly quan- tified by measuring changes in the velocity field associated with tissue shape changes and movement using particle image *Corresponding author. Email: gs714@cam.ac.uk (G.S.N.); c.j.weijer@dundee.ac.uk (C.J.W.) ‡Present address: Department of Genetics, University of Cambridge, Cambridge CB2 3EH, UK. 1 of 10 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E . 1. Characterization of gastrulation for control embryo and after FGF2 treatment. (A) Control, bright-field image of embryo at HH3+. (B) SNAI expression in bryo shown in (A). (C) Two sections taken at the white lines in (A) showing SNAI (green), FN1 (red), and actin (blue). Note that in the streak, the FN1 layer is fragmented ndicated by the white arrow. (D) Overview images of embryo from light sheet microscope at the end (HH3+) of the experiment overlaid with a deformation mesh. (E) ain rate tensor for stages shown in (D). Isotropic strain rate component shown as color blue for contraction and red for expansion. The anisotropic part of the strain rate sor is shown as black and white bars indicating the magnitude and direction of contraction. (F) Cells that will ingress through the streak [domain of attraction (DOA)] ing the experiment calculated from the dynamic morphoskeleton (DM). (G) Schematic summary of development (stages HH1, HH2, and HH3), extraembryonic area in e, mesendoderm precursors in yellow/red, and embryonic area in purple; black bars show the direction of cell intercalation, section taken at the broken white line in 3 embryo schematic in (E). Scale bars, 500 μm except in (C), which is 250 μm. For (E) and (L), the anisotropic scale bar length represents 0.5/hour for the anisotropic mponent of the strain rate. (H to N) Same layout as (A) to (F) for FGF2-treated embryo for the same time as the control embryo. FGF (50 μg/ml on the hypoblast; see terials and Methods)–treated embryos generate a circular primitive streak surrounding the epiblast (eight of eight treated embryos in the light sheet microscope). Downloaded from https://www.science.org at University of Dundee on January 06, 2023 Fig. 1. Characterization of gastrulation for control embryo and after FGF2 treatment. (A) Control, bright-field image of embryo at HH3+. (B) SNAI expression in embryo shown in (A). Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 Blocking ingression results in epiblast folding l y Previous work has established how cells in the sickle-shaped poste- rior territory form the primitive streak. Initially, cell-cell signaling between cells in the extra- and embryonic areas results in the induc- tion of mesendoderm precursors in a sickle-shaped domain in the posterior embryonic epiblast (20). These cells then start a myosin II–mediated directional cell-cell intercalation process that drives the convergence and extension of the sickle-shaped mesendoderm in the extending primitive streak (15, 21, 22). This developmental phase is characterized by large counterrotating vortical flows in the epiblast (fig. S1B), known as polonaise movements (23, 24). During this process, the mesendodermal precursor cells begin to contract apically in preparation for their ingression in the streak (16, 25). Following an epithelial-to-mesenchymal transition (EMT) and ingression, the cells migrate away as cohorts of loosely associated cells (26, 27) to form the various endodermal and meso- dermal structures in the embryo (Fig. 1C) (28–30). Blocking ingression results in epiblast folding Bone morphogenetic protein (BMP) signaling has been shown to occur in the anterior epiblast before the onset of gastrulation (37– 39). Furthermore, blocking BMP signaling can result in the induc- tion of mesoderm and the generation of ectopic primitive streaks (37, 38, 40). In line with these previous observations, we induced the expansion of the mesendoderm territory along the anterior marginal zone by blocking BMP signaling using the pan BMP type I receptor (ALK1, ALK2, ALK3, ALK6) inhibitor LDN- 193189 (41) (LDN; fig. S5, A to D), as seen by the expression of SNAI2+ cells in a circular domain (fig. S5, A to D). Therefore, block- ing BMP signaling is an alternative to FGF activation to generate SNAI2-expressing mesendoderm rings. The SNAI2-positive ectopic domains induced by LDN form rapidly and are more circu- lar than those formed after FGF addition, suggesting that relieving inhibition of mesoderm differentiation by LDN acts faster than the induction of mesendoderm by FGF. Blocking Nodal receptors ALK4, ALK5 by the inhibitor SB50124 does not block mesoderm formation (fig. S6). In LDN-treated embryos, concomitant with the formation of the mesoderm ring, we note the appearance of a circular fold at the boundary of the domain, which contracted over time, engulfing the internal epiblast (fig. S5, A to F). The observed tissue buckling after blocking BMP signaling results from the inhi- bition of EMT and cell ingression, even in the presence of exoge- nous FGF (figs. Quantitation of cell flows and behaviors Isotropic strain rate component shown as color blue for contraction and red for expansion. The anisotropic part of the strain rate tensor is shown as black and white bars indicating the magnitude and direction of contraction. (F) Cells that will ingress through the streak [domain of attraction (DOA)] during the experiment calculated from the dynamic morphoskeleton (DM). (G) Schematic summary of development (stages HH1, HH2, and HH3), extraembryonic area in blue, mesendoderm precursors in yellow/red, and embryonic area in purple; black bars show the direction of cell intercalation, section taken at the broken white line in HH3 embryo schematic in (E). Scale bars, 500 μm except in (C), which is 250 μm. For (E) and (L), the anisotropic scale bar length represents 0.5/hour for the anisotropic component of the strain rate. (H to N) Same layout as (A) to (F) for FGF2-treated embryo for the same time as the control embryo. FGF (50 μg/ml on the hypoblast; see Materials and Methods)–treated embryos generate a circular primitive streak surrounding the epiblast (eight of eight treated embryos in the light sheet microscope). velocimetry (PIV; fig. S1, A and B). We previously showed that the decomposed strain rates calculated from the velocity fields are a robust indicator of individual cell behaviors driving the tissue flows and allow the identification and quantification of cell behav- iors resulting from different perturbations (15, 17). Negative values of the isotropic part of the strain rate, associated with a decrease in tissue area, denote regions where cells are undergoing apical con- traction and cell ingression. Areas characterized by larger values in the anisotropic components of the strain rate, associated with shear deformation, indicate the convergence and extension of the tissue produced by directed cell intercalations (fig. S2 and movie S2). Our current observations on whole embryos confirm our earlier observations on embryo slices (15) that the primitive streak forms in a region undergoing convergent extension produced by apical contraction and cell ingression (negative values of the iso- tropic strain rate tensor) and cell intercalation (contracting aniso- tropic component of the strain rate tensor oriented perpendicular to the streak) (Fig. 1E). Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 2 of 10 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E Quantifying the scale and nature of these movements requires an effective description of the relative motions of cells. While Eulerian quantities (decomposed strain rates) correlate with distinct cell be- haviors, morphogenetic features arise from different processes inte- grated over space and time along cell trajectories. To account for this, we compute the dynamic morphoskeleton (DM) (18), which consists of Lagrangian attractors, their domain of attraction (DOA), and repellers that together organize tissue flows [figure 2 and movie 3 of the accompanying paper (19)]. The DOA consists of the back projection of the attractor on the initial stage of devel- opment and reveals the extent of the sickle-shaped posterior-located mesendoderm precursors fated to ingress through the streak (Fig. 1F). See Materials and Methods for details on DM computation. arrows). In vivo imaging shows that FGF-treated embryos exhibit a ring of tissue (movie S4) that presents the two characteristic signa- tures of the late primitive streak, isotropic contraction along the ring, indicating that cells in the ring are undergoing apical contrac- tion and ingression. Furthermore, shear contraction across the ring shows that cells intercalate in a direction perpendicular to the cir- cular streak (Fig. 1, K to M, and movie S5). These directional inter- calations generate convergent-extending flows along the ring (fig. S4). The resulting circular streak structure resembles the germ ring of teleost fish gastrulation, where cells undergo ingression and migrate toward the animal pole (33–36) [see also figure 3F and movie 9 of (19) for the corresponding DM]. Similarly, in FGF-treated embryos, cells that ingress in the circular streak migrate toward the center of the epiblast, the corresponding animal pole (movie S6). Downloaded from https://www.science.org at University of Dundee on January 06, 2023 These findings suggest that directed cell intercalation and cell in- gression in the mesendoderm territory drive the morphogenetic flows during avian gastrulation. To test this idea, we inhibited me- soderm differentiation by blocking fibroblast growth factor (FGF) signaling using the pan-FGF receptor inhibitor LY2874455 (31). The addition of LY2874455 resulted in the complete inhibition of mesoderm differentiation and the loss of the characteristic tissue flows associated with streak formation (movie S3). Analysis of the strain rate shows that there is little organized intercalation, confirm- ing that the presence of mesendoderm tissue is a prerequisite for streak formation (fig. S3). Blocking ingression results in epiblast folding S5, G to I, and S7). Closer inspection of the development of these LDN-treated embryos showed that they are noticeably smaller due to substantial inhibition of cell proliferation (fig. S8A). This prevented longer-term observation of the embryos in the light sheet microscope to follow the details of the closure of the ingression rings, because, in the absence of tissue growth but in the presence of extensive tissue buckling and involution, the embryo shrinks and detaches from the vitelline membrane, which is fixed in the light sheet imaging chamber (42). Mesoderm expansion produces a circular streak p p While inhibiting FGF signaling blocks the mesoderm differentia- tion, the addition of an excess of FGF results in the circular expan- sion of the SNAI2 (snail family transcriptional repressor 2) expression territory along the marginal zone (Fig. 1, H to J), con- firming previous observations that SNAI2, brachyury, and cChordin expression is activated by FGF (20). During normal development of SNAI2+, mesendoderm cells travel from the sickle-shaped territory where they are specified toward the primitive streak, where they undergo EMT and cell ingression (25, 32). This can be seen via the presence of SNAI2-expressing cells in the inner layers and the breakdown of the Fibronectin 1 (FN1) layer in the streak (Fig. 1, B and C, arrow). Embryos treated with excess FGF show that SNAI2+ cells have extended along a ring-shaped region along the marginal zone (Fig. 1, H to J), while the FN1 basal layer is degraded (Fig. 1J, We observed that the combination of the BMP signaling inhib- itor LDN and the glycogen synthase kinase 3 (GSK3) inhibitor CHIR-99021 (CHIR), conditions that have been used to produce self-renewing mesendoderm precursors in vitro from induced plu- ripotent stem and embryonic stem cells (43), successfully overcame the LDN-induced shrinking and detachment. Treated embryos still showed the formation of mesodermal precursor rings and inhibi- tion of EMT, resulting in buckling and involution of the central tissue (Fig. 2, A to G, and movie S7), but without significant inhi- bition of cell division (fig. S8B). Sectioning revealed that tissue in 3 of 10 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E haracterization of gastrulation in embryos after perturbation of the BMP/Wnt and Vegf signalling pathways. (A) Bright-field image of embryo treated with receptor inhibitor LDN (100 nM) and the GSK3 inhibitor CHIR (3 μM) for 16 hrs from stage HH1 onward (time after which control embryos reach stage HH3+). The generate a large circular invagination that engulfs the central epiblast. (B) SNAI2 expression in the embryo shown in (A). (C) Two sections taken at the white lines wing SNAI (green), FN1 (red), and actin (blue). Note that the FN1 layer is not fragmented at the site of invagination. Mesoderm expansion produces a circular streak (D) Lightsheet microscope overview image of yo at the end of the experiment overlaid with a deformation mesh (typical for four of five embryos in the light sheet microscope). (E) Strain rate tensor for the hown in (D). Isotropic strain rate component shown as color blue for contraction and red for expansion. The anisotropic part of the strain rate tensor is shown as d white bars indicating the magnitude and direction of contraction. (F) Cells that will ingress through the streak [domain of attraction (DOA)] during the exper- culated from the dynamic morphoskeleton (DM). (G) Schematic of the development of a contractile ring which in the absence of EMT results in the invagination ntral epiblast. (H to N) Same layout as (A) to (F) for embryos treated for 16 hrs with the vascular endothelial growth factor (VEGF) receptor inhibitor axitinib (100 Bright field image of axinitib treated embryo. (I) SNAI2 expression in embryo shown in (H). Axinitib treated embryos show limited breakdown of FN1 in the streak (J), an inhibition of the isotropic contraction in the region of the primitive streak (L), a pronounced folding in the anterior streak (K) and a reduced n of mesendoderm cells in the streak (M) (four of five in the light sheet microscope). (N) Schematic of the formation of the primitive streak, which in the of EMT buckles into the embryo. Downloaded from https://www.science.org at University of Dundee on January 06, 2023 h e s f e s - n 0 e d e Downloaded from https://www.science.org at University of Dundee on January 06, 2023 Fig. 2. Characterization of gastrulation in embryos after perturbation of the BMP/Wnt and Vegf signalling pathways. (A) Bright-field image of embryo treated with the BMP receptor inhibitor LDN (100 nM) and the GSK3 inhibitor CHIR (3 μM) for 16 hrs from stage HH1 onward (time after which control embryos reach stage HH3+). The embryos generate a large circular invagination that engulfs the central epiblast. (B) SNAI2 expression in the embryo shown in (A). (C) Two sections taken at the white lines in (A) showing SNAI (green), FN1 (red), and actin (blue). Note that the FN1 layer is not fragmented at the site of invagination. (D) Lightsheet microscope overview image of an embryo at the end of the experiment overlaid with a deformation mesh (typical for four of five embryos in the light sheet microscope). Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E generated in FGF2-treated embryos (Fig. 3, FGF2, and fig. S10, FGF2), further supporting the idea that this ring functions as a cir- cular primitive streak (Fig. 1, H to N). Xenopus, where tissue is invaginated as a continuous epithelial sheet (44) [see also figure 3H and movie 12 of (19) for the corresponding DM]. However, while Xenopus embryos invaginate their vegetal side during gastrulation, in LDN-treated chick embryos, the system resolves to engulf the central epiblast, corresponding to the animal side. This is most likely caused by the attachment of the extraembryonic edge of the embryo to the vitelline membrane functioning as an important mechanical constraint that prevents the invagination of the vegetal side (see also HNF3B expression in the next section). p g Instead, the ectopic ring of SNAI2 expression generated in LDN- treated embryos is not associated with large pores in the FN1 network (Fig. 3, LDN + CHIR) or the TNC fiber assembly (fig. S10, LDN), suggesting that cell ingression associated with EMT is blocked in LDN-treated embryos. Combined LDN and FGF treat- ment, which results in the formation of a large invagination, indi- cating that LDN dominates at the morphogenetical level (fig. S5, G to K), shows FN1 breakdown (fig. S5I, arrow) and assembly of TNC fibers (fig. S10, FGF2 + LDN), indicating that there could be a partial EMT and possibly some cell ingression. Last, axitinib- treated embryos show a less defined FN1 meshwork along the prim- itive streak (Fig. 3, axitinib), associated with a reduction of cell ingression. Last, we explored the effects of blocking growth factors that have been implicated in EMT and the migration of mesoderm cells of the platelet-derived growth factor and vascular endothelial growth factor (VEGF) families (45, 46). Cells in the posterior portion of the primitive streak express VEGFR2 and manifest a chemotactic response to several VEGF isoforms after ingression. We observed that the use of the VEGF signaling inhibitor axitinib (Fig. 2, H to N, and movie S9) (47) or competitive VEGFR2-Fc fragments (fig. S9) results in the formation of a short fat streak when added at early stages (2 to 3 hours of incubation). Light sheet microscopy of devel- oping embryos showed that cell intercalation was not significantly inhibited (Fig. 2L), and the vortical flow of tissue into the ventral midline of the embryo continued (movie S10). Aligned myosin cables direct intercalations g y Our experiments show that the shape of the mesendoderm, together with cell intercalation and ingression behaviors, determines the morphogenetic outcome of gastrulation. While cell intercalations drive the main in-plane tissue flows, the presence or absence of in- dividual cell ingressions determines whether the tissue will remain flat or will buckle. During normal development, the direction of in- tercalations of mesendoderm cells is strongly correlated with the di- rection of cables of phosphorylated myosin II that span several cells (15). At early gastrulation stages, these myosin cables extend from posterior to anterior, creating a semicircle between the embryonic and extraembryonic territories, and they correlate with the direction of intercalation, as shown by the intercalation (P) tensor (58) com- puted from cell tracking data (Fig. 4A), as described in detail in our previous work (15). We show that this strong correlation between mesendoderm expression domain and myosin planar cell polarity persists in the cases of circular mesendoderm domains, as verified with the FGF2 and LDN treatments (Fig. 4, B and C). As in normal early development, the myosin planar cell polarity is initially Mesoderm expansion produces a circular streak (E) Strain rate tensor for the embryo shown in (D). Isotropic strain rate component shown as color blue for contraction and red for expansion. The anisotropic part of the strain rate tensor is shown as black and white bars indicating the magnitude and direction of contraction. (F) Cells that will ingress through the streak [domain of attraction (DOA)] during the exper- iment calculated from the dynamic morphoskeleton (DM). (G) Schematic of the development of a contractile ring which in the absence of EMT results in the invagination of the central epiblast. (H to N) Same layout as (A) to (F) for embryos treated for 16 hrs with the vascular endothelial growth factor (VEGF) receptor inhibitor axitinib (100 nM). (H) Bright field image of axinitib treated embryo. (I) SNAI2 expression in embryo shown in (H). Axinitib treated embryos show limited breakdown of FN1 in the posterior streak (J), an inhibition of the isotropic contraction in the region of the primitive streak (L), a pronounced folding in the anterior streak (K) and a reduced ingression of mesendoderm cells in the streak (M) (four of five in the light sheet microscope). (N) Schematic of the formation of the primitive streak, which in the absence of EMT buckles into the embryo. S8). The tangential intercalation drives the contraction of the ring, and because these cells do not undergo EMT (Fig. 2C) sup- ported by the absence of apical contraction (movie S8), this contrac- tion does not result in ingression but instead in buckling. The measured negative isotropic strain rate at these stages of buckling indicates the disappearance of cells into the inside of the embryo rather than ingression. This situation reproduces some aspects of the tissue organization and flow during the closing blastopore in this internal domain had folded inside the embryo, and the mesen- doderm cells that internalized through the circular fold failed to undergo EMT, as shown by the cells staying connected in a contin- uous epithelial sheet and the presence of a continuous nondegraded layer of FN1 underlying the involuted epithelial cells (Fig. 2G). y y g p g Analysis of the anisotropic strain rate (Fig. 2E) in these embryos showed a pronounced contraction oriented tangentially to the ectopic mesendoderm ring produced by cell intercalation (movie Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 4 of 10 ownloaded from https://www.science.org at University of Dundee on January 06, 2023 , ) g ( g , ) The lack of cell ingression at the primitive streak in combination with an unaltered rate of cell division (fig. S9, C and D) results in more cells flowing into the midline than flowing out of it. When the epiblast cannot accommodate more tissue, the anterior portion of the streak can buckle inside (movie S11). Furthermore, the rate of cell proliferation is not affected by blocking VEGF signaling (fig. S9, C and D), suggesting that the buckling is not the result of reduced tissue fluidity produced by a reduced number of dividing cells (48). The resulting fold structure might resemble the blastoporal canal observed in chameleon gastrulation (49) [see also figure 3D and movie 7 of (19) for the corresponding DM]. Chameleon gastrula- tion is bimodal, they internalize part of the anterior mesoderm by involution through the blastoporal canal, and the posterior meso- derm is internalized by a limited amount by individual cell ingres- sions through the blastoporal plate (50). Intriguingly, the reduction of the individual cell ingression in the primitive streak of the chick embryo can originate a similar anterior invagination. SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E However, there was a distinct absence of ingression of cells in the streak evidenced by the absence of tissue contraction in the midline (Fig. 2L and fig. S9, A and B) and reduction of FN1 degradation (Fig. 3, axitinib). g In addition, we studied whether the ectopic rings of SNAI2+ cells generated with different treatments have a more endodermal or me- sodermal character. The tip of the primitive streak is a source of en- doderm precursors, and it is characterized by the expression of the transcription factor HNF3B, a marker of the definitive endoderm (55–57). We noted that cells in the tip of the streak showed reduced levels of SNAI2 expression concomitant with elevated HNF3B expression (figs. S11 and S12, control). It is noteworthy that scattered HNF3B-expressing cells are also found in the epiblast anterior to the streak (fig. S12, control). Among the ectopic ring of SNAI2-expressing cells in FGF-treated embryos, some scattered HNF3B+ cells can also be observed; however, their expression levels and density are not the same as in the tip of the streak rem- nants, suggesting that FGF treatment especially induces mesoderm precursors (fig. S12, FGF). In LDN-treated embryos, a large popu- lation of the SNAI2-expressing cells in the ring coexpresses HNF3B (fig. S12, LDN), which might indicate that these cells are shifted toward an endoderm precursor fate. Furthermore, LDN-treated embryos show elevated HNF3B expression across the central epi- blast (fig. S12, LDN), which will be engulfed by the contracting ring. Intriguingly, this makes this process more similar to Xenopus gastrulation, where the endoderm is also engulfed by the contracting blastopore. nloaded from https://www.science.org at University of Dundee on January 06, 2023 The data show that the presence or absence of ingression is reflected in the FN1 network structure. Scale bars, 500 μm (two top rows) and 100 μm (two bottom rows). Mode of ingression and cell type identities after perturbation SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E Fig. 4. Comparison of the patterns of mesendoderm differentiation, myosin alignment, and the direction of cell intercalation. Top row: Control embryo after 16 hours of incubation in EC culture as described in Materials and Methods. Second row: Embryo treated with FGF2 for 16 hours. Third row: Embryo treated with LDN (100 nM) for 16 hours (100 nM LDN + 3 μM CHIR for the P tensor). First column: SNAI2+ expression (red) and pMLC (green). The regions in the top and bottom white squares in (A) to (C) are shown at higher mag- nification on the right of these images as top and bottom images, respectively. Second main column: The quantitation of myosin cable alignment. Scale bars rep- resent ∼30% anisotropy. Third main column: The direction of intercalation calcu- lated as the anisotropic component of the intercalation (P) tensor (58). The scale bar represents a magnitude of ∼0.4 1/h. Regions of intercalating mesendoderm cells are highlighted in red. Scale bars, 500 μm (A to C) and 100 μm (small regions). Fig. 4. Comparison of the patterns of mesendoderm differentiation, myosin alignment, and the direction of cell intercalation. Top row: Control embryo after 16 hours of incubation in EC culture as described in Materials and Methods. Second row: Embryo treated with FGF2 for 16 hours. Third row: Embryo treated with LDN (100 nM) for 16 hours (100 nM LDN + 3 μM CHIR for the P tensor). First column: SNAI2+ expression (red) and pMLC (green). The regions in the top and bottom white squares in (A) to (C) are shown at higher mag- nification on the right of these images as top and bottom images, respectively. Second main column: The quantitation of myosin cable alignment. Scale bars rep- resent ∼30% anisotropy. Third main column: The direction of intercalation calcu- lated as the anisotropic component of the intercalation (P) tensor (58). The scale bar represents a magnitude of ∼0.4 1/h. Regions of intercalating mesendoderm cells are highlighted in red. Scale bars, 500 μm (A to C) and 100 μm (small regions). Fig. 4. Comparison of the patterns of mesendoderm differentiation, myosin alignment, and the direction of cell intercalation. Top row: Control embryo after 16 hours of incubation in EC culture as described in Materials and Methods. nloaded from https://www.science.org at University of Dundee on January 06, 2023 Fig. 3. Comparison of SNAI and FN1 expression after various perturbations. Top row: Bright-field images of embryos after 16 hours in EC culture after various treatments. Second row: Confocal images of the embryos in the top row showing the expression of SNAI (green) and FN1 (red). Third and fourth rows: Higher-mag- nification images taken in front of the streak (top white square) and at the tip of the streak (bottom white square). It can be seen in the control (first row) that the region anterior to the streak expresses high levels of FN1, forming a densely packed meshwork of fibrils. In the streak, FN1 shows a larger pore meshwork (arrows) probably associated with cell ingressions. Note that in the ring-shaped streak after FGF treatment (second column), FN1 shows the same large porous structure as in the streak, indicating that this is also a region where the cells ingress. After LDN and axitinib treatment, the anterior region shows a dense fibrillar network, while in the streak or ring cases, the porous network is less well developed as in the control or after FGF treatment. The data show that the presence or absence of ingression is reflected in the FN1 network structure. Scale bars, 500 μm (two top rows) and 100 μm (two bottom rows). Fig. 3. Comparison of SNAI and FN1 expression after various perturbations. Top row: Bright-field images of embryos after 16 hours in EC culture after various treatments. Second row: Confocal images of the embryos in the top row showing the expression of SNAI (green) and FN1 (red). Third and fourth rows: Higher-mag- nification images taken in front of the streak (top white square) and at the tip of the streak (bottom white square). It can be seen in the control (first row) that the region anterior to the streak expresses high levels of FN1, forming a densely packed meshwork of fibrils. In the streak, FN1 shows a larger pore meshwork (arrows) probably associated with cell ingressions. Note that in the ring-shaped streak after FGF treatment (second column), FN1 shows the same large porous structure as in the streak, indicating that this is also a region where the cells ingress. After LDN and axitinib treatment, the anterior region shows a dense fibrillar network, while in the streak or ring cases, the porous network is less well developed as in the control or after FGF treatment. Mode of ingression and cell type identities after perturbation The breakdown of the basement membrane is necessary for EMT to occur and cells to ingress through the primitive streak (5, 25, 51). To further characterize the mode of ingression associated with each chemical perturbation, we analyzed the breakdown of the basement membrane by inspecting the meshwork of FN1 (Fig. 3) and tenascin C (TNC; fig. S10) (25, 52, 53). During normal development, the an- terior epiblast shows a very dense FN1 layer (Fig. 3, control). The region of SNAI2 expression in the primitive streak, where cells undergo EMT and ingress, is associated with the formation of large pores in the FN1 network (Fig. 3, control), as has been ob- served in mice (54). Furthermore, this region is characterized by the assembly of oriented fibers of TNC (fig. S10, control). This sit- uation is fully reproduced in the ectopic ring of SNAI2 expression 5 of 10 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E Fig. 3. Comparison of SNAI and FN1 expression after various perturbations. Top row: Bright-field images of embryos after 16 hours in EC culture after various treatments. Second row: Confocal images of the embryos in the top row showing the expression of SNAI (green) and FN1 (red). Third and fourth rows: Higher-mag- nification images taken in front of the streak (top white square) and at the tip of the streak (bottom white square). It can be seen in the control (first row) that the region anterior to the streak expresses high levels of FN1, forming a densely packed meshwork of fibrils. In the streak, FN1 shows a larger pore meshwork (arrows) probably associated with cell ingressions. Note that in the ring-shaped streak after FGF treatment (second column), FN1 shows the same large porous structure as in the streak, indicating that this is also a region where the cells ingress. After LDN and axitinib treatment, the anterior region shows a dense fibrillar network, while in the streak or ring cases, the porous network is less well developed as in the control or after FGF treatment. The data show that the presence or absence of ingression is reflected in the FN1 network structure. Scale bars, 500 μm (two top rows) and 100 μm (two bottom rows). A mechanochemical model describes tissue flows: Predictions and implications To quantify these observations, we propose and explore a theoretical model for gastrulation as a mechanosensitive self-organizing process in the accompanying paper (19). Our continuum model couples the myosin-driven tension-sensitive active stress to tissue flow and cable ordering and leads to an interpretation of gastrula- tion as the result of an instability in the early embryo (fig. S13) (19). Our model can recapitulate these different gastrulation modes by varying the extension of the mesendoderm (an initial condition of our model) and a nondimensional parameter modulating the amount of active cell ingression [see figures 2 to 4 and movies 2 to 12 of (19)]. Notably, providing only the initial experimental dis- tribution of actomyosin cables’ orientation and active myosin inten- sity, our model predicts the observed orientation of the cable, the myosin intensity, cell velocities, and velocity divergence at later times without data fitting. The model predicts a strong correlation between the mesendoderm expression domain and myosin planar cell polarity, which persists in the cases of circular mesendoderm domains, as verified with the FGF2 and LDN treatments (fig. S14). By varying the initial mesendoderm domain shape in combi- nation with the degree of myosin-dependent ingression (p0, param- eter), this model can also generate the cell flows and DOAs after the various perturbations described here and thus provide a mechanis- tic understanding of the biophysical mechanisms and role of tissue ity of Dundee on January 06, 2023 oriented along the long axis of the mesendoderm expression domain, in the same direction of cell intercalations, which explains the contraction of the ring in the LDN + CHIR case. However, during normal streak extension at late stages (HH3 to HH4), cell intercalations orient perpendicular to the long axis of the mesendo- derm (the primitive streak) driven by cell ingression in the streak. The same occurs in embryos treated with FGF, where many cells ingress through the circular streak (negative values in the isotropic part of the strain rate), which results in the reorientation of the di- rection of intercalation of nearby cells perpendicular to the ring (di- rection of the shear). Presumably, in the LDN case, the cell intercalations continue to occur along the long axis of the mesen- doderm (tangential to the ring) because cell ingression and EMT are never activated. Mode of ingression and cell type identities after perturbation Second row: Embryo treated with FGF2 for 16 hours. Third row: Embryo treated with LDN (100 nM) for 16 hours (100 nM LDN + 3 μM CHIR for the P tensor). First column: SNAI2+ expression (red) and pMLC (green). The regions in the top and bottom white squares in (A) to (C) are shown at higher mag- nification on the right of these images as top and bottom images, respectively. Second main column: The quantitation of myosin cable alignment. Scale bars rep- resent ∼30% anisotropy. Third main column: The direction of intercalation calcu- lated as the anisotropic component of the intercalation (P) tensor (58). The scale bar represents a magnitude of ∼0.4 1/h. Regions of intercalating mesendoderm cells are highlighted in red. Scale bars, 500 μm (A to C) and 100 μm (small regions). A mechanochemical model describes tissue flows: Predictions and implications oriented along the long axis of the mesendoderm expression domain, in the same direction of cell intercalations, which explains the contraction of the ring in the LDN + CHIR case. However, during normal streak extension at late stages (HH3 to HH4), cell intercalations orient perpendicular to the long axis of the mesendo- derm (the primitive streak) driven by cell ingression in the streak. The same occurs in embryos treated with FGF, where many cells ingress through the circular streak (negative values in the isotropic part of the strain rate), which results in the reorientation of the di- rection of intercalation of nearby cells perpendicular to the ring (di- rection of the shear). Presumably, in the LDN case, the cell intercalations continue to occur along the long axis of the mesen- doderm (tangential to the ring) because cell ingression and EMT are never activated. Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 6 of 10 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E SC I EN C E A D VA N CE S ectopic structures and how different internalization modes might affect the cell fate. stresses and strains involved. For further details on the model and results, see (19). The experiments have also shown that directed intercalations along the long axis of the mesendoderm territory are a major part of the mechanism driving the formation of the internalization structure, which is a linear, a circular streak, or a partial or circular blastopore. These cell intercalations are driven by long myosin cables oriented along the long axis of the evolving mesendoderm territory (15). We have developed a mechanochemical model that, starting from an initial sickle- or ring-shaped mesendoderm precur- sor region containing a tension-dependent myosin recruitment mechanism, can explain the emergence of the tissue flows that drive the formation of the primitive streak and predict the dynamics of the domains of ingressing cells, which are in excellent agreement with those experimentally observed, both under normal and exper- imentally perturbed conditions (19). The model has highlighted and elucidated the importance of the mechanochemical feedback of the flow on myosin recruitment and organization of the epiblast tissue flow. DISCUSSION Here, we perturbed the mechanisms that control directed cell inter- calations and cell ingressions within the mesendoderm, which gen- erate the forces that drive the tissue flows that generate the primitive streak during chick gastrulation. By targeting FGF, WNT, BMP, and VEGF developmental signaling pathways, known critical regulators of early development, we modulated the size and shape of the mes- endoderm territory and uncoupled cell intercalation and ingression. Our results suggest that FGF may predominantly induce meso- derm precursors, while inhibition of BMP signaling could result in the formation of excess endoderm precursors. These findings could suggest that the differentiation status of the cells affects how these cells internalize. However, it remains unclear whether signaling pathways that control differentiation can modulate cell behaviors, such as cell ingression, independently. Another open question is what is the fate of the cells that are internalized through these Downloaded from https://www.science.org at University of Dundee on January 06, 2023 The purpose of gastrulation is to internalize the mesoderm and endoderm precursors. Unexpectedly, despite being an essential phase of early development, the exact morphology of gastrulation shows considerable variation across the evolutionary tree. Meso- derm and endoderm precursors are typically formed on the surface of the developing embryo from epithelial tissues. Besides Fig. 5. Changes in the extent of the mesendoderm territory and the degree of EMT could underlie the evolution of vertebrate gastrulation. (A) Evolutionary relationships of representative vertebrate morphologies recapitulated in the chick embryo via modification of key cell behaviors. Images of gastrulation showing the blastopoal canal in chameleon (28 days), the blastopore in Xenopus (12 hours), and the germ ring in zebrafish (5.7 hours) are adapted from (33, 44, 49). Scale bars, 250 μm. (B) SNAI2+ cell distributions in HH3+ chick embryos after different treatments show that manipulating the extent of the mesendoderm territory and the level of EMT controlling ingression in the chick embryo reproduce critical morphological structures of gastrulation in other vertebrates. Scale bars, 500 μm (chick embryos). w.science.org at University of Dundee on January 06, 2023 Fig. 5. Changes in the extent of the mesendoderm territory and the degree of EMT could underlie the evolution of vertebrate gastrulation. (A) Evolutionary relationships of representative vertebrate morphologies recapitulated in the chick embryo via modification of key cell behaviors. A mechanochemical model describes tissue flows: Predictions and implications However, the mechanisms that set the initial direction- ality of the myosin cable alignment during normal development and after the experimental perturbations described here resulting in the observed different mesendoderm territories morphologies, sickle or ring, remain to be elucidated. The manipulations described here are sufficient to induce large morphogenetic changes that recapitulate the hallmark dynamics and morphologies of different vertebrate gastrulation modes in a single organism, the chick embryo (Fig. 5). These results suggest that the modulation of the extension of the mesendoderm territory and the presence or absence of EMT controlling cell ingression largely determined the different morphologies observed during the evolution of vertebrate gastrulation. Image processing Surface focusing First, we transform the image volumes from the original 45° geom- etry to a 90° geometry. Once the image volumes have been trans- formed, we find the surface using a method based on the square gradient focusing algorithm (62, 63). Briefly, we tile the volume in 25 × 25 pixel columns and use the algorithm to find the position of the surface in each column, which constitutes the height map. Last, we apply a smoothing filter to the height map and use this to section the image volume to produce a two-dimensional image of the surface of the embryo (fig. S1A).i p Our experiments suggest that adjusting the extension of the mes- endoderm territory, from a ring to a crescent, and regulating the capacity of these cells to undergo cell ingression are enough to invoke large morphogenetic changes reminiscent of other gastrula- tion modes. Similarly, it is possible to induce an invaginating gas- trulation mode in a fly that normally gastrulates by stochastic individual cell ingression (Chironomus riparius) by expressing two genes (folded gastrulation and t48) that are known to control blastoderm invagination during gastrulation in another fly species (Drosophila melanogaster) (59). In yet another example, it is possi- ble to switch the gastrulation mode of the sea anemone Nematostel- la vectensis from invagination to individual cell ingression by dissociating and reaggregating the cells in the embryo (60). Togeth- er, these experiments suggest that different gastrulation modes could easily evolve because relatively small quantitative changes in signaling, gene expression, or the developmental context can drive major morphogenetic transitions through their effects on the mod- ulation of individual cell behaviors. yi The velocity fields at the surface of the embryo are computed by digital PIV (fig. S1B) using PIVLab vs1.32 for MATLAB with two passes of 64 × 64 pixels and 32 × 32 pixels with a 50% overlap as in (15). The gradient of the velocity can be used to compute the strain rate at the surface of the embryo. They can be decomposed into two parts: the isotropic term, which describes the local rate of change in the area, and the anisotropic term, which describes the local rate of shear deformation. Together, they can be used as an indirect but robust measure of cell behaviors. DISCUSSION Images of gastrulation showing the blastopoal canal in chameleon (28 days), the blastopore in Xenopus (12 hours), and the germ ring in zebrafish (5.7 hours) are adapted from (33, 44, 49). Scale bars, 250 μm. (B) SNAI2+ cell distributions in HH3+ chick embryos after different treatments show that manipulating the extent of the mesendoderm territory and the level of EMT controlling ingression in the chick embryo reproduce critical morphological structures of gastrulation in other vertebrates. Scale bars, 500 μm (chick embryos). Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 7 of 10 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E SC I EN C E A D VA N CE S covering roughly an area of 3.2 mm by 4.7 mm every 3 min. Embryos are cultured during imaging as described in (42). Under these conditions, chick embryos can develop for more than 24 hours at 37°C (480 time points, >2 terabytes). differences in the shape of the early embryos, spheres (frogs and fish) or sheets (amniotes), and size, there is a great diversity in the mode and geometry of the structures through which the cells internalize (1, 2). In spherical vertebrate embryos, the cells typically internalize through spherical structures such as the germ ring in fish and blastopores in amphibians, while in amniotes, cells internalize a structure known as the primitive streak. Furthermore, cells internal- ize either through ingression of individual cells, such as in amniotes, or through involution of cells in epithelial sheets, such as in Xenopus (44), or in a mixed mode as observed in reptiles (49). differences in the shape of the early embryos, spheres (frogs and fish) or sheets (amniotes), and size, there is a great diversity in the mode and geometry of the structures through which the cells internalize (1, 2). In spherical vertebrate embryos, the cells typically internalize through spherical structures such as the germ ring in fish and blastopores in amphibians, while in amniotes, cells internalize a structure known as the primitive streak. Furthermore, cells internal- ize either through ingression of individual cells, such as in amniotes, or through involution of cells in epithelial sheets, such as in Xenopus (44), or in a mixed mode as observed in reptiles (49). Chick lines and embryo culture Fertile eggs from white Leghorn chicks were obtained from Henry Stewart & Co., Lincolnshire, UK. Membrane GFP embryos were ob- tained from the National Avian Research facility at Edinburgh Uni- versity. Embryos were cultured using the Embryo Culture method according to published procedures (61). p y Myosin quantification To quantify the myosin concentration and alignment, the position of the apical epiblast surface needs to be identified. To achieve this more reliably, we used the stronger actin signal to find the precise location of the apical surface using the above-described method (fig. S1, E and F). This surface projection was analyzed for myosin anisotropy using a Fourier transform–based algorithm (fig. S1G) (64) to find cellular anisotropy and size using a subtiling of the surface in 256 × 256 pixels and a 50% overlap between tiles. on January 06, 2023 Chemical manipulations For in vivo imaging, the embryos were later transferred to imaging chambers as described in (65), and the chemical inhibitors were dis- solved in the 10-ml albumen that covers the embryos during imaging. For immunocytochemistry experiments, chemical inhibi- tors were dissolved in the EC culture substrate. Recombinant human FGF2 and VEGFR2-Fc were added as a 1-μl drop (50 μg/ ml) and carefully deposited on the hypoblast side of embryos in EC culture. For life imaging, the embryos were later transferred to Chemicals and other reagents g LDN, CHIR, axitinib, and SB50124 were acquired from Sigma- Aldrich. Hydromount was obtained from National Diagnostics. Ul- trapure grade was acquired from Merck/VWR. Recombinant human FGF2 (233-FB) and recombinant human VEGFR2/KDR- Fc (357-KD-050) were obtained from (R&D Systems). Primary an- tibodies phospho-myosin light chain 2 (S19) mouse monoclonal antibody (mAb) (3675), phospho-myosin light chain 2 (T18, S19) rabbit mAb (3674), and SNAI2 (C19G7) rabbit mAb (9585) were from Cell Signaling Technology. The TNC mouse mAb M1-B4 and the Hnf3β/FoxA2 mAb 4C7 were from the Developmental Studies Hybridoma Bank. Alexa fluorophore-conjugated secondary antibodies were obtained from Invitrogen. We developed a robust algorithm to detect and track individual cell divisions independently of image segmentation (63). During cyto- kinesis, epithelial cells round up on the surface, acquiring a very typical circular shape. We used the circular Hough transform im- plementation in MATLAB to find the characteristic circular shapes associated with dividing cells. The algorithm uses PIV to follow the dividing cells to avoid counting the same division event repeatedly between consecutive frames.i p Division detection Chemicals and other reagents Image processing Surface focusing In a confluent epithelial sheet, as the epiblast, the isotropic term is the product of the balance of changes in the cell area, cell divisions, and cell ingressions. In con- trast, the anisotropic part is the result of directed cell intercalation or coordinated asymmetric changes in cell shape (fig. S1C). The im- plementation details can be found in (15). Imaging To image the complete chick embryo during gastrulation, we use a chick line expressing GFP in the cell membranes (myr-GFP) and a dedicated light sheet microscope as described in detail in (15). To cover the entire embryo (~4 mm in diameter), we acquire images at intervals of 1.84 μm at 45° with respect to the surface in two over- lapping sequential scans (~5000 images of 2560 × 400 pixels), Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 8 of 10 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E Other Supplementary Material for this manuscript includes the following: Movies S1 to S11 imaging chambers maintaining the added protein on the hypoblast side of the embryo. All experiments were repeated two to three times, and each experiment typically contained six to eight embryos per treatment and six to eight internal control embryos to score variations in development success and timing between dif- ferent batches of fertilized eggs obtained at different periods during the year. View/request a protocol for this paper from Bio-protocol. View/request a protocol for this paper from Bio-protocol. Dynamic morphoskeletons 13. Y. Maroudas-Sacks, K. Keren, Mechanical patterning in animal morphogenesis. Annu. Rev. Cell Dev. Biol. 37, 469–493 (2021). ww.sc y Given a planar velocity field v(x, t), we compute the DM from the backward and forward finite-time Lyapunov exponents (FTLEs) (18). We compute the FTLE as 14. C. Collinet, T. Lecuit, Programmed and self-organized flow of information during mor- phogenesis. Nat. Rev. Mol. Cell Biol. 22, 245–265 (2021). 15. E. Rozbicki, M. Chuai, A. I. Karjalainen, F. Song, H. M. Sang, R. Martin, H. J. Knolker, M. P. MacDonald, C. J. Weijer, Myosin-II-mediated cell shape changes and cell intercalation contribute to primitive streak formation. Nat. Cell Biol. 17, 397–408 (2015). org at Un L tf t0ðx0Þ ¼ ln½l2ðx0Þ j tf t0 j ; xf :¼ F tf t0ðx0Þ ¼ x0 þ ðtf t0 v½Ft t0ðx0Þ; t ð1Þ 16. O. Voiculescu, L. Bodenstein, I. J. Lau, C. D. Stern, Local cell interactions and self-am individual cell ingression drive amniote gastrulation. eLife 3, e01817 (2014). where λ2(x0) denotes the highest singular value of the Jacobian of the flow map F tf t0ðx0Þ and the flow map describing the trajectories from their initial x0 to final xf positions. To compute the FTLE, we first calculate Ft t0ðx0Þ by integrating the cell velocity field v(x, t) using the MATLAB built-in Runge-Kutta solver ODE45 with an absolute and relative tolerance of 10−6, linear interpolation in space and time, and a uniform dense grid of the initial conditions. Then, denoting the ith component of the flow map Ft t0ðx0Þ by xiðx1 0; x2 0; t0; tÞ, we compute the deformation gradient rFt t0ðx0Þ using the finite-difference approximation where λ2(x0) denotes the highest singular value of the Jacobian of the flow map F tf t0ðx0Þ and the flow map describing the trajectories from their initial x0 to final xf positions. To compute the FTLE, we first calculate Ft t0ðx0Þ by integrating the cell velocity field v(x, t) using the MATLAB built-in Runge-Kutta solver ODE45 with an absolute and relative tolerance of 10−6, linear interpolation in space and time, and a uniform dense grid of the initial conditions. Then, denoting the ith component of the flow map Ft t0ðx0Þ by xiðx1 0; x2 0; t0; tÞ, we compute the deformation gradient rFt t0ðx0Þ using the finite-difference approximation 17. G. Serrano Nájera, C. J. Weijer, Cellular processes driving gastrulation in the avian embryo. Mech. Dev. 163, 103624 (2020). Immunocytochemistry 2. D. Arendt, K. Nübler-Jung, Rearranging gastrulation in the name of yolk: Evolution of gastrulation in yolk-rich amniote eggs. Mech. Dev. 81, 3–22 (1999). 2. D. Arendt, K. Nübler-Jung, Rearranging gastrulation in the name of yolk: Evolution of gastrulation in yolk-rich amniote eggs. Mech. Dev. 81, 3–22 (1999). After the chemical manipulations, the embryos were fixed in 4% paraformaldehyde/phosphate-buffered saline (PBS) containing 0.1% Tween 20 on ice for 3 hours and later washed three times in PBS. Later, antibody detection was performed as described (15) fol- lowed by staining with 4′,6-diamidino-2-phenylindole and Alexa 405 Phalloidin. Embryos were mounted in Hydromount and imaged using a wide-field microscope or a Leica TCS SP8 confocal microscope at ×10 to ×20 magnifications. Selected embryos were processed for sectioning; first, embryos were embedded in 7.5% gelatin/15% sucrose, followed by freezing in a dry ice/isopentane mixture, and sectioned on a Leica cryostat. Sections were mounted on lysine-coated microscope slides and mounted with Hydromount. 3. L. Solnica-Krezel, Gastrulation: From Embryonic Pattern to Form (Academic Press, 2020). 3. L. Solnica-Krezel, Gastrulation: From Embryonic Pattern to Form (Academic Press, 2020). 4. R. Keller, Shaping the vertebrate body plan by polarized embryonic cell movements. Science 298, 1950–1954 (2002). 4. R. Keller, Shaping the vertebrate body plan by polarized embryonic cell movements. Science 298, 1950–1954 (2002). 5. M. Williams, C. Burdsal, A. Periasamy, M. Lewandoski, A. Sutherland, Mouse primitive streak forms in situ by initiation of epithelial to mesenchymal transition without migration of a cell population. Dev. Dyn. 241, 270–283 (2012). 5. M. Williams, C. Burdsal, A. Periasamy, M. Lewandoski, A. Sutherland, Mouse primitive streak forms in situ by initiation of epithelial to mesenchymal transition without migration of a cell population. Dev. Dyn. 241, 270–283 (2012). 6. D. Kimelman, Mesoderm induction: From caps to chips. Nat. Rev. Genet. 7, 360–372 (2006). 7. T. Merle, E. Farge, Trans-scale mechanotransductive cascade of biochemical and biome- chanical patterning in embryonic development: The light side of the force. Curr. Opin. Cell Biol. 55, 111–118 (2018). 8. P.-F. Lenne, E. Munro, I. Heemskerk, A. Warmflash, L. Bocanegra-Moreno, K. Kishi, A. Kicheva, Y. Long, A. Fruleux, A. Boudaoud, Roadmap for the multiscale coupling of biochemical and mechanical signals during development. Phys. Biol. 18, 041501 (2021). 9. G. Sheng, A. M. Arias, A. Sutherland, The primitive streak and cellular principles of building an amniote body through gastrulation. Science 374, eabg1727 (2021). 10. M. Valet, E. D. Siggia, A. REFERENCES AND NOTES 1. C. D. Stern, Gastrulation: From Cells to Embryo (CSHL Press, 2004). 1. C. D. Stern, Gastrulation: From Cells to Embryo (CSHL Press, 2004). Confocal microscopy y To image the embryos, confocal tile scans were taken at ×10 and ×20 magnifications using sequential line scanning on a Leica SP8 con- focal microscope. To cover the thickness of the embryo, stacks of 20 to 40 slices were taken at 2- to 4-μm intervals along the z axis. 11. T. Brunet, A. Bouclet, P. Ahmadi, D. Mitrossilis, B. Driquez, A.-C. Brunet, L. Henry, F. Serman, G. Béalle, C. Ménager, F. Dumas-Bouchiat, D. Givord, C. Yanicostas, D. Le-Roy, N. M. Dempsey, A. Plessis, E. Farge, Evolutionary conservation of early mesoderm specifi- cation by mechanotransduction in Bilateria. Nat. Commun. , 1–15 (2013). ed from http 12. E. Hannezo, C.-P. Heisenberg, Mechanochemical feedback loops in development and disease. Cell 178, 12–25 (2019). Immunocytochemistry H. Brivanlou, Mechanical regulation of early vertebrate embryo- genesis. Nat. Rev. Mol. Cell Biol. 23, 169–184 (2022). Dynamic morphoskeletons ty of D 18. M. Serra, S. Streichan, M. Chuai, C. J. Weijer, Dynamic morphoskeletons in development. Proc. Natl. Acad. Sci. U.S.A. 117, 11444–11449 (2020). 19. M. Serra, G. Serrano Nájera, M. Chuai, V. Spandan, C. J. Weijer, L. Mahadevan, A mecha- nochemical model recapitulates distinct vertebrate gastrulation modes. bioRxiv 2021.10.03.462928 (2021).https://doi.org/10.1101/2021.10.03.462928. 20. C. Alev, Y. Wu, Y. Nakaya, G. Sheng, Decoupling of amniote gastrulation and streak for- mation reveals a morphogenetic unity in vertebrate mesoderm induction. Development 140, 2691–2696 (2013). uary 06, 2 21. O. Voiculescu, F. Bertocchini, L. Wolpert, R. E. Keller, C. D. Stern, The amniote primitive streak is defined by epithelial cell intercalation before gastrulation. Nature 449, 1049–1052 (2007). rFt t0ðx0Þ x1ðx1 0þd;x2 0;t0;tÞx1ðx1 0d;x2 0;t0;tÞ 2d x1ðx1 0;x2 0þd;t0;tÞx1ðx1 0;x2 0d;t0;tÞ 2d x2ðx1 0þd;x2 0;t0;tÞx2ðx1 0d;x2 0;t0;tÞ 2d x2ðx1 0;x2 0þd;t0;tÞx2ðx1 0;x2 0d;t0;tÞ 2d " # ð2Þ 22. M. Saadaoui, D. Rocancourt, J. Roussel, F. Corson, J. Gros, A tensile ring drives tissue flows to shape the gastrulating amniote embryo. Science 367, 453–458 (2020). 23. L. Gräper, Die Primitiventwicklung des Hühnchens nach stereokinematographischen Un- tersuchungen, kontrolliert durch vitale Farbmarkierung und verglichen mit der En- twicklung anderer Wirbeltiere. Dev. Genes Evol. 116, 382–429 (1929). ð2Þ where δ is the initial conditions’ grid spacing. After computing Ft t0ðx0Þ, we use Eq. 1 for computing the FTLE field. where δ is the initial conditions’ grid spacing. After computing Ft t0ðx0Þ, we use Eq. 1 for computing the FTLE field. 24. M. Chuai, W. Zeng, X. Yang, V. Boychenko, J. A. Glazier, C. J. Weijer, Cell movement during chick primitive streak formation. Dev. Biol. 296, 137–149 (2006). 25. Y. Nakaya, E. W. Sukowati, Y. Wu, G. Sheng, RhoA and microtubule dynamics control cell- basement membrane interaction in EMT during gastrulation. Nat. Cell Biol. 10, 765–775 (2008). SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E Schoenwolf, Specification of germ layer identity in the chick gastrula. BMC Dev. Biol. 7, 1–16 (2007). 33. C. B. Kimmel, W. W. Ballard, S. R. Kimmel, B. Ullmann, T. F. Schilling, Stages of embryonic development of the zebrafish. Dev. Dyn. 203, 253–310 (1995). 57. I. Burtscher, H. Lickert, Foxa2 regulates polarity and epithelialization in the endoderm germ layer of the mouse embryo. Development 136, 1029–1038 (2009). 34. J. P. Trinkaus, Ingression during early gastrulation of Fundulus. Dev. Biol. 177, 356–370 (1996). 58. F. Graner, B. Dollet, C. Raufaste, P. Marmottant, Discrete rearranging disordered patterns, part I: Robust statistical tools in two or three dimensions. Eur. Phys. J. E 25, 349–369 (2008). 35. L. A. Rohde, C.-P. Heisenberg, Zebrafish gastrulation: Cell movements, signals, and mech- anisms. Int. Rev. Cytol. 261, 159–192 (2007). 59. S. Urbansky, P. G. Avalos, M. Wosch, S. Lemke, Folded gastrulation and T48 drive the evo- lution of coordinated mesoderm internalization in flies. eLife 5, e18318 (2016). 36. D. Pinheiro, C.-P. Heisenberg, Zebrafish gastrulation: Putting fate in motion. Curr. Top. Dev. Biol. 136, 343–375 (2020). 60. K. Anastasia, G. Grigory, P. Ekaterina, D. Adrien, K. Yulia, T. Ulrich, Germ-layer commitment and axis formation in sea anemone embryonic cell aggregates. Proc. Natl. Acad. Sci. U.S.A. 115, 1813–1818 (2018). 37. A. Streit, K. J. Lee, I. Woo, C. Roberts, T. M. Jessell, C. D. Stern, Chordin regulates primitive streak development and the stability of induced neural cells, but is not sufficient for neural induction in the chick embryo. Development 125, 507–519 (1998). 61. S. C. Chapman, J. Collignon, G. C. Schoenwolf, A. Lumsden, Improved method for chick whole-embryo culture using a filter paper carrier. Dev. Dyn. 220, 284–289 (2001). 38. F. Bertocchini, C. D. Stern, Gata2 provides an early anterior bias and uncovers a global positioning system for polarity in the amniote embryo. Development 139, 4232–4238 (2012). 62. A. M. Eskicioglu, P. S. Fisher, Image quality measures and their performance. IEEE Trans. Commun. 43, 2959–2965 (1995). 63. G. Serrano Nájera, thesis, University of Dundee (2021). 39. H. C. Lee, H.-C. Lu, M. Turmaine, N. M. M. Oliveira, Y. Yang, I. De Almeida, C. D. Stern, Mo- lecular anatomy of the pre-primitive-streak chick embryo. Open Biol. 10, 190299 (2020). 64. M. Durande, S. Tlili, T. Homan, B. Guirao, F. Graner, H. Delanoe-Ayari, Fast determination of coarse-grained cell anisotropy and size in epithelial tissue images using Fourier transform. Phys. Rev. SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E 27. X. Yang, H. Chrisman, C. J. Weijer, PDGF signalling controls the migration of mesoderm cells during chick gastrulation by regulating N-cadherin expression. Development 135, 3521–3530 (2008). 50. M. J. Stower, F. Bertocchini, The evolution of amniote gastrulation: The blastopore-primi- tive streak transition. Wiley Interdiscip. Rev. Dev. Biol. 6, 1–17 (2017). 51. K. Mogi, R. Toyoizumi, Invasion by matrix metalloproteinase-expressing cells is important for primitive streak formation in early chick blastoderm. Cells Tissues Organs 192, 1–16 (2010). 28. M. A. Selleck, C. D. Stern, Fate mapping and cell lineage analysis of Hensen’s node in the chick embryo. Development 112, 615–626 (1991). 52. K. L. Crossin, S. Hoffman, M. Grumet, J. P. Thiery, G. M. Edelman, Site-restricted expression of cytotactin during development of the chicken embryo. J. Cell Biol. 102, 1917–1930 (1986). 29. X. Yang, D. Dormann, A. E. Munsterberg, C. J. Weijer, Cell movement patterns during gastrulation in the chick are controlled by positive and negative chemotaxis mediated by FGF4 and FGF8. Dev. Cell 3, 425–437 (2002). 53. Y. Nakaya, G. Sheng, An amicable separation: Chick’s way of doing EMT. Cell Adh. Migr. 3, 160–163 (2009). 30. A. Lawson, G. C. Schoenwolf, Epiblast and primitive-streak origins of the endoderm in the gastrulating chick embryo. Development 130, 3491–3501 (2003). 54. C. Kyprianou, N. Christodoulou, R. S. Hamilton, W. Nahaboo, D. S. Boomgaard, G. Amadei, 54. C. Kyprianou, N. Christodoulou, R. S. Hamilton, W. Nahaboo, D. S. Boomgaard, G. Amadei, I. Migeotte, M. Zernicka-Goetz, Basement membrane remodelling regulates mouse em- bryogenesis. Nature 582, 253–258 (2020). 31. G. Zhao, W. Li, D. Chen, J. R. Henry, H.-Y. Li, Z. Chen, M. Zia-Ebrahimi, L. Bloem, Y. Zhai, K. Huss, A novel, selective inhibitor of fibroblast growth factor receptors that shows a potent broad spectrum of antitumor activity in several tumor xenograft models. Mol. Cancer Ther. 10, 2200–2210 (2011). I. Migeotte, M. Zernicka-Goetz, Basement membrane remodelling regulates mouse em- bryogenesis. Nature 582, 253–258 (2020). 55. S. C. Chapman, F. R. Schubert, G. C. Schoenwolf, A. Lumsden, Analysis of spatial and temporal gene expression patterns in blastula and gastrula stage chick embryos. Dev. Biol. 245, 187–199 (2002). 32. H. Acloque, O. H. Ocana, A. Matheu, K. Rizzoti, C. Wise, R. Lovell-Badge, M. A. Nieto, Re- ciprocal repression between Sox3 and snail transcription factors defines embryonic terri- tories at gastrulation. Dev. Cell 21, 546–558 (2011). 56. S. C. Chapman, K. Matsumoto, Q. Cai, G. C. Acknowledgments 42. E. Rozbicki, M. Chuai, C. J. Weijer, Technique for liquid culture of early chick embryos suitable for long term live imaging (Research Square, 2015). Funding: This work was supported by the EASTBIO BBSRC PhD student training grant 1785593 (to G.S.N.); Swiss National Foundation, Schmidt Science Fellowship, and the Postdoc Mobility Fellowship (to M.S.); NSF-Simons Center for Mathematical and Statistical Analysis of Biology Award 1764269 (to L.M.); NIH 1R01HD097068 (to L.M.); Simons Foundation (to L.M.); Henri Seydoux Fund (to L.M.); BBSRC grants BB/N009789/1, BB/K00204X/1, BB/R000441/1, BB/ T006781/1 (to C.J.W.); and Wellcome Trust: Imaging equipment award 101468/Z/13/Z (to C.J.W.). Author contributions: Conceptualization: M.C., G.S.N., M.S., L.M., and C.J.W. Methodology: M.C., G.S.N., M.S., and C.J.W. Investigation: M.C., G.S.N., and M.S. Supervision: L.M. and C.J.W. Writing—original draft: G.S.N. and C.J.W. Writing—review and editing: M.C., G.S.N., M.S., L.M., and C.J.W. Competing interests: The authors declare that they have no competing interests. Data and materials availability: High-resolution images of confocal data and the original high-resolution images from the light sheet microscopy (movies S1, S4, S7, and S9) have been deposited in the BioImage Archive under accession number S-BIAD553 (www.ebi.ac. uk/biostudies/BioImages/studies/S-BIAD553). Higher-resolution versions of movies S1 to S11 are available at https://doi.org/10.15132/10000190. All other data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. 43. J. Chal, Z. Al Tanoury, M. Hestin, B. Gobert, S. Aivio, A. Hick, T. Cherrier, A. P. Nesmith, K. K. Parker, O. Pourquié, Generation of human muscle fibers and satellite-like cells from human pluripotent stem cells in vitro. Nat. Protoc. 11, 1833–1850 (2016). 44. D. R. Shook, E. M. Kasprowicz, L. A. Davidson, R. Keller, Large, long range tensile forces drive convergence during Xenopus blastopore closure and body axis elongation. eLife 7, e26944 (2018). 45. A. Eichmann, C. Corbel, V. Nataf, P. Vaigot, C. Bréant, N. M. Le Douarin, Ligand-dependent development of the endothelial and hemopoietic lineages from embryonic mesodermal cells expressing vascular endothelial growth factor receptor 2. Proc. Natl. Acad. Sci. U.S.A. 94, 5141–5146 (1997). 46. M. Chuai, D. Hughes, C. J. Weijer, Collective epithelial and mesenchymal cell migration during gastrulation. Curr. Genomics 13, 267–277 (2012). 47. J. Ma, D. J. Waxman, Modulation of the antitumor activity of metronomic cyclophospha- mide by the angiogenesis inhibitor axitinib. Mol. Cancer Ther. 7, 79–89 (2008). 48. J. Firmino, D. Rocancourt, M. Saadaoui, C. Moreau, J. Supplementary Materialsi Supplementary Materials This PDF file includes: Figs. S1 to S14 26. Y. Hatada, C. D. Stern, A fate map of the epiblast of the early chick embryo. Development 120, 2879–2889 (1994). y This PDF file includes: This PDF file includes: Figs. S1 to S14 This PDF file includes: Figs. S1 to S14 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 9 of 10 SC I EN C E A D VA N CE S \| RESE A RC H ARTI CL E E 99, 62401 (2019). 40. C. F. Arias, M. A. Herrero, C. D. Stern, F. Bertocchini, A molecular mechanism of symmetry breaking in the early chick embryo. Sci. Rep. 7, 15776 (2017). 65. E. Rozbicki, M. Chuai, C. J. Weijer, Liquid culture technique for early chick embryos suitable for long term live imaging. Nat. Protoc. Exchange2015). 10.1038/protex.2015.012. 41. G. Sanchez-Duffhues, E. Williams, M. J. Goumans, C. H. Heldin, P. ten Dijke, Bone mor- phogenetic protein receptors: Structure, function and targeting by selective small mole- cule kinase inhibitors. Bone 138, 115472 (2020). Science Advances (ISSN ) is published by the American Association for the Advancement of Science. 1200 New York Avenue NW, Washington, DC 20005. The title Science Advances is a registered trademark of AAAS. Copyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY). Use of this article is subject to the Terms of service Acknowledgments Gros, Cell division drives epithelial cell rearrangements during gastrulation in chick. Dev. Cell 36, 249–261 (2016). Submitted 3 December 2021 Accepted 28 November 2022 Published 4 January 2023 10.1126/sciadv.abn5429 Submitted 3 December 2021 Accepted 28 November 2022 Published 4 January 2023 10.1126/sciadv.abn5429 49. M. J. Stower, R. E. Diaz, L. C. Fernandez, M. W. Crother, B. Crother, A. Marco, P. A. Trainor, S. Srinivas, F. Bertocchini, Bi-modal strategy of gastrulation in reptiles. Dev. Dyn. 244, 1144–1157 (2015). 10 of 10 Chuai et al., Sci. Adv. 9, eabn5429 (2023) 4 January 2023 Reconstruction of distinct vertebrate gastrulation modes via modulation of key cell behaviors in the chick embryo Sci. Adv., 9 (1), eabn5429. • DOI: 10.1126/sciadv.abn5429 Use of this article is subject to the Terms of service Science Advances (ISSN ) is published by the American Association for the Advancement of Science. 1200 New York Avenue NW, Washington, DC 20005. The title Science Advances is a registered trademark of AAAS. Copyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution License 4.0 (CC BY).
https://openalex.org/W2035939515	https://zenodo.org/records/2118688/files/article.pdf	English	null	A SULPHOSALT OF THE ALIPHATIC CREOSOTE-ESTERS, AND ITS THERAPEUTIC USEFULNESS.	JAMA	1,900	public-domain	8,936	Sulphur stimulates the mucosa of the alimentary tract and its administration is frequently followed by a slight increase of intestinal secretion and peristalsis. Besides its valuable antifermentative properties, it seems to exhibit an especial affinity for one or the other components of the tissue albumin. Thus, I think, it facilitates the resorption or the deposition of the creo- sote derivative into the glandular system. fresh the overheated blood. Water is frequently repudi- ated by men and women, wholly unmindful that it has advantages for them beyond price. There are no known counterindications to water as a drink, and the quan- tity may be whatever the stomach and intestines are able to absorb. It is safe to say that pure water may be drunk at any time and with hardly any limitations save such as might appeal to anyone. In many cases, cover- ing seventeen years of observation, drinking of water freely with meals, immediately before or after them, has not been attended with bad effects, nor does it justify the numerous precautions generally given against mixing water and food in the stomach. Rightly considered, drinking with meals ought to be a benefit. It holds the food in better solution till acted on by the juices, sup- plying fluid to the blood-stream, and, most important, forming a -natural safeguard against gluttony. It is possible to keep well for a certain length of time without attention to physiology or dietetics, but it is a mistake which comes in for compound interest at the end. Moderately used, it is reasonably safe to indulge in coffee, tea and even alchoholics, but it is a hazard which a first-class life risk should avoid. Water is universally man's sweetest and safest drink, and rightly used would in itself largely help to extend his life well toward the century mark. Food tastes better and is more agreeably relished by the water drinker than by those who drink wine at table. Liquors confer no useful assistance in passing the dangers of life, and in self-interest it would be nearer to safety to let Nature's provision for drink have full credit, as being the best, and accept no tradesman's substitute. The best drink for man is pure water, and the ordinary drinking water of a country is always superior to any of the so-called health waters or bottle drinks. A SULPHOSALT OF THE ALIPHATIC CREO- SOTE-ESTERS, AND ITS THERAPEUTIC USEFULNESS. BY HEINRICH STERN, Ph.D., M.D. Professor of the Diseases of Metabolism, College of Physicians and Surgeons, St. Louis. Visiting Physician New York Red Cross Hospital. Attending Physician St. Elizabeth's Hospital. Member American Chemical Society, American Med- ical Association, Etc. NEW YORK CITY. p g y Calcium, finally, that is the form found in this salt, lessens the acid degree of the material contained in the alimentary organs, thereby preventing or allaying un- due irritation and neutralizing certain substances of toxic tendencies. Early in May, 1898, Dr. G. Wendt, of Berlin, the manufacturer of a series of sulpho-acid salts of the ali- phatic creosote-esters, through his American agents,1 placed some of his products in my hands for thera- peutic investigation. In responding to the kind invita- tion of your secretary, I grasped the opportunity to re- port at this meeting the therapeutic qualities of one of these salts. I k The eosolate of calcium considered as a guaiacol de- rivative has this formula: (C9H7S3012)2. Ca3 and contains, therefore, in the neighborhood of 25 per cent, of creosote. It is a grayish powder, feeling to the touch like finely pulverized pumice-stone. Its odor is slightly pungent and somewhat ethereal, its taste a little acrid and leathery. It is soluble in from eight to ten parts of cold, and in seven parts of hot water. It is very slightly soluble in alcohol, and insoluble in chloroform and turpentin, but is readily dissolved by hydrochloric and by citric, and by some other organic acids, while it dissolves only slowly in acetic acidJ Besides the lime salt, of which I shall presently speak, a salt of silver, argentum eosolicum, C6HOCH3 OC2H3OAG3(SO3)3, resp. C9, H7, Ag3, S3, O12 and one of quinin, chininum eosolicum, (C9H7S3O12) (C20H24N2O2)3, were given to me in order to deter- mine their therapeutic usefulness. Experimental work of another nature, however, prevented me from taking up the latter combinations on a larger scale. i t d , y y In a dog, weighing 16 kilog., 3 gm. of the eosolate of calcium produced severe vomiting and catarrhal condi- tion of the nasal and pharyngeal mucosa. In a dog weighing 7.5 kilog., 3 gm. of the salt brought on vom- iting, intense purgation and marked emaciation. In the healthy human organism 0.33 seemed to exert but little influence; 0.6 has produced a fulness in the epi- gastric region, slight constipation and diaphoresis; 1 gm. Presented to the Section on Materia Medica, Pharmacy and Therapeutics, at the Fiftieth Annual Meeting of the American Medical Association, held at Columbus, Ohio. June 6-9, 1899. 1 The Fischer Chemical Importing Co., New York City. g y The therapeutic value of creosote or its esters is too well known to be discussed here at length. Undoubt- edly creosote is gradually split up into more simple bodies, in which forms it is carried to the different or- gans. The neutralizing and antiputrefactive powers of creosote are comparatively limited if tested in the laboratory; in the body, however, it exerts these qual- ities in a greater measure than the majority of the more energetic so-called antiseptics. This fact I attribute to its non-coagulating of albumin. Mercuric corrosive chlorid, carbolic acid, salicylic acid, alcohol, beta- naphthol-sulphonate, resorcin, zinc chlorid, zinc sul- phate and other antiseptic agents are coagulants of albumin, and as such may become tissue destroyers. In consequence thereof great caution is exercised with most of these antiseptics when administered internally, and only small doses of them, in many instances insufficient for the purpose, are as a rule prescribed. Moreover, as is the case with corrosive sublimate for instance, Koch's "strongest antiseptic," the mercury coagulates and throws down the albumin occurring in the medium to be disinfected, combining to mercury albuminate, also possessing antiseptic qualities but which, by its forma- tion, deprives the supernatent fluid almost totally of its contents of mercury. The internal administration, therefore, of mercuric chlorid, and the antiseptics be- longing to its class, should. be restricted to certain specific and well-defined pathologic conditions, and their employment as disinfectants or germicides, if not other- wise combined, should be abandoned altogether, both on account of their toxicity and their relative ineffi- ciency in all those chronic affections, characterized by progressive systemic decline. 10 W. Forty-ninth Street. Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 p DIABETES INSIPIDUS. J l C y g , g , y July 11: Weight, 40 kilog.; the amount of urine voided 4200 c.c. Treatment: Mixed food and water ad lib., and cal- cium eosolicum continued. No. of Micturition. Time. Amount. 1.Mav 30, 8 p.m. 420 c.c. 2.May 30, 9:30 p.m. 420 c.c. 3.During night,i. 780 c.c. 4.During night. 570 c.c. 5.May 30, 7:20 a.m. 690 c.c. 6.May 30, 9:55 a.m. 960 c.c. 7.Mav 30, 12 m. 630 c.c. 8.May 30, 4 p.m. 930 c.c. 9.May 30, 6 p.m. 660 c.c. 10.May 30, 7:30 p.m. 630 c.c. 6690 c.c. July 20: Weight, 40.5 kilog.; rectal temperature, 37.1 C. (98.8 F.), with diuresis somewhat more pronounced. J l 23 W i h 40 kil t l C ( ), pronounced. July 23: Weight, 40 kilog.; rectal temperature, 37.2 C. (99 F.), diuresis declining. The patient felt very comfortable, with thirst diminished. Calcium eosolicum was advised to be continued and he was discharged, improving.3 ( ), p o ou ced July 23: Weight, 40 kilog.; rectal temperature, 37.2 C. (99 F.), diuresis declining. The patient felt very comfortable, with thirst diminished. Calcium eosolicum was advised to be continued and he was discharged, improving.3 g , p g Case 2.—E. R., male, aged 47, married, American; first con- sulted me on May 19, 1897. He had felt depressed for some time, and complained of anorexia, constipation, great thirst, polyuria and deterioration of sexual power, and of an eczema- tous condition on the neck and extremities. His weight was 68.5 kilog.; temperature (axilla), 37.1 C. (98.8 F.), with hydru- resis very pronounced. The urine was pale, limpid, sp. gr. 1002, neutral; no excess of phosphates, chlorids or sulphates; urea increased absolutely; neither albumin nor glucose. T t t N t i ti f di g , p g Case 2.—E. R., male, aged 47, married, American; first con- sulted me on May 19, 1897. He had felt depressed for some time, and complained of anorexia, constipation, great thirst, polyuria and deterioration of sexual power, and of an eczema- tous condition on the neck and extremities. His weight was 68.5 kilog.; temperature (axilla), 37.1 C. (98.8 F.), with hydru- resis very pronounced. The urine was pale, limpid, sp. gr. 1002, neutral; no excess of phosphates, chlorids or sulphates; urea increased absolutely; neither albumin nor glucose. p DIABETES INSIPIDUS. T t t N t i ti f di The specific gravity of the twenty-four hours' urine was 998.5.2 I mention this fact in particular, as I have never before seen a specimen of urine lighter than water, and as I, in the literature of this subject, failed to notice any references to urines of a specific gravity less than 1000. I must add that the urine had cooled before the specific gravity was ascertained: Two hydrometers standardized at 25 C. were employed; both indicated the same degree. The urinoglucosometer devised by me, one of the scales of which descends to 990 degrees, pointed out a density somewhat higher, but still below 999 degrees. This instrument being standardized at 15.6 C, it was necessary to cool off the urine to that temperature. Pycnometric meas- urements were not made. f f ilk y; g ucose Treatment.—No restriction of diet or water was imposed, but alcoholics were- prohibited, and antipyrin, 0.35 gm. given three times daily, with local medication for eczema. The symp- toms did not improve, and successively ergot, Fowler's solution, salol and strychnia were prescribed without effecting any change in the general condition. N b b l ilk di Di g g On November 16 an absolute milk diet was ordered. Dip- sesis decreased somewhat, as well as did diuresis. All medical agencies were dispensed with, excepting local applications. Toward the middle of January, 1898, the patient had to aban- don the milk regimen and return to the unrestricted diet. Treatment.—This was free ingestion of milk, to the total ex- clusion of all other nutrients and water; sitz baths slightly above body temperature; faradism, the ergot withdrawn and substituted by calcium eosolicum, 0.3 gm., three times daily. regimen the unrestricted diet. January 26: Weight 69.25 kilog., temperature normal. He was very irritable; thirst and diuresis slightly augmented. Th t t f th ti t i l i d lt d e y ; slightly augmented. The status of the patient, in general, remained unaltered, with no permanent improvement. Th i d y y The patient, as seen from the table, steadily gained in weight under this treatment. This was the more re- markable, as an increase in weight had not been noticed for years. At the same time the patient lost much of his nervousness and developed a healthy appearance. 2 I deem it irrelevant for the present purpose to give all the dates of the various very minute and exactive urinary examina- tions. I shall attempt to report this case in full at some other time. 3 The patient returned to his native village in a neighboring State. The last time he wrote to me he reported further im- provement. A SULPHOSALT OF THE ALIPHATIC CREO- SOTE-ESTERS, AND ITS THERAPEUTIC USEFULNESS. 7,050 9 783.33 1006 7,500 June 3-4,1898 From 8 p.m. to 6 :30 p.m. . . 6,600 11 600 1004 7,00035.75 June 4- 5,1898 From 6:30 P.M. to 6 p.m.. . 6,720 10 672 1007 8,000 36 June 5- 6,1898 From 6 p.m. to 8 P.M . . . 8,160 14 582.86 1004 8,000 June 6-7,1898 From 8 p.m. to 8 p.m . . . 6,000 10 600 1008 8,000 June 7- 8,1898 From 8 p.m. to 8 p.m . . . 6,600 10 660 1004 8.000 37 June 8- 9,1898 From 8 p.m. to 8 p.m . . . 5,400 10 540 1006 8.000 June 9-10,1898 From 8 p.m. to 8 P.M . . . 5,700 10 570 1006 8,000 37.5 June 10-11,1898 From 8 p.m. to 8 P.M . . . 6,000 10 600 1005 8,000 June 11-12,1898 From 8 p.m. to 8 P.M . . . 4,800 9 533.33 1008 8,000i June 12-13,1898 From 8 P.M. to 8 p.m . . . 7,500 12 625 1007 8,000 38.5 June 13-14,1898 From 8 p.m. to 8 p.m . . . 5,400 8 675 1005 7,000 June 14-15,1898 From 8 p.m. to 8 P.M . . . 5,250 11 477.27 1009 7,000 June 15-16,1898 From 8 p.m. to 8 p.m . . . 5,100 10 510 1008 7,000 June 16-17,1898 From 8 p.m. to 8 p.m . . . 4,500 9 500 1009 7,000 39.25 From June 1-2 to June 17. the following records were taken: Eosolate of calcium, in doses less than 1 gm., does not impart the greenish tint to the urine which gen- erally appears after the ingestion of phenols. When 2 gms were administered, a discoloration was noticeable in some instances. sta ces I have clinically experimented with calcium eosolicum in a variety of disorders. I report the following cases: A SULPHOSALT OF THE ALIPHATIC CREO- SOTE-ESTERS, AND ITS THERAPEUTIC USEFULNESS. Treatment: Diet, 4 liters of milk daih- (the amount had to be lowered as the patient developed a dislike for it) ; calcium eosolicum, 0.5 gm. four times daily. July 7: Weight, 40 kilog., hydruria diminished. July 11: Weight, 40 kilog.; the amount of urine voided 4200 c.c. Treatment: Mixed food and water ad lib., and cal- cium eosolicum continued. July 20: Weight, 40.5 kilog.; rectal temperature, 37.1 C. (98.8 F.), with diuresis somewhat more pronounced. July 23: Weight, 40 kilog.; rectal temperature, 37.2 C. (99 F.), diuresis declining. The patient felt very comfortable, with thirst diminished. Calcium eosolicum was advised to be continued and he was discharged, improving.3 Case 2.—E. R., male, aged 47, married, American; first con- sulted me on May 19, 1897. He had felt depressed for some time, and complained of anorexia, constipation, great thirst, polyuria and deterioration of sexual power, and of an eczema- tous condition on the neck and extremities. His weight was 68.5 kilog.; temperature (axilla), 37.1 C. (98.8 F.), with hydru- resis very pronounced. The urine was pale, limpid, sp. gr. 1002, neutral; no excess of phosphates, chlorids or sulphates; urea increased absolutely; neither albumin nor glucose. Treatment.—No restriction of diet or water was imposed, but alcoholics were- prohibited, and antipyrin, 0.35 gm. given three times daily, with local medication for eczema. The symp- toms did not improve, and successively ergot, Fowler's solution, salol and strychnia were prescribed without effecting any change in the general condition. On November 16 an absolute milk diet was ordered. Dip- sesis decreased somewhat, as well as did diuresis. All medical agencies were dispensed with, excepting local applications. Toward the middle of January, 1898, the patient had to aban- don the milk regimen and return to the unrestricted diet. January 26: Weight 69.25 kilog., temperature normal. He was very irritable; thirst and diuresis slightly augmented. The status of the patient, in general, remained unaltered, with t i t From June 1-2 to June 17. the following records were taken: 8 a °«8 o =5 . .S o.s u.SIf £-B :\| I ii II IsII P . a .a -s-rt ;.S o '5-s ate >>« 0> 3h "Jl ftW 3-S o« C z < oj C M June 1-2,1898 From 7:30 P.M to 6 :50 p.m. 7.800 10 780 11002 8,00035.25 June 2-3,1898 From 6:50p.m. to 8 P.M.. . p DIABETES INSIPIDUS. Case 1.—W. C, male, aged 16 years, an American with no occupation, was referred to me for treatment on May 22, 1898. A i H l h Th i ith occupation, y Anamnesis.—Healthy parentage. The patient met with a severe accident when 7 years old, from the effects of which he never fully recuperated. Hydruria or polyuria were present ever since, and the body development came to a standstill; he was under the care of a score of physicians. physicians. Examination.-—He to all appearances looked like a 9 or 10- year-old boy; height, 4 feet 8Vt inches, weight 35.5 kilog. Temperature, rectal, was 36.7 C. (98 F.). Pronounced emacia- tion and nervousness was evident. Physical examination of the heart and lungs revealed no disorder. The digestive organs and abdominal viscera were apparently normal. Mental acuteness was very keen, memory good, and there were no luetic symp- toms. d l June 19: Weight, 39.75 kilog.; rectal temperature, 37.2 C. (99 F.), and improvement plainly noticeable. Treatment: Diet, 5 liters of milk per diem; calcium eosolicum, 0.3 gm. four times daily. J (99 Urine (voided in my presence).-—570 c.c; watery, odorless, sp. gr. 1002, very clear, acid (degree 0.03); solids, 1.5 gm.; carbamid, 0.571 gm.; chlorids, 0.285 gm.; phosphates, see de- gree of acidity; no albumin, no dextrose. daily. June 26: Weight, 40 kilog.; rectal temperature, 37.2 C. (99 F.), the hydruric symptoms still improving. Treatment: Diet, 4 liters of milk daih- (the amount had to be lowered as the patient developed a dislike for it) ; calcium eosolicum, 0.5 gm. four times daily. gree acidity; , Treatment.—No dietetic restrictions were imposed, but a bath of 37.5 C. every other night; extract ergot fl., 3 gm., four times a day. On May 31 the patient was still more emaciated— weight 35'kilog.; rectal temperature 36.6 C. (97.8 F.). He had voided, since 8 p.m. of the previous day, the following amounts of urine: da y July 7: Weight, 40 kilog., hydruria diminished. J l W i h kil h f i da y July 7: Weight, 40 kilog., hydruria diminished. J l h f i id d July Weight, kilog., hydruria July 11: Weight, 40 kilog.; the amount of urine voided 4200 c.c. Treatment: Mixed food and water ad lib., and cal- cium eosolicum continued. Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 A SULPHOSALT OF THE ALIPHATIC CREO- SOTE-ESTERS, AND ITS THERAPEUTIC USEFULNESS. has produced griping pains in the intestines, and the ingestion of 2 gm. was followed by a profuse diar- rhea. up t e larger The calcium salt was the one most experimented with by me. Three distinct therapeutic elements enter into its composition, viz., sulphur, creosote-ester and calcium. From June 1-2 to June 17. the following records were taken: 8 a °«8 o =5 . .S o.s u.SIf £-B :\| I ii II IsII P . a .a -s-rt ;.S o '5-s ate >>« 0> 3h "Jl ftW 3-S o« C z < oj C M June 1-2,1898 From 7:30 P.M to 6 :50 p.m. 7.800 10 780 11002 8,00035.25 June 2-3,1898 From 6:50p.m. to 8 P.M.. . 7,050 9 783.33 1006 7,500 June 3-4,1898 From 8 p.m. to 6 :30 p.m. . . 6,600 11 600 1004 7,00035.75 June 4- 5,1898 From 6:30 P.M. to 6 p.m.. . 6,720 10 672 1007 8,000 36 June 5- 6,1898 From 6 p.m. to 8 P.M . . . 8,160 14 582.86 1004 8,000 June 6-7,1898 From 8 p.m. to 8 p.m . . . 6,000 10 600 1008 8,000 June 7- 8,1898 From 8 p.m. to 8 p.m . . . 6,600 10 660 1004 8.000 37 June 8- 9,1898 From 8 p.m. to 8 p.m . . . 5,400 10 540 1006 8.000 June 9-10,1898 From 8 p.m. to 8 P.M . . . 5,700 10 570 1006 8,000 37.5 June 10-11,1898 From 8 p.m. to 8 P.M . . . 6,000 10 600 1005 8,000 June 11-12,1898 From 8 p.m. to 8 P.M . . . 4,800 9 533.33 1008 8,000i June 12-13,1898 From 8 P.M. to 8 p.m . . . 7,500 12 625 1007 8,000 38.5 June 13-14,1898 From 8 p.m. to 8 p.m . . . 5,400 8 675 1005 7,000 June 14-15,1898 From 8 p.m. to 8 P.M . . . 5,250 11 477.27 1009 7,000 June 15-16,1898 From 8 p.m. to 8 p.m . . . 5,100 10 510 1008 7,000 June 16-17,1898 From 8 p.m. to 8 p.m . . . 4,500 9 500 1009 7,000 39.25 June 19: Weight, 39.75 kilog.; rectal temperature, 37.2 C. (99 F.), and improvement plainly noticeable. Treatment: Diet, 5 liters of milk per diem; calcium eosolicum, 0.3 gm. four times daily. June 26: Weight, 40 kilog.; rectal temperature, 37.2 C. (99 F.), the hydruric symptoms still improving. p DIABETES INSIPIDUS. The hydruric condition improved gradually, that is in a relative sense. The quantity of the twenty-four- hours urine for the total period became remarkably diminished, though occasional increases were recorded, and the amount of solids excreted, as seen by the urin- ary density, was far in excess of that prior to the institu- tion of this treatment. p p May 14: The milk regimen was resorted to again, and a fortnight later calcium eosolicum in 0.3 gm. doses four times daily was ordered to be taken. i h d kil daily June 12: Patient weighed <2 kilog. and felt very well. The eczema had disappeared; feeling of thirst and diuresis greatly improved—amount of twenty-four hours urine, 3800 c.c, while 5 liters of milk were taken. The milk regimen was dispensed with after another week and the dose of calcium eosolicum increased to 0.75 gm. four times daily. A decided and progressive improvement followed the administration of this medi- cine, which, in this dose even, was well borne by the stomach; an occasional slight constipation was caused July 24: Weight 97.5 kilog.; urine's sp. gr. 1028, glucose 2.1 per cent. by the calcium salt, but readily yielded to mild evacu- ants. I found the patient, when last seen in August, 189S, in high.spirits and he was reminded of his recent condition only by transitory attacks of dipsesis and subsequent hydruria by the calcium salt, but readily yielded to mild evacu- ants. I found the patient, when last seen in August, 189S, in high.spirits and he was reminded of his recent condition only by transitory attacks of dipsesis and subsequent hydruria perJuly 30: Weight 95.75 kilog.; sp. gr. of urine 1029.5, glu- cose 1.2 per cent. Treatment: Diet as above; calcium eosolicum,. 0.4 gm. four times daily. gm. daily. August 6: Weight 96.5 kilog.; sp. gr. of urine 1026, glucose 0.45 per cent. subsequent hydruria. Case 3.—B. I., female, aged 24, married, American, and mother of one child, suffered from polydipsia, hyperdiuresis, progressive emaciation and dysphoria for about six months, when she was referred to me for treatment on Sept. 26, 1898. Her weight was 55.5 kilog., her rectal temperature 37.2 C. (99 F.) ; bronchial irritation and cough were present, the lungs and abdominal organs apparently sound. The urine was pathognomonic of hydruria; excessive in quantity, of water- color, of low density and a faint acidity. CHRONIC ULCERATIVE PHTHISIS. July 14: Traces of glucose; diet continued, with calcium eosolicum, 0.5 gm. four times a day. T f i id t i t gm. every June 1: Weight 55.5 kilog., temperature (axilla), 6 p.m., 37.8 C. (100 F.) ; feels much stronger. J W i h gm. every June 1: Weight 55.5 kilog., temperature (axilla), 6 p.m., 37.8 C. (100 F.) ; feels much stronger. J W i h 56 25 kil ( ill ) , g y July 28: Traces of glucose, strict proteid nutriment; calcium eosolicum as before. l ( ) stronger. June 4: Weight 56.25 kilog., temperature (axilla), 4 p.m., 37.8 C. (100 F.); feels stronger. Treatment: Diet as above; fresh air; sun baths; calcium eosolicum, 0.6 gm. every three hours. August 2: No glucose; milk diet; calcium eosolicum dis- pensed with. August 2: No glucose; milk diet; calcium eosolicum dis- pensed with. di pensed with. August 6: No glucose; moderate mixed diet; no medication. Th i di i i d f i l pensed with. August 6: No glucose; moderate mixed diet; no medication. Th i di i i d f i l June 12: Weight 57.5 kilog., temperature (axilla), 3:30 p.m., 37.2 C. (99 F.). August 6: No glucose; ode ate ed diet; no medication. The patient, without further medication, continued fairly well, until the end of September, when she was taken with endocarditis to which she succumbed on October 3. G G p.m., ( ) June 18: Weight 58 kilog., temperature (axilla), 4 p.m., 37.2 C. (99 F.); feels-comfortable. Treatment: Diet, etc., as above; calcium eosolicum, 0.7 gm. every three hours. W i ht 58 25 Case 2.—Mrs. G., aged 47, German, had had the disease five years. h i h ; , g e e y June 25: Weight 58.25 kilog., temperature (axilla) 37.2 C. (99 F.) ; no cough; little perspiration; tubercle bacilli in the sputum. years. June 1, 1898: Under a diet largely albuminous, her weight was 72.5 kilog. The sp. gr. of the urine was 1026.5, glucose 2 per cent. Treatment: The same diet, of sufficient caloric value, was continued, and calcium eosolicum, 0.3 gm. three times daily. p The patient insisted on going home (Maryland); discharged—improving. When heard from last, im- provement persisted. The large doses of this salt of eosol, averaging for the last week 4.2 gm. daily, were very well borne by the patient. daily. p DIABETES MELLITUS. per October 18: Urine's sp. gr. 1022, glucose 0.45 per cent. Treatment: Milk regimen; calcium esolicum, 0.35 gm. three times a day I have tested calcium eosolicum in five diabetic pa- tients. 1 shall not endeavor to relate the complete clinical history of these cases, but will refer briefly only to the periods during which the medicine was taken. es day. October 20: Urine's sp. gr. 1021, glucose, 0.2 per ce N b 1 1022 5 f p g , g , p November 1: Urine's sp. gr. 1022.5, with traces of glucose. Treatment discontinued. CHRONIC ULCERATIVE PHTHISIS. taken. Case 1.—Mrs. B., aged 65, English, had been affected with diabetes mellitus since her 58th year, the quantity of urinary glucose under a milk regimen and a moderate mixed diet fluctuating between 0.25 and 2.5 per cent. Absolute proteid nourishment was not permissible, on account of excess of ace- tone and the presence of ethyldiacetic acid in the urine. f i h The action of calcium eosolicum in ten cases of this affection which were under my observation, during the course of the year, is recorded in the following: year, t e following: Case 1.—M. C. G., aged 26, female, single, American. The disease was first recognized 2% years previously. M h i j d f presence ethyldiacetic May 27, 1898: The amount of glucose in the urine was 2.25 per cent. Treatment: Milk diet and calcium eosolicum, 0.3 gm. three times daily. g y p y May 28, 1898: Patient having just returned from a sanator- ium, weight 56 kilog., axillary .temperature, 4 p.m., 38.1 C (100.5 F.). The disease was unilateral; coughing and ex- pectorating moderately; tubercle bacilli in the sputum; heart's action full and accelerated; complaints of drenching perspira- tion, of great weakness—patient hardly able to walk—but of little pain. Treatment: Overallmentation; calcium eosolicum, 0.5 gm. every three hours. g y p y May 28, 1898: Patient having just returned from a sanator- ium, weight 56 kilog., axillary .temperature, 4 p.m., 38.1 C (100.5 F.). The disease was unilateral; coughing and ex- pectorating moderately; tubercle bacilli in the sputum; heart's action full and accelerated; complaints of drenching perspira- tion, of great weakness—patient hardly able to walk—but of little pain. Treatment: Overallmentation; calcium eosolicum, 0.5 gm. every three hours. gm. daily. June 2: Traces of glucose; treatment continued. Gl b t t t t ti d g ucose; June 16: Glucose absent; treatment continued. J T f l T t t M d g ; June 16: Glucose absent; treatment continued. J T f l T M d t i ; June 25: Traces of glucose. Treatment: Moderate, mixed diet, calcium eosolicum, 0.4 gm. thrice daily. ; June 25: Traces of glucose. Treatment: Moderate, mixed diet, calcium eosolicum, 0.4 gm. thrice daily. , , gm. daily. July 6: Glucose present—0.25 per cent. Treatment: Diet continued; calcium eosolicum, 0.5 gm. thrice daily. i d continued; , gm. daily. p DIABETES INSIPIDUS. The patient was put on an exclusive milk diet and on calcium eosolicum, 0.5 gm. four times daily. She reported improvement on September 30. The milk regimen was continued (5 liters for the twenty-four hours), and the dose of the medicine increased to 0.75 gm. per August 13: Weight 96.5 kilog. The urine showed a sp. gr. of 1026, and traces of glucose. Treatment: Proteid diet, 100 gm. of carbohydrates allowed for the twenty-four hours, and the medicine continued. September 16: Weight 96 kilog.; sp. gr. of urine 1023.5, and' traces of glucose. The medicine was withdrawn. glucose. Case 4.—Mrs. S., aged 64, Hebrew, the duration of the dis- ease three years, was in charge of her family physician. S , aged , eb e , ease three years, was in charge of her family physicia A 5 1898 U i l 0 2 y , c a ge a y physician. Aug. 5, 1898: Urine, glucose 0.2 per cent, serum albumin in appreciable quantity, acetone present. Treatment: Dietary, milk; medicinally, calcium eosolicum, 0.3 gm. three times a day. day. August 6: Urine showed no glucose, serum albumin in appre- ciable quantity, and traces of acetone. hours), gm. October 6: Her weight amounted to 56.5 kilog., and her rectal temperature was 37.3 C. (99.1 F.); polydipsia and diu- resis were diminished. quantity, August 8 and 9: Urine showed no glucose, nor serum albu- min and no acetone. Diet and medicine were continued for some time afterward. were The urine voided from October 12, 11 a.m., to October 13, 11 a.m., was 3150 c.c; the milk ingested, 4500 c.c. Her weight, October 13, was 58 kilog.; rectal temperature 37.3 C. (99 F.). She continued to improve when seen last in November, 1898. Case 5.—Mrs. R., aged 47, Hebrew, apparently had aliment- ary diabetes of short duration. a y Oct. 14, 1898: Urine presented a sp. gr. of 1019.5, glucose 0.5 per cent. CHRONIC ULCERATIVE PHTHISIS. June 6: Weight 73 kilog. The urine showed a sp. gr. of 1024, and glucose 1.75 per cent. W i h kil i h f l , glucose per June 11: Weight 74 kilog., with sp. gr. of 1024, glucose 1.15 per cent. W i h kil 1021 l very by patient. Case 2.—M. M. D., aged 24, a female, married, Irish. The disease was first recognized six months previously. M 30 W i h perJune 19: Weight 74.75 kilog., urine's sp. gr. 1021.5, glucose 1 per cent. Treatment: Same diet continued; calcium eosoli- cum, 0.5 gm. three times a day. h i h d ecog ed p e ous y May 30, 1898: Weight 54 kilog., temperature (axilla), 3 p.m., 38.9 C. (102 F.) ; adhesions of pleura; constant pain beneath left scapula; dulness in left clavicle region; cavities at the apex; cough loose, expectoration, grayish-purulent; tubercle bacilli in sputum; nocturnal perspiration. Treat- ment: Overalimentation (milk and fats) ; fresh air. cum, gm. day. August 12: Weight 76 kilog.; the urine showed a sp. gr. of 1017, and traces of glucose. , glucose. August 25: Weight 77.5 kilog.; sp. gr. of the urine 1019, and no glucose. S W i h glucose. September 16: Weight 80.125 kilog.; sp. gr. of the urine 1019, with no glucose. The medicine was withdrawn. ( R-. Calcii eosolici. 7 5 Aquae font. 50 Extraeti hyoscyami fluidi. 2 Syr. pruni virginiana?. 60 Glyeerini q. s. ad. 150 ( R-. Calcii eosolici. 7 5 Aquae font. 50 Extraeti hyoscyami fluidi. 2 Syr. pruni virginiana?. 60 Glyeerini q. s. ad. 150 M. Sig. A teaspoonful every three hours. June 8: Weight 55.5 kilog., temperature (axilla), 3 p.m. 38.3 C. (101 F.) , glucose. Case 3.—Mrs. W., aged 53, German, presented herself, but the duration of the disease was unknown. W i h July 22, 1898: Weight was 98 kilog. The diet was unre- stricted, contrary to the advice of the family physician. The urine presented a sp. gr. of 1034, glucose 3.02 per cent. Treat- ment: Absolute proteid nourishment; hydrargyrum chloridum mite, 0.005 gm. every two hours. July 22, 1898: Weight was 98 kilog. The diet was unre- stricted, contrary to the advice of the family physician. The urine presented a sp. gr. of 1034, glucose 3.02 per cent. Treat- ment: Absolute proteid nourishment; hydrargyrum chloridum mite, 0.005 gm. every two hours. M. Sig. (99 F.) ; patient discontinued treatment. D b 2 A k f l i l ( ) September 6: Weight 85.75 kilog., rectal temperature 37.2 C. (99 F.) ; patient discontinued treatment. D b 2 A k f l i l Case 7.—D. T., aged 59, male, German; the duration of the disease as three years. ( ) ; pat e t December 27: Attack of la grippe; rectal temperature 40 C. (104 F.) ; acute affection lasting three weeks. y August 14, 1898: Weight 52.5 kilog., rectal temperature 37.8 C. (100 F.); repeated attacks of hemoptysis from exca- vation; cough loose; expectoration profuse; diarrhea; pro- nounced adynamia; tubercle bacilli in sputum. Treatment: Overalimentation; hypodermic injections of strychnia; calcium eosolicum, 0.3 gm. three times a day. A W i h 3 kil ( ) ; ast g Feb. 3, 1899: Weight 79 kilog., rectal temperature 37.5 C. (99.5 F.) Treatment: As above. ( ) R. Heroin . 0 15 Calcii eosolici. 8 Aquae . 50 Syr. tolutani Glycerini, aft. 60 M Si A t f l three ho rs , g day August 21: Weight 53 kilog., rectal temperature 37.7 C. (99.9 F.) Treatment: As above; calcium eosolicum, 0.4 gm. thrice daily, combined with sodium bicarbonate and charcoal. A t 28 W i ht 54 kil l 37 8 C Glycerini, M. Sig. A teaspoonful every three hours. Sig. teaspoonful every February 7: Weight 79.25 kilog., dyspeptic, symptoms. Treatment: Diet as above; mixture discontinued, instead of it: da y, August 28: Weight 54 kilog., rectal temperature 37.8 C. (100 F.) S t b W i ht kil l abo e; d sco t ued, R. Calcii eosolici Natrii biearbonici, aft. 013 Carbo animalis purificatus. 0\|75 Ft l d t l d t i i t Si O d f ( ) September 4: Weight 54 kilog.. rectal temperature 37.7 C. (99.9 F.) S t b 6 W i ht 53 5 kil l ( ) September 6: Weight 53.5 kilog., rectal temperature 37.2 C. (99 F.) Treatment: As above; calcium eosolicum, 0.3 gm. twice daily. S b ( ) September 6: Weight 53.5 kilog., rectal temperature 37.2 C. (99 F.) Treatment: As above; calcium eosolicum, 0.3 gm. twice daily. S b p \| Ft. pulv. d. tal. dos. triginta. Sig. One powder four times daily. (100 F.) ; physical signs of local disease very pro- nounced; hectic condition; asthenia, cough and expectoration loose; no tubercle bacilli in sputum. Treatment: Absolute milk regimen, starting with 4 liters a day; to spend twelve hours in bed, and the rest of the day in the open air; calcium eosolicum, 0.3 gm. four times daily. J 17 W i ht 8 kil ( ill ) p y June 5: Weight 77 kilog., temperature (axilla), 6 p.m., 37.8 C. (100 F.) ; physical signs of local disease very pro- nounced; hectic condition; asthenia, cough and expectoration loose; no tubercle bacilli in sputum. Treatment: Absolute milk regimen, starting with 4 liters a day; to spend twelve hours in bed, and the rest of the day in the open air; calcium eosolicum, 0.3 gm. four times daily. J 17 W i ht kil ( ill ) q a s , q. , M. Sig. A teaspoonful every three hours. January 27: Weight 56.25 kilog., temperature (axilla) 37.1 C. (98.8 F.) M h W i h ( ) March 1: Weight 59 kilog., temperature (axilla) 37.2 C. (99 F.) C 9 F 31 ( ) Case 9.—F. U., aged 31, female, married, American; dura- tion of disease 10 months. J 10 1899 Di ll , gm. y June 17: Weight 78.75 kilog., temperature (axilla), 6 p.m., 37.2 C. (99 F.) Treatment: Diet, 5 liters of milk daily; cal- cium eosolicum, 0.3 gm. three times daily. J l 3 W i ht 81 kil ( ill ) 11 Jan. 10, 1899: Disease well pronounced; thickening of pleura; cough very persistent; expectoration scanty; tubercle bacilli in sputum; temperature (axilla) 38.3 C. (101 F.) ; urine contains serum albumin, but no casts. Treatment: Milk diet. g y July 3: Weight 81 kilog., temperature (axilla), 11 a.m., normal. Treatment: Diet, G liters of milk daily; medicine as above. W i h 84 5 kil l July 25: Weight 84.5 kilog., temperature normal; asthenia and hectic condition greatly improved; very little cough and expectoration; discharged; improved. July 25: Weight 84.5 kilog., temperature normal; asthenia and hectic condition greatly improved; very little cough and expectoration; discharged; improved. R. Calcii eosolici. 10 Aquae . 70 Natrii benzoiei. 20 Syr, tolutani AquaB camphorae, aa. Case 5.—S. K., aged 32, female, American, presented, the duration of the disease five months. June 13: Weight 56 kilog., temperature (axilla), 3 p.m., 38.3 C. (101 F.) ( ill ) June 11, 1898: Weight 61.5 kilog., temperature (axilla), 37.9 C. (100.3 F.) ; apical disease very marked; cavity forma- tion on both sides; cough aggravated; expectoration profuse and purulent; great debility; nocturnal perspiration profuse and extremely exhausting; baeill us tuberculosis in sputum. Treatment: Overalimentation; salad oil, full cream pancrea- tin; calcium eosolicum, 0.3 gm. every three hours. J W i h ( 0 ) June 20: Weight 57 kilog., temperature (axilla), 2:30 p.m., 37.8 C. (100 F.) J ( ill ) ( ) June 28: Weight 57.75 kilog., temperature (axilla), 3 p.m., 37.2 C. (99 F.). Treatment: Mixed diet, milk in abundance; mist, calcii eosol. as above. Teaspoonful three times daily. ( ill ) Teaspoonful da y July 22: Weight 60.5 kilog., temperature (axilla), 4 p.m., 37.2 C. (99 F.). Treatment: Mixed diet, milk in abundance; as above; teaspoonful twice daily. S b W i h ( ill ) 3 30 ; , tin; calcium eosolicum, 0.3 gm. every three hours. ; , g y June 16: Weight 60.75 kilog., temperature (axilla), 38.3 C. (101 F.)23 W i h 60 kil ( ill ) C ; teaspoonful daily. September 16: Weight 64 kilog., temperature (axilla), 3:30 p.m., 37.2 C. (99 F.). Treatment as above; to continue medi- cine twice daily for two months. W i h kil 6 ( ) June 23: Weight 60 kilog., temperature (axilla), 37.5 C. (99.5 F.) W i h (100 ( ) June 29: Weight 00 kilog., temperature 37.8 C. (100 F.) Calcium eosolicum had to be withdrawn as the patient y March 8, 1899: Weight 66.5 kilog.; temperature (oxilla) 6 p.m., 37.2 C. (99 F.) ; no pain; very little cough and expectora- tion ; no tubercle bacilli in sputum; no nocturnal perspiration. C B d 37 l G h di June 29: Weight 00 kilog., temperature 37.8 C. (100 F.) Calcium eosolicum had to be withdrawn as the patient seemed to be idiosyncratic against it. C M A i h d i f Ca c u eoso cu patient seemed to be idiosyncratic against it. C 6 M A i h d i f h ; sputum; pe sp at o Case 3.—M. T. F b 14 W i ht 81 kil ( ill ) 37 1 da y September 8: The medicine was discontinued as the diarrhea seemed to increase under its administration, greatly weakening the patient. C G d daily. February 14: Weight 81 kilog., temperature (axilla), 37.1 C. (98.8 F.) ; dyspeptic symptoms subsiding. W i h kil ( ) ; dyspept c symptoms subs d g February 22: Weight 32.5 kilog., temperature (axilla), 37 C. (98.6 F.) M h W i ht 84 25 kil l patient. Case 8.—M. J. G., aged 33 female, married, American. The duration of the disease was 1% years. D b 23 1898 All 98.6 F.) March 1: Weight 84.25 kilog., temperature normal. W i h kil (98.6 F.) March 1: Weight 84.25 kilog., temperature normal. M h 10: Weight 85 kilog t t normal March 1: Weight 84.25 kilog., temperature normal. March 10: Weight 85 kilog., temperature normal. M 4 W i ht 88 5 kil t t l di i years. December 23, 1898: All symptoms characteristic of the dis- ease were present, including tubercle bacilli; bath lungs af- fected; temperature (axilla) 39^2 C. (102.5 F.) Treatment: Milk regimen; calcium eosolicum, 0.3 gm. three times a day with sodium bicarbonate. e g t og , temperature March 10: Weight 85 kilog., temperature normal. W i h kil l g g , p May 4: Weight 88.5 kilog., temperature normal; medication discontinued. Calcium eosolicum, plain, tended to purge the pa- tient. Together with syrup and glycerin, in the form of a cough mixture, indigestion followed its adminis- tration; combined with bicarbonate of soda and animal charcoal, it caused no unpleasant effects. December 28: Temperature (axilla), 37.5 C. (99.5 F.) December 30: Temperature (axilla) 37.7 C. (99.9 F.) Jan. 5, 1899: Temperature (axilla) 37 C. (98.6 F.) January 16: Weight 55.5 kilog., temperature (axilla) 37 g p ( ) R. Calcii eosolici. 5\| Aquae . 35 Ext. lobelias-infl. fl. 0 5 Aquae anisi, q. s., ad. 100\| " M Si A t f l th h p Case 4.—T. T., aged 34, male, American, with duration of disease unknown, probably nine months. J W i ht 77 kil ( ill ) 6 , p y June 5: Weight 77 kilog., temperature (axilla), 6 p.m., 37.8 C. Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 CHRONIC ULCERATIVE PHTHISIS. A teaspoonful every three hours. June 8: Weight 55.5 kilog., temperature (axilla), 3 p.m., 38.3 C. (101 F.) Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 B., aged 37, male, German, the disease was first recognized in December, 1897. J 1898 W i h 81 kil 38 5 C idiosyncratic against Case 6.—M. F., aged 44, male, Austrian; the duration of the disease 2% years, presented himself. J l 6 W i h kil t l 11 30 ecog ed , June 5, 1898: Weight 81 kilog., rectal temperature 38.5 C (101.3 F.) ; disease fully developed; tubercle bacilli in sputum; pallor exilmius; marked adynamia. Treatment: Forced feed- ing; 6 liters of milk per day; sun baths; lung gymnastics; calcium eosolicum, 0.3 gm. tablets, 3 per day. J l (101 y , p July 6, 1898: Weight 54 kilog., rectal temperature, 11:30 a.m., 38.5 C. (101.3 F.) ; disease far progressed; pulse very weak and frequent; edema of feet; dyspnea; pronounced hectic condition; asthenia; bacillus tuberculosis in sputum. Treat- ment: Mixed diet; olive-oil subcutaneously. strychnia intern- ally; calcium eosolicum, 0.3 gm. three times a day. , g , per day June 19: Weight 80 kilog., rectal temperature 38.3 C. (101 F.). Treatment: As above; calcium eosolicum, 0.3 gm. tab- lets, five per day. W i h kil l 37 1 , g , per y June 19: Weight 80 kilog., rectal temperature 38.3 C. (101 F.). Treatment: As above; calcium eosolicum, 0.3 gm. tab- lets, five per day. J l W i h 81 5 kil l 37 1 C ally; eosolicum, g day. July 7: Weight 54 kilog., rectal temperature, 4 p.m., 38.9 C. (102 F.) J l 8 W i ht 54 kil t l 38 3 C (101 , per day. July 1: Weight 81.5 kilog., rectal temperature 37.1 C. (98.8 F.) l , pe day. July 1: Weight 81.5 kilog., rectal temperature 37.1 C. (98.8 F.) W i h kil l 37 7 ( ) July 8: Weight 54 kilog., rectal temperature 38.3 C. (101 F.) J l 10 W i h 3 ( F.) July 14: Weight 83 kilog., rectal temperature 37.7 C. (99.9 F.) S b 6 W i h kil t l C ( F.) July 14: Weight 83 kilog., rectal temperature 37.7 C. (99.9 F.) S b 6 W i h kil l C ) July 10: Weight 53.75 kilog., rectal temperature 37.7 C. (99.9 F.) ; treatment discontinued; patient insisted on going to Colorado at once. ( ) September 6: Weight 85.75 kilog., rectal temperature 37.2 C. NEPHRITIS. My observations as to the therapeutic usefulness of calcium eosolicum in the various forms of nephritis are very limited. Only of late have I started to employ (his sulphocreosote salt in renal affections. As my cases are not long enough under the influence of this drug, I am not enabled to give any definite data at this early date. A study of the clinical reports tends to demonstrate at once that this sulphocreosote derivative can be ad- ministered, for long periods without effecting gastric or intestinal disturbances. It was well borne in nearly every instance, even in well-advanced cases, and also where dyspeptic and other symptoms of alimentary diffi- culty prevailed. y One of the noblest sentiments, the possession of which is characteristic of the civilized nations of the world, is that of a reverence for the dead which impels us to honor them after death, to erect monuments to perpetuate their memory, to beautify the places of interment, the cemeteries, and to regard as holy and sacred the partic- ular spot selected for the last resting-place of one's own dead, to be kept beautiful by constant care and by adorn- ment by flowers. This sentiment is most deeply rooted in the innermost recesses of the hearts of the best people of all cultured lands. The custom of acting on this sen- timent can never be changed until we learn that the presence of the urns containing the ashes of our most loved ones renders the chamber in which they are placed, in our own houses, or the columbaria in a cemetery, equally sacred places for communion, for reflection on the virtues of the dead, places to be beautified with flow- ers as the graves now are and having the advantage over them, especially when they are in one's own home, of constant access, regardless of weather or ill health. I hi b i i f y p The medicinal action of calcium eosolicum seems to depend on the first instance in its neutralizing qualities, which others may be tempted to call "antiseptic" or "germicidal." I doubt, however, the propriety of classifying calcium eosolicum as a germicide in the strict sense of the meaning; it does not destroy germs or micro-organisms, but it apparently effects a condi- tion of the system, especially of the fluids, which is averse to bacterial growth. Presented to the Section on State Medicine, at the Fiftieth Annual Meeting of the American Medical Association, held at Co- lumbus, Ohio, June 6-9, 1899. 75 M Si A t f l t h p ; g p The daily dose of calcium eosolicum had to be some- what reduced in the course of treatment on account of the loose bowels of the patient. Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 CREMATION. BY ROBERT MARSENA STONE, A.M., M.D. OMAHA, NEB. May 6: Weight 76 kilog., temperature normal; shallow and labored breathing; bronchial rales, somewhat crackling; very little cough and expectoration; no tubercle bacilli. 10 M N d 41 G coug expectoration; Case 10.—M. N. M., aged 41, male, German. The disease was first recognized about two years previously. J 21 1899 W i ht 60 kil ecog ed yea s previously. Jan. 21, 1899: Weight 60 kilog., temperature (axilla) nor- mal ; pain over scapula?; dyspeptic symptoms; infiltration of apices; cough persistent; expectoration mucopurulent, occa- sionally containing minute particles of blood; tubercle bacilli in sputum; drenching night sweats; emaciation; asthenia. Treatment: Milk diet, 200 c. c. every 1% hours; to facilitate digestion, sodium bicarbonate, nux vomica and hydrastis com- bined; calcium eosolicum, 0.3 gm. in capsules four times daily. J 25 W i ht 60 75 kil l thought passed. This particular study and address is due to a discus- sion of the general question of the disposition of the dead apart from any personal interest. It arises from a query presented: "Is it best that we bury our dead ?" "Is there good reason for a custom so long established, so well settled, so thoroughly accepted by the civilized world?" Is there any good reason why some other method of disposition should be adopted?" And still another question is put, and to the physician this is the hard- est of all: "Do the living suffer harm by reason of the almost universal custom of earth-burial?" Still another: "Are these few of whom we occasionally hear, who cremate their dead, fanatics, who without good reason have adopted that method, or are they wiser than the masses, have a deeper love for their fellows, an unwillingness to see them harmed, and have they as tender affections and sentiments as the rest of humanity?" h f hi h bined; , gm. capsules daily January 25: Weight 60.75 kilog., temperature normal. J 31 W i h 62 kil y g g , te pe atu e January 31: Weight 62 kilog., temperature normal; patient removed to Baltimore, whence he reported improvement in March. CREMATION. BY ROBERT MARSENA STONE, A.M., M.D. OMAHA, NEB. CREMATION. BY ROBERT MARSENA STONE, A.M., M.D. OMAHA, NEB. January 11: Cough looser; temperature (axilla) 37.5 C. (99.5 F.) ( ) January 12: Cough looser; expectoration increased; temper- ature (axilla) 37.6 C. (99.7 F.) l ( ) ( ) January 15: Cough and expectoration loose; temperature (axilla) 37.3 C. (99.1 F.) G l di i h A study of the customs of civilized nations of this and many past generations shows that we have blindly ac- cepted one custom and followed it for indefinite years, without much question. It is that of the disposition of our dead. Long-accepted custom has decreed that earth\x=req-\ burial is proper, and we have accepted it. Once in a great while we experience a rude shock as we learn of an event that took place in Cuba during the late war\p=m-\the necessary saturation of the bodies of the victims of yellow fever with oil, and their burning on vast pyres because the sanitation of a city demanded it\p=m-\butwe lapse into reacceptance of the old custom and continue to think earth-burial the only procedure for ordinary death. Few men care to think and reason about such matters until forced by the presence of grim death, and then grief holds sway and the time for dispassionate thought has passed. di ( ) (99 ) January 20: General condition much improved. F b W i h kil ( y p February 4: Weight 65.75 kilog., temperature (axilla) nor- mal. Treatment: Overalimentation with milk; calcium eoso- licum, 0.3 gm. four times daily. licum, gm. daily. February 7: Weight 66.25 kilog., temperature normal. F b W i h kil eb ua y e g t kilog., temperature February 11: Weight 67 kilog., temperature normal. F b 24 W i h 69 25 kil y g g , temperature February 24: Weight 69.25 kilog., temperature normal. M h 3 W i h 70 5 kil l y g g , p March 3: Weight 70.5 kilog., temperature normal; symptoms in general relieved; patient feels very comfortable. Treat- ment: As above; calcium eosolicum, 0.3 gm. twice daily. M W i h l h ll y g g , p March 3: Weight 70.5 kilog., temperature normal; symptoms in general relieved; patient feels very comfortable. Treat- ment: As above; calcium eosolicum, 0.3 gm. twice daily. M W i h l h ll ; , g daily. NEPHRITIS. If it is potent in the de- struction of the already formed toxic products of bacterial life, I can not say, but I maintain that the eosolate of calcium besides its apparent neutralizing power just mentioned, is a positve and harmless neu- tralizer of a number of toxic substances of non-bacterial origin. All o gAll the other medicinal qualities of calcium eosolicum seem to depend more or less on its action as a neutral izer and are therefore secondary to the latter. Its anti- pyretic, analgesic, emolient, stimulating, tonic and re- storative properties—sufficiently demonstrated by this report—seem to be nothing else but mere enunciations of its neutralizing force. g I am firmly convinced that this substitution of senti- ment, from the grave to the urn, this recognition that the urn with its ashes renders its home sacred, will sooner or later be accepted and will result in the transfer of our holy regard from the grave to the urn, also, beau- tified with constantly changed flowers, and result in the adoption of cremation by thinking, cultured people of all communities. Vaginal Irrigation for Leucorrhea. Vaginal Irrigation for Leucorrhea. R. Potassii chloratis.12 parts Vini opii.10 parts Aquae picis.300 parts M. Sig. Add 3ii-iii to a quart of warm water.—Lutaud. Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015 Downloaded From: http://jama.jamanetwork.com/ by a New York University User on 06/20/2015
https://openalex.org/W4237512502	http://www.scielo.cl/pdf/ejb/v5n3/a09.pdf	English	null	Spirulina platensis growth estimation by pH determination at different cultivations conditions	Electronic Journal of Biotechnology	2,002	cc-by	2,485	Electronic Journal of Biotechnology ISSN: 0717-3458 © 2002 by Universidad Católica de Valparaíso -- Chile Electronic Journal of Biotechnology ISSN: 0717-3458 © 2002 by Universidad Católica de Valparaíso -- Chile Vol.5 No.3, Issue of December 15, 2002 Received April 16, 2002 / Accepted November 29, 2002 SHORT COMMUNICATION Vol.5 No.3, Issue of December 15, 2002 Received April 16, 2002 / Accepted November 29, 2002 Spirulina platensis is a filamentous cyanobacterium that is Spirulina platensis growth estimation by pH determination at different cultivations conditions Lúcia Helena Pelizer Departamento Tecnología Bioquímico-Farmaceutica Universidade de Franca Av. Dr. Armando Salles Oliveira 201 Parque Universitario, CEP 14404-600 Franca/ Sao Paulo, Brasil Tel: 55 16 37118888 Fax: 55 16 37118889 E-mail: lucia.pelizer@uol.com.br J.C.M. Carvalho Departamento Tecnologia Bioquímico-Farmacêutica - FCF Universidade de São Paulo Sunao Sato Faculdade de Ciências Farmacêuticas Universidade de Sao Paulo Av. Lineu Prestes 580, Bl 16 Campus Universitario CEP 05508900 Sao Paulo, Brasil Tel: 55 11 38183691 E-mail: fcf@usp.br Iracema de Oliveira Moraes* ∗Laboratório de Bioprocessos Universidade Guarulhos Praça Thereza Cristina 01, Centro Guarulhos CEP 07023070 Sao Paulo, Brasil Tel: 55 19 32582001 Fax: 55 19 32582001 E-mail: iomoraes@hotmail.com Finanacial support: FAPESP. Iracema de Oliveira Moraes* ∗Laboratório de Bioprocessos Universidade Guarulhos Praça Thereza Cristina 01, Centro Guarulhos CEP 07023070 Sao Paulo, Brasil Tel: 55 19 32582001 Fax: 55 19 32582001 E-mail: iomoraes@hotmail.com Finanacial support: FAPESP. Finanacial support: FAPESP. Finanacial support: FAPESP. Keywords: biomass estimation, methodology, pH, protein content, Spirulina platensis. Spirulina platensis is a cyanobacterium that has a high protein content and therefore, a high nutritional value. It can be cultivated either in a liquid or in a solid culture. When cultivated in aqueous culture the cell growth can be determined by following the optical density. On the other hand, when produced by solid cultivation the growth can be determined only indirectly, such as, through determination of the protein content of the fermenting solids. In this work the possibility of estimating cell growth by pH determination was verified. From the results it was concluded that pH and protein production (solid or surface culture) or cell content (liquid culture) correlate well, therefore pH determination seems to be a good method to determine cell growth. biotechnologically important due its high nutritional value. The nutritional value derives from its high protein content (about 70%) and its type of lipids (γ-linolenic acid) (Ciferri and Tiboni, 1985; Henrikson, 1989). This microorganism also finds application in environmental technology (Pulz and Scheibenbogen, 1998). Generally it is produced in open ponds in liquid culture (Henrikson, 1989; Pulz and Scheibenbogen, 1998), but recently, its production in solid-state cultivation systems has been studied (Senecal et al. 1992; Cozza et al. 1999; Pelizer et al. 1999; Pelizer et al. 2000; Pelizer et al. 2002). During production in liquid cultivation systems cell growth is followed by measuring the optical density of the culture medium. In solid cultivation estimation of cell growth is made difficult by the problems of separating cells from the cultivated medium. As a result of these difficulties, biomass Spirulina platensis is a filamentous cyanobacterium that is ∗ Corresponding author Results and Discussion The protein produced during the various cultivation processes was correlated with the evolution of the pH and the equations for each experiment are shown in Table 1, Table 2, Table 3 and Table 4. For each type of culture medium one general equation was done plotting the results of all experiments together (Figure 1, Figure 2, Figure 3 and Figure 4). Reasonably high correlation coefficients were obtained for all experiments. The aim of this work was to verify the possibility of using pH determination to obtain rapid estimates of cell growth. Microorganism and inoculum For experiments done in solid or surface culture the equations obtained are similar and, except for experiments 06 and 07, they can be grouped by experiments done using the same luminance (luminosity). For experiments using sugar cane bagasse that had a moisture content of 95.8% the equations were similar but in this case experiments using other luminosities were not done. The microorganism used was Spirulina platensis. The inoculum was obtained by liquid cultivation using a mineral medium (Paoletti et al. 1975) in 500 mL Erlenmeyer flasks for seven days. The cultivation conditions were: agitation, 160 rpm, temperature, 30ºC and luminosity (luminance), 6.0 Klux. The inoculum concentration was determined by spectrofotometry using a standard curve. Experiments done with liquid medium presented almost the same equations and they were also carried out using only one luminosity. Concluding Remarks Liquid culture was done using mineral solution in 5 L tanks. In this case the only variable was inoculum concentration, as shown in Table 4. These results show that pH determination can be used to predict Spirulina platensis growth in the types of processes studied in the present work. This paper is available on line at http://www.ejbiotechnology.info/content/vol5/issue3/full/8 This paper is available on line at http://www.ejbiotechnology.info/content/vol5/issue3/full/8 Pelizer, L. et al. pH meter (SENTRON 1001). levels in solid-state fermentation systems are typically determined indirectly through the measurements of cell constituents (Hesseltine, 1972; Huang et al. 1978; Abdullahet al. 1985; Gutiérrez-Rojas et al. 1995). levels in solid-state fermentation systems are typically determined indirectly through the measurements of cell constituents (Hesseltine, 1972; Huang et al. 1978; Abdullahet al. 1985; Gutiérrez-Rojas et al. 1995). levels in solid-state fermentation systems are typically determined indirectly through the measurements of cell constituents (Hesseltine, 1972; Huang et al. 1978; Abdullahet al. 1985; Gutiérrez-Rojas et al. 1995). In solid media cell growth was determined by optical density. For surface and solid cultivation growth it was followed indirectly by determination of the protein content of the cultivated material by the Kjeldahl method (Aoac,1984). Desgranges et al. 1991a and Desgranges et al. 1991b, compared four methods for biomass estimation during growth of Beauveria bassiana in solid-state fermentation systems: they determined glucosamine, ergosterol, total sugar levels and determined the CO2 evolution rate. There was a good correlation between the cell growth and the indirect methods for biomass determination. Three types of experiments were done. These results show that pH determination can be used as an indicator of microbial growth. If an standard experiment is done to have a calibration curve for each process condition, especially luminosity (luminance), the cell growth can be rapidly estimated by culture pH. Solid culture was done in with 200 mL Erlenmeyer flasks containing 50 g of a solid medium containing sugar cane bagasse, mineral solution (Paoletti et al. 1975) and 8 g L-1 nutrient agar. Table 1 shows how the luminosity (provided with fluorescent lamps), the inoculum concentration and the moisture content were varied in these experiments. For liquid culture growth estimation can also be done very rapidly by spectrophotometer. In the case of solid cultivation it is very difficult to separate the cells from the cultivation medium, meaning that the optical density method cannot be used. Cell growth can be determined indirectly by the determination of cells constituents such as protein content as done in the present work. However these methods are mostly time-consuming and do not allow on- line monitoring the process. Tray fermentations were also done with 500 g of this culture medium. Table 2 shows how the luminosity and inoculum concentration were varied. The moisture content for all experiments was 95.8%. Surface culture was done in 200 mL Erlenmeyer flasks with 50 g of a medium consisting of mineral solution with 8 gL-1 nutrient agar. Table 3 shows how the luminosity and inoculum concentration were varied. Culture conditions Three types of experiments were done. References PELIZER, L.H.; CARVALHO, J.C.M.; SATO, S. and MORAES, I.O. Produción de Spirulina platensis utilizándose resíduo industrial sólido como soporte para el crecimiento. In: Memorias del VIII Congreso Nacional de Biotecnología y Bioingenería y IV Congreso Latinoamericano de Biotecnología y Bioingenería (25th – 29th July, 1999, Huatulco, México). 1999. p.277. ABDULLAH, A.L.; TENGERDY, R.P. and MURTHY, V.G. Optimization of solid substrate fermentation of wheat straw. Biotechnology and Bioengineering, 1985, vol. 27, p. 29-27. AOAC, ASSOCIATION OF OFFICIAL ANALYTICAL CHEMISTS. Official methods of analysis. 14 ed. Arlington, 1984. 500 p. ISBN 0-935584-24-2. PELIZER, L.H. Estudos de obtenção de Spirulina platensis, por fermentação em estado Sólido. Tese de Doutorado, USP/FCF/SP Brasil - Tecnologia das Fermentações. 2000, 143 p. CIFERRI, O. and TIBONI, O. The biochemistry and industrial potential of Spirulina. Annual Review of Microbiology, 1985, vol. 89, p. 503-526. PELIZER, L.H.; DANESI, E.D.G.; RANGEL, C.O.; SASSANO, C.E.N.; CARVALHO, J.C.M.; SATO, S. and MORAES, I.O. Influence of inoculum age and concentration in Spirulina platensis cultivation. Journal of Food Engineering. Elsevier Science Ltd. Oxford. England. In press (paper 01/1397). 2002. COZZA, K.L.; COSTA, J.A.V.; GONZALEZ, T.A.; OLIVEIRA, L. and BAISCH, A.L.M. Cultivo de Spirulina platensis em meio semi-sólido utilizando farinha de arroz. In: III Simpósio de Ciência e Tecnologia de Alimentos (19th – 25th September, 1999, Campinas, Sao Paulo, Brasil). Livro de programa e resumos, 1999, p. 97. PULZ, O. and SCHEIBENBOGEN, K. Photobioreactors: design and performance with respect to light energy imput. Advances in Biochemical Engineering / Biotechnology, 1998,vol. 38, p.123-152. DESGRANGES, C.; VERGOIGNAN, C.; GEORGES, M. and DURAND, A. Biomass estimation in solid state fermentation I. Manual biochemical methods. Applied Microbiology and Biotechnology, 1991a, vol. 35, p. 200- 205. SENECAL, K.J.; MANDELS, M. and KAPLAN, D.L. Biological conversion of inedible biomass to food. Biotechnology and Nutrition. Summary, 1992. DESGRANGES, C.; VERGOIGNAN, C.; GEORGES, M. and DURAND, A. Biomass estimation in solid state fermentation II. On-line measurements. Applied Microbiology and Biotechnology, 1991b, vol. 35, p. 206- 209. GUTIÉRREZ-ROJAS, M.; CÓRDOVA, J.; AURIA, R.; REVAH, S. and FAVELA-TORRES, E. Citric acid and polyols production by Aspergillus niger at high glucose concentration in solid state fermentation on inert support. Biotechnology Letters, 1995, vol. 17, no. 2, p. 219-224. HENRIKSON, R. Earth food Spirulina. California/USA. Ronore Enterprises, 1989. 180 p. HESSELTINE, C.W. Solid state fermentation. Biotechnology and Bioengineering, 1972, vol. 14, p. 517- 532. HUANG, S.Y.; WANG, H.; WEI, C.; MALANEY, G.W. and TANNER, R.D. Analytical procedures pH was determined directly in solid and liquid media by a To Dr David Mitchell of the Universidade Federal do 252 Spirulina platensis growth estimation by pH determination at different cultivations conditions Spirulina platensis growth estimation by pH determination at different cultivations conditions Paraná, Brasil, for helping us with the English expression in the manuscript. PAOLETTI, C.; PUSHPARAJ, B. and TOMASELLI, L. Ricerche sulla nutrizione minerale di Spirulina platensis. In: Atti Cong. Naz. Soc. Ital. Microbiol., 17, Padova, 1975. References Kinect response of the Koji solid state fermentation process. In: WISEMAN, A., ed. Topics in enzyme and fermentation biotechnology. Chichester, Ellis Horwood, 1978, vol. 10, chap. 4, p. 89-108. ISBN 0-85132- 052-8. 253 Pelizer, L. et al. APPENDIX Tables Table 1. Results of experiments carried out in Erlenmeyer flasks using sugar cane bagasse and mineral media with nutrient agar as the culture medium. APPENDIX Tables APPENDIX Tables Table 1. Results of experiments carried out in Erlenmeyer flasks using sugar cane bagasse and mineral media with nutrient agar as the culture medium. with nutrient agar as the culture medium. Experiments Luminosity Klux Inoculum g cells. Kg wet cult. media -1 Moisture % wet basis Equation Correlation coefficient 1 3.8 0.05 94.8 y = 1.6x - 15.6 0.979 2 3.8 0.25 94.8 y = 1.7x - 16.1 0.9607 3 6.0 0.05 94.8 y = 1.1x - 10.2 0.8518 4 6.0 0.15 94.8 y = 1.2x -11.1 0.99 5 6.0 0.25 94.8 y = 1.0x - 8.6 0.9608 6 8.2 0.05 94.8 y = 1.0x -9.5 0.99 7 8.2 0.25 94.8 y = 1.7x - 15.8 0.9907 8 6.0 0.05 95.8 y = 1.5x - 14.5 0.9788 9 6.0 0.15 95.8 y = 1.5x - 14.4 0.9632 10 6.0 0.25 95.8 y = 1.3x - 12.2 0.8618 General y = 1.3x - 12.4 0.8309 Table 2. Results of experiments carried out in tray type bioreactor using sugar cane bagasse and mineral media with nutrient agar as culture medium. Table 2. Results of experiments carried out in tray type bioreactor using sugar cane bagasse and mineral media with nutrient agar as culture medium. Experiments Luminosity Klux Inoculum g cells. Kg wet culture media -1 Equation Correlation coefficient 11 6.0 0.05 y = 5.6x - 53.9 0.8383 12 6.0 0.15 y = 4.1x - 39.2 0.9898 13 6.0 0.25 y = 4.9x - 46.4 0.9543 14 8.2 0.05 y = 2.4x - 22.0 0.7375 15 8.2 0.15 y = 2.8x - 26.7 0.9611 16 8.2 0.25 y = 2.8x - 26.0 0.9117 General y = 3.0x - 28.4 0.7263 254 Spirulina platensis growth estimation by pH determination at different cultivations conditions Table 3. Results of surface culture experiments carried out in Erlenmeyer flasks using mineral media with agar nutrient as the culture medium. Experiments Iuminosity Klux Inoculum g cells. Kg wet cell culture -1 Equation Correlation coefficient 17 3.8 0.05 y = 1.6x -14.0 0.6599 18 3.8 0.25 y = 1.9x - 16.2 0.6454 19 6.0 0.01 y = 1.0x - 8.5 0.6181 20 6.0 0.15 y = 1.4x - 11.9 0.8299 21 6.0 0.29 y = 1.3x - 11.6 0.7789 22 8.2 0.05 y = 2.8x - 26.6 0.9466 23 8.2 0.25 y = 2.4x - 22.4 0.9286 General y = 1.5x - 12.7 0.6024 Table 4. APPENDIX Tables Results of experiments carried out in mini-tanks using mineral solution as the culture medium. Experiments Inoculum g cells. Kg wet cell culture –1 Equation Correlation coefficient 24 0.05 y´ = 0.8x - 7.2 0.8868 25 0,10 y´ = 0.9x - 8.2 0.9241 26 0.15 y´ = 0.8x - 7.6 0.8995 27 0,20 y´ = 0.8x - 7.3 0.8526 General y´ = 0.8x - 6.9 0.7624 ults of experiments carried out in mini-tanks using mineral solution as the culture medium. Table 4. Results of experiments carried out in mini-tanks using mineral solution as the c 255 Pelizer, L. et al. Pelizer, L. et al. Figures Figure 1. General curve for experiments carried out in Erlenmeyer flasks using sugar cane bagasse and mineral media with nutrient agar as the culture medium. Figure 1. General curve for experiments carried out in Erlenmeyer flasks using sugar cane bagasse and mineral media with nutrient agar as the culture medium. Figue 2. General curve for experiments carried out in tray type bioreactor using sugar cane bagasse and mineral media with nutrient agar as the culture medium. Figue 2. General curve for experiments carried out in tray type bioreactor using sugar cane bagasse and mineral media with nutrient agar as the culture medium. 256 Spirulina platensis growth estimation by pH determination at different cultivations conditions Figue 3. General curve for surface culture experiments carried out in Erlenmeyer flasks using mineral media with agar nutrient as the culture medium. Figue 3. General curve for surface culture experiments carried out in Erlenmeyer flasks using mineral media with agar nutrient as the culture medium. Figue 4. General curve for experiments carried out in mini-tanks using mineral solution as the culture medium. Figue 4. General curve for experiments carried out in mini-tanks using mineral solution as the culture medium. 257
https://openalex.org/W2781852174	http://ejurnal.its.ac.id/index.php/teknik/article/download/23635/4702	Indonesian	null	Desain CNG Carrier dari Gresik ke Lombok untuk Mendukung Program Pembangkit Listrik 35000 MW	Jurnal Teknik/Jurnal Teknik ITS	2,017	cc-by-sa	3,335	G88 G88 JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) Kata Kunci—CNG, CNG Carrier, Gresik – Lombok. Kata Kunci—CNG, CNG Carrier, Gresik – Lombok. Made Dwi Ary Arjana Tusan dan Hesty Anita Kurniawati Departemen Teknik Perkapalan, Fakultas Teknologi Kelautan, Institut Teknologi Sepuluh Nopember (ITS) e-mail: tita@na.its.ac.id Made Dwi Ary Arjana Tusan dan Hesty Anita Kurniawati Departemen Teknik Perkapalan, Fakultas Teknologi Kelautan, Institut Teknologi Sepuluh Nopember (ITS) e-mail: tita@na.its.ac.id Abstrak—PLTGU Lombok Peaker merupakan pembangkit listrik tenaga gas dan uap yang menggunakan Compresed Natural Gas (CNG) sebagai bahan bakar. Di daerah Lombok tidak ada pasokan CNG untuk mendukung kebutuhan PLTGU tersebut, namun dengan adanya pembangunan CNG Plant di Gresik akan membantu dalam pasokan gas ke Lombok. Penelitian ini bermaksud memberikan solusi untuk menciptakan sarana distribusi gas alam seperti CNG sebagai bahan bakar pembangkit listrik khususnya di Lombok. Payload dari CNG carrier ini merupakan kebutuhan CNG yang digunakan sebagai bahan bakar PLTGU Lombok Peaker beserta tabung dan kontainernya. Ukuran utama kapal ditentukan berdasarkan penempatan tabung dan kontainer pada kapal. Setelah itu dilakukan perhitungan teknis berupa perhitungan berat, freeboard, trim dan stabilitas. Ukuran utama yang didapatkan adalah Lpp = 81.8 m; B = 14.7 m; H = 8m; T = 5m. Tinggi freeboard minimum sebesar 1074 mm, besarnya tonase kotor kapal adalah 2250 GT, dan kondisi stabilitas CNG carrier memenuhi kriteria Intact Stability (IS) Code Reg. III/3.1. Biaya pembangunan sebesar Rp51,298,798,739 dan biaya operasional sebesar Rp 26,888,561,985. tempat penyimpanan CNG lebih besar untuk jumlah massa yang sama dengan LNG. Murahnya produksi CNG ini membuat pemasarannya lebih ekonomis untuk lokasi-lokasi yang dekat dengan sumber gas alam. y g g g Dari kondisi tersebut pemerintah akan memanfaatkan CNG sebagai bahan bakar untuk pembangkit – pembangkit listrik. Pembangkit listrik yang didirikan tidak selalu dekat dengan daerah penghasil gas alam sehingga kondisi seperti ini menjadi kendala yang cukup serius dalam hal pengangkutannya. Salah satu pembangunan pembangkit listrik yaitu PLTGU Peaker 150 MW yang akan dibangun di daerah Lombok. Akan tetapi pasokan CNG pada daerah ini belum ada, sehingga harus dilakukan suplai gas dari daerah lain. Berhubungan dengan dibangunnya fasilitas CNG atau CNG Plant Gresik di lokasi Pembangkit Listrik Tenaga Gas Uap (PLTGU) Gresik, maka pemerintah akan mengirim pasokan CNG dari Gresik ke Lombok dan nantinya akan ditransfer ke daerah PLTGU yang di bangun. Dari latar belakang tersebut, maka akan dibuat desain CNG Carrier dari Gresik ke Lombok untuk mendukung program Pembangkit Listrik 35000 MW. A. Pembuatan Tabung CNG S EMAKIN berkembangnya teknologi di kehidupan ini, tentunya membuat kebutuhan listrik menjadi bertambah setiap tahunnya. Untuk menghindari krisis kelistrikan yang terjadi di Indonesia, maka pemerintah membuat program pembangkit listrik 35000 MW yang ditetapkan pada Peraturan Presiden Republik Indonesia Nomor 4 Tahun 2016 Tentang Percepatan Pembangunan Infrastruktur Ketenagalistrikan. Pada program ini pemerintah akan membangun banyak pembangkit listrik di berbagai daerah di Indonesia. S CNG atau gas alam terkompresi merupakan alternatif bahan bakar selain bensin atau solar. Bahan bakar ini dianggap lebih bersih bila dibandingkan bahan bakar minyak, karena emisi gas buangnya yang ramah lingkungan. CNG dibuat dengan melakukan kompresi metana (CH4) yang diekstrak dari gas alam. CNG disimpan dan didistribusikan dalam bejana tekan, biasanya berbentuk silinder. Gas CNG akan dimampatkan dengan tekanan antara 200 - 250 bar (2900 – 3600 psi)[1]. Gambar 1. Tabung CNG Tabung penyimpanan CNG terbuat dari baja atau aluminium yang mampu menahan tekanan hingga lebih dari 50 bar. P b t dil k k d dit l i d i Dengan menipisnya cadangan minyak bumi dan tingginya harga minyak bumi membuat banyak orang yang beralih ke sumber energi lain. Dalam memenuhi program pembangkit listrik 35000 MW, pemerintah memanfaatkan gas alam sebagai bahan bakar untuk pembangkit listrik. Gas alam yang digunakan yaitu Liquefied Natural Gas (LNG) atau Compressed Natural Gas (CNG). Antara LNG dan CNG terdapat perbedaan bentuk yang mendasar yaitu LNG merupakan gas alam yang berbentuk cair sedangkan CNG adalah gas alam yang terkompresi. Secara ekonomis produksi CNG lebih murah dibandingkan LNG yang membutuhkan pendinginan dan tangki kriogenik yang mahal. Akan tetapi Gambar 1. Tabung CNG Gambar 1. Tabung CNG Tabung penyimpanan CNG terbuat dari baja atau aluminium yang mampu menahan tekanan hingga lebih dari 50 bar. Pembuatannya dilakukan dengan cara ditempa, mulai dari JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) G89 PT Pelabuhan Indonesia III (PERSERO) cabang Lembar, menjadi salah satu tempat pengembangan usaha bagi PT BJTI PORT Surabaya di pulau Lombok sejak dua tahun yang lalu. Pelabuhan Lembar yang terletak di Kabupaten Lombok Barat – NTB, memiliki geografis: 080-43’-50,2” LS /1160-04’- 24,20” BT. Sebagai pelabuhan kelas III, Lembar memiliki luas perairan 481 hektar dengan luas daratan 156,50 hektar. Untuk alur pelayarannya sendiri memiliki panjang 1.490 meter dan lebar 60 meter. Rintangan bawah airnya meliputi lumpur, air serta batu karang. B. Transportasi CNG Metode pengiriman CNG menggunakan truck mounted CNG atau CNG trailer telah banyak diaplikasikan secara komersial di beberapa negara maju, terutama di Amerika Serikat dan Kanada. Sedangkan jalur laut (CNG Marine) menggunakan kapal dengan desain khusus, sayangnya hingga saat ini aplikasi komersial marine transportation CNG belum beroperasi secara komersial karena risiko yang terkait dengan teknologi baru[1]. Rute pelayaran kapal CNG dari Gresik ke Lombok dapat dilihat pada Gambar 3. Pelayaran tersebut melewati sebelah utara Pulau Madura untuk menghindari jembatan Suramadu. Jarak pelayaran sekitar 272 nautical miles atau 503744 m. Gambar 3. Rute Pelayaran Gresik - Lombok g Salah satu teknologi pengangkutan CNG di perairan dangkal seperti perairan Indonesia adalah yang dikembangkan oleh Enersea transport yaitu votrans (volume optimized transport and storage) seperti pada Gambar 2. Fitur utama dari operator VOTRANS CNG EnerSea adalah bahwa hal itu dapat membawa jumlah yang sama dari gas seperti sistem lain dengan sedikit pendinginan gas dengan tekanan yang relatif rendah daripada yang lain. Lebih khusus, sistem lain berisi gas pada suhu kamar dengan tekanan dari 250 sampai 300 bar, sedangkan metode VOTRANS menyimpan jumlah yang sama dari gas pada minus 30 derajat dan tekanan 120 bar, yang kurang dari setengah dari sistem lain. Kepadatan relatif dari gas pada kondisi yang kira-kira dua kali lipat dari sistem lain. Sebagai gas dapat diangkut sekitar setengah tekanan, berat total tangki juga dapat dikurangi dengan sekitar setengah dari sistem lain [2]. Gambar 3. Rute Pelayaran Gresik - Lombok A. Pembuatan Tabung CNG Khusus untuk kedalaman (LWS) memiliki alur pelayaran 19 m (rata-rata), dengan kolam pelabuhan 6,5 m (rata-rata), sementara di depan dermaga 6 m [3]. bentuk bongkahan baja utuh hingga terbentuk sebuah tangki tanpa sambungan. Contoh tabung yang digunakan dapat dilihat pada Gambar 1. Tabung yang dibuat harus lolos dari berbagai macam pengujian yang sesuai dengan code dan standar, mengingat isinya adalah gas dengan tekanan yang besar. Tiap negara memiliki code dan standar yang berbeda, misalnya di Amerika mengacu pada standar ASML, di Inggris mengacu pada standar BS dan sebagainya. A. Diagram Alir Diagram alir pengerjaan dapat dilihat pada Gambar 4. Gambar 2. Pengangkutan CNG Menggunakan VOTRANS CNG Gambar 2. Pengangkutan CNG Menggunakan VOTRANS CNG C. Tinjauan Wilayah Pelabuhan Gresik merupakan daerah industri yang memiliki beberapa segmen di setiap wilayah, Gresik memfokuskan pelabuhannya dalam 3 kategori yang sesuai dengan rute kegiatan dari masing-masing kepentingan. Pertama Pelabuhan Utama Gresik, adalah pelabuhan yang utama untuk arus barang dan penumpang, baik yang masuk ataupun yang keluar. Kedua, Pelabuhan Nelayan Gresik, merupakan pelabuhan utama khusus bagi nelayan yang akan pergi melaut. Sedangkan, yang ketiga Pelabuhan Gresik merupakan pelabuhan yang hanya khusus digunakan untuk kepentingan industri tertentu seperti Petrokimia, Plywood dan Semen. JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) G90 Gambar 4. Diagram Alir B. Penentuan Ukuran Utama B. Penentuan Ukuran Utama Ukuran utama ditentukan berdasarkan penempatan muatan di ruang muat kapal, sehingga diperoleh layout awal seperti pada Gambar 5. ngan ukuran utama kapal awal sebagai berikut: Dengan ukuran utama kapal awal sebagai berikut: LPP 81 8 Dengan ukuran utama kapal awal sebagai berikut: LPP 81 8 Dengan ukuran utama kapal awal sebagai berikut: LPP : 81.8 m LWL : 85.072 m B : 14.7 m H : 8 m T : 5 m Vs : 12 knot Gambar 4. Diagram Alir A. Penentuan Payload Payload diambil dari Keputusan Menteri Energi dan Sumber Daya Mineral Republik Indonesia Nomor 5899 K/20/MEM/2016 tentang Pengesahan Rencana Usaha Penyediaan Tenaga Listrik PT Perusahaan Listrik Negara (Persero) Tahun 2016 s.d 2025. Diperoleh muatan CNG yang dibutuhkan dalam sehari sebesar 5.4 bbtu. CNG carrier ini direncanakan beroperasi dalam 3 hari sekali dengan muatan yang dibawa sebanyak 4 hari kebutuhan gas PLTGU Lombok Peaker, sehingga muatan total sebesar 5.4 x 4 = 21.6 bbtu. Dari Tabel 1. diperoleh total muatan 2872.8 ton dan dilakukan pembulatan sehingga payload CNG carrier ini sebesar 2873 ton. IV. ANALISIS TEKNIS Gambar 5. Layout Awal D. Perhitungan Hambatan dan Propulsi D. Perhitungan Hambatan dan Propulsi Selanjutnya dilakukan perhitungan hambatan (Holtrop) dan propulsi kapal. p p p RT = 108.97 kN EHP = 672.67 kW DHP = 1094.75 kW BHP = 1297.60 kW p p p RT = 108.97 kN EHP = 672.67 kW DHP = 1094.75 kW BHP = 1297.60 kW RT = 108.97 kN EHP = 672.67 kW DHP = 1094.75 kW BHP = 1297.60 kW Setelah itu dilakukan pemilihan mesin induk kapal, mesin induk yang digunakan adalah sebagai berikut: Tipe : 6L20 Wartsila Daya [ kW ] = 1440 kW RPM = 1000 rpm Panjang = 3973 mm Lebar = 1756 mm Tinggi = 2424 mm Berat = 11 ton E. Perhitungan LWT dan DWT C. Perhitungan Koefisien Setelah didapatkan ukuran utama kapal awal, selanjutnya dilakukan perhitungan koefisien-koefisien kapal. Tabel 1. Perhitungan Payload Kebutuhan Gas PLTGU Lombok 1 scf = 0.02831685 m3 21.6 mmscf = 21600000 scf = 611643.96 m3 Tabung yg digunakan Type tabung = 1 Panjang = 10975 mm Diameter = 559 mm Tebal = 17.3 mm Tekanan = 250 bar Kapasistas CNG = 700 m3 Dimensi Kontainer Panjang = 12192 mm Lebar = 2438 mm Tinggi = 1435 mm Kapasitas tabung = 8 tabung Berat 1 kontainer dan 8 tabung = 25650 kg Jumlah tabung yang digunakan = total muatan/kapasitas tabung = 873.78 = 874 tabung Jumlah kontainer yang digunakan = jumlah tabung/kapasitas kontainer = 112 kontainer Berat total muatan = jumlah kontainer x berat total kontainer = 2872800 kg = 2872.8 ton Tabel 1. Perhitungan Payload Cb : 0.730 V : 4567.497 m3 Cm : 0.988 Δ : 4681.684 ton Cwp : 0.836 Cp : 0.739 E. Perhitungan LWT dan DWT Pada perhitungan LWT terdapat 3 komponen yaitu berat baja, berat peralatan dan berat permesinan. Sedangkan DWT terdapat komponen crew, consumable dan payload. Hasil perhitungan dapat dilihat pada Tabel 2. F. Perhitungan Titik Berat Kapal Titik berat kapal diperoleh dari titik berat LWT dan DWT. Titik berat LWT Tititk berat DWT KG = 7.296 m KG = 7.015 m Titik berat kapal diperoleh dari titik berat LWT dan DWT. Titik berat LWT Tititk berat DWT KG = 7.296 m KG = 7.015 m JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) G91 LCG (FP) = 39.4 m LCG (FP) = 41.045 m Titik berat total kapal KG = 7.2 m, LCG dari FP = 39.97 m. 4. Loadcase 4 merupakan kasus di mana muatan 85% dan bahan bakar 100%. 5. Loadcase 5 merupakan kasus di mana muatan 85% dan bahan bakar 50%. G. Perhitungan Tonase Kapal Tonase kapal merupakan perhitungan volume semua ruangan yang terletak di bawah geladak kapal ditambah dengan volume ruangan tertutup yang terletak di atas geladak ditambah dengan isi ruangan beserta semua ruangan tertutup yang terletak di geladak paling atas (superstructure). 6. Loadcase 6 merupakan kasus di mana muatan 85% dan bahan bakar 10%. 7. Loadcase 7 merupakan kasus di mana muatan 0% dan bahan bakar 100%. 8. Loadcase 8 merupakan kasus di mana muatan 0% dan bahan bakar 50%. GT = 2249.99 ton NT = 1405.506 ton GT = 2249.99 ton NT = 1405.506 ton 9. Loadcase 9 merupakan kasus di mana muatan 0% dan bahan bakar 10%. 9. Loadcase 9 merupakan kasus di mana muatan 0% dan bahan bakar 10%. Tabel 2. Hasil Perhitungan LWT dan DWT LWT Berat Baja = 1130.771 ton Berat Peralatan = 303.643 ton Berat Permesinan = 78.734 ton Total LWT = 1513.149 ton DWT Payload = 2873 ton Berat Bahan Bakar = 3.92 ton Berat Minyak Pelumas = 0.11 ton Berat Air Tawar = 5.25 ton Berat Provision = 0.20 ton Berat Orang dan Bawaan = 1.58 ton Total DWT = 2884.06 ton LWT dan DWT = 4397.206 ton Displasemen (Design) = 4681.684 ton Margin = 6.08 % H. Perhitungan Trim Tabel 2. I. Perhitungan Freeboard Perhitungan freeboard ini berdasarkan aturan yang terdapat pada International Convention on Load Lines 1966 and Protocol of 1988 (ICLL 1966). Perhitungan ini disesuaikan dengan tipe kapal, dimana untuk CNG carrier ini merupakan kapal tipe B. Nilai actual freeboard dapat dilihat pada Tabel 4. Tabel 4. Tabel 4. Rekapitulasi Freeboard Rekapitulasi Freeboard No Item Hasil Satuan 1 Tipe Kapal Tipe B 2 Freeboard Standard 933 mm Koreksi-Koreksi 3 Koreksi depth 429.216 mm Koreksi Cb 0 mm Koreksi Bangunan Atas -290 mm Koreksi Sheer 0 mm Freeboard Total 1074 mm Actual Freeboard 3000 mm J. Perhitungan Stabilitas Berikut loadcase yang harus dihitung kondisi trimnya: H. Perhitungan Trim Perhitungan trim dilakukan untuk mengetahui apakah kapal mengalami trim buritan atau trim haluan, dimana kondisi kapal akan berubah secara otomatis akibat perubahan kondisi pemuatan. Pemeriksaan trim ini mengacu pada SOLAS Reg. II/7, dimana kondisi trim maksimum yang diperbolehkan adalah 0.5% Lpp. Untuk hasil perhitungan trim dapat dilihat pada Tabel 3. K. Pembuatan Lines Plan Setelah semua perhitungan selesai dilakukan, selanjutnya adalah membuat Rencana Garis (Lines Plan). Lines Plan merupakan gambar pandangan atau gambar proyeksi badan kapal yang dilihat dari tampak depan (body plan), tampak samping (sheer plan) dan tampak atas (half breadth plan). Setiap proyeksi menggambarkan badan kapal yang terpotong- potong pada arah tertentu dengan jarak yang secara umum konstan. Lines Plan dapat dilihat seperti pada Gambar 6. Gambar Rencana Umum dibuat menggunakan gambar Rencana Garis yang ada, diambil bagian terluar dari Rencana Garis. Disamping itu juga didesain penempatan peralatan- peralatan, sistem-sistem dan perlengkapan bantu kapal sehingga diperoleh hasil seperti pada Gambar 7. potong pada arah tertentu dengan jarak yang secara umum konstan. Lines Plan dapat dilihat seperti pada Gambar 6. Gambar 6. Lines Plan CNG Carrier p , p g p p sehingga diperoleh hasil seperti pada Gambar 7. M. Pemodelan 3 Dimensi Setelah rencana umum dibuat, selanjutnya pemodelan 3D dapat dilakukan dengan pemroyeksian sesuai dengan Rencana Umum. 3D Model kapal ini dapat dilihat pada Gambar 8. Gambar 7. General Arrangement CNG Carrier L. Pembuatan General Arrangement L. Pembuatan General Arrangement Kondisi Diterima Diterima Diterima Diterima Diterima Diterima Rencana Umum adalah pembagian ruangan untuk semua kebutuhan dan perlengkapan. Ruangan yang dimaksud adalah ruang muat, ruang kamar mesin dan akomodasi atau disebut superstructure (bangunan atas). J. Perhitungan Stabilitas J. Perhitungan Stabilitas Pemeriksaan kondisi dilakukan guna mengetahui karakteristik kapal untuk setiap kondisi pemuatan yang berbeda (loadcase). Untuk mengetahui kriteria stabilitas dipenuhi atau tidak maka harus melebihi kriteria yang ditetapkan. Kriteria stabilitas yang digunakan dalam perhitungan adalah IS Code 2008. Banyaknya loadcase yang dihitung sama dengan loadcase pada perhitungan trim. Dimana hasil yang diperoleh seperti pada Tabel 5. Tabel 3. Rekapitulasi Batasan Trim No Kondisi Batasan Nilai Status 1 Loadcase 1 0.409 0.220 Diterima 2 Loadcase 2 0.409 0.322 Diterima 3 Loadcase 3 0.409 0.274 Diterima 4 Loadcase 4 0.409 0.197 Diterima 5 Loadcase 5 0.409 0.308 Diterima 6 Loadcase 6 0.409 0.259 Diterima 7 Loadcase 7 0.409 0.384 Diterima 8 Loadcase 8 0.409 0.277 Diterima 9 Loadcase 9 0.409 0.241 Diterima Tabel 3. Tabel 5. Rekapitulasi Stabilitas Kapal Data e0-30o (m.deg) e0-40o (m.deg) e30-40o (m.deg) h30o (m.deg) θmax (deg) GM0 (m) 1 14.47 26.75 12.27 1.40 45.60 1.55 2 13.97 25.71 11.73 1.32 44.70 1.48 3 15.36 27.97 12.61 1.42 45.60 1.66 4 13.21 24.06 10.84 1.20 44.70 1.39 5 13.20 23.73 10.52 1.15 43.80 1.39 6 13.67 24.32 10.65 1.16 44.70 1.46 7 24.81 43.52 18.70 2.19 50.90 2.92 8 24.48 42.87 18.39 2.14 50.90 2.88 9 24.80 43.40 18.59 2.17 50.90 2.92 Criteria ≥ 3.15 ≥ 4.87 ≥ 1.71 ≥ 0.2 ≥ 15 ≥ 0.15 Tabel 5. Rekapitulasi Stabilitas Kapal Tabel 5. Berikut loadcase yang harus dihitung kondisi trimnya: 1. Loadcase 1 merupakan kasus di mana muatan 100% dan bahan bakar 100%. 2. Loadcase 2 merupakan kasus di mana muatan 100% dan bahan bakar 50%. 3. Loadcase 3 merupakan kasus di mana muatan 100% dan bahan bakar 10%. JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) G92 M. Pemodelan 3 Dimensi G b 6 Li Pl CNG C i Gambar 6. Lines Plan CNG Carrier Setelah rencana umum dibuat, selanjutnya pemodelan 3D dapat dilakukan dengan pemroyeksian sesuai dengan Rencana Umum. 3D Model kapal ini dapat dilihat pada Gambar 8. Gambar 6. Lines Plan CNG Carrier Gambar 6. Lines Plan CNG Carrier Gambar 7. General Arrangement CNG Carrier Gambar 7. General Arrangement CNG Carrier Gambar 7. General Arrangement CNG Carrier JURNAL TEKNIK ITS Vol. 6, No. 2 (2017), 2337-3520 (2301-928X Print) G93 (a) (b) (c) Gambar 8. 3D Model CNG Carrier (c) (b) (b) (c) (a) Gambar 8. 3D Model CNG Carrier (Persero) Tahun 2016 s.d 2025. Dari data tersebut mengenai pasokan CNG yang dibutuhkan untuk pengoperasian PLTGU Lombok Peaker diperoleh payload sebesar 2873 ton. B. Biaya Operasional [1] G. Fathurahim, “Perancangan Barge Untuk Angkutan CNG (Compressed Natural Gas) Di Perairan Dangkal: Jalur Pelayaran Sembakung – Nunukan,” Publ. Ilm. Online Mhs., vol. 1, no. 1, 2013. [1] G. Fathurahim, “Perancangan Barge Untuk Angkutan CNG (Compressed Natural Gas) Di Perairan Dangkal: Jalur Pelayaran Sembakung – Nunukan,” Publ. Ilm. Online Mhs., vol. 1, no. 1, 2013. Biaya operasional merupakan biaya yang harus dikeluarkan pemilik kapal secara rutin setiap tahunnya. Untuk biaya operasional CNG carrier ini dapat dilihat pada Tabel 7. 2] Enersea, “CNG Technology,” Enersea, 2000. [Online]. Available: http://enersea.com/cng-technology/. 2] Enersea, “CNG Technology,” Enersea, 2000. [Online]. Available: http://enersea.com/cng-technology/. Tabel 7. Biaya Operasional TOTAL OPERATIONAL COST Biaya Nilai Masa Gaji Komplemen Rp 5,664,000,000 per tahun Biaya Perawatan Rp 5,129,879,874 per tahun Asuransi Rp 769,481,981 per tahun Biaya Kebutuhan Bahan Bakar Rp 13,253,091,871 per tahun Biaya Kebutuhan Air Bersih Rp 83,324,292 per tahun Biaya Perbekalan dan Perlengkapan Rp 1,988,783,967 per tahun Total Rp 26,888,561,985 per tahun Tabel 7. Biaya Operasional TOTAL OPERATIONAL COST K. E. dan S. D. M. (ESDM), Potensi dan Peluang Investasi Sektor Energi dan Sumber Daya Mineral. 2014. A. Biaya Pembangunan Biaya pembangunan diperoleh menggunakan ‘Pedoman Pembuatan Perkiraan Biaya (Cost Estimate), Direktorat Pengolahan, PERTAMINA’. Total biaya pembangunan dapat dilihat pada Tabel 6. 2. Ukuran Utama CNG Carrier yang didesain yaitu: 2. Ukuran Utama CNG Carrier yang didesain yaitu: • Lpp (Panjang) = 81.8 m • B (Lebar) = 14.7 m • H (Tinggi) = 8 m • T (Sarat) = 5 m Tabel 6. Biaya Pembangunan Detail % Cost ($) Hull Part 35.3 1,357,151.76 Machinery Part 23.0 884,263.19 Electric Part 8.2 315,259.05 Construction 20.0 768,924.51 Launching and Testing 1.0 38,446.23 Inspection, Survey and Certification 1.0 38,446.23 Design 3.0 115,338.68 Order 3.5 134,561.79 Margin 5.0 192,231.13 Total 100 3,844,622.55 Biaya total pembangunan sebesar Rp 51,298,798,739. Desain Rencana Garis dan Rencana Umum sudah dibuat dan dilampirkan pada lampiran. Desain Rencana Garis dan Rencana Umum sudah dibuat dan dilampirkan pada lampiran. p p p 3. Desain 3D Model sudah dibuat dan dilampirkan pada lampiran. 3. Desain 3D Model sudah dibuat dan dilampirkan pada lampiran. 4. Berdasarkan analisis ekonomis yang dilakukan, didapat biaya estimasi pembangunan CNG carrier ini sebesar Rp 51,298,798,739. Sedangkan untuk biaya operasional setiap tahunnya sebesar Rp 26,888,561,985. 4. Berdasarkan analisis ekonomis yang dilakukan, didapat biaya estimasi pembangunan CNG carrier ini sebesar Rp 51,298,798,739. Sedangkan untuk biaya operasional setiap tahunnya sebesar Rp 26,888,561,985. VI. KESIMPULAN Setelah dilakukan analisis maka didapat kesimpulan: p p 1. Berdasarkan data dari Keputusan Menteri Energi dan Sumber Daya Mineral Republik Indonesia Nomor 5899 K/20/MEM/2016 tentang Pengesahan Rencana Usaha Penyediaan Tenaga Listrik PT Perusahaan Listrik Negara 1. Berdasarkan data dari Keputusan Menteri Energi dan Sumber Daya Mineral Republik Indonesia Nomor 5899 K/20/MEM/2016 tentang Pengesahan Rencana Usaha Penyediaan Tenaga Listrik PT Perusahaan Listrik Negara 1. Berdasarkan data dari Keputusan Menteri Energi dan Sumber Daya Mineral Republik Indonesia Nomor 5899 K/20/MEM/2016 tentang Pengesahan Rencana Usaha Penyediaan Tenaga Listrik PT Perusahaan Listrik Negara
https://openalex.org/W2027222610	https://www.scielo.br/j/rbef/a/PMGNhcvS64gQD7J46mJ5QKk/?lang=pt&format=pdf	Portuguese	null	Dificuldades Conceituais em Física Apresentadas por Alunos Ingressantes em um Curso de Engenharia	Revista Brasileira de Ensino de Física	2,002	cc-by	12,996	I In tro du c ~ ao ram discutir a f sica em termos conceituais, evitando ao m aximo o uso de ferramen tas matem aticas [3]. No en tan to, no ensino de f sica para alunos das areas de ci ^ encias exatas, a habilidade para expressar matema- ticamen te os conceitos f sicos e t~ ao imp ortan te quan to o conhecimen to dos conceitos em si. T radicionalmen te nesses cursos, a disciplina de \C alculo" e oferecida em paralelo a disciplina de \F sica", e p or isso, m uitas v ezes o desconhecimen to do c alculo e considerado o culpado p elo fracasso de parte dos alunos em obter promo c~ ao nas disciplinas de f sica. Dev e-se aten tar para o fato, p or em, de que, al em das ferramen tas de c alculo, para que o aluno tenha sucesso em um curso de f sica no ensino sup erior, e preciso que ele domine os conceitos b asicos que s~ ao explorados p ela disciplina, b em como p ossua a habilidade para in terpretar e criar gr acos. O in cio de um curso de f sica b asica para o ensino sup erior normalmen te en v olv e o estudo de t opicos de mec^ anica de p on tos materiais. Em princ pio, os concei- tos explorados s~ ao aqueles que j a foram (ou dev eriam ter sido) aprendidos no ensino m edio, p or em com um aprofundamen to maior e com o uso de ferramen tas de c alculo diferencial e in tegral. Esse tip o de ab ordagem em espiral p or renamen tos sucessiv os (\spir al appr o- ach "), com a reexplica c~ ao de um dado assun to com um n v el de aprofundamen to maior, e uma t ecnica usual den tro de div ersas areas de ensino [1]. P ara que p ossa ser aplicada, no en tan to, e imp ortan te que o aluno te- nha tido sucesso na etapa an terior do pro cesso. 324 324 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Diculdades Conceituais em F sica Apresen tadas p or Alunos Ingressan tes em um Curso de Engenharia Conceptual Ph ysics diÆculties presen ted b y freshmen in an engineering course V agner Bernal Barb eta e Issao Y amamoto Dep artamento de F sic a, UNIFEI, Centr o Universit ario da FEI A v. Humb erto de A. C. Br anc o, 3972, 09850-901, S. B. Camp o, SP, Br as vb arb eta@fei.e du.br V agner Bernal Barb eta e Issao Y amamoto Dep artamento de F sic a, UNIFEI, Centr o Universit ario da FEI A v. Humb erto de A. C. Br anc o, 3972, 09850-901, S. B. Camp o, SP, Br asil vb arb eta@fei.e du.br ecebido em 19 de outubro, 2001. Man uscrito revisado em 30 de julho, 2002. Aceito em 6 de agosto, 2002 O ob jetiv o deste trabalho e o de apresen tar alguns resultados obtidos atra v es da aplica c~ ao de um teste en tre alunos ingressan tes no ciclo b asico de um curso de engenharia. Este teste visou lev an tar as principais diculdades conceituais em F sica, apresen tadas p or esses alunos, no t opico de mec^ anica cl assica. O teste aplicado foi uma adapta c~ ao do \Me chanics Baseline T est ", que tem sido utilizado com esta nalidade em algumas univ ersidades americanas. O que foi observ ado e que, em b ora os alunos tenham tido con tato pr evio com os t opicos explorados p elo teste, o n v el de amadurecimen to sobre o assun to e ainda p equeno, prev alecendo conceitos baseados em um senso com um impreciso. In this w ork w e sho w some results obtained through the application of a test among engineering freshmen. It w as in tended to rev eal the main conceptual diÆculties in classical mec hanics sho wn b y the studen ts. The test is a v ersion of the \Mec hanics Baseline T est", whic h has b een used in some American univ ersities to ev aluate the studen t bac kground in this area. W e w ere able to observ e that although the studen ts ha v e a previous kno wledge regarding the ph ysics concepts explored in the test, their lev el of maturit y in the sub ject is still unsatisfactory , prev ailing ideas based on an imprecise common sense. nica c~ ao. ou cujo conhecimen to do m undo f sico basea v a-se no seu senso com um e na exp eri ^ encia p essoal. O \Me cha- nics Baseline T est ", p or seu lado, utiliza quest~ oes que n~ ao p o deriam ser resp ondidas sem um conhecimen to dos conceitos b asicos de mec^ anica. Os dois testes se complemen tam, dando um b om p erl sobre o en tendi- men to da maioria dos conceitos b asicos newtonianos. O MBT foi escolhido p ois, em princ pio, o aluno ingres- san te no ensino sup erior j a tev e con tato com os concei- tos da mec^ anica newtoniana no ensino m edio. A tra v es da aplica c~ ao do teste, p o deria ser v ericado se real- men te esses conceitos s~ ao dominados de forma satis- fat oria p elos alunos. Quan to ao conhecimen to dos conceitos b asicos de f sica, p erceb e-se que tam b em n~ ao s~ ao dominados de forma satisfat oria e p o dem, p ortan to, con tribuir para uma maior diculdade na obten c~ ao de sucesso nessa disciplina de f sica. Em grande parte, essas diculda- des se caracterizam p ela exist ^ encia de concep c~ oes al- ternativ as, ou esp on t^ aneas, diferen tes daquelas aceitas cien ticamen te. Essas concep c~ oes esp on t^ aneas t ^ em sido ob jeto de estudo de p esquisadores de ensino de f sica no Brasil desde a d ecada de 80 [6, 7]. A imp ort^ ancia do seu estudo est a no fato de que a presen ca de tais con- cep c~ oes in uencia, de maneira imp ortan te, na forma p ela qual os alunos aprendem os no v os conceitos que lhe s~ ao ensinados. nica c~ ao. A p ersist ^ encia destas no c~ oes es- p on t^ aneas, mesmo ap os anos de educa c~ ao formal, indica a p ossibilidade de se p o der criar uma sup erp osi c~ ao en tre os conceitos cien t cos e aqueles esp on t^ aneos, tendo-se a um obst aculo para um aprendizado signicativ o da f sica. A escolha do esquema cien t co ou do esp on t^ aneo se d a aparen temen te p elo tip o de problema que se est a analisando. Isto e, quando o problema en v olv e m ui- tos conceitos formais, o esquema formal e utilizado, e quando en v olv e elemen tos do dia-a-dia, o esquema es- p on t^ aneo e escolhido [8]. Em um trabalho recen te [9], as concep c~ oes esp on t^ aneas t ^ em sido iden ticadas como uma das barreiras para a constru c~ ao de mo delos men- tais para a resolu c~ ao de problemas de f sica. O teste constitui-se de 26 quest~ oes de m ultipla es- colha, das quais ap enas 7 exigem a realiza c~ ao de algum tip o de c alculo matem atico (estas quest~ oes en v olv em basicamen te conhecimen tos de algebra elemen tar). P or- tan to, as quest~ oes t ^ em p or ob jetiv o v ericar a bagagem conceitual dos alunos em f sica, disp ensando o conheci- men to de um ferramen tal matem atico elab orado. Na T ab ela I apresen ta-se a distribui c~ ao das quest~ oes, feita p elos autores do MBT, segundo os con- ceitos en v olvidos. As quest~ oes en tre par ^ en teses indicam que outros conceitos s~ ao signicativ amen te explorados para a solu c~ ao do problema. Anexo, temos o conjun to completo das quest~ oes que comp~ oem o MBT. Este trabalho visou iden ticar o grau de en ten- dimen to conceitual dos v arios t opicos relativ os a ci- nem atica e a din^ amica de p on tos materiais, b em como a habilidade do aluno em in terpretar gr acos. nica c~ ao. Pro curou- se, ainda, v ericar at e que p on to a falta de ferramen tal matem atico e um obst aculo para um b om aprendizado de f sica b asica no n v el sup erior. P ara tan to, utilizou- se como instrumen to de p esquisa um teste, que foi apli- cado en tre os estudan tes ingressan tes em um curso de engenharia. Ele evita a necessidade de o aluno p ossuir conhecimen tos mais aprofundados de matem atica para a solu c~ ao dos problemas, bastando o conhecimen to de algebra elemen tar. O teste e uma adapta c~ ao do \Me- chanics Baseline T est " (MBT), que tem sido utilizado em p esquisas similares realizadas nos Estados Unidos [10]. Assunto Questão Cinemática Movimento linear aceleração constante 1, (2), (3) aceleração média (18), 23 velocidade média 25 deslocamento por integração 24 Movimento curvilíneo aceleração tangencial 4 aceleração normal 5, (8), (12) uso da expressão an = v2/r (9), (12) Princípios Gerais (Dinâmica) Primeira Lei (2) Segunda Lei (3), 8, (9), (12), (18) dependência com a massa 17, 21 Terceira Lei (12), (13), 14 Princípio da Superposição 7, (5), (13), 19 Trabalho-energia 20 Conservação da energia 10, 11 Impulso-momento linear 16, 22 Conservação do momento linear 15 Forças Específicas Queda livre gravitacional 6, 26 Atrito (9) Assunto Questão Cinemática Movimento linear aceleração constante 1, (2), (3) aceleração média (18), 23 velocidade média 25 deslocamento por integração 24 Movimento curvilíneo aceleração tangencial 4 aceleração normal 5, (8), (12) uso da expressão an = v2/r (9), (12) Princípios Gerais (Dinâmica) Primeira Lei (2) Segunda Lei (3), 8, (9), (12), (18) dependência com a massa 17, 21 Terceira Lei (12), (13), 14 Princípio da Superposição 7, (5), (13), 19 Trabalho-energia 20 Conservação da energia 10, 11 Impulso-momento linear 16, 22 Conservação do momento linear 15 Forças Específicas Queda livre gravitacional 6, 26 Atrito (9) I In tro du c ~ ao As diculdades enfren tadas p elos alunos no estudo de f sica do ensino sup erior s~ ao m uitas v ezes atribu das a deci ^ encia dos alunos em manipular o ferramen tal matem atico que e normalmen te exigido nestes cursos. P or essa raz~ ao, as disciplinas de f sica ministradas em cursos ligados a area de biom edicas, p or exemplo, s~ ao as v ezes concebidas de mo do a se utilizar p ouco ou ne- nh uma ferramen ta de c alculo [2]. Isso ca mais eviden te quando se analisam alguns livros did aticos que pro cu- Alguns trabalhos [4, 5] d~ ao uma indica c~ ao de que a in terpreta c~ ao de gr acos p or parte dos alunos ingres- san tes no ensino sup erior e decien te. Isso e bastan te problem atico, p ois o uso de gr acos na explica c~ ao de conceitos de f sica constitui-se em uma praxe, tornando- se uma linguagem esp ecial utilizada p elo professor, e que p o de n~ ao ser dominada p or uma parcela dos estu- dan tes, gerando-se assim um s erio problema de com u- V agner Bernal Barb eta e Issao Y amamoto V agner Bernal Barb eta e Issao Y amamoto 325 IV Resultados Na T ab ela I I, tem-se o resultado dos question arios apli- cados nos anos de 1999 e 2000. Nessa tab ela p o dem-se observ ar os ndices p ercen tuais para cada alternativ a escolhida, destacando-se em negrito os ndices corres- p ondentes as alternativ as corretas. O teste foi aplicado nas primeiras semanas de aula. I I Mec hanics Baseline T est (MBT) A m de a v aliar qualitativ amen te a compreens~ ao da maioria dos conceitos fundamen tais na mec^ anica, Hes- tenes e W ells elab oraram o \Me chanics Baseline T est " (MBT) [10], um conjun to de quest~ oes abrangendo con- ceitos de f sica in tro dut oria. Os dois autores, jun ta- men te com Sw ac khamer, j a ha viam elab orado outro teste, o \F or c e Conc ept Inventory " [11], destinado a es- tudan tes mesmo sem treinamen to formal em mec^ anica T ab ela I. Conceitos newtonianos do \Me chanics Baseline T est ". Cada conceito e explorado na quest~ ao indicada. Os n umeros en tre par ^ en teses indicam que outros conceitos s~ ao signicativ amen te explorados na quest~ ao. Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 326 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 I I I Amostragem e coleta de da- dos F oi solicitado aos alunos que resp ondessem a to das as quest~ oes, e optamos p ela inclus~ ao de um item \N~ ao sei resp onder" (alternativ a \f "), a m de p o dermos ter uma id eia dos assun tos que o aluno n~ ao tiv esse tido con- tato duran te o ensino m edio. N~ ao foi p ermitido o uso de calculadora, j a que os ev en tuais c alculos en v olviam ap enas conhecimen tos de algebra elemen tar. Duran te a aplica c~ ao do teste, foi extensiv amen te divulgado que o resultado n~ ao traria qualquer b enef cio ou preju zo ao aluno na obten c~ ao da promo c~ ao na disciplina, sendo que a participa c~ ao no teste foi op cional. O teste foi aplicado, inicialmen te, no segundo semestre de 1999, atingindo um total de 771 alunos matriculados em F sica I, sendo 494 do diurno e 277 do noturno. No segundo semestre de 2000, o teste foi no v amen te apli- cado, sendo p esquisados desta v ez 814 alunos, 520 do diurno e 294 do noturno. O curso analisado e dividido em p er o dos semestrais, sendo as disciplinas de f sica desen v olvidas em quatro semestres, iniciando-se no 1 o p er o do para os alunos do diurno, e no 2 o p er o do para os alunos do noturno. P elo fato de o curso ser semestral, h a turmas que iniciam o curso tan to no primeiro quan to no segundo semestre do ano. Nos totais apresen tados est~ ao inclu dos calouros b em como alunos dep enden tes em F sica I (que n~ ao conseguiram apro v a c~ ao e que foram considerados como alunos ingressan tes), abrangendo-se alunos dos cursos diurno e noturno. T ab ela I I. P ercen tagens das alternativ as escolhidas p elos alunos (agosto de 1999 e agosto de 2000) para cada uma das quest~ oes. Em negrito destacam-se as alternativ as corretas. T ab ela I I. P ercen tagens das alternativ as escolhidas p elos alunos (agosto de 1999 e agosto de 2000) para ca Em negrito destacam se as alternativ as corretas Em negrito destacam-se as alternativ as corretas. V agner Bernal Barb eta e Issao Y amamoto 327 Na T ab ela I I I apresen tam-se os ndices p ercen tu- ais de acerto obtidos p elos alunos no segundo semes- tre de 1999 (2s1999) e no segundo semestre de 2000 (2s2000), para as v arias quest~ oes do teste. A t tulo de compara c~ ao, transcrev eram-se tam b em os resultados obtidos p elos autores do MBT [10] em um col egio do Estado do Arizona e na Univ ersidade de Harv ard, nos Estados Unidos. Na coluna mais a direita, observ a-se tam b em o escore reduzido ou \escore z", calculado para compara c~ ao das duas turmas analisadas, para cada uma das quest~ oes. Esse par^ ametro e utilizado em an alises es- tat sticas para a v erica c~ ao da exist ^ encia ou n~ ao de di- feren ca signicativ a en tre 2 conjun tos de amostras [12]. F oi calculado tam b em o escore z para a compara c~ ao das m edias gerais de acertos en tre as duas turmas analisa- das, b em como para a compara c~ ao en tre as turmas do segundo semestre de 1999 e as turmas do Colegial Ari- zona e de Harv ard. Ainda nesta tab ela, \N" se refere ao n umero de alunos analisados e \D.P ." e o desvio- padr~ ao. As Figuras 1 e 2 fornecem uma compara c~ ao visual dos dados con tidos na T ab ela I I I. Questão Brasil 2s1999 Brasil 2s2000 Colegial Arizona Univ. T ab ela I I I. Resultados p ercen tuais obtidos na aplica c~ ao do \Me chanics Baseline T est " em uma escola de engenharia do Brasil, n um col egio do Arizona e na Univ ersidade de Harv ard. IV Resultados Agosto de 2000 a b c d e f q.1 26 66 2,3 1,2 1,2 3,2 q.2 6,9 8,8 12 45 17 10 q.3 6,1 13 8,4 29 19 24 q.4 0,7 2,9 76 14 4,1 2,5 q.5 3,4 47 14 4,5 26 4,5 q.6 2 29 39 5 18 6,6 q.7 2 3,7 14 65 5,3 9,5 q.8 7,1 17 27 10 14 25 q.9 1,8 6,9 2,8 6,4 6,3 75 q.10 62 2,3 8,2 7,9 14 4,9 q.11 13 9,6 5 9,6 4,3 58 q.12 17 7,5 2,8 1,8 8,1 62 q.13 17 20 6,3 25 2,3 29 q.14 17 40 1,8 12 1,2 27 q.15 15 5,5 5 14 30 30 q.16 23 45 3,8 2,8 12 12 q.17 0,5 2,2 46 26 4,3 20 q.18 11 6,8 8,4 8,1 6,8 59 q.19 6,1 33 26 13 7,9 14 q.20 22 39 15 5,3 1 17 q.21 64 9,2 15 3,1 0,9 6,6 q.22 23 25 23 3,9 0,9 23 q.23 9,8 13 17 24 9,7 26 q.24 20 10 13 5,8 24 28 q.25 16 16 11 8,1 15 33 q.26 9 6,1 15 42 6 21 Agosto de 1999 a b c d e f q.1 32 60 3 2,3 0,8 2,7 q.2 5,8 11 13 44 18 8,9 q.3 8,2 11 11 31 18 20 q.4 1,2 4 76 14 2,9 1,2 q.5 2,3 51 14 4,3 26 1,7 q.6 2,3 35 32 6,9 18 5,3 q.7 2,9 3,8 15 66 6,1 5,7 q.8 11 21 26 13 13 15 q.9 3,1 12 8,6 10 6,9 58 q.10 63 3,5 7,1 9,3 15 2,3 q.11 15 14 8,4 12 7 43 q.12 22 12 3,5 3 16 43 q.13 17 21 9,5 30 3,6 18 q.14 17 45 4,4 13 1,7 18 q.15 15 7 6,1 16 32 22 q.16 25 44 3,5 3,9 12 11 q.17 1,7 2,5 51 28 5,1 11 q.18 10 10 12 11 12 43 q.19 7,1 33 22 12 11 13 q.20 23 37 19 5,7 1 13 q.21 64 7,9 15 4,7 0,6 5,7 q.22 25 22 28 5,7 2,2 16 q.23 9,7 17 17 33 10 12 q.24 19 12 17 5,7 26 18 q.25 17 22 11 10 17 21 q.26 7,3 8,3 17 42 6 18 Em negrito destacam-se as alternativ as corretas. V agner Bernal Barb eta e Issao Y amamoto Harvard Escore z 1 60 66 54 78 -2,47 2 44 45 40 78 -0,40 3 18 19 29 93 -0,51 4 76 76 85 67 0,00 5 2 3 1 18 -1,27 6 32 39 45 87 -2,91 7 15 14 8 36 0,56 8 13 10 23 81 1,87 9 10 6 21 68 2,94 10 15 14 35 89 0,56 11 14 10 25 85 2,45 12 4 3 12 24 1,08 13 21 20 31 79 0,49 14 45 40 51 87 2,01 15 32 30 48 83 0,86 16 25 23 16 60 0,93 17 28 26 26 81 0,90 18 10 7 15 32 2,14 19 22 26 16 78 -1,86 20 19 16 25 46 1,57 21 64 62 62 89 0,82 22 22 26 56 32 -1,86 23 33 25 28 84 3,51 24 19 21 29 59 -0,99 25 17 16 25 61 0,54 26 6 8 13 53 1,56 N 771 814 600 183 Média (%) 26 25 32 66 D.P. (%) 11 11 11 14 328 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 V Discuss ~ ao e an alise agosto de 1999 e a de agosto de 2000), obt em-se um v alor de z = 1,81, o que indica que as diferen cas en tre as turmas n~ ao s~ ao signicativ as a um n v el de signi- c^ ancia de 0,05. F azendo-se uma an alise semelhan te en tre a turma de agosto de 1999 e a do col egio no Ari- zona, obt em-se um v alor de z = -10,02, o que p ermite concluir que existe uma diferen ca signicativ a en tre os dois grup os de alunos. P ara o caso da compara c~ ao com os alunos de Harv ard, obt em-se um v alor de z = -36,10, indicando que estes apresen tam diferen cas signicativ as com os alunos brasileiros, sendo os alunos de Harv ard signicativ amen te mais b em preparados. P ara os alu- nos de Harv ard, as quest~ oes mais f aceis foram a 3, 10 e 21, e as mais dif ceis a 5, 12 e 18. V.1. An alise dos resultados dos testes Analisando-se visualmen te as Figuras 1 e 2, n~ ao se notam grandes diferen cas en tre os resultados obtidos en tre os alunos que ingressaram no primeiro p er o do em agosto de 1999 e os que ingressaram em agosto de 2000, tan to nas quest~ oes consideradas mais f aceis, isto e, aquelas com maiores ndices de acertos (4, 21 e 1), quan to nas mais dif ceis, isto e, aquelas com menores ndices de acertos (5, 12 e 26). Ao calcular-se o es- core reduzido, ou escore z, para a compara c~ ao en tre a m edia de acertos das duas turmas analisadas (a de 0 10 20 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 11 12 13 número da questão percentagem de acertos Brasil- 2s1999 Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 1. Resultados p ercen tuais (quest^ oes 1 a 13) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 0 10 20 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 11 12 13 número da questão percentagem de acertos Brasil- 2s1999 Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 1. Resultados p ercen tuais (quest^ oes 1 a 13) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 0 10 20 30 40 50 60 70 1 2 3 4 5 6 7 8 9 10 11 12 13 número da questão percentagem de acertos Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 1. Resultados p ercen tuais (quest^ oes 1 a 13) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 0 10 20 30 40 50 60 70 80 90 100 14 15 16 17 18 19 20 21 22 23 24 25 26 número da questão percentagem de acertos Brasil- 2s1999 Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 2. V agner Bernal Barb eta e Issao Y amamoto P ara a correta an alise desse problema, e imp ortan te observ ar a m udan ca na tra jet oria do blo co e, p ortan to, notar que existe v aria c~ ao na dire c~ ao da v elo cidade v etorial, o que lev a a exist ^ encia da comp onen te normal da ace- lera c~ ao, fazendo com que as dire c~ oes da v elo cidade e da acelera c~ ao n~ ao coincidam. O conceito de comp o- nen te normal da acelera c~ ao est a inclu do no v amen te na quest~ ao 12, que tev e tam b em baix ssimo ndice de acer- tos. Como j a foi discutido, a m edia de acertos nas quest~ oes en v olv endo mo vimen to curvil neo, como e o caso da quest~ ao 12, foi de ap enas 7%. No v amen te a confus~ ao en tre os conceitos de v elo cidade e de acele- ra c~ ao p o de ser observ ada p ela p equena quan tidade de acertos obtidos na quest~ ao 26. Essa quest~ ao en v olv e ap enas o conhecimen to de que em queda livre a ace- lera c~ ao e constan te. A mistura en tre os conceitos de v elo cidade e acelera c~ ao se torna aparen te na medida em que a resp osta mais freq uen te foi a alternativ a \d" (42%), segundo a qual a acelera c~ ao de um ob jeto, ap os ser lan cado v erticalmen te para cima, v ai dimin uindo at e atingir o p on to de altura m axima, quando en t~ ao passa a ter o v alor zero. Esse e justamen te o comp ortamen to da v elo cidade. Como n~ ao se v ericou diferen ca signicativ a no de- semp enho global das duas turmas analisadas, para sim- plicar a discuss~ ao e a an alise dos resultados ser~ ao apre- sentados, a seguir, somen te os p ercen tuais relativ os ao teste aplicado em agosto de 1999. V agner Bernal Barb eta e Issao Y amamoto Na ultima coluna da T ab ela I I I, p o de ser observ ado o escore z, calculado para a compara c~ ao en tre as duas turmas brasileiras, para cada uma das quest~ oes do teste. P o de-se notar, p ela an alise desta tab ela, que existe di- feren ca signicativ a para um n v el de signic^ ancia de 0,05, somen te para as quest~ oes 1, 6, 9, 11, 14, 18 e 23. tos de v elo cidade e de acelera c~ ao. Essa confus~ ao en tre tais conceitos tem sido observ ada h a bastan te temp o, constituindo um tip o de concep c~ ao esp on t^ anea apresen- tada p or alunos de diferen tes n v eis [6, 7]. P o de-se notar, p ela an alise desta tab ela, que existe di- feren ca signicativ a para um n v el de signic^ ancia de 0,05, somen te para as quest~ oes 1, 6, 9, 11, 14, 18 e 23. p [ , ] A quest~ ao 4, que obtev e a maior p ercen tagem de acertos em to do o teste, necessita do conceito de comp o- nen te tangencial da acelera c~ ao. Essa foi a unica quest~ ao em que os alunos brasileiros tiv eram um ndice me- lhor do que aquele observ ado em Harv ard. T em-se a impress~ ao que os alunos resp onderam p or \in tui c~ ao" asso ciando-se a dire c~ ao da v elo cidade com a dire c~ ao da acelera c~ ao, o que e v alido para esta situa c~ ao. Isso e refor cado p ela an alise da quest~ ao 5, onde no v amen te a resp osta mais selecionada foi aquela em que a dire c~ ao da acelera c~ ao coincide com a da v elo cidade, o que n~ ao e v alido para este caso. Como conseq u ^ encia, tiv emos o menor ndice de acertos para essa quest~ ao. V agner Bernal Barb eta e Issao Y amamoto F azendo-se uma an alise global dos resultados, a m edia de acertos na parte de Cinem atica, mo vimen to linear, obtido atra v es dos resultados das quest~ oes 1, 2, 3, 18, 23, 24 e 25, foi de 28,7%. Como se sab e, no en- sino m edio, essa parte tem merecido mais aten c~ ao da grande maioria dos professores. P ortan to, esp era v a-se uma p ercen tagem maior de acertos para esse t opico. O mesmo ndice praticamen te se rep ete para os outros t opicos, exceto para o caso de mo vimen to curvil neo, o qual tev e uma m edia de acertos de cerca de 7 %, b em abaixo daquele observ ado para os outros casos. E in teressan te observ ar que os maiores ndices de resp ostas para a alternativ a \f " ("N~ ao sei resp onder") corresp ondem as quest~ oes 9, 11, 12 e 18. P ara a solu c~ ao dessas quest~ oes, al em do conceito f sico, era necess ario o conhecimen to de uma express~ ao matem atica para o c alculo da grandeza p edida. Basicamen te, o que se pre- cisa v a sab er para resolv er tais quest~ oes eram express~ oes relativ as a for ca de atrito, acelera c~ ao cen tr p eta, ener- gia p otencial e cin etica, e a 2 a lei de Newton. Outras quest~ oes que necessitam de algum c alculo, tam b em ti- v eram p ercen tagens apreci av eis de resp ostas para a al- ternativ a \f " (13, 14, 17, 23, 24 e 25). Nesse caso, necessita v a-se do conhecimen to das express~ oes da ci- nem atica (v elo cidade m edia e acelera c~ ao m edia) e da 2 a lei de Newton, al em da capacidade de in terpreta c~ ao gr aca. Uma an alise individual de algumas quest~ oes fornece alguns resultados tam b em in teressan tes, conforme ser a detalhado a seguir. V agner Bernal Barb eta e Issao Y amamoto Na quest~ ao 6, a maior p ercen tagem de resp ostas foi para a alternativ a \b", em que a dire c~ ao da acelera c~ ao foi no v amen te tomada como sendo a mesma da v elo ci- dade. Assim como no caso da quest~ ao 26, a sua solu c~ ao dep ende ap enas do conhecimen to do comp ortamen to de um corp o em queda livre. As quest~ oes 1 e 2 tiv eram ndices de acerto relativ a- men te altos. Na quest~ ao 1, a resp osta correta foi obtida p or 60% dos alunos. Os 32% que escolheram a alterna- tiv a \a", pro v a v elmen te n~ ao zeram um paralelo en tre o espa camen to do diagrama apresen tado no en unciado e as inclina c~ oes das retas do gr aco da v elo cidade, acer- tando ap enas o formato geral da curv a. Na quest~ ao 2, houv e 44% de acertos. Os 18% que optaram p ela alternativ a \e", no v amen te acertaram o asp ecto geral do gr aco, errando os v alores dos co ecien tes angulares obtidos a partir do gr aco da v elo cidade (quest~ ao 1). Na quest~ ao 7, a grande p ercen tagem de resp ostas atribu das a alternativ a errada \d" (66%) p ermite con- cluir que os conceitos de decomp osi c~ ao de for ca e de for ca resultan te n~ ao est~ ao presen tes na grande maioria dos alunos. Aqui p o de se observ ar um erro freq uen te, em que se atribui sempre a for ca normal aplicada p elo plano no corp o, o mesmo v alor da for ca p eso do corp o. Na quest~ ao 3, que ap enas 18% acertaram, a maioria das resp ostas (31%) foi para a alternativ a \d". V.1. An alise dos resultados dos testes Resultados p ercen tuais (quest^ oes 14 a 26) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. Figura 1. Resultados p ercen tuais (quest^ oes 1 a 13) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. Figura 1. Resultados p ercen tuais (questoes 1 a 13) obtidos na aplica cao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 0 10 20 30 40 50 60 70 80 90 100 14 15 16 17 18 19 20 21 22 23 24 25 26 número da questão percentagem de acertos Brasil- 2s1999 Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 2. Resultados p ercen tuais (quest^ oes 14 a 26) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 0 10 20 30 40 50 60 70 80 90 100 14 15 16 17 18 19 20 21 22 23 24 25 26 número da questão percentagem de acertos Brasil- 2s1999 Brasil- 2s2000 Colegial Arizona Univ. Harvard Figura 2. Resultados p ercen tuais (quest^ oes 14 a 26) obtidos na aplica c~ ao do Me chanics Baseline T est (agosto de 1999 e agosto de 2000), n um col egio do Arizona e na Univ ersidade de Harv ard. 329 V agner Bernal Barb eta e Issao Y amamoto Uma grande parte das resp ostas (63%) for para a alternativ a \a", em que te- mos uma curv atura do escorregador que lev a a uma \queda" mais brusca. Observ e-se que ap enas 2,3% as- sinalaram \N~ ao sei resp onder". Ainda neste t opico, te- mos a quest~ ao 20, sobre trabalho e energia cin etica, que s o obtev e 19% de acertos. Uma b oa p ercen tagem, 37%, escolheu a alternativ a \b", parecendo que estab e- leceram a prop or c~ ao de que maior massa corresp onde obrigatoriamen te a maior energia cin etica, esquecendo- se de que a v elo cidade tam b em e determinan te para a obten c~ ao dessa grandeza. A quest~ ao 21 foi a segunda em maior p ercen tagem de acertos do teste, 64%, e ad- mite solu c~ ao seja p or energia cin etica, seja atra v es da segunda lei de Newton. A solu c~ ao talv ez tenha sido feita tam b em de forma in tuitiv a, imaginando-se que, para a mesma for ca, o corp o com a menor massa c hega \mais r apido" ao destino. J a as resp ostas atribu das a quest~ ao 22 indicam que o conceito de quan tidade de mo vimen to n~ ao e claro para a maioria, j a que as alternativ as \a", \b" e \c" tiv eram praticamen te as mesmas p ercen tagens de resp ostas. Um gr aco da v elo cidade v ersus temp o est a no en unciado das quest~ oes 23, 24 e 25. A quest~ ao 23, so- bre c alculo de acelera c~ ao m edia, obtev e ap enas 33% de acertos. Existe, aparen temen te, um desconhecimen to do signicado gr aco da acelera c~ ao m edia (co ecien te angular da reta que liga os dois p on tos). A quest~ ao 24, em que se p edia o c alculo do espa co p ercorrido en tre dois instan tes, obtev e ap enas 19% de resp ostas corre- tas. V agner Bernal Barb eta e Issao Y amamoto T em-se a impress~ ao de que isso o correu devido ao estab eleci- men to de uma prop orcionalidade en tre for ca e v elo ci- dade, causada talv ez p or uma confus~ ao en tre os concei- Na quest~ ao 8 observ a-se no v amen te a grande di- culdade dos alunos na an alise de mo vimen tos em curv a e do en tendimen to do signicado da acelera c~ ao normal. P ara 47% dos alunos (que escolheram as alternativ as Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 330 sultan te ou sobre princ pio de sup erp osi c~ ao de for cas. A maior p ercen tagem dos alunos, 33%, escolheu a dire c~ ao da seta B como resp osta. A dire c~ ao correta e a da seta C, escolhida p or 22% dos alunos. \b" e \c"), se a v elo cidade e constan te, a acelera c~ ao e zero, isto e, eles n~ ao lev am em con ta que a acelera c~ ao e resultado da m udan ca n~ ao s o da in tensidade do v e- tor v elo cidade, mas tam b em de sua dire c~ ao. Ap enas 13% acertaram, sendo que 15% preferiram a alterna- tiv a \N~ ao sei resp onder". Os t opicos T rabalho-Energia e Momen to Linear (ou Quan tidade de Mo vimen to) comparecem com m edia de acertos de 26% e 32%, resp ectiv amen te. Note-se que o ndice de acertos para o conceito de momen to line- ar foi relativ amen te alto, equiv alen te ao de cinem atica. As p ercen tagens de acertos de 32% (quest~ ao 15) e 25% (quest~ ao 16) c hegam, p ortan to, a surpreender, uma v ez que s~ ao solicitados conhecimen tos de momen to linear, de subtra c~ ao de v etores e de impulso linear. A quest~ ao 10, que se enquadra no t opico de ener- gia, mostra que falta ao estudan te o conhecimen to da lei da conserv a c~ ao da energia mec^ anica, j a que o ndice de acerto foi de ap enas 15%. V agner Bernal Barb eta e Issao Y amamoto T alv ez grande parte dos alunos n~ ao asso cie espa co p ercorrido com a area sob o gr aco fornecido ou tenha tido diculdades em calcular essa area. Ap enas 17% deles resp onderam corretamen te a quest~ ao 25, em que se p edia a v elo cidade m edia n um in terv alo de temp o. Nessa quest~ ao era necess ario sab er calcular o deslo ca- men to atra v es do gr aco da v elo cidade v ersus temp o, obtido na quest~ ao an terior, para en t~ ao dividir este v alor p elo in terv alo de temp o. Boa parte daqueles que con- seguiram resolv er a quest~ ao 24 tam b em conseguiram resolv er a quest~ ao 25. Nas quest~ oes 13 e 14, com 21% e 45% de acertos, resp ectiv amen te, p erceb em-se falhas em rela c~ ao a se- gunda lei de Newton. A quest~ ao 13, em que o elev ador sob e com v elo cidade constan te, parece ter criado no aluno mais diculdades do que a 14, em que o mesmo elev ador est a estacion ario. A quest~ ao 17 tev e 28% de acertos, sendo que 51% escolheram como resp osta a alternativ a incorreta \c". Essa quest~ ao en v olvia basicamen te a aplica c~ ao da se- gunda lei de Newton, para dois sistemas com mesma for ca resultan te e massas diferen tes, para assim encon- trar a rela c~ ao de acelera c~ oes. A grande quan tidade de resp ostas para a alternativ a \c" denota um aparen te descuido na leitura do en unciado, p ois essa seria justa- men te a alternativ a correta caso as massas fossem m e 2m, quando o correto seria considerar m e 3m. V.2. Compara c ~ ao de amostras Uma outra an alise in teressan te se refere a com- para c~ ao do desemp enho no teste aplicado no segundo semestre de 1999, en tre calouros (418 alunos) e v ete- ranos (76 alunos dep enden tes, ou seja, que foram re- pro v ados p elo menos uma v ez em F sica I) do curso diurno. Calculando-se o p orcen tual m edio de acertos para calouros e v eteranos, obt em-se 25% e 27% res- p ectiv amen te, com desvios-padr~ ao de 11% e de 10 %. Calculando-se o escore reduzido, obt em-se z = -1,57. Isso mostra que den tro de uma signic^ ancia de 0,05 a diferen ca en tre as m edias de acertos de calouros e v e- teranos n~ ao e signicativ a. P o de-se notar visualmen te p ela Figura 3 que h a m uito p ouca diferen ca en tre as di- culdades enfren tadas p or am bas categorias de alunos, sendo os p ercen tuais de acerto bastan te semelhan tes. A quest~ ao 18 en v olv e conhecimen tos sobre acele- ra c~ ao (v alor e sinal) e aplica c~ ao da segunda e da terceira lei de Newton, sendo que ap enas 10% resp onderam cor- retamen te. In teressan te notar que as outras alternati- v as tam b em tiv eram apro ximadamen te essa p ercenta- gem de resp ostas, e 43% escolheram a alternativ a \N~ ao sei resp onder". Uma p ossibilidade para tal tip o de com- p ortamen to e que, em b ora conhe cam a form ula c~ ao da segunda (e da terceira) lei de Newton, os alunos t ^ em diculdades de aplic a-las quando a sua necessidade n~ ao est a expl cita no problema. A quest~ ao 19 en v olv e o conhecimen to sobre for ca re- V agner Bernal Barb eta e Issao Y amamoto 331 Comparação calouros x veteranos 0 10 20 30 40 50 60 70 80 90 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 número da questão % de acertos Calouros Veteranos Figura 3. V.2. Compara c ~ ao de amostras Gr aco comparativ o do desemp enho dos calouros (linha c heia) e v eteranos (linha tracejada) do segundo semestre de 1999, no teste Me chanics Baseline T est. Comparação calouros x veteranos Comparação calouros x veteranos Figura 3. Gr aco comparativ o do desemp enho dos calouros (linha c heia) e v eteranos (linha tracejada) do segundo semestre de 1999, no teste Me chanics Baseline T est. Curso diurno x curso noturno 0 10 20 30 40 50 60 70 80 90 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 número da questão % de acertos diurno noturno Figura 4. P ercen tagem de acertos obtida p elos alunos do curso de engenharia, p er o do diurno e noturno, segundo semestre de 1999, no teste Me chanics Baseline T est. Curso diurno x curso noturno Figura 4. P ercen tagem de acertos obtida p elos alunos do curso de engenharia, p er o do diurno e noturno, segundo semestre de 1999, no teste Me chanics Baseline T est. A Figura 4 p ermite comparar o desemp enho, no mesmo teste, dos alunos de F sica I dos cursos diurno (494) e noturno (277), do segundo semestre de 1999. Calculando-se a m edia de acertos, obteremos 25% para o aluno do diurno e 27% para o aluno do noturno, am- b os com desvio padr~ ao de 11%. Neste caso, z = -2,42, indicando que as diferen cas s~ ao signicativ as para um n v el de signic^ ancia de 0,05, p or em n~ ao s~ ao signica- tiv as para um ndice de signic^ ancia de 0,01. Assim, existe pro v a v elmen te diferen ca signicativ a en tre o de- semp enho das duas categorias de alunos, sendo os do noturno p ossiv elmen te um p ouco mais b em preparados. Isso p o deria ser explicado p elo fato de o aluno do no- turno cursar F sica I no segundo p er o do, tendo j a tido a op ortunidade de cursar um semestre de uma s erie de disciplinas de matem atica. Refer ^ encias solv er simples fra c~ oes, na mon tagem e resolu c~ ao de equa c~ oes, em geometria e em trigonometria. P or em, n~ ao e somen te a falta de ferramen tal matem atico o grande obst aculo para um b om desen v olvimen to desses alunos. Os resultados da aplica c~ ao do MBT conrmam a grande deci ^ encia em rela c~ ao aos conceitos b asicos de f sica. A presen ca de concep c~ oes esp on t^ aneas que lev am a uma vis~ ao restrita da natureza, parece ainda predomi- nar na men te da maioria dos estudan tes. A constru c~ ao de algumas dessas concep c~ oes, p or exemplo a prop or- cionalidade en tre for ca e v elo cidade, se dev e, em parte, ao fato de que div ersos fen^ omenos que nos cercam ne- cessitam da aplica c~ ao de for cas para que as v elo cidades sejam man tidas, isto e, as v ezes o mo vimen to de ob je- tos do dia-a-dia parece seguir melhor a descri c~ ao aris- tot elica do que a mec^ anica newtoniana. Isso tem sido rep ortado na literatura [13, 14, 15], inclusiv e com a pro- p osi c~ ao de que se discuta as hip ot eticas equa c~ oes que descrev eriam estes mo vimen tos, para en t~ ao se in tro du- zir a mec^ anica newtoniana [13]. A confron ta c~ ao en tre as concep c~ oes apresen tadas p elos alunos com aquelas acei- tas cien ticamen te { mostrando-se que estas ultimas ex- plicam, de forma mais ampla, os fen^ omenos observ ados { parece ser uma forma imp ortan te de se atacar esse problema. [1] V er p or exemplo o uso dessa t ecnica para o ensino de programa c~ ao em h ttp://mathserv.monmouth.edu/coursenotes/kun tz/ P ap ers/top do wn.h tm, acessado em 13/03/2001. [2] Um exemplo e o livro de Okuno, E., Caldas, I.L. e Cho w, C., F sic a p ar a ci ^ encias biol ogic as e biom edic as, Harbra, S~ ao P aulo (1982). Refer ^ encias [3] Hewitt, P .G., Conc eptual physics, Addison-W esley , Massac h usets, 8 a edi c~ ao (1998). [4] Beic hner, R. J., T esting student interpr etation of kine- matics gr aphs, Am. J. Ph ys. 62, 750 (1994). [5] Agrello, D.A. e Garg, R., Compr e ens~ ao de gr ac os de cinem atic a em f sic a intr o dut oria, Rev. Bras. Ens. F s. 21 (1), 103 (1999). [6] Zylb erszta jn, A., Conc ep c~ oes esp ont^ ane as em f sic a: exemplos em din^ amic a e implic a c~ oes p ar a o ensino, Rev. Ens. F s. 5 (2), 3 (1983). [7] Villani, A., P acca, J. L. A. e Hosoume, Y., Conc ep c~ ao esp ont^ ane a sobr e movimento, Rev. Ens. F s. 7 (1), 37 (1985). [8] Villani, A., P acca, J. L. A., Kishinami, R. I., Hosoume, Y., A nalisando o ensino de f sic a: c ontribui c~ oes de p es- quisas c om enfo ques difer entes, Rev. Ens. F s. 4 (1), 23 (1982). As deci ^ encias ap on tadas an teriormen te t ^ em que ser enfren tadas p elos professores dos p er o dos iniciais, ofe- recendo, ao mesmo temp o, condi c~ oes para que os es- tudan tes p ossam ampliar seus conhecimen tos, sua ca- pacidade de racio c nio, e consolidar os conceitos fun- damen tais. Conhecer a forma de p ensar dos alunos, trabalhar com as concep c~ oes esp on t^ aneas que trazem e planejar estrat egias para reelab or a-las e, p ois, de im- p ort^ ancia fundamen tal para que se p ossam minimizar as diculdades conceituais apresen tadas e assim maxi- mizar o pro cesso de aprendizagem. [9] Costa, S. S. C e Moreira, M. A., O p ap el da mo delagem mental dos enunciados na r esolu c~ ao de pr oblemas em f sic a, Rev. Bras. Ens. F s. 24 (1), 61 (2002). [10] Hestenes, D. e W ells, M., Me chanics Baseline T est, The Ph ys. T eac her 30, 159 (1992). [11] Hestenes, D., W ells, M. VI Considera c~ oes nais T em-se observ ado, ao longo dos anos, uma gradual dimin ui c~ ao na capacidade de uso de ferramen tal ma- tem atico dos alunos que ingressam no curso sup erior. V arios demonstram diculdades, p or exemplo, em re- Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 332 Refer ^ encias e Sw ac khamer, G., F or c e Con- c ept Inventory., The Ph ys. T eac her 30, 141 (1992). [12] Spiegel, M. R., Estat stic a, McGra w-Hill do Brasil, S~ ao P aulo (1974). Agradecimen tos [13] Ed Disy , A. D. N. e Garner, J., Hyp othetic al pr e- classic al e quations of motion, The Ph ys. T eac her 37, 42 (1999). Nossos agradecimen tos aos professores F abio do Prado, Jos e Maria Bec hara e Augusto Martins dos San- tos p elo ap oio, v aliosas discuss~ oes e sugest~ oes, e a to dos os professores que colab oraram na aplica c~ ao do teste. Gostar amos de agradecer tam b em as imp ortan tes su- gest~ oes e comen t arios do arbitro, que nos c hamou a aten c~ ao para uma s erie de p on tos falhos e omiss~ oes pre- sen tes no texto original. [14] W altham, C., Power r e quir ements for r ol lerblading and bicycling, The Ph ys. T eac her 37, 379 (1999). [15] Casp er, B. M., Galile o and the fal l of A ristotle: A c ase of historic al injustic e?, Am. J. Ph ys. 45, 325 (1977). 333 V agner Bernal Barb eta e Issao Y amamoto Anexo – Mechanics Baseline Test a t Anexo Mechanics Baseline Test Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 Considere o diagrama abaixo para responder às duas primeiras questões. O diagrama representa um objeto movendo-se ao longo de uma superfície horizontal. As posições indicadas no diagrama estão separadas por intervalos de tempo iguais. O primeiro ponto indica a posição em que o objeto começou a se mover e o último quando ele voltou a ficar em repouso. 1.- Qual dos gráficos seguintes representa melhor a velocidade do objeto em função do tempo? (f) Não sei responder 2.- Qual dos gráficos seguintes representa melhor a aceleração em função do tempo? (f) Não sei responder ------------------------------------------------------------------------ 3.- No gráfico seguinte é mostrada a velocidade de um objeto em função do tempo. v t (a) v t (b) v t (c) v t (d) v t (e) (b) a t (a) a t (c) a t (d) a t (e) Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 Considere o diagrama abaixo para responder às duas primeiras questões. O diagrama representa um objeto movendo-se ao longo de uma superfície horizontal. As posições indicadas no diagrama estão separadas por intervalos de tempo iguais. O primeiro ponto indica a posição em que o objeto começou a se mover e o último quando ele voltou a ficar em repouso. 1.- Qual dos gráficos seguintes representa melhor a velocidade do objeto em função do tempo? (f) Não sei responder 2.- Qual dos gráficos seguintes representa melhor a aceleração em função do tempo? v t (a) v t (b) v t (c) v t (d) v t (e) (f) Não sei responder 2.- Qual dos gráficos seguintes representa melhor a aceleração em função do tempo? v t (a) v t (b) v t (c) v t (d) v t (e) (e) (c) 2.- Qual dos gráficos seguintes representa melhor a aceleração em função do tempo? a t (f) Não sei responder ------------------------------------------------------------------------ 3.- No gráfico seguinte é mostrada a velocidade de um objeto em função do tempo. (b) a t (a) a t (c) a t (d) a t (e) a t (f) Não sei responder ------------------------------------------------------------------------ 3.- No gráfico seguinte é mostrada a velocidade de um objeto em função do tempo. (b) a t (a) a t (c) a t (d) a t (e) a t (b) a t (a) a t (c) a t (d) a t (e) (b) (e) (a) (c) (f) Não sei responder 3.- No gráfico seguinte é mostrada a velocidade de um objeto em função do tempo. v t v t Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 334 Qual dos gráficos abaixo mostra a relação entre a força resultante e o tempo? (f) Não sei responder ------------------------------------------------------------------------ Baseie-se no gráfico abaixo para responder às três questões seguintes. Este diagrama representa um bloco deslizando ao longo de uma rampa sem atrito. As oito setas numeradas representam direções a serem tomadas como referência para responder às questões. F t (a) F t (b) F t (c) F t (d) F t (e) Qual dos gráficos abaixo mostra a relação entre a força resultante e o tempo? Qual dos gráficos abaixo mostra a relação entre a força resultante e o tempo? Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 der F t (b) F t (c) F t (d) F t (e) (b) (c) (d) (e) Baseie-se no gráfico abaixo para responder às três questões seguintes. Este diagrama representa um bloco deslizando ao longo de uma rampa sem atrito. As oito setas numeradas representam direções a serem tomadas como referência para responder às questões. A B C 4.- A direção da aceleração do bloco, quando este estiver na posição A, é melhor A B C C 4.- A direção da aceleração do bloco, quando este estiver na posição A, é melhor representada por qual das setas no diagrama? (a) 1 (b) 2 (c) 4 (d) 5 (e) 6 ( f) Não sei responder ------------------------------------------------------------------------ 5.- A direção da aceleração do bloco, quando este estiver na posição B, é melhor representada por qual das seguintes setas no diagrama? (a) 1 (b) 3 (c) 5 (d) 7 (e) Nenhuma das setas, pois a aceleração é nula. ( f) Não sei responder ------------------------------------------------------------------------ 6.- A direção da aceleração do bloco (após deixar a rampa) na posição C, é melhor representada por qual das setas no diagrama? (a) 1 (b) 3 (c) 5 (d) 6 (e) Nenhuma das setas, pois a aceleração é zero. ( f) Não sei responder 4.- A direção da aceleração do bloco, quando este estiver na posição A, é melhor representada por qual das setas no diagrama? (a) 1 (b) 2 (c) 4 (d) 5 (e) 6 ( f) Não sei responder ------------------------------------------------------------------------ 5.- A direção da aceleração do bloco, quando este estiver na posição B, é melhor representada por qual das seguintes setas no diagrama? (a) 1 (b) 3 (c) 5 (d) 7 (e) Nenhuma das setas, pois a aceleração é nula. ( f) Não sei responder ------------------------------------------------------------------------ 6.- A direção da aceleração do bloco (após deixar a rampa) na posição C, é melhor representada por qual das setas no diagrama? (a) 1 (b) 3 (c) 5 (d) 6 (e) Nenhuma das setas, pois a aceleração é zero. ( f) Não sei responder 4.- A direção da aceleração do bloco, quando este estiver na posição A, é melhor representada por qual das setas no diagrama? (a) 1 (b) 2 (c) 4 (d) 5 (e) 6 ( f) Não sei responder ------------------------------------------------------------------------ 5.- A direção da aceleração do bloco, quando este estiver na posição B, é melhor representada por qual das seguintes setas no diagrama? Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 (a) 1 (b) 3 (c) 5 (d) 7 (e) Nenhuma das setas, pois a aceleração é nula. ( f) Não sei responder ------------------------------------------------------------------------ 6.- A direção da aceleração do bloco (após deixar a rampa) na posição C, é melhor representada por qual das setas no diagrama? (a) 1 (b) 3 (c) 5 (d) 6 (e) Nenhuma das setas, pois a aceleração é zero. ( f) Não sei responder 4.- A direção da aceleração do bloco, quando este estiver na posição A, é melhor representada l d t di ? 335 V agner Bernal Barb eta e Issao Y amamoto 7.- Uma pessoa empurra um bloco com velocidade constante através de uma superfície horizontal rugosa, aplicando uma força F. As setas no diagrama indicam corretamente as direções, mas não necessariamente as intensidades, das várias forças sobre o bloco. Qual das seguintes relações entre as intensidades das forças W, K, N e F deve ser verdadeira? (a) F = K (b) F = K e N > W (c) F > K e N < W (d) F > K e N = W (e) Nenhuma das alternativas anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 8.- Um pequeno cilindro metálico repousa sobre uma plataforma horizontal circular, a qual gira com velocidade constante como ilustrado no diagrama a seguir v N F W K W 8.- Um pequeno cilindro metálico repousa sobre uma plataforma horizontal circular, a qual gira com velocidade constante como ilustrado no diagrama a seguir. Qual dos seguintes conjuntos de vetores descreve melhor a velocidade, a aceleração e a força resultante agindo sobre o cilindro no ponto indicado no diagrama? ( f) Não sei responder ----------------------------------------------------------------------- 9.- Suponha que o cilindro metálico do problema anterior tenha massa de 0,10 kg e que o coeficiente de atrito estático entre a superfície e o cilindro seja 0,12. Se o cilindro está a 2,0 m do centro da plataforma, qual é aproximadamente a máxima velocidade com que o cilindro pode se mover ao longo de sua trajetória circular sem escorregar sobre a mesa? F v a (a) F v a = 0 (b) F v a = 0 (c) F v a (d) F v a (e) Qual dos seguintes conjuntos de vetores descreve melhor a velocidade, a aceleração e a força resultante agindo sobre o cilindro no ponto indicado no diagrama? Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 ( f) Não sei responder F v a (a) F v a = 0 (b) F v a = 0 (c) F v a (d) F v a (e) (a) (e) 9.- Suponha que o cilindro metálico do problema anterior tenha massa de 0,10 kg e que o coeficiente de atrito estático entre a superfície e o cilindro seja 0,12. Se o cilindro está a 2,0 m do centro da plataforma, qual é aproximadamente a máxima velocidade com que o cilindro pode se mover ao longo de sua trajetória circular sem escorregar sobre a mesa? 9.- Suponha que o cilindro metálico do problema anterior tenha massa de 0,10 kg e que o coeficiente de atrito estático entre a superfície e o cilindro seja 0,12. Se o cilindro está a 2,0 m do centro da plataforma, qual é aproximadamente a máxima velocidade com que o cilindro pode se mover ao longo de sua trajetória circular sem escorregar sobre a mesa? 9.- Suponha que o cilindro metálico do problema anterior tenha massa de 0,10 kg e que o coeficiente de atrito estático entre a superfície e o cilindro seja 0,12. Se o cilindro está a 2,0 m do centro da plataforma, qual é aproximadamente a máxima velocidade com que o cilindro pode se mover ao longo de sua trajetória circular sem escorregar sobre a mesa? 9.- Suponha que o cilindro metálico do problema anterior tenha massa de 0,10 kg e que o coeficiente de atrito estático entre a superfície e o cilindro seja 0,12. Se o cilindro está a 2,0 m do centro da plataforma, qual é aproximadamente a máxima velocidade com que o cilindro pode se mover ao longo de sua trajetória circular sem escorregar sobre a mesa? Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 336 (a) 0 (b) está entre 0,5 m/s e 1,0 m/s (c) 1,0 m/s (d) 1,5 m/s (e) 2,0 m/s ( f) Não sei responder ------------------------------------------------------------------------ 10.- Uma garota deseja escolher um dos escorregadores, abaixo ilustrados, que lhe dê a maior velocidade possível quando atingir o final do escorregador. Desprezando o atrito, qual dos escorregadores ela deverá escolher? (e) Tanto faz, pois a velocidade será a mesma em qualquer um deles. Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 ( f) Não sei responder 5 m 0,5 m (a) (b) (c) (d) (a) 0 (b) está entre 0,5 m/s e 1,0 m/s (c) 1,0 m/s (d) 1,5 m/s (e) 2,0 m/s ( f) Não sei responder ------------------------------------------------------------------------ 10.- Uma garota deseja escolher um dos escorregadores, abaixo ilustrados, que lhe dê a maior velocidade possível quando atingir o final do escorregador. Desprezando o atrito, qual dos escorregadores ela deverá escolher? 5 m 0,5 m (a) (b) (c) (d) (b) (d) (c) (a) (e) Tanto faz, pois a velocidade será a mesma em qualquer um deles. ( f) Não sei responder Consulte o diagrama a seguir para responder às duas questões seguintes: X e Z assinalam a mais alta e Y a mais baixa posições de um menino de 50,0 kg que balança para a direita conforme indicado no diagrama. 11.- Qual é aproximadamente a velocidade do garoto na posição Y? 1,0 m 5,0 m X Z 11.- Qual é aproximadamente a velocidade do garoto na posição Y 11.- Qual é aproximadamente a velocidade do garoto na posição Y? (a) 2,5 m/s (b) 4,5 m/s (c) 20 m/s (d) 0 (e) Nenhuma das anteriores. ( f) Não sei responder (a) 2,5 m/s (b) 4,5 m/s (c) 20 m/s (d) 0 (e) Nenhuma das anteriores. ( f) Não sei responder 337 V agner Bernal Barb eta e Issao Y amamoto 12. Qual é a força tensora no cabo na posição Y? (a) 250 N (b) 525 N (c) 7 x 102 N (d) 1,1 x 103 N (e) Nenhuma das anteriores ( f) Não sei responder ------------------------------------------------------------------------ Considere o diagrama a seguir para responder às duas questões seguintes. Os blocos A e B, cada um com massa de 1,0 kg estão suspensos no teto de um elevador pelas cordas 1 e 2. 13. Qual é a força exercida pela corda 1 sobre o bloco A, quando o elevador está subindo com uma velocidade constante de 2,0 m/s? A B 1 2 A 13. Qual é a força exercida pela corda 1 sobre o bloco A, quando o elevador está subindo com uma velocidade constante de 2,0 m/s? 13. Qual é a força exercida pela corda 1 sobre o bloco A, quando o elevador está subindo com uma velocidade constante de 2,0 m/s? Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 (a) 2 N (b) 10 N (c) 12 N (d) 20 N (e) 22 N ( f) Não sei responder ------------------------------------------------------------------------ 14. Qual é a força exercida pela corda 1 sobre o bloco B quando o elevador está estacionário? (a) 2 N (b) 10 N (c) 12 N (d) 20 N (e) 22 N ( f) Não sei responder ------------------------------------------------------------------------ Basear-se no diagrama a seguir para responder às duas próximas questões. O diagrama indica as trajetórias de duas bolas de aço, P e Q, que colidem. (a) 2 N (b) 10 N (c) 12 N (d) 20 N (e) 22 N ( f) Não sei responder ------------------------------------------------------------------------ 14. Qual é a força exercida pela corda 1 sobre o bloco B quando o elevador está estacionário? (a) 2 N (b) 10 N (c) 12 N (d) 20 N (e) 22 N ( f) Não sei responder ------------------------------------------------------------------------ Basear-se no diagrama a seguir para responder às duas próximas questões. O diagrama indica as trajetórias de duas bolas de aço, P e Q, que colidem. Basear-se no diagrama a seguir para responder às duas próximas questões. Basear-se no diagrama a seguir para responder às duas próximas questões. O diagrama indica as trajetórias de duas bolas de aço, P e Q, que colidem. Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 338 15. Qual das setas representa melhor a direção da variação do momento linear (quantidade de movimento) de cada bola? Q P Q P 15. Qual das setas representa melhor a direção da variação do momento linear (quantidade de movimento) de cada bola? 15. Qual das setas representa melhor a direção da variação do momento linear (quantidade de movimento) de cada bola? ( f) Não sei responder ------------------------------------------------------------------------ P (a) P Q (b) P Q (d) P Q (e) P Q (c) Q ( f) Não sei responder ------------------------------------------------------------------------ P (a) P Q (b) P Q (d) P Q (e) P Q (c) Q (e) 16. Qual das setas representa melhor a direção do impulso aplicado à bola Q pela bola P durante a colisão? 16. Qual das setas representa melhor a direção do impulso aplicado à bola Q pela bola P durante a colisão? ( f) Não sei responder ------------------------------------------------------------------------ (a) (b) (d) (e) (c) (e) (a) 17. Um carro tem uma aceleração máxima de 3,0 m/s2. Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 Qual será sua máxima aceleração quando estiver rebocando um outro carro com o dobro de sua massa? 17. Um carro tem uma aceleração máxima de 3,0 m/s2. Qual será sua máxima aceleração quando estiver rebocando um outro carro com o dobro de sua massa? 17. Um carro tem uma aceleração máxima de 3,0 m/s2. Qual será sua máxima aceleração quando estiver rebocando um outro carro com o dobro de sua massa? (a) 2,5 m/s2 (b) 2,0 m/s2 (c) 1,5 m/s2 (d) 1,0 m/s2 (e) 0,5 m/s2 ( f) Não sei responder ------------------------------------------------------------------------ 18. Uma mulher pesando 6,0 x 102 N está dentro de um elevador que sobe do primeiro para o sexto andar. Quando o elevador se aproxima do sexto andar, ele diminui sua velocidade de 8,0 m/s para 2,0 m/s em 3,0 s. Qual é a força média exercida pelo piso do elevador sobre a mulher durante este intervalo de tempo de 3,0 s? 339 V agner Bernal Barb eta e Issao Y amamoto (a) 120 N (b) 480 N (c) 600 N (d) 720 N (e) 1200 N ( f) Não sei responder ( f) Não sei responder ------------------------------------------------------------------------ 19. O diagrama a seguir representa um bloco que se move sobre uma superfície horizontal e sem atrito, no sentido indicado pela seta tracejada. Uma força constante F (horizontal para a direita), mostrada no diagrama, está agindo sobre o bloco. Para que o objeto fique sujeito a uma força resultante na direção da seta tracejada, em qual das direções indicadas por A, B, C, D, E, deverá estar agindo uma outra força? (a) (A) (b) (B) (c) (C) (d) (D) (e) (E) (f) Não sei responder ------------------------------------------------------------------------ Refira-se ao diagrama seguinte para responder às próximas três questões. O diagrama representa dois objetos sobre uma mesa horizontal sem atrito. O bloco B tem massa quatro vezes maior que o bloco A. Os dois blocos são empurrados, a partir do repouso, por duas forças iguais. 20. Qual dos blocos terá maior energia cinética ao atingir a linha de chegada? (a) O bloco A (b) O bloco B (c) Ambos os blocos terão a mesma energia cinética (d) Faltam informações para determinar (e) Não é possível determinar ( f) Não sei responder A B C D E F Chegada m 4m A B F F 19. Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 O diagrama a seguir representa um bloco que se move sobre uma superfície horizontal sem atrito, no sentido indicado pela seta tracejada. Uma força constante F (horizontal para direita), mostrada no diagrama, está agindo sobre o bloco. Para que o objeto fique sujeito a uma força resultante na direção da seta tracejada, em qual das direções indicadas por A, B C, D, E, deverá estar agindo uma outra força? (a) (A) (b) (B) (c) (C) (d) (D) (e) (E) (f) Não sei responder ------------------------------------------------------------------------ Refira-se ao diagrama seguinte para responder às próximas três questões A B C D E F A Chegada 340 Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 21. Qual dos blocos alcança primeiro a linha de chegada? Para responder às próximas três questões, baseie-se no gráfico da velocidade em função do tempo mostrado a seguir, que representa o movimento de um objeto em uma dimensão. tempo mostrado a seguir, que representa o movimento de um objeto em uma dimensão. 23. Qual é aproximadamente a aceleração média do objeto entre t = 0 s e t = 6,0 s? (a) 3,0 m/s2 (b) 1,5 m/s2 (c) 0,83 m/s2 (d) 0,67 m/ s2 (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 24. Qual o espaço percorrido pelo objeto entre t = 0 e t = 6,0 s? (a) 20 m (b) 8,0 m (c) 6,0 m (d) 1,5 m (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 0 2 6 12 15 10 5 3 5 1 v(m/s) t(s) p g q p j 23. Qual é aproximadamente a aceleração média do objeto entre t = 0 s e t = 6,0 s? (a) 3,0 m/s2 (b) 1,5 m/s2 (c) 0,83 m/s2 (d) 0,67 m/ s2 (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 0 2 6 12 15 10 5 3 5 1 v(m/s) t(s) 23. Qual é aproximadamente a aceleração média do objeto entre t = 0 s e t = 6,0 s? (a) 3 0 m/s2 0 2 6 12 15 10 5 3 5 1 v(m/s) t(s) 0 2 6 12 15 10 5 3 5 1 v(m/s) t(s) 23. Qual é aproximadamente a aceleração média do objeto entre t = 0 s e t = 6,0 s? Revista Brasileira de Ensino de F sica, v ol. 24, no. 3, Setem bro, 2002 Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 (a) 3,0 m/s2 (b) 1,5 m/s2 (c) 0,83 m/s2 (d) 0,67 m/ s2 (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 24. Qual o espaço percorrido pelo objeto entre t = 0 e t = 6,0 s? (a) 20 m (b) 8,0 m (c) 6,0 m (d) 1,5 m (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ 23. Qual é aproximadamente a aceleração média do objeto entre t = 0 s e t = 6,0 s? (a) 3 0 m/s2 341 V agner Bernal Barb eta e Issao Y amamoto 25. Qual é aproximadamente a velocidade média do objeto nos primeiros 6,0 s? 25. Qual é aproximadamente a velocidade média do objeto nos primeiros 6,0 s? (a) 3,3 m/s (b) 3,0 m/s (c) 1,8 m/s (d) 1,3 m/s (e) Nenhuma das anteriores. ( f) Não sei responder ------------------------------------------------------------------------ Refira-se ao diagrama abaixo para responder a questão seguinte. ------------------------------------------------------------------------ Refira-se ao diagrama abaixo para responder a questão seguinte. Refira se ao diagrama abaixo para responder a questão seguinte. A figura representa uma fotografia múltipla de uma bolinha sendo atirada verticalmente para cima. A mola, com a bolinha em cima, foi comprimida até o ponto marcado com X e então foi solta. A bola deixou a mola no ponto marcado com Y, e atingiu a altura máxima no ponto marcado com Z. Z Y X Z Y X A figura representa uma fotografia múltipla de uma bolinha sendo atirada verticalmente para cima. A mola, com a bolinha em cima, foi comprimida até o ponto marcado com X e então foi solta. A bola deixou a mola no ponto marcado com Y, e atingiu a altura máxima no ponto marcado com Z. 26. Assumindo que a resistência do ar é desprezível, podemos afirmar que: 26. Assumindo que a resistência do ar é desprezível, podemos afirmar que: (a) A maior aceleração da bola foi no momento em que atingiu o ponto Y (ainda em contato com a mola). (b) A aceleração da bola foi decrescendo em seu trajeto do ponto Y ao ponto Z. (c) A aceleração da bola foi zero no ponto Z. (d) Todas as respostas acima estão corretas. (e) A aceleração da bola foi a mesma em todos os pontos de sua trajetória, do ponto Y ao Z. Obs.: Quando necessário, utilize aceleração da gravidade g = 10 m/s2 ( f) Não sei responder (a) A maior aceleração da bola foi no momento em que atingiu o ponto Y (ainda em contato com a mola). co a o a) b) A aceleração da bola foi decrescendo em seu trajeto do ponto Y ao ponto Z. ) A l ã d b l f i Z (e) A aceleração da bola foi a mesma em todos os pontos de sua trajetória, do ponto Y ao Z. ( f) Não sei responder
https://openalex.org/W2548317149	https://europepmc.org/articles/pmc5363516?pdf=render	English	null	Genomic profiling is predictive of response to cisplatin treatment but not to PI3K inhibition in bladder cancer patient-derived xenografts	Oncotarget	2,016	cc-by	11,545	Genomic profiling is predictive of response to cisplatin treatment but not to PI3K inhibition in bladder cancer patient-derived xenografts Priority Research Paper Lei Wei1,, Sreenivasulu Chintala2,5,7,, Eric Ciamporcero2,, Swathi Ramakrishnan2,5, May Elbanna2,7, Jianmin Wang1, Qiang Hu1, Sean T. Glenn3, Mitsuko Murakami4, Lu Liu4, Eduardo Cortes Gomez4, Yuchen Sun4, Jacob Conroy4, Kiersten Marie Miles4, Kullappan Malathi6, Sudha Ramaiah6, Anand Anbarasu6, Anna Woloszynska- Read2,5, Candace S. Johnson2,5, Jeffrey Conroy4, Song Liu1, Carl D. Morrison4 and Roberto Pili2,7 Lei Wei1,, Sreenivasulu Chintala2,5,7,, Eric Ciamporcero2,, Swathi Ramakrishnan2,5, May Elbanna2,7, Jianmin Wang1, Qiang Hu1, Sean T. Glenn3, Mitsuko Murakami4, Lu Liu4, Eduardo Cortes Gomez4, Yuchen Sun4, Jacob Conroy4, Kiersten Marie Miles4, Kullappan Malathi6, Sudha Ramaiah6, Anand Anbarasu6, Anna Woloszynska- Read2,5, Candace S. Johnson2,5, Jeffrey Conroy4, Song Liu1, Carl D. Morrison4 and Roberto Pili2,7 Department of Biostatistics & Bioinformatics, Roswell Park Cancer Institute, Buffalo, NY, USA 1 Department of Biostatistics & Bioinformatics, Roswell Park Cancer Institute, Buffalo, NY, USA 1 Department of Biostatistics & Bioinformatics, Roswell Park Cancer Instit 2 Genitourinary Program, Roswell Park Cancer Institute, Buffalo, NY, USA Department of Cancer Genetics, Roswell Park Cancer Institute, Buffalo, NY, USA Center for Personalized Medicine, Roswell Park Cancer Institute, Buffalo, NY, USA Department of Pharmacology and Therapeutics, Roswell Park Cancer Institute, Buffalo, NY, USA 5 Department of Pharmacology and Therapeutics, Roswell Park Cancer Institute, Buffalo, NY, USA 6 Medical & Biological Computing Laboratory, School of Biosciences and Technology, VIT University, Vellore, Tamil Nadu, INDIA 6 Medical & Biological Computing Laboratory, School of Biosciences and Technology, VIT University, Vellore, Tamil Nadu, INDIA 6 Medical & Biological Computing Laboratory, School of Biosciences and Technology, VIT University, Vellore, Tamil Nadu, INDIA 7 Genitourinary Program, Indiana University-Simon Cancer Center, Indianapolis, IN, USA * These authors have contributed equally to this work Correspondence to: Roberto Pili, email: rpili@iu.edu Keywords: urothelial carcinoma, patient-derived xenograft, PI3KCA Received: October 15, 2016 Accepted: October 22, 2016 www.impactjournals.com/oncotarget/ Oncotarget, Vol. 7, No. 47 Genomic profiling is predictive of response to cisplatin treatment but not to PI3K inhibition in bladder cancer patient-derived xenografts Priority Research Paper www.impactjournals.com/oncotarget/ INTRODUCTION mTOR pathways in post-cisplatin bladder cancer has been tested, but has not been associated with improved clinical outcome [14]. Accordingly, more clinically and molecularly relevant models are necessary to better understand the molecular alterations associated with drug response, and to develop more effective personalized therapies for MIBC. Bladder cancer, with ~380,000 new cases per year and 15,000 deaths, stands as the ninth most common cancer worldwide. Histologically, more than 90% of the cases are urothelial carcinoma and at time of diagnosis 75- 85% of tumors are non-muscle invasive cancer (NMIBC). Approximately 60-70% of NMIBC recurs within one year and 10-20% will progress to muscle invasive disease. Muscle-invasive bladder cancer (MIBC) has the worst prognosis, with a five-year survival rate of less than 50%. Treatment options for MIBC remain cisplatin-based regimen [1]. Therefore, there is a need to develop more clinical relevant models to understand the biology and develop effective therapeutic options for patients with MIBC. In this study, we characterized two PDX tumors recently established in our lab by genomic profiling. As previously reported, BLCAb001 is less cisplatin responsive as compared to BLCAb002 [15], and carries specific cisplatin resistance markers, such as a caspase 8 mutation and over expression of the cystine transporter xCT. Genomic analysis also revealed that both BLCAb001 and BLCAb002 present common PIK3CA E542K and E545K driver mutations, respectively. However, the treatment response to the dual PI3K/mTOR inhibitor LY3023414 (LY414) was found to be significantly hampered in BLCAb001, suggesting the presence of alternative pathways. Overall, our data suggest that a comprehensive profiling, rather than solely mutational analysis, may predict response to PI3K/mTOR targeted therapies in bladder cancer. Patient derived tumor xenografts (PDX) have become accepted preclinical models because of their retained original tumor heterogeneity and genetic make- up, suggesting a more reliable drug development tool as compared to tumor cell lines [2, 3]. To date, there is a limited number of established bladder cancer PDX models [4] that are molecularly characterized and available for testing drug resistance and sensitivity. Recent high throughput genomic studies have revealed several gene and pathway alterations associated with MIBC [5, 6], including PI3K/AKT/mTOR and ERK/MEK/ RAS pathways as drivers of bladder cancer progression and potential targets for therapeutic interventions [3]. The genomic landscape in MIBC includes alteration of 9 oncogenes and 23 tumor suppressor genes. ABSTRACT Purpose: Effective systemic therapeutic options are limited for bladder cancer. In this preclinical study we tested whether bladder cancer gene alterations may be predictive of treatment response. Experimental design: We performed genomic profiling of two bladder cancer patient derived tumor xenografts (PDX). We optimized the exome sequence analysis method to overcome the mouse genome interference. Results: We identified a number of somatic mutations, mostly shared by the primary tumors and PDX. In particular, BLCAb001, which is less responsive to cisplatin than BLCAb002, carried non-sense mutations in several genes associated with cisplatin resistance, including MLH1, BRCA2, and CASP8. Furthermore, RNA-Seq analysis revealed the overexpression of cisplatin resistance associated genes such as SLC7A11, TLE4, and IL1A in BLCAb001. Two different PIK3CA mutations, E542K and E545K, were identified in BLCAb001 and BLCAb002, respectively. Thus, we tested whether the genomic profiling was predictive of response to a dual PI3K/mTOR targeting agent, LY3023414. Despite harboring similar PIK3CA mutations, BLCAb001 and BLCAb002 exhibited differential response, both in vitro and in vivo. Sustained target modulation was observed in the sensitive model BLCAb002 but not in BLCAb001, as well as decreased autophagy. Interestingly, computational modelling of mutant structures and affinity binding to PI3K revealed that E542K mutation was associated with weaker drug binding than E545K. Conclusions: Our results suggest that the presence of activating PIK3CA mutations may not necessarily predict in vivo treatment response to PI3K targeted therapies, while specific gene alterations may be predictive for cisplatin response in bladder cancer models and, potentially, in patients as well. www.impactjournals.com/oncotarget Oncotarget 76374 INTRODUCTION The reported top five mutated oncogenes are PIK3CA (9- 20%), FGFR3 (5-20%), ERBB3 (11%), RXRA (9%), and ERBB2 (8%) [5]. Among the tumor suppressors, the top 5 gene alterations include TP53 (24-56%), MLL (27%), ARID1A (25%), KDM6A (24%), and TSC1 (11-16%) [5]. The genetic characterization mutations reported in bladder cancer have contributed to the molecular subtyping of this disease: FGFR3 and TP53 mutations in UroA and UroB cluster [7], FGFR3 mutation in Cluster I [6], FGFR3 and TSC1 mutations in the basal and luminal phenotype [8, 9]. This molecular classification, combined with histopathology analysis, provides the opportunity to develop more effective personalized therapies for bladder cancer patients. Somatic mutations in primary tumors and PDXs The mutations with less than 20X coverage in primary or PDX tumor are highlighted in red. C. Circos plots [61] depic onsynonymous genetic alterations that were: 1) previously reported by ClinVar or COSMIC or 2) novel variant in a Cancer Gene Cen ene or other genes known to be recurrently mutated in bladder cancer. These two categories are distinguished by the color of connec etween gene symbol and chromosome: red = 1), blue = 2). From outer to inner components: gene symbol (red, truncating mutat ncluding nonsense and splice-site SNV, frameshift Indel; orange, alternating mutations including missense SNV and in-frame Ind hromosomes, variant allele frequency (VAF) bars for the corresponding point mutation (range = [0:1], the color of VAF stick indi Figure 1: Somatic mutations in primary tumor and PDX. A. Effects of mouse contamination on somatic mutation calling. Before (“Uncleaned”) and after (“Cleaned”) filtering out mouse contamination, the initial single nucleotide variation (SNV) calls from PDX samples (green) were compared with the matched primary tumor (blue). Top: “Uncleaned”, bottom: “Cleaned”; Left: BLCAb001, right: BLCAb002. The excessive amount of SNV calls in the “Uncleaned” PDX data likely reflects artifacts introduced by mouse contamination. B. Plot: variant allele fraction (VAF) defined as the fraction of reads harboring mutant allele for each mutation in the primary tumor and matched PDX. The mutations with less than 20X coverage in primary or PDX tumor are highlighted in red. C. Circos plots [61] depicting nonsynonymous genetic alterations that were: 1) previously reported by ClinVar or COSMIC or 2) novel variant in a Cancer Gene Census gene or other genes known to be recurrently mutated in bladder cancer. These two categories are distinguished by the color of connection between gene symbol and chromosome: red = 1), blue = 2). From outer to inner components: gene symbol (red, truncating mutations including nonsense and splice-site SNV, frameshift Indel; orange, alternating mutations including missense SNV and in-frame Indel), chromosomes, variant allele frequency (VAF) bars for the corresponding point mutation (range = [0:1], the color of VAF stick indicate coverage: light grey = 0-9X; grey:10-29X; black: > = 30X) in PDX (ring background = green) and the primary tumor (ring background = orange). Figure 1: Somatic mutations in primary tumor and PDX. A. Effects of mouse contamination on somatic mutation calling. Somatic mutations in primary tumors and PDXs We recently established two PDXs, BLCAb001 and BLCA002, from two patients undergoing cystectomy for urothelial carcinoma [15]. Based on the previously reported difference in cisplatin sensitivity between the two models, we decided to perform a genomic profiling of the original tumors and the derived PDXs. Using a high- throughput paired-end sequencing approach, we generated 84 to 330 million of 100-bp reads per sample. For non- PDX samples, over 98% of the reads were successfully mapped to the human reference by using BWA. For PDX samples, the mapping rates were 94.5% and 86.6% with human reference. After mapping to the human and mouse combined reference, the mapping rates for these two PDXs increased to 99.1% and 99.2%. All samples reached the designed goal of 80% of the targeted regions covered with at least 30X coverage (Table S1). Cisplatin based treatment options have improved the survival in bladder cancer. However, patients eventually develop resistance to treatment and disease progression. Several reports have revealed different potential mechanisms responsible for intrinsic and acquired drug resistance including cisplatin binding, metabolism, transport [10], and intracellular sequestration [11, 12]. As a potential marker for cisplatin resistance, differential expression of GSH synthesis regulating the cystine/glutamate exchanger protein, xCT, has also been reported in bladder cancer [13]. In addition, targeting Filtering out mouse contamination was a critical step in order to obtain accurate mutation calls in the PDX samples. In a test run on the unfiltered data, we identified 4,276 and 16,861 SNVs in BLCAb001 and BLCAb002, respectively (Figure 1A). The majority of these SNVs was not identified in the primary tumor and was likely caused by mouse contamination. After filtering out mouse reads, most of these suspicious mutation calls disappeared and the remaining mutations were highly consistent with the matched primary tumor. For BLCAb001, we identified www.impactjournals.com/oncotarget Oncotarget 76375 Figure 1: Somatic mutations in primary tumor and PDX. A. Effects of mouse contamination on somatic mutation calling. Be “Uncleaned”) and after (“Cleaned”) filtering out mouse contamination, the initial single nucleotide variation (SNV) calls from P amples (green) were compared with the matched primary tumor (blue). Top: “Uncleaned”, bottom: “Cleaned”; Left: BLCAb001, ri BLCAb002. The excessive amount of SNV calls in the “Uncleaned” PDX data likely reflects artifacts introduced by mouse contaminat B. Plot: variant allele fraction (VAF) defined as the fraction of reads harboring mutant allele for each mutation in the primary tumor matched PDX. Somatic mutations in primary tumors and PDXs Before (“Uncleaned”) and after (“Cleaned”) filtering out mouse contamination, the initial single nucleotide variation (SNV) calls from PDX samples (green) were compared with the matched primary tumor (blue). Top: “Uncleaned”, bottom: “Cleaned”; Left: BLCAb001, right: BLCAb002. The excessive amount of SNV calls in the “Uncleaned” PDX data likely reflects artifacts introduced by mouse contamination. B. Plot: variant allele fraction (VAF) defined as the fraction of reads harboring mutant allele for each mutation in the primary tumor and matched PDX. The mutations with less than 20X coverage in primary or PDX tumor are highlighted in red. C. Circos plots [61] depicting nonsynonymous genetic alterations that were: 1) previously reported by ClinVar or COSMIC or 2) novel variant in a Cancer Gene Census gene or other genes known to be recurrently mutated in bladder cancer. These two categories are distinguished by the color of connection between gene symbol and chromosome: red = 1), blue = 2). From outer to inner components: gene symbol (red, truncating mutations including nonsense and splice-site SNV, frameshift Indel; orange, alternating mutations including missense SNV and in-frame Indel), chromosomes, variant allele frequency (VAF) bars for the corresponding point mutation (range = [0:1], the color of VAF stick indicate coverage: light grey = 0-9X; grey:10-29X; black: > = 30X) in PDX (ring background = green) and the primary tumor (ring background = orange). www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76376 which were likely to be homozygous in the PDX (Figure 1B). 1,008 SNVs and 5 Indels from the primary and PDX and 1,101 SNVs and 14 Indels from BLCAb002. The identified mutations were then manually reviewed to ensure accuracy. After manual review, there were 919 mutations (917 SNVs and 2 Indels) left in BLCAb001 and 980 mutations (973 SNVs and 7 Indels) left in BLCAb002. i ) Among the identified somatic mutations, 13 alterations were previously found clinically relevant according to ClinVar or mutated in other cancers as summarized by COSMIC, including RS1 R209H, PIK3CA E542K, MLL3 R199, LRIG3 E576K, KLHL3 S410L, FANCD2 L1134V, DNAH7 R1957 (BLCAb001); and PIK3CA E545K, NCSTN S389C, MYD88 S219C, CREBBP W1472C, CFTR R1066C, CDKN2A E69(BLCAb002) (Figure 1C). Most of these mutations were present in both primary and PDXs except for CFTR R1066C, which was only present in the PDX. For other novel nonsynonymous mutations, 57 mutations occurred in Cancer Gene Census genes or other genes known to be recurrently mutated in bladder cancer genes [6, 36, 37], including 10 predicted loss-of-function mutations (BLCAb001: MLL2 Q1361, ARID2 L47fs; BLCAb002: ZFP36L1 F253fs, MYST4 E1398fs, MLL S2663, KRAS E3_splice, KDM6A Q958, ETV4 Q170, ELF3 D223fs, BMPR1A R361). The majority (55/57) of these potentially important mutations The mutation profiles were compared between the primary and PDX to determine similarity. In both cases, majority (91% in BLCAb001 and 82% in BLCAb002) of all mutations were shared by primary and PDX samples (Figure 1B). There also existed smaller numbers of sample-specific mutations (2% primary-unique and 7% PDX-unique in BLCAb001, 3% primary-unique and 15% PDX-unique in BLCAb002), which may reflect the tumor progression from primary to PDX tumors. In BLCAb001, the VAFs of the shared mutations were centered near primary = 0.25 and PDX = 0.5, which may indicate higher tumor purity in PDX than the primary tumor. Similarly, in BLCAb002, most shared mutations were around primary = 0.25 and PDX = 0.5. Additionally, there was a small group of mutations near primary = 0.4 and PDX = 1.0, Figure 2: Histological and molecular representation of BLCAb001 and BLCAb002. A. H&E staining of original cystectomy and PDX tumors. B. Expression of cytokeratine 20 (CK20), CK5/6 in BLCAb001 and BLCAb002. C. RNAseq analysis of luminal and basal phenotype gene signature in BLCAb001 and BLCAb002. Figure 2: Histological and molecular representation of BLCAb001 and BLCAb002. A. Overexpression of xCT is associated with cisplatin resistance All SNVs in one patient were segregated by their presence status in primary and PDX tumors into three groups: “Common”, “Unique to primary” and “Unique to PDX”. We analyzed mutation signature in every group. In all three groups in BLCAb001 and two groups (“Common” and “Unique to primary”) in BLCAb002, the mutation signatures were dominated by C > T transitions and C > G transversions. However, in BLCAb002 “Unique to PDX”, the two most prevalent patterns were C > T transition and C > A transversions. Additionally, this group also had elevated T > A transversions, which were not observed in any other groups (Table S2). Differential response to cisplatin treatment has also been associated with overexpression of the cystine transporter xCT (encoded product of SLC7A11) [10, 13]. Interestingly, BLCAb001 showed higher gene and protein expression of xCT, as compared to BLCAb002 (Figure 3C). RNA-Seq analysis confirmed the overexpression of SLC7A11 in BLCAb001 (Figure S2). Further, immunohistochemical evaluation also showed high expression of xCT in BLCAb001 (Figure 3D). Additionally, we evaluated xCT expression in bladder cancer tumors (n = 21) arranged in a tissue microarray (TMA) by immunohistochemistry. High expression of xCT (Figure 3E) was found in 67 % (14 out of 21) of tumors, including PDX BLCAb001. TCGA data analysis for xCT expression revealed the poorer survival of bladder cancer patients is associated with altered expression (Figure 3F). These results suggest that patients with urothelial cancer showing high expression of xCT may have shorter survival due to reduced response to cisplatin-based regimens. Interestingly, a recent study revealed the epigenetic alteration of microRNA-27A leading to xCT overexpression in bladder cancer [13]. To date, the effect of gene methylation on the expression of xCT has not been investigated. In order to correlate xCT expression and methylation, we determined the SLC7A11 (xCT) methylation status in bladder cancer patients (n = 52) and their matched non-tumor tissues (n = 106) and found significant hypomethylation of SLC7A11 in tumors compared to their matched non-tumor tissues (Figure 3G). Basal and luminal phenotype in BLCAb001 and BLCAb002, respectively BLCAb001 and BLCAb002 were established from cystectomy specimens and maintained the histological features of the original tumors (Figure 2A). Thus, we decided to better characterize the histological phenotypes. Immunohistochemical analysis revealed that BLCAb001 expresses lower levels of the tissue differentiation marker cytokeratin 20 (CK20) (Figure 2B), while had higher expression of cytokeratin CK5/6, as compared to BLCAb002. RNA-Seq analysis confirmed the upregulation of basal phenotype associated genes in BLCAb001 (Figure 2C) and the upregulation of luminal phenotype associated genes in BLCAb002 (Figure 2D). www.impactjournals.com/oncotarget H&E staining of original cystectomy and PDX tumors. B. Expression of cytokeratine 20 (CK20), CK5/6 in BLCAb001 and BLCAb002. C. RNAseq analysis of luminal and basal phenotype gene signature in BLCAb001 and BLCAb002. www.impactjournals.com/oncotarget Oncotarget 76377 were present in both the primary and PDX. Only two mutations, RANBP2 P1380R in BLCAb001, and RYR2 E1859K in BLCAb002, were present in the PDXs but not in the primary tumor (Table S2). ATP7A S1444* and CASP8 Q524. All other mutations were missense SNVs. In support of a genomic signature associated with cisplatin resistance, RNA-Seq analysis revealed overexpression of known cisplatin resistance associated genes in BLCAb001, including NRG1, EGFR, SLC7A11, TLE4, and IL1A (Figure S2) [39]. ATP7A S1444 and CASP8 Q524*. All other mutations were missense SNVs. In support of a genomic signature associated with cisplatin resistance, RNA-Seq analysis revealed overexpression of known cisplatin resistance associated genes in BLCAb001, including NRG1, EGFR, SLC7A11, TLE4, and IL1A (Figure S2) [39]. Alteration of cisplatin resistance associated genes in primary tumors and PDXs To evaluate the association between xCT overexpression and cisplatin resistance, human bladder cancer T24 cisplatin resistance cells were generated upon in vitro long drug exposure. As shown in Figure S4A, cisplatin resistant T24 cells presented a > 1-fold increase in IC50 as compared to the parental cells. The decreased sensitivity to cisplatin was associated with an increase in xCT expression (Figure S4B). CD44 expression has been reported to be involved in cisplatin resistance [40] and its isoform CD44V6 has a potential role in stabilization of xCT [41]. Thus, we investigated the expression of CD44 standard (CD44s) and its variants, CD44v6 and CD44v8, in the T24 and UMUC3 models. By q-PCR we observed that, while there was no difference in CD44s and CD44v8 gene expression between parental and cisplatin-resistant cell lines, there was a CD44v6 overexpression in cisplatin BLCAb001 and BLCAb002 had similar mutation burden, but exhibited a different response to increased doses of cisplatin (Figure 3A). We confirmed this differential response in primary tumor cells isolated from these tumors. Cells isolated from BLCAb001 were found to be less responsive to cisplatin treatment than BLCAb002 cells [15]. To determine whether specific genomic alterations were responsible for the observed different cisplatin sensitivity between BLCAb001 and BLCAb002, we compared the somatically genomic alterations in known cisplatin resistance associated genes [38]. Eight of such genes were found to be mutated (Figure 3B), six in BLCAb001 and two in BLCAb002. Two of these mutations were predicted to cause loss-of-function: www.impactjournals.com/oncotarget Oncotarget 76378 resistant T24 and UMUC3 cells as compared to the parental cells, confirming a possible role for xCT (Figure S4C). Next, we tested whether the use of a putative xCT inhibitor, sulfasalazine (SASP), was able to affect the response to cisplatin. Combination of SASP enhanced response to cisplatin in both cisplatin sensitive and cisplatin resistant cells. (Figure S4E and S4F). In addition, SASP effectively inhibited the colony formation of the cisplatin resistant cells, suggesting a role for lysosome function/biogenesis in the survival of xCT overexpressing T24 cells (Figure S4F). In addition, we performed our proliferation assay using the special RPMI media without cysteine, methionine, and glutamate. Supplementation of cysteine, but not methionine or glutamate, to the medium rescued the normal growth of T24 cells (Figures S5A and S5B). Taken together, these results suggest a potential role for targeting the cystine transporter to modulate cisplatin sensitivity in bladder cancer. PI3KCA mutation status does not correlate with response to a PI3K/mTOR dual inhibitor Logrank test p-value 0.0312. G. SLC7A11 gene methylation in human bladder cancer patients primary tumors (n = 22) d h h d i ( 106) Thi h h h f DNA h h l i i hi h SLC7A11 l E h ponse to cisplatin treatment in PDXs and associated mutational status and gene expression. Figure 3: Differential response to cisplatin treatment in PDXs and associated mutational status and gene expression. A. Cisplatin treatment (5 mg/kg and 10 mg/kg) had a differential effect on BLCAb001 and BLCAb002 in vivo growth. B. Mutational status of cisplatin resistance associated genes in BLCAb001 and BLCAb002. C. RT-PCR and Western blot analysis of SLC7A11 (xCT), a cisplatin resistance associated gene in BLCAb001 and BLCAb002 PDXs. D. Immunohistochemical evaluation of xCT in BLCAb001 and BLCAb002. E. Percent tumors with high and low xCT expression. Immunoscore determined as high (80-100% cells positive) and low (2-20% cells positive) [35]. F. TCGA data analysis of cBioportal showing the poor survival of patients with alteration (upregulation) of SLC7A11 (xCT). Logrank test p-value 0.0312. G. SLC7A11 gene methylation in human bladder cancer patients primary tumors (n = 22) and the matched non-tumor tissues (n = 106). This graph shows the presence of DNA hypomethylation within the SLC7A11 locus. Each data point represents the average methylation of 2 CG sites most significantly hypomethylated in tumors when compared to normal bladder tissues. Figure 3: Differential response to cisplatin treatment in PDXs and associated mutational status and gene expression. A. Cisplatin treatment (5 mg/kg and 10 mg/kg) had a differential effect on BLCAb001 and BLCAb002 in vivo growth. B. Mutational status of cisplatin resistance associated genes in BLCAb001 and BLCAb002. C. RT-PCR and Western blot analysis of SLC7A11 (xCT), a cisplatin resistance associated gene in BLCAb001 and BLCAb002 PDXs. D. Immunohistochemical evaluation of xCT in BLCAb001 and BLCAb002. E. Percent tumors with high and low xCT expression. Immunoscore determined as high (80-100% cells positive) and low (2-20% cells positive) [35]. F. TCGA data analysis of cBioportal showing the poor survival of patients with alteration (upregulation) of SLC7A11 (xCT). Logrank test p-value 0.0312. G. SLC7A11 gene methylation in human bladder cancer patients primary tumors (n = 22) and the matched non-tumor tissues (n = 106). This graph shows the presence of DNA hypomethylation within the SLC7A11 locus. PI3KCA mutation status does not correlate with response to a PI3K/mTOR dual inhibitor Each data point represents the average methylation of 2 CG sites most significantly hypomethylated in tumors when compared to normal bladder tissues. PI3KCA mutation status does not correlate with response to a PI3K/mTOR dual inhibitor The RNA-Seq analysis of BLCAb001 and BLCAb002 revealed that both patients harbored a PIK3CA hotspot mutation but on different residues: E542K and E545K in BLCAb001 and BLCAb002, respectively (Figure 4A). These mutations were present in the original Figure 3: Differential response to cisplatin treatment in PDXs and associated mutational status and gene expression. A. Cisplatin treatment (5 mg/kg and 10 mg/kg) had a differential effect on BLCAb001 and BLCAb002 in vivo growth. B. Mutational status of cisplatin resistance associated genes in BLCAb001 and BLCAb002. C. RT-PCR and Western blot analysis of SLC7A11 (xCT), a cisplatin resistance associated gene in BLCAb001 and BLCAb002 PDXs. D. Immunohistochemical evaluation of xCT in BLCAb001 and BLCAb002. E. Percent tumors with high and low xCT expression. Immunoscore determined as high (80-100% cells positive) and low (2-20% cells positive) [35]. F. TCGA data analysis of cBioportal showing the poor survival of patients with alteration (upregulation) of SLC7A11 (xCT). Logrank test p-value 0.0312. G. SLC7A11 gene methylation in human bladder cancer patients primary tumors (n = 22) and the matched non-tumor tissues (n = 106). This graph shows the presence of DNA hypomethylation within the SLC7A11 locus. Each data point represents the average methylation of 2 CG sites most significantly hypomethylated in tumors when compared to normal bladder tissues. ure 3: Differential response to cisplatin treatment in PDXs and associated mutational status a Figure 3: Differential response to cisplatin treatment in PDXs and associated mutational status and gene expression. A. Cisplatin treatment (5 mg/kg and 10 mg/kg) had a differential effect on BLCAb001 and BLCAb002 in vivo growth. B. Mutational status of cisplatin resistance associated genes in BLCAb001 and BLCAb002. C. RT-PCR and Western blot analysis of SLC7A11 (xCT), a Figure 3: Differential response to cisplatin treatment in PDXs and associated mutational status and gene expression. A. Cisplatin treatment (5 mg/kg and 10 mg/kg) had a differential effect on BLCAb001 and BLCAb002 in vivo growth. B. Mutational status of cisplatin resistance associated genes in BLCAb001 and BLCAb002. C. RT-PCR and Western blot analysis of SLC7A11 (xCT), a cisplatin resistance associated gene in BLCAb001 and BLCAb002 PDXs. D. Immunohistochemical evaluation of xCT in BLCAb001 and BLCAb002. E. Percent tumors with high and low xCT expression. Immunoscore determined as high (80-100% cells positive) and low (2-20% cells positive) [35]. F. TCGA data analysis of cBioportal showing the poor survival of patients with alteration (upregulation) of SLC7A11 (xCT). www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76379 tumors and in the established PDXs, and were confirmed in the derived cells lines (Table S2). Since PI3KCA helical hotspot mutations (E542K and E545K) are common (25%) in bladder cancers [42], we were interested to test whether this mutational status was associated to response to a targeted therapy. In vivo treatment with the dual PI3K/mTOR inhibitor LY3023414 (LY414) demonstrated a significant anti-tumor effect only in BLCAb002 and not in BLCAb001 (Figure 4A and 4B). The inhibition of tumor growth by LY414 was associated with inhibition of phosphorylated AKT (ser473) and phosphorylated S6 kinase (evaluated after 24h of drug treatment) in BLCAb002 but not in BLCAb001 (Figure 4C). While the initial tumor size is about 20% lower in BLCAb002 tumors compared to BLCAb001 tumors, the final size of the tumor at the end of treatment is about 3 fold (150%) lower in BLCAb002 tumors. We found there was no tumor size difference with the treatment in BLCAb001 tumors. We have repeated the experiment with same initial tumor size of BLCAb001 and BLCAb002 tumors and found significant tumor growth inhibition in BLACb002 tumors but not in BLCAb001 tumors (Figure S10). These results suggest that some additional molecular alterations or the cross talk of cisplatin resistant genes may contribute to the relative resistance of BLCAb001 to LY414 despite the presence of an activating PI3KCA mutation. Indeed, our RNAseq analysis revealed the overexpression of known genes associated with resistance to AKT inhibition such as phosphatidylinositol-4-phosphate 3-kinase C2 domain- containing gamma polypeptide (PIK3C2G), insulin receptor substrate 1 (IRS1) and serum- and glucocorticoid- regulated kinase (SGK1) (Figure S6) [43-45]. Thus, we evaluated the cytotoxic effects of LY414 on cells isolated from the two PDXs. Similarly to the in vivo results, BLCAb002 cells were found to be more sensitive to LY414 as compared to BLCAb001 (Figure 4D). The IC50 for LY414 was 143.94 nM and 45.06 nM in BLCAb001 and BLCAb002 cells, respectively. Figure 4: Differential response of PIK3CA mutated BLCAb001 and BLCAb002 to LY414. A. PI3KCA mutations in BLCAb001 (E452K) and BLCAb002 (E545K). Color codes- Yellow orange color indicates P13K alpha helical domain. B. Anti-tumor effects of LY414 on PDXs BLCAb001 and BLCAb002 tumors (upper panels), tumor weights at the end of treatment (EOT) (lower panels). C.. Western blot analysis of the effect of LY414 treatment on p-AKT, p-mTOR, and pS6 in BLCAb001 and BLCAb002 PDXs. D. www.impactjournals.com/oncotarget Effect of LY414 on BLCAb001 and BLCAb002 cells isolated from PDX cultured in DMEM. Statistical analysis of ANOVA was performed to determine the significance p≤ 0.001. E. Western blot analysis of the effect of LY414 treatment on p-AKT and p-S6K in BLCAb001 and BLCAb002 derived cells cultured in DMEM. Figure 4: Differential response of PIK3CA mutated BLCAb001 and BLCAb002 to LY414. A. PI3KCA mutations in BLCAb001 (E452K) and BLCAb002 (E545K). Color codes- Yellow orange color indicates P13K alpha helical domain. B. Anti-tumor effects of LY414 on PDXs BLCAb001 and BLCAb002 tumors (upper panels), tumor weights at the end of treatment (EOT) (lower panels). C.. Western blot analysis of the effect of LY414 treatment on p-AKT, p-mTOR, and pS6 in BLCAb001 and BLCAb002 PDXs. D. Effect of LY414 on BLCAb001 and BLCAb002 cells isolated from PDX cultured in DMEM. Statistical analysis of ANOVA was performed to determine the significance p≤ 0.001. E. Western blot analysis of the effect of LY414 treatment on p-AKT and p-S6K in BLCAb001 and BLCAb002 derived cells cultured in DMEM. www.impactjournals.com/oncotarget Oncotarget 76380 Interestingly, the target of LY414 p-AKT (ser473) was inhibited in both cells at an early time point, but only in BLCAb002 at 24 and 48 hrs (Figure 4E). Same results were observed for thr308 phosphorylation site (Figure S8A). We observed p-S6 inhibition in both the tumor cell models, also at later time points. When we tested selected agent for either mTOR (RAD001) or AKT (MK2206) inhibition, we observed target inhibition in both BLCAb001 and BLCAb002 (Figure S8B). The inhibition was maintained also at 24 and 48 hrs (Figure S8C). In order to confirm that BLCAB001 tumors in general are less sensitive to the drugs targeting P13K/mTOR inhibitors compared to BLCAb002 tumors, we have determined the p-AKT, downstream makers after treating the cells with BEZ235 (250 nM) and BKM120 (500nM). As the results shown with LY414 treatment, we found pronounced inhibition of p-AKT in BLCAb002 cells which are more sensitive to P13K/mTOR inhibitors compared to BLCAb001 cells (Figure 9SA, 9SB). Interestingly, when we combined LY414 with bromodomain inhibitor JQ, we observed inhibition of p-AKT in BLCAb001 cells, which was not observed with LY414 alone (Figure 9S C). Additional studies are warranted to investigate how JQ combination optimized the P13K/mTOR dual inhibitor LY414. www.impactjournals.com/oncotarget Interestingly, the differential cytotoxicity effect of LY414 between the two models was not observed when the cells were tested in insulin enriched RPMI F media (Figure S7A). The IC50 was 212.9 nM and 208 nM for BLCA001 and BLCAb002, respectively (Figure S7B). These results suggest that enriched media with insulin may compensate the AKT pathway inhibition. Oncotarget 76381 impactjournals com/oncotarget gure 5: Differential expression and modulation of autophagy in BLCAb001 and BLCAb002. A. RNA-Seq analysis data wing the differential expression of autophagy genes in BLCAb001 and BLCAb002. B. Differential effect of LY414 on autophagy and optosis markers in BLCAb001 and BLCAb002 derived cells. Figure 5: Differential expression and modulation of autophagy in BLCAb001 and BLCAb002. A. RNA-Seq analysis data Figure 5: Differential expression and modulation of autophagy in BLCAb001 and BLCAb002. A. RNA-Seq analysis data showing the differential expression of autophagy genes in BLCAb001 and BLCAb002. B. Differential effect of LY414 on autophagy and apoptosis markers in BLCAb001 and BLCAb002 derived cells. Figure 5: Differential expression and modulation of autophagy in BLCAb001 and BLCAb002. A. RNA-Seq analysis data showing the differential expression of autophagy genes in BLCAb001 and BLCAb002. B. Differential effect of LY414 on autophagy and apoptosis markers in BLCAb001 and BLCAb002 derived cells. E542 mutation is associated with weaker binding of LY414 to PI3K Finally, we examined whether modulation of autophagy could be responsible for the observed differential response to LY414 between the two models. Thus, in our RNA-Seq analysis we observed overexpression of BNIP3 and BCL11A, and downregulation of BCL2L14, ULK2, RAB11FIP4 and BAG1 in BLCAb001 tumors as compared to BLCAb002 (Figure 5A). Western blot analysis confirmed the persistence of autophagy protein expression in BLCAb001 cells treated with LY414, but downregulation of Beclin-1 and LC3B I/II in BLCAb002 cells, associated with cleaved PARP. Based on the different biological effect elicited by LY414, we investigated whether the mutation type could contribute by affecting the affinity of the compound to the substrate. Protein-ligand binding orientation was analyzed with the Sybyl-X 2.0 program. After docking, the best binding conformation of the compound with the protein was selected based on C score values. C score is the scoring system used to rank the binding affinity of ligands Figure 6: Effect of E542K and E545K mutations on LY414 binding to PI3K. A. Docking results for BLCAb001 and BLCAb002 with LY414 (Kcal/mol). The Crash score reveals inappropriate penetration into the binding site. The Polar score identifies the Region of the ligand. The D score defines the charge and van der waals interactions between the protein and ligand. The PMF score defines the Helmholtz free energies for protein-ligand atom pairs interactions. The Gscore identifies the Hydrogen bonding, ligand-protein and internal ligand-ligand energies. The Chem score identifies the points for hydrogen bonding, lipophilic contact and rotational entropy, along with an intercept term. Finally, the C-score is the consensus scoring which uses multiple types of scoring functions to rank the overall affinity of ligands. A higher C-score value denotes a greater binding affinity. The increased negative values of Crash score, Chem score, D-score, G-score and Potential Mean Force scores (PMF) indicate the high binding energy between the protein-ligand complexes. The low values indicate the least binding affinity of the compound towards the target. The binding energies are expressed in Kcal/mol unit. B. Binding mode in BLCAb001 (E542K mutation) and BLCAb002 (E545K mutation). Figure 6: Effect of E542K and E545K mutations on LY414 binding to PI3K. A. Docking results for BLCAb001 and BLCAb002 with LY414 (Kcal/mol). The Crash score reveals inappropriate penetration into the binding site. The Polar score identifies the Region of the ligand. The D score defines the charge and van der waals interactions between the protein and ligand. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76381 E542 mutation is associated with weaker binding of LY414 to PI3K The PMF score defines the Helmholtz free energies for protein-ligand atom pairs interactions. The Gscore identifies the Hydrogen bonding, ligand-protein and internal ligand-ligand energies. The Chem score identifies the points for hydrogen bonding, lipophilic contact and rotational entropy, along with an intercept term. Finally, the C-score is the consensus scoring which uses multiple types of scoring functions to rank the overall affinity of ligands. A higher C-score value denotes a greater binding affinity. The increased negative values of Crash score, Chem score, D-score, G-score and Potential Mean Force scores (PMF) indicate the high binding energy between the protein-ligand complexes. The low values indicate the least binding affinity of the compound towards the target. The binding energies are expressed in Kcal/mol unit. B. Binding mode in BLCAb001 (E542K mutation) and BLCAb002 (E545K mutation). Figure 6: Effect of E542K and E545K mutations on LY414 binding to PI3K. A. Docking results for BLCAb001 and BLCAb002 with LY414 (Kcal/mol). The Crash score reveals inappropriate penetration into the binding site. The Polar score identifies the Region of the ligand. The D score defines the charge and van der waals interactions between the protein and ligand. The PMF score defines the Helmholtz free energies for protein-ligand atom pairs interactions. The Gscore identifies the Hydrogen bonding, ligand-protein and internal ligand-ligand energies. The Chem score identifies the points for hydrogen bonding, lipophilic contact and rotational entropy, along with an intercept term. Finally, the C-score is the consensus scoring which uses multiple types of scoring functions to rank the overall affinity of ligands. A higher C-score value denotes a greater binding affinity. The increased negative values of Crash score, Chem score, D-score, G-score and Potential Mean Force scores (PMF) indicate the high binding energy between the protein-ligand complexes. The low values indicate the least binding affinity of the compound towards the target. The binding energies are expressed in Kcal/mol unit. B. Binding mode in BLCAb001 (E542K mutation) and BLCAb002 (E545K mutation). www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76382 and is based on different scoring functions, including crash score, polar score, D score, PMF score, G score and Chem score, ranging between 0 (low ) to 5 (high). The binding parameters for BLCAb001 and LY-3023414 are presented in Figure 6A. The BLCAb001 and LY-3023414 binding had a C-score value of 2.21. The LY-3023414 interacts with residues Ser 312 and Cys 838 as depicted in Figure 6B. BLCAb002 and LY-3023414 complex has a C-score value of 4.23 and it interacts with Glu 237 and Lys 678 residues of BLCAb002 (Figure 6B). The binding of LY- 414 to the wild type protein had a C score of 2.51 which is similar as BLCAb001 (data not shown). These results suggest that the E545K mutation observed in BLCAb002 may induce conformational changes in the PI3K protein that could lead to greater binding of LY414 and a stronger inhibitory effect. Lysine is a very dynamic amino acid in the proteins, being oxidized by lysyl oxidase (LOX) an enzyme whose functional significance on tumor progression and metastasis is known [46, 47]. Further, oxidation of specific position lysine molecule opens the protein-protein interactions that influence the tissue remodeling and drug response. Since the E545 mutation we found in BLCAb002 has a lysine at 545 position that may be involved in formation of drug binding pocket and more LY414 drug may bind to PI3K and inhibit the protein. original tumor and the derived PDXs as reported also in other tumor types, and support the clinical relevance of utilizing bladder cancer PDXs for drug development. Cisplatin resistance is one of the major challenges for the management of bladder cancer. The resistance mechanisms may be intrinsic or acquired [5, 49]. Several mechanisms have been proposed, including reduction of intracellular drug accumulation by either reduced uptake [38] or increased efflux by transmembrane pumps, such as the multidrug resistance 1 (MDR1)/p-glycoprotein and ATP binding cassette (ABC) protein [49]. Furthermore, overexpression of MDR1 has been shown to enhance DNA damage repair and to reduce induction of apoptosis [50]. In addition, solute carrier importers CTR1, the SLCs, AQP2, AQP9 and endocytic recycling compartment have been reported as cisplatin resistance regulators [51]. Upregulation of novel genes, including LUM, DCN, PDE3B, PDGF-C, NRG1, PKD2, IL1A have been shown to be overexpressed in cisplatin resistant cell lines [39]. www.impactjournals.com/oncotarget In our cisplatin resistant PDX model, BLCAb001, we observed genomic alterations of several cisplatin resistance associated genes, such as Casp8, SLC7A11, TLE4, and IL1A. We have also observed overexpression of the membrane bound cystine/glutamate exchanger xCT, a lysosome regulatory protein and its stabilizer CD44v6 in cisplatin resistant PDX and cells. Recent studies have reported the association of the cystine transporter xCT with cisplatin resistance in bladder cancer [13], and its potential prognostic role in hepatocellular carcinoma [52]. Drug repurposing of sulfasalazine, a xCT inhibitor, is currently in clinical testing [53]. Taken together, our findings suggest a role for xCT in the resistant phenotype and provide not only an additional potential predictive marker to be incorporated in a composite molecular signature for cisplatin resistance, but also a rational target for therapeutic interventions in patients with bladder cancer. i Whole exome sequencing (WES) Individual exome capture of each DNA sample was carried out using the SureSelectXT Reagent kit as per manufacturer’s recommendation (Agilent Technologies, Inc.). 3 µg of genomic DNA was fragmented to a size range of 150-200 bp followed by end repair, adaptor ligation, and low cycle (5) PCR. Libraries were purified and validated for appropriate size (225-275 bp) on a 2100 Bioanalyzer High Sensitivity DNA chip (Agilent Technologies, Inc.). 750 ng of purified library was then hybridized to the SureSelectXT Human All Exon 50 Mb library for 18 hours at 65°C. The captured regions were then bound to Streptavidin T1 magnetic beads (Life Technologies, Inc.) and washed to remove any non-specific bound products. Eluted library underwent a second 11 cycle PCR amplification using Herculase II Fusion Polymerase (Agilent Technologies, Inc.) to add sample specific barcodes necessary for multiplexing. Final libraries were purified, validated for size by a BioAnalyzer (250-350 bp), and quantitated using KAPA qPCR. Individual libraries were pooled (3-plex) in equimolar 2 nM final concentration. Each pool was normalized to 10 pM, loaded and clustered to individual lanes of a HiSeq Flow Cell using an Illumina cBot (TruSeq PE Cluster Kit v3), followed by 2 x 101 PE sequencing on a HiSeq2000 sequencer according to the manufacturer’s recommended protocol (Illumina Inc.). The presence of two similar but distinct hot-spot mutations in BLCAb001 and BLCAb002 lead us to explore the possibility that the differential mutational status could have an impact on the affinity for LY414. Interestingly, the docking results suggest that there might be a significant difference in the binding of the compound with the two mutants. Simulation of BLCAb002 and LY414 interaction suggest a higher binding affinity than the BLCAb001 and LY414 complex, as indicated by the differences in C scores. Thus, when we compared the binding affinity of LY414 with the mutants (BLCAb001 and BLCAb002) and the wild type, the best binding affinity for the compound was observed for BLCAb002. This observation may provide an additional mechanism responsible for the differential effect of LY414 in the two bladder cancer PDXs and raises the question whether different mutations may have different drug binding affinities and, consequently, different drug sensitivity. We can speculate that the difference in binding affinity due to conformational changes may be also in part responsible for the lack of sustained in vitro target modulation by LY414 observed in BLCAb001. DISCUSSION Precision medicine aimed to develop more tailored and effective cancer treatments is currently based on genomic profiling of either tumor tissues or liquid biopsies. The clinical challenge is whether the molecular findings are predictive of response to targeted therapies in cancer patients. Our preclinical observations suggest that, at least in bladder cancer, a composite genetic signature may be more predictive of response to cisplatin than a specific mutational status to a PI3K inhibitor. Genomic profiling of our PDX models showed two canonical PIK3CA mutations, which activate the PI3K/ AKT/mTOR signaling pathway. BLCAb001 possesses a PIK3CA activating mutation (E542K), which is less common than E545K present in BLCAb002 [54, 55]). Thus, we expected both tumors to respond to a targeted PI3K inhibitor [56, 57]). To our surprise, LY414 treatment did not significantly inhibit either the in vitro or in vivo growth of BLCAb001, and did not induce sustained target modulation, as compared to the BLCAb002 model. Several potential mechanisms may be responsible for this differential response. Our RNA-Seq data analysis revealed overexpression of several genes, including SGK1, IRS1, and PIK3C2G in BLCAb001 (Figure S6), which are known resistance markers for P13K/AKT targeted agents [43-45]. Interestingly, when we cultured BLCAb001 cells in enriched RPMI F medium with growth factors and insulin we were able, in part, to restore sensitivity to LY414 (Figure S7). Thus, we speculate that the in vitro Patient-derived xenografts (PDXs), along with genetically engineered mouse models (GEMMs), are becoming an important tool for drug development in several tumor types, including urothelial carcinoma [48]. In our exosome sequencing study, the PDXs maintained the same genetic make-up as the original tumors. In each case, 80-90% of all somatic mutations were present in both the primary and PDX tumors. There was only a small fraction (7% and 15% in BLCAb001 and BLCAb002, respectively) of PDX-specific mutations, and even a smaller number (2% and 3% in BLCAb001 and BLCAb002, respectively) of mutations present only in the primary tumor. These small differences may reflect either the addition or loss of mutations during tumor progression from primary to PDXs. Alternatively, some of these genetic differences may be explained by tumor heterogeneity. Overall, our results in bladder cancer models confirm the genomic similarity between the www.impactjournals.com/oncotarget Oncotarget 76383 presence of insulin may artificially hyper-activate PI3K signaling in BLCAb001 cells that is responsible for increased glycolysis and aldolase mobilization leading to increased vulnerability [58]. DISCUSSION As shown in the in vitro experiments, we observed LY414 induced inhibition of p-AKT, both at the ser473 and the thr308 phosphorylation sites, in BLCAb001 only at early time of exposure (30-90 minutes), but not at 24 and 48 hrs. If we used a selected AKT inhibitor we did not observed this difference. Thus, we hypothesize that in BLCAb001, but not in BLCAb002, inhibition of PI3K may induce a feed-back upregulation of receptor tyrosine kinases and, consequently, re- induction of pAKT and sustained autophagy, as reported in other tumor systems [59]. Recent studies have shown activation of autophagy as one of the critical molecular alterations that limits the anti-tumor effects of PI3K/ mTOR inhibitors [60]. Our RNAseq data showed BNIP3 and BCL11A upregulation, and ULK2, BAG1, RAB11FIP4 and BCLL14 downregulation in BLCAb001 tumors, mirroring the higher levels of LC3B I/II by Western blot analysis (Figure 5A and 5B). Interestingly, the treatment of sulfasalazine, an inducer of autophagic cell death, enhanced the efficacy of cisplatin resistant BLCAb001 cells. Taken together, these observations suggest that the autophagy may contribute to the observed resistance LY414 in BLCAb001 cells. identification of PIK3CA mutations through genomic profiling may not necessarily predict response to PI3K targeted therapies in mouse models and, likewise, in patients with urothelial carcinoma. Several “by-pass” mechanisms may influence treatment response. In our PDX models, we observed that a composite genetic and molecular signature may be more likely associated with cisplatin sensitivity/response. Additional molecular studies are warranted to identify specific gene mutations/ alterations responsible for the differential response to treatment in these PDX models. The development and validation of genomic signatures will help the clinical implementation of effective targeted therapies for bladder cancer. MATERIALS AND METHODS The RPCI Institutional Review Board gave approval for this study. Patients consented to remnant tissue procurement for next generation sequencing of their samples. Sequencing was performed according to a RPCI IRB approved investigator initiated protocol. Whole exome sequencing (WES) To our knowledge, there are no prior reports showing potential differential effects of E542K and E545K mutations on PI3K drug binding. These findings might provide a rationale for focusing the development of LY414 specifically for bladder cancer patients with a E545K mutation. Additional experimental validation studies are required to confirm these findings. In conclusion, our results suggest that the www.impactjournals.com/oncotarget Oncotarget 76384 and variant allele fraction (VAF). and variant allele fraction (VAF). RNA-Seq analysis To filter out reads caused by mouse stromal contamination in PDX, all reads from the PDXs were run through an in silico approach to determine the species of origin. More specifically, we first created a combined reference sequence containing the sequences of all chromosomes in the NCBI genome assemblies of human (GRCh37) and mouse (GRCm38), and then aligned reads from PDX to the combined reference sequence using BWA. Only reads classified of human but not mouse origin were kept in downstream analyses. Afterwards, standard somatic mutation calling was performed on PDX with the matched normal as described previously. For testing purpose, we also performed somatic mutation calling on the uncleaned PDX data to evaluate the effect of mouse contamination on the somatic mutation calling. Raw reads that passed quality filter from Illumina RTA were first pre-processed by using 1) FASTQC for sequencing base quality control and 2) cutadapt to remove adapter sequences if applicable. Those reads were then mapped to the latest mouse reference genome (mm10) and ENSEMBLE annotation database using Tophat [20] or STAR [21]. A second round of QC using RSeQC [22] was applied to mapped bam files to identify potential RNA-Seq library preparation problems. From the mapping results, reads that matched a single unique location in the genome were kept, allowing up to two mismatches for further analysis. The number of reads aligning to each gene were calculated using HTSeq [23]. Differentially expressed genes were identified using DESeq2 [24], a variance-analysis package developed to infer the statically significant difference in RNASeq data. A biological hypothesis was also tested using a generalized linear model implemented in DESeq2 by construct corresponding contrasts. Multiple testing corrections were applied. The list of differentially expressed genes (DEGs) were analyzed for enriched Gene Ontology and/or KEGG pathway term with the GAGE [25] Bioconductor package. GSAASeqSP [26]was also applied for pathway analysis that utilizes p-values from all genes instead of only DEGs. Sanger validation High quality paired-end reads passing Illumina RTA filter were aligned to the NCBI human reference genome (GRCh37) using BWA [16]. PCR duplicated reads were marked using Picard [17]. Putative SNVs and Indels were identified by running variation detection module of Bambino [18]. All putative SNVs were further filtered based on a standard set of criteria to remove the following common types of false calls: (1) the alternative allele was present in the matching normal sample and the contingency between the tumor and normal samples was not statistically significant; (2) the mutant alleles were only present in one stand and the strand bias was statistically significant; (3) the putative mutation occurred at a site with systematically reduced base quality scores; (4) the reads harboring the mutant allele were associated with poor mapping quality. Ambiguous calls were manually inspected to ensure accuracy. Putative indels were evaluated by a re-alignment process to filter out potential false calls introduced by unapparent germline events, mapping artifacts and homopolymer. All mutations were annotated using ANNOVAR [19] with NCBI RefSeq database. PCR amplicons targeting the PIK3CA and CASP8 regions were generated with gene specific primers (Table S3 and S4) using a touchdown PCR protocol with the following parameters: 94°C for 15 min, followed by 45 cycles of 94°C for 20 seconds, 68°C initial annealing for 30 seconds (followed by 1°C reduction of temperate per cycle to a final annealing temperature of 58°C for remaining 35 cycles), and 72°C for 1 min. Amplicons were purified using the QIAquick® PCR Purification Kit (QIAGEN Inc., Valencia, CA) as per manufacturer instructions. Ten microliter aliquots for each sample were run on a 2% agarose gel for 1 hour at 100V to confirm the correct amplified length (~250 bp). The products were tagged using Big Dye Terminator v3.1 Master Mix Kit (Life Technologies™, Carlsbad, CA) as per manufacturer instructions, and purified over hydrated Sephadex-G50 from Sigma-Aldrich (Sigma-Aldrich®, St. Louis, MO), in Multiscreen HV Plates (Thermo Fisher Scientific Inc., Waltham, MA). The eluted samples were placed on a 3130xl ABI Prism Genetic Analyzer and run on the default settings for 50 cm Array using POP-7 Polymer. The data was analyzed with Sequencing Analysis 5.2 software (Life Technologies™, Carlsbad, CA). Comparing mutations in xenograft and primary tumors All unique somatic mutations identified from the primary tumor or PDX were re-visited in both the primary tumor and matched PDX BAM files. The number of mutant and non-mutant reads at the site of each mutation in all BAM files were extracted using Mutation Reads Extractor (manuscript in submission) to calculate coverage Quantitative RT-PCR BLCAb001 and BLCAb002 cells were isolated from primary tumors of urothelial cell carcinoma of bladder and were originally authenticated by chromosome karyotyping [15]. Cells were cultured using enriched F-medium supplemented with ROCK inhibitor and insulin growth factor as described previously [33]. These cells were recently ( < 6 months) confirmed to be of human origin with the detection of the human specific Alu gene by RT- PCR. Human bladder cancer cells lines, T24 and UMUC3, were purchased from ATCC. No authentication was done. The cells were cultured in RPMI-1640 medium with 10% FBS. Cisplatin resistant ( > 10 fold) cells, T24-Cis and UMUC3-Cis, were generated by continuous treatment of cisplatin (gradual increase) for approximately 5-months. Gene expression at the mRNA level was determined by performing quantitative RT-PCR (qRT-PCR) as described earlier [34]. Briefly, RNA was isolated from tumor tissue and cells using Trizol reagent and prepared cDNA using the high efficiency cDNA synthesis kit (Life Technologies). Gene specific primers were utilized to determine expression levels of genes with SYBR Green PCR Master Mix (Life Technologies) with CFX96 Touch Real-Time PCR Detection system (Bio-Rad). β-actin was used as an internal control. Normalized fold expression of genes was determined using the CFX Manager Software. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76385 Western blot analysis The experimental structures for mutants BLCAb001 and BLCAb002 Phosphoinositide 3-kinase (PI3K) were not available and hence the mutant structures were modelled using the MODELLER 9.16 [27] using the the crystal structure of Pi3K alpha lipid kinase (PDB ID: 4YKN as template [28]. The constructed models were validated through PROCHECK [29]and ProSA [30] servers. The ligand structure (LY-3023414) was retrieved from the TOXNET’s Chem-IDplus database [31]. Sybyl-X 2.0 (Tripos international, USA) [32] was used for the present study. Protein extracts were isolated from tumor tissue using a polytron homogenizer and cells were extracted by sonication in lysis buffer. Forty micrograms of protein were separated by gel electrophoresis followed by transfer on to a PVDF membrane. Western blot analysis was performed using the primary antibodies for xCT (1:500 dilution, Abcam), mTOR (1:1000 dilution, Cell Signaling), p-mTOR (1:1000 dilution, Cell Signaling); Akt (1:1000 dilution, Cell Signaling) p-Akt (1:1000 dilution, Cell Signaling), p-PRAS40 (1:1000 dilution) p-S6 (1:1000 dilution, Cell Signaling), β-actin or GAPDH were used as loading controls. Colony formation assay Bladder cancer cells (3 x 102) of parental and cisplatin resistant T24 and UMUC3 cell lines were seeded in plates and allowed to grow overnight. Cells were treated with cisplatin, xCT inhibitor sulfasalazine alone and in combination with cisplatin for 24 to 48 h. Medium was removed and the cells were rinsed with PBS, fresh medium was added, and cells were allowed to grow for 3 to 4 weeks. Cells were then fixed and stained with methylene blue, photomicrographs were captured, and Modelling of mutant structures and affinity binding to PI3K colonies were counted. Tumor cell growth inhibition Two patient derived tumor xenografts (PDXs), BLCAb001 and BLCAb002, were generated using fresh primary tumors from bladder cancer patients as described by us earlier [15]. For evaluation of drug efficacy, small pieces of PDXs were surgically transplanted into SCID mice subcutaneously and allowed to establish. When the tumors reached approximately 100-200 mm3, mice were randomized into groups of 5-8 mice and treated with vehicle, LY414 (5 mg/kg BID by oral gavage), or cisplatin 5 mg/kg or 10 mg/kg weekly by intraperitoneal injection). Tumor sizes were blindly measured weekly along with the body weight of mice. At the end of the treatment, tumors were collected, weighed, and processed for formalin fixation and small pieces were frozen for protein extraction. Tumor cell growth inhibitory effect of cisplatin, sulfasalazine alone and in combination, was evaluated by SRB assay [34]. Effect of PI3K/mTOR/AKT inhibitors, LY414, RAD001, and MK2206, was determined by MTT cell proliferation assay kit as described by the manufacturer (ThermoFisher Scientific). LY414 was kindly provided by Eli Lilly and Company. The other compounds were purchased (Selleckchem, Houston TX). GRANT SUPPORT This study is supported by National Cancer Institute of National Institutes of Health grant P30 CA016056 (CSJ) and a grant provided by generous research donation from the DiVita family. 9. Damrauer JS, Hoadley KA, Chism DD, Fan C, Tiganelli CJ, Wobker SE, Yeh JJ, Milowsky MI, Iyer G, Parker JS and Kim WY. Intrinsic subtypes of high-grade bladder cancer reflect the hallmarks of breast cancer biology. Proceedings of the National Academy of Sciences of the United States of America. 2014; 111:3110-3115. Editorial note 11. Safaei R, Katano K, Larson BJ, Samimi G, Holzer AK, Naerdemann W, Tomioka M, Goodman M and Howell SB. Intracellular localization and trafficking of fluorescein- labeled cisplatin in human ovarian carcinoma cells. Clinical cancer research. 2005; 11:756-767. This paper has been accepted based in part on peer- review conducted by another journal and the authors’ response and revisions as well as expedited peer-review in Oncotarget. 12. Safaei R, Larson BJ, Cheng TC, Gibson MA, Otani S, Naerdemann W and Howell SB. Abnormal lysosomal trafficking and enhanced exosomal export of cisplatin in drug-resistant human ovarian carcinoma cells. Molecular cancer therapeutics. 2005; 4:1595-1604. Immunohistochemical evaluation We determined the expression of various molecular markers in the original cystectomy tumor samples and www.impactjournals.com/oncotarget Oncotarget 76386 538. their PDXs BLCAb001 and BLCAb002 along with human bladder cancer tumors arranged in tissue microarray (TMA). Standard immunohistochemical protocols were followed as previously described by us earlier [15, 35]. Briefly, formalin fixed paraffin embedded tumors were used to prepare sections (5 µm), deparaffinized followed by rehydration and antigen unmasking in sodium citrate buffer (pH 6.0). Primary antibodies xCT (1:400 dilution), AKT, p-AKT, mTOR, p-mTOR, and cytokeratin 5/6/20, were used for overnight incubation at 4oC, followed by their respective horseradish-conjugated secondary antibody for 1hr. Photomicrographs were captured using a Zeiss Axio (Peabody, MA) microscope. 4. Pan CX, Zhang H, Tepper CG, Lin TY, Davis RR, Keck J, Ghosh PM, Gill P, Airhart S, Bult C, Gandara DR, Liu E and de Vere White RW. Development and Characterization of Bladder Cancer Patient-Derived Xenografts for Molecularly Guided Targeted Therapy. PLoS One. 2015; 10:e0134346. 5. Knowles MA and Hurst CD. Molecular biology of bladder cancer: new insights into pathogenesis and clinical diversity. Nature reviews Cancer. 2015; 15:25-41. 6. Cancer Genome Atlas Research N. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature. 2014; 507:315-322. 7. Sjodahl G, Lauss M, Lovgren K, Chebil G, Gudjonsson S, Veerla S, Patschan O, Aine M, Ferno M, Ringner M, Mansson W, Liedberg F, Lindgren D and Hoglund M. A molecular taxonomy for urothelial carcinoma. Clinical cancer research. 2012; 18:3377-3386. ACKNOWLEDGMENTS We would like to acknowledge the RPCI Bioinformatics Shared Resource, Data Bank and BioRepository, Genomics Shared Resource, and Pathology Resource Network the CCSG Shared Resources for their support. 8. Choi W, Porten S, Kim S, Willis D, Plimack ER, Hoffman- Censits J, Roth B, Cheng T, Tran M, Lee IL, Melquist J, Bondaruk J, Majewski T, Zhang S, Pretzsch S, Baggerly K, et al. Identification of distinct basal and luminal subtypes of muscle-invasive bladder cancer with different sensitivities to frontline chemotherapy. Cancer cell. 2014; 25:152-165. CONFLICTS OF INTEREST 10. Drayton RM and Catto JW. Molecular mechanisms of cisplatin resistance in bladder cancer. Expert Rev Anticancer Ther. 2012; 12:271-281. The authors declare no conflicts of interest. REFERENCES Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R and Genome Project Data Processing S. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25:2078-2079. 32. Jain AN. Surflex: fully automatic flexible molecular docking using a molecular similarity-based search engine. J Med Chem. 2003; 46:499-511. 18. Edmonson MN, Zhang J, Yan C, Finney RP, Meerzaman DM and Buetow KH. Bambino: a variant detector and alignment viewer for next-generation sequencing data in the SAM/BAM format. Bioinformatics. 2011; 27:865-866. 33. Liu X, Ory V, Chapman S, Yuan H, Albanese C, Kallakury B, Timofeeva OA, Nealon C, Dakic A, Simic V, Haddad BR, Rhim JS, Dritschilo A, Riegel A, McBride A and Schlegel R. ROCK inhibitor and feeder cells induce the conditional reprogramming of epithelial cells. The American journal of pathology. 2012; 180:599-607. 19. Wang K, Li M and Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010; 38:e164. 34. Chintala S, Najrana T, Toth K, Cao S, Durrani FA, Pili R and Rustum YM. Prolyl hydroxylase 2 dependent and Von-Hippel-Lindau independent degradation of Hypoxia- inducible factor 1 and 2 alpha by selenium in clear cell renal cell carcinoma leads to tumor growth inhibition. BMC Cancer. 2012; 12:293. 20. Trapnell C, Pachter L and Salzberg SL. TopHat: discovering splice junctions with RNA-Seq. Bioinformatics. 2009; 25:1105-1111. 21. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M and Gingeras TR. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29:15-21. 35. Wang J, Papanicolau-Sengos A, Chintala S, Wei L, Liu B, Hu Q, Miles KM, Conroy JM, Glenn ST, Costantini M, Magi-Galluzzi C, Signoretti S, Choueiri T, Gallucci M, Sentinelli S, Fazio VM, et al. Collecting duct carcinoma of the kidney is associated with CDKN2A deletion and SLC family gene up-regulation. Oncotarget. 2016;7(21):29901- 15. doi: 10.18632/oncotarget.9093. 22. Wang L, Wang S and Li W. RSeQC: quality control of RNA-seq experiments. Bioinformatics. 2012; 28:2184- 2185. 23. Anders S, Pyl PT and Huber W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015; 31:166-169. 36. Morrison CD, Liu P, Woloszynska-Read A, Zhang J, Luo W, Qin M, Bshara W, Conroy JM, Sabatini L, Vedell P, Xiong D, Liu S, Wang J, Shen H, Li Y, Omilian AR, et al. Whole-genome sequencing identifies genomic heterogeneity at a nucleotide and chromosomal level in bladder cancer. REFERENCES 1. Soloway MS. Bladder cancer: Lack of progress in bladder cancer—what are the obstacles? Nat Rev Urol. 2013; 10:5- 6. 13. Drayton RM, Dudziec E, Peter S, Bertz S, Hartmann A, Bryant HE and Catto JW. Reduced expression of miRNA- 27a modulates cisplatin resistance in bladder cancer by targeting the cystine/glutamate exchanger SLC7A11. Clinical cancer research. 2014; 20:1990-2000. 2. Bernardo C, Costa C, Amaro T, Goncalves M, Lopes P, Freitas R, Gartner F, Amado F, Ferreira JA and Santos L. Patient-derived sialyl-Tn-positive invasive bladder cancer xenografts in nude mice: an exploratory model study. Anticancer research. 2014; 34:735-744. 2. Bernardo C, Costa C, Amaro T, Goncalves M, Lopes P, Freitas R, Gartner F, Amado F, Ferreira JA and Santos L. Patient-derived sialyl-Tn-positive invasive bladder cancer xenografts in nude mice: an exploratory model study. Anticancer research. 2014; 34:735-744. 14. Milowsky MI, Iyer G, Regazzi AM, Al-Ahmadie H, Gerst SR, Ostrovnaya I, Gellert LL, Kaplan R, Garcia-Grossman IR, Pendse D, Balar AV, Flaherty AM, Trout A, Solit DB and Bajorin DF. Phase II study of everolimus in metastatic urothelial cancer. BJU international. 2013; 112:462-470. 3. Cirone P, Andresen CJ, Eswaraka JR, Lappin PB and Bagi CM. Patient-derived xenografts reveal limits to PI3K/ mTOR- and MEK-mediated inhibition of bladder cancer. Cancer chemotherapy and pharmacology. 2014; 73:525- 3. Cirone P, Andresen CJ, Eswaraka JR, Lappin PB and Bagi CM. Patient-derived xenografts reveal limits to PI3K/ mTOR- and MEK-mediated inhibition of bladder cancer. Cancer chemotherapy and pharmacology. 2014; 73:525- 15. Ciamporcero E, Shen H, Ramakrishnan S, Yu Ku S, www.impactjournals.com/oncotarget Oncotarget 76387 Chintala S, Shen L, Adelaiye R, Miles KM, Ullio C, Pizzimenti S, Daga M, Azabdaftari G, Attwood K, Johnson C, Zhang J, Barrera G, et al. YAP activation protects urothelial cell carcinoma from treatment-induced DNA damage. Oncogene. 2015. hintala S, Shen L, Adelaiye R, Miles KM, Ullio C Proteins. 1992; 12:345-364. 30. Wiederstein M and Sippl MJ. ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res. 2007; 35:W407- 410. 16. Li H and Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009; 25:1754-1760. 31. Hochstein C, Arnesen S and Goshorn J. Environmental health and toxicology resources of the United States National Library of Medicine. Med Ref Serv Q. 2007; 26:21-45. 17. REFERENCES Proceedings of the National Academy of Sciences of the United States of America. 2014; 111:E672- 681. 24. Love MI, Huber W and Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome biology. 2014; 15:550. 25. Luo W, Friedman MS, Shedden K, Hankenson KD and Woolf PJ. GAGE: generally applicable gene set enrichment for pathway analysis. BMC bioinformatics. 2009; 10:161. 26. Xiong Q, Mukherjee S and Furey TS. GSAASeqSP: a toolset for gene set association analysis of RNA-Seq data. Scientific reports. 2014; 4:6347. 37. Cazier JB, Rao SR, McLean CM, Walker AK, Wright BJ, Jaeger EE, Kartsonaki C, Marsden L, Yau C, Camps C, Kaisaki P, Oxford-Illumina WGSC, Taylor J, Catto JW, Tomlinson IP, Kiltie AE, et al. Whole-genome sequencing of bladder cancers reveals somatic CDKN1A mutations and clinicopathological associations with mutation burden. Nature communications. 2014; 5:3756. 27. Fiser A and Sali A. Modeller: generation and refinement of homology-based protein structure models. Methods Enzymol. 2003; 374:461-491. 28. Yang H, Medeiros PF, Raha K, Elkins P, Lind KE, Lehr R, Adams ND, Burgess JL, Schmidt SJ, Knight SD, Auger KR, Schaber MD, Franklin GJ, Ding Y, DeLorey JL, Centrella PA, et al. Discovery of a Potent Class of PI3Kalpha Inhibitors with Unique Binding Mode via Encoded Library Technology (ELT). ACS Med Chem Lett. 2015; 6:531-536. 38. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, Castedo M and Kroemer G. Molecular mechanisms of cisplatin resistance. Oncogene. 2012; 31:1869-1883. 39. Yamano Y, Uzawa K, Saito K, Nakashima D, Kasamatsu A, Koike H, Kouzu Y, Shinozuka K, Nakatani K, Negoro K, Fujita S and Tanzawa H. Identification of cisplatin- resistance related genes in head and neck squamous cell 29. Morris AL, MacArthur MW, Hutchinson EG and Thornton JM. Stereochemical quality of protein structure coordinates. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76388 52. Kinoshita H, Okabe H, Beppu T, Chikamoto A, Hayashi H, Imai K, Mima K, Nakagawa S, Ishimoto T, Miyake K, Yokoyama N, Ishiko T and Baba H. Cystine/glutamic acid transporter is a novel marker for predicting poor survival in patients with hepatocellular carcinoma. Oncol Rep. 2013; 29:685-689. carcinoma. Int J Cancer. 2010; 126:437-449. 40. Cortes-Dericks L, Froment L, Boesch R, Schmid RA and Karoubi G. Cisplatin-resistant cells in malignant pleural mesothelioma cell lines show ALDH(high)CD44(+) phenotype and sphere-forming capacity. BMC Cancer. 2014; 14:304. 53. Shitara K, Doi T, Nagano O, Imamura CK, Ozeki T, Ishii Y, Tsuchihashi K, Takahashi S, Nakajima TE, Hironaka S, Fukutani M, Hasegawa H, Nomura S, Sato A, Einaga Y, Kuwata T, et al. Dose-escalation study for the targeting of CD44v cancer stem cells by sulfasalazine in patients with advanced gastric cancer (EPOC1205). Gastric cancer. 2016. 41. Ishimoto T, Nagano O, Yae T, Tamada M, Motohara T, Oshima H, Oshima M, Ikeda T, Asaba R, Yagi H, Masuko T, Shimizu T, Ishikawa T, Kai K, Takahashi E, Imamura Y, et al. CD44 variant regulates redox status in cancer cells by stabilizing the xCT subunit of system xc(-) and thereby promotes tumor growth. Cancer cell. 2011; 19:387-400. 54. Dickstein RJ, Nitti G, Dinney CP, Davies BR, Kamat AM and McConkey DJ. Autophagy limits the cytotoxic effects of the AKT inhibitor AZ7328 in human bladder cancer cells. Cancer biology & therapy. 2012; 13:1325-1338. 42. Platt FM, Hurst CD, Taylor CF, Gregory WM, Harnden P and Knowles MA. Spectrum of phosphatidylinositol 3-kinase pathway gene alterations in bladder cancer. Clinical cancer research. 2009; 15:6008-6017. 55. Knowles MA, Platt FM, Ross RL and Hurst CD. Phosphatidylinositol 3-kinase (PI3K) pathway activation in bladder cancer. Cancer Metastasis Rev. 2009; 28:305-316. 43. Martini M, Ciraolo E, Gulluni F and Hirsch E. Targeting PI3K in Cancer: Any Good News? Front Oncol. 2013; 3:108. 56. Brown KK and Toker A. The phosphoinositide 3-kinase pathway and therapy resistance in cancer. F1000Prime Rep. 2015; 7:13. 44. McCubrey JA, Steelman LS, Chappell WH, Abrams SL, Franklin RA, Montalto G, Cervello M, Libra M, Candido S, Malaponte G, Mazzarino MC, Fagone P, Nicoletti F, Basecke J, Mijatovic S, Maksimovic-Ivanic D, et al. Ras/ Raf/MEK/ERK and PI3K/PTEN/Akt/mTOR cascade inhibitors: how mutations can result in therapy resistance and how to overcome resistance. Oncotarget. 2012; 3:1068- 1111. doi: 10.18632/oncotarget.659. 57. Dienstmann R, Rodon J, Serra V and Tabernero J. www.impactjournals.com/oncotarget Picking the point of inhibition: a comparative review of PI3K/AKT/ mTOR pathway inhibitors. Molecular cancer therapeutics. 2014; 13:1021-1031. 58. Hu H, Juvekar A, Lyssiotis CA, Lien EC, Albeck JG, Oh D, Varma G, Hung YP, Ullas S, Lauring J, Seth P, Lundquist MR, Tolan DR, Grant AK, Needleman DJ, Asara JM, et al. Phosphoinositide 3-Kinase Regulates Glycolysis through Mobilization of Aldolase from the Actin Cytoskeleton. Cell. 2016; 164:433-446. 45. Sommer EM, Dry H, Cross D, Guichard S, Davies BR and Alessi DR. Elevated SGK1 predicts resistance of breast cancer cells to Akt inhibitors. Biochem J. 2013; 452:499- 508. 46. Lucero HA and Kagan HM. Lysyl oxidase: an oxidative enzyme and effector of cell function. Cell Mol Life Sci. 2006; 63:2304-2316. 59. Stratikopoulos EE, Dendy M, Szabolcs M, Khaykin AJ, Lefebvre C, Zhou MM and Parsons R. Kinase and BET Inhibitors Together Clamp Inhibition of PI3K Signaling and Overcome Resistance to Therapy. Cancer cell. 2015; 27:837-851. 47. Nilsson M, Adamo H, Bergh A and Halin Bergstrom S. Inhibition of Lysyl Oxidase and Lysyl Oxidase-Like Enzymes Has Tumour-Promoting and Tumour-Suppressing Roles in Experimental Prostate Cancer. Scientific reports. 2016; 6:19608. 60. Echeverry N, Ziltener G, Barbone D, Weder W, Stahel RA, Broaddus VC and Felley-Bosco E. Inhibition of autophagy sensitizes malignant pleural mesothelioma cells to dual PI3K/mTOR inhibitors. Cell Death Dis. 2015; 6:e1757. 48. Richmond A and Su Y. Mouse xenograft models vs GEM models for human cancer therapeutics. Dis Model Mech. 2008; 1:78-82. 61. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, Jones SJ and Marra MA. Circos: an information aesthetic for comparative genomics. Genome research. 2009; 19:1639-1645. 49. Dasari S and Tchounwou PB. Cisplatin in cancer therapy: molecular mechanisms of action. Eur J Pharmacol. 2014; 740:364-378. 50. Ohmichi M, Hayakawa J, Tasaka K, Kurachi H and Murata Y. Mechanisms of platinum drug resistance. Trends Pharmacol Sci. 2005; 26:113-116. 51. Shen DW, Pouliot LM, Hall MD and Gottesman MM. Cisplatin resistance: a cellular self-defense mechanism resulting from multiple epigenetic and genetic changes. Pharmacol Rev. 2012; 64:706-721. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 76389
https://openalex.org/W3165299396	https://hal.science/hal-03321921/file/Rosenbaum%20et%20al.%20Front%20Immunol%202021.pdf	English	null	Vaccine Inoculation Route Modulates Early Immunity and Consequently Antigen-Specific Immune Response	Social Science Research Network	2,020	cc-by	15,446	Vaccine Inoculation Route Modulates Early Immunity and Consequently Antigen-Specific Immune Response Pierre Rosenbaum, Nicolas Tchitchek, Candie Joly, André Rodriguez Pozo, Lev Stimmer, Sébastien Langlois, Hakim Hocini, Leslie Gosse, David Pejoski, Antonio Cosma, et al. To cite this version: Pierre Rosenbaum, Nicolas Tchitchek, Candie Joly, André Rodriguez Pozo, Lev Stimmer, et al.. Vac- cine Inoculation Route Modulates Early Immunity and Consequently Antigen-Specific Immune Re- sponse. Frontiers in Immunology, 2021, 12, pp.645210. 10.3389/fimmu.2021.645210. hal-03321921 Vaccine Inoculation Route Modulates Early Immunity and Consequently Antigen-Speciﬁc Immune Response Pierre Rosenbaum 1,2, Nicolas Tchitchek 1,2, Candie Joly 1,2, Andre´ Rodriguez Pozo 1,2, Lev Stimmer 3,4, Se´bastien Langlois 1,2, Hakim Hocini 2,5, Leslie Gosse 1,2, David Pejoski 1,2, Antonio Cosma 1,2, Anne-Sophie Beignon 1,2, Nathalie Dereuddre-Bosquet 1,2, Yves Levy 2,5, Roger Le Grand 1,2 and Fre´de´ric Martinon 1,2* 1 UMR1184 IMVA-HB, IDMIT Department, Universite´ Paris-Saclay – INSERM U1184 – CEA, Fontenay-aux-Roses, France, 2 Vaccine Research Institute, Henri Mondor Hospital, Cre´ teil, France, 3 INSERM, U1169, Kremlin-Biceˆ tre, France, 4 CEA – INSERM, MIRCen, UMS27, Fontenay-aux-Roses, France, 5 INSERM, U955, Team 16, Clinical and Infectious Diseases Department, Hospital Henri Mondor, University of Paris East, Cre´ teil, France Keywords: vaccine, innate & adaptive immune response, modiﬁed vaccinia Ankara (MVA), non human primates, administration routes, mass cytometry (CyTOF) Edited by: Jose´ Mordoh, IIBBA-CONICET Leloir Institute Foundation, Argentina Vaccination is one of the most efﬁcient public healthcare measures to ﬁght infectious diseases. Nevertheless, the immune mechanisms induced in vivo by vaccination are still unclear. The route of administration, an important vaccination parameter, can substantially modify the quality of the response. How the route of administration affects the generation and proﬁle of immune responses is of major interest. Here, we aimed to extensively characterize the proﬁles of the innate and adaptive response to vaccination induced after intradermal, subcutaneous, or intramuscular administration with a modiﬁed vaccinia virus Ankara model vaccine in non-human primates. The adaptive response following subcutaneous immunization was clearly different from that following intradermal or intramuscular immunization. The subcutaneous route induced a higher level of neutralizing antibodies than the intradermal and intramuscular vaccination routes. In contrast, polyfunctional CD8+ T-cell responses were preferentially induced after intradermal or intramuscular injection. We observed the same dichotomy when analyzing the early molecular and cellular immune events, highlighting the recruitment of cell populations, such as CD8+ T lymphocytes and myeloid-derived suppressive cells, and the activation of key immunomodulatory gene pathways. These results demonstrate that the quality of the vaccine response induced by an attenuated vaccine is shaped by early and subtle modiﬁcations of the innate immune response. In this immunization context, the route of administration must be tailored to the desired type of protective immune response. This will be achieved through systems vaccinology and mathematical modeling, which will be critical for predicting the efﬁcacy of the vaccination route for personalized medicine. Reviewed by: Sylvie Fournel, Universite´ de Strasbourg, France Stasya Zarling, Walter Reed Army Institute of Research, United States Correspondence: Fre´ de´ ric Martinon frederic.martinon@cea.fr Specialty section: This article was submitted to Vaccines and Molecular Therapeutics, a section of the journal Frontiers in Immunology Received: 22 December 2020 Accepted: 06 April 2021 Published: 20 April 2021 Edited by: Jose´ Mordoh, IIBBA-CONICET Leloir Institute Foundation, Argentina HAL Id: hal-03321921 https://hal.science/hal-03321921v1 Submitted on 18 Aug 2021 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. ORIGINAL RESEARCH published: 20 April 2021 doi: 10.3389/fimmu.2021.645210 Experiment Model and Subject Details In Vivo Animal Studies Vaccination is one of the most effective public health approach to prevent infectious diseases. However, for many vaccines, the mechanisms responsible for suitable immunity are still poorly understood. In particular, vaccination routes are known to affect the quality of the vaccine response (1–3), although the underlying molecular and cellular mechanisms at play at the injection site and in the draining lymphoid tissues are yet to be elucidated. Better knowledge of the mechanisms triggered during the early response are needed to aid the design of more efﬁcient vaccines. Thirty-ﬁveadultmalecynomolgusmacaques(Macacafascicularis), each weighing between 4 and 6 kg, imported from Mauritius, were housed in the animal facilities of the CEA (Fontenay-aux-Roses, France). The macaques were handled in accordance with national regulations (Commissariat à l’Energie Atomique et aux Energies Alternatives – Permit Number A 92-32-02), in compliance with the Standards for the Human Care and Use of Laboratory Animals of theOfﬁceforLaboratoryAnimalWelfare(OLAW)undertheOfﬁce for Laboratory Animal Welfare Assurance number A5826-01 and the European Directive (2010/63, recommendation N°9). This project received the authorization numbers 12-013 and 0201501281731916 (APAF1S#170).02 from the French Research Ministry, in accordance with French ethics policies. Data from animals used for adaptive response through subcutaneous route weresharedincommonwitharelatedstudy(22).Interventionsand follow-up of the animals (clinical examination, temperature, weight) were carried out by veterinarians and zootechnician staff of the “Animal Science and Welfare” core facility of the IDMIT infrastructure. Injections, tissue biopsies, and blood samples were performed after anesthesia with ketamine (10 mg/kg, Imalgen®, Rhone-Mérieux, Lyon, France) or tiletamin/zolazepam (5 mg/kg, Zolétil®, Virbac, Caros, France) for muscle biopsies. Intramuscular (IM) injection is the most commonly used administration route, along with the subcutaneous (SC) route, due to their easy access and safety. However, muscle tissue has a low density of immune cells (4) and is thus not expected to be an optimal site for immunization (5). Therefore, the efﬁcacy of the IM route requires a rapid diffusion of the vaccine antigens to draining lymph nodes (6). In contrast, skin tissue contains a broad array of antigen-presenting cells (APCs), which are likely to affect the magnitude, duration, and orientation of antigen- speciﬁc immune memory, and represents a potentially superior vaccine delivery site despite receiving less attention than classical administration routes due to practicality (7). Dose-sparing of up to ten times, without compromising immunogenicity, can be achieved via intradermal (ID) vaccination versus the SC and IM routes (8–10). In murine models, Mohanan et al. Citation: Rosenbaum P, Tchitchek N, Joly C, Rodriguez Pozo A, Stimmer L, Langlois S, Hocini H, Gosse L, Pejoski D, Cosma A, Beignon A-S, Dereuddre-Bosquet N, Levy Y, Le Grand R and Martinon F (2021) Vaccine Inoculation Route Modulates Early Immunity and Consequently Antigen-Speciﬁc Immune Response. Front. Immunol. 12:645210. doi: 10.3389/fimmu.2021.645210 April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 1 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. Experiment Model and Subject Details In Vivo Animal Studies (11) showed that the administration route strongly inﬂuences the Th1 response, with higher IgG2a titers and IFNg production favored by the IM and ID routes over the SC route. Frontiers in Immunology \| www.frontiersin.org Experimental Design Three groups of four NHPs were included in the study of adaptive response. Randomization was performed based on their weight and MHC haplotypes. The experiments and data analysis were performed following protocols which were previously validated during in-house studies (23). Animals were immunized two times with MVA at a 58-day interval. Blood samples were collected at baseline and at days 8, 14, 28, and 57 post-prime and days 8, 14, 28, 57, and 104 (month 4) post-boost three additional groups of six NHPs, of equivalent average weight and MHC proﬁle, were included to study the MVA-induced innate response. We evaluated the systemic response with blood sampling at baseline and at 3 h, 6 h, and 24 h after MVA immunization. An additional group of ﬁve cynomolgus macaques was used for studying the injection site after IM administration of MVA. y Here, we determined how the administration routes can affect the magnitude and quality of the antigen speciﬁc memory response in cynomolgus macaques, used as an animal model to study human vaccines. Indeed, non-human primates (NHP) and human immune systems are highly similar, allowing the direct use of human vaccines in highly predictive NHP immunization models without the need of experimental adaptation, as well as many cross-reactive reagents (12, 13). We compared vaccine- induced innate and adaptive responses following administration via the SC, IM, and ID routes after prime (ﬁrst immunization) and boost (a second injection two months after the prime), at the site of injection and at systemic level. We used the Modiﬁed Vaccinia virus Ankara (MVA), a living, attenuated, non- replicative vaccinia virus already safely used in humans as a 2nd and 3rd generation smallpox vaccine, as a model vaccine because of its capacity to elicit both strong B- and T-cell speciﬁc responses (14). This high immunogenicity combined with good safety has led MVA to be recently approved in Europe (market authorization EU/1/13/855/001) in case of malicious use of retained smallpox stocks for bioterrorism actions. Immunization Procedures The injection site was shaved and cleaned with 70% ethanol. Animals of the ID, SC, and IM groups simultaneously received injections containing a total of 4 x 108 PFU of MVA expressing HIV proteins Gag, Pol, and Nef [MVA HIV-B MVATG17401 Transgene SA, Illkirch-Graffenstaden, France (24)] in 2 ml. For the study of IM injection site responses, 200 ml phosphate buffer saline (PBS, Gibco®) or recombinant MVA expressing eGFP [1 x 108 PFU, MVATG15938, Transgene SA (25)] was injected into the biceps, triceps, or quadriceps muscles. In addition, MVA is currently considered as a promising vector for new recombinant vaccine candidates against cancer and infectious diseases in humans (15–21). Our results show that the administration route of MVA strongly affects the quality and magnitude of the early innate response, and the subsequent installation of the MVA-speciﬁc response. April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. Flow-Cytometry Staining of Blood Samples Flow-Cytometry Staining of Blood Samples 100 µL blood was stained for 30 min with 90 µL antibody mix in BD Horizon® staining buffer (BD) containing the panel of antibodies listed in Table S2 The red blood cells were removed with 1 mL BD FACs Lysing® (BD) buffer for 10 min at RT and the remaining cells washed twice with PBS. The design of the cytometry panel and rationale for the gating strategies were mostly based on previous studies (26). We standardized the ﬂow cytometry results based on complete blood count data to retain the absolute values in the blood, obtaining the number of cells per liter. The CBC-based number of cells per liter was calculated using the following formula: count obtained in ﬂow cytometry of a given cell population multiplied by the CBC number of leukocytes divided by the count of total blood cells obtained by ﬂow cytometry. Antibody Neutralization Assay Neutralization titers were determined using a modiﬁed version of a standard plaque inhibition assay, as previously described (23). Brieﬂy, wild type MVA (1 PFU/cell) was mixed with an equal volume of four-fold serial dilutions of decomplemented serum in assay medium (DMEM, 2% FCS), starting at 1:20. After 60 min of incubation at 37°C, 0.1 ml of the serum-virus mixture was transferred, in duplicate, to a 96-well plate containing subconﬂuent UMNSAH/DF-1 chicken ﬁbroblasts (ATCC). After 48 h of incubation at 37°C, cell viability was quantiﬁed using an MTS/PMS assay (CellTiter 96 AQueous Non- Radioactive Cell Proliferation Assay; Promega). Absorbance was measured at 492 nm using a spectrophotometer (Multiskan FC; Thermo Scientiﬁc) and the data analyzed using SoftMax Pro software (version 4.6; Molecular Devices). The sample dilution versus percentage viability was plotted (four- parameter logistic curve) to calculate the neutralizing concentration, corresponding to the sample dilution resulting in 50% neutralization of virus-mediated cell mortality. The viability of uninfected control cells and infected cells incubated with undiluted VIGIV (human polyclonal anti-vaccinia virus IgG; BEI Resources) was equivalent, as expected. Neutralization titers using VIGIV oscillated between 7 and 10 depending on the sample. Extraction of Cells From Tissue Biopsies p For biopsies after ID injection, subcutaneous tissue was removed to obtain only the epidermis and dermis, which were not separated afterwards. For biopsies after SC injection, all subcutaneous tissue was collected and the epidermis and dermis removed. Muscle biopsies after IM injection were cut into small pieces. All biopsy samples were then washed with PBS, cut into small pieces, weighed, and ﬁnally digested at 37°C under agitation for 60 min in 2 ml of a solution containing RPMI (Thermoﬁscher Scientiﬁc), 5% FCS (Lonza, Basel, Switzerland), 1% penicillin/streptomycin/neomycin (Thermoﬁscher Scientiﬁc), 10mM Hepes (Thermoﬁscher Scientiﬁc), 2 mg/ml collagenase D (Roche, Basel, Switzerland), and 0.02 mg/ml DNAse I (Roche). Tissues were then ﬁltered through a 70-µm cell strainer and centrifuged. The remaining skin sample was shredded with a GentleMACs® dissociator (Miltenyi Biotec, Paris, France). The cell suspension was then washed with PBS and stained for ﬂow cytometry analysis. Flow cytometry staining was performed from cell extractions coming from one single tissue biopsy. Tissue Biopsies in 1% w/v BSA in PBS for 2 h at RT, starting at 1:50 for IgG and 1:25 for IgM. Plates were then washed ﬁve times and peroxidase- conjugated goat anti-monkey H+L chain IgG (1:2,500) or IgM (1:1,000) (AbD Serotec) in 1% BSA (w/v) PBS was added and the plates incubated for 1 h at RT. Plates were washed ﬁve times, and 100 mL 3,3’,5,5’-tetramethylbenzidine (Thermo Scientiﬁc) added and the plates incubated for 30 min at RT in the dark. The reaction was stopped by adding 2N H2SO4. Absorbance was measured at 450 nm using a spectrophotometer (Multiskan FC, Thermo Scientiﬁc) and the data analyzed using SoftMax Pro software (version 4.6; Molecular Devices). Ab titers were calculated as described (23). Prior to biopsies, the site of injection was cleaned with a povidone iodate solution (Vetedine®, Vetoquinol SA, Lure, France). Skin biopsies were 8 mm in diameter. For the subcutaneous route, 5 x 1.5 cm (ellipse shape) deep skin biopsies were performed with a scalpel. Muscular biopsies were conducted after anaesthesia of the animal with tiletamin/zolazepam (Zolétil®, 5 mg/kg). Targeted muscles were the triceps, biceps, and quadriceps. Biopsies were surgically removed at 24 or 48 h after injection. Blood Sample Collection Blood samples were collected into K3-EDTA tubes (Greiner Bio- One, Frickenhausen, Germany) for complete blood counts (CBC; HmX, Beckman Coulter), plasma, and ﬂow cytometry analysis. For evaluation of anti-MVA binding antibody titers, serum was obtained from 2 mL of blood into serum clot activator tubes (Greiner Bio-One). Plasma and serum were collected after 10 min following centrifugation at 2000 x g. For transcriptomic analysis, a volume of 500 µL blood from Li- heparin tubes (Vacutainer BD) was added into a tempus RNA tube (Thermoﬁscher Scientiﬁc, Waltham,MA, USA) containing 1 mL of buffer. For CyTOF staining, 10 mL of blood was collected in Li-heparin. Measurement of Total Anti-MVA IgG and IgM in Serum Experimental procedures were previously described (23). Brieﬂy, wild type MVA (B. Verrier, Laboratoire de Biologie tissulaire et d’ingénierie thérapeutique, Institute of Biology and Chemistry of Proteins, Lyon, France) was used to coat 96-well MaxiSorp microplates (Nunc) at 105 PFU/well in coating buffer (200 mM NaHCO3, 80 mM Na2CO3, pH 9.5) overnight at 4˚C. Wells were washed ﬁve times with wash buffer (PBS, 0.1% Tween 20, 10 mM EDTA) and blocked for 1 h at room temperature (RT) with 3% w/v BSA (Sigma). Plates were washed ﬁve times and incubated with two- or four-fold serial dilutions of macaque serum diluted RNA Isolation and Microarray Sample Preparation Biopsies were immediately immersed in RLT-beta- mercaptoethanol 1/100 lysis buffer (Qiagen, Courtaboeuf Cedex, April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 3 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. France). Samples were then disrupted and homogenized using a TissueLyser LT (Qiagen, Courtaboeuf Cedex, France) and RNA puriﬁedusingtheQiagenRNeasyMicroKit,aspreviouslydescribed (23). Whole-blood RNA was recovered in tempus tubes (ThermoFisher scientiﬁc) and the RNA puriﬁed using the Tempus™Spin RNA isolation kit (ThermoFisher Scientiﬁc). Contaminating DNA was removed using the RNA cleanup step of the RNeasy Micro Kit for both puriﬁcations. Puriﬁed RNA was then quantiﬁed using an ND-8000 spectrophotometer (NanoDrop Technologies, Fisher Scientiﬁc, Illkirch Cedex, France) before checking for integrity on a 2100 BioAnalyzer (Agilent Technologies, Massy Cedex, France). cDNA was synthesized and labeled with biotin using Ambion Illumina TotalPrep RNA Ampliﬁcation Kits (Applied Biosystem/Ambion, Saint-Aubin, France). Labeled cRNA was hybridized to Illumina Human HT- 12V4 BeadChips. All steps were performed following the manufacturer’s protocol. CA), as indicated in the manufacturer’s preload method for 400 µg Ab. The element–Ab combinations are shown in Table S1. The element–Ab conjugates were adjusted to 1 mg/ml in Ab stabilizer buffer (Candor Bioscience, Wangen, Germany), supplemented with sodium azide (Santa Cruz Biotechnology) to a ﬁnal concentration of 0.01%, and stored at 4˚C under sterile conditions throughout the study. The element–Ab conjugates were titrated on PBMCs from healthy macaques in unstimulated and stimulated with PMA+ionomycin conditions to obtain optimal staining concentrations. Analysis of Transcriptomic Data Functional-enrichment analysis of the various lists of differentially expressed genes was performed using Ingenuity Pathway Analysis software (IPA®, Qiagen, http://www.qiagen.com/ingenuity). Upstream regulators with a q-value < 10-3 and belonging to the cytokine-signaling, humoral immune-response, or pathogen- inﬂuenced signaling categories are represented with the exception of FSH, LH, and prolactin regulators. Canonical pathways with a q- value < 10-3 and belonging to the cellular immune-response, cytokine-signaling, humoral immune-response, or pathogen- inﬂuenced signaling categories are represented. Raw and normalized microarray data are available in the ArrayExpress database (27) under accession number E-MTAB-10309. Multidimensional Scaling Representation MDS representations were generated using the SVD-MDS algorithm (28). MDS methods aim to represent the similarities between high-dimensional objects, generally in two or three dimensions. The Kruskal stress (29) indicated at the bottom of the representation quantiﬁes the quality of the MDS as the fraction of information lost during the dimensionality-reduction process. The MDS (30) representation in Figures 2 and 3 was generated basedoncellclusterproﬁlecalculatedfromSPADEanalysisforeach vaccination strategy. Each dot in the representation corresponds to one cluster from one administration route and distances between the dotsare proportionaltothe cumulativedifferenceofparameters between samples. Linear Discriminant Analysis LDA was performed based on cell-population abundances and transcript expression quantiﬁed in the blood for each vaccination strategy. The LDA (31) was parameterized to identify linear combinations of variables that best distinguish the immune response induced by the different administration routes. Preprocessing of Mass-Cytometry Data FCS ﬁles were normalized separately for each administration route experiment with the MATLAB Compiler software normalizer (33) using the signal from the 4-Element EQ beads (Fluidigm). Replicates were concatenated using R software. CD4+ and CD8+ T lymphocytes were manually gated using Cytobank® software (34) as illustrated in Figure S1. Beads, doublets, and dead cells were removed. HLA‑DR- and CD3+ cells were retained and considered to comprise mainly of Intracellular Cytokine Staining of Blood PBMCs and CyTOF Staining PBMC isolation was performed by overlaying a 1:1 (v/v) mixture of Li-heparin–collected whole blood and Dulbecco’s PBS (Invitrogen) on a mixture of 95% Lymphocyte Separation Media 1077 (GE Healthcare, PAA, Austria) and 5% PBS and centrifugation at 420 x g for 45 min at RT. PBMCs at the Separation Media–plasma interface were collected and washed twice in complete media (RPMI 1640 supplemented with 10% heat-inactivated FBS and 100 U/ml penicillin, streptomycin, and neomycin). Any remaining red blood cells were lysed for ﬁve minutes using a hypotonic solution and the pellet washed with PBS. Fresh PBMCs (5 x 106/well) were then stimulated in vitro with wild-type MVA (MVA TGN33.1, Transgène) for 4 h and then incubated overnight at 37°C with brefeldin A (10 µg ml−1) (Sigma). The next day, 1 µL rhodium (Fluidigm) was mixed into each well. Cells were then incubated an additional 15 min at 37°C. Cells were then washed twice in staining buffer (SB; BD Biosciences, Franklin Lakes, NJ) and incubated for 45 min at 4°C with the element tagged surface-staining Abs (Table S1) in a total volume of 50 µl SB. Cells were washed twice with SB and then ﬁxed in 2% paraformaldehyde (Electron Microscopy Sciences) in PBS for 45 min at 4°C. Cells were washed twice in SB, three times in permeabilization buffer (eBioscience), and then stained with intracellular Abs (Table S1) in 50 µl permeabilization buffer on ice for 30 min. Cells were washed twice with SB and resuspended in 0.25 mM iridium DNA intercalator (Fluidigm) in 200 µl PBS + 2% paraformaldehyde overnight at 4˚C, before washing once in SB, once in PBS, and three times in ddH2O, and ﬁltered through a 5-ml polystyrene round- bottom tube with a 35 µm cell strainer cap (BD Biosciences). Fifty microliters of 4-Element EQ Beads (Fluidigm) were added to each sample, which were acquired as two replicates in a mass cytometer (CyTOF; Fluidigm) following the standard procedure recommended by the manufacturer (32). Immunohistoﬂuorescence Skin and lymph-node biopsies were washed three times in PBS and then transferred into solution containing 14.6 g/L-lysine (Sigma Aldrich), 4% PFA (Sigma Aldrich), and 2 g/L sodium (meta)periodate (Sigma Aldrich) in 0.05 mol/l phosphate Buffer (made using 1 mol/L monobasic and dibasic potassium phosphate solutions in equal quantities (Sigma Aldrich) at 4°C for 8 h. Then, skin samples were washed in PBS and transferred to a 30% sucrose (Sigma-Aldrich) solution and incubated overnight. Tissues were then frozen embedded in optimal cutting temperature compound (OCT) in dry ice and acetone. Staining was performed according to the following steps (rinsing with PBS between steps). First, cells were permeabilized for 30 min at 37°C in PBS containing 0.3% Triton X-100 (Sigma Aldrich), followed by a saturation step for 30 min at 37°C in PBS containing 10% BSA (Sigma Aldrich) and 10% Macaque serum (In-lab production). Next, primary antibody staining was performed for 2 h at 37°C in 100 µl PBS with 0.2% BSA containing 1.5 µl of anti-E3L (NIH Biodefense and Emerging Infections Research Resources Repository, NIAID, NIH: Monoclonal Anti-Vaccinia Virus E3L, Clone TW2.3 (produced Antibody Conjugation for CyTOF Staining Pure (carrier-protein–free) mAbs or polyclonal Abs from various manufacturers were coupled to various elements using MAXPAR lanthanide labeling kits (Fluidigm, San Francisco, April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 4 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. NKG2a+ cells. Three main subtypes of monocytes were identiﬁed using CD14 and CD16 markers (37): non-classical monocytes as Lin (CD3, CD20)-, CD14-, CD16+; intermediate monocytes as Lin (CD3, CD20)-, CD14+, CD16+; and classical monocytes as Lin (CD3, CD20)-, CD14+, CD16-, HLA-DR+. pDCs were gated as Lin (CD3, CD20)-, CD14-, CD16-, HLA-DR+, CD123+ cells. Cells that were Lin (CD3, CD20)-, CD14-, CD16-, HLADRmid, CD123+ were considered to be remaining basophils, but could also contained HLADRlow pDCs. The CD33 marker allowed us to discriminate between two different DC subsets: Lin (CD3, CD20)-, CD14-, CD16-, HLA‑DR+, CD123-, CD33+ cells, named CD33+ DCs; and Lin (CD3, CD20)-, CD14-, CD16-, HLA‑DR+, CD123-, CD33-, named CD33- DCs. In accordance with the literature (38), MDSCs were identiﬁed as follows: Lin (CD3, CD20)-, CD14-, CD16-, HLADR-, CD123-, CD66low, CD11b+, and most likely CD33+ for Lin-MDSCs and Lin (CD3, CD20)-, CD14+, HLADR-, CD11b+, CD33+ for CD14+ MDSCs. Cells expressing no markers in the panel were named “Remaining Cells”. T lymphocytes. CD4+ T cells were considered to be CD4+ and CD8- whereas CD8+ T cells were considered to be CD4- and CD8+. The isolated CD4+ and CD8+ T cells were then analyzed separately using automatic gating tools (30). Immunohistochemistry Samples were ﬁxed at 4°C using 4% paraformaldehyde during 24h, then stored in PBS at 4°C. Parafﬁn inclusion was made using an automaton successively replacing PBS with alcohol, xylene, and parafﬁn. Tissues were cut afterward using microtome and set on slides. Hemalum Eosin coloration was then made using an automaton Microm HMS740 (Thermoﬁscher Scientiﬁc), alternating xylene, alcohol, Hematoxylin (Labonord, Templemars, France), Eosin (VWR international, Fontenay- sous-Bois, France) baths. Identiﬁcation of Cell Populations The Spanning-tree Progression Analysis of Density-normalized Events (SPADE) algorithm was used to identify CD4 and CD8 T- cell populations with similar marker intensities. SPADE was performed using the publicly available R/Bioconductor package (35). Six separate SPADE analyses were performed, one for each of the following conditions: CD8 SC, CD8 ID, CD8 IM, CD4 SC, CD4 ID, and CD4 IM (Figure S1). For each SPADE analysis, we ﬁrst randomly selected and concatenated an equal number of cells corresponding to the smallest sample uniformly for each sample to reduce bias during the SPADE down-sampling step caused by large variances between the numbers of cell events per macaque. We then performed the SPADE density-based down- sampling using a target of 30% of the preselected cells. SPADE analyses were based on 13 markers: Bcl2, GranzymeB, CD62L, CD69, CD154, CD197, IFNg, IL2, IL17, Ki67, MIP1b, Perforin, and TNFa. These markers were selected because of their ability to deﬁne T-cell functions. The number of cell clusters chosen for the ﬁnal analysis provided an optimal balance between phenotypically homogeneous clusters, which also contained adequate cell numbers for downstream analysis (22, 30). Flow cytometry data were scaled in the heatmap representations to have the same minimal and maximal expression values using R. Statistical analysis for the ﬂow cytometry data was performed using Prism 6.0 (Graphpad Software Inc., La Jolla, CA, USA). Two-sided Friedman tests, followed by Dunn’s post-test were used to compare each timepoint with the baseline. y The SPADEVizR (30) and CytoCompare (36) R packages were used to analyze the abundance and phenotypic similarity between the identiﬁed cell clusters. The phenotypic categorization of SPADE clusters in the heatmap representation was calculated using the 5th and 95th percentile range for each marker intensity and then dividing this range into ﬁve uniform categories. Categorization was computed based on the means of the individual SPADE expression medians for each marker. The categories represent negative, low, medium, high, and very high marker expression, using a color scale from white to dark red. The mean percentage of CD8+ or CD4+ T cells was computed for each SPADE cluster at each time point. Differentially abundant clusters (DACs) were identiﬁed using a two- sample t-test (p < 0.05) that compared the percentage of CD8+ and CD4+ T cells of a SPADE cluster with that at baseline at a given timepoint. Gating Strategy of Blood Cells by Flow Cytometry and Analysis The gating strategy is provided on Figure S3A. We optimized the panel using both NKG2a and CD33 markers on the same ﬂuorochrome. Indeed, these markers are not expressed by the same cell populations (Figure S3B). First, doublets and debris were excluded. SSC-A+, CD66+ cells were considered to be granulocytes, which comprised mainly of neutrophils as demonstrated by the recruitment kinetic data derived from complete blood counts. However, we cannot exclude the presence of eosinophils and basophils in this gate. B lymphocytes were gated as Lin (CD3, CD20)+,CD4-, HLA-DR+ cells; CD8+ T lymphocytes as Lin (CD3, CD20)+, HLA-DR-, CD8+ cells; CD4+ T lymphocytes as Lin (CD3, CD20)+,CD4-, HLA-DR- cells; and NK cells as Lin (CD3, CD20)-, CD16+, CD8+, April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 5 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. Quantiﬁcation and Statistical Analysis Quantiﬁcation and Statistical Analysis Statistical analysis for the Figures 1, 4 and 7 were performed using Prism 6.0 (Graphpad Software Inc., La Jolla, CA, USA). Statistical analysis for the Figures 2, 3, 6, and Figure S2 were performed using R software. Statistical analysis for Figure 5 was determined by Ingenuity Pathway analysis Software (Qiagen Inc. Hilden, Germany). Statistical analysis for the Figures 1, 4 and 7 were performed using Prism 6.0 (Graphpad Software Inc., La Jolla, CA, USA). Statistical analysis for the Figures 2, 3, 6, and Figure S2 were performed using R software. Statistical analysis for Figure 5 was determined by Ingenuity Pathway analysis Software (Qiagen Inc. Hilden, Germany). Figure 1: Four subjects per group. The bar represents the mean. Statistical signiﬁcances was determined by Mann-Whitney test Figure 1: Four subjects per group. The bar represents the mean. Statistical signiﬁcances was determined by Mann-Whitney test Figure 2/Figure 3/Figure S2: Four subjects per group. Statistical signiﬁcances was determined by comparing each group by t-test Figure 2/Figure 3/Figure S2: Four subjects per group. Statistical signiﬁcances was determined by comparing each group by t-test Figure 4A: Six subjects per group.Two-sided Friedman tests, followed by Dunn’s post-test. Figure 4A: Six subjects per group.Two-sided Friedman tests, followed by Dunn’s post-test. Figure 4B: Six subjects per group. The bar represents the mean. Statistical signiﬁcances was determined by Mann-Whitney test Figure 4B: Six subjects per group. The bar represents the mean. Statistical signiﬁcances was determined by Mann-Whitney test Figure 5: Six subjects per group. Figure 6: Six subjects per group. The bar represents the mean. Statistical signiﬁcances was determined by Mann-Whitney test A B FIGURE 1 \| Antibody responses induced by homologous MVA prime-boost administered by the SC/ID/IM routes. (A) Scatter plot of MVA-speciﬁc IgG titers over time, characterized by ELISA. Each dot corresponds to one single biological sample. The colors correspond to the different administration routes. (B) Scatter plot of MVA neutralization titers over time, characterized by plaque reduction assay. Each dot corresponds to one single biological sample. BL, baseline; D, day; PP, post- prime; PB, post-boost; M, month. A B B B FIGURE 1 \| Antibody responses induced by homologous MVA prime-boost administered by the SC/ID/IM routes. (A) Scatter plot of MVA-speciﬁc IgG titers over time, characterized by ELISA. Each dot corresponds to one single biological sample. The colors correspond to the different administration routes. RESULTS in vitro), NR-4547.), 0.5 µl of anti-CD66 (clone TET2, Miltenyi Biotec), and 6 µl of anti-CD163 (clone GHI/61, Biolegend). Finally, slices were stained for 1 h at 37°C with secondary antibodies (Thermoﬁscher Scientiﬁc). A post-ﬁxation step in 4% PFA (15 min) was then performed, followed by DAPI staining (Thermoﬁscher Scientiﬁc), and the samples mounted. Stained slides were then conserved at 4°C. The Route of Immunization Induced Distinct Proﬁles of Neutralizing Antibodies in Serum Three groups of four macaques were immunized two times with 4 x 108 PFU rMVA, injected ID, SC, or IM at 58-days intervals. MVA-speciﬁc IgG have similar kinetic proﬁles in all groups. The impact of administration route on antibody titers appeared to be limited (Figure 1A), although signiﬁcant at day 28 post-prime (D28PP) (SC vs ID, p = 0.029). A trend to maintain higher responses four months post-boost (M4PB) was also observed after IM route. In contrast, the SC route was signiﬁcantly more efﬁcient in inducing MVA-neutralizing activity early after the boost than the ID (p = 0.029 at day 8 post-boost (D8PB); p = 0.057 at day 14 post-boost (D14PB); or IM routes (p = 0.029 at D8PB) (Figure 1B). Interestingly, IM immunization appears to induce lower and delayed peaks of serum neutralizing responses, becoming signiﬁcantly higher when compared to SC by day 28 post-boost (p = 0.029 at D28PB and D57PB). However, these neutralizing activities did not persist over time. These changes in proﬁles of responses in spite of comparable binding IgG levels suggest that paths of antigen uptake, processing or presentation, may be driven by distinct inﬂammatory processes and populations of innate cells, accordingly to injections sites in a way that could modulate the quality (IgG isotypes, IgM and IgA, MVA recognized epitopes) of the speciﬁc antibodies and associated functionalities. Quantiﬁcation and Statistical Analysis FIGURE 2 \| Characterization of the impact of the SC/ID/IM routes on CD8+ T-cell phenotypic cluster diversity by mass cytometry. (A) Categorical marker expression heatmap representing the phenotypes of CD8+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set according to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into families by hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation showing the phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the mean signal intensity of the 13 SPADE clustering markers. Red square indicate the eight singular cell cluster (C) Charts showing changes in the percentage of CD8+ T cells over time for the selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. Quantiﬁcation and Statistical Analysis (B) Scatter plot of MVA neutralization titers over time, characterized by plaque reduction assay. Each dot corresponds to one single biological sample. BL, baseline; D, day; PP, post- prime; PB, post-boost; M, month. April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. A B C \| Characterization of the impact of the SC/ID/IM routes on CD8+ T-cell phenotypic cluster diversity by mass cytometry. (A) Categorical marker express presenting the phenotypes of CD8+ T cells for each administration route. SPADE analysis was performed for each administration route based on the en ng to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representati e phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the m sity of the 13 SPADE clustering markers. Red square indicate the eight singular cell cluster (C) Charts showing changes in the percentage of CD8+ T ce selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. A B C C FIGURE 2 \| Characterization of the impact of the SC/ID/IM routes on CD8+ T-cell phenotypic cluster diversity by mass cytometry. (A) Categorical marker expression heatmap representing the phenotypes of CD8+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set according to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into families by hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation showing the phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the mean signal intensity of the 13 SPADE clustering markers. Red square indicate the eight singular cell cluster (C) Charts showing changes in the percentage of CD8+ T cells over time for the selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells Phenotypic similarities were quantiﬁed based on the MSI of the 13 SPADE clustering markers (C) Charts showing changes in the percentage of CD4+ T cells over time for the selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. FIGURE 3 \| Characterization of the impact of the SC/ID/IM routes on CD4+ T-cell cluster diversity by mass cytometry. (A) Categorical marker expression heatmap representing the phenotypes of CD4+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set according to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into families by hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation showing the phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the MSI of the 13 SPADE clustering markers (C) Charts showing changes in the percentage of CD4+ T cells over time for the selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. FIGURE 3 \| Characterization of the impact of the SC/ID/IM routes on CD4+ T-cell cluster diversity by mass cytometry. (A) Categorical marker expression heatmap representing the phenotypes of CD4+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set according to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into families by hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation showing the phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the MSI of the 13 SPADE clustering markers (C) Charts showing changes in the percentage of CD4+ T cells over time for the selected marginal clusters. Black arrows indicate the days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. experiment timeframe and different antibody batches, one SPADE analysis per administration route was performed. Each SPADE analysis was performed using data from all animals and timepoints. Each SPADE contained 50 clusters. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells D8PP, D28PP, D8PB, and D28PB. Cells were stained with a panel of 29 antibodies coupled to heavy metals (Table S1) and the data acquired by mass cytometry. The SPADE algorithm (39, 40) was used on gated CD8+ T cells to identify cell clusters based on the expression of 13 functional and homing markers; Bcl2, Granzyme B, CD62L, CD69, CD154, CD197, IFNg, IL-2, IL-17, Ki67, MIP1b, Perforin, and TNFa (Figure S1). Due to different The MVA-speciﬁc CD8+ T-cell responses were characterized by the intracellular production of cytokines following PBMC stimulation ex vivo with MVA for 18 h. Assays were performed over the kinetics of immunization at baseline (BL), April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 7 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. A B C \| Characterization of the impact of the SC/ID/IM routes on CD4+ T-cell cluster diversity by mass cytometry. (A) Categorical marker expression heatmap g the phenotypes of CD4+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set o the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation e phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. Phenotypic similarities were quantiﬁed based on the MSI PADE clustering markers (C) Charts showing changes in the percentage of CD4+ T cells over time for the selected marginal clusters. Black arrows e days of vaccine injection. BL, baseline; D, day; PP, post-prime; PB, post-boost. A B C A B C A B C C FIGURE 3 \| Characterization of the impact of the SC/ID/IM routes on CD4+ T-cell cluster diversity by mass cytometry. (A) Categorical marker expression heatmap representing the phenotypes of CD4+ T cells for each administration route. SPADE analysis was performed for each administration route based on the entire cell set according to the gating strategy described in Figure S1. Markers in blue were used as clustering markers in the SPADE analyses. Cell clusters were organized into families by hierarchical clustering. The ﬁve-tiered color scale, from white to deep red, represents relative marker expression. (B) Two-dimensional MDS representation showing the phenotypic differences and similarities between cell clusters identiﬁed by each SPADE analysis. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells In addition, other key markers were heterogeneously expressed by these polyfunctional clusters, such as IL10 (expressed in two of four polyfunctional clusters of the IM route group) and CD278 (expressed moderately in the SC and strongly in the IM group). The phenotypic family with cytotoxic activity markers was deﬁned by the expression of perforin and Granzyme B. The proliferative family was deﬁned by the expression of Ki67 and the migrating family by the expression of homing receptors to secondary lymphoid organs, such as CD197 and CD62L. Finally, the remaining cell clusters were classiﬁed into phenotypic family named “various proﬁles”. Using multidimensional scaling (MDS) analysis (Figure 2B), we identiﬁed eight clusters with singular phenotypes all part of the polyfunctional families. Cells associated with SC-speciﬁc cell clusters with the polyfunctional phenotypes SC#50 (415 cells) and SC#41 (313 cells, p = 0.003 at D8PB) appeared to increase in only two macaques (Figure 2C) and at low magnitude. In contrast, abundance of the two polyfunctional clusters associated with ID route (clusters ID#44 and ID#50) mainly increased at D8PB (ID#44, p = 0.043) at D8PP (day 8 post- prime), but were differentially abundant at D8PB (ID#44, p = 0.028; ID#50, p = 0.015) and day D28PB (ID#44, p = 0.008; ID#50, p = 0.007) relative to the baseline. The IM route induced four polyfunctional clusters (IM#38, IM#48, IM#49, and IM#50), which also displayed changed kinetic proﬁles similar to the ID route, with signiﬁcant increases at D8PB (IM#38, p = 0.015; IM#50, p = 0.032) and D28PB (IM#50, p = 0.033). In addition, the IM route induced more heterogeneous marker expression. In particular, IL-10 was only expressed in two offour polyfunctional clusters of the IM route. CD8+ T cells of the cytotoxic group were not particularly more enriched after vaccination by the SC or IM route, whereas two clusters were more enriched at D8PB after vaccination by the ID route (Figure S2A). Clusters from the migrating-cells phenotypic family also became more abundant after immunization via the ID route. These results highlight that the MVA administration route alters the CD8+ response proﬁle. In particular, IFNg-secreting CD8+ T cells displayed a distinct polyfunctional proﬁle and followed different recruitment kinetics. Vaccination by the IM route induced a signiﬁcant increase in the number of antigen- speciﬁc polyfunctional CD8+ T cells associated with increased expression of other markers, including IL10. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells Phenotypes (Figure 2) and cell abundances (Figure S2) of each cluster were represented using categorical heatmaps based on the relative range of expression of each marker (ﬁve categories from no expression to very high expression). Hierarchical clustering at the cell cluster level was used to deﬁne groups of clusters with closely related phenotypes, which were designated as ‘phenotypic families’. We identiﬁed ﬁve phenotypic families. Clusters of polyfunctional cells expressed a backbone of markers April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 8 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. A B FIGURE 4 \| Impact of the SC/ID/IM routes on early systemic immunity. (A) Heatmap representation showing the changes in cell abundance (in log2 fold change) of populations characterized by ﬂow cytometry in the blood 3, 6, and 24 h after SC/ID/IM MVA injection. Populations were organized based on their cell abundance by hierarchical clustering and are represented by a dendrogram. Changes in cell abundance of each population were normalized and are represented by a color-gradient scale ranging from yellow to red. Signiﬁcant changes over time are indicated on the right (B) Scatter plot showing the kinetics of changes in cell abundance for selected cell populations. Each dot represents one single biological sample. ns, non-signiﬁcant; 0.05 > p > 0.01; 0.01 > p > 0.001; 0.001 > p > 0.0001; **p < 0.0001. A A B B FIGURE 4 \| Impact of the SC/ID/IM routes on early systemic immunity. (A) Heatmap representation showing the changes in cell abundance (in log2 fold change) of populations characterized by ﬂow cytometry in the blood 3, 6, and 24 h after SC/ID/IM MVA injection. Populations were organized based on their cell abundance by hierarchical clustering and are represented by a dendrogram. Changes in cell abundance of each population were normalized and are represented by a color-gradient scale ranging from yellow to red. Signiﬁcant changes over time are indicated on the right (B) Scatter plot showing the kinetics of changes in cell abundance for selected cell populations. Each dot represents one single biological sample. ns, non-signiﬁcant; 0.05 > p > 0.01; 0.01 > p > 0.001; 0.001 > p > 0.0001; ***p < 0.0001. related to T-cell activation, including CD69 and IFNg, and high levels of MIP1b, TNFa, and IL2. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells FIGURE 5 \| Transcriptomic analysis of systemic early responses induced by SC/ID/IM MVA administration. (A) Histogram of the number of signiﬁcantly up-and downregulated genes at 3, 6, and 24 after MVA injection in the whole blood relative to baseline (0 h) for each administration route. Downregulated genes are represented in green and upregulated genes in red. (B) Venn diagrams showing the overlaps between the differentially expressed genes for each administration route at 3, 6, and 24 h after MVA injection. (C) Heatmap representation showing functional enrichment scores for signiﬁcantly over-represented canonical pathways related to immune processing and cytokine signaling with a p-value < 10-3. The color scale corresponds to the Z-score, which is proportional to the upregulation (in red) or downregulation (in green) of the pathways. MVA Immunization Through the ID/IM Route Favors Polyfunctional Speciﬁc CD8+ T Cells In contrast, vaccination by the SC route induced limited MVA-speciﬁc CD8+ T-cell responses. Vaccination by the ID route appeared to induce an intermediate level of responses, with a moderate increase in polyfunctional activity, but also a cluster abundance increase in cytotoxic and migratory phenotypic families in comparison with other immunization routes. We surmise from these results that early molecular and cellular changes at the site of antigen exposure may signiﬁcantly impact on the initiation of antigen speciﬁc CD8+ T cell responses. April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 9 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. e CD4+ T-Cell Response Proﬁle Is ected by the Administration Route phenotypes (Figure 3A) and cell abundances (Figure S2B) of each cell cluster were analyzed. Five phenotypic families were A B C URE 5 \| Transcriptomic analysis of systemic early responses induced by SC/ID/IM MVA administration. (A) Histogram of the number of signiﬁcantly up-and nregulated genes at 3, 6, and 24 after MVA injection in the whole blood relative to baseline (0 h) for each administration route. Downregulated genes are esented in green and upregulated genes in red. (B) Venn diagrams showing the overlaps between the differentially expressed genes for each administration route 6, and 24 h after MVA injection. (C) Heatmap representation showing functional enrichment scores for signiﬁcantly over-represented canonical pathways related mmune processing and cytokine signaling with a p-value < 10-3. The color scale corresponds to the Z-score, which is proportional to the upregulation (in red) or nregulation (in green) of the pathways. A B B C C C FIGURE 5 \| Transcriptomic analysis of systemic early responses induced by SC/ID/IM MVA administration. (A) Histogram of the number of signiﬁcantly up-and downregulated genes at 3, 6, and 24 after MVA injection in the whole blood relative to baseline (0 h) for each administration route. Downregulated genes are represented in green and upregulated genes in red. (B) Venn diagrams showing the overlaps between the differentially expressed genes for each administration route at 3, 6, and 24 h after MVA injection. (C) Heatmap representation showing functional enrichment scores for signiﬁcantly over-represented canonical pathways related to immune processing and cytokine signaling with a p-value < 10-3. The color scale corresponds to the Z-score, which is proportional to the upregulation (in red) or downregulation (in green) of the pathways. The CD4+ T-Cell Response Proﬁle Is Affected by the Administration Route of MVA The CD4+ T-Cell Response Proﬁle Is Affected by the Administration Route of MVA phenotypes (Figure 3A) and cell abundances (Figure S2B) of each cell cluster were analyzed. Five phenotypic families were deﬁned for each SPADE analysis based on marker expression, similarly to that of the CD8+ T cells. Differences in MVA speciﬁc antibodies and CD8+ T cells functionalities and proﬁles may be associated to differences in CD4+ T cell compartment accordingly to immunization route. The characterization of CD4+ T cells was performed similarly to that of the CD8+ T cells (Figure S1), with the same antibody panel (Table S1). Three SPADE analyses were performed to identify 70 clusters for each administration route, and the MDS analysis identiﬁed three phenotypically distinct clusters (Figure 3B): one speciﬁc to the SC route (SC#70), one speciﬁc to the IM route (IM#70), and one speciﬁc to the ID route (ID#51). Two harbored a polyfunctional phenotype (SC#70 and IM#70) and expressed CD154 in addition to CD69, IFNg, TNFa, and IL-2, corresponding to a Th1 proﬁle. In contrast, the ID-associated April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 10 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. A B FIGURE 6 \| Integrative analysis of innate immune signatures after ID/IM/SC MVA immunization. (A) LDA representation of the innate dataset after ID, IM, and SC MVA immunization. Each dot corresponds to one biological sample. Each color corresponds to one administration route. Biological samples are represented on a two-dimension scale. Differential gene expression between samples was translated into the Euclidian distance between samples. (B) Top 20 identiﬁed genes involved in the level of expression of samples on the LD1 axis and the LD2 axis. A B B FIGURE 6 \| Integrative analysis of innate immune signatures after ID/IM/SC MVA immunization. (A) LDA representation of the innate dataset after ID, IM, and SC MVA immunization. Each dot corresponds to one biological sample. Each color corresponds to one administration route. Biological samples are represented on a two-dimension scale. Differential gene expression between samples was translated into the Euclidian distance between samples. (B) Top 20 identiﬁed genes involved in the level of expression of samples on the LD1 axis and the LD2 axis. cluster (ID#51), expressing high levels of TNFa, was classiﬁed in the cytotoxic family because of the high expression of perforin and granzyme B. The CD4+ T-Cell Response Proﬁle Is Affected by the Administration Route of MVA The kinetics of cell abundances within the clusters showed that the SC-speciﬁc cluster (SC#70) corresponded to cells originating mostly from one animal at D8PB (Figure 3C). However, the IM-speciﬁc cluster (IM#70) had signiﬁcantly more cells in all animals at D57PP (p = 0.029), D8PB (p = 0.025), and D28PB (p = 0.023) relative to baseline. The ID-speciﬁc cluster (ID#51) showed differential abundance at D8PP (p = 0.021) and D57PP (p = 0.032) only. CD4+ T cells exhibiting proliferation markers at D8PP in each administration route. This increase was however transient in the IM and ID groups. Overall, the CD4+ T-cell response was affected by the administration route, similarly to that of CD8+ T cells. A marked Th1 signature following IM and ID immunization, whereas the SC route induced a weaker Th1 response, with a lower abundance of polyfunctional cluster cells. MVA-Induced Systemic Inﬂammation Echoes the Adaptive Response The abundance of cytotoxic clusters expressing high levels of perforin and granzyme B signiﬁcantly increased over time following SC administration, in contrast to IM or ID administration (Figure S2B). MVA induced an increase of Since we observed that adaptive immune proﬁles were largely polarized depending on the administration route of the MVA, we April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. hypothesized that it was shaped by the early innate response following vaccine injection. Gene expression analysis showed that 22.4% of the differentially enriched transcripts were common between the three routes at 6 h (Figure 5B). Notably, more transcripts were in common between the IM and ID routes (37% at 3 h, 17.6% at 6 h, and 17.7% at 24 h) than between the ID and SC routes (0.2% at 3 h, 8.7% at 6 h, and 2.9% at 24 h) and the IM and SC routes (0.1% at 3 h, 0.1% at 6 h, and 12.5% at 24 h). We performed ﬂow cytometry analysis (Figure S3) at 3, 6, and 24 h after MVA injection in additional groups of six NHPs to decipher the MVA-induced early innate immune cell changes in the blood compartment (Figure 4). Myeloid cells populations ﬂuctuated globally similarly for the three routes, with a peak reached after 6 and 24 h post-MVA injection. The three routes of immunization induced signiﬁcant recruitment of granulocytes (p < 0.0001 for SC, ID, and IM) and intermediate monocytes (p < 0.01 for SC and ID and p = 0.05 for IM) over time, but with distinct frequencies accordingly to injection site. Intermediate monocytes (SC vs. ID after 6 h: p = 0.026; SC vs. ID after 24 h: p = 0.015), CD8+ T cells (SC vs. ID after 24 h: p = 0.0087) and pDCs (SC vs. ID after 6 h: p = 0.015; SC vs. ID after 24 h: p = 0.0022; SC vs. IM after 24 h: p = 0.026) were signiﬁcantly less recruited after vaccination by the ID and/or IM routes compared to the SC route. Functional enrichment analysis showed that the modulation of several immune-related pathways were shared between all three conditions (Figure 5C), such as innate functions associated with lL-6 and TREM1-signaling. MVA-Induced Systemic Inﬂammation Echoes the Adaptive Response Most of the signiﬁcantly immune-related increased pathways reached a peak at 3 h after vaccination by the IM and ID routes. Conversely, the same pathways plateaued 24 h after vaccination by the SC route. The magnitude of the pathways associated with the innate immune response was globally less, especially for the ID route (TREM1 signaling Z-score: z = 3.87 for SC, 3.46 for IM, and 2.5 for ID at the peak of the response). This is illustrated by a decrease of the inﬂammation-related pathways at 24 h, such as LPS-stimulated MAPK signaling, dendritic-cell (DC) maturation, and IL-8 signaling. In addition, several pathways related to the lymphoid response (such as iCOS-iCOSL signaling in T helper cells, calcium-induced T lymphocyte apoptosis, and CD28 signaling in T helper cells) were decreased after vaccination by the IM and ID route but not the SC route. Interestingly, these changes were associated to a signiﬁcant increase in the number of myeloid-derived suppressive phenotype (MDSCs), including CD14+ MDSCs (p < 0.0001 for SC and ID and p = 0.005 for IM), as well as Lin- MDSCs (p < 0.0001 for SC and IM and p = 0.005 for ID) (Figure 4A). This suggested that these cells play a physiological role in controlling early inﬂammation. A greater recruitment of Lin-MDSCs following vaccination by the IM route was observed compared to the ID route (ID vs. IM after 24 h: p = 0.015 after 24 h) (Figure 4B). Conversely, B lymphocytes (p < 0.01 for SC and IM and p = 0.05 for ID), CD8+ T lymphocytes (p < 0.001 for SC and IM and p = 0.0001 for ID), and CD4+ T lymphocytes (p < 0.001 for SC and IM and p = 0.0001 for ID) were reduced in the blood compartment between 6 and 24 h post-MVA injection (Figure 4B). As a result, there were subtle but clear differences in the early kinetics of key immune cell subsets observed. The response following the SC route differed from that of the ID and/or IM routes for three of the four cell populations: CD8- T cells, intermediate monocytes, and CD14+ MDSCs. Overall, transcriptomic proﬁling varied depending on the administration route. However, because of rapid and signiﬁcant changes occurring in cell subtypes proportions it is difﬁcult to discriminate the mRNA variations reﬂecting these changes from true gene expression modulations. MVA-Induced Systemic Inﬂammation Echoes the Adaptive Response However, canonical pathways associated with pathogen sensing, proinﬂammatory responses, and T cell regulation also followed different kinetics and differed in magnitude depending on the administration route. Dynamics of gene expression into these pathways categorized the SC group (peak at 24 h) separately to the IM and ID groups (peak at 3 h). This may be consistent with differences observed for antibody and CD8+ T cell responses induced, on the one hand, by IM and ID routes and, on the other hand, by SC route. Frontiers in Immunology \| www.frontiersin.org Integrative Analysis of the Innate Response Proﬁle Suggests That Early Innate Immune Effectors Drive the Adaptive-Response Proﬁle Some of them depicted a strong contribution but were not directly related to immune function, such as PLXDC2 (plexin domain containing 2, involved in the inhibition of angiogenesis) (41), SG5 (encoding a chaperone protein of a neuroendocrine enzyme) (42) and PER3 (involved in the circadian rhythm). Other factors were associated with innate immune processes, such as KLHL25 (associated with MHC I presentation) and TRAP1 (TNF response). Some of the identiﬁed genes were associated with T-cell maturation and regulation, such as Thy-1 (43) and GATA-3. Overall, the segregation proﬁles were more the sum of slightly modiﬁed signatures than the contribution of a few genes with speciﬁc immune functions. LD1 (Figure 6A). The important imbalance in the number of parameters integrated for each type of data (low numbers of variables for ﬂow cytometry in comparison with transcriptomics) bias the analysis and certainly explain that the main factors affecting this segregation were changes in transcripts enrichment rather than immune cells populations. Furthermore, values of the top 20 factors were quite homogeneous, suggesting a comparable contribution of these factors to drive SC speciﬁcity (Figure 6B). These factors included genes encoding proteins with very different functions. Some of them depicted a strong contribution but were not directly related to immune function, such as PLXDC2 (plexin domain containing 2, involved in the inhibition of angiogenesis) (41), SG5 (encoding a chaperone protein of a neuroendocrine enzyme) (42) and PER3 (involved in the circadian rhythm). Other factors were associated with innate immune processes, such as KLHL25 (associated with MHC I presentation) and TRAP1 (TNF response). Some of the identiﬁed genes were associated with T-cell maturation and regulation, such as Thy-1 (43) and GATA-3. Overall, the segregation proﬁles were more the sum of slightly modiﬁed signatures than the contribution of a few genes with speciﬁc immune functions. 24 h: p = 0.0022; ID route baseline vs. MVA 24 h: p = 0.0087, and buffer 24 h vs. MVA 24 h: p = 0.0022) (Figure 7C). Although the exact localization of the injection was difﬁcult to determine in muscle biopsies, we detected a trend for increased recruitment of macrophages characterized after IM immunization (Figure 7D). As a result, we observed an increase of granulocytes and/or macrophages 24h after any route of vaccine immunization. However, the localization and the amplitude of this response differed depending on the route. DISCUSSION Administration route can shape immune response proﬁles after vaccination (44, 45). Yet, crucial players of this innate and adaptive immune response remain to be identiﬁed. Here, we investigated the effect of the commonly used SC and IM in addition to ID administration routes on the stimulation of immune responses in an MVA vector-based, live attenuated NHP vaccine model. We combined a complete set of experimental data obtained by multiparameter techniques, at the cellular (ﬂow cytometry and mass cytometry) and molecular (transcriptomics) levels, with systems vaccinology approaches, providing a holistic view of the complex mechanisms of the vaccine response, to evaluate the impact of the administration route on the immune response. In conclusion, this integrative analysis of the blood innate immune response showed that early signatures induced by the IM and ID route differ from those induced by the SC route. Mechanisms associated with SC segregation appear to belong to various signaling pathways. These analyses of the innate immune response support the hypothesis that the early innate immune proﬁle can shape the adaptive response. Furthermore, these ﬁndings underlined the importance of investigating holistic analytical approaches. Indeed, in this study, parameters that were not known to be primarily involved in immune process could also be a part of an early signature that is associated with alteration of MVA-speciﬁc T cell and antibody responses. p p MVA administration induced substantial myeloid cell recruitment, whereas B and T lymphocytes were reduced in the blood compartment. Although the cell trafﬁcking was globally similar, in-depth analysis revealed that the early immune response shows route-speciﬁc differences. For instance, CD8+ T lymphocytes and pDCs, expanded more in between 6 to 24 h after SC immunization compared to after the IM and ID routes. In addition, the functional enrichment of differentially expressed genes was strongly associated with immune pathways peaking at 24 h for the SC route and 3 h post-injection for the IM and ID routes. These pathways were mostly associated with acute responses, i.e. IL-6, nitric oxide production, and triggering of receptors expressed on myeloid cells (46, 47). An integrative analysis based on the whole set of innate immune parameters showed the same polarization between the IM/ID and SC routes. These results reinforce the study of Gonçalves et al. with an inactivated inﬂuenza vaccine showing that gene signatures from the early innate immune response can be predictive of adaptive immune proﬁles (48). Integrative Analysis of the Innate Response Proﬁle Suggests That Early Innate Immune Effectors Drive the Adaptive-Response Proﬁle Overall, similarly to innate systemic signature, we observed shared phenomenon but also discrepancies in the local inﬂammatory response after SC, ID, or IM MVA administration that may be reﬂected in the blood. Integrative Analysis of the Innate Response Proﬁle Suggests That Early Innate Immune Effectors Drive the Adaptive-Response Proﬁle To complement characterization of cell changes, we additionally studied the early immune mechanisms following vaccination at the molecular level by performing transcriptomic proﬁling of blood samples at 3, 6, and 24 h after MVA immunization (Figure 5). Gene expression analysis highlighted different kinetic patterns, depending on the administration route (Figure 5A). Indeed, more mRNA expression appeared to be decreased after vaccination by the IM and ID route than the SC route. The kinetics of increased genes products differed between groups, despite a comparable magnitude (peak of approximately 800 increased transcripts). Furthermore, the peak for the SC route was observed after 6 h and persisted until 24 h, whereas the peak for the ID route was transient, reaching a maximum peaking at 6 h and then rapidly decreasing by 24 h. Following IM injection, the number of increased gene products was already high after 3 h and continued to increase up to 24 h. We performed an integrative analysis of all innate immune response data to evaluate the speciﬁcity of the routes. We used the data of cell populations identiﬁed by ﬂow cytometry and differentially enriched gene products after 24 h post-injection, which corresponds to the peak in terms of magnitude of cellular and transcriptomic signatures (Table S3). A linear discriminant analysis (LDA) was used to compare the different routes (Figure 6). LDA aims to identify linear combinations of variables (named Linear Discriminant) that best segregate samples from the different conditions. This approach allowed us to conﬁrm the dichotomy between the IM/ID conditions relative to the SC condition, mainly driven by April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 12 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. LD1 (Figure 6A). The important imbalance in the number of parameters integrated for each type of data (low numbers of variables for ﬂow cytometry in comparison with transcriptomics) bias the analysis and certainly explain that the main factors affecting this segregation were changes in transcripts enrichment rather than immune cells populations. Furthermore, values of the top 20 factors were quite homogeneous, suggesting a comparable contribution of these factors to drive SC speciﬁcity (Figure 6B). These factors included genes encoding proteins with very different functions. DISCUSSION In this work with an attenuated virus model, administration routes also appear to orientate the ensuing immune responses, which contribute to the body of knowledge describing the effects of innate immunity on adaptive immunity. Innate Response at the Site of Injection A scatter plot representation of local cellular movement to the site of injection at baseline and following injection with buffer or MVA by the SC or ID route is shown at the bottom. (B) HE staining of muscle, showing transversal sections of parafﬁn-embedded tissue cassettes at baseline and 24 h after injection with PBS and 24 h after MVA immunization. The scale bar is equal to 50 µm. On the bottom, scatter plot representation of local cellular movement to the site of injection at baseline and after injection with buffer or MVA after (C) SC or ID route and (D) IM route. Each dot represents one biological sample (n = 6 for the ID and SC groups and n = 5 for the IM group). Values correspond to the percentage of cells among CD45+ cells conforming to the gating strategy in Figure S5A for the SC and ID group and Figure S5B for the IM group. Statistical analysis consisted of a Mann-Whitney test test. A A B A D C C C D FIGURE 7 \| Local inﬂammatory response 24 h after MVA immunization by the SC or ID route. (A) HE staining of the injection site (skin + subcutaneous tissue) and draining inguinal lymph node, showing transversal sections of parafﬁn-embedded tissue cassettes 24 h after ID or SC injection of buffer or rMVA. The scale bar is equal to 50 µm. One representative experiment of two is shown. A scatter plot representation of local cellular movement to the site of injection at baseline and following injection with buffer or MVA by the SC or ID route is shown at the bottom. (B) HE staining of muscle, showing transversal sections of parafﬁn-embedded tissue cassettes at baseline and 24 h after injection with PBS and 24 h after MVA immunization. The scale bar is equal to 50 µm. On the bottom, scatter plot representation of local cellular movement to the site of injection at baseline and after injection with buffer or MVA after (C) SC or ID route and (D) IM route. Each dot represents one biological sample (n = 6 for the ID and SC groups and n = 5 for the IM group). Values correspond to the percentage of cells among CD45+ cells conforming to the gating strategy in Figure S5A for the SC and ID group and Figure S5B for the IM group. Innate Response at the Site of Injection p j A comprehensive understanding of the initiation of MVA vaccine-induced responses requires the characterization of molecular and cellular changes in tissues at site of injection, which may not be fully reﬂected by whole blood analysis. We thus compared granulocyte and macrophage recruitment between baseline, buffer immunization, and 24 h after MVA immunization by histology and ﬂow cytometry. Due to technical constraints, the analysis after vaccination by the IM route was performed under different conditions than that by the SC and ID route, detailed in Materials and Methods. The analysis of tissue slides by histology revealed that ID MVA immunization induced inﬂammatory cell inﬁltration, inﬁltrating mostly subcutaneous tissue, as well as the dermis after 24 h (Figure 7A). SC MVA immunization induced similar local inﬁltration, mostly located in the subcutaneous tissue (Figure 7A). There was no major local inﬂammation after IM MVA immunization (Figure 7B). Looking more precisely the nature of cell populations with ﬂow cytometry, MVA immunization induced clear local recruitment of granulocytes for both the SC and ID routes (SC route baseline vs. MVA 24 h: p = 0.041, buffer 24 h vs. MVA 24 h: p = 0.015, and ID route baseline vs. MVA 24 h: p = 0.0087), as well as macrophages (SC route baseline vs. MVA MVA administration through the IM and ID routes was associated with a Th1-oriented proﬁle, with an increase in MVA-speciﬁc CD8+ T lymphocytes producing IFNg, TNFa, MIP1b, and IL-2. These results conﬁrm that the ID vaccination route induces CD8+ T-cell responses (48–50). Moreover, the IM route was the only one to induce a polyfunctional Th1-like CD4+ T-cell cluster. In contrast, the SC route induced transient increases April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 13 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. f t li i tib di d l l f ti l CD4+ d CD8+ l ti h b t d d bl d d d bl t A C B D FIGURE 7 \| Local inﬂammatory response 24 h after MVA immunization by the SC or ID route. (A) HE staining of the injection site (skin + subcutaneous tissue) and draining inguinal lymph node, showing transversal sections of parafﬁn-embedded tissue cassettes 24 h after ID or SC injection of buffer or rMVA. The scale bar is equal to 50 µm. One representative experiment of two is shown. Frontiers in Immunology \| www.frontiersin.org Innate Response at the Site of Injection Although MVA does not require any speciﬁc receptor for infection (61), MVA-infected cells in the dermis are more abundant than in the SC tissue. This could lead to an increase in the local production of inﬂammatory mediators after ID injection (62). Furthermore, our results suggest that the time to reach the draining lymph node is dependent on the route of administration. Indeed, more abundant levels of E3L were detected in the subcapsular area of draining lymph nodes after SC injection, in accordance with literature on murine models (63, 64), whereas it was moderate after ID injection (Figure S4). Thus, the antigen-presenting cells that are mobilized might differ depending on the administration route. The persistence of the antigen in the skin after ID immunization suggests the possible uptake by local DC subsets, such as Langerhans cells or dermal DCs (57, 65). importance of immunization route for vaccination. Indeed, a major decrease of speciﬁc T cell response after intranasal immunization vs intravenous has been reported after BCG vaccination (66). Concomitantly, a decrease of humoral response, CD8+ T cell response, and protection has also been described after intramuscular immunization in comparison with intravenous using an attenuated malaria vaccine PfSPZ (44). Fowlpox-HIV vaccine has also been shown to differentially activate cytokine production of lung or muscle innate lymphoid cells according to the administration route (45). As a consequence the route of administration must be better considered in vaccine trials. Further investigation is also needed to better understand why a given administration route induces more suitable immunity relative to the others. Such understanding will require identiﬁcation of the key elements that drive the relationship between innate and adaptive responses. The use of systems vaccinology approaches combined with mathematical modeling to identify predictive biomarkers of adaptive immunity appear essential to harness and take advantage of these immune processes. ETHICS STATEMENT The animal study was reviewed and approved by French research Ministry (12-013 and 0201501281731916 (APAF1S#170).02). AUTHOR CONTRIBUTION Conceptualization: PR, YL, RG, and FM. Methodology: PR, ND-B, A-SB, DP, AC, RG, and FM. Validation: PR, RG, and FM. Formal analysis: PR, CJ, and NT. Investigation: PR, AR, CJ, LS, and HH. Resources: NT and HH. Writing—original draft: PR. Writing—review and editing: PR, NT, A-SB, DP, RG, FM, and CJ. Funding acquisition: YL and RG. Supervision: YL, RG, and FM. All authors contributed to the article and approved the submitted version. Conceptualization: PR, YL, RG, and FM. Methodology: PR, ND-B, A-SB, DP, AC, RG, and FM. Validation: PR, RG, and FM. Formal analysis: PR, CJ, and NT. Investigation: PR, AR, CJ, LS, and HH. Resources: NT and HH. Writing—original draft: PR. Writing—review and editing: PR, NT, A-SB, DP, RG, FM, and CJ. Funding acquisition: YL and RG. Supervision: YL, RG, and FM. All authors contributed to the article and approved the submitted version. Despite the data integration, this study did not identify key immune parameters that triggered the difference between the administration routes. Several relevant immune parameters were not included in this study such as NKT cells or innate lymphoid cells due to the design of the gating strategies. In addition, no direct correlation between innate and adaptive response could have been established because separate animal cohorts were used for each type of response due to injection site sampling requirements. The MVA was selected for this project as a model of live attenuated vaccine known to be highly immunogenic in addition to induce strong innate immune response. Other vaccine types such as inactivated, protein, polysaccharides or DNA/RNA are susceptible to use different pathways to activate the immune system. In this regard, further studies are necessary to extrapolate the impact of administration route to other vaccine responses. Frontiers in Immunology \| www.frontiersin.org DATA AVAILABILITY STATEMENT The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Innate Response at the Site of Injection Statistical analysis consisted of a Mann-Whitney test test. of neutralizing antibodies and less polyfunctional CD4+ and CD8+ T cells. Aside from the polyfunctional T-cell phenotypic group, CD4+ and CD8+ T-cell cytotoxic; proliferative, and migrating phenotypic families were also affected by the administration route of MVA to a lesser extent, amplifying the dichotomy between the SC and ID/IM routes. population has been reported as a double edged sword able to either positively or negatively impact pathogenicity (51). In vaccination context, poliovirus speciﬁc cytotoxic CD4+ T have been described to efﬁciently kill in vitro infected cells (52). In line with results of this study, Munier et al. also highlighted a prominent increase of cytotoxic effector CD4+ T phenotype early after vaccinia virus vaccination (53). Among those families, the unsupervised analysis on CD4+ T cells highlighted the presence of CD4+ harboring a cytotoxic phenotype producing perforin and granzyme B. This cell Among the factors that could explain the speciﬁcity of the SC route relative to the ID and IM routes are the tissue composition April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 14 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. and immune properties of the injection site. Indeed, subcutaneous tissue is dedicated to triglyceride storage and metabolic functions and is therefore mostly composed of adipocytes, with only a few macrophages and lymphocytes (54, 55). Nevertheless, the unique microvascularization of adipose tissue, which could facilitate the recruitment of cells from the blood, may amplify the magnitude and persistence of inﬂammation (56). In contrast, skin hosts a large diversity of cells dedicated to immune functions that are abundant at steady state (57). This could result in a better capacity to rapidly resolve the local inﬂammation induced by MVA (58) or conversely augment Ag presentation and the initiation of adaptive immunity. Normal muscle tissue contains only a few resident immune cells. However, IM injection of vaccines leads to signiﬁcant recruitment of immune cells, creating local inﬂammation (59, 60). Another factor linking the immunization site to the immune outcome could be the capacity of MVA to infect local cells or be rapidly driven to the draining lymph node. Innate Response at the Site of Injection Indeed, after ID injection, we observed E3L (an early expressed MVA protein) in the dermis but we did not detected its presence in the SC tissue after SC immunization (Figure S4), suggesting that tissue diffusion is faster after SC than ID administration of MVA. Although MVA does not require any speciﬁc receptor for infection (61), MVA-infected cells in the dermis are more abundant than in the SC tissue. This could lead to an increase in the local production of inﬂammatory mediators after ID injection (62). Furthermore, our results suggest that the time to reach the draining lymph node is dependent on the route of administration. Indeed, more abundant levels of E3L were detected in the subcapsular area of draining lymph nodes after SC injection, in accordance with literature on murine models (63, 64), whereas it was moderate after ID injection (Figure S4). Thus, the antigen-presenting cells that are mobilized might differ depending on the administration route. The persistence of the antigen in the skin after ID immunization suggests the possible uptake by local DC subsets, such as Langerhans cells or dermal DCs (57, 65). and immune properties of the injection site. Indeed, subcutaneous tissue is dedicated to triglyceride storage and metabolic functions and is therefore mostly composed of adipocytes, with only a few macrophages and lymphocytes (54, 55). Nevertheless, the unique microvascularization of adipose tissue, which could facilitate the recruitment of cells from the blood, may amplify the magnitude and persistence of inﬂammation (56). In contrast, skin hosts a large diversity of cells dedicated to immune functions that are abundant at steady state (57). This could result in a better capacity to rapidly resolve the local inﬂammation induced by MVA (58) or conversely augment Ag presentation and the initiation of adaptive immunity. Normal muscle tissue contains only a few resident immune cells. However, IM injection of vaccines leads to signiﬁcant recruitment of immune cells, creating local inﬂammation (59, 60). Another factor linking the immunization site to the immune outcome could be the capacity of MVA to infect local cells or be rapidly driven to the draining lymph node. Indeed, after ID injection, we observed E3L (an early expressed MVA protein) in the dermis but we did not detected its presence in the SC tissue after SC immunization (Figure S4), suggesting that tissue diffusion is faster after SC than ID administration of MVA. REFERENCES 11. Mohanan D, Slütter B, Henriksen-Lacey M, Jiskoot W, Bouwstra JA, Perrie Y, et al. Administration Routes Affect the Quality of Immune Responses: A Cross-Sectional Evaluation of Particulate Antigen-Delivery Systems. J Control Release (2010) 147:342–9. doi: 10.1016/j.jconrel.2010.08.012 1. Mullins DW, Sheasley SL, Ream RM, Bullock TNJ, Fu Y-X, Engelhard VH. Route of Immunization With Peptide-Pulsed Dendritic Cells Controls the Distribution of Memory and Effector T Cells in Lymphoid Tissues and Determines the Pattern of Regional Tumor Control. J Exp Med (2003) 198:1023–34. doi: 10.1084/jem.20021348 12. Bontrop RE. Non-Human Primates: Essential Partners in Biomedical Research. Immunol Rev (2001) 183:5–9. doi: 10.1034/j.1600-065X.2001.1830101.x 13. Hérodin F, Thullier P, Garin D, Drouet M. Nonhuman Primates are Relevant Models for Research in Hematology, Immunology and Virology. Eur Cytokine Netw (2005) 16:104–16. j 2. Henri P, Laurent PE. Intradermal Vaccine Delivery: Will New Delivery Systems Transform Vaccine Administration? Vaccine (2008) 26:3197–208. doi: 10.1016/j.vaccine.2008.03.095 14. Pittman PR, Hahn M, Lee HS, Koca C, Samy N, Schmidt D, et al. Phase 3 Efﬁcacy Trial of Modiﬁed Vaccinia Ankara as a Vaccine Against Smallpox. N Engl J Med (2019) 381:1897–908. doi: 10.1056/nejmoa1817307 3. Abadie V, Bonduelle O, Duffy D, Parizot C, Verrier B, Combadière B. Original Encounter With Antigen Determines Antigen-Presenting Cell Imprinting of the Quality of the Immune Response in Mice. PloS One (2009) 4:e8159. doi: 10.1371/journal.pone.0008159 15. Drexler I, Staib C, Sutter G. Modiﬁed Vaccinia Virus Ankara as Antigen Delivery System: How Can We Best Use Its Potential? Curr Opin Biotechnol (2004) 15:506–12. doi: 10.1016/j.copbio.2004.09.001 4. Wiendl H, Hohlfeld R, Kieseier BC. Immunobiology of Muscle: Advances in Understanding an Immunological Microenvironment. Trends Immunol (2005) 26:373–80. doi: 10.1016/j.it.2005.05.003 16. Vuola JM, Keating S, Webster DP, Berthoud T, Dunachie S, Gilbert SC, et al. Differential Immunogenicity of Various Heterologous Prime-Boost Vaccine Regimens Using DNA and Viral Vectors in Healthy Volunteers. J Immunol (2005) 174:449–55. doi: 10.4049/jimmunol.174.1.449 5. Laurent PE, Bonnet S, Alchas P, Regolini P, Mikszta JA, Pettis R, et al. Evaluation of the Clinical Performance of a New Intradermal Vaccine Administration Technique and Associated Delivery System. Vaccine (2007) 25:8833–42. doi: 10.1016/j.vaccine.2007.10.020 17. Schreiner B, Voss J, Wischhusen J, Dombrowski Y, Steinle A, Lochmüller H, et al. Expression of Toll-Like Receptors by Human Muscle Cells in Vitro and in Vivo: TLR3 is Highly Expressed in Inﬂammatory and HIV Myopathies, Mediates IL-8 Release and Up-Regulation of NKG2D-Ligands. FASEB J (2006) 20:118–20. doi: 10.1096/fj.05-4342fje 6. ACKNOWLEDGEMENTS Supplementary Figure 3 \| Gating strategies used in ﬂow cytometry to study the early innate response in blood. We warmly thank all the teams of IDMIT, as well as VRI, FlowCyTech, ANRS and ADITEC organizations for their crucial contribution and help on this project. Supplementary Figure 4 \| Skin and lymph node tissue after MVA immunization by the SC or ID route Supplementary Figure 5 \| Gating strategy for the ﬂow cytometry of tissue samples. FUNDING This work was supported by the IDMIT infrastructure and funded by the ANR via grant No ANR-11-INBS-0008. N.T. held a fellowship from the ANRS (France Recherche Nord&Sud Sida-HIV Hépatites). This work was also supported by the “Investissements d’Avenir” programs managed by the ANR under reference ANR-10-LABX-77-01 funding the Vaccine Research Institute (VRI), Créteil (ImMemory research program) and ANR-10-EQPX-02–01 funding the FlowCyTech facility Overall, we demonstrate that the administration route affects not only the magnitude but also the quality of the adaptive response to the MVA vaccine model. This modiﬁcation of the adaptive response proﬁle was probably shaped during the very early innate immune response. This work is in line with other studies highlighting the April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 15 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. Supplemental information provided with this manuscript contains 5 additional ﬁgures and 3 tables entitled as below: Supplemental information provided with this manuscript contains 5 additional ﬁgures and 3 tables entitled as below: (IDMIT, Fontenay-aux-Roses, France). Funds were also received from the European Commission: ADITEC, FP7-HEALTH-2011- 280873; Transvac, EU H2020 GA 730964; EHVA, EU H2020 GA 681032. (IDMIT, Fontenay-aux-Roses, France). Funds were also received from the European Commission: ADITEC, FP7-HEALTH-2011- 280873; Transvac, EU H2020 GA 730964; EHVA, EU H2020 GA 681032. Supplementary Figure 1 \| Analysis pipeline of the CyTOF dataset. Supplementary Figure 2 \| Overview of the kinetics of differentially enriched clusters. SUPPLEMENTARY MATERIAL Supplementary Table 1 \| Panel used for CyTOF staining Supplementary Table 1 \| Panel used for CyTOF staining The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/ﬁmmu.2021. 645210/full#supplementary-material Supplementary Table 2 \| Panel used for FlowCytometry staining Supplementary Table 3 \| Dataset of innate immune response REFERENCES Skin Immune Sentinels in Health and Disease. Nat Rev Immunol (2009) 9:679–91. doi: 10.1038/nri2622 58. Serhan CN, Brain SD, Buckley CD, Gilroy DW, Haslett C, O’Neill L a J, et al. Resolution of Inﬂammation: State of the Art, Deﬁnitions and Terms. FASEB J (2007) 21:325–32. doi: 10.1096/fj.06-7227rev 36. Platon L, Pejoski D, Gautreau G, Targat B, Le Grand R, Beignon AS, et al. A Computational Approach for Phenotypic Comparisons of Cell Populations in High-Dimensional Cytometry Data. Methods (2018) 132:66–75. doi: 10.1016/ j.ymeth.2017.09.005 59. Pillon NJ, Bilan PJ, Fink LN, Klip A. Cross-Talk Between Skeletal Muscle and Immune Cells: Muscle-Derived Mediators and Metabolic Implications. Am J Physiol Endocrinol Metab (2013) 304:E453–65. doi: 10.1152/ajpendo.00553.2012 37. Ziegler-Heitbrock L. The CD14+ CD16+ Blood Monocytes: Their Role in Infection and Inﬂammation. J Leukoc Biol (2007) 81:584–92. doi: 10.1189/ jlb.0806510 60. Liang F, Ploquin A, Hernández JD, Fausther-Bovendo H, Lindgren G, Stanley D, et al. Dissociation of Skeletal Muscle for Flow Cytometric Characterization of Immune Cells in Macaques. J Immunol Methods (2015) 425:69–78. doi: 10.1016/j.jim.2015.06.011 38. Sui Y, Hogg A, Wang Y, Frey B, Yu H, Xia Z, et al. Vaccine-Induced Myeloid Cell Population Dampens Protective Immunity to SIV. J Clin Invest (2014) 124:2538–49. doi: 10.1172/JCI73518DS1 j j 61. McFadden G. Poxvirus Tropism. Nat Rev Microbiol (2005) 3:201–13. doi: 10.1038/nrmicro1099 39. Qiu P, Simonds EF, Bendall SC, Gibbs KD, Bruggner RV, Linderman MD, et al. Extracting a Cellular Hierarchy From High-Dimensional Cytometry Data With SPADE. Nat Biotechnol (2011) 29:886–93. doi: 10.1038/ nbt.1991 62. Delaloye J, Roger T, Steiner-Tardivel QG, Le Roy D, Reymond MK, Akira S, et al. Innate Immune Sensing of Modiﬁed Vaccinia Virus Ankara (MVA) is Mediated by TLR2-TLR6, MDA-5 and the NALP3 Inﬂammasome. PloS Pathog (2009) 5:e1000480. doi: 10.1371/journal.ppat.1000480 40. Zaki MJ. SPADE: An Efﬁcient Algorithm for Mining Frequent Sequences. Mach Learn (2001) 42:31–60. doi: 10.1023/A:1007652502315 63. Kadoki M, Patil A, Thaiss CC, Brooks DJ, Pandey S, Deep D, et al. Organism- Level Analysis of Vaccination Reveals Networks of Protection Across Tissues. Cell (2017) 171:398–413.e21. doi: 10.1016/j.cell.2017.08.024 41. Cheng G, Zhong M, Kawaguchi R, Kassai M, Al-Ubaidi M, Deng J, et al. Identiﬁcation of PLXDC1 and PLXDC2 as the Transmembrane Receptors for the Multifunctional Factor PEDF. Elife (2014) 3:e05401. doi: 10.7554/eLife.05401 64. Reynoso GV, Weisberg AS, Shannon JP, McManus DT, Shores L, Americo JL, et al. Lymph Node Conduits Transport Virions for Rapid T Cell Activation. Nat Immunol (2019) 20:602–12. doi: 10.1038/s41590-019-0342-0 42. REFERENCES Phenotypes Using Mass Cytometry and a Two-Step Clustering Analysis. J Immunol (2016) 196:4814–31. doi: 10.4049/jimmunol.1502005 44. Ishizuka AS, Lyke KE, DeZure A, Berry AA, Richie TL, Mendoza FH, et al. Protection Against Malaria At 1 Year and Immune Correlates Following PfSPZ Vaccination. Nat Med (2016) 22:614–23. doi: 10.1038/nm.4110 Phenotypes Using Mass Cytometry and a Two-Step Clustering Analysis. J Immunol (2016) 196:4814–31. doi: 10.4049/jimmunol.1502005 24. Dereuddre-Bosquet N, Baron M-L, Contreras V, Gosse L, Mangeot I, Martinon F, et al. HIV Speciﬁc Responses Induced in Nonhuman Primates With ANRS HIV- Lipo-5 Vaccine Combined With RMVA-HIV Prime or Boost Immunizations. Vaccine (2015) 33:2354–9. doi: 10.1016/J.VACCINE.2015.03.032 45. Li Z, Jackson RJ, Ranasinghe C. Vaccination Route Can Signiﬁcantly Alter the Innate Lymphoid Cell Subsets: A Feedback Between IL-13 and IFN-g. NPJ Vaccines (2018) 3:1–9. doi: 10.1038/s41541-018-0048-6 46. Hodge DR, Hurt EM, Farrar WL. The Role of IL-6 and STAT3 in Inﬂammation and Cancer. Eur J Cancer (2005) 41:2502–12. doi: 10.1016/j.ejca.2005.08.016 25. Rosenbaum P, Tchitchek N, Joly C, Stimmer L, Hocini H, Dereuddre-Bosquet N, et al. Molecular and Cellular Dynamics in the Skin, the Lymph Nodes, and the Blood of the Immune Response to Intradermal Injection of Modiﬁed Vaccinia Ankara Vaccine. Front Immunol (2018) 9:870–89. doi: 10.3389/ﬁmmu.2018.00870 47. Colonna M, Facchetti F. TREM-1 (Triggering Receptor Expressed on Myeloid Cells): A New Player in Acute Inﬂammatory Responses. J Infect Dis (2003) 187 Suppl:S397–401. doi: 10.1086/374754 26. Terra M, Oberkampf M, Fayolle C, Rosenbaum P, Guillerey C, Dadaglio G, et al. Tumor-Derived TGF-b Alters the Ability of Plasmacytoid Dendritic Cells to Respond to Innate Immune Signaling. Cancer Res (2018) 78:3014–26. doi: 10.1158/0008-5472.CAN-17-2719. canres.2719.2017. 48. Gonçalves E, Bonduelle O, Soria A, Loulergue P, Rousseau A, Cachanado M, et al. Innate Gene Signature Distinguishes Humoral Versus Cytotoxic Responses to Inﬂuenza Vaccination. J Clin Invest (2019) 129:1960–71. doi: 10.1172/JCI125372 27. Kolesnikov N, Hastings E, Keays M, Melnichuk O, Tang YA, Williams E, et al. ArrayExpress Update-Simplifying Data Submissions. Nucleic Acids Res (2015) 43:D1113–6. doi: 10.1093/nar/gku1057 49. Liard C, Munier S, Joulin-Giet A, Bonduelle O, Hadam S, Duffy D, et al. Intradermal Immunization Triggers Epidermal Langerhans Cell Mobilization Required for CD8 T-Cell Immune Responses. J Invest Dermatol (2012) 132:615–25. doi: 10.1038/jid.2011.346 28. Bécavin C, Tchitchek N, Mintsa-Eya C, Lesne A, Benecke A. Improving the Efﬁciency of Multidimensional Scaling in the Analysis of High-Dimensional Data Using Singular Value Decomposition. Bioinformatics (2011) 27:1413–21. doi: 10.1093/bioinformatics/btr143 50. REFERENCES Sullivan SP, Koutsonanos DG, Del Pilar Martin M, Lee JW, Zarnitsyn V, Choi SO, et al. Dissolving Polymer Microneedle Patches for Inﬂuenza Vaccination. Nat Med (2010) 16:915–20. doi: 10.1038/nm.2182 29. Kruskal JB, Wish M. Multidimensional Scaling. SAGE Univ Pap Ser Quant Appl Soc Sci (1978). doi: 10.4135/9781412985130 51. Takeuchi A, Saito T. CD4 CTL, a Cytotoxic Subset of CD4+ T Cells, Their Differentiation and Function. Front Immunol (2017) 8:194. doi: 10.3389/ ﬁmmu.2017.00194 30. Gautreau G, Pejoski D, Le Grand R, Cosma A, Beignon AS, Tchitchek N. SPADEVizR: An R Package for Visualization, Analysis and Integration of SPADE Results. Bioinformatics (2017) 33:779–81. doi: 10.1093/ bioinformatics/btw708 52. Wahid R, Cannon MJ, Chow M. Virus-Speciﬁc CD4+ and CD8+ Cytotoxic T-Cell Responses and Long-Term T-Cell Memory in Individuals Vaccinated Against Polio. J Virol (2005) 79:5988–95. doi: 10.1128/jvi.79.10.5988-5995.2005 31. McLachlan GJ. Discriminant Analysis and Statistical Pattern Recognition. Hoboken, New Jersey: Wiley Interscience, John Wiley & Sons, Inc., (2004). 53. Munier CML, van Bockel D, Bailey M, Ip S, Xu Y, Alcantara S, et al. The Primary Immune Response to Vaccinia Virus Vaccination Includes Cells With a Distinct Cytotoxic Effector CD4 T-Cell Phenotype. Vaccine (2016) 34:5251–61. doi: 10.1016/j.vaccine.2016.09.009 32. Tricot S, Meyrand M, Sammicheli C, Elhmouzi-Younes J, Corneau A, Bertholet S, et al. Evaluating the Efﬁciency of Isotope Transmission for Improved Panel Design and a Comparison of the Detection Sensitivities of Mass Cytometer Instruments. Cytometry A (2015) 87:357–68. doi: 10.1002/cyto.a.22648 54. Ibrahim MM. Subcutaneous and Visceral Adipose Tissue: Structural and Functional Differences. Obes Rev (2010) 11:11–8. doi: 10.1111/j.1467- 789X.2009.00623.x 33. Finck R, Simonds EF, Jager A, Krishnaswamy S, Sachs K, Fantl W, et al. Normalization of Mass Cytometry Data With Bead Standards. Cytom Part A (2013) 83 A:483–94. doi: 10.1002/cyto.a.22271 55. Damouche A, Lazure T, Avettand-Fènoël V, Huot N, Dejucq-Rainsford N, Satie AP, et al. Adipose Tissue is a Neglected Viral Reservoir and an Inﬂammatory Site During Chronic HIV and SIV Infection. PloS Pathog (2015) 11:1–28. doi: 10.1371/journal.ppat.1005153 34. Kotecha N, Krutzik PO, Irish JM. Web-Based Analysis and Publication of Flow Cytometry Experiments. Curr Protoc Cytom (2010) 10:1–24. doi: 10.1002/0471142956.cy1017s53 56. Hausman GJ, Richardson RL. Adipose Tissue Angiogenesis. J Anim Sci (2004) 82:925–34. doi: 10.2527/2004.823925x 35. Linderman MD, Bjornson Z, Simonds EF, Qiu P, Bruggner RV, Sheode K, et al. CytoSPADE: High-Performance Analysis and Visualization of High- Dimensional Cytometry Data. Bioinformatics (2012) 28:2400–1. doi: 10.1093/ bioinformatics/bts425 57. Nestle FO, Di Meglio P, Qin J-Z, Nickoloff BJ. REFERENCES Neeland MR, Shi W, Collignon C, Taubenheim N, Meeusen ENT, Didierlaurent AM, et al. The Lymphatic Immune Response Induced by the Adjuvant AS01: A Comparison of Intramuscular and Subcutaneous Immunization Routes. J Immunol (2016) 197:2704–14. doi: 10.4049/ jimmunol.1600817 18. Gherardi MM, Esteban M. Recombinant Poxviruses as Mucosal Vaccine Vectors. J Gen Virol (2005) 86:2925–36. doi: 10.1099/vir.0.81181-0 7. Combadière B, Vogt A, Mahé B, Costagliola D, Hadam S, Bonduelle O, et al. Preferential Ampliﬁcation of CD8 Effector-T Cells After Transcutaneous Application of an Inactivated Inﬂuenza Vaccine: A Randomized Phase I Trial. PloS One (2010) 5:e10818. doi: 10.1371/journal.pone.0010818 19. Gómez CE, Perdiguero B, Garcı́a-Arriaza J, Cepeda V, Sánchez-Sorzano CÓ , Mothe B, et al. A Phase I Randomized Therapeutic MVA-B Vaccination Improves the Magnitude and Quality of the T Cell Immune Responses in HIV-1-Infected Subjects on HAART. PloS One (2015) 10:e0141456. doi: 10.1371/journal.pone.0141456 8. Roozbeh J, Moini M, Lankarani KB, Sagheb MM, Shahpoori S, Bastani B. Low Dose Intradermal Versus High Dose Intramuscular Hepatitis B Vaccination in Patients on Chronic Hemodialysis. Asaio J (2005) 51:242–5. doi: 10.1097/ 01.mat.0000161935.66158.97 20. Gómez CE, Perdiguero B, Garcı́a-Arriaza J, Esteban M. Poxvirus Vectors as HIV/AIDS Vaccines in Humans. Hum Vaccines Immunother (2012) 8:1192– 207. doi: 10.4161/hv.20778 21. Tscherne A, Schwarz JH, Rohde C, Kupke A, Kalodimou G, Limpinsel L, et al. Immunogenicity and Efﬁcacy of the COVID-19 Candidate Vector Vaccine MVA SARS 2 s in Preclinical Vaccination. bioRxiv (2021) 2021.01.09.426032. doi: 10.1101/2021.01.09.426032 9. Belshe RB, Newman FK, Wilkins K, Graham IL, Babusis E, Ewell M, et al. Comparative Immunogenicity of Trivalent Inﬂuenza Vaccine Administered by Intradermal or Intramuscular Route in Healthy Adults. Vaccine (2007) 25:6755–63. doi: 10.1016/j.vaccine.2007.06.066 10. Sheehan S, Harris SA, Satti I, Hokey DA, Dheenadhayalan V, Stockdale L, et al. A Phase I, Open-Label Trial, Evaluating the Safety and Immunogenicity of Candidate Tuberculosis Vaccines AERAS-402 and MVA85A, Administered by Prime-Boost Regime in BCG-Vaccinated Healthy Adults. PloS One (2015) 10:e0141687. doi: 10.1371/journal.pone.0141687 22. Palgen J-L, Tchitchek N, Elhmouzi-Younes J, Delandre S, Namet I, Rosenbaum P, et al. Prime and Boost Vaccination Elicit a Distinct Innate Myeloid Cell Immune Response. Sci Rep (2018) 8:3087. doi: 10.1038/s41598-018-21222-2 23. Pejoski D, Tchitchek N, Rodriguez Pozo A, Elhmouzi-Younes J, Yousﬁ-Bogniaho R, Rogez-Kreuz C, et al. Identiﬁcation of Vaccine-Altered Circulating B Cell April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 16 Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. 66. Darrah PA, Zeppa JJ, Hackney JA, Wadsworth MHI, Hughes TK, Pokkali S, et al. Prevention of Tuberculosis in Nonhuman Primates Following Intravenous BCG Immunization. Nat Press (2019) 577:95–102. doi: 10.1038/s41586-019-1817-8 Copyright © 2021 Rosenbaum, Tchitchek, Joly, Rodriguez Pozo, Stimmer, Langlois, Hocini, Gosse, Pejoski, Cosma, Beignon, Dereuddre-Bosquet, Levy, Le Grand and Martinon. 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. 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. Vaccine Inoculation Route Modulates Immunity Rosenbaum et al. April 2021 \| Volume 12 \| Article 645210 REFERENCES Mbikay M, Seidah NG, Chrétien M. Neuroendocrine Secretory Protein 7B2: Structure, Expression and Functions. Biochem J (2015) 357:329–42. doi: 10.1042/bj3570329 43. Sauzay C, Voutetakis K, Chatziioannou A, Chevet E, Avril T. CD90/Thy-1, a Cancer-Associated Cell Surface Signaling Molecule. Front Cell Dev Biol (2019) 7:66. doi: 10.3389/fcell.2019.00066 65. Kupper T, Fuhlbrigge R. Immune Surveillance in the Skin: Mechanisms and Clinical Consequences. Nat Rev Immunol (2004) 4:211–22. doi: 10.1038/ nri1310 April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 17 Vaccine Inoculation Route Modulates Immunity 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. April 2021 \| Volume 12 \| Article 645210 Frontiers in Immunology \| www.frontiersin.org 18
https://openalex.org/W2155708347	http://www.ajnr.org/content/ajnr/35/11/2039.full.pdf	English	null	Neuroimaging of Diving-Related Decompression Illness: Current Knowledge and Perspectives	American journal of neuroradiology	2,014	cc-by	4,543	ABSTRACT SUMMARY: Diving-related decompression illness is classiﬁed into 2 main categories: arterial gas embolism and decompression sickness. The latter is further divided into types 1 and 2, depending on the clinical presentation. MR imaging is currently the most accurate neuroimaging technique available for the detection of brain and spinal cord lesions in neurologic type 2 decompression sickness. Rapid bubble formation in tissues and the bloodstream during ascent is the basic pathophysiologic mechanism in decompression illness. These bubbles can damage the central nervous system through different mechanisms, namely arterial occlusion, venous obstruction, or in situ toxicity. Neuroimaging studies of decompression sickness have reported ﬁndings associated with each of these mechanisms: some typical results are summarized and illustrated in this article. We also review the limitations of previous work and make practical methodologic suggestions for future neuroimaging studies. type 2 decompression sickness, MR imaging clearly appears to be the most accurate for detecting pathologic changes in the brain and spinal cord.7,10 To date, several MR imaging studies of neu- rologic type 2 decompression sickness have been reported with different and often controversial results regarding the sensitivity and utility of this technique in patient evaluation and manage- ment.8,11-16 These discrepancies could be explained by lack of standardized imaging protocols, variable delays between symp- tom onset and image acquisition, and the high variability of the study design and technical devices. Our aim here is to provide a comprehensive review of neuroimaging studies of neurologic type 2 decompression sickness, discuss their limitations, and make proposals that could help improve the quality of future clinicora- diologic studies in the field. T he term “decompression illness” encompasses all clinical manifestations induced by a rapid decrease of environmental pressure, sufficient to cause the formation of inert gas bubbles previously loaded within tissues or blood as a soluble phase. This can occur in various circumstances, including ascent from diving, flying or climbing mountains immediately after a dive, and exer- cising in hyperbaric/hypobaric chambers.1-3 Diving-related de- compression illness is classified into 2 main categories: arterial gas embolism due to pulmonary decompression barotrauma and de- compression sickness.4 There are 2 clinically defined types of de- compression sickness. Type 1 refers to relatively mild symptoms such as joint pain, skin marbling, small patchy hemorrhages, and lymphatic obstruction. A. Kleinschmidt and M.I. Vargas J. Kamtchum Tatuene, R. Pignel, P. Pollak, K.O. Lovblad, of October 23, 2024. This information is current as REVIEW ARTICLE From the Neurology Division (J.K.T., P.P., A.K.), Department of Clinical Neurosci- ences; Hyperbaric Medicine Centre (R.P.), Department of Emergency Medicine; and Neuroradiology Division (K.O.L., M.I.V.), Department of Imaging and Medical Infor- mation Sciences, Geneva University Hospitals, Geneva, Switzerland. Please address correspondence to Maria Isabel Vargas, MD, Geneva University Hospitals, Division of Neuroradiology, DISIM, Rue Gabrielle-Perret-Gentil 4, 1211 Gene`ve 14, Switzerland; e-mail: Maria.I.Vargas@hcuge.ch Indicates open access to non-subscribers at www.ajnr.org Indicates article with supplement on-line table. http://dx.doi.org/10.3174/ajnr.A4005 AJNR Am J Neuroradiol 35:2039–44 Nov 2014 www.ajnr.org ABSTRACT Conversely, type 2 includes more serious and often life-threatening symptoms and can be further divided into 4 subtypes or syndromes according to the organ affected— namely the brain, spinal cord, inner ear, or lung.5,6 Spinal cord lesions represent the majority of central nervous system insults associated with type 2 decompression sickness.7-9 T Imaging Findings and Related Pathophysiologic Theories Reversiblethoracicspinalcordlesionduetodecompressionsickness.MRimagingperformed 24 hours after the diving accident shows a thoracic lesion appearing as a high signal on sagittal T2WI (A,whitearrows).Thelesionincreasesinsizeonthefollowingday(B,whitearrows)andsubsequently disappears on day 13 (C). This evolution may also be consistent with edema. (Figs 1, 2, and 3A). Second, in spinal cord decompression sickness, there is predomi- nant involvement of the thoracic cord seg- ments (Figs 2 and 3B), which may be ex- plained by the following: 1) their high fat content,allowinghighconcentrationsofni- trogen during diving, especially in the lat- eral and posterior columns5,7,14,15; and 2) theirrelativelylowmobilityandbloodflow, responsible for poor nitrogen bubble wash- out during ascent and hence a higher risk of ischemia.18-20 The exact mechanism through which nitrogen bubbles cause brain and spine damage is still highly debated, and there are currently 3 theories supporting 3 dif- ferent pathophysiologic mechanisms: ar- terial occlusion, venous infarction, and in situ nitrogen toxicity. The following sup- port the arterial occlusion theory: 1) the predominance of cerebral lesions in arte- rial gas embolisms with usually a clinical and radiologic stroke-like presentation (restricted diffusion appearing as hyperintensity on DWI with low values on the ADC map, high signal on T2WI, and usually no contrast enhancement) (Fig 3A).7,21,22 2) The finding that nitro- gen bubbles can interrupt arterial blood supply to the brain and spine either by direct obstruction of small capillaries or by activa- tion of pathologic clotting at the blood-bubble interface.23-26 This effect on clotting is increased in case of low hematocrit, explaining the higher risk of severe cardiopulmonary and neurologic decom- pression sickness in dehydrated divers.27-30 3) The higher preva- lence of a patent foramen ovale in patients experiencing decom- pression illness31-39 could favor paradoxic embolization of either nitrogen bubbles or pathologically formed thrombi with subse- quent ischemic infarcts in the brain40 or spine, where the collat- eral circulation network is less robust.41 4) The presence of exten- sive gray matter lesions in some neurologic decompression sickness cases—gray matter lesions being typically seen in arterial MRimagingperformed h signal on sagittal T2WI rows)andsubsequently FIG 2. Reversiblethoracicspinalcordlesionduetodecompressionsickness.MRimagingperformed 24 hours after the diving accident shows a thoracic lesion appearing as a high signal on sagittal T2WI (A,whitearrows).Thelesionincreasesinsizeonthefollowingday(B,whitearrows)andsubsequently disappears on day 13 (C). This evolution may also be consistent with edema. increasing amounts of nitrogen dissolve and accumulate in lipid components of tissues. As a diver ascends, there is a sequential well-orchestrated release of the nitrogen dissolved in tissues and blood. Imaging Findings and Related Pathophysiologic Theories Imaging Findings and Related Pathophysiologic Theories The fundamental reason why decompression illness occurs is that during ascent from depth, gas bubbles form more rapidly in tis- sues and the bloodstream than the body can eliminate through diffusion and perfusion processes.8 At this point, some basic physics principles will help in understanding the mechanism of bubble formation. With increasing depth, the pressure of the air found in pulmonary alveoli increases because of compression at relatively constant temperature as stated by Boyle’s law. There- fore, the partial pressure of nitrogen in pulmonary alveoli also increases according to Dalton’s law. This change leads to an in- crease in nitrogen diffusion into the blood through the alveoli membrane according to Henry’s principle of soluble gas pressure equilibration on either side of a permeable membrane. With time, Among the imaging modalities available to study neurologic From the Neurology Division (J.K.T., P.P., A.K.), Department of Clinical Neurosci- ences; Hyperbaric Medicine Centre (R.P.), Department of Emergency Medicine; and Neuroradiology Division (K.O.L., M.I.V.), Department of Imaging and Medical Infor- mation Sciences, Geneva University Hospitals, Geneva, Switzerland. Please address correspondence to Maria Isabel Vargas, MD, Geneva University Hospitals, Division of Neuroradiology, DISIM, Rue Gabrielle-Perret-Gentil 4, 1211 Gene`ve 14, Switzerland; e-mail: Maria.I.Vargas@hcuge.ch Indicates open access to non-subscribers at www.ajnr.org Indicates article with supplement on-line table. http://dx.doi.org/10.3174/ajnr.A4005 2039 FIG 1. Reversible cervical spinal cord lesion due to decompression sickness. MR imaging shows an extensive lesion causing enlargement of the cervical spinal cord. The lesion appears as a high signal on sagittal (A) and axial T2WI (D and E) and on DWI (B) without concomitant signal decrease on the ADC map (C). All these characteristics, with complete healing without a scar as shown on sagittal T2WI (F) obtained 2 weeks later, are consistent with vasogenic edema (referred to as an “ischemic-like” lesion in the On-line Table). FIG 1. Reversible cervical spinal cord lesion due to decompression sickness. MR imaging shows an extensive lesion causing enlargement of the cervical spinal cord. The lesion appears as a high signal on sagittal (A) and axial T2WI (D and E) and on DWI (B) without concomitant signal decrease on the ADC map (C). All these characteristics, with complete healing without a scar as shown on sagittal T2WI (F) obtained 2 weeks later, are consistent with vasogenic edema (referred to as an “ischemic-like” lesion in the On-line Table). FIG 2. 40 Kamtchum Tatuene Nov 2014 www.ajnr.org Imaging Findings and Related Pathophysiologic Theories This theory of in situ nitrogen toxicity, also called the “autoch- thonous theory,”14 may, at least partially, explainsymptomsrecordedinpatientswith type 1 decompression sickness, nitrogen narcosis,andtype2decompressionsickness without objective abnormalities on brain or spine MR imaging.48,49 Authors of MR imaging studies of de- compression illness frequently report that the different clinicoradiologic patterns mentioned above overlap in the same pa- tient, thereby suggesting that no single unifying pathophysiologic mechanism could completely explain this complex condition.14,16,47 Some have argued that there is a greater participation of arterial oc- clusion in brain lesions because they are more frequently described in cases of arte- rial gas embolism, while venous occlusion may play a greater role in spine lesions and may even be favored by local me- chanical cord compression.11,15 Whatever the preponderant mechanism involved, the treatment is urgent hyperbaric therapy to allow nitrogen to dissolve and be expelled via the lungs. ssionsickness(A,B,and of restricted diffusion hsides.Theyappearas 4, white arrows) T2WI trast enhancement on presentonthesebrain um (3 and 4). Lesions of now appear as “little of the central nervous FIG3. Examplesofirreversiblebrainandspinalcordlesionsduetodecompressionsickness(A,B,and C are taken from the same patient). A, On DWI (1, 2, 3, 4, 5, 6), the areas of restricted diffusion correspondtolesionsofthecorpuscallosumandfrontalwhitematteronbothsides.Theyappearas high signal on T2WI (7, 8, 9). B, Sagittal (1 and 2, white arrows) and axial (3 and 4, white arrows) T2WI shows a cervicothoracic (C7–T1) lesion appearing as high signal without contrast enhancement on T1WI(2).C,Onfollow-up1monthlater,thelesionsdescribedinAandBarestillpresentonthesebrain (1, 2, 3, 4) and spine (5, 6) images. Note the enhancement of the corpus callosum (3 and 4). Lesions of the right frontal deep white matter and corpus callosum have healed and now appear as “little cavities.” Similar features are usually seen in ischemic or necrotic lesions of the central nervous system. infarction processes.11,18,42 5) Hypoperfused areas compatible with embolic cerebral arterial occlusion are observed in some pa- tients with decompression illness when using SPECT with hexa- methylpropyleneamine oxime (HMPAO) marked with a meta- stable nuclear isomer of the radioisotope technetium Tc99m (the product is sometimes referred to as exametazime).43,44 Some general rules have been derived from available MR im- aging studies. Imaging Findings and Related Pathophysiologic Theories The nitrogen in the alveoli expands first and is eliminated; then, the nitrogen in blood is eliminated and its partial pressure in blood decreases, thus allowing saturated tissues to release nitro- gen back into the bloodstream for subsequent pulmonary elimi- nation. During fast ascents, the process of progressive nitrogen release does not take place properly and there is premature expan- sion of the nitrogen trapped in tissues and blood, leading to ex- cessive bubble formation and subsequent tissue damage.1,17 The critical role of nitrogen bubbles in tissue damage is highly supported by some typical radiologic findings. First, the high-fat content of the myelin sheath with high solubility of nitrogen in fat explains the preponderance of white matter lesions either in the brain or the spine of patients with neurologic decompression illness 2040 FIG3. Examplesofirreversiblebrainandspinalcordlesionsduetodecompressionsickness(A,B,and C are taken from the same patient). A, On DWI (1, 2, 3, 4, 5, 6), the areas of restricted diffusion correspondtolesionsofthecorpuscallosumandfrontalwhitematteronbothsides.Theyappearas high signal on T2WI (7, 8, 9). B, Sagittal (1 and 2, white arrows) and axial (3 and 4, white arrows) T2WI shows a cervicothoracic (C7–T1) lesion appearing as high signal without contrast enhancement on T1WI(2).C,Onfollow-up1monthlater,thelesionsdescribedinAandBarestillpresentonthesebrain onstrated venous obstruction and white matter congestion.45,46 Third, vasogenic edema, appearing as high signal on both DWI and ADC maps, was observed by Vollmann et al in 201147 in a case of spinal cord decompression sickness, also sug- gesting a venous rather than arterial pathologic process. The third theory of in situ nitrogen toxicity is more speculative and posits that nitrogen bubbles found in intercellu- lar spaces may have a direct toxic effect on neurons, causing alterations of the electri- cal properties of membranes and impair- ment of ion flow regulation with subse- quent cytotoxic edema and cell death. Areas of necrosis may be viewed as high- signal lesions on T2WI and as low signal on T1WI. Additionally, bubbles within the myelin sheath might cause alterations of nerve conduction. Imaging Findings and Related Pathophysiologic Theories First, any time a lesion is found on radiologic im- ages, there is good correspondence with the clinical symptoms or syndromes described.8,11 Second, normal MR imaging findings of the spinal cord do not rule out a diagnosis of decompression ill- ness.15,50,51 Third, in regard to evolution and prognosis, clinico- radiologic discrepancies have been observed (either worsening of le- sions on MR imaging despite clinical improvement as shown in Fig 2 or improvement in MR imaging lesions despite neurologic deterio- ration19). Furthermore, patients with hemorrhagic cord injury have theworstneurologicoutcome;patientswithmerespinalcordedema do better; and patients with normal-appearing spinal cords AJNR Am J Neuroradiol 35:2039–44 Nov 2014 www.ajnr.org 2041 On the other hand, various radiologic and histopathologic findings have been published to support the venous infarction theory. First, the more frequent occurrence of lateral and poste- rior column white matter lesions compared with gray matter le- sions in spinal cord decompression sickness is more likely a con- sequence of an obstruction of the slow-flowing epidural venous bed by nitrogen bubbles, which leads to vasogenic edema (Figs 1 and 2).5,8 In some cases, there may be additional venous infarc- tion (Fig 3B, -C). Similar lesions have been provoked experimen- tally in dogs.9 Second, several histopathologic studies have dem- FIG 3. Continued. brain or spine, which then would not be depicted by posttreatment imaging. However, several other limitations of available studies could have had a nega- tive impact on the ability to detect central nervous system lesions. First, none of those studies have used high-field mag- nets (3T). Most of the authors used a 1.5T system,11,47,49,55 and some even used a 0.5T system, which has a very low sensi- tivity.8,56 Second, with the exception of Vollmann et al,47 no previously published MR imaging studies of decompression sickness used DWI and ADC maps. The use of these sequences could conceivably increase MR imaging sensitivity and even provide additional data on the patho- physiology of brain damage in divers. The same is true for DTI, which has been used only once in animal models of decom- pression sickness.57 The third limitation is that only one56 of the previous MR im- aging studies included systematic imaging of both brain and spine. This is probably because they were all retrospective studies in which the part of the central nervous system explored was chosen in accor- dance with the clinical syndrome. Imaging Findings and Related Pathophysiologic Theories Because subclinical brain lesions have been re- ported in decompression sickness,58 we think that the predominance of symp- toms and signs related to spinal cord le- sions could have obscured the possibility of brain damage, thus leading to underes- timation of the overall prevalence of cen- tral nervous system lesions, with a nega- tive impact on MR imaging sensitivity. k d ll h Taking into consideration all the aforementioned methodologic limita- tions, we would recommend using higher field strengths (1.5 or 3T) and, systematically, joint brain and spine imaging in future MR imaging studies of decompression illness. We tentatively sug- gest the following imaging protocol for the brain: T1WI, T2WI, FLAIR, gradient-echo, DWI and ADC mapping, and contrast- enhanced T1WI. For spinal cord exploration, we would recom- mend the following sequences: sagittal fast spin-echo T2 and spin- echo T1 in the acute phase without and with contrast medium, which may be very useful in the subacute phase. We also encour- age clinicians and radiologists to use more sophisticated tech- niques such as DTI, which could provide deeper insight into pathophysiologic processes. FIG 3. Continued. FIG 3. Continued. consistently have a good outcome, though the lack of signal abnor- malities on MR imaging does not necessarily mean that the patient will not develop sequelae because the negative predictive value of an investigation with normal findings is 77%.15,16,52 Anecdotal findings have been reported in patients with de- compression illness. These include intracranial hemorrhages53 and extensive gas inclusions in CSF spaces.54 The On-line Table gives an overview of the major characteristics of lesions most fre- quently found on MR imaging in type 2 decompression sickness. 2042 Kamtchum Tatuene Nov 2014 www.ajnr.org REFERENCES 27. Fahlman A, Dromsky DM. Dehydration effects on the risk of severe decompression sickness in a swine model. Aviat Space Environ Med 2006;77:102–06 1. Elliott DH, Moon RE. Manifestations of the Decompression Disorders: The Physiology and Medicine of Diving. 4th ed. London: Saunders; 1993:481 28. Williams ST, Prior FG, Bryson P. Hematocrit change in tropical scuba divers. Wilderness Environ Med 2007;18:48–53 2. Faralli F. Decompression illness. In: Oriani G, Marroni A, Wattel F, eds. Handbook on Hyperbaric Medicine. Berlin: Springer-Verlag; 1996:135–82 29. Newton HB, Burkart J, Pearl D, et al. Neurological decompression illness and hematocrit: analysis of a consecutive series of 200 recre- ational scuba divers. Undersea Hyperb Med 2008;35:99–106 3. Pulley SA. Decompression sickness. Medscape. 2012 (updated June 14, 2012). http://emedicine.medscape.com/article/769717-overview. Accessed December 2, 2013 y 30. Gempp E, Blatteau JE, Pontier JM, et al. Preventive effect of pre-dive hydration on bubble formation in divers. Br J Sports Med 2009; 43:224–28 4. Melamed Y, Shupak A, Bitterman H. Medical problems associated with underwater diving. N Engl J Med 1992;326:30–35 31. Germonpre´ P, Dendale P, Unger P, et al. Patent foramen ovale and decompression sickness in sports divers. J Appl Physiol 1998; 84:1622–26 5. Manabe Y, Sakai K, Kashihara K, et al. Presumed venous infarction in spinal decompression sickness. AJNR Am J Neuroradiol 5. Manabe Y, Sakai K, Kashihara K, et al. Presumed venous infarction in spinal decompression sickness. AJNR Am J Neuroradiol 1998;19:1578–80 32. Bove AA. Risk of decompression sickness with patent foramen ovale. Undersea Hyperb Med 1998;25:175–78 6. Newton HB. Neurologic complications of scuba diving. Am Fam Physician 2001;63:2211–18 33. Wilmshurst PT, Nightingale S, Walsh KP, et al. Effect on migraine of closure of cardiac right-to-left shunts to prevent recurrence of de- compression illness or stroke or for haemodynamic reasons. Lancet 2000;356:1648–51 7. Warren LP Jr, Djang WT, Moon RE, et al. Neuroimaging of scuba diving injuries to the CNS. AJR Am J Roentgenol 1988;151:1003–08 8. Sparacia G, Banco A, Sparacia B, et al. Magnetic resonance findings in scuba diving-related spinal cord decompression sickness. MAGMA 1997;5:111–15 34. Schwerzmann M, Seiler C. Recreational scuba diving, patent foramen ovale and their associated risks. SwissMedWkly2001;131:365–74 9. Hallenbeck JM, Bove AA, Elliott DH. Mechanisms underlying spinal cord damage in decompression sickness. Neurology 1975;25:308–16 35. Schwerzmann M, Seiler C, Lipp E, et al. Relation between directly detected patent foramen ovale and ischemic brain lesions in sport divers. Ann Int Med 2001;134:21–24 10. Disclosures: Andreas Kleinschmidt—UNRELATED: Other: Senior Editor for the jour- nal Neuroimage (Elsevier). CONCLUSIONS 19. Yoshiyama M, Asamoto S, Kobayashi N, et al. Spinal cord decom- pression sickness associated with scuba diving: correlation of im- mediate and delayed magnetic resonance imaging findings with se- verity of neurologic impairment: a report on 3 cases. Surg Neurol 2007;67:283–87 Several neuroimaging studies of decompression illness have been conducted during the past decades. They have helped identify the spectrum of central nervous system lesions in neurologic decom- pression sickness and have provided useful data for understand- ing the related possible pathophysiologic mechanisms. However, due to technical and methodologic limitations of the available studies, many questions remain unanswered, especially those per- taining to the role of MR imaging in evaluating treatment proto- cols and estimating prognosis. Therefore, it seems obvious that well-designed prospective studies integrating new imaging tech- niques and better designed protocols should be conducted to ex- plore these aspects and eventually provide data that help improve the management of patients with decompression illness. 20. Hierholzer J, Tempka A, Stroszczynski C, et al. MRI in decompres- sion illness. Neuroradiology 2000;42:368–70 21. Moon RE. Gas embolism. In: Oriani G, Marroni A, Wattel F, eds. Handbook on Hyperbaric Medicine. Berlin: Springer-Verlag; 1996:229 22. Leitch DR, Green RD. Pulmonary barotrauma in divers and the treatment of cerebral arterial gas embolism. Aviat Space Environ Med 1986;57(10 pt 1):931–38 23. Hjelde A, Bergh K, Brubakk AO, et al. Complement activation in divers after repeated air/heliox dives and its possible relevance to DCS. J Appl Physiol 1995;78:1140–44 24. Huang KL, Lin YC. Activation of complement and neutrophils in- creases vascular permeability during air embolism. Aviat Space En- viron Med 1997;68:300–05 25. Dufka AJ, Francis TJ. Pathophysiology of decompression sickness. In: Bove AA, ed. Diving Medicine. 3rd ed. Philadelphia: Saunders; 1997:159–75 26. Boussuges A, Succo E, Juhan-Vague I, et al. Activation of coagulation in decompression illness. Aviat Space Environ Med 1998;69:129–32 Critical Appraisal of Available Imaging Studies and Suggested Future Directions Some authors have tried to identify prognostic factors in de- compression sickness.59,60 They have even proposed a prospec- tively validated clinical prognostic score.61-64 Added to that, the prognostic value of MR imaging findings has been independently evaluated, though in spinal cord decompression sickness only.15,16 However, there is currently no available study that tried To date, all neuroimaging studies have reported the low sensitivity of MR imaging for detecting central nervous system lesions in patients with decompression sickness.8,11,49 This could be ex- plained, in part, by the delay between the diving accident and the imaging study because urgent hyperbaric recompression therapy is the priority. The latter could reverse signal alterations in the p p plained, in part, by the delay between the diving accident and the imaging study because urgent hyperbaric recompression therapy is the priority. The latter could reverse signal alterations in the 2042 Kamtchum Tatuene Nov 2014 www.ajnr.org 2042 outcome in 45 divers with spinal cord decompression sickness. Aviat Space Environ Med 2008;79:1112–16 to integrate neuroimaging findings into a prognostic score. More- over, the importance of serial brain and spine imaging in the follow-up of patients with decompression sickness has not been studied. Such serial imaging could be used in comparative evalu- ations of the various treatment protocols currently available be- cause there is no consensus on which is the best.65 16. Hennedige T, Chow W, Ng YY, et al. MRI in spinal cord decompres- sion sickness. J Med Imaging Radiat Oncol 2012;56:282–88 17. McMullin AM. Scuba diving: What you and your patients need to know. Cleve Clin J Med 2006;73:711–12, 714, 716 passim 18. Aharon-Peretz J, Adir Y, Gordon CR, et al. Spinal cord decompres- sion sickness in sport diving. Arch Neurol 1993;50:753–56 AJNR Am J Neuroradiol 35:2039–44 Nov 2014 www.ajnr.org 204 NR Am J Neuroradiol 35:2039–44 Nov 2014 www.ajnr.org REFERENCES Hodgson M, Beran RG, Shirtley G. The role of computed tomogra- phy in the assessment of neurologic sequelae of decompression sickness. Arch Neurol 1988;45:1033–35 36. Cartoni D, De Castro S, Valente G, et al. Identification of profes- sional scuba divers with patent foramen ovale at risk for decom- pression illness. Am J Cardiol 2004;94:270–73 11. Reuter M, Tetzlaff K, Hutzelmann A, et al. MR imaging of the central nervous system in diving-related decompression illness. Acta Ra- diol 1997;38:940–44 37. Torti SR, Billinger M, Schwerzmann M, et al. Risk of decompression illness among 230 divers in relation to the presence and size of pat- ent foramen ovale. Eur Heart J 2004;25:1014–20 12. Yanagawa Y, Okada Y, Terai C, et al. MR imaging of the central nervous system in divers. Aviat Space Environ Med 1998;69:892–95 38. Harrah JD, O’Boyle PS, Piantadosi CA. Underutilization of echocar- diography for patent foramen ovale in divers with serious decom- pression sickness. Undersea Hyperb Med 2008;35:207–11 13. Hutzelmann A, Tetzlaff K, Reuter M, et al. Does diving damage the brain? MR control study of divers’ central nervous system. Acta Ra- diol 2000;41:18–21 14. Kei PL, Choong CT, Young T, et al. Decompression sickness: MRI of the spinal cord. J Neuroimaging 2007;17:378–80 39. Koch AE, Kirsch H, Reuter M, et al. Prevalence of patent foramen ovale (PFO) and MRI-lesions in mild neurological decompression sickness (type B-DCS/AGE). Undersea Hyperb Med 2008;35:197–205 15. Gempp E, Blatteau JE, Stephant E, et al. MRI findings and clinical 2043 AJNR Am J Neuroradiol 35:2039–44 Nov 2014 www.ajnr.org 40. Palmer AC, Calder IM, Yates PO. Cerebral vasculopathy in divers. Neuropathol Appl Neurobiol 1992;18:113–24 53. Josefsen R, Wester K. Cerebellar hemorrhage: a rare, but serious complication in decompression disease [in Norwegian]. Tidsskr Nor Laegeforen 1999;119:3901–02 p pp 41. Mastaglia FL, McCallum RI, Walder DN. Myelopathy associated with decompression sickness: a report of six cases. Clin Exp Neurol 1983;19:54–59 g f 54. Ozdoba C, Weis J, Plattner T, et al. Fatal scuba diving incident with massive gas embolism in cerebral and spinal arteries. Neuroradiol- ogy 2005;47:411–16 42. McCormac J, Mirvis SE, Cotta-Cumba C, et al. Spinal myelopathy resulting from decompression sickness: MR findings in a case and review of the literature. Emerg Radiol 2002;9:240–42 55. Gao GK, Wu D, Yang Y, et al. Cerebral magnetic resonance imaging of compressed air divers in diving accidents. Undersea Hyperb Med 2009;36:33–41 43. 2044 Kamtchum Tatuene Nov 2014 www.ajnr.org REFERENCES Macleod MA, Adkisson GH, Fox MJ, et al. 99Tcm-HMPAO single photon emission tomography in the diagnosis of cerebral baro- trauma. Br J Radiol 1988;61:1106–09 56. Todnem K, Nyland H, Skeidsvoll H, et al. Neurological long term consequences of deep diving. Br J Ind Med 1991;48:258–66 44. Adkisson GH, Macleod MA, Hodgson M, et al. Cerebral perfusion deficits in dysbaric illness. Lancet 1989;2:119–22 57. Hutchinson EB, Sobakin AS, Meyerand ME, et al. Diffusion tensor MRI of spinal decompression sickness.UnderseaHyperbMed2013;40:23–31 45. Royon M, Wolkiewiez J. Bilan scanographique et phle´bographique d’un accident me´dullaire de de´compression. Med Sub Hyp 1987; 4:117–20 58. Aksoy FG. MR imaging of subclinical cerebral decompression sickness: a case report. Acta Radiol 2003;44:108–10 59. Dick AP, Massey EW. Neurologic presentation of decompression sick- ness and air embolism in sportdivers.Neurology1985;35:667–71 46. Kim RC, Smith HR, Henbest ML, et al. Nonhemorrhagic venous infarction of the spinal cord. Ann Neurol 1984;15:379–85 60. Ball R. Effect of severity, time to recompression with oxygen, and re-treatment on outcome in forty-nine cases of spinal cord decom- pression sickness. Undersea Hyperb Med 1993;20:133–45 47. Vollmann R, Lamperti M, Magyar M, et al. Magnetic resonance im- aging of the spine in a patient with decompression sickness. Clin Neuroradiol 2011;21:231–33 yp 61. Boussuges A, Thirion X, Blanc P, et al. Neurologic decompression illness: a gravity score. Undersea Hyperb Med 1996;23:151–55 48. Levin HS, Goldstein FC, Norcross K, et al. Neurobehavioral and magnetic resonance imaging findings in two cases of decompres- sion sickness. Aviat Space Environ Med 1989;60:1204–10 62. Gempp E, Blatteau JE. Risk factors and treatment outcome in scuba divers with spinal cord decompression sickness. J Crit Care 2010;25:236–42 49. Grønning M, Risberg J, Skeidsvoll H, et al. Electroencephalography and magnetic resonance imaging in neurological decompression sickness. Undersea Hyperb Med 2005;32:397–402 63. Blatteau JE, Gempp E, Constantin P, et al. Risk factors and clinical out- comeinmilitarydiverswithneurologicaldecompressionsickness:in- fluenceoftimetorecompression.DivingHyperbMed2011;41:129–34 50. Tournebise H, Boucand MH, Landi J, et al. Paraplegia and decom- pression sickness. Paraplegia 1995;33:636–39 64. Blatteau JE, Gempp E, Simon O, et al. Prognostic factors of spinal cord decompressionsicknessinrecreationaldiving:retrospectiveandmul- ticentric analysis of 279 cases.NeurocritCare2011;15:120–27 51. Colineau X, Muyard B, Paoletti H, et al. Accidents neurologiques de de´compression: mise en e´vidence par IRM. Med Armees 1998; 26:371–79 65. Moon RE, Gorman DF. Treatment of decompression disorders. In: Bennett PB, Elliott DH, eds. The Physiology and Medicine of Diving. 4th ed. London: WB Saunders; 1993:506–41 52. WhiteML,El-KhouryGY.Neurovascularinjuriesofthespinalcord. REFERENCES Eur J Radiol 2002;42:117–26 2044 Kamtchum Tatuene Nov 2014 www.ajnr.org Kamtchum Tatuene Nov 2014 www.ajnr.org 2044
https://openalex.org/W3179678468	https://research.ulapland.fi/files/22215214/1_s2.0_S014362282100117X_main_1.pdf	English	null	Land-use and climate related drivers of change in the reindeer management system in Finland: Geography of perceptions	Applied geography	2,021	cc-by	17,633	University of Lapland This is a self-archived version of an original article. This version usually differs somewhat from the publisher’s final version, if the self-archived version is the accepted author manuscript. University of Lapland This is a self-archived version of an original article. This version usually differs somewhat from the publisher’s final version, if the self-archived version is the accepted author manuscript. University of Lapland This is a self-archived version of an original article. This version usually differs somewhat from the publisher’s final version, if the self-archived version is the accepted author manuscript. Land-use and climate related drivers of change in the reindeer management system in Finland Rasmus, Sirpa; Wallen, Henri; Turunen, Minna; Landauer, Mia; Tahkola, Juho; Jokinen, Mikko; Laaksonen, Sauli Published in: University of Lapland Citation for pulished version (APA): Rasmus, S., Wallen, H., Turunen, M., Landauer, M., Tahkola, J., Jokinen, M., & Laaksonen, S. (2021). Land-use and climate related drivers of change in the reindeer management system in Finland: Geography of perceptions. Applied Geography, 134, Article 102501. https://doi.org/10.1016/j.apgeog.2021.102501 Document License CC BY Download date: 24. Oct. 2024 University of Lapland University of Lapland This is a self-archived version of an original article. This version usually differs somewhat from the publisher’s final version, if the self-archived version is the accepted author manuscript. Sirpa Rasmus a,e,, Henri Wallen a,f, Minna Turunen a, Mia Landauer a,g, Juho Tahkola b, Mikko Jokinen c, Sauli Laaksonen d Sirpa Rasmus a,e,, Henri Wallen a,f, Minna Turunen a, Mia Landauer a,g, Juho Tahk Mikko Jokinen c, Sauli Laaksonen d a Arctic Centre, University of Lapland, P.O. Box 122, FI-96101, Rovaniemi, Finland b Reindeer Herders’ Association, P.O. Box 8168, FI-96101, Rovaniemi, Finland c Natural Resources Institute, Ounasjoentie 6, FI-96200, Rovaniemi, Finland d Department of Veterinary Biosciences, Faculty of Veterinary Medicine, University of Helsinki, P.O. Box 66, FI-00014, Helsinki, Finland e Department of Biological and Environmental Science, University of Jyv¨askyl¨a, University of Jyv¨askyl¨a, P.O.Box 35, FI-40014, Finland f Department of Archaeology, Faculty of Humanities, University of Oulu, P.O Box 8000, FI-90014, Oulu, Finland g Risk and Resilience Program, International Institute for Applied Systems Analysis (IIASA), Schlossplatz 1, A-2361, Austria 1. Introduction (Eide et al., 2017; Soppela & Turunen, 2017). Reindeer husbandry is considered regionally important as it employs people, keeps remote areas inhabited, and provides economic benefits. It also represents cul tural continuity and a way of life connected to traditions, indigenous rights, and trans-generational values (Helle & Jaakkola, 2008; Kumpula & Siitari, 2020; Turunen & Vuojala-Magga, 2013, 2014). A R T I C L E I N F O Keywords: Climate change Cumulative effects Land use Northern fennoscandia Practitioner knowledge Reindeer husbandry Drivers of change in the reindeer management system are rather well-known. But when developing the gover nance to support the traditional livelihoods, it is crucial to understand also practitioner perceptions. Systematic research on these is lacking. We analyzed the land-use and climate related drivers within the reindeer man agement area (RMA) in Finland, and, using a perception geography approach, studied the herder perceptions towards these. We conducted an on-site questionnaire survey with herders from 51 herding districts. Factors directly affecting the welfare of reindeer were perceived as crucial by herders, for example basal icing affecting the forage availability, and land-use related factors limiting the seasonal pasture access. Perceptions of herders on biophysical factors were rather homogeneous. The regional heterogeneities in perceptions towards land-use related factors could be explained by spatial differences in land-use and varying herding traditions. Cumulative land-use impacts raised particular concerns. Our approach can be utilized in the co-planning of the northern land-use and more widely in the co-management of natural resources. Land-use and climate related drivers of change in the reindeer management system in Finland R Si W ll H i T Mi L d Mi T hk l J h J ki DOI: 10.1016/j.apgeog.2021.102501 Published: 01.09.2021 Published: 01.09.2021 Document Version Publisher's PDF, also known as Version of record Citation for pulished version (APA): Rasmus, S., Wallen, H., Turunen, M., Landauer, M., Tahkola, J., Jokinen, M., & Laaksonen, S. (2021). Land-use and climate related drivers of change in the reindeer management system in Finland: Geography of perceptions. Applied Geography, 134, Article 102501. https://doi.org/10.1016/j.apgeog.2021.102501 Document License CC BY Applied Geography 134 (2021) 102501 Available online 8 July 2021 0143-6228/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1016/j.apgeog.2021.102501 Received 10 August 2020; Received in revised form 23 June 2021; Accepted 23 June 2021 * Corresponding author. Arctic Centre, University of Lapland, P.O. Box 122, FI-96101, Rovaniemi, Finland. E-mail address: sirpa.rasmus@ulapland.fi (S. Rasmus). 1 When referring to reindeer as a livelihood, we use the term “reindeer husbandry”. “Herding” refers to day-to-day practices (and it also appears in the term “reindeer herding district”), whereas husbandry considers reindeer as a resource and is related for example to the profits, breeding, and social mechanisms. “Reindeer management” is related to all of the practices pertaining to the keeping of reindeer, including governance (Forbes, 2006). Available online 8 July 2021 0143-6228/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Received 10 August 2020; Received in revised form 23 June 2021; Accepted 23 June 2021 y 0143-6228/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). * Corresponding author. Arctic Centre, University of Lapland, P.O. Box 122, FI-96101, Rovaniemi, Finland. E-mail address: sirpa.rasmus@ulapland.fi (S. Rasmus). 1.1. Reindeer management in Finland 1.1. Reindeer management in Finland Reindeer husbandry1 is one of the traditional livelihoods in northern Finland (Itkonen, 1948; Kortesalmi, 2007). The Reindeer management area (RMA) covers 36% of Finland’s total area. In this area, semi-domesticated reindeer (Rangifer tarandus tarandus) have a free grazing right which is not dependent on land ownership (Reindeer Herding Act 1990/848, Heikkinen et al., 2012). There is a vast cultural, Reindeer husbandry1 is one of the traditional livelihoods in northern Finland (Itkonen, 1948; Kortesalmi, 2007). The Reindeer management area (RMA) covers 36% of Finland’s total area. In this area, semi-domesticated reindeer (Rangifer tarandus tarandus) have a free grazing right which is not dependent on land ownership (Reindeer Herding Act 1990/848, Heikkinen et al., 2012). There is a vast cultural, historical and geographical diversity within the RMA. Both Indigenous Saami and Finnish people practice reindeer husbandry in Finland, unlike in Sweden and Norway where it is mainly an exclusive right of the Saami In a globalizing north, reindeer husbandry shares the same opera tional space with several other land-use forms. Economic development, such as industrial infrastructure projects and tourism activities, is growing in the Arctic (AMAP, 2017; Finger & Heininen, 2019; Forbes, 2006). In Fennoscandia, reindeer husbandry has conflicting interests with other land-use forms, mainly forestry (timber harvest), agriculture, exploration and extraction of natural resources, outdoor recreation and historical and geographical diversity within the RMA. Both Indigenous Saami and Finnish people practice reindeer husbandry in Finland, unlike in Sweden and Norway where it is mainly an exclusive right of the Saami i 1 When referring to reindeer as a livelihood, we use the term “reindeer husbandry”. “Herding” refers to day-to-day practices (and it also appears in the term “reindeer herding district”), whereas husbandry considers reindeer as a resource and is related for example to the profits, breeding, and social mechanisms. “Reindeer management” is related to all of the practices pertaining to the keeping of reindeer, including governance (Forbes, 2006). Applied Geography 134 (2021) 102501 S. Rasmus et al. area subsidy policy since reindeer husbandry in Finland belongs to the EU’s unified food and agriculture sector (Rees et al., 2008; Turunen & Vuojala-Magga, 2014). Other land use in the RMA is governed first and foremost by the National Land Use Guidelines (2017) which steer regional and municipal zoning in Finland. 1.1. Reindeer management in Finland Herders are consulted during planning of land-use projects affecting their livelihood (for example during the Environmental Impact Assessment procedures), but they do not necessarily have enough power to affect decision making (Landauer & Komendantova, 2018). In the Saami homeland, legislation on the rights of the Saami as an Indigenous people plays a central role in land-use planning (Markkula et al., 2019). i tourism, and energy production (Pape & L¨offler, 2012; Pettersson et al., 2017; Sarkki et al., 2018). Intensifying land use has led to fragmentation and deterioration of reindeer pastures, causing the available pasture area to shrink (Anttonen et al., 2011; Jaakkola et al., 2013; Kumpula et al., 2014). The warming climate poses additional challenges, among them increased risk of ice formation on the winter pastures (Rasmus et al., 2018; 2020a) and lengthening of the heat periods and increased insect harassment during the summer (Soppela & Turunen, 2017; Tur unen et al., 2016). , Reindeer husbandry is dependent on management decisions of external public authorities (Heikkil¨a, 2006). It is steered by the Ministry of Agriculture and Forestry of Finland. Furthermore, the European Union steers the livelihood by means of national livestock and arable Economic profitability of reindeer husbandry depends on several factors such as legislation, markets and historical legacies (Bernes et al., Table 1 The drivers of change in the reindeer management system in Finland and examples of their impacts on reindeer husbandry, based on national reviews and reports (1919–2020). The shading marks the drivers considered in our study (Grotenfelt, 1919; J¨a¨ask¨o, 2001; Pakkanen & Valkonen, 2012; Sonnenfeld, 1972; Working committee, 1992). Table 1 The drivers of change in the reindeer management system in Finland and examples of their impacts on reindeer husbandry, based on national reviews and reports (1919–2020). The shading marks the drivers considered in our study (Grotenfelt, 1919; J¨a¨ask¨o, 2001; Pakkanen & Valkonen, 2012; Sonnenfeld, 1972; Working committee, 1992). S. Rasmus et al. Applied Geography 134 (2021) 102501 2015). These are continuously evolving and diversifying. Still, the basis for successful reindeer husbandry is the productive herd of healthy an imals (Kynk¨a¨anniemi, 2020). The welfare of reindeer is largely deter mined by the quantity and quality of available forage (Helander-Renvall, 2014; Kitti et al., 2006). Forage adequacy ensures productivity of dams, which, in turn, affects the calving success and calf weight, both being pivotal to the profitability of the livelihood (Muut toranta et al., 2014). In addition to the area and state of pastures, limited access to pasture resources reduces forage availability: “It is more often a question of access to the forage than the amount of forage as such” (Kitti et al., 2006, p. 149). Factors affecting the accessibility of forage are mostly biophysical (Forbes, 2006; Rasmus et al., 2018); some are related to the management choices or governance of land use (e.g., Eira et al., 2018; Riseth et al., 2016). conditions, various snow types, and predator behavior (Porsanger & Guttorm, 2011; Turunen & Vuojala-Magga, 2014). Perception geography approach has been applied for example in studies on risk management (Ren et al., 2016) and urban planning (Bergeron et al., 2014). Using local perceptions together with the quantitative data from monitoring surveys has been as a tool to get to more holistic understanding of the changes in ecosystems and the nature-based livelihoods such as reindeer husbandry (Rasmus, Turunen, Luomaranta, et al., 2020), fishing (Coll et al., 2014) or forest use (Meijaard et al., 2013). Geographical distribution of perceptions has been studied using map-based methods, which enable the collection and presentation of information about local people’s views of their environment and important places, and can make their incorporation to decision-making easier (Nikula et al., 2020; Ren et al., 2016). Public Participation Geographic Information Systems (PPGIS) can be used to locate percep tions on a map and are designed to involve wide audience. Table 1 During recent years these have been applied in land use studies both in urban and rural areas (Brown & Kytt¨a, 2014; Kahila-Tani et al., 2016; Kantola et al., 2018; Nikula et al., 2020). Understanding people’s perceptions towards for example certain forms of land use may facilitate the management of conflicts between land users with different interests in the same resource (Brown et al., 2020). Understanding the spatial differences both in the drivers of change and in the perceptions towards them will help develop jurisdictional and institutional strategies to support governance and future of the northern livelihoods such as reindeer husbandry. Reindeer management represents an example of a social-ecological system (SES) (cf. Ostrom, 2009; McGinnis & Ostrom, 2014) in which biophysical and socio-economic factors are interacting as drivers of change (K¨ayhk¨o & Horstkotte, 2017). For reindeer management in Finland, these drivers and related impacts on the livelihood are rather well-known (Table 1). Already about a hundred years ago deterioration of pastures and conflicts between reindeer husbandry and other land-use forms were listed as factors causing pressure on reindeer husbandry (Table 1; Kortesalmi, 2007). A recent national report (Kumpula & Siitari, 2020) presented recommendations on how to support and develop “ecologically, economically, culturally and socially sustainable use of reindeer pastures, as well as the vitality of reindeer husbandry”. The fact that reindeer husbandry faces varying challenges in different areas within the RMA in Finland was strongly emphasized in the report. Still, local and regional differences are currently not sufficiently considered in the governance of the livelihood. Systematic research on herders’ per ceptions on factors affecting their livelihood is lacking. In addition, more knowledge is needed of the cumulative effects of various factors – ones resulting from the combined effects of multiple activities affecting reindeer management over space or time. Our work, for its part, ad dresses this gap. This work is based on a questionnaire survey conducted during a project “Reindeer health in the changing environment” (2016–2018, Laaksonen, 2016). The survey focused on herders’ perceptions on factors affecting reindeer welfare. This means factors related to drivers of change such as climate and weather, pasture conditions and land use (shaded in Table 1). Consequently, this paper also focuses on these concrete factors herders experience in their daily herding work; factors affecting the success of the livelihood through reindeer condition and available forage. Table 1 In reindeer management, some drivers, such as governance institutions, are of socio-economic or cultural nature (Table 1). In this study we do not focus on these drivers. Our research questions are: 1.2. Perception geography approach We use a perception geography approach (Downs, 1970, Sonnenfeld, 1972), basing our analysis on herder perceptions on and observations of their operational environment. The concept ‘Perception geography’ or ‘Perceptual geography’ emerged in the 1960s from the wider conceptual framework of behavioral geography (see Bunting & Guelke 1979; Tuan, 2003) and it provides tools and methods to discuss and quantify land scape research and is applicable also in natural resources management. 1) Which factors of the climatic, ecological and land-use related drivers of change are perceived by herders as important for reindeer welfare and why? 2) What kinds of regional differences are there in these perceptions within the RMA? 3) Can the regional differences be explained by regional variation in land-use patterns, climatic and ecological conditions, or varying herding traditions and practices? 3) Can the regional differences be explained by regional variation in land-use patterns, climatic and ecological conditions, or varying herding traditions and practices? Perceptual geography is characterized by a common idea that experience affects perception, which leads to the conclusion that per ceptions vary because individuals’ life experiences differ (Tuan, 2003). Perceptions are understood as points of contact between people and their environment and as a basis for spatial reasoning and decision making. Perception is the process that encodes the objective environ ment as a subjective one (Golledge & Stimson, 1997), with the subjective environment and past experiences influencing our behaviour and ac tions (Sonnenfeld, 1972; Guelke, 2003). 2.1. The study area The study area covers the RMA in Finland, situated between 64.5◦N and 70.1◦N (Fig. 1a, Fig. S1). The sparsely populated region is charac terized by boreal coniferous forests, mires, subarctic mountain birch woodlands and fells. For the time period from 2000 to 2019, the maximum number of reindeer over one year old allowed within the region has been set at 203,700. The numbers are set by the Ministry of Agriculture and Forestry every ten years. This is the size of the winter stock; furthermore, ca. 100,000 calves are born in spring (RHA 2018). Of this total number, ca. 80,000 calves and 20,000 over one-year-old reindeer are slaughtered in autumn. Finland has 54 herding districts. A reindeer herder can practice herding within one herding district at a time, and every district is responsible for reindeer herding within its area. The districts are regulated by the Reindeer Husbandry Act (1990/848). During the study period, there were approximately 4500 Herder perceptions are relevant when studying the reindeer man agement SES. Herders are central actors in the system: they experience and deal with drivers of change such as land use and climate change in their daily professional life (Landauer et al., 2021). Their perceptions affect the functioning of the system for example through decision making and risk preparedness. Perceptions also carry culture, and local and traditional knowledge (Forbes et al., 2020; Jaakkola et al., 2018). This means knowledge and practices, developed during centuries and handed down from generation to generation (Berkes 2008). Knowledge also accumates and develops through context-situated learning in new situations. Knowledge can consider for example reindeer behavior, forage and diseases, suitable herding practices during adverse weather 3 Applied Geography 134 (2021) 102501 S. Rasmus et al. State and privately owned land used for forestry and the present and planned land-use projects in the reindeer management area (RMA; region ensity in the northern area and close to the eastern border, or agricultural regions in the southern area not shown). b) The change in the annual ) in the number of warm weeks in the winter season in 1981–2010; locations with significant trends (at 5% significance level) are marked with b ta: Finnish Environment Institute (2020) and Finnish Meteorological Institute. Fig. 1. Table 2 Table 2 Factors related to climate and weather, usability of pasture resources, and other land use and disturbances, considered in the survey. Driver of change Factor considered Factor explained Climate and weather1,2,3,4 Timing of spring Cold and snow are harmful for newborn calves, and lactating reindeer benefit from natural fresh forage. Summer temperature Multiple impacts on reindeer welfare: affecting forage and water supply and insect harassment, causing heat stress. Cold spells in winter Long periods of very low temperatures deplete energy storages of reindeer Deep snow Makes reindeer difficult to access ground lichens; risk of predator attacks increase; herding work gets difficult. Icing events: -icy snow structure -basal ice on pastures Makes reindeer difficult to access ground lichens; decreased usability of pastures. Insect harassment in summer11 Increases stress, energy consumption and vector borne diseases of reindeer. Wild mushroom yield in autumn Important for reindeer to gain fat and energy storage for winter. Usability of pasture resources5,6,7 Summer pasture condition and availability Diverse and peaceful summer pastures needed for reindeer to recover from winter and improve the body condition, and for calves to grow. Winter pasture condition and availability Lichen pastures with adequate quantity and quality needed for winter survival, welfare, calving success and calf weight Fragmentation of pastures Fragmented pastures are difficult to utilize, managing the herds and controlling the grazing becomes difficult for herders. Availability of nature conservation areas Less human disturbance and other land use in conservation areas; diverse and peaceful seasonal pastures and increased winter forage availability because of old-growth forests. Other land use1,7,8,9,10 Forestry Forestry decrease the amount of old-growth forests important for reindeer as winter pastures; increases fragmentation of pastures; harvesting of dense forests can improve lichen growth. Mining Mining districts occupy pasture land and increase fragmentation of pastures; operations and transportation increase the risk of accidents, noise and dust impacts. Peat extraction Peat extraction destroys summer pastures and calving areas, reindeer may drown in deep dikes; areas can be utilized by reindeer to avoid insect harassment. Hunting/dogs Free-running dogs can kill or injure reindeer or cause extra work for herders by scattering the herd. The impacts are most common during rut, and can affect calf production. Disturbances by human activities Reduced forage availability through decreased access or usability of pastures; may disturb calving and grazing. 2.1. The study area a) State and privately owned land used for forestry and the present and planned land-use projects in the reindeer management area (RMA; regions with high predator density in the northern area and close to the eastern border, or agricultural regions in the southern area not shown). b) The change in the annual degree day sum and c) in the number of warm weeks in the winter season in 1981–2010; locations with significant trends (at 5% significance level) are marked with black check marks. Data: Finnish Environment Institute (2020) and Finnish Meteorological Institute. Fig. 1. a) State and privately owned land used for forestry and the present and planned land-use projects in the reindeer management area (RMA; regions with high predator density in the northern area and close to the eastern border, or agricultural regions in the southern area not shown). b) The change in the annual degree day sum and c) in the number of warm weeks in the winter season in 1981–2010; locations with significant trends (at 5% significance level) are marked with black check marks. Data: Finnish Environment Institute (2020) and Finnish Meteorological Institute. Applied Geography 134 (2021) 102501 S. Rasmus et al. 2.2. Data collection method According to regulation on land-use, the land in this area should not be used in a way detrimental to herding. The 13 northernmost districts belong to the Saami Homeland area (SHA) in Finland (Reindeer Husbandry Act 848/1990, N¨akk¨al¨aj¨arvi & Jaakkola, 2017). In the ASR, and especially in the SHA, the herds are generally larger and reindeer husbandry is more commonly the main source of livelihood, whereas in the southern districts, reindeer husbandry is traditionally more often combined with other livelihoods, particularly small-scale agriculture and forestry and reindeer-based tourism (Jaak kola et al., 2018; Soppela & Turunen, 2017). Supplementary winter feeding in enclosures is a more common practice in central and southern parts of the RMA. In the north, especially in the SHA, the livelihood is more based on herding the reindeer on natural pastures (Helle & Jaak kola, 2008; Turunen & Vuojala-Magga, 2014). For these reasons, we present some of the results separately for the northern area (ASR including the SHA) and the southern area (the rest of the RMA). The survey respondents were asked to estimate their perceptions towards 32 factors on a scale from −4 to 4, whereby −4 denotes a considerable negative effect, 0 denotes no effect and 4 denotes a considerable positive effect. They were asked to consider the latest ten- year period in their answers. Our study covers 94% of herding districts within the RMA (51 out of total of 54). The survey was conducted mainly on-site in the reindeer herding districts the herders represent by one of the co-authors (J.T.) between May 2016 and January 2017. In three cases herders were interviewed by phone. All respondents were male and full-time herders. Chiefs of districts (porois¨ant¨a) were selected to represent each district as key informants because they have the most holistic knowledge of the district they represent (purposive sampling; Bernard, 1995). i Topography, vegetation and climate vary within the RMA as well. Clear warming trends have been observed during the past decades (Fig. 1b and c). In reindeer management, impacts of changing climate are experienced through changing seasonal weather conditions. These impacts are already being observed across the area but in varying degree (Rasmus, Turunen, Luomaranta, et al., 2020). The key findings are presented in the form of maps, while all answers The key findings are presented in the form of maps, while all answers 2.2. Data collection method reindeer owners in Finland of which an estimated 900 are full-time herders (RHA 2018). Herding cultures, seasonal herding practices, reindeer numbers per district, as well as the intensity and type of other land use vary greatly within the RMA (Table S1, Fig. 1a). The 20 northernmost districts belong to the area specially intended for reindeer husbandry (ASR; “northern area”). According to regulation on land-use, the land in this area should not be used in a way detrimental to herding. The 13 northernmost districts belong to the Saami Homeland area (SHA) in Finland (Reindeer Husbandry Act 848/1990, N¨akk¨al¨aj¨arvi & Jaakkola, 2017). In the ASR, and especially in the SHA, the herds are generally larger and reindeer husbandry is more commonly the main source of livelihood, whereas in the southern districts, reindeer husbandry is traditionally more often combined with other livelihoods, particularly small-scale agriculture and forestry and reindeer-based tourism (Jaak kola et al., 2018; Soppela & Turunen, 2017). Supplementary winter feeding in enclosures is a more common practice in central and southern parts of the RMA. In the north, especially in the SHA, the livelihood is more based on herding the reindeer on natural pastures (Helle & Jaak kola, 2008; Turunen & Vuojala-Magga, 2014). For these reasons, we present some of the results separately for the northern area (ASR including the SHA) and the southern area (the rest of the RMA). Data for our analysis were gathered using an on-site questionnaire survey which was part of semi-structured interview conducted with 51 chiefs of herding districts in the RMA. In the survey factors affecting reindeer welfare were considered. The factors were selected by re searchers based on existing knowledge of external drivers of change affecting reindeer (Tables 1 and 2 and S1). Herding practices as internal drivers of change were discussed as well (Table 3). More information about the factors considered and their impacts on reindeer welfare and reindeer husbandry can be found in references in the tables. Reindeer management SES and the relationships between the key factors are thoroughly explained and graphically presented e.g. in K¨ayhk¨o & Horstkotte (2017) and Landauer et al. (2021). Herding cultures, seasonal herding practices, reindeer numbers per district, as well as the intensity and type of other land use vary greatly within the RMA (Table S1, Fig. 1a). The 20 northernmost districts belong to the area specially intended for reindeer husbandry (ASR; “northern area”). Table 2 Other land use (wind farms, hydropower, infrastructure like roads etc.) Land use occupies pasture land and increases fragmentation of pastures; reduced forage availability through decreased access or usability of pastures; grazing pressure on remaining pastures increases. Predation Wolverine (Gulo gulo), wolf (Canis lupus), brown bear (Ursus arctos), lynx (Lynx lynx) and golden eagle (Aquila chrysaetos) hunt and kill reindeer, which can cause severe damage to herds and extra work for herders though mitigating the damage and looking for carcasses (needed to get compensations). Decreases the usability of certain pastures. 1Pape & L¨offler 2012, 2Turunen et al.,2016, 3Jaakkola et al.,2018, 4Rasmus et al.,2020a, 5Kitti et al.,2006, 6Kivinen et al.,2012, 7Kumpula et al.,2014, 8Anttonen et al. Cold and snow are harmful for newborn calves, and lactating reindeer benefit from natural fresh forage. Multiple impacts on reindeer welfare: affecting forage and water supply and insect harassment, causing heat stress. Makes reindeer difficult to access ground lichens; risk of predator attacks increase; herding work gets difficult. i Makes reindeer difficult to access ground lichens; risk of predator attacks increase; herding work gets difficult. i i Makes reindeer difficult to access ground lichens; decreased usability of pastures. i Makes reindeer difficult to access ground lichens; decreased usability of pastures. Increases stress, energy consumption and vector borne diseases of reindeer. I f i d i f d f i Diverse and peaceful summer pastures needed for reindeer to recover from winter and improve the body condition, and for calves to grow. Diverse and peaceful summer pastures needed for reindeer to recover from winter and improve the body condition, and for calves to grow. Lichen pastures with adequate quantity and quality needed for winter survival, welfare, calving success and calf weight i Fragmented pastures are difficult to utilize, managing the herds and controlling the grazing becomes difficult for herders. i Less human disturbance and other land use in conservation areas; diverse and peaceful seasonal pastures and increased winter forage availability because of old-growth forests. i Less human disturbance and other land use in conservation areas; diverse and peaceful seasonal pastures and increased winter forage availability because of old-growth forests. Forestry decrease the amount of old-growth forests important for reindeer as winter pastures; increases fragmentation of pastures; harvesting of dense forests can improve lichen growth. Mining districts occupy pasture land and increase fragmentation of pastures; operations and transportation increase the risk of accidents, noise and dust impacts. 11not a climatic factor, but related to for example seasonal temperature, precipitation and wind conditions. Table 3 Due to poor winter pasture resources or limited access to forage, for example, due to icy snow, reindeer are fed with supplementary feeds in their natural pastures (often in forests); helps also controlling the herds and protecting reindeer from predators. Feeding in the pasture area to support active herding. On-the-spot management of the movement and foraging of herds; moving herds with or without the aid of hay from one pasture area to another, shepherding. Practice where certain pastures are reserved for certain seasons and natural seasonal behaviour of reindeer is supported by fences separating the pastures, and by active herding. re annually treated with antiparasitic medication, to improve the condition. In enclosure calving, reindeer give birth within a fenced pasture area. The calves are ear-marked immediately after their birth with the owner’s earmark i In free-ranging type of calving reindeer give birth in their natural calving regions and specific spots such as forested areas or the southern slopes of fells. Reindeer are gathered from pastures to summer round-ups, in which the calves are ear-marked with the owner’s reindeer earmark from midsummer onwards. Timing of slaughtering3,4,5 Impacts on amount of meat to sell and meat quality. The later the slaughter, the poorer the condition of reindeer generally is; they start loosing weight after the snow cover forms. Impacts on amount of meat to sell and meat quality. The later the slaughter, the poorer the condition of reindeer generally is; they start loosing weight after the snow cover forms. Pekkarinen et al., 2015, 2Horstkotte et al., 2020, 3Forbes 2006, 4Helle & Jaakkola 2008, 5K¨ayhk¨o & Horstkotte, 2017, 6Laaksonen et al., 2017. brmss (Bürkner, 2017) was used for fitting the GLMs and bayesplot (Gabry & Mahr, 2021) for creating visualizations from the models. i of the on-site questionnaire survey are shown in Tables S2-S5 – in the Supplementary material. The survey respondents had the possibility to comment on their answers (15 comments received), provide additional information on calving success of reindeer (37 answers received) and give suggestions on how to increase the welfare of reindeer (50 sug gestions received). Some excerpts of this free-form material are pre sented as part of the results to illustrate the themes of this article. Table 3 Table 3 Factors related to herding practices in use (internal drivers), considered in the survey. Driver of change Factor considered Factor explained Herding practices Supplementary winter feeding in enclosures1,2 Due to poor winter pasture resources or limited access to forage, for example, due to icy snow, reindeer are taken into enclosures and given supplementary feeds daily for several months; also protects reindeer from predators and keeps them off the roads and settlements. Supplementary winter feeding in the field1,2 Due to poor winter pasture resources or limited access to forage, for example, due to icy snow, reindeer are fed with supplementary feeds in their natural pastures (often in forests); helps also controlling the herds and protecting reindeer from predators. “Herding feeding”1,2 Feeding in the pasture area to support active herding. Active herding3,4,5 On-the-spot management of the movement and foraging of herds; moving herds with or without the aid of hay from one pasture area to another, shepherding. Pasture rotation3,4,5 Practice where certain pastures are reserved for certain seasons and natural seasonal behaviour of reindeer is supported by fences separating the pastures, and by active herding. Antiparasitic medication6 Reindeer are annually treated with antiparasitic medication, to improve the condition. Managing calving in enclosures3,4,5 In enclosure calving, reindeer give birth within a fenced pasture area. The calves are ear-marked immediately after their birth with the owner’s earmark Earmarking of calves during summer3,4,5 In free-ranging type of calving reindeer give birth in their natural calving regions and specific spots such as forested areas or the southern slopes of fells. Reindeer are gathered from pastures to summer round-ups, in which the calves are ear-marked with the owner’s reindeer earmark from midsummer onwards. Timing of slaughtering3,4,5 Impacts on amount of meat to sell and meat quality. The later the slaughter, the poorer the condition of reindeer generally is; they start loosing weight after the snow cover forms. 1Pekkarinen et al., 2015, 2Horstkotte et al., 2020, 3Forbes 2006, 4Helle & Jaakkola 2008, 5K¨ayhk¨o & Horstkotte, 2017, 6Laaksonen et al., 2017. Due to poor winter pasture resources or limited access to forage, for example, due to icy snow, reindeer are taken into enclosures and given supplementary feeds daily for several months; also protects reindeer from predators and keeps them off the roads and settlements. Table 3 To show the significance of cumulative effects of different land-use activities we calculated how many of the following factors were considered as harmful per district: forestry, mining, peat extraction, hunting/dogs, other disturbances by human activities (such as outdoor recreation), other land use (such as wind farms, hydropower, infra structure; see Table 2). We interpreted answers −3 or −4 as a harmful effect and 3 or 4 as a beneficial effect. The dataset was rescaled from range [-4,4] to [-1,1] for the regression models. Essentially all negative values were considered as negative were considered as negative effect (−1), zero values as neutral (0) and positive values as positive (1) effect. Perceptions with only negative and neutral (icing, predation, human disturbance, mining, hunting) or neutral and positive (herding, conser vation areas, mushroom yield) values were omitted from the analysis. Detailed background information on the herding practices of each district was also collected. Due to the quality and, in some cases, the confidentiality of the material, they are not shared in full as part of the original data set for this study, but referred to under the Results section to provide further insight into the commonness of and regional differ ences in certain herding practices. Our study also draws on articles published in the professional journal Poromies [Reindeer herder] in order to describe the changes herders have experienced in the pasture environment and the subsequent adoption of the supplementary winter feeding (Supplementary text S1). The journal has been published since 1931 by the Reindeer Herders’ Association in Finland, and it is a commonly used source of information in research on reindeer management (Kortesalmi, 2007; Helle & Jaak kola, 2008, Vuojala-Magga et al., 2011; Turunen et al., 2017). We used absence/presence data for analyzing the perceptions regarding peat production and mining. In these models the A/P variable was used as a predictor for the corresponding perception. Perceptions on forestry were analyzed using log-transformed ratio of private and state owned forests within each reindeer herding district as a predictor. Regression models were fitted using tight priors as constraints as sug gested by Gelman et al. (2020). Details are found in Table S9 in the Supplementary material. Table 2 eat extraction destroys summer pastures and calving areas, reindeer may drown in deep dikes; areas can e utilized by reindeer to avoid insect harassment. Peat extraction destroys summer pastures and calving areas, reindeer may drown in deep dikes; areas can be utilized by reindeer to avoid insect harassment. Free-running dogs can kill or injure reindeer or cause extra work for herders by scattering the herd. The impacts are most common during rut, and can affect calf production. Reduced forage availability through decreased access or usability of pastures; may disturb calving and grazing. Land use occupies pasture land and increases fragmentation of pastures; reduced forage availability through decreased access or usability of pastures; grazing pressure on remaining pastures increases. Wolverine (Gulo gulo), wolf (Canis lupus), brown bear (Ursus arctos), lynx (Lynx lynx) and golden eagle (Aquila chrysaetos) hunt and kill reindeer, which can cause severe damage to herds and extra work for herders though mitigating the damage and looking for carcasses (needed to get compensations). Decreases the usability of certain pastures. 1Pape & L¨offler 2012, 2Turunen et al.,2016, 3Jaakkola et al.,2018, 4Rasmus et al.,2020a, 5Kitti et al.,2006, 6Kivinen et a 2011, 9Rasmus et al.,2020b, 10Landauer et al.,2021 11not a climatic factor, but related to for example seasonal temperature, precipitation and wind conditions. 5 Applied Geography 134 (2021) 102501 S. Rasmus et al. S. Rasmus et al. 2.3. Analyses We divided the observations into two groups according to borders presented in Fig. 1a. The first group (n = 17) consists of herding districts north of the ASR -border, “northern area”, including the Saami Home land area (SHA). For the second group (n = 34) we combined the dis tricts within the rest of the RMA (“southern area”). Differences in herding cultures and practices between these groups, as well as the in tensity of other land use and the land-use regulations, justify this divi sion. Considering the SHA separately was not considered possible because of small number of survey respondents from that area (n = 10). We analyzed the group differences on perceptions using Bayesian ordinal regression framework as presented by Bürkner and Vuorre (2018). Cumulative models with probit -link were fitted using each perception as response variable and group as a predictor. The perceptions of the herders were analyzed by using exploratory data analysis methods and the differences of perceptions between the groups were analyzed using Bayesian generalized linear regression models (GLM). The survey data were combined with existing spatial data- sets on land use of the study region (RHA 2018; Mets¨ahallitus, 2019, Finnish Environmental Institute, 2020) to explain and discuss the results. Regional comparisons of the data were conducted by mapping the data into choropleth maps and compared visually. Data preparation, visual isations and analyses were done in R (R Core Team, 2020). The Tidyverse package (Wickham et al., 2019) was used for data preparation and visu alisations, the tmap package (Tennekes, 2018) for plotting the maps and the sf package (Pebesma, 2018) for spatial data operations. R-package S. Rasmus et al. S. Rasmus et al. Applied Geography 134 (2021) 102501 3. Results Depending on the factor studied, the distribution of answers shows agreement, disagreement and even polarization of perceptions. The general agreement on certain biophysical factors is clear (Fig. 3a and b). Perceptions on deep snow cover (Fig. 3c) give a good example of disagreement among the survey respondents. Generally, deep snow makes reindeer foraging more difficult, but in some districts with ample forest pastures it may ease grazing on arboreal lichen. Moreover, snow depth is not a significant factor in districts where most of the reindeer are fed in enclosures. 2.3. Analyses Also, some land-use related factors such as forestry were considered as neutral or even beneficial by some respondents, although forestry was generally seen as harmful for reindeer husbandry (Fig. 3d). All respondents perceived hunting negatively, although the importance of this factor varied (Fig. 3e). Three herders from the southern area explain the effects on reindeer: “Hunting dogs disturb rutting [reindeer], which has an impact on the calf percentage … Sometimes the [presence of]hunting dogs delays the rut so that calves are lighter when slaughtered … We should put an end to the barking of the elkhounds early in the autumn because it breaks up the herds, and the calves may become separated from the dams.” 3.2. Regional heterogeneity in perceptions Perceptions on the importance of factors related to climate and weather were rather homogeneous within the whole RMA (Table S2). Cold and rainy summers were mainly perceived as a harmful factor (Fig. 4a), but in some southern districts as a beneficial one. Few districts from the central region viewed also hot summers positively, although this factor is generally considered harmful to reindeer (Rasmus, Tur unen, Luomaranta, et al., 2020). Survey respondents explain how climate-related factors directly affect the welfare of the animals and, thus, the future calving success: “After a hot summer, calf production was very weak. The dams were not in heat” (Southern area); “Warm autumns led to a weak rutting period, the stags got lazy … Rutting is delayed in warm autumns. The dams need frosts to be in heat” (two herders from the southern area); “A cold winter may also cause dams to abort their calf” (Northern area). Despite the fact that forestry is practiced in most of the reindeer herding districts (Fig. 1a) and it is considered as a disturbing factor to reindeer husbandry by other studies (Table S1), perceptions of our survey respondents on forestry were diverse (Fig. 5c) and, overall, less negative than expected. Interestingly, perceptions on forestry were more often negative in regions where forests are mostly state owned (Fig. 5f), compared to regions where forest ownership is mostly private. When reading the results of a 30-year-old comparable survey together with ours (Supplementary text S1), the intensification and diversification of land use within the RMA during the past decades is clearly visible. Approximately 30 years ago, forestry stood out very clearly as the most harmful form of land use affecting reindeer hus bandry. Only a few other factors were mentioned by the respondents then, tourism and agriculture being the most common ones. Several other land-use related factors – hunting, human disturbance, peat extraction, other land use and mining – are nowadays seen as equally or more harmful than forestry in the districts in which forestry was considered as the most harmful factor in the 1980s. (Supplementary text S1, Table 4). Perceptions on factors related to pasture resources were heteroge neous (Table S4). Summer pastures were nearly unanimously viewed as beneficial for reindeer welfare. Less agreement was seen in the percep tions on winter pastures (Fig. 4b). 3.1. Importance of the drivers There was a strong agreement among the survey respondents that certain biophysical factors are harmful. For example, 75% of the herders perceived the impact of predators and 80% considered icy foraging conditions as harmful (Fig. 2a; See also Tables S2–S5). The respondents also agreed on benefits of certain biophysical factors such as abundance of mushroom (92% considered beneficial; Fig. 2b) and early arrival of spring (69% considered beneficial). As one herder from the northern area put it: “Dams’ success for becoming pregnant depends on mushroom yield”. Disturbance-related factors such as hunting were generally perceived as harmful. Perceptions on certain herding practices diverged the most. Feeding in enclosures was seen as an important, positive factor in the southern part of the RMA, but in the north it was mostly seen as a method which should not be preferred. Instead, herders in the northern part of the RMA preferred practices such as active herding (moving herds, shepherding) and feeding to support this (“herding feeding”). 7 Fig. 2. Factors perceived as harmful (a) for the welfare of reindeer; the percentage of respondents sharing the view (answer −3 or −4; N = 51); Factors perceived as beneficial (b) for the welfare of reindeer, the percentage of respondents sharing the view (answer 3 or 4; N = 51). Fig. 2. Factors perceived as harmful (a) for the welfare of reindeer; the percentage of respondents sharing the view (answer −3 or −4; N = 51); Factors perceived as beneficial (b) for the welfare of reindeer, the percentage of respondents sharing the view (answer 3 or 4; N = 51). 7 Applied Geography 134 (2021) 102501 S. Rasmus et al. Applied Geography 134 (2021) 1025 ig. 3. Distribution of herder perceptions on selected factors: a) basal ice on the pastures, b) abundance of mushrooms, c) deep snow cover, d) forestry, e) huntin ogs, f) peat extraction areas, g) pasture rotation, h) active herding (−4 denotes a considerable negative effect, 0 no effect and 4 a considerable positive effect Rasmus et al. Fig. 3. Distribution of herder perceptions on selected factors: a) basal ice on the pastures, b) abundance of mushrooms, c) deep snow cover, d) forestry, e) hunting/ dogs, f) peat extraction areas, g) pasture rotation, h) active herding (−4 denotes a considerable negative effect, 0 no effect and 4 a considerable positive effect). S. Rasmus et al. 3.1. Importance of the drivers Applied Geography 134 (2021) 102501 areas were considered beneficial for reindeer husbandry, also by herders from districts lacking such areas. Perceptions on peat extraction (Fig. 3f) give an example of localized views. Most of the survey respondents considered peat extraction as a neutral factor, but some perceived it as very harmful. Perceptions on some herding practices were also localized, pasture rotation being an example (Fig. 3g). It was perceived mostly as a neutral factor, except by districts benefiting from it in their own work. Active herding was most often perceived either neutrally or very positively (Fig. 3h). The importance of natural pastures was seen in many of the free-form comments of the respondents, as well as the need for actions to foster the pasture quality: “Grazing peace and diverse pastureland are needed” (Northern area); “Old-growth forests should not be cut. In late winter [they are] very important for reindeer” (Northern area); “Pasture regeneration would be important” (Southern area). Land-use related factors were generally perceived either as neutral or negative (Table S3). For example, perceptions on peat extraction (Fig. 5a) and mining (Fig. 5b) were localized. These forms of land use were considered harmful especially in areas where such activities had existed in the past, currently existed or were under planning (Fig. 5d and e, see also Fig. 1a and Table S1). Several herders from the southern area wanted to see the peat extraction areas restored: “Reindeer husbandry should be considered in the re-use of peatlands” (Southern area). Fig. 4. a) Perceptions of herders on factors “Cold and rainy summer” and b) “Winter pastures” (−4 denotes a considerable negative effect, 0 no effect and 4 de notes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 4. a) Perceptions of herders on factors “Cold and rainy summer” and b) “Winter pastures” (−4 denotes a considerable negative effect, 0 no effect and 4 de notes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 3.2. Regional heterogeneity in perceptions We assume that the perceptions are linked to the particular situation of each district: quality and usability of seasonal pastures and pasture accessibility. Good summer pastures are most often available, but in some regions lichen pastures for winter foraging are scarce and the quality of the remaining winter pastures is low. Especially interesting is a region in the middle of the RMA, where winter pastures were seen as a factor affecting reindeer welfare nega tively, most probably due to the impact of forestry. Nature conservation 9 Fig. 4. a) Perceptions of herders on factors “Cold and rainy summer” and b) “Winter pastures” (−4 denotes a considerable negative effect, 0 no effect and 4 de notes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Applied Geography 134 (2021) 102501 S. Rasmus et al. Perceptions of herders on the factor “Peat extraction”, the black dots show the peat extraction areas; b) “Mining”, the black dots show the locations of stricts; c) “Forestry” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect), d-e) Marginal effects of absenc of peat extraction/mining districts on the perceptions of herders on factors “Peat extraction and “Mining” (posterior mean with 89% credible intervals effect of forestry ownership ratio (private or state) on the perceptions of herders on factor “Forestry” (posterior mean with 89% credible intervals). See r the land used for forestry. Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red l order of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web versio e.) Fig. 5. 3.4. Perceptions on herding practices Antiparasitic medication as well as early slaughtering were generally stated as beneficial factors by the survey respondents (Table S5). Of the districts studied, 92% aimed at early slaughter annually during the study period and 94% gave antiparasitic medication to the majority of their animals. Then again, some herders expressed somewhat critical views of medication: “Medication should be developed as the reindeer will become immune or new diseases will appear” (Southern area); “Medication should be given only to [the reindeer] in poor condition” (Southern area). According to our data, in 90% of the districts studied majority of reindeer got some supplementary feed during a typical winter during the study period. Several forms of feeding (feeding in the forest/pastures, feeding in the enclosures, “herding feeding” to support moving the herds) were used depending on the need and situation in an individual district. At least some feeding in the forest/pastures was practiced in 55% of the districts (the main form of feeding in four of the studied districts). At least some “herding feeding” was practiced in 29% the districts (the main form of feeding in seven districts, six of them belonged to the Saami Homeland area). At least some reindeer were fed in enclosures in 88% of the districts (the main form of feeding in 34 districts). Interestingly, perceptions on some herding practices varied a great deal either locally or regionally (Table S5). Pasture rotation is a practice where certain grazing lands are reserved for certain seasons. Natural seasonal behaviour of reindeer is supported by fences separating the pastures, and by active herding of animals from one area to another. Pasture rotation may not be possible if grazing lands are fragmented due to competing land-use forms and the related infrastructure (Anttonen et al., 2011), or if some seasonal pasture types are missing from the area of the district. Also, low lichen biomass on winter pastures may hinder the use of these (Kumpula et al., 2014). Overall, 39% of the total number of the districts studied – and all districts situated in the northern part of the RMA, including the Saami Homeland area – use pasture rotation. Most respondents considered pasture rotation as beneficial (Fig. 7a), although the importance of this factor varied. Similarly, as feeding reindeer in enclosures seems to divide the RMA in distinct regions (Fig. 9a), also perceptions on it were rather polarized (Fig. 9b and c). Table 4 Table 4 Percentage of districts where certain land-use related factors were seen as harmful as or more harmful than forestry in our study (number of districts = 34; only those districts analyzed, which had seen forestry as the single most negatively affecting activity within their district during 1986–1987, see Supplementary text S1 for details). Factor Factor as harmful as or more harmful than forestry (% of districts) Hunting/dogs 79 Disturbance (tourism, traffic, etc.) 59 Peat extraction 29 Other land-usea 29 Mining 18 a including wind farms, hydropower, and related infrastructure (power transmission lines, roads etc.). Table 4 Percentage of districts where certain land-use related factors were seen as harmful as or more harmful than forestry in our study (number of districts = 34; only those districts analyzed, which had seen forestry as the single most negatively affecting activity within their district during 1986–1987, see Supplementary text S1 for details). Factor Factor as harmful as or more harmful than forestry (% of districts) Hunting/dogs 79 Disturbance (tourism, traffic, etc.) 59 Peat extraction 29 Other land-usea 29 Mining 18 a including wind farms, hydropower, and related infrastructure (power transmission lines, roads etc.). 3.3. Cumulative effect of land use enclosures in 49% of the districts studied. Half or more of the calves were born in enclosures only in three districts; these were located in the northern area. Earmarking of calves during summer was common, with 82% of the studied districts practicing this. Herders may also mark calves earlier in the spring in the case of enclosure calving, or marking may be postponed until autumn if there are problems with collecting the animals in the summer or if there is a risk of heat stress during hot pe riods (Rasmus, Turunen, Luomaranta, et al., 2020). At least one land-use related factor was considered harmful by 86% of the survey respondents. Many districts listed several factors as harmful. When the sum of these views (number of these factors per district) was mapped, some hot-spot areas of land use could be seen (Fig. 6a). Characteristic of these hot-spot areas is their location within the forestry region of the RMA, and close to the southern border of the RMA and/or roads with heavy traffic. The pasture lands of districts host active or planned mines and wind farms, hydropower reservoirs and peat extraction areas (Fig. 1b, Table S1). The districts perceiving several land-use related factors as harmful often also considered fragmentation of pastures as a problem (Fig. 6b). Some overlap is also seen with the region where the winter pasture situation was perceived negatively (Fig. 4b). Particularly heterogeneous perceptions were related to supplemen tary winter feeding. Some feeding of reindeer has been practiced in northern Fennoscandia for centuries (Helle & Jaakkola, 2008; Salmi et al., 2020). In the past, during difficult foraging conditions in winter lichen has been pulled off the trees, trees rich in lichen have been cut down, and hard snow cover has been broken to make digging easier for reindeer (Turunen & Vuojala-Magga, 2014). Annual feeding was adop ted especially in the southern districts in the 1970s. Due to the scarcity of forest pastures rich with ground and arboreal lichen, reindeer are nowadays provided with supplementary feed in the forest, or they are kept and fed in enclosures for some winter months or even throughout the winter (Helle & Jaakkola, 2008; Turunen & Vuojala-Magga, 2014; see also Supplementary text S1.) 3.2. Regional heterogeneity in perceptions a) Perceptions of herders on the factor “Peat extraction”, the black dots show the peat extraction areas; b) “Mining”, the black dots show the locations of the mining districts; c) “Forestry” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect), d-e) Marginal effects of absence or presence of peat extraction/mining districts on the perceptions of herders on factors “Peat extraction and “Mining” (posterior mean with 89% credible intervals), f) Marginal effect of forestry ownership ratio (private or state) on the perceptions of herders on factor “Forestry” (posterior mean with 89% credible intervals). See also Fig. 1a, for the land used for forestry. Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 10 Applied Geography 134 (2021) 102501 S. Rasmus et al. 3.4. Perceptions on herding practices According to the free-form answers of the survey re spondents, the northern districts emphasized the harmfulness of the intensive winter feeding: “Keeping reindeer in enclosures is harmful for them. Field feeding [of reindeer] should be controlled to avoid spoiling of the soil.” Also, several southern districts saw negative sides in enclosure feeding: “Reindeer will become lazy when they are kept in enclosures.” “In some regions the reindeer stag population is weak due to feeding in enclosures, because it keeps the animals in a restricted area.” However, enclosure feeding was considered necessary in the southern area: “There should be a shift from enclosure feeding into forest feeding, but predator pressure is too strong”; “We would feed [the reindeer]in the forests, if we had some [forests], but on private land it is not possible”. Several development needs and ideas about enclosure feeding were given by herders from the southern area: “If [reindeer] must be fed in enclosures, the reindeer should have enough space and clean area”. Active herding was a common practice within most of the RMA until the 1960s (Helle & Jaakkola, 2008). Now it is especially considered as a relevant part of the Saami herding tradition (Jaakkola et al., 2018). On the other hand, growing predator populations and increasing predation pressure on reindeer has increased the need for monitoring and con trolling of herds also in some of the southern districts (Turunen et al., 2017). In our study, active herding was locally seen as a beneficial factor (Fig. 7b). Enclosure calving (Fig. 8a) and summertime earmarking of calves (Fig. 8b) were perceived either as beneficial or harmful locally. Ac cording to one respondent from the northern area: “[The animals] should be handled only when necessary. Bringing reindeer to enclosures for calving is not good for reindeer health.” At least some calvings was managed in 11 Applied Geography 134 (2021) 102501 S. Rasmus et al. Applied Geography 134 (2021) 102501 4 Discussion and conclusions harmful were mainly climatic or related to land use factors Cumulative Fig. 6. a) Hot-spot areas of land-use related factors: the number of land-use forms perceived as harmful per district (forestry, mining, peat extraction, hunt ing, human disturbance, other land-use), b) Percep tions of herders on the factor “Fragmentation of pastures” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive ef fect). 3.4. Perceptions on herding practices Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 7. a) Perceptions of herders on the factors “Pasture rotation” and b) “Herding” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) S. Rasmus et al. Fig. 6. a) Hot-spot areas of land-use related factors: the number of land-use forms perceived as harmful per district (forestry, mining, peat extraction, hunt ing, human disturbance, other land-use), b) Percep tions of herders on the factor “Fragmentation of pastures” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive ef fect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 7. a) Perceptions of herders on the factors “Pasture rotation” and b) “Herding” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of hi i l ) Fig. 7. a) Perceptions of herders on the factors “Pasture rotation” and b) “Herding” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). 3.4. Perceptions on herding practices (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) 4. Discussion and conclusions harmful were mainly climatic or related to land-use factors. Cumulative land-use impacts raised particular concerns. 4.1. Herder perceptions on factors affecting reindeer welfare and reindeer husbandry 4.1. Herder perceptions on factors affecting reindeer welfare and reindeer husbandry (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) S. Rasmus et al. S. Rasmus et al. Fig. 8. a) Perceptions of herders on the “Managing calving in enclosures” and b) “Earm of calves during summer” (−4 denotes a consi negative effect, 0 no effect and 4 denotes a erable positive effect). Purple line: southern b the area specially intended for reindeer hu (ASR; “northern area” in this study). Red line: border of the Saami Homeland area (SHA interpretation of the references to colour in thi legend, the reader is referred to the Web ve this article.) Fig. 8. a) Perceptions of herd “Managing calving in enclosures” of calves during summer” (−4 den negative effect, 0 no effect and 4 erable positive effect). Purple line the area specially intended for (ASR; “northern area” in this study border of the Saami Homeland interpretation of the references to legend, the reader is referred to this article.) Fig. 8. a) Perceptions of herders on the factors “Managing calving in enclosures” and b) “Earmarking of calves during summer” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a consid erable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 9. a) Percentage of reindeer in enclosure feeding during a typical winter, b) Perceptions of herders on factors “Feeding in enclosures” and c) “Herding feeding (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). Purple line: southern border of the area specially intended fo reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 9. a) Percentage of reindeer in enclosure feeding during a typical winter, b) Perceptions of herders on factors “Feeding in enclosures” and c) “Herding feeding” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). 4.1. Herder perceptions on factors affecting reindeer welfare and reindeer husbandry According to the survey respondents, factors directly affecting the welfare of reindeer and the calving success are crucial. These are climate-related factors directly affecting the forage availability and grazing such as basal icing, and land-use related factors limiting the seasonal pasture access. For example, ample mushroom yield was considered as highly beneficial as it is associated with increasing the body condition of the reindeer by the onset of winter. Furthermore, in autumn, the presence of hunting dogs may disperse reindeer herds during the rutting period and thus disturb the calving success next spring. Nature conservation areas were perceived as beneficial, as they secure pasturelands from development activities. Indeed, nature con servation has prevented industrial land use on important grazing lands such as old-growth forests. However, the disadvantage for reindeer We studied perceptions of reindeer herders from Finland on factors affecting the welfare of reindeer and the consequent success of reindeer husbandry. These factors could be divided into three categories: 1) factors which were perceived unanimously positively (for example early spring) or unanimously negatively (for example predation), 2) factors on which perceptions were heterogeneous or even polarized (for example supplementary winter feeding), and 3) factors on which perceptions were localized (for example peat extraction). Factors that were most often considered as beneficial were mainly related to pasture resources or certain herding practices. Factors that were most often seen as 12 Applied Geography 134 (2021) 102501 Applied Geography 134 (2021) 102501 Fig. 8. a) Perceptions of herders on the factors “Managing calving in enclosures” and b) “Earmarking of calves during summer” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a consid erable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Fig. 9. a) Percentage of reindeer in enclosure feeding during a typical winter, b) Perceptions of herders on factors “Feeding in enclosures” and c) “Herding feeding” (−4 denotes a considerable negative effect, 0 no effect and 4 denotes a considerable positive effect). Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). 4.1. Herder perceptions on factors affecting reindeer welfare and reindeer husbandry Purple line: southern border of the area specially intended for reindeer husbandry (ASR; “northern area” in this study). Red line: souther border of the Saami Homeland area (SHA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) husbandry is that these areas are also habitats for predators (Turunen et al., 2017). Some land-use types raise concerns among the herders even if they do not yet exist but are planned in a particular area. Cumulative long-term effects of other land-use forms on reindeer husbandry have been studied earlier for example by looking at the ef fects of forestry actions (Kivinen, 2015) or infrastructure development (Nellemann et al., 2003; Vistnes et al., 2001). Combined local effects of individual tourist resorts (Nellemann et al., 2000) and extractive in dustries (Fohringer et al., unpublished) have also been studied. Tools to assess the cumulative effects of different drivers have been developed (AMAP, 2017) but they have not been empirically tested for reindeer management in northern Fennoscandia. Our analysis of cumulative land use showed the existence of land-use related hotspots as illustrated by herders’ negative perceptions. Some land-use activities are harmful during certain seasons (for example, land use in the proximity of calving grounds or on or along the route from one seasonal pasture to another, Anttonen et al., 2011) or in certain locations. The disturbing effects of some industrial land use, for example open-pit mines, can cover large areas, considerably larger than the spot where the activity takes place. We compared our results with those presented in a recent national report about the sustainability of reindeer husbandry in Finland (Table S1; Kumpula & Siitari, 2020). Also according to that report at least one land-use related factor was considered as “a moderate or considerable problem” to reindeer husbandry in nearly all herding dis tricts. Multiple land-use types were considered problematic in 63% of the districts. 4.4. Limits of the study We are aware of the fact that our analysis is lacking some important factors. Economic, socio-cultural, and governance aspects were not part of the questionnaire survey although these affect herder perceptions and give preconditions for decision-making and herding practices adopted. As perceptions cannot be objectively measured or observed, mis understandings and misinterpretations are possible both by those taking part and by those conducting the survey and further analyses. In our case, there is actually also an interesting latent level of interpretation in the study setting. We asked the survey respondents to consider either positive or negative effects of various factors on reindeer welfare and reindeer husbandry as a whole. What herders actually provided was not only their experiential knowledge of the subject, but also their inter pretation of “the preferences” of those herded – the reindeer. Thus, when providing their answers they also, whether consciously or uncon sciously, came to share their insights on what is good for the reindeer survival and reproduction from the animal point of view. How a certain factor is perceived among the herders seems to be connected also to historical land-use developments of the district. Forestry serves as a good example. Forestry measures are known to have unfavorable impacts on reindeer husbandry, beginning from the loss and fragmentation of pastures and ending with complicated work conditions (Berg et al., 2008; Helle & Jaakkola, 2008; Jaakkola et al., 2013; Kivinen et al., 2012, 2010; Moen & Keskitalo, 2010; Turunen et al., 2020). On the other hand, forestry as a form of land use has been present in the RMA for over 100 years and reindeer husbandry must have been adapting to it (Helle & Jaakkola, 2008; Turunen et al., 2020). Comparing our results to the results of the 30-year-old survey (Supple mentary text S1; Nieminen, 1988; Nieminen & Autto, 1989) provided interesting insights into the diversification and intensification of land use within the RMA. The majority of the survey respondents listed several other land-use pressures that nowadays override the effect of historical and present-day forestry. One might ask whether herders have, for example, got used to practicing herding in managed forests. Also, some framing effect cannot be ruled out due to the selection and phrasing of the factors considered in the questionnaire survey. 4.5. Challenges for land-use planning In nature-based livelihood SESs (K¨ayhk¨o & Horstkotte, 2017), one biophysical or socio-economic driver can affect another. The impact experienced by reindeer herders over a certain period of time is both the sum of impacts and their accumulation over time. This makes the governance of SESs difficult. Competing forms of land use, predation, degradation and fragmentation of pasture resources pose challenges to reindeer husbandry and give rise to conflicts with other land users (Hukkinen et al., 2003; K¨ayhk¨o & Horstkotte, 2017; Merist¨o et al., 2004; Pohjola & Valkonen, 2012; Soppela & Turunen, 2017). Industrial land use such as mining, wind farms and forest clear-cuts causes local but long-lasting impacts on reindeer husbandry. 4.4. Limits of the study The human memory tends to emphasize the most recent and unordinary conditions (Gray, 1955), so it is also possible that recent weather events or topical land-use projects affected some of the responses presented in this study. Only one person (although the key informant) was inter viewed per district in our study. This means that perception of one in dividual has been used to represent rather large land areas. This has limited also further data analyses. Developing detailed models to explain the regional differences in perceptions or studying spatial correlations between the actual land-use activities and perceptions in detail did not seem possible, based on our data. It would be very interesting to deepen the analysis by gathering more data per district. Also, it would be valuable to broaden the analysis to encompass other Nordic countries or even those parts of Russia where reindeer husbandry is practiced. p g p g g g Interestingly, also forest ownership within herding districts seem to affect perceptions towards forestry (Fig. 5f). Perceptions were negative in districts where state is the main owner of forest lands. Also, in Saami Homeland area forestry was seen as a harmful factor. This can be referred to environmental and political struggles between indigenous Saami people and state forestry (Jokinen 2014). Meanwhile, in the south-east part of the RMA in Finland forestry was often seen as a beneficial factor by herders. Most of the forest land there is privately owned. Many herders are also forest owners, and historically reindeer husbandry must have adapted to operate on private lands and with private forest owners. In social terms it is probably easier to express critics towards state-based forestry than local private forestry. Our re sults hint that herder perceptions towards forestry are not determined only by the ecological impacts on pastures, described above. They seem to be based on social, cultural and economic aspects as well. 4.2. Reasons for heterogeneity in herder perceptions Places and landscapes have various meanings which affect the per ceptions. What is perceived as important means that it has value for an individual; for biological survival, and for providing cultural good (Tuan, 1990). For herders, places and landscapes are both working en vironments and grazing lands for their herds. They also carry socio-cultural meanings and heritage. In the language of ecosystem 13 Applied Geography 134 (2021) 102501 S. Rasmus et al. have also shown that herder perceptions influence their actions (AMAP, 2017; Landauer et al., 2021). Local herding tradition and culture may carry traditional knowledge essential in coping during adverse condi tions. They may also limit the willingness to adapt certain new practices, as they may be perceived as harmful to reindeer or the livelihood. services: in addition to providing provisional and supporting services for them and for their herds, landscapes and places provide non-material benefits in the form of cultural ecosystem services (CAFF, 2015; Kettu nen et al., 2012; Markkula et al., 2019). Working as a herder means communality, social ties and identity (Heikkinen et al., 2012; Helle & Jaakkola, 2008; Kumpula & Siitari, 2020). Herding maintains important features of rural landscape, such as pastures and built structures such as reindeer fences and huts, as well as intangible assets of cultural heritage and tradition (Kumpula & Siitari, 2020). In the Saami Homeland area, reindeer livelihood is linked to vitality of the Saami languages and indigenous rights (Jaakkola et al., 2018; Markkula et al., 2019). Regional heterogeneity in perceptions, especially towards herding practices, can party be explained by varying herding traditions and culture in the northern study area (including the Saami Homeland area) and more southern herding districts. 4.3. From perceptions to action? One long-term question in perception geography is how perceptions translate into action (Bunting & Guelke 1979). Generally, our survey respondents considered herding practices used by them as beneficial. It may be that practices are valued because choises to use them have already been made and actions taken. On the other hand, developing and adopting herding practices are considered as ways to cope with changing conditions (Armitage et al., 2011; Turunen & Vuojala-Magga, 2014). Herding practices were in this study mainly considered as in ternal drivers (Table 3), affecting reindeer welfare from their part. On the other hand, they can be considered as coping strategies to mitigate conditions considered as harmful or to utilize the opportunities during conditions perceived beneficial (Table 2). This way the perception (what is harmful/beneficial to reindeer) indeed translates into action (which herding practices to use) and as potential to affect the welfare of rein deer and the success of the livelihood. Furthermore, climate-related risks affect the livelihood (Kumpula & Siitari, 2020; Peltonen-Sainio et al., 2017; Rasmus, Turunen, Luomar anta, et al., 2020; Turunen et al., 2016). Effects of climate change become visible through seasonal weather events which are stochastic and rather short-lived. The probability of extreme weather events such as hot summer periods, icing events and deep snow covers increases within the RMA in the warming climate (Abram et al., 2019; Jylh¨a et al., 2008; Rasmus, Turunen, Luomaranta, et al., 2020). During an extreme weather event, welfare of reindeer can be negatively affected. It all comes down to the sufficiency and diversity of pastures as well as pasture accessibility (Kitti et al., 2006) – or if needed, supplementary forage (L´epy et al., 2018; Pekkarinen et al., 2015). What is particularly detrimental to reindeer husbandry is the Earlier studies on climate change adaptation of reindeer husbandry 14 Applied Geography 134 (2021) 102501 S. Rasmus et al. combination of a harmful weather event and intensive land-use. Similar conclusions were also made about the situation in Norway in a recent review by Tyler et al. (2021). Herders need new strategies to adapt to the changes (Peltonen-Sainio et al., 2017; Rasmus, Turunen, Luomaranta, et al., 2020). 4.3. From perceptions to action? Their capacity to cope with extreme weather events is limited and climate vulnerability is increasing if there is no flexibility in the use of pasture resources, such as seasonal pasture rotation (Anttonen et al., 2011; Degteva et al., 2017; Eira et al., 2018; Pape & L¨offler, 2012). There is a need for more holistic regional land-use planning, which would take several overlapping and neighboring livelihoods into ac count. One solution would be to acknowledge the needs of reindeer husbandry by allocating space to ensure flexibility in pasture use (Kumpula & Siitari, 2020). However, planning should not be targeted at the mean conditions or even at the most probable event since extreme events tend to cause the most harm. just been completed. The number was set by the Ministry of Agriculture and Forestry of Finland, but it was negotiated within a stakeholder working group. The negotiations resulted also in a new process: putting together herding management plans for the pasture areas of every herding district. What this plan will contain in practice is not yet clear, and new biannual negotiations within the stakeholder group and with herding districts will soon begin. In this process, methods to bridge different knowledge sources will be needed. The approach and data presented in this work could be of use in this process, and also in other land-use planning processes aiming at genuine co-management. Perception is not only subjectively interprating the environment, but also acting accordingly. As Thomas and Thomas (1928) formulated: “If men define situations as real, they are real in their consequences”. Herders are central actors in the reindeer management SES. Their per ceptions translate into decision making, planning, and risk prepared ness. Interpretation of a situation, or in our case, perception on drivers of change, leads to actions and shapes the future of reindeer husbandry. In land-use planning, the needs of all land users should be under stood in order to be able to generate synergies, negotiate difficult trade- offs and manage conflicts. Environmental conflicts are mostly consid ered to be caused by differences in knowledge and irreconcilable values (Pettersson et al., 2017). Environmental conflicts can emerge and continue to persist because of a clash of diverging cultural models and frames that stakeholders carry in their individual and collective minds (Jokinen, 2019). These concepts come close to the concept of perception used in our study. Funding Financial support was provided by the Finnish Cultural Foundation (project “Gradual changes and abrupt crises - changing operational environment of Finnish reindeer herding”), Nordforsk (NCoE “Reindeer Husbandry in a Globalizing North – Resilience, Adaptations and Path ways for Actions”, project number 76915) and the Finnish Ministry of Agriculture and Forestry (MAKERA/2016 “Reindeer health in the changing environment” and “Sustainable bioeconomy on reindeer pas tures” projects). Declaration of competing interest None. 4.3. From perceptions to action? Improved understanding of local people’s perceptions could inform and shape political agendas regarding land use, sustain ability and people’s rights, and could lead to more equitable societal processes (Meijaard et al., 2013; Raymond et al., 2009). Author statement Individual contributions to the paper: Sirpa Rasmus: Conceptualization; Methodology; Writing - original draft; Writing - review & editing, Project administration. Henri Wallen: Formal analysis; Methodology; Software; Visualiza tion; Writing - review & editing. Minna Turunen: Conceptualization; Methodology; Writing - original draft; Writing - review & editing. Practitioner knowledge of herders is about local observations but also about interpretations and preferences. Presently, incorporating these types of facts in the environmental assessments and planning procedures is not easy (Chapman & Schott, 2020) and not adequately recognized by decision-makers and land use planners, but would be urgently needed. Local perceptions carry relevant information about the relationships between people and their environments. These subjective and intangible aspects are part of the knowledge of the experienced environment and cannot be excluded even from practical approaches. This is one argument for bringing the perspectives of local communities and livelihoods to the joint planning table. Mia Landauer: Methodology; Writing - original draft; Writing - re view & editing. Juho Tahkola: Investigation; Data curation; Validation; Writing - review & editing. Mikko Jokinen: Methodology; Software; Visualization; Writing - re view & editing. Sauli Laaksonen: Conceptualization, Supervision, Funding acquisi tion, Resources, Writing - review & editing. 4.6. Conclusions - contribution of this study to the participatory environmental governance We revisit the rather old approach of perception geography, where individual values towards, perceptions on, and observations of the environment are studied. Why study perceptions instead of just concentrating on proven land-use pressures or detrimental weather events? Some of these pressures and climate indicators are scarcely studied and poorly known. Understanding the perceptions is needed in managing the present-day and future environmental conflicts (Brown et al., 2020). Participatory environmental governance and public participation in environmental management are increasingly adopted (Adenskog, 2018; Huntington et al., 2019; J¨aske, 2018). Participatory decision making is believed to lead to more deliberate, inclusive and sustainable solutions. These processes have also been criticized for poor stakeholder involvement (e.g., Komendantova et al., 2015). Reindeer herders have experienced power imbalances in the negotiations with the governance of the livelihood, feeling that their herding practices do not get enough recognition and support and their voice is not heard (Lan dauer & Komendantova, 2018; Markkula et al., 2019). Our approach presents one tool which can be used to facilitate these processes. Local perceptions are needed as a relevant part of balanced discussion. They also carry valid local and traditional knowledge that can be bridged with scientific knowledge of the issues studied (Abu et al., 2019; Chapman & Schott, 2020). References https://doi.org/10.1007/s10113-019- 01478-8 Brown, G., & Kytt¨a, M. (2014). Key issues and research priorities for public participation GIS (PPGIS): A synthesis based on empirical research. Applied Geography, 46, 122–136. https://doi.org/10.1016/j. apgeog.2013.11.004 Brown, G., Reed, P., & Raymond, C. (2020). Mapping place values: 10 lessons from two decades of public participation GIS empirical research. Applied Geography, 116, 102156. Itkonen, T. I. (1948). Suomen lappalaiset. Osa 1. Suomen lappalaiset vuoteen 1945 [Part 1. Lapps of Finland until 1945] osa 2. Poronhoito [Part 2. Reindeer herding]. Porvoo: WSOY. Bürkner, P.-C. (2017). brms: An R package for Bayesian Multilevel Models using Stan. R package version 2.15.0. Journal of Statistical Software, 80(1), 1–28. https://doi.org/ 10.18637/jss.v080.i01 Jaakkola, L. M., Heiskanen, M. M., Lensu, A. M., & Kuitunen, M. (2013). Consequences of forest landscape changes for the availability of winter pastures to reindeer (Rangifer tarandus tarandus) from 1953 to 2003 in Kuusamo, Northeast Finland. Boreal Environment Research, 18, 459–472. j Bürkner, Paul-Christian, & Vuorre, Matti (2018). Ordinal Regression Models in Psychology: A Tutorial. https://doi.org/10.31234/osf.io/x8swp Jaakkola, J. J. K., Juntunen, S., & N¨akk¨al¨aj¨arvi, K. (2018). The holistic effects of climate change on the culture, well-being, and health of the Saami, the only Indigenous People in the European Union. Current Environmental Health Reports, 5, 401–417. J¨a¨ask¨o, O. (2001). Paliskunnan vuorovaikutteinen porotaloussuunnitelma. L¨aht¨okohtia ja ehdotuksia paliskunnan suunnittelun ja vuorovaikutuksen kehitt¨amiseen. Arctic Centre: Jaakkola, J. J. K., Juntunen, S., & N¨akk¨al¨aj¨arvi, K. (2018). The holistic effects of climate change on the culture, well-being, and health of the Saami, the only Indigenous People in the European Union. Current Environmental Health Reports, 5, 401–417. CAFF. (2015). The economics of ecosystems and biodiversity (TEEB) scoping study for the arctic. Conservation of arctic flora and fauna. Iceland: Akureyri. change on the culture, well-being, and health of the Saami, the only Indigenous People in the European Union. Current Environmental Health Reports, 5, 401–417. J¨a¨ask¨o, O. (2001). Paliskunnan vuorovaikutteinen porotaloussuunnitelma. L¨aht¨okohtia ja ehdotuksia paliskunnan suunnittelun ja vuorovaikutuksen kehitt¨amiseen. Arctic Centre: Rovaniemi: University of Lapland. l l Chapman, J. M., & Schott, S. (2020). Knowledge coevolution: Generating new understanding through bridging and strengthening distinct knowledge systems and empowering local knowledge holders. Sustainability Science. https://doi.org/ 10.1007/s11625-020-00781-2 J¨a¨ask¨o, O. (2001). Paliskunnan vuorovaikutteinen porotaloussuunnitelma. L¨aht¨okohtia ja ehdotuksia paliskunnan suunnittelun ja vuorovaikutuksen kehitt¨amiseen. Arctic Centre: Rovaniemi: University of Lapland. l J¨aske, M. (2018). Participatory innovations and maxi-publics: The influence of participation possibilities on perceived legitimacy at the local level in Finland. References Oslo: Arctic Monitoring and Assessment Programme (AMAP). N. Gunslay, & Y. Konstantinov (Eds.), Reindeer management in northernmost europe: Linking practical and scientific knowledge in social-ecological system (pp. 73–93). Berlin: Springer-Verlag: Ecological Studies 184. Anttonen, M., Kumpula, J., & Colpaert, A. (2011). Range selection by semi-domesticated reindeer (Rangifer tarandus tarandus) in relation to infrastructure and human activity in the boreal forest environment, northern Finland. Arctic, 64(1), 1–14. Heikkinen, H. I., Kasanen, M., & L´epy, ´E. (2012). Resilience, vulnerability and adaptation in reindeer herding communities in the Finnish-Swedish border area. Nordia Geographical Publications, 41(5), 107–121. Armitage, D., Berkes, F., Dale, A., Kocho-Schellenberg, E., & Patton, E. (2011). Co- management and the co-production of knowledge: Learning to adapt in Canada’s Arctic. Global Environmental Change, 21, 995–1004. https://doi.org/10.1016/j. gloenvcha.2011.04.006 Helander-Renvall, E. (2014). Relationships between Sami reindeer herders, lands, and reindeer. In G. Marvin, & S. McHugh (Eds.), Routledge handbook of human-animal studies (pp. 246–258). London and New York: Routledge. Bergeron, J., Paquette, S., & Poullaouec-Gonidec, P. (2014). Uncovering landscape values and micro-geographies of meanings with the go-along method. Landscape and Urban Planning, 122, 108–121. Helle, T., & Jaakkola, L. M. (2008). Transition in herd management of semi-domesticated reindeer in northern Finland. Annales Zoologici Fennici, 45(2), 81–101. k ´ ´ ll ( ) l f d d Horstkotte, T., L´epy, ´E., & Risvoll, C. (2020). Supplementary feeding in reindeer husbandry: Results from a workshop with reindeer herders and researchers from Norway, Sweden and Finland. Umeå: (REXSAC). https://www.rexsac.org/publicat ions/supplementary-feeding-reindeer-husbandry-results-workshop-reindeer-herders -researchers-norway-sweden-finland/. Berg, A., ¨Ostlund, L., Moen, J., & Olofsson, J. (2008). A century of logging and forestry in a reindeer herding area in northern Sweden. Forest Ecology and Management, 256, 1009–1020. Bernard, H. R. (1995). Research methods in anthropology: Qualitative and quantitative approaches. Walnut Creek, CA: Sage. i Hukkinen, J., J¨a¨ask¨o, O., Laakso, A., Müller-Wille, L., Nevalainen, S., & Raitio, K. (2003). Poronhoitokulttuurin arvo Suomessa – haasteet hallinnolle, ohjaukselle ja valvonnalle. Technology, society, environment 1/2003. Espoo: Helsinki University of Technology. Bernes, C., Bråthen, K. A., Forbes, B. C., Speed, J. D., & Moen, J. (2015). What are the impacts of reindeer/caribou (Rangifer tarandus L.) on arctic and alpine vegetation? A systematic review. Environmental Evidence, 4(1), 4. gy y p y gy Huntington, H. P., Carey, M., Apok, C., Forbes, B. C., Fox, S., Holm, L. K., & Stammler, F. (2019). Climate change in context: Putting people first in the arctic. Regional Environmental Change, 19(4), 1217–1223. Acknowledgements We would like to thank the survey respondents for their time and contribution. The Reindeer Herders’ Association is acknowledged for its collaboration during the work. We are grateful to Sanna Hast and Leena Valkeap¨a¨a for the valuable discussions during the preparation of this manuscript. We express our warm thanks to Sonja Kivinen, Vesa Nivala, Kari Oinonen and Jani R¨aih¨a for the technical help. We would also like to thank the researchers of the project “Sustainable bioeconomy on reindeer pastures”, especially Jouko Kumpula and Heli Saarikoski, for collaboration and access to the complementary material we needed. Financial support was provided by the Finnish Cultural Foundation (project “Gradual changes and abrupt crises - changing operational environment of Finnish reindeer herding”), Nordforsk (NCoE “Reindeer Husbandry in a Globalizing North – Resilience, Adaptations and While writing this paper, the process of setting the maximum allowed number of reindeer for the period 2020–2029 for the RMA has 15 S. Rasmus et al. Applied Geography 134 (2021) 102501 Pathways for Actions”, project number 76915) and the Finnish Ministry of Agriculture and Forestry (MAKERA/2016 “Reindeer health in the changing environment” and “Sustainable bioeconomy on reindeer pas tures” projects). Eira, I. M. G., Oskal, A., Hanssen-Bauer, I., & Mathiesen, S. D. (2018). Snow cover and the loss of traditional indigenous knowledge. Nature Climate Change, 8, 924–936. d h l b l h db k Finger, M., & Heininen, L. (Eds.). (2019). The global arctic handbook. Springer International Publishing. Finnish Environmental Institute. (2020). Open land-use database. https://liiteri.ympari sto.fi/. i Forbes, B. C. (2006). The challenges of modernity for reindeer management in northernmost Europe. In B. C. Forbes, M. B¨olter, L. Müller-Wille, J. Hukkinen, F. Müller, N. Gunslay, & Y. Konstantinov (Eds.), Reindeer management in northernmost europe: Linking practical and scientific knowledge in social-ecological system (pp. 11–25). Berlin: Springer-Verlag: Ecological Studies 184. References Gabry, J., & Mahr, T. (2021). bayesplot: Plotting for Bayesian Models. R package version 1.8.0, URL https://mc-stan.org/bayesplot/. Abram, N., Gattuso, J.-P., Prakash, A., Cheng, L., Chidichimo, M. P., Crate, S., Enomoto, H., Garschagen, M., Gruber, N., Harper, S., Holland, E., Kudela, R. M., Rice, J., Steffen, K., & von Schuckmann, K. (2019). Framing and context of the report. In H.-O. P¨ortner, D. C. Roberts, V. Masson-Delmotte, P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, & N. M. Weyer (Eds.), IPCC special report on the ocean and cryosphere in a changing climate. In press. Gelman, A., Vehtari, A., Simpson, D., Margossian, C. C., Carpenter, B., Yao, Y., Kennedy, L., Gabry, J., Bürkner, P.-C., & Modr´ak, M. (2020). Bayesian workflow. arXiv:2011.01808v1 [stat.ME]. Golledge, R. G., & Stimson, R. J. (1997). Spatial behavior – a geographic perspective. New York and London: The Guilford Press. Gray, P. G. (1955). The memory factor in social surveys. Journal of the American Statistical Association, 50(270), 344–363. Abu, R., Reed, M., & Jardine, T. (2019). Using two-eyed seeing to bridge Western science and Indigenous knowledge systems and understand long-term change in the Saskatchewan River Delta, Canada. International Journal of Water Resources Development. https://doi.org/10.1080/07900627.2018.1558050 Grotenfelt. (1919). Suomen poronhoito [Reindeer herding in Finland]. Helsinki: Otava. Guelke, L. (2003). Forms of life, history, and mind: An idealist proposal for integrating perception and behaviour in human geography. The behavioural environment. Routledge. Guelke, L. (2003). Forms of life, history, and mind: An idealist proposal for integrating perception and behaviour in human geography. The behavioural environment. Routledge. Heikkil¨a, L. (2006). The comparison of indigenous and scientific perceptions of reindeer management. In B. C. Forbes, M. B¨olter, L. Müller-Wille, J. Hukkinen, F. Müller, N. Gunslay, & Y. Konstantinov (Eds.), Reindeer management in northernmost europe: Linking practical and scientific knowledge in social-ecological system (pp. 73–93). Berlin: Springer-Verlag: Ecological Studies 184. ´ Adenskog, M. (2018). Democratic innovations in political systems - towards a systemic approach. ¨Orebro Studies in Political Science, 42. Heikkila, L. (2006). The comparison of indigenous and scientific perceptions of reindeer management. In B. C. Forbes, M. B¨olter, L. Müller-Wille, J. Hukkinen, F. Müller, N. Gunslay, & Y. Konstantinov (Eds.), Reindeer management in northernmost europe: Linking practical and scientific knowledge in social-ecological system (pp. 73–93). Berlin: Springer-Verlag: Ecological Studies 184. ´ AMAP. (2017). Adaptation actions for a changing arctic: Perspectives from the barents area. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.apgeog.2021.102501. Forbes, B. C., Turunen, M., Soppela, P., Rasmus, S., Vuojala-Magga, T., & Kitti, H. (2020). Changes in mountain birch forests and reindeer management: Comparing different knowledge systems in s´apmi, northern Fennoscandia. Polar Record. https://doi.org/ 10.1017/S0032247419000834 References Socio-economic importance of ecosystem services in the nordic countries, synthesis in the context of the economics of ecosystems and biodiversity (TEEB). Tema Nord 2012, 559. https://doi.org/10.6027/TN2012-559 i Nikula, A., Turunen, M., Bogad´ottir, R., Markkula, I., & Kantola, S. (2020). PPGIS for a better understanding of people’s values: Experiences from Finland and the Faroe Islands. In J. McDonagh, & S. Tuulentie (Eds.), Sharing Knowledge for land use management: Decision-making and Expertise in europe’s northern periphery, 70-85. Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Kitti, H., Gunslay, N., & Forbes, B. (2006). Defining the quality of reindeer pastures: The perspectives of s´ami reindeer herders. In B. C. Forbes, M. B¨olter, L. Müller-Wille, J. Hukkinen, F. Müller, N. Gunslay, & Y. Konstantinov (Eds.), Reindeer management in northernmost europe: Linking practical and scientific knowledge in social-ecological system (pp. 141–165). Berlin: Springer-Verlag: Ecological Studies 184. Ostrom, E. (2009). A general framework for analyzing sustainability of social-ecological systems. Science, 325(5939), 419–422. Pakkanen, A., & Valkonen, J. (2012). Porotalouden hyvinvointi ja tulevaisuuskuvat etel¨aisiss¨a paliskunnissa. In A. Pohjola, & A. Valkonen (Eds.), Poronhoitajien hyvinvoinnin uhat ja avun tarpeet. Rovaniemi: Lapland University Press. ll i (pp. 141–165). Berlin: Springer-Verlag: Ecological Studies 184 Kivinen, S. (2015). Many a little makes a mickle: Cumulative land cover changes and traditional land use in the Kyr¨o reindeer herding district, northern Finland. Applied Geography, 63, 204–211. ¨ Pape, R., & L¨offler, J. (2012). Climate change, land use conflicts, predation and ecological degradation as challenges for reindeer husbandry in Northern Europe: What do we really know after half a century of research? Ambio, 41(5), 421–434. https://doi.org/10.1007/s13280-012-0257-6 Kivinen, S., Berg, A., Moen, J., ¨Ostlund, L., & Olofsson, J. (2012). Forest fragmentation and landscape transformation in a reindeer husbandry area in Sweden. Environmental Management, 49(2), 295–304. Å Pebesma, E. (2018). Simple features for R: Standardized support for spatial vector data. The R Journal, 10(1), 439–446. https://doi.org/10.32614/RJ-2018-009 Kivinen, S., Moen, S., Berg, A., & Eriksson, Å. (2010). Effects of modern forest management on winter grazing resources for reindeer in Sweden. Ambio, 39, 269–278. Pekkarinen, A.-J., Kumpula, J., & Tahvonen, O. (2015). Reindeer management and winter pastures in the presence of supplementary feeding and government subsidies. Ecological Modelling, 312, 256–271. Komendantova, N., Vocciante, M., & Battaglini, A. (2015). Can the BestGrid process improve stakeholder involvement in electricity transmission projects? Energies, 8(9), 9407–9433. References Peltonen-Sainio, P., Sorvali, J., Müller, M., Huitu, O., Neuvonen, S., Nummelin, T., Rummukainen, A., et al. (2017). Sopeutumisen tila 2017 : Ilmastokest¨avyyden tarkastelut maa- ja mets¨atalousministeri¨on hallinnonalalla. Luonnonvara- ja biotalouden tutkimus 18/2017. Helsinki: Luonnonvarakeskus. Kortesalmi, J. J. (2007). Poronhoidon synty ja kehitys Suomessa [The origins and development of peasant reindeer management in Finland]. Suomalaisen Kirjallisuuden Seuran Toimituksia (Vol. 1149). Tampere: Tammer-Paino Oy. Pettersson, S., Hallikainen, V., Naskali, A., Rovanper¨a, S., & Tuulentie, S. (2017). Ymp¨arist¨okonfliktit suomessa: Mist¨a on kiistelty ja miksi? (Environmental conflicts in Finland: What issues have been disputed over and why?). Terra, 129(2), 87–107. P hj l A & V lk J (2012) P h it ji h i i i h t j t t Kumpula, J., Kurkilahti, M., Helle, T., & Colpaert, A. (2014). Both reindeer management and several other land use factors explain the reduction in ground lichens (Cladonia spp.) in pastures grazed by semi-domesticated reindeer in Finland. Regional Environmental Change, 14, 541–559. Pohjola, A., & Valkonen, J. (2012). Poronhoitajien hyvinvoinnin uhat ja avun tarpeet. Rovaniemi: Lapland University Press. i Rovaniemi: Lapland University Press. Porsanger, J., & Guttorm, G. (2011). Building up the field study and research on Sami traditional knowledge (Arbediehtu). Diedut, 1(11), 13–57. p y Porsanger, J., & Guttorm, G. (2011). Building up the field study and research on Sami traditional knowledge (Arbediehtu). Diedut, 1(11), 13–57. Kumpula, J., & Siitari, S. (Eds.). (2020). Kest¨av¨a biotalous porolaitumilla -hankkeen osaraportit, johtop¨a¨at¨okset ja toimenpide-ehdotukset. Luonnonvara- ja biotalouden tutkimus 4/2020. Helsinki: Luonnonvarakeskus. R Core Team. (2020). R version 4.0.0 (2020-04-24): A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/. Kynk¨a¨anniemi, S. (2020). The relationship between the reindeer (Rangifer tarandus tarandus) and the ectoparasitic deer ked (Lipoptena cervi): Reindeer welfare aspects. Acta Universitatis Ouluensis, 741, 2020 (Tampere: Punamusta). Rasmus, S., Kivinen, S., & Irannezhad, M. (2018). Basal ice formation in Northern Finland snow covers during 1948-2016. Environmental Research Letters, 13(2018), 114009. https://doi.org/10.1088/1748-9326/aae541 Laaksonen, S. (2016). Tunne poro : Poron sairaudet ja terveydenhoito. Kuusamo: Wazama Media Oy. Rasmus, S., Turunen, M., Luomaranta, A., Kivinen, S., Jylh¨a, K., & R¨aih¨a, J. (2020). Climate change and reindeer management in Finland: Co-analysis of practitioner knowledge and meteorological data for better adaptation. The Science of the Total Environment, 710(2020), 136229. https://doi.org/10.1016/j.scitotenv.2019.136229 Laaksonen, S., Oksanen, A., Kutz, S., Jokelainen, P., Holma-Suutari, A., & Hoberg, E. (2017). References Filarioid nematodes, threat to arctic food safety and security – bioinvasion of vector-borne filaroid nematodes in the arctic and boreal ecosystems. In P. Paulsen, A. Bauer, & F. J. M. Smulders (Eds.), Game meat hygiene: Food safety and security (pp. 101–120). Wageningen, the Netherlands: Wageningen Academic Publishers. i A. Bauer, & F. J. M. Smulders (Eds.), Game meat hygiene: Food safety and security (pp. 101–120). Wageningen, the Netherlands: Wageningen Academic Publishers. Rasmus, S., Turunen, M., Norberg, H., Kojola, I., Kumpula, J., & Ollila, T. (2020b). Mission impossible? Pursuing the co-existence of viable predator populations and sustainable reindeer husbandry in Finland. Journal of Rural Studies, 80, 135–148. Landauer, M., & Komendantova, N. (2018). Participatory environmental governance of infrastructure projects affecting reindeer husbandry in the Arctic. Journal of Environmental Management, 223, 385–395. Raymond, C. M., Bryan, B. A., MacDonald, D. H., Cast, A., Strathearn, S., et al. (2009). Mapping community values for natural capital and ecosystem services. Ecological Economics, 68, 1301–1315. Landauer, M., Rasmus, S., & Forbes, B. C. (2021). What drives reindeer management towards social and ecological tipping points? Regional Environmental Change. https:// doi.org/10.1007/s10113-021-01757-3 Rees, W. G., Stammler, F. M., Danks, F. S., & Vitebsky, P. (2008). Vulnerability of European reindeer husbandry to global change. Climatic Change, 87(1–2), 199. R X Ch Y Y K & T Y (2016) Ri k ti d bli t Rees, W. G., Stammler, F. M., Danks, F. S., & Vitebsky, P. (2008). Vulnerability of European reindeer husbandry to global change. Climatic Change, 87(1–2), 199. Ren, X., Che, Y., Yang, K., & Tao, Y. (2016). Risk perception and public acceptance toward a highly protested Waste-to-Energy facility. Waste Management, 48, 528–539. L´epy, A., Heikkinen, H. I., Komu, T., & Sarkki, S. (2018). Participatory meaning making of environmental and cultural changes in reindeer herding in the northernmost border area of Sweden and Finland. International Journal of Business and Globalisation, 20(2), 203. Ren, X., Che, Y., Yang, K., & Tao, Y. (2016). Risk perception and public acceptance toward a highly protested Waste-to-Energy facility. Waste Management, 48, 528–539. toward a highly protested Waste-to-Energy facility. Waste Management, 48, 528–539. RHA (Reindeer Herders’ Association). (2018). Reindeer statistics for Finland. Å Markkula, I., Turunen, M., & Kantola, S. (2019). Traditional and local knowledge in land use planning: Insights into the use of the akw´e: Kon guidelines in eanodat, Finnish s´apmi. Ecology and Society, 24(1), 20. https://doi.org/10.5751/ES-10735-240120 Riseth, J.Å., Tømmervik, H., & Bjerke, J. References European Journal of Political Research, 58(2), 603–630. Coll, M., Carreras, M., Ci´ercoles, C., Cornax, M. J., Gorelli, G., Morote, E., & Saez, R. (2014). Assessing fishing and marine biodiversity changes using Fishers’ perceptions: The Spanish mediterranean and gulf of cadiz case study. PloS One, 9(1), Article e85670. Jokinen, M. (2019). Lapin ymp¨arist¨okiistojen kulttuuriset tekij¨at. Dissertationes Forestales, 281. https://doi.org/10.14214/df.281 Degteva, A., Oskal, A., Mathiesen, S. D., & Burgess, P. (2017). Indigenous peoples’ perspectives. In M. Forsius, M. Olsen, S. Kalhok, M. Klint, O. M¨ah¨onen, H. Jensson, M. Kroglund, T. Lundeberg, Y. Tsaturov, J. M. Kuperberg, E. Krummel, J. Staffansson, & B. van Dijken (Eds.), Adaptation Actions for a changing arctic: Perspectives from the barents area (pp. 167–194). Oslo: Arctic Monitoring and Assessment Programme (AMAP). Jylh¨a, K., Fronzek, S., Tuomenvirta, H., Carter, T. R., & Ruosteenoja, K. (2008). Changes in frost, snow and Baltic sea ice by the end of the twenty-first century based on climate model projections for Europe. Climatic Change, 86(3–4), 441–462. Kahila-Tani, M., Broberg, A., Kytt¨a, M., & Tyger, T. (2016). Let the citizens map—public participation GIS as a planning support system in the helsinki master plan process. Planning Practice & Research, 31(2), 195–214. https://doi.org/10.1080/ 02697459.2015.1104203 Downs, R. (1970). Geographic space perception: Past approaches and future prospects. In C. Board, R. Chorley, P. Haggett, & D. Stoddart (Eds.), Progress in geography (pp. 65–108). London: Edward Arnold. Kantola, S., Uusitalo, M., Nivala, V., & Tuulentie, S. (2018). Tourism resort users’ participation in planning: Testing the public participation geographic information system method in Levi, Finnish Lapland. Tourism Management Perspectives, 27, 22–32. https://doi.org/10.1016/j.tmp.2018.04.001 Eide, W., Keskitalo, E. C., & Kovacs, K. M. (2017). Status of the natural and human environments. In M. Forsius, M. Olsen, S. Kalhok, M. Klint, O. M¨ah¨onen, H. Jensson, M. Kroglund, T. Lundeberg, Y. Tsaturov, J. M. Kuperberg, E. Krummel, J. Staffansson, & B. van Dijken (Eds.), Adaptation actions for a changing arctic: Perspectives from the barents area (pp. 5–46). Oslo: Arctic Monitoring and Assessment Programme (AMAP). K¨ayhk¨o, J., & Horstkotte, T. (2017). Reindeer husbandry under global change in the tundra region of Northern Fennoscandia. No. 1. In Publications from the Department of geography and geology, University of Turku (p. 73). Turku: Painosalama Oy. ISBN 978- 951-29-6703-2. 16 S. Rasmus et al. Applied Geography 134 (2021) 102501 Kettunen, M., Vihervaara, P., Kinnunen, S., D’Amato, D., Badura, T., Argimon, M., & Ten Brink, P. (2012). References W. (2016). 175 years of adaptation: North Scandinavian S´ami reindeer herding between government policies and winter climate variability (1835–2010). Journal of Forest Economics, 24, 186–204. https:// doi.org/10.1016/j.jfe.2016.05.002 ¨ McGinnis, M. D., & Ostrom, E. (2014). Social-ecological system framework: Initial changes and continuing challenges. Ecology and Society, 19(2), 30. Salmi, A.-K., Fjellstr¨om, M., ¨Aik¨as, T., Spangen, M., Nú˜nez, M., & Lid´en, K. (2020). Zooarchaeological and stable isotope evidence of S´ami reindeer offerings. Journal of Archaeological Science. Reports, 29, 102129. https://doi.org/10.1016/j. jasrep.2019.102129 Meijaard, E., Abram, N. K., Wells, J. A., Pellier, A. S., Ancrenaz, M., Gaveau, D. L., & Mengersen, K. (2013). People’s perceptions about the importance of forests on Borneo. PloS One, 8(9), Article e73008. Merist¨o, T., J¨arvinen, J., Kettunen, J., & Nieminen, M. (2004). Porotalouden tulevaisuus – ”Mitk¨a ovat mahdolliset maailmat”. Skenaarioluonnosten esittely. Kala- ja riistaraportteja nro 315. Helsinki: Riistan- ja kalatutkimus. Sarkki, S., Heikkinen, H., Herva, V., & Saarinen, J. (2018). Myths on local use of natural resources and social equity of land use governance: Reindeer herding in Finland. Land Use Policy, 77, 322–331. j j Mets¨ahallitus. (2019). Land-use database by mets¨ahallitus/national board of forestry in Finland. Sonnenfeld, J. (1972). Geography, perception and the behavioral environment. Man, Space and the Environment, 12(1), 244–251. Moen, J., & Keskitalo, E. C. H. (2010). Interlocking panarchies in multi-use boreal forests in Sweden. Ecology and Society, 15(3), 17. Soppela, P., & Turunen, M. (2017). Luku 4 Sopeutuuko porotalous kasautuvien muutosten paineessa? In M. Tennberg, J. Haapala, A. Hannukkala, J. P. Jaakkola, T. Jouttij¨arvi, K. Jylh¨a, S. Kauppi, et al. (Eds.), Barentsin alue muuttuu – Miten Suomi sopeutuu? Valtioneuvoston selvitys- ja tutkimustoiminnan julkaisusarja 31. Muuttoranta, K., Holand, Ø., Røed, K. H., Tapio, M., Nieminen, M., & M¨aki-Tanila, A. (2014). Genetic variation in meat production related traits in reindeer (Rangifer tarandus). Rangifer, 34(1), 21–36. Tennekes, M. (2018). tmap: Thematic maps in R. Journal of Statistical Software, 84(6), 1–39. https://doi.org/10.18637/jss.v084.i06 N¨akk¨al¨aj¨arvi, K., & Jaakkola, J. (2017). Saamelaiset ja muutos. In M. Tennberg, J. Haapala, A. Hannukkala, J. P. Jaakkola, T. Jouttij¨arvi, K. Jylh¨a, S. Kauppi, et al. (Eds.), Barentsin alue muuttuu – Miten Suomi sopeutuu? Valtioneuvoston selvitys- ja tutkimustoiminnan julkaisusarja 31. Thomas, W. I., & Thomas, D. S. (1928). The child in America: Behavior problems and programs (pp. 571–572). New York: Knopf. programs (pp. 571–572). New York: Knopf. Tuan, Y.-F. (1990). Topohilia (Morningside Edition). Columbia University Press. References T Y F (2003) P l d l l h A A l f h National Land Use Guidelines. (2017). Land-use guidelines in Finland. https://www.ym paristo.fi/en-US/Living_environment_and_planning/Land_use_planning_system/Na tional_land_use_guidelines. Tuan, Y.-F. (2003). Perceptual and cultural geography: A commentary. Annals of the Association of American Geographers, 93(4), 878–881. Turunen, M., Rasmus, S., Bavay, M., Ruosteenoja, K., & Heiskanen, J. (2016). Coping with increasingly difficult weather and snow conditions: Reindeer herders’ views on climate change impacts and coping strategies. Climate Risk Management, 11(2016), 15–36. https://doi.org/10.1016/j.crm.2016.01.002 Nellemann, C., Jordhøy, P., Støen, O.-G., & Strand, O. (2000). Cumulative impacts of tourist resorts on wild reindeer (Rangifer tarandus tarandus) during winter. Arctic, 53 (1), 9–17. Nellemann, C., Vistnes, I., Jordhøy, P., Strand, O., & Newton, A. (2003). Progressive impact of piecemeal infrastructure development on wild reindeer. Biological Conservation, 113, 307–317. Turunen, M., Rasmus, S., J¨arvenp¨a¨a, J., & Kivinen, S. (2020). Relations between forestry and reindeer husbandry in northern Finland: Perspectives of science and practice. Forest Ecology and Management, 117677. https://doi.org/10.1016/j. foreco.2019.117677 Nieminen, M. (1988). Porojen laitumet ja ruokinta poronhoitovuonna 1986-87. I Laitumet ja niiden kunto. Poromies, 6. Turunen, M., Rasmus, S., Norberg, H., Kumpula, J., Kojola, I., & Ollila, T. (2017). Porot ja pedot – kuinka poronhoidon sopeutuminen petoihin on muuttunut 90 vuodessa? Suomen Riista, 63, 19–42. Nieminen, M., & Autto, P. (1989). Porojen laitumet ja ruokinta poronhoitovuonna 1986- 87. II Rehut ja ruokinta. Poromies, 2. 17 S. Rasmus et al. Turunen, M., & Vuojala-Magga, T. (2014). Past and present winter feeding of reindeer in Finland: Herders adaptive learning of the practices. Arctic, 67(2), 173–188. https:// doi.org/10.14430/arctic4385 Turunen, M., & Vuojala-Magga, T. (2013). Porojen talviruokinta: Luppopuiden hakkuusta tarharuokintaan. [With English summary: Reindeer winter feeding: From lichen tree cuttings to pen feeding]. Suomen Riista, 59, 86–99. Turunen, M., & Vuojala-Magga, T. (2013). Porojen talviruokinta: Luppopuiden hakkuusta tarharuokintaan. [With English summary: Reindeer winter feeding: From lichen tree cuttings to pen feeding]. Suomen Riista, 59, 86–99. Turunen, M., & Vuojala-Magga, T. (2014). Past and present winter feeding of reindeer in Finland: Herders adaptive learning of the practices. Arctic, 67(2), 173–188. https:// doi.org/10.14430/arctic4385 Wickham, et al. (2019). Welcome to the tidyverse. Journal of Open Source Software, 4(43), 1686. https://doi.org/10.21105/joss.01686 Applied Geography 134 (2021) 102501 Vistnes, I., Nellemann, C., Jordhøy, P., & Strand, O. (2001). Wild reindeer: Impacts of progressive infrastructure development on distribution and range use. Polar Biology, 24, 531–537. Wickham, et al. (2019). Welcome to the tidyverse. Journal of Open Source Software, 4(43), 1686. https://doi.org/10.21105/joss.01686 Working committee. (1992). Poronhoitolain seurantaty¨oryhm¨an muistio. Rovaniemi: Poronhoitolain seurantaty¨oryhm¨a [Working committee for the follow-up of the herding legislation]. 18
https://openalex.org/W2280056530	https://europepmc.org/articles/pmc4949522?pdf=render	English	null	The anabolic action of intermittent parathyroid hormone on cortical bone depends partly on its ability to induce nitric oxide‐mediated vasorelaxation in BALB/c mice	Cell biochemistry and function	2,016	cc-by	10,309	cell biochemistry and function Cell Biochem Funct 2016; 34: 52–62. Published online 1 February 2016 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cbf.3164 Correspondence to: Stéphanie Gohin, Department of Comparative Bio- medical Sciences, Royal Veterinary College, London NW1 0TU, UK. E-mail: sgohin@rvc.ac.uk The copyright line has been corrected on 04 July2016 after ﬁrst publication. The anabolic action of intermittent parathyroid hormone on cortical bone depends partly on its ability to induce nitric oxide-mediated vasorelaxation in BALB/c mice key words—parathyroid hormone; bone blood ﬂow; nitric oxide; vasodilation; cortical bone key words—parathyroid hormone; bone blood ﬂow; nitric oxide; vasodilation; cortical bone Revised 17 December 2015 Accepted 1 January 2016 © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. 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. Received 19 October 2015 Revised 17 December 2015 Accepted 1 January 2016 The anabolic action of intermittent parathyroid hormone on cortical bone depends partly on its ability to induce nitric oxide-mediated vasorelaxation in BALB/c mice S Gohin1,3,4, A Carriero2, C Chenu4, AA Pitsillides4, TR Arnett3 and M Marenzana1,5 1Department of Bioengineering, Imperial College London, London, UK 2Department of Biomedical Engineering, Florida Institute of Technology, Florida, USA 3Department of Cell and Developmental Biology, University College London, London, UK 4Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK 5Kennedy Institute of Rheumatology, University of Oxford, Oxford, UK 1Department of Bioengineering, Imperial College London, London, UK 2Department of Biomedical Engineering, Florida Institute of Technology, Florida, USA 3Department of Cell and Developmental Biology, University College London, London, UK 4Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK 5Kennedy Institute of Rheumatology, University of Oxford, Oxford, UK 1Department of Bioengineering, Imperial College London, London, UK 2Department of Biomedical Engineering, Florida Institute of Technology, Florida, USA 3Department of Cell and Developmental Biology, University College London, London, UK 4Department of Comparative Biomedical Sciences, Royal Veterinary College, London, UK 5Kennedy Institute of Rheumatology, University of Oxford, Oxford, UK There is strong evidence that vasodilatory nitric oxide (NO) donors have anabolic effects on bone in humans. Parathyroid hormone (PTH), the only osteoanabolic drug currently approved, is also a vasodilator. We investigated whether the NO synthase inhibitor L-NAME might alter the effect of PTH on bone by blocking its vasodilatory effect. BALB/c mice received 28 daily injections of PTH[1–34] (80 μg/kg/day) or L-NAME (30 mg/kg/day), alone or in combination. Hindlimb blood perfusion was measured by laser Doppler imaging. Bone architecture, turnover and mechanical properties in the femur were analysed respectively by micro-CT, histomorphometry and three-point bending. PTH increased hindlimb blood ﬂow by >30% within 10 min of injection (P < 0.001). Co-treatment with L-NAME blocked the action of PTH on blood ﬂow, whereas L-NAME alone had no effect. PTH treatment increased femoral cortical bone volume and formation rate by 20% and 110%, respectively (P < 0.001). PTH had no effect on trabecular bone volume in the femoral metaphysis although trabecular thickness and number were increased and decreased by 25%, respectively. Co-treatment with L-NAME restricted the PTH-stimulated increase in cortical bone formation but had no clear-cut effects in trabecular bone. Co-treatment with L-NAME did not affect the mechanical strength in femurs induced by iPTH. These results suggest that NO-mediated vasorelaxation plays partly a role in the anabolic action of PTH on cortical bone. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. INTRODUCTION Both the vascular remod- elling and the osteoanabolic effects of iPTH have been shown to depend on the potent angiogenic factor, vascular endothelial growth factor (VEGF), whose production is directly stimulated by PTH in bone 23, and bone and endo- thelial cells 24,25. A recent report showed that VEGF signal- ling was involved in mediating PTH-induced vasorelaxation in isolated femoral principal nutrient arteries (PNAs), suggesting that it may also partly mediate the osteoanabolic effect of PTH in vivo 26. Interestingly, this study also revealed a strong dependency of PTH vasodilatory effect in PNAs on nitric oxide synthase (NOS) activity 26. Although these studies suggest that the osteoanabolic action of iPTH may depend on its vascular effects, it is still unknown whether the acute vasodilatory effect of iPTH could be involved in augmenting bone blood ﬂow during the course of treatment. Similarly, whilst the osteoanabolic effects of several NO donor drugs are well documented 27, it is not known whether the transient vasorelaxation induced daily by these drugs constitutes an essential component of their mech- anism of action. The physiological need for increasing blood supply in response to the metabolic demands of increased bone formation and turnover is well-documented 16. In the present study we investigated whether the acute vasodilatory effect of PTH contributes to limb blood ﬂow, bone perfusion and, ultimately, to its osteoanabolic action. To measure bone formation rates, mice were injected s.c. with calcein (15mg/kg), xylenol orange (90mg/kg) and calcein again, respectively before the beginning of treatments, 7days and 2days before the end of the study. Blood perfusion in one hind limb was monitored during the administration of treatments (acute effect) by laser Doppler imaging (LDI) under general anaesthesia. All mice underwent one LDI measurement on the ﬁrst day of dosing and then each mouse was imaged on at least two further occasions, throughout the duration of the study. The acute effect of PTH on intraosseous cortical bone perfusion was assessed by intravenous infusion of a bolus of a ﬂuorescent tracer in two additional groups of naïve mice (n = 4), injected either with human PTH (1–34) (80 μg/kg) or PBS. At the end of the study, mice were sacriﬁced by neck dislocation, left femurs were dissected, ﬁxed in 10% buffered formalin for 48h and stored in 70% ethanol for micro- computed tomography (microCT) imaging and undecalciﬁed histomorphometry. INTRODUCTION Right femurs were dissected and immedi- ately frozen for subsequent mechanical tests. All procedures were performed in accordance with the prin- ciples and guidelines established by the Animals Scientiﬁc Procedures Act and UK Home Ofﬁce regulations (PPL: 70/7444). INTRODUCTION Based on these results and those of a previous study in rats 23, the authors concluded that the combination of vascular remodelling and vasotonic action underpins the increase in bone perfusion induced by iPTH. Both the vascular remod- elling and the osteoanabolic effects of iPTH have been shown to depend on the potent angiogenic factor, vascular endothelial growth factor (VEGF), whose production is directly stimulated by PTH in bone 23, and bone and endo- thelial cells 24,25. A recent report showed that VEGF signal- ling was involved in mediating PTH-induced vasorelaxation in isolated femoral principal nutrient arteries (PNAs), suggesting that it may also partly mediate the osteoanabolic effect of PTH in vivo 26. Interestingly, this study also revealed a strong dependency of PTH vasodilatory effect in PNAs on nitric oxide synthase (NOS) activity 26. Although these studies suggest that the osteoanabolic action of iPTH may depend on its vascular effects, it is still unknown whether the acute vasodilatory effect of iPTH could be involved in augmenting bone blood ﬂow during the course of treatment. Similarly, whilst the osteoanabolic effects of several NO donor drugs are well documented 27, it is not known whether the transient vasorelaxation induced daily by these drugs constitutes an essential component of their mech- anism of action. The physiological need for increasing blood supply in response to the metabolic demands of increased bone formation and turnover is well-documented 16. In the present study we investigated whether the acute vasodilatory effect of PTH contributes to limb blood ﬂow, bone perfusion and, ultimately, to its osteoanabolic action. HCl containing 0.2% bovine serum albumin); (3) a group injected daily subcutaneously (s.c.) with L-NAME alone (30mg/kg/day); and (4) a group co-administered daily with PTH and L-NAME at the doses speciﬁed above. All groups were treated for 28days. The body weight of all mice was measured every day. vascular smooth muscle cells 20,21. PTH has recently been shown to increase blood perfusion of bone marrow in the mouse tibia in a manner that correlated strongly with bone remodelling parameters 22. The same study also showed that iPTH moderately increased intraosseous angiogenesis and blood vessel remodelling within the bone marrow. Based on these results and those of a previous study in rats 23, the authors concluded that the combination of vascular remodelling and vasotonic action underpins the increase in bone perfusion induced by iPTH. Measurements of whole hind limb and intraosseous cortical bone perfusion Two days before hind limb perfusion measurements, the inner side of the hind limb to be imaged was shaved with a depilatory cream under general anaesthesia. At the time of the measurement, mice were anesthetized with 2% isoﬂurane. Body temperature was maintained between at 36 and 37 °C with a feedback-controlled heating pad (507220F, Harvard Apparatus, UK). Hind limb perfusion was measured using a laser Doppler imaging (LDI) ﬂow meter (MoorLDI2 Imager, Moor Instruments Ltd, UK). The instrument measures the distribution of blood ﬂux (combination of velocity and concentration of red blood cells) in arbitrary units for a selected region of interest (19mm×27mm) with a special resolution of 100μm/pixel. The estimated maximum depth of detectable vessels (in soft connective tissue) was estimated to be 2–3mm (laser wavelength=830nm). Drugs and reagents Human PTH (1–34), bovine serum albumin (BSA), Nω-nitro- L-arginine methyl ester hydrochloride (L-NAME), procion red MX-5B, calcein and xylenol orange were purchased from Sigma (UK). Phosphate buffered saline solution (PBS) was purchased from VWR (UK). All drugs were used at a dose required to produce the desired pharmacological effects in mice, according to previous protocols 28,29. INTRODUCTION osteocytes and hypertrophic chondrocytes 8. Further evidence that the effect of PTH on bone is mediated by bone cells is provided from genetic studies where selective activation or disruption of PTH receptor type 1 (PTHR1) signalling in osteocytes modiﬁed the anabolic and catabolic responses induced by PTH in mice 9,10. One key aspect underpinning the complex in vivo role of PTH, which merits consideration when exploring its mode of action, is its pivotal dependence on mode of administration: although intermittent injections of PTH increase bone mass, continuous infusion stimulates osteo- clastic resorption and decreases bone mass11,12. Surprisingly, the direct effect of PTH on primary rodent osteoblasts in vitro is to strongly inhibit osteogenic differentiation and bone formation 13,14. This observation suggests that an additional indirect action of the hormone could contribute to the osteoanabolic effect of PTH on bone in vivo. The cardiovascular action of parathyroid hormone (PTH) was ﬁrst observed in 1925, soon after its discovery as a hypercalcaemic agent 1. Decades later, PTH was shown to decrease arterial blood pressure and increase cardiac output, with marked increases in coeliac, coronary and renal blood ﬂows and corresponding decreases in vascular resistance 2,3. PTH also has an opposing hypertensive action, secondary to its potent hypercalcemic effect 4. Since the discovery of the powerful osteoanabolic effect of intermittently administrated PTH (iPTH)5, research has focused mainly on its skeletal effects; iPTH remains the most potent bone anabolic drug available at present 6. There is good evidence that a key mechanism underlying the osteogenic action of PTH is the suppression of sclerostin production 7, an inhibitor of Wnt signalling that is predominantly produced by In recent years, considerable evidence has emerged in support of the critical role of the vascular supply in the control of bone formation and turnover 15–19. The possibility that some osteogenic effects of iPTH rely on its vascular function has yet, however, to be fully explored. PTHR1 is expressed by many cell types, including endothelial and VASODILATION, PTH AND BONE 53 vascular smooth muscle cells 20,21. PTH has recently been shown to increase blood perfusion of bone marrow in the mouse tibia in a manner that correlated strongly with bone remodelling parameters 22. The same study also showed that iPTH moderately increased intraosseous angiogenesis and blood vessel remodelling within the bone marrow. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. Experimental design For the perfusion measurements, reagents were adminis- tered in the same order as during the chronic study: ﬁrst PBS or L-NAME was injected subcutaneously and then PBS or PTH (1–34) was injected i.p. within 10 min after the ﬁrst injection. Perfusion was recorded continuously for 30 min (starting 10 min before the ﬁrst injection). Male BALB/c mice, aged 10–12weeks, were purchased from Charles River Laboratories, Ins (UK) and maintained under a controlled temperature (21°C) and lighting with 12h/12h light/dark cycle. They received a standard mouse diet and water ad libitum. Mice were randomized into four groups (8 mice/group): (1) a control group receiving daily intraperitone- ally (i.p.) injections of PBS; (2) a group injected daily i.p. with human PTH (1–34) alone (80μg/kg/day, dissolved in 1mM Vascular tracer uptake into the lacuno-canalicular porosity, an established method to measure intraosseous perfusion 30,31, Cell Biochem Funct 2016; 34: 52–62. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. S. GOHIN ET AL. 54 (RM2255, Leica, Germany). To determine the osteoclast surface per bone surface (Oc.S/BS), sections were stained for TRAP (Leucognost SP kit, Merck) and counterstained with Weigert’s haematoxylin. To determine the osteoid surface per bone surface (Ost.S/BS), sections were stained using a modiﬁed Goldner’s trichrome protocol. was assessed by injecting a bolus of procion red solution (0.8% in PBS), equal to 10% of the body weight, into the tail vein of naïve mice. The procion red bolus was given to two groups of restrained, conscious mice (n=4) immediately after i.p. injection of either PTH (1–34) or PBS. Mice were allowed to roam freely in their cages for 5min after the injection, and then sacriﬁced by neck dislocation. Femurs were immediately dissected and ﬁxed in 70% ethanol. Procion red uptake in cortical bone was examined in non-decalciﬁed histological sections of the femoral mid-diaphysis using the protocol described below. Fluorescent labels were visualized using a confocal laser scanning microscope (TCS SP5, Leica, Germany) equipped with 488nm (for calcein) and 561nm (for xynelol orange) lasers. Assessment of microstructure and volumetric bone mineral density by microCT The microstructure and volumetric bone mineral density (vBMD) of femurs were assessed by high-resolution microCT (5μm/pixel, 50kV, 200μA, Skyscan 1172, Belgium). Image reconstruction was performed by NRecon software (Skyscan) and bone parameters were analysed in 2D and 3D using CT. An software (Skyscan). The cortical and trabecular compart- ment within the femoral metaphysis and the vertebral body were determined by manual contouring. Speciﬁcally, the trabecular bone within the secondary spongiosa was analysed in the femoral metaphysis, a volume extending 1mm proxi- mally from the primary spongiosa limit (200 tomograms). In the vertebral body, the entire trabecular content extending between the two cortical plates was contoured and analysed. The threshold values for the microCT analysis of trabecular bone were chosen between 60 and 255, while they were between 100 and 255 for cortical bone. The following stan- dard histomorphometric parameters were evaluated in trabecular bone using a 3D-based analysis software (Ct. An, Skyscan): bone volume fraction (BV/TV, computed as the ratio of bone volume, BV, over tissue volume, TV), tra- becular thickness (Tb.Th), trabecular space (Tb.Sp), trabecular number (Tb.N) and 3D structure model index (SMI), trabecular pattern factor (Tb.Pf), trabecular connectivity density (Conn.D). Procion red uptake in the mid-shaft cortical bone was assessed in the thick sections (70–100μm) by confocal mi- croscopy (excitation laser=561nm, emission ﬁlter=610nm). Brieﬂy, a region of interest (ROI) was manually drawn within the postero-medial quadrant of the mid-diaphysis cortical bone. The average ﬂuorescence intensity of each ROI was normalized by the area of the ROI and was used a measure of the average concentration of the tracer per unit area of cortical bone. Assessment of bone mechanical properties For biomechanical analysis, right femurs were stored in gauze soaked in saline solution at 80 °C and thawed before analysis at room temperature. They were tested until fracture by three-point bending using an Instron testing machine (5866, Norwood, MA, USA). Femurs were loaded at the mid-diaphysis in their antero-posterior direction with a deﬂection rate of 1 μm/s while load was measured using a calibrated 50-N load cell. The loading span (i.e. the distance between the two rods supporting the femur shaft) was ﬁxed to 9mm. Femora were kept moist at all times. Force–deﬂection curves were analysed with a custom programme in Matlab (MathWorks Inc, MA, USA). Geometry parameters, including the mid femur cross-section dimensions were measured from the microCT images of the femur mid-diaphysis. Cortical bone in the mid-diaphysis was evaluated in 50 tomograms located at 50% of the total length of femur. The following parameters were evaluated in cortical bone using Ct.An and Matlab software: tissue volume (TV), bone volume (BV), average cross-sectional tissue area (Tt.Ar), average cross-sectional bone area (Ct.Ar), average cross- sectional cortical thickness (Ct.Th), maximum moment of inertia (Imax), minimum moment of inertia (Imin), polar moment of inertia (Ipol), volumetric bone mineral density (vBMD) and percentage porosity. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Experimental design For cortical bone, double calcein-labels were measured using an Osteometrics image analysis software system, and the following parameters were evaluated: mineralizing surfaces (MS/BS, %), computed as the last calcein labelled surfaces (MS, μm2) normalized to all bone surfaces (BS, μm2); mineral apposition rate (MAR, μm/day), measured as average distance between adjacent pairs of calcein labels divided by the number of days elapsed between the adminis- tration of both labels (26days); bone formation rates (BFR/ BS, μm3/μm2/d), derived by multiplying MS/BS by MAR. For trabecular bone, only mineral apposition rate was eval- uated using a dedicated open source plugin for Image J developed in our laboratory. We measured the distance between the xynelol orange label and the last calcein label (5 days between both labels). All histomorphometric parameters were quantiﬁed using a semi-automated pro- gramme, developed by Dr. van’t Hof 33. Assessment of microstructure and volumetric bone mineral density by microCT Non-decalciﬁed bone histomorphometry After microCT scanning, femurs were dehydrated and embedded in methyl methacrylate using an established cold embedding protocol which preserves tartrate resistant acid phosphatase (TRAP) enzymatic activity 32. Thicker cross sections of the femur mid-diaphysis (70–100μm) were obtained using a slow-speed rotating saw (Isomet, Buehler, USA) while thinner coronal sections of the femur distal metaphysis (8 μm thick) were cut with a microtome Bending stress (σ) was computed from the force according to the equation: σ =FLd/4Imin, where F is the force, L is the loading span, d is the antero-posterior diameter and Imin is the second moment of inertia from the lateromedial axis (πd^4⁄64); bending strain (e) was calculated according to the equation: e=deﬂection×6×d/(L^2); 34. The following set of parameters—regarded as extrinsic mechanical properties Cell Biochem Funct 2016; 34: 52–62. VASODILATION, PTH AND BONE 55 —was directly measured from the force–deﬂection curves: bending stiffness (S, slope of the linear elastic deformation), yield load (Fyield, force limit between the elastic and plastic de- formation), ultimate load (Fult, maximum force sustained) and breaking load (Fbreak, force at which the bone breaks). An additional set of parameters—regarded as intrinsic mechanical properties—was computed using the standard beam theory from the stress–strain curves derived from the force– deﬂection curves 34. These included: bone elastic modulus (E, slope of the linear elastic deformation in the stress- strain curve), yield stress (σyield, stress limit between the elastic and plastic deformation), ultimate stress (σult, maximum stress sustained) and breaking stress (σbreak, stress at which the bone breaks). Finally, total and plastic (yield) work to fracture, as well as their ratio, were calcu- lated as a measure of bone toughness. perfusion which are abolished by co-administration of L-NAME, suggesting that they are mediated by increases in NO production. This iPTH-induced hind limb perfusion might not necessar- ily be translated to changes in intra-cortical bone perfusion; therefore, in order to relate whole hind limb perfusion (mea- sured by LDI) to direct intracortical bone perfusion, we used confocal microscopy to assess uptake of procion red in corti- cal bone of the femoral mid-diaphysis. Our data revealed that PTH signiﬁcantly increases procion red average ﬂuorescence (proportional to its concentration) in femoral cortical bone within 5min of injection (Figure 2 a, b). This suggests that PTH injection induces both acute elevation in blood perfusion and a closely linked increase in tracer perfusion within the femoral cortical bone tissue. Statistical analysis Data are reported as mean±standard deviation (SD). Results were analysed by one-way ANOVA, followed by Dunnett’s post hoc test. Tukey’s post hoc test was used to compare all possible pairings in the four groups. The homoscedasticity requirement (equal variances across groups) for the applicabil- ity of one-way ANOVA was established by the Bartlett’s and the Brown–Forsythe test (which was not signiﬁcant for all sets of parameters tested). Statistical analysis was computed using Prism v.6 software (GraphPad, USA). A two-tailed P value less than 0.05 was regarded as statistically signiﬁcant. Cortical bone geometry and bone quality parameters in the femoral mid-diaphysis were measured by microCT. As expected, signiﬁcantly increased cortical bone volume (+12%) and average thickness (+10%) were observed in mice treated with iPTH compared to vehicle-treated mice (Figure 3 a, b). Moreover, cortical tissue volume was slightly higher in iPTH-treated mice compared with vehicle treated mice, although this was not signiﬁcant (P=0.059). These iPTH- induced changes were abrogated by co-administered L- NAME treatment, which effectively restrained cortical bone geometry to levels evident in control mice. L-NAME treat- ment alone had no effect (Figure 3 a, b). Additional measures of mid-shaft geometry (Table 1) were also signiﬁcantly ele- vated by iPTH compared with control, including Imin (+18%) and Ipol (+16%), and these were similarly reduced to basal control levels by co-administration of L-NAME. MicroCT analysis of bone quality indexes, including bone mineral density (BMD) and cortical porosity, showed that these were unchanged in the experimental groups (Table 1). © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. iPTH induces acute but not chronic increases in both intra-cortical and hind limb perfusion that are abolished by L-NAME co-administration Intermittent PTH increases bone blood ﬂow acutely in BALB/c mice. (a) Representative images of the mouse hind limb perfusion, obtained by laser Doppler imaging (LDI). Images acquired 10 min after administration of vehicle, PTH 1–34 (80μg/kg), L-NAME (30 mg/kg) or the combination PTH + L-NAME. The colours in the heat maps indicate the maximum (red) and the minimum (blue) level of perfusion expressed in arbitrary units representing the total blood ﬂux (combination of velocity and concentration of red blood cells). (b) Time course graph of the average hind limb perfusion monitored by LDI for PTH and L-NAME, alone or in combination. Arrows: 1—Vehicle or L-NAME injection; 2—vehicle or PTH. Values are presented as means ± standard errors of the mean (S.E.M.); n = 10–14 mice/group; P < 0.05 versus Control; ** P < 0.001 versus Control 56 S. GOHIN ET AL. S. GOHIN ET AL. 56 Figure 1. Intermittent PTH increases bone blood ﬂow acutely in BALB/c mice. (a) Representative images of the mouse hind limb perfusion, obtained by laser Doppler imaging (LDI). Images acquired 10 min after administration of vehicle, PTH 1–34 (80μg/kg), L-NAME (30 mg/kg) or the combination PTH + L-NAME. Figure 1. Intermittent PTH increases bone blood ﬂow acutely in BALB/c mice. (a) Representative images of the mouse hind limb perfusion, obtained by laser Doppler imaging (LDI). Images acquired 10 min after administration of vehicle, PTH 1–34 (80μg/kg), L-NAME (30 mg/kg) or the combination PTH + L-NAME. The colours in the heat maps indicate the maximum (red) and the minimum (blue) level of perfusion expressed in arbitrary units representing the total blood ﬂux (combination of velocity and concentration of red blood cells). (b) Time course graph of the average hind limb perfusion monitored by LDI for PTH and L-NAME, alone or in combination. Arrows: 1—Vehicle or L-NAME injection; 2—vehicle or PTH. Values are presented as means ± standard errors of the mean (S.E.M.); n = 10–14 mice/group; P < 0.05 versus Control; ** P < 0.001 versus Control Figure 2. Acute effect of PTH treatment on lacunar–canalicular perfusion. (a) Representative confocal micrographs of non-decalciﬁed sections of the femoral mid-diaphysis showing the uptake of ﬂuorescent tracer in cortical bone 5 min after injection of vehicle (control) and PTH 1–34 (80 μg/kg). Scale bar = 100 μm. (b) Average tracer ﬂuorescence within a region of interest manually drawn within the postero-medial quadrant of the mid-diaphysis cortical bone. iPTH induces acute but not chronic increases in both intra-cortical and hind limb perfusion that are abolished by L-NAME co-administration iPTH induces acute but not chronic increases in both intra-cortical and hind limb perfusion that are abolished by L-NAME co-administration The mean body weight of mice did not change during the 4- week treatment, and was similar among the four treatment groups (vehicle=25.3±0.4g, iPTH=25.5±0.5g, iPTH+L- NAME=25.4±0.3g, L-NAME=25±0.8g, mean±SD, n=10–14). Acute changes in hind limb perfusion were measur- able by LDI on the ﬁrst day of dosing, with PTH producing signiﬁcant elevation (over 30%; P< 0.001) in perfusion within 10–15min compared to the vehicle-treated controls, which was sustained for ≥30min (Figure 1 a, b). Notably, this acute response was reproduced in randomly selected mice after any of the daily doses of PTH throughout the 4-week period (ap- proximately two random measurements in each mouse, besides the initial one). While L-NAME alone had little decreases compared to the control, its co-administration completely blocked the PTH-related increases in vasorelaxation (Figure 1 a, b). However, at the end of the 4-week treatment, no measur- able chronic changes in hind limb blood perfusion were evident 24h after the last dosing in any groups compared to vehicle controls (vehicle=641±112, PTH=775±215, iPTH+L- NAME=734±117, L-NAME=678±122, perfusion expressed in arbitrary units, mean±SD, n=10–14). These data indicate that iPTH induces acute, but not chronic increases in hind limb Because of the differences in blood supply between the bone periosteal surfaces (normally well vascularized) and the endosteal surfaces (poorly supplied) 35, we investigated whether the modulation of vascular function induced by the different treatments might produce differing levels of osteoblastic activation in each of these two envelops. Quan- tifying bone formation rate in the whole bone and separately in periosteal and endosteal bone compartments revealed that the whole bone response more closely matched the perios- teal response (Figure 4 a–c). Intermittent PTH treatment signiﬁcantly increased mineral apposition rate (MAR) and bone formation rate (BFR/BS) on the periosteal surfaces compared to vehicle-treated controls (+92% and +130%, respectively) (Figure 4 a–c). These iPTH-induced increases in MAR and BFR/BS were signiﬁcantly suppressed by co-administration of L-NAME (by 40%). These levels were nevertheless still signiﬁcantly elevated when com- pared to either vehicle (+39%) or L-NAME-treated groups (+48%), despite the lack of any modiﬁcation in MAR or BFR/BS by L-NAME treatment alone (Figure 4 a–c). © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. Figure 1. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. mistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. iPTH induces acute but not chronic increases in both intra-cortical and hind limb perfusion that are abolished by L-NAME co-administration Values are mean ± SD, n = 4 mice/group; * P < 0.05 versus Control Figure 2. Acute effect of PTH treatment on lacunar–canalicular perfusion. (a) Representative confocal micrographs of non-decalciﬁed sections of the femoral mid-diaphysis showing the uptake of ﬂuorescent tracer in cortical bone 5 min after injection of vehicle (control) and PTH 1–34 (80 μg/kg). Scale bar = 100 μm. (b) Average tracer ﬂuorescence within a region of interest manually drawn within the postero-medial quadrant of the mid-diaphysis cortical bone. Values are mean ± SD, n = 4 mice/group; * P < 0.05 versus Control © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. 57 VASODILATION, PTH AND BONE Figure 3. L-NAME decreases the anabolic action of intermittent PTH on cortical bone volume. (a) Three-dimensional (3D) renderings of 2-mm segments of the femur mid-diaphysis imaged by microCT at the end of 28-day treatment with vehicle (control), iPTH and L-NAME, alone or in combination. (b) Structural parameters computed from the microCT scans: mid-diaphysis cortical bone volume (Ct.BV) and cortical bone thickness (Ct.Th). Values are mean ± SD, n = 7–8 mice/group; * P < 0.05 versus Control; ** P < 0.01 versus Control , re 3. L-NAME decreases the anabolic action of intermittent PTH on cortical bone volume. (a) Three-dimensional (3D) renderings of 2-mm segments of emur mid-diaphysis imaged by microCT at the end of 28-day treatment with vehicle (control), iPTH and L-NAME, alone or in combination. (b) Structural meters computed from the microCT scans: mid-diaphysis cortical bone volume (Ct.BV) and cortical bone thickness (Ct.Th). Values are mean ± SD, n = 7–8 /group; * P < 0.05 versus Control; ** P < 0.01 versus Control Figure 3. L-NAME decreases the anabolic action of intermittent PTH on cortical bone volume. (a) Three-dimensional (3D) renderings of 2-mm segments of the femur mid-diaphysis imaged by microCT at the end of 28-day treatment with vehicle (control), iPTH and L-NAME, alone or in combination. (b) Structural parameters computed from the microCT scans: mid-diaphysis cortical bone volume (Ct.BV) and cortical bone thickness (Ct.Th). iPTH induces acute but not chronic increases in both intra-cortical and hind limb perfusion that are abolished by L-NAME co-administration Values are mean ± SD, n = 7–8 mice/group; * P < 0.05 versus Control; ** P < 0.01 versus Control Conversely, BFR/BS was elevated on the endosteal surfaces in the iPTH (+120%) and L-NAME alone (+142%) groups compared to control, but not in the iPTH+L-NAME group (Figure 4e), where it was signiﬁcantly reduced in comparison with the iPTH group (66%; Figure 4e). and iPTH was evident in this highly vascularized compart- ment. In contrast to the anabolic effect of PTH in cortical bone, microCT analysis of the femoral metaphysis showed that iPTH treatment in BALB/c mice induced a decrease in trabecular number (Tb.N; 21%) and augmented trabecular thickness (+27%) without any modiﬁcation in total trabecular bone volume fraction (BV/TV; Figure 5 a–d). L-NAME treat- ment alone did not signiﬁcantly affect any trabecular bone parameter, compared to controls (Figure 5 a–e) but its co- administration with iPTH resulted in decreased BV/TV (15%), greater reduction in Tb.N (32%) and further en- hanced Tb.Th (+24%) compared to vehicle-treated controls (Figure 5 a–d). iPTH also elevated the structural model index (SMI; +20%), which was further increased by L-NAME co- administration (Figure 5 e). Additional measures of tra- becular bone parameters (Table 1) were also signiﬁcantly g p g In summary, L-NAME co-administration either partially or completed inhibited the increase in the rate of bone for- mation induced by iPTH, and this inhibition was dependent upon the cortical bone envelope (periosteal or endosteal) of the femur midshaft. chemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52– Co-administration of L-NAME with PTH reduces trabecular bone volume We then analysed trabecular bone in the femoral metaphysis in order to determine whether similar sensitivity to L-NAME Table 1. MicroCT analysis of femoral mid-diaphysis cortical bone and distal metaphysis trabecular bone. Femora were dissected at the end of the 28-day dosing and analysed by micro-CT. Bone parameters are shown for the four treatment groups (n = 7–8 mice/group): Vehicle (control), iPTH, iPTH + L- NAME and L-NAME alone Table 1. MicroCT analysis of femoral mid-diaphysis cortical bone and distal metaphysis trabecular bone. Femora were dissected at the end of the 28-day dosing and analysed by micro-CT. Bone parameters are shown for the four treatment groups (n = 7–8 mice/group): Vehicle (control), iPTH, iPTH + L- NAME d L NAME l NAME and L-NAME alone Cortical bone parameters Control iPTH iPTH + L-NAME L-NAME Tissue volume, TV (mm3) 0.028 ± 0.001 0.030 ± 0.001 0.028 ± 0.001 0.029 ± 0.001 Cross-sectional tissue area, Tt.Ar (mm2) 0.105 ± 0.009 0.112 ± 0.004 0.107 ± 0.004 0.107 ± 0.007 Cross-sectional bone area, Ct.Ar (mm2) 0.064 ± 0.005 0.072 ± 0.003a 0.068 ± 0.002 0.066 ± 0.005 Imax (mm4) 0.23 ± 0.04 0.27 ± 0.03 0.24 ± 0.02 0.24 ± 0.03 Imin (mm4) 0.10 ± 0.02 0.12 ± 0.01a 0.11 ± 0.01 0.11 ± 0.01 Ipol (mm4) 0.0016 ± 0.0003 0.0019 ± 0.0001a 0.0017 ± 0.0001 0.0017 ± 0.0002 vBMD (g/cm3) 1.269 ± 0.017 1.278 ± 0.010 1.276 ± 0.013 1.279 ± 0.010 Cortical porosity (%) 0.04 ± 0.02 0.04 ± 0.05 0.02 ± 0.03 0.05 ± 0.05 Trabecular bone parameters Control iPTH iPTH + L-NAME L-NAME Trabecular space, Tb.Sp (mm) 0.18 ± 0.01b 0.21 ± 0.01abc 0.23 ± 0.01ac 0.19 ± 0.01b Trabecular pattern factor, Tb.Pf (mm1) 20.9 ± 2.5b 21.3 ± 1.8b 24.5 ± 2.2a 22.7 ± 1.9 Connectivity density, Conn.D (mm-3) 290.4 ± 35.5b 183.4 ± 25.0abc 144.8 ± 18.6ac 273.7 ± 37.7b Values are mean ± SD; aP < 0.05 versus control; bP < 0.05 versus iPTH + L-NAME; cP < 0.05 versus L-NAME. Cortical bone parameters © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. Figure 4. L-NAME reduces iPTH-induced increases in cortical bone formation rate. (a) Representative confocal micrographs of non-decalciﬁed sections of the femoral mid-diaphysis, showing the calcein double labels (26 days between both labels) marking bone apposition during the 26 days of dosing with vehicle (control), iPTH and L-NAME, alone or in combination. Scale bar = 50 μm. (b) Bar graphs of the average periosteal mineral apposition rate (MAR) and (c) periosteal bone formation rate (BFR/BS) for the four groups. (d) Bar graphs of the average endosteal mineral apposition rate (MAR) and (e) endosteal bone formation rate (BFR/ BS) for the four groups. Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH + L-NAME, d P < 0.05 versus L-NAME 58 S. GOHIN ET AL. 58 S. GOHIN ET AL. Figure 4. L-NAME reduces iPTH-induced increases in cortical bone formation rate. (a) Representative confocal micrographs of non-decalciﬁed sections of the femoral mid-diaphysis, showing the calcein double labels (26 days between both labels) marking bone apposition during the 26 days of dosing with vehicle (control), iPTH and L-NAME, alone or in combination. Scale bar = 50 μm. (b) Bar graphs of the average periosteal mineral apposition rate (MAR) and (c) periosteal bone formation rate (BFR/BS) for the four groups. (d) Bar graphs of the average endosteal mineral apposition rate (MAR) and (e) endosteal bone formation rate (BFR/ BS) for the four groups. Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH + L-NAME, d P < 0.05 versus L-NAME modiﬁed by iPTH and by L-NAME. Tb.Sp was increased by iPTH compared to control (+16%) and further increased by L-NAME co-administration. iPTH co-administrated with L-NAME increased Tb.Pf compared to controls (+17%) and iPTH alone (+15%). Conn.Dn was reduced by PTH treatment compared to control (37%), and greater reductions were observed with L-NAME co-administration (50%; Table 1). Similar microstructural changes were observed in the trabecular bone of L5 vertebrae (data not shown), indicating a compartment-speciﬁc (trabecular versus cortical bone) but not a site-speciﬁc (appendicular versus axial skeleton) sensi- tivity to iPTH and L-NAME treatment. Cortical bone parameters bone remodelling by altering the thickness and number of tra- beculae without inducing a ﬁnal increase in trabecular bone volume fraction compared to control. Co-administration of L-NAME did not reduce iPTH-induced trabecular thickening but did further decrease the number of trabeculae which resulted in a ﬁnal reduction of the trabecular bone volume fraction. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. L-NAME co-administration does not modify iPTH-enhanced bone mechanical strength In order to explore whether these changes in perfusion and associated modiﬁcations in bone formation and geometry engender a structural advantage, we measured whole-bone mechanical properties by three-point bending. This showed that iPTH signiﬁcantly increased several extrinsic bone material properties compared to control (Table 2), including yield load (+17%) and ultimate load (+15%), which measure whole femur strength. Surprisingly, this iPTH-induced enhancement in bone strength was not reduced by co- administration of L-NAME. Moreover, iPTH and L-NAME in combination exerted unexpected effects on some of the intrinsic material properties of bone, including Young’s modulus, yield stress and ultimate stress, which were signif- icantly increased compared with bones from both control and also with iPTH-treated groups (Table 2). Conversely, iPTH did not affect any of the intrinsic material properties y In order to determine osteoblast and osteoclast activities, we also examined histomorphometric parameters of trabecular bone in the distal femur metaphysis. The percentage of resorp- tion surfaces was not signiﬁcantly different among groups (Figure 6 a–c). The proportion of osteoid surfaces was signif- icantly greater in both iPTH (+167%) and PTH+L-NAME treated groups (+133%) than control (Figure 6 d–f). Although L-NAME co-administration appeared to reduce the elevation of osteoid surfaces compared to iPTH alone, this did not reach statistical signiﬁcance. L-NAME treatment alone did not affect any of the cellular parameters. Despite the signiﬁcant increase in osteoid surfaces induced by iPTH, the bone appo- sition rate (MAR) in trabecular bone was not affected by any of the treatments (Figure 6 g–i). Thus, iPTH appear to regulate Cell Biochem Funct 2016; 34: 52–62. 59 VASODILATION, PTH AND BONE Figure 5. Effects of PTH and L-NAME on trabecular bone architecture in femurs. (a) Three-dimensional (3D) rendering of microCT scans of trabecular bone in secondary spongiosa of the distal femur in mice treated with vehicle (control), iPTH and L-NAME, alone or in combination for 28 days. The bar graphs show some of the parameters obtained from the 3D analysis of the trabecular architecture, including (b) the bone volume fraction (BV/TV), (c) the number of trabeculae (Tb.N), (d) the trabecular thickness (Tb.Th) and (e) the structure model index (SMI). Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH+L-NAME, d P < 0.05 versus L-NAME Figure 5. Effects of PTH and L-NAME on trabecular bone architecture in femurs. L-NAME co-administration does not modify iPTH-enhanced bone mechanical strength (a) Three-dimensional (3D) rendering of microCT scans of trabecular bone in secondary spongiosa of the distal femur in mice treated with vehicle (control), iPTH and L-NAME, alone or in combination for 28 days. The bar graphs show some of the parameters obtained from the 3D analysis of the trabecular architecture, including (b) the bone volume fraction (BV/TV), (c) the number of trabeculae (Tb.N), (d) the trabecular thickness (Tb.Th) and (e) the structure model index (SMI). Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH+L-NAME, d P < 0.05 versus L-NAME ﬁnding that PTH-induced acute vasodilation in rat femoral principal nutrient arteries in vitro was almost totally inhibited by L-NAME 26, and suggest a signiﬁcant role for NO-mediated arterial vasorelaxation in the osteoanabolic actions of PTH. Bone cells also signal through NO 36,37, and thus speciﬁc interactions between PTH and NOS in bone cells cannot be ruled out, although in vitro evidence suggests that NO production is not signiﬁcantly altered by PTH in human and murine osteoblasts 38,39. Our results showed that L-NAME treatment alone did not signiﬁcantly affect perfusion or bone mass, suggesting that NOS function may not play an important role in bone homeostasis in unchallenged adult mice. This is consistent with another report showing that L-NAME administration for 18 days did not signiﬁcantly change trabecular bone formation rate in rats 40. of cortical bone compared to control. Finally, L-NAME alone did not affect either extrinsic or intrinsic material properties of bone compared with control. These data indi- cate that the combined effect of opposing modiﬁcations in structural versus material properties induced by iPTH versus those in the presence of PTH+ L-NAME caused the two treatment groups to achieve a similar improvement in femur mechanical strength compared with control. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. DISCUSSION Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH + L-NAME, d P < 0.05 versus L-NAME 60 S. GOHIN ET AL. 0 S. GOHIN ET AL. 60 Figure 6. Effects of PTH and L-NAME on trabecular bone turnover. (a) Representative brightﬁeld micrographs of non-decalciﬁed sections of secondary spongiosa in the distal femur stained for TRAP (a) or Goldner’s trichrome (d) with respective magniﬁed regions demonstrating (b) TRAP-positive osteoclasts (arrow) or (e) positively stained osteoid (arrow) lining trabeculae (Tb). (g) Representative confocal micrographs of a non-decalciﬁed section of the secondary spongiosa of the distal femur, showing the ﬂuorescent labels (calcein, green and xylenol orange, red, 5 days between both labels) marking bone apposition during the last 5 days of the 28 day dosing experiment. Scale: bars = 500 μm in top row images and bars = 50 μm in second row images. Bar graphs (C, F, I) summarize the different histomorphometric parameters in trabecular bone at the end of the 28-day treatment with vehicle (control), PTH 1–34 (80μg/kg/d) and L-NAME (30 mg/kg/d), alone or in combination including: (c) percentage of osteoclast surface per bone surface, (f) percentage of osteoid surface per bone surface and (i) mineral apposition rate (MAR). Values are mean ± SD, n = 7–8 mice/group; a P < 0.05 versus control; b P < 0.05 versus iPTH; c P < 0.05 versus iPTH + L-NAME, d P < 0.05 versus L-NAME Table 2. Mechanical properties of the femur measured by the three-point bending test. DISCUSSION The results of the present study show that PTH (1–34) causes acute stimulation of limb blood ﬂow in BALB/c mice, and that this response is required for PTH to achieve its anabolic effect on cortical bone. PTH treatment however did not increase trabecular bone volume in BALB/c mice. Our data also shows that the vasomotor effect of PTH is not diminished by repeated daily dosing. This observation suggests that the transient surge in blood perfusion follow- ing daily PTH administration may be integral to the mode of action of intermittent PTH on bone. Because our laser Doppler imaging measurements only incorporated perfusion near the surface of mouse hindlimbs (to a maximum penetration depth in skin/muscle of about 2 mm), we tested whether acute vasodilation also induced an increase in intraosseous cortical perfusion. The reference method to measure bone blood ﬂow is the intravascular injection of labelled microspheres41. However, a lot of studies demonstrated a positive correlation between the standardized measure of blood ﬂow with labelled microspheres and blood perfusion measured by the Laser Doppler Imaging42, We demonstrate that blockade of nitric oxide synthase activity by the inhibitor L-NAME greatly reduced PTH- stimulated blood ﬂow and also restricted the anabolic effect of PTH on cortical bone. These results ﬁt with the earlier Cell Biochem Funct 2016; 34: 52–62. Figure 6. Effects of PTH and L-NAME on trabecular bone turnover. (a) Representative brightﬁeld micrographs of non-decalciﬁed sections of secondary spongiosa in the distal femur stained for TRAP (a) or Goldner’s trichrome (d) with respective magniﬁed regions demonstrating (b) TRAP-positive osteoclasts (arrow) or (e) positively stained osteoid (arrow) lining trabeculae (Tb). (g) Representative confocal micrographs of a non-decalciﬁed section of the secondary spongiosa of the distal femur, showing the ﬂuorescent labels (calcein, green and xylenol orange, red, 5 days between both labels) marking bone apposition during the last 5 days of the 28 day dosing experiment. Scale: bars = 500 μm in top row images and bars = 50 μm in second row images. Bar graphs (C, F, I) summarize the different histomorphometric parameters in trabecular bone at the end of the 28-day treatment with vehicle (control), PTH 1–34 (80μg/kg/d) and L-NAME (30 mg/kg/d), alone or in combination including: (c) percentage of osteoclast surface per bone surface, (f) percentage of osteoid surface per bone surface and (i) mineral apposition rate (MAR). DISCUSSION Bone mechanical properties were measured in femurs dissected at the end of the 28-day dosing for the four treatment groups (n = 7-8 mice/group): Vehicle (control), iPTH, iPTH + L-NAME and L-NAME alone Group Control iPTH iPTH + L-NAME L-NAME Diameter femurs (mm) 1.16 ± 0.06 1.22 ± 0.04 1.18 ± 0.03 1.19 ± 0.05 Yield load (N) 12.0 ± 1.3b 14.1 ± 1.0a 14.6 ± 1.5ac 12.5 ± 1.3b Ultimate load (N) 13.5 ± 1.0b 15.5 ± 0.7a 16.6 ± 1.1ac 14.9 ± 1.7b Breaking load (N) 10 ± 2 12 ± 2 11 ± 2 10 ± 4 Yield extension (mm) 0.18 ± 0.02 0.19 ± 0.03 0.18 ± 0.01 0.17 ± 0.03 Ultimate extension (mm) 0.26 ± 0.05 0.25 ± 0.03 0.24 ± 0.03 0.24 ± 0.05 Breaking extension (mm) 0.5 ± 0.1 0.4 ± 0.2 0.5 ± 0.1 0.4 ± 0.1 Stiffness (N/mm) 70 ± 6 80 ± 7 87 ± 10a 79 ± 11 Elastic modulus (MPa) 4763 ± 945 4530 ± 654b 5580 ± 516 5037 ± 363 Yield stress (MPa) 69 ± 9b 72 ± 6 83 ± 10ac 71 ± 8b Ultimate stress (MPa) 78 ± 10b 79 ± 5b 93 ± 6a 84 ± 8 Breaking stress (MPa) 58 ± 9 61 ± 13 60 ± 7 57 ± 22 Yield strain 0.015 ± 0.002 0.017 ± 0.003 0.016 ± 0.001 0.015 ± 0.002 Ultimate strain 0.022 ± 0.005 0.023 ± 0.002 0.021 ± 0.003 0.021 ± 0.003 Breaking strain 0.04 ± 0.01 0.04 ± 0.02 0.04 ± 0.01 0.04 ± 0.01 Work to fracture (mJ) 4.3 ± 1.3 4.3 ± 2.2 5.1 ± 0.7 4.0 ± 1.7 Work yield to fracture (mJ) 3.3 ± 1.3 3.0 ± 2.2 3.8 ± 0.7 3.0 ± 1.7 Values are mean ± SD; aP < 0.05 versus control; bP < 0.05 versus iPTH + L-NAME; cP < 0.05 versus L-NAME. Table 2. Mechanical properties of the femur measured by the three-point bending test. Bone mechanical properties were measured in femurs dissected at the end of the 28-day dosing for the four treatment groups (n = 7-8 mice/group): Vehicle (control), iPTH, iPTH + L-NAME and L-NAME alone Table 2. Mechanical properties of the femur measured by the three-point bending test. Cell Biochem Funct 2016; 34: 52–62. chemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52– 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell B DISCUSSION However, because of technical limitations, we were not able to measure cortical perfusion at the end of our study, and thus cannot exclude the possibility that iPTH might have a chronic effect on this parameter. Our ﬁnding that the anabolic effect of iPTH on cortical bone was not completely blocked by L-NAME suggests that other bone and/or vascular responses are triggered by the activation of PTHR1 receptor independently of NOS activ- ity. One possibility is the involvement of VEGF signalling, because a VEGF-blocking antibody has been reported to abolish the anabolic effect of PTH and impair bone blood vessel remodelling in vivo 23, as well as reducing vasodila- tion in vitro 26. Alternatively, and irrespectively of the changes in perfusion, PTH may act on osteocytes to decrease production of the Wnt signalling and bone forma- tion inhibitor, sclerostin 7. Our observation that L-NAME inhibition of the osteoanabolic effect (MAR and BFR/BS) of iPTH on cortical bone was more prominent on endosteal surfaces ﬁts the view that the vascular effect of PTH is particularly marked in more hypoxic areas of bone such as the endosteum and less essential at more oxygen-replete periosteal surfaces 35. This differential stimulation of bone formation in response to the different treatments on cortical bone external and internal envelops suggests a complex remodelling of the mineral surfaces of these envelops, which may contribute to changes in the material properties and, ultimately, the mechanical strength of bone. In conclusion, by demonstrating that the acute vasorelax- ation induced by PTH modulates its anabolic effect on cortical bone in BALB/c mice, the current study provides new evidence for a novel mode of action for iPTH in bone, through an indirect, transient vasomotor stimulation which intermittently increases bone perfusion. Our simplest explanation of this action is that the daily boost of oxygen and nutrients induced by iPTH in bone might be responsible for accelerating bone cell metabolism and, ultimately, increasing bone formation. In addition, changes in intraosseous perfusion are bound to alter gradients for hypoxia in the bone marrow with the potential of activating quiescent osteoblasts and/or mobilizing osteogenic progenitor cells 50. Our results support the notion that daily pharmacological stimulation of vasodilation could either boost the effect of osteoanabolic drugs or by itself increase bone formation and bone mass. Notwithstanding its inhibition of PTH-induced cortical osteogenesis, L-NAME co-administration did not affect the mechanical strength of the femurs. DISCUSSION Bone mechanical properties were measured in femurs dissected at the end of the 28-day dosing for the four treatment groups (n = 7-8 mice/group): Vehicle (control), iPTH, iPTH + L-NAME and L-NAME alone © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. VASODILATION, PTH AND BONE 61 given that several reports have documented a strong increase in trabecular bone mass in C57Bl/6 mice administered similar doses of iPTH for 15 to 28days 22,45,46. It was however shown that the responsiveness to anabolic actions of PTH is mouse strains dependent, 47 and it is likely that our results may reﬂect responses that are speciﬁc for the BALB/c mouse strain, consis- tent with a previous report 48. It is also possible that our lack of effect of iPTH on trabecular BV/TV may be related to the age of the animals studied, because iPTH has been shown to increase metaphyseal trabecular bone in younger BALB/c mice 49. The lack of BV/TV elevation by iPTH in our study did not appear to reﬂect the complete absence of an effect but rather an homeostatic response where the increase in trabecular thickness (Tb.Th) was compensated by a decrease in the number of trabeculae (Tb.N). Moreover, the proportion of trabecular surfaces covered by osteoid was increased by iPTH and unaffected by co-treatment with L-NAME, suggesting that iPTH augmented trabecular bone matrix formation independently of NOS activity. This increase was not matched, however, by an accrual in mineralized bone matrix. Additionally, the reduc- tion in Tb.N in the iPTH treated mice does not ﬁt with the lack of change in osteoclast surfaces. The reasons for the lack of concordance of histomorphometric and microCT parameters in trabecular bone are unclear. demonstrating that the whole hindlimb blood ﬂow is a good in- dicator of bone blood ﬂow. We found that daily PTH treatment increased cortical bone perfusion (assessed by procion red staining), consistent with previous studies reporting a PTH- induced elevation of radioactive microspheres uptake in the tibiae of rats 43. Our results are also consistent with the recent ﬁndings of Roche et al. 22, who showed that tibial perfusion and blood vessel area increased in response to daily PTH injec- tion. In contrast, though, we did not observe any chronic changes in mouse hind limb perfusion, measured by laser Doppler, following prolonged daily administration of PTH. DISCUSSION The reasons for this are not completely clear but it could be that the reduction in PTH-induced cortical osteogenesis by L-NAME was com- pensated by other changes in the intrinsic mechanical prop- erties of the bones in order to achieve the same overall mechanical strength 44. Indeed, we observed that L-NAME treatment improved the intrinsic mechanical properties of the cortical bone compared to control, while PTH alone had no effect. However, we were unable to detect any changes in either BMD or cortical porosity by microCT in PTH and L- NAME-treated groups. Alternatively, it is possible that any changes in mechanical strength are below the powers of detec- tion of this methodology. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. ACKNOWLEDGEMENTS We thank Mr Yitong Shang and Ms Patricia Das Neves Borges (from the Department of Bioengineering, Imperial College London, London, UK) for their collaboration with the develop- ment of the ImageJ plug-in for the analysis of bone ﬂuorescent labels. This work was funded by Arthritis Research UK. CONFLICT OF INTEREST None of the authors have a conﬂict of interest. None of the authors have a conﬂict of interest. We did not observe any effect of iPTH on the trabecular bone volume fraction (BV/TV) in the distal femoral metaphysis. We noted a similar lack of effect of iPTH in the trabecular bone of the L5 vertebra (data not shown), suggesting that this was not a site-speciﬁc phenomenon. This lack of effect was surprising, 1. Collip JB, Clark EP. Further studies on the physiological action of parathy- roid hormone. J Biol Chem 1925; 64: 485–507. REFERENCES Ralston SH, Todd D, Helfrich M, Benjamin N, Grabowski PS. Human osteoblast-like cells produce nitric oxide and express inducible nitric oxide synthase. Endocrinology 1994; 135(1): 330–336. 17. Colleran PN, Wilkerson MK, Bloomﬁeld SA, Suva LJ, Turner RT, Delp MD. Alterations in skeletal perfusion with simulated microgravity: a possible mechanism for bone remodeling. J Appl Physiol 1985 2000; 89(3): 1046–1054. 40. Turner CH, Owan I, Jacob DS, McClintock R, Peacock M. Effects of nitric oxide synthase inhibitors on bone formation in rats. Bone 1997; 21(6): 487–490. y ( ) 18. Prisby RD, Dominguez JM, 2nd, Muller-Delp J, Allen MR, Delp MD. Aging and estrogen status: a possible endothelium-dependent vascular coupling mechanism in bone remodeling. PLoS One 2012; 7(11): e48564. 41. Anetzberger H, Thein E, Becker M, Zwissler B, Messmer K. Microspheres accurately predict regional bone blood ﬂow. Clin Orthop Relat Res 2004; 424: 253–265. 19. Prisby RD, Ramsey MW, Behnke BJ, et al. Aging reduces skeletal blood ﬂow, endothelium-dependent vasodilation, and NO bioavailability in rats. J Bone Miner Res 2007; 22(8): 1280–1288. 42. Shymkiw RC, Zernicke RF, Forrester KR, Bray RC. Evaluation of laser-Doppler perfusion imaging for measurement of blood ﬂow in cortical bone. J Appl Physiol 1985 2001; 90(4): 1314–1318. pp y 43. Kapitola J, Zak J. Effect of parathormone on bone blood ﬂow in rats— possible role of NO. Sb Lek 2003; 104(2): 133–137. 20. Jiang B, Morimoto S, Yang J, Niinoabu T, Fukuo K, Ogihara T. Expression of parathyroid hormone/parathyroid hormone-related pro- tein receptor in vascular endothelial cells. J Cardiovasc Pharmacol 1998; 31(Suppl 1): S142–S144. 44. Turner CH. Biomechanics of bone: determinants of skeletal fragility and bone quality. Osteoporos Int 2002; 13(2): 97–104. 21. Nickols GA, Metz MA, Cline WH, Jr. Endothelium-independent linkage of parathyroid hormone receptors of rat vascular tissue with increased adenosine 3′,5′-monophosphate and relaxation of vascular smooth muscle. Endocrinology 1986; 119(1): 349–356. 45. Iida-Klein A, Zhou H, Lu SS, et al. Anabolic action of parathyroid hor- mone is skeletal site speciﬁc at the tissue and cellular levels in mice. J Bone Miner Res 2002; 17(5): 808–816. muscle. Endocrinology 1986; 119(1): 349–356. 46. Sugiyama T, Saxon LK, Zaman G, et al. Mechanical loading enhances the anabolic effects of intermittent parathyroid hormone (1–34) on trabecular and cortical bone in mice. Bone 2008; 43(2): 238–248. 22. Roche B, Vanden-Bossche A, Malaval L, et al. REFERENCES Aging Cell 2010; 9(2): 147–161. 9. O’Brien CA, Plotkin LI, Galli C, et al. Control of bone mass and remodeling by PTH receptor signaling in osteocytes. PLoS One 2008; 3(8): e2942. g g g 32. Chappard D, Palle S, Alexandre C, Vico L, Riffat G. Bone embedding in pure methyl methacrylate at low temperature preserves enzyme activities. Acta Histochem 1987; 81(2): 183–190. 10. Saini V, Marengi DA, Barry KJ, et al. Parathyroid hormone (PTH)/ PTH-related peptide type 1 receptor (PPR) signaling in osteocytes regulates anabolic and catabolic skeletal responses to PTH. J Biol Chem 2013; 288(28): 20122–20134. ( ) 33. Van’t Hof RJ, Landao-Basonga E. Semi-automated histomorphometry of bone resorption parameters. Bone 2011; 48: S135–S136. 11. Poole KE, Reeve J. Parathyroid hormone—a bone anabolic and catabolic agent. Curr Opin Pharmacol 2005; 5(6): 612–617. 34. Turner CH, Burr DB. Basic biomechanical measurements of bone: a tutorial. Bone 1993; 14(4): 595–608. 12. Robling AG, Kedlaya R, Ellis SN, et al. Anabolic and catabolic regimens of human parathyroid hormone 1–34 elicit bone- and envelope-speciﬁc attenuation of skeletal effects in Sost-deﬁcient mice. Endocrinology 2011; 152(8): 2963–2975. 35. Spencer JA, Ferraro F, Roussakis E, et al. Direct measurement of local oxygen concentration in the bone marrow of live animals. Nature 2014; 508(7495): 269–273. 36. van’t Hof RJ, Ralston SH. Nitric oxide and bone. Immunology 2001; 103(3): 255–261. 13. Bellows CG, Ishida H, Aubin JE, Heersche JN. Parathyroid hormone reversibly suppresses the differentiation of osteoprogenitor cells into functional osteoblasts. Endocrinology 1990; 127(6): 3111–3116. 37. Zaman G, Pitsillides AA, Rawlinson SC, et al. Mechanical strain stim- ulates nitric oxide production by rapid activation of endothelial nitric oxide synthase in osteocytes. J Bone Miner Res 1999; 14(7): 1123–1131. 14. Wang BL, Dai CL, Quan JX, et al. Parathyroid hormone regulates osterix and Runx2 mRNA expression predominantly through protein kinase A signaling in osteoblast-like cells. J Endocrinol Invest 2006; 29(2): 101–108. y y 38. Bakker AD, Joldersma M, Klein-Nulend J, Burger EH. Interactive effects of PTH and mechanical stress on nitric oxide and PGE2 produc- tion by primary mouse osteoblastic cells. Am J Physiol Endocrinol Metab 2003; 285(3): E608–E613. 15. Lafage-Proust MH, Roche B, Langer M, et al. Assessment of bone vascularization and its role in bone remodeling. Bonekey Rep 2015; 4: 662. 16. Marenzana M, Arnett TR. The key role of the blood supply to bone. Bone Res 2013; 3: 203–215. 39. REFERENCES 1. Collip JB, Clark EP. Further studies on the physiological action of parathy- roid hormone. J Biol Chem 1925; 64: 485–507. 6 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. Cell Biochem Funct 2016; 34: 52–62. S. GOHIN ET AL. 62 2. Charbon GA. A rapid and selective vasodialtor effect of parathyroid hormone. Eur J Pharmacol 1968; 3(3): 275–278. 2. Charbon GA. A rapid and selective vasodialtor effect of parathyroid hormone. Eur J Pharmacol 1968; 3(3): 275–278. 26. Prisby R, Menezes T, Campbell J. Vasodilation to PTH (1–84) in bone arteries is dependent upon the vascular endothelium and is mediated partially via VEGF signaling. Bone 2013; 54(1): 68–75. 3. Nickols GA, Metz MA, Cline WH, Jr. Vasodilation of the rat mesenteric vascu- lature by parathyroid hormone. J Pharmacol Exp Ther 1986; 236(2): 419–423. 27. Jamal SA, Reid LS, Hamilton CJ. The effects of organic nitrates on osteoporosis: a systematic review. Osteoporos Int 2013; 24(3): 763–770. y p y p 4. Bukoski RD, Kremer D. Calcium-regulating hormones in hypertension: vascular actions. Am J Clin Nutr 1991; 54(1 Suppl): 220S–226S. 28. Budziszewska B, Leskiewicz M, Jaworska-Feil L, Lason W. The effect of N-nitro-L-arginine methyl ester on cocaine-induced hormonal changes in mice. Exp Clin Endocrinol Diabetes 1998; 106(4): 340–345. 5. Potts JT, Gardella TJ. Progress, paradox, and potential: parathyroid hormone research over ﬁve decades. Ann N Y Acad Sci 2007; 1117: 196–208. 29. Jeyabalan J, Shah M, Viollet B, et al. Mice lacking AMP-activated protein kinase alpha1 catalytic subunit have increased bone remodelling and modiﬁed skeletal responses to hormonal challenges induced by ovariectomy and intermittent PTH treatment. J Endocrinol 2012; 214(3): 349–358. 6. Esbrit P, Alcaraz MJ. Current perspectives on parathyroid hormone (PTH) and PTH-related protein (PTHrP) as bone anabolic therapies. Biochem Pharmacol 2013; 85(10): 1417–1423. 7. Baron R, Kneissel M. WNT signaling in bone homeostasis and disease: from human mutations to treatments. Nat Med 2013; 19(2): 179–192. 30. Knothe Tate ML and U. Knothe, An ex vivo model to study transport processes and ﬂuid ﬂow in loaded bone. J Biomech 2000; 33(2): 247–254. 8. Paszty C, Turner CH, Robinson MK. Sclerostin: a gem from the genome leads to bone-building antibodies. J Bone Miner Res 2010; 25(9): 1897–1904. 31. Weinstein RS, Wan C, Liu Q, et al. Endogenous glucocorticoids decrease skeletal angiogenesis, vascularity, hydration, and strength in aged mice. © 2016 The Authors. Cell Biochemistry and Function published by John Wiley & Sons, Ltd. Cell Biochem Funct 2016; 34: 52–62. REFERENCES Parathyroid hormone 1–84 targets bone vascular structure and perfusion in mice: impacts of its administration regimen and of ovariectomy. J Bone Miner Res 2014; 29(7): 1608–1618. 47. Iwaniec UT, Wronski TJ, Liu J, et al. PTH stimulates bone formation in mice deﬁcient in Lrp5. J Bone Miner Res 2007; 22(3): 394–402. 23. Prisby R, Guignandon A, Vanden-Bossche A, et al. Intermittent PTH (1–84) is osteoanabolic but not osteoangiogenic and relocates bone marrow blood vessels closer to bone-forming sites. J Bone Miner Res 2011; 26(11): 2583–2596. 48. Lynch MA, Brodt MD, Stephens AL, Civitelli R, Silva MJ. Low-magni- tude whole-body vibration does not enhance the anabolic skeletal effects of intermittent PTH in adult mice. J Orthop Res 2011; 29(4): 465–472. 49. Stanislaus D, Devanarayan V, Hock JM. In vivo comparison of activated protein-1 gene activation in response to human parathyroid hormone (hPTH)(1–34) and hPTH(1–84) in the distal femur metaphyses of young mice. Bone 2000; 27(6): 819–826. 24. Rashid G, Bernheim J, Green J, Benchetrit S. Parathyroid hormone stimulates the endothelial expression of vascular endothelial growth factor. Eur J Clin Invest 2008; 38(11): 798–803. 50. Kusumbe AP, Ramasamy SK, Adams RH. Coupling of angiogenesis and osteogenesis by a speciﬁc vessel subtype in bone. Nature 2014; 507(7492): 323–328. 25. Wang DS, Yamazaki K, Nohtomi K, et al. Increase of vascular endothelial growth factor mRNA expression by 1,25-dihydroxyvitamin D3 in human osteoblast-like cells. J Bone Miner Res 1996; 11(4): 472–479.
https://openalex.org/W2969585532	https://hal.archives-ouvertes.fr/hal-02393147/file/124-GagnonERC2019-all.pdf	English	null	The carbon sink due to shrub growth on Arctic tundra: a case study in a carbon-poor soil in eastern Canada	Environmental research communications	2,019	cc-by	10,057	To cite this version: Mikael Gagnon, Florent Dominé, Stéphane Boudreau. The carbon sink due to shrub growth on Arctic tundra: a case study in a carbon-poor soil in eastern Canada. Environ. Res. Commun, 2019, 1 (9), pp.091001. 10.1088/2515-7620/ab3cdd. hal-02393147 Distributed under a Creative Commons Attribution 4.0 International License Abstract The microbial respiration of ancient permafrost carbon represents a positive feedback to climate warming. However, warming-induced shrub expansion in circumpolar latitudes may partly offset these emissions, due to greater biomass and litter inputs than that of primary tundra vegetation. Quantifying this carbon sink is challenging as the simultaneous mineralization of ancient carbon renders the attribution of changes in soil carbon stocks uncertain. We measured the contribution of shrubs to the terrestrial carbon reservoir in a Low-Arctic region where ancient carbon stocks are among the lowest in the Arctic. The study site near the eastern shore of Hudson Bay is experiencing rapid Betula glandulosa Michx. expansion throughout lichen tundra. We ﬁnd that the terrestrial carbon stocks (i.e. soil and vegetation) under a cover of low to medium-size shrubs is increased by 3.9±1.3 kg m−2 , regardless of shrub cover age. Along water tracks, taller shrubs and the transition to moss understories provide an even greater increase in terrestrial carbon (6.5±3.5 kg m−2). Using published maps of vegetation change from 1994 to 2010, we estimate that the carbon sink associated to shrub expansion in our study area (5.228 km2) has been 2.4±0.8 Gg or 29±9 g m−2 yr−1. Extrapolating this result to the Arctic requires additional studies in representative environments. The carbon sink due to shrub growth on Arctic tundra: a case study in a carbon-poor soil in eastern Canada Mikael Gagnon1,2,3, Florent Domine1,2,3,4 and Stéphane Boudreau2,5 1 Takuvik Joint International Laboratory, Université Laval (Canada) and CNRS-INSU (France), Québec City, QC, G1V 0A6, Canada 2 Centre d’études nordiques (CEN), Université Laval, Québec City, QC, G1V 0A6, Canada 3 Department of Chemistry, Université Laval, Québec City, QC, G1V 0A6, Canada 4 Department of Geography, Université Laval, Québec City, QC, G1V 0A6, Canada 5 Department of Biology, Université Laval, Québec City, QC, G1V 0A6, Canada E-mail: ﬂorent.domine@gmail.com Keywords: carbon, tundra, shrub, Arctic, permafrost Supplementary material for this article is available online Mikael Gagnon1,2,3, Florent Domine1,2,3,4 and Stéphane Boudreau2,5 1 Takuvik Joint International Laboratory, Université Laval (Canada) and CNRS-INSU (France), Québec City, QC, G1V 0A6, Canada 2 Centre d’études nordiques (CEN), Université Laval, Québec City, QC, G1V 0A6, Canada 3 Department of Chemistry, Université Laval, Québec City, QC, G1V 0A6, Canada 4 Department of Geography, Université Laval, Québec City, QC, G1V 0A6, Canada 5 Department of Biology, Université Laval, Québec City, QC, G1V 0A6, Canada E-mail: ﬂorent.domine@gmail.com Keywords: carbon, tundra, shrub, Arctic, permafrost Supplementary material for this article is available online Keywords: carbon, tundra, shrub, Arctic, permafrost Supplementary material for this article is available online HAL Id: hal-02393147 https://hal.science/hal-02393147v1 Submitted on 4 Dec 2019 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. Distributed under a Creative Commons Attribution 4.0 International License https://doi.org/10.1088/2515-7620/ab3cdd Environ. Res. Commun. 1 (2019) 091001 © 2019 The Author(s). Published by IOP Publishing Ltd 830 Pg; (IPCC 2013) Currently warming induced permafrost thaw is releasing part of this ancient (i e perennially OPEN ACCESS RECEIVED 31 May 2019 REVISED 18 July 2019 ACCEPTED FOR PUBLICATION 20 August 2019 PUBLISHED 2 September 2019 Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. LETTER The carbon sink due to shrub growth on Arctic tundra: a case study in a carbon-poor soil in eastern Canada Mikael Gagnon1,2,3, Florent Domine1,2,3,4 and Stéphane Boudreau2,5 1 Takuvik Joint International Laboratory, Université Laval (Canada) and CNRS-INSU (France), Québec City, QC, G1V 0A6, Canada 2 Centre d’études nordiques (CEN), Université Laval, Québec City, QC, G1V 0A6, Canada 3 Department of Chemistry, Université Laval, Québec City, QC, G1V 0A6, Canada 4 Department of Geography, Université Laval, Québec City, QC, G1V 0A6, Canada 5 Department of Biology, Université Laval, Québec City, QC, G1V 0A6, Canada E-mail: ﬂorent.domine@gmail.com Keywords: carbon, tundra, shrub, Arctic, permafrost Supplementary material for this article is available online Abstract The microbial respiration of ancient permafrost carbon represents a positive feedback to climate warming. However, warming-induced shrub expansion in circumpolar latitudes may partly offset these emissions, due to greater biomass and litter inputs than that of primary tundra vegetation. Quantifying this carbon sink is challenging as the simultaneous mineralization of ancient carbon renders the attribution of changes in soil carbon stocks uncertain. We measured the contribution of shrubs to the terrestrial carbon reservoir in a Low-Arctic region where ancient carbon stocks are among the lowest in the Arctic. The study site near the eastern shore of Hudson Bay is experiencing rapid Betula glandulosa Michx. expansion throughout lichen tundra. We ﬁnd that the terrestrial carbon stocks (i.e. soil and vegetation) under a cover of low to medium-size shrubs is increased by 3.9±1.3 kg m−2 , regardless of shrub cover age. Along water tracks, taller shrubs and the transition to moss understories provide an even greater increase in terrestrial carbon (6.5±3.5 kg m−2). Using published maps of vegetation change from 1994 to 2010, we estimate that the carbon sink associated to shrub expansion in our study area (5.228 km2) has been 2.4±0.8 Gg or 29±9 g m−2 yr−1. Extrapolating this result to the Arctic requires additional studies in representative environments. 1. Introduction Arctic land masses represent one ﬁfth of the emerged surface of the Earth and contain ca. 1300 Pg of frozen carbon in their permafrost (Hugelius et al 2014), a reservoir 1.5 times larger than the atmospheric one (ca. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. 1. Introduction Arctic land masses represent one ﬁfth of the emerged surface of the Earth and contain ca. 1300 Pg of frozen carbon in their permafrost (Hugelius et al 2014), a reservoir 1.5 times larger than the atmospheric one (ca. 830 Pg; (IPCC 2013). Currently, warming-induced permafrost thaw is releasing part of this ancient (i.e. perennially frozen) carbon into the atmosphere through microbial respiration. These CO2 and CH4 emissions have the potential to trigger a strong positive feedback to climate warming (Schuur et al 2008). However, these emissions may be partly offset by the simultaneous increase in primary productivity (i.e. plant growth) throughout the Arctic. The causes of shrub expansion have been detailed by (Myers-Smith et al 2011) and the main driver is climate warming and its consequences such as changes in snow cover and permafrost thaw. Permafrost thaw in particular favors nutrient recycling which contributes to shrub growth. Consequences of shrub expansion and growth include increases in both plant and soil carbon stocks through enhanced biomass production and subsequent soil organic matter (SOM) accumulation (Myers-Smith et al 2011). Quantifying the balance between ancient carbon microbial respiration and increased primary productivity is difﬁcult in many Arctic regions as long-term ﬁeld experiments on the impacts of shrub growth on soil carbon often generated contradictory results. For example, in their two decade-long experiment in the Alaskan tundra, Sistla et al (2013) found that, although simulated Arctic summer warming increased shrub growth and litter © 2019 The Author(s). Published by IOP Publishing Ltd Environ. Res. Commun. 1 (2019) 091001 Environ. Res. Commun. 1 (2019) 091001 inputs, it had no impact on total soil organic carbon (SOC) stocks, a result likely associated to the fact that surface (vegetation) inputs were compensated by ancient carbon mineralization. Nearby, Mack et al (2004), in their long-term (>20yr) fertilization experiment to simulate the positive effect of warmer temperatures on nutrient availability through an increase in soil microbial activity, observed a net loss of ca. 2 kg m−2 of carbon although ecosystem primary productivity doubled over the course of the experiment. The results of such studies may be difﬁcult to extrapolate throughout the Arctic as ancient carbon mineralization rates vary both spatially and temporally. 2.1. Study area Our study took place in Nunavik, a region currently experiencing one of the strongest greening trends in North America (Ju and Masek 2016). This greening trend is mainly driven by the expansion of dwarf birch (Ropars and Boudreau 2012, Tremblay et al 2012, Provencher-Nolet et al 2014). Based on dendrochronological data, Ropars et al (2015) argued that this greening trend likely occurred over the last two decades, a timeframe coinciding with the strong warming trend observed in Nunavik since the 1990s (Bhiry et al 2011). Our study area is located four kilometers east of the village of Umiujaq, in the upper part of the Tasiapik Valley (ﬁgure 1(a); 56.56° N, 76.48° W). Located on the eastern shore of Hudson Bay, the upper valley is comprised of post-glacial ﬂuvio-marine sandy sediments of the Tyrell Sea deposited during the last deglaciation (Lavoie et al 2012) and is characterized by discontinuous permafrost (Lemieux et al 2016). Soils are relatively young, poorly developed podzols and emerged from the sea ca. 7 000 to 6 000 years ago due to isostatic uplift (Lavoie et al 2012). As a result, SOC content in the upper valley is likely among the lowest in the Arctic (Hugelius et al (2014). (Provencher-Nolet et al 2014) have mapped the vegetation cover changes in the Tasiapik Valley from 1994 to 2010, showing a transition from lichen to shrub tundra. During our 2017 ﬁeld campaign, small shrubs (ca. 50 cm in height) were heterogeneously found throughout the lichen tundra in the upper valley (ﬁgure 1(b)). The tallest shrubs (ca. 1 m in height) were located within the main water tracks (ﬁgure 1(c)), consisting of channels where enhanced rates of water drainage allow greater primary productivity (Curasi et al 2016). These features often occur on hilly terrain and in valleys, particularly in permafrost regions where the frozen soil acts as a barrier for water drainage (Trochim et al 2016). Although little permafrost is found in the upper valley (Lemieux et al 2016), thaw ponds can be found lower down the valley, within 200 m of our sampling plots (ﬁgure 1(c)). These ponds could already be seen in the oldest aerial photograph of the valley dating back to 1957 (Natural Ressources Canada 1957) and were already surrounded by tall shrubs, suggesting that the establishment of these shrubs preceded the shrub expansion observed throughout Nunavik since the 1990s. 1. Introduction In contrast, Day et al (2008) observed a net total increases in soil and vegetation carbon stocks of about 300 g m−2 near Palmer Station (Antarctica) in a four-year warming experiment at a tundra site dominated by vascular plants. In their case, simulated warming did not alter SOC stocks at depths as soils were permafrost- free, had no ancient carbon and had only a thin (< 5 cm) organic layer overlaying glacial drift. Surface inputs to the terrestrial carbon reservoir remain one of the largest sources of uncertainty in Earth system models (Todd-Brown et al 2014, Koven et al 2015, Crowther et al 2016, McGuire et al 2018). Some modeling studies (McGuire et al 2018) indicate that future increases in plant biomass could offset ancient carbon mineralization, while others reached an opposite conclusion (Abbott et al 2016). Field studies aiming to unambiguously quantify the contribution of plant biomass on the terrestrial carbon reservoir are thus needed. In order to contribute to ﬁlling this knowledge gap, we conducted a study at a low Arctic tundra site near Umiujaq (Nunavik, 56.55° N, 76.55° W) where the impact of shrub growth on SOC accumulation could be quantiﬁed. The maps of (Hugelius et al 2014) suggest that SOC in the ﬁrst meter of soil in this region is <5 kg m−2, a result corroborated by our preliminary measurements in lichen tundra sites in the region (SOC<2 kg m−2). As dwarf birch (Betula glandulosa Michx.) has progressively been colonising the lichen tundra over the last decades, our study region appears suitable to quantify carbon stock changes in plant biomass and soils, with minimal noise associated to the mineralization of ancient carbon. 2.1. Study area In September-October 2017, sixteen 1 m2 plots were delimited in the upper Tasiapik Valley. The average elevation of our sampling plots is 132±10 m, as determined by DGPS using a nearby geodesic point. Plots were placed on pure lichen mats (n=3, hereafter referred to as lichen tundra), on small to medium shrubs (30 cm<height<80 cm) with lichens (n=8; hereafter referred to as medium shrub tundra) and on tall shrubs (ca. 1 m) with mosses, (n=5; hereafter referred to as high shrub tundra). Some herbs and other shrub species (Rhododendron groenlandicum Oeder, Vaccinium sp.) were also present in low abundance. Those shrubs were in general not as tall as birches, about 50 cm high at the most for R. groenlandicum and less than 30 cm for Vaccinium sp. In wetter areas shrubs such as Alnus viridis subsp. crispa Turrill and Salix planifolia Pursh. were also observed but these were essentially absent from our plots. Lichens present were mostly Cladonia spp., mostly C. 2 2 Environ. Res. Commun. 1 (2019) 091001 Figure 1. Location of the study area. (a) Map of the Tasiapik Valley region, modiﬁed from (NRCAN 2016). Contour lines are 20 m in elevation apart. Green areas are dominated by trees or shrubs. White areas are lichen tundra or rock. (b) Illustration of the recent expansion of dwarf birch throughout the lichen tundra of the upper Tasiapik Valley, Nunavik. The green vegetation is dwarf birch shrubs which have grown in recent decades while the yellow/beige vegetation is the original lichen vegetation. (c) Extent of the upper valley region studied (red), along with the main water tracks (blue). Thaw ponds can be seen further down in the valley, followed by forest-tundra with black spruce (Picea mariana Mill.) trees (Payette et al 2001). The yellow star shows the location from where photograph (b) was taken. Image (c) was taken in October 2017, when the dwarf birch had lost most of its foliar biomass. As a result, only the largest shrub canopies of the upper valley can be seen, revealing the main water tracks, outlined in blue. Figure 1. Location of the study area. (a) Map of the Tasiapik Valley region, modiﬁed from (NRCAN 2016). Contour lines are 20 m in elevation apart. Green areas are dominated by trees or shrubs. White areas are lichen tundra or rock. 2.1. Study area (b) Illustration of the recent expansion of dwarf birch throughout the lichen tundra of the upper Tasiapik Valley, Nunavik. The green vegetation is dwarf birch shrubs which have grown in recent decades while the yellow/beige vegetation is the original lichen vegetation. (c) Extent of the upper valley region studied (red), along with the main water tracks (blue). Thaw ponds can be seen further down in the valley, followed by forest-tundra with black spruce (Picea mariana Mill.) trees (Payette et al 2001). The yellow star shows the location from where photograph (b) was taken. Image (c) was taken in October 2017, when the dwarf birch had lost most of its foliar biomass. As a result, only the largest shrub canopies of the upper valley can be seen, revealing the main water tracks, outlined in blue. stellaris Opiz and C. rangiferina L. with some C. bellidiﬂora and C. coccifera. Other lichen species included Flavocetraria nivalis L., Alectoria ochroleuca Hoffm., and Stereocaulon paschale L. Snow depth was recorded at each plot on April 9, 2018. 2.2. Soil carbon stocks Soil pits were dug at each plot until reaching the C mineral horizon (ﬁgure 2). The upper 30 cm of the soil proﬁle was sampled wth a 5 cm resolution (i.e. 0–5 cm, 5–10 cmK, 25–30 cm) with a spatula and placed in 50 ml centrifuge tubes. Dead mosses, decaying lichens and ﬁne roots were included in the soil sampling. Soil density was measured using a 200 cm3 cylinder (4.78 cm in diameter) using the same 5 cm depth intervals. Soil samples were frozen and brought back to the laboratory, where subsamples were weighed, dried at 70 °C for 72 h and weighed a second time in order to determine their gravimetric water content. The subsamples were sieved at 2 mm, ground in a ball mill grinder and analysed for their total carbon content using a Leco TruMac CNS (carbon, nitrogen, sulfur) Macro Analyzer, operating under oxygen ﬂow at 1400 °C. Two additional subsamples per plot were used to measure the inorganic carbon content of soil horizons A and B. These samples were dried, ground and heated at 400°C for four hours before being placed in the CNS analyser. The organic carbon content of the soil was thus obtained by subtracting inorganic carbon from total carbon content. The uncertainty of the soil organic carbon stocks (kg m−2) was estimated to be in the range of 11 to 13%, based on analytical error (about 5%) and on error in soil density measurement. Most of the error was caused by the density measurement, about 10%. This is due to errors in determining the size of the sampler (negligible), the sample weight (±0.2 g) and to difﬁculties in obtaining a perfect soil sample. This is actually the main cause of error, and could be due to the presence of roots or rocks, and sometimes difﬁculties in perfectly ﬁlling the sampler with soil. 2.3. Soil characterization Using the same soil samples, granulometric analyses of soil horizons A and B were performed at each plot using the Horiba LA-950 Laser Particle Size Analyser. Prior to analysis, soil samples were dried at 400 °C, sieved at 2 mm and agitated in a deﬂocculating solution of 0.01 M sodium hexametaphosphate. For soil classiﬁcation, the pyrophosphate-extractible Al and Fe contents of the B horizon were measured at two plots for each vegetation type (lichen, medium and high shrub tundra). 3 Environ. Res. Commun. 1 (2019) 091001 Figure 2. Three types of vegetation covers selected in the upper Tasiapik Valley. (a) Lichen tundra. The beige lichen patch is surrounded by dwarf birch with leaves in fall colors; (c) Medium shrub tundra with visible lichen understory; (e) High shrub tundra with visible green moss understory. On the second row, their respective soil proﬁles (b), (d), and (f) are shown in the same order. The horizontal lines on images b, d, and f are at the 10 cm depth. Figure 2. Three types of vegetation covers selected in the upper Tasiapik Valley. (a) Lichen tundra. The beige lichen patch is surrounded by dwarf birch with leaves in fall colors; (c) Medium shrub tundra with visible lichen understory; (e) High shrub tundra with visible green moss understory. On the second row, their respective soil proﬁles (b), (d), and (f) are shown in the same order. The horizontal lines on images b, d, and f are at the 10 cm depth. 2.5. Tree-ring analysis In September 2017, a minimum of three dwarf birch samples were taken at each plot to determine the age of the shrub cover. Although samples were taken at the junction between above- and below-ground biomass, the shapes of the junctions suggest they mostly originated from clonal growth rather than from seed germination. Therefore, the oldest shrub sampled at each plot would only provide a minimum age of the shrub cover. Analyses were performed in the dendrochronology laboratory of the Centre for northern studies (CEN) at Université Laval. The samples were boiled for a minimum of four hours, enabling thin sections of ca. 50 microns to be sliced using a microtome. The thin sections were then dyed in a safranin solution (1%) and glued on microscope slides using a 66% toluene solution (SHUR/MountTM Liquid mounting medium). Samples were then photographed using a binocular-mounted camera (Olympus sZ61 with a SC100 camera) and dated by visual ring counting. 2.4. Vegetation carbon stocks Total biomass (dwarf birch, lichens and mosses) was determined at each plot, with the exception of graminoids, as their cover was too low. Above- and below-ground dwarf birch biomass was harvested within 1 m2 plots (ﬁgure 2). Most of the root biomass was conﬁned to the top 40 cm of soil, although ﬁner roots could reach depths of 60 cm. The birch biomass was air-dried and weighed in Umiujaq. However, as dwarf birch samples were not fully dried, subsamples of small, medium and tall shrubs were brought back to Université Laval where they were thoroughly dried at 70°C for 72 h. Moisture correction factors, ranging from 30 to 50% (mass fraction) were then applied to the air-dried samples. To quantify lichen biomass, 20×20 cm sections of live lichen mats were hand-collected. Moss biomass was hand-collected over a 31.6×31.6 cm (0.1 m2) area. Lichens and mosses were dried at 70°C for 72 h. Back in the laboratory, dried lichen, moss and birch samples were ground in liquid nitrogen and analysed for their carbon content using the CNS analyser. The error was estimated from the precision of the 1 m2 sampling, the extrapolation of dry mass from the subsamples to the whole samples and from the analytical accuracy. The overall error (table S2 is available online at stacks.iop.org/ERC/1/091001/ mmedia) was in the range 9 to 16%. 3.2. Snow depth Snow depth in April 2018 was found to be relatively constant throughout the upper valley. Lichen tundra, medium shrub tundra and high shrub tundra had snow depths of 105±30, 113±8 and 115±8 cm, respectively, resulting in a combined average of 110±10 cm for all 16 plots. An automatic snow gauge and 2 time-lapse cameras revealed that snow depth in early April was within 15 cm of the peak depth reached on 14 May 2018. Detailed snow observations have been made in spring since 2012 (Domine et al 2015) and the snow depth recorded in spring 2018 was the highest observed. 3.3. Soil carbon content Vegetation cover greatly inﬂuenced the visible amount of black organic matter in the uppermost 20 cm of the soil proﬁles (ﬁgure 2). Lichen tundra plots (ﬁgure 2(a)) only had a very thin layer of decomposing lichens within the ﬁrst two centimeters of the sandy horizon (ﬁgure 2(b)). Their highest SOC content recorded between depths of 5 to 30 cm was on average 0.30% (mass fraction of dry soil; ﬁgure 3). Fallen leaves from nearby shrubs were often found in the lichen mats and likely contributed to the SOC content of these sites. Medium shrub tundra plots (ﬁgure 2(c)) contained a thicker layer of still distinguishable, decomposing organic matter in the uppermost 5 cm of the soil proﬁles (ﬁgure 2(d)). Compared to the lichen tundra, their SOC contents were slightly greater at depths, reaching 1.27% below 5 cm. The greatest SOC contents (ﬁgure 3) were however recorded at high shrub tundra plots (ﬁgure 2(e)), where soils had a thick organic horizon which almost reached 10 cm in depth and mainly consisted of decomposing mosses (ﬁgure 2(f)), so that between 5 and 10 cm the organic carbon content was 5.72%. 3.1. Soil characterization Such classiﬁcation results from their low mass content of organic carbon (0.31±0.18%), Fe (0.05±0.04%) and Al (0.10±0.09%) in the Bm horizon (Supporting information, table S1). Soils were acidic, with an average CaCl2 pH of 4.10±0.10 in the Bm horizon. Group 1998). Such classiﬁcation results from their low mass content of organic carbon (0.31±0.18%), Fe (0.05±0.04%) and Al (0.10±0.09%) in the Bm horizon (Supporting information, table S1). Soils were acidic, with an average CaCl2 pH of 4.10±0.10 in the Bm horizon. 3.1. Soil characterization Granulometric analyses (supporting information, ﬁgure S1) revealed that the upper valley consisted of sandy soils throughout the A and B horizons, with the exception of a loamy sand layer (ca. 89% sand, 11% silt, 0% clay) found within the ﬁrst 5 to 10 cm of the A horizon at 10 of the 16 sites. A 10 cm thick surface layer of sandy loam (69.8% sand, 30.2% silt, 0% clay) was also found at one site. Soils were podzolic in appearance and were classiﬁed as brunisolic, according to the Canadian system of soil classiﬁcation (Soil Classiﬁcation Working 4 Figure 3. Average soil organic carbon content according to the type vegetation cover: Lichen tundra (n=3), Medium shrub tundra (n=8) and High shrub tundra (n=5). Compared to the lichen tundra, both shrub tundra types have slightly greater SOC contents at depths (10–30 cm). The inserted graph shows a blown-up view of SOC content from a depth of 10 cm. Error bars represent standard deviations. The colored bars are overlain and all start at 0% carbon. Environ. Res. Commun. 1 (2019) 091001 Environ. Res. Commun. 1 (2019) 091001 Figure 3. Average soil organic carbon content according to the type vegetation cover: Lichen tundra (n=3), Medium shrub tundra (n=8) and High shrub tundra (n=5). Compared to the lichen tundra, both shrub tundra types have slightly greater SOC contents at depths (10–30 cm). The inserted graph shows a blown-up view of SOC content from a depth of 10 cm. Error bars represent standard deviations. The colored bars are overlain and all start at 0% carbon. Figure 3. Average soil organic carbon content according to the type vegetation cover: Lichen tundra (n=3), Medium shrub tundra (n=8) and High shrub tundra (n=5). Compared to the lichen tundra, both shrub tundra types have slightly greater SOC contents at depths (10–30 cm). The inserted graph shows a blown-up view of SOC content from a depth of 10 cm. Error bars represent standard deviations. The colored bars are overlain and all start at 0% carbon. Group 1998). Such classiﬁcation results from their low mass content of organic carbon (0.31±0.18%), Fe 0.05±0.04%) and Al (0.10±0.09%) in the Bm horizon (Supporting information, table S1). Soils were acidi with an average CaCl2 pH of 4.10±0.10 in the Bm horizon. Group 1998). 3.4. Total terrestrial carbon stocks P=0.001; ﬁgure 5). As expected, the relationship was weaker (F1,14=2.88, P>0.05) when tall shrub tundra plots were included. 3.4. Total terrestrial carbon stocks The total terrestrial carbon stocks (TTCS) of the upper Tasiapik Valley (ﬁgure 4) include the carbon stocked in the shrubs, in the understory vegetation and in the soil. Detailed data on terrestrial carbon stocks are given in the Supporting information, table S2. The lichen tundra contained the lowest TTCS (2.3±0.4 kg m−2), with SOC contributing only 1.4±0.2 kg m−2 (ﬁgure 4). In comparison, total terrestrial carbon stocks in the medium shrub tundra averaged 6.2±1.2 kg m−2 (ﬁgure 4) and most of this carbon was contained in the soil (4.3±0.5 kg m−2). Medium shrub carbon was 1.3±0.3 kg m−2, which compensated for the lower contribution of lichens at these plots. Lastly, the high shrub tundra held the greatest total terrestrial carbon stocks (8.8±3.5 kg m−2). Their SOC stocks of 6.2±0.7 kg m−2 were also the greatest of the three vegetation covers, partly due to the thick layer of decomposing mosses found in the litter. Shrub biomass was also the greatest and contained on average 2.3±0.7 kg m−2 of carbon. Green moss dry biomass varied greatly between plots and its carbon content ranged from 0.02 to 0.53 kg m−2, adding on average 0.2±0.2 kg m−2 of carbon to the terrestrial reservoir (ﬁgure 4). SOC content was signiﬁcantly associated with the carbon stock in the aboveground vegetation (shrubs, lichens, mosses), particularly when considering only lichen tundra and medium shrub plots (F1,9=23.18, 5 Environ. Res. Commun. 1 (2019) 091001 P=0.001; ﬁgure 5). As expected, the relationship was weaker (F1,14=2.88, P>0.05) when tall shrub tundra Figure 4. Distribution of total terrestrial carbon stocks (TTCS) in different compartments (soil, lichens, mosses and shrubs) of terrestrial ecosystems found in the upper valley. Plots are classiﬁed according to their vegetation cover. Figure 5. Relationship between soil organic carbon (SOC) content and plant live biomass carbon stock. Closed symbols represent plots sampled in the lichen tundra and in the medium shrub tundra vegetation type. Triangles are for lichen tundra and circles for medium shrub tundra. Open symbols represent plots sampled in the tall shrub tundra type. The regression line is based on data from the lichen tundra and medium shrub tundra plots. As the regression including also the tall shrub tundra was not signiﬁcant, it is not shown here. Its equation is y= 1.49 × +1.58, R2=0.17. Figure 4. 3.5. Tree ring analysis g y Dendrochronological data revealed that the minimal time elapsed since the establishment of the shrub covers ranged from 26 to 79 years (relative to sampling year 2017). Detailed data on shrub age are given in the Supporting information, table S3. A stem section micrograph used in age determination is shown in ﬁgure S2. For medium shrubs, the data indicate no signiﬁcant correlation between terrestrial carbon stocks and shrub age (R2=0.22, p=0.24; ﬁgure 6). Hence, the transition from lichen to medium shrub tundra resulted in a net gain of 3.9±1.3 kg m−2 of carbon, independent of shrub age within the age range sampled. The slope of the correlation of ﬁgure 6 may actually suggest that carbon stocks slightly decreased as shrubs aged, but the trend is not signiﬁcant. High shrub tundra plots are omitted from the regression because moss growth represents a different carbon sink than lichens. Due to insufﬁcient data (n=5), we did not conduct a separate regression for these plots. Dendrochronological data revealed that the minimal time elapsed since the establishment of the shrub covers ranged from 26 to 79 years (relative to sampling year 2017). Detailed data on shrub age are given in the Supporting information, table S3. A stem section micrograph used in age determination is shown in ﬁgure S2. For medium shrubs, the data indicate no signiﬁcant correlation between terrestrial carbon stocks and shrub age (R2=0.22, p=0.24; ﬁgure 6). Hence, the transition from lichen to medium shrub tundra resulted in a net gain of 3.9±1.3 kg m−2 of carbon, independent of shrub age within the age range sampled. The slope of the l i f ﬁ 6 ll h b k li h l d d h b d b h d i Dendrochronological data revealed that the minimal time elapsed since the establishment of the shrub covers ranged from 26 to 79 years (relative to sampling year 2017). Detailed data on shrub age are given in the Supporting information, table S3. A stem section micrograph used in age determination is shown in ﬁgure S2. For medium shrubs, the data indicate no signiﬁcant correlation between terrestrial carbon stocks and shrub age (R2=0.22, p=0.24; ﬁgure 6). Hence, the transition from lichen to medium shrub tundra resulted in a net gain of 3.9±1.3 kg m−2 of carbon, independent of shrub age within the age range sampled. 3.4. Total terrestrial carbon stocks Distribution of total terrestrial carbon stocks (TTCS) in different compartments (soil, lichens, mosses and shrubs) of terrestrial ecosystems found in the upper valley. Plots are classiﬁed according to their vegetation cover. Figure 4. Distribution of total terrestrial carbon stocks (TTCS) in different compartments (soil, lichens, mosses and shrubs) of terrestrial ecosystems found in the upper valley. Plots are classiﬁed according to their vegetation cover. Figure 4. Distribution of total terrestrial carbon stocks (TTCS) in different compartments (soil, lichens, mosses and shrubs) of terrestrial ecosystems found in the upper valley. Plots are classiﬁed according to their vegetation cover. Figure 5. Relationship between soil organic carbon (SOC) content and plant live biomass carbon stock. Closed symbols represent plots sampled in the lichen tundra and in the medium shrub tundra vegetation type. Triangles are for lichen tundra and circles for medium shrub tundra. Open symbols represent plots sampled in the tall shrub tundra type. The regression line is based on data from the lichen tundra and medium shrub tundra plots. As the regression including also the tall shrub tundra was not signiﬁcant, it is not shown here. Its equation is y= 1.49 × +1.58, R2=0.17. Figure 5. Relationship between soil organic carbon (SOC) content and plant live biomass carbon stock. Closed symbols represent plots sampled in the lichen tundra and in the medium shrub tundra vegetation type. Triangles are for lichen tundra and circles for medium shrub tundra. Open symbols represent plots sampled in the tall shrub tundra type. The regression line is based on data from the lichen tundra and medium shrub tundra plots. As the regression including also the tall shrub tundra was not signiﬁcant, it is not shown here. Its equation is y= 1.49 × +1.58, R2=0.17. Figure 5. Relationship between soil organic carbon (SOC) content and plant live biomass carbon stock. Closed symbols represent plots sampled in the lichen tundra and in the medium shrub tundra vegetation type. Triangles are for lichen tundra and circles for medium shrub tundra. Open symbols represent plots sampled in the tall shrub tundra type. The regression line is based on data from the lichen tundra and medium shrub tundra plots. As the regression including also the tall shrub tundra was not signiﬁcant, it is not shown here. Its equation is y= 1.49 × +1.58, R2=0.17. P=0.001; ﬁgure 5). As expected, the relationship was weaker (F1,14=2.88, P>0.05) when tall shrub tundra plots were included. 4.1. The carbon stocks of tundra biomes Our results indicate that the total terrestrial carbon stock in the upper Tasiapik Valley is a function of the complexity of the vegetation structure. Lichen tundra contained the lowest carbon reservoir (2.3±0.4 kg m−2), followed respectively by medium (6.2±1.2 kg m−2) and high shrub tundra (8.8±3.5 kg m−2). Thus, the observed transition from lichen tundra to medium shrub tundra in the upper valley has the potential to increase the terrestrial carbon stocks by 3.9±1.3 kg m−2. Although comparative SOC data is limited in Eastern Canada, particularly in Nunavik, as shown by the absence of data in the soil carbon map of Hugelius et al (2014), our SOC data are generally at the lower end of data obtained in other parts of North America. For example, our SOC stocks in the high shrub tundra are roughly similar to the one of a mossy shrub tundra site near the northwestern Alaskan treeline (Wilmking et al 2006). There, the tundra consisted of an old ﬂoodplain with silt and sand deposits underlain by coarse material and permafrost at a depth of 40 cm. The surface organic horizon reached an average depth of 10.7±7.1 cm and contained 6.21±3.73 kg m−2 of carbon. The mineral horizon making up the remainder of the 30 cm of the soil proﬁle contained an additional 3.2 kg m−2 of carbon. Radiocarbon dating revealed that at least part of this carbon was of modern origin. However, the shrub carbon sink could not be fully assessed as isotopically enriched carbon makes it difﬁcult to evaluate the proportion of ancient carbon found in the mineral horizon (Levin et al 2010). In a similar study at Tulemalu Lake, in the continuous permafrost terrain of the Central Canadian Low Arctic, Hugelius et al (2010) compared the carbon stocks of various ecosystem types. The shrub tundra (mainly Salix spp. and Betula spp.) consisted of turbic cryosols made up of loamy glacial till and sandy glacioﬂuvial deposits with varying levels of peat. In the upper 30 cm of soil, SOC stocks were respectively 4.8±3.5 kg m−2 for lichen tundra, 6.0±3.0 kg m−2 for dry shrub tundra, 9.8±5.2 kg m−2 for moist shrub tundra and 20.6±4.1 kg m−2 for wet shrub tundra. Such differences between ecosystems appear to be driven by moisture, a result corroborating our ﬁndings in the Tasiapik valley. 3.5. Tree ring analysis The slope of the correlation of ﬁgure 6 may actually suggest that carbon stocks slightly decreased as shrubs aged, but the trend is not signiﬁcant. High shrub tundra plots are omitted from the regression because moss growth represents a different carbon sink than lichens. Due to insufﬁcient data (n=5), we did not conduct a separate regression for these plots. 6 Environ. Res. Commun. 1 (2019) 091001 Figure 6. Plot of the terrestrial carbon stocks of the medium and tall shrub tundra plots (kg m−2) as a function of the age of the oldest shrub sampled in the plot. The least square ﬁt is just for the medium shrub tundra, and the correlation is not signiﬁcant, p=0.24. For comparison, the lichen tundra holds an average terrestrial carbon stock of 2.2±0.4 kg m−2 (ﬁgure 4). Figure 6. Plot of the terrestrial carbon stocks of the medium and tall shrub tundra plots (kg m−2) as a function of the age of the oldest shrub sampled in the plot. The least square ﬁt is just for the medium shrub tundra, and the correlation is not signiﬁcant, p=0.24. For comparison, the lichen tundra holds an average terrestrial carbon stock of 2.2±0.4 kg m−2 (ﬁgure 4). 4.2. The carbon sink in the tasiapik valley Based on the results of (Provencher-Nolet et al 2014), shrub expansion from 1994 to 2010 led to a 12% (0.609 km2) increase in shrub cover in the valley (5.228 km2). As the majority of this expansion occurred by the growth of medium shrubs on the lichen tundra, it is possible to extrapolate our results to a valley-wide scale. Given a terrestrial carbon stock gain of 3.9 kg m−2 for the vegetation transition over 0.609 km2, the areal increase of the medium shrub tundra within this 16yr-period induced a carbon sink of 2.4±0.8 Gg of carbon throughout the entire valley, or an addition of 28±9 gC m−2 yr−1 averaged over the valley area. As suggested by (Lemay et al 2018), shrub expansion in the valley is likely to continue in the future. Using land-cover change analysis and modeling, they projected that shrub cover should increase from ca. 60% in 2010 to ca. 70% in 2030, when it should essentially cease as shrubs would have colonized most of the available area. This future shrub expansion throughout the 5.228 km2 valley would then add an additional 2.0±0.7 Gg, resulting in a carbon sink of 4.4±1.0 Gg from 1994 to 2030. g This case study is clearly insufﬁcient to extrapolate to the circumpolar Arctic, but it does represent a useful case to help constrain future estimates of the permafrost carbon budget. (McGuire et al 2018) reported simulations of circumpolar permafrost carbon evolution until the year 2299 using 5 models. Under the RCP4.5 projection, simulated soil carbon stocks varied from gains of 70 Pg C to losses of 67 Pg C, while carbon gains in vegetation were also widely scattered, 69±70 Pg C, showing that there are huge uncertainties in our understanding of the fate of permafrost carbon and in the contribution of vegetation changes to these stocks. To place our local results in perspective, the gain of 3.9 kg m-2 of carbon due to the lichen to shrub tundra transition observed here would translate into a 6.8 Pg C gain if extrapolated to the 1 733 000 km2 of High-Arctic tundra (Walker et al 2005). Unfortunately, (Walker et al 2005) do not map lichen tundra separately so we cannot evaluate how the Arctic-wide transformation of lichen tundra into medium shrub tundra would affect the Arctic carbon stocks. Our last extrapolation then has limited value. 4.1. The carbon stocks of tundra biomes Deeper in the soil proﬁle (from 30 to 100 cm), ancient carbon stocks, as revealed by radiocarbon dating, ranged from 7.0±6.0 to 19.8±10.9 kg m−2. Again, the presence of both ancient and modern carbon (bomb carbon) in the superﬁcial horizons clouds the shrub contribution to the SOC stocks. Our study in the Tasiapik valley presents a much simpler situation. First, the generalized very low level of SOC in the mineral horizons (ﬁgure 3) allows us to associate the change in SOC content to plant community shifts, at least from lichen tundra to medium shrub tundra. Second, the transition from lichen tundra to medium shrub tundra has been well documented in the Tasiapik valley (Provencher-Nolet et al 2014), with most of the expansion believed to be recent, i.e. post 1990, as observed in other parts of Nunavik (Ropars et al 2015). The strong positive association between SOC content and above-ground plant carbon stock on well-drained sites suggest that SOC content is closely associated with the development of an erect shrub cover (ﬁgure 5). However, the lack of a signiﬁcant relationship between SOC content and the age of the oldest stem in the plot (ﬁgure 6) was not expected. Such a result may indicate that a steady state between organic matter accumulation and degradation is reached in a short period, less than 30 years, perhaps even less than 20 (ﬁgure 6). That a steady state is reached for soil and terrestrial organic carbon is generally valid, and forms the basis of soil carbon modeling, see e.g. (Wang et al 2017) or (Nicoloso et al 2016). Regarding the time to reach steady state we are not 7 Environ. Res. Commun. 1 (2019) 091001 Environ. Res. Commun. 1 (2019) 091001 aware of any relevant Arctic study but (Nicoloso et al 2016) modeled agricultural land in Kansas and found an equilibration time of about 30 years, in line with local measurements. While this mid-latitude study obviously took place under vastly different conditions, it may indicate that our observed equilibration time is not unreasonable. Alternatively, it could also mean that stem clonal renewal is such that the age of the oldest stem found in a plot is not a suitable indicator of shrub age. Lastly, since we observed that some stems were more than 75 yr-old (ﬁgure 6 and table S3), i.e. 4.1. The carbon stocks of tundra biomes they preceded shrub expansion by a few decades, it could be concluded that shrub age is not a valid indicator of shrub expansion at our study site. In any case, tree-ring dating here did not yield any conclusive results. The relationship between SOC content and above-ground plant carbon stock was much weaker (and not signiﬁcant) when tall tundra plots were added. Such a result is believed to be mainly associated with different edaphic conditions between lichen/medium shrub tundra sites, found mainly on well-drained sites, and tall shrub tundra sites found in water tracks. As a result, it appears highly unlikely that SOC content of medium shrub tundra site will be comparable to the one in tall shrub tundra sites sampled in this study due to signiﬁcant edaphic differences. However, as the shrub cover will continue to grow vertically and get denser, mosses will eventually replace lichens underneath the shrub canopy (Schuur et al 2007, Paradis et al 2016). This process is likely to contribute to local increases in SOC content, although there might be an important variability between sites. In fact, green moss carbon stocks throughout tall shrub tundra varied greatly, from 0.021 to 0.528 kg m-2 (Table S2). The strength of this potential carbon sink thus cannot be determined without further study. Preferential snow accumulation in shrub covers has been suggested as a having a positive impact on shrub growth, as the thermal insulation provided by the greater snow depth would prevent frost damage and increased nutrient availability via greater soil microbial activity (Sturm et al 2005). Here, we did not observe signiﬁcant differences in snow depth between lichen tundra, medium shrub tundra and tall shrub tundra, even though (Paradis et al 2016) found a signiﬁcant relationship between snow cover thickness and shrub height. Snow depth in spring 2018 was the greatest recorded since 2012 and it was greater than shrub height, as we saw no protruding branches. (Domine et al 2016) observed that shrubs increased snow depth only up to their own height, so that under the spring 2018 conditions, any impact of shrubs on snow accumulation could not be observed. Shrub effect on snow accumulation could nevertheless have taken place earlier in the season, as observed elsewhere (Paradis et al 2016, Barrere et al 2018). Acknowledgments This work was funded by the Natural sciences and Engineering Research Council of Canada (NSERC), the BNP Paribas Foundation, the Institut polaire français (IPEV), the INTAROS project funded by the ERC, and the Northern Scientiﬁc Training Program (NSTP). The Centre for northern studies (CEN) provided us with the research facilities in Umiujaq. Special thanks to Alain Brousseau for performing numerous soil analyses, as well as to Ann Delwaide and Élizabeth Demeule for their assistance with the dendrochrological studies. We thank Serge Payette for encouragements and many helpful discussions. Xavier Morel, Marie-José Naud and Claire Brutel-Vuilmet assisted in preliminary exploration of the potential of the site for this work. 4.2. The carbon sink in the tasiapik valley In conclusion, (Turetsky et al 2019) evaluated research priorities to understand the contribution of permafrost emissions to future warming and stressed that ‘we need to identify the extent to which plant growth will offset the carbon that is released by permafrost’. Many studies similar to that conducted here are therefore required for a reliable assessment of the evolution of the permafrost carbon budget. 8 Environ. Res. Commun. 1 (2019) 091001 Social media abstract Shrub expansion at a low Arctic site results in a carbon sink of about 4 kg per m2. ORCID iDs Florent Domine https://orcid.org/0000-0001-6438-6879 Florent Domine https://orcid.org/0000-0001-6438-6879 References 14 1827–43 Domine F, Barrere M and Morin S 2016 The growth of shrubs on high Arctic tundra at Bylot Island: impact on snow physical properties and permafrost thermal regime Biogeosciences 13 6471–86 Domine F, Barrere M and Morin S 2016 The growth of shrubs on high Arctic tundra at Bylot Island: impact on snow physical properties and permafrost thermal regime Biogeosciences 13 6471–86 p g g Domine F, Barrere M, Sarrazin D, Morin S and Arnaud L 2015 Automatic monitoring of the effective thermal conductivity of snow in a low- Arctic shrub tundra The Cryosphere 9 1265–76 Domine F, Barrere M, Sarrazin D, Morin S and Arnaud L 2015 Automatic monitoring of the effective thermal conductivity of snow in a low- Arctic shrub tundra The Cryosphere 9 1265–76 y p Hugelius G, Kuhry P, Tarnocai C and Virtanen T 2010 Soil organic carbon pools in a periglacial landscape: a case study from the Central Canadian Arctic Permafrost and Periglacial Processes 21 16–29 Canadian Arctic Permafrost and Periglacial Processes 21 16–29 Hugelius G et al 2014 Estimated stocks of circumpolar permafrost carbon with quantiﬁed uncertainty ranges and identiﬁed data gaps Biogeosciences 11 6573–93 g IPCC 2013 Climate Change 2013: the physical science basis Contribution of Working group I to the ﬁfth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press) Ju J C and Masek J G 2016 The vegetation greenness trend in Canada and US Alaska from 1984-2012 Landsat data Remote Sens Environ 176 IPCC 2013 Climate Change 2013: the physical science basis Contribution of Working group I to the ﬁfth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press) Ju J C and Masek J G 2016 The vegetation greenness trend in Canada and US Alaska from 1984-2012 Landsat data Remote Sens. Environ. 176 1–16 Intergovernmental Panel on Climate Change (Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press) u J C and Masek J G 2016 The vegetation greenness trend in Canada and US Alaska from 1984-2012 Landsat data Remote Sens. Environ. 17 1–16 Ju J C and Masek J G 2016 The vegetation greenness trend in Canada and US Alaska from 1984-2012 Landsat data Remote Sens. References Enviro 1–16 Koven C D, Chambers J Q, Georgiou K, Knox R, Negron-Juarez R, Riley W J, Arora V K, Brovkin V, Friedlingstein P and Jones C D 2015 Controls on terrestrial carbon feedbacks by productivity versus turnover in the CMIP5 Earth system models Biogeosciences 12 5211–28 Lavoie C, Allard M and Duhamel D 2012 Deglaciation landforms and C-14 chronology of the Lac Guillaume-Delisle area, eastern hudson bay: a report on ﬁeld evidence Geomorphology 159 142–55 Lemay M-A, Provencher-Nolet L, Bernier M, Lévesque E and Boudreau S 2018 Spatially explicit modeling and prediction of shrub cover increase near Umiujaq, Nunavik Ecological Monographs 0 Lemay M-A, Provencher-Nolet L, Bernier M, Lévesque E and Boud increase near Umiujaq, Nunavik Ecological Monographs 0 Lemieux J M, Fortier R, Talbot-Poulin M C, Molson J, Therrien R, Ouellet M, Banville D, Cochand M and Murray R 2016 Groundwat occurrence in cold environments: examples from Nunavik, Canada Hydrol. J. 24 1497–513 Levin I, Naegler T, Kromer B, Diehl M, Francey R J, Gomez-Pelaez A J, Steele L P, Wagenbach D, Weller R and Worthy D E 2010 Observations and modelling of the global distribution and long-term trend of atmospheric 14CO2 Tellus Series B-Chemical and Physical Meteorology 62 26–46 y gy Mack M C, Schuur E A G, Bret-Harte M S, Shaver G R and Chapin F S 2004 Ecosystem carbon storage in arctic tundra reduced by long-term nutrient fertilization Nature 431 440–3 Mack M C, Schuur E A G, Bret-Harte M S, Shaver G R and Chapin F S 2004 Ecosystem carbon storage in arctic tundra reduced by long-term nutrient fertilization Nature 431 440–3 McGuire A D et al 2018 Dependence of the evolution of carbon dynamics in the northern permafrost region on the trajectory of climate McGuire A D et al 2018 Dependence of the evolution of carbon dynamics in the northern permafrost region on the trajectory of climate change PNAS 115 3882–7 Myers-Smith I H et al 2011 Shrub expansion in tundra ecosystems: dynamics, impacts and research priorities Environ. Res. Lett. 6 045509 Myers-Smith I H et al 2011 Shrub expansion in tundra ecosystems: dynamics, impacts and research priorities Environ. Res. Lett. References Abbott B W et al 2016 Biomass offsets little or none of permafrost carbon release from soils, streams, and wildﬁre: an expert assessment Environ. Res. Lett. 11 034014 Barrere M, Domine F, Belke-Brea M and Sarrazin D 2018 Snowmelt events in autumn can reduce or cancel the soil warming effect of snow– vegetation interactions in the Arctic J. Clim. 31 9507–18 Q y g g p g Curasi S R, Loranty M M and Natali S M 2016 Water track distribution and effects on carbon dioxide ﬂux in an eastern Siberian upland tundra landscape Environ. Res. Lett. 11 045002 Day T A, Ruhland C T and Xiong F S 2008 Warming increases aboveground plant biomass and C stocks in vascular-plant-dominated Antarctic tundra Global Change Biol. References 6 045509 Natural Ressources Canada 1957 National earth observation data framework catalogue https://www.eodms-sgdot.nrcan-rncan.gc.ca/ y y y Natural Ressources Canada 1957 National earth observation data framework catalogue https://www.eodms-sgdot.nrcan-rncan.gc.ca/ index_en.jsp Nicoloso R S, Rice C W and Amado T J C 2016 Kinetic to Saturation Model for Simulation of Soil Organic Carbon Increase to Steady State Soil Sci. Soc. Am. J. 80 147–56 NRCAN 2016 The atlas of Canada https://www.nrcan.gc.ca/maps-tools-and-publications/maps/atlas-canada/10784 Paradis M, Lévesque E and Boudreau S 2016 Greater effect of increasing shrub height on winter versus summer soil temperature Environ. Res. Lett. 11 085005 9 Environ. Res. Commun. 1 (2019) 091001 Payette S, Fortin MJ and Gamache I 2001 The subarctic forest–tundra: The structure of a biome in a changing climate BioScience 51 709-718 Provencher-Nolet L, Bernier M and Levesque E 2014 Quantiﬁcation of recent changes to the forest- tundra ecotone through numerical Payette S, Fortin MJ and Gamache I 2001 The subarctic forest–tundra: The structure of a biome in a changing climate BioScience 51 709-718 Provencher-Nolet L, Bernier M and Levesque E 2014 Quantiﬁcation of recent changes to the forest- tundra ecotone through numerical analysis of aerial photographs Ecoscience 21 419–33 Ropars P and Boudreau S 2012 Shrub expansion at the forest-tundra ecotone: spatial heterogeneity linked to local topography Environ. Res. Lett. References 7 015501 Ropars P, Levesque E and Boudreau S 2015 How do climate and topography inﬂuence the greening of the forest-tundra ecotone in northern Quebec ?a dendrochronological analysis of Betula glandulosa Journal of Ecology 103 679–90 Ropars P, Levesque E and Boudreau S 2015 How do climate and topography inﬂuence the greening of the forest-tundra ecotone in northern Quebec ?a dendrochronological analysis of Betula glandulosa Journal of Ecology 103 679–90 Schuur E A G et al 2008 Vulnerability of permafrost carbon to climate change: Implications for the global carbon cycle Bioscience 58 701–14 Schuur E A G, Crummer K G, Vogel J G and Mack M C 2007 Plant species composition and productivity following permafrost thaw and thermokarst in Alaskan tundra Ecosystems 10 280–92 y p g p g y Schuur E A G, Crummer K G, Vogel J G and Mack M C 2007 Plant species composition and productivity following permafrost thaw and thermokarst in Alaskan tundra Ecosystems 10 280–92 Sistla S A, Moore J C, Simpson R T, Gough L, Shaver G R and Schimel J P 2013 Long-term warming restructures Arctic tundra without changing net soil carbon storage Nature 497 615–8 g g g Soil Classiﬁcation Working Group 1998 The Canadian System of soil Classiﬁcation (Ottawa: Agriculture and agri-food Canada publication) 1646ISBN 0-660-17404-9 Soil Classiﬁcation Working Group 1998 The Canadian System of soil Classiﬁcation (Ottawa: Agriculture and agri-food Canada publication) 1646ISBN 0-660-17404-9 p ) Sturm M, Schimel J, Michaelson G, Welker J M, Oberbauer S F, Liston G E, Fahnestock J and Romanovsky V E 2005 Winter biological processes could help convert arctic tundra to shrubland Bioscience 55 17–26 p ) Sturm M, Schimel J, Michaelson G, Welker J M, Oberbauer S F, Liston G E, Fahnestock J and Romanovsky V E 2005 Winter biological processes could help convert arctic tundra to shrubland Bioscience 55 17–26 p ) Sturm M, Schimel J, Michaelson G, Welker J M, Oberbauer S F, Liston G E, Fahnestock J and Romanovsky V E 2005 Winter biological ld h l d h bl d Sturm M, Schimel J, Michaelson G, Welker J M, Oberbauer S F, Liston G E, Fahnestock J and Romanovsky V E 2005 Winter biological processes could help convert arctic tundra to shrubland Bioscience 55 17–26 Todd-Brown K E O et al 2014 Changes in soil organic carbon storage predicted by Earth system models during the 21st century Biogeosciences 11 2341–56 Todd-Brown K E O et al 2014 Changes in soil organic carbon storage predicted by Earth system models during the 21st century Biogeoscienc 11 2341–56 blay B, Levesque E and Boudreau S 2012 Recent expansion of erect shrubs in the Low arctic: evidence from eastern Nunavik Environ. References Res. Lett. 7 Tremblay B, Levesque E and Boudreau S 2012 Recent expansion of erect shrubs in the Low arctic: evidence from eastern Nunavik Envi Res. Lett. 7 Trochim E D, Jorgenson M T, Prakash A and Kane D L 2016 Geomorphic and biophysical factors affecting water tracks in northern Alaska Earth and Space Science 3 123–41 Trochim E D, Jorgenson M T, Prakash A and Kane D L 2016 Geomorphic and biophysical factors affecting water tracks in northern Alaska Earth and Space Science 3 123–41 Turetsky M R et al 2019 Permafrost collapse is accelerating carbon release Nature 569 32–4 Turetsky M R et al 2019 Permafrost collapse is accelerating carbon release Nature 569 32–4 Walker D A et al 2005 The Circumpolar Arctic vegetation map Journal of Vegetation Science 16 267–82 Walker D A et al 2005 The Circumpolar Arctic vegetation map Journal of Vegetation Science 16 267–82 Walker D A et al 2005 The Circumpolar Arctic vegetation map Journal o l 2005 The Circumpolar Arctic vegetation map Journal of Vegetation S g Z Q, Yang Y, Li J L, Zhang C B, Chen Y Z, Wang K, Odeh I and Qi J G 2017 Simulation of terrestrial carbon equilibrium state by using a detachable carbon cycle scheme Ecol. Indic. 75 82–94 Wilmking M, Harden J and Tape K 2006 Effect of tree line advance on carbon storage in NW Alaska Journal of Geophysical Research- Biogeosciences 111 10 Wilmking M, Harden J and Tape K 2006 Effect of tree line advance on carbon storage in NW Alaska Journal of Geophysical Research- Biogeosciences 111 10 10 Supplementary material Table S1: Chemical characterization of soil horizon Bm. Iron and aluminum pyrophosphate- extractable content (mass fraction) and soil pH measured in a 0.01 M CaCl2 solution with a soil to water mass fraction of 1:2. Vegetation Site % Al % Fe % Corg pH CaCl2 Lichen L1 0.04 0.02 0.18 4.08 L2 0.10 0.04 0.20 4.02 Birch + Lichen BL2 0.04 0.01 0.28 4.06 BL6 0.30 0.13 0.30 4.24 BL7 0.11 0.06 0.65 4.09 Birch + moss BM1 0.05 0.02 0.11 4.22 BM2 0.08 0.04 0.45 3.98 1 1 Table S2. Detailed carbon data for the 16 plots studied in the upper Tasiapik Valley in September 2017. ailed carbon data for the 16 plots studied in the upper Tasiapik Valley in September 2017. Vegetation Site Lichen carbon stocks (kg m-2) Birch carbon stocks (kg m-2) Moss carbon stocks (kg m-2) SOC stocks (kg m-2) Surface carbon stocks (kg m-2) Lichen L1 1.13 ± 0.11 0 0 1.07 ± 0.13 2.20 ± 0.25 L2 0.46 ± 0.05 0 0 1.47 ± 0.17 1.93 ± 0.22 L3 0.94 ± 0.09 0 0 1.73 ± 0.20 2.67 ± 0.29 Birch + lichen BL1 0.21 ± 0.02 1.08 ± 0.15 0 3.12 ± 0.36 4.41 ± 0.53 BL2 0.52 ± 0.05 1.08 ± 0.13 0 3.16 ± 0.35 4.76 ± 0.54 BL3 0.59 ± 0.06 1.55 ± 0.19 0 3.53 ± 0.39 5.67 ± 0.64 BL4 0.18 ± 0.02 1.53 ± 0.19 0 4.18 ± 0.49 5.89 ± 0.69 BL5 0.63 ± 0.06 0.92 ± 0.15 0 4.7 ± 0.52 6.25 ± 0.74 BL6 0.82 ± 0.08 1.49 ± 0.24 0 5.25 ± 0.68 7.55 ± 1.00 BL7 0.33 ± 0.03 1.55 ± 0.21 0 5.25 ± 0.59 7.12 ± 0.83 BL8 0.73 ± 0.07 1.6 ± 0.23 0 5.45 ± 0.67 7.77 ± 0.97 Birch + moss BM1 0 1.88 ± 0.20 0.528 ± 0.016 3.01 ± 0.32 5.13 ± 0.67 BM2 0 3.27 ± 0.29 0.123 ± 0.004 3.96 ± 0.48 7.47 ± 0.92 BM3 0 2.02 ± 0.25 0.244 ± 0.007 4.25 ± 0.46 6.51 ± 0.86 BM4 0 2.87 ± 0.27 0.021 ± 0.001 7.91 ± 0.84 11.02 ± 1.26 BM5 0 1.7 ± 0.17 0.096 ± 0.003 12.10 ± 1.35 14.04 ± 1.66 2 2 Table S3. Detailed dendrochronological data for the 16 plots studied in the upper Tasiapik Valley in September 2017. Supplementary material Vegetation Site All measured birch ages Max. birch age (yrs) Age ratio (youngest /oldest) Canopy height (cm) Lichen L1 - - - 0 L2 - - - 0 L3 - - - 0 Birch + lichen BL1 21, 24, 35, 38 45 45 0.47 54 ± 5 BL2 31, 34, 46, 54, 79 79 0.39 44 ± 5 BL3 18, 25, 25, 27, 38, 46 46 0.39 46 ± 2 BL4 38, 39, 42, 51, 74 74 0.51 76 ± 6 BL5 20, 33, 42 42 0.48 31 ± 5 BL6 21, 21, 23, 23, 26, 26 26 0.81 35 ± 5 BL7 32, 51, 55 55 0.58 56 ± 4 BL8 36, 39, 44, 47 47 0.77 46 ± 6 Birch + moss BM1 39, 51, 64 64 0.61 62 ± 3 BM2 26, 31, 31, 48, 48, 50 50 0.52 94 ± 12 BM3 30, 33, 37, 41 41 0.73 104 ± 11 BM4 23, 28, 30, 33, 42 42 0.55 102 ± 7 BM5 21, 33, 36, 40, 40, 47 47 0.45 85 ± 11 Table S3. Detailed dendrochronological data for the 16 plots studied in the upper Tasiapik Valley in September 2017. 3 3 Figure S1. A typical example of the granulometric distribution of the soils found in the upper Tasiapik Valley. (a) the surface of the A horizon from a depth of 0 to 5 cm; (b) the surface of the B horizon at a depth of 25 to 30 cm. 4 4 Figure S2. Micrograph of a thin section of birch stems dyed in safranin and used for age determination. Scale bar: 500 µm Figure S2. Micrograph of a thin section of birch stems dyed in safranin and used for age determination. Scale bar: 500 µm Figure S2. Micrograph of a thin section of birch stems dyed in safranin and used for age determination. Scale bar: 500 µ 5 5
https://openalex.org/W1984083541	https://link.springer.com/content/pdf/10.1245/s10434-012-2387-2.pdf	English	null	Lymphatic Drainage Patterns from Melanomas on the Shoulder or Upper Trunk to Cervical Lymph Nodes and Implications for the Extent of Neck Dissection	Annals of surgical oncology	2,012	cc-by	4,869	ABSTRACT Conclusions. Only a minority of the patients with upper limb or trunk melanomas demonstrated lymphatic drainage to cervical lymph node basins, with preferential drainage to levels IV and Vb. Our current dissection protocol of levels II–V, with or without extension to the suboccipital region, in those patients with involved cervical sentinel nodes seems sufﬁcient. Purpose. To determine the incidence and pattern of cer- vical lymphatic drainage in patients with melanomas located on the upper limb or trunk, and to evaluate our current neck dissection protocol for those patients with a N? neck. Methods. Of 1192 melanoma patients who underwent sentinel node biopsy, 631 were selected with a primary tumor on the upper limb or trunk. All lymphoscintigrams, Methods. Of 1192 melanoma patients who underwent sentinel node biopsy, 631 were selected with a primary tumor on the upper limb or trunk. All lymphoscintigrams, SPECT/CT images and operative reports were reviewed to determine the exact locations of sentinel nodes visualized preoperatively and dissected during operation. Sentinel node biopsy is increasingly used to stage patients with melanoma and has evolved into a routine procedure for tumors originating at different sites of the body.1 The head and neck region is known for its complex lymphatic drainage, with melanomas draining to multiple, bilateral and sometimes even contralateral sentinel nodes.2–5 The high number of lymph nodes and lymphatic vessels in a small anatomic area may contribute to this phenomenon.6 However, despite the complex drainage pathways, head and neck melanomas usually seem to fol- low a somewhat predictable pattern, with preferred drainage to the parotid and levels I–V for lesions located anterior of the so-called watershed line and to the suboc- cipital region and level II–V for tumors located posterior of this line.7–9 On the basis of the primary tumor location, O’Brien et al.10 deﬁned the extent of several therapeutic neck dissections. SPECT/CT images and operative reports were reviewed to determine the exact locations of sentinel nodes visualized preoperatively and dissected during operation. p p y g p Results. Thirty-nine (6.2 %) of 631 patients with a mela- noma on the upper limb or trunk showing cervical lymph node drainage were identiﬁed. In 34 (87 %) of 39 patients, sentinel nodes were excised from level IV or Vb, and in 30 of those 39 patients simultaneous from the axilla. Lymphatic Drainage Patterns from Melanomas on the Shoulder or Upper Trunk to Cervical Lymph Nodes and Implications for the Extent of Neck Dissection Hidde J. Veenstra, MD1, W. Martin C. Klop, MD, PhD1,2, Maarten J. Speijers, MD4, Peter J. F. M Omgo E. Nieweg, MD, PhD3, Harald J. Hoekstra, MD, PhD4, and Alfons J. M. Balm, MD, PhD Hidde J. Veenstra, MD1, W. Martin C. Klop, MD, PhD1,2, Maarten J. Speijers, MD4, Peter J. F. M. Lohuis, MD, PhD1,2, Omgo E. Nieweg, MD, PhD3, Harald J. Hoekstra, MD, PhD4, and Alfons J. M. Balm, MD, PhD1,2 1Department of Head and Neck Surgery and Oncology, The Netherlands Cancer Institute – Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands; 2Department of Otolaryngology, Academic Medical Centre, Amsterdam, The Netherlands; 3Department of Surgery, The Netherlands Cancer Institute—Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands; 4Department of Surgery, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands Ann Surg Oncol (2012) 19:3906–3912 DOI 10.1245/s10434-012-2387-2 Ann Surg Oncol (2012) 19:3906–3912 DOI 10.1245/s10434-012-2387-2 ORIGINAL ARTICLE – MELANOMAS The Author(s) 2012. This article is published with open access at Springerlink.com First Received: 21 August 2011; Published Online: 11 May 2012 H. J. Veenstra, MD e-mail: h.veenstra@nki.nl Surgical and Pathological Procedures Lymphoscintigrams and SPECT/CT images were dis- cussed with the surgeon before the sentinel nodes were excised. The next day, patent blue dye (Laboratoire Guerbet, Aulnay-Sous-Bois, France) was administered intradermally in a mean volume of 1.0 ml, completely surrounding the scar of the reexcision site. Ten minutes after injection all nodal basins identiﬁed by lymphoscin- tigraphy were explored surgically through limited incisions. Surgical dissection was guided by a handheld gamma ray detection probe (Neoprobe 1000 and 1500, Johnson & Johnson Medical, Hamburg, Germany) and by looking for blue-stained afferent lymphatic vessels that led to blue stained sentinel lymph nodes. Once the sentinel node had been excised, the probe was used to search the resection bed to ensure that there were no residual sentinel nodes. The search for additional sentinel nodes was ceased if there was no residual activity or if this was less than ten percent of the most active node in the same area. After sentinel node biopsy, wide local excision was performed of the primary melanoma site with a 1 or 2 cm margin, depending on the Breslow thickness. All sentinel nodes were formalin-ﬁxed, bisected, parafﬁn-embedded, and cut at a minimum of six levels at 50 to 150 lm intervals. Pathologic evaluation included hematoxylin and eosin and immunohistochemical staining (S-100 and MART-1). With the exception of one patient harboring a small tumor bur- den in the sentinel node (Starz classiﬁcation I), patients with a tumor-positive sentinel node underwent therapeutic node dissection.16–18 Patients received a level I–III axillary dissection for tumor-positive sentinel nodes located in the axilla. In case a cervical positive sentinel node was col- lected, derived from an anterior shoulder or anterior trunk melanoma, a selective neck dissection of the levels II–V was performed. The dissection was extended to the sub- occipital region in case a tumor-positive sentinel node was found in level II–V or the suboccipital region derived from a melanoma on the posterior shoulder or posterior trunk. Extension to the suboccipital region consists of the ‘so-called’ hockey-stick incision more posteriorly for a posterolateral neck dissection (levels II–V including the suboccipital region). During suboccipital dissection a thin layer of fat and lymph nodes between the ligamentum ABSTRACT In the remaining ﬁve patients (13 %), sentinel nodes were col- lected from level IIb, level III or the suboccipital region. All collected sentinel nodes were located in the intended dissection area for N? patients. Thirteen patients (33 %) had a total of 22 tumor-positive sentinel nodes in either the axilla (n = 10), level IV (n = 2), Vb (n = 9) or suboc- cipital (n = 1). Occasionally, primary melanomas located on the upper limb or trunk drain to cervical lymph nodes, but studies on that subject are scarce.11–13 Uren et al. presented 1735 patients with melanomas located on the upper limb or upper trunk where lymphatic drainage was visualized to a total number of 321 cervical sentinel nodes.13 3907 Melanomas on the Shoulder or Upper Trunk A completion neck dissection in patients with a histologi- cally proven tumor-positive sentinel node originating from a melanoma on the upper limb or trunk is usually advised. The extent of therapeutic neck dissection hereby is based on the expected drainage pathways and generally includes levels II–V (with or without the suboccipital region). However, in these speciﬁc cases the actual extent of the neck dissection is still a matter of ongoing debate, because scientiﬁc evidence for this approach is lacking. Therefore, the aim of this study is to determine the incidence of lymphatic drainage to cervical lymph nodes derived from melanomas located on the upper limb or trunk in our patient group, and to evaluate our current neck dissection protocol of the levels II–V (with or without the suboccipital region) in these patients with a N? neck. A completion neck dissection in patients with a histologi- cally proven tumor-positive sentinel node originating from a melanoma on the upper limb or trunk is usually advised. conventional images were difﬁcult to interpret and for research purposes.14 SPECT/CT was additionally per- formed in eight patients, all treated in the Netherlands Cancer Institute. A sentinel node was deﬁned as a lymph node upon which the primary tumor drains directly.15 Both conventional lymphoscintigrams and SPECT/CT images were used to mark the locations of the sentinel nodes on the skin with indelible ink. Patients From 1192 unselected melanoma patients who under- went sentinel node biopsy at the Netherlands Cancer Institute or the University Medical Centre Groningen between December 1991 and July 2010 a total of 631 patients with a primary melanoma on the upper limb or trunk were identiﬁed. The anterior trunk melanomas were located on the trunk from the umbilicus up to the clavicle and on the anterior side of the shoulder, which is divided by an imaginary line over the deltoid region. The posterior trunk melanomas were located on the trunk from the level of the umbilicus up to vertebra C7 and on the posterior side of the shoulder posterior from the imaginary line. Sentinel node biopsy was carried out if the primary tumor had a Breslow thickness of at least 1 mm or a Clark level IV. Lymphoscintigraphy and SPECT/CT Both conventional lymphoscintigraphy and SPECT/CT visualized a total of 112 sentinel nodes in 39 patients. Forty-four sentinel nodes were depicted in the axilla, 67 sentinel nodes in cervical lymph node basins and one sentinel node was seen on the lateral border of the left scapula. In 31 (80 %) of 39 patients lymphatic drainage from the shoulder or upper trunk was seen to level Vb, of which 27 patients had simultaneous drainage to the axilla. Two patients (5 %) showed lymphatic drainage to level IV of which one also to the axilla. In the remaining six patients (15 %), lymphatic drainage was depicted to the parotid gland, level Va, level III, level IIa, level IIb or suboccipital. nuchae (medial border) and the fascia of trapezius muscle (deep border) is removed. The supreme nuchal line and external occipital protuberance represent the cranial border of the dissection and an imaginary line connecting the two mastoid processes the inferior border. The suboccipital dissection is performed in continuity with the selective neck dissection (II–V).19,20 Dissection of both the axilla and neck was performed if metastases were present at each site in the same patient. All lymph nodes collected from regional node dissection specimens were examined in 4.0 mm sections stained with hematoxylin and eosin. Sentinel Node Biopsy In 39 patients a total of 121 sentinel nodes were excised (Table 2). Forty-four sentinel nodes were collected from the axilla, 76 sentinel nodes from cervical lymph node basins and one sentinel node from the lateral border of the left scapula (Figs. 1, 2). In 32 (82 %) of 39 patients, a total of 65 sentinel nodes were excised from level Vb. In 28 of those 39 patients, 44 sentinel nodes were also excised from the axilla or lateral border of the left scapula. In two patients (5 %), ﬁve sentinel nodes were collected from level IV, of which one also had a sentinel node in the axilla. In the remaining ﬁve patients (13 %), two sentinel nodes were excised from level IIb, one sentinel node from level III and three sentinel nodes from the suboccipital region. In two patients, these suboccipital nodes were not visualized preoperatively, but the handheld gamma probe detected them. All collected suboccipital sentinel nodes were found in patients with a primary tumor located on the upper posterior trunk or posterior shoulder. Not all Imaging On the day before operation, technetium-99m-labeled nanocolloid (Nanocoll, Amersham Cygne, Eindhoven, The Netherlands) was injected intradermally around the biopsy site in a mean volume of 0.4 ml and a mean dosage of 63 MBq (2 mCi). The mean time interval between injec- tion of the technetium-99m-labeled nanocolloid and the actual operation is approximately 18 hours. Static images were performed and were preceded by a dynamic study of ten minutes. Both anterior and lateral images were rou- tinely made. Immediate and three-hour postinjection lymphoscintigrams were evaluated with regard to the location, number and sequence of appearance of sentinel nodes. At The Netherlands Cancer Institute, hybrid single- photon emission computed tomography with CT (SPECT/ CT) was introduced in 2006 and was performed when H. J. Veenstra et al. 3908 TABLE 1 Characteristics of 39 patients with a melanoma on the upper limb or trunk Characteristic Value Age (y), mean (range) 56 (15–89) Gender (M:F) 27:12 Primary tumor location Anterior shoulder 3 Posterior shoulder 12 Anterior trunk 5 Posterior trunk 19 Breslow thickness (mm), median (range) 2.9 (1.0–12) Follow-up (mo), median (range) 54 (4–162) identiﬁed with lymphatic drainage to at least one lymph node basin in the head and neck area. All melanomas on the upper limb were located on the anterior or posterior shoulder. Overall patient characteristics are shown in Table 1. Data Analysis To determine the precise location of sentinel nodes visualized preoperatively and collected during operation all lymphoscintigrams, SPECT/CT images and operative reports were retrospectively reviewed. All patients were monitored with a median follow-up of 54 months. Patients Thirty-nine (6.2 %) of 631 patients with a primary melanoma located either on the upper limb or trunk were TABLE 2 Locations of sentinel nodes excised Melanoma site No. of patients Location Axilla Level IIa Level IIb Level III Level IV Level Va Level Vb Suboccipital Other Anterior shoulder 3 4 (1) – – – – – 5 (1) – Anterior trunk 5 3 (1) – – 1 (0) 2 (0) – 9 (0) – Posterior shoulder 12 15 (2) – 1 (0) – – – 20 (5) 1 (1) Posterior trunk 19 22 (6) – 1 (0) – 3 (2) – 31 (3) 2 (0) Scapula Total 39 44 (10) – 2 (0) 1 (0) 5 (2) – 65 (9) 3 (1) 1 (0) The total number of tumor-positive sentinel nodes found is indicated in parentheses 3909 Melanomas on the Shoulder or Upper Trunk suboccipital (Table 3). One patient did not receive any further treatment, because only a minor tumor burden (Starz I) was found in the involved sentinel node.16–18 In two (17 %) of 12 patients who underwent a regional lymph node dissection, additional tumor-positive lymph nodes were collected. The ﬁrst patient with a melanoma on the posterior shoulder underwent a posterolateral dissection with extension to the suboccipital region and one additional tumor-positive lymph node was found in level Vb. The second patient with a melanoma on the scapula underwent both an axillary dissection and a posterolateral dissection with extension to the suboccipital region. Two additional tumor-positive lymph nodes were excised from the axilla and two additional tumor-positive lymph nodes from level Vb (Table 3). Both patients received adjuvant radiotherapy. Follow-up In gray the dissection area is visualized in case a tumor- positive sentinel node is collected preoperatively visualized sentinel nodes were collected. A sentinel node in the parotid gland, in level Va and in level IIa was not found during operation, because the handheld gamma probe did not detect any radioactivity. Follow-up FIG. 1 Locations of eight primary melanomas on the anterior shoulder and anterior trunk and 24 sentinel nodes draining these primary lesions. In gray the dissection area is visualized in case a tumor-positive sentinel node is collected Three of the 26 patients with a tumor-negative sentinel node developed a recurrence, with an average follow-up of 44 (range 19–162) months. The ﬁrst patient, who had a primary tumor on the left shoulder, developed a subcuta- neous metastasis on the left wrist and simultaneously pulmonary metastases after 107 months and is currently being treated with systemic therapy. The second patient with a primary melanoma on the upper posterior trunk developed pulmonary metastases after 32 months and died due to these metastases 3 months later. The third patient with a primary tumor on the right shoulder and tumor- negative sentinel nodes removed from level IIb and Vb developed a lymph node recurrence after 5 months in the previously biopsied level Vb and this procedure was thus considered falsely negative. After therapeutic neck dis- section no other recurrences developed in the next 5 years. Eight of 12 tumor-positive sentinel node patients who underwent therapeutic dissection remained disease free, with an average follow-up of 59 months (range, 17–150). Four of 12 patients with a tumor-positive sentinel node, who received therapeutic lymph node dissection, devel- oped a local recurrence. In none of these latter four patients additional tumor-positive lymph nodes were found in the dissection specimen. The ﬁrst patient developed extensive distant metastases after 31 months and died from these recurrences. The second patient developed brain and oro- pharynx metastases after 24 months and is still alive 4 months later. In the third patient liver metastases were found 9 months after neck dissection and this patient died 3 months later due to these metastases. The fourth devel- oped patient a local recurrence in the scar of the primary tumor excision 48 months after sentinel node biopsy, which was reexcised and no further recurrences developed. The patient with a tumor-positive sentinel node who FIG. 2 Locations of 31 primary melanomas on the posterior shoulder and posterior trunk and 97 sentinel nodes draining these primary lesions. In gray the dissection area is visualized in case a tumor- positive sentinel node is collected FIG. 2 Locations of 31 primary melanomas on the posterior shoulder and posterior trunk and 97 sentinel nodes draining these primary lesions. Pathology In 13 (33 %) of the 39 patients, a total of 22 tumor- positive sentinel nodes were excised. The involved sentinel nodes were located in the axilla, level IV, level Vb or 3910 H. J. Veenstra et al. TABLE 3 Data of 12 patients who underwent therapeutic lymph node dissection Patient no. Tumor location Location of positive sentinel nodes Type of therapeutic node dissection Positive lymph nodes in dissection specimen 1 Posterior trunk Axilla Axillary dissection 0/15 2 Posterior shoulder Axilla; level Vb Axillary dissection; posterolateral dissection level II–V including the suboccipital region 0/17; 0/13 3 Anterior shoulder Level Vb Axillary dissection; selective dissection level II–V 0/35 4 Posterior trunk Axilla; suboccipital Axillary dissection; posterolateral dissection level II–V including the suboccipital region 0/17; 0/19 5 Posterior trunk Axilla Axillary dissection 0/11 6 Posterior trunk Axilla; level Vb Axillary dissection; posterolateral dissection level II–V including the suboccipital region 2/16; 2/53 (level Vb) 7 Posterior shoulder Axilla; level Vb Axillary dissection; selective dissection level II–V 0/42; 0/84 8 Posterior shoulder Level Vb Posterolateral dissection level II–V including the suboccipital region 0/13 9 Posterior shoulder Axilla; level Vb Axillary dissection; posterolateral dissection level II–V including the suboccipital region 0/14; 0/21 10 Posterior shoulder Level Vb Posterolateral dissection level II–V including the suboccipital region 1/37 (level Vb) 11 Posterior trunk Level IV Posterolateral dissection level II–V including the suboccipital region 0/32 12 Posterior trunk Axilla; level Vb Axillary dissection; posterolateral dissection level II–V including the suboccipital region 0/14; 0/16 et al. mapped lymphatic drainage of 5239 melanoma patients.12,21 In total, 2406 patients had melanomas located on the trunk (above the waist or umbilicus) or upper limb. Although the incidence of patients with cervical lymph drainage was not reported, it could be calculated that a total of 569 sentinel nodes were visualized in cervical lymph node basins. A total of 508 (89 %) of 569 sentinel nodes were visualized in level Vb (n = 330) and level Va (n = 178), 30 sentinel nodes (5.5 %) in level IV, 15 sen- tinel nodes (2.5 %) in level III and 11 sentinel nodes (2 %) in level II. In contrast to our ﬁndings, two sentinel nodes (0.5 %) were visualized in level I derived from melanomas located just infraclavicularly. Three sentinel nodes (0.5 %) were found in the suboccipital region, originating from melanomas located on the upper posterior trunk and pos- terior shoulder. Pathology In the present study 76 cervical sentinel nodes were identiﬁed in the 39 included patients. It has to be emphasized that the results of Reynold’s study were based on preoperative lymphoscintigrams without conﬁrmation by additional dissection, as we did. The deﬁnite locations of the sentinel node might therefore be different, as has been shown by Jansen et al.5 In their study, a discrepancy between sentinel node locations depicted by preoperative imaging and collected during operation was shown in 18 (60 %) of 30 patients. The current study demonstrated this received no further treatment (Starz I) remained disease free with a follow-up of 25 months. No lymph node recurrences developed beyond the dissected neck levels in patients who underwent a neck dissection. DISCUSSION This study demonstrates 39 patients with a melanoma on the shoulder or upper trunk with drainage to at least one cervical lymph node basin. Thirty-four (87 %) of these 39 patients had lymphatic drainage to levels IV and Vb, and the axilla. In the remaining ﬁve patients (13 %), sentinel nodes were collected from level IIb, level III or from the suboccipital region. As our current dissection protocol for patients with a tumor-positive sentinel node consist of level II–V for melanomas located on the anterior shoulder and anterior trunk and of level II–V with extension to the suboccipital region for melanomas on the posterior shoul- der and posterior trunk, 100 % of the collected cervical sentinel nodes are covered. The lymphatic drainage pattern from primary melano- mas located on the shoulder or trunk to cervical lymph node basins can’t be compared with the lymphatic drainage pattern of head and neck melanomas. So far, no studies have been published dealing with this speciﬁc subject. Reynolds Melanomas on the Shoulder or Upper Trunk 3911 inconsistency in 6 (15 %) of 39 patients. For example, in one patient a sentinel node was depicted in the parotid preoperatively, but was collected from level IIb. increased risk of ﬂap necrosis (a thin skin ﬂap has to be elevated because the suboccipital nodes are mainly located just under the skin) makes it a more delicate procedure. The suboccipital dissection can be omitted if the risk of morbidity is considered to exceed the potential beneﬁt. Strict follow-up of this speciﬁc area by ultrasound is then recommended. Melanoma cells detected in cervical sentinel nodes derived from tumors on the shoulder or trunk are consid- ered as regional metastases and therefore, outside a clinical trial, completion neck dissection is advised. The extent of neck dissection, as described in the methods section, is based on the principle that only those lymph node basins considered at risk for metastases should be removed. A total of twelve patients with a primary melanoma located on the shoulder or trunk underwent completion lymph node dissection (Table 3) and in two patients (17 %) additional tumor-positive lymph nodes were found in level Vb and the axilla. CONFLICT OF INTEREST None. OPEN ACCESS This article is distributed under the terms of the Creative Commons Attribution License which permits any use, dis- tribution, and reproduction in any medium, provided the original author(s) and the source are credited. DISCUSSION The percentage of additional melanoma containing lymph nodes in the head and neck region is comparable with the percentages reported by other authors of 10–30 %.8,22–24 The safety of our current neck dissection protocol for patients with tumor-positive cervical sentinel nodes derived from shoulder or trunk melanoma is also conﬁrmed by the lack of regional recurrences developing outside the operation ﬁeld. This even could raise the question if a more selective approach guided by the lymph node drainage pattern is appropriate, for example, by including only the ﬁrst- and second echelon nodes in the dissection area. Comparable to Reynolds et al., we dem- onstrated that the vast majority of sentinel nodes are located in level IV and Vb. Although only one tumor- positive sentinel node was found outside these levels in the suboccipital region, lymphatic drainage was also shown to level II and III. On the basis of these results, and the knowledge that anatomic boundaries of the cranial border of level IV and Vb are hard to deﬁne, it seems adequate to continue our current neck dissection protocol including level II–V (with or without extension to the suboccipital region) in the N? patients with primary melanomas located on the shoulder or trunk.25 In conclusion, in our study, 39 patients are described with a primary melanoma located on the shoulder or upper trunk with lymphatic drainage to cervical lymph node basins. The preferred drainage pattern was to level IV, level Vb and the axilla, but in some cases lymphatic ﬂow was seen up to level II, level III or the suboccipital region. Our current dissection protocol of the levels II–V, in case a tumor-positive cervical sentinel node is collected in patients with a melanoma on the anterior shoulder or anterior upper trunk, appears to be sufﬁcient. For patients with an involved cervical sentinel node originating from a primary melanoma on the posterior upper trunk or posterior shoulder, a level II–V dissection with extension to the suboccipital region is advised. REFERENCES 1. Balch CM, Morton DL, Gershenwald JE, et al. Sentinel node biopsy and standard of care for melanoma. J Am Acad Dermatol. 2009;60:872–5. 2. De Wilt JH, Thompson JF, Uren RF, et al. Correlation between preoperative lymphoscintigraphy and metastatic nodal disease sites in 362 patients with cutaneous melanomas of the head and neck. Ann Surg. 2004;239:544–52. 3. O’Brien CJ, Uren RF, Thompson JF, et al. Prediction of potential metastatic sites in cutaneous head and neck melanoma using lymphoscintigraphy. Am J Surg. 1995;170:461–6. As in our study, Reynolds et al.12 showed lymphatic drainage to the suboccipital region in three patients with a melanoma located on the upper posterior trunk or posterior shoulder. In the current study one patient had a tumor- positive lymph node in the suboccipital area and underwent a level II–V dissection with extension to the suboccipital region. There were no involved nonsentinel nodes in the suboccipital dissection specimens, but drainage to the suboccipital region was seen in three of the 31 patients (10 %) with a posteriorly located melanoma. This dem- onstrates that the suboccipital region remains a potential drainage site for metastatic cells. In our opinion, this observation justiﬁes treatment of the suboccipital lymph node basin for melanomas of the posterior shoulder and upper back in the N? neck. Nevertheless the absence of a clear surgical boundary of this region and relatively 4. Chao C, Wong SL, Edwards MJ, et al. Sentinel lymph node biopsy for head and neck melanomas. Ann Surg Oncol. 2003; 10:21–6. 5. Jansen L, Koops HS, Nieweg OE, et al. Sentinel node biopsy for melanoma in the head and neck region. Head Neck. 2000;22: 27–33. 6. Pan WR, Suami H, Taylor GI. Lymphatic drainage of the superﬁcial tissues of the head and neck: anatomical study and clinical implications. Plast Reconstr Surg. 2008;121:1614–24. 7. Klop WM, Veenstra HJ, Vermeeren L, et al. Assessment of lymphatic drainage patterns and implications for the extent of neck dissection in head and neck melanoma patients. J Surg Oncol. 2011;103:756–60. 8. Lin D, Franc BL, Kashani-Sabet M. Lymphatic drainage patterns of head and neck cutaneous melanoma observed on lympho- scintigraphy and sentinel lymph node biopsy. Head Neck. 2006; 28:249–55. 3912 H. J. Veenstra et al. 9. Pathak I, O’Brien CJ, Petersen-Schaeffer K, et al. Do nodal metastases from cutaneous melanoma of the head and neck fol- low a clinically predictable pattern? Head Neck. 2001;23:785–90. 18. REFERENCES Starz H, Siedlecki K, Balda BR. Sentinel lymphonodectomy and s-classiﬁcation: a successful strategy for better prediction and improvement of outcome of melanoma. Ann Surg Oncol. 2004; 11:162S–8S. 10. O’Brien CJ, Petersen-Schaefer K, Ruark D, et al. Radical, mod- iﬁed, and selective neck dissection for cutaneous malignant melanoma. Head Neck. 1995;17:232–41. 19. Diaz EM Jr, Austin JR, Burke LI, Goepfert H. The posterolateral neck dissection. Technique and results. Arch Otolaryngol Head Neck Surg. 1996;122:477–80. 11. Wagner JD, Park HM, Coleman JJ III, et al. Cervical sentinel lymph node biopsy for melanomas of the head and neck and upper thorax. Arch Otolaryngol Head Neck Surg. 2000;126:313–21. 20. Thiel W. Photographic atlas of practical anatomy II. New York: Springer; 1999. p. 104–5. 21. Reynolds HM, Smith NP, Uren RF, et al. Three-dimensional visualization of skin lymphatic drainage patterns of the head and neck. Head Neck. 2009;31:1316–25. 12. Reynolds HM, Dunbar PR, Uren RF, et al. Three-dimensional visualisation of lymphatic drainage patterns in patients with cutaneous melanoma. Lancet Oncol. 2007;8:806–12. 13. Uren RF, Howman-Giles R, Thompson JF. Patterns of lymphatic drainage from the skin in patients with melanoma. J Nucl Med. 2003;44:570–82. 22. De Rosa N, Lyman GH, Silbermins D, et al. Sentinel node biopsy for head and neck melanoma: a systematic review. Otolaryngol Head Neck Surg. 2011;145:375–82. 23. Gomez-Rivera F, Santillan A, McMurphey AB, et al. Sentinel node biopsy in patients with cutaneous melanoma of the head and neck: recurrence and survival study. Head Neck. 2008;30:1284– 94. 14. Van der Ploeg IM, Valde´s Olmos RA, Kroon BBR, et al. The yield of SPECT/CT for anatomical lymphatic mapping in patients with melanoma. Ann Surg Oncol. 2009;16:1537–42. g 15. Nieweg OE, Tanis PJ, Kroon BBR. The deﬁnition of a sentinel node. Ann Surg Oncol. 2001;8:538–41. 24. Carlson GW, Murray DR, Lyles RH, et al. Sentinel lymph node biopsy in the management of cutaneous head and neck mela- noma. Plast Reconstr Surg. 2005;115:721–8. 16. Van der Ploeg IM, Kroon BBR, Antonini N, et al. Is completion lymph node dissection needed in case of minimal melanoma metastasis in the sentinel node? Ann Surg. 2009;249:1003–7. 25. Lohuis PJ, Klop WM, Tan IB, et al. Effectiveness of therapeutic (N1, N2) selective neck dissection (levels II to V) in patients with laryngeal and hypopharyngeal squamous cell carcinoma. Am J Surg. 2004;187:295–9. 17. Van der Ploeg IM, Kroon BBR, Antonini N, et al. REFERENCES Comparison of three micromorphometric pathology classiﬁcations of melanoma metastases in the sentinel node. Ann Surg. 2009;250:301–4.
https://openalex.org/W3164849643	https://www.frontiersin.org/articles/10.3389/fdata.2021.622106/pdf	English	null	Search Engine Gender Bias	Frontiers in big data	2,021	cc-by	11,664	Search Engine Gender Bias This article discusses possible search engine page rank biases as a consequence of search engine proﬁle information. After describing search engine biases, their causes, and their ethical implications, we present data about the Google search engine (GSE) and DuckDuckGo (DDG) for which only the ﬁrst uses proﬁle data for the production of page ranks. We analyze 408 search engine screen prints of 102 volunteers (53 male and 49 female) on queries for job search and political participation. For job searches via GSE, we ﬁnd a bias toward stereotypically “female” jobs for women but also for men, although the bias is signiﬁcantly stronger for women. For political participation, the bias of GSE is toward more powerful positions. Contrary to our hypothesis, this bias is even stronger for women than for men. Our analysis of DDG does not give statistically signiﬁcant page rank differences for male and female users. We, therefore, conclude that GSE’s personal proﬁling is not reinforcing a gender stereotype. Although no gender differences in page ranks was found for DDG, DDG usage in general gave a bias toward “male-dominant” vacancies for both men and women. We, therefore, believe that search engine page ranks are not biased by proﬁle ranking algorithms, but that page rank biases may be caused by many other factors in the search engine’s value chain. We propose ten search engine bias factors with virtue ethical implications for further research. INTRODUCTION Search engines enable their users to receive relevant search results in the enormous mass of data on the Web. The most used and best known search engine is the Google search engine, named GSE in the rest of this article. By collecting data about its users, GSE creates personal proﬁles of its users for targeted advertising service (Thurman and Schifferes, 2012). With the same process, GSE also personalizes search outcomes and page ranks for optimizing a search outcomes relevance experience for its users (Rogers 2013; De Corniere and Taylor, 2014). Personal proﬁle processing, as part of search query processing, may result in a high use value experience but also can keep important less relevant information away from its user (Helberger et al., 2018). This “ﬁlter bubble effect” (Pariser 2011) also comes with beneﬁts, as people lack the capacity to process all the zettabytes of information available on the Internet. This also implies that people with different proﬁles who enter the same query into the GSE receive a different list of relevant Web pages. This ability of semantic search enables GSE to differentiate the meaning of ambiguous terms in a query. For example, while a lawyer would refer to the keyword “code” as a set of rules or the law, an information technology specialist may be more interested in “code” referring to software, and thus a proﬁle aware search engine will serve both differently. This subjective relevance increasing process may, in a subtle way, result in inﬂuencing users to certain viewpoints and opinions (Epstein and Robertson 2015), which is a process also named nudging (Gal et al., 2020). Correspondence: Fons Wijnhoven a.b.j.m.wijnhoven@utwente.nl Specialty section: This article was submitted to Data Mining and Management, a section of the journal Frontiers in Big Data Received: 27 October 2020 Accepted: 04 May 2021 Published: 26 May 2021 Edited by: Mascha Kurpicz-Briki, Bern University of Applied Sciences, Switzerland Reviewed by: Jahna Otterbacher, Open University of Cyprus, Cyprus Jason Millar, University of Ottawa, Canada Reviewed by: Jahna Otterbacher, Open University of Cyprus, Cyprus Jason Millar, University of Ottawa, Canada Correspondence: Fons Wijnhoven a.b.j.m.wijnhoven@utwente.nl ORIGINAL RESEARCH published: 26 May 2021 doi: 10.3389/fdata.2021.622106 Keywords: personalization, Google, DuckDuckGo, ﬁlter bubble, gender bias, job search, political participation search BACKGROUND, THEORY, AND HYPOTHESES Personalized search results are customized search results based on the digital proﬁle of an individual user. When using a personal proﬁle aware search engine such as GSE or Bing, search results will differ depending on who is supplying the query to the search engine. This proﬁle is supposed to represent the interests of the individual user (Harvey et al., 2013). GSE’s updated privacy policy of 2012 states that information is being shared on all sorts of Google tailors, which include Gmail, YouTube, and Google Maps for further proﬁling of the search engine user (Voss 2014). Also, search location indicators (Hannak et al., 2013) and browsing histories can combine with user attributes to produce personalized outcomes (Goel et al., 2012). A manipulation effect occurs if the search engine can inﬂuence users’ attitudes, beliefs, or decisions based on the order of the search engine results. A manipulation effect may result from search results that are biased toward one view. By providing one- sided search results, GSE could either reinforce an existing opinion or it may turn an undecided person to be nudged toward one opinion direction (Epstein and Robertson 2015; Wijnhoven and Brinkhuis 2015). There is a substantial volume of research outcomes regarding the difference in the use of the Internet between men and women (Colley and Maltby 2008; Yom-Tov and Elad, 2019), but there is little knowledge on what the role of the medium is in this story. Hargittai and Shafer (2006) suggest that the supply-side of content—due to its structure and presentation—is in itself male-biased. This would imply that gender bias of search engine page ranks is not necessarily the result of personalization but could be the result of biased supplies of information on the Internet as well. Unfortunately, it is difﬁcult to identify what GSE knows about you, we only know that personal proﬁle processing is a key part of GSE’s proprietary way of delivering relevant content to its user (Bozdag 2013; Rogers 2013). Such a personal proﬁling algorithm is lacking in privacy-aware search engine DuckDuckGo (DDG), and thus comparing page ranks of male and female users of both using GSE and DDG for the same queries would be able to determine if the search engine is the cause of page rank bias or if alternative reasons exist for page rank bias. Citation: Wijnhoven F and van Haren J (2021) Search Engine Gender Bias. Front. Big Data 4:622106. doi: 10.3389/fdata.2021.622106 Wijnhoven F and van Haren J (2021) Search Engine Gender Bias. Front. Big Data 4:622106. doi: 10.3389/fdata.2021.622106 May 2021 \| Volume 4 \| Article 622106 1 Frontiers in Big Data \| www.frontiersin.org Wijnhoven and van Haren Search Engine Bias A common nudge in computer environments is the setting of defaults, which are preset courses of action that take effect if nothing is speciﬁed by the decision maker (Cronqvist et al., 2018). But nudging can also take more elaborate and sinister forms. For example, in 2017, it became known that advertisers on Facebook could target advertisements at teenagers during moments of psychological vulnerability. This kind of nudging can also be employed by organizations through the use of algorithmic systems to encourage “appropriate” behaviors. For instance, Kaptein et al. (2015) have shown that emails and Web site usage nudges can move people to more healthy behavior. But these practices can also be manipulative and ethically questionable when they are aimed at covertly subverting people’s decision-making capacity by exploiting their psychological, cognitive, or emotional vulnerabilities to change their beliefs, thoughts, or behaviors. Such manipulative practices limit people’s capacity for voluntary action and learning. Nudging may not only be unethical from a utilitarian perspective (Tavani 2012), that is, the diminishing value of what is received for the search engine user, but also from a virtue ethics perspective as it disallows the user to identify the source of bias, learn from that, and reduces the freedom of voluntary action of what to believe or not (Gal et al., 2020). Following Gal et al. (2020), we choose for a virtue ethics approach for the analysis of search engine bias. Virtue ethics, based on the philosophical work of Aristotle, highlights personal characteristics in determining the ethical nature of individuals and their actions. Virtue ethics focus on the virtuous agent rather than on the right actions to reach certain values (i.e., the utilitarian approach) or on what anyone should do given certain rules (i.e., the deontological approach to ethics). reproduce, or reinforce biases with virtue ethical implications for other social segments. BACKGROUND, THEORY, AND HYPOTHESES For knowing if this bias has virtue ethical consequence, we will take queries that may have important consequences for women to learn and develop themselves. The ranking of search results is especially important because users tend to pay more attention to the results that rank higher on the search results page (Bozdag 2013). The actual impact of search engine and page rank biases is also named as the search engine manipulation effect. A slight manipulation of a page rank can have an impact on the outcome of elections by providing search results that are in favor of one candidate (Epstein and Robertson, 2015). This eventually causes a user to be less likely to discover new topics on the Web or to see documents with different views on a topic that contradicts his or her values (Wilson et al., 2012; Burger et al., 2016). The purpose of this study was to gain further insights into possible search engine bias toward one-sided search results with ethical implications. This leads to the research question: To what extent does the Google search engine creates a biased page rank for its users with negative virtual ethical consequences? To answer this question, this article focuses on search engine gender bias and its virtue ethical implications for women. Besides the commitment of the authors to equal opportunities of women in our society, gender characteristics are one of the most common and easily traced search engine proﬁle features (Kosinski et al., 2013). A search engine’s gender bias is the inﬂuence of gender features in a search engine proﬁle on gender bias of a search engine’s outcomes and page rank (Lopes et al., 2016). The insight from a gender-bias study can produce generalizable understandings on how search engines may produce, This section ﬁrst goes into more details about causes and consequences of search engine bias. Then it will set out the concerns regarding gender inequalities and our hypotheses. Frontiers in Big Data \| www.frontiersin.org Jobs According to Scherer (2004), the labor market entry and a successful transition from school to work are of crucial importance for subsequent career chances, it is a good start to look at the possibly different opportunities here for men and women. Deﬂoor et al. (2015) found that the quality of a person’s ﬁrst job is largely dependent on personal effort, although circumstances, for example, gender do have a considerable inﬂuence on these efforts. This would, therefore, suggest that a difference of jobs between men and women is due to a difference in effort spent between men and women when searching for their ﬁrst job. Equality of opportunity besides efforts is also highly debated in relation to gender. Equality of opportunity means that outcomes experienced by a population depend only on factors for which persons can be considered to be responsible (Roemer and Trannoy 2015). As people are increasingly searching for jobs online (Nikolaou 2014), gender indicators in personal proﬁles may disturb the equality of ﬁnding the same job opportunities on the Internet, although Chen et al. (2018) did not ﬁnd evidence for gender bias in recruiting site algorithms. As Pariser (2011) pointed out, the ﬁlter bubble not only reﬂects a person’s identity but it also selectively shows the related options a person has. This may be positive, because some options may be less relevant for people with a speciﬁc proﬁle, for example, jobs for university alumni have less relevance for those with higher vocational education. In defense of personal proﬁle processing, one may see this as the effective working of relevance. Alternatively, however, personal proﬁle processing search engines can replicate and reinforce current gender stereotypes on the job market. The existence of these stereotypical outcomes as a consequence of the usage of user proﬁle information has been identiﬁed before by Otterbacher et al. (2017), but one may ask the question if a privacy-aware search engine like DDG will be free of such bias. Table 1 gives a list of various occupations classiﬁed as male-dominated, mixed, or female-dominated based on data from 12 EU countries. Page rank bias has ethical implications, because search engine outcomes may impact people’s opinions, choices, and actions, and search engines may be intentionally used by stakeholders to inﬂuence society by attempts of censorship and search engine optimization (Fu and Karan 2015; Baye et al., 2016). Jobs The inﬂuence of search engine should not be understated and thus identifying possible page rank biases is important from an ethical perspective (Flaxman et al., 2016; Haim et al., 2018), and its page rank outcomes can nudge people toward certain opinions and decision (Epstein and Robertson 2015) and reinforce possible stereotypical views on people and gender (Kay et al., 2015; Otterbacher et al., 2017). Recent experiences have shown that GSE sometimes intentionally nudges its users toward Web pages that are commercially more interesting from Google’s own perspective (Blanckenburg 2018). Google has been punished for this by the European Committee and corrections have been implemented, but also for nonshopping purposes, search biases can easily happen because of a lack of transparency and the proprietary nature of GSE services (van Drunen et al., 2019). The virtue ethical implications of search engines can be very diverse. For example, it may result in a lack of capability of voters to judge and choose among election candidates (Epstein et al., 2017), or the reinforcement of racist views and behavior (Noble 2018), and the lack of critical thinking (Heersmink 2018). We choose for a focus on virtue ethics from a gender perspective, that is, the abilities of women to emancipate from stereotypical roles and achieve gender equality on the workplace and society. Furthermore, regarding working sectors, women make up almost 80% of those employed in health and social work, over 70% of those employed in education, and over 60% of those working in retailing. In contrast, only 8% of those employed in construction and 14% of those in land transport are women. This pattern is visible throughout all EU member states. We thus expect the following from a proﬁle-aware search engine: Gender Inequalities q According to the United Nations Development Program (UNDP, United Nations Europe and Central Asia, 2018), there are still signiﬁcant inequalities between women and men in Europe and Central Asia, particularly when it comes to jobs and income, political participation, and the distribution of unpaid domestic and care work. Similarly, gender stereotypes are prevalent, hindering women’s access to opportunities. Furthermore, men are more likely to gain promotion to top management positions and prestigious leadership roles than women. Moreover, women are more likely to have insecure jobs, no contract or regular salary, or part-time jobs (UNDP, United Nations Europe and Central Asia, 2018). Search engine outcomes, like any message or piece of information, can reproduce or reinforce views on and the reality of gender inequalities related to jobs and leadership roles in society (Otterbacher et al., 2017; Chen et al., 2018). We, therefore, H1: Women are mainly shown vacancies for female-dominant jobs in their GSE search results. H2: Men are mainly shown vacancies for male-dominant jobs in their GSE search results. Causes and Virtue Ethical Consequences of Search Engine Bias Search engine bias includes a selective presentation of Web documents to a search engine user. This selectivity can be May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 2 Wijnhoven and van Haren Search Engine Bias will go in more detail on jobs and leadership roles in society as two gender-speciﬁc virtue ethical topics. will go in more detail on jobs and leadership roles in society as two gender-speciﬁc virtue ethical topics. selectivity of sources, content, views, and page ranks (Bozdag 2013). The causes for these biases can be explicit by censorship that disables users to access certain sources, content, viewpoints, and algorithms, that is, will make it difﬁcult for users to ﬁnd and see certain content, or the outcome of optimization actions of content suppliers, that is, through search optimization. Finally, search engine bias may also be the result of account and behavior information that is delivered to the search engine, and by which the search aims at increasing its service level may increase in the perception of the user. If this user proﬁle information is used for page ranking, it increases the probability that users will access less diverse sources, content, and viewpoints, which is called the ﬁlter bubble (Pariser 2011; Flaxman et al., 2016). Political Participation Cabeza-García et al. (2018) found differences between men and women regarding political participation. They found evidence that participation in bodies of power and institutions is predominantly masculine, whereas voluntary associations, organizations, and “informal” community politics tend to be May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 3 Search Engine Bias Wijnhoven and van Haren TABLE 1 \| Gender occupation stereotypes (Bettio and Verashchagina 2009). Occupation Number of countries where occupation is dominantly. . . Male Mixed Female Armed forces 10 0 0 Legislators, senior ofﬁcials, and managers 7 5 0 Corporate managers 8 4 0 Managers of small enterprises 5 7 0 Physical, mathematical, and engineering science professionals 12 0 0 Life science and health professionals 0 7 5 Teaching professionals 0 2 10 Other professionals 0 11 1 Physical and engineering science associate professionals 11 1 0 Life science and health associate professionals 0 0 12 Teaching associate professionals 0 2 8 Other associate professionals 0 10 2 Ofﬁce clerks 0 0 12 Customer services clerks 0 0 12 Personal and protective services workers 0 2 10 Models, salespersons, and demonstrators 0 1 11 Skilled agricultural and ﬁshery workers 8 4 0 Extraction and building trades workers 12 0 0 Metal, machinery, and related trades workers 12 0 0 Precision, handicraft, craft printing, and related trades workers 5 7 0 Stationary plant and related operators 11 1 0 Machine operators and assemblers 5 6 1 Drivers and mobile plant operators 12 0 0 Sales and services elementary occupations 0 0 12 Agricultural, ﬁshery, and related laborers 3 6 2 Laborers in mining, construction, manufacturing, and transport 12 0 0 As explained above, women are more likely to engage in certain political practices while men in other, which could possibly be reﬂected and reinforced through personalization techniques used by GSE. We believe that this difference between men and women in politics can be interpreted as involvement of women in less powerful political activities and involvement of men in more powerful activities, consistent with previously identiﬁed stereotypes of competent men and warm women as found on an analysis of Internet images (Otterbacher et al., 2017). This leads to the following hypotheses: led by women. Moreover, the number of women as member of governments was only 16% in 127 countries, and not one of the women was in a top parliamentary management post. Political Participation Furthermore, the women who had a post in local, municipal, or national government occupied posts with a more social and cultural nature and thus with less political importance. Brandtzaeg (2017) studied the difference between men and women in expressions of civic engagement on Facebook and found that millennial women more often engage with posts related to children and the environment when compared to men of the same age. Moreover, it was found that patterns in the ofﬂine world regarding the differences among men and women in civic engagement were reproduced and reinforced rather than equalized on Facebook. H3: Women are mainly shown political involvement opportunities for positions of low power and inﬂuence in their GSE results. H4: Men are mainly shown political involvement opportunities for positions of high power and inﬂuence in their GSE results. Pfanzelt and Spies (2019) studied the difference between young German men and women regarding political participation. They distinguished between three types of political participation, namely, institutional, noninstitutional, and expressive. The ﬁrst one covers activities that mainly address the state via participation in elections or actively running for or holding ofﬁce. Noninstitutional participation refers to protest activities like boycotts and expressive participation includes giving voice to political aims and intentions of citizens. The Internet is a main medium for expressive participation. They found that young men are more likely to participate in institutional and expressive forms, while young women tend toward noninstitutional, protest-oriented activities. Participants’ Actions p We received 144 respondents; however, not all of these respondents could be used due to certain requirements set for this research. One of these requirements is related to cookies. Participants were asked when they removed their cookies for the last time. A cookie is a small piece of data sent from the Web site and stored on the user’s browser that is sent back to the Web site every time the user returns (Coey and Bailey 2016). The information stored on the cookies is used to create a user’s proﬁle by GSE. This has led to the decision of including all participants who removed their cookies two weeks ago or later as the frequent use of the Internet and the instant installation of cookies gives GSE an impression of the users’ proﬁles in these 2 weeks. Other requirements include the use of a personal computer or laptop instead of a phone, being 18 or older, and the use of GSE as the main search engine. This left 102 suitable participants for this study, 53 males and 49 females. All of the participants are students following a higher vocational education, bachelor, or master. The age ranges from 18 to 29 years, with most participants in the range of 20–23 years. The three most occurring nationalities are Dutch (60.8%), Mexican (8.8%), and German (7.8%). This section describes the methodology for observing and analyzing if people’s search results are biased. DuckDuckGo DuckDuckGo (DDG) is said to not gather personal data from its users for its page ranks (Hannak et al., 2013). Instead, DDG bases their search results on expert advice. This means that they focus only on the search term and its semantics and what documents an expert would recommend to access (Jansen and Spink 2006; Jatwani et al., 2020). Thus, considering the method used by DDG, it is expected that search results for every individual user will be nearly the same. This leads to the following hypotheses: H5: There is no gender bias in the vacancies shown by DuckDuckGo to men and women. May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 4 Wijnhoven and van Haren Search Engine Bias H6: There is no gender bias in the political involvement opportunities shown on DuckDuckGo to men and women. outcomes, one of them who is also the co-author of this article. Furthermore, it has been shared in the WhatsApp groups of these two students. Research Design DDG and GSE are the units of analysis in this study. In total, 144 students participated in this study. All participants needed to be at least 18 years of age to assure that they are legally allowed to accept the terms of this research study and that their data are being used for research purposes. Participants should use GSE as their main search engine and should have also not removed their browser cookies too recently so that a rich personal proﬁle could be created by GSE. Participants have been excluded from the research if they have removed their browser cookies in the last 2 weeks. Furthermore, individuals need to participate in the experiment on their own personal laptops or computers to be suitable for the study. Two search queries have been used that address issues that can cause people to have different page ranks. The selected search queries are about highly different gender virtue issue to avoid possible carry over effects, that is, one search query inﬂuencing another one in the same study. Analysis To analyze the collected data, a coding scheme is designed for determining if a query result is biased toward one side. The data analysis in this research is a directed content analysis because this research has started with an explanation of theories related to the research question and from this the coding scheme has followed (Hsieh and Shannon 2005). Only the ﬁrst four delivered screen shots of organic search results have been analyzed, as the ﬁfth and the sixth result were not always visible on all screenshots. This makes a total of 408 screen shots to be coded. The points given to a screenshot range from −6, “only male-biased search results are shown,” to 6, “only female-biased search results are shown.” The ﬁrst two results have been given extra weight and score −2 for male-biased and 2 for female-biased, as according to (Epstein et al., 2017) people generally believe that the top results are presented ﬁrst. The third and the fourth result are given either −1, for male-biased, or 1 for female-biased. This way, this study measures page ranking bias. If there is no bias present, zero points are given. For example, when the ﬁrst three search results were determined to be female-biased and the fourth was determined male-biased, the screenshot would receive a score of 2 + 2 + 1 − 1 4. The detailed coding scheme is given in Supplementary Appendix Tables S1, S2. Frontiers in Big Data \| www.frontiersin.org May 2021 \| Volume 4 \| Article 622106 Data Collection (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 6.286 96 0.000 1.361 0.93 1.79 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 7.757 0.006 −1.875 95 0.064 −0.802 0.428 −1.652 0.047 Not assumed −1.845 81.795 0.069 −0.802 0.435 −1.667 0.063 One sample test including only male participants T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 3.951 50 0.000 0.980 0.48 1.48 Wijnhoven and van Haren Search Engine Bias Search Engine Bias Wijnhoven and van Haren distributed and therefore for all four variables (“Jobs Google,” “Politics Google,” “Jobs DDG,” and “Politics DDG”), a Shapiro–Wilk test has been conducted to test for normality. The Shapiro–Wilk test uses the null hypothesis that the population is normally distributed, which is rejected whenever the p-value falls below a certain alpha value, in our case when the p-value is lower than 0.05. For all four variables, the null hypothesis of normal distribution is rejected as the p-value is TABLE 2 \| SPSS output for “Job openings near me Google.” Sample statistics for “Job openings near me Google” Mean Std. deviation Std. error mean N 97 1.36 2.132 0.216 Male 51 0.98 1.772 0.248 Female 46 1.78 2.421 0.357 One-sample test T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 6.286 96 0.000 1.361 0.93 1.79 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 7.757 0.006 −1.875 95 0.064 −0.802 0.428 −1.652 0.047 Not assumed −1.845 81.795 0.069 −0.802 0.435 −1.667 0.063 One sample test including only male participants T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 3.951 50 0.000 0.980 0.48 1.48 TABLE 3 \| SPSS output for “How to become involved in politics Google.” Sample statistics for “How to become involved in politics Google” Mean Std. deviation Std. error mean N 93 −1.89 2.656 0.275 Male 46 −1.48 2.689 0.397 Female 47 −2.30 2.587 0.377 One-sample test T df Sig. Data Collection Regarding the gender issues, participants of the survey are asked to type in the two search queries into GSE and the same two into DDG. These two search queries are “job openings near me” and “how to become involved in politics.” These two search queries are based on the literature about the gender gap in occupations and political participation as described in Background, Theory, and Hypotheses section before. To be able to measure the construct of bias, the study consisted of ﬁve parts. The ﬁrst part asked questions to determine the suitability of a participant. The second part asked general questions regarding demographics and behavior on GSE of the respondent, for example, age, educational level, and the language mostly used to conduct searches. The third and the fourth parts asked the participants to upload screenshots of their ﬁrst six organic search results appearing, when searching with “job openings near me” and “how to become involved in politics.” These organic search outcomes are the nonpaid search outcomes of the search engine, normally found below the paid search outcomes (i.e., advertisements) of the search engine. The third part of the study asked this for GSE, while the fourth part asked to do this for DDG. The ﬁfth part is the analysis of the results. After the data have been coded, it has been statistically analyzed. Independent-samples t tests have been chosen for comparing the statistical signiﬁcance of mean for the two GSE and DDG male and female users (sub)samples. One-sample t tests are conducted for determining the differences of the means relative to neutral values for the GSE and DDG page rank outcomes. Both tests require the data to be normally To gather enough participants, the link of the answer form has been shared on the personal Facebook page of two of the author’s students who also participated in the coding of the May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 5 TABLE 2 \| SPSS output for “Job openings near me Google.” Sample statistics for “Job openings near me Google” Mean Std. deviation Std. error mean N 97 1.36 2.132 0.216 Male 51 0.98 1.772 0.248 Female 46 1.78 2.421 0.357 One-sample test T Df Sig. Data Collection (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper −2.816 93 0.006 −0.255 −0.44 −0.08 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 0.006 0.939 −0.055 92 0.956 −0.010 0.183 −0.373 0.353 Not assumed −0.055 91.232 0.956 −0.010 0.182 −0.372 0.352 TABLE 4 \| Detailed SPSS output for “Job openings near me DDG.” below 0.05. Hence, a further analysis of Q-Q plots is used to check the normality of the data distribution by a visual inspection of the line that the data make with the theoretically ideal normal distribution. These plots show straight upward lines, thus accepting the data for further analysis (Jaccard and Becker 2002) (data are available on request). below 0.05. Hence, a further analysis of Q-Q plots is used to check the normality of the data distribution by a visual inspection of the line that the data make with the theoretically ideal normal distribution. These plots show straight upward lines, thus accepting the data for further analysis (Jaccard and Becker 2002) (data are available on request). below 0.05. Hence, a further analysis of Q-Q plots is used to check the normality of the data distribution by a visual inspection of the line that the data make with the theoretically ideal normal distribution. These plots show straight upward lines, thus accepting the data for further analysis (Jaccard and Becker 2002) (data are available on request). May 2021 \| Volume 4 \| Article 622106 Data Collection (2-tailed) Mean difference 95% conﬁdence interval of the difference −6.872 92 0.000 −1.892 Lower −2.44 Upper −1.35 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of difference Equal variance F Sig. T df Sig. (2-tailed) Mean difference Std. error, difference Lower Upper Assumed 2.421 0.123 1.498 91 0.138 0.820 0.547 −0.267 1.906 Not assumed 1.497 90.667 0.138 0.820 0.547 −0.268 1.907 Frontiers in Big Data \| www.frontiersin.org May 2021 \| Volume 4 \| Article 622106 6 Levene’s test for equality of variances T 1.48 population is normally distributed, which is rejected whenever the p-value falls below a certain alpha value, in our case when the p-value is lower than 0.05. For all four variables, the null hypothesis of normal distribution is rejected as the p-value is distributed and therefore for all four variables (“Jobs Google,” “Politics Google,” “Jobs DDG,” and “Politics DDG”), a Shapiro–Wilk test has been conducted to test for normality. The Shapiro–Wilk test uses the null hypothesis that the TABLE 3 \| SPSS output for “How to become involved in politics Google.” Sample statistics for “How to become involved in politics Google” Mean Std. deviation Std. error mean N 93 −1.89 2.656 0.275 Male 46 −1.48 2.689 0.397 Female 47 −2.30 2.587 0.377 One-sample test T df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference −6.872 92 0.000 −1.892 Lower −2.44 Upper −1.35 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of difference Equal variance F Sig. T df Sig. (2-tailed) Mean difference Std. error, difference Lower Upper Assumed 2.421 0.123 1.498 91 0.138 0.820 0.547 −0.267 1.906 Not assumed 1.497 90.667 0.138 0.820 0.547 −0.268 1.907 Frontiers in Big Data \| www.frontiersin.org May 2021 \| Volume 4 \| Article 622106 6 TABLE 3 \| SPSS output for “How to become involved in politics Google.” Levene’s test for equality of variances 6 Search Engine Bias Wijnhoven and van Haren below 0.05. Hence, a further analysis of Q-Q plots is used to check the normality of the data distribution by a visual inspection of the line that the data make with the theoretically ideal normal distribution. These plots show straight upward lines, thus accepting the data for further analysis (Jaccard and Becker 2002) (data are available on request). Data Collection Reliability The two persons who helped in the data collection independently performed the coding of the screenshots after discussing the coding scheme of Supplementary Appendix Tables S1, S2 with the ﬁrst author. For testing the interrater reliability of the coding, we used Krippendorff’s alpha (Hayes and Krippendorff TABLE 4 \| Detailed SPSS output for “Job openings near me DDG.” One-sample statistics for “Job openings near me DDG” N Mean Std. Deviation Std. Error mean N 94 −0.26 0.879 0.091 Male 50 −0.26 0.899 0.127 Female 44 −0.25 0.866 0.131 One-sample test T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper −2.816 93 0.006 −0.255 −0.44 −0.08 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 0.006 0.939 −0.055 92 0.956 −0.010 0.183 −0.373 0.353 Not assumed −0.055 91.232 0.956 −0.010 0.182 −0.372 0.352 TABLE 5 \| Detailed SPSS output for “How to become involved in politics DDG.” One-sample statistics “How to become involved in politics” DDG N Mean Std. deviation Std. error mean N 91 0.03 0.314 0.033 Male 48 0.00 0.000 0.000 Female 43 0.07 0.457 0.070 One-sample test T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 1.000 90 0.320 0.033 −0.03 0.10 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 4.692 0.033 −1.057 89 0.293 −0.070 0.066 −0.201 0.061 Not assumed −1 000 42 000 0 323 −0 70 0 070 −0 211 0 071 Wijnhoven and van Haren Search Engine Bias b l f h l f l d h k R li bili TABLE 4 \| Detailed SPSS output for “Job openings near me DDG.” One-sample statistics for “Job openings near me DDG” N Mean Std. Deviation Std. Error mean N 94 −0.26 0.879 0.091 Male 50 −0.26 0.899 0.127 Female 44 −0.25 0.866 0.131 One-sample test T Df Sig. Hypothesis H1: Women are mainly shown vacancies for female related jobs in their GSE search results H1: Women are mainly shown vacancies for female related jobs in their GSE search results H2: Men are mainly shown vacancies for male related jobs in their GSE search results H3: Women are mainly shown political involvement opportunities for positions of low power and inﬂuence in their GSE results H4: Men are mainly shown political involvement opportunities for positions of high power and inﬂuence in their GSE results H5: There is no gender bias in the vacancies shown on DDG to men and women True. The bias toward female-oriented jobs is 1.36. This bias is signiﬁcantly larger than for men Not true. Men also receive a bias (0.98) toward female-oriented jobs Not true. On the contrary women are presented more powerful opportunities than men, but the difference between men and women is not statistically signiﬁcant True True, no signiﬁcant difference between men and women, but a bias toward male- dominated vacancies for both men and women T i iﬁ diff b d H6: There is no gender bias in the political involvement opportunities shown on DuckDuckGo to men and women True, no signiﬁcant difference between men and women 2007). Krippendorff’s alpha is a statistical measure that displays the extent of agreement between coders on the values given to a variable on the basis of a coding scheme. This test is suitable for ratio data which we collected for this study. With the scores assigned to the screenshots a meaning is given to the data ranging from −6 (fully male-biased) to 6 (fully female- biased), the intervals are of equal distance and there is an absolute zero point as assigning a zero means that there is no bias. Before checking for reliability, the results of coding for GSE and DDG have been combined per variable. This has been done because the same coding scheme has been used for “Jobs Google” and “Jobs DDG” as well as for “Politics Google” and “Politics DDG.” For both categories, Krippendorff’s alpha were higher than 0.67, namely 0.7682 for “Jobs” and 0.7725 for “Politics,” which means that the coding schemes for both political participation and jobs were sufﬁciently reliable (Hayes and Krippendorff 2007). indeed signiﬁcantly different than zero (one-sample t test 3.951, df 50, p < 0.000). RESULTS We present the results for the GSE and DDG jobs and political participation queries per hypothesis in the following subsections. Hypothesis This means that they are shown one- sided views on the search query “job openings near me”; however, it was not in the expected direction. Therefore, the second hypothesis is not supported. The SPSS output for these tests is given in Table 2. Political Participation GSE (H3 and H4) Starting with a one-sample t test to determine if there is a bias to be found in the search results for the query “how to become involved in politics,” the test gives that the results differ signiﬁcantly from zero (one-sample t test −6.872, df 92, p < 0.000). The general bias is in the direction of political positions with high power and inﬂuence, as the mean is −1.89. Hypothesis 3 states that women would have signiﬁcantly higher scores than men and that women would have a positive score. This hypothesis is rejected because women have a mean score −2.30, even higher than men who have an average score of −1.48. Clearly, men have a negative score, as expected, and hypothesis 4 is not rejected, but the t test has no signiﬁcant difference between men and women (independent-samples t test 1.498, df 91, p 0.069). The SPSS output is given in Table 3. Jobs DuckDuckGo (H5) ( ) A one-sample t test is used to test if there is no gender bias in the search results of DDG, thus the average score should not differ signiﬁcantly from zero. The one-sample t test shows a signiﬁcant difference from zero, which means that there is a general bias in the search results (one-sample t test −2.816, df 93, p 0.006). When looking at the average score of this sample, it shows that the direction of this bias is toward male-dominated vacancies, as the mean is −0.26. The test does not show signiﬁcantly different results between men and women and therefore supports hypothesis 5 (independent-samples t test −0.055, df 92, p 0.478). The SPSS output is given in Table 4. A one-sample t test has been conducted to test if there is a bias toward either side. The results differ signiﬁcantly from zero (one-sample t test 6.286, df 96, p < 0.000), meaning that the search results for “job openings near me” are biased toward one side as zero means neutral results. This bias is in the direction of female-dominated jobs as the average score of the screenshots is 1.36. With the coding scheme used in this study, the hypothesis is that women will have a signiﬁcantly higher score than men. An independent-samples t test supports this ﬁrst hypothesis: “women are mainly shown vacancies for female-related jobs in their GSE search results” (independent-samples t test −1.845, df 81.795, p 0.0345). Surprisingly, the mean score of men is 0.98, which also implies that on average men were shown more female-dominated jobs instead of male- dominated jobs. Before being able to say with evidence that the second hypothesis is not supported, a one-sample t test only for men has been conducted. The results from men are Reliability Th t y The two persons who helped in the data collection independently performed the coding of the screenshots after discussing the coding scheme of Supplementary Appendix Tables S1, S2 with the ﬁrst author. For testing the interrater reliability of the coding, we used Krippendorff’s alpha (Hayes and Krippendorff TABLE 5 \| Detailed SPSS output for “How to become involved in politics DDG.” One-sample statistics “How to become involved in politics” DDG N Mean Std. deviation Std. error mean N 91 0.03 0.314 0.033 Male 48 0.00 0.000 0.000 Female 43 0.07 0.457 0.070 One-sample test T Df Sig. (2-tailed) Mean difference 95% conﬁdence interval of the difference Lower Upper 1.000 90 0.320 0.033 −0.03 0.10 Independent-samples test Levene’s test for equality of variances t test for equality of means 95% conﬁdence interval of the difference Equal variance F Sig. t Df Sig. (2-tailed) Mean difference Std. error difference Lower Upper Assumed 4.692 0.033 −1.057 89 0.293 −0.070 0.066 −0.201 0.061 Not assumed −1.000 42.000 0.323 −0.70 0.070 −0.211 0.071 Frontiers in Big Data \| www.frontiersin.org May 2021 \| Volume 4 \| Article 622106 7 TABLE 5 \| Detailed SPSS output for “How to become involved in politics DDG.” Levene’s test for equality of variances May 2021 \| Volume 4 \| Article 622106 7 Search Engine Bias Wijnhoven and van Haren Frontiers in Big Data \| www.frontiersin.org DISCUSSION This study tried to answer the question to what extent GSE biases its users by showing one-sided search results with possible virtue ethical gender implications. Overall, GSE gives a bias, but not necessarily against a stronger female participation in more powerful jobs and political participation opportunities. As far as GSE could nudge people toward certain choices and behavior, it also can nudge people toward virtues by showing opportunities of emancipation. A nonproﬁle aware search engine like DDG may be nonbiased from a proﬁle, in our case, gender perspective, but is, therefore, not free from virtue ethical implications. Our study shows that DDG gives male-dominant job outcomes, and this may have virtue ethical implications if this results in a larger attractiveness or social value for such jobs, and thus an undervaluation of female-dominant jobs. We do not want to say that DDG bears any responsibility for this outcome, because proﬁle- free search engines may just give a representation of a biased society or information market. Therefore, it is important that Internet users know of the existence of such a bias and the risk that it can bring (Epstein et al., 2017). Epstein et al. (2017) state that a bias awareness tool may help search engine users to decide for themselves what they want to accept as valid, credible, or just, but developing a critical view on search engine outcomes is often difﬁcult (Lazer et al., 2018; Flanagin et al., 2014). Nobias.com iscurrently developingasa bias-alertingtool.Search engine services could also develop dialectic search outcomes, with which we mean a search engine that gives contrasting search outcomes and alerts users to information sources with alternative views on a subject (Wijnhoven and Brinkhuis 2015). Regarding the political participation results for Google, the result of the tests does not support the hypothesis that women would be given higher page ranks for less powerful activities. The general bias toward more powerful positions could support the view that the supply side of content on the Web—due to its structure and presentation—is in itself male-biased (Hargittai and Shafer 2006). DDG shows different patterns as for jobs, there is a slight but signiﬁcant bias to more male-oriented job outcomes and for political participation no bias toward more or less powerful activities exists. For both queries, the differences between men and women are not signiﬁcant, as expected. Political Participation DDG (H6) Hypothesis 6 states the absence of DDG gender bias in its search results for men and women regarding political participation. The scores given to the screenshots give no signiﬁcant difference from zero (one-sample t test 1.000, df 90, p 0.320). An independent-samples t test of the average scores of men and May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 8 Wijnhoven and van Haren Search Engine Bias is needed for more clearly concluding about its implications. We also have not separated participants in those who search via a GSE account (i.e., explicit proﬁle) and those who do not. Because GSE will likely know more about the ﬁrst group, the ﬁlter bubble effect may be greater for them than for those who search without an account. A third limitation is the queries we use. The number of possible queries is of course unlimited and as researchers we needed to focus. We selected only two topics that are highly relevant for students searching for a social position and have been highly debated in the literature on women equality. We leave it to the readers to redo our study for other topics. Finally, our study is limited to only two search engines. The difference between GSE and DDG, however, is theoretically and practically relevant and found back in the evidence of this study. However, we cannot generalize over all search engines with personalization or not. Comparative studies of search engines remain important from the perspective of knowing their biases and how these biases could be canceled out via searches with multiple search engines (Jansen and Spink 2006; Nguyen et al., 2012). women gives no signiﬁcant difference between the results shown to men and women (independent-samples t test −1.000, df 42.000, p 0.1615), which supports hypothesis 6. The SPSS output is given in Table 5. DISCUSSION For DDG, the bias found is thus independent of the gender of its users and only for job opportunities. This bias thus is not caused by proﬁle processing but possibly by indexing algorithms and the effects of supply and demand of content. Conclusion We summarize our conclusions regarding the hypotheses in Table 6. In this study, we ﬁnd that men have higher page ranks for female-dominated jobs as GSE query outcomes. This is not in line with our expectations. One explanation for this is that GSE applied student (our sample) proﬁle data and with that the page ranked ﬂexible jobs higher. Student employment includes any form of paid work during the academic year or during summer, but especially ﬂexible jobs (Baert et al., 2016). As in this study, ﬂexible work has been seen as an aspect of a female job following the UNDP, United Nations Europe and Central Asia, 2018; this may give male users higher scores for female-stereotyped jobs. Moreover, as this study has been conducted around the time that people were searching for summer jobs, GSE may even generate more outcomes toward female stereotype jobs (Burger et al., 2016). REFERENCES Cabeza-García, L., Del BrioDel Brio, E., and Oscanoa-Victorio, M. (2018). Gender Factors and Inclusive Economic Growth: The Silent Revolution. Sustainability 10 (121), 2–14. doi:10.3390/su10010121 Andersen, J. (2018). Archiving, Ordering, and Searching: Search Engines, Algorithms, Databases, and Deep Mediatization. Media, Cult. Soc. 40 (8), 1135–1150. doi:10.1177/0163443718754652 Chen, L., Ma, R., Hannák, A., and Wilson, C. (2018). “Investigating the Impact of Gender on Rank in Resume Search Engines,” in Proceedings Of the 2018 CHI Conference On Human Factors In Computing Systems. Editors R. Mandryk and M. Hancock (Montreal QB, Canada: Australian National University P ) Baert, S., Rotsaert, O., Verhaest, D., and Omey, E. (2016). Student Employment and Later Labour Market Success: No Evidence for Higher Employment Chances. Kyklos 69 (3), 401–425. doi:10.1111/kykl.12115 and M. Hancock (Montreal QB, Canada: Australian National University Press), 1–14. Coey, D., and Bailey, M. (2016). “People and Cookies,” in 25th International World Wide Web Conference. WWW. Baye, M. R., De los Santos, B., and Wildenbeest, M. R. (2016). Search Engine Optimization: What Drives Organic Trafﬁc to Retail Sites?. J. Econ. Manag. Strategy 25 (1), 6–31. doi:10.1111/jems.12141 Colley, A., and Maltby, J. (2008). Impact of the Internet on Our Lives: Male and Female Personal Perspectives. Comput. Hum. Behav. 24 (5), 2005–2013. doi:10. 1016/j.chb.2007.09.002 Strategy 25 (1), 6–31. doi:10.1111/jems.12141 Bettio, F., and Verashchagina, A. (2009). Gender Segregation in the Labour Market: Root Causes, Implications and Policy Responses in the EU. Brussels, Belgium; Luxembourg: European Commission. Cronqvist, H., Thaler, R. H., and Yu, F. (2018). When Nudges Are Forever: Inertia in the Swedish Premium Pension Plan. AEA Pap. Proc. 108 (May), 153–158. doi:10.1257/pandp.20181096 Blanckenburg, K. V. (2018). Google Search Abuses Dominant Position to Illegally Favour Google Shopping: An Economic Review of the EU Decision. Dprg 20 (3), 211–224. doi:10.1108/DPRG-05-2017-0020 De Corniere, A., and Taylor, G. (2014). Integration and Search Engine Bias. RAND J. Econ. 45 (3), 576–597. doi:10.1111/1756-2171.12063 Favour Google Shopping: An Economic Review of the EU Decision. Dprg 20 (3), 211–224. doi:10.1108/DPRG-05-2017-0020 Bozdag, E. (2013). Bias in Algorithmic Filtering and Personalization. Ethics Inf. Technol. 15 (3), 209–227. doi:10.1007/s10676-013-9321-6 Deﬂoor, B., Van Ootegem, L., and Verhofstadt, E. (2015). A Good or Bad Transition from School to Work: Who Is Responsible? Int. J. Manpower 36 (8), 1207–1226. doi:10.1108/IJM-06-2014-0137 Brandtzaeg, P. B. (2017). Facebook Is No “Great Equalizer”. Soc. Sci. Comp. Rev. 35 (1), 103–125. doi:10.1177/0894439315605806 Epstein, R., Robertson, R. E., Lazer, D., and Wilson, C. (2017). Limitations Our study has a few important limitations that need further research. One limitation is our sample. Although the sample size as a whole and for the subsamples of men and women is above 40 and allows the tests we did, the sample is restricted to students from Western countries and even with mainly Dutch and German origins. Samples including other age-groups, different occupations, and different regions could result in alternative insights. For example, the differences between men and women in Asian societies are stated to be larger than in Western societies (Hofstede et al., 2010). A second limitation is our operationalization of GSE’s proﬁle. GSE is not transparent to us regarding the content of the proﬁles they use, but search proﬁles include more than gender indicators alone (Pan et al., 2007; Scheitle 2011; Kosinski et al., 2013). Consequently, the actual bias is not only a ﬁlter on gender. A further study of other proﬁle characteristics and their search engine bias effects GSE and DDG both can deliver biased search results, but no evidence in our study is found of a ﬁlter bubble, and thus we need a more complete overview of possible sources of search engine bias. For such an analysis, one may go for a full analysis of the search engine value chain (Andersen 2018), starting with publishing biases on the Internet, that is, different volumes of certain content available on the Internet and thus a higher chance of being retrieved. Search engines May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 9 Wijnhoven and van Haren Search Engine Bias DATA AVAILABILITY STATEMENT Other biases further on the road to the information end consumer are biases in query composition based on the user’s world view (language and interests). Possible query support tools could help users avoiding unwanted biases or to be aware of their search biases. The algorithm by which a search engine matches queries with documents is the search matching bias. When applying personal proﬁling, proﬁle biases may include the search engine’s understanding of the user’s subjective perceived need, but search engines also may have classiﬁcation biases related to their understanding of documents and actual needs. Finally, also browsers may present search outcomes in different ways and users may interpret outcomes in different ways. To summarize, we therefore identify the following search engine biases: 1) publishing biases, 2) crawling and indexing biases, 3) query biases, 4) search support biases, 5) algorithm biases, 6) proﬁle biases, 7) content matching biases, 8) content classiﬁcation biases, 9) browser presentation biases, and 10) interpretation biases. This study only uncovered some part of possible proﬁle biases and more work is to be done to gain a fuller understanding of search engine biases. Search engine ﬁrms may have a tendency to optimize proﬁle biases for creating a highest subjective satisfaction of their users, thus optimizing utility ethics, but this can be at the cost of virtue ethics, that is, understanding and opportunities of voluntary acting (Stucke and Ezrachi 2016; Gal et al., 2020). In contrast, search engine services should be free from the need to optimize user subjective satisfaction, and instead search engine services should be developed that can be free from commercial or political funding, serving the users’ virtues of personal skills, wisdom, and voluntary action (Fuchs 2011; Zuboff 2019). The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. ETHICS STATEMENT The studies involving human participants were reviewed and approved by the University of Twente, faculty of Behavioral, Management and Social Sciences. The patients/participants provided their written informed consent to participate in this study. ACKNOWLEDGMENTS We thanks Rebecca Schaefer for her support in the data collection and data coding. AUTHOR CONTRIBUTIONS FW designed the study, analyzed the data, wrote the article and communicated with reviewers and the publisher. JH co-designed the study, collected the data and did a ﬁrst raw analysis. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fdata.2021.622106/ full#supplementary-material DATA AVAILABILITY STATEMENT use crawlers to index this content in which they necessarily have to be selective because of the excessive size of the Internet. In this selection process, crawlers are, therefore, biased to certain criteriaand the setting of these selection criteria may help avoiding an unethical bias in the search engine’s index. Other biases further on the road to the information end consumer are biases in query composition based on the user’s world view (language and interests). Possible query support tools could help users avoiding unwanted biases or to be aware of their search biases. The algorithm by which a search engine matches queries with documents is the search matching bias. When applying personal proﬁling, proﬁle biases may include the search engine’s understanding of the user’s subjective perceived need, but search engines also may have classiﬁcation biases related to their understanding of documents and actual needs. Finally, also browsers may present search outcomes in different ways and users may interpret outcomes in different ways. To summarize, we therefore identify the following search engine biases: 1) publishing biases, 2) crawling and indexing biases, 3) query biases, 4) search support biases, 5) algorithm biases, 6) proﬁle biases, 7) content matching biases, 8) content classiﬁcation biases, 9) browser presentation biases, and 10) interpretation biases. This study only uncovered some part of possible proﬁle biases and more work is to be done to gain a fuller understanding of search engine biases. Search engine ﬁrms may have a tendency to optimize proﬁle biases for creating a highest subjective satisfaction of their users, thus optimizing utility ethics, but this can be at the cost of virtue ethics, that is, understanding and opportunities of voluntary acting (Stucke and Ezrachi 2016; Gal et al., 2020). In contrast, search engine services should be free from the need to optimize user subjective satisfaction, and instead search engine services should be developed that can be free from commercial or political funding, serving the users’ virtues of personal skills, wisdom, and voluntary action (Fuchs 2011; Zuboff 2019). use crawlers to index this content in which they necessarily have to be selective because of the excessive size of the Internet. In this selection process, crawlers are, therefore, biased to certain criteriaand the setting of these selection criteria may help avoiding an unethical bias in the search engine’s index. REFERENCES A Virtue Epistemology of the Internet: Search Engines, Intellectual VirtuesandEducation.Soc. Epistemol.32 (1), 1–12.doi:10.1080/02691728.2017.1383530 Tavani, H. (2012). Search Engines and Ethics, Stanford Encyclopedia of Philosophy. Tavani, H. (2012). Search Engines and Ethics, Stanford Encyclopedia of Philosophy. Stanford, CA, United States: The Metaphysics Lab, Stanford University. Stanford, CA, United States: The Metaphysics Lab, Stanford University. Helberger, N., Karppinen, Kari., Karppinen, K., and D’Acunto, L. (2018). Exposure Diversity as a Design Principle for Recommender Systems. Inf. Commun. Soc. 21 (2), 191–207. doi:10.1080/1369118x.2016.1271900 Thurman, N., and Schifferes, S. (2012). The Future of Personalization at News Websites. Journalism Stud. 13 (5–6), 775–790. doi:10.1080/1461670x.2012.664341 Hofstede, G., Hofstede, G. J., and Minkov, M. (2010). Cultures and Organizations, Software of the Mind. ed. London: McGraw-Hill. UNDP, United Nations Europe and Central Asia (2018). United Nations Development Programme Focus Gender Equality. Available at: http://www. eurasia.undp.org/content/rbec/en/home/gender-equality.html Hsieh, H.-F., and Shannon, S. E. (2005). Three Approaches to Qualitative Content Analysis. Qual. Health Res. 15 (9), 1277–1288. doi:10.1177/1049732305276687 eurasia.undp.org/content/rbec/en/home/gender-equality.html van Drunen, M. Z., Natali, H., and Bastian, M. (2019). Know Your Algorithm: What Media Organizations Need to Explain to Their Users about News Personalization. Int. Data Privacy L. 9 (4), 220–235. Jaccard, J., and Becker, M. A. (2002). Statistics for the Behavioral Sciences. Boston, MA, United States: Wadsworth Publishing Company Voss, W. G. (2014). European Union Data Privacy Law Developments. The Business Lawyer 70 (1), 253–260. Jansen, B. J., and Spink, A. (2006). How Are We Searching the World Wide Web? A Comparison of Nine Search Engine Transaction Logs. Inf. Process. Manag. 42 (1), 248–263. doi:10.1016/j.ipm.2004.10.007 Wijnhoven, F., and Brinkhuis, M. (2015). Internet Information Triangulation: Design Theory and Prototype Evaluation. J. Assn Inf. Sci. Tec 66 (4), 684–701. doi:10.1002/asi.23203 Jatwani, P., Tomar, P., and Dhingra, V. (2020). Comparative Performance Evaluation of Keyword and Semantic Search Engines Using Different Query Set Categories. Racsc 13 (5), 1057–1070. doi:10.2174/2213275912666190328202153 Wilson, R. E., Gosling, S. D., and Graham, L. T. (2012). A Review of Facebook Research in the Social Sciences. Perspect. Psychol. Sci. 7 (3), 203–220. doi:10. 1177/1745691612442904 Kaptein, M., Markopoulos, P., de Ruyter, B., and Aarts, E. (2015). Personalizing Persuasive Technologies: Explicit and Implicit Personalization Using Persuasion Proﬁles. Int. J. Human-Computer Stud. 77, 38–51. doi:10.1016/j. ijhcs.2015.01.004 Yom-Tov, E. (2019). Demographic Differences in Search Engine Use with Implications for Cohort Selection. Inf. Retrieval J. 22 (6), 570–580. doi:10.1007/s10791-018-09349-2 Kay, M., Matuszek, C., and Munson, S. A. (2015). REFERENCES Suppressing the Search Engine Manipulation Effect (SEME). Proc. ACM Hum.-Comput. Interact. 1, 1–22. doi:10.1145/3134677 Burger, V., Hirth, M., Tobias, H., and Tran-Gia, P. (2016). “Principles of Information Neutrality and Counter Measures against Biased Information.” in Lecture Notes In Informatics (LNI), Proceedings - Series of the Gesellschaft Fur Informatik (GI)Editors H. Mayer and M. Pinzger. Springer. Epstein, R., and Robertson, R. E. (2015). The Search Engine Manipulation Effect (SEME) and its Possible Impact on the Outcomes of Elections. Proc. Natl. Acad. Sci. USA 112 (33), E4512–E4521. doi:10.1073/pnas.1419828112 May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 10 Wijnhoven and van Haren Search Engine Bias Flanagin, A. J., Metzger, M. J., Pure, R., Markov, A., and Hartsell, E. (2014). Mitigating Risk in Ecommerce Transactions: Perceptions of Information Credibility and the Role of User-Generated Ratings in Product Quality and Purchase Intention. Electron. Commer Res. 14 (1), 1–23. doi:10.1007/s10660-014-9139-2 Nguyen, D., Demeester, T., Trieschnigg, D., and Hiemstra, D. (2012). “Federated Search in the Wild: The Combined Power of over a Hundred Search Engines,” in Proceedings Of the 21st ACM International Conference On Information And Knowledge Management. New York, New York, USA: ACM, 1874–1878. g y ectron. Commer Res. 14 (1), 1–23. doi:10.1007/s10660-014-9139-2 Flaxman, S., Goel, S., and Rao, J. M. (2016). Filter Bubbles, Echo Chambers, and Online News Consumption. Pubopq 80 (S1), 298–320. doi:10.1093/poq/nfw006 Nikolaou, I. (2014). Social Networking Web Sites in Job Search and Employee Recruitment. Int. J. Select Assess. 22 (2), 179–189. doi:10.1111/ijsa.12067 Noble, S. U. (2018). Algorithms of Oppression: How Search Engines Reinforce Racism. New York: New York University Press. doi:10.2307/j.ctt1pwt9w5 Fu, T., and Karan, K. (2015). How Big Is the World You Can Explore? A Study of Chinese College Students’ Search Behavior via Search Engines. Proced. - Soc. Behav. Sci. 174, 2743–2752. doi:10.1016/j.sbspro.2015.01.961 Otterbacher, J., Bates, J., and Clough, P. (2017). “Competent Men and Warm Women: Gender Stereotypes and Backlash in Image Search Results,” in Proceedings Of the 2017 Chi Conference On Human Factors In Computing Systems. Editors G. Mark Fuchs, C. (2011). A Contribution to the Critique of the Political Economy of Google. Fast Capitalism 8 (1), 1–24. doi:10.32855/fcapital.201101.006 p g y and S. Fussell (Denver, CO, United States: ACM New York), 6620–6631. Fussell (Denver, CO, United States: ACM New York), 6620–6631. Gal, U., Jensen, T. B., and Stein, M.-K. (2020). REFERENCES Breaking the Vicious Cycle of Algorithmic Management: A Virtue Ethics Approach to People Analytics. Inf. Organ. 30 (2), 100301. doi:10.1016/j.infoandorg.2020.100301 Pan, B., Hembrooke, H., Joachims, T., Lorigo, L., Gay, G., and Granka, L. (2007). In Google We Trust: Users’ Decisions on Rank, Position, and Relevance. J. Computer- Mediated Commun. 12 (3), 801–823. doi:10.1111/j.1083-6101.2007.00351.x Pan, B., Hembrooke, H., Joachims, T., Lorigo, L., Gay, G., and Granka, L. (2007). In Google We Trust: Users’ Decisions on Rank, Position, and Relevance. J. Computer- Mediated Commun. 12 (3), 801–823. doi:10.1111/j.1083-6101.2007.00351.x Goel, S., Hofman, J. M., and Sirer, M. I. (2012). Who Does what on the Web: A Large-Scale Study of Browsing Behavior. ICWSM), 130–137. Pariser, E. (2011). The Filter Bubble: What the Internet Is Hiding from You. London, United Kingdom: Penguin Haim, M., Graefe, A., and Brosius, H.-B. (2018). Burst of the Filter Bubble? Digital Journalism 6 (3), 330–343. doi:10.1080/21670811.2017.1338145 Pfanzelt, H., and Spies, D. C. (2019). The Gender Gap in Youth Political Participation: Evidence from Germany. Polit. Res. Q. 72 (1), 34–48. doi:10. 1177/1065912918775249 Hannak, A., Sapiezynski, P., Kakhki, A. M., Krishnamurthy, B., Lazer, D., Mislove, A., et al. (2013). “Measuring Personalization of Web Search,” in Proceedings Of the 22nd International Conference On World Wide Web. ACM, 527–538. Roemer, J. E., and Trannoy., A. (2015). Equality of Opportunity, Handbook of Income Distribution, 2. , 217–300. doi:10.1016/B978-0-444-59428-0.00005-9 Hargittai, E., and Shafer, S. (2006). Differences in Actual and Perceived Online Skills: The Role of Gender. Social Sci. Q 87 (2), 432–448. doi:10.1111/j.1540-6237.2006.00389.x Rogers, R. (2013). Digital Methods. Boston, MA, United States: MIT press. doi:10.7551/ mitpress/8718.001.0001 Harvey, M., Crestani, F., and Carman, M. J. (2013). “Building User Proﬁles from Topic Models for Personalised Search,” in Proceedings Of the 22nd ACM International Conference On Conference On Information & Knowledge Management. New York, United States: ACM, 2309–2314. Scheitle, C. P. (2011). Google’s Insights for Search: A Note Evaluating the Use of Search Engine Data in Social Research. Soc. Sci. Q. 92 (1), 285–295. doi:10. 1111/j.1540-6237.2011.00768.x Hayes, A. F., and Krippendorff, K. (2007). Answering the Call for a Standard Reliability Measure for Coding Data. Commun. Methods Measures 1 (1), 77–89. doi:10.1080/19312450709336664 Scherer, S. (2004). Stepping-Stones or Traps?. Work, Employment Soc. 18 (2), 369–394. doi:10.1177/09500172004042774 Stucke, M. E., and Ezrachi, A. (2016). When Competition Fails to Optimize Quality: A Look at Search Engines. Yale JL Tech. 18, 70. doi:10.2139/ssrn.2598128 Heersmink, R. (2018). When is it one-sided for women? When is it one-sided for women? • When there is a vacancy mentioned in either the header, abstract, or link of the search result that relates to a part-time job, insecure contract, and/or the following female dominated occupations: • When there is a vacancy mentioned in either the header, abstract, or link of the search result that relates to a part-time job, insecure contract, and/or the following female dominated occupations: • Teaching professionals g p • Life science and health associate professionals • Teaching associate professionals REFERENCES “Unequal Representation and Gender Stereotypes in Image Search Results for Occupations,” in Proceedings Of the 33rd Annual ACM Conference On Human Factors In Computing Systems. Editors B. Begole and J. Kim (Seoul Korea: ACM New York), 3819–3828. Zuboff, S. (2019). Surveillance Capitalism and the Challenge of Collective Action. New Labor Forum 28 (1), 10–29. doi:10.1177/1095796018819461 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. Kosinski, M., Stillwell, D., and Graepel, T. (2013). Private Traits and Attributes Are Predictable from Digital Records of Human Behavior. Proc. Natl. Acad. Sci. 110 (15), 5802–5805. doi:10.1073/pnas.1218772110 Lazer, D. M. J., Baum, M. A., Benkler, Y., Berinsky, A. J., Berinsky, G. K. M., Greenhill, F. M., et al. (2018). The Science of Fake News. Science 359 (6380), 1094–1096. doi:10.1126/science.aao2998 Copyright © 2021 Wijnhoven and van Haren. 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. Lopes, C., Cabral, B., and Bernardino, J. (2016). “Personalization Using Big Data Analytics Platforms,” in Proceedings Of the Ninth International C* Conference On Computer Science & Software Engineering - C3S2E ’16. Editors E. Desai. New York, New York, USA: ACM Press, 131–132. May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 11 Search Engine Bias Wijnhoven and van Haren of search result First result mentions male dominated occupation(s) 2nd result mentions male dominated occupation(s) 3rd result mentions male dominated occupation(s) 4th result mentions male dominated occupation(s) First result mentions female dominated occupation(s) 2nd result mentions female dominated occupation(s) 3rd result mentions female dominated occupation(s) 4th result mentions female dominated occupation(s) When no occupations are mentioned in the results When is it one-sided for men? When is it one-sided for men? • When there is a vacancy mentioned in either the header, abstract, or link of the search result that relates to a full-time job, and/or the following male dominated occupations SpargWhen is it one-sided for men? • Armed forces • Armed forces • Legislators, senior ofﬁcials, and managers • Corporate managers • When the search result mention in either the header, abstract, or link, relate to high power and inﬂuence positions regarding politics like: • Regarding economics • Foreign and internal affairs • Defense and justice • Direct involvement (e.g., run for ofﬁce, get in touch with politicians) • As well as if somewhere it mentions explicitly male/man etc. • Physical, mathematical, and engineering science professionals • Physical and engineering science associate professionals • Skilled agricultural and ﬁshery workers • Extraction and building trades workers • Metal, machinery, and related trades workers • Stationary-plant and related operators • Drivers and mobile plant operators • Laborers in mining, construction, manufacturing, and • Laborers in mining, construction, manufacturing, and transport Content of search result Code First result mentions high power and inﬂuence position(s) -2 2nd result mentions high power and inﬂuence position(s) -2 3rd result mentions high power and inﬂuence position(s) -1 4th result mentions high power and inﬂuence position(s) -1 First result mentions low power and inﬂuence position(s) 2 2nd result mentions low power and inﬂuence position(s) 2 3rd result mentions low power and inﬂuence position(s) 1 4th result mentions low power and inﬂuence position(s) 1 No particular position(s) of power and inﬂuence are mentioned 0 APPENDIX 2: CODING SCHEME FOR POLITICAL PARTICIPATION. • Customer services clerks • Personal and protective services (e.g., security) workers • Models, salespersons, and demonstrators When is it one-sided for women? • Sales and services elementary occupations • When the search result mention in either the header, abstract, or link, relate to examples of low power and inﬂuence positions regarding politics, like: • Furthermore, when it is explicitly mentioned that the vacancy is for a woman than it is also showing a one- sided result. • Social and cultural activities Content of search result • Furthermore, when it is explicitly mentioned that the vacancy is for a man than it is also showing a one-sided result. ! If a part time or full-time job is mentioned in the results as well as occupations, occupations are determining the bias, however if for example two female dominated occupations are mentioned and one male + full-time then neutral! ! If the occupation(s) does not relate to some in the above lists than it’s neutral! ! If the occupation(s) does not relate to some in the above lists than it’s neutral! Coding system: ! When there are more occupations mentioned (for example two relate to male and two to female), then also neutral (0)! ! If the position(s) does not relate to some in the above lists than it’s neutral! ! When there are more positions mentioned (for example two relate to male and two to female), then also neutral (0)! ! When the screenshot cannot be used, use code 999! Coding system: ! When the screenshot cannot be used, use code 999! Coding system: May 2021 \| Volume 4 \| Article 622106 Frontiers in Big Data \| www.frontiersin.org 12
https://openalex.org/W3139020192	https://biblio.ugent.be/publication/8703325/file/8703331	English	null	Setting Priorities in River Management Using Habitat Suitability Models	Water	2,021	cc-by	13,720	water water Academic Editors: Rui Cortes and Thomas Hein Received: 25 January 2021 Accepted: 22 March 2021 Published: 24 March 2021 Keywords: habitat suitability modeling; decision support; ecological water quality Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Citation: Bennetsen, E.; Gobeyn, S.; Everaert, G.; Goethals, P. Setting Priorities in River Management Using Habitat Suitability Models. Water 2021, 13, 886. https://doi.org/ 10.3390/w13070886 Elina Bennetsen 1,, Sacha Gobeyn 2 1 Aquatic Ecology Research Unit, Department of Animal Sciences and Aquatic Ecology, Ghent University, 9000 Ghent, Belgium; peter.goethals@ugent.be 2 Fluves, 9000 Ghent, Belgium; sacha@ﬂuves.com 3 Flanders Marine Institute, 8400 Ostend, Belgium; gert.everaert@vliz.be Correspondence: elinabennetsen@gmail.com 1 Aquatic Ecology Research Unit, Department of Animal Sciences and Aquatic Ecology, Ghent University, 9000 Ghent, Belgium; peter.goethals@ugent.be 2 Fluves, 9000 Ghent, Belgium; sacha@ﬂuves.com 3 Flanders Marine Institute, 8400 Ostend, Belgium; gert.everaert@vliz.be * Correspondence: elinabennetsen@gmail.com g 3 Flanders Marine Institute, 8400 Ostend, Belgium; gert.everaert@vliz.be * Correspondence: elinabennetsen@gmail.com Abstract: Worldwide river systems are under pressure from human development. River managers need to identify the most important stressors in a stream basin, to propose effective management interventions for river restoration. In the European Union, the Water Framework Directive proposes the ecological status as the management endpoint for these interventions. Many decision support tools exist that use predictive water quality models to evaluate different river management scenarios, but only a few consider a river’s ecological status in this analysis explicitly. This paper presents a novel method, which combines abiotic monitoring data and biological monitoring data, to provide information and insight on why the ecological status does not reach the good status. We use habitat suitability models as a decision support tool, which can identify the most important stressors in river systems to deﬁne management scenarios. To this end, we disassemble the ecological status into its individual building blocks, i.e., the community composition, and we use habitat suitability models to perform an ecological gap analysis. In this paper, we present our method and its underlying ecological concepts, and we illustrate its beneﬁts by applying the method on a regional level for Flanders using a biotic index, the Multimetric Macroinvertebrate Index Flanders (MMIF). To evaluate our method, we calculated the number of correctly classiﬁed instances (CCI = 47.7%) and the root-mean-square error (RMSE = 0.18) on the MMIF class and the MMIF value. Furthermore, there is a monotonic decreasing relationship between the results of the priority classiﬁcation and the ecological status expressed by the MMIF, which is strengthened by the inclusion of ecological concepts in our method (Pearson’s R2 −0.92 vs. −0.87). In addition, the results of our method are complementary to information derived from the legal targets set for abiotic variables. Thus, our proposed method can further optimize the inclusion of monitoring data for the sake of sustainable decisions in river management. water water Article Elina Bennetsen 1,*, Sacha Gobeyn 2 , Gert Everaert 3 and Peter Goethals 1 1. Introduction Worldwide river systems are under pressure from human development [1]. Pollution from agriculture, households or industry changes water quality, and urban development or navigation often require hydromorphological alterations [2]. To preserve freshwater biodiversity and maintain the ecosystem services that rivers provide, pollution mitigation and river restoration are necessary [3]. River managers have taken various actions to restore rivers to their original status, but often river water quality remains low [4]. Copyright: © 2021 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/). To increase the success of river restoration actions, it is necessary to understand which aspects of a river system, physical–chemical or hydromorphological, contribute to the impairment of water quality. Elosegi et al. [5] argue that river managers should perform a differential diagnosis, as a doctor would diagnose a patient. This means not only that they should monitor individual quality elements and compare them to their legal targets, but they should also interpret these results using a systems perspective [6]. These targets are typically only background values representing the minimum requirements [7]. For https://www.mdpi.com/journal/water Water 2021, 13, 886. https://doi.org/10.3390/w13070886 Water 2021, 13, 886 2 of 19 2 of 19 example, the legal target for dissolved oxygen concentration in river ecosystems can be interpreted as a minimal allowable lower limit. They thus serve as an orientation for what aspects of the river system are changed and could be considered stressors, but it is not necessarily so that river aspects that deviate from their legal targets always pose a stress on the river system. Therefore, Voulvoulis et al. [6] argue that river management should not focus on regulating individual monitored pollutants and hydromorphological changes, but that individual aspects of river system should be regulated from a systems perspective. For this, the ecological status needs to be central to an analysis that focuses on the causes of impairment [6]. However, at this stage it is unclear whether the classiﬁcation of the ecological status provides enough information to identify the causes of river impairment. Within the European Water Framework Directive (WFD), a river’s ecological status is classiﬁed based on the ecological quality ratio. The ecological quality ratio consists of a set of biotic indices that show how far river water quality deviates from the so-called reference condition. 1. Introduction Several authors argue that the ecological quality ratios, albeit being powerful quality assessment tools, do not provide the explanatory power to identify the causes of river impairment [5,8–10]. Rather, they are qualitative indicators [6] that indicate how large the need is for restoring water quality. This means it is not always clear on which aspect river managers should focus, given the observed ecological status. g g g The ecological status classiﬁcation provides data on the ecological status and how much it deviates from the reference condition, the ecological gap [6], but little information and insight into what causes the impairment of river systems. A “moderate” status can represent many conditions in the ﬁeld. Furthermore, many European countries use multi- metric indices to calculate the ecological quality ratio [11]. These indices aggregate different metrics into one index. Due to this, they have a risk of information loss when aggregating multiple dimensions into one metric [12]. Although they have proven to be more robust than single metric indices to assess quality status, they may be subject to systematic bias from the conversion of raw data into categorical scores [12,13]. However, the building blocks of the biotic indices that measure ecological quality, being the taxonomic and func- tional composition of the biotic communities in a water body, do provide information and insight into what aspects of the river system have a negative impact on the ecological status and can thus be considered stressors [14]. By considering the relationship between the abiotic condition and the presence of individual taxa, rather than the aggregated index, river managers can better understand what the cause of the impairment is [15,16]. To this end, simulation models can be used. Many decision support tools will consider the ecological status as the endpoint for the models [17–19] in a predictive framework with which the effect of management scenarios can be evaluated. These frameworks are very valuable, but often they cannot be used as the explanatory tools [20] that we need for the diagnostic analysis that we want to perform. To date, there is a need for a novel an analysis that uses the ecological status and its building blocks as a starting point. To evaluate the ecological status and the community composition, habitat suitabil- ity models can be used. Typically, data-driven and/or expert based habitat suitability modeling is used to express the relationship between species presence and stressors. 1. Introduction The in- tegration of these models into an analysis of the ecological status is a challenge, as the data requirements for describing the habitat suitability of large biotic communities are very high. Speciﬁcally, sensitive taxa that typically have a low occurrence in existing datasets [21] can be hard to model. This is a key issue, as they are often the taxa of interest for remediating water quality. However, recent improvements in the development of habitat suitability models, such as the integration of ecological theory in modeling frameworks, now make it possible to build species speciﬁc methods for many species, including rarer taxa [15,22–24]. Yet, the question remains how results from habitat suitability models can serve as reliable and applicable information for a decision maker [25]. To increase the potential of these models for use as diagnostic tools, it is necessary to relate their results to the ecological quality status. Water 2021, 13, 886 3 of 19 3 of 19 In this paper, we present a method to perform an ecological gap analysis using a system’s thinking approach, as suggested by Voulvoulis et al. [6]. The objective is to combine data on the biological status and the state of the supporting elements, being the physical–chemical status and hydromorphological status, to provide information for river managers on what aspects of a river system are stressors and have a need for management to restore the ecological status. The presented method can be used as part of a pressure and impact analysis following the requirements of the Water Framework Directive. To this end, the method analyses the state of a water body in terms of its ecological gap. This ecological gap can be interpreted as the deviation of the current biological status to its desired situation, being the good ecological status. This can be represented by a classiﬁcation of the ecological status, but in this method, we go a step further and disassemble the ecological gap into its individual building blocks, being the community composition. The method thus compares the current community composition to an operational Leitbild [26] (good ecological state) to identify taxa that represent the ecological gap to the good ecological status. For each missing taxon, the method then evaluates the potential impact of elements of the abiotic state of the river system using habitat suitability models as presented in Bennetsen et al. [22]. 1. Introduction This results in a classiﬁcation of each abiotic variable in terms of their need for management to restore the ecological status. The objective of the paper is to present the method and its individual elements as developed for water bodies in Flanders, Belgium. We illustrate our method with an application on macroinvertebrate monitoring data from Flanders, using the Multimetric Macroinvertebrate Index Flanders (MMIF) [27]. We demonstrate that our method is sensitive enough to detect different types of impairments, that there is a good relationship with the quality status as deﬁned by the MMIF and that our method allows ﬂexibility when deﬁning targets for river restoration. In this paper, we present a method to perform an ecological gap analy system’s thinking approach, as suggested by Voulvoulis et al. [6]. The objectiv bine data on the biological status and the state of the supporting elements, bein ical–chemical status and hydromorphological status, to provide informatio managers on what aspects of a river system are stressors and have a need for m to restore the ecological status. The presented method can be used as part of and impact analysis following the requirements of the Water Framework Di this end, the method analyses the state of a water body in terms of its ecologica ecological gap can be interpreted as the deviation of the current biological s desired situation, being the good ecological status. This can be represented b cation of the ecological status, but in this method, we go a step further and d the ecological gap into its individual building blocks, being the community co The method thus compares the current community composition to an operation [26] (good ecological state) to identify taxa that represent the ecological gap t ecological status. For each missing taxon, the method then evaluates the poten of elements of the abiotic state of the river system using habitat suitability mod sented in Bennetsen et al. [22]. This results in a classification of each abiotic terms of their need for management to restore the ecological status. The obje paper is to present the method and its individual elements as developed for w in Flanders, Belgium. We illustrate our method with an application on macroin monitoring data from Flanders, using the Multimetric Macroinvertebrate Inde (MMIF) [27]. 1. Introduction We demonstrate that our method is sensitive enough to detect diff of impairments, that there is a good relationship with the quality status as def MMIF and that our method allows flexibility when defining targets for river r 2.1. Overview of the Method 2.1. Overview of In our method, we use a set of suitability models for 92 macroinvertebrates as devel- oped in Bennetsen et al. [22] in combination with biotic monitoring data, to classify aspects of river system according to their priority for management to restore the ecological status (Figure 1). In our method, we use a set of suitability models for 92 macroinvertebrate oped in Bennetsen et al. [22] in combination with biotic monitoring data, to clas of river system according to their priority for management to restore the ecolo (Figure 1). verview of our method. The method classifies aspects of a river system in 4 priority classes by iden from the reference lists and revealing their most important stressors through the calculation of the sui c environment for these taxa. The abiotic state is classified into priority classes according to their res on for obtaining a good ecological status. Figure 1. Overview of our method. The method classiﬁes aspects of a river system in 4 priority classes by identifying missing taxa from the reference lists and revealing their most important stressors through the calculation of the suitability of the abiotic environment for these taxa. The abiotic state is classiﬁed into priority classes according to their respective need for action for obtaining a good ecological status. erview of our method. The method classifies aspects of a river system in 4 priority classes by iden from the reference lists and revealing their most important stressors through the calculation of the suit environment for these taxa. The abiotic state is classified into priority classes according to their resp Figure 1. Overview of our method. The method classiﬁes aspects of a river system in 4 priority classes by identifying missing taxa from the reference lists and revealing their most important stressors through the calculation of the suitability of the abiotic environment for these taxa. The abiotic state is classiﬁed into priority classes according to their respective need for action for obtaining a good ecological status. Water 2021, 13, 886 4 of 19 In a ﬁrst step, we compared the biological community observed in a river to a list of taxa describing the operational Leitbild [26] to identify missing taxa that represent the ecological gap. Secondly, for all taxa in the operational Leitbild that are not present in the community currently observed, we calculate the suitability of the current abiotic conditions with their respective habitat suitability models. 2.2. Deﬁning an Operational Leitbild 2.2. Deﬁning an Operational Leitbild 2.1. Overview of the Method 2.1. Overview of Input data for these models are physical– chemical and hydromorphological variables (Table SA3). All variables describe different aspects of a river system of which alterations of the natural condition could potentially pose a negative impact on the state of the water body. We then classify the resulting habitat suitabilities of each abiotic variables into different priority classes according to their need for management actions and to what degree they can be considered stressors. g g y In what follows, we first describe the construction of the operational Leitbild (Section 2.2) that the method uses to select the habitat suitability models to perform the analysis. This Leitbild represents a description of the desired situation per water body type, being the good ecological status. It consists of two elements: a description of the expected community in terms of ecological groups (Section 2.2.1), and a taxa list containing taxa that are members of these ecological groups (Section 2.2.2). We describe this operational Leitbild in terms of ecological groups because it allows one to reduce the number of Type I errors, where a taxon is classiﬁed as missing, even though another taxon sharing a similar ecological niche is present (thus indicating that the habitat is suitable for the taxon considered). For example, if a taxon A on the reference list is missing from the community observed and no other taxon from its ecological group is found in a sample, this taxon A is considered missing. Next, we present the algorithm to apply the method (Section 2.3) and its evaluation (Section 2.4). We then describe the example application for the region of Flanders, Belgium (Section 2.5). 2.2. Deﬁning an Operational Leitbild 2.2.1. Deﬁning Ecological Groups As the goal of our method is to provide information on the need for action to reach a good ecological status, it is not necessary that a current community exactly matches the ecological assemblage on our taxa lists. We consider these lists to be descriptions of an assemblage of species archetypes, in which each taxon on the list represents a speciﬁc ecological niche. Our goal was to reduce the number of Type I errors, where a taxon would be classiﬁed as missing, even though another taxon sharing a similar ecological niche was present. This could, for example, be the case as a different taxon sharing a similar niche would have a larger presence in the stream basin of the water body, than the taxon on our taxa list. Functionally, they would represent the same group of taxa, but taxonomically they differ. Using this method, we also avoided focusing on taxonomical composition alone. It can be seen as an approximation of functional diversity [28,29]. For this purpose, we deﬁned a set of ecological groups, using data on ecological preferences from Verberk and Nijboer [30]. We therefore conducted a data-driven cluster analysis to identify clusters of taxa, similar to the analysis presented in Usseglio-Polatera et al. [31]. The aim of these clusters was to deﬁne ecological groups of taxa that have similar ecological preferences. The dataset contained autecological information for 2325 species for the following variables: trophic state, saprobic state, depth, velocity, substrate, salinity, water surface area, intermittent ﬂow and acidity. This selection of ecological traits also align with the suggested categories presented by Mueller et al. [28]. Only taxa that are part of the taxa list considered in the MMIF calculation [27] were taken into consideration. We used divisive analysis clustering with a Gower coefﬁcient as the dissimilarity index to construct a hierarchical clustering tree of species [32]. We speciﬁcally selected the Gower coefﬁcient, as it can handle mixed data. The data on ecological traits are ordinal (fuzzy-coded). We set the maximum number of clusters to 20. As the data in Verberk and Nijboer [30] were deﬁned on the species level, and our data were deﬁned on the genus or the family level, we had to aggregate the results to a higher taxonomic level. 2.2.2. Constructing the Taxa Lists Describing the Operational Leitbild 2.2.2. Constructing the Taxa Lists Describing the Operational Leitbild The reference condition for macroinvertebrates in Flanders has been deﬁned based on expert judgment on the expected values of each metric in the MMIF by Gabriëls et al. [27]. Due to this, there are no taxa lists describing the expected community composition in each river type under reference conditions. Therefore, there are also no taxa lists describing the expected community at a good ecological status of each water body type. To overcome this, we developed lists of taxa for seven river types in Flanders, which describe the expected community composition under a good ecological status as deﬁned by the MMIF. As the WFD requires member states the good ecological status the statutory requirement for water bodies, this is the desired situation and management endpoint for practitioners. Our taxa reference lists make a distinction between “base taxa”—generalist taxa that are generally present in this water body even at a lower than good ecological status—and “critical taxa”— sensitive taxa that only occur when a good ecological status is reached. The idea is that a good ecological status can only be reached with speciﬁc management that will lead to the (re)introduction of certain critical taxa in addition to these base taxa. These critical taxa are assumed to mainly occur in the good and high ecological status water bodies. This means that our benchmark describes a “least disturbed” or “best attainable condition” [33] rather than a “natural” or “historical” condition. We developed the taxa lists using a dataset from the biological monitoring network of the Flanders Environment Agency. It contains 2910 samples between 2003 and 2012 describing the community composition observed at 714 sites in seven river types in Flanders. It also contains a classiﬁcation of each sample according to the MMIF. Macroinvertebrate monitoring data were collected by kick sampling following De Pauw and Vanhooren [34]. The identiﬁcation of the macroinvertebrates was carried out according to the taxonomic levels deﬁned by Gabriels et al. [27], i.e., according to family, genus or another intermediate taxonomic level. A standardized list of taxa is provided in Gabriels et al. [27]. A full description of the datasets and the water quality status in the study area can be found in Tables SA1 and SA2. 2.2.1. Deﬁning Ecological Groups For each genus or family on our list of taxa, we determined the Water 2021, 13, 886 5 of 19 relative proportion of species in this taxon that was part of each group. We then assigned each taxon to the group with the highest proportion. To account for a loss of information in transforming these groups from a species level to a higher taxonomic level, we consulted three macroinvertebrate experts for Flanders. They were asked to check the proposed groups against the ﬁrst criterion: similar ecological needs, but no assumed co-occurrence. The Flemish experts could also propose corrections to the original group compositions. We checked their propositions against the evidence that was provided by the cluster analysis (cfr. above). During the discussion, taxa were assigned to different groups by consensus. 2.2.2. Constructing the Taxa Lists Describing the Operational Leitbild 6 of 19 Water 2021, 13, 886 Algorithm 1 Deﬁning an operational Leitbild Algorithm 1 Deﬁning an operational Leitbild g g p Data: dataset with macroinvertebrate samples and their respective MMIF score from 2003 to 2012 Result: taxa list describing the good ecological status per water body type initialization; FOREACH water body type W FOREACH taxon t Calculate number of occurrences O IF O > 40 Calculate Fidelity F for taxon t in all samples with MMIF ≥0.7 END END FOREACH species group G Calculate median M and mode Mo of the number of representatives of group G in all samples with MMIF ≥0.7 Select M taxa with the highest F from G and add to taxa list END Calculate MMIF mmifcal with taxa list using RA WHILE mmifcal <0.7 FOREACH metric that is too low for a good status Select species group Gm that contributes to metric with the highest Mo and of which the number of taxa on the existing list is lower than its respective Mo Select taxon t with the next highest F from Gm and add to taxa list END Calculate MMIF mmifcal with taxa list using RA END END Data: dataset with macroinvertebrate samples and their respective MMIF score from 2003 to 2012 Result: taxa list describing the good ecological status per water body type initialization; FOREACH water body type W FOREACH taxon t Calculate number of occurrences O IF O > 40 Calculate Fidelity F for taxon t in all samples with MMIF ≥0.7 END END FOREACH species group G Calculate median M and mode Mo of the number of representatives of group G in all samples with MMIF ≥0.7 Select M taxa with the highest F from G and add to taxa list END Calculate MMIF mmifcal with taxa list using RA WHILE mmifcal <0.7 FOREACH metric that is too low for a good status Select species group Gm that contributes to metric with the highest Mo and of which the number of taxa on the existing list is lower than its respective Mo Select taxon t with the next highest F from Gm and add to taxa list END Calculate MMIF mmifcal with taxa list using RA END END In a second step, we made sure that the taxa lists actually described a good ecological status, by calculating the expected MMIF with these taxa present (Section SB1). 2.2.2. Constructing the Taxa Lists Describing the Operational Leitbild We developed the taxa lists in two steps (Algorithm 1) to overcome the bias towards bad and poor ecological status in the dataset (Table SA1) and to avoid classifying sites focused on taxonomical composition alone. In the ﬁrst step, we described the operational Leitbild by the species groups as were deﬁned in Section 2.2.1. For this, we calculated the median and the mode of the number of individual taxa that were present per group at each sample with good or high ecological status. The median represents the typical number of taxa within each group that would be present at good ecological status. We then developed a list of candidate taxa with a data-driven analysis. To develop the initial list of candidate taxa, we calculated the ﬁdelity score of each taxon to sites of a river type with a good or high ecological status. Fidelity is the degree to which a taxon prefers a water body type with a good or high ecological status, deﬁned as the ratio of the relative frequency of a taxon in the subset of samples of sites in a river type with a good or high ecological status and their overall relative frequency [35]. We ranked the taxa according to their ﬁdelity score, and for each river type we added the number of taxa per group as was calculated in the ﬁrst step, starting with the taxon with the highest ﬁdelity. This resulted in a ﬁrst draft of the taxa list. 2.2.2. Constructing the Taxa Lists Describing the Operational Leitbild The MMIF is calculated using 5 submetrics: the number of taxa, the number of sensitive taxa, the mean tolerance score, the Shannon–Wiener index and the number of taxa from the group of Ephemeroptera, Plecoptera or Trichoptera [27]. If the MMIF for the proposed list of taxa did not reach 0.7 (the cutoff for a good ecological status) for a water body type, we reviewed the submetric values to assess which metric caused the lower MMIF value. If necessary we added one extra taxon from an ecological group to the list for that water body type, which would correct this submetric. When multiple groups were possible, we chose the group for which the mode of this group in this water body type was higher than the median as calculated above. We then iteratively went through this procedure until the MMIF associated with the taxa list for each water body type reached minimally 0.7. Furthermore, only taxa that had more than 40 occurrences in the dataset were considered, as there were no models for taxa with lower occurrences [22]. Taxa with a tolerance score lower than 6 (not very sensitive to pollution) were consid- ered base taxa. Taxa with a tolerance score of 6 or higher were considered critical taxa. This threshold coincides with the threshold related to the submetric number of sensitive taxa in the MMIF calculation [27]. 2.3. Classifying Aspects of a River System into Priority Classes Our method uses a set of suitability models as presented in Bennetsen et al. [22]. We based the model development on ecological theory and calibrated the models using Flemish monitoring data and databases with ecological knowledge. We quantiﬁed the added value of incorporating ecological theory and knowledge on ecological preferences. One model for one taxon consists of different non-symmetric unimodal trapezoid functions describing the univariate suitability per abiotic variable. This simpliﬁcation of the typical bell-shaped curve makes it possible to describe various types of distributions with different degrees of skewness and kurtosis [36]. For a detailed description of the model development, we refer to Bennetsen et al. [22]. A list of all input variables can be found in Table SA3. All input variables are variables describing the state of a water body. All variables included in Water 2021, 13, 886 7 of 19 Table iables 7 of 19 Table iables the models have legal targets deﬁned. When they deviate from the legal target and can be related to pressures in the water body system, they can potentially be considered a stressor. All hydromorphological variables included are deﬁned as ordinal variables describing the degree of alteration of its state from the expected natural condition [22]. ered a stressor. All hydromorphological variables included are defined as ordinal varia- bles describing the degree of alteration of its state from the expected natural condition [22]. To define the priority class for each stressor (Algorithm 2), the current community is g p To deﬁne the priority class for each stressor (Algorithm 2), the current community is ﬁrst compared to the taxa lists deﬁned in Section 2.2. To identify a taxon as missing, we ﬁrst compare the community observed in terms of ecological groups. If in the community observed the number of taxa in an ecological group is lower than the number of taxa in this ecological group on the reference list, this amount of taxa is considered missing. This number represents the ecological gap. In a second step, we selected from the taxa list this number of taxa per ecological group. These are referred to as the missing taxa. These will be the taxa for which we used the suitability models in the priority classiﬁcation. These selected taxa cannot be present in the community observed. For each missing taxon, we then calculated the suitability index SI for each abiotic variable. 2.3. Classifying Aspects of a River System into Priority Classes This SI was quantiﬁed as a score between 0 and 1, which we divided into four priority classes. Each of these classes represents the need for management for each abiotic variable (Figure 2). In the ﬁrst class (SI = 0, “Out of range, needs action”), a variable is out of range of suitability for a taxon, meaning that it poses an important pressure on the biological status and can be considered a stressor. The good ecological status cannot be reached unless this stressor is addressed. In the second class (0 < SI ≤cutoff value, “Consider action”), the site has low habitat suitability for this abiotic variable. In this range, the degree to which this aspect of a river system can be considered a stressor depends on the interaction between different variables of lower suitability for this taxon. The cutoff for this suitability value is derived from the so-called interaction threshold as calibrated in Bennetsen et al. [22]. This interaction threshold represents the idea that if the suitability is low according to many variables, the likelihood that a taxon can be present decreases. In the third class (cutoff value < SI < 1, “Flagged”), a variable is suboptimal, which indicates that this variable probably does not pose an important stressor and should not be addressed with a high priority even if it deviates from its legal target. Still, there may be a risk for rapid deterioration when its value changes due to an incident or unexpected effected of management interventions. In the fourth case (SI = 1, “No action needed”), the habitat, according to this variable, is considered as suitable, which means that it is not a stressor and that it should not be taken into account when planning management actions in this river for macroinvertebrates. first compared to the taxa lists defined in Section 2.2. To identify a taxon as missing, we first compare the community observed in terms of ecological groups. If in the community observed the number of taxa in an ecological group is lower than the number of taxa in this ecological group on the reference list, this amount of taxa is considered missing. This number represents the ecological gap. In a second step, we selected from the taxa list this number of taxa per ecological group. These are referred to as the missing taxa. 2.3. Classifying Aspects of a River System into Priority Classes These will be the taxa for which we used the suitability models in the priority classification. These selected taxa cannot be present in the community observed. For each missing taxon, we then calculated the suitability index SI for each abiotic variable. This SI was quantified as a score between 0 and 1, which we divided into four priority classes. Each of these classes represents the need for management for each abiotic variable (Figure 2). In the first class (SI = 0, “Out of range, needs action”), a variable is out of range of suitability for a taxon, meaning that it poses an important pressure on the biological status and can be considered a stressor. The good ecological status cannot be reached unless this stressor is addressed. In the second class (0 < SI ≤ cutoff value, “Consider action”), the site has low habitat suit- ability for this abiotic variable. In this range, the degree to which this aspect of a river system can be considered a stressor depends on the interaction between different varia- bles of lower suitability for this taxon. The cutoff for this suitability value is derived from the so-called interaction threshold as calibrated in Bennetsen et al. [22]. This interaction threshold represents the idea that if the suitability is low according to many variables, the likelihood that a taxon can be present decreases. In the third class (cutoff value < SI < 1, “Flagged”), a variable is suboptimal, which indicates that this variable probably does not pose an important stressor and should not be addressed with a high priority even if it deviates from its legal target. Still, there may be a risk for rapid deterioration when its value changes due to an incident or unexpected effected of management interventions. In the fourth case (SI = 1, “No action needed”), the habitat, according to this variable, is con- sidered as suitable, which means that it is not a stressor and that it should not be taken into account when planning management actions in this river for macroinvertebrates. Figure 2. Visual representation of the translation of the suitability index (SI) into priority classes. The variable (e.g., oxygen concentration) is shown on the x-axis. The y-axis shows the suitability Figure 2. Visual representation of the translation of the suitability index (SI) into priority classes. The variable (e.g., oxygen concentration) is shown on the x-axis. 2.3. Classifying Aspects of a River System into Priority Classes = 1 Set priority P to 1 ELSE IF SI < 1 Set priority P to 2 ELSE IF SI = 1 Set priority P to 3 END END FOREACH abiotic variable AV Priority Class PC = ﬂoor(weighted harmonic mean (P)) with weight = 1\|2 for respectively base and critical taxa END END END FOREACH sample S in D IF MMIF ≥0.7 Set number of missing taxa to 0 NMT Set priority P for all abiotic variables to 0 ELSE Identify missing taxa TM Count number of missing taxa NMT FOREACH TM and abiotic variable AV Calculate suitability index SI IF SI = 0 Set priority P to 0 ELSE IF SI < cutoff value CV & CV! = 1 Set priority P to 1 ELSE IF SI < 1 Set priority P to 2 ELSE IF SI = 1 Set priority P to 3 END END FOREACH abiotic variable AV Priority Class PC = ﬂoor(weighted harmonic m END END END To obtain a priority class for each variable per water body, we aggregated the priority classes per variable and per missing taxon using a weighted harmonic mean. Critical taxa receive double the weight of base taxa. As we used the harmonic mean, taxa for which the variable has a lower suitability would have more impact on the ﬁnal value for classiﬁcation [37]. The aggregated priority scores were rounded to their lowest integer value and were coded into the same four classes as the individual classes. This results in a classiﬁcation of each aspect of the river system of each water body in the dataset. 2.4. Evaluation of Our Method Each individual suitability model was validated and tested as described in Bennetsen et al. [22]. We further tested our method, by calculating the expected MMIF by executing a simulation of the models developed in Bennetsen et al. [22] for the taxa that were included in the reference list, using a test dataset that was not used in model development, with data gathered from 2003 to 2012. For this we used the method described in Section B1 in supportive information. We compared this simulated MMIF value with the true MMIF, by constructing the confusion matrix on the classiﬁed MMIF and calculating Cohen’s Kappa and the accuracy. We also compared the accuracy to the no information rate (NIR), which is the proportion of the most prevalent class in the test dataset. The accuracy should be higher than this NIR to be acceptable [38]. For the MMIF value, we also calculated the RMSE. We also calculated the Pearson correlation of the number of missing taxa and the number variables with a priority class of 0 or 1 to the MMIF score. 2.3. Classifying Aspects of a River System into Priority Classes The y-axis shows the suitability index ranging from 0 to 1, for a missing taxon. In case the suitability is 0, the variable is classiﬁed as “out of range” (red). In case suitability is between 0 and the interaction threshold, the variable is classiﬁed as “consider action” (orange). If the suitability is between the interaction threshold and 1, the variable is classiﬁed as “ﬂagged” (yellow) and in case the suitability equals 1 it is classiﬁed as “no action required” (green). Figure 2. Visual representation of the translation of the suitability index (SI) into priority classes. The variable (e.g., oxygen concentration) is shown on the x-axis. The y-axis shows the suitability Figure 2. Visual representation of the translation of the suitability index (SI) into priority classes. The variable (e.g., oxygen concentration) is shown on the x-axis. The y-axis shows the suitability index ranging from 0 to 1, for a missing taxon. In case the suitability is 0, the variable is classiﬁed as “out of range” (red). In case suitability is between 0 and the interaction threshold, the variable is classiﬁed as “consider action” (orange). If the suitability is between the interaction threshold and 1, the variable is classiﬁed as “ﬂagged” (yellow) and in case the suitability equals 1 it is classiﬁed as “no action required” (green). Water 2021, 13, 886 8 of 19 Algorithm 2 Deﬁning priority classes for individual abiotic variables Algorithm 2 Deﬁning priority classes for individual abiotic variables Data: dataset with macroinvertebrate samples, their respective abiotic conditions and their respective MMIF score from 2010–2012 D; taxa list describing a reference condition per water body type TL; a set of habitat suitability models HSM Result: description of missing taxa; classiﬁcation of priorities initialization; p g p initialization; FOREACH sample S in D IF MMIF ≥0.7 Set number of missing taxa to 0 NMT Set priority P for all abiotic variables to 0 ELSE Identify missing taxa TM Count number of missing taxa NMT FOREACH TM and abiotic variable AV Calculate suitability index SI IF SI = 0 Set priority P to 0 ELSE IF SI < cutoff value CV & CV! 3.1.2. Taxa Lists The ﬁnal taxa lists (Table SC2) describe community assemblages with 25–30 taxa. Of these taxa, between 35% and 45% of taxa were identiﬁed as critical taxa, with a high ﬁdelity score to the good water quality class. The distinction between critical taxa and base taxa is reﬂected in the tolerance scores as deﬁned by Gabriëls et al. [27]. Critical taxa have a tolerance score of 6 and above, indicating high sensitivity to pollution. 3.1.1. Ecological Groups Overview of the composition of the ecological groups in terms of species order (%) and numb 3.1. Deﬁning the Operational Leitbild 3.1.1. Ecological Groups 3.1. Deﬁning the Operational Leitbild 2.5. Example Application: The Case of Flanders We illustrate our method by applying it to a dataset of sampling data from 381 mon- itoring points collected between 2010 and 2012 in Flemish water bodies and local water Water 2021, 13, 886 9 of 19 9 of 19 bodies of the 1st order. To compile the abiotic input dataset, we coupled the biotic moni- toring data to the statistical derivative of the physical–chemical abiotic data, as used for assessing water quality against the water quality norms in Flanders (Table SA2) and the hydromorphological observations reported for each water body. The statistical derivatives represent an overall view of the water quality per water body for the full year. All data were obtained from the WFD monitoring network run by the Flanders Environment Agency [39]. A full description of the dataset and the water quality status in the study areas can be found in Table SA2. 3.1.1. Ecological Groups The ecological groups represents groups of taxa that have similar habitat needs, but do not necessarily co-occur. Based on the hierarchical cluster analysis and the expert consultation, we identiﬁed 16 ecological groups (Table SC1). Table 1 gives an overview of the ﬁnal taxonomic composition of each group, and how many taxa were assigned to each group. The groups are mainly clustered on the ecological preferences, and the tolerance scores as deﬁned in Gabriëls et al. [27]. However, some taxonomical clustering is also observed. e composition of the ecological groups in terms of species order (%) and number of taxa per group Table 1. Overview of the composition of the ecological groups in terms of species order (%) and number of taxa per group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Coleoptera 0 0 0 56 0 7 0 0 0 0 0 0 0 0 0 0 Crustacea 0 0 0 0 8 0 0 0 7 9 0 0 100 0 0 0 Diptera 0 0 0 0 75 21 0 0 0 18 0 0 0 100 0 0 Ephemeroptera 0 0 0 0 0 7 38 67 0 0 0 0 0 0 0 0 Hemiptera 0 0 0 31 0 0 8 0 86 0 0 0 0 0 0 0 Hirudinea 0 0 0 0 0 0 0 0 0 0 0 67 0 0 0 0 Megaloptera 0 0 0 0 0 0 0 0 0 0 0 0 0 0 50 0 Mollusca 100 100 100 0 0 0 0 0 0 55 0 0 0 0 50 0 Odonata 0 0 0 6 0 0 15 0 0 0 100 0 0 0 0 0 Oligochaeta 0 0 0 6 8 0 0 0 0 18 0 0 0 0 0 0 Plathelminthes 0 0 0 0 0 0 0 0 0 0 0 33 0 0 0 0 Plecoptera 0 0 0 0 0 29 0 0 0 0 0 0 0 0 0 0 Polychaeta 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 Trichoptera 0 0 0 0 0 36 38 33 7 0 0 0 0 0 0 100 Number of taxa 8 9 5 16 12 14 13 3 14 11 17 12 10 8 2 4 Table 1. 3.3. Example Application: The Case of Flanders The application of our method on the regional level of Flanders illustrates how the ecological status can serve as a classiﬁer to determine the priority of management of different aspects of a river system (Table 2). Overall, hydromorphological variables were almost never classiﬁed as being out of range, except variables related to the proﬁle (width and depth) and the presence of sediment banks. Physical–chemical variables were classiﬁed in on average 15% of cases as being out of range. The variables with the highest priority in Flanders are electrical conductivity (28%), Kjeldahl nitrogen concentration (25%) and nitrate concentration (19%). Some variables are mostly classiﬁed as “consider action”, because of lower suitability, such as the presence of plants (46%) and width variation (50%). The variables that are most often classiﬁed as suboptimal are: presence of dead wood (84%), variables related to stream patterns (pool-rifﬂe pattern (69%) and stream variation (64%)). Only depth is to a high degree classiﬁed as having no need for action (60%). Table 2. Overview of the classiﬁcation of each stressor according to their management need (% samples). The last two columns describe the overlap between stressors classiﬁed as low to no priority that do not comply with their legal targets, and stressors classiﬁed as moderate to high priority that comply with their legal targets. For some variables no speciﬁc legal targets exist, which is indicated by NA in the last two columns. 3.2. Evaluation of the Method Using the suitability models deﬁned in Bennetsen et al. [22] for the taxa on the lists describing the operational Leitbild, we validated the predictive power of the models for the MMIF classiﬁcation and score on a dataset independent from the dataset used in model development. Of all samples 47.7% were correctly classiﬁed, Cohen’s Kappa was 0.26 (p = 0.05) and the RMSE value on the MMIF score was 0.18. This accuracy was higher than the no information rate 32.96% (p = 0.008). There is a monotonic decreasing relationship between the results of the priority classiﬁcation and the ecological status expressed as the MMIF. The number of variables Water 2021, 13, 886 10 of 19 10 of 19 that have a low to high priority was negatively correlated with the MMIF score of the water body (r = −0.72, p < 0.001). The number of missing taxa (median count 14) was negatively correlated (r = −0.92, p < 0.001) with the MMIF value of the water body. If no ecological groups were considered, more missing taxa (median count 18) were identiﬁed. In this case the correlation of the number of missing taxa to the MMIF score would drop to −0.86 (Pearson correlation coefﬁcient, p < 0.001). 3.3. Example Application: The Case of Flanders Variable Out of Range Consider Action Flagged No Action Required Not good Status with Low Priority Good Status with High Priority Bank structure 0.0 22.1 61.3 16.6 2.1 5.8 Biological oxygen demand 12.3 32.8 40.2 14.7 0.0 25.4 Chemical oxygen demand 9.7 29.7 37.0 23.6 0.0 1.3 Chloride concentration 15.8 6.8 61.2 16.3 0.0 5.1 Curvature erosion 0.0 50.1 29.7 20.2 NA NA Depth 9.2 5.3 25.2 60.4 28.9 0.4 Electrical conductivity 28.0 6.3 50.1 15.6 0.0 7.3 Kjeldahl nitrogen concentration 24.7 16.3 40.2 18.9 0.0 17.7 Lateral connectivity 0.3 18.6 66.9 14.2 4.7 13.4 Micromeandering 0.0 4.7 81.1 14.2 NA NA Nitrate-nitrogen concentration 18.7 28.9 28.6 23.9 0.0 30.5 Orthophosphate- phosphorus concentration 9.2 15.2 58.8 16.8 11.0 7.0 Overhanging vegetation 0.0 17.3 67.7 15.0 2.0 5.5 Oxygen concentration 14.2 25.5 39.6 20.7 0.8 8.0 Oxygen saturation 7.6 18.9 52.8 20.7 1.3 14.4 pH 14.7 2.4 56.7 26.3 0.0 12.0 Pool-rifﬂe pattern 0.0 15.2 69.3 15.5 0.0 7.5 Presence of algae 0.0 69.0 8.5 22.6 6.0 4.9 Presence of dead wood 0.3 1.3 84.2 14.2 30.6 1.5 Presence of macrophytes 0.0 45.5 34.2 20.3 15.4 2.4 Presence of sediment banks 2.9 2.1 79.3 15.8 30.8 0.9 Presence of water-related elements 0.0 30.8 53.4 15.8 18.3 1.0 Water 2021, 13, 886 11 of 19 Table 2. Cont. Variable Out of Range Consider Action Flagged No Action Required Not good Status with Low Priority Good Status with High Priority Sinuosity 0.0 5.4 68.3 26.3 28.1 2.0 Sludge layer depth 0.0 41.8 39.5 18.7 12.7 4.9 Stream variation 0.3 20.1 65.0 14.7 0.3 11.0 Substrate 1.7 14.2 63.9 20.3 15.2 0.4 Suspended solids 12.6 23.1 45.7 18.7 0.0 15.8 Temperature 11.0 10.0 57.0 22.1 0.0 19.0 Total phosphorus concentration 15.8 7.6 53.3 23.4 6.7 0.0 Width 4.5 20.7 38.1 36.8 19.0 1.1 Width erosion 0.0 42.3 40.4 17.3 NA NA Width variation 0.0 49.7 34.4 15.9 12.7 0.0 Table 2. Cont. Table 2. Cont. The legal targets set for physical–chemical and hydromorphological variables can differ from the results of the priority classiﬁcation (Table 2). For example, in 25% of samples the biological oxygen demand complies with its legal target, but according to our model results, there is still a need for action. 3.3. Example Application: The Case of Flanders On the other hand, there are water bodies whose proﬁle has been modiﬁed, resulting in a non-compliance for the width-to-depth ratio with the legal target, yet the width and depth of these water bodies do not require action in respectively 19% and 29% of those samples for macroinvertebrates. p y p On average, 13 missing taxa were identiﬁed, and 9 variables were classiﬁed as in need of action (Figure 3A,B). There is a gradient in the number of missing taxa and in the number of limiting variables from west to east, which coincides with the gradient of the MMIF from west to east (Figure SD1). There are regional differences between the ways in which each stressor is classiﬁed according to their need for management (Figure 4A–C). Width variation is classiﬁed as “consider action” in smaller subsidiaries of larger water bodies. Biological oxygen demand can be of high priority all throughout Flanders, while or- thophosphate phosphorus concentration speciﬁcally needs to be addressed in the westside of the region. 12 of 19 12 of 19 Water 2021, 13, 886 Figure 3. Geographical representation of the number of missing taxa (A) and the number of limiting variables (B). The visualizations are made on the stream basins of each individual water body in Flanders. Figure 3. Geographical representation of the number of missing taxa (A) and the number of limiting variables (B). The visualizations are made on the stream basins of each individual water body in Flanders. gure 3 Geographical representation of the number of missing taxa (A) and the number of limiting variables (B) Th gure 3. Geographical representation of the number of missing taxa (A) and the number of limiting variables (B). The Figure 3. Geographical representation of the number of missing taxa (A) and the number of limiting variables (B). The visualizations are made on the stream basins of each individual water body in Flanders. Figure 3. Geographical representation of the number of missing taxa (A) and the number of limiting variables (B). The visualizations are made on the stream basins of each individual water body in Flanders. 13 of 19 13 Water 2021, 13, 886 Water 2020 12 igure 4. Geographical representation of the priority classification of width variation (A), biological oxygen nd orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each b d i Fl d 4. 4.1. Method Development In this paper, we presented a method for classifying individual aspects of a river system according to their need for management, combining insights from biological moni- toring and ecological modeling. The goal was to increase the value of individual biological measurements as explanatory tools, when planning for management actions in river restora- tion. To this end, we disassembled the biotic index for macroinvertebrates, the MMIF, into its individual building blocks. g The results of our method had a good correlation with MMIF values. Both the number of missing taxa identiﬁed from the reference lists, and the number of stressors identiﬁed have a good correlation with the MMIF. The monotonic negative correlation of the MMIF with the number of missing taxa identiﬁed strengthens from −0.86 to −0.92 when ecological groups are included in this step. Hence, the inclusion of ecological groups reduces the number of type I errors in our classiﬁcation exercise. This reduces the risk of considering an aspect of a river system as a stressor, when no management action is needed, a reduction that can result in more cost-effective management [40]. Furthermore, the composition of the ecological groups is consistent with ecological preferences and to a lesser extent with taxonomical composition, but the latter is not a prerequisite for the purpose of this tool [41]. The combination of the taxa reference lists with the ecological groups can be linked to the idea of species surrogates for a good ecological status [42]. For inclusion of candidate taxa on the taxa lists describing the operational Leitbild, we considered both samples with a good and high ecological status in the dataset. This increased the applicability of the integrated model, as, in practice, a good ecological status will be considered the ﬁrst management endpoint for seeking compliance with the WFD [6]. Bennetsen et al. [22] indicate that there is a good relationship between the performance of the suitability models used and the inclusion of ecological knowledge in the construction of the models. For example, compared to the use of threshold indicator taxa analysis (TITAN) to detect change points [43], both the average sensitivity (0.57 vs. 0.48) and speciﬁcity (0.87 vs. 0.67) for the prediction of species occurrence are higher. Furthermore, using the habitat suitability models from Bennetsen et al. [22], we assessed the predictive power for the MMIF of the models used. 4.1. Method Development The accuracy of our models (47.7%) is within of the interval of accuracy (45–57%) presented when classifying the MMIF score directly with classiﬁcation trees [44]. Major advantages of the presented method are the insights that are gained in the shifts in community aspects (diversity, tolerance and speciﬁc groups), which types of invertebrates are likely to occur or disappear as a consequence of changes in abiotic conditions and moreover an improved insight related to the uncertainty in the models and their predictions. 3.3. Example Application: The Case of Flanders Geographical representation of the priority classiﬁcation of width variation (A), biological oxygen demand ( hosphate phosphorus concentration (C). The visualizations are made on the stream basins of each individual Figure 4. Geographical representation of the priority classification of width variation (A), biological oxygen dem and orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each ind water body in Flanders. re 4. Geographical representation of the priority classiﬁcation of width variation (A), biological oxygen demand (B) a ophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each individual wa y in Flanders. 4 G hi l i f h i i l iﬁ i f id h i i (A) bi l i l d d (B) Figure 4. Geographical representation of the priority classification of width variation (A), biological oxygen demand and orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each indivi water body in Flanders. Figure 4. Geographical representation of the priority classiﬁcation of width variation (A), biological oxygen demand (B) and orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each individual water body in Flanders. Figure 4. Geographical representation of the priority classification of width variation (A), biological oxygen deman and orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each indiv water body in Flanders. Figure 4. Geographical representation of the priority classiﬁcation of width variation (A), biological oxygen demand (B) and orthophosphate phosphorus concentration (C). The visualizations are made on the stream basins of each individual water body in Flanders. Water 2021, 13, 886 14 of 19 14 of 19 4. Discussion 4.1. Method Development 4.3. Relationship of Our Method to Legal Targets The results of our integrated model indicate that taxon speciﬁc responses have a different sensitivity to identify stressors than indicated by the legal targets set for physical– chemical and hydromorphological variables. The norms for these quality elements are based on supporting the ecological status [6], but the number of sites that are selected as being priority sites for management differ from the number of sites that do not comply with the norms in the WFD [39]. Hence, for reaching a good ecological status, river managers should not only focus on the compliance of individual elements, but rather consider a systems approach to the catchment [6]. This is possible as there are 48 sites that have a “good” MMIF status, whose physical–chemical status is poor to moderate, and there are 10 sites whose physical–chemical status is “good” and MMIF is only moderate. This is corroborated by studies such as Sundermann et al. [7], who argue that these norms should be considered ecological background values, not change point values for biological groups. This does not mean that these legal targets have no basis, as they are supportive of more biological groups than macroinvertebrates alone. g g p Our method is not the only method with which taxon-speciﬁc tipping points could be deﬁned, there are examples with taxon-speciﬁc analyses using the TITAN algorithm [7]. Even the breakpoints in the classiﬁcation trees relating directly to the MMIF [44,49] could inform on change points speciﬁc to macroinvertebrates. Our method differs from these other change point detecting method by assigning a fuzzy classiﬁcation to these change points. When a suitability index is either 0 or 1, the classiﬁcation for the need of manage- ment is clear-cut, but when habitat suitability is between 0 and 1, this makes for a grayer zone. There might or might not be a need for management. We introduce a more risk-based approach in which aspects of a river system that are altered to a degree that they result in a lower habitat suitability, are designated as “consider action”, and variables with a higher suitability are “ﬂagged”. This informs river managers on uncertainty and allows for expert judgment on these aspects in local cases, using knowledge of stakeholders and river managers that cannot easily be captured in desktop analyses. This could increase the acceptance of our method by stakeholders and river managers [50]. 4.3. Relationship of Our Method to Legal Targets Yet it still makes it possible to screen different aspects of river systems on regional differences in terms of their negative impact on the biological status, without further expert judgment. 4.2. Example Application: The Case of Flanders Our method is successful in identifying the most important stressors in water bod- ies. It also allows the detection of regional differences. For example, the importance of orthophosphate concentration decreases when going from west to east in Flanders. This could be explained by a decreasing proportion of agricultural and urban land use from the west to the east (Section SA). Furthermore, the results also show a distinction between small and large water bodies in terms of priorities. In smaller water bodies, hydromorphological variables such as the presence of width variation have a higher importance than in larger water bodies. The presence of width variation can be considered an indicator of the pres- ence of diversity in microhabitats, like in, for example, a sequence of slow and fast ﬂowing segments, and is expected to be higher in smaller water bodies, than in larger ones [45]. The diversity in microhabitats has been shown to be important for macroinvertebrates in streams [46]. Overall, physical–chemical variables seem to require a higher priority than hydromorphological variables, which coincides with insights in the literature [9,47,48]. Yet, many hydromorphological variables, such as the presence of water-related elements, the presence of macrophytes, the presence of natural erosion processes and the proﬁle Water 2021, 13, 886 15 of 19 15 of 19 of a stream (width-to-depth ratio and width variation) are classiﬁed as “consider action” in between 30 and 50% of all water bodies. Many water bodies in Flanders are artiﬁcial or have been heavily modiﬁed, which is reﬂected in the moderate priority of channel processes that reﬂect habitat diversity in a watercourse, such as erosion processes and bedding vegetation. 4.5. Outlook and Future Development Our method is able to identify different types of stressors, has a good relationship with the ecological status and is ﬂexible in terms of target setting for river restoration. As each individual aspect of a river system is classiﬁed individually, it would for example also be possible to run these models on incomplete data, with the indication that on these aspects no information is available. Yet its univariate character is a potential disadvantage of our method, as no interaction between stressors is identiﬁed in our calculations. However, as the interaction value, used to designate variables either as “consider action” or “ﬂagged”, is based on the multivariate interaction between different stressors [22], interactions are nevertheless considered, albeit indirectly. Yet a possible improvement to the method would be an input variable selection, for example using genetic algorithms [36], to reduce the number of input variables, which would make the consideration of interactions between stressors a less complex problem. p p The method presented in this paper used an operational Leitbild developed using data from 2003 to 2012 for an application on a dataset from 2010 to 2012. To apply the method in the future on more recent data, or in different regions, it is advised to update this Leitbild with more recent biological data or data from that speciﬁc region to ensure that it describes the expected community under a good ecological status in terms of taxa that have a strong presence in the region and period considered. This will ensure that the results of the classiﬁcation are the most relevant to the most important taxa in the area at that time. Furthermore, dispersal or the recolonization potential is not considered in the method presented in this paper. Yet model performance improved signiﬁcantly when a simple dispersal metric was included in the initial model development [22]. Recently, several authors have indicated the importance of recolonization potential in planning ecological river restoration [54,55]. There are also water bodies in our example application with large ecological gaps, indicating many missing taxa for a good water ecological status, but where water quality is nevertheless good. Perhaps a better explanation than individual stressors would be the absence of source populations for these water bodies [55]. Furthermore, by combining our method with that proposed in Dahm et al. 4.5. Outlook and Future Development [56], water bodies that have a high potential of reaching good water quality status immediately after river restoration could be identiﬁed, and thus prioritized spatially. 4.4. Possible Applications of the Method The most important application of our method is that of turning a bulk of measurement data into structured information for river managers. This allows for setting regional priorities and local priorities: which areas should we address ﬁrst, and what aspects? Our method has similarities to that proposed by De Zwart et al. [51]. A more typical use of the habitat suitability models as presented in Bennetsen et al. [22] would be to assess the effect of different water quality management scenarios, by integrating them as the last set of models in an integrated modeling chain with other water quality models. However, we think that such an integrated modeling framework overlooks a crucial step in planning and decision support: how can we design management scenarios? To test a great number of different scenarios with water quality models can be very time-consuming, either owing to computational time and requirements, or to the sheer complexity of preparing the various input data. To reduce the number of scenarios that should be tested with water quality models, a crucial ﬁrst step is something that can be referred to as solution scanning [52]. By ﬁrst analyzing the system and making a diagnosis [5] of the impairment at hand, river managers can specify a limited set of management options [53] that could be tested in Water 2021, 13, 886 16 of 19 scenario calculations with perhaps more complex water quality models. Due to this, we argue to not only put the ecological models as the last model in an integrated modeling framework, but to use them also as a ﬁrst step in these frameworks. There is of course logic to putting these models last, as they typically represent the management endpoint within integrated water management. However, as other authors and we [7] have illustrated, ecological responses do not always align with information derived from legal targets. Performing an ecological data analysis ﬁrst makes it possible to perform a gap analysis sensitive to ecological responses. Thus, by putting the ecological models ﬁrst, our method aligns with the argument for a true systems approach in integrated water management [6]. 5. Conclusions In this study, we aimed to combine suitability models with information on the ecologi- cal status to classify aspects of a river system as potential stressors according to their need for management. We showed that disassembling the ecological status into its individual building blocks has great potential to integrate the ecological status from a systems perspec- tive into models supporting the WFD implementation and integrated water management in general. Speciﬁcally, our method considers the ecological status not only as an endpoint in decision support modeling to evaluate scenarios, but also considers it the starting point for a diagnostic analysis related to the pressure and impact analysis. More in particular, we found that the results of our method correlate well with other biotic index studies, and that our method adds new knowledge on the following points: (1) our explicit inclusion of Water 2021, 13, 886 17 of 19 17 of 19 ecological theory and knowledge makes it possible to disassemble biotic indices into their individual building blocks and (2) this approach allows a fuzzy classiﬁcation of aspects of a river system as potential stressors and their priority for management, which allows expert judgment in local cases. Thus, our method is beneﬁcial to currently used methods directly relating information to the biotic index. The method presented results in a solid classiﬁcation of individual aspects of a river system according to their priority for management. It can be supporting of the pressure- impact analyses of water bodies and in the design of a river restoration plan. The inclusion of ecological knowledge and expert insight improves the ecological soundness of the indi- vidual elements and the reliability of the method. It also results in good conﬁdence in the prioritization of stressors, despite a moderate predictive performance of the underlying models. We presented the application of our model for Flanders and only for macroinverte- brates, but the principles of the method can be implemented in any region and with other underlying models. Supplementary Materials: The following are available online at https://www.mdpi.com/2073-4 441/13/7/886/s1. 5. Conclusions Table SA1: Number of samples per river type and per ecological quality class, Table SA2: Percentage of samples per quality class for the ecological status based on macroinverte- brates, the physical-chemical status and the hydromorphological status, Table SA3: Overview of the input variables of the suitability models, Figure SB1: Macroinvertebrate taxa in Flanders ranked according to their 95th percentile abundance value in the monitoring data, Table SB1: Overview of the abundance classes, Table SC1: Composition of ecological groups, Table SC2: Taxa reference lists with base (B) and critical (C) taxa, Figure SD: Spatial visualization of all the priority classes of all variables considered in the case study of Flanders. Author Contributions: Conceptualization, E.B., S.G., G.E. and P.G.; methodology, E.B., S.G., G.E. and P.G.; software, S.G. and E.B.; validation, E.B., S.G. and P.G.; formal analysis, E.B. and S.G.; writing—original draft preparation, E.B.; writing—review and editing, P.G., S.G., G.E. and E.B.; visualization, E.B. and S.G.; supervision, P.G.; project administration, P.G.; funding acquisition, P.G. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Flanders Environment Agency, study, N◦VMM/SGB/002/2012. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Restrictions apply to the availability of these data. Data was obtained from the Flanders Environment Agency and are available at https://www.integraalwaterbeleid.be/ nl/geoloket/geoloket-stroomgebiedbeheerplannen with the permission of the Flanders Environment Agency. Acknowledgments: The authors would like to acknowledge Rob Laethem and Tom D’heygere for the fruitful discussions during the research project. Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. 5. Elosegi, A.; Gessner, M.O.; Young, R.G. River Doctors: Learning from Medicine to Improve Ecosystem Management. Sci. Total Environ. 2017, 595, 294–302. [CrossRef] [PubMed] 4. Bernhardt, E.S.; Palmer, M.A. River Restoration: The Fuzzy Logic of Repairing Reaches to Reverse Catchment Scale Degradation. Ecol. Appl. 2011, 21, 1926–1931. [CrossRef] 3. Jansson, R.; Nilsson, C.; Malmqvist, B. Restoring Freshwater Ecosystems in Riverine Landscapes: The R Recovery Processes. Freshw. Biol. 2007, 52, 589–596. [CrossRef] [ ] 3. Jansson, R.; Nilsson, C.; Malmqvist, B. Restoring Freshwater Ecosystems in Riverine Landscapes: The Roles of Connectivity and Recovery Processes. Freshw. Biol. 2007, 52, 589–596. [CrossRef] y y [ ] 2. Allan, J.D. Landscapes and Riverscapes: The Inﬂuence of Land Use on Stream Ecosystems. Annu. Rev. Ecol. Evol. Syst. 2004, 35, 257–284. [CrossRef] References 1. Vorosmarty, C.J.; McIntyre, P.B.; Gessner, M.O.; Dudgeon, D.; Prusevich, A.; Green, P.; Glidden, S.; Bunn, S.E.; Sullivan, C.A.; Liermann, C.R.; et al. Global Threats to Human Water Security and River Biodiversity. Nature 2010, 467, 555–561. [CrossRef] 2. Allan, J.D. Landscapes and Riverscapes: The Inﬂuence of Land Use on Stream Ecosystems. Annu. Rev. Ecol. Evol. Syst. 2004, 35, 257–284. [CrossRef] 1. Vorosmarty, C.J.; McIntyre, P.B.; Gessner, M.O.; Dudgeon, D.; Prusevich, A.; Green, P.; Glidden, S.; Bunn, S.E.; Sullivan, C.A.; Liermann, C.R.; et al. Global Threats to Human Water Security and River Biodiversity. Nature 2010, 467, 555–561. [CrossRef] y y 2. Allan, J.D. Landscapes and Riverscapes: The Inﬂuence of Land Use on Stream Ecosystems. Annu. Rev. Ecol. Evol. Syst. 2004, 35, 257–284. [CrossRef] [ ] 3. Jansson, R.; Nilsson, C.; Malmqvist, B. Restoring Freshwater Ecosystems in Riverine Landscapes: The Roles of Connectivity and Recovery Processes. Freshw. Biol. 2007, 52, 589–596. [CrossRef] 4. Bernhardt, E.S.; Palmer, M.A. River Restoration: The Fuzzy Logic of Repairing Reaches to Reverse Catchment Scale Degradation. Ecol. Appl. 2011, 21, 1926–1931. [CrossRef] 5. Elosegi, A.; Gessner, M.O.; Young, R.G. River Doctors: Learning from Medicine to Improve Ecosystem Management. Sci. Total Environ. 2017, 595, 294–302. [CrossRef] [PubMed] 18 of 19 18 of 19 Water 2021, 13, 886 6. Voulvoulis, N.; Arpon, K.D.; Giakoumis, T. The EU Water Framework Directive: From Great Expectations to Problems with Implementation. Sci. Total Environ. 2017, 575, 358–366. [CrossRef] p 7. Sundermann, A.; Leps, M.; Leisner, S.; Haase, P. Taxon-Speciﬁc Physico-Chemical Change Points for Stream Benthic Invertebrates. Ecol. Indic. 2015, 57, 314–323. [CrossRef] 8. Bouleau, G.; Pont, D. Did You Say Reference Conditions? Ecological and Socio-Economic Perspectives on the European Water Framework Directive. Environ. Sci. Policy 2015, 47, 32–41. [CrossRef] Bouleau, G.; Pont, D. Did You Say Reference Conditions? Ecological and Socio-Economic Perspectives on Framework Directive. Environ. Sci. Policy 2015, 47, 32–41. [CrossRef] Bouleau, G.; Pont, D. Did You Say Reference Conditions? Ecological and Socio-Economic Perspectives on Framework Directive. Environ. Sci. Policy 2015, 47, 32–41. [CrossRef] y 9. Dahm, V.; Hering, D.; Nemitz, D.; Graf, W.; Schmidt-Kloiber, A.; Leitner, P.; Melcher, A.; Feld, C.K. Effects of Physico-Chemistry, Land Use and Hydromorphology on Three Riverine Organism Groups: A Comparative Analysis with Monitoring Data from Germany and Austria. Hydrobiologia 2013, 704, 389–415. [CrossRef] y y g 10. References Friberg, N.; Bonada, N.; Bradley, D.C.; Dunbar, M.J.; Edwards, F.K.; Grey, J.; Hayes, R.B.; Hildrew, A.G.; Lamouroux, N.; Trimmer, M.; et al. Biomonitoring of Human Impacts in Freshwater Ecosystems. In Advances in Ecological Research; Elsevier: Amsterdam, The Netherlands, 2011; Volume 44, pp. 1–68. ISBN 978-0-12-374794-5. pp 11. Birk, S.; Bonne, W.; Borja, A.; Brucet, S.; Courrat, A.; Poikane, S.; Solimini, A.; van de Bund, W.; Zampoukas, N.; Hering, D. Three Hundred Ways to Assess Europe’s Surface Waters: An Almost Complete Overview of Biological Methods to Implement the Water Framework Directive. Ecol. Indic. 2012, 18, 31–41. [CrossRef] 12. Brown, E.D.; Williams, B.K. Ecological Integrity Assessment as a Metric of Biodiversity: Are We Mea Biodivers. Conserv. 2016, 25, 1011–1035. [CrossRef] ms, B.K. Ecological Integrity Assessment as a Metric of Biodiversity: Are We Measuring What We Say We Are? 016, 25, 1011–1035. [CrossRef] 13. Carré, C.; Meybeck, M.; Esculier, F. The Water Framework Directive’s “Percentage of Surface Water Bodies at Good Status”: Unveiling the Hidden Side of a “Hyperindicator”. Ecol. Indic. 2017, 78, 371–380. [CrossRef] g yp 14. Verberk, W.; van Noordwijk, C.; Hildrew, A. Delivering on a Promise: Integrating Species Traits to Transform Descriptive Community Ecology into a Predictive Science. Freshw. Sci. 2013, 32, 531–547. [CrossRef] 14. Verberk, W.; van Noordwijk, C.; Hildrew, A. Delivering on a Promise: Integrating Species Traits to Transform Descriptive C i E l i P di i S i F h S i 1 32 531 547 [C R f] , ; j , ; , g g g p Community Ecology into a Predictive Science. Freshw. Sci. 2013, 32, 531–547. [CrossRef] y gy 15. Schuwirth, N.; Reichert, P. Bridging the Gap between Theoretical Ecology and Real Ecosystems: Modeling Invertebrate Commu- nity Composition in Streams. Ecology 2013, 94, 368–379. [CrossRef] [PubMed] 16. Clews, E.; Ormerod, S.J. Improving Bio-Diagnostic Monitoring Using Simple Combinations of Standard Biotic Indices. River Res. Appl. 2009, 25, 348–361. [CrossRef] pp 17. Holguin Gonzalez, J.E.; Everaert, G.; Benedetti, L.; Amerlinck, Y.; Goethals, P.L.M. Use of Habitat Suitability Modeling in the Integrated Urban Water System Modeling of the Drava River (Varazdin, Croatia). In Proceedings of the 9th International Symposium on Ecohydraulics (ISE 2012), Vienna, Austria, 17–21 September 2012. y p y p 18. Klauer, B.; Rode, M.; Schiller, J.; Franko, U.; Mewes, M. Decision Support for the Selection of Measures According to the Requirements of the EU Water Framework Directive. Water Resour. Manag. References 2012, 26, 775–798. [CrossRef] 19. Villeneuve, B.; Souchon, Y.; Usseglio-Polatera, P.; Ferréol, M.; Valette, L. Can We Predict Biological Condition of Stream Ecosystems? A Multi-Stressors Approach Linking Three Biological Indices to Physico-Chemistry, Hydromorphology and Land Use. Ecol. Indic. 2015, 48, 88–98. [CrossRef] 20. Bizzi, S.; Surridge, B.W.J.; Lerner, D.N. Structural equation modelling: A novel statistical framework for exploring the spatial distribution of benthic macroinvertebrates in riverine ecosystems: Benthic macroinvertebrate distributions through SEM. River Res. Appl. 2013, 29, 743–759. [CrossRef] 21. Kuemmerlen, M.; Stoll, S.; Sundermann, A.; Haase, P. Long-Term Monitoring Data Meet Freshwater Species Distribution Models: Lessons from an LTER-Site. Ecol. Indic. 2016, 65, 122–132. [CrossRef] 22. Bennetsen, E.; Gobeyn, S.; Goethals, P.L.M. Species Distribution Models Grounded in Ecological Theory for Decision Support in River Management. Ecol. Model. 2016, 325, 1–12. [CrossRef] 23. Jähnig, S.C.; Kuemmerlen, M.; Kiesel, J.; Domisch, S.; Cai, Q.; Schmalz, B.; Fohrer, N. Modelling of Riverine Ecosystems by Integrating Models: Conceptual Approach, a Case Study and Research Agenda. J. Biogeogr. 2012, 39, 2253–2263. [CrossRef] 24. D’Amen, M.; Mateo, R.G.; Pottier, J.; Thuiller, W.; Maiorano, L.; Pellissier, L.; Ndiribe, C.; Salamin, N Spatial Predictions of Taxonomic, Functional and Phylogenetic Diversity. J. Ecol. 2018, 106, 76–86. [Cro 25. Guisan, A.; Tingley, R.; Baumgartner, J.B.; Naujokaitis-Lewis, I.; Sutcliffe, P.R.; Tulloch, A.I.T.; Regan, T.J.; Brotons, L.; McDonald- Madden, E.; Mantyka-Pringle, C.; et al. Predicting Species Distributions for Conservation Decisions. Ecol. Lett. 2013, 16, 1424–1435. [CrossRef] [PubMed] 26. Jungwirth, M.; Muhar, S.; Schmutz, S. Re-Establishing and Assessing Ecological Integrity in Riverine Landscapes. Freshw. Biol. 2002, 47, 867–887. [CrossRef] 27. Gabriels, W.; Lock, K.; De Pauw, N.; Goethals, P.L.M. Multimetric Macroinvertebrate Index Flan Assessment of Rivers and Lakes in Flanders (Belgium). Limnol. Ecol. Manag. Inland Waters 2010, 40 ck, K.; De Pauw, N.; Goethals, P.L.M. Multimetric Macroinvertebrate Index Flanders (MMIF) for Biolog Assessment of Rivers and Lakes in Flanders (Belgium). Limnol. Ecol. Manag. Inland Waters 2010, 40, 199–207. [CrossRef] 28. Mueller, M.; Pander, J.; Geist, J. Taxonomic Sufﬁciency in Freshwater Ecosystems: Effects of Taxonomic Resolution, Functional Traits, and Data Transformation. Freshw. Sci. 2013, 32, 762–778. [CrossRef] g g 28. Mueller, M.; Pander, J.; Geist, J. Taxonomic Sufﬁciency in Freshwater Ecosystems: Effects of Taxonomic Resolution, Functional Traits, and Data Transformation. Freshw. Sci. 2013, 32, 762–778. [CrossRef] , , , [ ] 29. Péru, N.; Dolédec, S. From Compositional to Functional Biodiversity Metrics in Bioassessment: A Case Study Using Stream Macroinvertebrate Communities. Ecol. Indic. 31. Usseglio-Polatera, P.; Bournaud, M.; Richoux, P.; Tachet, H. Biological and Ecological Traits of Benthic Freshwater Macroinverte- brates: Relationships and Deﬁnition of Groups with Similar Traits. Freshw. Biol. 2000, 43, 175–205. [CrossRef] References Taxonomic Relatedness Does Not Matter for Species Surrogacy in the Assessment of Community Responses to Environmental Drivers. J. Appl. Ecol. 2012, 49, 357–366. [CrossRef] 42. Rodrigues, A.S.; Brooks, T.M. Shortcuts for Biodiversity Conservation Planning: The Effectiveness of Surrogates. Annu. Rev. Ecol. Evol. Syst. 2007, 38, 713–737. [CrossRef] 43. Leps, M.; Leisner, S.; Haase, P.; Sundermann, A. Prediction of Taxon Occurrence: A Test on Taxon-Speciﬁc Change Point Values of Stream Benthic Invertebrates. Freshw. Biol. 2016, 61, 1773–1786. [CrossRef] Everaert, G.; Pauwels, I.; Bennetsen, E.; Goethals, P.L.M. Development and Selection of Decision Trees fo Impact of Data Preprocessing Algorithms and Settings AI Commun 2016 29 711 723 [CrossRef] ert, G.; Pauwels, I.; Bennetsen, E.; Goethals, P.L.M. Development and Selection of Decision Trees for Water 44. Everaert, G.; Pauwels, I.; Bennetsen, E.; Goethals, P.L.M. Development and Selection of Decision T Impact of Data Preprocessing, Algorithms and Settings. AI Commun. 2016, 29, 711–723. [CrossRef] Impact of Data Preprocessing, Algorithms and Settings. AI Commun. 2016, 29, 711–723. [CrossRef] 45. Flanders Environment Agency. Meetstrategie En Methodiek Hydromorfologie (Measurement Strategy and Methodology for Hydromor- phology); Vlaamse Milieumaatschappij: Aalst, Belgium, 2015. p gy pp j g 46. Verdonschot, R.C.M.; Kail, J.; McKie, B.G.; Verdonschot, P.F.M. The Role of Benthic Microhabitats in Determining the Effects of Hydromorphological River Restoration on Macroinvertebrates. Hydrobiologia 2016, 769, 55–66. [CrossRef] 47. Kail, J.; Brabec, K.; Poppe, M.; Januschke, K. The Effect of River Restoration on Fish, Macroinvertebrates and Aquatic Macrophytes: A Meta-Analysis. Ecol. Indic. 2015, 58, 311–321. [CrossRef] y 48. Verdonschot, P.F. Impact of Hydromorphology and Spatial Scale on Macroinvertebrate Assemblage Composition in Streams. Integr. Environ. Assess. Manag. 2009, 5, 97–109. [CrossRef] 49. Everaert, G.; Pauwels, I.S.; Goethals, P.L.M. Development of Data-Driven Models for the Assessment of Macroinvertebrates in Rivers in Flanders. In Proceedings of the 5th International Congress on Environmental Modelling and Software, Ottawa, ON, Canada, 5–8 July 2010. 50. Junier, S.; Mostert, E. A Decision Support System for the Implementation of the Water Framework Directive in the Netherlands: Process, Validity and Useful Information. Environ. Sci. Policy 2014, 40, 49–56. [CrossRef] 51. de Zwart, D.; Posthuma, L.; Gevrey, M.; von der Ohe, P.C.; de Deckere, E. Diagnosis of Ecosystem Impairment in a Multiple-Stress Context—How to Formulate Effective River Basin Management Plans. Integr. Environ. Assess. Manag. 2009, 5, 38–49. [CrossRef] [PubMed] [ ] 52. References 2010, 10, 1025–1036. [CrossRef] 30. Verberk, W.C.; Nijboer, R.C. Milieu-En Habitatpreferenties van Nederlandse Zoetwatermacrofauna; STOWA: Engelsbrand, Germany, 2012. 31. Usseglio-Polatera, P.; Bournaud, M.; Richoux, P.; Tachet, H. Biological and Ecological Traits of Benthic Freshwater Macroinverte- brates: Relationships and Deﬁnition of Groups with Similar Traits. Freshw. Biol. 2000, 43, 175–205. [CrossRef] 19 of 19 19 of 19 Water 2021, 13, 886 32. Pavoine, S.; Vallet, J.; Dufour, A.-B.; Gachet, S.; Daniel, H. On the Challenge of Treating Various Types of Variables: Application for Improving the Measurement of Functional Diversity. Oikos 2009, 118, 391–402. [CrossRef] p g y 33. Stoddard, J.L.; Larsen, P.; Hawkins, C.P.; Johnson, R.K.; Norris, R.H. Setting Expectations for the Ecological Condition of Running Waters: The Concept of Reference Condition. Ecol. Appl. 2006, 16, 1267–1276. [CrossRef] ard, J.L.; Larsen, P.; Hawkins, C.P.; Johnson, R.K.; Norris, R.H. Setting Expectations for the Ecological Condit s: The Concept of Reference Condition. Ecol. Appl. 2006, 16, 1267–1276. [CrossRef] 34. De Pauw, N.; Vannevel, R.; Beyens, J.; Dobbelaere, F.; Gardeniers, J.-P.; Goddeeris, B.; Mercken, L.; Neven, B.; Stroot, P.; Tolkamp, H.; et al. Macro-Invertebraten En Waterkwaliteit: Determineersleutels Voor Zoetwatermacro-Invertebraten En Methoden Ter Bepaling van de Waterkwaliteit; Stichting Leefmilieu: Antwerpen, Belgium, 1991. g p g 35. Nijboer, R.C.; Verdonschot, P.F.M.; Van Der Werf, D.C. The Use of Indicator Taxa as Representatives of Communities in Bioassessment. Freshw. Biol. 2005, 50, 1427–1440. [CrossRef] 36. Gobeyn, S.; Volk, M.; Dominguez-Granda, L.; Goethals, P.L.M. Input Variable Selection with a Simple Genetic Algorithm for Conceptual Species Distribution Models: A Case Study of River Pollution in Ecuador. Environ. Model. Softw. 2017, 92, 269–316. [CrossRef] 37. Langhans, S.D.; Hermoso, V.; Linke, S.; Bunn, S.E.; Possingham, H.P. Cost-Effective River Rehabilitation Planning: Optimizing for Morphological Beneﬁts at Large Spatial Scales. J. Environ. Manag. 2014, 132, 296–303. [CrossRef] [PubMed] 38. Kuhn, M.; Johnson, K. Applied Predictive Modelling; Springer: New York, NY, USA; Heidelberg, Germany; Dordrecht, The Netherlands; London, UK, 2013. 39. Flanders Environment Agency Geoloket Stroomgebiedbeheerplannen. Available online: https://www.integraalwaterbeleid.be/ nl/geoloket (accessed on 24 January 2021). 40. Psaltopoulos, D.; Wade, A.J.; Skuras, D.; Kernan, M.; Tyllianakis, E.; Erlandsson, M. False Positive and False Negative Errors in the Design and Implementation of Agri-Environmental Policies: A Case Study on Water Quality and Agricultural Nutrients. Sci. Total Environ. 2017, 575, 1087–1099. [CrossRef] [ ] 41. Bevilacqua, S.; Terlizzi, A.; Claudet, J.; Fraschetti, S.; Boero, F. References Sutherland, W.J.; Gardner, T.; Bogich, T.L.; Bradbury, R.B.; Clothier, B.; Jonsson, M.; Kapos, V.; Lane, S.N.; Möller, I.; Schroeder, M.; et al. Solution Scanning as a Key Policy Tool: Identifying Management Interventions to Help Maintain and Enhance Regulating Ecosystem Services. Ecol. Soc. 2014, 19, 3. [CrossRef] 53. von der Ohe, P.C.; de Zwart, D.; Semenzin, E.; Apitz, S.E.; Gottardo, S.; Harris, B.; Hein, M.; Marcomini, A.; Posthuma, L.; Schäfer, R.B.; et al. Monitoring Programmes, Multiple Stress Analysis and Decision Support for River Basin Management. In Risk-Informed Management of European River Basins; Brils, J., Brack, W., Müller-Grabherr, D., Négrel, P., Vermaat, J.E., Eds.; Springer: Berlin/Heidelberg, Germany, 2014; Volume 29, pp. 151–182. ISBN 978-3-642-38597-1. 54. de Vries, J.; Kraak, M.H.S.; Verdonschot, P.F.M. A Conceptual Model for Simulating Responses of Freshwater Macroinvertebrate Assemblages to Multiple Stressors. Ecol. Indic. 2020, 117, 106604. [CrossRef] g p 55. Tonkin, J.D.; Stoll, S.; Sundermann, A.; Haase, P. Dispersal Distance and the Pool of Taxa, but Not Colonisation of Restored River Reaches by Benthic Invertebrates. Freshw. Biol. 2014, 59, 1843–1855. [Cr 56. Dahm, V.; Hering, D. A Modeling Approach for Identifying Recolonisation Source Sites in River Restoration Planning. Landsc. Ecol. 2016, 31, 2323–2342. [CrossRef]
https://openalex.org/W2513817050	https://europepmc.org/articles/pmc5012232?pdf=render	English	null	Assessment of the impact of quality improvement interventions on the quality of sick child care provided by Health Extension Workers in Ethiopia	Journal of global health	2,016	cc-by	8,074	journal of health global journal of health global Electronic supplementary material: Electronic supplementary material: Electronic supplementary material: The online version of this article contains supplementary material. Electronic supplementary material: The online version of this article contains supplementary material. 1 Institute for International Programs, Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA 2 United Nations Children’s Fund (UNICEF) Ethiopia Country Office, Addis Ababa, Ethiopia 3 ABH Services, PLC, Addis Ababa, Ethiopia Correspondence to: Nathan P Miller Institute for International Programs Johns Hopkins Bloomberg School of Public Health 615 N Wolfe St. Baltimore, MD 21205 USA nmiller@jhu.edu Assessment of the impact of quality improvement interventions on the quality of sick child care provided by Health Extension Workers in Ethiopia VIEWPOI PAPERS Nathan P Miller1, Agbessi Amouzou1, Elizabeth Hazel1, Hailemariam Legesse2, Tedbabe Degefie2, Mengistu Tafesse3, Robert E Black1, Jennifer Bryce1 Nathan P Miller1, Agbessi Amouzou1, Elizabeth Hazel1, Hailemariam Legesse2, Tedbabe Degefie2, Mengistu Tafesse3, Robert E Black1, Jennifer Bryce1 Background Ethiopia has scaled up integrated community case man- agement of childhood illness (iCCM), including several interventions to improve the performance of Health Extension Workers (HEWs). We assessed associations between interventions to improve iCCM quality of care and the observed quality of care among HEWs. Methods We assessed iCCM implementation strength and quality of care provided by HEWs in Ethiopia. Multivariate logistic regression analyses were performed to assess associations between interventions to improve iCCM quality of care and correct management of iCCM illnesses. Findings Children who were managed by an HEW who had attend- ed a performance review and clinical mentoring meeting (PRCMM) had 8.3 (95% confidence interval (CI) 2.34–29.51) times the odds of being correctly managed, compared to children managed by an HEW who did not attend a PRCMM. Management by an HEW who received follow–up training also significantly increased the odds of correct management (odds ratio (OR) = 2.09, 95% CI 1.05–4.18). Supervision on iCCM (OR = 0.63, 95% CI 0.23–1.72) did not sig- nificantly affect the odds of receiving correct care. Conclusions These results suggest PRCMM and follow–up training were effective interventions, while implementation of supportive su- pervision needs to be reviewed to improve impact. Approximately 6.9 million children younger than five years of age died in 2011, with the vast majority of these deaths occurring in low– and middle–income countries [1]. Pneumonia, diarrhea, and malaria are among the leading causes of under–five mortality, with malnutrition as an important underlying cause [2]. Most under–five deaths can be pre- vented with available and cost–effective interventions [3]. Antibiotic treat- ment for pneumonia [4–6], oral rehydration salts and zinc for diarrhea [6–8], and artemisinin–based combination therapy for malaria [9,10] are effective interventions for preventing death from these diseases among children under five. However, poor and inequitable access to primary health care [11–13] and low quality of care [14–17] are important barri- Approximately 6.9 million children younger than five years of age died in 2011, with the vast majority of these deaths occurring in low– and middle–income countries [1]. Pneumonia, diarrhea, and malaria are among the leading causes of under–five mortality, with malnutrition as an important underlying cause [2]. Assessment of the impact of quality improvement interventions on the quality of sick child care provided by Health Extension Workers in Ethiopia Most under–five deaths can be pre- vented with available and cost–effective interventions [3]. Antibiotic treat- ment for pneumonia [4–6], oral rehydration salts and zinc for diarrhea [6–8], and artemisinin–based combination therapy for malaria [9,10] are effective interventions for preventing death from these diseases among children under five. However, poor and inequitable access to primary health care [11–13] and low quality of care [14–17] are important barri- Correspondence to: Nathan P Miller Institute for International Programs Johns Hopkins Bloomberg School of Public Health 615 N Wolfe St. Baltimore, MD 21205 USA nmiller@jhu.edu December 2016 • Vol. 6 No. 2 • 020404 1 www.jogh.org • doi: 10.7189/jogh.06.020404 Miller et al. ers. Integrated community case management of childhood illness (iCCM, see Box 1 for definitions of key terms as ap- plied in this study) is increasingly promoted as a key strat- egy to increase access to appropriate treatments for child- hood illnesses, to reduce child mortality, and to improve equity in access to health services [18–20]. Through iCCM, community health workers (CHWs) are trained and equipped to manage common childhood illnesses (usually pneumonia, diarrhea, and malaria) in the community. through training on iCCM, supportive supervision, provi- sion of essential commodities, and enhanced monitoring and evaluation. Following the initial six–day clinical train- ing on iCCM, a number of interventions were implement- ed within the iCCM program to ensure continued quality of care. These interventions were: follow–up training, per- formance review and clinical mentoring meetings (PRC- MM), and supportive supervision. A description of the iCCM quality improvement interventions is provided in Table 1. Since 2011, Ethiopia has been implementing iCCM, includ- ing treatment of uncomplicated pneumonia, diarrhea, ma- laria, malnutrition, and measles through Health Extension Workers (HEWs) in most regions of the country. All HEWs are literate women with at least a tenth–grade education, who receive a one–year pre–service training. The pre–ser- vice training included management of childhood diarrhea, malaria, and malnutrition; pneumonia was not part of the package prior to the introduction of iCCM. Following the training, they are recruited as government employees and deployed to work out of health posts at the kebele (sub–dis- trict) level. There are typically two HEWs working at one health post, which serves approximately 5000 people, the average population of a kebele. Through the iCCM pro- gram, HEWs treat children 2–59 months of age; sick chil- dren under two months are referred to a health facility [21]. Assessment of the impact of quality improvement interventions on the quality of sick child care provided by Health Extension Workers in Ethiopia Ensuring quality of services provided through an iCCM pro- gram is essential to achieving mortality reductions. Assess- ments of IMCI and iCCM suggest that quality of care is vari- able and depends on the strength of program implementation, which is defined here as the level of deliv- ery of key components of a program, such as training, su- pervision, and provision of commodities [22–29]. Program processes such as training and supportive supervision are presumed to lead to improved health worker performance, but, in practice, this is not always the case [30–35]. Most of the evidence available on the impact of program activities on quality of care pertains to health facility–based workers and is limited to formal training and supervision. A wide range of approaches to training and supervision activities have been implemented for iCCM [36], but little is known about the extent to which iCCM quality improvement ac- tivities are associated with the quality of care provided by CHWs. As part of an independent evaluation of the iCCM program in the Oromia Region of Ethiopia, we conducted a survey of health posts to assess iCCM program implemen- tation strength and the quality of iCCM services provided by HEWs. These data provide an opportunity to assess as- sociations between iCCM quality improvement interven- tions and quality of care provided by HEWs at health posts. The iCCM program was designed to strengthen the capac- ity of HEWs to assess, classify, and treat childhood pneu- monia, diarrhea, malaria, malnutrition, and measles Box 1. Definitions of key terms Box 1. Definitions of key terms Correct management: This indicator of quality of care was used as the outcome variable in this analysis. Correct man- agement is defined for this study as: The proportion of sick children for whom all HEW treatments matched gold stan- dard treatments, including correct dose, duration, and fre- quency, and HEW referral matched the gold standard clas- sification for referral for all major iCCM illnesses (pneumonia, diarrhea, malaria, malnutrition, and measles). METHODS Performance review and clinical mentoring meet- ings (PRCMM) Two–day meeting held every six months (originally planned to be quarterly) at the woreda (district) level, during which approximately 20 HEWs met with supervisors, woreda health officials, health center staff. Zonal and regional focal persons sometimes also participated. Woreda iCCM facilita- tors and staff from implementing partner NGOs facilitated the meetings. On the first day, facilita- tors abstracted data from iCCM patient registers, reviewed registers with HEWs, and discussed is- sues related to quality of care and utilization of services. The second day was dedicated to clinical practice for HEWs in a health facility with feedback from facilitators. PRCMM guidelines are pre- sented in Appendix S1 in Online Supplementary Document. Supportive supervision Standardized supportive supervision on iCCM in the health post was performed on a quarterly basis. Supervisors were usually implementing partner NGO staff, and sometimes health center staff or woreda health officials. Supervisors used a standardized supervision checklist (Appendix S2 in Online Supplementary Document), which included sections on case management, preventive services, supply of commodities, data collection, knowledge assessment, feedback, and next steps. Supervisors were instructed to either observe the HEWs conducting sick child consultations or review iCCM patient registers for completeness and consistency between recorded signs/symptoms, classification, and treatment. HEW – health extension worker, iCCM – integrated community case management of childhood illness, PRCMM – performance review and clinical men- toring meeting Refresher training in the health post within eight weeks of the iCCM training. This was a half–day to one–day visit by an iCCM trainer from the district or from an implementing partner agency. The purpose of the visit was to reinforce knowledge and skills learned during the initial iCCM training. They also carried out observation of sick child consultations, if possible, as well as reviewed sick child registers and administered case scenarios to assess HEW performance. Supervisors identified the HEW’s skills gaps and then focused on improving these during the visit. Two–day meeting held every six months (originally planned to be quarterly) at the woreda (district) level, during which approximately 20 HEWs met with supervisors, woreda health officials, health center staff. Zonal and regional focal persons sometimes also participated. Woreda iCCM facilita- tors and staff from implementing partner NGOs facilitated the meetings. METHODS Assuming an average of 1.5 HEWs and two children observed per health post, the sample sizes were expected to give estimates of health post– and HEW–level indicators with precision of +/– 10% per- centage points and +/– 9 points for patient–level indicators. sification. Data collectors also conducted caretaker exit interviews, inspected commodities and patient registers, and interviewed HEWs. Data were entered directly into tablet computers using Open Data Kit (ODK) [39] as the data capture software and were stored in a Research Elec- tronic Data Capture (REDCap) database [40]. Further de- tails on the survey methods are available elsewhere [41]. Within health posts, all HEWs providing clinical services were included in the study. Children had to meet the fol- lowing criteria to be included: 1) 2–59 months of age, 2) having at least one complaint consistent with an eligible iCCM illness, and 3) this was the initial consultation for the current illness episode. Eligible children were included in the study if they presented for care at the health post ei- ther spontaneously or because they were mobilized to seek care by the HEWs. If fewer than two children presented at the health post within one hour of the health post opening, the team supervisor, along with an HEW or community volunteer, recruited sick children from nearby households. Ethical approval was obtained from the Institutional Re- view Boards of the Oromia Regional Health Bureau and the Johns Hopkins University Bloomberg School of Public Health. Informed consent forms were read in local lan- guage to all participating HEW and caregivers of sick chil- dren and oral consent was obtained from all participants. METHODS Data on iCCM implementation strength and quality of care were obtained through a cross–sectional survey of rural health posts in Jimma and West Hararghe Zones of Oromia Region, Ethiopia in which HEWs work. Oromia is the larg- est region in Ethiopia, with a population of approximately 30 million [37]. The survey was conducted in May and June 2012, about one year after initiation of iCCM imple- mentation and during the low malaria season. The survey instruments were adapted from a survey of Health Surveil- lance Assistants in Malawi [26] and the World Health Or- ganization (WHO) Health Facility Survey Tool [38]. Implementation strength: This term refers to the quantity of effective program activities carried out to reduce child mortality. These activities include training, supportive super- vision, and continued availability of essential iCCM com- modities and supplies. Integrated community case management (iCCM): In gen- eral, iCCM refers to the concurrent management of child- hood pneumonia, diarrhea, and malaria in the community. iCCM in Ethiopia is integrated management by an HEW at the health post or community level of the following child- hood illnesses: pneumonia, diarrhea, malaria, malnutrition, and measles. We randomly selected 104 health posts from the 490 health posts that were implementing iCCM. To determine the sample size we assumed that proportions of key indicators were 50%, confidence level was 95%, non–response was Quality of care: For this study, quality of care refers to cor- rect assessment, classification, treatment, and referral of pa- tients according to iCCM clinical guidelines. www.jogh.org • doi: 10.7189/jogh.06.020404 December 2016 • Vol. 6 No. 2 • 020404 www.jogh.org • doi: 10.7189/jogh.06.020404 2 Interventions to improve quality of care Table 1. Description of Ethiopia iCCM quality improvement activities Activity Description Follow–up training Refresher training in the health post within eight weeks of the iCCM training. This was a half–day to one–day visit by an iCCM trainer from the district or from an implementing partner agency. The purpose of the visit was to reinforce knowledge and skills learned during the initial iCCM training. They also carried out observation of sick child consultations, if possible, as well as reviewed sick child registers and administered case scenarios to assess HEW performance. Supervisors identified the HEW’s skills gaps and then focused on improving these during the visit. www.jogh.org • doi: 10.7189/jogh.06.020404 METHODS On the first day, facilita- tors abstracted data from iCCM patient registers, reviewed registers with HEWs, and discussed is- sues related to quality of care and utilization of services. The second day was dedicated to clinical practice for HEWs in a health facility with feedback from facilitators. PRCMM guidelines are pre- sented in Appendix S1 in Online Supplementary Document. Standardized supportive supervision on iCCM in the health post was performed on a quarterly basis. Supervisors were usually implementing partner NGO staff, and sometimes health center staff or woreda health officials. Supervisors used a standardized supervision checklist (Appendix S2 in Online Supplementary Document), which included sections on case management, preventive services, supply of commodities, data collection, knowledge assessment, feedback, and next steps. Supervisors were instructed to either observe the HEWs conducting sick child consultations or review iCCM patient registers for completeness and consistency between recorded signs/symptoms, classification, and treatment. 5% for health posts, 5% for HEWs, and 10% for patients, design effect to account for clustering of HEWs and pa- tients within health posts was 1.3. Assuming an average of 1.5 HEWs and two children observed per health post, the sample sizes were expected to give estimates of health post– and HEW–level indicators with precision of +/– 10% per- centage points and +/– 9 points for patient–level indicators. Within health posts, all HEWs providing clinical services were included in the study. Children had to meet the fol- lowing criteria to be included: 1) 2–59 months of age, 2) having at least one complaint consistent with an eligible iCCM illness, and 3) this was the initial consultation for the current illness episode. Eligible children were included in the study if they presented for care at the health post ei- ther spontaneously or because they were mobilized to seek care by the HEWs. If fewer than two children presented at the health post within one hour of the health post opening, the team supervisor, along with an HEW or community volunteer, recruited sick children from nearby households. 5% for health posts, 5% for HEWs, and 10% for patients, design effect to account for clustering of HEWs and pa- tients within health posts was 1.3. Indicators The primary indicators of implementation strength and quality of care were adapted from the WHO Health Facil- ity Survey Tool [38] and the iCCM Global Indicators [42]. Table 2 shows the list of indicators considered for use in this analysis. The outcome of interest was the proportion of children correctly managed for major iCCM illnesses (correct management), with a sick child as the unit of anal- ysis. This is a summary indicator that includes manage- ment of pneumonia, diarrhea, malaria, malnutrition, and measles. A child was considered correctly managed if 1) the child received all recommended treatments and no un- necessary treatments; 2) the child was referred if and only if referral was indicated; and 3) the HEW prescribed the correct dose, duration, and frequency for all treatments. Data collectors were health professionals who had worked as iCCM trainers or supervisors. Survey personnel were trained for seven days, and all observers and re–examiners achieved at least 90% concordance with gold standard cli- nicians on three consecutive role play examinations. During the survey, data collectors observed HEWs’ consul- tations with sick children and recorded details of the HEW’s assessment, classification, treatment, referral, and counsel- ing. A separate data collector then carried out a re–exami- nation of the same child to obtain the gold standard clas- December 2016 • Vol. 6 No. 2 • 020404 www.jogh.org • doi: 10.7189/jogh.06.020404 3 Miller et al. Table 2. Selected indicators of iCCM quality of care and implementation strength and their proportions (or means (standard devia- tion), Jimma and West Hararghe Zones, Oromia Region, Ethiopia, 2012* Variable No. % or mean 95% CI or SD Outcome: Child correctly managed for iCCM illnesses 257 64.2 57.4, 70.5 Predictors: HEW received follow–up training within 8 weeks of iCCM training 134 46.3 36.7, 56.1 HEW attended performance review and clinical mentoring meeting 137 89.1 82.1, 93.5 Health post received at least one supervision on iCCM in the previous three months 100† 87.0 78.8, 92.9 CI – confidence interval, SD – standard deviation, HEW – health extension worker *Indicators of quality of care and implementation strength were previously published by Miller et al [41]. †Three health posts excluded because HEWs reported not being present for majority of previous three months. VIEWPOI Papers the HEWs, or recruited by the survey team), and the other two predictors of interest. Indicators The predictors of interest were the training– and supervi- sion–related interventions presented in Table 1 that were intended to improve health worker performance (1. HEW received follow–up training within eight weeks of iCCM training, 2. HEW attended PRCMM, 3. Health post re- ceived at least one supervision on iCCM in the previous three months). We used purposeful selection of variables to assess poten- tial confounders [43]. In the first step, we conducted bi- variate logistic regressions of the outcome on potential con- founding variables. Potential confounders with a P–value 0.25 from a Wald test were retained for further assess- ment. Starting with the variable with the weakest associa- tion, these variables were assessed as confounders for each of the models. We then compared a full model, with the predictor and all of the variables retained in step one above, to a null model with one of the potential confounders re- moved. If removing the variable changed the coefficient of the predictor of interest by 15%, the variable was includ- ed in the final model as a confounder. This was repeated for each of the potential confounders. Child characteristics, such as age or severity of illness, can be important predictors of quality of care. However, be- cause the objective of this analysis was to assess the rela- tionships between the interventions to improve HEW per- formance and correct management, rather than to assess all possible predictors of correct management, we did not include child characteristics as predictors of interest in the analysis. The role of child characteristics as potential con- founders is discussed below. Finally, we performed multivariate logistic regression anal- yses using two–level random intercept models with child and health post levels to assess associations between each of the quality improvement interventions and the outcome, controlling for the retained confounders. An alpha level of 0.05 was used for tests of statistical significance. All analy- ses were carried out in Stata 13 [44]. Model selection and data analysis Analyses for this manuscript were limited to HEWs who had received the standardized clinical training on iCCM. We assessed whether each of the three predictors of interest was associated with the outcome of correct management of childhood illness. Separate multivariate logistic regression models were developed for each predictor. For each of the three models (one for each predictor), we assessed potential confounders of the relationship between each predictor and the outcome. The potential confounders assessed were: zone, distance from the health post to the nearest referral health facility, availability of iCCM commodities, malaria risk of the health post catchment area, the number of sick child consultations at the health post in the previous month (caseload), HEW age, years of experience as an HEW, whether the HEW was from the kebele in which she worked, whether the HEW currently lived in the kebele in which she worked, whether the HEW intended to continue working as an HEW through the coming year (as a proxy for motivation), child sex, child age, whether the child had a severe illness, recruitment method that brought the child to the health post (spontaneous consultation, mobilized by December 2016 • Vol. 6 No. 2 • 020404 RESULTS All but one of the selected 104 health posts implementing iCCM were successfully surveyed, giving a final sample of 103 health posts. All 137 HEWs encountered in health posts and 257 sick children were included in the survey. Sample characteristics and detailed results on iCCM imple- mentation strength and quality of care are available else- where [41]. Bivariate analyses showed that PRCMM (OR = 6.45, 95% CI 2.22–18.73) and follow–up training (OR = 2.25, 95% CI 1.24–4.07) were significantly associated with the outcome. Supervision on iCCM (OR = 1.15, 95% CI 0.44–2.99) was not significantly associated with correct management. December 2016 • Vol. 6 No. 2 • 020404 www.jogh.org • doi: 10.7189/jogh.06.020404 4 Interventions to improve quality of care The final models for each predictor and the results of the multivariate logistic regressions are shown in Table 3. At- tendance at PRCMM had the largest association with correct management, controlling for other covariates. Children who were managed by an HEW who had attended a PRCMM had about eight times the odds of being correctly managed (OR = 8.3, 95% CI 2.34–29.51), compared to children man- aged by an HEW who had not attended a PRCMM. The HEW receiving follow–up training also significantly in- creased the odds of the child receiving correct management (OR = 2.09, 95% CI 1.05–4.18). The HEW receiving super- vision on iCCM (OR = 0.63, 95% CI 0.23–1.72) was not sig- nificantly associated with a child receiving correct care. CHWs. The evidence from this small number of studies is also mixed. Some found positive associations between quality of care and training [55,56] and supervision [56], while others showed no significant associations with train- ing [57,58] or supervision [58]. Two recent studies in Ethi- opia showed associations between supportive supervision [59] and PRCMM [60] and improved consistency in iCCM patient registers between HEWs’ assessment, classification, and treatment of sick children. We found that HEW attendance at a PRCMM and follow– up training were both significantly associated with the out- come. PRCMM, in particular, had a very strong association with correct management. These are important findings because these interventions are not always included in quality improvement programs. DISCUSSION These findings should be interpreted with caution, as there is the potential for unmeasured confounding. It is possible that the HEWs who provided lower quality care were less motivated and did not attend a PRCMM. This would spu- riously inflate the observed association between PRCMM and the outcome. During the survey, a small number of HEWs (16) reported that they did not plan to continue working as an HEW through the coming year. Because this variable may be related to HEW motivation, we assessed whether there may be a negative association between at- tending a PRCMM and intention to continue as an HEW. In fact, all of the HEWs who reported an intention to leave their jobs as HEWs had attended a PRCMM. This is far from an ideal proxy for motivation and a very small sample size, but it does somewhat weaken the argument that HEW motivation confounded the relationship between correct management and PRCMM. Previous studies that looked at associations between pro- gram inputs and health worker quality of care in low–in- come countries have typically focused on training and su- pervision of health workers in health facilities. These studies of health facility workers have had mixed results [30,45], with some showing significant or near–significant associations between indicators of quality of care and train- ing [46–52] and others showing little or no association [32–34]. Similarly, supervision has been shown to improve health worker performance in some studies [53] and to have no or little effect in others [33,34,54,55]. Reviews of interventions to improve health worker performance found that quality improvement interventions with multifaceted interventions, such as training plus supervision, generally have a greater effect and that supervision with feedback is usually beneficial [30,45]. Routine supervision for iCCM was not associated with the outcome. PRCMM and follow–up training may have been Only a few studies have assessed the effect of quality im- provement interventions on quality of care provided by Table 3. Final models and results of multivariate analyses of associations between iCCM quality improvement interventions and correct management of major iCCM illnesses, controlling for selected covariates, in children 2–59 months, Jimma and West Hararghe Zones, Oromia Region, Ethiopia, 2012 Predictor variable Covariates Number. of children (N = 257) % children correctly treated OR (95% CI) P–value Child managed by HEW who attended PRCMM: – Yes 233 68.2 8.3 (2.34, 29.51) <0.001 No 24 25.0 Ref. DISCUSSION Child managed by HEW who received follow–up training within 8 weeks of iCCM training: Child with severe illness Yes 116 74.1 2.09 (1.05, 4.18) 0.037 No 141 56.0 Ref. Child managed in health post that received at least one supervision on iCCM in the previous three months PRCMM; follow–up training; child with severe illness Yes 225 65.3 0.63 (0.23, 1.72) 0.369 No 29 62.1 Ref. OR – odds ratio, CI – confidence interval, HEW – health extension worker, iCCM – integrated community case management of childhood illness. PRC- MM – performance review and clinical mentoring meeting d / h 5 Table 3. Final models and results of multivariate analyses of associations between iCCM quality improvement interventions and correct management of major iCCM illnesses, controlling for selected covariates, in children 2–59 months, Jimma and West Hararghe Zones, Oromia Region, Ethiopia, 2012 December 2016 • Vol. 6 No. 2 • 020404 5 5 Miller et al. Miller et al. This study has several limitations. First, the observational design of the study limits the conclusions about causality that can be drawn. The absence of a comparison group makes it impossible to rule out unmeasured variables that may have confounded the observed associations between the interventions and correct management. Second, health workers may have performed better under observation than they would under normal circumstances [62,63]. Third, data on the predictors were based on HEW recall, which may be biased toward having received the interven- tions, although Hazel et al. found that HSAs in Malawi ac- curately reported implementation strength indicators [64]. Finally, because HEWs receive relatively more training and compensation than many CHW cadres in other countries, these results may not be generalizable to other contexts where iCCM is implemented. more effective than routine iCCM supervision because they were longer and more intensive, and were more specifi- cally focused on reviewing case management guidelines and clinical practices. On the other hand, the observed lack of association between supervision and correct manage- ment may reflect a lack of focused improvement of clinical skills, with more emphasis on data collection and register review. The effect of supervision may also have been re- duced in a context where a large majority of HEWs had received PRCMM. Supervision may have had a limited ad- ditional effect after the strong effect of PRCMM. DISCUSSION Funding: This work was supported by The United Nations Children’s Fund (award 1000017212) and the Government of Canada (award 7056791). NPM received funding from the United States National Institutes of Health Training Grant for Maternal and Child Health (HD046405). Authorship declarations: All authors were involved in the conception and design of the study or data collection. NPM, AA, EH, HL, TD, REB, and JB conceived of the study and designed the ex- periments. NPM, AA, EH, and MT carried out data collection. NPM conducted data analysis. NPM, AA, EH, HL, TD, MT, REB, JB participated in drafting and reviewing the manuscript. Competing interests: All authors have completed the Unified Competing Interest form at www. icmje.org/coi_disclosure.pdf (available on request from the corresponding author). HL and TD were staff of UNICEF, which was an iCCM implementing partner, at the time of the study. Acknowledgments: The authors thank Ato Shallo Dhaba and Dr Zelalem Habtamu of the Oromia Regional Health Bureau and Dr Luwei Pearson of the United Nations Children’s Fund (UNICEF) Ethiopia Country Office for their strong support of this research. Thanks to ABH Services, PLC for implementation of the survey. We also thank the Ethiopian Federal Ministry of Health, the JSI Re- search and Training Institute, Inc./Last 10 Kilometers Project (JSI/L10K), the Integrated Family Health Program (JSI/IFHP), and UNICEF New York for their support and assistance. Funding: This work was supported by The United Nations Children’s Fund (award 1000017212) and the Government of Canada (award 7056791). NPM received funding from the United States National Institutes of Health Training Grant for Maternal and Child Health (HD046405). Funding: This work was supported by The United Nations Children’s Fund (award 1000017212) and the Government of Canada (award 7056791). NPM received funding from the United States National Institutes of Health Training Grant for Maternal and Child Health (HD046405). Authorship declarations: All authors were involved in the conception and design of the study or data collection. NPM, AA, EH, HL, TD, REB, and JB conceived of the study and designed the ex- periments. NPM, AA, EH, and MT carried out data collection. NPM conducted data analysis. NPM, AA, EH, HL, TD, MT, REB, JB participated in drafting and reviewing the manuscript. Competing interests: All authors have completed the Unified Competing Interest form at www. icmje.org/coi_disclosure.pdf (available on request from the corresponding author). DISCUSSION VIEWPOIN Papers The discrepancy between these results and the previous study showing a significant association between supervision and quality of care among HEWs [59] is likely because of the use of different outcome variables. Ameha et al. used consistency between classification and treatment from pa- tient registers as the outcome variable. We have previously shown that register review overestimates quality of care com- pared to estimates derived from observation of consultations and re–examination of sick children [61]. Because supervi- sions focus on register review, it is likely that they improve recording of cases, but this may neglect improvement of as- sessment skills, which can lead to incorrect classification. The results of this study, as well as large variations in impact from quality improvement interventions studied in other contexts, suggest that it may be the quality of implementa- tion of interventions, more than their inherent characteris- tics, which determines their effectiveness. Quality improve- ment interventions for health workers can lead to improvements in quality of care, but they do not necessarily do so. Therefore, efforts should be focused not only on achieving high coverage of quality improvement interven- tions, but also on the content and quality of these interven- tions to ensure that they have the expected impact. The pos- itive associations between follow–up training and especially PRCMM suggest that these interventions should be consid- ered as complements to standard trainings and supportive supervision to improve the quality of iCCM services. Further research is needed on how to improve the quality and effectiveness of training and supervision interventions to achieve maximum impact. For example, supportive su- pervision may be more effective if it includes a component of case observation with feedback, rather than relying en- tirely on register review. However, the feasibility of doing this in a setting of small numbers of consultations needs to be assessed. Acknowledgments: The authors thank Ato Shallo Dhaba and Dr Zelalem Habtamu of the Oromia Regional Health Bureau and Dr Luwei Pearson of the United Nations Children’s Fund (UNICEF) Ethiopia Country Office for their strong support of this research. Thanks to ABH Services, PLC for implementation of the survey. We also thank the Ethiopian Federal Ministry of Health, the JSI Re- search and Training Institute, Inc./Last 10 Kilometers Project (JSI/L10K), the Integrated Family Health Program (JSI/IFHP), and UNICEF New York for their support and assistance. DISCUSSION Zinc for the treatment of diarrhoea: effect on diarrhoea morbidity, mortality and incidence of future episodes. Int J Epidemiol. 2010;39 Suppl 1:i63-9. Medline:20348128 doi:10.1093/ije/dyq023 q 8 Walker CL, Black RE. Zinc for the treatment of diarrhoea: effect on diarrhoea morbidity, mortality and incidence f f d I J E d l 2010 39 S l 1 63 9 M dl 20348128 d 10 1093/ /d 023 9 Ogbonna A, Uneke CJ. Artemisinin–based combination therapy for uncomplicated malaria in sub–Saharan Af- rica: the efficacy, safety, resistance and policy implementation since Abuja 2000. Trans R Soc Trop Med Hyg. 2008;102:621-7. Medline:18499204 doi:10.1016/j.trstmh.2008.03.024 j 10 Nosten F, White NJ. Artemisinin–based combination treatment of falciparum malaria. Am J Trop Med Hyg. 2007;77(6 Suppl):181-92. Medline:18165491 11 Victora CG, Wagstaff A, Schellenberg JA, Gwatkin D, Claeson M, Habicht JP. Applying an equity lens to child health and mortality: more of the same is not enough. Lancet. 2003;362:233-41. Medline:12885488 doi:10.1016/ S0140-6736(03)13917-7 ( ) 12 Gwatkin DR, Bhuiya A, Victora CG. Making health systems more equitable. Lancet. 2004;364:1273-80. Med- line:15464189 doi:10.1016/S0140-6736(04)17145-6 13 Schellenberg JA, Victora CG, Mushi A, de Savigny D, Schellenberg D, Mshinda H, et al. Inequities among the very poor: health care for children in rural southern Tanzania. Lancet. 2003;361:561-6. Medline:12598141 doi:10.1016/S0140-6736(03)12515-9 14 Waiswa P, Kallander K, Peterson S, Tomson G, Pariyo GW. Using the three delays model to understand why newborn babies die in eastern Uganda. Trop Med Int Health. 2010;15:964-72. Medline:20636527 doi:10.1111/ j.1365-3156.2010.02557.x j 15 Terra de Souza AC, Peterson KE, Andrade FM, Gardner J, Ascherio A. Circumstances of post–neonatal deaths in Ceara, Northeast Brazil: mothers' health care–seeking behaviors during their infants' fatal illness. Soc Sci Med. 2000;51:1675-93. Medline:11072887 doi:10.1016/S0277-9536(00)00100-3 16 Källander K, Hildenwall H, Waiswa P, Galiwango E, Peterson S, Pariyo G. Delayed care seeking for fatal pneu- monia in children aged under five years in Uganda: a case–series study. Bull World Health Organ. 2008;86:332- 8. Medline:18545734 doi:10.2471/BLT.07.049353 17 Sodemann M, Jakobsen MS, Molbak K, Alvarenga IC Jr, Aaby P. High mortality despite good care–seeking be- haviour: a community study of childhood deaths in Guinea–Bissau. Bull World Health Organ. 1997;75:205-12. Medline:9277007 18 Haines A, Sanders D, Lehmann U, Rowe AK, Lawn JE, Jan S, et al. Achieving child survival goals: potential con- tribution of community health workers. Lancet. 2007;369:2121-31. Medline:17586307 doi:10.1016/S0140- 6736(07)60325-0 19 WHO/UNICEF. Joint Statement: Integrated Community Case Management (iCCM). DISCUSSION HL and TD were staff of UNICEF, which was an iCCM implementing partner, at the time of the study. December 2016 • Vol. 6 No. 2 • 020404 6 www.jogh.org • doi: 10.7189/jogh.06.020404 Interventions to improve quality of care 1 UNICEF. UN–Inter–agency Group for Child Mortality Estimation. Levels and Trends of Child Mortality. The 2012 Report. New York: UNICEF, 2012. 2 Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, et al. Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet. 2012;379:2151-61. Medline:22579125 doi:10.1016/S0140-6736(12)60560-1 3 Jones G, Steketee RW, Black RE, Bhutta ZA, Morris SS. Bellagio, Child Survival Study Group. How many child deaths can we prevent this year? Lancet. 2003;362:65-71. Medline:12853204 doi:10.1016/S0140- 6736(03)13811-1 4 Sazawal S, Black RE. Meta–analysis of intervention trials on case–management of pneumonia in community set- tings. Lancet. 1992;340:528-33. Medline:1354286 doi:10.1016/0140-6736(92)91720-S 4 Sazawal S, Black RE. Meta–analysis of intervention trials on case–management of pneumonia in community set- tings. Lancet. 1992;340:528-33. Medline:1354286 doi:10.1016/0140-6736(92)91720-S 5 Sazawal S, Black RE; Pneumonia Case Management Trials Group. Effect of pneumonia case management on mortality in neonates, infants, and preschool children: a meta–analysis of community–based trials. Lancet Infect Dis. 2003;3:547-56. Medline:12954560 doi:10.1016/S1473-3099(03)00737-0 5 Sazawal S, Black RE; Pneumonia Case Management Trials Group. Effect of pneumonia case management on mortality in neonates, infants, and preschool children: a meta–analysis of community–based trials. Lancet Infect Dis. 2003;3:547-56. Medline:12954560 doi:10.1016/S1473-3099(03)00737-0 5 Sazawal S, Black RE; Pneumonia Case Management Trials Group. Effect of pneumonia case management on mortality in neonates, infants, and preschool children: a meta–analysis of community–based trials. Lancet Infect Dis. 2003;3:547-56. Medline:12954560 doi:10.1016/S1473-3099(03)00737-0 6 Bhutta ZA, Das JK, Walker N, Rizvi A,Campbell H, Rudan I ,et al. Interventions to address deaths from child- hood pneumonia and diarrhoea equitably: what works and at what cost? Lancet. 2013;381:1417-29. Med- line:23582723 doi:10.1016/S0140-6736(13)60648-0 7 Munos MK, Walker CL, Black RE. The effect of oral rehydration solution and recommended home fluids on di- arrhoea mortality. Int J Epidemiol. 2010;39 Suppl 1:i75-87. Medline:20348131 doi:10.1093/ije/dyq025 7 Munos MK, Walker CL, Black RE. The effect of oral rehydration solution and recommended home fluids on di- h li I J E id i l 2010 39 S l 1 i75 87 M dli 20348131 d i 10 1093/ij /d 025 y p pp j yq 8 Walker CL, Black RE. DISCUSSION Geneva and New York: World Health Organization and United Nations Children's Fund, 2012. 20 Young M, Wolfheim C, Marsh DR, Hammamy D. World Health Organization/United Nations Children's Fund Joint Statement on Integrated Community Case Management: An Equity–Focused Strategy to Improve Access to Essential Treatment Services for Children. Am J Trop Med Hyg. 2012;87(5 Suppl):6-10. Medline:23136272 doi:10.4269/ajtmh.2012.12-0221 21 Legesse H, Degefie T, Hiluf M, Sime K, Tesfaye C, Abebe H. National scale–up of integrated community case management in rural Ethiopia: implementation and early lessons learned. Ethiop Med J. 2014;52 Suppl 3:15- 26. Medline:25845070 22 Bryce J, Victora CG, Habicht JP, Black RE, Scherpbier RW. Programmatic pathways to child survival: results of a multi–country evaluation of Integrated Management of Childhood Illness. Health Policy Plan. 2005;20 Suppl 1:i5-17. Medline:16306070 doi:10.1093/heapol/czi055 23 Curtale F, Siwakoti B, Lagrosa C, LaRaja M, Guerra R. Improving skills and utilization of community health vol- unteers in Nepal. Soc Sci Med. 1995;40:1117-25. Medline:7597465 doi:10.1016/0277-9536(94)00172-P 23 Curtale F, Siwakoti B, Lagrosa C, LaRaja M, Guerra R. Improving skills and utilization of community health vol- unteers in Nepal. Soc Sci Med. 1995;40:1117-25. Medline:7597465 doi:10.1016/0277-9536(94)00172-P 7 December 2016 • Vol. 6 No. 2 • 020404 www.jogh.org • doi: 10.7189/jogh.06.020404 Miller et al. 24 Bhutta Z, Lassi ZS, Pariyo G, Huicho L. Global Experience of Community Health Workers for Delivery of Health Related Millenium Development Goals: A Systematic Review, Country Case Studies, and Recommendations for Integration into National Health Systems. Geneva: Global Health Workforce Alliance and World Health Orga- nization, 2010. 25 Management OP. Lady Health Worker Programme: External Evaluation of the National Programme for Family Planning and Primary Health Care, Summary of Results. Oxford: Oxford Policy Management, 2009. 26 Gilroy KE, Callaghan–Koru JA, Cardemil CV, Nsona H, Amouzou A, Mtimuni A, et al. Quality of sick child care delivered by Health Surveillance Assistants in Malawi. Health Policy Plan. 2013;28:573-85. Medline:23065598 doi:10.1093/heapol/czs095 p 27 Rogers S, Paija S, Embiap J, Pust RE. Management of common potentially serious paediatric illnesses by aid post orderlies at Tari, Southern Highlands Province. P N G Med J. 1991;34:122-8. Medline:1750253 VIEWPOI Papers 28 Beracochea E, Dickson R, Freeman P, Thomason J. Case management quality assessment in rural areas of Papua New Guinea. Trop Doct. 1995;25:69-74. Medline:7778198 29 Kelly JM, Osamba B, Garg RM, Hamel MJ, Lewis JJ, Rowe SY, et al. Community health worker performance in the management of multiple childhood illnesses: Siaya District, Kenya, 1997–2001. DISCUSSION Am J Public Health. 2001;91:1617-24. Medline:11574324 doi:10.2105/AJPH.91.10.1617 30 Rowe AK, de Savigny D, Lanata CF, Victora CG. How can we achieve and maintain high–quality performance of health workers in low–resource settings? Lancet. 2005;366:1026-35. Medline:16168785 doi:10.1016/S0140- 6736(05)67028-6 31 Bryce J, el Arifeen S, Pariyo G, Lanata C, Gwatkin D, Habicht JP. Reducing child mortality: can public h deliver? Lancet. 2003;362:159-64. Medline:12867119 doi:10.1016/S0140-6736(03)13870-6 32 Ofori–Adjei D, Arhinful DK. Effect of training on the clinical management of malaria by medical assistants in Ghana. Soc Sci Med. 1996;42:1169-76. Medline:8737435 doi:10.1016/0277-9536(95)00389-4 33 Rowe AK, Hamel MJ, Flanders WD, Doutizanga R, Ndoyo J, Deming MS. Predictors of correct treatment of children with fever seen at outpatient health facilities in the Central African Republic. Am J Epidemiol. 2000;151:1029-35. 34 Rowe AK, Onikpo F, Lama M, Cokou F, Deming MS. Management of childhood illness at health facilities in Be- nin: problems and their causes. Am J Public Health. 2001;91:1625-35. Medline:11574325 doi:10.2105/ AJPH.91.10.1625 35 Rowe AK, Onikpo F, Lama M, Deming MS. Risk and protective factors for two types of error in the treatment of children with fever at outpatient health facilities in Benin. Int J Epidemiol. 2003;32:296-303. Medline:12714553 doi:10.1093/ije/dyg063 36 Bosch–Capblanch X, Marceau C. Training, supervision and quality of care in selected integrated community case management (iCCM) programmes: a scoping review of programmatic evidence. J Glob Health. 2014;4:020403. Medline:25520793 37 Office of the Population and Housing Census Commission. Summary and Statistical Report of the 2007 Popu- lation and Housing Census: Population Size by Age and Sex. Addis Ababa, Ethiopia: Federal Democratic Re- public of Ethiopia, Population Census Commission, 2011. 38 WHO. Health Facility Survey: Tool to evaluate the quality of care delivered to sick children attending outpatient facilities. Geneva: Department of Child and Adolescent Health and Development, World Health Organization, 2003. 39 Hartung C, Anokwa Y, Brunette W, Lerer A, Tsent C, Borriello G. Open Data Kit: Tools to Build Information Ser- vices for Developing Regions. 4th ACM/IEEE Intl Conf on Information and Communication Technologies and Development (ICTD); 2010. 40 Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)––a metadata–driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377-81. Medline:18929686 doi:10.1016/j.jbi.2008.08.010 41 Miller NP, Amouzou A, Tafesse M, Hazel E, Legesse H, Degefie T, et al. Integrated community case management of childhood illness in Ethiopia: implementation strength and quality of care. DISCUSSION Medline:19954537 doi:10.1186/1475-2875-8-275 53 Rowe AK, Onikpo F, Lama M, Osterholt DM, Rowe SY, Deming MS. A multifaceted intervention to improve health worker adherence to integrated management of childhood illness guidelines in Benin. Am J Public Health. 2009;99:837-46. Medline:19299681 doi:10.2105/AJPH.2008.134411 54 Osterholt DM, Rowe AK, Hamel MJ, Flanders WD, Mkandala C, Marum LH, et al. Predictors of treatment error for children with uncomplicated malaria seen as outpatients in Blantyre district, Malawi. Trop Med Int Health. 2006;11:1147-56. Medline:16903878 doi:10.1111/j.1365-3156.2006.01666.x 55 Zeitz PS, Harrison LH, Lopez M, Cornale G. Community health worker competency in managing acute res tory infections of childhood in Bolivia. Bull Pan Am Health Organ. 1993;27:109-19. Medline:8339109 56 Hadi A. Management of acute respiratory infections by community health volunteers: experience of Ba Rural Advancement Committee (BRAC). Bull World Health Organ. 2003;81:183-9. Medline:1276451 57 Rowe SY, Kelly JM, Olewe MA, Kleinbaum DG, McGowan JE Jr, McFarland DA, et al. Effect of multiple inter- ventions on community health workers' adherence to clinical guidelines in Siaya district, Kenya. Trans R Soc Trop Med Hyg. 2007;101:188-202. Medline:17064747 doi:10.1016/j.trstmh.2006.02.023 58 Ashwell HE, Freeman P. The clinical competency of community health workers ince of Papua New Guinea. P N G Med J. 1995;38:198-207. Medline:9522859 59 Ameha A, Karim AM, Erbo A, Ashenafi A, Hailu M, Hailu B, et al. Effectiveness of supportive supervision on the consistency of integrated community cases management skills of the health extension workers in 113 districts of Ethiopia. Ethiop Med J. 2014;52 Suppl 3:65-71. Medline:25845075 59 Ameha A, Karim AM, Erbo A, Ashenafi A, Hailu M, Hailu B, et al. Effectiveness of supportive supervision on the consistency of integrated community cases management skills of the health extension workers in 113 districts of Ethiopia. Ethiop Med J. 2014;52 Suppl 3:65-71. Medline:25845075 60 Mengistu B, Karim AM, Eniyew A, Yitabrek A, Eniyew A, Tsegaye S, et al. Effect of performance review and clin- ical mentoring meetings (PRCMM) on recording of community case management by health extension workers in Ethiopia. Ethiop Med J. 2014;52 Suppl 3:73-81. Medline:25845076 60 Mengistu B, Karim AM, Eniyew A, Yitabrek A, Eniyew A, Tsegaye S, et al. Effect of performance review and clin- ical mentoring meetings (PRCMM) on recording of community case management by health extension workers in Ethiopia. Ethiop Med J. 2014;52 Suppl 3:73-81. Medline:25845076 60 Mengistu B, Karim AM, Eniyew A, Yitabrek A, Eniyew A, Tsegaye S, et al. DISCUSSION Am J Trop Med Hyg. 2014;91:424- 34. Medline:24799369 doi:10.4269/ajtmh.13-0751 34. Medline:24799369 doi:10.4269/ajtmh.13-0751 42 ICCM Task Force CCM Central: Integrated Community Case Management of Childhood Illness. Available: http:// ccmcentral.com/benchmarks-and-indicators/indicators/. Accessed: 27 April 2016. Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Sour Code Biol Med. 2008;3:17. 44 StataCorp. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP, 2013. 45 Ross–Degnan D, Laing R, Santoso B, Ofori–Adjei D, Lamoureux C, Hogerzell H. Improving Pharmaceutical Use in Primary Care in Developing Countries: A Critical Review of Experience and Lack of Experience: Internation- al Network for Rational Use of Drugs. ICIUM 1997: Proceedings of the International Conference on Improving Use of Medicines; 1997 1-4 Apr; Chiang Mai, Thailand. 46 Zurovac D, Rowe AK, Ochola SA, Noor AM, Midia B, English M, et al. Predictors of the quality of health work- er treatment practices for uncomplicated malaria at government health facilities in Kenya. Int J Epidemiol. 2004;33:1080-91. Medline:15256523 doi:10.1093/ije/dyh253 December 2016 • Vol. 6 No. 2 • 020404 8 www.jogh.org • doi: 10.7189/jogh.06.020404 Interventions to improve quality of care 47 Armstrong Schellenberg J, Bryce J, de Savigny D, Lambrechts T, Mbuya C, Mgalula L, et al. The effect of Inte- grated Management of Childhood Illness on observed quality of care of under–fives in rural Tanzania. Health Policy Plan. 2004;19:1-10. Medline:14679280 doi:10.1093/heapol/czh001 y p 48 Amaral J, Gouws E, Bryce J, Leite AJ, Cunha AL, Victora CG. Effect of Integrated Management of Childhood Ill- ness (IMCI) on health worker performance in Northeast–Brazil. Cad Saude Publica. 2004;20 Suppl 2:S209-19. Medline:15608935 doi:10.1590/S0102-311X2004000800016 49 Naimoli JF, Rowe AK, Lyaghfouri A, Larbi R, Lamrani LA. Effect of the Integrated Management of Childhood Ill- ness strategy on health care quality in Morocco. Int J Qual Health Care. 2006;18:134-44. Medline:16423842 doi:10.1093/intqhc/mzi097 q 50 Pariyo GW, Gouws E, Bryce J, Burnham G. Improving facility–based care for sick children in Uganda: training is not enough. Health Policy Plan. 2005;20 Suppl 1:i58-68. Medline:16306071 doi:10.1093/heapol/czi051 51 Arifeen SE, Hoque DM, Akter T, Rahman M, Hoque ME, Begum K, et al. Effect of the Integrated Management of Childhood Illness strategy on childhood mortality and nutrition in a rural area in Bangladesh: a cluster ran- domised trial. Lancet. 2009;374:393-403. Medline:19647607 doi:10.1016/S0140-6736(09)60828-X 52 Rowe AK, de Leon GF, Mihigo J, Santelli AC, Miller NP, Van–Dunem P. Quality of malaria case management at outpatient health facilities in Angola. Malar J. 2009;8:275. www.jogh.org • doi: 10.7189/jogh.06.020404 DISCUSSION Effect of performance review and clin- ical mentoring meetings (PRCMM) on recording of community case management by health extension workers in Ethiopia. Ethiop Med J. 2014;52 Suppl 3:73-81. Medline:25845076 61 Miller NP, Amouzou A, Hazel E, Degefie T, Legesse H, Tafesse M, et al. Assessing the quality of sick child care provided by Community Health Workers. PLoS ONE. 2015;10:e0142010. Medline:26551035 doi:10.1371/ journal.pone.0142010 62 Rowe AK, Lama M, Onikpo F, Deming MS. Health worker perceptions of how being observed influe practices during consultations with ill children. Trop Doct. 2002;32:166-7. Medline:12139161 63 Leonard K, Masatu MC. Outpatient process quality evaluation and the Ha 2006;63:2330-40. Medline:16887245 doi:10.1016/j.socscimed.2006.06.003 63 Leonard K, Masatu MC. Outpatient process quality evaluation and the Hawthorne Effect. Soc Sci Med. 2006;63:2330-40. Medline:16887245 doi:10.1016/j.socscimed.2006.06.003 64 Hazel E, Amouzou A, Park L, Banda B, Chimuna T, Guenther T, et al. Real–time assessments of the strength of program implementation for community case management of childhood illness: validation of a mobile phone– based method in Malawi. Am J Trop Med Hyg. 2015;92:660-5. Medline:25582691 doi:10.4269/ajtmh.14-0396 December 2016 • Vol. 6 No. 2 • 020404 www.jogh.org • doi: 10.7189/jogh.06.020404 9
https://openalex.org/W2914648888	https://jurnal.ugm.ac.id/jurnal-humaniora/article/download/40374/23726	English	null	English Blends in Indonesian Context: Their Formation Processes and Meanings	Humaniora	2,019	cc-by-sa	7,725	HUMANIORA HUMANIORA Vol. 31, No. 1 (February 2019) https://doi.org/10.22146/jh.v31i1.40374 page 21—32 https://jurnal.ugm.ac.id/jurnal-humaniora https://jurnal.ugm.ac.id/jurnal-humaniora English Blends in Indonesian Context: Their Formation Processes and Meanings Rio Rini Diah Moehkardi Department of Language and Literature, Universitas Gadjah Mada, Indonesia Email: riomoehkardi@ugm.ac.id Rio Rini Diah Moehkardi Department of Language and Literature, Universitas Gadjah Mada, Indonesia Email: riomoehkardi@ugm.ac.id ABSTRACT This research aims at observing the blending process, the patterns of blending, and the possible new meanings of the English-influenced blends found in the Indonesian context. The influence of English and Indonesian not only in terms of the source words involved, but also of the morphological processes of the blends and also to see whether there are some semantic changes, anticipating the cultural influence of the blend creators. Data are collected from advertisement of schools events, brand name of food and beverage products, also names of cafes and restaurants found mostly in Yogyakarta. The data are analyzed using Mattiello’s formula (2013) in categorizing the process of blending: morphotactical, morphonological, and morphosemantic. The results show that morphotactically the data can be classified into total and partial blends of which there are more blends whose first source words are kept full and followed by clipped second source words. In morphopohonological process, there is no non-overlapping blends; phonological overlap also takes place, but not orthographical one. Morphosemantically, attributive blends are more frequent than the coordinative ones. Keywords: blending; formation process; meaning; Indonesian context Keywords: blending; formation process; meaning; Indonesian context Vol. 31, No. 1 (February 2019) https://doi.org/10.22146/jh.v31i1.40374 page 21—32 Vol. 31, No. 1 (February 2019) https://doi.org/10.22146/jh.v31i1.40374 page 21—32 INTRODUCTION Kreidler’s definition of blending (1994 in Fandrych, 2008): “Sometimes two words are clipped simultaneously and united to form a ‘blend’. The two source words may be syntagmatically related or paradigmatically related.” Kubozono (1990, in Mattiello, 2013, p. 113) sees the relations of the source words, of blends that are in paradigmatic relations, such as chofa (chair + sofa) or syntagmatic relations, such as porta-light (portable + light). Other linguists (Cannon, 2000; Plag, 2003; and Bat El, 200 in Mattiello, 2013, p. 114) also see that the source words often exhibit some semantic similarity but “ rarely synonyms, ” such as brunch (breakfast + lunch); usually “belong to the same syntactic category” (Kubozono, 1990, in Mattiello, 2013, p. 114), such as adjective + adjective : ginormous (gigantic + enormous) or show some “phonological similarity” (Cannon, 2000, in Mattiello, 2013, p. 114), such as hesiflation (hesitation + inflation). Unlike, many Indonesian-word blends, English blends found in Yogyakarta, Indonesia, are mostly found in advertisement. The popular creation of blends in Indonesia, Yogyakarta in particular, seems to resonance Steinmetz and Kipfer’s sentiment in favor of acronyms, clipping and blending as more popular word formation processes due to their catchiness, economy, and humour compared to compounding and derivation (2006 in Fadrych, 2008) and echoing Algeo’s (1977, p. 61) that blends “ are coined not alone for their usefulness, but partly, and in some cases principally, for their cleverness.” It is the clever and funny aspect of blending that renders an attractive process, not only to linguists, but also to advertising executives, script writers…(in Danks, 2003, p. 2). Danks furthermore suggests that most blends “are ephemereal; they are coined for particular purposes and once the purpose has gone then the word is no longer needed.” (Danks, 2003, p. 3). Blends may live shortly, yet many new ones are born in many more creative ways. Whereas, Plag (2003) offers rather narrow definition of blend that “it is always the first part of the first element that is combined with the second part of the second element” that is similar to that of Bat El (2006, in Mattiello, p. INTRODUCTION Bahasa Indonesia (BI) is an official language introduced to preschoolers and still taught to university students, used at homes to most households in urban areas in Indonesia, yet as a relatively young and dynamic language, especially the vocabulary of Bahasa Indonesia is very much influenced by many other languages, both outside and inside the country. which is blending. Blends can be found almost anywhere in Indonesian context, both written and spoken, in the national and indigenous languages. Some blends are regarded informal, and some others are so well accepted that they are used in more formal situations. For example in Bahasa Indonesia, blends: pemilu (pemilihan + umum--general election) and pilkada (pemilihan + kepala +daerah--regional head election) are among well accepted blends and regarded as formal. People use the blends and not the full phrases in any situation. Bahasa Indonesia has many loanwords in its lexicon. To name only few here are some foreign words borrowed in Bahasa Indonesia: adakala (BI) meaning sometimes is from Hindi kȃda kalȃ; akad (BI) meaning promise is from Arabic aqd and many others from local languages: cengang (BI) meaning amazed is from Minang language cangang; anjlok (BI) meaning drop is from Javanese anjlok. Blending is one of word formation processes naturally occurs in many languages, such as English (workaholic from work+a(lco)holic); Bahasa Indonesia (Orba from orde + baru -- new order; pemprov from pemerintah + provinsi -- provincial government) even Javanese, one of Indonesian local indigenous languages also practices this word formation (nasgitel from panas As a dynamic language BI practices word- formation processes to enrich its vocabulary, among of Humaniora, Vol. 31, No. 1 (February 2019) + legi + kentel meaning hot + sweet + strong; indicating strong, sweet and hot tea or coffee). l blending involves at least 2 or more source words, non- morphematic formation, clipping of the source words, coordinative relation between source words, phonic integration, and overlap (2014: appendix, table 1). Her confirmation is based on previous studies in the area by various linguists. i English also gives some influence in BI, not only loanwords but also many blends found in Yogyakarta, Indonesia are of English influence. Although there are many English loan blends in Bahasa Indonesia, such as teenlit, vlog, workaholic, chocolicious, etc. there are many blends created by Indonesian speakers; many of them are names of school events, music performance, cafes or restaurants, food and beverages products. INTRODUCTION 113) “blends refer only to cases where the inner edges are truncated,” Gries (2004) observes more in the process of blending that “involves the coinage of a new lexeme by fusing parts of at least two other source words of which either one is shortened in the fusion and or where there is some form of phonemic or graphemic overlap of the source words.” Overlapping graphemes and phonemes in blend actually distinguish blend from acronym. The following is the distinction between blending and other similar word formations: acronym, clipping and compounding. This paper aims at analyzing the formation-process of English blends in Indonesian context and their meaning will be observed to see the influence of both English and BI in the blends, not only in terms of the source words involved, but also of the morphological process of the blends and also to see the relations between the source words involvement in the blend and the semantic change, anticipating the cultural influence of the blend creators. THEORETICAL FRAMEWORK Furthermore, Bauer notes that splinters refer to word parts that have been frequently used in forming blends, such as (-holic; -nomic), and, some other linguists regard them as ‘bound morpheme’ (Lehrer, 1998), combining form (Lehrer, 1988; B Warren, 1990), or bound splinter (Fandrych 2008a in Believa, 2014, p. 49). Believa (2014, p. 49-50) proposes that splinters are parts of words that make the blends, on condition that they are not full words and not morphemes or bound morphs accepted in dictionaries. Similar to compounding that can be written as a simple word (paperclip), with a space (paper aeroplane) or include a hyphen (oil-paper), although most blends are written as simple words, some blends are found with hyphen (rap-sploitation, porta-light). Furthermore, although rarely, they may be written with a space (docu drama). Both compound and blend, once formed, become the base to which affixes attach: blue-eyed, bookkeeper, sky-diving in compounding and sexploiting, chortled, electrocution in blending (Danks, 2003, pp. 48-49). Unlike compound, which is self-defining, some blends do not explain themselves when out of context, for example: cyxploitation and probot (Danks 2003, p. 50). Moreover, whereas compounding reproduces the whole component words, in blending at least one word is reproduced partially and can come from non-words too, for example aquarobics blends splinter and neo-classsical combining forms (aqua + aerobics). Soudek (1978 in Fandrych, 2008) develops the term into: “initial splinter” and “ final splinter,” referring to the first or second element of the blend for the former, and only the second element of the blend for the latter. Rua (2002 in Fandrych, 2008) defines “splinter” as graphic and phonemic sequence which are neither inflectional nor derivational morphemes, nor combining forms, and whose length generally allows their identification as belonging to a previous word. Splinters tend to be syllable or larger than syllables in their sources. When they are shorter than syllables, their constituents are a. the syllable onset (the prevocalic consonant or consonant); b. the onset and the nucleus (prevocalic consonants and vowel); c. the rhyme (vowel + postvocalic consonant or coda) (Rua, 2002 in Fandrych, 2008). Danks (2003, p. 19) confirms that the term “splinter” is used to describe “bits” from the source words truncated that appear in the blends. However, she reminds that blending may not only use splinter, but may also two words overlapped, such as Japanimation (Japan + animation). Ingo Plag classifies blends (2002, pp. THEORETICAL FRAMEWORK Acronym and blending are often confused as of the same processes (Stockwell and Minkova, 2001 in Fandrych, 2008; Permata and Suyudi, 2011), both processes in creating a new lexeme require at least two lexemes, both undergo cut in some part of the member lexemes. However, in acronym, the new lexeme consists of the first phoneme or grapheme of the source words, although sometimes in order the new lexeme can be pronounced as a word, not only does the initial sounds are taken but sometimes the next first consonant and first vowel. In blending, usually there are more graphemes or phonemes of the source words involved in forming a Although blends have been in use further back in Shakespeare’s time and have been studied extensively since the twentieth century (Algeo, 1977, p. 47), definitions of blending and blending categories have not yet set. The term “blending” has been used in a number of ways, usually to denote a word formation process which combines two source lexemes, at least one of which has been shortened in the combination, sometimes with a graphic and /or phonological overlap (Mattiello, 2013, p. 112). Beliaeva confirms in her research that 22 Moehkardi.- English Blends in Indonesian Context pronounceable blend (Fandrych, 2008). putting two words together to form a third” (Bauer, 1983 in Danks, 2003, p. 47). Lehrer (2007 in Mattiello, 2013, p. 113) states that “blends are underlying compounds which are composed of one word and part of another, or parts of two (occasionally three) other words.” Carstairs- McCarthy (2002, p. 65) says that blends “are a kind of compound where at least one component is reproduced only partially.” Furthermore, Kemmer (2003 in Mattiello, 2013, p. 114) points out that what distinguish blends from compounds is that “they (blends) combine parts of lexical source words, rather than the whole source words”. Whereas truncation in clipping is rather clear cut, that in acronyms and blends are often less than a syllable. Hozzeinzadeh (2014) regards the parts clipped or overlapped as bits (“… new blend created by using the first bit of the first word, the last bit of the second one”), however, some other linguists, such as Berman (1961), Adam (1973) in Fandrych (2008), and Bauer (2006 in Believa, 2014, p. 49) regard them as splinters. THEORETICAL FRAMEWORK 155-160) in two types: abbreviated compound and proper blend. The former is originally a compound whose meaning is mostly determined by the second element (the head), for example: breath analyzer—breathalyzer—is a kind of analyzer; science fiction—sci-fi, a kind of fiction; when they are shortened, they retain their meaning as compound. This type, in other classification, is referred to as syntagmatic origin (Bauer, 2012 in Believa 2014, p. 30), one of the source words, usually the second or the right one, is the head and the first is the modifier. However, the latter’s source words, for examples, boat + hotel; breakfast + lunch, when they are shortened – boatel and brunch, they denote the referents of both source words, resembling the copulative compound, or considered as paradigmatic origin (Bauer, 2012 in Believa 2014, p. 30): both words are heads. Mattiello (2013, pp. 123-125) refers abbreviated compound to attributive blend of which its second word source functions as head and the first as modifier; whereas, the proper blend, is called coordinate blend of which the two word sources are related syntactically and semantically. As the parts shortened in the source words in blending process often “include unpredictable splinters instead of existing morphemes” (Soudek, 1978 in Danks 2003, p. 11), blending usually is classified as the non- morphematic word-formation (Fandrych, 2008), and thus labelled as “extra grammatical morphology,” a term introduced by Dressler and Merlini Barbressi (1994 in Mattiello 2013, p. 1) referring to word formations “through which the process obtained are not identifiable and the input does not allow a prediction of regular input.” The fact that it takes at least two source words in blending reminds us of compounding: “the process of 23 Humaniora, Vol. 31, No. 1 (February 2019) resulted from blending (Algeo, 1977, p. 52), such as – cast that has the noun sense of broadcast, -athon from marathon. However, the term ‘source-form’ is used for non-words, such as the use of affix (Belieava, 2014, p. 4), the present research also uses this term for abbreviation, such as FIB and MUF. The term ‘splinter’ is used in this current discussion rather than ‘part’ (Quirk, 1985) or ‘bit’ (Hozzainzadech, 2014). THEORETICAL FRAMEWORK In presenting the data, the researcher uses italic for the blend; symbol + indicating the source words, the origin of the blend; ( ) indicating the part of the word that is clipped, shortened, or cut; bold indicating overlap, for example: biolympic – biol(ogy) + olympic and underline indicating embedded splinter in one of the source word: autopathography--auto(bio) graphy + patho(logy). Syntactically, the source words paradigmatically belong to the same syntactic category. Semantically, the source words are usually co-hyponyms of a more general term (capcin—cappuccino + cincau) or combination of synonyms (attractivating—attractive + captivating). Furthermore, Mattiello (2013, p. 124) also suggests that most coordinate blends are endocentric. i Not only are definitions various, earlier linguists (including Algeo, 1977; Soudek, 1978; Cannon, 2000; Kemmer, 2003; Ronneberger-Sibold, 2006, Lehrer, 1996 and 2007 in Mattiello, 2013, p. 118), Quirk (1985) and more recent ones such as, Fandrych, (2008), Mattiello (2013) and Hozzainzadech (2014) have proposed various classifications demonstrating many parameters in combining patterns of blending. Whereas Quirk (1985), Fandrych (2008) and Hozzainzadech (2014) focus on structural classification of blend formation, Algeo (1977) also includes systematic categories and Mottiello (2013) morphosemantic categories referring to the semantic relations of the source words. This research follows Mattiello’s formula in categorizing blending (2013). It distinguishes blends into three categories of which 2 are in the formation processes : a. morphotactical classifies blends into total blends where all source words are reduced in splinters, and partial blends where only one source word is reduced; and, b. morphonological and graphical systems distinguish blends into overlapping and non-overlapping; and another category on morphosemantic observing the semantic relation between the source words: attributive and coordinate blends Algeo (1977, pp. 49-55 and 55-61) classifies blending into structural categories, dealing with how blends are formed and systemic categories dealing with the relationship of the source words. The structural categories include: blends with overlapping, blends with clipping, clipping at morpheme (syllable) boundaries, blends with clipping and overlapping. The systemic categories include syntagmatic blends, associative blends: synonymic, paradigmatic, jumble, indefinite composites, telescope and portmanteau. Whereas, Algeo (1977) includes the significance of relationship of the source words, Quirk (1985), Fandrych (2008) and Hozzainzadech (2014) focus on structural categories of blending which are more or less similar. THEORETICAL FRAMEWORK There is not any data in the current research that fit 1 sub-categories of total blends in morphotactical classification: blends made from both splinters that are the end of the source words (Kongfrontation—(King) Kong + (con)frontation. Furthermore, there is not any data that fit 1 sub-category of overlapping in morphonological and phonological category: the constituents overlap orthographically but not phonologically, for example: smog—smo(ke) + (f)og the shared letter /o/ is pronounced /әʊ/ in smoke but /ɒ/in fog. DATA AND METHODOLOGY The data of this research are taken from advertisement of school events, name of food and beverage products, name of restaurants or cafes found mostly in Yogyakarta, but some are also found in magazine, stores (especially, the names of food and beverage products). The data (80 blends) were collected from April – November 2017. Including blends with similar morphemes and formation process are counted individually, for example there are 6 blends with –licious morpheme following the unshortened first words, as each in fact has different initial source word. The term ‘source words’ is used to refer to the lexical units in the blend, including the bound roots or combining forms (Carstairs-McCarthy, 2002, p. 66), such as anthropo- and –logy also new morphemes DISCUSSION Mottiello divides blend morphotactically into two types: a) total blends (all source words are reduced in splinter) and b) partial blends (only one source word is reduced). The following is how the current data are categorized. Morphotactical 1. Total blends Total blends are those in which all source words are reduced to splinters; there are 4 sub-categories of this. a. The beginning of one word is followed by the end of another: Oxbridge-- Ox(ford) + (Cam)bridge 24 Moehkardi.- English Blends in Indonesian Context blends Constituents and process Chocoberry Choco(late) + (straw)berry Robominton Robo(t) + (bad)minton Fooshion (market) Foo(d) + (fa)shion Fashound Fash(ion) + (s)ound Chitato Chi(ps) + (po)tato Lovamil* Lov(e) + (h)amil Jogjacation Jogja(karta) + (edu)cation Idenesia Ide(a) + (Indo)nesia Sociophoria* Socio(logy) + (eu)phoria Sociopreneur Socio + (entre)preneur Ebotec E(LINS) + (Ro)bot + eC(ompetition Chernival Ch(emical) e-(ca)r (car)nival Pasco Pas(sion) + (cho)co(late) Instafluences Insta(gram) + (in)fluences Agriventor Agri(culture) +( in)ventor Anthropocene Anthropo(logy) + (s)cene Goviesta Gov(ernment) + (F)iesta Tusning* Tues(day) + (eve)ning source bound root anthropo- is from ‘anthropology’. There are 2 blends in this table that undergoes multiple processes. The blend ebotec consists of 3 sources, one of which is an acronym ELINS, Electronics and Instrumentation, a program study at Universitas Gadjah Mada. This formation process could fall into acronym because the one retained in the first and the third source are only the first graphemes, /e/ from ELINS and /c/ from ‘competition’. However, the fact that the second syllable from the second source is kept /bot/ keeps this formation a blend. In order the new formation pronounceable the vowel /e/ is inserted between the second clipped form and the grapheme /c/. Almost similar is Chernival, this blend undergoes blending process twice: the first syllable “Cher” is from ch(emical) e(-ca)r which is Chem(ical) E(ngineering) Car Competition held at Institut Teknologi Surabaya. Besides competition, they also used shortened “carnival” to name the event. i Instead of combining the final splinter, pasco uses the second syllable and results in Indonesian pronunciation rather than English: /pastjo/ not /pasko/. Although tusning follows the rule of initial splinter combined with final splinter, the process has changed the orthography of Tues to Tus by dropping grapheme /e/ as they are pronounced almost the same in Bahasa Indonesia. All of the blends of this sub-type are nouns. From the data above, only one blend (Lovamil) has BI word source: Lov(e) + (h)amil meaning pregnant. All of the blends of this sub-type are nouns. From the data above, only one blend (Lovamil) has BI word source: Lov(e) + (h)amil meaning pregnant. Morphotactical 1. Total blends The splinter –amil convinces that it is not from ‘milk’ despite the fact that Lovamil is a milk brand for pregnant women, as it lacks of grapheme /a/ in neither love nor milk. b. Both splinters are the beginning of words: acetal— acet(yl) + al(cohol) blends Constituents and process Capcin Cap(pucino) + cin(cau )* Buavita Bua(h)* + vita(min) Sadis Sa(turday) + dis(hes) Cheepo Chee(sy) + po(tato) Fantasteen Fantas(y) + teen(ager) Femfest Fem(inist) + fest(ival ) Robocon Robo(t) + con(test) In this sub-type, there are 2 blends containing BI source word: ‘cincau’ (grass jelly) and ‘buah’ (fruit). The pronunciation of ‘capcin’ is / ʧʌpʧɪn/, moreover, its second syllable is pronounced the same with that from the third of ‘cappucino’. The splinter vita in ‘buavita’ is from ‘vitamin’ and pronounced Indonesian way /vɪtʌ/. It is also found 1 English-like Indonesian word sadis (from English loanword ‘sadistic’) which is actually a blend whose source words are both English: ‘Saturday’ and ‘dish.’ The meaning of the b. Both splinters are the beginning of words: acetal— acet(yl) + al(cohol) blends Constituents and process Capcin Cap(pucino) + cin(cau )* Buavita Bua(h)* + vita(min) Sadis Sa(turday) + dis(hes) Cheepo Chee(sy) + po(tato) Fantasteen Fantas(y) + teen(ager) Femfest Fem(inist) + fest(ival ) Robocon Robo(t) + con(test) b. Both splinters are the beginning of words: acetal— acet(yl) + al(cohol) b. blends Constituents and process Capcin Cap(pucino) + cin(cau )* Buavita Bua(h)* + vita(min) Sadis Sa(turday) + dis(hes) Cheepo Chee(sy) + po(tato) Fantasteen Fantas(y) + teen(ager) Femfest Fem(inist) + fest(ival ) Robocon Robo(t) + con(test) Blends fooshion and fashound, names of event, are interesting in the order of the source words. Both use source word ‘fashion’ yet the order is different, last and initial word, respectively. The order of the source words indicates the order of importance; the initial words represent events held, the second words suggest as supplementary. The source words of proper name, Jogjakarta and Indonesia, in Jogjacation and Idenesia, are also put in different order. The pronunciation of the second blend will be awkward should Indonesia become the initial source words. In this sub-type, there are 2 blends containing BI source word: ‘cincau’ (grass jelly) and ‘buah’ (fruit). The pronunciation of ‘capcin’ is / ʧʌpʧɪn/, moreover, its second syllable is pronounced the same with that from the third of ‘cappucino’. The splinter vita in ‘buavita’ is from ‘vitamin’ and pronounced Indonesian way /vɪtʌ/. Morphotactical 1. Total blends Another sample of noun + adjective is winetastic. i In this sub-type, there are 5 noun + adjective blends with the over-use adjective splinter –licious, which is even regarded as a morpheme (Algeo 1977: 52, Mottiello, 2013: 117) and thus, its combination is regarded more as compound not blends. However these compounds do not follow the rule of English adjective + noun but rather of Indonesian. In the above data, all of the final source words are adjective ‘delicious’. Another sample of noun + adjective is winetastic. i + patho(logy) There are three data belonging to this sub-class of which a splinter of the first source words are replaced by the second source word whose sounds, to Indonesians, are similar to the clipped splinters: /at/ and /art/; /nat/ and /night/; and /vo/ and /food/ which actually are distinct from each other. In terms of duplicating, final splinters –venture, -preneur, -stalgia are among popular final splinters found in this research. In the data, the initial word ‘art’ is often combined with other words. Most of them are in initial position, there is one though in final position (see the next table: dedicart). This shows how blend formation is easily duplicated. Morphotactical 1. Total blends It is also found 1 English-like Indonesian word sadis (from English loanword ‘sadistic’) which is actually a blend whose source words are both English: ‘Saturday’ and ‘dish.’ The meaning of the In this sub-type, there are 2 blends containing BI source word: ‘cincau’ (grass jelly) and ‘buah’ (fruit). The pronunciation of ‘capcin’ is / ʧʌpʧɪn/, moreover, its second syllable is pronounced the same with that from the third of ‘cappucino’. The splinter vita in ‘buavita’ is from ‘vitamin’ and pronounced Indonesian way /vɪtʌ/. It is also found 1 English-like Indonesian word sadis (from English loanword ‘sadistic’) which is actually a blend whose source words are both English: ‘Saturday’ and ‘dish.’ The meaning of the In this sub-class, there are 2 blends with Latin source form: socio- and anthropo- Whereas in the sociophoria, the clipped form socio- is from Sociology, a department in a Political Science Faculty that holds the happy event, the source form socio- in blend sociopreneur, is a prefix or combining form. The 25 Humaniora, Vol. 31, No. 1 (February 2019) Folkamartani Folk + (mino)martani Youthgether Youth + (to)gether Earthernity Earth + (et)ernity Coffeenection Coffee + (con)nection Wine’tastic Wine’ + (fan)tastic heteroddicted Hetero + (a)ddicted Hydrococo hydro + coco(nut) blend is completely different from the meaning of the compound. This blend is intentionally coined to be posted on a poster to be eye-catching. Passer-by then can connect the message of the word/blend sadis with the picture of the poster, displaying a chef holding a big and sharp knife cutting steak, the special dish offered on Saturday evenings at a hotel restaurant. c. Either the beginning or the end of splinter is embedded in a discontinuous splinter (with some reduction in either splinter): autopathography – auto(bio)graphy Either the beginning or the end of splinter is embedded in a discontinuous splinter (with some reduction in either splinter): autopathography – auto(bio)graphy + patho(logy) blends Constituents and process Narture N(at)ure + art Imaginightion Imagi(nat)ion + night Refoodlution Re(vo)lution + food In this sub-type, there are 5 noun + adjective blends with the over-use adjective splinter –licious, which is even regarded as a morpheme (Algeo 1977: 52, Mottiello, 2013: 117) and thus, its combination is regarded more as compound not blends. However these compounds do not follow the rule of English adjective + noun but rather of Indonesian. In the above data, all of the final source words are adjective ‘delicious’. 2. Partial blends Partial blends in which only one source word is reduced a. The full word is followed by a splinter : bloggerific – blogger + (terr)ific Only one blend that complies with English adjective + noun combination: with the noun as the head: creativepreneur. In this sub-type, there is one blend with bound root hetero- functioning like prefix to the adjective ‘addicted.’ blends Constituents and process Sweetylicious Sweety + (de)licious Snacklicious Snack + (de)licious Rubylicious Ruby + (de)licious Cafelicious Café + (de)licious Selfielicious Selfie + (de)licious Artscape Art + (e)scape or (land)scape Artspiration Art + (a)spiration Art-griculture Art + (a)griculture Artventure Art + (ad)venture Eggventure Egg + (ad)venture Soundsations Sound + (sen)sations Creativepreneur Creative + (entre)preneur Foodpreneur Food + (entre)preneur Popstalgia Pop + (no)stalgia Rockstalgia Rock + (no)stalgia blends In forming blend folkamartani vowel /a/ is added in between the full intial word ‘folk’ and the splinter martani a clipped name of a place in Yogya: Minomartani. b. The full word is preceded by a splinter b. The full word is preceded by a splinter blend Constituents + process Kalimilk Kali(urang) + milk Ngayogbook Ngayog(yokarto) + book Ngayogjazz Ngayog(yokarto) + jazz Metamormovies Metamor(phosis) + movies Dedicart Dedic(ation) + art Two initial source words are proper names of place: Kaliurang and Ngayogyakarta, other name of Yogyakarta or Jogjakarta. The clipped parts are mostly syllables. Interestingly, Ngayogyakarta is clipped in 26 Moehkardi.- English Blends in Indonesian Context the same part in the two blends. the same part in the two blends. in guitarun, /l/ specialive, /t/ in artourism, /n/ in iconinety and /k/ in Jogjarockarta to several: two phonemes /ma/in aMayzing and maygic, a syllable / ter/ in lobsterrific and soulotions /soul/ that sounds similar with the overlapped constituent of /sol/. The initial full word of supersemarch is actualy an Indonesian acronym: Surat Perintah Sebelas Maret ( a decree signed by the first Indonesian president, on 11 March, that led Suharto, the second one, came to power as his successor). c. The full word is intercalated within a discontinuous splinter : cewebrity—ce(le)brity + web; but also adorkable—adorable + dork blends Constituents and process AMAYzing Amazing + May JogjaRockarta Jogjakarta + rock Narture N(at)ure + art Imaginightion Imagi(nat)ion + night Refoodlution Re(vo)lution + food b. The constituent overlap both graphically and phonologically with shortening at least one of them: compfusion—comp(uter) + co(n)fusion The second source words in this sub-type are inserted in the initial with or without a cut in the initial. In Jogjarockarta, there is no cut in either first nor second resource word, but overlap graphame /k/. In imaginightion, the vowel /a/ in the third syllable of the initial is clipped. In refoodlution, the consonant /v/ is deleted overlapped fully with second source word. Generally the constituents overlapped are single similar graphemes (see table 2); however, there are two blends with different grapheme overlapped /f/ and /v/ and /sh/ and /s/; but phonetically they are regarded the same. Yet, /at/ in ‘nature’ and that in ‘art’ are almost similar grapheme, to Indonesia they are hardly pronounced differently. Although, ‘tolerance’ and ‘run’ share grapheme /r/, the pronounciation of splinter –rance is replaced the morpheme of ‘run’ . Bahasa Indonesia has the assimilated English tolerance ‘toleran’ whose final syllable is pronounced /-rʌn/ similar to that of ‘run’. There are 3 adjective blends: ‘sempolicious’, ‘FIBulous’ and ‘sluurprising’ two of which with noun initials: ‘sempol’ ( a local snack) and FIB ( a proper name). b. The full word is preceded by a splinter The interesting thing is the initial source word of sluurprising that gives onomatopaeic emphasis on its original word by giving extra vowel /u/. The blend is a combination of verb and Morphonological and graphical formation 1. Overlapping a. Overlap both graphically and phonologically with no other shortening: anecdotage—anecdot + dotage Except lobsterrific which is a noun-adjective combination, the other blends in this sub-type are noun-noun combinations. See table 1. a. Overlap both graphically and phonologically with no other shortening: anecdotage—anecdot + dotage Except lobsterrific which is a noun-adjective combination, the other blends in this sub-type are noun-noun combinations. See table 1. Except lobsterrific which is a noun-adjective combination, the other blends in this sub-type are noun-noun combinations. See table 1. It can be seen that the extent of constituents overlapped is varied, from a single phoneme: /r/ Table 1. blends Constituents and process Constituent overlapped AMAYzing Amazing + May /ma/ Maygic (night) May + magic /ma/ Guitarun Guitar + run /r/ Specialive Special + live /l/ Jogjarockarta Jogjakarta + rock /k/ Soulutions* Soul + solutions /Soul/ and /sol/ to Indonesian are almost phonologically similar Artourism Art + tourism /t/ Motivasinger Motivasi + singer /si/ Supersemarch Supersemar + march /mar/ Iconinety (9) Icon + ninety /n/ Lobsterrific Lobster + terrific /ter/ Table 1. 27 Humaniora, Vol. 31, No. 1 (February 2019) Table 2. blends Constituents and process Constituent overlapped Constituent shortened Medicare Medic(al) + care /c/ -al Biolympic Biol(ogy) + olympic /o/ and /l/ -ogy Javabica Java + (ar)abica /a/ ar- Bassaurus Bass + (dino)saurus /s/ dino- Soundrenaline Sound + (a)drenaline- /d/ a- Lobstreet Lobst(er) + street /st/ -er Sempolicious Sempol + (de)licious /l/ de- FIBulous FIB + (fa)bulous /b/ fa- sluurprising Sluurp + (su)rprising /rp/ su- MUFaganza MUF + (extra)vaganza /v/→/f/ extra- Fashound Fash(ion) + sound /sh/ and /s/→ /ʃ/ -ion Narture Nature + art /eɪ/ → /ɑːt/ and /t/ -r- Tolerun Toler(ance) + Run /r/ /-rǝns/ →/rʌn/ -ance Table 3. blend Formation process Phonological overlap Cultourism Culture + tourism /ture/ and /tour/ / tʃər/ and /tʊər/ Eggsperience Eggs + experience /gs/ and /ex/ /egs/ and /ɪkˈs/ Coolinary Cool + culinary /cool/ and /cul/ /ku:l/ and /kʌl/ Soulutions Soul + solutions /soul/ and /sol-/ /səʊl/ and /səˈl/ Salebrate Sale + celebrate /sale/ and /cele-/ /seɪl/ and /səˈl/ Eatereview Eatery + review /ry/ and /re/ /ri/ and /rɪ/ Table 2. c. The constituent overlap phonologically but not orthographically: buyography—buy + biography Moehkardi.- English Blends in Indonesian Context The following is some samples of left-headed and right-headed blends found in the research. / tʃər/ rather than /tʊər/. To Indonesian, English single vowel is possibly easier than dipthongs.The bold pronunciation of the overlapped constituents are possibly the ones chosen to pronounce the blends. To Indonesian learners of English, pronunciation gives them problems and this sentiment is used by the current research to draw this preliminary conclusion, this part of the study needs further research. In the data, there are 5 noun blends consisting of noun + noun combination whose first noun functions as semantic heads and the second noun , on the right, as the modifier. There are 6 adjective blends, besides sweetylicious, using adjective ‘delicious’ and there are other 4 adjective blends with all the adjective in the right and the heads are nouns in the left side. This formation type (noun + adjective) is a common formation of Indonesian adjective-noun phrase: “ FIB keren” (FIB fabulous) rather than “keren FIB” (fabulous FIB). There is also a verb-like salebrate and an adverb-like youthgether. 1. Attributive blend This type of blends is similar to endocentric compound in which the second word source is the head and its relation is transparent. However, in this research, the original compounds of the blends are not always right-headed. The following is the right-headed and left-headed blends found in the research. a. Left-headed attributive blends. a. Left-headed attributive blends. Blends Constituents Chitato Chips + potato Idenesia Idea + Indonesia Iconinety9 Icon + ninety9 maygic May + magic Bassaurus Bass + dinosaurus Sweetylicious* Sweety + delicious FIBulous FIB + fabulous sluurprising Sluurp + surprising Wine’tastic Wine + fantastic lobsterrific Lobster + terrific salebrate Sale + celebrate youthgether Youth + together b. Right-headed attributive blends Morphonological and graphical formation 1. Overlapping blends Constituents and process Constituent overlapped Constituent shortened Medicare Medic(al) + care /c/ -al Biolympic Biol(ogy) + olympic /o/ and /l/ -ogy Javabica Java + (ar)abica /a/ ar- Bassaurus Bass + (dino)saurus /s/ dino- Soundrenaline Sound + (a)drenaline- /d/ a- Lobstreet Lobst(er) + street /st/ -er Sempolicious Sempol + (de)licious /l/ de- FIBulous FIB + (fa)bulous /b/ fa- sluurprising Sluurp + (su)rprising /rp/ su- MUFaganza MUF + (extra)vaganza /v/→/f/ extra- Fashound Fash(ion) + sound /sh/ and /s/→ /ʃ/ -ion Narture Nature + art /eɪ/ → /ɑːt/ and /t/ -r- Tolerun Toler(ance) + Run /r/ /-rǝns/ →/rʌn/ -ance Table 2. Table 3. blend Formation process Phonological overlap Cultourism Culture + tourism /ture/ and /tour/ / tʃər/ and /tʊər/ Eggsperience Eggs + experience /gs/ and /ex/ /egs/ and /ɪkˈs/ Coolinary Cool + culinary /cool/ and /cul/ /ku:l/ and /kʌl/ Soulutions Soul + solutions /soul/ and /sol-/ /səʊl/ and /səˈl/ Salebrate Sale + celebrate /sale/ and /cele-/ /seɪl/ and /səˈl/ Eatereview Eatery + review /ry/ and /re/ /ri/ and /rɪ/ Although the overlap is not phonologically the same, they are very close and to Indonesians (see table 3), they are regarded as similar, for example /cool/ and /cul/ in ‘culinary’. However, the choice of which splinter’s pronunciation to be chosen in the new blend is not always the same; the easier pronunciation of the splinter the bigger chance to be chosen as the sound of the blend. For example, the blend soulutions, the first syllable is possibly pronounced by Indonesian as the first splinter of the second word source as / səˈl/ rather than /səʊl/ ; however, for cultourism the second syllable of the blend will likely be pronounced adjective, which is actually uncommon combination, resulting in a look alike present participle adjective. There are 2 blends with acronym initials FIB (Fakultas Ilmu Budaya, Faculty of Cultural Sciences at Universitas Gadjah Mada) followed by an adjective (fabulous) and MUF (Mandiri Utama Finance) followed by noun (extravaganza). Originally, FIB and MUF are abbreviations, but for this blending purpose, they are pronounced as words /fɪb/ and /muf/. c. The constituent overlap phonologically but not orthographically: buyography—buy + biography 28 Moehkardi.- English Blends in Indonesian Context 2. Non-overlapping Neither phonological nor orthographical overlap takes place in the blends: Calexico—Cal(ifornia) + (M)exicoi The data classified under this can be observed in part 1.a-b, and 2.a-b above. The nature of most of these blends is endocentric, for example chitato is ‘chips from potato,’ and bassaurus is music event using powerful bass guitars, as powerful as a dinosaurus. Whereas, wine’tastic is understood as ‘fantastic wine’ or ‘wine that is fantastic’ whose two source words exhibit endocentric relationship, sluurprising blend, however, exhibits exocentric relationship whose semantic head ‘beverage’ is outside. Another exocentric blend is Maygic whose semantic head ‘night’ is outside. In terms of form, even when it is pronounced by Indonesian, salebrate is like the verb ‘celebrate’ , but shoppers most likely understand it ‘a sale to celebrate a particular holiday’ rather than ‘celebrate a sale.’ Almost similar is the adverb-like youthgether that likely means ‘young people get together’. b. Right-headed attributive blends Blends Constituents Jogjacation Jogjakarta + education sociophoria Sociology + euphoria sociopreneur Socio + entrepreneur ebotec ELINS + Robot + e-Competition Chernival Chemical + e-car + carnival pasco Passion + chocolate instafluences Instagram + influences agriventor Agriculture + inventor anthropocene Anthropology + scene Goviesta Government + Fiesta Tusning Tuesday + evening artscape Art + escape artspiration Art + aspiration This type of blends is similar to endocentric compound in which the second word source is the head and its relation is transparent. However, in this research, the original compounds of the blends are not always right-headed. Many of the blends are left-headed; it is understandable, because most of Indonesian compounds are left-headed, for example: keripik (chips) + kentang (potato) and not kentang keripik. 29 Humaniora, Vol. 31, No. 1 (February 2019) artgriculture Art + agriculture eggsperience Eggs + experience eggventure Egg + adventure foodpreneur Food + entrepreneur soundsations Sound + sensations soundrenaline Sound + adrenalin popstalgia Pop + nostalgia Javabica Java + arabica Folkamartani Folk + Minomartini Earthernity Earth + eternity Coffeenection Coffee + connection Kalimilk Kaliurang + Milk Tolerun Tolerance + Run Ngayogbook Ngayogyokarto + book Ngayogjazz Ngayogyokarto + jazz metamormovies Metamorphosis + movies cultourism Culture + tourism biolympic Biology + Olympic dedicart Dedication + art lobstreet Lobster + street eatereview Eatery + review heteroddicted Hetero + addicted specialive Special + live medicare Medical+ care creativepreneur Creative + entrepreneur Motivasinger Motivasi + singer Lovamil Love + hamil narture Nature + art imaginightion Imagination + night refoodlution Revolution + food and ‘hamil’ meaning pregnant. The data also have proper names of place as one of the source word: Ngayogyokarto/Jogjakarta, Kaliurang, and Minomartani. Semantically, they are all attributive, for example Jogjarockarta; rock concert in Jogja, another name Yogyakarta, similar to that of Ngayogjazz another attributive blend meaning ‘Jazz concert in Yogyakarta’ which sometimes is called, Ngayogya shortened from Ngayogyakarta Most attributive right-headed blends in the table exhibit endocentric relationship between the source words of which the second word functions as a semantic head and the first one as the modifier. The meaning of the noun + noun blends are transparent, for example: chernival means a carnival of e-car held by Chemistry department at a Surabaya-based engineering university; coffeenection may mean establishing connection with people who share the same interest in coffee or javabica means coffee (of Arabica type) that comes from Java. b. Right-headed attributive blends Whereas creativepreneur means entrepreneur who is creative, foodpreneur can be understood as entrepreneur focussing on food. i There are 3 blends whose first word source is ‘art’ in contrast with one blend whose second source word is ‘art’. Whereas artspiration, for example, may mean aspiration people get from art, dedicart may mean art for dedication or that dedicated to someone. Furthermore, artscape can be interpreted as escape from something to enjoy art, and artgriculture can be understood the agriculture students hold art performance. The above possible meaning of these blends indicate the word creator’s intention of playing with words to create funny, attractive, memorable blends. Like those using ‘delicious’, ‘entrepreneur, ’ the use of “art, street, live’, and some others either as heads or as modifier in creating blends, despite its lack of creativity, seem common in neologism, creating new words. Compared to that of the left-headed blends, there are more of right-headed blends in the data. Most of them are noun + noun combinations; however, there are some of those with adjectives, for example: the semantic right head addicted is preceded by a combining form ‘hetero’ . Unlike the combination noun + adjective shown in the previous table, in this sub-class, to semantically mean the same, the combination used is adjective + noun: creativepreneur, specialive, medicare. It is clear the right heads syntactically determine the word class of the blends. Fashound Fashion + sound Capcin Cappuccino + cincau * Artventure Art + adventure Artourism Art + tourism Motivasinger* Motivator + singer Fashound Fashion + sound Capcin Cappuccino + cincau * Artventure Art + adventure Artourism Art + tourism Motivasinger* Motivator + singer seems intentional: cappucinno with ‘cincau’ (jelly grass) and vitamin made from ‘buah’ (fruit). Few data fit the sub- category in which either the initial or the end of splinter is embedded in a discontinuous splinter. In partial blends, there are more blends whose first source words are kept full and followed by clipped second source word. In morphonological and graphical formation, there is no non-overlapping blend. There are graphically and phonologically overlapping blends with no shortening and that with shortening. The constituents overlapped are from single phoneme to a syllable. In this sub-category, there are a few blends with Indonesian initials: ‘sempol’, and 2 abbreviations that turn into acronyms in the blends: FIB and MUF. Phonological, but not orthographical, overlap also takes place, especially in syllables whose English and Indonesian’ sounds are similar, for example to Indonesia /cool/ and /cul sound similar, therefore phonological overlap is possible. There is one blend motivasinger whose source words could possibly be motivasi + singer, or motivator + singer. Whereas, the former blend includes an assimilated word, the latter use both English source word, despite the fact that ‘motivator’ is also borrowed in Bahasa Indonesia. If the latter is the case, the blend is a blend: someone who is a singer and at the same a motivator (appositional compound, Plag 2002, p. 188) . In this category, there is another blend using an Indonesian word ‘cincau’ grass-jelly. When motivasinger is from motivasi + singer, it is endocentric, the other blends are exoteric whose heads are outside the blend. For example: ‘Fooshion market’ and Robominton competition; a competition of making robot and of badminton.f Morpho-Semantically, attributive blends are more frequent than the coordinative ones. Although understandably, there are more right-headed blends, there are also some left-headed blends that could be influenced by Indonesian word combination rule. The semantic relations of these right-headed blends are endocentrics, whereas in the left-headed there could also be also exocentric blends, especially when the head words are adjective. The coordinate blends whose source words are equal in importance as heads are also likely exocentric. Despite the different order of the source words, semantic explanation of fooshion and fashound is that they are events of food and fashion in fooshion and fashion and sound representing musical performance in fashound. Fashound Fashion + sound Capcin Cappuccino + cincau * Artventure Art + adventure Artourism Art + tourism Motivasinger* Motivator + singer The reason behind this possibly because each event has its own emphasis, the former on food, and the latter on fashion. Besides, each blend phonologically sounds similar to the words that relate to each event each blend represents. The pronunciation of fooshion, to Indonesian, is close to ‘fusion’ which reminds us of fusion food; whereas fashound sounds similar to ‘fashion’. The source words ‘art’, however, are positioned at the left in both blends and have equal importance to the right source words. REFERENCES Algeo, J. (1977). Blends, a structural and systemic view. American Speech, 52(1-2), 47-64. Belieava, N. (2014). A study of English Blend: from strcture to meaning and back again. Word Structures, 7(1), 29-54. Belieava, N. (no date). The power of slanguage: form and meaning of English blends (ppt). Url: www2.uni- siegen.de/-engspra/draem/Beliaeva-ppt.pdf 2. coordinate blends Like coordinate compound whose source words are heads of equal importance, so does the coordinate blend. chocoberry Chocolate + strawberry robominton Robot + badminton Fooshion Food+ fashion There are 2 blends using Indonesian words ‘motivasi’ the assimilated version of ‘motivation’ 30 Moehkardi.- English Blends in Indonesian Context CONCLUSION Carstairs-McCarthy, A. (2002). An Introduction to English Morphology: Words and Their Structure. Edinburgh: Edinburgh University Press. This research found that morphotactically, the data can be classified into total and partial blend. In the sub- classification of total blend, the data can be classified into initial splinter of the first source word + final splinter of the second source, with one exception of Pasco, whose final splinter is the mid splinter of the three-syllable word ‘chocolate’. Some data can be classified into total blends consisting of both initials of the source words. In this sub-category, there are two blends with Indonesian noun source word in each, ‘cincau’ the second source word of capcin, and ‘buah’, the first source word of buavita . This different order of Indonesian words in the combinations Danks, D. (2003). Separating Blends: A Formal Investigation of the Blending Process in English nd Its Relationship to Associated Word Formation Processes (a Thesis). University of Liverpool. Fandrych, Ingrid. (2008). Submorphemic elements in the formation of acronyms, blends and clippings. Lexis: Journal in English Lexicology, 2, 105 – 123. Gries, S.T. (2004). Some characteristics of english morphological blends. A paper downloaded from http://citeseercx.ist.psu.edu/viewdoc/downoad?doi 31 Humaniora, Vol. 31, No. 1 (February 2019) +10.1.1.78.1181&rep=1&type=pdf Pedrosa-Trias, S. (2013). Are there coordinate compound. On-line Proceedings of Mediterranean Morphology Meeting, Centre de Linguistica Teorica, UAB. Retrieved from http://www.lilec.it/mmm/wp/wp- content/uploads/2013/98-111-Padrosa-Trias.pdf. Gries, S.T. (2004). Shouldn’t it be breakfunch. Linguistics, 42(3), 639-667. Hosseinzadeh, N. M. (2014). New blends in English Language. International Journal of English Language and Linguistics Research, 2(2), 15 – 26. Permatasari and Suyudi (2011). The Patterns of Indonesian Blends. Proceeding of Uzbek_Indonesian Joint International Conference, 23-26. Gunadarma University, Jakarta 18 October 2011. Kemmer, R. Types of word formation processes (Ling/Engl 215 course information). Retrieved from http://www. ruf.rice.edu/kemmer/words/wordtypes.html. Plag, I. (2002). Word Formation in English. Cambridge University Press: a draft version. Mattiello, E. (2008). An Introduction to English Slang: a Description of its Morphology, Semantics, and Sociology. Milano: Polimetrica. Quirk, R., Greenbaum, S., Leech, G. and Svartvik, J. (1985). A Comprehensive Grammar of Englsh Grammar. Longman. Mattiello, E. (2013). Extra-Grammatical Morphology in English. Berlin: De Gruyter Mouton. 32
https://openalex.org/W1999362200	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0027089&type=printable	English	null	Disturbances in Body Ownership in Schizophrenia: Evidence from the Rubber Hand Illusion and Case Study of a Spontaneous Out-of-Body Experience	PloS one	2,011	cc-by	8,909	Abstract Background: A weakened sense of self may contribute to psychotic experiences. Body ownership, one component of self- awareness, can be studied with the rubber hand illusion (RHI). Watching a rubber hand being stroked while one’s unseen hand is stroked synchronously can lead to a sense of ownership over the rubber hand, a shift in perceived position of the real hand, and a limb-specific drop in stimulated hand temperature. We aimed to assess the RHI in schizophrenia using quantifiable measures: proprioceptive drift and stimulation-dependent changes in hand temperature. Methods: The RHI was elicited in 24 schizophrenia patients and 21 matched controls by placing their unseen hand adjacent to a visible rubber hand and brushing real and rubber hands synchronously or asynchronously. Perceived finger location was measured before and after stimulation. Hand temperature was taken before and during stimulation. Subjective strength of the illusion was assessed by a questionnaire. Results: Across groups, the RHI was stronger during synchronous stimulation, indicated by self-report and proprioceptive drift. Patients reported a stronger RHI than controls. Self-reported strength of RHI was associated with schizotypy in controls Proprioceptive drift was larger in patients, but only following synchronous stimulation. Further, we observed stimulation- dependent changes in skin temperature. During right hand stimulation, temperature dropped in the stimulated hand and rose in the unstimulated hand. Interestingly, induction of RHI led to an out-of-body experience in one patient, linking body disownership and psychotic experiences. Conclusions: The RHI is quantitatively and qualitatively stronger in schizophrenia. These findings suggest that patients have a more flexible body representation and weakened sense of self, and potentially indicate abnormalities in temporo-parietal networks implicated in body ownership. Further, results suggest that these body ownership disturbances might be at the heart of a subset of the pathognomonic delusions of passivity. Citation: Thakkar KN, Nichols HS, McIntosh LG, Park S (2011) Disturbances in Body Ownership in Schizophrenia: Evidence from the Rubber Hand Illusion and Case Study of a Spontaneous Out-of-Body Experience. PLoS ONE 6(10): e27089. doi:10.1371/journal.pone.0027089 Editor: Michael H. Herzog, Ecole Polytechnique Federale de Lausanne, Switzerland Copyright: 2011 Thakkar 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. Abstract Funding: This work was supported in part by National Institute of Mental Health Grants F31-MH085405-01 (KNT), R01-MH073028 (SP), National Alliance for Research on Schizophrenia and Depression (SP), and National Institute of Child Health and Human Development P30 HD15052. 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: sohee.park@vanderbilt.edu * E-mail: sohee.park@vanderbilt.edu Disturbances in Body Ownership in Schizophrenia: Evidence from the Rubber Hand Illusion and Case Study of a Spontaneous Out-of-Body Experience Katharine N. Thakkar1, Heathman S. Nichols1, Lindsey G. McIntosh1, Sohee Park1,2* 1 Department of Psychology, Vanderbilt University, Nashville, Tennessee, United States of America, 2 Department of Psychiatry, Vanderbilt University Medical College, Nashville, Tennessee, United States of America October 2011 \| Volume 6 \| Issue 10 \| e27089 Materials Subjects were seated in front of a two-compartment, open- ended box (Figure 1). One compartment had a transparent cover, and the other had an opaque cover. Subjects placed one hand in the opaque compartment and the other hand behind a barrier on the table beside the box, rendering both hands hidden from the subject’s view. A lifelike rubber hand was positioned in the transparent compartment. Although visual similarity between the rubber hand and participant’s hand does not have an effect on the RHI [17], a neutral density filter was placed over the cover to de- emphasize the color of the rubber hand. A cape was placed around the subject to cover his or her arms and the end of the rubber hand. The RHI is typically measured with self-report questionnaires, and the shift in perceived hand location before and after stimulation (‘‘proprioceptive drift’’). A physiological correlate of the RHI involves a limb-specific drop in hand temperature [8], with cooling being observed in the stimulated, but not unstimu- lated, hand. Cooling in the stimulated hand was related to self- reported RHI strength, and importantly, hand temperature was unchanged during asynchronous tactile stimulation. A recent study [9] showed that the RHI is intensified and has a more rapid onset in schizophrenia. However, these authors only administered synchronous stimulation and did not measure proprioceptive drift, so it is unclear whether these findings reflect a more liberal threshold to report illusion onset and/or differences in response biases on the questionnaire in patients. Skin temperature measurement was modeled after Moseley, et al. [8] and obtained using a handheld non-contact thermometer (Fluke Corporation, Everett, WA) from three points on the each hand (below the second and fifth digits and on the wrist), marked with washable marker. The major aim of the present study was to systematically investigate disturbed body ownership with the RHI in schizo- phrenia and its relationship to delusions and hallucinations. The strength of the RHI was measured using self-reported intensity of perceptual experiences, proprioceptive drift, and changes in hand temperature. In healthy controls, we also examined RHI strength in relation to psychometric schizotypy. Schizotypy refers to the personality traits that are related to symptoms of schizophrenia and imply a latent liability for the disorder [10]; it was assessed using a self-report measure. Introduction Self-awareness can be separated into body ownership and agency. Body ownership refers to ‘‘the perceptual status of one’s own body, which makes bodily sensations seem unique to oneself’’ [4], and it contributes to a sense of self and a developmental basis for a psychological identity [5]. Agency refers to the subjective experience of being the initiator of one’s actions. Impairments in a sense of agency are thought to underlie several psychotic symptoms. Yet, based on patient report, body ownership is presumed to be largely intact. However, experimental support for intact body ownership in schizophrenia is limited, and clinical observations do not preclude subtle disturbances in body ownership that may contribute to an anomalous sense of agency. ‘‘I felt like an alien that was being used to manipulate humans by other aliens. It felt like someone was looking through my eyes without my consent.’’ –Study participant ‘‘I felt like an alien that was being used to manipulate humans by other aliens. It felt like someone was looking through my eyes without my consent.’’ –Study participant Early descriptions of schizophrenia stress disturbances in self- processing [1], and some posit that altered sense of self is central to both positive and negative symptomatology [2]. Anomalous self- awareness is evident in passivity phenomena, in which the patient does not experience himself as the agent of his actions, instead attributing them to an external source. Disturbances in self-processing in schizophrenia have been discussed in the psychoanalytic tradition [3], but do not easily lend themselves to empirical study. A recent surge of interest in the cognitive neuroscience of self- awareness has cultivated empirical investigations of body owner- ship, for example with the rubber hand illusion (RHI). Watching a PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 1 October 2011 \| Volume 6 \| Issue 10 \| e27089 Body Ownership in Schizophrenia matched for age, sex, and handedness. Although HC had a higher estimated IQ and more education, both groups were well within the normal IQ range and, on average, SZ had completed one year of college. Demographic characteristics and SPQ scores are outlined in Table 1. All subjects were paid for their participation. Introduction rubber hand being stroked while one’s own unseen hand is stroked simultaneously often leads to a sense of ownership over the rubber hand and a shift in perceived position of the real hand toward the rubber hand [6]. The RHI is reduced or absent when tactile stimulation of the real and rubber hands is asynchronous or when the rubber hand is spatially incongruent with the real hand [7]. Thus, visuo-tactile correlation is necessary, but not sufficient, to induce the RHI; it is also dependent on the representation of one’s own body. Participants Twenty-four schizophrenia outpatients (SZ) were recruited from a psychiatric facility in Nashville, TN. Diagnoses were made according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria using structured clinical interviews (SCID-IV). All patients were medicated. Symptoms were assessed with the Brief Psychiatric Rating Scale (BPRS) [11], the Scale for the Assessment of Positive Symptoms (SAPS) [12], and the Scale for the Assessment of Negative Symptoms (SANS) [13]. Twenty-one healthy control participants (HC) without a history of DSM-IV Axis I disorder or use of psychotropic medications were recruited from the same commu- nity by advertisements. The Schizotypal Personality Questionnaire (SPQ) [14] was administered to HC. All subjects were screened for neurological disorders, drug use and past head injury. Finally, the subject was asked to provide an open-ended description of the experience and was given a standard questionnaire [6] to rate the occurrence of nine perceptual effects (Figure 2) on a 7-point Likert scale from -3 to 3. This procedure was performed four times, for both synchrony conditions in both hands. Trial order was counterbalanced across subjects. Finally, the subject was asked to provide an open-ended description of the experience and was given a standard questionnaire [6] to rate the occurrence of nine perceptual effects (Figure 2) on a 7-point Likert scale from -3 to 3. This procedure was performed four times, for both synchrony conditions in both hands. Trial order was counterbalanced across subjects. Procedure Prior to tactile stimulation, four measurements of perceived index finger location of the hidden hand under the opaque cover were taken. A ruler was placed on top of the box, and subjects were asked to verbally indicate their perceived index finger position. Different rulers, each offset by a random length, were used to prevent subjects from reciting the same number for each measurement. Baseline skin temperature was recorded at the three locations on the stimulated and unstimulated hand. Following pre- stimulation measurements, the experimenter brushed the index finger of the rubber hand and that of the subject’s invisible hand with a paintbrush at approximately one stroke per second. The subject was instructed to watch the rubber hand and report any change in sensation they might experience. The fingers were brushed for 3 minutes, either synchronously or asynchronously (180 degrees out-of-phase). During stimulation, a second experi- menter recorded skin temperature at the three locations of each hand at 1, 2, and 3 minutes. Following stimulation, four measurements were taken of the subject’s perceived index finger location of the hidden hand. See Figure 1 for experimental set-up. In addition, a movie of the procedure is available at http:// vanderbilt.edu/parklab/Projects.html. Ethics Statement This study was conducted according to the principles expressed in the Declaration of Helsinki. The Vanderbilt Institutional Review Board approved the study protocol and informed consent procedure. After complete description of the study to the subjects, written informed consent was obtained. Materials We hypothesized that a stronger RHI in schizophrenia would be consistent with the notion of a weaker or more flexible sense of body ownership. PLoS ONE \| www.plosone.org Analysis Proprioceptive Drift. Proprioceptive drift was quantified as the difference between mean perceived index finger location before and after tactile stimulation. Positive numbers indicate drift toward the rubber hand. Skin Temperature. Temperature was averaged across the three locations on both stimulated and unstimulated hands at each of the three measurement time points. The change in temperature for each hand in each condition was quantified as the temperature difference between baseline and the average of the three Skin Temperature. Temperature was averaged across the three locations on both stimulated and unstimulated hands at each of the three measurement time points. The change in temperature for each hand in each condition was quantified as the temperature difference between baseline and the average of the three Intelligence was assessed with the Adult North American Reading Test (ANART) [15]. Handedness was assessed using the Modified Edinburgh Handedness Inventory [16]. All subjects had normal or corrected-to-normal vision. The two groups were Intelligence was assessed with the Adult North American Reading Test (ANART) [15]. Handedness was assessed using the Modified Edinburgh Handedness Inventory [16]. All subjects had normal or corrected-to-normal vision. The two groups were PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 October 2011 \| Volume 6 \| Issue 10 \| e27089 2 Body Ownership in Schizophrenia Table 1. Demographic characteristics of the patient and control groups. Patients (n = 24) Mean (s.d.) Controls (n = 21) Mean (s.d.) t p Age 41.7 (8.3) 40.1 (9.1) 0.6 0.53 Sex 9F/15 M 10 F/11 M Phi = 0.51 0.49 Handedness2 60.0 (57.5) 82.6 (43.0) 1.5 0.15 Estimated IQ (ANART) 101.1 (9.9) 106.5 (6.3) 2.1 0.04 Years of education 13.0 (2.3) 15.8 (2.3) 3.8 0.0004 SPQ-Positive Syndrome N/A 3.0 (3.2) SPQ-Negative Syndrome N/A 5.2 (4.2) SAPS 14.3 (9.6) N/A SANS 25.0 (17.2) N/A BPRS 14.1 (8.1) N/A Chlorpromazine (CPZ) equivalent dose3 362.8 (318.9) N/A Psychotropic Medications 4 typical antipsychotic drugs N/A 19 atypical antipsychotic drugs 1 antidepressant Duration of illness (years) 20.4 (10.1) N/A 1The Phi value is the result of a Fishers exact test. 2Edinburgh handedness inventory : 2100 completely sinistral to +100 completely dextral. 3Chlorpromazine (CPZ) equivalent dose of antipsychotic medication (mg/kg/day). N/A = not applicable. doi:10.1371/journal.pone.0027089.t001 within-subjects variable. Spearman rank-correlation coefficients were used to evaluate the association between RHI measures and the severity of symptoms in SZ, and SPQ subscale scores in HC. stimulation time points. Analysis Negative scores indicate cooling. We collapsed across time points because illusion strength varies across the stimulation period and the strength of the illusion was found to correspond with temperature modulation [8]. Since it was not feasible to track RHI strength at each time point, we averaged across them. Temperature data from one patient was lost due to recording error. Self-report RHI Questionnaire There was a main effect of synchrony (F(1,43) = 16.5, p = 0.0002). Paired t-tests indicated that, in both groups, the strength of the RHI was greater during synchronous than asynchronous stimulation (t(44) = 4.0, p = 0.0002). Although there was a significant effect of group (F(1,43) = 5.7, p = 0.02), there was no group-by-synchrony interaction effect (F(1,43) = 0.7, p = 0.4). SZ endorsed RHI-related perceptual effects more strongly than HC in both synchronous (t(43) = 2.0, p = 0.05) and asynchronous (t(43) = 2.7, p = 0.01) conditions. There were significant item Statistical Analyses. Separate repeated-measures ANOVAs were conducted on self-reported illusion strength from the questionnaire, proprioceptive drift, and temperature modulation, with diagnosis entered as a between-group variable and stimulated hand (right or left) and synchrony condition (synchronous or asynchronous) entered as within-subject variables. For questionnaire data analysis, item was also entered as a within- group variable. For temperature modulation, stimulation condition (stimulated or unstimulated) was also entered as a Figure 1. Experimental set-up. doi:10.1371/journal.pone.0027089.g001 PLoS ONE \| www.plosone.org 3 October 2011 \| Volume 6 \| Issue 10 \| e27089 Figure 1. Experimental set-up. Figure 1. Experimental set-up. doi:10.1371/journal.pone.0027089.g00 Figure 1. Experimental set-up. doi:10.1371/journal.pone.0027089.g00 Figure 1. Experimental set-up. doi:10.1371/journal.pone.0027089.g001 PLoS ONE \| www.plosone.org 3 October 2011 \| Volume 6 \| Issue 10 \| e27089 Body Ownership in Schizophrenia Figure 2. Mean rating for each item in the self-report questionnaire for each group following synchronous and asynchronous tactile stimulation. doi:10.1371/journal.pone.0027089.g002 Figure 2. Mean rating for each item in the self-report questionnaire for each group following synchronous and asynchronous tactile stimulation. d i 10 1371/j l 0027089 002 Figure 2. Mean rating for each item in the self-report questionnaire for each group following synchronous and asynchronous tactile stimulation. doi:10.1371/journal.pone.0027089.g002 (F(8,344) = 22.4, p,0.0001) and synchrony condition-by-item effects (F(8,344) = 5.6, p,0.0001). After Bonferroni correction for multiple comparisons, only items 1–3 and 6 were rated more strongly in the synchronous condition (Figure 2). Please note that these first three items have been found to be most reliably endorsed following synchronous stimulation and our results replicate those of others [6]. temperature revealed a significant effect of both hand (F(1,42) = 41.1, p,0.0001) and stimulation condition (F(1,42) = 30.1, p, 0.0001). Temperature was higher in the left hand, and baseline temperature in the stimulated hand was greater than in the unstimulated hand. Clinical symptoms and schizotypy We examined correlations between clinical symptoms and the following measures in patients: self-reported illusion vividness for questions 1–3 for both synchrony conditions (since these item scores were significantly different between synchrony conditions in the current and previous studies), mean proprioceptive drift in the two synchrony conditions, and mean change in temperature in the stimulated and unstimulated hand during right hand stimulation (since no stimulation effect on temperature was observed when the left hand was stimulated). In HC, we examined the relationship between these measures and schizotypy. In patients, we explored the relationship between these measures and BPRS, SAPS, and SANS. Because we were specifically interested in the relationship between body ownership and hallucinations and delusions, we also investigated correlations between RHI measures and individual items from the SAPS subscales. SAPS subscale and individual item scores were missing from one patient. Self-report RHI Questionnaire However, see Data S1 for results from a control experiment that indicated the limb-specific temperature modula- tions observed in the main experiment were not caused by idiosyncratic parameters of the experimental environment, but instead, were related to RHI induction. Proprioceptive drift Results are displayed in Figure 3. There were significant main effects of group (F(1,43) = 5.1, p = 0.03) and synchrony condition (F(1,43) = 7.5, p = 0.009), with proprioceptive drift being greater in SZ and larger following synchronous than asynchronous stimula- tion. Importantly, there was also a significant group-by-synchrony interaction (F(1,43) = 5.4, p = 0.02). Greater proprioceptive drift was observed in SZ following synchronous (t(43) = 2.8, p = 0.009), but not asynchronous stimulation (t(43) = 0.9, p = 0.39). Paired t- tests indicated that although proprioceptive drift was greater following synchronous than asynchronous stimulation in SZ (t(23) = 3.3, p = 0.003), this effect was not observed in HC (t(20) = 0.3, p = 0.74). These differences in proprioceptive drift cannot be explained by differences in perceived baseline location, as the initial error did not differ between groups (t(43) = 0.4, p = 0.71) or synchrony conditions (t(44) = 1.1, p = 0.28). Temperature Modulation Further, skin temperature increase in the unstimulated hand during right hand stimulation was related to increased delusions of there were no significant correlations with the overall delusions subscale score, self-reported illusion strength in the synchronous condition was related to increased delusions of reference (rs = 0.65, p = 0.0008) and delusions of control (rs = 0.53, p = 0.009), and reduced scores for somatic delusions (rs = 20.44, p = 0.03). Further, skin temperature increase in the unstimulated hand during right hand stimulation was related to increased delusions of Temperature Modulation Results are displayed in Figure 4. A significant effect of stimulation on temperature change was observed (F(1, 42) = 4.7, p = 0.04), such that temperature dropped in the stimulated hand and rose in the unstimulated hand. However, this effect was moderated by hand of stimulation, indicated by a stimulation-by- hand interaction (F(1,42) = 21.8, p,0.0001). Significant heating of the unstimulated hand (t(43) = 2.3, p = 0.02) and cooling of the stimulated hand (t(43) = 2.8, p = 0.008) were only observed when the right hand was brushed. However, an analysis of baseline In HC, self-reported vividness of the illusion during both synchronous and asynchronous stimulation was associated with increased positive (synchronous: rs = 0.61, p = 0.003; asynchro- nous: rs = 0.49, p = 0.03) and negative (synchronous: rs = 0.59, p = 0.005; asynchronous: rs = 0.62, p = 0.003) schizotypy. In SZ, self-reported RHI strength in the synchronous, but not asynchronous, condition was related to increased score on the hallucinations subscale of the SAPS (rs = 0.48, p = 0.02). Although In SZ, self-reported RHI strength in the synchronous, but not asynchronous, condition was related to increased score on the hallucinations subscale of the SAPS (rs = 0.48, p = 0.02). Although PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 4 Body Ownership in Schizophrenia Figure 3. Mean proprioceptive drift for each group following synchronous and asynchronous tactile stimulation. Vertical bars denote standard error of the mean. doi:10.1371/journal.pone.0027089.g003 Figure 3. Mean proprioceptive drift for each group following synchronous and asynchronous tactile stimulation. Vertical bars denote standard error of the mean. doi:10.1371/journal.pone.0027089.g003 thought insertion (rs = 0.41, p = 0.05). No significant correlations with BPRS and SANS scores were observed. there were no significant correlations with the overall delusions subscale score, self-reported illusion strength in the synchronous condition was related to increased delusions of reference (rs = 0.65, p = 0.0008) and delusions of control (rs = 0.53, p = 0.009), and reduced scores for somatic delusions (rs = 20.44, p = 0.03). Body Ownership in Schizophrenia Body Ownership in Schizophrenia minute tactile stimulation period, he reported that he had returned to his body. During both hands of the asynchronous condition, he reported that the feeling of the rubber hand was his own fluctuated and was less strong than during synchronous stimulation. He also reported that he sometimes felt that the unstimulated hand was in the same location as the rubber hand. spontaneous OBE in the laboratory is extremely rare, we decided to conduct a case report in depth. We refer to this participant as R.M. throughout the case report. R.M. agreed to return for an additional visit to investigate whether his OBE was replicable and to obtain a more detailed timeline of his subjective experience. to obtain a more detailed timeline of his subjective experience. Experimental set-up and procedure. Set-up and procedure were identical to that outlined above. Since R.M. reported an OBE during the synchronous condition for the left hand during the initial experiment, we repeated the experimental procedure on a separate day. In the follow-up procedure, synchronous stimulation was only administered to the left hand, and the stimulation period was extended to 10 minutes. After the follow-up session, R.M. was asked a series of questions about the phenomenological details regarding previous OBEs and the OBEs he experienced during the rubber hand illusion. These questions were based on a published report of the phenomeno- logical correlates of OBEs in neurological patients [18]. Demo- graphics and medical, psychiatric, and social history were also obtained. A brief, comprehensive neuropsychological examination was administered. The testing battery included the Repeatable Battery for the Assessment of Neuropsychological Status [19], which includes five subscales: Immediate Memory, Visuospatial/Con- struction, Language, Attention, and Delayed Memory. Trails A and B and the Stroop (Golden Version) were administered to assess processing speed and executive functions. Demographically- normalized scores were computed. Experimental set-up and procedure. Set-up and procedure were identical to that outlined above. Since R.M. reported an OBE during the synchronous condition for the left hand during the initial experiment, we repeated the experimental procedure on a separate day. In the follow-up procedure, synchronous stimulation was only administered to the left hand, and the stimulation period was extended to 10 minutes. During the follow-up session where we synchronously stimulat- ed the left hand for approximately ten minutes, the subject reported a fluctuating sensation of ownership of the rubber hand. Body Ownership in Schizophrenia 2:23 Sees his own wrinkles in the rubber hand 2:49 Reports: ‘‘Really feels like my hand’’ 3:47 Reports imagining things he could do with the hand, like play the violin 4:04 Reports that the rubber hand, ‘‘looks like rubber hand again’ 4:04 Reports that the rubber hand, looks like rubber hand again 4:20 Reports that the rubber hand doesn’t feel like his hand anymore 4:04 Reports that the rubber hand, looks like rubber hand again 4:20 Reports that the rubber hand doesn’t feel like his hand anymore Participant. R.M. is a 55-year-old, right-handed, Caucasian male who was diagnosed with schizoaffective disorder at the age of 23. Since his first hospitalization in 1979, he estimated the number of psychiatric hospitalizations to be around 15–20, with the most recent in May of 2009. R.M. was free of neurological disorder, history of seizures, and history of ECT. He subject reported one head injury with loss of consciousness at the age of 16, but he sustained no internal bleeding and there was no evidence of cognitive sequelae following the injury. R.M. did not meet criteria for any alcohol or substance abuse or dependence and reports smoking marijuana only twice in his lifetime. Participant. R.M. is a 55-year-old, right-handed, Caucasian male who was diagnosed with schizoaffective disorder at the age of 23. Since his first hospitalization in 1979, he estimated the number of psychiatric hospitalizations to be around 15–20, with the most recent in May of 2009. R.M. was free of neurological disorder, history of seizures, and history of ECT. He subject reported one head injury with loss of consciousness at the age of 16, but he sustained no internal bleeding and there was no evidence of cognitive sequelae following the injury. R.M. did not meet criteria for any alcohol or substance abuse or dependence and reports smoking marijuana only twice in his lifetime. Body Ownership in Schizophrenia He also reported an OBE, which evolved into a perception of himself and the experimenter hovering above the table. Below is his timeline (in minutes and seconds) of his self-reported experience: After the follow-up session, R.M. was asked a series of questions about the phenomenological details regarding previous OBEs and the OBEs he experienced during the rubber hand illusion. These questions were based on a published report of the phenomeno- logical correlates of OBEs in neurological patients [18]. Demo- graphics and medical, psychiatric, and social history were also obtained. After the follow-up session, R.M. was asked a series of questions about the phenomenological details regarding previous OBEs and the OBEs he experienced during the rubber hand illusion. These questions were based on a published report of the phenomeno- logical correlates of OBEs in neurological patients [18]. Demo- graphics and medical, psychiatric, and social history were also obtained. 0:48 Reports: ‘‘This is my hand; it looks like my hand’’ 1:07 Reports feeling of ownership over rubber hand disappear- ing 1:20 Report that the rubber hand feels and looks like his hand 1:40 Reports fluctuating feeling of ownership over rubber hand 1:46 Reports: ‘‘This is my hand; looks like my hand’’ 2:10 Reports that the knuckles of the rubber hand sometimes look like his own A brief, comprehensive neuropsychological examination was administered. The testing battery included the Repeatable Battery for the Assessment of Neuropsychological Status [19], which includes five subscales: Immediate Memory, Visuospatial/Con- struction, Language, Attention, and Delayed Memory. Trails A and B and the Stroop (Golden Version) were administered to assess processing speed and executive functions. Demographically- normalized scores were computed. A brief, comprehensive neuropsychological examination was administered. The testing battery included the Repeatable Battery for the Assessment of Neuropsychological Status [19], which includes five subscales: Immediate Memory, Visuospatial/Con- struction, Language, Attention, and Delayed Memory. Trails A and B and the Stroop (Golden Version) were administered to assess processing speed and executive functions. Demographically- normalized scores were computed. Body Ownership in Schizophrenia y 5:02 Reports ‘‘strange feeling; that out of body deal’’ y 5:02 Reports ‘‘strange feeling; that out of body deal’’ 5:37 Reports that he and the experimenter are levitating 5:37 Reports that he and the experimenter are levitating 6:40 Reports that he and the experimenter are ‘‘turning in a circle; rubber hand looks like my hand’’ y 7:07 Reports, ‘‘feels like we’re a foot off the floor, turning in a circle’’ 7:07 Reports, ‘‘feels like we’re a foot off the floor, turning in a circle’’ 7:39 Reports, ‘‘feels like we’re coming back down; felt like there wasn’t a floor beneath my feet’’ 7:39 Reports, ‘‘feels like we’re coming back down; felt like there wasn’t a floor beneath my feet’’ 8: 27 Reports being back on ground 9: 04 Reports that the rubber hand looks like rubber hand, not his hand 9: 04 Reports that the rubber hand looks like rubber hand, not his hand R.M. holds a Master’s degree and was a reporter until 2002. Since then he has been employed in restaurants, and recently has been volunteering as an archivist and seeking employment as a freelance writer. Description of RHI-induced OBEs. Visual phenome- nology: In both his past and RHI-induced OBEs, he described his visuo-spatial perspective in a prone-positioned, elevated second body outside the physical body. He described them as vivid and veridical, and he reported seeing his whole body and other objects in the room in their actual positions. However, he only saw the people and objects in his immediate space (i.e. the experimenter performing the tactile stimulation and the table). In both RHI- induced OBEs, he reported his back to the ceiling, looking at himself seated. Although he reported only seeing from one visuo- spatial perspective during the first OBE and then ‘‘snapping out of it’’ after about 1 minute, he said there was a ‘‘duality’’ about the second OBE episode, where he alternated between his seated and elevated visuo-spatial perspectives and gradually came back to his body. Vision and hearing were adequate for testing. Recently, R.M. received eye surgery for glaucoma and indicated normal acuity at the time of the experiment. He also reported Tarsal tunnel syndrome in his left ankle; no additional significant medical history was reported. His current medications included thiothixine (3 mg daily), fenofibrate for cholesterol, omeprazole for stomach acid, and travoprost ophthalmic solution and betaxolol eye drops to manage his glaucoma. Case Report One individual with schizophrenia reported an out-of-body experience (OBE) during RHI induction with synchronous stimulation applied to the left hand. Because observation of a Figure 4. Mean change in skin temperature in both the unstimulated and stimulated hand during right and left hand stimulation for patients and controls. Vertical bars denote standard error of the mean. doi:10.1371/journal.pone.0027089.g004 Figure 4. Mean change in skin temperature in both the unstimulated and stimulated hand during right and left hand stimulation for patients and controls. Vertical bars denote standard error of the mean. doi:10.1371/journal.pone.0027089.g004 October 2011 \| Volume 6 \| Issue 10 \| e27089 October 2011 \| Volume 6 \| Issue 10 \| e27089 PLoS ONE \| www.plosone.org 5 Body Ownership in Schizophrenia On the other hand, the OBEs he experiences while awake begin in a similar manner, but then become qualitatively different. He described the feeling of elevation that progresses into religious delusions in which he talks to angels and demons in the sky. He noted that he eventually feels a greater sense of ownership over his ‘second body’ and described this ‘second body’ as a puppeteer and his real body as a marionette. He said these OBEs come and go for about 4-5 days, until he is eventually hospitalized. Thus, R.M. described an explicit link between body disownership and psychosis. Although self-reported intensity of the illusion was stronger in both synchrony conditions, there was a group difference in proprioceptive drift only for the synchronous condition. Principal components analyses of self-report questions indicate that ownership and location shifts are dissociable components of the RHI [22], a finding which has been supported by recent experimental data [23]. Functional MRI data also suggest that ownership and location shifts are dissociable, and point to a relationship between ventral premotor activation and subjective report [24] and inferior parietal lobule (IPL) activation during recalibration of perceived hand position [24,25]. Since the self- report questionnaire contains items related to ownership, it is possible that activity in premotor cortex corresponds with feelings of ownership over the rubber hand, and IPL activation is related to proprioceptive drift [24,25]. This idea is bolstered by the finding that TMS applied over right IPL reduced proprioceptive drift, but not subjective report of RHI intensity [26]. In SZ, body ownership could be even more flexible than perceived location of the body, resulting in a strong sub- jective sense of the illusion, even when tactile stimulation is asynchronous. After the completion of this experiment and interview, he requested, and was given further information about the OBE, including peer-reviewed journal articles. He then recognized that his experiences may have a tangible cause and interestingly he did not have another psychotic episode even though he had experienced an awake OBE during this experiment. At the time of the final preparation of this article, 10 months after he had participated in this study, he was still in remission, was working part-time, and had a paper accepted for publication in an academic journal. Thus, in this particular case, gaining insight into the etiology of anomalous experiences appears to be therapeutic. However, it must be noted that R.M. Body Ownership in Schizophrenia These data support the finding of increased RHI vividness in SZ [20]. Across groups, the RHI was subjectively stronger during synchronous stimulation, but we also observed that the RHI was greater in SZ compared with HC during both synchronous and asynchronous stimulation. Greater vividness of the illusion in SZ compared with HC in the asynchronous condition could indicate increased response bias or suggestibility in the patients. However, an alternative explanation is that asynchronous stimulation may be perceived as less temporally asynchronous, or even synchronous, by patients due to a coarser window of temporal integration. There is evidence to indicate that SZ have a longer time window in which they perceive two temporally separate unimodal and bimodal visual and auditory events as simultaneous [21]. Additional evidence that group differences in the RHI are not driven by response biases or suggestibility comes from the novel analysis of proprioceptive drift. Patients and controls showed equal proprioceptive drift in the asynchronous condition, but drift in the synchronous condition was nearly three times as large in SZ. Thus, data from both self-report and proprioceptive drift support a more robust RHI in schizophrenia. anything. It is possible that during OBEs, auditory processing is suppressed, although we are not aware of other reports of this phenomenon. He could still feel the tactile sensation of the brush. He reported that although his second body was complete, he did not really think about or acknowledge that body. anything. It is possible that during OBEs, auditory processing is suppressed, although we are not aware of other reports of this phenomenon. He could still feel the tactile sensation of the brush. He reported that although his second body was complete, he did not really think about or acknowledge that body. OBE history. During the extended interview, R.M. reported that has had too many OBEs to count. Typically, they occur before he falls asleep and last approximately 10 minutes. He has had about 12 OBEs while he was awake and alert, and they have always preceded a psychotic episode. Those OBEs last approximately 15 minutes at first, but get longer as his psychotic episode progresses. He reported that his first OBE happened when he was at church camp at age 16. R.M. indicated that his RHI-induced OBEs felt more like the OBEs he has while falling asleep. Body Ownership in Schizophrenia is a very high-functioning individual with exceptional verbal skills and insight into his conditions. Furthermore, had we known that there was a link between awake OBEs and psychotic episodes for R.M., we certainly would not have conducted this case study. To be cautious, future studies of RHI in schizophrenia-spectrum should consider screening out those who are prone to OBEs if their OBEs are correlated with psychotic episodes. Although we did not observe group differences in skin temperature modulation, our finding of cooling in the stimulated right hand during RHI induction partly replicated the findings of Moseley, et al. [8]. However, contrary to prior findings, no differences in temperature modulation between synchrony conditions were observed. Additionally, temperature in the stimulated hand was warmer than the unstimulated hand at baseline (i.e., before brushing occurs). The cause of this baseline difference is unclear, and might be due to arousal, or orienting effects [27]; data from a control experiment rule out environ- ment-specific factors (See Data S1). Data from this control experiment also indicate that the observed cooling of the stimulated right hand and warming of the unstimulated left hand, are not caused by environmental confounds. Moseley and colleagues [8] posit that skin cooling is related to a sense of disownership of the stimulated hand; however, temperature modulation might also result from changes in arousal or orienting. These hypotheses are not necessarily mutually exclusive. Lateralization of skin temperature changes is curious; it might be related to hemispheric contributions to body illusions [18,28,29] and should be explored in future neuroimaging or lesion studies. Body Ownership in Schizophrenia Vision and hearing were adequate for testing. Recently, R.M. received eye surgery for glaucoma and indicated normal acuity at the time of the experiment. He also reported Tarsal tunnel syndrome in his left ankle; no additional significant medical history was reported. His current medications included thiothixine (3 mg daily), fenofibrate for cholesterol, omeprazole for stomach acid, and travoprost ophthalmic solution and betaxolol eye drops to manage his glaucoma. Overall, R.M’s neurocognitive functioning was within normal range. He had significant psychomotor slowing, which is consistent with a schizophrenia diagnosis and neuroleptic use. Memory, visuospatial processing, construction, language, and executive functions (inhibition, context switching) were grossly intact. His estimated premorbid IQ was in the superior range at 120. Non-visual phenomenology: In the second RHI-induced OBE, he reported numbness in his feet and hands that preceded the OBE, but this did not occur for the first RHI-induced OBE, nor his past OBEs. For the RHI-induced OBEs, he reported a feeling of elevation but no experience of rotating into the inverted position. Although during the first OBE he felt fearful because he thought he might be going into a psychotic episode (see section on OBE history) and was relieved after he came out of the OBE, he reported the second OBE was pleasant and that he wanted the feeling to come back. Interestingly, he conveyed that for both OBEs, there was ‘‘a silence about it’’, and he could not hear Open-ended description of RHI. In the initial rubber hand experiment, following synchronous stimulation of the right hand, R.M. reported the sense that the rubber hand felt like his hand, and eventually that his real hand moved to the position of the rubber hand at into the trial. He reported the same sensation during synchronous stimulation of the left hand. After a while, he reported that he felt he was ‘‘out of his body’’, and was hovering above and looking down on himself. Before the end of the 3- PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 6 Body Ownership in Schizophrenia Acknowledgments Another limitation of the current study is that the illusion was not very robust in HC. Although we did observe drift towards the rubber hand in healthy controls, this effect was small during both synchronous (Cohen’s d = 0.29) and asynchronous (Cohen’s d = 0.19) brushing. One probable reason for this finding is the relatively short period of tactile stimulation compared to prior studies [6]. We chose the briefest possible duration of stimulation from past reports of RHI in order to minimize the total testing time for our participants. Nevertheless, even with the brief stimulation duration, psychometric schizotypy was associated with strength of the RHI in healthy participants. It is also evident from Figure 2 that the mean We would like to thank Ms. Channing Cochran for her assistance with data collection, Dr. Carissa Cascio, Dr. Peter Brugger and Ms. Leonie Hill for methodological and conceptual guidance, and Mr. Holden Bitner for helping to make the video of the procedure. Supporting Information Data S1 Supplementary Data. Methods, results, and discus- sion of a control experiment for temperature modulation during RHI induction. (DOCX) Potential Mechanisms self-reported RHI intensity for many questions fell below zero in both groups. However, it is important to note that a rating of 0 does not mean that the particular perceptual effect was absent. Rather, a score of 23 (definitely disagree) indicates the absence of the effect. self-reported RHI intensity for many questions fell below zero in both groups. However, it is important to note that a rating of 0 does not mean that the particular perceptual effect was absent. Rather, a score of 23 (definitely disagree) indicates the absence of the effect. With regard to temperature data, our task parameters might not be optimal for measuring subtle temperature modulations. Moseley, et al. [8] used a longer stimulation period and began measuring skin temperature only after subjects reported the onset of the illusion. Further, they measured temperature at shorter time intervals. Future studies investigating RHI-induced physiological changes in schizophrenia should take these methodological considerations into account. Based on current findings, several non-mutually exclusive hypotheses for mechanisms underlying a stronger RHI in SZ emerge. The RHI may arise from bottom-up multisensory integration of conflicting signals, with visual information overwrit- ing the relatively weaker proprioceptive input. However, the RHI is significantly modulated in a top-down fashion by the internal model of the body, such that the illusion is typically not observed when the rubber hand is spatially incongruent with the real hand [7] or when non-corporeal objects are used [7,30]. One possibility is that disturbances in multisensory integration lead to an exaggerated RHI in SZ, such that visual information is overweighed because of altered proprioception [31]. However, this hypothesis is inconsistent with findings of reduced influence of visual information on proprioceptive [32] and auditory [32,33] signals in schizophrenia. Alternatively, SZ might have a weaker or more flexible internal model of their body, making them more susceptible to the illusion. This would be consistent with the anecdotal observations in this and a previous study [20], that some patients experience the illusion even before tactile stimulation begins. With regard to temperature data, our task parameters might not be optimal for measuring subtle temperature modulations. Moseley, et al. [8] used a longer stimulation period and began measuring skin temperature only after subjects reported the onset of the illusion. Further, they measured temperature at shorter time intervals. Future studies investigating RHI-induced physiological changes in schizophrenia should take these methodological considerations into account. Implications Correlations between schizotypy and RHI strength suggest a weaker sense of body ownership in those who may share latent liability for schizophrenia, although we cannot rule out response bias effects. In SZ, correlations between RHI strength and positive, but not negative symptoms, suggest that disturbed body ownership might contribute to psychotic symptoms. Indeed, one patient reported an explicit link between body disownership and psychotic episode onset. Since our sample included fairly asymptomatic patients, future studies should examine the RHI in patients with more severe symptoms. Nevertheless, these results present an intriguing potential link between body ownership disturbances and the agency abnormalities characteristic of many positive symptoms, which may have implications for treatment. For example, yoga, which promotes body awareness, has been shown to be effective in attenuating positive and negative symptoms and increasing social functioning in schizophrenia, exceeding the beneficial effects of aerobic exercise [42,43]. Regarding neural mechanisms, fMRI [25] and TMS [34] data implicate the right temporoparietal junction (TPJ) and IPL in body ownership. Further, somatoparaphenia, a neurological syndrome associated with dis-ownership of a body part, is most often associated with lesions in these regions [29], and right TPJ seizure focus is associated with OBEs [18,28]. Although there is relatively little data on IPL and TPJ function and structure in schizophrenia, behavioral and neuroimaging studies implicate possible IPL and TPJ disturbance in schizophrenia spectrum disorders [35,36]. A stronger RHI in schizophrenia might also be related to NMDA glutamate hypofunction. The glutamate hypothesis of schizophrenia is promising, due to the observation that NMDA antagonists mimic both positive and negative symptoms and can account for the developmental profile of schizophrenia [37]. Interestingly, a recent study found that ketamine administration to healthy volunteers led to an increase in self-reported RHI vividness and proprioceptive drift in both synchronous and asynchronous stimulation conditions [38], and frequency of ketamine use has been found to correlate more strongly with OBE frequency than other drugs [39]. Limitations One limitation of this study is that we cannot eliminate potential effects of antipsychotic medication, but it seems unlikely that neuroleptic effects drive these findings. Ketamine administration results in a stronger RHI in healthy volunteers [38]. Pretreatment with an atypical antipsychotic blocks ketamine-induced metabolic activation in the rodent brain [40] and attenuates ketamine- induced increases in positive symptoms in schizophrenia [41]. Furthermore, CPZ equivalent dose was not related to any of the RHI measures in our patients (all p’s .0.10). Finally, self-reported RHI strength correlated with schizotypy in unmedicated healthy participants, a finding that cannot be attributed to medication. Conclusions Patients with schizophrenia showed a stronger RHI, indexed by self-report and mislocalization of their own hand, suggesting that body ownership is disturbed in schizophrenia. Correlations between RHI strength and clinical symptoms and psychometric schizotypy suggest that disturbed body ownership is close to the heart of the schizophrenic experience. Moreover, our in-depth investigation of the spontaneous out-of-body experience induced by the RHI in R.M. provides rich descriptions of a context in which flexible and anomalous body ownership and representation could foster passivity delusions. These findings expand our etiological understanding of the most pathognomonic symptoms of schizophrenia and point towards innovative behavioral interventions. Discussion The RHI was stronger in individuals with schizophrenia than in healthy participants. On a self-report questionnaire, patients reported this illusion as more vivid than healthy controls, both in synchronous and asynchronous stimulation conditions. Further, self-reported vividness of the illusion was associated with increased positive and negative schizotypy in HC. In patients, vividness of the illusion, during synchronous stimulation only, was associated with elevated hallucinations, consistent with previous findings [9] as well as delusions of reference, and delusions of control, and counter intuitively, reduced report of somatic delusions. Addition- ally, SZ experienced greater proprioceptive drift during synchro- nous, but not asynchronous, stimulation. Finally, we observed stimulation-dependent changes in skin temperature. During right hand stimulation, temperature dropped in the stimulated hand and rose in the unstimulated left hand. This effect did not differ across groups or synchrony conditions. Together, these findings are consistent with a weaker or more flexible sense of body ownership in schizophrenia. PLoS ONE \| www.plosone.org PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 7 Body Ownership in Schizophrenia PLoS ONE \| www.plosone.org References 25. Ehrsson HH, Holmes NP, Passingham RE (2005) Touching a rubber hand: feeling of body ownership is associated with activity in multisensory brain areas. J Neurosci 25: 10564–10573. 1. Bleuler E (1911) Dementia Praecox or the Group of Schizophrenias. Zinkin J, ed. translator New York, NY: International Universities Press. 2. Sass LA, Parnas J (2003) Schizophrenia, consciousness, and the self. Schizophr Bull 29: 427–444. 26. Kammers MP, Verhagen L, Dijkerman HC, Hogendoorn H, De Vignemont F, et al. (2009) Is this hand for real? Attenuation of the rubber hand illusion by transcranial magnetic stimulation over the inferior parietal lobule. J Cogn Neurosci 21: 1311–1320. 3. Pollack WS (1989) Schizophrenia and the self: contributions of psychoanalytic self-psychology. Schizophr Bull 15: 311–322. 4. Tsakiris M, Hesse MD, Boy C, Haggard P, Fink GR (2007) Neural signatures of body ownership: a sensory network for bodily self-consciousness. Cereb Cortex 17: 2235–2244. 27. Sokolov EN (1963) Higher nervous functions; the orienting reflex. Annu Rev Physiol 25: 545–580. 28. Blanke O, Mohr C (2005) Out-of-body experience, heautoscopy, and autoscopic hallucination of neurological origin Implications for neurocognitive mechanisms of corporeal awareness and self-consciousness. Brain research Brain research reviews 50: 184–199. 5. Gallagher II (2000) Philosophical conceptions of the self: implications for cognitive science. Trends Cogn Sci 4: 14–21. 6. Botvinick M, Cohen J (1998) Rubber hands ‘feel’ touch that eyes see. Nature 391: 756. 29. Vallar G, Ronchi R (2009) Somatoparaphrenia: a body delusion. A review of the neuropsychological literature. Exp Brain Res 192: 533–551. 7. Tsakiris M, Haggard P (2005) The rubber hand illusion revisited: visuotactile integration and self-attribution. J Exp Psychol Hum Percept Perform 31: 80–91. 30. Tsakiris M, Carpenter L, James D, Fotopoulou A (2010) Hands only illusion: multisensory integration elicits sense of ownership for body parts but not for non- corporeal objects. Exp Brain Res 204: 343–352. 8. Moseley GL, Olthof N, Venema A, Don S, Wijers M, et al. (2008) Psychologically induced cooling of a specific body part caused by the illusory ownership of an artificial counterpart. Proc Natl Acad Sci U S A 105: 13169–13173. p j p 31. Chang BP, Lenzenweger MF (2005) Somatosensory processing and schizophre- nia liability: proprioception, exteroceptive sensitivity, and graphesthesia performance in the biological relatives of schizophrenia patients. J Abnorm Psychol 114: 85–95. 9. Peled A, Pressman A, Geva AB, Modai I (2003) Somatosensory evoked potentials during a rubber-hand illusion in schizophrenia. Schizophr Res 64: 157–163. References 32. Williams LE, Ramachandran VS, Hubbard EM, Braff DL, Light GA (2010) Superior size-weight illusion performance in patients with schizophrenia: evidence for deficits in forward models. Schizophr Res 121: 101–106. 10. Raine A, Lencz T, Mednick SA (1995) Schizotypal Personality NY: Cambridge University Press. 11. Overall JE, Gorham DR (1962) The Brief Psychiatric Rating Scale. Psychological Report 10: 799–812. p 33. de Gelder B, Vroomen J, Annen L, Masthof E, Hodiamont P (2003) Audio- visual integration in schizophrenia. Schizophr Res 59: 211–218. visual integration in schizophrenia. Schizophr Res 59: 211–218. 12. Andreasen NC (1984) The Scale for the Assessment of Positive Symptoms (SAPS). Iowa CityIA: The University of Iowa. 34. Tsakiris M, Costantini M, Haggard P (2008) The role of the right temporo- parietal junction in maintaining a coherent sense of one’s body. Neuropsycho- logia 46: 3014–3018. 13. Andreasen NC (1983) The Scale for the Assessment of Negative Symptoms (SANS). Iowa City, IA: The University of Iowa. g 35. Torrey EF (2007) Schizophrenia and the inferior parietal lobule. Schizophr Res 97: 215–225. ( ) y y 14. Raine A (1991) The SPQ: a scale for the assessment of schizotypal personality based on DSM-III-R criteria. Schizophr Bull 17: 555–564. 36. Arzy S, Mohr C, Michel CM, Blanke O (2007) Duration and not strength of activation in temporo-parietal cortex positively correlates with schizotypy. NeuroImage 35: 326–333. 15. Nelson HE (1982) National Adult Reading Test (NART) Test Manual. Berkshire: Windsor. g 37. Olney JW, Newcomer JW, Farber NB (1999) NMDA receptor hypofunction model of schizophrenia. J Psychiatr Res 33: 523–533. 16. Oldfield RC (1971) The assessment and analysis of handedness: the Edinburgh inventory. Neuropyschologia 9: 97–113. 38. Morgan HL, Turner DC, Corlett PR, Absalom AR, Adapa R, et al. (2011) Exploring the impact of ketamine on the experience of illusory body ownership. Biol Psychiatry 69: 35–41. 17. Longo MR, Schuur F, Kammers MP, Tsakiris M, Haggard P (2009) Self awareness and the body image. Acta Psychol (Amst) 132: 166–172. awareness and the body image. Acta Psychol (Amst) 132: 166–172. 18. Blanke O, Landis T, Spinelli L, Seeck M (2004) Out-of-body experience an autoscopy of neurological origin. Brain 127: 243–258. 39. Wilkins LK, Girard TA, Cheyne JA (2011) Ketamine as a primary predictor of out-of-body experiences associated with multiple substance use. Consciousness and cognition 20: 943–950. autoscopy of neurological origin. Brain 127: 243–258. 19. Author Contributions Conceived and designed the experiments: KNT SP. Performed the experiments: KNT HSN LGM. Analyzed the data: KNT SP. Contributed reagents/materials/analysis tools: KNT SP HSN LGM. Wrote the paper: KNT SP HSN LGM. PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 October 2011 \| Volume 6 \| Issue 10 \| e27089 8 Body Ownership in Schizophrenia Body Ownership in Schizophrenia References Randolph C, Tierney MC, Mohr E, Chase TN (1998) The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): preliminary clinical validity. Journal of clinical and experimental neuropsychology 20: 310–319. 40. Duncan GE, Leipzig JN, Mailman RB, Lieberman JA (1998) Differential effects of clozapine and haloperidol on ketamine-induced brain metabolic activation. Brain Res 812: 65–75. 20. Peled A, Ritsner M, Hirschmann S, Geva AB, Modai I (2000) Touch feel illusion in schizophrenic patients. Biol Psychiatry 48: 1105–1108. 41. Malhotra AK, Adler CM, Kennison SD, Elman I, Pickar D, et al. (1997) Clozapine blunts N-methyl-D-aspartate antagonist-induced psychosis: a study with ketamine. Biol Psychiatry 42: 664–668. 21. Foucher JR, Lacambre M, Pham BT, Giersch A, Elliott MA (2007) Low time resolution in schizophrenia Lengthened windows of simultaneity for visual, auditory and bimodal stimuli. Schizophr Res 97: 118–127. y y 42. Behere RV, Arasappa R, Jagannathan A, Varambally S, Venkatasubramanian G, et al. (2010) Effect of yoga therapy on facial emotion recognition deficits, symptoms and functioning in patients with schizophrenia. Acta Psychiatr Scand. y p 22. Longo MR, Schuur F, Kammers MP, Tsakiris M, Haggard P (2008) What is embodiment? A psychometric approach. Cognition 107: 978–998. 22. Longo MR, Schuur F, Kammers MP, Tsakiris M, Hagga embodiment? A psychometric approach. Cognition 107: 978–99 23. Rohde M, Di Luca M, Ernst MO (2011) The Rubber Hand Illusion: feeling of ownership and proprioceptive drift do not go hand in hand. PloS one 6: e21659. 43. Duraiswamy G, Thirthalli J, Nagendra HR, Gangadhar BN (2007) Yoga therapy as an add-on treatment in the management of patients with schizophrenia–a randomized controlled trial. Acta Psychiatr Scand 116: 226–232. 24. Ehrsson HH, Spence C, Passingham RE (2004) That’s my hand! Activity in premotor cortex reflects feeling of ownership of a limb. Science 305: 875–877. PLoS ONE \| www.plosone.org October 2011 \| Volume 6 \| Issue 10 \| e27089 9
https://openalex.org/W3095093340	https://cot.unhas.ac.id/journals/index.php/epiije/article/download/762/614	English	null	Design and Manufacture of 6 Axis Force and Moments Transducers for Seaplane Floaters Test in Towing Tank	EPI international journal of engineering	2,020	cc-by-sa	3,816	1. Introduction Seaplanes will run at high speed from start to 60 knots when taking off, especially on the water so that the hydrodynamic forces will work due to the friction of the floater with water. Because of the hydrodynamic forces so it will be important to know the amount of resistance and the amount of lift force of the floater and the moments that work that will facilitate the control of the seaplane. With the existence of these forces, in addition to simulations and aerodynamic tests [3], the simulation and testing of the floater model to determine the hydrodynamic forces are very important to do. One of the proper transportation models developed in Indonesia as an archipelago is a transportation that can be used both on land, air, and water such as seaplanes. Seaplanes are airplanes that can air from land or from the surface of the water and can return to land or to the surface of the water. One type of seaplane is a floatplane or seaplane where the aircraft uses additional components in the form of a floater to float while the fuselage is above the surface of the water. This transportation system will be very flexible because it can cover the land (airstrips), large lakes and rivers, bays, seas, and waters among islands that have calm seas, which allow for takeoff and landing for amphibious aircraft throughout the year. This transportation is much easier and cheaper than building airports in general. In testing the seaplane floater model in the Towing tank, a measuring instrument should capture the hydrodynamic forces acting on the floater. These forces include resistance (Fx), side forces (Fy), lift (Fz), as well as moments in the three-axis directions. A method of measuring the force of 6 degrees of freedom (6 Degree of Freedom, 6 DOF) requires a special design to obtain an optimal value following their needs [4]. For this reason, it is necessary to design a transducer by utilizing six S type load cells which are arranged in such a way as to form six-component transducers. With six-component transducers, it is expected that the forces of Fx, Fy, Fz, and the moments of Mx, My, In the design and development of seaplanes, they should not be careless because they are related to safety so an analysis of the forces that work both on the fuselage and the floater is needed. Abstract To develop seaplanes as a means of inter-island transportation, it is necessary to have a simulation, testing, and analysis of force measurements that work so that the aircraft can be designed optimally in terms of function and safety. To fulfill one type of test, the seaplane floater model is pulled in the Towing Tank to determine the hydrodynamic forces acting on the floater which include resistance (Fx), side force (Fy), lift force (Fz), and moments in all three axes. A method of measuring the force of 6 axis force and moment or 6 degrees of freedom (6 Degree of Freedom, 6 DOF) was built and designed by combining several single load cells so that these forces can be known optimally. From the results of the 6 DOF transducer design, it is proven that it can be used well in measuring 6 forces and moments with force measurement errors ranging from 1.38%. The distance between the 6 DOF transducer capture point and the floater force capture point will affect the measured moment transformation. Keywords: 6 degree of freedom; floater; towing tank; transducer design pISSN 2615-5109 eISSN 2621-0541 pISSN 2615-5109 eISSN 2621-0541 EPI International Journal of Engineering Volume 3, Number 1, February 2020, pp. 84-89 DOI: 10.25042/epi-ije.022020.12 pISSN 2615-5109 eISSN 2621-0541 EPI International Journal of Engineering Volume 3, Number 1, February 2020, pp. 84-89 DOI: 10.25042/epi-ije.022020.12 * Corresponding author. Tel.: +62811328567 Jl. Hidrodinamika, Kompleks ITS, Sukolilo, Surabaya, 60111 Nurhadia,, Mochammad Nasirb, Chandra Permanac, Endah Suwarnid Nurhadia,, Mochammad Nasirb, Chandra Permanac, Endah Suwarnid aLaboratory for Hydrodynamics Technology, BPPT. Email: nrhadi@gmail.com bLaboratory for Hydrodynamics Technology, BPPT. Email: nasir08.lhi@gmail.com cLaboratory for Hydrodynamics Technology, BPPT. Email: chandra.pzm@gmail.com dLaboratory for Hydrodynamics Technology, BPPT. Email: endahsuwarni@gmail.com 1. Introduction To know the forces, of course, it can be done using simulation [1] and through measurement analysis using force sensors [2]. 84 EPI International Journal of Engineering, Vol. 3 No. 1, Feb 2020, pp. 84-89 R24 increases to R + R R13 increases to R + R R14 is reduced to R - R R23 is reduced to R - R So that 𝑖24 = 𝐸 𝑅24+𝑅14 = 𝐸 𝑅+𝜕𝑅+𝑅−𝜕𝑅= 𝐸 2𝑅 (3) 𝑉24 = 𝑖24.𝑅24 = 𝐸 2𝑅(𝑅+ 𝜕𝑅) (4) 𝑉= 𝑉23 −𝑉24 = 𝐸𝑅−𝐸𝜕𝑅 2𝑅 − 𝐸𝑅+𝐸𝜕𝑅 2𝑅 (5) 𝑉= ± 𝐸𝜕𝑅 𝑅 (6) Mz that occur on the floater when pulled will be known optimally. 𝑆𝑡𝑟𝑎𝑖𝑛(𝜀) = 𝑐ℎ𝑎𝑛𝑔𝑒 𝑜𝑓 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑡ℎ𝑒 𝑙𝑜𝑎𝑑 (𝜕𝑙) 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ (𝑙) (2) (2) Based on the principle of changes in strain values due to forces, by using 4 strain gages which are connected in the full-bridge, the value of a strain that occurs due to changes in force acting can be measured. 2.1. Strain and sensors applications So that So that that 𝑖24 = 𝐸 𝑅24+𝑅14 = 𝐸 𝑅+𝜕𝑅+𝑅−𝜕𝑅= 𝐸 2𝑅 (3) 𝑉24 = 𝑖24.𝑅24 = 𝐸 2𝑅(𝑅+ 𝜕𝑅) (4) 𝑉= 𝑉23 −𝑉24 = 𝐸𝑅−𝐸𝜕𝑅 2𝑅 − 𝐸𝑅+𝐸𝜕𝑅 2𝑅 (5) 𝑉= ± 𝐸𝜕𝑅 𝑅 (6) A strain sensor is a passive transducer that uses the concept of changing the electrical resistance in a metal wire to change every pull produced by the force that hits the wire. The concept of this change can be used to measure weight, pressure, and mechanical force. Strain sensors are made of one long wire, folded, and placed on a soft or flexible material [5]. Any force, strain, applied to the area where the strain gauges are attached will cause a change in the length and cross-sectional area of the wire as shown in Fig. 1. This change will result in a change in the resistance of the wire itself through the connection: (6) By comparing the output voltage generated due to the forces acting on the load cell will be obtained by a factor as a basis for calculating the forces that occur in the test model. 𝑅= 𝜌.𝑙 𝐴 (1) (1) R = wire resistance (Ω)  = wire resistance coefficient (Ω -m) l = wire length (m) A = wire cross-sectional area (m2) R = wire resistance (Ω)  = wire resistance coefficient (Ω -m) l = wire length (m) A = wire cross-sectional area (m2) 2.2. The forces acting on the floater test model The forces acting on the floater are the same as the forces acting on ships that moving on the surface of the water at a certain speed. As a result of water shifting with the floater body, hydrodynamic forces include the drag force (Fx), side force (Fy), lift (Fz), and moments in the x, y, and z directions to the point catch floater. Illustration of these styles can be seen in Fig. 3. The relationship between strain and force (load) can be calculated by the formula: 𝑆𝑡𝑟𝑎𝑖𝑛(𝜀) = 𝑐ℎ𝑎𝑛𝑔𝑒 𝑜𝑓 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑡ℎ𝑒 𝑙𝑜𝑎𝑑 (𝜕𝑙) 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ (𝑙) (2) 𝑆𝑡𝑟𝑎𝑖𝑛(𝜀) = 𝑐ℎ𝑎𝑛𝑔𝑒 𝑜𝑓 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑡ℎ𝑒 𝑙𝑜𝑎𝑑 (𝜕𝑙) 𝑤𝑖𝑟𝑒 𝑙𝑒𝑛𝑔𝑡ℎ (𝑙) (2) 4.1. 6 DOF sensor system design For the 6 DOF load cell design, 6 S type load cells are needed. S type load cells can be made from aluminum alloy material as shown in Fig. 6. For dimensions adjusted to the design of the required load cell capacity. In the calculation of dimensions can be used software ‘Transducers’ made by Marin. Based on the results of the software for 100 N capacity obtained dimensions as in Table 1, the spring element dimensions for load cell type S with a capacity of 100 N, so that if desired can be made according to the required capacity. Figure 5 shows how the measurement method to be used where 6 channels of 6 DOF sensors will be connected to signal conditioning which functions as a supply of voltage, signal output amplifier, and filter. The output of the signal condition is connected to the NI Data Acquisition which is connected to a PC for recording data and processing the output data from the 6 DOF sensor. Figure 4. Methodology Figure 5. The Measurement method Figure 6. S Type of load cell (shear) Figure 6. S Type of load cell (shear) Table 1. The Output of transducers software for 100 N [7] Table 1. The Output of transducers software for 100 N [7] MATERIAL MEASURING RANGE ALUMINUM SENSITIVITY ELECTRICAL OUTPUT FX 8.64 uR/N FX 116 N FX 17 uV/V/N FY 8.74 uR/N FY 114 N FY 17 uV/V/N FZ 1.3 uR/N FZ 797 N FZ 2 uV/V/N MX 1747 uR/NM MX 0.57 NM MX 3511 uV/V/NM MY 343 uR/NM MY 2.91 NM MY 690 uV/V/NM MZ 6217 uR/NM MZ 0.16 NM MZ 12496 uV/V/NM MATERIAL MICRO STRAIN FULL BRIDGE STRAIN GAGE (Z-AXIS IS LENGTH AXIS) INPUT DATA MILLIMETERS HEIGHT IN X-DIRECTION A= 10.00 MM HEIGHT IN Y-DIRECTION B= 5.00 MM SLOT LENGTH C= 10.00 MM THICKNESS SHEAR PLATE F= 2.00 MM SLOT HEIGHT H= 8.00 MM The design of manufacture is based on the numerical data of the force to be measured so we need a load cell with a capacity of 2 pieces of 50 N and 4 pieces of 100 N. The capacity of 50 N is used as an X-axis load cell so that the maximum Fx force can be measured up to 100 N, Fy 100 N is being FZ to 300 N. 3. Methodology To solve the problems faced, a method was made to facilitate the design of a measuring instrument 6 degrees of freedom of force and moment. In brief, this method can be seen in the flow chart in Fig. 4. In Fig. 2, if the load cell has a force F, there will be a strain on the R-value of the strain gages mounted with the change in the value of R is: There are three main stages in the design of buildings measuring 6 DOF. The first is the design phase where at this stage it is necessary to know the range of force and moment measurements that will work on the measuring instrument, the determination, and design of a single load cell, the calculation of the weight of the material used. The second stage is the manufacturing stage of design. At this stage, a material for the spring load cell and a framework for the 6 DOF sensor is made. Single load cell calibration is also Figure 1. Principle of strain gages sensor [5] Figure 2. Strain sensor application with the full-bridge relationship [5] E V F Figure 1. Principle of strain gages sensor [5] Figure 1. Principle of strain gages sensor [5] Figure 3. The styles in the floater test model [6] Figure 3. The styles in the floater test model [6] Figure 2. Strain sensor application with the full-bridge relationship [5] E V F F E F Figure 2. Strain sensor application with the full-bridge relationship [5] Figure 3. The styles in the floater test model [6] 85 EPI International Journal of Engineering, Vol. 3 No. 1, Feb 2020, pp. 84-89 performed to determine the value of the single load cell calibration and repeatability performance. The last step was assembling the prototype 6 DOF sensor and testing the load force Fx, Fy, Fz to determine the accuracy of the signal and the measurement uncertainty that occurred. [6] 4.1. 6 DOF sensor system design Because of the time and capacity considerations of 50 N and 100 N are widely available in the market, for the S type load cell, the manufacturer's Load cell "Dacell", Korean, with the UU model [8]. Figure 4. Methodology Figure 5. The Measurement method Figure 4. Methodology The calibration process is carried out by individually calibrating both the drag and press of each type S load cell used. Calibration is carried out by giving a load in stages, as shown in Fig. 7, the resulting output is measured so that the calibration factor is known that shows the relationship between the load or force (kg) and the output voltage (volts). Figure 5. The Measurement method 86 EPI International Journal of Engineering, Vol. 3 No. 1, Feb 2020, pp. 84-89 Figure 8. Calibration output of S1 load cell Figure 7. Pull and push calibration of single load cell measurement results of S1 - S6 load cell, then the magnitude of force and moment can be determined by the formula: Figure 7. Pull and push calibration of single load cell Figure 7. Pull and push calibration of single load cell 𝐹𝑥= 𝑆1 + 𝑆2 (7) 𝐹𝑦= 𝑆3 (8) 𝐹𝑧= 𝑆4 + 𝑆5 + 𝑆6 (9) 𝑀𝑥= 1 2 𝑐(𝑆6 −𝑆5) (10) 𝑀𝑦= 1 2 𝑏(𝑆4 −𝑆5 −𝑆6) (11) 𝑀𝑧= 1 2 𝑎(𝑆1 + 𝑆2) (12) 𝐹𝑥= 𝑆1 + 𝑆2 (7) 𝐹𝑦= 𝑆3 (8) 𝐹𝑧= 𝑆4 + 𝑆5 + 𝑆6 (9) 𝑀𝑥= 1 2 𝑐(𝑆6 −𝑆5) (10) 𝑀𝑦= 1 2 𝑏(𝑆4 −𝑆5 −𝑆6) (11) 𝑀𝑧= 1 2 𝑎(𝑆1 + 𝑆2) (12) Figure 7. Pull and push calibration of single load cell Figure 7. Pull and push calibration of single load cell Figure 8. Calibration output of S1 load cell Figure 7. Pull and push calibration of single load cell (12) Figure 8. Calibration output of S1 load cell 4.2. Design measurement model test To be able to know the magnitudes of the forces and moments caused by the hydrodynamic forces on the amphibious plane floater, the measurement design is made as shown in Fig. 10. In Fig. 10 it is shown that the transformation of the measured force and moment from the 6 DOF sensor axis to the specified measuring point for example, from A to B. In Fig. 10, the magnitude of the measured force and moment is at point A which is so that it will be given the notation Fxa, Fya, Fza, and Mxa, Mya, Mza. The results of the measured quantities need to be transformed to the desired center point in this case at point B. If the distance between points A and B is equal to m, the formula for force and moment transformation is obtained as follows: Figure 8. Calibration output of S1 load cell An example of a calibration graph output can be seen in Fig. 8. The x-axis shows the input load (kg) and the y-axis is the resulting output (volts) so the trend line graph is known as a multiplier factor for the output signal load cell amplifier that read by the data acquisition system. The mechanical configuration of the 6 DOF sensor system and the design of the 6 DOF sensor frame can be seen in Fig. 9. In Fig. 9 (a). There are 6 load cells arranged in such a way that a 6 DOF (6 degree of freedom) sensor is formed. On the 6 DOF sensor, there are six load cells S1 - S6 which are used to measure the magnitude of the force Fz, Fy, Fz, and moments Mx, My, Mz. Figure 9 (b) shows the 6 DOF sensor circuit in 3 dimensions, how the placement of load cells S1 - S6 refers to Fig. 9 (a). From the series of Fig. 9 (a), it is known that to measure the Fx force 2 load cells are used, Fy 1 load cell and Fz 3 load cells are used. From the 𝐹𝑥𝑏= 𝐹𝑥𝑎 (13) 𝐹𝑦𝑏= 𝐹𝑦𝑎 (14) 𝐹𝑧𝑏= 𝐹𝑧𝑎 (15) 𝑀𝑥𝑏= 𝑀𝑥𝑎+ 𝐹𝑦𝑎. 𝑚 (16) 𝑀𝑦𝑏= 𝑀𝑦𝑎+ 𝐹𝑥𝑎. 𝑚 (17) 𝑀𝑧𝑏= 𝑀𝑧𝑎 (18) 𝐹𝑥𝑏= 𝐹𝑥𝑎 (13) 𝐹𝑦𝑏= 𝐹𝑦𝑎 (14) 𝐹𝑧𝑏= 𝐹𝑧𝑎 (15) 𝑀𝑥𝑏= 𝑀𝑥𝑎+ 𝐹𝑦𝑎. 𝑚 (16) 𝑀𝑦𝑏= 𝑀𝑦𝑎+ 𝐹𝑥𝑎. 𝑚 (17) 𝑀𝑧𝑏= 𝑀𝑧𝑎 (18) 𝐹𝑥𝑏= 𝐹𝑥𝑎 (13) 𝐹𝑦𝑏= 𝐹𝑦𝑎 (14) 𝐹𝑧𝑏= 𝐹𝑧𝑎 (15) 𝑀𝑥𝑏= 𝑀𝑥𝑎+ 𝐹𝑦𝑎. 4.3. Test and measurement results Figure 12. Example force and moment result from the output testing with 6 DOF V uji (m/s) Time (s) 0 50 100 02468 10 Fx (kg) Time (s) 0 50 100 -200 0 200 400 600 Fy (kg) Time (s) 0 50 100 -100 0 100 Fz (kg) Time (s) 0 50 100 -500 0 Mx (kgm) Time (s) 0 50 100 -1000 -500 0 500 My (kgm) Time (s) 0 50 100 -1000 0 1000 2000 3000 Mz (kgm) Time (s) 0 50 100 0 50 100 The model testing was carried out on the floater model which was connected with the 6 DOF measuring device as the measurement design of Figure 10, while the shape of the aircraft was not modeled because the fuselage was not immersed in water. The floater model is made on a certain scale by considering the maximum speed of the tug train in the Towing Tank. In the floater test, the model is pulled with a certain speed, and with the measuring instrument, 6 DOF measured the forces that work so that the influence of hydrodynamic forces on the floater model is known. Before testing, the measurement is carried out by weighing the measuring instrument 6 DOF of 1300 gr towards the Z-axis (Fz). The measurement results can be seen in Fig. 11. From Fig. 11 it can be seen that the dominant values are found in the outputs of S4, S5, and S6 as constituents of the FZ style [9]. By adding the three in which S4 (345 gr), S5 (387 gr), and S6 (550 gr) the average measured force is 1,282 gr, so there is a difference of 18/1300 or 1.38%. Before testing, the measurement is carried out by weighing the measuring instrument 6 DOF of 1300 gr towards the Z-axis (Fz). The measurement results can be seen in Fig. 11. From Fig. 11 it can be seen that the dominant values are found in the outputs of S4, S5, and S6 as constituents of the FZ style [9]. By adding the three in which S4 (345 gr), S5 (387 gr), and S6 (550 gr) the average measured force is 1,282 gr, so there is a difference of 18/1300 or 1.38%. Figure 12. Example force and moment result from the output testing with 6 DOF 5. Conclusion From the design of the 6 DOF transducer for the measurement of the hydrodynamic forces of the seaplane floater, it can be concluded that: Model testing is done after testing the DOF 6 gauge and testing the road to find out the performance of the model and the measuring instrument, the floater test model is pulled in several speed points with variations of the draft on the model.  6 DOF transducer can be made from 6 S type load cells which are arranged in such a way that the direction x, y, and z, so that it can be used to measure the force Fx, Fy, Fz, Mx, My and Mz. In Fig. 12, the output of one of the seaplane floater models is shown. In the picture, there are 7 channels wherein channel 1 shows the test speed of the floater model, channel 2 (Fx), channel 3 (Fy), channel 4 (Fz), channel 5 (Mx), channel 6 (My), and channel 7 (Mz). These values are obtained by processing the output signal according to the formula [13-18].  Load cell calibration and calibration factor values are very important to know correctly because this will affect the accuracy of the measurement results. In testing the measurement of the Fz force, an error of 1.38% is obtained so that the sensor is proper to use.  In testing the hydrodynamic forces using a 6 DOF transducer it is necessary to transform the force and moment produced by the sensor at the desired point, so the distance between that point and the center of the 6 DOF sensor needs to be known. On the test graph in Fig. 12, when the condition of the test speed has stabilized on channel 1 (V) it is known that the average test speed is 9.19 m/s. For the magnitude of the force and moment, there are fluctuations in the measurement results that can be caused by the vibration of the floater when pulled. From the Fig. 12, we get the average values of Fx (336.21 kg), Fy (6.40 kg), Fz (583.43 kg), Mx (200.64 kg meter), My (1013.85 kg meter) and Mz (78.12 kg meter) where this value has been multiplied by the scale factor of the model test. 4.2. Design measurement model test 𝑚 (16) 𝑀𝑦𝑏= 𝑀𝑦𝑎+ 𝐹𝑥𝑎. 𝑚 (17) 𝑀𝑧𝑏= 𝑀𝑧𝑎 (18) Figure 10. Transformation from point A to B Figure 10. Transformation from point A to B Figure 9. (a) Configuration of 6 DOF (b) Frame design of 6 DOF Figure 10. Transformation from point A to B Figure 9. (a) Configuration of 6 DOF (b) Frame design of 6 DOF 87 EPI International Journal of Engineering, Vol. 3 No. 1, Feb 2020, pp. 84-89 5. Conclusion  To get high accuracy, the capacity of the type S load cell must be adjusted according to the needs according to the prediction of the magnitude of the force to be measured. [5] J. Vaughan, Application of B&K Equipment to Strain Measurement. Copenhagen: Bruel & Kjaer, 1975. [8] “Load Cell.” [Online]. Available: http://www.dacell.com/en/load cell/show/view/cno/10/pno/40/page/3/id/748. [Accessed: 22-Nov- 2019]. [6] K. Ito, T. Daenen, Y. Hirakawax, T. Hirayama, and T. Sakurai, “Longitudinal Stability Augmentation of Seaplanes in Planing,” J. Aircr., vol. 53, no. 3, pp. 1332–1342, 2016. [7] “Marine Transducers Software,” 2008. EPI International Journal of Engineering, Vol. 3 No. 1, Feb 2020, pp. 84-89 [7] “Marine Transducers Software,” 2008. References Figure 11. Load test of Fz at 6 DOF transducers S1 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 S2 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 S3 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 S4 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 S5 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 S6 (kg) Time (s) 0 2 4 6 -0.8 -0.4 0 [1] M. M. Lubis, “Analisis Aerodinamika Airfoil Naca 2412 Pada Sayap Pesawat Model Tipe Glider Dengan Menggunakan Software Berbasis Computional Fluid Dinamic Untuk Memperoleh Gaya Angkat Maksimum,” J. e-Dinamis, vol. 2, no. 2, pp. 23–33, 2012. [2] J. Ma and A. Song, “Fast Estimation of Strains for Cross-Beams Six- Axis Force/Torque Sensors by Mechanical Modeling,” Sensors Actuators A. Phys, vol. 13, pp. 6669–6686, 2013. [3] D. F. Rendra, W. B. Setiawan, and R. Mareta, “Aplikasi Pengukur 3 Derajat Kebebasan Gaya Aerodinamika Pada Water Tunnel,” J. Sains dan Teknol., vol. 6, no. 2, pp. 258–267, 2017. [4] S. A. Liu and H. L. Tzo, “A Novel Six-component Force Sensor of Good Measurement Isotropy and Sensitivities,” Sensors Actuators A 100, pp. 223–230, 2002. [5] J. Vaughan, Application of B&K Equipment to Strain Measurement. Copenhagen: Bruel & Kjaer, 1975. Figure 11. Load test of Fz at 6 DOF transducers 88 89
https://openalex.org/W2000252897	https://hal.science/hal-01389907/document	English	null	A numerical analysis of chaos in the double pendulum	Chaos, solitons & fractals/Chaos, solitons and fractals	2,006	cc-by	2,816	To cite this version: Tomasz Stachowiak, Toshio Okada. A numerical analysis of chaos in the double pendulum. Chaos, Solitons & Fractals, 2006, 29 (2), pp.417 - 422. 10.1016/j.chaos.2005.08.032. hal-01389907 Distributed under a Creative Commons Attribution 4.0 International License 1. Introduction The planar double pendulum, as shown in Fig. 1, is one of the mechanical systems that exhibits chaotic behaviour, despite their simple construction. As it is a Hamiltonian system, its energy is a conserved quantity determining the glo bal properties of motion at the two limits of zero and inﬁnite energy, the systems motion is regular. (And in the inﬁ nite limit, the total angular momentum might be considered as the second conserved quantity.) It is the interval between these limits that presents typical chaotic features. Although there has been no rigorous proof of the system being non integrable, the analysis of chaos has given many results indicating such a possibility (see [1] for further references). In this paper, we concentrate on three chaotic aspects. In Section 2 we set the notation and derive the equations used, together with the parameter values chosen for simulations. The subsequent sections contain a presentation of Poincare´ sections given for increasing values of energy, bifurcation diagrams for regular, quasi regular and chaotic regions, and ﬁnally, calculation of the largest Lyapunov characteristic exponents. A numerical analysis of chaos in the double pendulum Tomasz Stachowiaka, Toshio Okadab a Astronomical Observatory, Jagiellonian University, Orla 171, 30 244 Krako´w, Poland b Tokyo University of Agriculture and Technology, Naka machi, Koganei shi, Tokyo 184 8588, Japan a Astronomical Observatory, Jagiellonian University, Orla 171, 30 244 Krako´w, Poland b Tokyo University of Agriculture and Technology, Naka machi, Koganei shi, Tokyo 184 8588, Japan We analyse the double pendulum system numerically, using a modiﬁed mid point integrator. Poincare´ sections and bifurcation diagrams are constructed for certain, characteristic values of energy. The largest Lyapunov characteristic exponents are also calculated. All three methods conﬁrm the passing of the system from the regular low energy limit into chaos as energy is increased. HAL Id: hal-01389907 https://hal.science/hal-01389907v1 Submitted on 30 Oct 2016 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. Distributed under a Creative Commons Attribution 4.0 International License 2. Main equations he notation used in Fig. 1, we can immediately obtain the Lagrange function for the system: With the notation used in Fig. 1, we can immediately obtain the Lagrange function for the system: L ¼ 1 2 l2 1ðm1 þ m2Þ _u2 1 þ l2 2m2 _u2 2 þ 2l1l2m2 cosðu1 u2Þ _u1 _u2 þ gl1ðm1 þ m2Þ cos u1 þ gl2m2 cos u2; L ¼ 1 2 l2 1ðm1 þ m2Þ _u2 1 þ l2 2m2 _u2 2 þ 2l1l2m2 cosðu1 u2Þ _u1 _u2 þ gl1ðm1 þ m2Þ cos u1 þ gl2m2 cos u2; ð1Þ 1 1 ϕ1 ϕ2 l1 l2 m1 m2 Fig. 1. The physical setup of the system. Fig. 1. The physical setup of the system. where the dot denotes the derivative with respect to time. We choose not to employ the Hamiltons formalism, and use as two additional variables the angular velocities instead of momenta. Thus, the second order equations of motion: where the dot denotes the derivative with respect to time. We choose not to employ the Hamiltons formalism, and use as two additional variables the angular velocities instead of momenta. 2. Main equations Thus, the second order equations of motion: l1ðm1 þ m2Þ€u1 þ l2m2 cosðu1 u2Þ€u2 þ sinðu1 u2Þ _u2 2 þ gðm1 þ m2Þ sin u1 ¼ 0; l1ðm1 þ m2Þ€u1 þ l2m2 cosðu1 u2Þ€u2 þ sinðu1 u2Þ _u2 2 þ gðm1 þ m2Þ sin u1 ¼ 0; l1ðm1 þ m2Þ€u1 þ l2m2 cosðu1 u2Þ€u2 þ sinðu1 u2Þ _u2 2 þ gðm1 þ m2Þ sin u1 ¼ 0; l2€u2 þ l1 cosðu1 u2Þ€u1 sinðu1 u2Þ _u2 1 þ g sin u2 ¼ 0 ð2Þ l1ðm1 þ m2Þ€u1 þ l2m2 cosðu1 u2Þ€u2 þ sinðu1 u2Þ _u2 2 þ gðm1 þ m2Þ sin u1 ¼ 0; ð2Þ l1ðm1 þ m2Þu1 þ l2m2 cosðu1 u2Þu2 þ sinðu1 u2Þu2 þ gðm1 þ m2Þ sin u1 ¼ 0; l2€u2 þ l1 cosðu1 u2Þ€u1 sinðu1 u2Þ _u2 1 þ g sin u2 ¼ 0 ð2Þ are reduced to a system of four equations of the ﬁrst order: ð2Þ l2€u2 þ l1 cosðu1 u2Þ€u1 sinðu1 u2Þ _u2 1 þ g sin u2 ¼ 0 are reduced to a system of four equations of the ﬁrst order: _u1 ¼ x1; _u2 ¼ x2; _x1 ¼ sinðu1 u2Þ½l1 cosðu1 u2Þx2 1 þ x2 2 2l1½1 þ m1 cos2ðu1 u2Þ ð1 þ 2m1Þ sin u1 þ sinðu1 2u2Þ l1½1 þ m1 cos2ðu1 u2Þ ; _x2 ¼ sinðu1 u2Þ ð1 þ m1Þðcos u1 þ l1x2 1Þ þ cosðu1 u2Þx2 2 1 þ m1 cos2ðu1 u2Þ ; ð3Þ _u1 ¼ x1; _u2 ¼ x2; ð3Þ where m2 = 1 is the unit of mass, l2 = 1 of length and g = 1, which eﬀectively sets a unit of time. For practical appli cation, we choose particular values of the remaining parameters as: m1 = 3 and l1 = 2. where m2 = 1 is the unit of mass, l2 = 1 of length and g = 1, which eﬀectively sets a unit of time. For practical appli cation, we choose particular values of the remaining parameters as: m1 = 3 and l1 = 2. Fig. 2. Poincare´ section for E = E1 = E01 + 0.01; regular behaviour. Fig. 2. Poincare´ section for E = E1 = E01 + 0.01; regular behaviour. 3. Poincare´ sections As the whole phase space is four dimensional, it is convenient to look at its sections, i.e., the values of the coordi nates whenever one of them assumes a certain value. Here the sections were constructed with the surface u1 = 0, and x1 ¼ _u1 > 0, and restricted to the plane of u2 and x2. 1 The ﬁrst one (Fig. 2) at the energy level of E = E1 presents a very regular image, as would be expected from the regular limit E = E01. The situation becomes more complex in Fig. 3, when the energy is increased to E = E2, although without chaotic regions, the invariant tori are visibly deformed. Finally, in Fig. 4 almost all of the regular orbits decay into a global chaotic region. 2. Main equations 2 2 There are four critical (or stationary) points: There are four critical (or stationary) points: ðu1; u2; x1; x2Þ ¼ ð0; 0; 0; 0Þ; E01 ¼ 1 l1ð1 þ m1Þ ¼ 9; ð0; p; 0; 0Þ; E02 ¼ þ1 l1ð1 þ m1Þ ¼ 7; ðp; 0; 0; 0Þ; E03 ¼ 1 þ l1ð1 þ m1Þ ¼ 7; ðp; p; 0; 0Þ; E04 ¼ þ1 þ l1ð1 þ m1Þ ¼ 9; 8 > > > > > < > > > > > : ð4Þ ðu1; u2; x1; x2Þ ¼ ð0; 0; 0; 0Þ; E01 ¼ 1 l1ð1 þ m1Þ ¼ 9; ð0; p; 0; 0Þ; E02 ¼ þ1 l1ð1 þ m1Þ ¼ 7; ðp; 0; 0; 0Þ; E03 ¼ 1 þ l1ð1 þ m1Þ ¼ 7; ðp; p; 0; 0Þ; E04 ¼ þ1 þ l1ð1 þ m1Þ ¼ 9; 8 > > > > > < > > > > > : ð4Þ ð4Þ with E being the corresponding energy values. The phase space sections and Lyapunov exponents have been obtained for three values of energy, close to the above, namely: E1 = E01 + 0.01, E2 = E02 0.01 and E3 = E03 0.01. The numerical analysis was performed using the Bulirsch Stoer method with Richardson extrapolation as described in [2]. with E being the corresponding energy values. The phase space sections and Lyapunov exponents have been obtained for three values of energy, close to the above, namely: E1 = E01 + 0.01, E2 = E02 0.01 and E3 = E03 0.01. The numerical analysis was performed using the Bulirsch Stoer method with Richardson extrapolation as described with E being the corresponding energy values. The phase space sections and Lyapunov exponents have been obtained for three values of energy, close to the above, namely: E1 = E01 + 0.01, E2 = E02 0.01 and E3 = E03 0.01. for three values of energy, close to the above, namely: E1 = E01 + 0.01, E2 = E02 0.01 and E3 = E03 0.01. The numerical analysis was performed using the Bulirsch Stoer method with Richardson extrapolation as described in [2]. The numerical analysis was performed using the Bulirsch Stoer method with Richardson extrapolation as described in [2]. 4. Bifurcation diagrams In order to detect periodic orbits and their number, regular windows in chaos, etc., a bifurcation diagram can be constructed. It depicts the dependence of a suitably chosen quantity reﬂecting the chaotic behaviour on some parameter of the system. As we ﬁxed the values for mass and length, we used energy for that purpose. The other quan tity, which should be diﬀerent for diﬀerent orbits, if they exist, could be the amplitude of one of the variables. Choosing Fig. 3. Poincare´ section for E = E2 = E02 0.01; regular trajectories begin to deform. Fig. 3. Poincare´ section for E = E2 = E02 0.01; regular trajectories begin to deform. 3 Fig. 4. Poincare´ section for E = E3 = E03 0.01; regular regions decay with the appearance of a global, completely chaotic region. Fig. 4. Poincare´ section for E = E3 = E03 0.01; regular regions decay with the appearance of a global, completely chaotic region. Fig. 5. The ﬁrst appearance of global chaos, and ‘‘windows’’ that are magniﬁed in Fig. 6. Fig. 5. The ﬁrst appearance of global chaos, and ‘‘windows’’ that are magniﬁed in Fig. 6. u2 as this variable, we eﬀectively needed to construct Poincare´ sections with respect to _u2 ¼ x2, with _x2 < 0, so that the maximum values of u2 were obtained. Fi 5 h h b h i f h l h ﬁ f l b l h Th i Fig. 5 shows the behaviour from the lowest energy E01, up to the ﬁrst appearance of global chaos. There are regions, shown magniﬁed in Fig. 6, where the orbits maxima tend to cluster at evenly spaced intervals of u2, and ‘‘diverge’’ around such places, densely occupying the whole admissible region. Fig. 5 shows the behaviour from the lowest energy E01, up to the ﬁrst appearance of global chaos. There are regions, shown magniﬁed in Fig. 6, where the orbits maxima tend to cluster at evenly spaced intervals of u2, and ‘‘diverge’’ around such places, densely occupying the whole admissible region. In Fig. 7 contains a window between two completely chaotic regions, showing smaller regions, of behaviour similar to Fig. 6, where there are only four possible values of u2, which then chaotically increase. In Fig. 7 contains a window between two completely chaotic regions, showing smaller regions, of behaviour similar to Fig. 4. Bifurcation diagrams 6, where there are only four possible values of u2, which then chaotically increase. In Fig. 7 contains a window between two completely chaotic regions, showing smaller regions, of behaviour similar to Fig. 6, where there are only four possible values of u2, which then chaotically increase. 4 4 Fig. 6. A ‘‘window’’ showing convergence of admissible values of u2max; this is thus a regular behaviour. Fig. 6. A ‘‘window’’ showing convergence of admissible values of u2max; this is thus a regular behaviour. Fig. 7. A gap between two globally chaotic regions, showing windows, and regular regions with bifurcations quickly leading to chaos. Fig. 7. A gap between two globally chaotic regions, showing windows, and regular regions with bifurcations quickly leading to chaos. 5. Lyapunov exponents We limited our investigations to the largest Lyapunov exponents, as it eﬀectively gives us the information on the divergence of two close trajectories. We used the simple method described, in [3]. The main formula is We limited our investigations to the largest Lyapunov exponents, as it eﬀectively gives us the information on the divergence of two close trajectories. We used the simple method described, in [3]. The main formula is v ¼ 1 tN X N i 1 ln jdiðtÞj jdið0Þj ; ð5Þ v ¼ 1 tN X N i 1 ln jdiðtÞj jdið0Þj ; ð5Þ v ¼ 1 tN X N i 1 ln jdiðtÞj jdið0Þj ; ð5Þ where v denotes the Lyapunov exponents, the index i, consecutive initial positions, and d is the separation between two close trajectories (Fig. 8). where v denotes the Lyapunov exponents, the index i, consecutive initial positions, and d is the separation between two close trajectories (Fig. 8). where v denotes the Lyapunov exponents, the index i, consecutive initial positions, and d is the separation between two close trajectories (Fig. 8). The simulations were performed for the same values of energy as in the case of the Poincare´ sections and the results are in very good agreement. For both E1 and E2, the exponents tend to zero, indicating practically regular behaviour. The simulations were performed for the same values of energy as in the case of the Poincare´ sections and the results are in very good agreement. For both E1 and E2, the exponents tend to zero, indicating practically regular behaviour. 5 Fig. 8. The largest Lyapunov exponents for the system. The regular behaviour of the two ﬁrst cases, E = E1 and E = E2, makes the thin solid line, and the dashed line, respectively, almost coincide. The third line, corresponding to the chaotic case of E3, indicates a positive value of the exponents. Fig. 8. The largest Lyapunov exponents for the system. The regular behaviour of the two ﬁrst cases, E = E1 and E = E2, makes the thin solid line, and the dashed line, respectively, almost coincide. The third line, corresponding to the chaotic case of E3, indicates a positive value of the exponents. 5. Lyapunov exponents In the case of E3, deﬁnite limit is not clearly visible, but it is obvious that the exponents stay positive in the vicinity of 0.179, thus corresponding to the chaos clearly visible in the third phase space section. 6. Summary In this article we present further numerical results concerning the planar double pendulum system. Its chaotic fea tures are fully exposed by all the tools used, namely, Poincare´ sections, bifurcation diagrams and Lyapunov character istic exponents. They reveal regular behaviour at the zero energy limit, which, as energy is increased, transforms into quasi regular and, ﬁnally, globally chaotic motion. We note that the existence and decay of the invariant torii, is visible on the phase space secti exponent analysis is not as evident. Together with the other aspects, though, it could be consid We note that the existence and decay of the invariant torii, is visible on the phase space sections, but the Lyapunov exponent analysis is not as evident. Together with the other aspects, though, it could be considered as a valid result. The bifurcation diagrams are not standard, in the sense that energy was used as the parameter, instead of a quantity explicitly appearing in the equations of motion. However, it is with energy that the characteristics of the system change drastically, as the obtained diagrams proved. Thus, further evidence of chaotic behaviour such as windows and bifur cations were found, providing yet another image of chaos for the double pendulum. The bifurcation diagrams are not standard, in the sense that energy was used as the parameter, instead of a quantity explicitly appearing in the equations of motion. However, it is with energy that the characteristics of the system change drastically, as the obtained diagrams proved. Thus, further evidence of chaotic behaviour such as windows and bifur cations were found, providing yet another image of chaos for the double pendulum. [1] OhlhoﬀA, Richter PH. Forces in the double pendulum. Z Angew Math Mech 2000;80(8):517 34. [2] Press WH, Flannery BP, Teukolsky SA, Vetterling WT. Numerical recipes in C. Cambridge: Cambridge University Press; 1992. [3] Wolf A, Swift JB, Swinney HL, Vastano JA. Determining Lyapunov exponents from a time series. Physica D 1985;16:285 315. hlhoﬀA, Richter PH. Forces in the double pendulum. Z Angew Math Mech 2000;80(8):517 34. [2] Press WH, Flannery BP, Teukolsky SA, Vetterling WT. Numerical recipes in C. Cambridge: Cambridg [3] Wolf A, Swift JB, Swinney HL, Vastano JA. Determining Lyapunov exponents from a time series. Phy References 6
https://openalex.org/W4384405943	https://drops.dagstuhl.de/opus/volltexte/2022/16553/pdf/LIPIcs-SEA-2022-19.pdf	English	null	An Experimental Evaluation of Semidefinite Programming and Spectral Algorithms for Max Cut	ACM journal of experimental algorithmics	2,023	cc-by	8,907	An Experimental Evaluation of Semidefinite Programming and Spectral Algorithms for Max Cut Renee Mirka ! Cornell University, Ithaca, NY, USA David P. Williamson ! Cornell University, Ithaca, NY, USA © Renee Mirka and David P. Williamson; licensed under Creative Commons License CC-BY 4.0 20th International Symposium on Experimental Algorithms (SEA 2022). Editors: Christian Schulz and Bora Uçar; Article No. 19; pp. 19:1–19:14 Leibniz International Proceedings in Informatics Schloss Dagstuhl – Leibniz-Zentrum für Informatik, Dagstuhl Publishing, Germany Abstract We experimentally evaluate the performance of several Max Cut approximation algorithms. In particular, we compare the results of the Goemans and Williamson algorithm using semidefinite programming with Trevisan’s algorithm using spectral partitioning. The former algorithm has a known .878 approximation guarantee whereas the latter has a .614 approximation guarantee. We investigate whether this gap in approximation guarantees is evident in practice or whether the spectral algorithm performs as well as the SDP. We also compare the performances to the standard greedy Max Cut algorithm which has a .5 approximation guarantee and two additional spectral algorithms. The algorithms are tested on Erdős-Renyi random graphs, complete graphs from TSPLIB, and real-world graphs from the Network Repository. We find, unsurprisingly, that the spectral algorithms provide a significant speed advantage over the SDP. In our experiments, the spectral algorithms return cuts with values which are competitive with those of the SDP. 2012 ACM Subject Classification Theory of computation →Approximation algorithms analysis Keywords and phrases Max Cut, Approximation Algorithms Keywords and phrases Max Cut, Approximation Algorithms Digital Object Identifier 10.4230/LIPIcs.SEA.2022.19 Digital Object Identifier 10.4230/LIPIcs.SEA.2022.19 Supplementary Material Software (Source Code): https://github.com/rmirka/max-cut- experiments; archived at swh:1:dir:eb13652be65db33c0ea45e66314475a4327cae0d Supplementary Material Software (Source Code): https://github.com/rmirka/max-cut- experiments; archived at swh:1:dir:eb13652be65db33c0ea45e66314475a4327cae0d Funding This work was supported by the National Science Foundation [NSF CCF-2007009]. Exp Eval of SDP and Spectral Algs for Max Cut to O∗(2(1−3/(∆+1))n) [8]. Results from Hrga et al., Hrga and Povh, Krislock, Malick, and Roupin, and Rendl, Rinaldi, and Wiegele utilize branch and bound techniques to produce further exact solvers [10, 11, 14, 19]. However, due to the lack of an efficient (polynomial-time) algorithm, researchers have also considered finding good approximation algorithms. An α-approximation algorithm is a polynomial time algorithm which guarantees a solution with a value at least an α fraction of the optimal solution. As one of the most well-studied problems in theoretical computer science, there is a breadth of known approximation algorithms for Max Cut varying in runtime and approximation guarantee quality. The simplest randomized approximation algorithm assigns a vertex v ∈V to either S or T with equal probability. In expectation, this is a .5-approximation algorithm. Another .5-approximation can be achieved through a simple greedy algorithm presented by Sahni and Gonzalez [21]. In this algorithm, start with S, T = ∅. While there are still unassigned vertices, any unassigned vertex v is chosen and the quantities cS(v) = P u∈S:(u,v)∈E w(u,v) and cT (v) = P u∈T :(u,v)∈E w(u,v) are computed. If cS(v) > cT (v), v is assigned to T and otherwise to S. The .5-approximation guarantee was the best known until Goemans and Williamson [7] presented a .878-approximation algorithm, which is the best possible guarantee assuming the Unique Games Conjecture [13]. Their algorithm relies on a semidefinite programming (SDP) relaxation of the Max Cut problem to find a high-value cut. While the approximation guarantee likely cannot be surpassed by another polynomial-time algorithm, solving the SDP can be quite costly in practice. More recently, Trevisan [23] introduced a simple .531-approximation for Max Cut based on spectral partitioning. Soto [22] improved this guarantee to .614. Though the approximation guarantees are weaker than the SDP algorithm, the spectral techniques are much cheaper to implement. In theory, there is a trade off between the computational speed and solution quality of Goemans and Williamson’s SDP algorithm versus Trevisan’s spectral algorithm. This paper seeks to determine whether this trade off exists in practice or if Trevisan’s algorithm returns solutions competitive with those of the SDP. Several previous papers have experimentally compared Max Cut algorithms and heuristics. Bertoni, Campadelli, and Grossi compare cuts computed by their .39-approximation Lorena algorithm, inspired by Goemans-Williamns SDP, to the SDP and a neural .5-approximation algorithm [3]. 1 Introduction Given as input a graph G = (V, E) and weights we ∈R+ for all e ∈E, the Max Cut problem asks to partition V into two sets such that the sum of the weights of the edges crossing the partition is maximized. In particular, a cut is given by a pair of sets (S, T) such that V = S ∪T and S ∩T = ∅. The value of this cut is X (s,t)∈E:s∈S,t∈T w(s,t), and Max Cut seeks to find a cut maximizing this quantity. and Max Cut seeks to find a cut maximizing this quantity. and Max Cut seeks to find a cut maximizing this quantity.i Max Cut is a problem of vast theoretical and practical significance. It is polynomial solvable for certain classes of graphs, e.g. planar graphs [9, 15], and is well-known to be NP- hard in general; it appears on Karp’s original list of NP-complete problems [12]. Additionally, Max Cut has applications in fields such as data clustering [16], circuit design, and statistical physics [1]; see Poljak and Tuza for a comprehensive survey [17]. Many researchers have made improvements towards exact solvers for Max Cut. For general graphs of unbounded average degree, Williams presented a Max Cut algorithm using exponential space to exactly solve (and count the number of optimal solutions) in O(m32ωn/3) time where ω < 2.376 [24]. Croce, Kaminski, and Paschos introduced an algorithm to find a Max Cut in graphs with bounded maximum degree, ∆, running in O∗(2(1−2/∆)n) time where O∗() suppresses polynomial factors [4]. Golovnev improved this 19:2 2.2 Trevisan’s Algorithm The next algorithm for finding a large cut in a graph is Trevisan’s spectral algorithm. Trevisan proved a .531 approximation ratio for the algorithm, while Soto improved the analysis to .614. Here, we describe Soto’s presentation of the algorithm. Given a graph G = (V, E) with \|V \| = n, the adjacency matrix A = (aij) is given by aij = 1 if (i, j) ∈E and 0 otherwise. Then the normalized adjacency matrix A is given by A = D−1/2AD−1/2 where D = diag(d) for d(i) the degree of vertex i. In our implementation of the algorithm, we compute the eigenvector, x, corresponding to the minimum eigenvalue of I + A. After normalizing x so that maxi \|xi\| = 1, a number t2 is drawn uniformly at random from [0,1]. We let L = {v : xv ≤−t}, R = {v : xv ≥t}, and V ′ = V \ (L ∪R). L = {v : xv ≤−t}, R = {v : xv ≥t}, and V ′ = V \ (L ∪R). Now (L, R) represents a partial cut of the vertices with V ′ being the vertices yet to be partitioned. Now (L, R) represents a partial cut of the vertices with V ′ being the vertices yet to be partitioned. Given L, R, and V ′, we compute C = total weight of the edges between L and R, C = total weight of the edges between L and R, X = total weight of the edges between L ∪R and V ′, and M = total weight of all edges −total weight of the edges between vertices of V M = total weight of all edges −total weight of the edges between vertices of V ′. If C + X/2 −M/2 < 0, we use the greedy algorithm to partition the vertices instead of t as the expected value of the cut is worse than that of greedy. If C + X/2 −M/2 > 0, we keep the partial cut and recurse to find a cut of the vertices in V ′ given by (L′, R′). Finally, we return the larger of the cuts (L ∪L′, R ∪R′) and (L ∪R′, R ∪L′). Exp Eval of SDP and Spectral Algs for Max Cut They found, on average, Lorena provided larger cuts on random graphs in significantly less time than the SDP and comparable time to the neural algorithm. Dolezal, Hofmeister, and Lefmann compare cuts from six algorithms, including the SDP, on random graphs concluding that the computationally-cheap random .5-approximation algorithm provides the best tradeoff between runtime and cut quality [5]. Goemans and Williamson also included computational results in their initial paper demonstrating the SDP often outperforms its .878 approximation guarantee. Berry and Goldberg tested several graph partitioning heuristics against each other and against the SDP, finding the heuristics consistently produce larger cuts than the SDP [2]. Dunning, Gupta, and Silberholz performed a systematic review of Max Cut heuristics and computationally tested 19 of them [6]. As far as we are aware there are no previously published results comparing Trevisan’s spectral algorithm. In this paper, we evaluate the performances of the SDP, spectral, and greedy algorithms on a variety of graphs. Section 2 provides more complete descriptions of the five algorithms considered. Section 3 describes the experiments and presents the results of the algorithms on different classes of graphs. Finally, Section 4 concludes with a summary of the performances and introduces a few possible directions for future theoretical study. 19:3 R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson 2 Algorithms This section describes the five algorithms that we implemented for Max Cut. Section 2.1 de- scribes the benchmark greedy .5-approximation algorithm for Max Cut. Section 2.2 describes Trevisan’s spectral algorithm for Max Cut, while Section 2.3 describes two simplifications of this algorithm. Finally Section 2.4 describes the SDP algorithm. 2.1 Greedy Algorithm The first Max Cut algorithm we consider is the standard greedy algorithm. This fast and simple algorithm provides a benchmark for the speed and cut value of the others. For a graph G = (V, E), we return a cut (L, R) by greedily assigning vertices to either L or R one at a time. We start with L, R = ∅. In each step of the algorithm, we choose a vertex v ∈V yet to be assigned to L or R. Then we add v to L or R by choosing the larger of the two cuts (L ∪{v}, R) and (L, R ∪{v}) at this step. Exp Eval of SDP and Spectral Algs for Max Cut X/2 is the expected value added to this cut from the edges of the second type if the remaining vertices are greedily assigned, and M/2 is the expected value of the greedy cut due to edges of both of the first two types. Therefore, if C + X/2 > M/2, the partial cut being considered is performing better than the greedy algorithm would be in expectation. The greater the difference between C + X/2 and M/2, theoretically the better Trevisan’s algorithm is performing compared to greedy. It is not obvious that this is the best heuristic, but it does allow the algorithm to test several random values quickly. Alternatively, we also experimented with running Trevisan’s algorithm for several iter- ations and maintaining the best cut that was found. The advantage here as opposed to the previous modification is we do not have to determine which t value to keep. However, the runtime is slower because the entire algorithm is run several times instead of adding additional quick random draws.i The results of these modifications for two of the tested graphs are provided in Figure 1. In each figure, the line represents the results from trials of running the algorithm multiple times (1, 2, 5, 10, 20, 35, and 50 times). Note that the line is not monotonically increasing. This is because each group of runs was unique and not a cumulative total. For example, when considering how well Trevisan’s algorithm performs when running 10 iterations, we run 10 new iterations and do not build off of the 5 from the previous data point. The single dot represents the average runtime and cut value of the best result when implementing the first modification of multiple t2 values. The experiments were run on a variety of graphs and these are a representative sample. In general, it seems running the algorithm multiple times is more effective in increasing cut quality than choosing multiple t2 values. However, the number of iterations needed is not obvious, though it appears at least 5 are beneficial. Due to this observation, we use this method of running Trevisan’s algorithm 5 times and keeping the best cut for the experiments presented in this paper. 2.2 Trevisan’s Algorithm Since the results of Trevisan’s algorithm are highly reliant on the t2 value chosen, one could ask if there are ways to modify the algorithm to increase the likelihood of choosing a good t2 value. We tested two methods. The first chooses more than one t2 value at each stage of the algorithm. In particular, T = max(5, n/k) values were chosen where k was tested for k = 1, 2, 5, 10, 15, 25, 50, 100. The idea here was that choosing many random numbers should increase the probability of choosing a “good” random number. The major roadblock is deciding which partial cut corresponding to one of the t2 values the algorithm should SEA 2022 19:4 Exp Eval of SDP and Spectral Algs for Max Cut recurse on. We tested the greedy choice. More specifically, C, X, and M were computed for each drawn t2 value as above, and we kept the partial cut maximizing C + X/2 −M/2. We made this selection because it represents the partial cut that is currently performing better than the greedy algorithm by the largest margin. In particular, given a partial cut due to a partial assignment of vertices, we can consider three types of edges: edges with both endpoints assigned, edges with exactly one endpoint assigned, and edges with neither endpoint assigned. The third type are not affected by the partial cut and aren’t considered in this iteration. However, C computes the value of the cut in Trevisan’s algorithm due to the first type of edge. X/2 is the expected value added to this cut from the edges of the second type if the remaining vertices are greedily assigned, and M/2 is the expected value of the greedy cut due to edges of both of the first two types. Therefore, if C + X/2 > M/2, the partial cut being considered is performing better than the greedy algorithm would be in expectation. The greater the difference between C + X/2 and M/2, theoretically the better Trevisan’s algorithm is performing compared to greedy. It is not obvious that this is the best heuristic, but it does allow the algorithm to test several random values quickly. recurse on. We tested the greedy choice. More specifically, C, X, and M were computed for each drawn t2 value as above, and we kept the partial cut maximizing C + X/2 −M/2. We made this selection because it represents the partial cut that is currently performing better than the greedy algorithm by the largest margin. In particular, given a partial cut due to a partial assignment of vertices, we can consider three types of edges: edges with both endpoints assigned, edges with exactly one endpoint assigned, and edges with neither endpoint assigned. The third type are not affected by the partial cut and aren’t considered in this iteration. However, C computes the value of the cut in Trevisan’s algorithm due to the first type of edge. 2.3 Simple Spectral and Sweep Cuts Algorithms The simple spectral algorithm is a modification of Trevisan’s algorithm described in the previ- ous section. Instead of drawing a random number t in [0, 1], we return the cut corresponding to t = 0. In particular, let x be the eigenvector corresponding to the smallest eigenvalue as before. Since scaling numbers by a positive factor does not change their sign, we may skip the normalizing step for x. We let L = {v : xv < 0} and R = V \ L and return the cut (L, R). This modified simple spectral algorithm has no known approximation guarantee.f The sweep cuts algorithm works in a similar fashion. Here, we consider n −1 different cuts and return the best. Given the smallest eigenvector x, we sort the entries so that xi1 ≤xi2 ≤· · · ≤xin. Then we calculate the sweep cut value for Lj = {i1, . . . , ij} and Rj = V \ Lj for j = 1, . . . , n −1. The sweep cuts algorithm returns the cut (Lj, Rj) of maximal value. 19:5 R. Mirka and D. P. Williamson (a) johnson16-2-4. (b) ca-netscience. Figure 1 Plots depicting the effects on runtime and returned cut quality of running Trevisan’s algorithm multiple times on the johnnson16-2-4 and ca-netscience graphs. The X and Y axes are the time in seconds and the cut value, respectively. The light gray ‘x’ is presented for comparison and is the result of running Trevisan’s algorithm once, but testing many random values. (a) johnson16-2-4. (b) ca-netscience. (b) ca-netscience. Figure 1 Plots depicting the effects on runtime and returned cut quality of running Trevisan’s algorithm multiple times on the johnnson16-2-4 and ca-netscience graphs. The X and Y axes are the time in seconds and the cut value, respectively. The light gray ‘x’ is presented for comparison and is the result of running Trevisan’s algorithm once, but testing many random values. It is worth noting that the sweep cuts algorithm will always perform at least as well as the simple spectral algorithm in terms of cut value since one of the sweep cuts will be the same as the t = 0 cut. However, it is interesting to see how much better the sweep cuts algorithm performs since it is also guaranteed to have a slower runtime for the same reasons. 2.4 SDP Algorithm Goemans and Williamson introduced a .878 approximation algorithm for Max Cut. Instead of directly solving MaxCut(G) = max xi∈{1,−1} 1 2 X i<j wij(1 −xixj) where again wij is the weight of edge (i, j), they relax this program to one solvable by a semidefinite program. In particular, instead of requiring xi ∈{1, −1}, they require vi ∈Rn to be unit vectors and replace xixj with ⟨vi, vj⟩. Given a solution to this SDP relaxation, they draw a random vector r ∈Rn uniformly from the unit sphere and partition the vertices according to where again wij is the weight of edge (i, j), they relax this program to one solvable by a semidefinite program. In particular, instead of requiring xi ∈{1, −1}, they require vi ∈Rn to be unit vectors and replace xixj with ⟨vi, vj⟩. Given a solution to this SDP relaxation, they draw a random vector r ∈Rn uniformly from the unit sphere and partition the vertices according to L = {i : ⟨r, vi⟩≤0} and R = {i : ⟨r, vi⟩> 0}. L = {i : ⟨r, vi⟩≤0} and R = {i : ⟨r, vi⟩> 0}. This gives the .878 approximation in expectation. For our testing purposes, we draw 100 random vectors instead of 1. In terms of computation time, this is a cheap modification as the SDP does not have to be rerun. We return the maximum cut resulting from these 100 random vectors. 3.1 Erdős-Renyi Random Graphs The first class of graphs tested was Erdős-Renyi random graphs. An Erdős-Renyi random graph G(n, p) is a graph on n vertices where each possible edge is included independently with probability p. We tested random graphs with n = 50, 100, 200, 350, 500 and p = .1, .25, .5, .75. In our model, each included edge was given an edge weight of 1.i In terms of speed, the simple spectral algorithm significantly outperformed the other algorithms on all but three tested random graphs (where greedy was faster). On the other end of the spectrum, the SDP was far slower than the alternative algorithms. The time statistics are presented in Table 1. The plots in Figure 2a and Figure 2b illustrate how the computation times of each algorithm grow as the number of vertices increases. For these plots, we use the data from Table 1 with p = .5 fixed. Table 1 The time in seconds each algorithm took to compute a cut of an Erdős-Renyi random graph. Table 1 The time in seconds each algorithm took to compute a cut of an Erdős-Renyi random graph. 3.1 Erdős-Renyi Random Graphs Graph Greedy Trevisan Simple Spectral Sweep Cuts SDP G(50,0.1) 5.560 × 10−3 5.560 × 10−3 5.560 × 10−3 2.503 × 10−1 5.192 × 10−2 3.485 × 10−2 5.556 × 10−1 G(50,0.25) 7.600 × 10−4 1.533 × 10−2 6.600 × 10−4 6.600 × 10−4 6.600 × 10−4 2.410 × 10−3 4.711 × 10−1 G(50,0.5) 1.280 × 10−3 2.354 × 10−2 7.800 × 10−4 7.800 × 10−4 7.800 × 10−4 4.760 × 10−3 4.502 × 10−1 G(50,0.75) 1.870 × 10−3 1.741 × 10−2 8.000 × 10−4 8.000 × 10−4 8.000 × 10−4 8.690 × 10−3 9.727 × 10−1 G(100,0.1) 2.000 × 10−3 2.000 × 10−3 2.000 × 10−3 3.597 × 10−2 2.380 × 10−3 1.106 × 10−2 2.929 G(100,0.25) 3.860 × 10−3 6.945 × 10−2 2.340 × 10−3 2.340 × 10−3 2.340 × 10−3 1.849 × 10−2 3.440 G(100,0.5) 7.330 × 10−3 1.021 × 10−1 2.370 × 10−3 2.370 × 10−3 2.370 × 10−3 3.206 × 10−2 7.235 G(100,0.75) 1.064 × 10−2 1.162 × 10−1 9.960 × 10−3 9.960 × 10−3 9.960 × 10−3 2.653 × 10−1 5.823 G(200,0.1) 1.222 × 10−2 1.222 × 10−2 1.222 × 10−2 2.464 × 10−1 3.299 × 10−2 6.941 × 10−2 2.575 × 101 G(200,0.25) 2.963 × 10−2 2.444 × 10−1 8.650 × 10−3 8.650 × 10−3 8.650 × 10−3 1.892 × 10−1 2.942 × 101 G(200,0.5) 5.428 × 10−2 6.949 × 10−1 1.266 × 10−2 1.266 × 10−2 1.266 × 10−2 3.853 × 10−1 3.848 × 101 G(200,0.75) 7.809 × 10−2 6.463 × 10−1 9.900 × 10−3 9.900 × 10−3 9.900 × 10−3 4.740 × 10−1 4.945 × 101 G(350,0.1) 6.192 × 10−2 1.022 2.407 × 10−2 2.407 × 10−2 2.407 × 10−2 4.184 × 10−1 1.216 × 102 G(350,0.25) 1.737 × 10−1 1.201 3.009 × 10−2 3.009 × 10−2 3.009 × 10−2 1.138 1.726 × 102 G(350,0.5) 3.013 × 10−1 1.718 3.848 × 10−2 3.848 × 10−2 3.848 × 10−2 2.342 2.058 × 102 G(350,0.75) 4.438 × 10−1 2.015 3.324 × 10−2 3.324 × 10−2 3.324 × 10−2 3.015 2.798 × 102 G(500,0.1) 1.668 × 10−1 1.875 7.049 × 10−2 7.049 × 10−2 7.049 × 10−2 1.622 3.355 × 102 G(500,0.25) 3.937 × 10−1 2.239 5.936 × 10−2 5.936 × 10−2 5.936 × 10−2 3.325 3.919 × 102 G(500,0.5) 7.859 × 10−1 4.587 6.472 × 10−2 6.472 × 10−2 6.472 × 10−2 6.598 5.669 × 102 G(500,0.75) 1.260 5.263 7.195 × 10−2 7.195 × 10−2 7.195 × 10−2 9.837 8.116 × 102 The spectral algorithms also performed the best in terms of the returned cut quality for random graphs. 3 Experiments All algorithms were implemented in Julia. They were run on a machine with a 2 GHz Intel Core i5 processor with 8 GB 1867 MHz LPDDR3 memory. The SDP algorithm was computed with the JuMP modeling language for Julia and the SCS package providing the splitting cone solver. The LinearAlgebra package was used for the eigenvector computations of the spectral algorithms. SEA 2022 19:6 Exp Eval of SDP and Spectral Algs for Max Cut We measured the algorithms’ performance with three types of test data. We used 20 Erdős-Renyi random graphs with 50-500 vertices, 16 complete graphs from TSPLIB [18] with 29-280 vertices (average 124), and 17 sparser graphs from the Network Repository [20] with 39-1133 vertices (average 327). R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson (a) All tested algorithms excluding the SDP. (b) All tested algorithms. Figure 2 Plots depicting the effects on runtime of increasing the number of vertices of an Erdős-Renyi graph with p = .5. The X and Y axes are the number of vertices and the computation time in seconds, respectively. (a) All tested algorithms excluding the SDP. (b) All tested algorithms. (b) All tested algorithms. (a) All tested algorithms excluding the SDP. Figure 2 Plots depicting the effects on runtime of increasing the number of vertices of an Erdős-Renyi graph with p = .5. The X and Y axes are the number of vertices and the computation time in seconds, respectively. Figure 2 Plots depicting the effects on runtime of increasing the number of vertices of an Erdős-Renyi graph with p = .5. The X and Y axes are the number of vertices and the computation time in seconds, respectively. Figure 2 Plots depicting the effects on runtime of increasing the number of vertices of an Erdős-Renyi graph with p = .5. The X and Y axes are the number of vertices and the computation time in seconds, respectively. Table 2 The value of the cut each algorithm returned for an Erdős-Renyi random graph. 3.1 Erdős-Renyi Random Graphs The SDP returned the best result for three graphs but one of the cuts was matched by Trevisan’s algorithm. Trevisan’s algorithm provided the best cut for 5 graphs, and the sweep cuts algorithm was the second best option for all of these, in addition to being the best for 14 graphs. These results are provided in Table 2. The spectral algorithms also performed the best in terms of the returned cut quality for random graphs. The SDP returned the best result for three graphs but one of the cuts was matched by Trevisan’s algorithm. Trevisan’s algorithm provided the best cut for 5 graphs, and the sweep cuts algorithm was the second best option for all of these, in addition to being the best for 14 graphs. These results are provided in Table 2. The spectral algorithms also performed the best in terms of the returned cut quality for random graphs. The SDP returned the best result for three graphs but one of the cuts was matched by Trevisan’s algorithm. Trevisan’s algorithm provided the best cut for 5 graphs, and the sweep cuts algorithm was the second best option for all of these, in addition to being the best for 14 graphs. These results are provided in Table 2. 19:7 Exp Eval of SDP and Spectral Algs for Max Cut Table 3 The time in seconds each algorithm took to compute a cut of a complete graph from TSPLIB. Table 3 The time in seconds each algorithm took to compute a cut of a complete graph from TSPLIB. Table 3 The time in seconds each algorithm took to compute a cut of a complete graph from TSPLIB. Graph Greedy Trevisan Simple Spectral Sweep Cuts SDP bayg29 4.300 × 10−4 5.040 × 10−3 3.100 × 10−4 3.100 × 10−4 3.100 × 10−4 9.700 × 10−4 1.945 × 10−1 bays29 7.500 × 10−4 9.660 × 10−3 6.900 × 10−4 6.900 × 10−4 6.900 × 10−4 1.160 × 10−3 3.002 × 10−1 berlin52 2.190 × 10−3 2.190 × 10−3 2.190 × 10−3 3.058 × 10−2 2.540 × 10−3 7.580 × 10−3 9.291 × 10−1 bier127 3.674 × 10−2 3.520 × 10−1 1.370 × 10−2 1.370 × 10−2 1.370 × 10−2 1.125 × 10−1 6.832 brazil58 3.260 × 10−3 3.260 × 10−3 3.260 × 10−3 3.237 × 10−2 4.370 × 10−3 8.490 × 10−3 1.229 brg180 9.482 × 10−2 9.482 × 10−2 9.482 × 10−2 1.223 2.028 × 10−1 3.072 × 10−1 1.548 × 101 ch130 3.371 × 10−2 3.355 × 10−1 1.221 × 10−2 1.221 × 10−2 1.221 × 10−2 1.352 × 10−1 7.503 ch150 5.701 × 10−2 9.769 × 10−1 1.811 × 10−2 1.811 × 10−2 1.811 × 10−2 1.745 × 10−1 1.209 × 101 d198 1.094 × 10−1 1.402 3.997 × 10−2 3.997 × 10−2 3.997 × 10−2 4.869 × 10−1 3.844 × 101 eil101 1.912 × 10−2 5.535 × 10−1 8.870 × 10−3 8.870 × 10−3 8.870 × 10−3 7.724 × 10−2 3.982 gr120 2.376 × 10−2 4.412 × 10−1 1.050 × 10−2 1.050 × 10−2 1.050 × 10−2 1.153 × 10−1 1.372 × 101 gr137 4.262 × 10−2 6.078 × 10−1 1.493 × 10−2 1.493 × 10−2 1.493 × 10−2 1.665 × 10−1 1.452 × 101 gr202 1.194 × 10−1 2.471 2.662 × 10−2 2.662 × 10−2 2.662 × 10−2 4.779 × 10−1 3.067 × 101 gr96 1.632 × 10−2 3.044 × 10−1 4.690 × 10−3 4.690 × 10−3 4.690 × 10−3 5.674 × 10−2 4.753 kroA100 1.880 × 10−2 2.037 × 10−1 7.720 × 10−3 7.720 × 10−3 7.720 × 10−3 5.644 × 10−2 3.285 a280 2.988 × 10−1 4.439 6.127 × 10−2 6.127 × 10−2 6.127 × 10−2 1.534 1.555 × 102 3.2 Complete Graphs The algorithms were also tested on 16 complete graphs from TSPLIB, an online library of sample instances for the Travelling Salesman Problem and related graph problems. The performance in regards to time largely mirrored that of the random graphs. The simple spectral algorithm was significantly faster than the rest of the algorithms on the vast majority of graphs, followed by the greedy, Trevisan’s, and sweep cuts algorithms with relatively quick computation times, and the SDP with a massive slowdown. This data is presented in Table 3. Again, the spectral algorithms most frequently returned the highest quality cut; these results are summarized in Table 4. For 15 16 (93.75%) of these graphs, the best cut was found by either the simple spectral algorithm (5 times), Trevisan’s algorithm (3 times) or the sweep cuts algorithm (12 times). Furthermore, for the graph d198 where the SDP computed the best cut, the loss in quality from the spectral solutions was quite small. These values are given in Table 5. In Figure 3a, Figure 3b, Figure 4a, and Figure 4b, we provide a representative sample of the trade-off between runtime and returned cut value of the algorithms using the a280, ch150, and eil101 graphs. R. Mirka and D. P. Williamson Graph Greedy Trevisan Simple Spectral Sweep Cuts SDP G(50,0.1) 8.700 × 101 9.600 × 101 9.600 × 101 9.600 × 101 9.400 × 101 9.500 × 101 9.200 × 101 G(50,0.25) 1.970 × 102 2.060 × 102 2.060 × 102 2.080 × 102 2.100 × 102 2.100 × 102 2.100 × 102 G(50,0.5) 3.480 × 102 3.600 × 102 3.600 × 102 3.600 × 102 3.560 × 102 3.600 × 102 3.600 × 102 3.600 × 102 3.600 × 102 3.600 × 102 3.600 × 102 G(50,0.75) 5.140 × 102 5.140 × 102 4.990 × 102 5.190 × 102 5.240 × 102 5.240 × 102 5.240 × 102 G(100,0.1) 3.210 × 102 3.290 × 102 3.420 × 102 3.430 × 102 3.430 × 102 3.430 × 102 3.290 × 102 G(100,0.25) 7.640 × 102 7.830 × 102 7.850 × 102 7.880 × 102 7.880 × 102 7.880 × 102 7.860 × 102 G(100,0.5) 1.351 × 103 1.363 × 103 1.346 × 103 1.375 × 103 1.375 × 103 1.375 × 103 1.361 × 103 G(100,0.75) 2.019 × 103 2.024 × 103 2.020 × 103 2.026 × 103 2.026 × 103 2.026 × 103 2.016 × 103 G(200,0.1) 1.212 × 103 1.250 × 103 1.250 × 103 1.250 × 103 1.234 × 103 1.242 × 103 1.211 × 103 G(200,0.25) 2.795 × 103 2.859 × 103 2.847 × 103 2.861 × 103 2.861 × 103 2.861 × 103 2.778 × 103 G(200,0.5) 5.388 × 103 5.420 × 103 5.412 × 103 5.423 × 103 5.423 × 103 5.423 × 103 5.326 × 103 G(200,0.75) 7.784 × 103 7.855 × 103 7.831 × 103 7.875 × 103 7.875 × 103 7.875 × 103 7.815 × 103 G(350,0.1) 3.556 × 103 3.582 × 103 3.639 × 103 3.651 × 103 3.651 × 103 3.651 × 103 3.611 × 103 G(350,0.25) 8.378 × 103 8.544 × 103 8.583 × 103 8.585 × 103 8.585 × 103 8.585 × 103 8.236 × 103 G(350,0.5) 1.623 × 104 1.627 × 104 1.643 × 104 1.649 × 104 1.649 × 104 1.649 × 104 1.603 × 104 G(350,0.75) 2.356 × 104 2.378 × 104 2.378 × 104 2.378 × 104 2.374 × 104 2.374 × 104 2.353 × 104 G(500, .1) 7.155 × 103 7.155 × 103 7.303 × 103 7.329 × 103 7.329 × 103 7.329 × 103 7.097 × 103 G(500, .25) 1.673 × 104 1.697 × 104 1.712 × 104 1.714 × 104 1.714 × 104 1.714 × 104 1.652 × 104 G(500, .5) 3.272 × 104 3.275 × 104 3.313 × 104 3.314 × 104 3.314 × 104 3.314 × 104 3.311 × 104 G(500, .75) 4.820 × 104 4.852 × 104 4.852 × 104 4.852 × 104 4.847 × 104 4.849 × 104 4.813 × 104 SEA 2022 19:8 R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson Table 4 The value of the cut each algorithm returned for a complete graph from TSPLIB. R. Mirka and D. P. Williamson Graph Greedy Trevisan Simple Spectral Sweep Cuts SDP bayg29 3.837 × 104 4.225 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 4.269 × 104 bays29 4.831 × 104 5.393 × 104 5.369 × 104 5.399 × 104 5.399 × 104 5.399 × 104 5.386 × 104 berlin52 4.532 × 105 4.616 × 105 4.465 × 105 4.681 × 105 4.681 × 105 4.681 × 105 4.522 × 105 bier127 2.162 × 107 2.300 × 107 2.322 × 107 2.330 × 107 2.330 × 107 2.330 × 107 2.320 × 107 brazil58 2.319 × 106 2.319 × 106 2.319 × 106 2.319 × 106 2.319 × 106 2.319 × 106 2.315 × 106 2.315 × 106 2.180 × 106 brg180 4.118 × 107 4.616 × 107 4.616 × 107 4.616 × 107 4.531 × 107 4.551 × 107 4.330 × 107 ch130 1.777 × 106 1.885 × 106 1.888 × 106 1.888 × 106 1.888 × 106 1.888 × 106 1.888 × 106 1.888 × 106 1.887 × 106 ch150 2.500 × 106 2.521 × 106 2.526 × 106 2.526 × 106 2.526 × 106 2.526 × 106 2.526 × 106 2.526 × 106 2.434 × 106 d198 9.635 × 106 1.286 × 107 1.292 × 107 1.293 × 107 1.293 × 107 1.293 × 107 1.293 × 107 eil101 1.052 × 105 1.070 × 105 1.070 × 105 1.070 × 105 1.063 × 105 1.064 × 105 1.058 × 105 gr120 2.123 × 106 2.147 × 106 2.156 × 106 2.157 × 106 2.157 × 106 2.157 × 106 2.154 × 106 gr137 2.241 × 107 3.044 × 107 3.066 × 107 3.070 × 107 3.070 × 107 3.070 × 107 3.070 × 107 gr202 1.372 × 107 1.533 × 107 1.559 × 107 1.593 × 107 1.593 × 107 1.593 × 107 1.581 × 107 gr96 8.967 × 106 1.156 × 107 1.166 × 107 1.166 × 107 1.166 × 107 1.166 × 107 1.166 × 107 1.166 × 107 1.157 × 107 kroA100 5.848 × 106 5.850 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 5.897 × 106 a280 2.447 × 106 3.151 × 106 3.21 × 106 3.21 × 106 3.21 × 106 3.21 × 106 2.970 × 106 Table 4 The value of the cut each algorithm returned for a complete graph from TSPLIB. 3.3 Sparser Graphs The third group of graphs is composed of a variety of graphs from the Network Repository, an online and interactive collection of network graph data coming from a variety of sources and applications. Though more structured than a random graph, these 17 graphs are sparser than the complete graphs tested in Section 3.2 and were chosen from a range of real-world scenarios. Unsurprisingly, the relationships between relative computation times remains unchanged. The simple spectral and greedy algorithms each accounted for about half of the fastest times while the SDP was consistently considerably slower (Table 6). 19:9 19:10 Exp Eval of SDP and Spectral Algs for Max Cut 10 Exp Eval of SDP and Spectral Algs for Max Cut (a) ch150. (b) eil101. Figure 4 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the ch150 and eil101 graphs. The X and Y axes are the runtime in seconds and the Figure 4 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the ch150 and eil101 graphs. The X and Y axes are the runtime in seconds and the returned cut value, respectively. For this group of graphs, the algorithms’ relative cut quality is more varied than with the previous. Of the 17 graphs tested, the SDP returned the best cut for 7 instances whereas the spectral algorithms combined for 11 best (with one instance of a tie between the SDP and simple spectral) (Table 7). In Figure 5a, Figure 5b, Figure 6a, and Figure 6b, we provide a representative sample of the trade-off between runtime and returned cut value of the algorithms using the graphs ia-infect-dublin, email-enron-only, and soc-dolphins. R. Mirka and D. P. Williamson Table 5 The percent decrease in cut value from the SDP to the spectral cuts. Graph Trevisan Simple Spectral Sweep Cuts d198 ∼.6% ∼.1% ∼.06% (a) All tested algorithms excluding the SDP. (b) All tested algorithms. Figure 3 Plots depicting the computation time and returned cut values of algorithms on the a280 graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. Table 5 The percent decrease in cut value from the SDP to the spectral cuts. Graph Trevisan Simple Spectral Sweep Cuts d198 ∼.6% ∼.1% ∼.06% (a) All tested algorithms excluding the SDP. (b) All tested algorithms. Figure 3 Plots depicting the computation time and returned cut values of algorithms on the a280 graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. Table 5 The percent decrease in cut value from the SDP to the spectral cuts. Table 5 The percent decrease in cut value from the SDP to the spectral cuts. (a) All tested algorithms excluding the SDP. (b) All tested algorithms. (a) All tested algorithms excluding the SDP. Figure 3 Plots depicting the computation time and returned cut values of algorithms on the a280 graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. Figure 3 Plots depicting the computation time and returned cut values of algorithms on the a280 graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. SEA 2022 9:10 Exp Eval of SDP and Spectral Algs for Max Cut (a) ch150. (b) eil101. Figure 4 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the ch150 and eil101 graphs. The X and Y axes are the runtime in seconds and the returned cut value, respectively. 19:10 Exp Eval of SDP and Spectral Algs for Max Cut 19:10 Exp Eval of SDP and Spectral Algs for Max Cut R. Mirka and D. P. Williamson Table 6 The time in seconds each algorithm took to compute a cut of a graph from the Network Repository arising in the real-world. Table 6 The time in seconds each algorithm took to compute a cut of a graph from the Network Repository arising in the real-world. Table 6 The time in seconds each algorithm took to compute a cut of a graph from the Network Repository arising in the real-world. R. Mirka and D. P. Williamson Table 7 The value of the cut each algorithm returned for a graph from the Network Repository. 4 Conclusion The goal of this paper was to compare Max Cut algorithms with varying approximation guarantees in practice. In particular, we know the SDP has the provably best approximation guarantee; however, it is also the costliest in terms of computational space and time. This raises the question of whether or not the “cheaper” spectral Max Cut algorithms can perform competitively to the SDP in practice. Furthermore, if yes, can the approximation guarantees be improved? As demonstrated, the spectral and greedy algorithms provide a significant speed advantage over the SDP. Additionally, they often compute cuts better than or comparable to the cuts returned by the SDP, despite the disparity in approximation guarantees. The results of this experiment appear to illustrate spectral algorithms are competitive with the SDP algorithm in practice. This suggests that the investigation into approximation guarantees is a direction for further theoretical study. In terms of practical implementations, for the graphs that the SDP seems to perform better on, one could consider running Trevisan’s algorithm for even more than 5 iterations and choosing the best cut returned. The magnitude of the speed advantage of Trevisan’s algorithm would allow for many runs before being as costly as the SDP, especially since the initial eigenvector only needs to be computed once. Additionally, finding a viable heuristic to use when choosing multiple t2 values would also provide implementation benefits. We attempted to improve Trevisan’s algorithm through drawing additional random t2 values and greedily choosing one. However, it is not obvious that this choice in heuristic is optimal. 19:11 R. Mirka and D. P. Williamson R. Mirka and D. P. Williamson Graph # vertices # edges Greedy Trevisan Simple Spectral Sweep Cuts SDP ENZYMES8 88 133 1.370 × 10−3 1.370 × 10−3 1.370 × 10−3 4.342 × 10−2 1.776 × 10−1 1.015 × 10−2 2.356 × 101 eco-stmarks 54 356 5.900 × 10−4 5.900 × 10−4 5.900 × 10−4 2.146 × 10−2 1.993 × 10−1 2.597 × 10−2 6.939 johnson16-2-4 120 5460 2.519 × 10−2 1.603 × 10−1 2.600 × 10−3 2.600 × 10−3 2.600 × 10−3 8.544 × 10−2 8.178 × 10−1 hamming6-2 64 1824 3.540 × 10−3 5.081 × 10−2 1.190 × 10−3 1.190 × 10−3 1.190 × 10−3 1.558 × 10−2 9.926 × 10−1 ia-infect-hyper 113 2196 8.920 × 10−3 1.248 × 10−1 3.280 × 10−3 3.280 × 10−3 3.280 × 10−3 4.649 × 10−2 6.279 soc-dolphins 62 159 4.600 × 10−4 4.600 × 10−4 4.600 × 10−4 1.845 × 10−2 8.900 × 10−4 1.690 × 10−3 2.464 email-enron-only 143 623 5.960 × 10−3 1.174 × 10−1 5.650 × 10−3 5.650 × 10−3 5.650 × 10−3 1.575 × 10−2 5.681 × 101 dwt_209 209 976 1.349 × 10−2 3.012 × 10−1 8.380 × 10−3 8.380 × 10−3 8.380 × 10−3 4.641 × 10−2 7.073 × 101 inf-USAir97 332 2126 5.780 × 10−2 5.780 × 10−2 5.780 × 10−2 2.944 7.350 × 10−2 2.258 × 10−1 3.361 × 102 ca-netscience 379 914 2.590 × 10−2 2.590 × 10−2 2.590 × 10−2 5.124 × 10−1 8.440 × 10−2 1.146 × 10−1 3.584 × 102 ia-infect-dublin 410 2765 6.480 × 10−2 9.720 × 10−1 4.770 × 10−2 4.770 × 10−2 4.770 × 10−2 2.387 × 10−1 6.438 × 102 road-chesapeake 39 170 4.000 × 10−4 4.000 × 10−4 4.000 × 10−4 8.000 × 10−3 5.000 × 10−4 1.200 × 10−3 2.759 × 10−1 Erdos991 492 1417 4.490 × 10−2 4.490 × 10−2 4.490 × 10−2 2.634 6.090 × 10−2 1.933 × 10−1 5.143 × 102 dwt_503 503 3265 7.240 × 10−2 2.039 6.640 × 10−2 6.640 × 10−2 6.640 × 10−2 3.471 × 10−1 1.081 × 103 p-hat700-1 700 60999 1.264 1.009 × 101 1.591 × 10−1 1.591 × 10−1 1.591 × 10−1 9.273 1.270 × 103 DD687 725 2600 9.390 × 10−2 9.390 × 10−2 9.390 × 10−2 4.332 3.816 × 10−1 6.521 × 10−1 3.320 × 103 email-univ 1133 5451 2.081 × 10−1 2.081 × 10−1 2.081 × 10−1 1.345 × 101 1.179 2.703 7.572 × 103 Table 7 The value of the cut each algorithm returned for a graph from the Network Repository. R. Mirka and D. P. Williamson Graph # vertices # edges Greedy Trevisan Simple Spectral Sweep Cuts SDP ENZYMES8 88 133 1.170 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 1.260 × 102 eco-stmarks 54 356 8.891 × 102 1.190 × 103 1.190 × 103 1.190 × 103 9.354 × 102 9.354 × 102 9.601 × 102 johnson16-2-4 120 5460 3.036 × 103 3.036 × 103 3.036 × 103 3.036 × 103 3.036 × 103 3.036 × 103 2.958 × 103 2.986 × 103 2.918 × 103 hamming6-2 64 1824 9.920 × 102 9.920 × 102 9.920 × 102 9.920 × 102 9.920 × 102 9.920 × 102 9.680 × 102 9.690 × 102 9.760 × 102 ia-infect-hyper 113 2196 1.213 × 103 1.233 × 103 1.233 × 103 1.233 × 103 1.227 × 103 1.227 × 103 1.211 × 103 soc-dolphins 62 159 1.120 × 102 1.120 × 102 1.190 × 102 1.210 × 102 1.210 × 102 1.210 × 102 1.150 × 102 email-enron-only 143 623 3.920 × 102 4.130 × 102 4.130 × 102 4.130 × 102 3.710 × 102 3.800 × 102 3.960 × 102 dwt_209 209 976 5.250 × 102 5.270 × 102 5.250 × 102 5.270 × 102 5.400 × 102 5.400 × 102 5.400 × 102 inf-USAir97 332 2126 9.661 × 101 9.820 × 101 8.184 × 101 9.337 × 101 1.074 × 102 1.074 × 102 1.074 × 102 ca-netscience 379 914 5.830 × 102 5.880 × 102 5.270 × 102 5.270 × 102 6.110 × 102 6.110 × 102 6.110 × 102 ia-infect-dublin 410 2765 1.648 × 103 1.659 × 103 1.550 × 103 1.558 × 103 1.664 × 103 1.664 × 103 1.664 × 103 road-chesapeake 39 170 1.230 × 102 1.230 × 102 1.210 × 102 1.230 × 102 1.250 × 102 1.250 × 102 1.250 × 102 Erdos991 492 1417 9.330 × 102 9.340 × 102 7.350 × 102 7.580 × 102 9.240 × 102 dwt_503 503 3265 1.822 × 103 1.822 × 103 1.921 × 103 1.921 × 103 1.909 × 103 p-hat700-1 700 60999 3.261 × 104 3.269 × 104 3.215 × 104 3.305 × 104 3.304 × 104 DD687 725 2600 1.669 × 103 1.671 × 103 1.616 × 103 1.617 × 103 1.680 × 103 email-univ 1133 5451 3.546 × 103 3.546 × 103 3.341 × 103 3.344 × 103 3.264 × 103 SEA 2022 19:12 Exp Eval of SDP and Spectral Algs for Max Cut 19:12 Exp Eval of SDP and Spectral Algs for Max Cut (a) All tested algorithms excluding the SDP. (b) All tested algorithms. Figure 5 Plots depicting the computation time and returned cut values of algorithms on the ia-infect-dublin graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. (a) All tested algorithms excluding the SDP. (b) All tested algorithms. (b) All tested algorithms. (a) All tested algorithms excluding the SDP. Figure 5 Plots depicting the computation time and returned cut values of algorithms on the ia-infect-dublin graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. Figure 5 Plots depicting the computation time and returned cut values of algorithms on the ia-infect-dublin graph. The X and Y axes are the runtime in seconds and the returned cut value, respectively. (a) email-enron-only. (b) soc-dolphins. Figure 6 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the email-enron-only and soc-dolphins graphs. The X and Y axes are the runtime in seconds and the returned cut value, respectively. (a) email-enron-only. (b) soc-dolphins. (a) email-enron-only. (b) soc-dolphins. Figure 6 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the email-enron-only and soc-dolphins graphs. The X and Y axes are the runtime in seconds and the returned cut value, respectively. Figure 6 Plots depicting the computation time and returned cut values of algorithms excluding the SDP on the email-enron-only and soc-dolphins graphs. The X and Y axes are the runtime in seconds and the returned cut value, respectively. 19:13 R. Mirka and D. P. Williamson In particular, perhaps it is more useful to draw a fixed number of t2 values but finish the algorithm’s entire partitioning instead of estimating at that point in time. The magnitude by which the spectral algorithms are faster than the SDP allows this to be a reasonable option. It is also worth noting the performances of the simple spectral and sweep cuts algorithms. Particularly for large graphs, these two algorithms along with the greedy algorithm are much faster than even Trevisan’s algorithm, with the simple spectral almost always being several times faster than greedy (and sweep cuts being slightly slower than greedy). It is known that the greedy algorithm has a .5 approximation guarantee, but to the best of our knowledge, there is no known approximation guarantee for the simple spectral or sweep cuts algorithms. This raises the question of whether any approximation guarantee can be proven for either of these algorithms. A desired guarantee would be greater than greedy’s .5; given the performance results presented here, it seems possible that this is achievable. Relatedly, there is no indication that Soto’s .614 approximation guarantee for Trevisan’s algorithm is tight. It is clear that the algorithm often far surpasses this in practice. Can the analysis of this algorithm be improved? References 1 Francisco Barahona, Martin Grötschel, Michael Jünger, and Gerhard Reinelt. An application of combinatorial optimization to statistical physics and circuit layout design. Operations Research, 36(3):493–513, 1988. doi:10.1287/opre.36.3.493. 2 Jonathan W. Berry and Mark K. Goldberg. Path optimization for graph partitioning problems. Discrete Appl. Math., 90(1–3):27–50, January 1999. doi:10.1016/S0166-218X(98)00084-5. 3 Alberto Bertoni, Paola Campadelli, and Giuliano Grossi. An approximation algorithm for the maximum cut problem and its experimental analysis. Discrete Applied Mathematics, 110:3–12, 2001. doi:10.1016/S0166-218X(00)00299-7. 4 F. Della Croce, M.J. Kaminski, and V.Th. Paschos. An exact algorithm for MAX-CUT in sparse graphs. Operations Research Letters, 35(3):403–408, 2007. doi:10.1016/j.orl.2006.04.001. 5 Oliver Dolezal, Thomas Hofmeister, and Hanno Lefmann. A comparison of approximation algorithms for the maxcut-problem, May 2000. doi:10.17877/DE290R-5013. 6 Iain Dunning, Swati Gupta, and John Silberholz. What works best when? A systematic evaluation of heuristics for max-cut and QUBO. INFORMS Journal on Computing, 30(3):608– 624, 2018. doi:10.1287/ijoc.2017.0798. 7 Michel X. Goemans and David P. Williamson. Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming. J. ACM, 42(6):1115– 1145, November 1995. doi:10.1145/227683.227684. 8 Alexander Golovnev. New upper bounds for MAX-2-SAT and MAX-2-CSP w.r.t. the average variable degree. In Dániel Marx and Peter Rossmanith, editors, Parameterized and Exact Computation, pages 106–117, Berlin, Heidelberg, 2012. Springer Berlin Heidelberg. 9 F. Hadlock. Finding a maximum cut of a planar graph in polynomial time. SIAM Journal on Computing, 4(3):221–225, 1975. doi:10.1137/0204019. 10 Timotej Hrga, Borut Lužar, Janez Povh, and Angelika Wiegele. BiqBin: Moving boundaries for NP-hard problems by HPC. In Ivan Dimov and Stefka Fidanova, editors, Advances in High Performance Computing, pages 327–339, Cham, 2021. Springer International Publishing. 11 Timotej Hrga and Janez Povh. MADAM: A parallel exact solver for max-cut based on semidefinite programming and ADMM. Comput. Optim. Appl., 80(2):347–375, November 2021. doi:10.1007/s10589-021-00310-6. 12 Richard Karp. Reducibility among combinatorial problems. In Complexity of Computer Computations, volume 40, pages 85–103, January 1972. doi:10.1007/978-3-540-68279-0_8. SEA 2022 19:14 Exp Eval of SDP and Spectral Algs for Max Cut 19:14 13 Subhash Khot, Guy Kindler, Elchanan Mossel, and Ryan O’Donnell. Optimal inapproximability results for MAX-CUT and other 2-variable CSPs? SIAM J. Comput., 37(1):319–357, April 2007. doi:10.1137/S0097539705447372.i 14 Nathan Krislock, Jérôme Malick, and Frédéric Roupin. BiqCrunch: A semidefinite branch- and-bound method for solving binary quadratic problems. ACM Trans. Math. Softw., 43(4), January 2017. doi:10.1145/3005345. 15 G.I. Orlova and Ya. G. Dorfman. Finding the maximal cut in a graph. Engineering Cybernetics, 10(3):502–506, 1972. 16 Jan Poland and Thomas Zeugmann. Clustering pairwise distances with missing data: Maximum cuts versus normalized cuts. In Ljupčo Todorovski, Nada Lavrač, and Klaus P. Jantke, editors, Discovery Science, pages 197–208, Berlin, Heidelberg, 2006. Springer Berlin Heidelberg. 17 Svatopluk Poljak and Zsolt Tuza. Maximum cuts and largest bipartite subgraphs. In DIMACS Series in Discrete Mathematics and Theoretical Computer Science, volume 20, pages 181–244, 1995. doi:10.1090/dimacs/020/04. 18 Gerhard Reinelt. TSPLIB – A traveling salesman problem library. ORSA Journal on Computing, 3(4):376–384, 1991. 19 Franz Rendl, Giovanni Rinaldi, and Angelika Wiegele. Solving max-cut to optimality by intersecting semidefinite and polyhedral relaxations. Mathematical Programming, 121:307–335, February 2010. doi:10.1007/s10107-008-0235-8. 20 Ryan A. Rossi and Nesreen K. Ahmed. The network data repository with interactive graph analytics and visualization. In AAAI, 2015. URL: https://networkrepository.com.i 21 Sartaj Sahni and Teofilo Gonzalez. P-Complete approximation problems. J. ACM, 23(3):555– 565, July 1976. doi:10.1145/321958.321975. 22 José A. Soto. Improved analysis of a max-cut algorithm based on spectral partitioning. SIAM Journal on Discrete Mathematics, 29(1):259–268, 2015. doi:10.1137/14099098X. 22 José A. Soto. Improved analysis of a max-cut algorithm based on spectral partitioning. SIAM Journal on Discrete Mathematics, 29(1):259–268, 2015. doi:10.1137/14099098X. 23 Luca Trevisan. Max cut and the smallest eigenvalue. SIAM Journal on Computing, 41(6):1769– 1786, 2012. doi:10.1137/090773714. 23 Luca Trevisan. Max cut and the smallest eigenvalue. SIAM Journal on Computing, 41(6):1769– 1786, 2012. doi:10.1137/090773714. 24 Ryan Williams. A new algorithm for optimal 2-constraint satisfaction and its implications. Theoretical Computer Science, 348(2):357–365, 2005. Automata, Languages and Programming: Algorithms and Complexity (ICALP-A 2004). doi:10.1016/j.tcs.2005.09.023.
https://openalex.org/W2955015181	https://cyberleninka.ru/article/n/determination-of-student-decision-factors-in-choosing-study-programs-in-the-faculty-of-public-health-at-andalas-university-indonesia/pdf	English	null	DETERMINATION OF STUDENT DECISION FACTORS IN CHOOSING STUDY PROGRAMS IN THE FACULTY OF PUBLIC HEALTH AT ANDALAS UNIVERSITY, INDONESIA	Russian journal of agricultural and socio-economic sciences	2,019	cc-by	8,118	RJOAS, 7(91), July 2019 RJOAS, 7(91), July 2019 DOI 10.18551/rjoas.2019-07.06 ABSTRACT Selection of study programs for students requires several considerations before making a decision in determining the place of study. This study aims to identify the student's decision to choose a study program at the Faculty of Public Health Unand, analyze the factors that influence the student's decision to choose a study program at the Faculty of Public Health Andalas University, and formulate managerial implications related to strategies to increase the number of applicants and those who apply to Unand FKM. This study was conducted on students of Public Health and Nutrition Sciences in 2017 and 2018 who actively carry out lectures with a total sample of 391. The variables of this study are product, people, process, promotion, physical evidence, campus image and employment. The data collection method uses a questionnaire with a Likert scale, analyzed using SEM. The results showed that the factors of product, process and promotion had a significant direct influence on the decision of students to choose a study program at the Faculty of Public Health Unand. Amanah Siti Department of Science in Communication and Community Development, Faculty of Human Ecology, Bogor Agricultural Institute, Indonesia Amanah Siti Department of Science in Communication and Community Development, Faculty of Human Ecology, Bogor Agricultural Institute, Indonesia DETERMINATION OF STUDENT DECISION FACTORS IN CHOOSING STUDY PROGRAMS IN THE FACULTY OF PUBLIC HEALTH AT ANDALAS UNIVERSITY, INDONESIA Armeilia Dessy* Business School, Bogor Agricultural Institute, Indonesia Simanjuntak Megawati Department of Family and Consumer Sciences, Faculty of Human Ecology, Bogor Agricultural Institute, Indonesia Simanjuntak Megawati Department of Family and Consumer Sciences, Faculty of Human Ecology, Bogor Agricultural Institute, Indonesia RJOAS, 7(91), July 2019 images and future work have a significant influence on students choosing to study at the University of Taiwan (Yao-Chuan Tsai (2017). A good image of an organization is an asset that has an impact on consumer perceptions can be seen from the organization's communication and operations in various ways, this has an influence on the decision to choose an accounting department (Risnawati 2012), followed by interest factors, shared decision factors, and the availability of employment opportunities for accounting graduates. The development of the world of education in Indonesia, there are many public universities and private universities that open public health faculties, including STIKES which also establishes public health science and nutrition science study programs that make competition even higher. The faculty and study program must strive to improve the quality and provide the best education for recipients of services, namely students. Students applying for and being accepted at the Unand Faculty of Public Health experienced quite volatile changes from 2016 to 2018. The ratio of the number of applicants and the number received in 2016 was: 1: 30.55, in 2017 amounting to 1: 24.21 and in 2018 is 1: 29.17. This means that student tightness / competition chooses FKM is quite high. This must be improved to get quality students, to produce quality and useful graduates in the labor market. Based on the above problems this study takes the title of the determinants of student decisions in choosing a study program at the Faculty of Public Health Unand. This study aims to identify the student's decision to choose a study program at the Faculty of Public Health Unand, analyze the factors that influence student decisions to choose a study program at the Faculty of Public Health Unand, and formulate recommendations for strategies to increase the number of applicants and those who apply to Unand FKM. KEY WORDS Partial least square, campus image, employment, marketing mix, decision to choose, study program. Fulfilling education needs in Indonesia is one of the nation's goals set out in Law Number 12 of 2012, concerning higher education. Higher education as part of the national education system has a strategic role in educating the nation's life and advancing science and technology by paying attention to and applying the values of humanities and the empowerment of a sustainable nation of Indonesia. The selection of study programs for students requires several considerations before making a decision in determining the place of study later. Engel et al. (1995) states that consumer decisions in choosing products / services are influenced by three things: 1. Internal environment / individual consists of consumer resources (time, money, and attention), motivation and involvement, knowledge, attitude, personality, lifestyle and demography; 2. The external environment consists of cultural factors, social class, personal, family and situation influences; and 3. Psychology consists of information processing, learning, attitude and behavior changes. p g g g There are six important factors when choosing a study program, namely factors related to family / parents, peers, school of origin, individual personality, campus image and employment prospects (Saputro 2017). Opinions from the family and individual interests are the main factors in choosing a place to study (Yamamoto 2006). Product, price, promotion, place, people, process and physical evidence have an influence on the decisions of students choosing the MIPA UNMUL Faculty of Computer Science majors (Merni 2011). Institutional 46 RJOAS, 7(91), July 2019 RJOAS, 7(91), July 2019 The results of his research show that the product has a significant and positive effect on the decision to choose students. This study uses the interview method with a questionnaire. The sample is a student at the Faculty of Economics, University of Bojonegoro with a number of respondents of 100 students. The main objective of students choosing study programs is to obtain products / knowledge that are in accordance with their needs and desires, so that students can be said to enter the product in the form of knowledge they need (Merni 2011). g y ( ) 2. People Factor. The application of better marketing mix elements, especially people, has a significant influence on students' decisions to choose study programs at Covenant University Nigeria (Ogunnaike 2014). The sample is a student of the Business Management Department with 226 students. People's variables in the research conducted by Merni (2011) state that lecturers and administrative staff play an important role in the decision of students to choose majors in higher education. The sample is a student of the Department of Computer Science, Faculty of Mipa Unmul with 195 students. Raharjo et al. (2015) and Fosu at al (2017) identify that people have a positive influence on the decision of students to choose a place to study. The decision to choose a place to study becomes an important thing because it involves the future for the students themselves. 3. Process Factors. Muhyidin (2014) states that process variables have a significant influence on the decision of students to choose a place of study. This study uses descriptive quantitative methods and questionnaires. The sample is new students in 15 PTS in West Java who actively study in the odd semester of the 2013/2014 academic year with a total of 341 students. The process in education services greatly determines the quality of a college graduate. Similar results were found in the Suryani et al. (2013) that the process variable significantly influences the decision of students to choose the place of birth. Ogunnaike (2014), Kurniawati (2013), Handayani (2017) and Raharjo et al. (2016) identifying process variables have a significant influence on students' decision to choose a place of study. g p y 4. Promotion factor. Promotion factor is a way that makes it easier to convey information related to education services. RJOAS, 7(91), July 2019 products, namely: intangibility, inseparability, variability and perishability. Service marketing is different from product marketing. products, namely: intangibility, inseparability, variability and perishability. Service marketing is different from product marketing. The image of higher education as defined by Kotler and Fox (1995) are images, impressions and beliefs that a person has towards an object and object in the form of an organization. The good image given by an organization has become an asset because the image itself is the impact of consumer perceptions that come from communication and operations of the organization in various fields. The determinants of students in choosing study programs are factor product, people, process, promotion, physical evidence of campus image and employment. Following are the results of previous research studies based on these factors: 1. Product Factors. Choosing a study program for students requires some consideration before making a decision. According to Rodic et al. (2016) the product has a positive influence on students' decisions to choose faculties at University Novi Sad, Serbia. This study uses the interview method with a questionnaire. The sample is students in six faculties at this university with 783 students as respondents. This is reinforced by Handayani (2017) who conducted research on marketing mix analysis of students' decision to choose study programs. The results of his research show that the product has a significant and positive effect on the decision to choose students. This study uses the interview method with a questionnaire. The sample is a student at the Faculty of Economics, University of Bojonegoro with a number of respondents of 100 students. The main objective of students choosing study programs is to obtain products / knowledge that are in accordance with their needs and desires, so that students can be said to enter the product in the form of knowledge they need (Merni 2011). 1. Product Factors. Choosing a study program for students requires some consideration before making a decision. According to Rodic et al. (2016) the product has a positive influence on students' decisions to choose faculties at University Novi Sad, Serbia. This study uses the interview method with a questionnaire. The sample is students in six faculties at this university with 783 students as respondents. This is reinforced by Handayani (2017) who conducted research on marketing mix analysis of students' decision to choose study programs. LITERATURE REVIEW According to Kotler (2009) and Engel et al. (1995) decision making on purchasing products / services is influenced by several factors, namely: C C f p y y 1. Cultural Factors. Culture is a value, thought and symbol that influences a person's behavior, attitudes, beliefs and habits; 1. Cultural Factors. Culture is a value, thought and symbol that influences a person's behavior, attitudes, beliefs and habits; 2. Social Factors. Consumer behavior is influenced by social factors such as reference groups, families and the role and status of consumers; g p 3. Personal Factors. Personal factors are individual habits that are influenced by the closest environment in making choices, and then expressed in an action (Purimahua 2005); 4. Psychological factors. Psychological factors are encouragement from someone who influences the selection of something based on the flexibility of the product used, greater desire and ease of use of the product compared to others. Marshall and Jhonston (2010) state that the marketing of traditional marketing mix concepts consists of 7 P, namely:  Product. Product services are all things that can be offered by producers to be considered, requested, sought, bought, used or consumed by the market as meeting the needs or desires of the relevant market;  Price. Price is the amount of money customers pay for a particular product;  Place. Location is a place where services are provided. Location determination includes consideration of how to deliver services to customers where services must be placed;  Promotion. Promotion is an activity carried out by a company or organization to communicate and promote its products to the target market; p p g  People. People are an important element in producing and delivering services;  Physical evidence. Physical Evidence is the physical environment in which services are delivered, and there is an interaction between service providers and consumers;  Process. Process is a procedure, mechanism, and series of service delivery activities. Service marketing is carried out because services have different characteristics from 47 RJOAS, 7(91), July 2019 RJOAS, 7(91), July 2019 2009 to 2012 with a sample of 460 students. Ogunnaike (2014), Merni (2011), Handayani (2017), Raharjo et al. (2016) and Ade (2016) found the same results of research that physical evidence had a significant effect on students' decisions to study. 2009 to 2012 with a sample of 460 students. Ogunnaike (2014), Merni (2011), Handayani (2017), Raharjo et al. (2016) and Ade (2016) found the same results of research that physical evidence had a significant effect on students' decisions to study. p y g y 6. Campus Image Factors. Campus image has a big role in influencing decision making for students. Yao-Chuan Tsai (2017) states that institutional images have a significant influence on students choosing to study at Taiwan University. This study uses the method of direct interview and questionnaire. The sample is international students studying in Taiwan with 210 students as respondents. The results of this study are reinforced by Fakhri at al (2017) and Saputro (2017), who concluded that there was a significant effect of campus image on the decision of students to choose a place to study. This indicates that even though students do not know the product or brand of a campus, students can choose through the image of the campus. 6. Campus Image Factors. Campus image has a big role in influencing decision making for students. Yao-Chuan Tsai (2017) states that institutional images have a significant influence on students choosing to study at Taiwan University. This study uses the method of direct interview and questionnaire. The sample is international students studying in Taiwan with 210 students as respondents. The results of this study are reinforced by Fakhri at al (2017) and Saputro (2017), who concluded that there was a significant effect of campus image on the decision of students to choose a place to study. This indicates that even though students do not know the product or brand of a campus, students can choose through the image of the campus. g p 7. Factors of Employment. Employment is one of the main concerns for students in determining the place of study. The results obtained from Risnawati's research (2012) is that there is a significant influence on the employment prospects of the decision to choose students. The sample is a student of accounting major in 2011 STIE Perbanas Surabaya with 435 students. RJOAS, 7(91), July 2019 Yao-Chuan Tsai (2017) future work has a significant influence on students choosing to study at Taiwan University. Likewise, the results of research conducted by Saputro (2007) state that employment has a positive effect on student decisions in choosing a place to study. The sample is a student of the 2015/2016 and 2016/2017 Academic Year Program with a total sample of all students in the academic year. RESEARCH HYPOTHESIS Figure 1 – Framework for thinking This study consists of 7 hypotheses, namely H1: Product has a significant effect on the decision to choose a study program, H2: People has a significant effect on the decision to choose a study program, H3: Process has a significant effect on the decision to choose a study program, H5: Physical evidence has a significant effect on the decision to choose a study program, H6: Campus image has a significant effect on the decision to choose a study program, and H7: Employment has a significant effect on the decision to choose a study program. Employment Student Decision to Choose Campus image Product Physical Evidence People Process Promotion RESEARCH HYPOTHESIS Figure 1 – Framework for thinking Employment Student Decision to Choose Campus image Product Physical Evidence People Process Promotion RESEARCH HYPOTHESIS Employment Campus image Product People Student Decision to Choose Process Promotion Physical Evidence Figure 1 – Framework for thinking This study consists of 7 hypotheses, namely H1: Product has a significant effect on the decision to choose a study program, H2: People has a significant effect on the decision to choose a study program, H3: Process has a significant effect on the decision to choose a study program, H5: Physical evidence has a significant effect on the decision to choose a study program, H6: Campus image has a significant effect on the decision to choose a study program, and H7: Employment has a significant effect on the decision to choose a study program. RJOAS, 7(91), July 2019 Osman (2013) found that promotion had a positive and significant influence on students' decision to choose study programs at Malaysian Universities. The sample is all middle school students with 373 students. Samat et al. (2017) in his research stated that promotion variables have a significant influence on student decisions in choosing a place to study. The sample is students who study in five universities in Palembang City with a total of 110 students. Ade (2016) states that promotion variables in the marketing mix have a significant influence on the decision of students to choose majors in higher education. g 5. Physical evidence factor. Suryani et al. (2013) examined the factors that influence students' decisions in choosing departments in the faculty. His research found that evidence physical variables influence the decision of students to choose a place to study. The sample is students of the Faculty of Economics, Islamic University of North Sumatra, Al Munawaroh, g 5. Physical evidence factor. Suryani et al. (2013) examined the factors that influence students' decisions in choosing departments in the faculty. His research found that evidence physical variables influence the decision of students to choose a place to study. The sample is students of the Faculty of Economics, Islamic University of North Sumatra, Al Munawaroh, 48 RESULTS AND DISCUSSION Decision Making Process. The results showed that the highest consideration in choosing the Unand Faculty of Public Health was the 60% offered and 13% reputation offered. With this situation, FKM Unand must improve its existence, especially related to the field of science offered and reputation. Rudhumbu (2017) also found that study programs and locations were the highest factors in choosing universities in Botswana. The most widely used source of information is through mass media (websites, brochures, newspapers) with a percentage of 28%, a nuclear family of 19% and teachers (teachers in schools, at tutoring places) of 19%. This shows that respondents are easier to get information about FKM Unand, especially through mass media, nuclear families and teachers. Osman (2013) in his research also found that students choosing institutions were mainly promoted through radio and television in Malaysia. The giver of influence in choosing the Unand Faculty of Public Health showed the highest results were obtained from the students' own interest of 61%, while the influence of the family (parents, siblings) was 18% and teachers (teachers in the school, teachers on the guidance place) were 7%. Family opinion and student interests are the most important factors that influence student choice (Yamamoto 2006). Description of Variable Indicators. The results showed that the availability of expertise and skills in accordance with employment became an attraction for respondents in choosing FKM Unand study program, which amounted to 60.4%, while for science that was in accordance with needs / desires 57.3% and knowledge that made it easy to get a job after graduating 52.7%. More can be seen in Table 1. This illustrates that respondents really need expertise, skills and knowledge in accordance with their needs and desires to compete in the world of work. Merni (2011) also suggested that students enter the computer science department of F.MIPA UNMUL to obtain products / knowledge according to their needs and desires. Respondents want competent lecturers in the field of science to be taught at 70.8% and administrative staff in providing good services to students by 33.2%. More can be seen in Table 1. The results of this Unand FKM need to improve the knowledge and skills of lecturers and administrative staff in order to produce qualified and competent graduates. Internal environmental factors including people are important for students in choosing majors at FEB Undiksha (Ayu 2014). RJOAS, 7(91), July 2019 promotion, physical evidence, campus image and employment and student decisions to choose study programs. As for the sample of all S1 students who are actively studying at the Unand Faculty of Public Health in 2017 and 2018, there are 391 active students. The sampling technique is carried out by census or saturated samples. Saturated census or sample is a sampling technique if all members of the population are used as a sample (Sugiyono 2015). Data and information were collected by distributing closed questionnaires to respondents, filling out questionnaires conducted self-administered by being guided directly by researchers. This study used descriptive analysis with SPSS24 and SEM-PLS analysis. METHODS OF RESEARCH The study was conducted at the Faculty of Public Health of Andalas University, with student research subjects in 2017 and 2018 who actively conducted lectures. Research data collection was conducted from February to March 2019. The sources and types of data used were primary data. Primary data used by researchers include product, people, process, 49 RJOAS, 7(91), July 2019 RESULTS AND DISCUSSION The completeness of the media and teaching and learning facilities in PTS is a benchmark for students in making a decision (Muhyidin 2014). Having an accredited BAN-PT / LAM-PTKES status is the most important thing for respondents to choose a study program of 80.3%. For indicators of good quality / quality of education in higher education, having a proud academic achievement in higher education and an IKM / Nutrition study program is the favorite choice to be the next thing chosen by respondents. More can be seen in Table 1. This shows the accredited status of a campus is very necessary to get jobs for the respondents later. Risnawati (2012) also noted that the accredited status of study programs was the main thing for students at STIE Perbanas Surabaya. Table 1 – Percentage of student assessment based on product indicators, people, process, promotion, physical evidence, campus image, and employment on student decisions to choose study programs Table 1 – Percentage of student assessment based on product indicators, people, process, promotion, physical evidence, campus image, and employment on student decisions to choose study programs No Indicator TS N S Indicator Product 1 Having knowledge in accordance with needs / desires 6,4 36,3 57,3 2 Expertise and skills in accordance with existing employment 2,8 36,8 60,4 3 Providing knowledge that makes it easy to get a job after graduation 4,6 42,7 52,7 Indicator People 4 Have competent lecturers. 4,4 24,8 70,8 5 Have administrative staff providing good service. 27,2 39,6 33,2 Indicator Process 6 Having student activities increases skill 2,0 22,3 75,7 7 Produce graduates on time 3,6 39,4 57,0 8 Providing curriculum / syllabus / lecture methods 10,0 28,1 61,9 9 Produce competent graduates 7,7 56,3 36,0 Indicator Promotion 10 Actively carry out promotions through brochures / leaflets 22,7 53,5 23,8 11 Online promotion through internet / website 16,6 52,4 30,9 12 Dissemination to schools 13,8 46,8 39,4 Indicator Physical evidence 13 Laboratory facilities 33,3 41,7 25,0 14 Have books in the library 18,7 43,2 38,1 15 Has a building (lecture room, practice room and prayer room) 44,5 31,2 24,3 16 Has media (In focus / sound system) 7,9 31,7 60,3 17 It has very good IT (internet) facilities 14,9 39,6 45,5 Indicator Campus Image 18 Have academic achievements (research and community service) 3,9 41,4 54,7 19 Have good quality / quality of education in college. RESULTS AND DISCUSSION Lecturers are an important factor in changing students' perceptions of an institution (Osman 2013). p p ( ) In the process variables, more than 50% of students choose student activity indicators to increase the soffskill, provide clear curriculum / syllabus / lecture methods and produce graduates on time, while indicators produce competent graduates with a short work waiting period of under 50%. More can be seen in Table 1. This shows tense when graduates get a job long enough, because the cooperation carried out by FKM with companies and hospitals has not been well established, as is the alumni network that has not been well organized. Muhyidin (2014) also argues that the better the process will affect the increasing decision of students to choose certain universities. Promotional activities through dissemination to schools are the main choices of respondents, amounting to 39.4%, promotion via the internet / website by 30.9% and promotion through brochures / leaflets of 23.8%. More can be seen in Table 1. This shows 50 RJOAS, 7(91), July 2019 that promotions carried out through socialization to schools will quickly arrive or be accepted by prospective students. Likewise through the internet / website because the age range of 17-19 is very active in using the internet to obtain the information they need. Promoting school visits led by alumni from the school will be the main attraction for prospective students (Ivy 2008). that promotions carried out through socialization to schools will quickly arrive or be accepted by prospective students. Likewise through the internet / website because the age range of 17-19 is very active in using the internet to obtain the information they need. Promoting school visits led by alumni from the school will be the main attraction for prospective students (Ivy 2008). ( y ) Indicators have a very complete media (infocus / sound system) teaching and learning being the highest indicator chosen by respondents at 60.3%, while indicators have very good IT facilities, have complete books in the library, have very good laboratory facilities complete and have a very adequate building below 60.3%. More can be seen in Table 1. This shows the supporting facilities for teaching and learning activities in Unand FKM are the availability of media (infocus and sound systems) and internet networks. For other facilities University facilities are used together. RJOAS, 7(91), July 2019 Alumni indicators have the highest chance to work in companies and hospitals having the highest score of 57%, compared to the other three indicators, namely indicators of getting jobs that are in accordance with the field of specialization and easy to find work both in hotels / offices / schools / consultants. More can be seen in Table 1. This shows that the employment opportunities obtained by graduates are not necessarily in accordance with the knowledge they have because currently the market share not only sees the knowledge they have mastered but also the skills they have. Job opportunities are important for students in Taiwan in choosing a place to study (Yao-Chuan Tsai 2017). Decision to Choose a Study Program. In this case, the highest choice of respondents choosing study programs is because the desired specialization is 60.1%. Furthermore, the ease of getting a job, campus reputation and invitation from friends / relatives / acquaintances. More can be seen in Table 1. This explains that the respondents chose the study program because the field of specialization was desired. The field of knowledge desired or offered is an attraction for Bogor City Senior High Schools to choose IPB (Haryanti 2016, Solikhah 2016). ) Analysis of Measurement Model Compatibility Test. Based on the convergent validity test, which is the correlation value between indicators and latent variables, the loading factor and AVE values of each indicator from a latent variable are more than 0.5, but there are two indicators with a loading factor of less than 0.5, the X5.5 indicator on the physical evidence variable and indicator Y4 on the decision decision variable, the indicator must be discarded and re-analyzed. The AVE analysis results obtained AVE values from the process variable indicator, X3.1, less than 0.5, this indicator must be discarded and re-analyzed. The results of confirmatory analysis after removing invalid indicators can be seen in Figure 2. Figure 2 – SEM Results For discriminant validity test, how far the difference in the value of the validity of a variable when compared with other variables, by comparing the value of the square root AVE and the value of correlation between verbs. Obtained AVE square root values of all variables and indicators greater than 0.7 and comparison of AVE root values greater than the correlation value between other variables. RESULTS AND DISCUSSION 2,0 32,2 65,7 20 The Unand IKM / Nutrition study program is a favorite choice 10,0 46,5 43,4 21 Has an accredited BAN-PT / LAM-KES status 1,6 18,2 80,3 Indicator Employment 22 Alumni have the opportunity to work in a company / hospital 3,9 38,9 57,3 23 Graduates get jobs in accordance with the field of specialization 4,9 55,5 39,6 24 Graduates easily get jobs 6,9 56,0 37,0 Indicator Decision to Choose a Study Program 25 In accordance with the field of specialization 10,3 29,7 60,1 26 The ease of getting a job 9,5 46,8 43,7 27 Campus reputation 25,0 39,6 35,4 28 Invite friends / relatives / acquaintances 65,5 17,4 17,1 Description: TS = disagree; N = neutral; S = agree. Description: TS = disagree; N = neutral; S = agree. 51 RJOAS, 7(91), July 2019 RJOAS, 7(91), July 2019 The reliability values of each latent variable, namely product, people, process, promotion, physical evedent, campus image, employment and student decisions have values above 0.7. Thus all latent variables have good composite reliability. The results of the processed data are complete in Table 2. Table 2 – Value of research variable reliability Table 2 – Value of research variable reliability Variable Composite reliability Product 0,850 People 0,734 Prices 0,805 Promotion 0,844 Physical Evidence 0,823 Campus Image 0,821 Employment 0,829 Student decisions 0,759 Testing the hypothesis in this study can be seen from the magnitude of the critical value of the statistical t-test. The hypothesis is said to be significant at t-value> 1.96. Based on the results of hypothesis testing in Table 3, three variables were obtained which had a significant positive direct effect. Table 3 – Hypothesis test results Effects of Variable T Statistics Description Campus image -> Student decisions 1,953 Not significant Employment -> Student decision 0,286 Not significant People -> Student decision 0,926 Not significant Physical evidence-> Student decision 0,441 Not significant Process-> Student decision 2,113 Significant Product -> Student decision 7,386 Significant Promotion -> Student decision 3,384 Significant Table 3 – Hypothesis test results Table 3 – Hypothesis test results Based on data processing, it is known that the R² value of the student's decision to choose a study program at FKM Unand is 0.388, meaning that the latent variable decision to choose students can be explained by latent product variables, people, process, physical evidence, campus image and employment at 38.8%, the remainder is explained by other variables not examined in this study. These variables are price variables, place, interests, internal environment and external environment. Product Effect on Choosing Decisions. The one hypothesis states that product has a significant effect on the decision to choose a study program. Based on Table 3, SEM results show the value of t count at 7.386. The value of t count is greater than the value of t table so that the product has a positive and significant effect on the decision to choose a study program. The results of this study are reinforced by previous studies conducted by Rodic et al. (2016) which states that there is a positive and significant relationship between the product and the decision to choose a study program in the faculty of the University of Novi Sad. RJOAS, 7(91), July 2019 Likewise the cross loading value between variables with indicators is greater than the correlation value of other latent variables with the indicator variable. It can be interpreted that all latent variables in this study have good discriminant validity. Figure 2 – SEM Results Figure 2 – SEM Results For discriminant validity test, how far the difference in the value of the validity of a variable when compared with other variables, by comparing the value of the square root AVE and the value of correlation between verbs. Obtained AVE square root values of all variables and indicators greater than 0.7 and comparison of AVE root values greater than the correlation value between other variables. Likewise the cross loading value between variables with indicators is greater than the correlation value of other latent variables with the indicator variable. It can be interpreted that all latent variables in this study have good discriminant validity. For discriminant validity test, how far the difference in the value of the validity of a variable when compared with other variables, by comparing the value of the square root AVE and the value of correlation between verbs. Obtained AVE square root values of all variables and indicators greater than 0.7 and comparison of AVE root values greater than the correlation value between other variables. Likewise the cross loading value between variables with indicators is greater than the correlation value of other latent variables with the indicator variable. It can be interpreted that all latent variables in this study have good discriminant validity. For discriminant validity test, how far the difference in the value of the validity of a variable when compared with other variables, by comparing the value of the square root AVE and the value of correlation between verbs. Obtained AVE square root values of all variables and indicators greater than 0.7 and comparison of AVE root values greater than the correlation value between other variables. Likewise the cross loading value between variables with indicators is greater than the correlation value of other latent variables with the indicator variable. It can be interpreted that all latent variables in this study have good discriminant validity. 52 RJOAS, 7(91), July 2019 Composite reliability test to measure the consistency and accuracy of internal measuring instruments that have a standard value above 0.7. RJOAS, 7(91), July 2019 (2014) which found that the application of marketing mix elements, especially people, had a significant influence on the students' decision to choose study programs at Covenant University Nigeria. Osman (2013), Raharjo et al. (2015), Handayani (2017), also has different results, where people have a positive and significant influence on the decision of students to choose a study program in college. The results of this study indicate that students are not too influential in the presence of a lecturer or educational staff in the education process. Currently lectures can be done through e-learning, where this is one factor of the student's decision. Effect of Process on Decision to Choose. The third hypothesis states that the process has a significant effect on the decision to choose a study program. Based on Table 3, SEM results show a value of t count of 2.113. The value of t count is greater than the value of t table so that the product has a positive and significant effect on the decision to choose a study program. The results of this study are in line with the research conducted by Suryani et al. (2013), Kurniawati (2013) and Handayani (2017) which states that the process has a significant effect on the decision to choose students. Likewise the research conducted by Muhyidin (2014) states that the process has a significant effect on the decision of students to choose a place to study. The process in education services greatly determines the quality of a college graduate. g g Effect of Promotion on Decision to Choose. The fourth hypothesis states that promotion has a significant effect on the decision to choose a study program. Based on Table 3, SEM results show the value of t count of 3.384. The value of t count is greater than the value of t table so that the product has a positive and significant effect on the decision to choose a study program. The results of this study are reinforced by previous research conducted by Osman (2013) which states that promotion has a positive and significant influence on the decision of students to choose study programs at universities in Malaysia. Rudhumbu (2017) and Ivy (2008) state that promotion strategies carried out by universities have a significant effect on students' decisions about university choices. RJOAS, 7(91), July 2019 Other research shows the same results conducted by Hestiningtyas (2015), Merni (2011) which states that the products offered by universities have a positive and significant effect on the decision of students to choose study programs. Rahardjo et al (2015) concluded that there was a positive and significant influence between the products on student motivation in choosing the LP3I Banjarmasin Business College. This means that the knowledge and skills offered by universities are a consideration for students in deciding to choose a place to study. p y Effects of People on Decisions to Choose. The second hypothesis states that people have a significant effect on the decision to choose a study program. Based on Table 3, SEM results show the value of t count is 0.926 smaller than the value of t table so that people have a negative effect and are not significant towards the decision to choose a study program. The results of this study are different from the results of the Ogunnaike study 53 RJOAS, 7(91), July 2019 RJOAS, 7(91), July 2019 Influence of Employment on Decision to Choose. The seventh hypothesis states that employment has a significant effect on the decision to choose a study program. Based on Table 3, SEM results show the value of t count of 0.286 smaller than the value of t table so that employment has a positive and not significant effect on the decision to choose a study program. The results of this study are not in line with the research conducted by Yao-Chuan Tsai (2017) suggesting that future work has a significant influence on students choosing to study at Taiwan University. The same results were also found in Risnawati (2012) and Saputro (2007) employment opportunities that had a positive and significant effect on students' decisions in choosing a place to study. Managerial Implications for Institutions: 1. The results of this study found that the product variable contributed the most to the students' decisions in choosing study programs at FKM Unand. For this reason, good and organized management must be carried out on study programs (IKM and Nutrition), so that the knowledge, skills and expertise possessed are able to compete and meet the needs of the labor market; 2. The health department is very popular with the community. To get information related to public health schools can get it through promotions carried out by the school. Variable promotion which influences students to choose study programs, through this, FKM Unand can promote excellence in the field of expertise possessed by IKM and Nutrition by involving various social media networks, with the aim of getting students who really have the basic knowledge and talent so that the quality of graduates will be better; ; 3. The educational process carried out by higher education determines the quality of graduates. The process of teaching and learning carried out by FKM Unand has been well seen from the graduates produced not much different when registering. Different and increasing results are needed for the quality of later graduates, such as creating a harmonious academic atmosphere between lecturers and students. Providing and explaining syllabus / method of lectures before the teaching and learning process begins to give direction to students. Alumni network strengthened and collaborated with hospitals and private companies in the health sector related to graduate placement. RJOAS, 7(91), July 2019 Rahardjo et al (2015) found that there was a positive and significant influence between promotions on the motivation of students in choosing the LP3I Banjarmasin Business College. g j g Effect of Physical Evidence on Decision to Choose. The fifth hypothesis states that physical evidence has a significant effect on the decision to choose a stud program. Based on Table 3, SEM results show a value of t count of 0.441 smaller than the value of t table so that physical evidence has a negative effect and is not significant for the decision to choose a study program. The results of this study are not in line with the research conducted by Suryani et al. (2013) shows that physical evidence has a significant influence on the decision to choose students in the Faculty of Economics, Islamic University of North Sumatra Al Munawaroh. Ogunnaike (2014), Merni (2011), Handayani (2017) also found different results, where physical evidence had a significant effect on students' decisions to study. The results of this study indicate that students are not too concerned about buildings, libraries, laboratories because this can be done in a shared lecture hall provided by the university. The center of the laboratory and university library is adjacent to the faculty. Effect of Campus Image on Decision to Choose. The sixth hypothesis states that the campus image has a significant effect on the decision to choose a study program. Based on Table 3, SEM results show a t value of 1.953 smaller than the value of t table so that the campus image has a positive and not significant effect on the decision to choose a study program. Several studies have different results from this study, namely research conducted by Saputro (2017) and Susilowati (2008) suggests that there is a significant effect of campus image on the decision of students to choose a place to study. The same research was also conducted by Risnawati (2012) who found that campus image had a significant effect on the decision to choose students at STIE Perbanas Surabaya. The brand image owned by IPB has an influence on the interest of students choosing IPB (Solikhah 2016). This is because FKM is part of a university that already has A accreditation, brand image and big name in the eyes of the public, making respondents no longer pay attention to the campus image in the faculty. 54 1. Ayu Juli AD. 2014. Factors Affecting Student Decisions in Choosing Departments at the Faculty of Economics and Business (Feb) University of Education Ganesha (Undiksha) as a Lecture Place. Department of Management 2 (1): 1-8. CONCLUSION AND SUGGESTIONS Students decide to choose a study program at FKM Unand because of the ease of getting a job, the field of specialization offered in accordance with their wants and needs and FKM has a good campus reputation. Product, process and promotion have the greatest influence on the decision to choose a study program at FKM Unand and show a positive and significant relationship with the decision to choose. People, physical evidence, campus image and employment have a positive, negative and not significant relationship to student decisions in choosing study programs at Unand FKM. The need for an academic atmosphere to create an optimal teaching and learning process to achieve the mission of FKM Unand as a center for the development of public health human resources through the implementation of quality oriented education, competitiveness and character and professionalism. In making student decision making, it is also influenced by internal and external factors, while in this study it did not analyze the influence of these two factors, so for the next study to analyze the internal and external factors so that the influence of the decision to study program is known. RJOAS, 7(91), July 2019 2. Engel J, Blackwell FRD, Miniard PW. 1995. Consumer Behavior. 2nd volume. Translation from Consumer Behavior. Jakarta (ID): Bina Rupa Aksara. ( ) p 3. [FKM] Faculty of Public Health Unand. 2016. Vision, Mission and work program of the Faculty of Public Health Unand in 2016. Padang (ID): Unand Faculty of Public Health. 4. Handayani A. 2017. Analysis of Marketing Mix against Student Decisions Choosing Study. y 5. Programs at the Faculty of Economics, University of Bojonegoro [thesis]. Surakarta (ID): 6 Muhammadiyah University of Surakarta 5. Programs at the Faculty of Economics, University of Bojonegoro [thesis]. Surakarta (ID) 6. Muhammadiyah University of Surakarta. y y 7. Haryanti. 2016. Analysis of factors that influence the high school students of Bogor city choose Bogor Agricultural Institute [thesis]. Bogor (ID): Bogor Agricultural Institute. 8. Hestiningtyas W. 2015. The effect of the marketing mix on the decisions of students choosing a special field of expertise in the education of the FKS UNS trade system. Proceedings of the National Seminar on Economic & Business Education. ISBN: 978- 602-8580-19-9. http://snpe.fkip.uns.ac.id. 9. Ivy J. 2008. A new higher education marketing mix: the 7ps for MBA marketing. Journal of International Management 22 (4): 288-299. g 10. Jogiyanto HM. 2011. Concepts and Applications of Structural Equation Modeling: Variant Based in Business Research. Yogyakarta (ID): UPP STIM YKPN. gy ( ) 11. [Kemenristekdikti] Ministry of Research, Technology and Higher Education. 2012. Minister of Technology Research and Higher Education Regulation (Permen) Number 12 of 2012 concerning higher education. Jakarta (ID): Kemenristekdikti. g g ( ) 12. Kurniawati D. 2013. The effect of marketing mix strategies on student decisions in choosing Widya Mandala Catholic University Madiun. Widya Warta Journal No. 01 Year XXXV II / January 2013: 65-80. y 13. Merni IP. 2011. Analysis of the factors that influence student decision (Decision Making) choose the computer science department of the Faculty of MIPA Mulawarman University Using SPSS Software. Mulawarman Informatics Journal 6 (2): 86-94. 14. Muhyidin U. 2014. The effect of the marketing mix of educational services on students' decisions in choosing private universities in West Java. Proceedings of the multi- disciplinary national seminar & call for papers at UNISBANK. ISBN: 978-979-3649-81-8. p y p p 15. Ogunnaike O O. 2014. Empirical analysis of marketing mix strategy and student loyalty in education marketing. Mediterranean Journal of Social Sciences 5 (23): 2039-2117. 16. Osman M Zain. 2013. REFERENCES 1. Ayu Juli AD. 2014. Factors Affecting Student Decisions in Choosing Departments at the Faculty of Economics and Business (Feb) University of Education Ganesha (Undiksha) as a Lecture Place. Department of Management 2 (1): 1-8. 55 RJOAS, 7(91), July 2019 Factors influencing students' decisions in choosing private institutions of higher education in Malaysia: Asian Academy of Management Journal 18 (1): 75–90. 17. Purimahua. 2005. Factors that influence the behavior of students in choosing the Department of Development Economics at the Faculty of Economics, Maluku Christian University in Ambon. Journal of Consumer Behavior. Th. IX. No. 2. May: 541-551. 18. Rahardjo Sis, Firdaus, Reza M. 2015. Analysis of the effect of the marketing mix on socio-economic status and motivation on students' decisions in choosing courses (Study at the Education Institutions of Banjarmasin Business College). Journal of Management Insights 3 (1): 45-59. g ( ) 19. Risnawati E. 2012. Factor analysis of student decision-making to choose an accounting department at STIE Perbanas Surabaya. Journal of The Indonesia Accounting Review 2 (1): 63-72. doi: 10.14414 / tiar.v2i01.318. g ( ) 19. Risnawati E. 2012. Factor analysis of student decision-making to choose an accounting department at STIE Perbanas Surabaya. Journal of The Indonesia Accounting Review 2 (1): 63-72. doi: 10.14414 / tiar.v2i01.318. ( ) 20. Rodic V, Nemanja. 2016. Application of marketing mix concept in student recruitment strategies: Evidence from University of Novi Sad. (SB). Scientific original paper 13 (3): 183-202. ( ) 20. Rodic V, Nemanja. 2016. Application of marketing mix concept in student recruitment strategies: Evidence from University of Novi Sad. (SB). Scientific original paper 13 (3): 183-202. 21. Rudhumbu N. 2017. A choice of students of auniversity: a case af Botho University in Botswana. International Journal of Learning and Development 7 (2): 27-37. 21. Rudhumbu N. 2017. A choice of students of auniversity: a case af Botho University in Botswana. International Journal of Learning and Development 7 (2): 27-37. 22. Samat, Marnisah L. 2017. Marketing mix to the decision of students to choose private universities in the city of Palembang. Management journal and Sriwijaya Business 15 (4): 221-229. 22. Samat, Marnisah L. 2017. Marketing mix to the decision of students to choose private universities in the city of Palembang. Management journal and Sriwijaya Business 15 (4): 221-229. 23. Saputro M. 2017. Analisis faktor-faktor yang mempengaruhi keputusan mahasiswa dalam memilih program studi. Jurnal Pendidikan Informatika and Sains 6 (1): 83-94. 23. Saputro M. 2017. Analisis faktor-faktor yang mempengaruhi keputusan mahasiswa dalam memilih program studi. Jurnal Pendidikan Informatika and Sains 6 (1): 83-94. 56 RJOAS, 7(91), July 2019 24. Solikhah A. 2016. Pengaruh Kepribadian, motivasi, citra merek and lingkungan terhadap minat siswa SMA memilih Institut Pertanian Bogor. [tesis]. Bogor: Institut Pertanian Bogor. g 5. Sugiyono. 2015. Statistik Nonparametris untuk Peneliti.Yogyakarta (ID): CV. Alfabeta 26. Susilowati T. 2008. Factors Affecting Student Decision Making (Individual Internal Factor Factors, External Environment and Student Communication Strategy of Economic Education Study Program on Decision Making in Choosing the Special Field of Expertise in Office Administration of FKIP UNS). [thesis]. Semarang (ID): Sebelas Maret University 27. Suriyani W, Ginting P. 2013. Factors influencing the decision of students to choose the economics faculty of the Islamic University of North Sumatra, Al Munawaroh, Medan. Modernization Journal 9 (1): 33-48. 28. Widjaja D. 2106. Effects of Service Marketing Mix on Decisions in Choosing Higher Education (Study of UKRIDA & UNTAR Students). Business Management Scientific Journal 16 (2): 119-216. ( ) 29. Yamamoto GT. 2006. University Evaluation-Selection: a Turkish Case. International Journal of Educational Management 20 (7): 559-569. g ( ) 30. Yao-Chuan Tsai. 2017. Factors influencing international students to study at Universities in Taiwan. International Journal for Innovation Education and Research 5 (1): 4-11. 57 57
https://openalex.org/W2469201796	https://europepmc.org/articles/pmc5020840?pdf=render	English	null	Understanding negative feedback from South Asian patients: an experimental vignette study	BMJ open	2,016	cc-by	6,034	Understanding negative feedback from South Asian patients: an experimental vignette study Jenni Burt,1 Gary Abel,2 Natasha Elmore,1 Cathy Lloyd,3 John Benson,4 Lara Sarson,5 Anna Carluccio,5 John Campbell,6 Marc N Elliott,7 Martin Roland1 To cite: Burt J, Abel G, Elmore N, et al. Understanding negative feedback from South Asian patients: an experimental vignette study. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016- 011256 Open Access Open Access Research Strengths and limitations of this study 1Cambridge Centre for Health Services Research, Institute of Public Health, Forvie Site, University of Cambridge School of Clinical Medicine, Cambridge, UK 2University of Exeter Medical School, Exeter, UK 3Faculty of Health and Social Care, The Open University, Milton Keynes, UK 4Primary Care Unit, Institute of Public Health, Forvie Site, University of Cambridge School of Clinical Medicine, Cambridge, UK 5Ipsos MORI, London, UK 6University of Exeter Medical School, Exeter, UK 7RAND Corporation, Santa Monica, California, USA Correspondence to Dr Jenni Burt; jab35@medschl.cam.ac.uk ▪To enable the same vignettes to be viewed by all participants, the study was conducted in English, limiting our ability to understand evaluations by those with low English language proficiency. Main outcome measure: Mean differences in communication score (on a scale of 0–100) between White British and Pakistani participants, estimated from linear regression. Results: Pakistani participants, on average, scored consultations 9.8 points higher than White British participants (95% CI 8.0 to 11.7, p<0.001) when viewing the same consultations. When adjusted for age, gender, deprivation, self-rated health and video, the difference increased to 11.0 points (95% CI 8.5 to 13.6, p<0.001). The largest differences were seen when participants were older (>55) and where communication was scripted to be poor. To cite: Burt J, Abel G, Elmore N, et al. Understanding negative feedback from South Asian patients: an experimental vignette study. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016- 011256 Strengths and limitations of this study Objectives: In many countries, minority ethnic groups report poorer care in patient surveys. This could be because they get worse care or because they respond differently to such surveys. We conducted an experiment to determine whether South Asian people in England rate simulated GP consultations the same or differently from White British people. If these groups rate consultations similarly when viewing identical simulated consultations, it would be more likely that the lower scores reported by minority ethnic groups in real surveys reflect real differences in quality of care. ▪This is the first study to exclusively use a video vignette approach to assess the extent to which ethnic differences in reported patient experience of primary care reflect real differences rather than differences in expectations, perceptions or in the use of scales. ▪Our experimental design enables us to control the content of the consultations being rated by respondents in order to explore how differences in reporting may or may not explain the dispar- ities in minority ethnic experience in real-life surveys. ▸Prepublication history and additional material is available. To view please visit the journal (http://dx.doi.org/ 10.1136/bmjopen-2016- 011256). Design: Experimental vignette study. Trained fieldworkers completed computer-assisted personal interviews during which participants rated 3 video recordings of simulated GP–patient consultations, using 5 communication items from the English GP Patient Survey. Consultations were shown in a random order, selected from a pool of 16. ▪While our in-home face-to-face recruitment approach ensured access to a wide range of respondents, respondents who agreed to partici- pate in this study may differ in a number of unidentified ways from the population as a whole. Received 22 January 2016 Revised 18 May 2016 Accepted 19 May 2016 Setting: Geographically confined areas of ∼130 households (output areas) in England, selected using proportional systematic sampling. Setting: Geographically confined areas of ∼130 households (output areas) in England, selected using proportional systematic sampling. ▪Our study involved face-to-face interviews in which consultations were viewed and rated: this differs from the postal mode of the national GP Patient Survey, and in completing questionnaire items via an interviewer rather than independ- ently, social desirability bias may be an issue. Participants: 564 White British and 564 Pakistani adults recruited using an in-home face-to-face approach. Open Access Pakistani groups providing the lowest scores.9 Around half of the difference in these scores is explained by the concentration of South Asian patients in low-scoring primary care practices.7 The remaining difference cur- rently remains unexplained. groups of participants, we produced a series of 16 vign- ettes for this study. We sought to manipulate the vign- ettes on three key domains: (1) the presenting symptom; (2) the quality of the communication within the consult- ation (poor or good) and (3) the ethnic backgrounds of the doctor and patient (South Asian or White British). Following published recommendations for the produc- tion of vignettes, we based ours on real-life GP–patient consultations recorded as part of another study.14 We undertook an extensive process of script development, roleplaying and rating prior to ﬁlming the vignettes with professional actors (ﬁgure 1). y p Several potential explanations have, however, been proposed for the lower ratings given by South Asian respondents. Broadly, these relate to whether South Asian patients receive lower quality care, or whether they receive similar care, but rate this more negatively.8 For example, differences in the use of questionnaire response scales10 may lead to South Asian groups being less likely to endorse the most positive options when asked to evaluate a doctor’s communication skills. Alternatively, there may be systematic variations in eva- luations of consultations because South Asian respon- dents vary in their expectations of, or preferences for, care. Vignettes covered four different clinical scenarios: per- sistent cough, perforated ear drum, painful elbow and generalised numbness. We devised two different scripts for each scenario: one designed to illustrate poor Understanding why minority ethnic groups often give poorer evaluations of care is critical to forming an effect- ive response in policy and practice. One approach is to use item response theory to explore whether items receive systematically different responses by the ethnic group. METHODS We undertook an experimental vignette study in which videos of simulated GP–patient consultations were shown to two groups of people, who were asked to rate the quality of the communication within each consult- ation. The primary outcome of interest was communica- tion score (on a scale of 0–100). Figure 1 The vignette development process This study aimed to examine whether people from a Pakistani background rate the communication within simulated GP consultations differently than White British people do. If these groups rate simulated consul- tations similarly when viewing identical video vignettes, it is more likely that the lower scores observed among South Asian people in national patient experience surveys reﬂect real differences in quality of communica- tion within consultations. INTRODUCTION Communication between doctors and patients is a core component of patient experience.1 Patients’ evaluations of doctors’ interpersonal skills are widely used in assess- ments of the quality of care, with an increas- ing focus on the public reporting of patient feedback.2 In the USA and the UK, certain minority ethnic groups report lower patient experience scores compared to the majority population.3–8 For example, analysis of the English General Practice Patient Survey found that South Asian groups report par- ticularly low scores compared to the White British majority, with Bangladeshi and Conclusions: Substantial differences in ratings were found between groups, with Pakistani respondents giving higher scores than White British respondents to videos showing the same care. Our findings suggest that the lower scores reported by Pakistani patients in national surveys represent genuinely worse experiences of communication compared to the White British majority. Correspondence to Dr Jenni Burt; jab35@medschl.cam.ac.uk 1 Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 Open Access Open Access Recent analysis of the GP Patient Survey suggests that this is not the case for differences observed between South Asian and White British groups.11 However, a more robust approach to determining whether differ- ences in evaluations of care reﬂect real differences is to ask respondents to rate standardised clinical scenarios (‘vignettes’).12 A US study using primarily written vign- ettes in an online survey concluded that score variations observed on national surveys among African-American, Latino and white respondents were likely to reﬂect true differences in real-life experiences, at least for items using an ‘Always-to-Never’ Response Scale.10 13 Simulated consultations To ensure generalisability and to avoid the chance inclu- sion of a characteristic or event which, unknown to us, might systematically be rated differently by our two Figure 1 The vignette development process. 2 Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 Open Access communication by the doctor, and one designed to illus- trate good communication. Standards of communication were formulated according to the Global Consultation Rating Scale (GCRS), an observer-rated measure of com- munication competence derived from the widely used Calgary-Cambridge guide to the medical interview.15–17 however, recruitment was based on an ‘over 55’ cut point. Fieldworkers were also provided with one or two neighbouring output areas to recruit from if necessary. Snowball recruitment (eg, known neighbours) and add- itional household interviews were allowed. To recruit White British participants, we ﬁrst excluded output areas with low proportions of White British residents (<90%) and residents aged over 50. The remaining output areas were ranked by social grade (the percentage of people who were Social Grade A/B according to 2011 Census data) and geography. Output areas to approach were then selected using proportional systematic sampling. g y g g The GCRS covers 12 domains including ‘initiating the session’, ‘gathering information’, ‘building the relation- ship’ and ‘achieving a shared understanding’. The ‘poor’ and ‘good’ versions of the four clinical scenarios were then used to ﬁlm two sets of vignettes, ﬁrst with White British actors playing the GP and patient, and then with South Asian actors playing the GP and patient. The GP role was acted throughout by either one White British or one South Asian actor; eight different actors (four White British and four South Asian) role- played patients, each participating in one clinical scen- ario. The restriction of vignettes to the same-ethnicity pairings, rather than including mixed pairings, is a func- tion of wishing to introduce some variation to ensure generalisability while keeping the number of vignettes to a manageable number. The ﬁnal 16 videos were each scored by 3 clinical raters using the GCRS to assess com- munication quality in relation to professionally deﬁned norms.15 The ‘poor’ communication vignettes had mean GCRS scores between 0.6 and 2.4 (out of 10) while the ‘good’ communication vignettes mean scores were between 5.1 and 8.4. Data collection We worked with a UK market research company, Ipsos MORI, to collect the data. We aimed to recruit 1120 adult respondents who self-identiﬁed as either Pakistani or White British, across a broad age range. Each respondent was asked to rate three vignettes. Our sample size calculation was based on data from the national GP Patient Survey, as we used the same commu- nication questions for our respondents as are used in this national survey. Inclusion of 560 Pakistani respon- dents and 560 White British respondents gave over 80% power to detect a 3.1-point difference (on a 0–100 scale) seen between these two groups after controlling for age, gender, deprivation, self-rated health and practice. As ethnic disparities are largest in older ages, we aimed to recruit equal numbers above and below the age of 55 within each ethnic group.9 Simulated consultations g p p y p g Potential participants were screened by ﬁeldworkers for ethnicity (using the Ofﬁce for National Statistics 18-group categorisation) and English language compe- tency (using a screening question regarding self- reported conﬁdence in understanding short videos in English). Eligible respondents who consented then com- pleted a computer-assisted personal interview (CAPI) using a standardised script. Each participant viewed three simulated consultation videos. Following each video, the participant was asked to rate the communica- tion within the consultation using ﬁve items taken from the national GP–Patient Survey (table 1). Videos were assigned so each participant saw three different present- ing conditions, with at least one of the videos featuring each of the two ethnic GP/patient pairings, and at least one of the videos scripted to feature each of the two levels of GP–patient communication. The selection of videos shown to each participant was such that approxi- mately equal numbers of all possible combinations were used, given the restrictions described. Participants also completed basic sociodemographic questions (age, self- rated health, whether born in the UK, language spoken most often at home). All interview questions and ratings were completed verbally, with responses recorded by interviewers directly onto the CAPI software. An area- based measure of socioeconomic deprivation (Index of Multiple Deprivation) was calculated. Analysis We scored each participant’s rating of each consultation by linearly scaling the response options between 0 (very poor) and 100 (very good) and averaging all informative answers when at least three of the ﬁve items were com- pleted. Linear regression was used to model the mean difference between White British and Pakistani partici- pants’ ratings of doctor–patient communication. We esti- mated the unadjusted difference in ratings, as well as the difference adjusting for patient age, gender, self- rated health, deprivation and a set of 15 indicator vari- ables for the video. No analysis of interaction terms was originally planned. However, the effect size found was much larger than that anticipated, and so interactions were investigated between participant ethnicity and the following variables: (1) relating to the video: ethnicity of GP/patient and quality of GP–patient communication, and (2) relating to the participant: age, gender and We used different recruitment strategies for the differ- ent ethnic groups. To recruit Pakistani respondents, geo- graphically conﬁned areas in England of ∼130 households (output areas) were selected in which at least 35% of the populations were identiﬁed as Pakistani in 2011 Census data. These were then ranked according to the proportion of the population aged over 50. Trained ﬁeldworkers recruited participants within these areas using an in-home face-to-face approach, starting in the output areas with the highest proportion of residents over the age of 50. Available Census age categories drove our use of ‘over 50’ rather than ‘over 55’ at sampling: 3 Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 Table 1 GP–patient communication items Thinking about the doctor you have just seen in the video, how good was the doctor at: Very good Good Neither good nor poor Poor Very poor Doesn’t apply* Giving enough time……………………………… □ … □ … □ … □ … □ … □ Listening…………………………………………….. □ … □ … □ … □ … □ … □ Explaining tests and treatments…………….. □ … □ … □ … □ … □ … □ Involving in decisions about care …………….. □ … □ … □ … □ … □ … □ Treating with care and concern…………….. □ … □ … □ … □ … □ … □ *Considered to be uninformative for the purposes of our analysis. p Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 RESULTS Participants deprivation. When modelling interactions, only variables for the video attributes were used (rather than using indicator variables for all videos). For interactions involv- ing age, the oldest two age groups were combined and a continuous version of age groups was used in the inter- action term only. CIs and p values were estimated using bootstrapping with 500 replications (given non-normal data), clustered by participants (with each participant supplying three communication scores). A sensitivity analysis that clustered the bootstrap resampling by output area rather than by participants (to account for multiple sampling in households and small geographic areas) made only trivial changes to SEs, so we do not report it here. A total of 1128 participants were recruited, 564 (50%) self-identiﬁed White British and 564 (50%) self- identiﬁed Pakistani participants. The sociodemographic proﬁle of participants appears in table 2. While the sam- pling restriction that half of participants in each group should be aged 55 or above increased the similarity of the groups’ age distribution, Pakistani participants were younger than the White British participants within each age stratum. Pakistani participants were also more likely to be man (58% vs 52%); to be in fair or poor health (38% vs 26%) and to live in the most deprived areas (82% vs 14%). The geographic locations where Table 2 Sociodemographic profile of study participants All White British Pakistani n Per cent n Per cent n Per cent Age (years) 18–24 88 7.8 40 7.1 48 8.5 25–34 154 13.7 56 9.9 98 17.4 35–44 151 13.4 70 12.4 81 14.4 45–54 175 15.5 118 20.9 57 10.1 55–64 267 23.7 94 16.7 173 30.7 65–74 179 15.9 109 19.3 70 12.4 75–84 95 8.4 63 11.2 32 5.7 85 or over 19 1.7 14 2.5 5 0.9 Gender Male 583 51.7 255 45.2 328 58.2 Female 545 48.3 309 54.8 236 41.8 Self-rated health Excellent 132 11.7 82 14.5 50 8.9 Very good 289 25.6 181 32.1 108 19.1 Good 348 30.9 157 27.8 191 33.9 Fair 207 18.4 86 15.2 121 21.5 Poor 152 13.5 58 10.3 94 16.7 Deprivation 1—least deprived 108 9.6 100 17.7 8 1.4 2 137 12.1 137 24.3 0 0.0 3 122 10.8 111 19.7 11 2.0 4 221 19.6 138 24.5 83 14.7 5—most deprived 540 47.9 78 13.8 462 81.9 4 Burt J, et al. BMJ Open 2016;6:e011256. RESULTS Participants doi:10.1136/bmjopen-2016-011256 4 Open Access Figure 2 Geographic locations of the census-based output areas where White British and Pakistani participants were recruited. Open Access he census-based output areas where White British and Pakistani participants were recruited. Figure 2 Geographic locations of the census-based output areas where White British and Pakistani participants were recruited. participants were recruited are shown in ﬁgure 2. While the White British participants were recruited from a wide range of geographic locations, the Pakistani participants were located from a small number of geographically con- ﬁned locations. Between 202 and 220 participants scored each of the video vignettes for GP–patient communication (full details in online supplementary material table S1). communication scores from Pakistani participants were typically higher than those from White British partici- pants. The mean communication score from Pakistani participants was 67.3 out of 100, 9.9 points higher (95% CI 8.0 to 11.7, p<0.001) than the mean score from White British participants (57.4 out of 100). In a regres- sion model (full output shown in online supplementary material table S2) adjusting for participant age, gender, self-rated health, deprivation and video, there was a slightly larger difference between the two ethnicities: 11.0 points (95% CI 8.5 to 13.5, p<0.001). Main results The distribution of communication scores for White British and Pakistani participants is shown in ﬁgure 3. The data are skewed in both groups, with high commu- nication scores given more often; however, the Analysis of interactions As the difference in scores between Pakistani and White British participants was considerably larger than that expected, we were able to explore interactions between ethnicity and other variables. There was no evidence that the difference in scores between Pakistani and White British participants varied by patient gender (p=0.92), deprivation (p=0.68) or by the ethnicity of the doctor/patient pair shown in the videos (p=0.53). There was strong evidence that the difference in scores between Pakistani and White British participants was larger for older participants (p=0.001) and consultations scripted to contain poorer doctor–patient communica- tion (p<0.001). Table 3 shows the mean difference in age by good/poor scripted communication strata, esti- mated from a model containing all main effects, plus (1) ethnicity and age interactions, (2) ethnicity and good/poor communication interactions and (3) the three-way interaction between those variables (p<0.001 Figure 3 Box plots showing the distribution of GP communication scores recorded by White British and Pakistani participants. Figure 3 Box plots showing the distribution of GP communication scores recorded by White British and Pakistani participants. Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 5 Open Access differ from the Pakistani population as a whole: for example, recent research suggests that minority ethnic populations in lower ethnic density areas may report higher satisfaction with healthcare.18 Table 3 Adjusted difference in communication scores for age group by good/poor scripted communication between White British and Pakistani participants Scripted communication Age Good Poor 18–24 −1.31 (−5.38, 2.76) 10.29 (5.00, 15.57) 25–34 −0.15 (−3.58, 3.27) 13.32 (9.10, 17.54) 35–44 1.01 (−1.96, 3.97) 16.34 (12.91, 19.77) 45–54 2.17 (−0.62, 4.95) 19.37 (16.24, 22.50) 55–64 3.33 (0.39, 6.27) 22.40 (18.94, 25.86) 65–74 4.49 (1.11, 7.87) 25.42 (21.16, 29.69) 75 and over 5.65 (1.64, 9.66) 28.45 (23.11, 33.79) A positive difference implies Pakistani patients gave, on average, higher (more favourable) scores. g Our study involved face-to-face interviews in which consultations were viewed and rated: this differs from the postal mode of the national GP Patient Survey. In completing questionnaire items via an interviewer rather than independently, social desirability bias may become an issue. However, the magnitude of social desirability bias would have to be substantially different between White British and Pakistani respondents to have a large impact on our ﬁndings. Analysis of interactions In these ‘poor’ consultations, the difference in scores increases with rising age of participants. For example, ratings of consultations with poor communica- tion are 10.29 points higher (95% CI 5.00 to 15.57) for Pakistani participants aged 18 to 24 than White British participants of the same age. This difference increases to 28.45 points (95% CI 23.11 to 33.79) for over the age of 75 years. Analysis of interactions Additionally, ratings of consulta- tions by ‘analogue patients’ (members of the public asked to rate care received by a third party), such as our participants, are commonly more critical than patients commenting on their own care.19 In our study, across both groups, low scores were used more often than in the national GP Patient Survey: for example, only 2.6% of answers to the GP communication questions in the most recent GP Patient Survey were given as poor or very poor, compared to 26.6% of answers in this study.20 To enable the same vignettes to be viewed by all partici- pants, the study was conducted in English, limiting our ability to understand evaluations by those with low English language proﬁciency (and who might, eg, respond to the GP Patient Survey in other languages). In the USA, ethnic minorities preferring languages other than English generally show response tendencies that are in the same direction as English-preferring members of the same ethnic minority, but to a greater extent, perhaps reﬂecting a continuum of accultur- ation.6 However, it was not possible to produce equiva- lent vignettes in other languages, and as 99.8% of respondents to the GP Patient Survey respond in English, our ability to extrapolate to the wider popula- tion remains high. for three-way interaction). Interactions between present- ing condition and ethnicity were omitted for clarity. The difference between scores given by younger (under 55 years) White British and Pakistani participants to con- sultations containing ‘good’ communication is small and not statistically signiﬁcant. However, larger and statistic- ally signiﬁcant differences are seen for older patients and for consultations portraying ‘poor’ communication at all ages. In these ‘poor’ consultations, the difference in scores increases with rising age of participants. For example, ratings of consultations with poor communica- tion are 10.29 points higher (95% CI 5.00 to 15.57) for Pakistani participants aged 18 to 24 than White British participants of the same age. This difference increases to 28.45 points (95% CI 23.11 to 33.79) for over the age of 75 years. for three-way interaction). Interactions between present- ing condition and ethnicity were omitted for clarity. The difference between scores given by younger (under 55 years) White British and Pakistani participants to con- sultations containing ‘good’ communication is small and not statistically signiﬁcant. However, larger and statistic- ally signiﬁcant differences are seen for older patients and for consultations portraying ‘poor’ communication at all ages. DISCUSSION doi:10.1136/bmjopen-2016-011256 Open Access study builds on previous vignette research by using mul- tiple video vignettes manipulating several key attri- butes.12 13 Video vignettes have so far been little employed in this ﬁeld, in spite of evidence of viewers perceiving them as realistic and enabling immersion in the situation at hand, although well-crafted vignettes are essential to ensure good construct validity.14 In the USA, Weinick et al, reported no evidence of differences among White, African-American and Latino evaluations of doctor–patient communication in vignettes when using an ‘Always-to-Never’ Response Scale; they con- cluded that variations within national surveys on such items for these groups were likely to reﬂect differences in real-life experiences.13 In our study, however, we found substantially more positive ratings by Pakistani in comparison to White British respondents. factors including language barriers, health literacy, provider-side discrimination and system-level failures combine to inhibit good communication within individ- ual consultations. Twitter Follow Jenni Burt at @jenniaburt Acknowledgements The authors thank the members of the public who kindly agreed to participate in this study and without whom the study would not have been possible. The authors thank all those whose contributions were crucial to the set up and conduct of this work, including those involved in the development, acting and filming of the clinical vignettes; the clinical raters who assessed the vignettes and the whole team at Ipsos MORI for their enthusiastic and efficient collection of data. They also thank the Improve Advisory Group for their input and support throughout this study. Finally, they thank the three reviewers whose comments helped to improve the clarity and rigour of this paper. Contributors JBu designed the study, managed the production of the vignettes, oversaw the conduct of the study and drafted and revised the paper. GA designed the study, wrote the statistical analysis plan, conducted the analysis and drafted and revised the paper. NE assisted with the conduct of the study, produced the vignettes and commented on draft versions of the paper. CL contributed to the design of the study and commented on draft versions of the paper. JBe contributed to the design of the study, assisted with the production of the vignettes and commented on draft versions of the paper. LS and AC contributed to the design of the study, oversaw data collection and commented on draft versions of the paper. DISCUSSION JC contributed to the design of the study and commented on draft versions of the paper. MNE contributed to the design of the study, contributed to the statistical analysis plan and revised the paper. MR designed the study, oversaw the conduct of the study and commented on draft versions of the paper. MR is the guarantor. This study was designed to explore whether people from a Pakistani background rate the communication within simulated GP consultations differently from ratings provided by White British people. Similar ratings of simulated consultations from both ethnic groups would have suggested that the low scores observed in national surveys from Pakistani respondents reﬂect real differences in the quality of communication experienced by these patients in comparison to White British patients. The substantially more positive ratings from Pakistani respondents that we observed in our experi- mental study suggest that not only are there differences in the quality of communication in real-life consulta- tions, but also that these differences are even greater than those identiﬁed in real-life surveys, such as those we have previously reported from ﬁndings using the GP Patient Survey.7 9 We suggest that Pakistani patients experience genuinely worse standards of communica- tion. However, while we can be conﬁdent that differ- ences in experience exist, it is difﬁcult to extrapolate our vignette-derived data to estimate the magnitude of difference in real life. Poor communication for these groups may arise from system-level, provider-level and/ or patient-level factors.21 For example, language barriers within consultations may lead to more negative experi- ences of care for doctors and patients.22 23 Levels of acculturation may be linked with a patient’s ability to navigate the healthcare system, with consequent impacts on patient experiences of care.24 Discrimination and bias are sensitive and challenging topics – whether at the level of the system or provider. However, they need to be considered as key contributors to inequalities in care.25 Funding This work was funded by the National Institute for Health Research Programme Grants for Applied Research (NIHR PGfAR) Programme (RP-PG-0608-10050). The study was sponsored by Cambridgeshire and Peterborough Clinical Commissioning Group. Disclaimer The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health. The researchers confirm their independence from the study funders, the National Institute for Health Research. Competing interests All authors have completed the ICMJE uniform disclosure form at http://www.icmje.org/coi_disclosure.pdf. DISCUSSION Our experimental study found that respondents from a Pakistani background rated communication in simulated GP consultations signiﬁcantly more positively than their White British counterparts. These differences were largest for consultations depicting poor doctor–patient communication and for older respondents. The differ- ences we observed were in the opposite direction to those seen repeatedly in the national GP Patient Survey, which relates to a patients’ most recent consultation with a GP. In the national survey, Pakistani respondents give signiﬁcantly lower scores for communication than their White British counterparts. g Previous examinations of inequalities in patient experience between ethnic groups have commonly relied on real-world data such as those generated through surveys, in which it is difﬁcult to distinguish whether differences are attributable to variations in care or variations in the reporting of that care.3–9 Large-scale video recording of actual GP–patient consultations, an external assessment of their communication quality and the comparison of this to reported patient experiences of care would enable us to develop a more robust ‘real- world’ understanding of the drivers of variations in reported experience, but the utility of such an undertak- ing must be balanced against its many challenges. Our experimental design enables us to control the content of the consultations being rated by respondents in order to efﬁciently explore how differences in reporting may explain the disparities in minority ethnic experience in real-life surveys. We chose to focus on communication as this is a key component of quality of care, yet one where certain minority ethnic groups report consistently poor experience of their interactions with clinicians.7–9 The p Our in-home face-to-face recruitment approach ensured access to a wide range of respondents, inde- pendent of the GP practice they were registered with. However, respondents who agreed to participate in this research may differ in a number of unidentiﬁed ways from the population as a whole. For example, to ensure efﬁcient recruitment to the study, we focused our efforts on high-density Pakistani areas, which also have high levels of deprivation (the 82% of participants living in areas in the most deprived quintile compared to 51% nationally). The sampled population may, therefore, 6 Burt J, et al. BMJ Open 2016;6:e011256. DISCUSSION MR, JC, JBe, JBu, GA and NE declare support from the National Institute for Health Research for the submitted work. Ethics approval The Central Office for Research Ethics Committee (COREC) advised that the survey does not require formal medical research ethical approval, but it adheres to the Market Research Society code of ethics. Data sharing statement The data set is available on request from the authors: contact Jenni Burt ( jab35@medschl.cam.ac.uk). Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http:// creativecommons.org/licenses/by/4.0/ 1. Ahmed F, Burt J, Roland M. Measuring patient experience: concepts and methods. Patient 2014;7:235–41. 2. Fung CH, Lim YW, Mattke S, et al. Systematic review: the evidence that publishing patient care performance data improves quality of care. Ann Intern Med 2008;148:111–23. Open Access Open Access 15. Burt J, Abel G, Elmore N, et al. Assessing communication quality of consultations in primary care: initial reliability of the Global Consultation Rating Scale, based on the Calgary-Cambridge Guide to the Medical Interview. BMJ Open 2014;4:e004339. 4. Price RA, Haviland AM, Hambarsoomian K, et al. Do experiences with Medicare managed care vary according to the proportion of same-race/ethnicity/language individuals enrolled in One’s Contract? Health Serv Res 2015;50:1649–87. ; 5. Goldstein E, Elliott MN, Lehrman WG, et al. Racial/ethnic differences in patients’ perceptions of inpatient care using the HCAHPS survey. Med Care Res Rev 2010;67:74–92. 16. Kurtz S, Silverman J, Benson J, et al. Marrying content and process in clinical method teaching: enhancing the Calgary-Cambridge guides. Acad Med 2003;78:802–9. 6. Weech-Maldonado R, Morales LS, Elliott M, et al. Race/ethnicity, language, and patients’ assessments of care in Medicaid managed care. Health Serv Res 2003;38:789–808. g 17. Kurtz SM, Silverman JD. The Calgary-Cambridge Referenced Observation Guides: an aid to defining the curriculum and organizing the teaching in communication training programmes. Med Ed 1996;30:83–9. 7. Lyratzopoulos G, Elliott M, Barbiere JM, et al. Understanding ethnic and other socio-demographic differences in patient experience of primary care: evidence from the English General Practice Patient Survey. BMJ Qual Saf 2012;21:21–9. 18. Bécares L, Das-Munshi J. Ethnic density, health care seeking behaviour and expected discrimination from health services among ethnic minority people in England. Health Place 2013;22:48–55. y 8. Mead N, Roland M. Understanding why some ethnic minority patients evaluate medical care more negatively than White patients: a cross sectional analysis of a routine patient survey in English general practices. BMJ 2009;339:b3450. 19. van Vliet LM, van der Wall E, Albada A, et al. The validity of using analogue patients in practitioner–patient communication research: systematic review and meta-analysis. J Gen Intern Med 2012;27:1528–43. ; 20. NHS England. GP Patient Survey: Survey and Reports 2014. https:// gp-patient.co.uk/surveys-and-reports#july-2014 (accessed 27 Jul 2015). g p 9. Burt J, Lloyd CE, Campbell J, et al. Variations in GP–patient communication by ethnicity, age and gender: evidence from a national patient survey. Br J Gen Pract 2016;66:e47–52. ) 21. Institute of Medicine. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: The National Academic Press, 2003. 10. Elliott MN, Haviland AM, Kanouse DE, et al. Adjusting for subgroup differences in extreme response tendency in ratings of health care: impact on disparity estimates. Health Serv Res 2009;44:542–61. 22. CONCLUSIONS 1. Ahmed F, Burt J, Roland M. Measuring patient experience: concepts and methods. Patient 2014;7:235–41. Our ﬁndings add substantial weight to the likelihood that inequalities affecting South Asian people in national surveys reﬂect systematic variations in the quality of communication within consultations. While there is a body of research into the drivers of inequal- ities in care, future research needs to focus on how 2. Fung CH, Lim YW, Mattke S, et al. Systematic review: the evidence that publishing patient care performance data improves quality of care. Ann Intern Med 2008;148:111–23. 3. Weech-Maldonado R, Elliott MN, Adams JL, et al. Do racial/ethnic disparities in quality and patient experience within Medicare plans generalize across measures and racial/ethnic groups? Health Serv Res 2015;50:1829–49. Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 7 Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256 Open Access Parsons JA, Baker NA, Smith-Gorvie T, et al. To ‘Get by’ or ‘get help’? A qualitative study of physicians’ challenges and dilemmas when patients have limited English proficiency. BMJ Open 2014;4: e004613. ; 11. Setodji CM, Elliott MN, Abel G, et al. Evaluating differential item functioning in the English General Practice Patient Survey: comparison of South Asian and White British subgroups. Med Care 2015;539:809–17. 23. Ahmed F, Abel GA, Lloyd CE, et al. Does the availability of a South Asian language in practices improve reports of doctor–patient communication from South Asian patients? Cross sectional analysis of a national patient survey in English general practices. BMC Fam Pract 2015;16:55. 12. King G, Murray CJ, Salomon JA, et al. Enhancing the validity and cross-cultural comparability of measurement in survey research. Am Polit Sci Rev 2004;98:191–207. 13. Weinick RM, Elliott MN, Volandes AE, et al. Using standardized encounters to understand reported racial/ethnic disparities in patient experiences with care. Health Serv Res 2011;46:491–509. 24. Hasnain M, Schwartz A, Girotti J, et al. Group UICEoCP. Differences in patient-reported experiences of care by race and acculturation status. J Immigr Minor Health 2013;15:517–24. 14. Hillen MA, van Vliet LM, de Haes HCJM, et al. Developing and administering scripted video vignettes for experimental research of patient–provider communication. Patient Educ Couns 2013;91:295–309. g 25. Burgess DJ. Addressing racial healthcare disparities: how can we shift the focus from patients to providers? J Gen Intern Med 2011;26:828–30. 8 Burt J, et al. BMJ Open 2016;6:e011256. doi:10.1136/bmjopen-2016-011256
https://openalex.org/W4380989988	https://www.mdpi.com/1424-8220/23/12/5615/pdf?version=1686827612	English	null	Toward Energy-Efficient Routing of Multiple AGVs with Multi-Agent Reinforcement Learning	Sensors	2,023	cc-by	10,004	sensors sensors sensors sensors Citation: Ye, X.; Deng, Z.; Shi, Y.; Shen, W. Toward Energy-Efﬁcient Routing of Multiple AGVs with Multi-Agent Reinforcement Learning. Sensors 2023, 23, 5615. https:// doi.org/10.3390/s23125615 A d Ed Al Al d d Xianfeng Ye 1 , Zhiyun Deng 1, Yanjun Shi 2 and Weiming Shen 1,* Xianfeng Ye 1 , Zhiyun Deng 1, Yanjun Shi 2 and Weiming Shen 1,* 1 School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China; xfye@hust.edu.cn (X.Y.); dengzy@hust.edu.cn (Z.D.) 2 Department of Mechanical Engineering, Dalian University of Technology, Dalian 116023, China; syj@dlut.edu.cn * Correspondence: shenwm@hust.edu.cn 1 School of Mechanical Science and Engineering, Huazhong University of Science and Technology Wuhan 430074, China; xfye@hust.edu.cn (X.Y.); dengzy@hust.edu.cn (Z.D.) yj * Correspondence: shenwm@hust.edu.cn Abstract: This paper presents a multi-agent reinforcement learning (MARL) algorithm to address the scheduling and routing problems of multiple automated guided vehicles (AGVs), with the goal of minimizing overall energy consumption. The proposed algorithm is developed based on the multi-agent deep deterministic policy gradient (MADDPG) algorithm, with modiﬁcations made to the action and state space to ﬁt the setting of AGV activities. While previous studies overlooked the energy efﬁciency of AGVs, this paper develops a well-designed reward function that helps to optimize the overall energy consumption required to fulﬁll all tasks. Moreover, we incorporate the ϵ-greedy exploration strategy into the proposed algorithm to balance exploration and exploitation during training, which helps it converge faster and achieve better performance. The proposed MARL algorithm is equipped with carefully selected parameters that aid in avoiding obstacles, speeding up path planning, and achieving minimal energy consumption. To demonstrate the effectiveness of the proposed algorithm, three types of numerical experiments including the ϵ-greedy MADDPG, MADDPG, and Q-Learning methods were conducted. The results show that the proposed algorithm can effectively solve the multi-AGV task assignment and path planning problems, and the energy consumption results show that the planned routes can effectively improve energy efﬁciency. Keywords: automated guided vehicles; multi-agent reinforcement learning; task assignment; path planning; energy consumption Article Toward Energy-Efﬁcient Routing of Multiple AGVs with Multi-Agent Reinforcement Learning Xianfeng Ye 1 , Zhiyun Deng 1, Yanjun Shi 2 and Weiming Shen 1,* 1. Introduction Automated guided vehicles (AGVs) are autonomous portable robots that navigate predetermined paths using various sensing technologies [1,2]. They play a vital role in modern manufacturing and logistics systems by facilitating the transportation of raw materials and ﬁnished products [3,4]. To optimize their performance, it is necessary to integrate the scheduling control system of AGVs with existing production management systems, such as manufacturing execution systems (MES), enterprise resource planning (ERP), warehouse management systems (WMS), logistics control systems (LCS), and Radio- Frequency IDentiﬁcation (RFID) [5–8]. The AGV scheduling system receives request messages from the MES, dispatches AGVs to transport raw materials or ﬁnished products, and designs routes for AGVs to follow. However, the task assignment and path planning problem for multiple AGVs is challenging, as the number of decision variables and safety- related constraints grows signiﬁcantly with the number of AGVs [9]. Academic Editor: Alex Alexandridis Received: 26 April 2023 Revised: 4 June 2023 Accepted: 11 June 2023 Published: 15 June 2023 Received: 26 April 2023 Revised: 4 June 2023 Accepted: 11 June 2023 Published: 15 June 2023 2. Literature Review Reinforcement learning (RL) has become a promising solution to the AGV scheduling and routing problems, while many researchers have carried out much pioneering work with the application of RL [23]. pp For example, the Markov decision process (MDP) formulation was combined with the asynchronous deep Q network (DQN) to solve the routing problem in real time and obtain high-quality solutions [24]. A decentralized framework for multiple AGVs was proposed in [25] for multi-task allocation with attention (MTAA), which uses the DNN network and the A3C and MTAA-DQN path planning techniques to achieve task assignment equilibrium. Aside from this application, RL was used to solve the routing problem in a bidirectional transport network for the purpose of avoiding deadlocks and obtaining collision-free trajectories [26]. The deep Q network (DQN) was used in [27] to learn a transportation strategy with breakpoint continuation and hierarchical feedback, which can calculate and further modify a transportation schedule in a short time to accommodate dynamic factors. That aside, the authors of [28] tried to teach a neural network to allocate transportation duties to AGVs and design routes for them in accordance with the rewards computed by the network. An enhanced DQN was suggested in [29] to ﬁnd appropriate navigational approaches for certain current road circumstances, which limits the Q output of speciﬁc actions and incorporates their outcomes using calculations based on experience- based pooling. Moreover, a state space ﬁlter was proposed in [30] to improve the negotiation rules between different agents that adjust their routes when probable collisions are identiﬁed. Li et al. [31] proposed a deep learning approach that concurrently addresses task assignment and path planning concerns, and it uses the Markov decision chain to formulate the challenge of ﬁnding the shortest path without running afoul of other AGVs. In addition, De Ryck et al. [1] gave a general overview of the control algorithms and methods applied to the ﬁrst-generation and latest AGV systems. Xue et al. [32] used an RL approach to solve a multi-AGV ﬂow-shop scheduling problem, where AGVs communicate comprehensive data about each machine’s current state and running jobs. In other words, users are able to make decisions based on knowledge of the entire ﬂow shop. Nagayoshi et al. 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/). g g y To address this issue, several optimization-based approaches have been proposed, such as integer programming [10–12], heuristic algorithms [13–15], and metaheuristics [7,16–19]. However, these methods have limitations in dealing with the dynamic and uncertain nature of the industrial environment. Therefore, machine learning-based methods, particularly reinforcement learning (RL), have emerged as a promising approach to solving the prob- lem [20,21]. RL is a subﬁeld of machine learning that involves an agent learning from its https://www.mdpi.com/journal/sensors Sensors 2023, 23, 5615. https://doi.org/10.3390/s23125615 Sensors 2023, 23, 5615 2 of 14 interactions with the environment to maximize a cumulative reward signal. Multi-agent reinforcement learning (MARL) is an extension of RL that involves multiple agents learning to coordinate with each other to achieve a common objective [22]. MARL has been shown to be effective in solving complex problems that involve coordination and competition among multiple agents [8]. Compared with existing studies, the key contributions of this paper are summarized as follows: (1) We propose an ϵ-greedy MADDPG algorithm which is able to converge faster and achieve better performance during training by balancing exploration and ex- ploitation. (2) Modiﬁcations are made to the action and state space to ﬁt the setting of AGV activities, while a well-designed reward function is incorporated into the proposed algorithm that optimizes energy consumption while fulﬁlling all tasks. (3) The effectiveness of the proposed algorithm is demonstrated through numerical experiments, which show that it outperforms other methods in improving energy efﬁciency while addressing the multi-AGV task assignment and path planning problem. The remainder of this paper is organized as follows. Section 2 presents a literature review about AGV scheduling algorithms with MARL. Section 3 presents the background of reinforcement learning, followed by a description of the proposed algorithm and model in Section 4. Section 5 presents the simulation results and analyses. Section 6 concludes this paper and discusses some open issues and future work. 2. Literature Review [33] presented a decentralized autonomous strategy for controlling a large number of AGVs in response to ambiguous delivery requests, where the AGVs are equipped with transportation route plans that are intended to save travel time while avoiding collisions. Sierra-Garcia et al. [34] presented an intelligent hybrid control scheme that combines RL-based control (RLC) with Sensors 2023, 23, 5615 3 of 14 conventional PI regulators, where the RLC allows the AGVs to learn how to improve trajectory tracking adaptively. conventional PI regulators, where the RLC allows the AGVs to learn how to improve trajectory tracking adaptively. j y g p y As suggested in [35], the multi-agent reinforcement learning (MARL) policy is capable of (1) scaling to a large number of agents in a real-world setting with an ofﬂine response time within acceptable levels and (2) outperforming existing algorithms with lower path lengths and faster solution times. Takahashi et al. [36] provided a multi-agent deep deterministic policy gradient (MADDPG) approach for managing several AGVs using DRL, where simulated experiments demonstrated that the suggested method learns optimal or nearly optimal solutions from prior knowledge. Aside from that, several numerical tests were carried out in [8] to conﬁrm the effectiveness of the RL method. The authors of [37] used the MARL method to deal with the increased ﬂexibility and complexity introduced by the increased use of AGVs. In addition, Li et al. [38] proposed a reward-shaping technique based on the potential information ﬁeld which offers stepwise incentives and implicitly directs the AGVs to various targets to address the problem of reward sparsity. Moreover, Lu et al. [24] presented a DRL technique to address the AGV routing issue, where the conﬂict vectors are created from the retrieved embeddings and then processed using the LSTM network. From the above work, it can be seen that RL is very effective for solving the AGV scheduling and routing problem. However, the existing DQN algorithm in RL has some limitations, since it cannot solve continued questions directly. Moreover, it does not consider how to solve the path planning problem with the application of RL. 3.2. Multi-Agent Reinforcement Learning Model 3.2. Multi-Agent Reinforcement Learning Model In the MARL model, there are at least two agents existing in the same environ- ment and interacting with each other as shown in Figure 2 [39]. We take into account the multi-agent Markov decision process extension known as Markov games [40]. A set of states S, action sets for each of N agents A1, . . . , AN, a state transition function T : S × A1 × . . . AN−> P(S) which speciﬁes the probability distribution over the possible next states, given the current state and actions for each agent, and a reward function for each agent that also depends on the overall state and actions of all agents Ri : S × A1 . . . AN−> R deﬁne them. This means that all agents choose actions ai simultaneously after watching the system’s state s and receiving each agent’s individual reward ri. In the case of multiple AGVs in a warehouse, each AGV can be considered an agent, and the entire warehouse map can be viewed as the environment. To enable effective cooperation and competition between AGVs, researchers have developed various MARL algorithms that can learn the best strategies for multiple agents in the same environment. For example, one approach is to use a centralized training and decentralized execution (CTDE) architecture in which a central controller learns a joint policy for all agents during training, and each agent executes its own policy during execution. This approach has been shown to be effective in scenarios where there is a strong interdependence between agents, such as in a convoy of AGVs transporting a large item. Another approach is to use independent reinforcement learning (IRL), in which each agent learns its own policy independently without any communication or coordination with other agents. This approach can be useful when the actions of different agents do not have a signiﬁcant impact on each other, such as in scenarios where AGVs are used to transport different items to different locations. MARL has the potential to improve the efﬁciency and effectiveness of AGV systems in various industrial applications, and ongoing research in this area is expected to lead to even more sophisticated and effective algorithms in the future. Figure 2. The multi-agent architecture in the reinforcement learning system. Figure 2. The multi-agent architecture in the reinforcement learning system. 3.1. Single-Agent Reinforcement Learning Model 3.1. Single-Agent Reinforcement Learning Model RL is a framework for learning how an agent can take action in an environment to max- imize a cumulative reward signal. This framework can be expressed as a system consisting of an agent and an environment, as illustrated in Figure 1 [39]. The environment produces information that describes the state of the system, while the agent interacts with the en- vironment by observing the state and then selecting an action to perform. Subsequently, the environment accepts the action and transitions into the next state while returning a reward to the agent. This reward denotes the feedback signal from the environment, indicating whether it is beneﬁcial for the agent to adopt a certain strategy at a certain step. The agent’s objective is to learn a policy that maps states to actions to maximize the expected future cumulative reward. That is to say, the agent outputs the action At, observes the system’s state St, and receives the reward Rt from the system. Figure 1. The single agent of the reinforcement learning system. Figure 1. The single agent of the reinforcement learning system. In the context of AGVs, the vehicle can be formulated as an agent, since it can capture information with onboard sensors and perform an action with its actuators. The environ- ment can be the map of a manufacturing factory or an automated warehouse where the AGV operates. The action space of the AGV agent can be represented by go forward, turn left, turn right, and stopping operations, for example. During the training phase, the agent interacts with the environment until the terminal conditions are met. After that, the agent can learn to perform actions without human guidance to maximize its expected future reward in a certain state. Sensors 2023, 23, 5615 4 of 14 4. Methodology gy 4.1. The Multi-Agent Deep Deterministic Policy Gradient (MADDPG) Algorithm 4.1. The Multi-Agent Deep Deterministic Policy Gradient (MADDPG) Algorithm In this paper, we propose an energy-efﬁcient scheduling and routing algorithm based on the multi-agent deep deterministic policy gradient (MADDPG) algorithm and the path planning D* Lite algorithm. The D* Lite algorithm’s basic idea is to plan the global optimal path from the destination point to the beginning point based on available environmental information, treating the unknown portion as free space [41]. However, in this paper, we combine the D* Lite algorithm with energy consumption computation to optimize energy efﬁciency. gy y In RL, the deep deterministic policy gradient (DDPG) algorithm is a model-free, off- policy, and policy-based method suitable for solving such problems [42,43]. The DDPG Sensors 2023, 23, 5615 5 of 14 5 of 14 algorithm uses a deterministic policy, which means that when the policy and observed state are given, the action is uniquely determined. This is in contrast to classical RL algorithms, which use a stochastic policy that performs actions based on a probability distribution. DDPG follows the idea of ﬁxing the target network that is used in the DQN algorithm, resulting in only two networks that need to be learned: the policy network and the value network [44,45]. In DDPG, each network is subdivided into a current network and a target network, and the updating process for these two networks is different. Under the actor-critic framework, the policy network is referred to as an actor network that outputs a deterministic action, while the value network is referred to as a critic network that ﬁts the value function Qπ(s, a). Multi-agent in DDPG, an extension of DDPG, means that decentralized agents learn a centralized critique based on their collective observations and actions, resulting in a multi-agent policy gradient algorithm. It generates learned policies that, during execution, only use local information (i.e., their own observations), does not require a differentiable model of the dynamics of the environment or any particular structure on the method of communication between agents, and is applicable to competitive or mixed interactions involving both physical and communicative behavior. The critic possesses additional knowledge about the practices of other agents, but the actor just has access to local information. Once trained, only locally based actors who work independently are used throughout the execution phase. 4. Methodology The estimated policy network of the actor is θ(s), where θ is the parameter of the neural network. The actor also has another target network that is used to update the value of the critic network. Both networks have the same structure and output corresponding actions, but the parameters within the neural networks are different. In terms of the critic network, there are also two networks: an estimation network and a target network. Both networks output the Q value of the current state, but they differ in terms of their input. For instance, the input of the critic’s target network has two parameters, which are the observation of the current state and the action of the actor’s target network output. In contrast, the input of the critic’s estimation network is the action of the current actor’s estimation network output. The target network is used to calculate Qtarget. The update of the value network is based on the gradient descent of the TD-error. The critic, which acts as a judge, does not initially know whether the actor’s action is good enough and needs to learn step by step to provide accurate scoring. With the help of the value Qπ in the next moment, ﬁtted by the target network and the actual gain r, we can obtain Qtarget, which is then subtracted from the current Q to ﬁnd the mean squared deviation, allowing us to construct the loss function. In terms of the policy network, its update is based on gradient ascent. Since the goal of the actor is to ﬁnd an action A that maximizes the value Q of the output, optimizing the gradient of the policy network is for maximizing this Q value of the output of the value network. The loss function then adds a negative sign to facilitate minimizing the error. The parameters of n agents are identiﬁed as θ = [θ1, . . . , θn], and the policies of n agents are identiﬁed as π = [π1, . . . , πn] [42]. Therefore, the accumulated reward for a certain agent i and the expected reward gradient for a deterministic policy µθi can be represented as follows: J(θi) = Es∼ρπ,a∼πθi [ ∞ ∑ t=0 γtri,t] (1) L(θi) = 1 s ∑ j (yj −Qµ i (xj, aj 1, . . . , aj N))2 (2) = Ex,a∼D[∇θiµi(ai\|oi)∇aiQµ i (x, a1, . . . 4. Methodology , an)\|ai = µi(oi)] (3) J(θi) = Es∼ρπ,a∼πθi [ ∞ ∑ t=0 γtri,t] (1) L(θi) = 1 s ∑ j (yj −Qµ i (xj, aj 1, . . . , aj N))2 (2) (1) (2) ∇θi J(µi) = Ex,a∼D[∇θiµi(ai\|oi)∇aiQµ i (x, a1, . . . , an)\|ai = µi(oi)] (3) ∇θi J(µi) = Ex,a∼D[∇θiµi(ai\|oi)∇aiQµ i (x, a1, . . . , an)\|ai = µi(oi)] 4.2. Multi-Agent Model for AGV Operations 4.2. Multi-Agent Model for AGV Operations This paper presents a formal deﬁnition of the action space of the automated guided vehicle (AGV) agent, denoted by a(v, ω). The velocity of the AGV v is variable and can range from −1 m/s to 1 m/s, while the angular velocity ω is restricted to values between −1 rad/s and 1 rad/s. Consequently, the AGV agent is capable of performing ﬁve distinct actions, namely moving forward, moving backward, turning left, turning right, and halting. The reward function is deﬁned as follows: This paper presents a formal deﬁnition of the action space of the automated guided vehicle (AGV) agent, denoted by a(v, ω). The velocity of the AGV v is variable and can range from −1 m/s to 1 m/s, while the angular velocity ω is restricted to values between g g y −1 rad/s and 1 rad/s. Consequently, the AGV agent is capable of performing ﬁve distinct actions, namely moving forward, moving backward, turning left, turning right, and halting. The reward function is deﬁned as follows: + ki2 × Cv × v × cos(ω) + ki3 × Ce × (Etarget −Ei) + ki4 × CAGV + ki5 × Cobstacles (4) ri = ki1 × Dposition + ki2 × Cv × v × cos(ω) + ki3 × Ce × (Etarget −Ei) + ki4 × CAGV + ki5 × Co (4) In the above equation, ki1, ki2, ki3, ki4, and ki5 are the weight parameters, while Dposition represents the reward value based on the current position relative to the previous position. A positive reward is given if the current position is closer to the destination than the previous position, and vice versa. This incentivizes the AGV to approach the target site, using the distance reward as guidance. The equation D = p (x2 −x1)2 + (y2 −y1)2 calculates the distance between the current position and the destination point. The value of the award will be negative if Dcurrent is greater than Dprevious. The value of the award will be negative if there is an AGV collision. 4.2. Multi-Agent Model for AGV Operations Let us deﬁne the previous and current position for AGVi as Pos1(x0, y0) and Pos2(x1, y1) for AGVj, respectively; that is, Pos1(x2, y2) and Pos2(x3, y3) if p (x1 −x3)2 + (y1 −y3)2) < 0.5 are satisﬁed by (y1 −y0) × (y3 −y2) < 0 or (x1 −x0) × (x3 −x2) < 0, which means that if they travel in the opposite direction, then they will collide, and the reward value will be negative. The result of q ((xi −xj)2 + (yi −yj)2) is less than 0.1 if an AGV collides with an obstruction due to the line’s narrow width. CAGV and Cobstacles are both deﬁned as −50 when there is a collision; otherwise, they are 50. y The speed reward guides the AGV to complete the task with the least number of rotations and the most signiﬁcant amount of linear speed achievable. Cv is the velocity coefﬁcient and is used to scale this reward item. The third reward item is related to energy consumption, denoted by Ei, which is computed differently based on whether the AGV is stationary or in motion. When the AGV is not stationary, the energy consumption is computed as the average energy consumption per time step. Otherwise, the corresponding energy consumption is multiplied by a parameter factor of 0.3, which signiﬁes that the AGV consumes less energy when in a stationary state compared with when it is in a normal driving state. The value of Etarget is computed using the route path determined by the D* Lite algorithm, while Ce is the energy coefﬁcient, which is used to ﬁt this reward item with other items. The last two reward items are related to collision avoidance and incentivize the AGV to avoid path conﬂicts with other AGVs or obstacles. These reward items are critical to ensuring the safe and efﬁcient operation of the AGV. In addition, it is worth noting that the parameters in the reward function play a crucial role in numerical simulation and will be discussed further in another paper. ∇θi J(µi) = Ex,a∼D[∇θiµi(ai\|oi)∇aiQµ i (x, a1, . . . , an)\|ai = µi(oi)] (3) where oi is the observation of the agent i, x = [oi, . . . , on] is the observation value, Qµ i (x, a1, . . . , an) is the action and state function, ∇θiµi(ai\|oi) is the gradient of the pol- icy network at θi, and ∇aiQµ i (x, a1, . . . , an) is the gradient of the value network at x and action sets (a1, . . . , an). Sensors 2023, 23, 5615 6 of 14 4.3. MADDPG with ϵ-Greedy Exploration and exploitation are very prominent problems in reinforcement learning, and they are also the focus of determining whether the reinforcement learning system can obtain an optimal solution. Exploration would allow an agent to enhance its knowledge about its action, which may lead to long-term beneﬁts. Improving the accuracy of action value estimation enables an agent to make more informed decisions. Exploitation uses the greedy action to acquire the greatest reward through exploiting the agent’s action value estimates. However, being greedy may not lead to the greatest reward and in fact may lead to suboptimal results. While exploration may ﬁnd more accurate action value estimates, exploitation may obtain more rewards. However, it is not possible to do both at the same time. Exploration is the right way to maximize the expected return at the present moment, while exploitation is the right way to maximize the total return in the long run. Unfortunately, in a certain state, the agent can only perform one action: either Sensors 2023, 23, 5615 7 of 14 hich is i i 7 of 14 hich is i i exploration or exploitation. The two cannot be carried out at the same time, and thus this is the contradiction that accentuates the emphasis of reinforcement learning and how to balance exploration and exploitation. the total return in the long run. Unfortunately, in a certain state, the agent can only perform one action, either exploration or exploitation, the two cannot be carried out at the same time, so this is the contradiction that emphasizes the emphasis of reinforcement learning how to balance exploration and exploitation. p p The ϵ-greedy policy is a popular strategy for balancing exploration and exploita- tion [46,47]. This policy selects the best action with a probability 1 −ϵ and a random action with a probability ϵ. The parameter ϵ determines the degree of exploration versus exploitation, where a high value of ϵ results in more exploration and a low value of ϵ results in more exploitation [48]. However, the ϵ-greedy method has a limitation in that it selects random actions uniformly, even though certain actions may be better than others. To address this limitation, softmax policies have been proposed, which select random actions with probabilities proportional to their current values [49].In the context of AGV route selection, the ϵ-greedy policy can be used to balance exploration and exploitation by randomly selecting between exploration and exploitation. 4.3. MADDPG with ϵ-Greedy When one AGV explores its action, the other AGVs can exploit that action to their advantage. However, the optimal action for one AGV may not be the optimal action for all AGVs, as route conﬂicts may require different actions. Therefore, the challenge in AGV route selection is to ﬁnd a balance between exploration and exploitation that maximizes the overall performance of the system. The ϵ-greedy policy is a straightforward strategy for balancing discovery and exploitation, where the parameter ϵ controls the degree of exploration versus exploitation. The algorithm of MADDPG with ϵ-greedy for AGVs is illustrated in Algorithm 1. o to ba a ce e p o at o a d e p o tat o The ϵ −greedy policy is a popular strategy for balancing exploration and exploitation[47] This policy selects the best action with probability 1 −ϵ and a random action with prob- ability ϵ. The parameter ϵ determines the degree of exploration versus exploitation, where a high value of ϵ results in more exploration and a low value of ϵ results in more exploitation[49]. However, the ϵ −greedy method has a limitation in that it selects random actions uniformly, even though certain actions may be better than others. To address this limitation, softmax policies have been proposed, which select random actions with probabilities proportional to their current values[50]. In the context of AGV route selection, the ϵ −greedy policy can be used to balance exploration and exploitation by randomly selecting between exploration and exploitation. When one AGV explores its action, the other AGVs can exploit that action to their advantage. However, the optimal action for one AGV may not be the optimal action for all AGVs, as the route conflicts may require different actions. Therefore, the challenge in AGV route selection is to find a balance between exploration and exploitation that maximizes the overall performance of the system. The ϵ −greedy policy is a straightforward strategy for balancing discovery and exploitation, where the parameter ϵ controls the degree of exploration versus exploitation. The algorithm of MADDPG with ϵ −greedy for AGVs is illustrated in Algorithm 1. Algorithm 1: An algorithm of MADDPG with the ϵ-greedy policy for AGVs. 4.3. MADDPG with ϵ-Greedy Algorithm 1 An algorithm of MADDPG with ϵ −greedy for AGVs for j=1 to max-episode do Initialization of the parameters for t=1 to M do for i=1 to N do n= random number if n < ϵ then execute any action(a) else execute the action which maximizes Qt(a)with1 −ϵ end if ai = µθi(oi)+Nt a=(ai, ..., aN) ri = ki1 × Dposition + ki2 × Cv × v × cos(ω) + ki3 × Ce × (Etarget −Ei) + ki4 × CAGV + ki5 × Cobstacles end for for agent i=1 to N do yj = rj i + γQµ′ i(x′ j, a′ 1, ...a′ N) L(θi) = 1 s ∑j(yj −Qµ i (xj, aj 1, ..., aj N))2 ∇θiJ ≃1 S ∑j θiµi(oj i)aiQµ i (xj, aj 1, ...ai, ..., aj N)\|ai = µi(oj i) end for for i=1 to N do θ′ i < −τθi + (1 −τ)θ′ i end for end for end for Algorithm 1: An algorithm of MADDPG with the ϵ-greedy policy for AGVs. Algorithm 1 An algorithm of MADDPG with ϵ −greedy for AGVs 4.4. Three Algorithms in This Experiment 5.1. Test Scenarios The experimental test scenarios presented in this paper aim to evaluate the perfor- mance of an AGV-based system in a warehouse environment. The state of the AGV includes the position, velocity, and distance between the starting point and the ﬁnal destination. The warehouse, as illustrated in Figure 3, has a total area of 82 × 66 m2 and is equipped with a variety of facilities, including a power center, a repair center, and an ofﬁce room for workers. The warehouse is divided into 39 racks, denoted by BLOCK, which are strictly off limits for AGVs. The AGVs were programmed to stop at the park center when no tasks were assigned. Additionally, there were designated halt positions, such as A02 and B12, where the AGVs could temporarily pause while processing an order. There were three types of goods in this warehouse: delivered goods, transported goods, and relocated goods. Delivered goods were loaded at the door and unloaded at another position, whereas transported goods were moved from a speciﬁc rack to the door, and relocated goods were transferred from one rack to another. These tasks were challenging, as the AGVs had to navigate through the warehouse while avoiding obstacles, as shown in Figure 3. To evaluate the effectiveness of the proposed algorithm, these tasks were selected as benchmark problems, and they are described in Table 1. The experimental scenarios aimed to examine the AGVs’ ability to perform these transfer tasks efﬁciently and accurately. The proposed algorithm’s performance was evaluated based on various metrics, including completion time, task efﬁciency, and AGV utilization. These experiments provided valuable insights into the AGV’s performance in a real-world warehouse environment, and they may lead to improvements in AGV-based systems’ efﬁciency and effectiveness. 4.4. Three Algorithms in This Experiment In this study, we focus on evaluating the performance of three popular reinforce- ment learning algorithms in a speciﬁc scenario. The algorithms we considered were Q-learning, MADDPG, and enhanced MADDPG with the epsilon-greedy policy, which are described below. First, Q-learning is a widely used algorithm that employs off-policy reinforcement learning to maximize rewards. The algorithm updates a Q table that stores the expected Sensors 2023, 23, 5615 8 of 14 reward of each state-action pair. Although Q-learning is a model-independent technique, it can be prone to taking risks in real-world applications. Our study aims to investigate the strengths and limitations of Q-learning in the given scenario. reward of each state-action pair. Although Q-learning is a model-independent technique, it can be prone to taking risks in real-world applications. Our study aims to investigate the strengths and limitations of Q-learning in the given scenario. g g g MADDPG, on the other hand, is a centralized, critic-based actor-critic approach that allows one to consider various reward functions. The algorithm has a critic for each agent, which increases the amount of data used in the learning process. We evaluated the performance of MADDPG and compared it with Q-learning to gain insights into the strengths and limitations of these two approaches. g pp Finally, we introduce enhanced MADDPG with the epsilon-greedy policy, which is a straightforward strategy for balancing exploration and exploitation. By randomly selecting between exploration and exploitation, the algorithm can achieve a balance between the two strategies. We compared the performance of enhanced MADDPG with the epsilon-greedy policy with that of Q-learning and MADDPG to evaluate the effectiveness of this approach in the given scenario. Our study contributes to the existing literature on reinforcement learning by evaluating the performance of three popular algorithms in a speciﬁc scenario. The ﬁndings of this study can help researchers and practitioners select the most appropriate algorithm for a given problem and improve the overall performance of reinforcement learning algorithms. 5.2. Numerical Results In this study, we analyzed the numerical results of the AGV system, which involved 30 AGVs and 3 types of goods categorized into 30 tasks, as illustrated in Table 1. Speciﬁcally, there were 12 tasks for delivered goods (Tasks 01–12), 12 tasks for transported goods (Tasks 13–24), and 6 tasks for relocated goods (Tasks 25–30). Each AGV was assigned 1 task, which required the 30 AGVs to complete all tasks simultaneously. For instance, AGV-01 was responsible for completing Task 01 by loading the cargo at the door and unloading it at position A24. The other tasks followed a similar pattern. Table 1 summarizes the three types of goods that needed to be processed. Two different types of software tools were used for the numerical simulations. A piece of the MADDPG algorithm was taken from [42] Sensors 2023, 23, 5615 9 of 14 9 of 14 and speciﬁcally altered for our needs to yield the transportation line for the AGVs with parameters in Table 2. Our unique Java-written AGV-scheduling program was used in the left part for visualization of transportation. and speciﬁcally altered for our needs to yield the transportation line for the AGVs with parameters in Table 2. Our unique Java-written AGV-scheduling program was used in the left part for visualization of transportation. Figure 3. Map of the warehouse. Figure 4 illustrates the convergence of the reward function in route selection based on the Q-learning technique. The function showed a gradual rise and eventual stabilization to a ﬁnal state as the number of episodes increased. However, both the native multi-agent deep deterministic policy gradient (MADDPG) and ϵ-greedy MADDPG techniques converged more quickly to the steady state. When comparing the two MADDPG techniques, the ϵ- greedy strategy yielded better beneﬁts. In the MADDPG algorithm, an ϵ value of 0.1 provided superior results to values of 0.01 and 0.05. This ﬁnding was consistent with most simulations of the ϵ-greedy approach, which demonstrated that an optimal value was attained around ϵ = 0.1. However, determining the ideal value of ϵ for the current AGV environment requires further investigation. Figure 3. Map of the warehouse. Figure 3. Map of the warehouse. Figure 4 illustrates the convergence of the reward function in route selection based on the Q-learning technique. The function showed a gradual rise and eventual stabilization to a ﬁnal state as the number of episodes increased. 5.2. Numerical Results However, both the native multi-agent deep deterministic policy gradient (MADDPG) and ϵ-greedy MADDPG techniques converged more quickly to the steady state. When comparing the two MADDPG techniques, the ϵ- greedy strategy yielded better beneﬁts. In the MADDPG algorithm, an ϵ value of 0.1 provided superior results to values of 0.01 and 0.05. This ﬁnding was consistent with most simulations of the ϵ-greedy approach, which demonstrated that an optimal value was attained around ϵ = 0.1. However, determining the ideal value of ϵ for the current AGV environment requires further investigation. Table 1. Thirty transfer tasks in three projects. Transfer Type Task List Load Cargo Position Unload Cargo Position Goods Delivered Task 01 Door A24 Task 02 Door A17 Task 03 Door A16 Task 04 Door A09 Task 05 Door A08 Task 06 Door A01 Task 07 Door E24 Task 08 Door E17 Task 09 Door E16 Task 10 Door E09 Task 11 Door E08 Task 12 Door E01 Table 1. Thirty transfer tasks in three projects. 10 of 14 10 of 14 Sensors 2023, 23, 5615 Table 1. Cont. Transfer Type Task List Load Cargo Position Unload Cargo Position Goods Transported Task 13 I24 Door Task 14 I17 Door Task 15 I16 Door Task 16 I09 Door Task 17 I08 Door Task 18 I01 Door Task 19 M24 Door Task 20 M17 Door Task 21 M16 Door Task 22 M09 Door Task 23 M08 Door Task 24 M01 Door Goods Relocated Task 25 Q24 Z01 Task 26 Q17 Z08 Task 27 Q16 Z09 Task 28 S24 W01 Task 29 S17 W08 Task 30 S16 W09 Task 20 M17 Door Task 21 M16 Door Task 22 M09 Door Task 23 M08 Door Task 24 M01 Door Goods Relocated Task 25 Q24 Z01 Task 26 Q17 Z08 Task 27 Q16 Z09 Task 28 S24 W01 Task 29 S17 W08 Task 30 S16 W09 Table 2. Parameter settings for simulations. Description Notation and Value Weight Parameters ki1 = 0.2, ki2 = 0.1, ki3 = 0.5, ki4 = 0.1, ki5 = 0.1 Reward Value Dposition Velocity Coefﬁcient Cv Energy Consumption Ei Target Coefﬁcient Etarget Energy Coefﬁcient Ce Collision Parameter between AGVs CAGV Collision Parameter between AGV and Obstacle CObstacle Learning Rate 0.15 Discount Factor 0.99 α 0.01 β 0.01 γ 0.95 τ 0.01 (a) Goods delivered. (b) Goods transported. (c) Goods relocated. Figure 4. Three kinds of goods: rewards. 5.2. Numerical Results It was observed that the rewards ﬂuctuated more when the AGVs traversed multiple obstacles, whereas the rewards were more stable when the AGVs encountered fewer obstacles. 5.2. Numerical Results From this, it is clear that the route planned by Q-learning used a signiﬁcant amount of energy, followed by the route planned by the Sensors 2023, 23, 5615 11 of 14 11 of 14 primeval MADDPG algorithm and the route planned by the ϵ-greedy MADDPG algorithm, which used the least amount of energy. primeval MADDPG algorithm and the route planned by the ϵ-greedy MADDPG algorithm which used the least amount of energy. primeval MADDPG algorithm and the route planned by the ϵ-greedy MADDPG algorithm, which used the least amount of energy. (c) Goods relocated. (b) Goods transported. (a) Goods delivered. (b) Goods transported. (c) Goods relocated. Figure 5. Three kinds of goods: energy consumption. (a) Goods delivered. (a) Goods delivered. (b) Goods transported. (c) Goods relocated. Figure 5. Three kinds of goods: energy consumption. Figure 5. Three kinds of goods: energy consumption. The path of the AGV is shown in Figure 6. The shelves’ green tint indicates that they can hold stock. The shelf’s yellow tint shows that there is merchandise there. The shelves’ crimson tint indicates that they are completely full. When an AGV is in transit, it is shown by a yellow AGV, and when it is idle, it is indicated by a green AGV. From the ﬁgure, we can see that due to route conﬂicts, the AGV transporting items stopped for a period of time in the middle so that the transport AGV could pass smoothly. At the same time, some AGVs chose another path to avoid route conﬂicts. Figure 6. Illustration of AGVs’ routes. Furthermore, the simulation results indicate that energy consumption should be taken into account when selecting the path for the AGVs. It was observed that the rewards ﬂuctuated more when the AGVs traversed multiple obstacles, whereas the rewards were more stable when the AGVs encountered fewer obstacles. l i Figure 6. Illustration of AGVs’ routes. Figure 6. Illustration of AGVs’ routes. Furthermore, the simulation results indicate that energy consumption should be taken into account when selecting the path for the AGVs. It was observed that the rewards ﬂuctuated more when the AGVs traversed multiple obstacles, whereas the rewards were more stable when the AGVs encountered fewer obstacles. Furthermore, the simulation results indicate that energy consumption should be taken into account when selecting the path for the AGVs. 5.2. Numerical Results In this article, we used the AGV’s energy consumption per second as a unit, the AGV’s total energy once it arrived at its destination as an indicator, and the total energy of all AGV vehicles, as indicated in Figure 5. From this, it is clear that the route planned by Q-learning used a signiﬁcant amount of energy, followed by the route planned by the Table 2. Parameter settings for simulations. Description Notation and Value Weight Parameters ki1 = 0.2, ki2 = 0.1, ki3 = 0.5, ki4 = 0.1, ki5 = 0.1 Reward Value Dposition Velocity Coefﬁcient Cv Energy Consumption Ei Target Coefﬁcient Etarget Energy Coefﬁcient Ce Collision Parameter between AGVs CAGV Collision Parameter between AGV and Obstacle CObstacle Learning Rate 0.15 Discount Factor 0.99 α 0.01 β 0.01 γ 0.95 τ 0.01 (a) Goods delivered. (b) Goods transported. (c) Goods relocated. Figure 4. Three kinds of goods: rewards. In this article, we used the AGV’s energy consumption per second as a unit, the AGV’s total energy once it arrived at its destination as an indicator, and the total energy of all AGV vehicles, as indicated in Figure 5. From this, it is clear that the route planned by Q-learning used a signiﬁcant amount of energy, followed by the route planned by the Table 2. Parameter settings for simulations. Description Notation and Value Weight Parameters ki1 = 0.2, ki2 = 0.1, ki3 = 0.5, ki4 = 0.1, ki5 = 0.1 Reward Value Dposition Velocity Coefﬁcient Cv Energy Consumption Ei Target Coefﬁcient Etarget Energy Coefﬁcient Ce Collision Parameter between AGVs CAGV Collision Parameter between AGV and Obstacle CObstacle Learning Rate 0.15 Discount Factor 0.99 α 0.01 β 0.01 γ 0.95 τ 0.01 (c) Goods relocated. (a) Goods delivered. (b) Goods transported. (a) Goods delivered. (b) Goods transported. (c) Goods relocated. Figure 4. Three kinds of goods: rewards. Figure 4. Three kinds of goods: rewards. In this article, we used the AGV’s energy consumption per second as a unit, the AGV’s total energy once it arrived at its destination as an indicator, and the total energy of all AGV vehicles, as indicated in Figure 5. 6. Conclusions In this paper, we proposed a multi-agent reinforcement learning (MARL) algorithm to address the problem of scheduling and routing multiple AGVs with the aim of minimizing Sensors 2023, 23, 5615 12 of 14 12 of 14 the overall energy consumption. The proposed algorithm was built upon the multi-agent deep deterministic policy gradient (MADDPG) algorithm, with modiﬁcations made to the action space and state space to suit the speciﬁc activities of AGVs. While prior studies have overlooked the energy efﬁciency of AGVs, we designed a reward function that helps to optimize the overall energy consumption of the system during task completion. To enhance the performance of our proposed MARL algorithm, we selected suitable parameters that facilitated obstacle avoidance, speedy path planning, and energy conserva- tion. We conducted numerical experiments to evaluate the performance of the algorithm, and the results demonstrate its effectiveness in solving the multi-AGV task assigning and path planning problem. This leads to a reduction in the total energy consumption of AGV transportation, which increases as the number of operational AGVs increases. During the simulation, the outcomes of the simulation were inﬂuenced by a range of parameters. We adopted identical settings for all AGVs in the reward function, although the parameters for each AGV’s reward function were slightly different due to variations in position and the actions taken in response to those positions. These differences arose due to the unique nature of each AGV’s activities. For instance, an AGV tasked with transporting heavier goods may use more energy than one carrying lighter items. Additionally, we observed that the optimal ϵ value for the MADDPG model in the current environment has not yet been established using the ϵ-greedy approach. y g g y pp In our future work, we plan to develop an end-to-end learning framework that fo- cuses on the direct control inputs of AGVs, rather than utilizing the desired velocity and angular velocity as the action space. We anticipate that this approach will enhance the performance of our proposed algorithm even further. By integrating a more comprehensive understanding of AGVs’ behaviors and their interactions with the environment, the algo- rithm will be better equipped to adapt to the diverse requirements of various tasks and settings. Ultimately, we are conﬁdent that our research will contribute signiﬁcantly to the advancement of energy-efﬁcient AGV systems, which are of growing importance in contemporary logistics and transportation applications. Author Contributions: Conceptualization, X.Y. 6. Conclusions and W.S.; methodology, X.Y., Z.D., Y.S. and W.S.; software, X.Y.; validation, X.Y. and Z.D.; formal analysis, X.Y.; investigation, X.Y., Z.D., Y.S. and W.S.; resources, W.S.; data curation, X.Y.; writing—original draft preparation, X.Y.; writing—review and editing, Z.D., Y.S. and W.S.; visualization, X.Y.; supervision, W.S.; project administration, Z.D.; funding acquisition, W.S. All authors have read and agreed to the published version of the manuscript. Funding: The work presented in this paper was supported in part by the National Key R&D Program of China under Grant No.2022YFE0114200. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy. Conﬂicts of Interest: The authors declare no conﬂict of interest. pp y 6. Alzahrani, A.; Sajjad, K.; Hafeez, G.; Murawwat, S.; Khan, S.; Khan, F.A. Real-time energy opt buildings integrated with renewable microgrid. Appl. Energy 2023, 335, 120640. [CrossRef] References [CrossRef] 20 H H Ji X H Q F S Li K D i f t l i b d AGV l ti h d li ith i d l f ﬂ ibl p p p g f [ ] 19. Xie, J.; Gao, L.; Peng, K.; Li, X.; Li, H. Review on ﬂexible job shop scheduling. IET Collab. Intell. Manuf. 2019, 1, 67–77. [CrossRef] 20. Hu, H.; Jia, X.; He, Q.; Fu, S.; Liu, K. Deep reinforcement learning based AGVs real-time scheduling with mixed rule for ﬂexible shop ﬂoor in industry 4.0. Comput. Ind. Eng. 2020, 149, 106749. [CrossRef] 21. Melesse, T.Y.; Di Pasquale, V.; Riemma, S. Digital Twin models in industrial operations: State-of-the directions. IET Collab. Intell. Manuf. 2021, 3, 37–47. [CrossRef] 22. Zhou, T.; Tang, D.; Zhu, H.; Zhang, Z. Multi-agent reinforcement learning for online scheduling in smart factories. Robot. Comput.-Integr. Manuf. 2021, 72, 102202. [CrossRef] 23. Russell, S.J.; Norvig, P. Artiﬁcial Intelligence: A Modern Approach; Pearson Education Limited: Berkeley, 23. Russell, S.J.; Norvig, P. Artiﬁcial Intelligence: A Modern Approach; Pearson Education Limited: Berkeley, CA, USA , 2016. 24. Lu, C.; Long, J.; Xing, Z.; Wu, W.; Gu, Y.; Luo, J.; Huang, Y. Deep Reinforcement Learning for Solving AGVs Routing Problem. In 23. Russell, S.J.; Norvig, P. Artiﬁcial Intelligence: A Modern Approach; Pearson Education Limited: Berkeley, CA, USA , 2016. 24 L C L J Xi Z W W G Y L J H Y D R i f L i f S l i AGV R i P bl I 24. Lu, C.; Long, J.; Xing, Z.; Wu, W.; Gu, Y.; Luo, J.; Huang, Y. Deep Reinforcement Learning for Solving AGVs Routing Problem. In Proceedings of the International Conference on Veriﬁcation and Evaluation of Computer and Communication Systems, Xi’an, China, 26–27 October 2020; Springer: Berlin/Heidelberg, Germany, 2020; pp. 222–236. Proceedings of the International Conference on Veriﬁcation and Evaluation of Computer and Communication Systems, Xi’an, China, 26–27 October 2020; Springer: Berlin/Heidelberg, Germany, 2020; pp. 222–236. 25. Yin, Z.; Liu, J.; Wang, D. Multi-AGV Task allocation with Attention Based on Deep Reinforcement Learning. Int. J. Pattern Recognit. Artif. Intell. 2022, 36, 2252015. [CrossRef] 26. Chujo, T.; Nishida, K.; Nishi, T. A Conﬂict-Free Routing Method for Automated Guided Vehicles Using Reinforcement Learning. In Proceedings of the International Symposium on Flexible Automation, Virtual, Online, 8–9 July 2020; American Society of Mechanical Engineers: New York, NY, USA, 2020; Volume 83617, p. References 1. De Ryck, M.; Versteyhe, M.; Debrouwere, F. Automated guided vehicle systems, state-of-the-art control algorithms and techniques. J. Manuf. Syst. 2020, 54, 152–173. [CrossRef] 1. De Ryck, M.; Versteyhe, M.; Debrouwere, F. Automated guided vehicle systems, state-of-the-art control algorithms and techniques. J. Manuf. Syst. 2020, 54, 152–173. [CrossRef] 2. Shi, Y.; Han, Q.; Shen, W.; Zhang, H. Potential applications of 5G communication technologies in collaborative intelligent manufacturing. IET Collab. Intell. Manuf. 2019, 1, 109–116. [CrossRef] 2. Shi, Y.; Han, Q.; Shen, W.; Zhang, H. Potential applications of 5G communication technologies in collaborative intelligent manufacturing. IET Collab. Intell. Manuf. 2019, 1, 109–116. [CrossRef] g f 3. Yoshitake, H.; Kamoshida, R.; Nagashima, Y. New automated guided vehicle system using real- warehouse picking. IEEE Robot. Autom. Lett. 2019, 4, 1045–1052. [CrossRef] 3. Yoshitake, H.; Kamoshida, R.; Nagashima, Y. New automated guided vehicle system using real-time holonic scheduling for warehouse picking. IEEE Robot. Autom. Lett. 2019, 4, 1045–1052. [CrossRef] h dd b d b h ll b ll f h l 5. Ahmed, S.U.; Affan, M.; Raza, M.I.; Hashmi, M.H. Inspecting Mega Solar Plants through Computer Vision and Drone Technolo- gies. In Proceedings of the 2022 International Conference on Frontiers of Information Technology (FIT), Islamabad, Pakistan, 12–13 December 2022; pp. 18–23; IEEE: Piscataway, NJ, USA, 2022. 5. Ahmed, S.U.; Affan, M.; Raza, M.I.; Hashmi, M.H. Inspecting Mega Solar Plants through Computer Vision and Drone Technolo- gies. In Proceedings of the 2022 International Conference on Frontiers of Information Technology (FIT), Islamabad, Pakistan, 12–13 December 2022; pp. 18–23; IEEE: Piscataway, NJ, USA, 2022. pp y 6. Alzahrani, A.; Sajjad, K.; Hafeez, G.; Murawwat, S.; Khan, S.; Khan, F.A. Real-time energy optimization and scheduling of buildings integrated with renewable microgrid. Appl. Energy 2023, 335, 120640. [CrossRef] 13 of 14 Sensors 2023, 23, 5615 13 of 14 7. Xin, J.; Wei, L.; D’Ariano, A.; Zhang, F.; Negenborn, R. Flexible time–space network formulation and hybrid metaheuristic for conﬂict-free and energy-efﬁcient path planning of automated guided vehicles. J. Clean. Prod. 2023, 398, 136472. [CrossRef] gy p p g g J [ ] 8. Hu, H.; Yang, X.; Xiao, S.; Wang, F. Anti-conﬂict AGV path planning in automated container terminals based on multi-agent reinforcement learning. Int. J. Prod. Res. 2023, 61, 65–80. [CrossRef] g 9. Lian, Y.; Yang, Q.; Xie, W.; Zhang, L. Cyber-physical system-based heuristic planning and scheduling method for multiple automatic guided vehicles in logistics systems. IEEE Trans. Ind. References Inform. 2020, 17, 7882–7893. [CrossRef] 10. Goli, A.; Tirkolaee, E.B.; Aydın, N.S. Fuzzy integrated cell formation and production scheduling considering automated guided vehicles and human factors. IEEE Trans. Fuzzy Syst. 2021, 29, 3686–3695. [CrossRef] 11. Hu, H.; Chen, X.; Wang, T.; Zhang, Y. A three-stage decomposition method for the joint vehicle dispatchi problem in automated container terminals. Comput. Ind. Eng. 2019, 129, 90–101. [CrossRef] 11. Hu, H.; Chen, X.; Wang, T.; Zhang, Y. A three-stage decomposition method for the joint vehicle dispatching and storage allocation problem in automated container terminals. Comput. Ind. Eng. 2019, 129, 90–101. [CrossRef] Y O ﬁ d h d l f d bl f d l ll d G f 12. Yue, L.; Fan, H.; Ma, M. Optimizing conﬁguration and scheduling of double 40 ft dual-trolley qu improving container terminal services. J. Clean. Prod. 2021, 292, 126019. [CrossRef] p g 13. Fransen, K.; van Eekelen, J. Efﬁcient path planning for automated guided vehicles using A*(Astar) algorithm incorporating turning costs in search heuristic. Int. J. Prod. Res. 2023, 61, 707–725. [CrossRef] g 14. Nishi, T.; Akiyama, S.; Higashi, T.; Kumagai, K. Cell-based local search heuristics for guide path design of automated guided vehicle systems with dynamic multicommodity ﬂow. IEEE Trans. Autom. Sci. Eng. 2019, 17, 966–980. [CrossRef] 15. Kabir, Q.S.; Suzuki, Y. Comparative analysis of different routing heuristics for the battery management of automated guided vehicles. Int. J. Prod. Res. 2019, 57, 624–641. [CrossRef] 16. Zhong, M.; Yang, Y.; Dessouky, Y.; Postolache, O. Multi-AGV scheduling for conﬂict-free path planning in automated container terminals. Comput. Ind. Eng. 2020, 142, 106371. [CrossRef] p g 17. Zou, W.Q.; Pan, Q.K.; Meng, T.; Gao, L.; Wang, Y.L. An effective discrete artiﬁcial bee colony alg dispatching problem in a matrix manufacturing workshop. Expert Syst. Appl. 2020, 161, 113675. [Cross g K.; Meng, T.; Gao, L.; Wang, Y.L. An effective discrete artiﬁcial bee colony algorithm for multi-AGVs in a matrix manufacturing workshop. Expert Syst. Appl. 2020, 161, 113675. [CrossRef] 18. Xiao, X.; Pan, Y.; Lv, L.; Shi, Y. Scheduling multi–mode resource–constrained tasks of automated guided vehicles with an improved particle swarm optimization algorithm. IET Collab. Intell. Manuf. 2021, 3, 93–104. [CrossRef] p p p g f [ ] 19. Xie, J.; Gao, L.; Peng, K.; Li, X.; Li, H. Review on ﬂexible job shop scheduling. IET Collab. Intell. Manuf. 2019, 1, 67–77. References V001T04A001. g p 27. Yan, J.; Liu, Z.; Zhang, T.; Zhang, Y. Autonomous decision-making method of transportation process for ﬂexible job shop scheduling problem based on reinforcement learning. In Proceedings of the 2021 International Conference on Machine Learning and Intelligent Systems Engineering (MLISE), Chongqing, China, 9–11 July 2021; pp. 234–238. g y g g gq g y pp 28. Zhang, H.; Luo, J.; Lin, X.; Tan, K.; Pan, C. Dispatching and Path Planning of Automated Guided Vehicles based on Petri Nets and Deep Reinforcement Learning. In Proceedings of the 2021 IEEE International Conference on Networking, Sensing and Control (ICNSC), Xiamen, China, 3–5 December 2021; Volume 1, pp. 1–6. pp 29. Liu, H.; Hyodo, A.; Akai, A.; Sakaniwa, H.; Suzuki, S. Action-limited, Multimodal Deep Q Learning for AGV Fleet Route Planning. In Proceedings of the 5th International Conference on Control Engineering and Artiﬁcial Intelligence, Sanya, China, 14–16 January 2021; pp. 57–62. 30. Nagayoshi, M.; Elderton, S.; Sakakibara, K.; Tamaki, H. Adaptive Negotiation-rules Acquisition Methods in Decentralized AGV Transportation Systems by Reinforcement Learning with a State Space Filter. In Proceedings of the International Conference on Artiﬁcial Life and Robotics, ICAROB 2017, Miyazaki, Japan, 19–22 January 2017. 31. Li, M.P. Task Assignment and Path Planning for Autonomous Mobile Robots in Stochastic Warehouse Systems. Ph.D. Thesis, Rochester Institute of Technology, Rochester, NY, USA, 2021. 32. Xue, T.; Zeng, P.; Yu, H. A reinforcement learning method for multi-AGV scheduling in manufacturin 2018 IEEE International Conference on Industrial Technology (ICIT), Lyon, France, 19–22 February 201 gy y y pp 33. Nagayoshi, M.; Elderton, S.J.; Sakakibara, K.; Tamaki, H. Reinforcement Learning Approach for Adaptive Negotiation-Rules Acquisition in AGV Transportation Systems. J. Adv. Comput. Intell. Intell. Inform. 2017, 21, 948–957. [CrossRef] Sensors 2023, 23, 5615 14 of 14 14 of 14 34. Sierra-Garcia, J.E.; Santos, M. Combining reinforcement learning and conventional control to improve automatic guided vehicles tracking of complex trajectories. Expert Syst. 2022, e13076. [CrossRef] 35. Zhang, Y.; Qian, Y.; Yao, Y.; Hu, H.; Xu, Y. Learning to cooperate: Application of deep reinforcement learning for online AGV path ﬁnding. In Proceedings of the 19th International Conference on Autonomous Agents and Multiagent Systems, Auckland, New Zealand, 9–13 May 2020; pp. 2077–2079. y pp 36. Takahashi, K.; Tomah, S. Online optimization of AGV transport systems using deep reinforcement learning. Bull. Netw. Comput. Syst. Softw. 2020, 9, 53–57. y f , , 37. Popper, J.; Yfantis, V.; Ruskowski, M. References Simultaneous production and agv scheduling using multi-agent deep reinforcement learning. Procedia CIRP 2021, 104, 1523–1528. [CrossRef] 38. Li, M.; Guo, B.; Zhang, J.; Liu, J.; Liu, S.; Yu, Z.; Li, Z.; Xiang, L. Decentralized Multi-AGV Task Allocation based on Multi-Agent Reinforcement Learning with Information Potential Field Rewards. In Proceedings of the 2021 IEEE 18th International Conference on Mobile Ad Hoc and Smart Systems (MASS), Denver, CO, USA, 4–7 October 2021; pp. 482–489. y ( ) pp 39. Zhang, K.; Yang, Z.; Ba¸sar, T. Multi-agent reinforcement learning: A selective overview of theories and algorithms. In Handbook of Reinforcement Learning and Control; Springer: Cham, Switzerland, 2021 ; pp. 321–384. f g p g pp 40. Littman, M.L. Markov games as a framework for multi-agent reinforcement learning. In Machine Learning Proceedings 1994; Elsevier: Amsterdam, The Netherlands, 1994; pp. 157–163. pp 41. Koenig, S.; Likhachev, M. Fast replanning for navigation in unknown terrain. IEEE Trans. Robot. 20 41. Koenig, S.; Likhachev, M. Fast replanning for navigation in unknown terrain. IEEE Trans. Robot. 2005, 21, 354–363. [CrossRef] 42. Lowe, R.; Wu, Y.; Tamar, A.; Harb, J.; Abbeel, P.; Mordatch, I. Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments. In Proceedings of the Neural Information Processing Systems (NIPS), Long Beach, CA, USA, 6–9 December 2017. g p g g [ ] 42. Lowe, R.; Wu, Y.; Tamar, A.; Harb, J.; Abbeel, P.; Mordatch, I. Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments. In Proceedings of the Neural Information Processing Systems (NIPS), Long Beach, CA, USA, 6–9 December 2017. 42. Lowe, R.; Wu, Y.; Tamar, A.; Harb, J.; Abbeel, P.; Mordatch, I. Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments. In Proceedings of the Neural Information Processing Systems (NIPS), Long Beach, CA, USA, 6–9 December 2017. 43. Mordatch, I.; Abbeel, P. Emergence of Grounded Compositional Language in Multi-Agent Populations. arXiv 2017, arXiv:1703.04908. 44. Silver, D.; Lever, G.; Heess, N.; Degris, T.; Wierstra, D.; Riedmiller, M. Deterministic policy gradient algorithms. In Proceedings of the International Conference on Machine Learning, Beijing, China, 21–26 June 2014; pp. 387–395. g g y ( ) g 43. Mordatch, I.; Abbeel, P. Emergence of Grounded Compositional Language in Multi-Agent Populations. arXiv 2017, arXiv:1703.04908. g p g g g p 44. Silver, D.; Lever, G.; Heess, N.; Degris, T.; Wierstra, D.; Riedmiller, M. Deterministic policy gradient algo the International Conference on Machine Learning, Beijing, China, 21–26 June 2014; pp. 387–395. g j g pp 45. 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. References Lillicrap, T.P.; Hunt, J.J.; Pritzel, A.; Heess, N.; Erez, T.; Tassa, Y.; Silver, D.; Wierstra, D. Continuous control with deep reinforcement learning. arXiv 2015, arXiv:1509.02971. 45. Lillicrap, T.P.; Hunt, J.J.; Pritzel, A.; Heess, N.; Erez, T.; Tassa, Y.; Silver, D.; Wierstra, D. Continuous control with deep reinforcement learning. arXiv 2015, arXiv:1509.02971. g 46. Wunder, M.; Littman, M.L.; Babes, M. Classes of multiagent q-learning dynamics with epsilon-greedy exploration. In Proceedings of the 27th International Conference on Machine Learning (ICML-10), Haifa, Israel, 21–24 June 2010; pp. 1167–1174. g 46. Wunder, M.; Littman, M.L.; Babes, M. Classes of multiagent q-learning dynamics with epsilon-greedy exploration. In Proceedings of the 27th International Conference on Machine Learning (ICML-10), Haifa, Israel, 21–24 June 2010; pp. 1167–1174. 47. Dann, C.; Mansour, Y.; Mohri, M.; Sekhari, A.; Sridharan, K. Guarantees for epsilon-greedy reinforcement learning with function approximation. In Proceedings of the International Conference on Machine Learning, Baltimore, MD, USA, 17–23 July 2022; pp. 4666–4689. 47. Dann, C.; Mansour, Y.; Mohri, M.; Sekhari, A.; Sridharan, K. Guarantees for epsilon-greedy reinforcement learning with function approximation. In Proceedings of the International Conference on Machine Learning, Baltimore, MD, USA, 17–23 July 2022; pp. 4666–4689. pp 48. Wadhwania, S.; Kim, D.K.; Omidshaﬁei, S.; How, J.P. Policy distillation and value matching in multiagent reinforcement learning. In Proceedings of the 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Macau, China, 3–8 November 2019; pp. 8193–8200. 48. Wadhwania, S.; Kim, D.K.; Omidshaﬁei, S.; How, J.P. Policy distillation and value matching in multiagent reinforcement learning. In Proceedings of the 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Macau, China, 3–8 November 2019; pp. 8193–8200. 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/W2096956129	https://refubium.fu-berlin.de/bitstream/fub188/15417/1/fnhum-08-00798.pdf	English	null	ECoG high gamma activity reveals distinct cortical representations of lyrics passages, harmonic and timbre-related changes in a rock song	Frontiers in human neuroscience	2,014	cc-by	13,766	Irene Sturm 1,2,3, Benjamin Blankertz 2,4, Cristhian Potes 5,6, Gerwin Schalk 5,6,7,8,9,10 and Gabriel Curio 1,3,4 1 Berlin School of Mind and Brain, Humboldt Universität zu Berlin, Berlin, Germany 1 Berlin School of Mind and Brain, Humboldt Universität zu Berlin, Berlin, Germany 2 Neurotechnology Group, Department of Electrical Engineering and Computer Science, Berlin Institute of Technology, Berlin, Germany 3 Neurophysics Group, Department of Neurology and Clinical Neurophysiology, Charité — University Medicine Berlin, Berlin, Germany 4 Bernstein Focus: Neurotechnology, Berlin, Germany 5 National Resource Center for Adaptive Neurotechnologies, Wadsworth Center, New York State Department of Health, Albany, NY, USA 6 Department of Electrical and Computer Engineering, University of Texas at El Paso, El Paso, TX, USA 7 Department of Neurosurgery, Washington University in St. Louis, St. Louis, MO, USA 8 Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA 9 Department of Neurology, Albany Medical College, Albany, NY, USA 10 Department of Neurosurgery, Washington University in St. Louis, St. Louis, MO, USA 1 Berlin School of Mind and Brain, Humboldt Universität zu Berlin, Berlin, Germany 2 Neurotechnology Group, Department of Electrical Engineering and Computer Science, Berlin Institute of Technology, Berlin, Germany 3 Neurophysics Group, Department of Neurology and Clinical Neurophysiology, Charité — University Medicine Berlin, Berlin, Germany 4 Bernstein Focus: Neurotechnology, Berlin, Germany gy, , y 5 National Resource Center for Adaptive Neurotechnologies, Wadsworth Center, New York State Department of Health, Albany, NY, USA 6 Department of Electrical and Computer Engineering, University of Texas at El Paso, El Paso, TX, USA 7 Department of Neurosurgery, Washington University in St. Louis, St. Louis, MO, USA 8 Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA 9 Department of Neurology, Albany Medical College, Albany, NY, USA 10 Department of Neurosurgery, Washington University in St. Louis, St. Louis, MO, USA 10 Department of Neurosurgery, Washington University in St. Louis, St. Louis, MO, USA Listening to music moves our minds and moods, stirring interest in its neural underpinnings. A multitude of compositional features drives the appeal of natural music. How such original music, where a composer’s opus is not manipulated for experimental purposes, engages a listener’s brain has not been studied until recently. Here, we report an in-depth analysis of two electrocorticographic (ECoG) data sets obtained over the left hemisphere in ten patients during presentation of either a rock song or a read-out narrative. Irene Sturm 1,2,3, Benjamin Blankertz 2,4, Cristhian Potes 5,6, Gerwin Schalk 5,6,7,8,9,10 and Gabriel Curio 1,3,4 First, the time courses of ﬁve acoustic features (intensity, presence/absence of vocals with lyrics, spectral centroid, harmonic change, and pulse clarity) were extracted from the audio tracks and found to be correlated with each other to varying degrees. In a second step, we uncovered the speciﬁc impact of each musical feature on ECoG high-gamma power (70–170 Hz) by calculating partial correlations to remove the inﬂuence of the other four features. In the music condition, the onset and offset of vocal lyrics in ongoing instrumental music was consistently identiﬁed within the group as the dominant driver for ECoG high-gamma power changes over temporal auditory areas, while concurrently subject-individual activation spots were identiﬁed for sound intensity, timbral, and harmonic features. The distinct cortical activations to vocal speech-related content embedded in instrumental music directly demonstrate that song integrated in instrumental music represents a distinct dimension in complex music. In contrast, in the speech condition, the full sound envelope was reﬂected in the high gamma response rather than the onset or offset of the vocal lyrics. This demonstrates how the contributions of stimulus features that modulate the brain response differ across the two examples of a full-length natural stimulus, which suggests a context-dependent feature selection in the processing of complex auditory stimuli. HUMAN NEUROSCIENCE published: 13 October 2014 doi: 10.3389/fnhum.2014.00798 ECoG high gamma activity reveals distinct cortical representations of lyrics passages, harmonic and timbre-related changes in a rock song ORIGINAL RESEARCH ARTICLE bli h d 3 O b 20 ORIGINAL RESEARCH ARTICLE bli h d 3 O b 20 HUMAN NEUROSCIENCE Frontiers in Human Neuroscience Edited by: Edited by: Edmund C. Lalor, Trinity College Dublin, Ireland Reviewed by: Nai Ding, New York University, USA Rebecca Schaefer, University of California, Santa Barbara, USA Correspondence: Irene Sturm, Neurotechnology Group, Fakultät IV, Berlin Institute of Technology, Sek. MAR 4-3, Marchstr. 23, Berlin 10587, Germany e-mail: irene.sturm@charite.de Keywords: music processing, natural music, electrocorticography (ECoG), high gamma, acoustic features Edited by: Edmund C. Lalor, Trinity College Dublin, Ireland Reviewed by: Nai Ding, New York University, USA Rebecca Schaefer, University of California, Santa Barbara, USA Correspondence: Irene Sturm, Neurotechnology Group, Fakultät IV, Berlin Institute of Technology, Sek. MAR 4-3, Marchstr. 23, Berlin 10587, Germany e-mail: irene.sturm@charite.de Keywords: music processing, natural music, electrocorticography (ECoG), high gamma, acoustic features 1. INTRODUCTION 2012), timbre (Deike et al., 2004; Goydke et al., 2004; Caclin et al., 2006, 2007), melodic contour (Trainor et al., 2002), key (Janata et al., 2002), mode (Halpern et al., 2008), scale properties (Brattico et al., 2006), music-syntactic congruity (Koelsch et al., 2002; Sammler et al., 2011, 2013; Jentschke et al., 2014; Kim et al., 2014) and rhythmic aspects (Jongsma et al., 2004; Snyder and Large, 2005; Grahn and Rowe, 2009; Abrams et al., 2011; Schaefer et al., 2011a). Typically, these approaches rely on carefully selected or speciﬁcally designed stimulus material that allows to examine one aspect of music in an isolated manner while controlling for other inﬂuences. This approach has provided a large corpus of The appreciation for music is a universal human capacity that plays an inspiring role in our individual and social lives. Music is processed in a cascade of steps that lead from the segregation within the auditory stream, the extraction and integration of a variety of acoustic features, to cognitive, memory-related pro- cesses that induce personal, often emotional, experiences. Critical structural components of music have been analyzed in studies addressing the processing of pitch (Hyde et al., 2008; Kumar et al., 2011; Nan and Friederici, 2013; Plack et al., 2014), sensory disso- nance (Regnault et al., 2001; Perani et al., 2010; Daikoku et al., October 2014 \| Volume 8 \| Article 798 \| 1 Frontiers in Human Neuroscience www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. processing (Pasley et al., 2012; Kubanek et al., 2013; Leonard and Chang, 2014; Martin et al., 2014) while studies on music percep- tion are still rare. A ﬁrst example how the time course of sound intensity of a naturalistic music stimulus can be tracked in ECoG features was provided by Potes et al. (2012). Speciﬁcally, this study revealed that high-gamma band (70–170 Hz) ECoG activity in the superior temporal gyrus as well as on the dorsal precentral gyrus is highly correlated with the time course of sound inten- sity in a continuous stream of natural music. A subsequent study by Kubanek et al. (2013) found that high-gamma ECoG activ- ity also tracks the temporal envelope of speech and compared it to the activations related to music, identifying different levels of speciﬁcity in an auditory network constituted by the audi- tory belt areas, the superior temporal gyrus (STG) and Broca’s area. Keywords: music processing, natural music, electrocorticography (ECoG), high gamma, acoustic features Very recently, a new analysis of the same data set identi- ﬁed spatial and causal relationships between alpha and gamma ECoG activity related to the processing of sound intensity (Potes et al., 2014). Considering that sound intensity (a technical proxy for perceived loudness) was tracked in ECoG features with signif- icant robustness, the same data set appears highly promising for a further investigation that takes into account the variety of features available in this natural music stimulus, a rock song. evidence about associations between speciﬁc aspects of music and brain areas. Notably, by design, it does not directly address the conﬂuence of the multitude of musical features and their intrinsic relations. Including this integrative aspect of musical composi- tions could contribute to a comprehensive and veridical picture of brain responses to music. Brain responses to naturalistic stimula- tion may differ from those related to controlled stimulation with simpliﬁed stimuli, as suggested by evidence from both the visual and the auditory domain (Hasson et al., 2010 or Abrams et al., 2013, respectively). Abrams et al. provided (to our knowledge) the ﬁrst direct evidence that the between-subject synchronization of a large-scale distributed network including auditory midbrain and thalamus, auditory cortex, parts of frontal and parietal cortex, and motor planning regions was signiﬁcantly higher when listen- ers were presented with complex musical stimuli lasting minutes than when they listened to shorter pseudo-musical contexts. Accordingly, a stimulus should be sufﬁciently long to repre- sent a complex musical context, and the music material should be naturalistic and free of manipulations to approximate ecological validity. These requirements imply a complex, often unbalanced, stim- ulus material and the single presentation of one (long) stimulus without a-priori deﬁned chances for repetition and signal aver- aging. In spite of this challenge for data analysis, the interest in the processing of natural music has recently grown consid- erably. A number of studies using naturalistic music stimuli examine relations between brain signals and behavioral measures, such as autobiographic salience (Janata, 2009), expressive per- formance (Chapin et al., 2010), emotion ratings (Mikutta et al., 2012, 2013), or ratings of perceived tension (Lehne et al., 2014). Keywords: music processing, natural music, electrocorticography (ECoG), high gamma, acoustic features Several approaches that combine neuroimaging and acoustic fea- ture extraction directly investigate the relation between brain signals and the multi-dimensional structure of music (Alluri et al., 2012, 2013; Toiviainen et al., 2014) or investigate the inter- subject synchronization of brain responses to naturalistic music (Abrams et al., 2013; Potes et al., 2014). Only few of these stud- ies used the electroencephalogram (EEG), which due to its high temporal resolution is suitable for investigating the dynamics of music on a ﬁne-grained time scale but typically relies on averag- ing. Apart from (Mikutta et al., 2012, 2013) where EEG recordings are related to behavioral measures, one novel approach to analyze ongoing EEG elicited by natural music stimuli has been proposed in Cong et al. (2012). It allows to identify EEG components that are common to the majority of subjects and, subsequently, com- pares the time course of these components to music features. Common to all approaches mentioned above is that they are only sensitive to effects that occur with a certain degree of (spatial) consistency within the group of subjects. The goal of the present follow-up analysis was to explore whether music-related variables other than sound intensity can be tracked in ECoG and, if so, how respective areas of cortical acti- vation compare to those associated with the processing of sound intensity in Potes et al. (2012). Because a naturalistic music stimu- lus contains different perceptual dimensions that are intrinsically related, it was a critical challenge to differentiate these in the brain response. In addition to the feature of sound intensity that was investigated in the previous studies, we chose four features that relate to different aspects of music. These include the moment- to-moment distinction vocals on/off, a continuous measure of harmonic change probability, a measure related to timbral aspects (spectral centroid), and a rhythm-related measure (pulse clarity) (for details see Materials and Methods). Frontiers in Human Neuroscience 2. MATERIALS AND METHODS 2.1. SUBJECTS AND DATA COLLECTION We analyzed data from ten subjects (for patient’s clinical proﬁles see Table 1 in the Supplemental Data). These 10 subjects included seven of the eight subjects who were analyzed in our previous study (Potes et al., 2012) where patients with epilepsy (4 women, 4 men) were instructed to listen attentively (without any other task) to a single presentation of the rock song “Another Brick in the Wall - Part 1” (Pink Floyd, Columbia Records, 1979) while ECoG activity was recorded. We added to this dataset data from three additional subjects who followed the same protocol. In all patients in the present analysis the electrode grid was in the left hemisphere. All subjects gave informed consent to participate in the study, which was approved by the Institutional Review Board of Albany Medical College. None of the subjects had a history of hearing impairment. The total numbers of implanted electrodes were 96, 83, 109, 58, 120, 58, 59, 112, 134, and 98 for subjects S1 to S10, respectively. After removal of channels containing environ- mental or other artifacts, 86, 82, 104, 56, 108, 57, 53, 93,110, and 92 channels were left for analysis. Grid placement and duration Electrocorticographic recordings (ECoG) from the brain sur- face provide additional beneﬁts since their superior signal-to- noise ratio is advantageous for the analysis of single stimulus presentations at the level of single subjects. They combine high temporal resolution with high spatial resolution. Thus, they offer a much higher level of spatial speciﬁcity and an extended fre- quency range compared to scalp-recorded EEG. In the ﬁeld of speech perception research, ECoG has emerged as a new tech- nique to study the functional cortical organization of speech October 2014 \| Volume 8 \| Article 798 \| 2 www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. of ECoG monitoring were based solely on the requirements of the clinical evaluation without any consideration of this study. Each subject had postoperative anterior–posterior and lateral radio- graphs, as well as computer tomography (CT) scans to verify grid locations. The song was 3:10 min long, digitized at 44.1 kHz in waveform audio ﬁle format, and binaurally presented to each subject using in-ear monitoring earphones (12–23.5 kHz audio bandwidth, 20 dB isolation from environmental noise). 2.4. EXTRACTION OF MUSIC FEATURES Sound intensity was calculated in Matlab (The MathWorks Inc., Natick, Massachusetts). Vocals on/off was determined manually. All other features were extracted using freely available software (see below). We used the ﬁrst 125 s of Pink Floyd’s The Wall - part 1 in the analysis since the last minute of the song is an instrumental afterlude passage with considerably less varia- tion, in particular without any vocal parts.The ﬁve features were calculated as described in the following sections. 2.2. EXTRACTION OF ECoG FEATURES Our analysis focused on the high-gamma band. ECoG activity in the high gamma band has generally been associated with func- tional activation of the cortex in different domains (Crone et al., 2006). For auditory and speech perception, numerous studies have shown that ECoG high-gamma power modulations over auditory areas provide important information about the spatio- temporal dynamics of sound processing (Edwards et al., 2005; Towle et al., 2008; Pei et al., 2011; Pasley et al., 2012; Potes et al., 2014). We extracted ECoG high-gamma power using the same method as in Potes et al. (2012): high-gamma (70–170 Hz) ampli- tudes were extracted by ﬁrst applying a 0.1 Hz high-pass ﬁlter and then a common average reference (CAR) spatial ﬁlter to the ECoG signals. For every 50 ms window, we estimated a power spectrum from the time-series ECoG signal using an autoregressive (AR) model. Spectral magnitudes were averaged for all frequency bins between 70 and 115 and between 130 and 170 Hz (omitting line noise at 120 Hz). 2.4.2. Vocals on/off The presence of vocals was annotated manually in the audio ﬁle. This annotation resulted in a binary function that contained the value 1 for passages with lyrics and 0 otherwise. In the music stimulus there are seven passages with vocal lyrics with aver- age duration of 4.22 s (±0.77) that are separated by at least 5 s of purely instrumental music. In a complementary analysis, we applied a similar procedure to the speech stimuli. Here, 0 was assigned to passages of silence within the story that exceeded the duration of 400 ms, such as pauses between sentences or phrases, while 1 denoted ongoing speech. In the speech stimulus the dura- tion of speech passages was shorter (mean duration 1.65 s ±0.55) and vocals on/off changes occurred more frequently (30 changes in 100 s). In both stimuli the analyzed data start with the ﬁrst tone of the song or with the ﬁrst sentence of the narration, respectively, not including a silent pre-stimulus period. 2. MATERIALS AND METHODS 2.1. SUBJECTS AND DATA COLLECTION ECoG signals were referenced to an electrocorticographically silent elec- trode (i.e., a location that was not identiﬁed as eloquent cortex by electrocortical stimulation mapping), digitized at 1200 Hz, syn- chronized with stimulus presentation, and stored with BCI2000 (Schalk et al., 2004; Schalk and Mellinger, 2010). In addition, we analyzed data from the same subjects where they listened to the presentation of four narrated stories that are part of the Boston Aphasia Battery (Goodglass et al., 1983) (details see Kubanek et al., 2013). Lartillot et al. (2008b) and since then has been used in numerous studies (Eerola et al., 2009; Zentner, 2010; Higuchi et al., 2011; Alluri et al., 2012; Burger et al., 2013). Since an essential char- acteristic of the music stimulus is the presence of song (lyrics), the ﬁfth feature, vocals on/off, captures the change between purely instrumental passages and passages with vocal lyrics content. In summary, we chose a description of the audio signal that relates to important basic variables of the perception of music: loudness, timbre, and rhythm. With harmonic change, it encom- passes also an abstract high-level property related to the rules of Western major-minor harmony. Finally, with vocals on/off, it allows also to address the impact of vocals with lyrics in music. For comparison, in a complementary analysis, the identical anal- ysis was applied to the sound ﬁles of the speech stimuli. 2.4.1. Sound intensity The sound intensity of the audio signal was calculated as the aver- age power derived from 50 ms segments of the audio waveform overlapping by 50%. The resulting time course was downsampled to match the sampling rate of 20 Hz of the extracted ECoG high gamma power. Frontiers in Human Neuroscience 2.3. SELECTION OF MUSIC FEATURES From the large number of potential features that characterize a music audio signal, we chose a set of ﬁve features that capture salient dynamic features of the stimulus and cover a broad spec- trum of structural categories of music. Since the results of Potes et al. (2012) revealed a strong correlation of ECoG high-gamma power ﬂuctuations with the sound intensity of the continuous music stimulus, sound intensity was chosen as ﬁrst feature. It is a temporal feature that can be extracted directly from the raw audio signal and can be considered as an approximate measure of loudness. The second feature was the logarithmic spectral cen- troid, which is perceptually related to the complex property of timbre. More speciﬁcally, it has been related to perceived bright- ness of sound in Schubert et al. (2004) and to perceived pitch level in Coutinho and Cangelosi (2011). The third feature was proba- bility of harmonic change, which relates to higher-level musical structure, i.e., to harmonic progression and musical syntax. Pulse clarity as fourth feature indicates how easily listeners perceive the underlying rhythmic or metrical pulsation of a piece of music. This feature has been introduced and perceptually validated in 2.4.4. Pulse clarity Within the framework of linear regression analysis, the partial correlation coefﬁcient can be derived as the correlation of the residuals that are produced if the interfering variable z (that is to be eliminated) is used as a regressor to predict each of the two variables of interest x and y (Abdi, 2007). The partial correlation coefﬁcient is related to multiple linear regression analysis (MLR), which was applied in Schaefer et al. (2009) in a similar setting to decompose EEG responses into evoked response components that relate to speciﬁc aspects of music stimuli. Furthermore, regression-based models have been applied in a natural speech context in Power et al. (2012) where the auditory evoked spread spectrum analysis (AESPA) method gives a precise account of the temporal dynamics of the transformation of the speech envelope into an EEG signal in single trials. In Ding and Simon (2012) this is extended to an approach that differentiates further between a range of modulation frequencies of the stimulus signal, and, sub- sequently, identiﬁes the slow temporal modulations of speech in a broad spectral region (below 1 Hz) as features that are represented best in the brain response. Importantly, the partial correlation coefﬁcient differs in one important aspect from the semi-partial correlation/regression coefﬁcient of the multiple linear regres- sion framework: The partial correlation coefﬁcient eliminates the inﬂuence of the interfering factor from both variables of inter- est, not only from one (in the framework of MLR: from the regressor). As a consequence, using the partial correlation coef- ﬁcient, shared variance that does not cover a large proportion of the total variance, but may still reﬂect speciﬁc relations, is also detected. In a different context, partial correlation has been applied previously in connectivity analysis of EEG recordings: In Marrelec et al. (2006) it was used as a simple but effec- tive method to identify connections between brain areas while accounting for the effects of volume conduction between elec- trodes. In contrast, here we examine how much each of the ﬁve features of music contributes to the sensor-level ECoG record- ings in a manner that is independent from the remaining four features. Pulse clarity was calculated for windows of 3 s with a 33% overlap using the MIRtoolbox (Lartillot et al., 2008b), then interpolated to match the ECoG sampling frequency of 20 Hz. 2.4.3. Spectral centroid 2.4.3. Spectral centroid The centroid of the log-scaled frequency spectrum was calculated for 50% overlapping windows of 50 ms using the implementation in the MIRtoolbox (Lartillot et al., 2008b). The spectral centroid is the amplitude-weighted mean frequency in a window of 50 ms. It is an acoustic measure that indicates where the “mass” of the spectrum is located. The log-scaled centroid was downsampled October 2014 \| Volume 8 \| Article 798 \| 3 www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. rxy.z = rxy −rxzryz (1 −rxz)2(1 −ryz)2 (1) to match the sampling rate of 20 Hz of the extracted ECoG high gamma power. (1) 2.4.5. Harmonic change h h h f The harmonic change function measures the probability of a har- monic change and detects chord changes. We derived this metric using the Queen Mary plugin for the sonic visualizer (del Bimbo et al., 2010), which implements an algorithm that was proposed and validated on a selection of rock songs in Harte et al. (2006). The algorithm comprises a segmentation of the audio signal into 50 ms windows, spectral decomposition of each window, assign- ment of chroma and a tonal centroid to each window. After that, the tonal distance between consecutive frames is calculated based on a hypertoroid model of tonal space proposed by Chew (2000). Figure 2 gives a visual representation of each stimulus’ spec- trogram, an annotation of lyrics and chords or text and the time courses of the ﬁve extracted music features for a 12 s-segment. 2.4.4. Pulse clarity Within the framework of linear regression analysis, the partial correlation coefﬁcient can be derived as the correlation of the residuals that are produced if the interfering variable z (that is to be eliminated) is used as a regressor to predict each of the two variables of interest x and y (Abdi, 2007). The partial correlation coefﬁcient is related to multiple linear regression analysis (MLR), which was applied in Schaefer et al. (2009) in a similar setting to decompose EEG responses into evoked response components that relate to speciﬁc aspects of music stimuli. Furthermore, regression-based models have been applied in a natural speech context in Power et al. (2012) where the auditory evoked spread spectrum analysis (AESPA) method gives a precise account of the temporal dynamics of the transformation of the speech envelope into an EEG signal in single trials. In Ding and Simon (2012) this is extended to an approach that differentiates further between a range of modulation frequencies of the stimulus signal, and, sub- sequently, identiﬁes the slow temporal modulations of speech in a broad spectral region (below 1 Hz) as features that are represented best in the brain response. Importantly, the partial correlation coefﬁcient differs in one important aspect from the semi-partial correlation/regression coefﬁcient of the multiple linear regres- sion framework: The partial correlation coefﬁcient eliminates the inﬂuence of the interfering factor from both variables of inter- est, not only from one (in the framework of MLR: from the regressor). As a consequence, using the partial correlation coef- ﬁcient, shared variance that does not cover a large proportion of the total variance, but may still reﬂect speciﬁc relations, is also detected. In a different context, partial correlation has been applied previously in connectivity analysis of EEG recordings: In Marrelec et al. (2006) it was used as a simple but effec- tive method to identify connections between brain areas while accounting for the effects of volume conduction between elec- trodes. In contrast, here we examine how much each of the ﬁve features of music contributes to the sensor-level ECoG record- ings in a manner that is independent from the remaining four features. 2.4.4. Pulse clarity Pulse clarity is a measure of how strong rhythmic pulses and their periodicities can be perceived by the listener. It is based on the relative Shannon entropy of the ﬂuctuation spectrum (Pampalk et al., 2002) and has been perceptually validated as being strongly related to lis- tener’s perception of the degree of rhythmicity in a piece of music in Lartillot et al. (2008a). 2.5. ANALYSIS The ﬁve features that we used to describe the music stimulus are not independent of each other, but are correlated with each other to variable degrees (see Figure 1). Only by accounting for this correlation, one can attribute a particular ECoG signal to one par- ticular music feature (Kendall et al., 1973). This post-hoc approach is a way to exert statistical control over variables in a setting where experimental control on the different aspects that are to be inves- tigated is ruled out by design. The partial correlation coefﬁcient is given by Equation (1). FIGURE 1 \| Correlation between ﬁve stimulus features: left: music stimulus, right: speech stimulus. FIGURE 1 \| Correlation between ﬁve stimulus features: left: music stimulus, right: speech stimulus. Frontiers in Human Neuroscience October 2014 \| Volume 8 \| Article 798 \| 4 www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. Sturm et al. of signiﬁcant correlations at the group-level, we visualized the results as following: for each subject, we determined electrodes with signiﬁcant correlation and projected their positions onto the MNI brain. To detect activated electrodes in similar regions, each of these electrodes was represented by a round patch of activation with radius 10 mm centered around its position. These represen- tations were added up for the 10 subjects, resulting in a map showing the topographical overlap of the presence of signiﬁcant correlation within the group of subjects. Values range from zero (no signiﬁcant correlation in all ten subjects) to ten (signiﬁcant correlation in all ten subjects). The degree of overlap is deter- mined by the radius around an electrode (10 mm). Since grid placement was determined by clinical requirements and, conse- quently, varied between patients, we needed to account for the fact that the maximal number of subjects who can contribute to the group-level overlap of activation also varies between brain regions. Therefore, we determined the group-level overlap of grid coverage on the MNI brain, referred to as grid coverage index in the following, for all electrodes. Using the grid coverage index, a normalized group-level overlap in a speciﬁc cortical area can be obtained by dividing the (unnormalized) group-level overlap by the grid coverage index for each vertex. However, even the nor- malized group-level overlap values cannot be used for inferring group-level statistics, for instance to assess differences between brain areas. 2.5. ANALYSIS Nonetheless, this does not affect the primary goal of the present analysis, which is to explore potential differences in one location between features and also between the conditions music and pure speech. For distinct foci of high degree of group- level overlap, we determined representative coordinates on the MNI brain manually, and derived the corresponding Brodmann areas using the Talairach Atlas daemon1. Owing to the variance introduced by the projection of each subject’s individual brain onto the MNI brain and to the blurring effect that the above men- tioned procedure of determining group-level overlap may cause, this procedure yields only an approximate localization of cor- tical activation. Notwithstanding, on the scale of the Brodman area, this level of exactness appear appropriate for comparing the present results with the literature. It is important to recognize that both ECoG features and the extracted music features have an autocorrelation, i.e., subsequent samples are not independent of each other. This fact violates the assumptions that underlie the standard tests for signiﬁcance of correlation. To account for this issue, we assessed the signiﬁcance of the partial correlation coefﬁcients by applying randomized per- mutation tests with surrogate data as proposed in Theiler et al. (1992). For each music feature, we generated a surrogate target function by transforming the time domain signal into the fre- quency domain, randomly permuting its phase spectrum, and reconstructing the time domain signal using the original spectral amplitudes and the permuted phases. After that, we calculated the correlation coefﬁcient between the ECoG feature and this surro- gate target function. We repeated this process 1000 times, which resulted in a distribution of correlation coefﬁcients for the sur- rogate data. We then asked how likely the observed correlation coefﬁcient was to be produced by this surrogate distribution of correlation coefﬁcients. The resulting p-values were corrected for multiple compar- isons within all electrodes [false discovery rate (FDR), q < 0.05]. We then plotted the negative logarithm of the corrected p-values for each electrode on each subject’s brain model as an indicator of how much brain activity at a particular site was related to a speciﬁc acoustic feature. Since we did not observe negative cor- relation coefﬁcients, there was no need to distinguish between negative and positive correlation. Naturally, one would expect that a cortical brain response that tracks features of an auditory stimulus will not respond instan- taneously, but delayed. 1http://www.talairach.org/daemon.html. Frontiers in Human Neuroscience 2.5. ANALYSIS Accordingly, we examined the channel- wise partial correlation coefﬁcients with time lags up to 300 ms. However, this resulted in cross-correlation sequences that varied only on a very small scale over time and were not conclusive with respect to an optimal time lag, suggesting that a time lag between stimulus and brain response may be evened out by our sampling rate of 20 Hz. For instance, selecting a biologically plausible time lag of 100 ms, based on Kubanek et al. (2013) where the opti- mal (averaged) time lag for tracking the speech envelope ranged between 86.7 and 89.9 ms, had only an marginal effect on the signiﬁcance of correlation coefﬁcients, although the magnitude of correlation coefﬁcients varied slightly (but not systematically). An overview of the group-level results for different time lags is depicted in Figure S3. On these grounds it would have been arbi- trary to deﬁne a ﬁxed time lag for the analysis and, moreover, a chosen time lag would not have been informative. Therefore, we decided to calculate instantaneous correlation coefﬁcients in the present analysis, using this is a neutral or ’null’ hypothesis given that no signiﬁcant estimate of a biologically plausible time lag was obtainable for this data set. For a detailed analysis of latencies, in particular with respect to differences in the processing of differ- ent aspects of music, as suggested in Schaefer et al. (2011b), our approach is not appropriate since the dependencies between the ﬁve features play a role in calculating the partial correlation coef- ﬁcients for one music feature and the ECoG signal. This could be a topic for a future investigation, for instance applying methods proposed in Bießmann et al. (2010) or Power et al. (2012). 3. RESULTS Correlation with spectral centroid is signiﬁcant only in subject S5 and S10 and distributed similarly to the feature vocals on/off, but spatially less extended. For harmonic change, signiﬁcant correlation is present in four subjects (subject S3, S5, S9, and S10) on the posterior STG and in subject S3 in frontal areas. The correlation with pulse clarity reaches signiﬁcance in only one subjects (S6) in a small region on the precentral cortex. FIGURE 3 \| Grid coverage index: Overlap of grid coverage on MNI brain. Figure 5 depicts the cortical distribution of signiﬁcant par- tial correlation of ECoG high-gamma features with each of the ﬁve acoustic features for the natural speech stimuli at the level of each individual patient. Differing from the music condition, the feature that is reﬂected most consistently within the group is sound intensity with signiﬁcant correlation in 6/10 subjects (S1, S2, S3, S4, S5, S9, and S10). In all of them, the focus of correla- tion is located on the pSTG. Beyond that, signiﬁcant correlation is present on the inferior/medial temporal gyrus (S1, S2), on the inferior frontal gyrus (S3) and on the precentral cortex (S5). For the feature spectral centroid, signiﬁcant correlations are present only in three subjects on the superior and medial temporal gyrus. Of these, subject S10 is the only subject with signiﬁcant corre- lation for spectral centroid in, both the music and the speech condition. For harmonic change, signiﬁcant correlation is present only in subject S4 on the inferior frontal gyrus (IFG). For pulse clarity, no signiﬁcant correlation with ECoG high gamma features is present. FIGURE 3 \| Grid coverage index: Overlap of grid coverage on MNI brain. superior temporal gyrus and the ventral parts of the precentral and postcentral gyri. gy Figure 4 shows the signiﬁcance values of partial correlation of ECoG high-gamma features with each of the ﬁve music features for each individual patient. Signiﬁcant high-gamma correlations with vocals on/off are present in 9/10 of the subjects, and exceed in spatial extent those of all other acoustic features. In all of these nine patients, signiﬁcant positive correlations are present in audi- tory areas around the Sylvian ﬁssure, either conﬁned to a region on the posterior superior temporal gyrus (pSTG) (S1, S2, S4, S6, and S8), or extending also to the anterior part of the STG and dor- sally from the Sylvian ﬁssure (S3, S5, S9, and S10). 3. RESULTS Figure 1 shows a confusion matrix. For each element in this matrix, the brightness gives the correlation between two par- ticular music features. In the music stimulus vocals on/off is strongly correlated with spectral centroid (r = 0.69) and intensity (r = 0.37), which conﬁrms the necessity for calculating partial correlations. Figure 2 gives a visual representation of each stimulus’ spec- trogram, an annotation of lyrics and chords or text and the time courses of the ﬁve extracted music features for a 12 s-segment as well as the time course of ECoG high gamma power, measured at one representative electrode in one subject. Figure 3 documents the overlap of grid coverage (grid cov- erage index) within the group of patients. The regions covered in all of the 10 subjects comprise the posterior part of the Since these measures of signiﬁcance cannot be directly aver- aged across subjects, to examine the topographical distribution October 2014 \| Volume 8 \| Article 798 \| 5 www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. FIGURE 2 \| Spectrogram of a segment (12 s) of the music/speech recording, lyrics/text, and chord annotations and time courses of the ﬁve analyzed features. For comparison with the time course of the music features the time course of ECoG high gamma power, measured at one representative electrode of subject S5 was added below. The location of the electrode is indicated on the brain model on the right panel. FIGURE 2 \| Spectrogram of a segment (12 s) of the music/speech recording, lyrics/text, and chord annotations and time courses of the ﬁve analyzed features. For comparison with the time course of the music FIGURE 2 \| Spectrogram of a segment (12 s) of the music/speech recording, lyrics/text, and chord annotations and time courses of the ﬁve analyzed features. For comparison with the time course of the music features the time course of ECoG high gamma power, measured at one representative electrode of subject S5 was added below. The location of the electrode is indicated on the brain model on the right panel. FIGURE 3 \| Grid coverage index: Overlap of grid coverage on MNI brain. parameters, including vocals on/off) is low, reaching signiﬁcance only in subject S4, S5, S7, and S10) and is detected only in a smaller region on the posterior STG. Frontiers in Human Neuroscience Frontiers in Human Neuroscience 3. RESULTS In addition, signiﬁcant correlation in an isolated area at the dorsal precentral cortex is present in three subjects (S3, S5, and S9). Compared to the effect related to vocals on/off, correlation with sound intensity (after calculating the partial correlation and thereby rendering it now independent from ﬂuctuations in the other four acoustic The top row of Figure 6 shows the group-level overlap of sig- niﬁcant “standard” correlation (Pearson’s correlation coefﬁcient without partialing out the other features) of high-gamma ECoG features with each of the ﬁve music features, i.e., including inﬂu- ences of the other features, on the MNI brain. Common to all patterns except pulse clarity is a focus of signiﬁcant correlation in peri-Sylvian areas that is present in all ten subjects for the features sound intensity, vocals on/off, spectral centroid, in six subjects for harmonic change. The pattern for pulse clarity is most extended, and shows a large spatial variability of activation October 2014 \| Volume 8 \| Article 798 \| 6 www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. 4 \| Single subjects (individual brain models), music condition: distribution of signiﬁcant correlation with each of the ﬁve features after removing the inﬂuence of the remaining four features by calculating partial correlation coefﬁcients. A value of corresponds to a p-value of 0.01. Correlation coefﬁcients determined signiﬁcant by permutation tests ranged between r = 0.07 and r = 0 features by calculating partial correlation coefﬁcients. A value of 2 corresponds to a p-value of 0.01. Correlation coefﬁcients determined as signiﬁcant by permutation tests ranged between r = 0.07 and r = 0.26. FIGURE 4 \| Single subjects (individual brain models), music condition: Cortical distribution of signiﬁcant correlation with each of the ﬁve acoustic features after removing the inﬂuence of the remaining four October 2014 \| Volume 8 \| Article 798 \| 7 Frontiers in Human Neuroscience www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. 5 \| Single subjects (individual brain models), speech condition: distribution of signiﬁcant correlation with each of the ﬁve c features after removing the inﬂuence of the remaining four features by calculating partial correlation coefﬁcients. A value of corresponds to a p-value of 0.01. Correlation coefﬁcients determined signiﬁcant by permutation tests ranged between r = 0.06 and r = 0. features by calculating partial correlation coefﬁcients. A value of 2 corresponds to a p-value of 0.01. Frontiers in Human Neuroscience 3. RESULTS Correlation coefﬁcients determined as signiﬁcant by permutation tests ranged between r = 0.06 and r = 0.16. FIGURE 5 \| Single subjects (individual brain models), speech condition: Cortical distribution of signiﬁcant correlation with each of the ﬁve acoustic features after removing the inﬂuence of the remaining four FIGURE 5 \| Single subjects (individual brain models), speech condition: Cortical distribution of signiﬁcant correlation with each of the ﬁve acoustic features after removing the inﬂuence of the remaining four October 2014 \| Volume 8 \| Article 798 \| 8 Frontiers in Human Neuroscience Frontiers in Human Neuroscience www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. FIGURE 6 \| Number of participants with effects visualized on the MNI brain. The color code indicates the degree of group-level overlap. Top: Music, “standard” correlation. Middle: Music, partial correlation. Bottom: Speech, partial correlation. FIGURE 6 \| Number of participants with effects visualized on the MNI brain. The color code indicates the degree of group-level overlap. Top: Music, “standard” correlation. Middle: Music, partial correlation. Bottom: Speech, partial correlation. that do not overlap in more than three patients. In general, at a descriptive level, the similarity between cortical overlap pat- terns mirrors the correlation matrix of the music features in that they mainly document the interdependence of musical fea- tures rather than allowing to differentiate between processing of speciﬁc dimensions of music. on the middle part of the STG, corresponding to the respec- tive location in the music condition, but with ﬁve contributing subjects, compared to three subjects in the music condition. However, for all other features the degree of overlap does not exceed two subjects in any location. Figure S2 shows the group-level overlap depicted in Figure 6, normalized with respect to the grid coverage index depicted in Figure 3. We included only cortical sites with a minimum grid coverage of 2 subjects. This representation demonstrates that the characteristic patterns of the group-level overlap representation (Figure 5) do not merely reﬂect the distribution of the grid cov- erage index, but that the distribution of signiﬁcant correlation has features that are consistently present in a large proportion of the subjects in which grid coverage is given. Frontiers in Human Neuroscience 3. RESULTS The middle row of Figure 6 gives the group-level overlap of signiﬁcant correlation of high-gamma ECoG features with each of the ﬁve music features after the inﬂuence of the remaining four other features has been removed by calculating partial correla- tions (see Materials and Methods). The highest degree of overlap is present in the feature vocals on/off with signiﬁcant correla- tion of high-gamma power with vocals on/off in more than seven subjects around the Sylvian ﬁssure, covering the posterior and middle part of the superior temporal gyrus and of the middle temporal gyrus. The point of most consistently detected activa- tions in the present group of subjects is the posterior part of the superior temporal gyrus (9/10 subjects). Furthermore, over- lap of signiﬁcant correlation is present in the precentral gyrus in three subjects. For all other features, the group-level overlap is considerably less: for sound intensity, there is a common focus of activation in the anterior peri-Sylvian area in three patients. Locations of signiﬁcant correlation for harmonic change vary along the STG, amounting to a number of three overlapping subjects at the maximum. Signiﬁcant correlation with spectral centroid is distributed around the Sylvian ﬁssure, however with minimal inter-individual overlap. Figure S3 shows the group-level overlap of signiﬁcant correla- tion for delays of 0, 50, 100, 150, 200, 250, and 300 ms of the time course of the ECoG high gamma power and the music features. 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI Considering that speech and music can both be characterized by similar concepts relating to temporal structure (onset struc- ture and syllable structure), pitch-related structure (melody and prosody) and timbral aspects, it is an interesting question whether an identical description of a speech stimulus leads to similar reﬂection of extracted acoustic features in the high-gamma band. Therefore, we applied the same analysis to ECoG recordings of stimulation with natural speech [four narratives which are part of the Boston Aphasia Battery (Goodglass et al., 1983), for details see Kubanek et al., 2013 where this data set was analyzed previously]. In the speech condition, the feature vocals on/off was reﬂected much less than in the music condition, in two subjects on the superior temporal gyrus (STG) and in one subject in the posterior part of the middle lateral temporal lobe. Contrastingly, reﬂections of sound intensity were present in seven subjects with a common focus on the middle part of the STG and individually different distributed effects on the posterior part of the STG, on the MTG, on the dorsal precentral cortex, and on the inferior frontal gyrus. One the one hand, the present distinct cortical reﬂection of temporal information in speech agrees with the essential role of the sound envelope in speech understanding that has been Considering that speech and music can both be characterized by similar concepts relating to temporal structure (onset struc- ture and syllable structure), pitch-related structure (melody and prosody) and timbral aspects, it is an interesting question whether an identical description of a speech stimulus leads to similar reﬂection of extracted acoustic features in the high-gamma band. Therefore, we applied the same analysis to ECoG recordings of stimulation with natural speech [four narratives which are part of the Boston Aphasia Battery (Goodglass et al., 1983), for details see Kubanek et al., 2013 where this data set was analyzed previously]. An activation of the dorsal precentral cortex in auditory per- ception of singing/speaking as well as in covert production has been observed before (Callan et al., 2006; Sammler et al., 2010) and was associated either with internal singing, or, more generally, with an activation of (pre)motor codes upon perception of song or speech. Cortical representations of a rock song in electrocorticography Sturm et al. Beyond the impact of vocals on/off, a speciﬁc reﬂection of the features spectral centroid and harmonic change is present on the STG. For the spectral centroid, foci of signiﬁcant correlation are present in 2/10 subjects (Figure 4). In both of them, they are located on the posterior part of the STG, which includes a part of Heschl’s gyrus (BA 41 and 42). This particular area has been related to auditory processing in general, but speciﬁcally also to frequency (Liebenthal et al., 2003), pitch (Patterson et al., 2002), and harmonic tones (Zatorre and Krumhansl, 2002). The reﬂec- tion of the spectral centroid on the STG is in line with Alluri et al. (2012) where the ﬂuctuation of brightness (a component of timbre) in a natural music stimulus correlated with the BOLD response in the STG/MTG. For harmonic change, the focus of signiﬁcant correlation is distributed similarly to spectral centroid extending also to Wernicke’s area (posterior part of BA 22). those with lyrics content) exceeded by far any of the other features in spatial extent and was consistently present within the group of subjects. Furthermore, distinct cortical patterns of signiﬁcant cor- relation with the features sound intensity, spectral centroid and harmonic change were present in single subjects. Frontiers in Human Neuroscience 4.1. REFLECTION OF ASPECTS OF MUSIC IN ECoG HIGH GAMMA POWER These results demonstrate that in this example of a rock song, the change between purely instrumental passages and those with vocal lyrics content is the compositional feature that exerts the most marked effect on the electrocorticographic brain response in the high-gamma frequency band. In nine of ten patients, sig- niﬁcant correlation of the high-gamma ECoG features with vocals on/off was present on the superior temporal gyrus. This led to a core region of high group-level overlap of signiﬁcant corre- lation along the middle and posterior superior temporal gyrus (including Brodman areas 22, 41, and 42, see Figure 6, second column), particularly, extending posteriorly toward Wernicke’s area (posterior BA 22 and BA 40) and also dorsally from the Sylvian ﬁssure. In three subjects, signiﬁcant correlation was also present on the dorsal precentral cortex (BA 4) and in two sub- jects on the inferior frontal gyrus near Broca’s area (BA 45). Thus, the very onset of speech-related vocal content within a stream of music has a major impact on the brain response, as it affects stimulus-related neural activity distributed across several distinct brain regions. Considering that the partial correlation approach has removed the inﬂuence of the co-ﬂuctuating four other factors that relate to sound intensity, to timbral (spectral centroid) or rhythmic characteristics (pulse clarity), and to harmonic struc- ture, the remaining signiﬁcant correlation could be related to speech-related aspects not addressed speciﬁcally in the analysis, such as linguistic-semantic aspect or the presence of the human voice that has been found to effect ECoG gamma activity even for (unintelligible) reversed speech in Brown et al. (2014). A more speciﬁc interpretation of this effect cannot be derived in the present context of a naturalistic complex music stimulus. The topography of the speech-related neural activity during listening to music is in line with Merrill et al. (2012), where the left supe- rior temporal gyrus was found to code for the difference between (normal) speech and speech without words (hummed speech prosody) as well as for the difference between song with lyrics and song without lyrics (hummed). Furthermore, a differential BOLD response of the STG (bilaterally) for music with lyrics vs. instrumental music was observed in Brattico et al. (2011). The left mid-superior temporal sulcus was found to reﬂect (variable degrees of) integrated processing of lyrics and tunes (Sammler et al., 2010). For sound intensity, signiﬁcant correlation is present in 4/10 subjects. 4. DISCUSSION The present study explored the processing of complex natu- ral music by examining relations between ECoG high-gamma band power and ﬁve features of music and thereby extends ﬁnd- ings by Potes et al. (2012). To varying degree, these features co-ﬂuctuate in original (unmodiﬁed) music and were found to produce similar cortical distributions of signiﬁcant correlation in the high-gamma band. To address this issue, we calculated par- tial correlation coefﬁcients to assess the unique impact of each of the ﬁve features (for comparison of standard and partial correla- tion see Figure 6, top and middle). Signiﬁcant correlation of the high-gamma band ECoG features with the feature vocals on/off (indicating the change between purely instrumental passages and The bottom row of Figure 6 shows the group-level overlap of signiﬁcant correlation for complementary analysis of speech-only stimuli. The overlap of signiﬁcant correlation with sound inten- sity is distributed around the Sylvian ﬁssure with highest values October 2014 \| Volume 8 \| Article 798 \| 9 www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography 4.1. REFLECTION OF ASPECTS OF MUSIC IN ECoG HIGH GAMMA POWER The maximal group-level overlap of signiﬁcant correla- tion is located on the anterior STG. Since pulse clarity was found to be reﬂected in the listener’s BOLD response very clearly in the bi-lateral STG, insula and sup- plementary motor areas in Alluri et al. (2012), detecting a related reﬂection in the ECoG high gamma response would in principle have been possible with the present setup. Even though pulse clar- ity is (relatively within our feature set) uncorrelated with the other four features, signiﬁcant correlation was present in only one sub- ject and was conﬁned to a small area. This suggests that changes in pulse clarity did not have a speciﬁc reﬂection in the ECoG high gamma response. A speculative explanation for the absence of rhythm-related effects in the brain response might include that acoustic features that have been found to be typical for “high- groove” music, such as clear pulses, high energy in low frequency bands, and high beat salience (Madison et al., 2011) may not be very salient in Pink Floyd’s The Wall, part 1 (not to be confused with Pink Floyd’s The Wall, part 2 that has strong drums). 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI In the speech condition, for spectral centroid the group-level overlap of signiﬁcant correlation is located on the pSTG, simi- larly to that in the music condition, but (apart from subject S10) different subjects contribute. This suggests that ﬂuctuation of the spectral centroid is reﬂected on the posterior STG consistently in both conditions in single subjects, but, at the same time, that it is individually different for which stimulus and in whom these reﬂections reach signiﬁcance. The present results complement those of Kubanek et al. (2013) where high-gamma ECoG activity was reported to track the tem- poral envelope of natural speech stimuli in non-primary areas of the auditory cortex: in the superior temporal gyrus (STG) and on the inferior frontal gyrus (IFG) near Broca’s area. On the other hand, the temporal envelope of music including song (condition “lyrics,” represented by the concatenated periods of singing) and purely instrumental music (condition “melody,” represented by the purely instrumental periods) was tracked considerably weaker by high-gamma ECoG activity and only in the auditory belt area. This suggested a speciﬁcity of STG and IFG for speech-related processing and conﬁrmed the importance of temporal infor- mation in speech. The present results help to further elucidate the previous results insofar that they demonstrate that not only the sound envelope is encoded weaker in the music condition, but that the alternating presence/absence of vocals is represented predominantly. In the speech condition, a reﬂection of the feature harmonic change is present in ECoG high-gamma power in one subject only while in the music condition 4/10 subjects showed a focus of sig- niﬁcant correlation on the pSTG. The presence of an effect in the music condition demonstrates that the model-based extrac- tion of harmonic change in music indeed has a physiological reﬂection. Notably, the harmonic change algorithm is tailored to measure the distance of chords within a tonal space, a metric that is not necessarily applicable to speech sounds. However, since speech also contains harmonic content, we hypothesized that, e.g., changes in the voice fundamental (F0) are extracted by the algorithm to some extent and may lead to a similar representation in the ECoG high gamma-power. Accordingly, the application of the harmonic change algorithm to the speech stimulus resulted in a time course showing a variance comparable to that in the music condition. 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI The absence of such a representation in ECoG high gamma response (with exception of one subject) suggests that the music-speciﬁc extraction of harmonic features from the speech signal does not convey information triggering co-varying cortical processing. Since research on neural processing of natural music is het- erogeneous with respect to data-recording techniques, stimulus material, music features, and methods of data analysis, it is difﬁ- cult to directly compare results of different studies. With respect to physiology, the present data reveal that in addition to the alpha and theta frequency bands, which have been found to reﬂect dynamic rhythmic features of music (Cong et al., 2012), the high-gamma band carries information about the music stimulus. To summarize, in this differential analysis of ﬁve features of a natural music stimulus, we found the on/offset of vocal lyrics to be the dominant driver of ECoG high-gamma power on the STG (mostly in the posterior part) and in peri-Sylvian areas consistently within the group of subjects. In parallel, in single subjects, sound intensity, harmonic change, and spectral cen- troid produced speciﬁc high gamma reﬂections in the same brain area. In the speech condition, topographically similar effects for sound intensity were present most consistently, and in single sub- jects for vocals on/off, harmonic change and spectral centroid. In general, these ﬁndings are in line with the assumed involve- ment of the pSTG in the intermediate stage of auditory processing (Pasley et al., 2012), more speciﬁcally in the selective extraction of spectro-temporal features relevant for auditory object recogni- tion. Thus, the observed different activations between music and speech may demonstrate differences in the relative importance of Our analysis was focused on the high gamma frequency range, based on initial results that were not informative in lower fre- quency bands. Negative correlation between ECoG power in lower frequency bands (8–12 and 18–24 Hz) and sound inten- sity was reported in Potes et al. (2012). Recently, Potes et al. (2014) showed in a detailed analysis that envelope-related high gamma activity (70–110 Hz) in areas close to primary auditory cortex, in peri-Sylvian and superior pre-motor areas precedes and predicts envelope-related alpha band activity (8–12 Hz) near pri- mary auditory areas that have been found to be the target of afferent auditory projections from the thalamus. In light of the hypothesis of Kumar et al. (2007) and Zatorre et al. 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI In particular, since the presentation of a full-length well-known rock song resembles a natural listening experience at least with respect to the stimulus material, one may speculate that patients might have silently sung along. One the one hand, the present distinct cortical reﬂection of temporal information in speech agrees with the essential role of the sound envelope in speech understanding that has been October 2014 \| Volume 8 \| Article 798 \| 10 www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. established by clinical results (Rosen, 1992; Drullman et al., 1994; Zeng et al., 1999; Lorenzi et al., 2006). On the other hand, the considerably weaker reﬂection in music suggests that, if speech- related content is embedded as song in music, the impact of this feature may be overruled by the change from instrumen- tal to vocal/lyrics sound. Together with the fact that the contrast between silence and speech in the speech condition effected less change-related high gamma activity, this may indicate that the pSTG (where vocals on/off in the music condition was reﬂected most consistently), is responsive to vocals/lyrics in the context of music. Thus, it may be associated with identifying vocal speech- related content within the context of other complex sounds, reﬂecting a speciﬁc aspect of auditory scene analysis. the features in both stimuli. Tentatively, one may explain the inter- individual differences with respect to the presence of effects for harmonic change, spectral centroid and sound intensity with dif- ferent contributions of the ﬁve features to an individual’s listening experience. The present results differentiate further the pioneering work of Potes et al. (2012) where ECoG high-gamma features were found to trace sound intensity in two distinct regions, in the posterior STG and in the inferior frontal gyrus. The present follow-up anal- ysis helps to attribute this effect mainly to the presence of vocal speech in the stimulus while the effect of sound intensity proper is found much weaker and conﬁned to a smaller region on the STG. In addition, with spectral centroid and harmonic change, we identiﬁed two further aspects speciﬁc for music that have an impact on the high-gamma ECoG response in some subjects. Notwithstanding, in these single subjects, these effects are highly signiﬁcant and derived from one presentation of the stimulus. Frontiers in Human Neuroscience 4.3. METHODOLOGY 4.3. METHODOLOGY With respect to methodology, we approach a typical problem that arises when assessing the relationship between brain recordings and natural auditory stimuli. We address the problem of interde- pendent features of natural complex stimuli that complicate the correlation-based analysis of the relation between brain signals and stimulus features. Recently, in Alluri et al. (2013), this chal- lenge has been faced by applying principal component regression modeling where the interrelated multi-dimensional description of a music signal is transformed into a lower-dimensional space of uncorrelated components that are subsequently perceptually eval- uated. Here, operating on the original features, we demonstrate that a partial correlation approach as an extension of multiple linear regression analysis (Schaefer et al., 2009) can be used to differentiate between the processing of aspects of natural music. Typically, in a naturalistic setting, multi-channel measure- ments are related (e.g., by correlation measures) to a multi- dimensional description of music, a situation which is prone to produce false positive effects such as spurious correlations. One way of constraining the solution is to assume inter-individual consistent spatial distribution of neural activity, e.g., by averag- ing the EEG time course across subjects (Schaefer et al., 2009) or by selecting components that are common to the majority of sub- jects (Alluri et al., 2012; Cong et al., 2012). The present results are an important complement to previous studies, as they were obtained at the single-subject level and for one single stimulus presentation owing to the ECoG’s characteristics of offering both high temporal and spatial resolution. This sensitivity helped to reveal a considerable variability between subjects with respect to reﬂected features of music, an insight that suggests that assuming within-group consistency might neglect some of these individual effects. With respect to methodology, we approach a typical problem that arises when assessing the relationship between brain recordings and natural auditory stimuli. We address the problem of interde- pendent features of natural complex stimuli that complicate the correlation-based analysis of the relation between brain signals and stimulus features. Recently, in Alluri et al. (2013), this chal- lenge has been faced by applying principal component regression modeling where the interrelated multi-dimensional description of a music signal is transformed into a lower-dimensional space of uncorrelated components that are subsequently perceptually eval- uated. 4.4. CURRENT EXPERIMENTAL LIMITATIONS Obviously, there are limitations of what can be achieved with this approach. Typical for ECoG recordings, the data were recorded from epilepsy patients whose physical and cognitive conditions were impaired to different degrees and whose brains may not be comparable to that of the healthy population in function and neuroanatomy. Furthermore, grid coverage was determined by clinical reasons and thus varied between subjects. Important issues, such as hemispheric specialization for speech and music, cannot be addressed with the present data set of left-hemispheric recordings. Another important issue is, that information about the patients’ music preference, cultural background and musical training that could give valuable clues for interpreting inter- personal differences is not available in this follow-up analysis. ACKNOWLEDGMENT h k This work was supported by grants from the US Army Research Ofﬁce (W911NF-07-1-0415, W911NF-08-1-0216, W911NF-12- 1-0109) and the NIH/NIBIB (EB006356 and EB000856). We gratefully acknowledge Dr. Ritaccio for his help with patient inter- actions, Drs. Peter Brunner and Aysegul Gunduz for their help with data collection, and the cooperation of the subjects in this study. Irene Sturm’s work was funded by the Berlin School of Mind and Brain and also by the Christiane Nüsslein-Volhard foundation. SUPPLEMENTARY MATERIAL 5. CONCLUSION related to cortical extraction of complex auditory features in non- primary auditory areas. According to this model, the high gamma frequency range would be suitable starting point for a differen- tial analysis of the processing of higher-level auditory features. However, extending the present analysis to lower frequency bands in the future could provide additional information. In a natural speech context, in particular phase information in low-frequency bands of the brain signal has been found to be informative in Zion Golumbic et al. (2013) and Ding and Simon (2012). However, our analysis is an example of what can be achieved within these limits and contributes to the growing body of methodological approaches for research on the processing of nat- ural music. Partial correlation, proposed here as one solution for inter-dependence of stimulus features, has detected speciﬁc reﬂections of music features in the ECoG high-gamma response. However, it has to be kept in mind that this method gives a differ- ential picture of each music feature’s impact on the brain response showing cortical reﬂections that are unique to this feature beyond all others in the feature set. Thus, for a given feature, the por- tion of independent variance from the other features is crucial for the detectability of its reﬂection in the brain response. It should be kept in mind that the present approach provides a differential view on brain responses to aspects of a natural music stimulus, not a comprehensive decomposition of the brain signal. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: http://www.frontiersin.org/journal/10.3389/fnhum. 2014.00798/abstract Frontiers in Human Neuroscience 4.2. COMPLEMENTARY ANALYSIS OF SPEECH STIMULI (2007) the alpha band activity has been associated with relay mechanisms that govern the transfer of auditory information from the thala- mus to core auditory areas. The high gamma activity has been October 2014 \| Volume 8 \| Article 798 \| 11 www.frontiersin.org www.frontiersin.org www.frontiersin.org Cortical representations of a rock song in electrocorticography Sturm et al. 4.3. METHODOLOGY Here, operating on the original features, we demonstrate that a partial correlation approach as an extension of multiple linear regression analysis (Schaefer et al., 2009) can be used to differentiate between the processing of aspects of natural music. Naturally, when comparing two different stimuli, such as in our case in the speech and music condition, the individual inter- dependence of stimulus features is not the same, nor can the stimulus features themselves be balanced between both stim- uli. Our results, therefore, have to be regarded as highly speciﬁc cortical imprints of two different, naturally unbalanced exam- ples of natural auditory stimulation from two sound categories, not as general ﬁndings on the processing of music or speech. Nonetheless, the present differentiated picture of brain responses at the level of single subjects and a single presentation is a valu- able complement of the recent series of investigations in natural music processing research. p g p Typically, in a naturalistic setting, multi-channel measure- ments are related (e.g., by correlation measures) to a multi- dimensional description of music, a situation which is prone to produce false positive effects such as spurious correlations. One way of constraining the solution is to assume inter-individual consistent spatial distribution of neural activity, e.g., by averag- ing the EEG time course across subjects (Schaefer et al., 2009) or by selecting components that are common to the majority of sub- jects (Alluri et al., 2012; Cong et al., 2012). The present results are an important complement to previous studies, as they were obtained at the single-subject level and for one single stimulus presentation owing to the ECoG’s characteristics of offering both high temporal and spatial resolution. This sensitivity helped to reveal a considerable variability between subjects with respect to reﬂected features of music, an insight that suggests that assuming within-group consistency might neglect some of these individual effects. Cortical representations of a rock song in electrocorticography Sturm et al. Goydke, K., Altenmüller, E., Möller, J., and Münte, T. (2004). Changes in emotional tone and instrumental timbre are reﬂected by the mismatch negativity. Cogn. Brain Res. 21, 351–359. doi: 10.1016/j.cogbrainres.2004.06.009 Alluri, V., Toiviainen, P., Lund, T. E., Wallentin, M., Vuust, P., Nandi, A. K., et al. (2013). From vivaldi to beatles and back: predicting lateralized brain responses to music. Neuroimage 83, 627–636. doi: 10.1016/j.neuroimage.2013.06.064 Grahn, J. A., and Rowe, J. B. (2009). Feeling the beat: premotor and striatal inter- actions in musicians and nonmusicians during beat perception. J. Neurosci. 29, 7540–7548. doi: 10.1523/JNEUROSCI.2018-08.2009 Bießmann, F., Meinecke, F. C., Gretton, A., Rauch, A., Rainer, G., Logothetis, N. K., et al. (2010). Temporal kernel cca and its application in multimodal neuronal data analysis. Mach. Learn. 79, 5–27. doi: 10.1007/s10994-009-5153-3 Halpern, A., Martin, J., and Reed, T. (2008). An ERP study of major-minor classiﬁ- cation in melodies. Music Percept. 25, 181–191. doi: 10.1525/mp.2008.25.3.181 Brattico, E., Alluri, V., Bogert, B., Jacobsen, T., Vartiainen, N., Nieminen, S., et al. (2011). A functional MRI study of happy and sad emotions in music with and without lyrics. Front. Psychol. 2:308. doi: 10.3389/fpsyg.2011.00308 Harte, C., Gasser, M., and Mark, S. (2006). “Detecting harmonic change in musical audio,” in AMCMM ’06 Proceedings of The 1st ACM Workshop on Audio and Music Computing Multimedia (Santa Barbara, CA), 21–26. Brattico, E., Tervaniemi, M., Naatanen, R., and Peretz, I. (2006). Musical scale properties are automatically processed in the human auditory cortex. Brain Res. 1117, 162–174. doi: 10.1016/j.brainres.2006.08.023 Hasson, U., Malach, R., and Heeger, D. J. (2010). Reliability of corti- cal activity during natural stimulation. Trends Cogn. Sci. 14, 40–48. doi: 10.1016/j.tics.2009.10.011 Brown, E. C., Muzik, O., Rothermel, R., Juhász, C., Shah, A. K., Fuerst, D., et al. (2014). Evaluating signal-correlated noise as a control task with language-related gamma activity on electrocorticography. Clin. Neurophysiol. 125, 1312–1323. doi: 10.1016/j.clinph.2013.11.026 Higuchi, M., Fornari, J., Del Ben, C., Graeff, F., and Leite, J. P. (2011). Reciprocal modulation of cognitive and emotional aspects in pianistic performances. PLoS ONE 6:e24437. doi: 10.1371/journal.pone.0024437 Burger, B., Thompson, M. R., Luck, G., Saarikallio, S., and Toiviainen, P. (2013). Inﬂuences of rhythm-and timbre-related musical features on characteristics of music-induced movement. Front. psychol. 4:183. doi: 10.3389/fpsyg.2013.00183 Hyde, K. L., Peretz, I., and Zatorre, R. J. (2008). Evidence for the role of the right auditory cortex in ﬁne pitch resolution. Neuropsychologia 46, 632–639. Cortical representations of a rock song in electrocorticography doi: 10.1037/a0024700 Kubanek, J., Brunner, P., Gunduz, A., Poeppel, D., Schalk, G., and Rodriguez- Fornells, A. (2013). The tracking of speech envelope in the human cortex. PLoS ONE 8:e53398. doi: 10.1371/journal.pone.0053398 Kumar, S., Sedley, W., Nourski, K. V., Kawasaki, H., Oya, H., Patterson, R. D., et al. (2011). Predictive coding and pitch processing in the auditory cortex. J. Cogn. Neurosci. 23, 3084–3094. doi: 10.1162/jocn_a_00021 Crone, N. E., Sinai, A., and Korzeniewska, A. (2006). High-frequency gamma oscil- lations and human brain mapping with electrocorticography. Prog. Brain Res. 159, 275–295. doi: 10.1016/S0079-6123(06)59019-3 Daikoku, T., Ogura, H., and Watanabe, M. (2012). The variation of hemodynam- ics relative to listening to consonance or dissonance during chord progression. Neurol. Res. 34, 557–563. doi: 10.1179/1743132812Y.0000000047 Kumar, S., Stephan, K. E., Warren, J. D., Friston, K. J., and Grifﬁths, T. D. (2007). Hierarchical processing of auditory objects in humans. PLoS Comput. Biol. 3:e100. doi: 10.1371/journal.pcbi.0030100 Deike, S., Gaschler-Markefski, B., Brechmann, A., and Scheich, H. (2004). Auditory stream segregation relying on timbre involves left auditory cortex. Neuroreport 15, 1511–1514. doi: 10.1097/01.wnr.0000132919.12990.34 Lartillot, O., Eerola, T., Toiviainen, P., and Fornari, J. (2008a). “Multi-feature mod- eling of pulse clarity: design, validation and optimization,” in Proceedings of ISMIR (Philadelphia, PA: Citeseer), 521–526. Lartillot, O., Toiviainen, P., and Eerola, T. (2008b). “A matlab toolbox for music information retrieval,” in Studies in Classiﬁcation, Data Analysis, and Knowledge Organization, eds C. Preisach, H. Burkhardt, L. Schmidt-Thieme, and R. Decker (Berlin; Heidelberg: Springer Berlin Heidelberg), 261–268. del Bimbo, A., Chang, S.-F., Smeulders, A., Cannam, C., Landone, C., and Sandler, M. (2010). “Sonic visualiser,” in Proceedings of The International Conference on Multimedia - MM ’10 (Firenze: ACM Press), 1467. Ding, N., and Simon, J. Z. (2012). Neural coding of continuous speech in auditory cortex during monaural and dichotic listening. J. Neurophysiol. 107, 78–89. doi: 10.1152/jn.00297.2011 Lehne, M., Rohrmeier, M., and Koelsch, S. (2014). Tension-related activity in the orbitofrontal cortex and amygdala: an fmri study with music. Soc. Cogn. Affect. Neurosci. 9, 1515–1523. doi: 10.1093/scan/nst141ion-related Drullman, R., Festen, J. M., and Plomp, R. (1994). Effect of reducing slow tem- poral modulations on speech reception. J. Acoust. Soc. Am. 95, 2670. doi: 10.1121/1.409836 Leonard, M. K., and Chang, E. F. (2014). Dynamic speech representations in the human temporal lobe. Trends Cogn. Sci. 18, 472–479. doi: 10.1016/j.tics.2014. 05.001 Edwards, E., Soltani, M., Deouell, L. Y., Berger, M. S., and Knight, R. T. (2005). Cortical representations of a rock song in electrocorticography doi: 10.1016/j.neuropsychologia.2007.09.004 Caclin, A., Brattico, E., Tervaniemi, M., Näätänen, R., Morlet, D., Giard, M.-H., et al. (2006). Separate neural processing of timbre dimensions in auditory sensory memory. J. Cogn. Neurosci. 18, 1959–1972. doi: 10.1162/jocn.2006.18.12.1959 Janata, P. (2009). The neural architecture of music-evoked autobiographical mem- ories. Cereb. Cortex 19, 2579. doi: 10.1093/cercor/bhp008 Janata, P., Birk, J., van Horn, J., Leman, M., Tillmann, B., and Bharucha, J. (2002). The cortical topography of tonal structures underlying Western music. Science 298, 2167. doi: 10.1126/science.1076262 Caclin, A., Giard, M.-H., Smith, B. K., and McAdams, S. (2007). Interactive process- ing of timbre dimensions: a garner interference study. Brain Res. 1138, 159–170. doi: 10.1016/j.brainres.2006.12.065 Jentschke, S., Friederici, A. D., and Koelsch, S. (2014). Neural correlates of music- syntactic processing in two-year old children. Dev. Cogn. Neurosci. 9, 200–208. doi: 10.1016/j.dcn.2014.04.005 Callan, D. E., Tsytsarev, V., Hanakawa, T., Callan, A. M., Katsuhara, M., Fukuyama, H., et al. (2006). Song and speech: brain regions involved with perception and covert production. Neuroimage 31, 1327–1342. doi: 10.1016/j.neuroimage.2006.01.036 Jongsma, M. L., Desain, P., and Honing, H. (2004). Rhythmic context inﬂuences the auditory evoked potentials of musicians and nonmusicians. Biol. Psychol. 66, 129–152. doi: 10.1016/j.biopsycho.2003.10.002 Chapin, H., Jantzen, K., Kelso, J., Steinberg, F., and Large, E. (2010). Dynamic emo- tional and neural responses to music depend on performance expression and listener experience. PLoS ONE 5:e13812. doi: 10.1371/journal.pone.0013812 Kendall, M. G., Stuart, A., and Ord, J. K. (1973). Inference and Relationship, Volume 2 of The Advanced Theory of Statistics, 3rd Edn. London: Grifﬁn. Chew, E. (2000). Towards a Mathematical Model of Tonality. Ph.D. thesis, Massachusetts Institute of Technology. Kim, C. H., Lee, S., Kim, J. S., Seol, J., Yi, S. W., and Chung, C. K. (2014). Melody effects on eranm elicited by harmonic irregularity in musical syntax. Brain Res. 1560, 36–45. doi: 10.1016/j.brainres.2014.02.045 Cong, F., Phan, A. H., Zhao, Q., Nandi, A. K., Alluri, V., Toiviainen, P., et al. (2012). “Analysis of ongoing EEG elicited by natural music stimuli using non- negative tensor factorization,” in Signal Processing Conference (EUSIPCO), 2012 (Bucharest), 494–498. Koelsch, S., Gunter, T. C., von Cramon, D. Y., Zysset, S., Lohmann, G., and Friederici, A. D. (2002). Bach speaks: a cortical “language-network” serves the processing of music. Neuroimage 17, 956–966. doi: 10.1006/nimg.2002.1154 Coutinho, E., and Cangelosi, A. (2011). Musical emotions: predicting second-by- second subjective feelings of emotion from low-level psychoacoustic features and physiological measurements. Emotion 11, 921–937. REFERENCES Abdi, H. (2007). “Part (semi partial) and partial regression coefﬁcients,” in Encyclopedia of Measurement and Statistics, ed N. J. Salkind (SAGE Publications), 736–740. Abrams, D. A., Bhatara, A., Ryali, S., Balaban, E., Levitin, D. J., and Menon, V. (2011). Decoding temporal structure in music and speech relies on shared brain resources but elicits different fine-scale spatial patterns. Cereb. Cortex 21, 1507– 1518. doi: 10.1093/cercor/bhq198 Abrams, D. A., Ryali, S., Chen, T., Chordia, P., Khouzam, A., Levitin, D. J., et al. (2013). Inter-subject synchronization of brain responses during natural music listening. Eur. J. Neurosci. 37, 1458–1469. doi: 10.1111/ejn.12173 Alluri, V., Toiviainen, P., Jääskeläinen, I. P., Glerean, E., Sams, M., and Brattico, E. (2012). Large-scale brain networks emerge from dynamic pro- cessing of musical timbre, key and rhythm. Neuroimage 59, 3677–3689. doi: 10.1016/j.neuroimage.2011.11.019 October 2014 \| Volume 8 \| Article 798 \| 12 www.frontiersin.org 12 Cortical representations of a rock song in electrocorticography Cortical representations of a rock song in electrocorticography S., and Large, E. W. (2005). Gamma-band activity reﬂects the met- ric structure of rhythmic tone sequences. Cogn. Brain Res. 24, 117–126. doi: 10.1016/j.cogbrainres.2004.12.014 Pampalk, E., Rauber, A., and Merkl, D. (2002). “Content-based organization and visualization of music archives,” in Proceedings of The Tenth ACM International Conference on Multimedia (Juan les Pins), 570–579. doi: 10.1145/641007.641121 Theiler, J., Eubank, S., Longtin, A., Galdrikian, B., and Doyne Farmer, J. (1992). Testing for nonlinearity in time series: the method of surrogate data. Physica D 58, 77–94. doi: 10.1016/0167-2789(92)90102-S Pasley, B. N., David, S. V., Mesgarani, N., Flinker, A., Shamma, S. A., Crone, N. E., et al. (2012). Reconstructing speech from human auditory cortex. PLoS Biol. 10:e1001251. doi: 10.1371/journal.pbio.1001251 Toiviainen, P., Alluri, V., Brattico, E., Wallentin, M., and Vuust, P. (2014). Capturing the musical brain with lasso: dynamic decoding of musical features from fmri data. Neuroimage 88, 170–180. doi: 10.1016/j.neuroimage.2013.11.017 Patterson, R. D., Uppenkamp, S., Johnsrude, I. S., and Grifﬁths, T. D. (2002). The processing of temporal pitch and melody information in auditory cortex. Neuron 36, 767–776. doi: 10.1016/S0896-6273(02)01060-7 Towle, V. L., Yoon, H.-A., Castelle, M., Edgar, J. C., Biassou, N. M., Frim, D. M., et al. (2008). Ecog gamma activity during a language task: differentiating expres- sive and receptive speech areas. Brain 131, 2013–2027. doi: 10.1093/brain/ awn147 Pei, X., Leuthardt, E. C., Gaona, C. M., Brunner, P., Wolpaw, J. R., and Schalk, G. (2011). Spatiotemporal dynamics of electrocorticographic high gamma activ- ity during overt and covert word repetition. Neuroimage 54, 2960–2972. doi: 10.1016/j.neuroimage.2010.10.029 Trainor, L., McDonald, K., and Alain, C. (2002). Automatic and controlled process- ing of melodic contour and interval information measured by electrical brain activity. J. Cogn. Neurosci. 14, 430–442. doi: 10.1162/089892902317361949 Perani, D., Saccuman, M., Scifo, P., Spada, D., Andreolli, G., Rovelli, R., et al. (2010). Functional specializations for music processing in the human newborn brain. Proc. Natl. Acad. Sci. U.S.A. 107, 4758. doi: 10.1073/pnas.0909074107 Zatorre, R., Chen, J., and Penhune, V. (2007). When the brain plays music: auditory–motor interactions in music perception and production. Nat. Rev. Neurosci. 8, 547–558. doi: 10.1038/nrn2152 Plack, C. J., Barker, D., and Hall, D. A. (2014). Pitch coding and pitch processing in the human brain. Hear. Res. 307, 53–64. doi: 10.1016/j.heares.2013.07.020 Zatorre, R., and Krumhansl, C. (2002). Mental models and musical minds. Science 298, 2138. doi: 10.1126/science.1080006 Potes, C., Brunner, P., Gunduz, A., Knight, R. T., and Schalk, G. (2014). Cortical representations of a rock song in electrocorticography Sturm et al. Madison, G., Gouyon, F., Ullén, F., and Hörnström, K. (2011). Modeling the tendency for music to induce movement in humans: ﬁrst correlations with low-level audio descriptors across music genres. J. Exp. Psychol. Hum. Percept. Perform. 37, 1578. doi: 10.1037/a0024323 Schaefer, R., Vlek, R., and Desain, P. (2011a). Decomposing rhythm process- ing: electroencephalography of perceived and self-imposed rhythmic patterns. Psychol. Res. 75, 95–106. doi: 10.1007/s00426-010-0293-4 y Schaefer, R. S., Desain, P., and Suppes, P. (2009). Structural decomposition of EEG signatures of melodic processing. Biol. Psychol. 82, 253–259. doi: 10.1016/j.biopsycho.2009.08.004 Marrelec, G., Krainik, A., Duffau, H., Pélégrini-Issac, M., Lehéricy, S., Doyon, J., et al. (2006). Partial correlation for functional brain inter- activity investigation in functional MRI. Neuroimage 32, 228–237. doi: 10.1016/j.neuroimage.2005.12.057 Schaefer, R. S., Farquhar, J., Blokland, Y., Sadakata, M., and Desain, P. (2011b). Name that tune: decoding music from the listening brain. Neuroimage 56, 843–849. doi: 10.1016/j.neuroimage.2010.05.084 Martin, S., Brunner, P., Holdgraf, C., Heinze, H.-J., Crone, N. E., Rieger, J., et al. (2014). Decoding spectrotemporal features of overt and covert speech from the human cortex. Front. Neuroeng. 7:14. doi: 10.3389/fneng.2014.00014 Schalk, G., McFarland, D. J., Hinterberger, T., Birbaumer, N., and Wolpaw, J. R. (2004). BCI2000: a general-purpose brain-computer interface (bci) sys- tem. IEEE Trans. Biomed. Eng. 51, 1034–1043. doi: 10.1109/TBME.2004. 827072 Merrill, J., Sammler, D., Bangert, M., Goldhahn, D., Lohmann, G., Turner, R., et al. (2012). Perception of words and pitch patterns in song and speech. Front. Psychol. 3:76. doi: 10.3389/fpsyg.2012.00076 Schalk, G., and Mellinger, J. (2010). Human-Computer Interaction: Practical Guide to Brain-Computer Interfacing with BCI2000: General-Purpose Software for Brain-Computer Interface Research, Data Acquisition, Stimulus Presentation, and Brain Monitoring. London, UK: Springer. Mikutta, C., Altorfer, A., Strik, W., and Koenig, T. (2012). Emotions, arousal, and frontal alpha rhythm asymmetry during beethoven’s 5th symphony. Brain Topogr. 25, 423–430. doi: 10.1007/s10548-012-0227-0 Mikutta, C. A., Schwab, S., Niederhauser, S., Wuermle, O., Strik, W., and Altorfer, A. (2013). Music, perceived arousal, and intensity: psychophysiological reactions to chopin’s tristesse. Psychophysiology 50, 909–919. doi: 10.1111/psyp.12071 Schubert, E., Wolfe, J., and Tarnopolsky, A. (2004). “Spectral centroid and timbre in complex, multiple instrumental textures,” in Proceedings of The International Conference on Music Perception and Cognition, North Western University, Illinois 2004 (Chicago, IL), 112–116. Nan, Y., and Friederici, A. D. (2013). Differential roles of right temporal cortex and broca’s area in pitch processing: evidence from music and mandarin. Hum. Brain Mapp. 34, 2045–2054. doi: 10.1002/hbm.22046 Snyder, J. Cortical representations of a rock song in electrocorticography Spatial and temporal relationships of electrocorticographic alpha and gamma activity during auditory processing. Neuroimage 97, 188–195. doi: 10.1016/j.neuroimage.2014.04.045 Zeng, F.-G., Oba, S., Garde, S., Sininger, Y., and Starr, A. (1999). Temporal and speech processing deﬁcits in auditory neuropathy. Neuroreport 10, 3429–3435. doi: 10.1097/00001756-199911080-00031 Potes, C., Gunduz, A., Brunner, P., and Schalk, G. (2012). Dynamics of electro- corticographic (ecog) activity in human temporal and frontal cortical areas during music listening. Neuroimage 61, 841–848. doi: 10.1016/j.neuroimage. 2012.04.022 Zentner, M. (2010). Homer’s prophecy: an essay on music’s primary emotions. Music Anal. 29, 102–125. doi: 10.1111/j.1468-2249.2011.00322.x Zion Golumbic, E. M., Ding, N., Bickel, S., Lakatos, P., Schevon, C. A., McKhann, G. M., et al. (2013). Mechanisms underlying selective neuronal track- ing of attended speech at a ¨cocktail party¨. Neuron 77, 980–991. doi: 10.1016/j.neuron.2012.12.037 Power, A. J., Foxe, J. J., Forde, E.-J., Reilly, R. B., and Lalor, E. C. (2012). At what time is the cocktail party? A late locus of selective attention to natural speech. Eur. J. Neurosci. 35, 1497–1503. doi: 10.1111/j.1460-9568.2012.08060.x Conﬂict of Interest Statement: The authors declare that the research was con- ducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Regnault, P., Bigand, E., and Besson, M. (2001). Different brain mechanisms medi- ate sensitivity to sensory consonance and harmonic context: evidence from auditory event-related brain potentials. J. Cogn. Neurosci. 13, 241–255. doi: 10.1162/089892901564298 Received: 22 July 2014; accepted: 19 September 2014; published online: 13 October 2014. Rosen, S. (1992). Temporal information in speech: acoustic, auditory and lin- guistic aspects. Philos. Trans. R. Soc. Lond. B Biol. Sci. 336, 367–373. doi: 10.1098/rstb.1992.0070 Citation: Sturm I, Blankertz B, Potes C, Schalk G and Curio G (2014) ECoG high gamma activity reveals distinct cortical representations of lyrics passages, harmonic and timbre-related changes in a rock song. Front. Hum. Neurosci. 8:798. doi: 10.3389/ fnhum.2014.00798 Sammler, D., Baird, A., Valabrègue, R., Clément, S., Dupont, S., Belin, P., et al. (2010). The relationship of lyrics and tunes in the processing of unfamil- iar songs: a functional magnetic resonance adaptation study. J. Neurosci. 30, 3572–3578. doi: 10.1523/JNEUROSCI.2751-09.2010 This article was submitted to the journal Frontiers in Human Neuroscience. This article was submitted to the journal Frontiers in Human Neuroscience. Copyright © 2014 Sturm, Blankertz, Potes, Schalk and Curio. This is an open- access article distributed under the terms of the Creative Commons Attribution License (CC BY). Cortical representations of a rock song in electrocorticography High gamma activity in response to deviant auditory stimuli recorded directly from human cortex. J. Neurophysiol. 94, 4269–4280. doi: 10.1152/jn. 00324.2005 Liebenthal, E., Ellingson, M. L., Spanaki, M. V., Prieto, T. E., Ropella, K. M., and Binder, J. R. (2003). Simultaneous ERP and fMRI of the auditory cortex in a passive oddball paradigm. Neuroimage 19, 1395–1404. doi: 10.1016/S1053- 8119(03)00228-3 Eerola, T., Lartillot, O., and Toiviainen, P. (2009). “Prediction of multidimensional emotional ratings in music from audio using multivariate regression models,” in Proceedings of ISMIR (Kobe), 621–626. Lorenzi, C., Gilbert, G., Carn, H., Garnier, S., and Moore, B. C. (2006). Speech perception problems of the hearing impaired reﬂect inability to use tem- poral ﬁne structure. Proc. Natl. Acad. Sci. U.S.A. 103, 18866–18869. doi: 10.1073/pnas.0607364103 Goodglass, H., Edith, K., and Barbara, B. (1983). BDAE: The Boston Diagnostic Aphasia Examination. Philadelphia, PA: Lea and Febiger. October 2014 \| Volume 8 \| Article 798 \| 13 Frontiers in Human Neuroscience www.frontiersin.org Cortical representations of a rock song in electrocorticography Cortical representations of a rock song in electrocorticography The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor 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. Sammler, D., Koelsch, S., Ball, T., Brandt, A., Grigutsch, M., Huppertz, H.-J., et al. (2013). Co-localizing linguistic and musical syntax with intracranial EEG. Neuroimage 64, 134–146. doi: 10.1016/j.neuroimage.2012.09.035 Sammler, D., Koelsch, S., and Friederici, A. D. (2011). Are left fronto-temporal brain areas a prerequisite for normal music-syntactic processing? Cortex 47, 659–673. doi: 10.1016/j.cortex.2010.04.007 October 2014 \| Volume 8 \| Article 798 \| 14 Frontiers in Human Neuroscience www.frontiersin.org 14
https://openalex.org/W2407194178	https://www.scientiaplena.org.br/sp/article/download/3096/1502	Portuguese	null	Concepções sobre serpentes entre jovens estudantes do ensino médio: um diálogo entre ciência e cultura	Scientia plena	2,016	cc-by	7,151	A. W. P. da Silva1; S. M. V. de Castro2; M. D. B. Silva3; P. H. G. de Castro4; J. B. Costa5 4 Médico Veterinário do Serviço de Ecologia e Manejo de Primatas do Centro Nacional de Primatas,67.020- 380, Ananindeua, Brasil 5Membro do grupo de Pesquisa Ciência, Tecnologia, Meio Ambiente e Educação não- formal, Centro de Ciências Planetário do Pará,66.623- 590, Belém do Pará, Brasil O presente estudo desenvolveu-se com o objetivo de identificar as concepções que permeiam a visão etnoherpetológica, em relação às serpentes, entre alunos do ensino médio do município de Santa Izabel do Pará/ PA. Por serem animais considerados perigosos o desconhecimento de sua biologia provoca matança indiscriminada, o que futuramente poderá acarretar em um desequilíbrio ambiental. A partir de tal problemática o trabalho realizou o levantamento das concepções dos alunos em relação às serpentes através, de questionário prévio aplicado na escola e, posteriormente como ação educativo- ambiental de intervenção, uma palestra expositiva e dialogada, a fim de contribuir para a formação de consciência na conservação da fauna ofídica. A visita foi realizada no “Criadouro Comercial de Fauna Silvestre – Sítio Xerimbabo”. Por fim para análise da reflexão, referentes aos assuntos discutidos na palestra, onde a avaliação foi realizada através da observação dos indivíduos, registros em caderno de campo e gravações de áudio e imagem. Assim os resultados alcançados foram dispostos em categorias, onde a primeira parte da pesquisa apresentou-se com predomínio a categoria medo em relação a esses animais pela generalização de que todos são perigosos por serem peçonhentos, outra categoria freqüente mencionada pelos alunos foi a curiosidade despertada pelas peculiaridades apresentada por esse grupo. Palavras-chave: Etnoherpetologia, Serpentes, Educação ambiental. The study was developed in order to identify the concepts that permeate etnoherpetológica vision in relation to snakes, among high school students in the city of Santa Izabel do Para / PA. Because they are animals considered dangerous ignorance of their biology causes indiscriminate killing, which ultimately could result in an environmental imbalance. From this problematic work conducted the survey of students' conceptions regarding snakes through prior questionnaire in school and later as environmental educative action intervention, expository and dialogued lecture in order to contribute to the formation awareness on conservation of ophidian fauna. The visit was made in the "Breeding Fauna Silvestre Commercial - Site Xerimbabo". Finally for the analysis of reflection, regarding the issues discussed in the lecture, where the assessment was carried out through observation of individuals, field book records and audio and image recordings. VOL. 12, NUM. 06 2016 www.scientiaplena.org.br doi: 10.14808/sci.plena.2016.069925 Concepções sobre serpentes entre jovens estudantes do ensino médio: um diálogo entre ciência e cultura Concepts about snakes among young students of high school: a dialogue between science and culture A. W. P. da Silva1; S. M. V. de Castro2; M. D. B. Silva3; P. H. G. de Castro4; J. B. Costa5 1Membro do grupo de Pesquisa Ciência, Tecnologia, Meio Ambiente e Educação não- formal, Centro de Ciências e Planetário do Pará, 66.623- 590, Belém do Pará, Brasil 2Professor Adjunto I, Universidade do Estado do Pará,66.050- 540,Belém do Pará, Brasil 3Professor Assistente IV, Universidade do Estado do Pará,66.050- 540,Belém do Pará, Brasil 4 Médico Veterinário do Serviço de Ecologia e Manejo de Primatas do Centro Nacional de Primatas,67.020- 380, Ananindeua, Brasil 5Membro do grupo de Pesquisa Ciência, Tecnologia, Meio Ambiente e Educação não- formal, Centro de Ciências e Planetário do Pará,66.623- 590, Belém do Pará, Brasil awpsbio@gmail.com (Recebido em 08 de abril de 2016; aceito em 10 de maio de 2016) VOL. 12, NUM. 06 2016 doi: 10.14808/sci.plena.2016.069925 1. INTRODUÇÃO Estima-se que no Brasil há 371 espécies de serpentes catalogadas, o que corresponde ao percentual de 10% da fauna mundial [1]. Deste total a Amazônia abriga cerca de 150 espécies [2]. Os consideráveis valores mencionados revelam a riqueza da biodiversidade existente. No entanto o estereótipo negativo agregado a esses animais resultantes de um conjunto de fatores que contribuem para o potencial conflito entre humanos e serpentes. Assim Souza e Souza [3] destacam que se um animal é culturalmente percebido como feio e/ou nojento, provavelmente pouco saberá a seu respeito, isso porque haverá sempre um bloqueio sobre o conhecer sua biologia, devido tais percepções estarem negativamente construídas, interferindo no processo de convivência harmônica e sustentável. Nessa perspectiva a pesquisa realizada direcionou o estudo das concepções sobre as serpentes, a partir da ótica da etnoherpetologia, na busca pela compreensão da representação construída pelos indivíduos a partir do ambiente em que estão inseridos e relações estabelecidas no que “diz respeito ao seu conhecimento, utilização, classificação e convivência com os répteis [4]”. Partindo do pressuposto que, através de suas concepções antropológicas o indivíduo interage ou não com a fauna que o rodeia, implicando diretamente na conservação de sua biodiversidade local, sua integração estará intrinsecamente relacionada na forma de como se estabelece as relações seguindo suas concepções culturais, sociais, religiosas que constroem seu imaginário. Assim, dispondo de uma fundamentação teórica a pesquisa aborda os aspectos etnoherpetológicos que permeiam a visão dos indivíduos e refletem em suas interações com a fauna local, tais como cultura, religião e categorizações antropocêntricas. Em busca da valorização e respeito às serpentes, devem ser tratados conceitos sobre educação e conservação ambiental, exprimindo sua importância para o equilíbrio ecológico e pesquisas referentes à área da saúde. Para isso a educação ambiental se torna uma ferramenta importante no que versa sobre a conscientização da função ecológica das serpentes. Moura [5] destaca que indivíduos com menos informação tendem considerar todos ofídios nocivos o que causa mais intolerância com as espécies. Tendo em vista contribuir com a conservação da fauna ofídica, o desenvolvimento de ações educativas no intuito de difundir o conhecimento em relação às serpentes e suas interações com o ambiente configura uma eficiente estratégia conservacionista. Dodd [6] alerta que a falta de conhecimento e o preconceito em relação a essas espécies podem ocasionar uma extinção herpetológica local resultante da matança indiscriminada. A. W. P. da Silva1; S. M. V. de Castro2; M. D. B. Silva3; P. H. G. de Castro4; J. B. Costa5 Thus the results were arranged in categories, where the first part of the research presented predominantly with the fear category of such animals by the generalization that all are dangerous because they are venomous, another frequent category mentioned by the students was curiosity aroused by peculiarities presented by this group. K d Eth h t l S t E i t l d ti 069925 – 1 A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 2 2.1. Local da pesquisa O referido trabalho apresentou como lócus de execução o Sítio Xerimbabo, que se trata de um Criadouro Comercial de Fauna Silvestre, localizado no Km12, na Rua do Ficado, no município de Santo Antônio do Tauá/PA, acesso pela rodovia BR 316 até o município de Santa Izabel do Pará/PA, direcionando-se à PA 140. Nas imagens a seguir pode-se ver na figura 1A o mapa de localização do Sítio Xerimbabo disponível no site do Criadouro (www.sitioxerimbabo.com.br), a imagem 1B por sua vez, mostra a localização via satélite indicando as direções espaciais: latitude 01º 09’ 07” sul e longitude 48º 07’ 46” oeste. Figura 1A: Ilustração disponível no site do Sítio Xerimbabo na área de localização, a figura mostra o trajeto saindo de Belém a caminho ao Sítio passando pelos municípios descritos e as rodovias BR 316 em seguida a PA 140; Figura 1B: Imagem de satélite indicando o local do criadouro, após deixar a PA 140 indicada pela seta, entra-se na Rua do Ficado, onde o Sítio está localizado no ponto vermelho. 1A 1B 1B 1A 1A 1B Figura 1A: Ilustração disponível no site do Sítio Xerimbabo na área de localização, a figura mostra o trajeto saindo de Belém a caminho ao Sítio passando pelos municípios descritos e as rodovias BR 316 em seguida a PA 140; Figura 1B: Imagem de satélite indicando o local do criadouro, após deixar a PA 140 indicada pela seta, entra-se na Rua do Ficado, onde o Sítio está localizado no ponto vermelho. Especializado na criação, reprodução e comercialização de serpentes da família Boidae, dentre as quais se criam as seguintes espécies: jiboias amazônicas (Boa constrictor constrictor), periquitamboias (Corallus caninus), suaçubóias (Corallus hortulanus) e salamantas (Epicrates cenchria cenchria) descritas por Gray em 1825. Especializado na criação, reprodução e comercialização de serpentes da família Boidae, dentre as quais se criam as seguintes espécies: jiboias amazônicas (Boa constrictor constrictor), periquitamboias (Corallus caninus), suaçubóias (Corallus hortulanus) e salamantas (Epicrates cenchria cenchria) descritas por Gray em 1825. Em abril de 2002, o sítio obteve sua homologação pelo Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA) [7], com o seguinte número de registro: 434325, amparado pela Portaria Normativa nº 117 de 1997, que permite a comercialização legal de animais da fauna silvestre brasileira. 1. INTRODUÇÃO Visando atingir as metas propostas, o trabalho além de levantar a concepção de serpentes entre os estudantes de ensino médio, apresentou como intervenção uma ação educativa em um ambiente caracterizado como espaço de ensino informal, o Sítio Xerimbabo, localizado no município de Santo Antônio do Tauá/ PA, visto que, atividades de educação-ambiental podem fomentar a conscientização quanto à importância das serpentes, como reforça Moura [5]. Por fim, para analisar o nível de compreensão dos conceitos abordados e mudanças em sua postura diante as questões referente às serpentes aplicou-se um jogo, pelo fato deste constituir- se um método capaz de despertar uma práxis reflexiva perante as ações anteriormente desempenhadas. Desta forma, a transversalidade da Educação Ambiental contribui para o fomento de práticas, seja em espaços formais, não - formais e informais, diferenciando - se em suas estruturas físicas, organizacionais e de planejamento, os ambientes extra escolares proporcionam um espaço desenvolto para que os alunos participem de forma espontânea, visto que, não se configuram em uma exigência curricular. Como subsidio o estudo apresenta os elementos que contribuem para a construção da percepção dos indivíduos diante as serpentes, no que se refere a valores culturais, religiosos e antropocêntricos e a partir disso como se estabelecem as relações homem e fauna ofídica. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 3 2.4 Instrumentos de coleta de dados Para a coleta de dados optou-se pelo uso de questionários na fase inicial da pesquisa, esta técnica segundo Marconi e Lakatos [11] configuram uma “observação direta extensiva constituída por uma série de perguntas que devem ser respondidas por escrito e sem a presença do pesquisador”, composto por perguntas fechadas e abertas, dentre as quais seguiram em diferentes análises. As perguntas fechadas auxiliam a traçar um perfil estatístico do público e as perguntas abertas “concedem uma gama de significados o que através da análise de conteúdo proporcionaram mais qualidade das respostas [12]. Como método de intervenção desenvolveu-se um jogo que possibilitou a coleta de dados final, “através de observação assistemática, no qual todos os comportamentos e colocações dos indivíduos foram registrados [12]”. 2.2 Sujeitos da pesquisa Para a realização deste estudo optou-se pelo público alvo, alunos do 3º ano do Ensino Médiode uma escola da rede particular de ensino do município de Santa Izabel do Pará. A turma constituída por 60 alunos regularmente matriculados, com faixa etária entre 16 a 21 anos. A opção por alunos desse nível de ensino e faixa etária levou em consideração que a partir deste nível o educando é capaz de desenvolver “consciência plena de suas responsabilidades e conhecimentos mais amplos e abstratos, que correspondam a uma cultura geral e a uma visão de mundo [8]”. Outro ponto a se considerar é a localidade, pois o município atualmente se enquadra na área metropolitana de Belém, mas que, no entanto possui separação entre zona rural e urbana. Assim, pode-se considerar que para os indivíduos que residem na zona rural a “visita” de certas espécies animais se torna comum, pois neste “ambiente que vivem e realizam suas atividades diárias, a presença de répteis não é estranha [4]”. Contudo, o crescimento urbano tem ocorrido de forma acelerada no município, sendo outro fator que pode ser considerado, uma vez que este vem acarretando a destruição do ambiente natural propiciando maior proximidade com animais silvestres devido à disponibilidade de alimento e abrigo [9]. Dessa forma se torna de grande valia desenvolver um estudo de educação ambiental para a conservação da fauna, em especial serpentes, devido os sujeitos da pesquisa apresentarem proximidade com maior variedade de espécies. 2.1. Local da pesquisa Configura-se como um criadouro de pequeno porte, com produção pequena, implantado e mantido com recursos próprios, no qual o lucro das vendas dos animais é revestido em melhoria na área do sistema de criação. Dispõe de recintos como baías de reprodução, laboratório e alojamento, além de um biotério para produção própria de alimento para as serpentes através da criação de camundongos e preás. O Sítio Xerimbabo, apesar de ser um criadouro comercial, foi escolhido para o desenvolvimento da pesquisa por se tratar de um espaço diferenciado, onde seus objetivos perpassam pela busca por parcerias com órgãos de fiscalização, recebendo animais de apreensões e oferecendo tratamento veterinário a vítimas de maus tratos de traficantes ou criadores ilegais; contribuir para a diminuição do comércio ilegal; oferecer treinamento de contenção e manuseio e o mais importante desenvolver um programa de educação ambiental junto à comunidade com o acolhimento de escolas, promovendo um trabalho de conscientização focando na importância das serpentes no ambiente. p p Percebeu-se também o resgate do regionalismo amazônida no próprio nome do criadouro, onde Xerimbabo de origem tupi “xereimbawa” significa animal de estimação, assim trazendo a imagem da serpente para algo mais próximo e menos hostil. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 4 2.3 Tipo de pesquisa No que se refere aos objetivos à pesquisa é definida como descritiva, visto que se direciona em descrever as características de determinada população e de fenômenos, a fim de estudar opiniões, atitudes e crenças da população [10]. 2.5 Aplicação da pesquisa A pesquisa almejou como objetivo principal: identificar as concepções que permeiam a visão etnoherpetológica de alunos do Ensino Médio, sobre as serpentes. A pergunta norteadora do problema perpassa em: Como se constitui a concepção dos indivíduos frente às serpente?. Através da detecção dessa percepção, foi traçado o plano de trabalho afim de: realizar levantamento sobre os conhecimentos que os sujeitos possuem sobre as serpentes; promover ações educativo-ambiental, contribuindo para a conservação da fauna ofídica; analisar a possibilidade de desenvolvimento de práticas educativas ambientais a serem realizadas em A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 5 espaços, como criadouros, caracterizando-os como espaços informais de ensino, sendo os pontos elencados acima constituintes os objetivos específicos. p j p A aplicação realizou-se em três momentos distintos. No primeiro momento a proposta de trabalho foi apresentada, para que após devidamente esclarecidos os estudantes assinassem “Termo de Consentimento Livre e Esclarecido”. A etapa seguinte correspondeu à coleta de dados, representada pela aplicação do questionário inicial, composto por sete questões, onde duas primeiras tinham por objetivo caracterizar os alunos, coletando dados sobre idade e sexo, para a análise das questões abertas, nos casos em que se fez necessária identificação dos sujeitos, utilizou-se de códigos para que fosse garantido o anonimato dos estudantes. Os códigos foram definidos a partir de um número correspondente a ordem de entrega do questionário, associado a letra inicial do gênero correspondente, por exemplo, 4F, número 4, sexo feminino, exceto para o sujeito que não assinalou o sexo sendo a este atribuído a letra x. As demais questões dividiam- se em três objetivas (Você já viu uma serpente?; Se sua resposta foi sim, onde foi? ; Ao ver uma serpente qual foi, ou qual seria sua reação?) e duas subjetivas (Você já viu uma serpente? ; Se sua resposta foi sim, onde foi? ; Ao ver uma serpente qual foi, ou qual seria sua reação?). Nessa etapa 51 alunos contribuíram para a pesquisa com suas respostas prévias. Posteriormente realizou-se a visita ao “Sítio Xerimbabo”. Durante a visita ao Criadouro (Figura 2), da qual participaram 47 alunos, foi realizada uma roda de conversa mediada pelo médico veterinário e proprietário do Sítio Xerimbabo, abordando as temáticas referentes à: anatomia, fisiologia, papel ecológico, desmistificações empíricas, legalidade na aquisição de animais silvestres; abordagens estas que posteriormente, foram discutidas no jogo. 3.1. Concepções de estudantes de ensino médio sobre serpentes A primeira etapa da pesquisa realizada a partir da aplicação de um questionário possibilitou traçar um perfil dos sujeitos e categorizar suas representações acerca das serpentes. Ao se traçar um breve perfil dos 51 alunos que participaram da atividade, foi possível identificar que 51% eram do sexo feminino, 47% masculino e 2% indivíduos que optaram em não assinalar a opção. No que se refere à faixa etária observou-se uma distribuição entre 16 e 21 anos. Essa análise se faz importante, pois a faixa etária denota o seu nível de conhecimento [13] e suas experiências. O primeiro questionamento realizado destinou-se verificar se os alunos teriam tido algum tipo de contato com serpentes, nessa análise constatou-se que 16% dos participantes não apresentaram nenhuma aproximação com o referido animal e 84% responderam positivamente a questão, o que exprime a proximidade com o animal o fato de uma significativa parcela residir em zona rural. Aos participantes que indicaram já haver estabelecido algum tipo de contato com serpentes, foi indagado o tipo de local em que ocorreu o referido contato. Para tal indagação a opção, zoológico/ jardim botânico/ museu apresentou o maior percentual com 37%, seguida de casa com 21% e rua e mata/floresta, cada uma com 16 % (Figura 3). A referência ao contato com esse tipo de animal através de espaços de entretimento e lazer, como zoológicos, expressa uma concepção utilitarista de uso indireto, ou seja, que se utiliza de uma exposição do animal para fins recreacionais [14]. Figura 3: Em relação aos questionários a primeira pergunta das questões objetivas foi se os alunos teriam visto alguma serpente, em seguida a próxima questão pedia para que eles dissessem onde teria sido a imagem mostra a porcentagem dos locais onde mais se viu os animais em estudo. Figura 3: Em relação aos questionários a primeira pergunta das questões objetivas foi se os alunos teriam visto alguma serpente, em seguida a próxima questão pedia para que eles dissessem onde teria sido a imagem mostra a porcentagem dos locais onde mais se viu os animais em estudo. 2.5 Aplicação da pesquisa A dinâmica adotada permitiu que medida em que os temas eram abordados, os alunos partilhavam suas dúvidas, observando atentamente as estruturas da serpente que estava sendo manipulada. Figura 2: Segundo dia de aplicação, os alunos em visita ao Sítio para o início da atividade onde serão abordados pontos relativos a senso comum, fisiologia, comportamento em relação às serpentes, as faces estão desfocadas para preservar a identidade dos alunos. Figura 2: Segundo dia de aplicação, os alunos em visita ao Sítio para o início da atividade onde serão abordados pontos relativos a senso comum, fisiologia, comportamento em relação às serpentes, as faces estão desfocadas para preservar a identidade dos alunos. As etapas da pesquisa ocorreram em dias distintos e em uma fase no contra turno dos estudantes, o que justifica a oscilação do número de participantes. Para os registros foram utilizadas mídias: como fotos de todas as etapas da pesquisa e gravações em áudio onde os alunos relatavam o que a prática teria modificado sobre o conhecimento deles referente a biologia das serpentes. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 6 3.1. Concepções de estudantes de ensino médio sobre serpentes Seguindo a análise, os alunos que entraram em contato com o réptil em suas casas cabem à porcentagem de 21%, vale ressaltar que o município de Santa Izabel do Pará/PA atualmente passa por um intenso processo de crescimento urbano, fator que vem interferindo na constituição do habitat de várias espécies animais, o que podemos inferir que os16% referidos na opção rua, podem valer-se da mesma justificativa. Correspondem 16% a opção em seu habitat natural (mata), devido a atividades de lazer comumente praticadas no município e a alternativa em outros locais satisfaz o percentual de 10%, nessa categoria foram mencionadas áreas como: sítio, campos de futebol, fazenda e programas televisivos. Ao se analisar as respostas referentes à pergunta qual foi/seria a reação dos alunos ao ver uma serpente, a maioria das respostas se concentrou entre medo/susto e interesse/curiosidade. Anderson [15] pondera que o fator emocional é quem direciona a percepção e a quantidade de informações disponíveis sobre determinado objeto. 7 A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 7 A sensação de medo obteve o maior percentual com 41% (Figura 4), esse resultado reflete a forma de como as pessoas concebem o referido animal e o significado que o representa, pois permeia no imaginário a imagem da serpente como traiçoeira, relembrando os preceitos bíblicos, a visão de que todas são peçonhentas e nocivas ao homem, a literatura reforça a perpetuação da repulsa atribuindo ações não exercidas por serpentes, presentes na obra de Rosa [16], como por exemplo, cobra que chicoteia, voadeira e bicho imundo Os indivíduos que reagiram de forma a despertar sua curiosidade correspondem aos 35%, considerando o fato de a curiosidade ser aguçada, justamente por esses animais não serem comuns em nosso cotidiano e apresentarem características particulares exclusiva de sua classe, como por exemplo: rastejar-se devido à ausência de membros e língua bifurcada. Com 7% apresentaram-se simultaneamente as opções nojo e atacaria/mataria para se defender. O nojo é uma percepção construída a partir do aspecto visual desses animais, uma vez sua pele brilhosa passa a impressão errônea de um aspecto úmido. No que diz respeito à alternativa atacaria/mataria, adotamos a interpretação de Moura [5] ao considerar que o extermínio de espécies de seres vivos pode ocorrer de forma indiscriminada devido à falta de conhecimento que determinada comunidade apresenta. 3.1. Concepções de estudantes de ensino médio sobre serpentes Os indivíduos que se mostraram indiferentes diante do animale que não esboçaram nenhuma reação correspondem à porção de 5%, a parcela dos 5% restantes destacou que fugiriam/ permitiriam que o animal seguisse em seu caminho sem interferência. Figura 4: Uma questão importante para o estudo é realizar o levantamento das sensações que os alunos tiveram ao ver uma serpente, a figura apresenta a porcentagem das reações, a maioria apresentou medo/ susto o que corresponde a 41% justamente pelo não conhecimento e imaginário construído. Figura 4: Uma questão importante para o estudo é realizar o levantamento das sensações que os alunos tiveram ao ver uma serpente, a figura apresenta a porcentagem das reações, a maioria apresentou medo/ susto o que corresponde a 41% justamente pelo não conhecimento e imaginário construído. Figura 4: Uma questão importante para o estudo é realizar o levantamento das sensações que os alunos tiveram ao ver uma serpente, a figura apresenta a porcentagem das reações, a maioria apresentou medo/ susto o que corresponde a 41% justamente pelo não conhecimento e imaginário construído. O item 4 solicitava que os alunos descrevessem resumidamente suas opiniões sobre as serpentes, dentre as respostas foram elencadas algumas categorias, descritas a seguir. A categoria que agregou maior frequência entre os respondentes foi a que identificava as serpentes como animais perigosos. Nesta categoria estão presentes as respostas que enquadram as serpentes como animais perigosos e nocivos à saúde humana, como pode- se constatar nas falas dos indivíduos 12F, 16F e 35M, percebeu- se que nessas respostas há a expressão de que todas as espécies de serpentes são ‘venenosas’, considerando este o fator principal a atribuição ao seu alto nível de periculosidade. “as serpentes são animais que são consideradas altamente perigosas, pelo fato de possuírem um veneno mortal, capaz de matar um ser humano” (12F) “é um animal muito perigoso pois ao ser surpreendida ela ataca sua presa, libera um veneno que leva a morte”(16F) “é um animal muito perigoso pois ao ser surpreendida ela ataca sua presa, libera um veneno que leva a morte”(16F) “é um animal muito perigoso pois ao ser surpreendida ela ataca sua presa, libera um veneno que leva a morte”(16F) “são perigosas, odeio, tenho nojo, mataria todas”(35M) “são perigosas, odeio, tenho nojo, mataria todas”(35M) A.W.P. 3.1. Concepções de estudantes de ensino médio sobre serpentes A consciente percepção sobre o crescimento urbano como fator de interferência nos habitats animais apresentam as seguintes descrições: “no geral as cobras não estão invadindo nosso ambiente, mas sim nós que não respeitamos seu habitat”(11F) “no geral as cobras não estão invadindo nosso ambiente, mas sim nós que não respeitamos seu habitat”(11F) ” (17F) 3.1. Concepções de estudantes de ensino médio sobre serpentes da Silva et al.,Scientia Plena 12, 069925 (2016) 8 Assim dentro desta categoria pôde-se inferir que os sujeitos desconheçam a dicotomia entre um animal peçonhento e venenoso, para essa classificação o Instituto Butantan [17], ressalta que a peçonha é uma substância tóxica sintetizada por glândulas especializadas, introduzida em determinado tecido por um aparato específico, como as presas das serpentes, por exemplo. E veneno trata-se de toda e qualquer substância, seja de origem animal, vegetal ou mineral, capaz de provocar algum dano ao organismo, se inalado, absorvido ou ingerido, segundo o Instituto Butantan [17]. Desta forma, nenhuma serpente seria “venenosa”, mas algumas, e não todas seriam peçonhentas. A diferenciação de serpentes peçonhentas e não peçonhentas pode ser definida em sua grande maioria pela presença da fosseta loreal, estrutura presente somente nas espécies peçonhentas, com exceção da coral verdadeira. Sobre a importância do papel ecológico das serpentes no ambiente, a seguir estão representadas as descrições mais significativas referente a categoria ecologia, na fala dos alunos. “elas são importantes para o controle de sua população e de outras espécies no meio” (26M) “elas são importantes para o controle de sua população e de outras espécies no meio” (26M) “elas são importantes para o controle de sua população e de outras espécies no meio” (26M) “ajudam no equilíbrio ecológico ao participarem efetivamente da cadeia alimentar. O equilíbrio de sua espécie é de suma importância para a homeostase ambiental”(32F) cadeia alimentar. O equilíbrio de sua espécie é de suma importância para a homeostase ambiental”(32F) Um ponto frequentemente descrito nos questionários foi à importância das serpentes para o equilíbrio ambiental, contempladas nas falas dos sujeitos mencionados na tabela acima, ressaltando a fala da aluna 32F, utilizando o termo homeostase, comumente empregado às questões fisiológicas dos organismos, atribuindo-lhe uma aplicabilidade mais ampla no sentindo ambiental. Os ecossistemas são conexões que formam uma teia complexa de relações interdependentes, no qual qualquer perturbação pode afetar vários pontos dessa teia, causando desequilíbrios que podem ser irreversíveis. Assim Moura [5] destaca que as serpentes são de suma importância para a manutenção do equilíbrio ecológico, atuando no controle do número de animais vetores de doenças. Comportamento:“as serpentes sempre andam juntos, o macho e fêmea, para que um defenda o outro”(20F) Descrição sobre as lendas da cobra grande e do boitatá foram as narrativas predominantes nos questionários. A cultura na região é fortemente perceptível nas falas dos indivíduos 7M e 24M, transcritas no quadro acima. Estendendo-se para os mitos, histórias desenvolvidas a partir da observação de fatos desprovidos de rigor científico, consideramos o relato do individuo 20F, e pondera-se que as serpentes não possuem cuidado parental, tão pouco possui relações sociais, encontrando-se somente no período do acasalamento. “as serpentes estão cada vez mais entrando no espaço urbano” (17F) Um dos fatores a se considerar sobre o extermínio indiscriminado das serpentes, contribuindo com o aumento no número da população de vetores de doenças é o veemente crescimento urbano, em direção a áreas até então pouco habitadas, com significativa porcentagem nas opções rua e casa, na fala da aluna 11F, no entanto a aluna 17F em sua fala desconsidera a urbanização como fator de maior aproximação com as serpentes, como a degradação do seu habitat e maior disponibilidade de recursos nas áreas habitáveis esses animais tendem a melhor se adaptar. A categoria interessante aborda aspectos anteriormente discutidos, como a resposta do aluno 50M no que diz ser as serpentes animais diferente dos outros, no que alude sobre suas características exclusivas da espécie, sua de alimentação, onde a presa é deglutida inteira, seu formato cilíndrico sem membros, mas que possibilita deslocar-se. Distinto das respostas vistas A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 9 até o momento o aluno 19M discorreu sobre a beleza do animal referente à sua forma e cores específicas para determinadas espécies. “são animais curiosos e bonitos devido suas formas e cores”(19M) “é um animal interessante talvez por ser diferente dos outros”(50M) A seguir as percepções relatadas em menor frequência nos questionários, mas que configuram grande importância para a discussão da pesquisa, por contribuírem para a compreensão da imagem construída em relação às serpentes. Utilitarismo: “seu veneno pode ser usado em pesquisas para cura de doenças”(7M) Religioso: “é do capeta, são traiçoeiras e tem que morrer” (15M) Perigoso x Papel ecológico: “as serpentes são animais muito perigosos, porém faz um papel importante na natureza, principalmente na cadeia alimentar”(40F) O primeiro ponto aborda a categoria utilitarista, que se encaixa no uso direto [18], destina-se a utilização dos metabólitos animais para produção de fármacos com finalidades terapêuticas. A imagem da serpente no âmbito religioso possui uma carga de negatividade, traição, desde a gênese da criação, concepção presente ainda na descrição da fala de 15M, e como reforço a essa figuração, as iconografias cristãs retratam serpentes aos pés dos santos, simbolismo referente à vitória do bem sobre o mal [19]. A categoria que apresenta a contraposição entre perigoso e seu desempenho ecológico (40F), percebeu-se a generalização de que todas as serpentes são perigosas, assim muitas vezes desconsidera-se as interações tróficas que realizam com os demais animais [20]. A fim de identificar os valores culturais e experiências vivenciadas pelos indivíduos, a questão 5º do questionário solicitava a descrição de alguma situação presenciada ou relato de histórias através de lendas e mitos sobre serpentes que estejam presentes em seu grupo social. Lendas- Cobra-grande:“dorme embaixo de Belém, seu rabo esta localizada embaixo da basílica de Nazaré e seu corpo percorre o mesmo traslado do círio”(7M) Lendas- Cobra-grande:“dorme embaixo de Belém, seu rabo esta localizada embaixo da basílica de Nazaré e seu corpo percorre o mesmo traslado do círio”(7M) Boitatá:“é um animal folclórico da região amazônica que matava os caçadores na floresta e tinha olhos que pegavam fogo”(24M) Lendas- Cobra-grande:“dorme embaixo de Belém, seu rabo esta localizada embaixo da basílica de Nazaré e seu corpo percorre o mesmo traslado do círio”(7M) Boitatá:“é um animal folclórico da região amazônica que matava os caçadores na floresta e tinha olhos que pegavam fogo”(24M) Boitatá:“é um animal folclórico da região amazônica que matava os caçadores na floresta e tinha olhos que pegavam fogo”(24M) 3.2. Ação educativa: aproximações entre ciência e cultura A visita ao Criadouro Sítio Xerimbabo contou com o quantitativo de 47 alunos no qual participaram de uma palestra proferida pelo médico veterinário do sítio, tendo como objetivo A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 10 apresentar aos visitantes aspectos relativos ao meio de vida das serpentes, desmistificação de conceitos incorporados as concepções dos estudantes. Nos primeiros momentos da atividade é perceptível a postura de guarda por parte dos alunos, o primeiro contato com a serpente ainda que de forma visual suscite espanto, no entanto ao decorrer da palestra essas sensações transformaram-se em um olhar de curiosidade e busca por compreender a biologia do animal. Os visitantes foram convidados a conhecer alguns aspectos da biologia das serpentes, principalmente sua evolução, morfologia, fisiologia e importância ecológica, isso partindo dos conhecimentos construídos em sua vivência, desmistificando conceitos e explorando bases científicas. Decorrido o impacto inicial, os alunos se mostraram participativos a apresentação, demonstrando curiosidade em tocar no animal (Figura 5), entender como suas características tão peculiares lhes possibilitam sobrevivência, conforme demonstra observação realizada por uma aluna: “como um bicho sem patas consegue andar”, o interessante nessa fala é perceber como a ação andar esta diretamente ligada aos movimentos de membros, e as serpentes desprovidas de tal são capazes de locomover-se. Figura 5: Ao decorrer da atividade o veterinário do Sítio abordou pontos sobre as peculiaridades do grupo em estudo, ao explanar sobre fisiologia foi mostrado aos alunos serpentes de diferentes tamanhos, após o “espanto” inicial podemos ver a interação e curiosidade durante essa parte do diálogo, as faces estão desfocadas para preservar a identidade dos alunos. Figura 5: Ao decorrer da atividade o veterinário do Sítio abordou pontos sobre as peculiaridades do grupo em estudo, ao explanar sobre fisiologia foi mostrado aos alunos serpentes de diferentes tamanhos, após o “espanto” inicial podemos ver a interação e curiosidade durante essa parte do diálogo, as faces estão desfocadas para preservar a identidade dos alunos. Outros questionamentos foram levantados pelos estudantes durante a palestra, destacaram-se principalmente o que era relativo à fisiologia e morfologia, tais como: “a pele das cobras é composta de que?”, o médico veterinário responsável pela palestra buscou realizar analogia entre a pele da serpente e as unhas dos alunos, visto que ambos são constituídos do mesmo composto, a queratina, e por isso o aspecto brilhante das serpentes, o que muitas pessoas pensam ser algo escorregadio e úmido. 3.2. Ação educativa: aproximações entre ciência e cultura Em relação à troca de pele, abordada na visita houve indagação: “o ‘desenho’ da pele muda com a troca?”. Neste caso, o “desenho” ao qual o aluno se refere está relacionado ao que pode ser considerada a impressão digital da serpente, conforme destacou o palestrante. Novamente recorrendo à analogia, ressaltou que assim como nós a impressão do animal é única e exclusivamente própria e não se altera ao processo de muda. Destaca-se que o uso de analogias empregadas pelo palestrante constitui como uma importante estratégia pedagógica que versa por diminuir a distância entre o conceito e o objeto, Oliveira [21] destaca que “a partir do momento em que o objeto do conhecimento se mostra inacessível à experiência, uma relação analógica pode sugerir seu esclarecimento. Relacionado às serpentes as analogias promovem uma relação de proporcionalidade direcionando a uma compreensão onde os indivíduos percebam as diferentes estratégias adaptativas que os seres convergiram ao longo do processo evolutivo. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 11 Um aspecto relativo à fisiologia surgiu a partir da indagação de outro estudante: “as cobra respiram embaixo da água?” As serpentes não possuem guelras, são animais pulmonares, destacou o palestrante, o que as possibilita explorar ambientes aquáticos é o fato de um de seus pulmões serem reduzidos, e o lado desenvolvido responsável pelas trocas gasosas possuir maior extensão sendo capaz de armazenar significativa quantidade de ar. Seu papel ecológico foi posto em foco na palestra, pois como se constatou nas análises dos questionários os indivíduos pontuavam satisfatoriamente sua importância no ambiente como controladores de espécies inclusive de sua própria espécie, e participação efetiva na cadeia alimentar. A visita seguiu com a exploração dos espaços presentes no sítio, como laboratório, recintos baia de criação de preás, alimento das serpentes. Ao final da atividade o que era espanto, tornou-se efetivamente um olhar de curiosidade e compreensão relativo às serpentes, onde os alunos perceptivelmente mudaram sua concepção primária de animal perigoso fortemente empregado nas diversas culturas, inclusive no meio em que vivem. Assim, até quem relatou ter pavor, passou a vislumbrar as serpentes sob uma nova óptica (Figura 6), pois o conhecer nos permite abandonar ideais que comumente são vivenciadas sem a devida fundamentação científica. 5. REFERÊNCIAS BIBLIOGRÁFICAS 1. Franco FL. Origem e diversidade das serpentes. In: Cardoso, J. L. C. et al. Animais peçonhentos no Brasil biologia, clínica e terapêutica dos acidentes.2003.São Paulo: Sarvier. 13-32 p. 1. Franco FL. Origem e diversidade das serpentes. In: Cardoso, J. L. C. et al. Animais peçonhentos no Brasil biologia, clínica e terapêutica dos acidentes.2003.São Paulo: Sarvier. 13-32 p. 2. Bernarde PS, Albuquerque S, Barros TO, Turci LCB. Serpentes do estado de Rondônia, Brasil. Biota Neotropica, 2012;12(3):1- 29. 2. Bernarde PS, Albuquerque S, Barros TO, Turci LCB. Serpentes do estado de Rondônia, Brasil. Biota Neotropica, 2012;12(3):1- 29. 3. Souza CEP, Souza JG. (RE)conhecendo os animais peçonhentos: diferentes abordagens para a acompreensão da dimensão histórica, sócio-ambiental e cultural das ciências da natureza. In: V Encontro Nacional de Pesquisa em Educação em Ciências. Associação Brasileira de Pesquisa em Educação em Ciências. Atas. 2006. Nº 5. Disponível em: <http://www.nutes.ufrj.br/abrapec/venpec/conteudo/artigos/1/pdf/p847.pdf>.Acesso em: 03. Set. 2014. 4. Barbosa AR. Os humanos e répteis da Mata: Uma abordagem Etnoecológica de São José da Mata Paraíba. Revista de Biologia e Ciências da Terra. 2007; 7(2):117- 123. g 5. Moura MR, Costa HC, São-Pedro VA, Fernandes VD, . Feio RN. O relacionamento entre pessoas e serpentes no leste de Minas Gerais, sudeste do Brasil. Biota Neotropica. 2010; 10(4):133- 141. 6. Dodd Jr CK. Strategies for snake conservation. In: Siegel R.A, Collins J.T. (Eds). Snakes.ecology an behavior. New York: McGraw-Hill. 1993. 363-393 p. 7. BRASIL. Portaria IBAMA nº 117, de 15 de outubro de 1997. Disponível em<http://licenciamento.cetesb.sp.gov.br/ legislacao/federal/portarias/1997_Port_IBAMA_117.pdf > acesso: 20. out. 2014. 8. BRASIL. Ministério da Educação. Secretaria de Educação Média e Tecnológica. PCN + Ensino Médio: orientações educacionais complementares aos parâmetros curriculares nacionais: ciências da natureza, matemática e suas tecnologias. 2002. Brasília. g 9. Soares SC, Ruiz CM, Rocha DV, Jorge KM, Senkowski STVS, Ortêncio Filho H, Magalhães Júnior CAO. Percepção dos Moradores de Goioerê - PR, sobre a Fauna Silvestre Urbana. Arquivos do MUDI 2011;15(1/2/3):17- 30. 10. Gil AC. Como elaborar projetos de pesquisa. São Paulo: Atlas. 2002. 176 p. 11. Marconi MdeA, Lakatos EM. Fundamentos da Metodologia Cientifica. 2007. 6.ed. São Paulo: Edito Atlas S. A. 12. Appolinário F. Metodologia da ciência: filosofia e prática da pesquisa. 2009. São Paulo. Cengage Learning. g 13. Tavares CC, Tavares MR. Jogos didáticos no ensino de ciências: ensinando e aprendendo através do lúdico. [Trabalho de Conclusão de Curso]. Universidade do Estado do Pará. 2013. 69 p. 14. 4. CONCLUSÃO A valia da pesquisa se concretizou no alcance de seus objetivos, referente a mudança de postura dos alunos o que antes era visto por um olhar de medo, atualmente configura-se como respeito a biologia das serpentes, pontuamos também a (re)construção das visões do senso comum construídas ao longo da vida, sendo estas de caráter mítico, religioso e barreira pessoal. Relativo à prática de educação ambiental o Sítio demonstrou ser um ambiente onde a educação informal pode ser desenvolvida de maneira satisfatória, pois o planejamento direcionado para o público alvo da pesquisa atingiu os objetivos propostos, proporcionando ação reflexiva nos indivíduos Tocar na serpente já não era visto como atitude de coragem, mas sim como uma forma de participar da ação, e claro todos os cuidados necessários foram tomados, visto que se trata de um animal silvestre e que precisa ter seu espaço respeitado. q p p ç p Consideramos que trabalhos dessa natureza são de importância para a prática de conscientização ambiental para a manutenção da sobrevivência de espécies com valor “desmerecido” que são construídos ao através de conhecimentos incipientes cientificamente. q p O conhecimento ainda é o principal meio para que o equilíbrio e convívio harmônico entre as espécies se estabeleçam, pois não se protege o que não se conhece. O conhecimento ainda é o principal meio para que o equilíbrio e convívio harmônico entre a espécies se estabeleçam, pois não se protege o que não se conhece. 3.2. Ação educativa: aproximações entre ciência e cultura Figura 6: A imagem de finalização da atividade mostra que o conhecer muda a concepção e o comportamento do que se tinha antes, a referida aluna da imagem participou do diálogo de forma receosa e distante, ao término não demonstrava medo como antes, face desfocada para preservar sua identidade. Figura 6: A imagem de finalização da atividade mostra que o conhecer muda a concepção e o comportamento do que se tinha antes, a referida aluna da imagem participou do diálogo de forma receosa e distante, ao término não demonstrava medo como antes, face desfocada para preservar sua identidade. Desta forma, o Sítio Xerimbabo pode ser considerado um ambiente apto ao ensino informal por desenvolver práticas educativas de maneira plena versando sobre os princípios básicos da educação ambiental, relativos à conservação da fauna ofídica. Desta forma, o Sítio Xerimbabo pode ser considerado um ambiente apto ao ensino informal por desenvolver práticas educativas de maneira plena versando sobre os princípios básicos da educação ambiental, relativos à conservação da fauna ofídica. Pois de acordo com Brasil [22] o Artigo 13, da Lei n.º 9.795, a EA (Educação Ambiental) e as práticas educativas devem ser voltadas à “sensibilização da coletividade sobre as questões ambientais e à sua organização e participação na defesa da qualidade do meio ambiente”. Reis [23] salienta que a informação é à base da educação ambiental, por esse motivo se faz necessário apresentar as questões abordadas partindo das premissas empiristas dos sujeitos, visto que a fragmentação que separa o cultural e o natural presente ensino formal [24], pode ser o fator que distancia sujeito e objeto. Na visita constatou-se a abertura que os alunos mantiveram com o palestrante, sem o constrangimento de estarem certos ou errados, configurando um caráter mais exploratório da temática, e a proximidade do objeto (serpente), propiciando a unidade de ser humano e natureza. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 12 5. REFERÊNCIAS BIBLIOGRÁFICAS Almeida GVL, Santos EM. A Salamanta (Epicratesassisi Machado, 1945) É um animal venenoso: percepção de algumas comunidades do sertão de Pernambuco. In: Seabra, Giovanni; Mendonça, I. (org.). Educação ambiental: responsabilidade para a conservação da sociobiodiversidade. 2010. João Pessoa. 36-41 p. A.W.P. da Silva et al.,Scientia Plena 12, 069925 (2016) 13 15. Anderson EN. Ecologies of the heart: emotion, belief, and the environment. Anthrozoös, 1996;15(2):119-138. 16. Rosa G. Grande Sertão: veredas. Editora Nova Aguilar. vol. 2. 1994. 477 p. 16. Rosa G. Grande Sertão: veredas. Editora Nova Aguilar. vol. to Butantan. Série Didática. 2006. São Paulo. 7 ed. 18. Langanke R. Conservação para o ensino médio. SNUC – Sistema Nacional de Unidades de Conservação. 2010 Disponível em: <http://eco.ib.usp.br/lepac/conservscao/ensino/bioma_snuc.htm>. Acesso em: 29 out. 2010. 19. Leite MCVA. Cobras e sapos: esses bichos malditos! Um estudo sobre a relação entre saberes populares e saberes acadêmicos na Educação Ambiental. 2004. [Dissertação de Mestrado] - Faculdade de Psicologia e de Ciências da Educação da Universidade do Porto. populares e saberes acadêmicos na Educação Ambiental. 2004. [Dissertação de Mestrado] - Faculd de Psicologia e de Ciências da Educação da Universidade do Porto. 20. Feitosa RA, Abílio FJP. Dizendo Cobras e Lagartos: Uma experiência de Educação Ambiental co futuros professores de Biologia. Experiências em Ensino de Ciências. 2012; 7(3):1-13. 21. Oliveira HR. Argumentação no Ensino de Ciências: o uso de Analogias como Recurso para a Construção do Conhecimento. [Dissertação]. Universidade Federal de Juiz de Fora. 2012. 131 p. ç ç p 22. BRASIL. Lei n° 9.795, de 27 de abril de 1999. Disponível em :<https://www.planalto.gov.br/ccivil_03/leis/l9795.htm>. Acesso em: 14. Out. 2014. 23. Reis LCL, Semêdo LTAS, Gomes RC. Conscientização ambiental: da educação formal a não formal. Revista Fluminense de Extensão Universitária. Universidade Severino Sombra, Vassouras. 2012; 2(1):47-60. 24. Bonotto DMB, Santos JR. Educação ambiental e animais não humanos: linguagens e valores atribuídos por professoras do ensino fundamental. Pesquisa em Educação Ambiental. 2012;7(1):09- 27.
https://openalex.org/W4250766098	https://revistadeeducacaofisica.emnuvens.com.br/revista/article/download/344/372	Portuguese	null	CONTRIBUIÇÕES DA MISSÃO MILITAR FRANCESA PARA O DESENVOLVIMENTO DO DESPORTO NO EXÉRCITO BRASILEIRO. Parte 2	Revista de Educação Física	1,970	cc-by	3,648	THE CONTRIBUTION OF THE MILITARY FRENCH MISSION TO THE DEVELOPMENT OF SPORTS IN THE BRAZILIAN ARMY (100 Year-celebration on the beginning of orientation in that mission) Part two A d é M d Rib i 1 André Morgado Ribeiro1 1 Instituto de Pesquisa da Capacitação Física do Exército (IPCFEx) – Rio de Janeiro – RJ – Brasil Resumo: O presente artigo foi idealizado no intuito de descrever as importantes contribuições da Missão Militar francesa no campo das práticas esportivas, na promoção do aprimoramento do condicionamento físico e a difusão, no público civil, de sua importância para a saúde e a disciplina no Exército Brasileiro. Dessa forma, primeiramente abordou-se a importância do treinamento físico militar no Brasil e suas inﬂ uências. Em seguida, foi realizada uma pesquisa sobre os contextos históricos brasileiros da Educação Física e o Exército, sob o enfoque da inﬂ uência francesa. Por ﬁ m, tratou-se dos esportes e das atividades físicas como instrumentos da formação cívica e do desenvolvimento da disciplina entre seus praticantes. p Palavras-chaves: Exército. Educação Física. Missão Militar francesa. p Palavras-chaves: Exército. Educação Física. Missão Militar francesa. Abstract: The present study has the objective of describing the important contributions of the French Military Mission to sports, to the promotion of the improvement of physical conditions, and the diffusion among civilians of its importance to health and discipline in the Brazilian Army. This way, the importance of military physical training in Brazil and its inﬂ uence were approached. Next, a research on the historical background of physical education and the Army was carried out, with great emphasis on the French inﬂ uence. Then, sports and physical activities as instruments for civic formation and the development of discipline among people were signiﬁ cantly treated as well. Keywords: Army. Physical Education. French Military Mission. Aceito em 11/09/2009 - Rev. Educ. Fís. 2009 - 9-15. Rio de Janeiro - RJ - Brasil Aceito em 11/09/2009 - Rev. Educ. Fís. 2009 - 9-15. Rio de Janeiro - RJ - Brasil INTRODUÇÃO conhecimentos e as criações intelectual, artística e profissional. Dessa interação surgiu uma nova identidade, uma rica e universal cultura que se consolida e prospera ao longo de mais de trezentos anos(1). As origens do Exército brasileiro remontam à primeira metade do século XVII, quando brancos, negros e índios do nordeste se uniram para expulsar o invasor holandês. Nessa ocasião foi utilizado, pela primeira vez, no Brasil, o termo pátria. O patrimônio imaterial da Força terrestre tem sua gênese em Guararapes. Diferentes culturas interagiram e se fundiram por meio de inúmeros fatores entre os quais destacamos os traços biológicos transmitidos, a transmissão de Nessa fusão de raças está a história do nosso Exército, instituição que se preocupa com a valorização do homem, agente criador que se manifesta em diversas atividades culturais da pátria. O Exército tem sido no Brasil muito mais do que uma simples instituição armada prevista na organização do estado para desempenhar 9 Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil 10 não é uma característica apenas brasileira. É, na realidade, aspecto atemporal que se faz presente em vários países(1). Não existem referências relativas às práticas relacionadas à atividade física do Exército português no Brasil, no momento da vinda da família real para o país. Tais práticas, caso registradas, demonstrariam com clareza o legado de Portugal, antes que houvesse a influência de outras escolas ou métodos aprendidos a posteriori. Assim, sabe-se apenas que D. João VI, apesar do receio de educar a população da colônia, terminou por fundar a primeira escola superior do Brasil, que posteriormente deu origem à Academia Militar das Agulhas Negras. Apesar de, mais uma vez, inexistirem registros, admite-se que a prática de atividades físicas provavelmente faria parte da rotina dos militares da Academia. Tais atividades incluiriam a esgrima, a equitação e a natação(2). sua destinação constitucional: a defesa da pátria assegurando a preservação da soberania nacional, na garantia dos poderes constitucionais, na manutenção da lei e da ordem e de operações humanitárias e de cooperação internacional. Neste artigo especificamente, há de se ressaltar o importante papel do Exército na evolução e difusão dos esportes e, principalmente, da formação dos recursos humanos em Educação Física, no Brasil, como fator incontestável na valorização do homem em diversas partes do território nacional. Ao longo da história muitas foram as influências sofridas pela Instituição. INTRODUÇÃO Nas décadas de 1920 e 1930, integrou-se na Força Terrestre uma Missão Militar Francesa composta de oficiais, subalternos e civis assemelhados. O acordo dos Franceses para servir ao Exército deveu-se por duas razões: o despreparo do Exército e a evolução da 1ª Grande Guerra, na qual a França foi vitoriosa. A Missão Militar Francesa contribuiu para que ocorressem expressivas mudanças na área esportiva e de treinamento físico no Exército brasileiro, com relevantes reflexos na História do Brasil. (A Missão Militar Francesa no Brasil, 1983, p.6) Datam dessa época, publicações importantes na área da educação física, que lançadas em Portugal, acabaram por chegar ao Brasil. São elas: “Tratado de Educação Física e Moral” de Luiz Carlos Muniz Barreto — 1787; “Tratado de Educação Física dos Meninos”, de Francisco Melo Franco — 1890 e o “Tratado de Educação Física”, de Francisco José de Almeida — 1891. Apesar de tudo que foi descrito acima, permanece a necessidade de mais informações para a compreensão do exercício do papel da Missão Militar Francesa no Exército brasileiro e sua contribuição para a interpretação do Esporte nacional. É evidente que esta interpretação sempre foi relacionada aos contextos históricos de cada época. A partir de 1820 pode-se perceber uma mudança na vida e nos costumes dos brasileiros porém, após a independência, iniciativas são tomadas com o objetivo de tornar o Brasil uma nação desenvolvida(3). As mudanças na estrutura econômica nacional devido ao desenvolvimento da indústria, ao crescimento da área urbana e ao surgimento de um empresariado brasileiro deram origem a um contexto favorável à implementação da prática esportiva no país(3). OS CONTEXTOS HISTÓRICOS BRASILEIROS DA EDUCAÇÃO FÍSICA E O EXÉRCITO A forma como as atividades físicas eram desenvolvidas no Exército, não ficou imune a todas essas modificações. O decreto nº 2.116 de 1858 estabeleceu que as práticas escolares incluíssem a esgrima e a natação para os cursos de Infantaria e Cavalaria da Escola Militar. Nesse mesmo ano, a esgrima, a ginástica e a natação se tornam práticas obrigatórias na Escola da Marinha. A história da educação física no Brasil se confunde com a história dos militares e um melhor entendimento do desenvolvimento do esporte no Brasil apenas se torna possível através de uma análise cuidadosa da participação do Exército brasileiro nesse contexto. A ligação entre educação física e militarismo 11 Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil 11 regular(3). Nessa época, as publicações sobre o tema começam a surgir no país. Surgia o interesse pelos esportes náuticos e a capoeira era muito praticada na época do Brasil Império, ainda sob a égide da escravidão. O esporte mais praticado no período imperial, entretanto, foi o remo, inclusive com a organização e disputa de importantes provas(2). Como não existiam cursos de formação de professores de educação física no país nesse período, que ficou sob a responsabilidade dos militares durante o século XIX(5). Tal responsabilidade não se restringia apenas às Escolas Militares, uma vez que sua atuação se estendia ao meio civil. Os esportes náuticos recebem grande impulso com o surgimento da Escola Militar na Praia Vermelha. Um clube foi organizado visando angariar fundos para que os cadetes pudessem praticar a canoagem. A pesca era outra prática de grande receptividade na época, assim como as escaladas ao Pão de Açúcar e Morro da Urca, que se constituíam em provas de tenacidade, coragem e resistência para seus praticantes(4). O método alemão, de forte caráter militar, embasava a prática da época, visando manter o povo forte e saudável. Impregnado de civismo e patriotismo, aspectos relacionados ao movimento nacionalista alemão, acabou por criar forte oposição, pois era percebido como algo negativo, que estimularia hábitos e costumes bárbaros, considerados comuns aos brasileiros na época. Acreditava-se que tais características reforçariam esses aspectos negativos da personalidade brasileira. FIGURA 1 EXAMES MÉDICOS PARA A COLÔNIA DE FÉRIAS DA URCA. Os entusiastas consideravam a disseminação do esporte no Brasil como um aspecto positivo, visto que auxiliaria na educação de muitos indivíduos, pois contribuiria para o controle de instintos e paixões. OS CONTEXTOS HISTÓRICOS BRASILEIROS DA EDUCAÇÃO FÍSICA E O EXÉRCITO Foto do arquivo digital: Museu do Desporto do Exército O desenvolvimento da educação física e do esporte tem continuidade dentro da Força Terrestre. Através do decreto nº 330 de 1890, estabelecendo que as Escolas Militares da capital federal e do Rio Grande do Sul passariam a contar com uma sala d’armas, campo de exercício, linha de tiro, picadeiro, barca e outros aparelhos. No mesmo ano, o decreto nº 371 determinava que os alunos do Colégio Militar, visando o seu desenvolvimento físico e intelectual, receberiam instruções quanto à nomenclatura e manejo de armas, tiro ao alvo, esgrima, evoluções militares, natação, ginástica e música. Foto do arquivo digital: Museu do Desporto do Exército As alterações no ensino militar, embasadas pelo avanço da ciência e das artes, tinha como principal entusiasta o Marechal Mallet, ministro da Guerra de Campos Salles. Mallet acreditava que o ensino deveria ser contínuo, intensivo e rigoroso. Queria adotar a didática do “fazer para aprender” e não ficar apenas baseado em livros. Apesar da educação física e a prática de esportes serem temas de interesse e discussão na arena do ensino nacional desde 1823, a implementação de atividades físicas nas escolas públicas sob a responsabilidade do governo não se deu de forma tranqüila, pois existiam resistências ao caráter e natureza das mesmas. Somente a partir da segunda metade do século XIX é que a educação física passa a fazer parte do currículo escolar de forma mais O processo de modernização do Exército teve continuidade. Coube ao Marechal Hermes da Fonseca defender a criação de grandes Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil 12 existentes foram superados pelo espírito de união resultante da farda(6). unidades militares, a instituição do serviço militar obrigatório e a criação dos tiros de guerra. Durante o mesmo ano, surge o projeto prevendo a criação de duas escolas de educação física, uma civil e outra militar, baseado no método de ginástica sueca, que embora aprovado, não foi implementado(2). Com a derrota alemã na primeira Grande Guerra, as missões militares francesas passaram a ser a maior fonte de influência para a modernização do Exército brasileiro. As metas traçadas foram: a adoção e aplicação da lei do serviço militar, ampliação dos efetivos como conseqüência do serviço militar obrigatório; renovação do armamento e melhoria das instalações, implantação de uma nova estrutura organizacional e reforma completa do ensino com ênfase na formação profissional. OS CONTEXTOS HISTÓRICOS BRASILEIROS DA EDUCAÇÃO FÍSICA E O EXÉRCITO O método alemão adotado no país manteve a sua supremacia até 1912, sendo substituído pelo método francês. A modificação foi devido à atuação da Missão Militar Francesa no país, em 1907, que teve como desdobramento a criação da primeira Escola de Educação Física no país em 1909, na Força Pública de São Paulo e que formou os primeiros instrutores e mestres de esgrima ainda no mesmo ano(2). A escola francesa lançava mão de jogos, da ginástica, dos esportes individuais e dos esportes coletivos(6). Em 1921 é aprovado o Regulamento de Instrução Física Militar, com base no método Hébert e adaptando as teorias de Joinville Le Pont. Com a aprovação do Regulamento, a escola francesa substitui o método alemão de forma deﬁ nitiva. Vários militares participavam das equipes de futebol dos principais clubes do Rio de Janeiro. Eram tão numerosos que acabaram por constituir uma seleção militar que enfrentou a seleção do Chile. Esses militares organizavam equipes e competições amistosas na Vila Militar, onde serviam. Utilizavam para as disputas o campo localizado ao lado do 1º Regimento de Artilharia Montado. A construção do campo foi iniciativa do Tenente Francisco Mendes, atleta do Fluminense Football Club. Todo esse movimento resultou na fundação da Liga Militar de Futebol em 1915, passando a se denominar Liga de Sports do Exército, em 1920. Em 1922 é criado o Centro Militar de Educação Física, destinado a difundir o novo método de educação física e suas aplicações práticas. Com essas modificações, um novo modelo de oficial passa a ser formado nas Escolas Militares. Era um oficial que empregava seu tempo na sua preparação técnica, dedicava- se ao estudo, tinha forte senso de disciplina e hierarquia e compreendia a relação entre corpo e espírito. Além disso, tentava compreender os problemas nacionais e elaborava teorias relativas ao futuro do país. Havia aprendido com os franceses o que era uma força armada moderna, eficaz tecnicamente, com senso de unidade e disciplina(1). A Liga, juntamente com a Comissão Militar Sportiva e as comissões especiais designadas para os Jogos Latino- Americanos se reúnem para coordenar os esportes hípicos, o atletismo, o tiro, a esgrima, o futebol e o pentatlo moderno, que pela primeira vez seria disputado em um país da América do Sul. As disciplinas de caráter teórico voltam a ser valorizadas, sendo ministradas em combinação com as atividades práticas. ATIVIDADE FÍSICA – DÉCADA DE 1930 ATIVIDADE FÍSICA – DÉCADA DE 1930 Foto do arquivo digital: Museu do Desporto do Exército O Centro Militar de Educação Física tinha como objetivos dirigir, coordenar e difundir o novo método de Educação Física militar e suas aplicações esportivas. Oferecia cursos de Educação Física para oficiais, sargentos e um curso de preparação para oficiais. No ano de 1929, é baixada uma determinação para que providências imediatas sejam tomadas no sentido de criar um Curso Provisório de Educação Física. Uma nova missão francesa chega ao país, trazendo Pierre de Seguir, responsável pelo desenvolvimento de metodologias para diversas práticas, em especial as lutas. Uma turma de dez oficiais, 28 sargentos e 20 professores de escolas públicas do Distrito Federal são selecionados para formar a primeira turma de Educação Física do Brasil. O método aplicado era eminentemente científico, com bases pedagógicas, anátomo- fisiológicas e psicológicas da Escola Francesa de Joinville-le-Pont. Foto do arquivo digital: Museu do Desporto do Exército Foto do arquivo digital: Museu do Desporto do Exército A prática sistematizada de exercícios físicos pelos militares visava a manutenção da boa forma dos combatentes e o aperfeiçoamento da disciplina da tropa. Os exercícios de ordem unida são considerados úteis por favorecerem a disciplina, a coesão e a absoluta obediência às ordens recebidas. Os exercícios objetivam preparar a tropa para executar movimentos com ordem e rapidez, em concordância com o indicado pelo comandante, a despeito de perigos, dificuldades e privações(7). Um periódico chamado “O Jornal” de 1º de janeiro de 1930 descreveu a solenidade de formatura dos primeiros técnicos de Educação Física formados pelo Exército: “Findo o juramento, o Ministro da Guerra e todos os presentes entraram no grande salão de suas aulas da Escola. Estava vistosamente ornamentado. O seu tecto era formado por uma Bandeira Nacional, trabalho em laços de papel de seda. Em lugares destacados ﬁ caram os convidados. O General Sezefredo e altas autoridades tomaram assento à mesa que presidiu a sessão solene comemorativa do acto. Foram, então, proclamados pelo Capitão Caldas os nomes das 62 praças que concluíram o curso da Escola de Sargentos e foram promovidas a inferiores do Exército. Em nome da turma falou um dos alumnos (sic). OS CONTEXTOS HISTÓRICOS BRASILEIROS DA EDUCAÇÃO FÍSICA E O EXÉRCITO São produções teóricas da época os seguintes textos: “Regulamento de Instrução Física”, “Noções de Higiene” e “Noções de Anatomia e Fisiologia”(3). Após os Jogos Latino-Americanos, fica patente a inferioridade esportiva brasileira, em virtude da deficiência do preparo físico dos atletas. Para resolver o problema, a Escola de Sargentos de Infantaria, contando com o auxílio de mestres franceses, edita um manual de Educação Física que passou a ser referência para todos os instrutores de educação física no país. A partir de 1931, a Escola Militar de Realengo passa por um período de mudanças que se basearam nos moldes das escolas militares americana, francesa e inglesa. É criado o Departamento de Educação Física, visando centralizar as funções de planejamento e execução(2). Com o objetivo de se modernizar, o Exército se dividia entre a escola alemã e a francesa. Entretanto, quaisquer conflitos porventura Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil 13 é apresentada a nova Organização do Exército, havendo a previsão da criação de uma escola de educação física. Tal aspiração se concretiza com a fundação do Centro Militar de Educação Física (CMEF), na Escola de Sargentos de Infantaria, Vila Militar. ATIVIDADE FÍSICA – DÉCADA DE 1930 Seguiu-se a entrega dos diplomas aos sargentos, em número de 28, que concluíram o curso de educação physica (sic), bem como os primeiros ofﬁ ciaes (sic) que ﬁ zeram.” A influência militar na educação física brasileira fica ainda mais evidente à medida que a mesma se desenvolve. A Escola de Esgrima, sediada no Batalhão de Caçadores, representa o esforço para a criação de uma instituição voltada para a formação de professores e instrutores de educação física. Em 1919 se inicia movimento na Escola Militar, que busca criar e consolidar a Escola de Educação Física do Exército, capitaneado pelo tenente Newton Cavalcanti, mais tarde diretor da escola(8). Essas aspirações são levadas ao Presidente da República, através do manifesto da União Atlética da Escola Militar. Porém, o boletim do Exército nº 431, é considerado o marco inicial do movimento que objetivava a inclusão da Educação Física no Exército. Nesse documento 14 Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil de Desportos do Exército. Não diferente foi a criação do Centro Militar de Educação Física, passando a ser chamado de Escola de Educação Física. Sua cooperação para o desenvolvimento do esporte nacional é incansável. Nos atuais dias, desde a criação do Instituto de Pesquisa da Capacitação Física, coordena e executa projetos e trabalhos de pesquisas voltados para a operacionalidade da Força Terrestre e à saúde de seus integrantes, mantendo o compromisso com as ciências dos esportes. Em 1930 é realizado um segundo curso. Os primeiros colocados são aproveitados como instrutores das futuras turmas. Essa prática possibilitou uma rápida ampliação do quadro de profissionais(8). As instalações destinadas à parte náutica, localizada na Escola de Sargentos, eram extremamente precárias e necessitavam ser transferidas para um local mais apropriado(1). Por razões econômicas e de localização, o Forte São João é selecionado como nova sede, recebendo, assim, o recém criado Centro Militar de Educação Física (CMEF), em 1930. Em alguns momentos, fez-se pioneiro, ao buscar a formação profissional, a organização de competições e o desenvolvimento e apresentação de novos métodos no campo da Educação Física. Em outros, mostrou-se capaz de buscar, junto à sociedade civil, os meios de manter-se em constante aperfeiçoamento e assegurar sua posição como centro de excelência no cenário esportivo nacional. No mesmo ano, inicia-se a expansão do ensino da educação física no país, com a criação de centros similares em São Paulo e Minas Gerais(3). ATIVIDADE FÍSICA – DÉCADA DE 1930 Ao término de 1930, o Brasil já possuía 170 profissionais de educação física formados no país. Contava também com alguns que haviam concluído a sua formação no exterior. Desse total, 150 eram militares e 20, civis. Vale ressaltar, que esses vinte civis tiveram a sua formação orientada por militares(3). REFERÊNCIAS BIBLIOGRÁFICAS Em 19 de outubro de 1933, Getúlio Vargas, através do decreto 23252, transforma o Centro Militar de Educação Física em Escola de Educação Física do Exército (EsEFEx), dando- lhe nova organização, atualizando os seus currículos e ampliando os seus objetivos(9). 1. Neto AF. A Pedagogia no Exército e na Escola: a educação física brasileira (1880-1950). Aracruz: Facha, 1999. 2. Tubino MG. Metodologia cientifica do treinamento desportivo. São Paulo: Ibrasa, 1984. 7. Facó J. Princípios e Métodos de Instrução. A Defesa Nacional. Rio de Janeiro: V. 12. n. 139. p. 187–190. jul/ago. 1952. 9. Revista de Educação Física. Rio de Janeiro: n. 12, 1933. p. 2. 15 Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil CONCLUSÃO 3. Melo VA. História da educação física e do esporte no Brasil: panorama e perspectivas. São Paulo: Ibrasa, 1999. A partir desse estudo, pode-se perceber o quão significativa e variada é a contribuição da CDE, do IPCFEx e da EsEFEx e, conseqüentemente, do Exército Brasileiro para o esporte nacional. Tal contribuição mostrou- se sempre adequada à situação política e à necessidade demonstradas pelo corpo esportivo nacional. 4. Ramos JJ. Os exercícios físicos na história e na arte: do homem primitivo aos nossos dias. São Paulo: Ibrasa, 1982. 5. Melo VA. Relação Teoria & Prática e Formação profissional na Educação Física Brasileira: Apontamentos na história. Florianópolis: Revista Motrivivência n.º 8, Ano 7; 1995. p.103–115. Imbuído do espírito de pioneirismo e sempre buscando a excelência, incentivado pela Missão Militar Francesa, procurou o intercâmbio com outros países. 6. Neto MD. Influência estrangeira e luta interna no exército; 1889-1930. O Exército inicia o estímulo à Educação Física nos primórdios do século passado com a criação da Liga Militar de Futebol passando a denominar-se Liga de Sports do Exército, posteriormente conhecido como Comissão 7. Facó J. Princípios e Métodos de Instrução. A Defesa Nacional. Rio de Janeiro: V. 12. n. 139. p. 187–190. jul/ago. 1952. 15 Revista de Educação Física 2009; 9-15. Rio de Janeiro (RJ) - Brasil 8. Molina A. Escola de Educação Física do Exército. Sua atuação em prol da Educação Física nacional. Rio de Janeiro: Revista de Educação Física. n. 25. 8. Molina A. Escola de Educação Física do Exército. Sua atuação em prol da Educação Física nacional. Rio de Janeiro: Revista de Educação Física. n. 25. 9. Revista de Educação Física. Rio de Janeiro: n. 12, 1933. p. 2. Endereço para correspondência: Avenida João Luís Alves s/nr – Urca - Rio de Janeiro – RJ – CEP 22.291-090 Telefone: (21) 2586-2269 ou (21) 8585-3042 Fax:(21) 2295-5340 e-mail: andre.morgado@globo.com
https://openalex.org/W4313166265	https://www.nomos-elibrary.de/10.5771/9783748935544-177.pdf?download_full_pdf=1&page=1	English	null	Synopsis	Nomos Verlagsgesellschaft mbH & Co. KG eBooks	2,022	cc-by	286	787 ‘Various conflicting interpretations have arisen, hindering the quest for a co- herent approach to occupation law. Citing these central provisions, occupying powers have often justified a very large scope of authority over occupied ter- ritories. In other cases, foreign administrators have invoked the obligation to respect local laws in order to minimize their authority and evade their respon- sibilities under occupation law. This situation is unsatisfactory and work has to be done in this regard to avoid discrepancies in the interpretation and imple- mentation of the law in contemporary contexts of occupation’ (Tristan Ferraro (ed), ICRC Expert Meeting Report: Occupation and Other Forms of Adminis- tration of Foreign Territory (International Committee of the Red Cross 2012) 54).f yf 789 8.4 Reconciliation with the Munitions of War Rule, 166ff. ) 788 8.1.1 Priority over Occupation Law, 157ff.f 788 8.1.1 Priority over Occupation Law, 157ff. 789 8 4 Reconciliation with the Munitions of War Rule 166ff Synopsis Contrasting the conclusions of Part I and Part II, it becomes clear that Chapter XI can shift the legal paradigm for military occupations. This paradigm shift away from traditional occupation law is warranted.787 The economic leeway that the occupant enjoyed under occupation law is removed by Chapter XI and replaced with the interests of the inhab- itants.788 While occupation law created an economic incentive to stay in foreign territory by force, Chapter XI not only eliminates this incen- tive, but deters occupation. Under Chapter XI, States are still allowed to defend themselves militarily, but they must not use foreign territory eco- nomically.789 Chapter XI could thus end not only historic colonialism, but all forcible stay in foreign territory. f 789 8.4 Reconciliation with the Munitions of War Rule, 166ff. 177
https://openalex.org/W2972053557	https://zenodo.org/records/3392572/files/13850__1_223270_LE_333668.pdf	Dutch	null	In principe geregeld, of in de regel principieel?	MAB	2,007	cc-by	1,347	Roger Dassen in de loop van de tijd. Toegepast op verslaggevings- regels constateert hij dat ‘principles-based’ stan- daarden hun kracht ontlenen aan de consistentie met een conceptueel raamwerk, waardoor de behoeft e aan voorbeelden, uitzonderingen of nog meer gedetail- leerde regels aanzienlijk wordt beperkt. Van der Tas be spreekt de voor- en nadelen van principles-based standards setting, en constateert dat de tendens naar meer gedetailleerde standaarden onvermijdelijk is in een verjuridiserende samenleving. Niet verbazing- wekkend dus dat ook IFRS – met zijn bijna 2500 pagina’s – volgens hem niet kan gelden als principles- based. Niettemin ziet hij reële mogelijkheden voor een meer principles-based benadering binnen IFRS, onder een aantal voorwaarden, waaronder een kwali- tatief hoogwaardig Framework. Van der Tas laat daarmee op uiterst constructieve wijze de openingen zien naar een trendbreuk die wellicht a-typisch is in de maatschappelijke ontwikkeling, maar die wel consistent is met het streven van de IASB. Een citaat: ‘Th e Financial Accounting Standards Board (FASB) strongly and actively supports the internationaliza- tion of accounting standards. As part of its mission, the FASB seeks to promote the international comparability of accounting standards concur- rent with improving the quality of fi nancial reporting’. Een interessant, en ogenschijnlijk recent citaat, passend binnen het streven van onder meer de SEC om op termijn IFRS te accepteren als verslagge- vingstandaard voor ter (Amerikaanse) beurze geno- teerde ondernemingen. Toch is het citaat elf jaar oud, en afk omstig uit ‘Th e IASC-US Comparison Project (FASB, 1996), een eerste dappere poging van de FASB om overeenkomsten en verschillen tussen IAS/IFRS en US GAAP in kaart te brengen. Inderdaad, een dappere poging, want al op de eerste pagina’s wordt duidelijk gemaakt dat de meer principiële IAS/IFRS zich lastig laten vergelijken met het in meer gedetail- leerde regels gegoten US GAAP. Het tweede artikel is van de hand van Pheijff er, die als lid van de International Auditing and Assurance Standards Board vanuit de eerste hand commentaar geeft op de ontwikkelingen rondom controlestan- daarden. Ook hij constateert dat de ‘rules-based’ benadering terrein wint, niet in de laatste plaats door het zogenaamde ‘clarity project’. Pheijff er meent dat de controlestandaarden door dit project minder complex, minder uitgebreid en concreter worden, maar dat door de strakkere verwoording het gezag van de standaarden als meer dwingend zal worden ervaren. Ik deel zijn visie overigens maar ten dele. THEMA Roger Dassen Momenteel wordt aan de herziening hiervan gewerkt, en ik ben zeer geïnte- resseerd in de principiële insteek die hierbij wordt gekozen. Sommigen vrezen dat deze aanpassing uiteindelijk zal resulteren in het verlaten van de stewardship-functie van de jaarrekening ten faveure van de waarderelevantie. Ofwel, de reductie van de ‘value gap’ (tussen boekwaarde en marktwaarde/ beurswaarde) zou prevaleren boven het belang van de jaarrekening als objectief en toetsbaar verantwoor- dingsdocument. Wat mij betreft de verkeerde weg. De value gap is een gegeven, en deze moet niet worden bestreden door een steeds verder door- schieten in fair value accounting, maar door gebalan- ceerde guidance ten aanzien van de opname van niet- fi nanciële prestatie-indicatoren in de verslaggeving over de daadwerkelijke ‘value drivers’. Bij een derge- lijk model blijven objectiviteit en toetsbaarheid op een hoog niveau, en komt niettemin de informatie beschikbaar die beslissers nodig hebben om hun waarderingsmodellen te kunnen maken. Hommen neemt het laatste artikel voor zijn rekening, geschreven vanuit zijn perspectief van commissaris bij een aantal vooraanstaande (veelal beursgeno- teerde) ondernemingen. Hij deelt met ons zijn zorgen over mogelijke excessen in het toezicht op accoun- tants en de impact die dat zou kunnen hebben op wat hij noemt de ‘angstcultuur’. Hij roept op tot een herwaardering van het belang van de professionele oordeelsvorming door accountants, en pleit voor een driedeling in het toezicht op accountants, waarbij de zwaarte van het toezicht afh ankelijk is van de mate van ‘openbaar belang’. Een interessante gedachte, die bij de grote internationale accountantsorganisaties overigens zou leiden tot een redelijk ingrijpende split- sing tussen hun nationale en internationale prak- tijken, met mogelijk uiteenlopende kwaliteitspro- fi elen. Food for thought… Hommen is overigens positief over de introductie van IFRS in de Europese markt. Wel spreekt hij zich uit voor een ontwikkeling naar een meer principles-based benadering van IFRS, met Van der Tas plaatst ook hij vraagtekens bij het volgens sommigen ‘principled-based’ karakter van de lijvige IFRS bundel. Daarbij snijdt hij kort ook het thema aan van interpretaties van IFRS. Een tweede factor van eminent belang is de aan - wezigheid van een goed functionerende interpretatie- commissie die erin slaagt om interpretaties op een - duidige en navolgbare wijze te herleiden tot dat Framework, zodat de interpretatie het principe en het Framework bekrachtigt en niet onderuithaalt. Roger Dassen Terwijl onmiskenbaar de tekst er qua helderheid op vooruit is gegaan door de strikte toepassing van een aantal conventies, ben ik niet gerust op de splitsing van de huidige standaarden in ‘hoofdtekst’ en ‘application material’. Ik zie, overigens met Pheijff er, een risico dat het ‘application material’ door profes- sionals en door nationale standards setters als minder relevant en niet verplicht wordt ervaren, waardoor de robuustheid van het standaardenstelsel juist zou kunnen teruglopen ten opzichte van het huidige stelsel. Dat brengt ons bij de kern van dit themanummer. Is ‘rules-based’ versus ‘principles-based’ na de ingrij- pende veranderingen in zowel verslaggevings- als controlestandaarden nog altijd een herkenbaar ver - schilpunt tussen de toonaangevende standaarden? Waar liggen de fundamentele verschillen? Is het moge- lijk om op conceptuele gronden een voorkeur voor een van de benaderingen te hebben? In vier zeer uiteenlopende bijdragen zal getracht worden op deze en andere vragen een antwoord te geven. g g De spits wordt afgebeten door de bijdrage van Van der Tas. In deze uiterst lezenswaardige bijdrage gaat hij in op de ontwikkeling van het begrip ‘principles’ Prof . dr. R.J.M. Dassen is bestuursvoorzitter van Deloitte Nederland en tevens hoogleraar Auditing aan de Vrije Universiteit Amsterdam. MAB 584 MAB 584 d e c e m b e r 2 0 0 7 d e c e m b e r 2 0 0 7 Het derde artikel is van Mouthaan. Hij biedt een verfrissend perspectief op het vraagstuk regels versus principes binnen het corporate governance-domein, door een beschouwing te geven van spectaculaire ‘corporate failures’ uit de recente historie. Zijn con - clusie is dat deze niet het gevolg waren van tekort- komingen in de formele structuren, maar van falend toezicht en van menselijk tekortschieten. Gedeelde waarden en een bedrijfscultuur van onvoorwaarde- lijke integriteit zijn naar de mening van Mouthaan de belangrijkste waarborgen voor een goede governance, maar terecht stelt hij vast dat er bij toezichthouders, accountants en wetenschappers nog te weinig inzicht bestaat in de vraag hoe ethiek, integriteit en ‘normen en waarden’ objectief gemeten en gemonitord kunnen worden. de dynamiek van de tijd onderworpen is zoals de ge- schiedenis van ‘standardsetting’ ons leert (Ter Hoeven, 2005). Van eminent belang is, zoals ook Van der Tas al aangaf, in dit kader de beschikbaarheid van een zeer hoogwaardig ‘Framework’. Financial Accounting Standards Board, The IASC-US Comparison Project, Norwalk, 1996 Hoeven, R.L. ter (2005), Principle-based versus Rule-based Standards; waar ligt de balans?, Maandblad voor Accountan cy en Bedrijfsecono mie, jg. 79, no. 10 (oktober), pp. 475-485. Roger Dassen In prin- cipe heeft IASB het zo geregeld, maar daarmee is men helaas in de praktijk nog niet in de regel principieel! ■ d e c e m b e r 2 0 0 7 Literatuur p En daarmee toont Hommen bewust of onbewust het dilemma aan van dit thema, ‘rules’ versus ‘principles’. Want tegenover het risico van uiteenlopende inter- pretatie van IFRS als gevolg van het ontbreken van uitvoerige voorbeelden, uitzonderingen en detailin- terpretaties – en daarmee het risico van het ontstaan van IFRS-dialecten, staat het risico dat de interpreta- ties een zodanige autoriteit gaan krijgen dat deze gemakkelijk als nadere ‘rules’ kunnen worden opge- pakt. De IASB ontkomt er niet aan om als internatio- nale standsetter in dit krachtenveld naar een soort Pareto-optimum te zoeken. Daarbij geldt ook nog dat de balans tussen rules en principles nadrukkelijk aan MAB 585 d e c e m b e r 2 0 0 7 585
https://openalex.org/W2971515279	https://nottingham-repository.worktribe.com/preview/2447963/Cohort%20survey%20-%20support%20interest%20use%20and%20attitudes%20v3.1_accepted%20%28002%29.pdf	English	null	Interest in and Use of Smoking Cessation Support Across Pregnancy and Postpartum	Nicotine & tobacco research	2,019	cc-by	8,080	Title: Interest in and use of smoking cessation support across pregnancy and postpartum Title: Interest in and use of smoking cessation support across pregnancy and postpartum Title: Interest in and use of smoking cessation support across pregnancy and postpartum Felix Naughton PhD1, Luis Reeves Vaz PhD2, Tim Coleman MD2, Sophie Orton PhD2, Katharine Bowker PhD2, Jo Leonardi-Bee PhD3, Sue Cooper PhD2, Laura Vanderbloemen PhD4, Stephen Sutton PhD5, And Michael Ussher PhD6 1 School of Health Sciences, University of East Anglia, Norwich, NR4 7TJ, UK. 2 Division of Primary Care, U.K. Centre for Tobacco and Alcohol Studies and National Institute for Health Research School for Primary Care Research, University of Nottingham, Nottingham, UK Health Research School for Primary Care Research, University of Nottingham, Nottingham, UK 3 Division of Epidemiology and Public Health, University of Nottingham, Nottingham, UK 3 Division of Epidemiology and Public Health, University of Nottingham, Nottingham, UK 4 Department of Primary Care and Public Health, Imperial College London, London, UK 5 Behavioural Science Group, Institute of Public Health, University of Cambridge, CB2 0SR, UK 6 Population Health Research Institute, St Georges, University of London, London, UK and Institute for Social Marketing and Health, University of Stirling, Stirling, UK Corresponding author: Felix Naughton, School of Health Sciences, University of East Anglia, Norwich, NR4 7TJ, UK; +44 (0) 1603 593459, e-mail: f.naughton@uea.ac.uk Keywords: smoking cessation support, pregnancy, self-help, Word count = 4,105 1 1 ABSTRACT Background: Limited research exists on interest in and use of smoking cessation support in pregnancy and postpartum. Methods: A longitudinal cohort of pregnant smokers and recent ex-smokers were recruited in Nottinghamshire, UK (N=850). Data were collected at 8-26 weeks gestation, 34-36 weeks gestation and 3 months postpartum and used as three cross-sectional surveys. Interest and use of cessation support and belief and behaviour measures were collected at all waves. Key data were adjusted for non-response and analysed descriptively, and multiple regression used to identify associations. Results: In early and late pregnancy, 44% (95% CI 40-48%) and 43% (95% CI 37-49%) of smokers respectively were interested in cessation support, with 33% (95% CI 27-39%) interested postpartum. In early pregnancy, 43% of smokers reported discussing cessation with a midwife and in late pregnancy 27% did so. Over one-third (38%) did not report discussing quitting with a health professional during pregnancy. Twenty-seven percent of smokers reported using any NHS cessation support and 12% accessed NHS Stop Smoking Services during pregnancy. Lower quitting confidence (self-efficacy), higher confidence in stopping with support, higher quitting motivation and higher age were associated with higher interest in support (ps≤0.001). A recent quit attempt and greater interest in support was associated with speaking to a health professional about quitting and use of NHS cessation support (ps≤0.001). Conclusions: When asked in early or late pregnancy, about half of pregnant smokers were interested in cessation support, though most did not engage. Cessation support should be offered throughout pregnancy and after delivery. Implications: • There is relatively high interest in cessation support in early and late pregnancy and postpartum among smokers, however a much smaller proportion of pregnant or postpartum women access any cessation support, highlighting a gap between interest and engagement. 2 2 • Reflecting women’s interest, offers of cessation support should be provided throughout pregnancy and after delivery. • Reflecting women’s interest, offers of cessation support should be provided throughout pregnancy and after delivery. • Increasing motivation to quit and confidence in quitting with assistance may enhance interest in support and promoting the discussion of stopping smoking between women and health practitioners may contribute to higher support engagement rates. • Increasing motivation to quit and confidence in quitting with assistance may enhance interest in support and promoting the discussion of stopping smoking between women and health practitioners may contribute to higher support engagement rates. 3 3 INTRODUCTION Reducing smoking rates in pregnancy remains a global public health priority.1 While smoking in pregnancy rates have continued to reduce over the last decade,2 rates remain relatively high in many European and American nations. The United Kingdom is among the countries with the highest smoking in pregnancy rates.2 Based on routinely collected data at the time of delivery in England in 2018,3 recent reductions of smoking in pregnancy rates appears to have stalled, remaining at 11%. Increased efforts and new approaches are likely needed in order to reach the English national ambition of no more than 6% of women smoking in pregnancy by the end of 2022.4 One key approach to reducing smoking in pregnancy is the provision of smoking cessation support. Guidance from the UK National Institute for Health and Care Excellence (NICE) recommends that all pregnant women are carbon monoxide (CO) breath tested, with all those identified as smokers provided with risk information and referred to local National Health Service (NHS) Stop Smoking Services (SSS) for specialist behavioural support and Nicotine Replacement Therapy (NRT) where appropriate.5 This ‘opt-out’ referral pathway can increase both access rates to the English SSS, from 11% to 18% in one site,6 and abstinence rates.6,7 Use of other smoking cessation support among pregnant smokers is largely unknown. This includes discussions about stopping smoking with health professionals, the use of the NHS telephone helpline and use of NRT outside of the SSS other than in primary care, where NRT prescription rates around the time of pregnancy are estimated to be 11%.8 4 A likely important factor in accessing or accepting the offer of cessation support in pregnancy is an individual’s interest in the support being offered. There have been very few assessments of interest in cessation support among pregnant smokers and, when undertaken, this has been only for a limited selection of interventions. INTRODUCTION An English cross-sectional study conducted in 2004 found that 60% of pregnant smokers interviewed in early pregnancy indicated an interest in receiving help with stopping, with interest highest for in-person behavioural support followed closely by self-help materials.9 We are not aware of any longitudinal studies examining interest in cessation support over time during pregnancy, nor for a wider selection of intervention types, though findings have shown that quit 4 4 attempts continue throughout pregnancy and postpartum suggesting interest may be maintained over time.10 Findings from such work could help prioritise which type of support to offer and when. For example, in recent years, self-help has emerged as a promising and low cost approach to supporting cessation in pregnancy. Reviews have demonstrated that self-help11 and digital self-help12 cessation interventions are effective in pregnancy. Currently, though, we do not know how pregnant smokers view these types of interventions and how interested they are in using them. Little is also known about which characteristics of pregnant smokers are associated with interest in or the uptake of cessation support. Existing studies have shown that interest in cessation counselling among pregnant smokers is associated with being older, having a lower income, having a significant other who advises quitting and lower quitting confidence.13 Accepting a referral or accessing specialist cessation support is associated with having a mental health problem, when pregnant women with and without mental health problems were compared,14 and, among postpartum smokers, higher education.15 However, these cross-sectional studies have only investigated a relatively narrow range of potential predictors. Identifying predictors of interest and uptake of a variety of cessation support in this population will help guide and inform interventions that aim to increase support engagement. The primary aim of this study was to assess the interest in, use of and attitudes towards smoking cessation support during pregnancy and the immediate postpartum period among current or recent ex- smokers. Our secondary aims were to identify predictors of interest in and use of cessation support during pregnancy and to identify perceived barriers to using self-help cessation support over the pregnancy and postpartum period. Design Three cross-sectional surveys, taken from a longitudinal cohort of pregnant and postpartum women (the Pregnancy Lifestyle Survey), were used for this study.10,16 Data were collected at 8-26 weeks gestation (wave 1), 34-36 weeks gestation (wave 2) and 3 months after childbirth (wave 3). 5 5 Participants Women aged 16 years or above, between 8 and 26 weeks pregnant, and who self-reported being either current smokers (self-reported occasional smokers or daily smokers) or having smoked in the 3 months prior to becoming pregnant were eligible for participation. Women who were unable to understand study procedures sufficiently to provide consent or were unable to read or understand the written questionnaires in English were excluded. Procedure Recruitment to the Pregnancy Lifestyle Survey took place between August 2011 and August 2012 at two antenatal clinics within Nottingham University Hospitals NHS Trust (City Hospital and Queen’s Medical Centre). To ensure representative sampling, researchers attended on average five different clinic sessions per week. All women self-reporting to be between 8 and 26 weeks gestation attending routine antenatal appointments at these clinics were invited by a researcher or a member of clinic staff to complete an anonymous screening questionnaire to determine study eligibility. Those who met the criteria were directed to read a participant information sheet describing the study, and, if willing, to then complete a baseline questionnaire; women could also seek further information from the researcher/staff member. On completion of the baseline questionnaire, women were offered a £5 high street shopping voucher as recognition for the time taken to complete the questionnaire. Written informed consent was obtained from those who wished to complete the two follow-up questionnaires, who made up the sample of the study. All participants were posted a second questionnaire at 34-36 weeks gestation. In addition, participants who provided an email address were emailed a link to a web-based version of the questionnaire, and sent one email reminder. Non-respondents were sent one postal/email reminder letter and then contacted by telephone; if no response was received, participants were texted a reminder. Participants who were successfully contacted via telephone were invited to complete the questionnaire during the call. 6 Participants were sent the final questionnaire 3 months after their delivery using the same method as described above for follow-up in later pregnancy. All participants who completed follow-up questionnaires were sent a £5 shopping voucher for each follow-up. Participants were sent the final questionnaire 3 months after their delivery using the same method as described above for follow-up in later pregnancy. All participants who completed follow-up questionnaires were sent a £5 shopping voucher for each follow-up. A more detailed description of the procedure for enrolment and data collection and sample characteristics is provided elsewhere.16 The study was approved by Derbyshire Research Ethics Proportionate Review Sub-Committee (11/EM/0078). Measures Background and health predictors included gestation, general health (general health rating from excellent-poor, and whether they had a longstanding physical or mental illness or disability),19 depression (during the past month bothered by feeling down, depressed or hopeless, or having little interest or pleasure in doing things),20 the Perceived Stress Scale 4 (PSS-4),21 ethnicity, age and index of multiple deprivation.22 Smoking-related predictors included general smoking behaviour,23,24 urges to smoke,25 partner’s smoking status, nicotine dependence using the ‘Heaviness of Smoking Index’,26 intentions and determination to quit smoking and confidence (self-efficacy) in achieving this,27 support for stopping smoking from friends/family, beliefs about the harm of smoking during pregnancy27 and whether participant had talked to a health professional (midwife, nurse or GP) about quitting. Smoking-related norm predictors included injunctive norm (‘people important to me think I should stop smoking’) and descriptive norm (knowing others who smoked throughout pregnancy),28. A broad range of potential demographic and psychosocial predictors were included based on previous evidence in predicting interest or uptake in cessation support13-15 and cessation,17,18 although as relatively little is known about this, we had no prior hypotheses regarding potential associations. Background and health predictors included gestation, general health (general health rating from excellent-poor, and whether they had a longstanding physical or mental illness or disability),19 depression (during the past month bothered by feeling down, depressed or hopeless, or having little interest or pleasure in doing things),20 the Perceived Stress Scale 4 (PSS-4),21 ethnicity, age and index of multiple deprivation.22 Smoking-related predictors included general smoking behaviour,23,24 urges to smoke,25 partner’s smoking status, nicotine dependence using the ‘Heaviness of Smoking Index’,26 intentions and determination to quit smoking and confidence (self-efficacy) in achieving this,27 support for stopping smoking from friends/family, beliefs about the harm of smoking during pregnancy27 and whether participant had talked to a health professional (midwife, nurse or GP) about quitting. Smoking-related norm predictors included injunctive norm (‘people important to me think I should stop smoking’) and descriptive norm (knowing others who smoked throughout pregnancy),28. Sample size and analysis Sample size and analysis The sample size calculation for the cohort survey was based on estimating the number of quit attempts initiated during pregnancy among smokers and is reported elsewhere.16 The sample size calculation for the cohort survey was based on estimating the number of quit attempts initiated during pregnancy among smokers and is reported elsewhere.16 Measures Copies of the questionnaires used at each wave can be found in a separate publication.10 The questions used in the current study are described below and used a range of response formats including yes/no responses, multiple choices and five-point Likert-type scales for attitudinal items. Interest, use and attitudes towards smoking cessation support Interest, use and attitudes towards smoking cessation support At all three waves, participants were asked to rate their interest in receiving help with stopping smoking in general and their interest in, difficulty in using and perceived usefulness of nine different types of smoking cessation support using a 1-5 scale (‘not at all’ to ‘extremely’). These were split into health practitioner-orientated support (telephone helpline, group sessions and one-to-one sessions) and self-help support (booklet, DVD, website, text messages, email and smartphone/digital device application). Participants were also asked to indicate whether they had accessed any of the cessation support offered by the NHS, though not necessarily delivered by the NHS, since finding out they were pregnant (wave 1) or since completing the last survey (waves 2 and 3). These included talking to a GP/nurse or midwife about stopping smoking, attending an NHS Stop Smoking Service group or individual session, calling a stop smoking helpline or using NRT (from any source). At baseline only, participants were asked about their access to electronic/digital devices for using self-help. In all three waves participants were asked to indicate from a list of statements, informed by prior work,13 any potential barriers towards accessing or using self-help cessation support. 7 7 Pre-determined predictor variables of interest and use of cessation support A broad range of potential demographic and psychosocial predictors were included based on previous evidence in predicting interest or uptake in cessation support13-15 and cessation,17,18 although as relatively little is known about this, we had no prior hypotheses regarding potential associations. Analysis Analysis 8 Descriptive statistics were used to report interest in and use of smoking cessation services as well as attitudes to smoking cessation support and barriers to self-help use for all three waves. For the prevalence estimates of interest in and attitudes towards smoking cessation support among smokers, responses were dichotomised by grouping the five possible responses: not at all/a little (no interest) versus moderately/very much/extremely (interest), with the same thresholds for determining low vs. high perceived usefulness and difficulty accessing/using. Multiple imputation using chained equations 8 was performed using 20 iterations29 for the percentage who were interested in cessation support at late pregnancy and postpartum. Missing data on interest in and attitudes towards smoking cessation support was imputed using a logit imputation method based on the following baseline factors: maternal age, ethnicity, highest educational qualification, gestation of pregnancy, smoking status, urge to smoke, general health status, depression status, parity, and smoking status in previous pregnancy. For the exploratory analysis investigating potential correlates of general interest in cessation support at baseline, discussion of stopping smoking with a health professional at baseline and use of NHS cessation support (telephone helpline, group sessions, one-to-one sessions or NRT) in late pregnancy, we used linear regression and logistic regression respectively to test the univariable associations with baseline characteristics. Interest was measured on a 1-5 scale and use was coded as a binary “use” or “non-use” variable. Following this, all predictors that were associated with interest in or use of cessation support in the univariable analyses at p<0.1 were included in multivariable models,30 providing they were not found to be collinear (variance inflation factor ≥10). Missing data for the predictor variables (i.e. not having answered survey questions at either baseline or late pregnancy follow-up), were addressed through including missing data as either an additional category for categorical variables or using a dummy variable to indicate missingness for continuous variables. As two out of three of the predictor analyses were cross-sectional and they were part of a secondary aim of the study, we did not perform multiple imputation for these. RESULTS 9 At the late pregnancy and postpartum follow-ups, 509 (60%) and 476 (56%) of baseline participants completed a questionnaire respectively. Table 1 presents the baseline sample characteristics (N=850), full details of which are reported elsewhere.16 On average, participants were 26 years old and 16 weeks gestation at baseline. One-third reported that the current pregnancy was their first and, among those who had been pregnant before, around two thirds smoked during their last pregnancy. Fifty- seven percent reported being current smokers and, among these, over half had a partner who smoked. 9 9 Almost all respondents owned a mobile phone (97%) and seventy-one percent owned a smartphone. Almost all respondents owned a mobile phone (97%) and seventy-one percent owned a smartphone. Interest and use of cessation support and perceived barriers to accessing self-help At baseline, during early pregnancy, 44% (95% CI 40-48%) of current smokers (table 2) and 9% (95% CI 6-12%) of recent ex-smokers (not shown in table) reported being interested in receiving help with stopping smoking. Among smokers, the specific support types rated of most interest were one-to- one sessions (42%), followed by self-help booklets (39%). Imputed survey results showed interest in support to stop smoking, among smokers, changed little in late pregnancy (43%; 95% CI 37-49%) relative to early pregnancy but dropped to 33% (95% CI 27-39%) three months postpartum. Self-help booklets were of most interest (49%) followed by self-help websites (45%) in late pregnancy. Across all time points, one-to-one sessions had the highest perceived usefulness and self-help support had the lowest levels of perceived difficulty to access and use. In early pregnancy, less than half of smokers (43%) reported having talked to a midwife about stopping smoking and fewer had spoken to a GP or nurse (27%) about this (table 3). Between early and late pregnancy 27% of smokers reported speaking to a midwife about stopping. Across the whole of pregnancy, over one-third (38%) of smokers did not report having talked with either a midwife, GP or nurse about stopping smoking (not shown in table). Out of those participants who smoked at all during pregnancy, 12% reported accessing Stop Smoking Service support (group or one-to-one support) at some point in their pregnancy (not shown in table). Across the whole of pregnancy 17% of participants reported accessing at least one form of NHS provided cessation support, which rose to 27% when restricted to those reporting smoking at both baseline and end of pregnancy (not shown in table). The most common type of support used was NRT without additional behavioural support (table 3). 10 In early pregnancy, the most common barriers for using self-help cessation support was preferring to receive support from a health professional (19% smokers, 6% recent ex-smokers) and thinking self- help would not be much help with quitting (14% smokers, 7% recent ex-smokers). These remained 10 the two most common barriers at the two follow up time points. DISCUSSION This study found a substantial minority of pregnant smokers are interested in getting smoking cessation support and that this level of interest is as high at the end of pregnancy and drops only a modest amount by three months postpartum. While there is relatively high interest in cessation support, a much smaller proportion of women accessed any cessation support in pregnancy, This study found a substantial minority of pregnant smokers are interested in getting smoking cessation support and that this level of interest is as high at the end of pregnancy and drops only a modest amount by three months postpartum. While there is relatively high interest in cessation support, a much smaller proportion of women accessed any cessation support in pregnancy, highlighting a gap between interest and engagement. Our study indicates that speaking to a healthcare professional about stopping smoking, being motivated to stop and having low confidence in doing so without assistance could influence interest in NHS provided cessation support. Interest in support, in turn, prospectively predicted use of NHS cessation support. However, women who had not tried to quit early on in their pregnancy were less likely to have used support later on in pregnancy compared with those who had tried to quit, independent of quitting motivation and interest in support. This suggests that the process of trying and failing to quit may be helpful for some by prompting them to seek assistance. highlighting a gap between interest and engagement. Our study indicates that speaking to a healthcare professional about stopping smoking, being motivated to stop and having low confidence in doing so without assistance could influence interest in NHS provided cessation support. Interest in support, in turn, prospectively predicted use of NHS cessation support. However, women who had not tried to quit early on in their pregnancy were less likely to have used support later on in pregnancy compared with those who had tried to quit, independent of quitting motivation and interest in support. This suggests that the process of trying and failing to quit may be helpful for some by prompting them to seek assistance. the two most common barriers at the two follow up time points. Factors associated with interest in cessation support among smokers in early pregnancy Among smokers, 11 out of 23 pre-specified variables were univariable correlates of interest in receiving cessation support in general at baseline (p<0.05) (table 4). When these were entered into a multivariable model, the following five predictors remained statistically significant: higher determination to quit (B=0.42, 95% CI 0.31-0.54), lower confidence in stopping until the end of pregnancy (B=-0.23, 95% CI −0.35-−0.12), higher confidence in stopping with health professional support (B=0.22, 95% CI 0.11-0.33), having spoken to a health professional about stopping (B=0.77, 95% CI 0.55-0.98) and higher age (B=0.03, 95% CI 0.01-0.05). Factors associated with discussing stopping smoking with a health professional among smokers in early pregnancy Ten out of 23 pre-specified baseline variables were univariable correlates (p<0.05) of whether or not smokers reported having had a discussion with a health professional (midwife, nurse or GP) about quitting smoking in early pregnancy (supplementary table S1). Of these, two remained statistically significant in the multivariable analysis: having previously tried to quit during the current pregnancy (odds ratio [OR] 3.0, 95% CI 2.0-4.6) and interest in support (OR 5.4, 95% CI 3.6-8.2). Predictors of use of NHS cessation support among smokers by late pregnancy Nine baseline variables were associated with use of NHS provided cessation support by late pregnancy (p<0.05) among baseline smokers (table 4). Three variables remained statistically significant in the multivariable model: having previously tried to quit during the current pregnancy (OR 2.7, 95% CI 1.3-5.5), older age (OR 1.1, 95% CI 1.0-1.1) and interest in support (OR 1.8 95% CI 11 1.4-2.3). Strengths and limitations We believe this was the first longitudinal cohort study looking at smoking cessation support interest and use over pregnancy and into the post-partum period. We examined women’s views and use of a broad range of support types, across the three time-points in a large cohort. Although participants were recruited from one area, the sample as a whole demonstrated similar characteristics to pregnant smokers in national cohort studies16 and so key findings may be generalisable to pregnant smokers across the UK. Furthermore, very few prior studies have examined a wide range of potential predictors of interest and use of cessation support. Common to many cohort studies, we experienced moderate attrition at follow-ups. Although not excessive and we minimised the impact of this by performing multiple imputation for key variables, 12 12 this is a limitation. A further limitation is that we did not collect data during the second trimester of participant’s pregnancies, and so there may have been fluctuations in variables of interest that we were unable to capture. Also, we did not collect data on views and use of e-cigarettes in this survey, as when the survey was carried out e-cigarette usage was still relatively low. We have now undertaken research to explore this in pregnancy in several separate recent studies.31,32 Similarly, the use of smartphone apps were somewhat lower when the study was undertaken compared to current usage, and interest in this may have grown in recent years. While our analysis investigating predictors of use of support was prospective, the correlates of interest in cessation support and discussion about stopping smoking analyses were cross-sectional and so we are limited in our ability to indicate possible directions of causality. Findings in context Our finding that just under half of pregnant smokers were interested in receiving cessation support in early pregnancy corresponds closely with Ussher et al9 who found in 2004 that 60% of women were interested in cessation support in early pregnancy. As with this prior study, our data also showed similar interest levels between self-help and one-to-one cessation support, with low interest in group support. However, in our study the point estimates for difficulty accessing one-to-one support were higher across all waves than the self-help options, suggesting perceived access could inhibit uptake, particularly in late pregnancy. No studies have previously explored the relative interest in different types of self-help among pregnant women. Booklets were of most interest across all waves and with mostly the highest perceived usefulness compared to other estimates for self-help types. Though there were only minor differences between the self-help types, email appeared to be of least interest and perceived usefulness. Despite relatively high rates of interest in one-to-one cessation support among baseline smokers (42%), only a small proportion (12%) of our cohort who smoked at all during pregnancy reported using the NHS Stop Smoking Services. This very closely matches the 11% access rate identified from NHS records in another study in the same region, prior to implementation of an opt-out pathway.6 13 Low rates of access to specialist cessation support among pregnant smokers has been reported historically33,34 and can be increased using an opt-out referral pathway,6,7 which also increases abstinence. The most common type of cessation support used by women in our cohort that the NHS provides was NRT. This is despite safety concerns and reluctance to recommend NRT among many practitioners.35,36 However, use of NRT without guidance or behavioural support is unlikely to be effective in pregnancy as even with guidance and support it is only of borderline efficacy37 and use of over-the-counter NRT is not associated with abstinence in the general population.38 Sub-optimal adherence39,40 and increased nicotine metabolism41 indicate higher doses of NRT combined with adherence support is likely required to make NRT more effective in pregnancy. In line with a further study,13 we found higher age and lower quitting self-efficacy associated with interest in receiving support. We also found a positive association between confidence in quitting with a health professional’s help and support interest, which had not been examined before. Findings in context Prior international research has found that while most practitioners working with pregnant women ask about smoking status,42,43 far fewer discuss cessation and only around one-third discuss treatment options.42-44 Our findings broadly support this from the UK pregnant women’s perspective – fewer than two-thirds of smokers in our sample reported discussing stopping smoking with a practitioner. UK research undertaken over the last two decades report similar findings suggesting that little has changed in this regard.13,45 Implications for practice One key implication is that, based on findings related to interest levels, the offer of cessation support should be provided throughout pregnancy and postpartum. In the UK at least, currently policy relating to the offer of cessation support is weighted towards early pregnancy, even if strictly following NICE guidance.5 This is reflected by our findings, where we found far fewer smokers reporting a discussion with a health professional about stopping smoking in mid to late pregnancy compared with early pregnancy. This matches other research showing less than one quarter of health professionals follow 14 up women after an initial discussion about smoking,14,44 which can inadvertently reassure women that quitting smoking may not be a priority.46 To promote interest in cessation support, which was predictive of accessing support, our findings suggest that increasing motivation to stop smoking, ensuring health professionals discuss stopping smoking and enhancing women’s confidence that their chances of stopping smoking are higher if they receive professional support are important targets for interventions. Practitioners often do not initiate discussions about quitting smoking with pregnant clients because of low levels of confidence, feelings of being under-skilled and concern about damaging the client relationship.35 The use of carbon monoxide monitoring may help with facilitating a discussion about stopping smoking, from both the perspective of practitioners35,47 and pregnant women,48 and enhance motivation to quit. When combined with an opt-out referral it can increase the uptake of support.6,7 Training health practitioners can also increase the assessment of smoking and support provision49 and the use of prompts in maternity paperwork and electronic systems may also facilitate appropriate discussions.35 In terms of factors influencing uptake of support, being younger was associated with a reduced likelihood of using practitioner orientated cessation support. This suggests a potential targeting opportunity. Our findings also reinforce previous research9 by highlighting the popularity of self-help cessation support. Current UK guidance only recommends promoting self-help to those who accept the offer of formal cessation support but who are struggling to engage.5 This may be a missed opportunity, given recent evidence of effectiveness of digital self-help approaches for pregnant smokers,12 particularly when delivered by SMS text message.27,50 One less prominent but nonetheless important finding was that some of the most predictive factors in determining smoking and a failure to quit smoking in pregnancy i.e. DECLARATION OF INTERESTS None. Implications for practice socioeconomic status/deprivation, nicotine dependence and having a partner that smokes,17 did not predict interest in or use of support in our cohort. From a public health perspective, this is positive and suggests these factors may not inhibit the seeking of and acceptance of cessation support. 15 FUNDING This article presents independent research funded by the National Institute for Health Research (NIHR) under the Programme Grants for Applied Research programme RP-PG-0109-10020. The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. Conclusion Almost half of pregnant smokers in our cohort were interested in help to stop smoking and this changed little from early to late pregnancy. Yet rates of discussing stopping smoking with a health professional reduced after early pregnancy and a substantial minority of pregnant smokers did not report having a discussion with a health professional about stopping at any point in pregnancy. With one-quarter of smokers accessing any NHS provided support and less than half of these accessing stop smoking services, the gap between support interest and access indicates a missed opportunity. Our findings indicate that increasing motivation to quit to enhance interest in support, the discussion of stopping smoking with health practitioners and confidence in quitting with cessation support may contribute to higher support use rates. In addition, non-routine forms of cessation support, including self-help, should be promoted given evidence of effectiveness, low cost and their popularity among this population. ACKNOWLEDGEMENTS The authors would like to thank Nottingham University Hospital NHS Trust for facilitating this research. We also thank Rachel Whitemore for invaluable assistance in setting up this research. Tim Coleman acknowledges the support of the East Midlands Collaboration for Leadership in Applied Health Research and Care (CLARHC). Professor Coleman is an NIHR Senior Investigator. 16 Table 1. Characteristics of the cohort (N=850) Characteristic Mean (SD) Gestation (weeks) (n=806) 15.6 (4.1) Age (n=847) 25.8 (5.6) Perceived stress scale (PSS-4) (n=819) 10.6 (3.5)* n (%) Qualifications: GCSEs or similar* 594/850 (69.9) Home ownership 166/846 (19.6) Cars or vans available for use in household 446/839 (53.2) In paid work 383/850 (45.1) Ethnicity: Caucasian 783/844 (92.8) Baseline smoker 488/850 (57.4) First pregnancy 275/839 (32.8) If pregnant before, smoked during last pregnancy 368/561 (65.6) Owns a mobile phone 776/797 (97.4) Owns a smartphone 547/774 (70.7) Mobile phone bundle includes free text messages 610/778 (78.4) * Maximum score on PSS-4 is 16 Numbers may not add up to 850 (total sample size) due to missing data * GCSEs (General Certificate of Secondary Education) are a qualification/exams taken by school students in the UK (except Scotland) usually when aged between 14-16 years old Table 1. Characteristics of the cohort (N=850) * Maximum score on PSS-4 is 16 Numbers may not add up to 850 (total sample size) due to missing data Numbers may not add up to 850 (total sample size) due to missing data * GCSEs (General Certificate of Secondary Education) are a qualification/exams taken by school students in the UK (except Scotland) usually when aged between 14-16 years old 17 17 Table 2. Interest in and attitudes towards smoking cessation support among smokers (n=488) Responses were dichotomised by grouping the 5 possible responses: not at all/a little (no interest) versus moderately/very much/extremely (interest). The same approach was used for perceived usefulness (low vs. Responses were dichotomised by grouping the 5 possible responses: not at all/a little (no interest) versus moderately/very much/extremely (interest). The same approach was used for perceived usefulness (low vs. high perceived usefulness) and difficulty accessing/using (low vs/ high difficulty) Missing data were generated using multiple imputation by chained equations for dichotomised interest in cessation support amongst smokers in early and late pregnancy and in postpartum period ACKNOWLEDGEMENTS high perceived usefulness) and difficulty accessing/using (low vs/ high difficulty) Missing data were generated using multiple imputation by chained equations for dichotomised interest in cessation support amongst smokers in early and late pregnancy and in postpartum period Type of support* Early pregnancy (8- 26 weeks gestation) Late pregnancy (34-36 weeks gestation) Postpartum (3 months) n (%, 95% CI) % (95% CI) % (95% CI) Interest in receiving help with stopping 212 (44.0, 39.5-48.4) 42.8 (37.0-48.6) 32.9 (26.9-38.8) Health professional telephone helpline Interest 92 (20.5) 25.2 21.8 Perceived usefulness 112 (24.7) 24.6 25.5 Difficulty accessing/using 153 (34.3) 30.9 31.9 Health professional group sessions Interest 62 (14.1) 14.9 12.2 Perceived usefulness 128 (28.3) 28.0 24.9 Difficulty accessing/using 179 (40.3) 39.5 43.0 Health professional one-to-one Interest 190 (41.7) 43.9 38.1 Perceived usefulness 225 (49.2) 49.8 48.3 Difficulty accessing/using 155 (34.9) 33.0 34.0 Self-help booklet Interest 177 (39.3) 48.9 39.3 Perceived usefulness 182 (40.1) 46.0 36.6 Difficulty accessing/using 135 (30.1) 21.5 20.3 Self-help DVD Interest 154 (34.8) 37.9 29.0 Perceived usefulness 163 (36.1) 39.8 29.8 Difficulty accessing/using 130 (29.3) 20.5 24.8 Self-help website Interest 148 (33.9) 44.7 37.0 Perceived usefulness 160 (35.5) 42.2 32.1 Difficulty accessing/using 132 (30.1) 24.2 27.7 Self-help text messages Interest 133 (30.5) 40.5 30.4 Perceived usefulness 138 (30.5) 40.2 30.3 Difficulty accessing/using 128 (28.8) 23.1 22.6 Self-help emails Interest 112 (25.7) 32.8 24.2 Perceived usefulness 124 (27.5) 32.2 24.7 Difficulty accessing/using 138 (31.0) 25.8 24.3 Self-help phone app Interest 146 (33.2) 40.2 36.3 Perceived usefulness 159 (35.3) 43.3 37.0 Difficulty accessing/using 136 (30.7) 25.4 22.7 Table 2. Interest in and attitudes towards smoking cessation support among smokers (n=488 Responses were dichotomised by grouping the 5 possible responses: not at all/a little (no interest) versus moderately/very much/extremely (interest). The same approach was used for perceived usefulness (low vs. high perceived usefulness) and difficulty accessing/using (low vs/ high difficulty) Missing data were generated using multiple imputation by chained equations for dichotomised interest in cessation support amongst smokers in early and late pregnancy and in postpartum period 18 Table 3. Reported use of NHS provided smoking cessation support split by smoking status Non-cumulative support usage. Rates in each time period do not include usage in earlier time period and represent what was reported in each questionnaire. ** Period between the early pregnancy questionnaire and the late questionnaire. ACKNOWLEDGEMENTS Includes 10 women who had given birth prior to completing the 34-week questionnaire Type of support* Early pregnancy Early-late pregnancy Postpartum Recent ex- smoker (n=362) Smoker (n=488) Recent ex- smoker (n=265)* Smoker (n=256)*** Recent ex- smoker (n=196) Smoker (n=280) n (%) n (%) n (%) n (%) n (%) n (%) Talked to GP/Nurse about giving up 22 (6.1) 132 (27.0) 3 (1.1) 28 (10.9) 3 (1.5) 36 (12.9) Talked to midwife about giving up 36 (9.9) 211 (43.2) 16 (6.0) 68 (26.6) 3 (1.5) 24 (8.6) Accessed/used any NHS smoking cessation support 29 (8.0) 90 (18.4) 21 (7.9) 66 (25.8) 6 (3.1) 38 (13.6) Attended NHS stop smoking service group session 3 (0.8) 21 (4.3) 1 (0.4) 10 (3.9) 0 (0.0) 5 (1.8) Attended solo/individual NHS stop smoking service session 10 (2.8) 25 (5.1) 15 (5.7) 23 (9.0) 1 (0.5) 7 (2.5) Called a stop smoking telephone helpline 5 (1.4) 18 (3.7) 3 (1.1) 7 (2.7) 0 (0.0) 5 (1.8) Used NRT 25 (6.9) 64 (13.1) 16 (6.0) 57 (22.3) 6 (3.1) 31 (11.1) Reported use of NHS provided smoking cessation support split by smoking status 19 19 Table 4: Univariable and multivariable models for correlates/predictors among smokers of a) general interest in cessation support at baseline and b) use of NHS cessation support in late pregnancy Table 4: Univariable and multivariable models for correlates/predictors among smokers of a) general interest in cessation support at baseline and b) use of NHS cessation support in late pregnancy Baseline variables DV: Interest in cessation support at baseline (n=482) DV: Use of cessation support in late pregnancy (n=263) Univariable model Multivariable model Univariable model Multivariable model Beta p 95% CI Beta p 95% CI OR p 95% CI OR p 95% CI Smoked in a prior pregnancy -0.126 0.336 -0.38,0.13 0.89 0.690 0.51,1.56 Heaviness of Smoking Index 0.039 0.396 -0.05,0.13 1.08 0.486 0.87,1.34 Urges to smoke 0.009 0.791 -0.06,0.07 1.00 0.972 0.88,1.15 Has tried to quit (y/n) 0.716 <0.001 0.47,0.96 3.03 0.001 1.60,5.73 2.70 0.006 1.34,5.46 Number of 24 hour quit attempts -0.022 0.036 -0.04,0.00 0.97 0.465 0.89,1.06 Seriously planning to quit 0.491 <0.001 0.40,0.58 1.65 <0.001 1.30,2.08 Determination to stop till baby born 0.507 <0.001 0.42,0.59 0.423 <0.001 0.31,0.54 1.71 <0.001 1.32,2.20 Confidence to stop till baby born 0.217 <0.001 0.12,0.31 -0.231 <0.001 -0.35,-0.12 1.24 0.049 1.00,1.54 Confidence to stop alone 0.026 0.596 -0.07,0.12 0.89 0.339 0.71,1.13 Confidence to stop with health professional help 0.425 <0.001 0.33,0.52 0.216 <0.001 0.11,0.33 1.36 0.008 1.08,1.70 Beliefs on harms of smoking to baby 0.774 <0.001 0.43,1.11 3.05 0.026 1.14,8.11 Knows others who smoked in pregnancy -0.035 0.863 -0.43,0.36 1.40 0.484 0.54,3.61 People important to me think I should stop 0.605 <0.001 0.33,0.88 1.69 0.132 0.85,3.33 Has support to stop 0.304 0.025 0.04,0.57 1.73 0.104 0.89,3.36 Partner smokes 0.041 0.760 -0.22,0.30 0.90 0.729 0.50,1.62 Spoken to health professional about stopping 1.212 <0.001 0.99,1.43 0.765 <0.001 0.55,0.98 3.29 <0.001 1.80,6.01 General health -0.206 0.244 -0.55,0.14 1.67 0.283 0.66,4.24 Depression 0.169 0.276 -0.14,0.47 0.94 0.877 0.46,1.95 Perceived Stress Scale 0.008 0.706 -0.04,0.05 0.99 0.871 0.90,1.10 Gestation (weeks) -0.022 0.152 -0.05,0.01 0.97 0.386 0.91,1.04 Ethnic minority -0.143 0.609 -0.69,0.41 0.76 0.640 0.24,2.39 Age 0.032 0.006 0.01,0.05 0.030 0.001 0.01,0.05 1.09 0.001 1.03,1.14 1.08 0.006 1.02,1.15 Deprivation (Index of Multiple Deprivation; IMD) -0.004 0.274 -0.01,0.00 1.00 0.638 0.98,1.01 Interest in support at baseline - - - - - - 1.91 <0.001 1.53,2.37 1.80 <0.001 1.43,2.27 20 References 1. Ussher M, Naughton F, Notley C, Bauld L. Smoking and Pregnancy: Risk Factors, Women’s Experiences, and Interventions. In: Ussher J, Chrisler J, Perz J, eds. Routledge International Handbook of Women’s Sexual and Reproductive Health. London: Routledge; in press. 2. Lange S, Probst C, Rehm J, Popova S. National, regional, and global prevalence of smoking during pregnancy in the general population: a systematic review and meta-analysis. Lancet Glob Health. 2018;6(7):e769-e776. doi: 10.1016/S2214-109X(18)30223-7. 3. Health and Social Care Information Centre. Statistics on Women’s Smoking Status at Time of Delivery, England - 2017-18. 2018; http://content.digital.nhs.uk/catalogue/PUB20899/stat- wome-smok-time-deli-eng-q4-15-16-rep.pdf. Accessed 05/26, 2018. 4. Department of Health and Social Care. Tobacco Control Delivery Plan 2017 - 2022. London: Department of Health & Social Care; 2018. 5. National Institute for Health and Care Excellence. Quitting smoking in pregnancy and following childbirth. NICE public health guidance 26;2010. 6. Campbell KA, Cooper S, Fahy SJ, et al. 'Opt-out' referrals after identifying pregnant smokers using exhaled air carbon monoxide: impact on engagement with smoking cessation support. Tob Control. 2017;26(3):300-306. doi: 10.1136/tobaccocontrol-2015-052662. 7. Bell R, Glinianaia SV, Waal ZV, et al. Evaluation of a complex healthcare intervention to increase smoking cessation in pregnant women: interrupted time series analysis with economic evaluation. Tob Control. 2018;27(1):90-98. doi: 10.1136/tobaccocontrol-2016- 053476. 8. Dhalwani NN, Szatkowski L, Coleman T, Fiaschi L, Tata LJ. Prescribing of nicotine replacement therapy in and around pregnancy: a population-based study using primary care data. Br J Gen Pract. 2014;64(626):e554-560. doi: 10.3399/bjgp14X681361. 8. Dhalwani NN, Szatkowski L, Coleman T, Fiaschi L, Tata LJ. Prescribing of nicotine replacement therapy in and around pregnancy: a population-based study using primary care data. Br J Gen Pract. 2014;64(626):e554-560. doi: 10.3399/bjgp14X681361. 21 21 9. Ussher M, West R, Hibbs N. A survey of pregnant smokers' interest in different types of smoking cessation support. PatientEducCouns. 2004;54(1):67-72. doi: 10.1016/S0738- 3991(03)00197-6. 10. Cooper S, Orton S, Leonardi-Bee J, et al. Smoking and quit attempts during pregnancy and postpartum: a longitudinal UK cohort. BMJ Open. 2017;7(11):e018746. doi: 10.1136/bmjopen-2017-018746. 11. Naughton F, Prevost AT, Sutton S. Self-help smoking cessation interventions in pregnancy: a systematic review and meta-analysis. Addiction. 2008;103(4):566-579. doi: 10.1111/j.1360- 0443.2008.02140.x. 12. Griffiths SE, Parsons J, Naughton F, Fulton EA, Tombor I, Brown KE. Are digital interventions for smoking cessation in pregnancy effective? A systematic review and meta-analysis. Health Psychol Rev. 2018;12(4):333-356. doi: 10.1080/17437199.2018.1488602. 13. Ussher M, Etter JF, West R. References Perceived barriers to and benefits of attending a stop smoking course during pregnancy. Patient Educ Couns. 2006;61(3):467-472. doi: 10.1016/j.pec.2005.06.021. 14. Howard LM, Bekele D, Rowe M, Demilew J, Bewley S, Marteau TM. Smoking cessation in pregnant women with mental disorders: a cohort and nested qualitative study. BJOG. 2013;120(3):362-370. doi: 10.1111/1471-0528.12059. 15. Roske K, Hannover W, Kelbsch J, Thyrian JR, John U, Hapke U. The readiness of women, after they have given birth to children, to participate in individualized counselling for smoking cessation. Gesundheitswesen. 2004;66(10):697-702. doi: 10.1055/s-2004-813621. 16. Orton S, Bowker K, Cooper S, et al. Longitudinal cohort survey of women's smoking behaviour and attitudes in pregnancy: study methods and baseline data. BMJ Open. 2014;4(5):e004915. doi: 10.1136/bmjopen-2014-004915. 22 17. Riaz M, Lewis S, Naughton F, Ussher M. Predictors of smoking cessation during pregnancy: a systematic review and meta-analysis. Addiction. 2018;113(4):610-622. doi: 10.1111/add.14135. 18. Orton S, Coleman T, Coleman-Haynes T, Ussher M. Predictors of Postpartum Return to Smoking: A Systematic Review. Nicotine Tob Res. 2018;20(6):665-673. doi: 10.1093/ntr/ntx163. 19. National Centre for Longitudinal Studies. Millennium Cohort Study, First Survey: CAPI Questionnaire Documentation. London 2003. 20. National Institute for Health and Care Excellence. Antenatal and postnatal mental health; Clinical management and service guidance. 2007. 21. Cohen S, Williamson GM. Perceived stress in a probability sample of the United States. Soc Psychol Health. 1988;1:31-67. 22. Noble M, McLennan D, Wilkinson K, et al. The English Indices of Deprivation 2007. Whetherby 2007. 23. Coleman T, Thornton J, Britton J, et al. Protocol for the smoking, nicotine and pregnancy (SNAP) trial: double-blind, placebo-randomised, controlled trial of nicotine replacement therapy in pregnancy. BMC Health Serv Res. 2007;7:2. doi: 10.1186/1472-6963-7-2. 24. Mullen PD, Carbonari JP, Tabak ER, Glenday MC. Improving disclosure of smoking by pregnant women. Am J Obstet Gynecol. 1991;165(2):409-413. doi: 10.1016/0002- 9378(91)90105-z. 25. West R, Hajek P. Evaluation of the mood and physical symptoms scale (MPSS) to assess cigarette withdrawal. Psychopharmacology (Berl). 2004;177(1-2):195-199. doi: 10.1007/s00213-004-1923-6. 26. Heatherton TF, Kozlowski LT, Frecker RC, Rickert W, Robinson J. Measuring the heaviness of smoking: using self-reported time to the first cigarette of the day and number of cigarettes 23 23 smoked per day. British Journal of Addiction. 1989;84:791-799. doi: 10.1111/j.1360- 0443.1989.tb03059.x. smoked per day. British Journal of Addiction. 1989;84:791-799. doi: 10.1111/j.1360- 0443.1989.tb03059.x. 27. Naughton F, Prevost AT, Gilbert H, Sutton S. Randomised controlled trial evaluation of a tailored leaflet and SMS text message self-help intervention for pregnant smokers (MiQuit). References Nicotine Tob Res. 2012;14(5):569-577. doi: 10.1093/ntr/ntr254. 28. Ajzen I, Fishbein M. The Influence of attitudes on behaviour. In: Albarracin D, Johnson BT, Zanna MP, eds. The handbook of attitudes. Mahwah, NJ: Lawrence Erlbaum Associates Publishers; 2005:173-221. 29. Royston P, White IR. Multiple Imputation by Chained Equations (MICE): Implementation in Stata. Journal of Statistical Software. 2011;45(4):1-20. doi: 10.18637/jss.v045.i04 30. Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med. 2008;3:17. doi: 10.1186/1751-0473-3-17. 31. Bowker K, Orton S, Cooper S, et al. Views on and experiences of electronic cigarettes: a qualitative study of women who are pregnant or have recently given birth. BMC Pregnancy Childbirth. 2018;18(1):233. doi: 10.1186/s12884-018-1856-4. 32. Cooper S, Orton S, Campbell KA, et al. Attitudes to E-Cigarettes and Cessation Support for Pregnant Women from English Stop Smoking Services: A Mixed Methods Study. Int J Environ Res Public Health. 2019;16(1). doi: 10.3390/ijerph16010110. 33. Taylor T, Hajek P. Smoking cessation services for pregnant women. London: Health Development Agency; 2001. 34. Klerman LV, Spivey C, Raykovich KT. Smoking reduction activities in a federal program to reduce infant mortality among high risk women. Tob Control. 2000;9 Suppl 3:III51-55. doi: 10.1136/tc.9.suppl_3.iii51. 35. Naughton F, Hopewell S, Sinclair L, McCaughan D, McKell J, Bauld L. Barriers and facilitators to smoking cessation in pregnancy and in the post-partum period: The health care 24 professionals' perspective. Br J Health Psychol. 2018;23(3):741-757. doi: 36. Flemming K, Graham H, McCaughan D, Angus K, Sinclair L, Bauld L. Health professionals' perceptions of the barriers and facilitators to providing smoking cessation advice to women in pregnancy and during the post-partum period: a systematic review of qualitative research. BMC Public Health. 2016;16:290. doi: 10.1186/s12889-016-2961-9. 37. Coleman T, Chamberlain C, Davey MA, Cooper SE, Leonardi-Bee J. Pharmacological interventions for promoting smoking cessation during pregnancy. Cochrane Database Syst Rev. 2015(12):CD010078. doi: 10.1002/14651858.CD010078.pub2. 38. Kotz D, Fidler J, West R. Factors associated with the use of aids to cessation in English smokers. Addiction. 2009;104(8):1403-1410. doi: 10.1111/j.1360-0443.2009.02639.x. 39. Fish LJ, Peterson BL, Namenek Brouwer RJ, et al. Adherence to nicotine replacement therapy among pregnant smokers. Nicotine Tob Res. 2009;11(5):514-518. doi: 10.1093/ntr/ntp032. 40. Coleman T, Cooper S, Thornton JG, et al. A randomized trial of nicotine-replacement therapy patches in pregnancy. N Engl J Med. 2012;366(9):808-818. doi: 10.1056/NEJMoa1109582. 41. Dempsey D, Jacob P 3rd, Benowitz NL. Accelerated metabolism of nicotine and cotinine in pregnant smokers. J Pharmacol Exp Ther. References 2002;301(2):594-598. doi: 10.1124/jpet.301.2.594. 42. Jordan TR, Dake JR, Price JH. Best practices for smoking cessation in pregnancy: do obstetrician/gynecologists use them in practice? J Womens Health. 2006;15(4):400-441. doi: 10.1089/jwh.2006.15.400. 43. Okoli CT, Greaves L, Bottorff JL, Marcellus LM. Health care providers' engagement in smoking cessation with pregnant smokers. J Obstet Gynecol Neonatal Nurs. 2010;39(1):64- 77. doi: 10.1111/j.1552-6909.2009.01084.x. 44. Abatemarco DJ, Steinberg MB, Delnevo CD. Midwives' knowledge, perceptions, beliefs, and practice supports regarding tobacco dependence treatment. J Midwifery Womens Health. 2007;52(5):451-457. doi: 10.1016/j.jmwh.2007.03.019 25 25 45. Owen L, Penn G. Smoking and Pregnancy: A Survey of Knowledge, Attitudes and Behaviour 1992 - 1999. London: Health Education Authority;1999. 46. Naughton F, Eborall H, Sutton S. Dissonance and disengagement in pregnant smokers: a qualitative study. Journal of Smoking Cessation. 2013;8(1):24-32. doi: 10.1017/jsc.2013.4. 47. Campbell KA, Bowker KA, Naughton F, Sloan M, Cooper S, Coleman T. Antenatal Clinic and Stop Smoking Services Staff Views on "Opt-Out" Referrals for Smoking Cessation in Pregnancy: A Framework Analysis. Int J Environ Res Public Health. 2016;13(10). doi: 10.3390/ijerph13101004. 48. Sloan M, Campbell KA, Bowker K, et al. Pregnant Women's Experiences and Views on an "Opt-Out" Referral Pathway to Specialist Smoking Cessation Support: A Qualitative Evaluation. Nicotine Tob Res. 2016;18(5):900-905. doi: 10.1093/ntr/ntv273. 49. Althabe F, Aleman A, Berrueta M, et al. A Multifaceted Strategy to Implement Brief Smoking Cessation Counseling During Antenatal Care in Argentina and Uruguay: A Cluster Randomized Trial. Nicotine Tob Res. 2016;18(5):1083-1092. doi: 10.1093/ntr/ntv276. 50. Naughton F, Cooper S, Foster K, et al. Large multi-centre pilot randomized controlled trial testing a low-cost, tailored, self-help smoking cessation text message intervention for pregnant smokers (MiQuit). Addiction. 2017;112(7):1238-1249. doi: 10.1111/add.13802. 26
https://openalex.org/W3113429441	https://journals.openedition.org/pontourbe/pdf/9067	Portuguese	null	Internet e trabalho de campo antropológico: dois relatos etnográficos	Ponto Urbe	2,020	cc-by	9,566	Núcleo de Antropologia Urbana da Universidade de São Paulo Núcleo de Antropologia Urbana da Universidade de São Paulo Cristina Marins Cristina Marins Internet e trabalho de campo antropológico: dois relatos etnográficos Internet and anthropology: an account on ethnographic fieldwork Internet and anthropology: an account on ethnographic fieldwork Ponto Urbe Revista do núcleo de antropologia urbana da USP 27 \| 2020 Ponto Urbe 27 Ponto Urbe Revista do núcleo de antropologia urbana da USP 27 \| 2020 Ponto Urbe 27 Refêrencia eletrónica Cristina Marins, «Internet e trabalho de campo antropológico: dois relatos etnográficos», Ponto Urbe [Online], 27 \| 2020, posto online no dia 28 dezembro 2020, consultado o 08 agosto 2024. URL: http:// journals.openedition.org/pontourbe/9067 ; DOI: https://doi.org/10.4000/pontourbe.9067 Este documento foi criado de forma automática no dia 8 de agosto de 2024. Apenas o texto pode ser utilizado sob licença CC BY 4.0. Outros elementos (ilustrações, anexos importados) são "Todos os direitos reservados", à exceção de indicação em contrário. Internet e trabalho de campo antropológico: dois relatos etnográficos 1 NOTA DO EDITOR Versão original recebida em / Original Version 13/04/2020 Aceitação / Accepted 20/06/2020 Versão original recebida em / Original Version 13/04/2020 Aceitação / Accepted 20/06/2020 Internet and anthropology: an account on ethnographic fieldwork Internet and anthropology: an account on ethnographic fieldwork Cristina Marins Internet e trabalho de campo antropológico: dois relatos etnográficos Introdução Este artigo busca problematizar a incorporação de ferramentas digitais de pesquisa no trabalho de campo antropológico a partir de duas experiências de pesquisa distintas. A primeira serviu de base a minha tese de doutorado sobre o trabalho de fotógrafos e foi realizada durante os anos de 2016 e 2018 entre profissionais especializados nos registros de ritos matrimoniais (Marins 2018). A segunda pesquisa, realizada durante estágio pós-doutoral, tomou como objeto a atuação de deputados federais eleitos no pleito de 2018 nos primeiros seis meses de seus mandatos. Ainda que à primeira vista a correlação entre as duas investigações não seja evidente, ambas tomaram como foco de observação processos de construção de reputação – ora num dado universo profissional, ora no âmbito da política institucional. Mais importante, considerando o propósito específico deste artigo, as duas pesquisas se apoiaram fortemente sobre ferramentas digitais, em especial, encontraram nas redes sociais. Na primeira experiência, a utilização de tecnologias de interação digital como estratégia privilegiada de pesquisa não foi propriamente premeditada, mas determinada pelo modo como os interlocutores que encontrei ao longo do trabalho de Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 2 campo lidavam eles próprios com a internet. Após uma primeira aproximação do grupo que conformava meu objeto de interesse em um congresso de natureza profissional, gradualmente me dei conta de que a internet era, naquele universo, uma importante fonte de informação e aprendizado, além de um espaço de sociabilidade, bem como uma ferramenta de marketing e vendas. Já o projeto da segunda pesquisa considerava a experiência da tese de doutorado. Naquele ponto, eu tinha maior clareza sobre as potencialidades da internet para a pesquisa antropológica. Interessada nos processos de construção de reputação de parlamentares que ocupavam um cargo eletivo pela primeira vez, me propus a acompanhar suas atividades de maneira sistemática durante seus primeiros seis meses de mandato, tomando como ponto de observação fundamental suas postagens em redes sociais (em especial, no Instagram, no Twitter e no Facebook). Com menor regularidade, observei também a presença dos parlamentares em canais jornalísticos tradicionais e, pontualmente, também em eventos presenciais. Ainda assim, a pesquisa apresentava peculiaridades que exigiam determinados cuidados até então inéditos para mim. Introdução Num contexto de intensas discussões sobre o papel da antropologia nas últimas décadas, marcadas por questões que incluem “globalização, industrialização e desindustrialização, comércio, fundamentalismo religioso, burocratização, conflitos étnicos, escassez, urbanização, desemprego, falta de moradia e a transformação da família e dos padrões de parentesco” (Ahmed apud. Ortner 1999: 984, tradução minha), reflexões em torno de novas formas de realização de pesquisa etnográfica e pesquisa de campo se fazem presentes, ao menos, desde o final da década de 1990. Como chama atenção Sherry Ortner em seu “Some futures of anthropology” pesquisadores vinham, já naquele momento, repensando e reelaborando a prática etnográfica” (Ortner 1999). Neste sentido, destaco que alguns trabalhos apontavam desde então para a necessidade de desenvolver ferramentas de pesquisa antropológica inovadoras, considerando a tradição da disciplina que implicitamente privilegia campos geograficamente bem demarcados1. O florescimento dos questionamentos em torno do papel da antropologia num contexto de transformações ocorreu ao mesmo tempo em que emergiam trabalhos que tomam como foco a investigação da chamada cibercultura. Escobar, em artigo que preconizava uma ampliação no campo antropológico das pesquisas dedicadas às transformações na sociedade decorrentes de novas tecnologias, defendia a articulação de um campo antropológico “particularmente preocupado com as construções e reconstruções culturais nas quais novas tecnologias são baseadas e, por sua vez, ajudam a configurar” (Escobar 1994: 211, tradução minha). Em 1999, a prestigiada editora Routledge publicou “Cyborgs@Cyberspace?” (Hakken 1999) com o sugestivo subtítulo “um etnógrafo olha para o futuro”. No ano 2000, foi publicado o livro de Christine Hine, “Virtual ethnography”, trabalho que mais tarde se desdobraria em frutífera agenda de pesquisa (Hine, 2015, Rufas e Hine 2018 e Hine, 2020). Ao longo das duas últimas décadas, com a internet se tornando onipresente no quotidiano de grande parcela da população mundial, solidificou-se o campo da antropologia dedicado ao tema. Uma abordagem etnográfica da internet pode ser verificada, por exemplo, nos trabalhos realizados por Daniel Miller e sua equipe no já consolidado campo da Antropologia Digital (Miller 2011, 2016, Miller et al 2016, Miller e Sinanan 2017, Miller 2017, Miller e Venkatraman 2018, Miller 2019). Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 3 Também no Brasil, o campo dos estudos da cibercultura tem rendido discussões tão numerosas quanto profícuas. Introdução Exemplos de grupos de trabalho dedicados ao tema são o pioneiro Grupo de Pesquisas em Ciberantropologia (GrupCiber), fundado em 1996 no Programa de Pós-Graduação em Antropologia Social da Universidade Federal de Santa Catarina e o Núcleo de Estudos da Modernidade (NEMO), sediado no Programa de Pós- Graduação em Antropologia da Universidade Federal Fluminense, que desde 2003 realiza estudos no campo da cibercultura, com especial foco no consumo. A partir de um esforço de mapeamento da produção científica sobre tecnologias digitais no Brasil, Freitas e Gomes (2015) observam que a dimensão etnográfica se encontra ainda ausente em parcela significativa das publicações. Já Segata (2016), ao delinear a trajetória dos estudos da cibercultura no país, pondera que, em seu estágio incipiente, o campo era marcado por desconfiança e acusações de esvaziamento humano da etnografia (p. 38) tendo sido, posteriormente, fortalecido com um investimento etnográfico substancial. Com o amadurecimento do campo de estudos, assistimos a uma gradual redução da distância entre “mundo real” e “mundo virtual”. Também foram criticadas as abordagens da internet como neutras e autoexplicativas, ao invés de considerar que ela é feita “de atores que reconfiguram tanto as tecnologias quanto seus ‘contextos’, um em relação ao outro” (Miller e Slater 2004: 42). Neste contexto, o exercício que apresento aqui vai ao encontro da proposta de Segata que, em entrevista, chamou atenção para o desafio de pensar “as particularidades de um trabalho de campo em que as novas tecnologias digitais, particularmente aquelas da internet e seus dispositivos, produzem o encontro etnográfico” (2017: 318). Pretendo, com este artigo, apresentar um registro de minha experiência de realização de pesquisas de campo costuradas pela utilização de ferramentas digitais, atentando para as potencialidades, impasses e limitações que minhas estratégias de investigação impuseram. Procurarei argumentar ainda que a categoria circuito – originada no campo dos estudos urbanos e presente em diversas reflexões de Magnani2 – pode abarcar também plataformas digitais que, mais do que meios de comunicação, são espaços nos quais as pessoas socializam (Miller et al. 2016). Uma vez que tendem a se ampliar ainda mais os os estudos antropológicos atravessados pelo universo digital, teço algumas considerações ao final do texto que visam contribuir para o debate acadêmico sobre novos desafios metodológicos. Inauguração de um trabalho de campo e os primeiro impasses Eu entrava no quarto semestre de doutorado e, de acordo com o cronograma previsto pelo Programa de Pós-Graduação no qual eu estava inscrita, deveria dar andamento à etapa da tese que convencionamos chamar de trabalho de campo. Inicialmente interessada em refletir sobre rituais de casamento a partir de suas fotografias, eu procurava traçar estratégias que me permitissem encontrar profissionais especializados no registro destes eventos que pudessem vir a se tornar interlocutores. Minha primeira providência, neste sentido, foi a mais banal que naquele momento poderia me ocorrer: digitei no Google uma expressão da qual não me recordo com exatidão, mas que suponho ter sido algo como “fotógrafos de casamentos”. Quase instantaneamente, surgiu na tela um resultado de busca que se revelaria decisivo para os desdobramentos da pesquisa: era anunciada, para dali a poucas semanas, a realização de um congresso sobre fotografia de casamento, adjetivado como “o maior da América Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos Latina”. Acessei com senso de urgência o endereço ali indicado e, diante de seu conteúdo, julguei que o evento era promissor. Se um dos fatores que pesaram em minha decisão de iniciar o trabalho de campo em um evento de natureza profissional foi a possibilidade de acessar potenciais interlocutores, já no primeiro dia do congresso revisei meu planejamento. O ritmo das atividades era intenso e difícil de seguir, de modo que procurei me concentrar na programação de palestras que se revelava mais interessante do que eu pudera imaginar. Disto decorreu que mal troquei palavras com os demais participantes do evento, com exceções que se revelariam importantes em fases posteriores da pesquisa. Duas delas foram fotógrafas que abordei logo após suas respectivas apresentações. Aqui, as chamarei de Letícia e Luciana3. Logo após a palestra de Letícia, provocada pelo conteúdo de sua fala, me juntei a outros espectadores que aguardavam para lhe dirigir algumas palavras. Em momento que julguei oportuno, me apresentei rapidamente e falei sobre minhas intenções de uma conversa posterior. Ao me ouvir afirmar que morava no Rio de Janeiro, Letícia ponderou um pouco e disse que visitaria a cidade dali a um mês e meio, propondo um encontro pessoal. Antes de encerrar a conversa, sugeriu que eu a “adicionasse no Facebook” para que pudéssemos combinar os detalhes da entrevista mais adiante. O encontro com Luciana seguiu um roteiro semelhante. Inauguração de um trabalho de campo e os primeiro impasses Observando as imagens que ela compartilhava, bem como os textos que as acompanhavam, me parecia evidente que o Instagram5 era utilizado naquele contexto como uma ferramenta de divulgação de seu trabalho. No caso de Luciana, era visível o notável cuidado com que tratavam esta rede social. Refletindo sobre as condições de realização de seu trabalho de campo entre habitantes de Sirkanda, no Himalaia, Gerald Berreman (1975) pondera que a primeira tarefa do etnógrafo ao chegar no campo é o confronto com sua própria apresentação diante do grupo que pretende estudar. Neste sentido, o autor argumenta que a interação entre pesquisador e pesquisados envolve controle de impressões, estas construídas “a partir do que os indivíduos fazem, assim como do que dizem, tanto em público, isto é, quando sabem que estão sendo observados, quanto privadamente, isto é, quando pensam que não estão sendo observados. As tentativas de dar a impressão desejada de si próprio, e de interpretar com precisão o comportamento e as atitudes dos outros são uma componente inerente de qualquer interação social e são cruciais para a pesquisa etnográfica.” (BERREMAN 1975: 125) Confiando em meu treinamento antropológico, bem como em minha experiência prévia de pesquisa, me julgava relativamente preparada para o jogo de controle de impressões do qual trata Berreman. Embora eu tivesse noção de que, tal qual sublinha este autor, o etnógrafo em campo é observado por seus interlocutores tanto quanto os observa, supus — erroneamente, como mais tarde se tornaria claro — que minhas atividades em redes sociais estavam inscritas em uma esfera de minha vida privada que passaria ao largo de minhas atividades de pesquisa. Contudo, logo me vi às voltas com as dificuldades e complexidades no gerenciamento de impressões no contexto das redes sociais, sobre os quais tratou Polivanov (2014). A primeira experiência de relacionamento online com meus interlocutores deixava clara a existência de uma assimetria entre nós quanto aos distintos usos de redes sociais, em especial no que diz respeito ao controle das impressões, para retomar o termo empregado por Berreman. Se meu uso particular do Instagram se aproximava daquele originalmente previsto por seus criadores6, fotógrafos como Letícia e Gabriel haviam-no convertido em ferramenta de marketing. Se, para eles, colocar suas contas de Facebook ou Instagram à disposição da etnógrafa parecia ser ato banal, meu constrangimento colocava em jogo a delicada questão das fronteiras entre vida pessoal e profissional, entre o público e o privado. Inauguração de um trabalho de campo e os primeiro impasses Assisti à sua palestra suspeitando que uma conversa mais detida com ela seria particularmente proveitosa para meu trabalho. Neste caso também, aguardei o final da apresentação para então me aproximar, falar de minha pesquisa e de meu interesse em ouvi-la mais detidamente. Meu pedido de entrevista foi acolhido com simpatia e entusiasmo, de modo que restava apenas combinar os detalhes do encontro. Antes que eu pudesse propor um contato posterior por e-mail, Luciana sugeriu, com smartphone em riste: “Me adiciona no seu Instagram ou no seu Facebook.” Aquelas tinham sido as primeiras vezes de muitas em que ouviria, partindo de potenciais interlocutores, frases como “_me adiciona”, “me add”, “qual o seu Instagram?”, “vou te procurar no Facebook”. Se para Luciana e Letícia (e, posteriormente, para outros interlocutores de pesquisa) esta aproximação via redes sociais parecia um passo natural no relacionamento que começava a se estabelecer, de minha parte, estas solicitações causaram certo constrangimento. Refletindo a respeito, constatei que este incômodo decorria de um certo receio quanto aos possíveis efeitos que o eventual compartilhamento do conteúdo veiculado por mim em redes sociais poderia provocar sobre a pesquisa. Naquele momento eu estava inscrita no Facebook e no Instagram, sendo que minha participação em cada uma destas mídias possuía características distintas. Embora eu não me considerasse exatamente uma usuária assídua no Facebook, minhas esparsas manifestações naquela rede social diziam respeito, basicamente, a meus posicionamentos sobre a conjuntura política do país. Minha hesitação em permitir que meus interlocutores acessassem tal conteúdo tinha origem no sentimento de que, diante de um cenário de “polarização política” (Machado e Miskolci 2019), minha identificação com uma certa linha ideológica viesse a dificultar o meu acesso a interlocutores que adotassem posicionamentos políticos opostos4. Já no Instagram, rede social que eu utilizava com maior frequência, eu veiculava prioritariamente imagens de minha família, produzidas com o intuito de circulação entre um grupo bastante Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 5 restrito. Conceder aos meus interlocutores o acesso a estas imagens, acarretava, para mim, certo prejuízo à minha privacidade. Diferente de meu uso pessoal das redes sociais, o de Luciana (e de boa parte de meus interlocutores, conforme eu viria a descobrir mais tarde) parecia ser calculado para comportar suas atividades profissionais. Inauguração de um trabalho de campo e os primeiro impasses Se o tema demandava reflexão cuidadosa – e, portanto, mais demorada do que o trabalho de campo permitiria – ela impunha, de imediato, um impasse a ser resolvido para que eu prosseguisse com a pesquisa, a saber: como lidar então com minhas redes sociais? Sem desenvolver uma estratégia clara nos primeiros meses de pesquisa, encaminhava a questão de maneira algo atabalhoada. Ao primeiro pedido realizado por Luciana, mencionado há pouco, reagi de maneira evasiva. De modo improvisado, tentei desviar o assunto sugerindo que talvez fosse mais simples uma comunicação pelo aplicativo de trocas de mensagens WhatsApp. Minha sugestão foi prontamente aceita e me vi livre do constrangimento de permitir acesso a minhas atividades nas redes sociais. Mais adiante, no entanto, durante um segundo encontro com Luciana, a solicitação de acesso às minhas redes sociais voltou a ocorrer. Uma vez que simplesmente negar o pedido Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 6 não me parecia uma opção razoável, optei – a contragosto – por compartilhar o acesso ao meu Instagram. Já no caso de Letícia, não vi remédio senão aceitar a conexão mediada pelo Facebook e, mais tarde, respirei aliviada ao ver que o conteúdo que minha interlocutora veiculava não se chocava com minhas próprias visões de mundo expostas naquele espaço. Por algum tempo, às sugestões do tipo “me add” feitas pessoalmente por interlocutores, eu respondia com sorrisos amarelos e desculpas mal elaboradas como “eu não uso muito o Facebook”. Quando conversava sobre meu impasse com colegas, algumas vezes ouvi deles a sugestão de criar “um perfil para a pesquisa” – possibilidade que desconsiderei antecipando os transtornos que a necessidade de gerenciar duas contas poderia acarretar. Enquanto isso, os novos pedidos de amizades em redes sociais surgiam e eu, ainda um pouco sem jeito, passei a aceitá-los, tomando o cuidado de reduzir minhas atividades na rede social, que já eram poucas, a quase nulas. Aquilo que, a meu ver, se aproximou de uma espécie de “saída honrosa” surgiu apenas quando decidi eu mesma aprender a fotografar e, com o intuito de pôr em prática meu recém-adquirido conhecimento teórico, criei uma conta no Instagram com o compromisso de postar fotografias feitas por mim, como uma espécie de mecanismo de autodisciplina. Minha nova conta acabou gerando alguns benefícios suplementares. Inauguração de um trabalho de campo e os primeiro impasses Já que eu tinha criado um perfil em rede social que me sentia à vontade para fornecer aos meus interlocutores, estava solucionado o meu incômodo em relação ao compartilhamento de meus perfis privados em redes sociais, ainda que eu não tivesse previsto esta possibilidade. Ela acarretou ainda um vínculo com meus interlocutores de natureza inesperada: alguns deles passaram a submeter espontaneamente minhas imagens a críticas, contribuindo, portanto, com ensinamentos valiosos, tanto para a pesquisa quanto para minha condição de fotógrafa neófita. Um desfecho para impasses iniciais e a atenção para novas potencialidades da pesquisa nas redes As mensagens surgiam na tela meu aparelho celular em momentos os mais diversos: enquanto eu participava de uma reunião de trabalho, no momento que eu preparava o meu jantar ou estava sentada no bar conversando com amigos. Boa parte de nossas conversas começava por iniciativa de Letícia. Muitas vezes ela me enviava algum tipo de material como, por exemplo, fotografias suas ou de colegas, links para entrevistas, blogs ou podcasts, enfim, algo que imaginava ser de interesse para minha tese. Outras vezes, Letícia simplesmente narrava algum episódio que lhe tinha acontecido, compartilhava fofocas ou exibia seus trabalhos dentro ou fora da fotografia de casamento. A partir desta interação que se desdobrava online, entre os anos de 2016 e 2017, Letícia e eu também arranjamos alguns reencontros presenciais. Em viagens, tivemos dias de intensa convivência e em suas vindas a minha cidade de residência, nos encontrávamos. Em suma, em diversos momentos ao longo de minha empreitada de pesquisa, Letícia me apresentou terminologias, equipamentos, técnicas, pessoas e grupos – numa convivência que, assim como ocorrera com diversos interlocutores, foi inteiramente costurada pelo uso de mídias digitais. Embora jamais tenha concebido a possibilidade de realizar uma pesquisa sem a utilização de instrumentos tais como mecanismos de busca e aplicativos de comunicação, quando a pesquisa era ainda um projeto, eu não podia imaginar a importância que o universo digital mais amplo assumiria no dia-a-dia do trabalho de campo. Com efeito, uma parcela muito significativa da pesquisa foi realizada online e, antes que eu pudesse ter me dado conta, acumulei horas de acompanhamento e interação com meus interlocutores em redes sociais, integrei rotinas de trocas de mensagens através de aplicativos de telefone celular, frequentei cursos transmitidos via internet, ouvi podcasts e tive acesso a um vasto conteúdo disponibilizado em plataformas de vídeo. Este modo de me mover pelo campo foi, em grande parte, fruto da relação que alguns interlocutores tinham com a internet. Logo no início da pesquisa, já era possível perceber que a rede funcionava para eles como importante fonte de informação (e formação), espaço de sociabilidade, ferramenta de divulgação do trabalho e também de comércio. Transitar eu mesma neste espaço se apresentava, portanto, como uma oportunidade de acompanhar e interagir com fotógrafos de casamento, não raro de maneiras inesperadas. Um desfecho para impasses iniciais e a atenção para novas potencialidades da pesquisa nas redes Se as demandas por relacionamentos vias redes sociais apresentava impasses inesperados, elas também geravam oportunidades que eu não fora capaz de prever, a exemplo das relações que se desenrolavam a partir de minha nova conta de Instagram. A sequência do meu primeiro encontro com Letícia deixava isto evidente. Uma vez conectadas pelo Facebook, acertamos através da ferramenta de troca de mensagens daquela rede social de nos encontrarmos para um café no bairro de Copacabana, na Zona Sul carioca, onde ela se hospedaria durante sua visita à cidade. A entrevista com Letícia teve a duração de quase quatro horas e partiu de mim a conclusão do encontro já no início da noite, uma vez que eu precisava dar conta ainda de alguns compromissos pessoais. Naquele mesmo dia, às 23h28, recebi uma notificação de mensagem do Facebook. Era Letícia retomando um dos temas de nossa conversa. Na semana que sucedeu a entrevista, Letícia e eu trocamos mensagens diariamente. Nelas, minha interlocutora desenvolvia alguns assuntos suscitados em nosso encontro e que lhe pareciam pendentes. De minha parte, além de ouvi-la atentamente, aproveitei a disponibilidade de Letícia lhe dirigindo novas perguntas, fazendo provocações e mesmo expondo alguns de meus primeiros esboços analíticos na expectativa de observar suas reações. Continuamos a nos falar com certa regularidade nos meses seguintes. Em momento que considerei o meu trabalho de campo concluído, achei por bem fazer um Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 7 back up das conversas que tive com Letícia através do aplicativo do Facebook. Assim como no corpo da tese, utilizei fonte Times New Roman, tamanho 12. As mensagens trocadas preencheram 252 páginas de um documento em Word. Mesmo depois disso, permanecemos em contato, embora com menor frequência. Diminuir as conversas com Letícia foi parte de um esforço de deixar para trás a pesquisa de campo para avançar em novas etapas da tese – algo não tão simples uma vez que as redes sociais seguiam estimulando o vínculo com esta e com outros interlocutores. Eu achava curiosa esta forma de interação e demorou algum tempo para que eu me desse conta de que “bater papo” com Letícia pelo aplicativo do Facebook era uma maneira de fazer pesquisa. A segunda pesquisa e o amadurecimento de estratégias No dia 4 de fevereiro de 2019, Kim Kataguiri, um dos parlamentares que eu acompanhava como parte de minha pesquisa de pós-doutorado, subiu na tribuna da Câmara dos Deputados para seu primeiro discurso. Dentro dos cinco minutos destinados à sua fala, agradeceu aos eleitores do estado de São Paulo pelos 465.310 votos recebidos no pleito de 2018, afirmou seu propósito de “trazer a renovação e a esperança na política que se perdeu”, fez menção de sua origem social. Também anunciou qual seria a pauta prioritária de seu mandato, agradeceu ao movimento político que ajudara a fundar anos antes e, no minuto final de sua fala, visivelmente comovido, fez menção a sua família. Nos meses que seguiram o primeiro discurso do jovem deputado, eu ganharia familiaridade com parte de sua rotina. O gabinete de Kataguiri, com uma parede verde ao fundo de sua mesa pouco a pouco transformada (em alguns meses, ela ganharia notícias de jornais e revistas estampadas com fotos do deputado emolduradas) tornar- se-ia logo um cenário conhecido, assim como os rostos daqueles que compartilhavam aquele o espaço. Embora no ano da realização de minha pesquisa, discussões e votações do Plenário e das comissões da Câmara dos Deputados já viessem sendo transmitidas há pouco mais de duas décadas pela TV Câmara7, através dos canais alimentados diretamente por parlamentares e seus auxiliares tais como Instagram e Facebook, eu ganhava outro tipo de acesso ao dia a dia da câmara dos deputados, incluindo seus bastidores. As redes sociais exibiam ainda fragmentos da intimidade dos deputados: imagens de suas residências, de seus animais de estimação e das refeições que faziam, por exemplo. Ainda que eu entendesse que aquela era uma edição do cotidiano cuidadosamente construída pelos atores políticos, buscar os critérios utilizados por eles para eleger os elementos publicizados nas redes sociais, por si só, já parecia exercício bastante promissor para uma etnógrafa em busca de compreender como a internet e as redes sociais vinham transformando a política institucional. Fato é que não fosse pelas redes sociais, o projeto de investigar processos de construção de reputação de alguns deputados federais que obtiveram votações expressivas no cenário brasileiro seria pouco provável. Como um projeto que se propunha a investigar um fenômeno que se desdobrava em uma instância que integra a elite política do país, era de se esperar que o acesso aos parlamentares apresentasse “barreiras etnográficas”. Um desfecho para impasses iniciais e a atenção para novas potencialidades da pesquisa nas redes Por exemplo, frequentando com regularidade grupos de Facebook que tinham a fotografia de casamento como foco principal, eu acompanhava diariamente algumas dezenas de posts sobre os mais variados aspectos do dia-a-dia dos fotógrafos. Posto que em alguns destes grupos, eram corriqueiras e assumiam posição de destaque as chamadas “tretas” (que eu definiria de maneira um tanto quanto simplificada como discórdias públicas entre alguns de seus integrantes) eu ganhava acesso quase que Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 8 diário aos conflitos que ficavam de fora de meu campo de visão em outras situações da pesquisa. O universo digital não serviu, portanto, apenas para mediar encontros com interlocutores “de carne e osso” — muito embora eu reconheça que este tenha sido um de seus papéis fundamentais. Nas redes sociais pude testemunhar conflitos, brincadeiras e conversas que a princípio soavam triviais, mas que, conforme eu avançava na pesquisa, se revelavam significativas. Em resumo, minha vida online garantia o quinhão diário de “convivência com os nativos”, se quisermos fazer alusão aos fundamentos da pesquisa etnográfica expostos nos escritos célebres de Malinowski (1978[1922]). Parecia uma maneira bastante conveniente de acessar informações que de outra forma não estariam ao meu alcance. A segunda pesquisa e o amadurecimento de estratégias Nos termos de Teixeira, as dificuldades de acesso poderiam incluir “portas fechadas, Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 9 interpelação de seguranças, espaços restritos, normas estritas sobre quando, onde e como circular nestes espaços e com quem se está autorizado a conversar por e quanto tempo” (2014: 34). O acesso aos parlamentares via redes sociais, por outro lado, não apresentava barreiras significativas. Além disso, acompanhar tais redes sistematicamente, ao menos a princípio, parecia projeto condizente com uma pesquisa cuja duração não deveria ultrapassar um ano. Assim, fiz do modo como os parlamentares lidavam com suas redes sociais o eixo central de minha pesquisa. Com efeito, a própria escolha dos atores que eu observaria tomava este elemento como ponto de partida. Como Kataguiri, outros parlamentares que eu observava dedicavam consideráveis esforços a nutrir suas redes sociais e encontraram nelas uma via expressa para eleição. Tendo em conta minha experiência de pesquisa com a utilização de mídias digitais até então, procurei tomar algumas providências que tornariam a investigação mais proveitosa. A primeira delas consistiu na criação de contas/perfis específicos para o trabalho, com o intuito principal de separar minhas atividades de pesquisa de atividades outras (informação, comunicação ou entretenimento, por exemplo) que eu pudesse vir a desempenhar nas mesmas redes sociais. Foram três as escolhidas: primeiro o Instagram, rede social que com a qual eu já tinha familiaridade e, conforme já mencionado, fora bastante utilizada em pesquisa anterior; o Facebook, também uma rede social que eu já utilizava desde 2010, ainda que com certa irregularidade e o Twitter, rede em que minha experiência enquanto usuária era quase nula. Com exceção do Facebook8, cuja proposta, segundo esclarece Ramos, é de fazer “convergir as identificações e autoidentificações dentro e fora da rede de computadores” funcionando como “lugar por excelência da apresentação de um eu tomado como unidade” (2015: 72), criei contas que seriam utilizadas exclusivamente para a pesquisa. As disparidades em meu histórico de utilização das diversas redes sociais somada a particularidades de cada uma das plataformas fez com que eu sentisse a necessidade de produzir um registro minucioso desta etapa do trabalho. A segunda pesquisa e o amadurecimento de estratégias Minha preocupação em registrar não só a criação dos perfis da pesquisa, mas, posteriormente, também material observado nas redes sociais, relacionava-se em parte com a constatação – respaldada, aliás, pela observação de outros pesquisadores do Ponto Urbe, 27 \| 2020 10 Internet e trabalho de campo antropológico: dois relatos etnográficos digital9 – do caráter efêmero, dinâmico, de constante transformação do conteúdo que circulava por ali10. Conforme já expus neste artigo, em meu trabalho anterior, lidei com as redes sociais de maneira improvisada. Isto fez com que eu deixasse de fazer registros supondo, erroneamente, que imagens e textos permaneceriam disponíveis online até o ponto em que eu realizaria a análise dos dados de pesquisa. Posto que não foram poucas as situações de perda do material ou ainda de inutilização de registros apressados, passei a empregar maior cuidado não só no registro da criação das contas, mas também no uso do print screen (a captura da tela do computador ou celular). Assim, abandonei a utilização indiscriminada de tal recurso, para registrar detalhes do contexto das capturas, incluindo os critérios de seleção para análise posterior. O cuidado que sobreveio certos lapsos anteriores permitiu a formulação de um caderno de campo digital cuja função primordial era registrar o contexto da construção dos dados, tornando mais produtiva a etapa de análise. Em outros termos, tratava-se de reconhecer que a seleção e o armazenamento dos dados deveriam conferir um caráter afirmativo ao trabalho de acompanhar as redes sociais dos atores que compunham meu universo de pesquisa. Deste modo, eu poderia construir ativamente uma leitura das redes sociais no lugar de simplesmente coletar dados de modo automático, seguindo certa tradição etnográfica (Geertz 2011 [1973]). E nesta etapa de análise de dados, a noção de circuito se tornou um elemento-chave para as duas diferentes pesquisas que transitavam entre universos online e offline. A segunda pesquisa e o amadurecimento de estratégias Tratei, assim, de criar caderno de campo – não nos moldes do caderno clássico problematizado por Magnani, “com espiral, formato 100 x 140 mm., próprio para a jaqueta ou o bolso de trás da calça jeans” (1997, sem paginação), no artigo “O (velho e bom) caderno de campo”. O caderno aqui em questão consistia numa versão digital do “velho e bom” caderno, num arquivo no formato .docx, armazenado simultaneamente em meu HD e em servidores virtuais. Minhas primeiras anotações no caderno de campo digital traziam os registros do processo de criação das novas contas, como pode ser observado no seguinte trecho: ...criei algumas contas para a minha nova pesquisa. (...) comecei criando uma conta do Twitter (@maisumapesquisa) num processo que não durou mais do que alguns minutos. Utilizei minha conta de e-mail, precisei inserir alguns dados incluindo número de telefone celular, confirmei duas vezes as informações através de códigos enviados primeiro para o meu e-mail, depois para o meu telefone. O Twitter então criou um nome de usuário genérico baseando-se em meu nome e sobrenome, mas após buscas rápidas aprendi a mudar meu nome de usuário e alterei minha foto de perfil. Imediatamente o Twitter sugeriu outros usuários para que eu seguisse a partir da busca por Kim Kataguiri. (...) Outras sugestões incluíram políticos do clã Bolsonaro, o empresário Luciano Hang e a Deputada Estadual Janaína Paschoal. ...criei algumas contas para a minha nova pesquisa. (...) comecei criando uma conta do Twitter (@maisumapesquisa) num processo que não durou mais do que alguns minutos. Utilizei minha conta de e-mail, precisei inserir alguns dados incluindo número de telefone celular, confirmei duas vezes as informações através de códigos enviados primeiro para o meu e-mail, depois para o meu telefone. O Twitter então criou um nome de usuário genérico baseando-se em meu nome e sobrenome, mas após buscas rápidas aprendi a mudar meu nome de usuário e alterei minha foto de perfil. Imediatamente o Twitter sugeriu outros usuários para que eu seguisse a partir da busca por Kim Kataguiri. (...) Outras sugestões incluíram políticos do clã Bolsonaro, o empresário Luciano Hang e a Deputada Estadual Janaína Paschoal. O circuito como unidade de análise Posto de outro modo, na perspectiva dos trabalhos de etnografia urbana, os pesquisadores se movem por diferentes pontos da cidade que se tornam uma unidade de análise a partir das práticas dos interlocutores de pesquisa – estas apreendidas em campo. As pesquisas que embasam a escrita deste artigo revelam que a validade da noção de circuito excede o contexto da etnografia urbana e pode ser proveitosamente apropriado por etnógrafos cujas pesquisas não se desenrolam exclusivamente online, mas são fortemente costuradas pelo universo digital. Assim, a internet torna o circuito uma categoria profícua para pensar para além da metrópole. No caso das pesquisas que servem de base a este artigo, circuito tem se mostrado útil para pensar uma categoria profissional cujas atividades não se encerram dentro de limites geográficos bem definidos, assim como para investigar os processos que redundam na emergência de um novo fenômeno político-eleitoral. Minha pesquisa de doutorado, cujo foco era um universo de profissionais que encontravam nas redes sociais um espaço de sociabilidade, formação, negócios e construção de reputação exigia o desenvolvimento de estratégias de pesquisa distintas daquelas apresentadas na maior parte das obras antropológicas que fizeram parte de minha formação, isto é, etnografias situadas dentro de limites geográficos/físicos mais ou menos bem demarcados. Posto que a pesquisa de campo fora inaugurada num encontro de natureza profissional cujos participantes residiam em pontos diversos do país – mas que se mantinham em estreita relação através de aplicativos de trocas de mensagem e redes sociais – mapear os circuitos percorridos por meus interlocutores seguindo seu trânsito online e off-line se revelou estratégia proveitosa. Mais tarde, foi esta categoria que possibilitou o projeto de minha pesquisa entre os parlamentares que seria de outro modo pouco provável. Neste caso, como na primeira pesquisa, as redes sociais foram cruciais para que eu pudesse refletir sobre o processo construção de reputação dos deputados que conformavam meu universo de pesquisa. Vale assinalar que a estratégia que adotei não é exatamente nova e que, nos últimos anos, me deparei com alguns trabalhos de pesquisa que apresentam variados graus de diálogo com trabalhos do “ciberespaço” que parecem se utilizar de pesquisar circuitos, ainda que nem sempre realizando referências diretas à categoria tomada do campo de estudos urbanos. Esta impressão parte, por exemplo, da leitura da dissertação de Faria (2014) dedicada a investigar as condições para o surgimento das chamadas “webcelebridades”. O circuito como unidade de análise Presente em diversos trabalhos realizados no âmbito do Laboratório do Núcleo de Antropologia Urbana (LabNAU/USP), a noção de circuito foi objeto de reflexão detida do antropólogo José Guilherme Magnani. Segundo aponta este autor, ela surgiu na continuidade de um estudo de práticas de lazer na periferia de São Paulo. Atrelado a outras categorias tais como mancha, pedaço e trajeto, circuito possui o traço distintivo de “vincular domínios não necessariamente marcados pela contiguidade espacial” (2014: 15), de conectar pontos descontínuos e distantes no tecido urbano, mas que integram um conjunto mais amplo. Como chama atenção Magnani, circuito é uma categoria abrangente porque “ao mesmo tempo que possibilita identificar e construir totalidades analíticas mais consistentes e coerentes com os objetos de análise, permite também extrapolar o espaço físico” (2010: 17), proporcionando assim recortes não restritos ao território da metrópole. Para citar algumas aplicações da noção de circuitos elaborados pela equipe do LabNAU/USP, esta se mostrou proveitosa, por exemplo, para uma investigação sobre os straight edge (Mantese 2006); em trabalho que enfoca baladas black e rodas de samba na cidade de São Paulo (Macedo 2007); ou para uma pesquisa que focaliza chamado “forró universitário” (Alfonsi 2008)11. Na análise de Magnani, o circuito é “determinado pelo recorte, pela pergunta, pelos objetivos da pesquisa em curso” (2014: 15). A ilustração fornecida pelo autor em seu esforço de delimitação da categoria é clara e merece ser revisitada: o campinho que serve para uma pelada de fim de semana na margem do rio Solimões pode ser incluído no circuito do futebol ou do lazer de um pesquisador interessado em formas de lazer vicinal. Se, por outro lado, o foco do interesse recair sobre jogos que integram um calendário de algum torneio, então o campinho é deixado fora do circuito. Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 11 11 Conforme evidenciado no artigo “De perto e de dentro”, circuito é uma noção que emerge num contexto de consolidação da etnografia urbana. Trata-se de um esforço de superação da ideia de “tentação da aldeia” (2002: 19), já célebre entre antropólogos. Nos trabalhos etnográficos que informam a reflexão de Magnani, a dimensão espacial não está de todo ausente, mas a noção de circuito marca a dispensa de contiguidade entre os diversos pontos da cidade. O circuito como unidade de análise Elas se inserem num contexto de necessidade de revisão de marcos metodológicos da antropologia que não descarta a teoria que nos antecedeu, mas reconhece que um mundo transformado exige de nós também estratégias novas de pesquisa. Neste sentido, a noção de circuito, tomada de empréstimos da antropologia urbana, parece se apresentar como caminho rentável. O circuito como unidade de análise Na dissertação em que a narrativa e história de vida são combinadas com dados construídos como “uma bricolagem de encontros (...) informais face a face e virtuais” (Faria 2014: 9), a autora se debruça sobre um vasto material que inclui postagens no Twitter e observações realizadas em um evento presencial. Este trabalho exibe diversas referências a circuitos (de formação, de produção e consumo), embora não sejam acompanhadas de uma reflexão cuidadosa sobre esta noção. Também a dissertação de Fernandes (2017) faz uma rápida menção a existência de um circuito fitness e, ainda que sem referência explícita à categoria notabilizada por Magnani, parece operar com esta ao acompanhar os usuários dentro e fora da rede social Ponto Urbe, 27 \| 2020 12 Internet e trabalho de campo antropológico: dois relatos etnográficos 12 Instagram, foco de sua investigação. Já a tese de doutorado de Mesquita, dedicada ao consumo de maquiagem por mulheres consideradas “aficionadas”, toma a categoria circuito como referência central para traçar estratégia metodológica. Neste sentido, Mesquita acompanhou consumidoras de maquiagem durante seu trabalho de campo, percorrendo estabelecimentos e eventos especializados, bem como redes sociais em declarada inspiração em trabalhos da antropologia urbana, em especial de Magnani (Mesquita 2020: 61). Produções como as de Faria, Fernandes e Mesquita indicam que a categoria analítica explorada nos artigos de Magnani já produz rendimentos analíticos em diversos campos de pesquisa antropológica. No caso de meus próprios empreendimentos investigativos, os circuitos permitiram uma abordagem de meus objetos de pesquisa de um ângulo particular. Tomando como eixo central os processos de construção de reputação ora de fotógrafos profissionais, ora de parlamentares, observei a relação de continuidade – e não de sobreposição – entre universos off-line e online. Em termos concretos, isto significa que ações e discursos que eu observava presencialmente adquiriam novos sentidos quando minha análise levava em conta também o cotidiano de meus interlocutores observado nas redes sociais. Mais do que isso, seguir o trânsito de meus interlocutores entre espaços online e off-line que frequentavam, permitiu que eu identificasse padrões e redes de relação pouco evidentes em estágio incipiente das investigações. Se são cada vez mais raros os contextos de pesquisa que não são de alguma forma atravessados pelo universo online, procuro levantar alguns tópicos que lhes dizem respeito na próxima seção que encerra este artigo. Considerações finais: em busca de uma renovada atenção à metodologia Como chamam atenção diversos pesquisadores que vêm se debruçando sobre redes sociais, há tempos não podemos mais pensar em “universo virtual” como algo que separa espaços online do “mundo real”. Antes, as mídias sociais devem ser encaradas como parte essencial de nosso cotidiano, do mesmo modo que encaramos uma conversa telefônica como parte da vida off-line (Miller et al. 2016). Uma vez que o universo digital é parte integrante do cotidiano de nossos interlocutores, aos etnógrafos resta dedicar atenção às implicações deste fenômeno em nossas pesquisas – e isto inclui suas potencialidades, mas também certos limites e desafios. Dentre os principais desafios implicadas nas investigações que realizei, para além dos impasses já mencionados neste artigo, gostaria de mencionar o caráter opaco das plataformas digitais. Entendendo que a etnografia é mais do que a coleção de dados in loco, mas a explicitação de sua construção com discussões críticas, considero que a prática etnográfica é dificultada à medida que parte dos rumos da pesquisa é regida por algoritmos e manipulações com as quais, até pouco tempo atrás, não precisávamos lidar. A influência de algoritmos no desenvolvimento da pesquisa etnográfica pode ser verificada, por exemplo, na situação já mencionada no início deste artigo. Refiro-me às Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 13 condições de minha entrada no campo, a partir dos resultados de busca do Google que indicaram um evento de natureza profissional que acabou se tornando definidor dos passos de pesquisa subsequentes12. Já os diversos tipos de manipulação nas redes sociais dificultam, em particular, a análise que leva em conta certas métricas disponíveis pelas redes sociais tais como números de seguidores e contagem de likes. No caso de minha pesquisa de pós-doutorado, na qual eu acompanhava de maneira sistemática as atividades de parlamentares recém-eleitos tomando como ponto de observação fundamental suas postagens em redes sociais, esta se tornou uma questão especialmente sensível. A investigação realizada por Lindquist (2018) que procura entender o funcionamento de um mercado ilícito de serviços digitais é reveladora da importância alcançada pelas manipulações nas redes sociais. O autor demonstra que adulterações nos números de seguidores, visualizações, likes e rankings de popularidade são mercadorias em circulação num mercado volátil e altamente competitivo. Considerações finais: em busca de uma renovada atenção à metodologia Lindquist afirma ainda que, nos últimos anos, tornou-se cada vez mais evidente a existência de um mercado consumidor global das chamadas “fábricas de seguidores” ou “fazendas de cliques”13, constando entre os grupos de compradores lojas online, os chamados “influenciadores” e políticos. Tenho algumas razões para suspeitar que os circuitos que pesquisei não estavam isentos deste tipo de manipulação. Com efeito, em minha pesquisa de doutorado, cuja duração excedeu em muito a pesquisa superior e, portanto, permitiu que a relação com interlocutores fosse estreitada, soube por um deles que a contratação de bots, isto é, aplicativos de software que executam scripts automatizados com o intuito de angariar seguidores fazia parte das estratégias de construção de reputação dos fotógrafos em redes sociais14. Contudo, mapear uma prática que não apenas é considerada ilegítima, mas também é potencialmente condenada do ponto de vista moral em universos de pesquisa diversos, constitui tarefa inglória. De todo modo, chamo atenção para a necessidade, ao integrarmos as redes sociais como ferramenta e locus da pesquisa etnográfica, de exercitarmos uma insistente desconfiança em relação à suposta neutralidade destas tecnologias. Levando em conta que há fortes chances de mecanismos de inteligência artificial abalarem, ao menos em algum grau, os rumos de nossas pesquisas, é importante que atentemos também para este aspecto, renovando a vocação de nossa disciplina que, nos termos de Lévi-Strauss, consiste numa “busca inquieta e fervorosa, que atormenta o investigador com questionamentos morais tanto quanto científicos” (2013 [1973]: 41). Se, por um lado, a internet nos convida a ampliar nossa atenção em relação aos dados que construímos em campo, ela também apresenta instrumentos valiosos ao etnógrafo. As ferramentas de busca disponíveis nas redes sociais podem ser largamente vantajosas, uma vez que elas permitem ao pesquisador realizar consultas filtradas por temas de seu interesse sem as restrições evidentes do recorte temporal da realização do trabalho de campo. A possibilidade de participação em fóruns, grupos de aplicativos de mensagens, bem como os registros das conversas travadas com interlocutores nas mídias digitais vem sendo igualmente úteis em pesquisas que abarcam o universo online. O grande volume de informações ao alcance de poucos cliques pode ser muito proveitoso. Enfim, com uma estratégia clara e renovada atenção aos aspectos metodológicos da pesquisa, a prática etnográfica recebe fôlego renovado. Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 14 Internet e trabalho de campo antropológico: dois relatos etnográficos BIBLIOGRAFIA ALFONSI, Daniela. 2008. Para todos os gostos: um estudo sobre classificações, bailes e circuitos de produção do forró. Dissertação de Mestrado. São Paulo: FFLCH-USP. BERREMAN, Gerald D. 1975. “Etnografia e controle de impressões em uma aldeia do Himalaia”. In: GUIMARÃES, Alba Zaluar (org.) Desvendando máscaras sociais. Rio de Janeiro: Francisco Alves. pp. 123-174. ESCOBAR, Arturo. 1994. Welcome to Cyberia: notes on the anthropology of cyberculture. Current Anthropology 35(3): 211-231. FARIA, Louise Scoz Pasteur de. 2014. A arte da fama: Victor Calazans e a construção de (@)Hebe Camargo enquanto uma webcelebridade. Dissertação de mestrado em Antropologia Social. Porto Alegre: IFCH-UFRGS. FERNANDES, Ana Carolina Machado. 2015. #Musas do Instafit: as construções de corpos e reputações mediadas pela rede social Instagram. 201p. Dissertação de mestrado em Antropologia. Niterói: ICHF-UFF. FREITAS, Eliane T. e GOMES, Laura G. 2015. Uma Antropologia da cibercultura. Dossiê Antropologia da Cibercultura. Vivência: revista de antropologia. UFRN/DAN/PPGAS v. I. GEERTZ, Clifford. 2011 [1973]. Uma descrição densa: por uma teoria interpretativa da cultura. In: A interpretação das culturas. Rio de Janeiro: LTC. GOFFMAN, Erving. 1985 [1956]. A representação do eu na vida cotidiana. 10. ed. Petrópolis: Vozes. GUPTA, Akhil e FERGUSON, James (org.). 1997. Anthropological locations: boundaries and grounds of a field science. Berkeley/Los Angeles/Londres: University of California Press. HAKKEN, David. 1999. Cyborgs@Cyberspace? An Ethnographer looks to the Future. Londres: Routledge. HINE, Christine. 2000. Virtual ethnography. Londres: Sage. HINE, Christine. 2015. Ethnography for the Internet: Embedded, Embodied and Everyday. Christine Hine. Bloomsbury Academic. HINE, Christine. 2020. The evolution and diversification of Twitter as a cultural artefact in the British press 2007–2014. Journalism Studies 21 (5) pp. 678-696, Taylor and Francis. LEITÃO, Débora K. e GOMES, Laura Graziela. 2017. Etnografia em ambientes digitais: perambulações, acompanhamentos e imersões. Antropolítica, n. 42: 41-65. LÉVI-STRAUSS, Claude.2013 [1973]. Antropologia estrutural dois. São Paulo: Cosac Naify. LUPTON, Deborah. 2015. Digital Sociology. Londres: Routledge. LINDQUIST, Johan. 2019. Illicit Economies of the Internet: Click Farming in Indonesia and Beyond. In.: FRANCESCHINI, Ivan and LOUBERE, Nicholas (org.). Dog Days: A Year of Chinese Labour, Civil Society, and Rights. ANU Press, The Australian National University, Canberra, Australia. MACHADO, Jorge e MISKOLCI, Richard. 2019. Das Jornadas de Junho à cruzada moral: o papel das redes sociais na polarização política brasileira. Sociologia & Antropologia, vol.9, n.3. Ponto Urbe, 27 \| 2020 15 Internet e trabalho de campo antropológico: dois relatos etnográficos MACEDO, Marcio. 2007. BIBLIOGRAFIA Baladas black e rodas de samba da Terra da Garoa. In: MAGNANI, José G. C. e MANTESE, Bruna. (orgs.) Jovens na Metrópole: etnografias dos circuitos de lazer, encontro e sociabilidade. São Paulo, Editora Terceiro Nome: 189-224. MAGNANI, José Guilherme Cantor. 1997. O velho e bom caderno de campo. Revista Sexta Feira, n. 1: 8-12. MAGNANI, José Guilherme Cantor. 2010. Os circuitos dos jovens urbanos. Sociologia: Revista do Departamento de Sociologia da FLUP, Vol. XX. MAGNANI, José Guilherme Cantor. 2013. Da periferia ao centro, cá e lá: seguindo trajetos, construindo circuitos. Anuário Antropológico, II, 53-72. MAGNANI, José Guilherme Cantor. 2014. O Circuito: proposta de delimitação da categoria. Ponto Urbe (online), 15. MALINOWSKI, Bronislaw. 1978 [1922]. Argonautas do Pacífico Ocidental: um relato do empreendimento e da aventura dos nativos nos arquipélagos da Nova Guiné melanésia. São Paulo: Abril Cultural. MANTESE, Bruna. 2006. Os straight edges e suas relações com a alteridade na cidade de São Paulo. Dissertação de mestrado. São Paulo: USP. MARCUS, George E. 1995. Ethnography in/of the World System: the emergence of multi-sited ethnography. Annual Review of Anthropology, 24: 95-117. MARKOWITZ, Lisa. 2001. Finding the field: notes on the ethnography of NGOs. Human Organization, 60(1): 40-46. MARINS, Cristina Teixeira. 2018. Entre palcos e flashes: reflexão etnográfica sobre trabalho, construção de reputação e circuitos de consagração de fotógrafos de casamento. Tese de doutorado em Antropologia. Niterói: ICHF-UFF. MESQUITA, Marcia. 2020. “Um reboco é um reboco”: maquiagem como performance de gênero. Tese de doutorado em Antropologia. Niterói: ICHF-UFF. MILLER, Daniel and SLATER, Don. 2004. Etnografia on e off-line: cibercafés em Trinidad. Horizontes antropológicos [online], vol.10, n.21: 41-65. MILLER, Daniel. 2011. Tales from Facebook. Cambridge: Polity. MILLER, Daniel. 2011. Tales from Facebook. Cambridge: Polity. MILLER, Daniel. 2016. Social media in an English village. London: UCL Press. MILLER, Daniel; COSTA, Elisabetta; HAYNES, Nell; MCDONALD, Tom; NICOLESCU, Razvan; SINANAN, Jolynna; SPYER, Juliano; VENKATRAMAN, Shriram and WANG, Xinyuan. 2016. How the World Changed Social Media. Londres: UCL Press. MILLER, Daniel e SINANAN, Jolynna. 2017. Visualising Facebook: a comparative perspective. London: UCL Press. MILLER, Daniel. 2017. The ideology of friendship in the era of Facebook. HAU: Journal of Ethnographic Theory, 7 (1) pp. 377-395. MILLER, Daniel. e VENKATRAMAN, Shriram. 2018. Facebook Interactions: An Ethnographic Perspective. Social Media and Society, 4 (3). MILLER, Daniel. 2019. Anthropological studies of mobile phones. Technology and Culture, 60 (4) pp. 1093-1097. ORTNER, Sherry B. 1999. Some Futures of Anthropology. BIBLIOGRAFIA American Ethnologist 26(4): 984-991, 1999. Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 16 PIERRO, Bruno de. 2018. O mundo mediado por algoritmos. Revista Pesquisa FAPESP, n. 266: 18-25. POLIVANOV, Beatriz. 2014. Dinâmicas identitárias em sites de redes sociais: estudo com participantes de cenas de música eletrônica no Facebook. Rio de Janeiro: Multifoco. RAMOS, Jair. 2015. Subjetivação e poder no ciberespaço: da experimentação à convergência identitária na era das redes sociais. Revista Vivência, n. 45: 57-76. RUFAS, Alix e HINE, Christine. 2018. Everyday connections between online and offline: Imagining others and constructing community through local online initiatives. New Media & Society, 20(10), 3879–3897. SANJEK, Roger. 2013. Ethnography in Today's World: Color Full Before Color Blind. Philadelphia: University of Pennsylvania Press. SEGATA, Jean. 2016. Um efeito ciber na antropologia. Revista Florestan, Ano 2, n.04. SEGATA, Jean. 2017. Cibercultura, imagem e ética na pesquisa. [Entrevista concedida a] Lorena Tamyres Trindade da Costa. Revista Visagem, Natal e Belém, v.03, n.2: 314-331. TEIXEIRA, Carla Costa. 2014. Pesquisando instâncias estatais: reflexões sobre o segredo e a TEIXEIRA, Carla Costa. 2014. Pesquisando instâncias estatais: reflexões sobre o segredo e a mentira. In: CASTILHO, Sérgio R.R.; SOUZA LIMA, Antonio Carlos e TEIXEIRA, Carla C. (Org.). Antropologia das Práticas do Poder: reflexões etnográficas entre burocratas, elites e corporações. 1ed. Rio de Janeiro: Contracapa, v. 1, pp. 33-42. ZIEWITZ, Malte. 2016. Governing Algorithms: Myth, Mess, and Methods. Science, Technology, & Human Values, 41(1): 3-16. Internet e trabalho de campo antropológico: dois relatos etnográficos 17 Internet e trabalho de campo antropológico: dois relatos etnográficos 9. Ver, por exemplo, Leitão e Gomes (2017) e Lupton (2015). 9. Ver, por exemplo, Leitão e Gomes (2017) e Lupton (2015). 10. Duas mudanças ocorridas em meio à realização das pesquisas que servem de base a este artigo ilustram este argumento: a primeira, foi a inserção no Instagram de um recurso que permite aos usuários postar imagens, textos e vídeos que desaparecem em 24 horas, alterando bastante a dinâmica da rede. A segunda, observada também no Instagram, foi a eliminação da contagem de “curtidas” que deixou longe dos olhos do público (e também da pesquisadora) um indicativo importante da repercussão das postagens. 11. Estas e outras etnografias do campo da antropologia urbana que se valem da noção de circuitos é apresentada no próprio artigo de Magnani (2010: 18-32). 12. Ao leitor não familiarizado com o termo, esclareço que os algoritmos são sequências de etapas para resolver um problema ou realizar uma tarefa de forma automática. Tratam-se de receitas escritas em linguagem de programação computacional, capazes de filtrar em segundos bilhões de páginas na web. Para uma abordagem introdutória algoritmos e seus crescentes impactos na vida social, recomendo a leitura de Pierro, 2018. Para uma reflexão mais aprofundada sobre o tema, ver: Ziewitz, 2015. 13. Respectivamente, “follower factories” e “click farms”, no texto original. 14. Este episódio foi relatado em minha tese, já mencionada (Marins 2018: 194). NOTAS 3. Como de praxe em textos etnográficos, os nomes de meus interlocutores são fictícios. 4. Em acordo com a teoria goffmaniana sobre as representações de si (1985 [1956]), eu deduzia que apresentar aos meus interlocutores o conteúdo de meu perfil do Facebook significava veicular certas informações a meu respeito que, naquele momento, eu preferia resguardar. Talvez a questão se colocasse em outros termos caso meus interlocutores propusessem a relação nas redes sociais em outro momento que não o início da interlocução de pesquisa. 3. Como de praxe em textos etnográficos, os nomes de meus interlocutores são fictícios. 4. Em acordo com a teoria goffmaniana sobre as representações de si (1985 [1956]), eu deduzia que apresentar aos meus interlocutores o conteúdo de meu perfil do Facebook significava veicular certas informações a meu respeito que, naquele momento, eu preferia resguardar. Talvez a questão se colocasse em outros termos caso meus interlocutores propusessem a relação nas redes sociais em outro momento que não o início da interlocução de pesquisa. 5. Para uma análise mais detida do uso desta rede social em pesquisa etnográfica, recomendo a leitura de Fernandes (2015). 5. Para uma análise mais detida do uso desta rede social em pesquisa etnográfica, recomendo a leitura de Fernandes (2015). 6. “O Instagram é uma rede social que inicialmente pretendia ser um mural de fotos ou de memórias, como antes era o álbum de fotografias que contava as histórias de famílias, lembranças de viagens, festas, momentos para serem guardados e revisitados de tempos em tempos” (Fernandes, 2015, p.22) 6. “O Instagram é uma rede social que inicialmente pretendia ser um mural de fotos ou de memórias, como antes era o álbum de fotografias que contava as histórias de famílias, lembranças de viagens, festas, momentos para serem guardados e revisitados de tempos em tempos” (Fernandes, 2015, p.22) 7. Emissora estatal criada em 1998 para transmitir as discussões e votações do Plenário e das comissões da Câmara dos Deputados. 8. Inicialmente, minha intenção era criar uma nova conta também no Facebook. Abandonei o plano, contudo, ao perceber que ele implicava um limite ético, já que segundo informava o site da empresa, manter mais de uma conta pessoal naquela rede social contrariava os “Padrões da Comunidade do Facebook”. Ponto Urbe, 27 \| 2020 RESUMOS Ponto Urbe, 27 \| 2020 Internet e trabalho de campo antropológico: dois relatos etnográficos 18 ÍNDICE Keywords: ethnography, methodology, circuits, internet, social networks Palavras-chave: etnografia, metodologia, circuitos, internet, redes sociais RESUMOS Neste artigo, busco problematizar a realização de trabalho de campo antropológico com o auxílio de redes sociais a partir de duas experiências de pesquisa. Embora as duas investigações tenham se desenrolado em campos distintos, ambas foram desenvolvidas online em grande medida. Pretendo, assim, registrar e analisar as experiências de pesquisas de campo costuradas pela utilização de ferramentas digitais, atentando para as potencialidades, impasses e limitações de minhas estratégias de investigação. Travando diálogo com outras pesquisas etnográficas recentes, procuro argumentar, ainda de maneira incipiente, que a categoria circuito, originada no campo dos estudos urbanos, pode ser proveitosamente apropriada por etnógrafos cujas pesquisas não se desenrolam exclusivamente online, mas abrangem o universo digital. Considerando que os estudos antropológicos atravessados pela internet tendem a se ampliar, é propósito deste artigo contribuir para o debate acadêmico em torno dos novos desafios metodológicos neste contexto. Neste artigo, busco problematizar a realização de trabalho de campo antropológico com o auxílio de redes sociais a partir de duas experiências de pesquisa. Embora as duas investigações tenham se desenrolado em campos distintos, ambas foram desenvolvidas online em grande medida. Pretendo, assim, registrar e analisar as experiências de pesquisas de campo costuradas pela utilização de ferramentas digitais, atentando para as potencialidades, impasses e limitações de minhas estratégias de investigação. Travando diálogo com outras pesquisas etnográficas recentes, procuro argumentar, ainda de maneira incipiente, que a categoria circuito, originada no campo dos estudos urbanos, pode ser proveitosamente apropriada por etnógrafos cujas pesquisas não se desenrolam exclusivamente online, mas abrangem o universo digital. Considerando que os estudos antropológicos atravessados pela internet tendem a se ampliar, é propósito deste artigo contribuir para o debate acadêmico em torno dos novos desafios metodológicos neste contexto. In this article, I present an account of ethnographic fieldwork of two separate studies carried out simultaneously online and offline. Starting from my own ethnographic research experiences, I expect to abbreviate the existing gap concerning potentialities and pitfalls of online ethnographic tools to study markedly fluid and heterogeneous groups in terms of their configuration. I argue that the category so called “circuit”, originated in the field of urban studies, can be successfully appropriated by ethnographers whose researches take place online and offline at once. It is my goal to contribute to the refinement of debates supporting anthropological investigations in a context of rapid and profound transformations shaped by digital technology. CRISTINA MARINS Doutora pela Universidade Federal Fluminense E-mail: ctmarins@gmail.com ORCID: https://orcid.org/0000-0002-9016-8215 Ponto Urbe, 27 \| 2020
https://openalex.org/W2540599737	https://jeccr.biomedcentral.com/track/pdf/10.1186/s13046-016-0447-3	English	null	A panel of four genes accurately differentiates benign from malignant thyroid nodules	Journal of experimental & clinical cancer research	2,016	cc-by	6,136	© The Author(s). 2016 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. A panel of four genes accurately differentiates benign from malignant thyroid nodules Qing-Xuan Wang1†, En-Dong Chen1†, Ye-Feng Cai1†, Quan Li1, Yi-Xiang Jin1, Wen-Xu Jin1, Ying-Hao Wang1, Zhou-Ci Zheng1, Lu Xue2, Ou-Chen Wang1 and Xiao-Hua Zhang1* Keywords: Thyroid nodules, Diagnostic panel, Biomarkers As a result of the increased frequency of medical exami- nations, such as ultrasonic examination, clinicians are con- fronted with an increasing number of patients with thyroid nodules. However, the majority of these thyroid nodules are benign, and only a few cases are malignant [7, 8]. Therefore, accurately diagnosing thyroid nodules preopera- tively is important for patients with thyroid cancer to receive timely treatment appropriately. Also, those with be- nign thyroid nodules can avoid unnecessary treatments, such as diagnostic surgery. Preoperative diagnosis of thy- roid nodules can also help to reduce medical burden of pa- tients and the government to a large extent, particularly in developing countries. * Correspondence: zxhoncology@sina.com †Equal contributors 1Department of Oncology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province 325000, China Full list of author information is available at the end of the article Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 DOI 10.1186/s13046-016-0447-3 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 DOI 10.1186/s13046-016-0447-3 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 DOI 10.1186/s13046-016-0447-3 Open Access Abstract Background: Clinicians are confronted with an increasing number of patients with thyroid nodules. Reliable preoperative diagnosis of thyroid nodules remains a challenge because of inconclusive cytological examination of fine-needle aspiration biopsies. Although molecular analysis of thyroid tissue has shown promise as a diagnostic tool in recent years, it has not been successfully applied in routine clinical use, particularly in Chinese patients. Methods: Whole-transcriptome sequencing of 19 primary papillary thyroid cancer (PTC) samples and matched adjacent normal thyroid tissue (NT) samples were performed. Bioinformatics analysis was carried out to identify candidate diagnostic genes. Then, RT-qPCR was performed to evaluate these candidate genes, and four genes were finally selected. Based on these four genes, diagnostic algorithm was developed (training set: 100 thyroid cancer (TC) and 65 benign thyroid lesions (BTL)) and validated (independent set: 123 TC and 81 BTL) using the support vector machine (SVM) approach. Results: We discovered four genes, namely fibronectin 1 (FN1), gamma-aminobutyric acid type A receptor beta 2 subunit (GABRB2), neuronal guanine nucleotide exchange factor (NGEF) and high-mobility group AT-hook 2 (HMGA2). A SVM model with these four genes performed with 97.0 % sensitivity, 93.8 % specificity, 96.0 % positive predictive value (PPV), and 95.3 % negative predictive value (NPV) in training set. For additional independent validation, it also showed good performance (92.7 % sensitivity, 90.1 % specificity, 93.4 % PPV, and 89.0 % NPV). Conclusions: Our diagnostic panel can accurately distinguish benign from malignant thyroid nodules using a simple and affordable method, which may have daily clinical application in the near future. Conclusions: Our diagnostic panel can accurately distinguish benign from malignant thyroid nodules using a simple and affordable method, which may have daily clinical application in the near future. Keywords: Thyroid nodules, Diagnostic panel, Biomarkers Background Thyroid cancer (TC) is the most common endocrine malig- nancy [1]. Recently, the number of thyroid carcinoma cases annually had increased by 4 % globally [2] and became the fastest growing type of cancer in many countries [1, 3]. China also suffers a large burden from thyroid cancer [4, 5]. China accounts for a large portion of thyroid cancer pa- tients around the world. Papillary thyroid cancer (PTC) ac- counts for 80–85 % of all types of thyroid carcinomas [6]. * Correspondence: zxhoncology@sina.com †Equal contributors 1Department of Oncology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province 325000, China Full list of author information is available at the end of the article Page 2 of 8 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Page 2 of 8 Page 2 of 8 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Cytological examination by ultrasound-guided fine- needle aspiration biopsy (US-FNAB) is the most accur- ate and gold-standard preoperative diagnostic method currently [9, 10]. However, about 17 % (10–26 %) of FNAB are reported as indeterminate, and 6 % (1–11 %) were nondiagnostic; while 72 % median (range 62–85 %) of undertaken FNAB were benign, and 5 % (1–8 %) were malignant [11]. As a consequence, most of these patients with indeterminate result undergo unnecessary diagnos- tic surgery, making them suffer from lifelong replace- ment therapy of thyroid hormones and corresponding surgical complications [12]. Technologies) according to the manufacturer’s instructions. The cDNA libraries were then processed for sequencing using Illumina Hiseq 2500 according to the commercially available protocols. Before reads mapping, clean reads were obtained from the raw reads by removing the adaptor sequences, reads with > 5 % ambiguous bases, and low- quality reads. The clean reads were then aligned to the human genome (version: GRCH37) using the MapSplice program (v2.1.6, University of Kentucky, Lexington, KY, USA). We applied Ebseq algorithm to screen out the differ- ently expressed genes using the following criteria: 1) fold change (FC) > 2 for up or downregulation; 2) false discov- ery rate (FDR) < 0.05 and 3) P-Value < 0.05. g p In recent years, molecular analysis of thyroid tissue has shown promise as a diagnostic tool [8, 13–15]. A microarray-based test, combining the expression levels of 167 genes (The Afirma) can diagnose thyroid nodules with 92 % sensitivity and 52 % specificity [13]. Background Another panel, named ThyroSeq v2 assay, which consists of more than 50 gene mutations and gene fusions (more than 1,000 hotspots), makes accurate diagnosis with 90 % sen- sitivity, 93 % specificity, 83 % positive predictive value (PPV), and 96 % negative predictive value (NPV) by using next-generation sequence [8]. Other studies also performed with impressive PPV or NPV [14–17]. Never- theless, studies aforementioned have their limitations. Some studies contain large amounts of gene alterations, such as The Afirma and ThyroSeq v2 assay, which are ex- pensive and difficult to promote, particularly in less- developed countries. Some studies failed to show both optimal sensitivity and specificity in validation analysis. Moreover, majority of these panels were based on patients from Western countries. Since genetic background is dif- ferent between Asians and Western populations, whether these panels could be successfully applied to Chinese pa- tients still remain unknown. RNA isolation and RT-qPCR RNA isolation and RT-qPCR Total RNA was isolated using TRIZOL Reagent (Invitrogen) and reverse transcription (TOYOBO, Japan) was performed according to the manufacturer’s instructions. Quantitative real-time PCR (RT-qPCR) analysis was performed in tripli- cate on the ABI prism 7500 sequence detection system (Ap- plied Biosystems, USA) using the THUNDERBIRD SYBR qPCR Mix (TOYOBO, Japan) according to manufacturer’s instructions. The GAPDH mRNA level was used for normalization. Primer sequences were as follows: FN1: 5′- TGGCCCCCACCTTCTTG -3′ (forward) and 5′- TGCGG GAAAAATCCCTTCTA -3′ (reverse); GABRB2: 5′- ACT Patients and tissue collection Primary TC samples or benign thyroid lesions and their matched adjacent normal thyroid tissue samples were obtained at the time of initial surgery from the First Af- filiated Hospital of Wenzhou Medical University. Inclu- sion and exclusion criteria are as follows, inclusion criteria: patients with surgical indication; tumor size lar- ger than 5 mm; papillary thyroid cancer or follicular thy- roid cancer or anaplastic thyroid cancer or medullary thyroid cancer or benign thyroid lesions; with accurate postoperative pathological diagnosis; with informed con- sent from patients. Exclusion criteria: tumor size less than or equal to 5 mm; with inconclusive postoperative pathological diagnosis; with other malignant tumors; pa- tients with surgical contraindications. Samples were snap- frozen in liquid nitrogen immediately after surgical resec- tion and subsequently stored at a −80 °C freezer. All the training and validation samples set used in this study are snap frozen tissues. Histopathological slides were reviewed retrospectively for all cases to confirm the histological diagnosis and to ensure abundant cancer content of the tumor by two pathologists. Detailed information about these patients are shown in Additional files 1 and 2. In- formed consent for the scientific use of biological material was obtained from each patient and the work was ap- proved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. Therefore, in the present study, we performed whole- transcriptome sequencing of 19 paired-PTC tissue samples and bioinformatics analysis to investigate global mRNA ex- pression. By using molecular approaches and statistical models, we successfully established a diagnosis model that combines only four genes: fibronectin 1 (FN1), gamma-aminobutyric acid type A receptor beta 2 sub- unit (GABRB2), neuronal guanine nucleotide exchange factor (NGEF), and high-mobility group AT-hook 2 (HMGA2). This diagnosis model could discriminate malignant from benign thyroid nodules with both high sensitivity and specificity in Chinese patients. Statistical analysis Data on normal distribution were expressed as mean ± standard variation (SD) and were compared with t-test. Categorical variables were expressed as percentage and were compared with Chi-square test or Fisher’s exact test, as appropriate. All P values were two sided, and a P value of < 0.05 was considered statistically significant. Statistical analysis was performed with SPSS software version 19.0 (SPSS, Chicago, IL, USA). GraphPad Prism 5 (GraphPad Software, USA) was used for graphs. Model development We evaluated several methodologies for building statistical models that can predict benign versus malignant thyroid nodules based on expression level of the four genes. Fi- nally, the SVM (support vector machine) was used for classification. With the modeling platform of MATLAB 2014a, we establish the SVM model for the classification of benign and malignant thyroid nodules by using Libsvm 3.20. During the development of SVM model, using the training data, we chose the C-SVC and RBF kernel func- tions and used the grid search method to debug the model. The grid c bound was −8 to 8, the grid c step was 0.5, the grid g bound was −8 to 8, and the grid g step was 0.5. The fold number of cross validation was 5. Methods Next generation sequencing and bioinformatics analysis Total RNA was extracted from tissue samples using TRI- ZOL Reagent (Invitrogen) according to the manufacturer’s protocol. After the quality test, The cDNA libraries were prepared using Ion Total RNA-Seq Kit v2.0 (Life Page 3 of 8 Page 3 of 8 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 GCCGTGTCACATCCTCTAA -3′ (forward) and 5′- TGG AACTGTCAACTTGCTTCAAA -3′ (reverse); NGEF: 5′- GAGCTGGAAATGGTGGTGAAG -3′ (forward) and 5′- TTCCGTGCGGCTCATTTT -3′ (reverse); HMG A2: 5′- GACCCAAAGGCAGCAAAAAC -3′(forward) and 5′- AGTGGCTTCTGCTTTCTTTTGAG -3′(reverse); GAPDH: 5′-GGTCGGAGTCAACGGATTTG-3′(forward) and 5′-ATGAGCCCCAGCCTTCTCCAT -3′(reverse). GCCGTGTCACATCCTCTAA -3′ (forward) and 5′- TGG AACTGTCAACTTGCTTCAAA -3′ (reverse); NGEF: 5′- GAGCTGGAAATGGTGGTGAAG -3′ (forward) and 5′- TTCCGTGCGGCTCATTTT -3′ (reverse); HMG A2: 5′- GACCCAAAGGCAGCAAAAAC -3′(forward) and 5′- AGTGGCTTCTGCTTTCTTTTGAG -3′(reverse); GAPDH: 5′-GGTCGGAGTCAACGGATTTG-3′(forward) and 5′-ATGAGCCCCAGCCTTCTCCAT -3′(reverse). Filtrating differentially-expressed genes in TC by whole- transcriptome sequence We performed whole-transcriptome sequencing of 19 paired-PTC tissue samples to filtrate differentially- expressed genes. We identified a total of 212 differentially- expressed genes from the 19 paired-PTC tissue samples (log2FC > 1.0, FDR < 0.05 and P-Value < 0.05) (Fig. 1), in- cluding 99 overexpressed and 113 downregulated genes, Fig. 1 Differentially expressed mRNAs in PTC tissue samples and adjacent normal tissue samples are analyzed using hierarchical clustering. Each row represents a single mRNA and each column represents one tissue sample. Red indicates high relative expression and green indicates low relative expression Fig. 1 Differentially expressed mRNAs in PTC tissue samples and adjacent normal tissue samples are analyzed using hierarchical clustering. Each row represents a single mRNA and each column represents one tissue sample. Red indicates high relative expression and green indicates low relative expression Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Page 4 of 8 GABRB2, NGEF and HMGA2). Results revealed that the panel was able to differentiate thyroid nodules excel- lently. 95.8 % (158/165 samples correctly classified) of the samples were correctly classified (Sensibility: 97.0 %, Specificity: 93.8 %, PPV: 96.0 %, NPV: 95.3 %). GABRB2, NGEF and HMGA2). Results revealed that the panel was able to differentiate thyroid nodules excel- lently. 95.8 % (158/165 samples correctly classified) of the samples were correctly classified (Sensibility: 97.0 %, Specificity: 93.8 %, PPV: 96.0 %, NPV: 95.3 %). and then the 99 overexpressed genes were given focus. We raised the FC cut-off value to a higher level (log2FC > 4.0) to obtain more limited genes, which are significantly over- expressed. Only 33 overexpressed genes were identified. To obtain genes that are highly associated with thyroid tumor, we screened out the intersection of these 33 overexpressed genes and adenocarcinoma related genes, including only 12 genes (Fig. 2). Then, we investigated the expression of these 12 overexpressed genes in our 19 paired PTC samples with sequencing data (19 PTC and 19 NT). As shown in Fig. 2, all 12 genes showed significant differential expression be- tween TC and NT, with FN1, GABRB2, NGEF and HMGA2 showing the best diagnostic value. Finally, these four genes with the most diagnostic value were filtrated. The flowchart of our study design is presented in Fig. 3. Training set: establishing the diagnostic panel g g g p To establish diagnostic panel, mRNA expression levels of the four genes were assessed in 165 samples (100 TC and 65 BTL) using RT-qPCR. Result of each gene is shown in Fig. 4a. Each of four genes demonstrated de- cent ability to differentiate malignancy and benignity. Using SVM statistical model, we successfully established diagnostic panel that combines the four genes (FN1, Validation set: validating the panel in external independent samples To validate panel performance, mRNA expression levels of the four genes were measured by RT-qPCR in another 204 samples (123 TC and 81 BTL). Similar to previous data, relative expression level of the four genes were sig- nificantly upregulated (Fig. 4b). The panel also achieved high accuracy in validation set, with 91.7 % accuracy (187/204 samples correctly classified). Only nine malig- nant thyroid tumors (4.4 %) and 8 benign thyroid lesions (3.9 %) were misclassified among all the 204 samples. Panel performance based on RT-qPCR was satisfactory, with 91.7 % sensitivity, 92.7 % sensitivity, 90.1 % specifi- city, 93.4 % PPV and 89.0 % NPV. Discussion Thyroid cancer, particularly PTC, has become the fastest growing type of cancer in recent years [1, 3]. Owing to the Fig. 2 Expression of the 12 overexpressed genes in our sequencing data (19 PTC and 19 NT). P < 0.05; P < 0.001 ig. 2 Expression of the 12 overexpressed genes in our sequencing data (19 PTC and 19 NT). P < 0.05; P < 0.001 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Page 5 of 8 Page 5 of 8 Fig. 3 Flowchart of the study design. Totally 212 differential expressed genes were identitied by using whole-transcriptome sequencing of 19 paired PTC tissue samples. Only 33 genes were identitied when log2FC > 4 and we only focusd on those 33 over expressed genes. Then we got the intersection of those 33 over expressed genes and adenocarcinoma related genes, including 12 genes. we investigated the expression of these 12 overexpressed genes in our 19 paired PTC samples with sequencing data, 4 genes with the most diagnostic value were picked out (FN1, GABRB2, NGEF and HMGA2). High accuracy of diagnosis was both realized in training set and validation set not cost-effective outside of North America [20]. Another panel, named ThyroSeq v2 assay, which consists of more than 50 gene mutations and gene fusions (more than 1,000 hotspots), performed with 90 % sensitivity, 93 % spe- cificity, 83 % PPV, and 96 % NPV by using next-generation sequence [8]. Other studies also achieved encouraging ef- fects [14–19]. However, these diagnostic panels are too complex and costly to apply in clinical practice. Also, they have not yet been proven to be effective in Asian popula- tion. Many studies have shown that genetic background is different between Asians and Western populations. For ex- ample, a gene expression classifier (Afirma) showed dis- tinctly different sensitivity and specificity in different populations partly because of different genetic background [13, 20, 22–25]. What is more, Xing M et al. reported BRAF mutation occurred in about 45 % of PTC patients in the USA [26] while the rate reported in Asian was much higher, 63.7 % in the study from Jin L et al. [27] and 58 % in the study from Lee JH [28]. However, the rate of TERT promoter mutation was much higher in Western popula- tions than Asian population (about 10 % vs 4.1 %) [27, 29]. Discussion Moreover, the incidence of thyroid cancer is also different between Asians and Western populations [3, 4]. In the present study, we developed a panel that presents excellent performance in discriminating benign from ma- lignant thyroid nodules by combining only four genes (FN1, GABRB2, NGEF and HMGA2). Its performance was further validated by independent external samples. To fur- ther confirm its performance in FNAB samples, we also tested the panel in 20 FNAB samples with indeterminate result, which correctly recognized 17 of 20 with 85 % pre- diction accuracy. In spite of the limited sample size, the panel also showed good discriminating ability in FNAB samples. Fig. 3 Flowchart of the study design. Totally 212 differential expressed genes were identitied by using whole-transcriptome sequencing of 19 paired PTC tissue samples. Only 33 genes were identitied when log2FC > 4 and we only focusd on those 33 over expressed genes. Then we got the intersection of those 33 over expressed genes and adenocarcinoma related genes, including 12 genes. we investigated the expression of these 12 overexpressed genes in our 19 paired PTC samples with sequencing data, 4 genes with the most diagnostic value were picked out (FN1, GABRB2, NGEF and HMGA2). High accuracy of diagnosis was both realized in training set and validation set Fig. 3 Flowchart of the study design. Totally 212 differential expressed genes were identitied by using whole-transcriptome sequencing of 19 paired PTC tissue samples. Only 33 genes were identitied when log2FC > 4 and we only focusd on those 33 over expressed genes. Then we got the intersection of those 33 over expressed genes and adenocarcinoma related genes, including 12 genes. we investigated the expression of these 12 overexpressed genes in our 19 paired PTC samples with sequencing data, 4 genes with the most diagnostic value were picked out (FN1, GABRB2, NGEF and HMGA2). High accuracy of diagnosis was both realized in training set and validation set FN1 encodes fibronectin involved in cell adhesion and migration processes, including embryogenesis, wound healing, blood coagulation, host defense and metastasis. This gene was reported to be differentially expressed in human cancers [30]. Some research showed that FN1 could help distinguish benign from malignant thyroid tumors [31, 32], which is consistent with our findings. frequency of medical examinations, clinicians are now faced with an even increasing number of patients with thyroid nodules, most of which, however, are benign [7, 8]. Discussion Avoid- ing unnecessary surgery, especially total thyroidectomies, thereby reducing surgery-associated risks and subsequent treatments, is still an intractable challenge for clinicians. Thus, accurately distinguishing benign from malignant thy- roid nodules preoperatively is important. GABRB2 encodes gamma-aminobutyric acid A recep- tor, beta 2 subunit. This gene was found to be related to mental diseases, like schizophrenia [33, 34]. A recent study revealed that GABRB2 is highly upregulated in thyroid cancer [35]. In our study, GABRB2 was one of the most significant upregulated genes in 203 differential expression genes. Considering this challenge, previous studies have devel- oped several diagnostic panels through high-throughput technologies, like next-generation sequence and micro- array [8, 13–15, 18, 19]. A prospective study evaluated diagnostic molecular test consisting of 167 genes in inde- terminate nodules, which presented high sensitivity (92 %), but low specificity (52 %) [13]. Some additional research reports consistent results in using the classifier [20, 21], while some reported only 17 % malignancy rate confirm- ation in indeterminate nodules [22]. Moreover, the test is NGEF has been found to be associated with myopia and obesity-related diseases [36, 37]. However, it is not yet well studied in human cancer. A study from Korea had demon- strated that NGEF is a differentially-expressed gene in PTC [38]. We also found that NGEF is upregulated and can be used for diagnosis. Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Page 6 of 8 Fig. 4 RT-qPCR analysis of the four genes. a. RT-qPCR analysis of the four genes in training set (100 TC and 65 BTL). b. RT-qPCR analysis of the four genes in validation set (123 TC and 81 BTL). P < 0.05; *P < 0.001 Fig. 4 RT-qPCR analysis of the four genes. a. RT-qPCR analysis of the four genes in training set (100 TC and 65 BTL). b. RT-qPCR analysis of the four genes in validation set (123 TC and 81 BTL). P < 0.05; ***P < 0.001 Additional files HMGA2 encodes protein that belongs to non-histone chromosomal high mobility group (HMG) protein family. This protein contains structural DNA-binding domains and may act as a transcriptional regulating factor. Many studies have confirmed that HMGA2 is important in tumorigenesis and tumor progression [39–43], including thyroid cancer [35, 44]. Chiappetta et al. reported that the mRNA expression of HMGA2 is associated with malignant phenotype in human thyroid neoplasias [44]. HMGA2 was also used as a diagnostic marker in previous studies [14, 45]. In our study, HMGA2 was significantly up- regulated in primary TC tissue, compared with benign and normal thyroid nodules. Additional file 1: Table S1. Clinicopathologic characteristics of 223 thyroid cancer patients enrolled in the study. (DOCX 16 kb) Additional file 2: Table S2. Clinicopathologic characteristics of 146 benign thyroid lesions patients enrolled in the study. (DOCX 14 kb) Additional file 1: Table S1. Clinicopathologic characteristics of 223 thyroid cancer patients enrolled in the study. (DOCX 16 kb) Additional file 2: Table S2. Clinicopathologic characteristics of 146 benign thyroid lesions patients enrolled in the study. (DOCX 14 kb) Conclusions XHZ and OCW contribute to conception and design. QXW, EDC, YFC and LX contribute to data analysis and manuscript writing. QXW, YXJ, WXJ and YHW performed experiments. QL and ZCZ performed data acquisition. All authors reviewed the manuscript and approved the final authorship. QXW, EDC and YFC have contributed equally to this article. In summary, this study successfully established a diagnostic panel by using a simple and affordable method, which uti- lized only four genes with excellent performance in Chinese patients. We believe this panel could form the basis for a robust, practical, and affordable molecular diagnostic tool for clinical use in the near future. Funding u d g This work was supported by a grant from the National High Technology Research and Development Program of 863 project of China (NO. 2012AA02A210) and Major Science and Technology Projects of Zhejiang Province (2015C03G2090019). Abbreviations BTL: Benign thyroid lesions; FC: Fold change; FDR: False discovery rate; FN1: Fibronectin 1; GABRB2: Gamma-aminobutyric acid type A receptor beta 2 subunit; HMGA2: High-mobility group AT-hook 2; NGEF: Neuronal guanine nucleotide exchange factor; NPV: Negative predictive value; NT: Normal thyroid tissues; PPV: Positive predictive value; PTC: Papillary thyroid cancer; SVM: Support vector machine; TC: Thyroid cancer; US-FNAB: Ultrasound- guided fine-needle aspiration biopsy Although we had successfully established a diag- nostic panel with excellent performance, the current study has several limitations. First, based on frozen tissue samples, our panel was only tested in a small FNAB size, so it needs to be validated in FNAB sam- ples with large-sample size. Second, our study was a single-center study; multi-center prospective studies are needed. References Jin L, Chen E, Dong S, Cai Y, Zhang X, Zhou Y, Zeng R, Yang F, Pan C, Liu Y, Wu W, Xing M, Zhang X, Wang O. BRAF and TERT promoter mutations in the aggressiveness of papillary thyroid carcinoma: a study of 653 patients. Oncotarget. 2016;7:18346–55. 8. Nikiforov YE, Carty SE, Chiosea SI, Coyne C, Duvvuri U, Ferris RL, Gooding WE, Hodak SP, LeBeau SO, Ohori NP, Seethala RR, Tublin ME, Yip L, Nikiforova MN. Highly accurate diagnosis of cancer in thyroid nodules with follicular neoplasm/suspicious for a follicular neoplasm cytology by ThyroSeq v2 next-generation sequencing assay. Cancer. 2014;120:3627–34. 28. Lee JH, Lee ES, Kim YS, Won NH, Chae YS. BRAF mutation and AKAP9 expression in sporadic papillary thyroid carcinomas. Pathology. 2006;38:201–4. 29. Liu X, Qu S, Liu R, Sheng C, Shi X, Zhu G, Murugan AK, Guan H, Yu H, Wang Y, Sun H, Shan Z, Teng W, Xing M. TERT promoter mutations and their association with BRAF V600E mutation and aggressive clinicopathological characteristics of thyroid cancer. J Clin Endocrinol Metab. 2014;99:E1130–6. 9. Schneider DF, Chen H. New developments in the diagnosis and treatment of thyroid cancer. CA Cancer J Clin. 2013;63:374–94. 9. Schneider DF, Chen H. New developments in the diagnosis and treatment of thyroid cancer. CA Cancer J Clin. 2013;63:374–94. 10. Xing M, Haugen BR, Schlumberger M. Progress in molecular-based management of differentiated thyroid cancer. Lancet. 2013;381:1058–69. 10. Xing M, Haugen BR, Schlumberger M. Progress in molecular-based management of differentiated thyroid cancer. Lancet. 2013;381:1058–69. 30. Helleman J, Jansen MP, Ruigrok-Ritstier K, van Staveren IL, Look MP, Meijer-van Gelder ME, Sieuwerts AM, Klijn JG, Sleijfer S, Foekens JA, Berns EM. Association of an extracellular matrix gene cluster with breast cancer prognosis and endocrine therapy response. Clin Cancer Res. 2008;14:5555–64. 11. Wang CC, Friedman L, Kennedy GC, Wang H, Kebebew E, Steward DL, Zeiger MA, Westra WH, Wang Y, Khanafshar E, Fellegara G, Rosai J, Livolsi V, Lanman RB. A large multicenter correlation study of thyroid nodule cytopathology and histopathology. Thyroid. 2011;21:243–51. 31. Fryknas M, Wickenberg-Bolin U, Goransson H, Gustafsson MG, Foukakis T, Lee JJ, Landegren U, Hoog A, Larsson C, Grimelius L, Wallin G, Pettersson U, Isaksson A. Molecular markers for discrimination of benign and malignant follicular thyroid tumors. Tumour Biol. 2006;27:211–20. 12. References 1. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM. Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res. 2014;74:2913–21. 22. McIver B, Castro MR, Morris JC, Bernet V, Smallridge R, Henry M, Kosok L, Reddi H. An independent study of a gene expression classifier (Afirma) in the evaluation of cytologically indeterminate thyroid nodules. J Clin Endocrinol Metab. 2014;99:4069–77. 2. Chen AY, Jemal A, Ward EM. Increasing incidence of differentiated thyroid cancer in the United States, 1988–2005. Cancer. 2009;115:3801–7. 2. Chen AY, Jemal A, Ward EM. Increasing incidence of differentiated thyroid cancer in the United States, 1988–2005. Cancer. 2009;115:3801–7. 3. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. 2014;64:9–29. 23. Brauner E, Holmes BJ, Krane JF, Nishino M, Zurakowski D, Hennessey JV, Faquin WC, Parangi S. Performance of the Afirma Gene Expression Classifier in Hurthle Cell Thyroid Nodules Differs from Other Indeterminate Thyroid Nodules. Thyroid. 2015;25:789–96. 4. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, Jemal A, Yu XQ, He J. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66:115–32. 5. Xie SH, Chen J, Zhang B, Wang F, Li SS, Xie CH, Tse LA, Cheng JQ. Time trends and age-period-cohort analyses on incidence rates of thyroid cancer in Shanghai and Hong Kong. BMC Cancer. 2014;14:975. y 24. Lastra RR, Pramick MR, Crammer CJ, LiVolsi VA, Baloch ZW. Implications of a suspicious afirma test result in thyroid fine-needle aspiration cytology: an institutional experience. Cancer Cytopathol. 2014;122:737–44. g g g 6. Xing M. Molecular pathogenesis and mechanisms of thyroid cancer. Nat Rev Cancer. 2013;13:184–99. 25. Yang SE, Sullivan PS, Zhang J, Govind R, Levin MR, Rao JY, Moatamed NA. Has Afirma gene expression classifier testing refined the indeterminate thyroid category in cytology? Cancer Cytopathol. 2016;124:100–9. 7. Gharib H, Papini E, Paschke R, Duick DS, Valcavi R, Hegedus L, Vitti P, Nodules AAETFoT. American Association of Clinical Endocrinologists, Associazione Medici Endocrinologi, and European Thyroid Association Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules: executive summary of recommendations. Endocr Pract. 2010;16:468–75. y g y y gy y p 26. Xing M. BRAF mutation in thyroid cancer. Endocr Relat Cancer. 2005;12:245–62. 27. References Hundahl SA, Cady B, Cunningham MP, Mazzaferri E, McKee RF, Rosai J, Shah JP, Fremgen AM, Stewart AK, Holzer S, Hundahl SA, Cady B, Cunningham MP, Mazzaferri E, McKee RF, Rosai J, Shah JP, Fremgen AM, Stewart AK, Holzer S. Initial results from a prospective cohort study of 5583 cases of thyroid carcinoma treated in the united states during 1996. U.S. and German Thyroid Cancer Study Group. An American College of Surgeons Commission on Cancer Patient Care Evaluation study. Cancer. 2000;89:202–17. 32. da Silveira Mitteldorf CA, de Sousa-Canavez JM, Leite KR, Massumoto C, Camara-Lopes LH. FN1, GALE, MET, and QPCT overexpression in papillary thyroid carcinoma: molecular analysis using frozen tissue and routine fine- needle aspiration biopsy samples. Diagn Cytopathol. 2011;39:556–61. 33. Tsang SY, Zhong S, Mei L, Chen J, Ng SK, Pun FW, Zhao C, Jing B, Chark R, Guo J, Tan Y, Li L, Wang C, Chew SH, Xue H. Social cognitive role of schizophrenia candidate gene GABRB2. PLoS One. 2013;8, e62322. 13. Alexander EK, Kennedy GC, Baloch ZW, Cibas ES, Chudova D, Diggans J, Friedman L, Kloos RT, LiVolsi VA, Mandel SJ, Raab SS, Rosai J, Steward DL, Walsh PS, Wilde JI, Zeiger MA, Lanman RB, Haugen BR. Preoperative diagnosis of benign thyroid nodules with indeterminate cytology. N Engl J Med. 2012;367:705–15. g , g , , , g , , , g , , Guo J, Tan Y, Li L, Wang C, Chew SH, Xue H. Social cognitive role of schizophrenia candidate gene GABRB2. PLoS One. 2013;8, e62322. 34. Srivastava S, Cohen J, Pevsner J, Aradhya S, McKnight D, Butler E, Johnston M, Fatemi A. A novel variant in GABRB2 associated with intellectual disability and epilepsy. Am J Med Genet A. 2014;164A:2914–21. 34. Srivastava S, Cohen J, Pevsner J, Aradhya S, McKnight D, Butler E, Johnston M, Fatemi A. A novel variant in GABRB2 associated with intellectual disability and epilepsy. Am J Med Genet A. 2014;164A:2914–21. 14. Barros-Filho MC, Marchi FA, Pinto CA, Rogatto SR, Kowalski LP. High Diagnostic Accuracy Based on CLDN10, HMGA2, and LAMB3 Transcripts in Papillary Thyroid Carcinoma. J Clin Endocrinol Metab. 2015;100:E890–9. 35. Wu ZY, Wang SM, Chen ZH, Huv SX, Huang K, Huang BJ, Du JL, Huang CM, Peng L, Jian ZX, Zhao G. MiR-204 regulates HMGA2 expression and inhibits cell proliferation in human thyroid cancer. Cancer Biomark. 2015;15:535–42. 15. Consent for publication 16. Cerutti JM, Delcelo R, Amadei MJ, Nakabashi C, Maciel RM, Peterson B, Shoemaker J, Riggins GJ. A preoperative diagnostic test that distinguishes benign from malignant thyroid carcinoma based on gene expression. J Clin Invest. 2004;113:1234–42. Informed consent for the scientific use of biological material was obtained from each patient. 17. Zheng B, Liu J, Gu J, Lu Y, Zhang W, Li M, Lu H. A three-gene panel that distinguishes benign from malignant thyroid nodules. Int J Cancer. 2015;136:1646–54. Competing interests Th h d l h The authors declare that they have no competing interests. The authors declare that they have no competing interests. Page 7 of 8 Page 7 of 8 Page 7 of 8 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 Received: 29 July 2016 Accepted: 22 October 2016 21. Duick DS, Klopper JP, Diggans JC, Friedman L, Kennedy GC, Lanman RB, McIver B. The impact of benign gene expression classifier test results on the endocrinologist-patient decision to operate on patients with thyroid nodules with indeterminate fine-needle aspiration cytopathology. Thyroid. 2012;22:996–1001. The work was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. 18. Vriens MR, Weng J, Suh I, Huynh N, Guerrero MA, Shen WT, Duh QY, Clark OH, Kebebew E. MicroRNA expression profiling is a potential diagnostic tool for thyroid cancer. Cancer. 2012;118:3426–32. Author details 1 f 1Department of Oncology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province 325000, China. 2Department of 1Department of Oncology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, Zhejiang Province 325000, China. 2Department of 19. Liu Y, Cope L, Sun W, Wang Y, Prasad N, Sangenario L, Talbot K, Somervell H, Westra W, Bishop J, Califano J, Zeiger M, Umbricht C. DNA copy number variations characterize benign and malignant thyroid tumors. J Clin Endocrinol Metab. 2013;98:E558–66. Otolaryngology Head and Neck Surgery, Xinhua Hospital, Shanghai Jiaotong Otolaryngology Head and Neck Surgery, Xinhua Hospital, Shanghai Jiaotong University, School of Medicine, Shanghai 200000, China. Otolaryngology Head and Neck Surgery, Xinhua Hospital, Shanghai Jiaotong University, School of Medicine, Shanghai 200000, China. University, School of Medicine, Shanghai 200000, China. 20. Alexander EK, Schorr M, Klopper J, Kim C, Sipos J, Nabhan F, Parker C, Steward DL, Mandel SJ, Haugen BR. Multicenter clinical experience with the Afirma gene expression classifier. J Clin Endocrinol Metab. 2014;99:119–25. Received: 29 July 2016 Accepted: 22 October 2016 Received: 29 July 2016 Accepted: 22 October 2016 Ethics approval and consent to participate The work was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. References Keutgen XM, Filicori F, Crowley MJ, Wang Y, Scognamiglio T, Hoda R, Buitrago D, Cooper D, Zeiger MA, Zarnegar R, Elemento O, Fahey 3rd TJ. A panel of four miRNAs accurately differentiates malignant from benign indeterminate thyroid lesions on fine needle aspiration. Clin Cancer Res. 2012;18:2032–8. 36. Guo L, Frost MR, Siegwart Jr JT, Norton TT. Scleral gene expression during recovery from myopia compared with expression during myopia development in tree shrew. Mol Vis. 2014;20:1643–59. Page 8 of 8 Page 8 of 8 Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 37. Kim HJ, Park JH, Lee S, Son HY, Hwang J, Chae J, Yun JM, Kwon H, Kim JI, Cho B. A Common Variant of NGEF Is Associated with Abdominal Visceral Fat in Korean Men. PLoS One. 2015;10, e0137564. 38. Chung KW, Kim SW, Kim SW. Gene expression profiling of papillary thyroid carcinomas in Korean patients by oligonucleotide microarrays. J Korean Surg Soc. 2012;82:271–80. 39. Shi Z, Li X, Wu D, Tang R, Chen R, Xue S, Sun X. Silencing of HMGA2 suppresses cellular proliferation, migration, invasion, and epithelial- mesenchymal transition in bladder cancer. Tumour Biol. 2016;37:7515–23. 40. Zhao XP, Zhang H, Jiao JY, Tang DX, Wu YL, Pan CB. Overexpression of HMGA2 promotes tongue cancer metastasis through EMT pathway. J Transl Med. 2016;14:26. 41. D’Armiento J, Shiomi T, Marks S, Geraghty P, Sankarasharma D, Chada K. Mesenchymal Tumorigenesis Driven by TSC2 Haploinsufficiency Requires HMGA2 and Is Independent of mTOR Pathway Activation. Cancer Res. 2016;76:844–54. 42. Kumar MS, Armenteros-Monterroso E, East P, Chakravorty P, Matthews N, Winslow MM, Downward J. HMGA2 functions as a competing endogenous RNA to promote lung cancer progression. Nature. 2014;505:212–7. 43. Natarajan S, Begum F, Gim J, Wark L, Henderson D, Davie JR, Hombach- Klonisch S, Klonisch T. High Mobility Group A2 protects cancer cells against telomere dysfunction. Oncotarget. 2016;7:12761–82. 44. Chiappetta G, Ferraro A, Vuttariello E, Monaco M, Galdiero F, De Simone V, Califano D, Pallante P, Botti G, Pezzullo L, Pierantoni GM, Santoro M, Fusco A. HMGA2 mRNA expression correlates with the malignant phenotype in human thyroid neoplasias. Eur J Cancer. 2008;44:1015–21. 45. Lappinga PJ, Kip NS, Jin L, Lloyd RV, Henry MR, Zhang J, Nassar A. HMGA2 gene expression analysis performed on cytologic smears to distinguish benign from malignant thyroid nodules. Cancer Cytopathol. 2010;118:287–97. Wang et al. Journal of Experimental & Clinical Cancer Research (2016) 35:169 References • 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: • 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: Submit your next manuscript to BioMed Central and we will help you at every step:
https://openalex.org/W4220969813	https://iris.unipa.it/bitstream/10447/549233/2/Antibiotics%202022b.pdf	English	null	Antimicrobial Prophylaxis in Neonates and Children Undergoing Dental, Maxillo-Facial or Ear-Nose-Throat (ENT) Surgery: A RAND/UCLA Appropriateness Method Consensus Study	Antibiotics	2,022	cc-by	14,083	Citation: Rigotti, E.; Bianchini, S.; Nicoletti, L.; Monaco, S.; Carrara, E.; Opri, F.; Opri, R.; Caminiti, C.; Donà, D.; Giuffré, M.; et al. Antimicrobial Prophylaxis in Neonates and Children Undergoing Dental, Maxillo-Facial or Ear-Nose-Throat (ENT) Surgery: A RAND/UCLA Appropriateness Method Consensus Study. Antibiotics 2022, 11, 382. https://doi.org/10.3390/ antibiotics11030382 Academic Editor: Albert Figueras Received: 4 February 2022 Accepted: 10 March 2022 Published: 13 March 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Copyright: © 2022 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/). 4 Research and Innovation Unit, University Hospital of Parma, 43126 Parma, Italy; ccaminiti@ao.pr.it 5 Division of Paediatric Infectious Diseases, Department for Woman and Child Health, University of Padua, 35100 Padua, Italy; daniele.dona@unipd.it Citation: Rigotti, E.; Bianchini, S.; Nicoletti, L.; Monaco, S.; Carrara, E.; Opri, F.; Opri, R.; Caminiti, C.; Donà, D.; Giuffré, M.; et al. Antimicrobial Prophylaxis in Neonates and Children Undergoing Dental, Maxillo-Facial or Ear-Nose-Throat (ENT) Surgery: A RAND/UCLA Appropriateness Method Consensus Study. Antibiotics 2022, 11, 382. https://doi.org/10.3390/ antibiotics11030382 Academic Editor: Albert Figueras Received: 4 February 2022 Accepted: 10 March 2022 Published: 13 March 2022 6 Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties “G. D’Alessandro”, University of Palermo, 90134 Palermo, Italy; mario.giuffre@unipa.it 7 General Surgery Department, Bambino Gesu Children’s Hospital, Istituto di Ricerca e Cura a Carattere Scientiﬁco (IRCCS), 00165 Rome, Italy; alessandro.inserra@opbg.net 7 General Surgery Department, Bambino Gesu Children’s Hospital, Istituto Scientiﬁco (IRCCS), 00165 Rome, Italy; alessandro.inserra@opbg.net 8 Paediatric and Infectious Disease Unit, Academic Department of Pediatrics, IRCCS Bambino Gesù Children’s Hospital, 00165 Rome, Italy; laura.lancella@opbg.net (L.L.); alberto.villani@opbg.net (A.V.) p , , y; p g ( ); p g ( ) 9 Department of Neurosciences, Psychology, Drug Research and Child Health, Section of Pharmacology and Toxicology, University of Florence, Viale G. Antimicrobial Prophylaxis in Neonates and Children Undergoing Dental, Maxillo-Facial or Ear-Nose-Throat (ENT) Surgery: A RAND/UCLA Appropriateness Method Consensus Study Erika Rigotti 1, Sonia Bianchini 2, Laura Nicoletti 2, Sara Monaco 2, Elena Carrara 3, Francesca Opri 1 , Roberta Opri 1 , Caterina Caminiti 4 , Daniele Donà 5, Mario Giuffré 6 , Alessandro Inserra 7 , Laura Lancella 8, Alessandro Mugelli 9, Giorgio Piacentini 1, Nicola Principi 10 , Simonetta Tesoro 11, Elisabetta Venturini 12 , Annamaria Staiano 13, Alberto Villani 8 , Enrico Sesenna 14, Claudio Vicini 15, Susanna Esposito 2,* and on behalf of the Peri-Operative Prophylaxis in Neonatal and Paediatric Age (POP-NeoPed) Study Group † 1 Pediatric Unit, Department of Surgical Sciences, Dentistry, Gynecology and Pediatrics, University of Verona, 37124 Verona, Italy; erika.rigotti@aovr.veneto.it (E.R.); opri.francesca@gmail.com (F.O.); roberta.opri@gmail.com (R.O.); giorgio.piacentini@univr.it (G.P.) 2 Pediatric Clinic, University Hospital, Department of Medicine and Surgery, University of Parma, 43126 Parma, Italy; bianchini.sonia@outlook.it (S.B.); laura.nicoletti@studenti.unipr.it (L.N.); s.monaco1410@gmail.com (S.M.) 3 Infectious Diseases Section, Department of Diagnostics and Public Health, University of Verona, 37134 Verona, Italy; elena.carrara@univr.it 4 antibiotics antibiotics antibiotics antibiotics antibiotics antibiotics 1. Introduction Surgical site infections (SSIs) represent a potential complication in any type of surgical procedure, being associated with prolonged hospital stays and increased postoperative mortality rates, and consequently have a signiﬁcant medical, social and economic impact [1]. This explains why many if not all surgical procedures have been associated with the administration of antibiotics potentially effective against bacteria that could be responsible for SSIs. In fact, the purpose of reducing the risks of developing SSIs for many years far outweighed the consideration of how effective prophylaxis actually was in various surgical situations and the risks associated with antibiotic use [2]. Hence, the wide use of surgical antibiotic prophylaxis (SAP) in conditions in which this is unnecessary, the use of drugs not suitable for the bacteria potentially present and the prolongation of prophylaxis for much longer than necessary contribute to the emergence of bacterial resistance and excessive health care costs. Moreover, the abuse and misuse of antibiotics for SAP could have consequences that in some cases are more severe than the risk of infection and the emergence of major medical problems which might also prolong hospital stays. The onsets of acute renal failure and antibiotic-associated diarrhoea (i.e., Clostridium difﬁcile colitis) remain some of the most signiﬁcant examples in this regard [3]. The awareness of these problems has led many experts to revise the methods of administrating SAP and to draw up guidelines to rationalize its use in each type of surgery. However, at present, not all aspects of the problem have been precisely deﬁned. For many types of surgery, there is a lack of data derived from controlled clinical trials necessary to deﬁne with certainty whether and how to perform SAP. This explains why medical behaviors remain very heterogeneous and often different from what is suggested by ofﬁcial recommendations [4,5]. gg y Many of the recommendations that can be debated concern surgery that is considered clean/contaminated because it involves organs that are normally colonized by bacteria. Dental, maxillo-facial and ear-nose-throat (ENT) surgeries are among these because the mouth and the ﬁrst respiratory tracts are normally colonized by a rich bacterial ﬂora, includ- ing potential pathogens that can result in SSIs [6,7]. The known limitations are even more evident for children. Pediatric studies of SAP are extremely hard to come by. Pieraccini, 6, 50139 Florence, Italy; alessandro.mugelli@uniﬁ.it 10 Università degli Studi di Milano, 20122 Milan, Italy; nicola.principi@unimi.it sion of Anesthesia, Analgesia, and Intensive Care, Department of Surgical and Biomedical Sciences, 11 Division of Anesthesia, Analgesia, and Intensive Care, Department of Surgical and Biomedical Sciences, University of Perugia, 06129 Perugia, Italy; simonettatesoro@gmail.com Division of Anesthesia, Analgesia, and Intensive Care, Department of Surg University of Perugia, 06129 Perugia, Italy; simonettatesoro@gmail.com 12 Pediatric Infectious Disease Unit, Meyer Children’s Hospital, 50139 Florence, Italy; elisabetta.venturini@meyer.it 13 Department of Translational Medical Science, Section of Pediatrics, University of Naples “Federico II”, 80138 Naples, Italy; staiano@unina.it 14 Maxillo-Facial Surgery Unit, Head and Neck Department, University Hospital of Parma, 43126 Parma, Italy; enrico.sesenna@unipr.it 15 Head-Neck and Oral Surgery Unit, Department of Head-Neck Surgery, Otolaryngology, Morgagni Piertoni Hospital, 47121 Forli, Italy; claudio@claudiovicini.com Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. * Correspondence: susannamariaroberta.esposito@unipr.it; Tel.: +39-0521-903524 † Membership of the Peri-Operative Prophylaxis in Neonatal and Paediatric Age (POP-NeoPed) Study Group is provided in the Acknowledgements. Abstract: Surgical site infections (SSIs) represent a potential complication in surgical procedures, mainly because clean/contaminated surgery involves organs that are normally colonized by bacteria. Dental, maxillo-facial and ear-nose-throat (ENT) surgeries are among those that carry a risk of SSIs because the mouth and the ﬁrst respiratory tracts are normally colonized by a bacterial ﬂora. The aim of this consensus document was to provide clinicians with recommendations on surgical antimicrobial prophylaxis in neonates (<28 days of chronological age) and pediatric patients (within the age range of 29 days–18 years) undergoing dental, maxillo-facial or ENT surgical procedures. These included: (1) dental surgery; (2) maxilla-facial surgery following trauma with fracture; (3) temporo-mandibular surgery; (4) cleft palate and cleft lip repair; (5) ear surgery; (6) endoscopic paranasal cavity surgery Copyright: © 2022 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/). https://www.mdpi.com/journal/antibiotics Antibiotics 2022, 11, 382. https://doi.org/10.3390/antibiotics11030382 Antibiotics 2022, 11, 382 2 of 19 and septoplasty; (7) clean head and neck surgery; (8) clean/contaminated head and neck surgery and (9) tonsillectomy and adenoidectomy. Due to the lack of pediatric data for the majority of dental, maxillo-facial and ENT surgeries and the fact that the recommendations for adults are currently used, there is a need for ad hoc studies to be rapidly planned for the most deﬁcient areas. This seems even more urgent for interventions such as those involving the ﬁrst airways since the different composition of the respiratory microbiota in children compared to adults implies the possibility that surgical antibiotic prophylaxis schemes that are ideal for adults may not be equally effective in children. Keywords: dental surgery; ENT surgery; head and neck surgery; maxilla-facial surgery; surgical antimicrobial prophylaxis 1. Introduction Moreover, the respiratory microbiota in children is different from that of adults and susceptibility to respiratory infection varies signiﬁcantly in different pediatric ages [8]. Therefore, the aim of this consensus document was to provide clinicians with recommendations on SAP in neonates (<28 days of chronological age) and pediatric patients (within the age range of 29 days–18 years) undergoing dental, maxillo-facial or ENT surgical procedures. This consensus may lead to two advantages: to clarify if and how to perform SAP in these subjects and to indicate which ﬁelds deserve signiﬁcant in-depth studies and new research. 2.1. RAND/UCLA Appropriateness Method The consensus document was realized using the Research and Development Corpo- ration (RAND) and the University of California—Los Angeles (UCLA) appropriateness Antibiotics 2022, 11, 382 3 of 19 method. The RAND/UCLA method consists of the evaluation of the appropriateness of diagnostic and therapeutic procedures with sub-optimal scientiﬁc evidence by a panel of experts [9]. According to the RAND method, a procedure is deﬁned as “appropriate” if the expected beneﬁts outweigh the expected negative consequences, with a wide margin that justiﬁes it, regardless of the costs. On the contrary, a procedure whose expected risks outweigh the expected beneﬁts is considered as “inappropriate”. According to the RAND deﬁnition, the expert who makes an appropriateness/inappropriateness judgment must consider the clinical beneﬁts and not be inﬂuenced by economic considerations. There- fore, the appropriateness accounts for the evaluation of the risk/beneﬁt ratio of a list of management and therapeutic procedures [10]. For a heterogeneous topic such as surgical antimicrobial prophylaxis on which randomized controlled trials in pediatrics are lacking, the application of methods aiming to increase the homogeneity of behaviors by neonatolo- gists, infectious diseases specialists, pediatric surgeons and anesthetists appeared useful and appropriate. For this reason, the RAND/UCLA approach was chosen instead of the GRADE methodology. Through the RAND method, the participants discussed different clinical scenarios and elaborated statements on the basis of the literature and their clinical experience. The group of experts did not consider it appropriate to combine the GRADE method with the RAND/UCLA approach because the absence of randomized studies represents a bias in deﬁning the strength of the recommendations and in representing a consensus reached for real-life situations. 2.2. Recruitment of the Expert Panel A multidisciplinary group of experts belonging to the main Italian scientiﬁc societies dealing with anti-infective therapy in pediatric ages was selected. The following scientiﬁc societies were involved: the Italian Society of Pediatrics (SIP), the Italian Society of Neona- tology (SIN), the Italian Society of Pediatric Infectious Diseases (SITIP), the Italian Society of Infectious and Tropical Diseases (SIMIT), the Italian Society of Pediatric Surgery (SICP), the Italian Society of Microbiology (SIM), the Italian Society of Pharmacology (SIF), the Italian Society of Anesthesia and Neonatal and Pediatric Resuscitation (SARNEPI) and the Italian Society of Childhood Respiratory Diseases (SIMRI). The panel of experts was made up of 52 medical doctors with at least 5-years of experience: pediatricians (n = 20), neonatologists (n = 6), infectious diseases specialists (n = 5), pediatric surgeons (n = 3), a maxillo-facial surgeon (n = 1), an otolaryngologist surgeon (n = 1), anesthetists (n = 8), pharmacologists (n = 5) and microbiologists (n = 3). 2.3. Generation of Scenarios Initially, a literature search was performed with the selection of documents including randomized studies, systematic reviews of the literature, meta-analyses and guidelines on peri-operative prophylaxis for the prevention of SSI in neonatal and pediatric den- tal, maxillo-facial or ENT surgery. The literature search was carried out on the PubMed Database, with only articles published in English from the year 2000 to 2020 being cho- sen. The following key terms were used: “antimicrobial prophylaxis” OR “antibiotic prophylaxis” AND “dental surgery” OR “dental” OR “teeth” OR “gum tissue” OR “oral mucosa” OR “maxillo-facial” OR “mandible” OR “maxillary” or “zygomatic” OR “temporo- manidibular” OR “cleft palate” OR “cleft lip” OR “ear surgery” OR “tympanostomy” OR “tympanoplasty” OR “stapedectomy” OR “ear tube placement” OR “cochlear implant” “paranasal surgery” OR “rhinosinus surgery” OR “head surgery” OR “neck surgery” OR “thyroidectomy” OR “parathyroidectomy” OR “salivary gland surgery” OR “parotidec- tomy” OR “submandibular gland excision” OR “laryngectomy” OR “pharyngectomy” OR “tracheotomy” OR “neck dissection” OR “lymphangiomas exeresis” OR “neck cysts excision” or “ﬁstulas excision” OR “laser airway surgery” OR “tonsillectomy” OR “ade- noidectomy” OR “adenotonsillectomy” OR “septoplasty” AND “neonate” OR “new- born” OR “paediatric” OR “pediatric” OR “children” OR “adolescent”. Subsequently, using the Patient/Problem/Population-Intervention-Comparison/Control/Comparator- Antibiotics 2022, 11, 382 4 of 19 4 of 19 Outcome (PICO) model (i.e., deﬁning a clinical question in terms of the speciﬁc patient problem), a questionnaire was created on SAP in neonatal and pediatric dental, maxillo- facial and ENT surgery, and was divided into nine clinical scenarios. All neonatal and pediatric dental, maxillo-facial and ENT surgical procedures were considered. Before ad- ministration, the questionnaire was tested twice with a one-week interval to a convenience sample of four pediatricians, two neonatologists, one infectious diseases specialist, one pe- diatric surgeon, one maxillo-facial surgeon, one otolaryngologist surgeon, one anesthetist, one pharmacologist and one microbiologist. Then, 26 out of 52 experts were selected by the scientiﬁc societies and the questionnaire was administered to 11 pediatricians, 3 neona- tologists, 2 infectious diseases specialists, 1 pediatric surgeon, 1 maxillo-facial surgeon, 1 otolaryngologist surgeon, 4 anesthetists, 2 pharmacologists and 1 microbiologist. 2.4. Two-Round Consensus Process On the basis of the scenarios, the questionnaire was submitted to the experts on the “REDCap” online platform. Each question included the clinical scenario and possible answers relating to whether or not SAP was recommended for the scenario, and, in case of its recommendation, a list of all the antibiotics available on the EU market was included so that the expert could select the antibiotics that he/she considered as their ﬁrst choice. The selected bibliographic material was made available to all panel members, who were instructed on how to ﬁll in the panel. The experts answered the questionnaire anonymously and their judgments were expressed on a 1–9 scale, where “1” was considered deﬁnitely inappropriate, “5” was considered uncertain and 9 was considered deﬁnitely appropriate. Intermediate values corresponded to different modulations in the judgment in terms of inappropriateness (“2” and “3”), uncertainty (from “4” to “6”) and appropriateness (“7” and “8”), respectively. When evaluating each indication, each expert could refer to both his/her own experience and clinical judgment and the available scientiﬁc evidence. A free space was provided for any annotations or comments. p p y The ﬁrst round of the questionnaire was blind to other panel members. The results of the survey were discussed in a collegial meeting in order to ﬁnd agreements and reduce eventual disagreements. Clariﬁcations, adaptations and reﬁnements of the indications and appropriateness ratings were made. A total of nine recommendations were developed. Participants were asked to approve the recommendations in a second round during the following four weeks. During round two, the level of consensus within the panel for each scale for each scenario was calculated in real-time. Mean values and disagreements were classiﬁed in terms of three levels of appropriateness (appropriate: between ‘7’ and ‘9’, without disagreement; uncertain: between ‘4’ and ‘6’ or any median with disagreement; inappropriate: between ‘1’ and ‘3’, with agreement). Agreement was reached when at least 75% of participants ranked within the same level of appropriateness. 3.1. SCENARIO #1. Dental Surgery For years, it was believed that dental surgery, in addition to carrying the risk of local infections at the site of surgery, could lead to bacteremia and, consequently, could promote the development of distant infections [11]. The pathogens responsible for these problems were identiﬁed mainly in viridans streptococcal species, followed by Staphylococcus aureus and Enterococcus spp. [12]. Over time, it has been shown that the risk of bacteremia is very low, can be taken into account only for subjects with cardiological or orthopedic problems and can be considered marginal in all other cases. The acquisition of new information has modiﬁed the recommendations of scientiﬁc societies on SAP for subjects undergoing dental surgery, becoming increasingly restrictive. With regard to the prevention of bacterial endocarditis in subjects with heart disease, it is now accepted by most scientiﬁc societies that the risk of developing endocarditis following dental surgery is lower than that which occurs when brushing teeth, chewing gum for hours or using toothpicks [13,14]. Moreover, it seems to be well established Antibiotics 2022, 11, 382 5 of 19 5 of 19 that the beneﬁts of administering antibiotics prior to dental surgery are minimal or even non-existent and may, in any case, not justify the harm associated with the use of these drugs [13,14]. A study conducted in Taiwan that monitored the development of bacterial endocarditis in the entire population over 10 years and attempted to correlate it with the performance of dental surgery was unable to demonstrate any relationship between the two variables [15]. Hence, the authors recommended that the proper and continuous cleaning of the mouth and teeth was maintained as a basic element to reduce the risk of bacterial endocarditis, thus avoiding any form of SAP in dental surgery. However, uncertainties remain with regard to a selected group of patients undergoing speciﬁc forms of dental surgery. The American Heart Association [16] and the American Academy of Pediatric Dentistry [17] suggest that SAP may be considered when the surgery involves the manipulation of the gum tissue or periapical region of the teeth or involves the perforation of the oral mucosa. 3.1. SCENARIO #1. Dental Surgery This is indicated, in particular, if the subject has already suffered from endocarditis, has already been operated on with the application of prosthetic material, has a cyanogenic congenital heart disease that has not yet completely repaired, has a congenital heart disease that has already been repaired with the application of prosthetic material in the 6 months following surgery and has been transplanted and has developed valvulopathy. When necessary, it is recommended that SAP should be conducted prior to the performance of surgery without further continuation of antibiotic administration. Amoxicillin per os, ampicillin or cefazolin ev or, in subjects with a penicillin allergy, cephalexin per os or clindamycin ev are considered the drugs of choice [13]. gy p p y g The problem of SAP is equally or even more controversial in dental procedures among patients with orthopedic problems, particularly those with joint prostheses. Also, in cases such as these, the increased risk of infection and the usefulness of SAP for the prevention of prosthetic infections have been overestimated for years. In fact, the most recent and best-performed studies seem to deny any relationship between dental surgery and the development of prosthetic infection [18–20]. This explains why the most recent guidelines [21–24] do not recommend SAP for dental procedures in individuals with joint prostheses. Despite this, the majority of dentists and orthopedists continue to use antibiotic prophylaxis, ignoring the recommendations of scientiﬁc societies [25,26]. Recommendation 1. In the case of a pediatric patient undergoing dental surgery, no perioperative antibiotic prophylaxis is recommended. Oral amoxicillin or ampicillin ev (50 mg/kg for both) should be administered during the 30 min before surgery if the operation involves the manipulation of gum tissue or the periapical region of the teeth or involves the perforation of the oral mucosa and the subject has already suffered from endocarditis, has already been operated on with the application of prosthetic material, has a cyanogenic congenital heart disease not yet fully repaired, has a congenital heart disease already repaired with the application of prosthetic material in the 6 months following surgery or has been transplanted and has developed valvulopathy. No prophylaxis is recommended in subjects with prosthetic implants. 3.2. Maxillo-Facial Surgery Several forms of surgery fall under the category of maxillo-facial surgery. The use- fulness of SAP in these cases can vary considerably depending on the type of surgical procedure [27]. 3.2.1. SCENARIO #2. SURGERY Following Trauma with Fracture 3.2.2. SCENARIO #3. Temporo-Mandibular Surgery Arthroplasty, condylectomy and, in recent years in pediatrics, total joint replacement surgery all fall into this group. The risk of SSIs is calculated at 1.5–4.5%, with S. aureus, S. epidermidis and Peptostreptococcus spp. as the main pathogens. Despite lacking reliable demonstrations of the beneﬁts of SAP, some authors suggest the administration of an- tibiotics active against these bacteria at least pre-operatively [30–32]. Intravenous (EV) cefazolin is recommended in these cases. Orthognathic surgery includes reconstructive surgery, maxillomandibular advance- ment and surgical correction of facial asymmetry. Corticotomy, osteotomy and the place- ment of distraction devices for the treatment of Pierre Robin syndrome are the best examples of children [33]. Studies regarding prophylactic antibiotic use in this setting are few and the results are conﬂicting. Those with the lowest risk of bias seem to indicate that preoperative antibiotics can be effective, whereas postoperative antibiotics have no role to play [34,35]. Recommendation 3. In the case of a pediatric patient undergoing temporo-mandibular surgery, pre-operative antibiotic prophylaxis with cefazolin in a single dose of 30 mg/kg (maximum dose 2 g) EV is recommended in the 30 min before surgery. 3.2.1. SCENARIO #2. SURGERY Following Trauma with Fracture A number of studies have shown that the risk of SSIs depends on the type and site of the fracture, with it being greater if the bone fracture is in communication with the oral cavity or skin surface and involves the mandible rather than the maxilla. Studies evaluating the usefulness of SAP have shown that the pre-operative administration of antibiotics may be useful in surgeries involving the mandible, while there are no beneﬁts in those involving the upper and middle portions of the face [28,29]. In any case, post-operative prophylaxis is unnecessary. These conclusions derive from two meta-analyses of studies conducted Antibiotics 2022, 11, 382 6 of 19 almost exclusively in adults. In the ﬁrst one, which analyzed four studies published before 2006, it was shown that the infectious risk is three times lower in subjects undergoing surgery for fracture of the mandible who receive pre-operative SAP in a single dose or continued administration for 24 h compared to those not treated [28]. In contrast, SAP does not result in reduced infectious risk in mandibular condyle, maxilla or zygomatic fracture. Although several antibiotics have been shown to be effective, the one most often used was oral amoxicillin. In the second meta-analysis, which included 13 studies published before 2019 and compared pre-operative prophylaxis or continued administration for 24 h with SAP maintained for several days after surgery, it was shown that the latter did not result in any signiﬁcant reduction in the risk of SSIs (relative risk [RR]: 1.11, 95% conﬁdence intervals [CI]: 0.86–1.44; p > 0.1) [28]. No advantage was demonstrated when the analysis was restricted to the site of surgery or the type of injury (mandibular fracture, RR: 1.22; 95% CI: 0.92–1.62; maxilla fracture RR, 1.02, 95% CI: 0.62–1.67). With pediatric studies lacking, recommendations in children are those followed in adults. Recommendation 2. In the case of a pediatric patient undergoing maxillo-facial fracture surgery, pre-operative antibiotic prophylaxis with oral amoxicillin 50 mg/kg is recommended within 30 min prior to surgery when the surgery involves the mandible. Prophylaxis is not recommended in case of maxillary or zygomatic surgery. 3.2.2. SCENARIO #3. Temporo-Mandibular Surgery 3.2.3. SCENARIO #4. Cleft Palate and Cleft Lip Repair The development of SSIs after cleft lip or cleft palate surgery can cause clinical prob- lems of extreme importance both immediately and in long term. Besides the risk of bacteremia and the infection of distant organs and systems, there is the local risk of the relevant dehiscence of the wound, with serious repercussions on facial aesthetics, speech development and, in case of cleft lip, the creation of palatal ﬁstulas, all conditions that may require further surgery [36]. Nevertheless, few studies have exactly quantiﬁed this risk and the possible beneﬁts derived from the use of SAP. Some data have been collected in patients undergoing cleft palate repair, but the results are far from conclusive. A retrospec- tive study conducted by the American Cleft Palate-Craniofacial Association, in which the postoperative course of 311 patients was analyzed, of whom only 173 had received SAP, showed that surgery itself carries a signiﬁcant risk of delayed wound healing and palatal ﬁstula development and that the administration of SAP does not reduce this risk [37]. The development of delayed wound healing was demonstrated in 16.8% of subjects on SAP and in 15.2% of those without (p = 0.71). The appearance of palatal ﬁstulas was observed in 2.9% versus 1.4% of the cases, respectively (p = 0.47). Results apparently in favor of SAP were collected in a prospective, randomized, double-blind, placebo-controlled study conducted in India [38]. However, even in this case, in which a slightly lower incidence of SSIs was Antibiotics 2022, 11, 382 7 of 19 7 of 19 documented in subjects undergoing short-term and long-term antibiotic prophylaxis (13.8% versus 8.7% for SSIs in short-term and 17.1% versus 10.7% for ﬁstulas in long-term), the differences between the two groups were not statistically signiﬁcant (p = 0.175 and p = 0.085, respectively) and the authors’ conclusions were uncertain regarding the efﬁcacy of SAP. The scarcity of speciﬁc studies and the lack of results capable of deﬁnitively estab- lishing the importance of SAP explain why there are currently no internationally accepted guidelines indicating the most appropriate approach to cleft lip or cleft palate repair and why, in clinical practice, surgeons in various centers have very different behaviors, even if, in general, they are in favor of SAP. 3.2.3. SCENARIO #4. Cleft Palate and Cleft Lip Repair In the study of the American Cleft Palate-Craniofacial Association, it is reported that SAP was not applied in only 15% of operated patients, while the SAP schemes used in 85% of treated subjects and consisted of a single pre-operative dose in 26% of treated patients and multiple administrations in the others [38]. Speciﬁcally, antibiotic administration continued for 24 h beyond the end of surgery in 23% of cases, for 25–72 h in 12%, for 4–5 days in 16% and for 6–10 days in 12%. A ﬁrst-generation cephalosporin was used in 64% of the cases, ampicillin/sulbactam in 13%, clindamycin in 8% and penicillin in 5%. All of this is in contrast to what is expected based on the considera- tion that cleft lip and cleft palate repair surgeries should be considered clean surgeries and, therefore, should not require antibiotic prophylaxis. On the other hand, the widespread fear of SSI development with dramatic consequences may explain the widespread use of SAP in these cases. Microbiological data show that, in general, SSIs following these surgeries are sustained by the same bacteria present in the oral cavity [36] and that the preoperative presence of S. pyogenes and S. aureus seems to be a signiﬁcant risk factor for the development of SSIs [39]. The antibiotics recommended must, therefore, be effective against these bacteria. In this regard, the American Cleft Palate-Craniofacial Association suggests the use of ampicillin/sulbactam. With a lack of convincing data on the use of multiple doses, the administration of a single pre-operative dose is recommended [38]. Recommendation 4. In the case of a neonatal or pediatric patient undergoing cleft lip or cleft palate correction surgery, peri-operative prophylaxis with ampicillin/sulbactam at a dose of 50 mg/kg (as ampicillin) EV is recommended to be administered within 30 min before surgery. 3.3.1. SCENARIO #5. Ear Surgery There are numerous surgical procedures that fall within ear surgery. All those that are performed in the absence of ongoing infectious processes, such as those involving the insertion of tympanostomy tubes, tympanoplasty and stapedectomy in subjects without infection of the ear canal and/or middle ear, are among the so-called clean procedures that, by deﬁnition, are associated with a low or no risk of SSI development [40]. On the other hand, surgeries performed in subjects with chronic infectious middle ear disease with or without cholesteatoma or involving drainage from an infected site, including transtympanic tube placements, are considered clean/contaminated or simply contaminated surgeries and are generally associated with a high risk of SSI development [40]. Studies that quantiﬁed the true magnitude of this risk have shown that in clean surgeries, less than 5% of operated subjects experience SSIs, whereas in clean/contaminated or contaminated surgeries, this value rises to more than 10% [40]. The pathogens involved are, in addition to S. aureus, all those commonly involved in the determination of otologic infections, including those that may play a role in chronic suppurative otitis media, such as Pseudomonas aeruginosa [41]. Given this, it would be easy to infer whether and what SAP to use in the various forms of ear surgery. However, the available studies, often burdened by important methodological limitations, do not permit any conclusions to be drawn. The major limitation is represented by the fact that in many studies, the evaluation of the usefulness of SAP includes subjects with different pathologies, thus adding cases of clean intervention to cases of contaminated intervention. Moreover, the comparison between SAP limited to a single pre-operative administration and long-term prophylaxis after surgery is often conducted with different Antibiotics 2022, 11, 382 8 of 19 drugs and dosages [42–47]. A systematic review of the literature available up to the end of 2009, in which the use of SAP in clean and clean/contaminated otologic surgery was evaluated, showed that SAP in these conditions is completely unnecessary [48]. The incidence of SSIs was generally low in all cases, with no differences between those who had received the placebo, those who received peri-operative prophylaxis and those treated with prolonged postoperative prophylaxis [48]. In the absence of ﬁrm data, the rationale generally followed in clean versus clean/contaminated or simply contaminated surgeries prevails. 3.3.1. SCENARIO #5. Ear Surgery In the former, prophylaxis is not recommended; in the latter, the use of pre- operative prophylaxis with cefazolin EV is recommended [49]. p p p y [ ] The use of SAP in the case of cochlear implant surgery is a separate issue. Initially considered possible in up to 40% of cases, infections secondary to cochlear implantation are now limited to 1.4–8.2% of cases in many patients months after surgery [50]. Improved surgical techniques and materials used for implantation are considered the main reasons for this change. However, although relatively uncommon, infections following cochlear implantation can have dramatic consequences, far greater than those that can result from other forms of ear surgery. Infections can result in the need to remove the device and perform a second surgery [51], and infection can spread to the interior of the skull with the development of meningitis and abscesses [52]. Theoretically, cochlear implant placement is a clean procedure that may not require prophylaxis [53–55]. However, the risk of dramatic complications dictates a careful evaluation of the importance of SAP. Unfortunately, there are no randomized clinical trials performed with appropriate methods that can clarify this point. A systematic review of the literature published in this regard identiﬁed only three retrospective studies, all burdened by considerable heterogeneity and not negligible methodological limitations [56]. These studies seem to indicate that no form of SAP is useful in modifying the low tendency to develop SSIs in subjects undergoing cochlear implantation, suggesting no use of SAP in these cases [56]. In reality, these data are not convincing, as demonstrated by the fact that some authors believe they should suggest a different choice depending on the characteristics of the individual patient and some prestigious scientiﬁc institutions recommend, for caution, a systematic peri-operative prophylaxis with cefazolin EV [57]. Also discussed is how to perform SAP, whether with a single pre-intervention dose or with a more or less protracted antibiotic administration after surgical wound closure. A recent French study showed that a short treatment is ideal for adults, while a protracted one is more effective in children [57]. However, our panel of experts considered pre-operative antibiotic prophylaxis with a single dose of cefazolin EV as more appropriate, as recommended for clean/contaminated or contaminated ear surgery [49]. Recommendation 5. 3.3.1. SCENARIO #5. Ear Surgery In the case of a neonatal or pediatric patient undergoing ear surgery, peri-operative antibiotic prophylaxis is not recommended for clean surgery, whereas it is recommended in cases of clean/contaminated or contaminated operation and for cochlear implant placement. When antibiotic prophylaxis is indicated, it is rec- ommended to administer cefazolin as a single dose of 30 mg/kg (maximum dose 2 g) EV within 30 min before surgery. 3.3.2. SCENARIO #6. Endoscopic Paranasal Cavity Surgery and Septoplasty 3.3.2. SCENARIO #6. Endoscopic Paranasal Cavity Surgery and Septoplasty Endoscopic surgery of the rhinosinus cavities by definition falls into clean-contaminated surgeries or even into contaminated ones in cases with bacterial rhinosinusitis [58]. It should, therefore, be among the surgical procedures for which SAP should be provided. Several studies have shown that patients undergoing endoscopic rhinosinus cavity surgery frequently harbor pathogens such as S. aureus, anaerobes and S. pneumoniae in the examined sinuses [58–61]. Moreover, the procedure may be followed by bacteremia in a number of cases (7%) [62]. This suggests the possibility of developing sepsis and other signiﬁcant infections at a distance from the site of surgery with very clinically relevant outcomes. Nevertheless, in the guidelines prepared by the American Society of Health-System Phar- macists, the Infectious Diseases Society of America, the Surgical Infection Society and the Antibiotics 2022, 11, 382 9 of 19 9 of 19 Society for Healthcare Epidemiology of America, this form of surgery, along with tonsillec- tomy, is excluded from the recommendations for SAP [63]. In reality, there are no ﬁrm data indicating the reasons for this choice. Studies quantifying the risk of developing SSIs are practically absent. In addition, they do not provide information on pre-operative prophy- laxis, considering only the administration of antibiotics for days or weeks after surgery. In three studies, administration of cefuroxime for 10 days [64], amoxicillin/clavulanic acid for 3 weeks [65] or amoxicillin for 4 weeks [66] did not yield results different from those seen in patients who had received a placebo. In contrast, in another study [67], the use of amoxi- cillin/clavulanic acid for 2 weeks allowed better results in terms of endoscopic ﬁndings at both 5 and 12 days. However, the use of SAP in endoscopic surgery is widely practiced by surgeons. A survey conducted among members of the American Rhinologic Society showed that 20.6% routinely performed preoperative prophylaxis, 54.4% intraoperative prophylaxis and 62.3% postoperative prophylaxis [68]. There are, however, no deﬁnitive data available on the actual efﬁcacy of SAP in pediatric subjects undergoing endoscopic rhi- nosinus surgery [69]. Pending speciﬁc studies, and given the potential risk of development of infectious complications, our expert panel believed that pre-operative antibiotic pro- phylaxis with a single dose of cefazolin EV may be recommended in children undergoing rhinosinus endoscopy surgery. Regarding septoplasty, studies demonstrating the necessity or effectiveness of antibiotic prophylaxis are few. However, available data seem to indicate that if septoplasty increases S. 3.4. Head and Neck Surgery 3.4. Head and Neck Surgery 3.3.2. SCENARIO #6. Endoscopic Paranasal Cavity Surgery and Septoplasty aureus colonization and reduces normal ﬂora, pre-operative antibiotic administration does not protect against potential pathogen colonization and contributes to a further decrease in normal rhinopharyngeal microbiota [70]. Therefore, SAP is not recommended in septoplasty [71]. Recommendation 6. In the case of a pediatric patient undergoing endoscopic surgery of the rhinosinus cavities, it is recommended to administer peri-operative antibiotic pro- phylaxis with cefazolin 30 mg/kg (maximum dose 2 g) EV within the 30 min before surgery. No antibiotic prophylaxis is recommended in septoplasty. 3.4.1. SCENARIO #7. Head and Neck Clean Interventions 3.4.1. SCENARIO #7. Head and Neck Clean Interventions The majority of head and neck surgeries are considered clean surgeries and are fol- lowed by the development of SSIs in less than 1% of cases [72]. These include thyroidectomy, parathyroidectomy, salivary gland surgeries, the removal of lymphangiomas and the ex- cision of lateral and medial neck cysts and ﬁstulas. For these, no SAP is recommended, also because the available studies, all referring to adult patients, seem to indicate that the administration of antibiotics either pre-, intra- or post-operatively does not reduce the already low frequency of the occurrence of SSIs [72,73]. A separate evaluation is suggested by some authors for neck dissection surgery that, although classiﬁed as clean surgery, is associated with a slightly higher risk of SSIs because it involves a higher degree of tissue exposure. However, studies aimed at quantifying the true risk of SSIs in this type of surgery [74,75] and those aimed at measuring the impact of perioperative antibiotic prophylaxis [76,77] do not deﬁnitively clarify the characteristics of this type of surgery. In fact, the data collected from a small number of studies performed with different (and sometimes contradictory) methods are largely contrasting. The problem remains open, although some authors indicate that in these conditions, the use of pre-operative antibiotic prophylaxis with post-operative prolongation after surgical wound closure for less than 24 h could be recommended [78]. In any case, data speciﬁc to pediatric aged patients are lacking. Therefore, the recommendations provided for adults are considered valid also for children and our expert panel agreed to not recommend SAP for this type of surgery. Recommendation 7. No perioperative antibiotic prophylaxis is recommended in the case of a neonatal or pediatric patient undergoing clean head and neck surgery (i.e., thy- roidectomy, parathyroidectomy, salivary gland surgeries, the removal of lymphangiomas and the excision of lateral and medial neck cysts and ﬁstulas). Antibiotics 2022, 11, 382 10 of 19 10 of 19 3.4.2. SCENARIO #8. Head and Neck Clean-Contaminated Interventions 3.4.2. SCENARIO #8. Head and Neck Clean-Contaminated Interventions All surgeries on head and neck structures that involve the opening of the airway or gastrointestinal tract (i.e., oral cavity resection, laryngectomy, pharyngectomy, tracheotomy and maxillary of upper airways tumor masses) are considered clean-contaminated [79]. It has been shown that such surgeries are followed by SSIs in 25–85% of cases [80]. 3.4.1. SCENARIO #7. Head and Neck Clean Interventions The bacte- ria most often responsible are the same ones that normally colonize the mouth and pharynx, with the highest frequencies found for Streptococcus spp. (aerobes and anaerobes), S. aureus, Bacteroides spp. (with the exception of B. fragilis), Fusobacterium spp., Peptostreptococcus spp. and Veillonella spp. [79,80]. In addition, it has been demonstrated that SAP is extremely effective in reducing subsequent infections. A meta-analysis of 12 studies conducted before 1991 had already quantiﬁed the reduction in the frequency of SSIs as 43.7% when peri- operative antibiotic prophylaxis was used, with the greatest advantage associated with long-term prophylaxis over single-dose prophylaxis [81]. More recent studies have con- ﬁrmed the usefulness of peri-operative prophylaxis, although they have not fully clariﬁed which antibiotics and which form of prophylaxis might be the most effective. In particular, peri-operative prophylaxis is recommended for patients undergoing parotid gland surgery, and intravenous antibiotics during the post-operative course are highly suggested in case of patients with a histories of previous acute parotid infection and drain output ≥50 mL in the ﬁrst 24 h [82]. As for the drug(s), a huge number of studies are available. The relative quality of many of them and the low number of subjects enrolled in others make it impossible to indicate which single antibiotic or combination might be recommended. On the other hand, it is not uncommon for similar studies to have yielded conﬂicting results, with there being a further difﬁculty in identifying the ideal antibiotic prophylaxis. What seems to be established is that antibiotic prophylaxis should be implemented with drugs or associations that are active on Gram-positive and Gram-negative bacteria and have good coverage against anaerobes. Clindamycin alone or in combination with other compounds active on both Gram-positive and Gram-negative bacteria [83–85], cefazolin alone or with metronidazole [86–88], other cephalosporins [89–91] and the combinations of amoxicillin/clavulanic acid [92,93] and ampicillin/sulbactam [94–96] are the most widely tested forms of antibiotic prophylaxis with no clear superiority. With regard to the duration of administration, it is not possible to draw deﬁnitive conclusions about the efﬁcacy of a single dose of antibiotics before the start of surgery because studies in this regard are too limited. On the contrary, it seems certain that a prophylaxis that extends beyond the closure of the intervention site is useful, even if it is not clear how long the administration should be prolonged. 3.4.1. SCENARIO #7. Head and Neck Clean Interventions Indeed, there are no signiﬁcant differences in prevention implemented with a 24-h or a 3-, 5- or 7-day prolongation [97–99]. This has led to recommendations for the use of cefazolin or cefuroxime associated with metronidazole or ampicillin/sulbactam as the ﬁrst choice to be administered before surgery and immediately afterwards for no more than 24 h. Because all the studies were conducted almost exclusively on adults, recommendations for children can only be derived from that evidence. y Recommendation 8. In the case of a neonatal or pediatric patient undergoing clean- contaminated ENT surgery (i.e., oral cavity resection, laryngectomy, pharyngectomy, tra- cheotomy or the removal of upper airways tumor masses), peri-operative antibiotic pro- phylaxis with cefazolin 30 mg/kg (maximum dose 2 g) EV administered within 30 min before surgery combined with metronidazole 15 mg/kg (max 500 mg) is recommended. 3.5. SCENARIO #9. Tonsillectomy and Adenoidectomy For many years since the beginning of the antibiotic era, it was believed that SAP was an essential measure to reduce the risk of post-operative problems, including SSIs, in patients undergoing tonsillectomy. In fact, a survey of US otolaryngologists in 2004 showed that nearly 80% of them prescribed antibiotics to subjects scheduled for tonsillectomy [100]. More recent research and the demonstration that many of the studies that had led to the use of SAP were burdened by severe methodological limitations have completely reversed the initial assessments, leading to completely different recommendations [101–108]. Cur- Antibiotics 2022, 11, 382 11 of 19 11 of 19 rent knowledge is well summarized by the results of the meta-analysis of 10 randomized controlled trials conducted before 2012, which clearly highlighted that SAP does not reduce postoperative pain, the need for pain medication or the risk of bleeding [101]. In the few studies in which the administration of antibiotics seemed to be somewhat effective, the beneﬁts were extremely modest. Subsequent studies conﬁrmed these results, pointing out that SAP had no advantage even in reducing emergency room admissions or hospitaliza- tion [102–108]. All of this explains why the American Academy of Otolaryngology-Head and Neck Surgery Foundation recently reiterated in its guidelines for tonsillectomy in otherwise healthy children that SAP administration in these subjects should not be used at all [106]. Exceptions may be made for subjects at high risk of the onset of serious infec- tious problems (i.e., subjects with pre-existing cardiologic pathology already identiﬁed as requiring antibiotic prophylaxis in case of surgery). q g p p y g y Overlapping conclusions can be made for adenoidectomy, alone or in association with tonsillectomy. A number of studies have shown that in both of these conditions, surgery can be associated with bacteremia and, therefore, with the potential risk of sepsis or localization of the infection at a distance from the oral cavity [107]. However, while the risk of bacteremia is undeniable, with Haemophilus inﬂuenzae, viridans streptococcal species, S. pneumoniae and S. aureus as the most common pathogens [108,109], this seems entirely transient and not remotely followed by the development of major infectious issues. 3.5. SCENARIO #9. Tonsillectomy and Adenoidectomy A study comparing subjects undergoing adenoidectomy with and without SAP showed that antibiotics were markedly helpful in reducing the risk of bacteremia at 30 s after the end of surgery (3.9% in treated versus 32.7%; p < 0.001), but that this difference was no longer signiﬁcant in controls performed at 20 min after surgery (3.9% versus 14.3%; p = 0.089). In addition, both in the short- and long-term after surgery, the risk of complications of any kind, including acute otitis media, proved to be extremely low in each case and not different in the two groups [110]. g p Recommendation 9. In the case of a pediatric patient undergoing tonsillectomy, adenoidectomy or both, no antibiotic prophylaxis is recommended. 4. Discussion Many clinical conditions requiring surgical procedures are relatively uncommon in pediatrics. This explains why in these cases, studies on the necessity and efﬁcacy of SAP to reduce SSIs are very few or completely absent and the recommendations for the use of this preventive measure for pediatric aged patients are simply derived from those provided for adults. This seems to be somewhat different in the case of dental, maxillo- facial or ENT surgery because some of the surgical procedures in these areas are extremely common (Table 1). This is the case in cleft palate and cleft lip correction, tonsillectomy, adenoidectomy and transtympanic tube placement. In reality, recommendations are well- deﬁned and shared by all experts in only very few conditions. Generally, studies are few and far between, methodologically questionable and provide different results. In these cases, SAP is recommended because the intervention could be at risk of serious complications even if it is not clear that it is really necessary. Typical examples in this regard are given by SAP for cleft palate or cleft lip correction surgery and for cochlear implant placement. In these cases, conclusive data are lacking, and the fear of serious complications ends up being the main motivation for the systematic use of antibiotics and very poor adherence to suggested recommendations in everyday surgical practice. More precise and deﬁnitive recommendations can only be given for tonsillectomy and adenoidectomy procedures because the in-depth study of the infectious risks associated with these procedures have been extensively deﬁned. In these cases, the recommendation to not perform SAP is precisely supported by the evidence that the bacteremia following surgery is of very short duration and not followed by further localization. The speciﬁc scenarios developed are intended to guide healthcare professionals in practice, so as to ensure the improved and standardized management of neonatal and pediatric patients. The strengths of the work are an updated literature review, the use of a Antibiotics 2022, 11, 382 12 of 19 rigorous analysis method (RAND/UCLA), the involvement of a large number of exponents of the most important Italian scientiﬁc societies and the speciﬁc consideration of neonatal aged patients. The potential limitation of the work is the scarcity of data in the literature, which is partly overcome by the involvement of numerous and selected experts. 4. Discussion Oral amoxicillin or ampicillin ev (50 mg/kg for both) should be administered during the 30 min before surgery if the operation involves the manipulation of gum tissue or the periapical region of the teeth or involves the perforation of the oral mucosa and the subject has already suffered from endocarditis, has already been operated with the application of prosthetic material, has a cyanogenic congenital heart disease which has not yet fully repaired, has a congenital heart disease already repaired with the application of prosthetic material in the 6 months following surgery or has been transplanted and has developed valvulopathy. No prophylaxis is recommended in subjects with prosthetic implants. 4. Discussion On the other hand, the lack of pediatric studies on the selected topics did not permit the use of the GRADE methodology and the complexity of the topics required an online one-to-one meeting with all the participants. Table 1. Main maxillo-facial and ear-nose-throat (ENT) surgery procedures, divided into clean or clean/contaminated and/or frankly contaminated. Clean Procedures Clean/Contaminated and/or Frankly Contaminated Procedures Cleft lip and cleft palate repair Insertion of tympanostomy tubes * Tympanoplasty * Stapedectomy * Cochlear implant placement Septoplasty Thyroidectomy Parathyroidectomy Neck dissection Salivary gland surgeries Removal of lymphangiomas Excision of lateral and medial neck cysts and ﬁstulas Tonsillectomy Adenoidectomy Insertion of tympanostomy tubes Tympanoplasty Stapedectomy ** Rhinosinus endoscopic surgery Oral cavity resection Laryngectomy Pharyngectomy Tracheotomy Removal of upper airways tumor masses * In subjects without infection of the ear canal and/or middle ear; in subjects with chronic infectious middle ear disease with or without cholesteatoma or involving drainage from an infected site. Table 1. Main maxillo-facial and ear-nose-throat (ENT) surgery procedures, divided into clean or clean/contaminated and/or frankly contaminated. Clean/Contaminated and/or Frankly Contaminated Procedures Insertion of tympanostomy tubes Tympanoplasty Stapedectomy Rhinosinus endoscopic surgery Oral cavity resection Laryngectomy Pharyngectomy Tracheotomy Removal of upper airways tumor masses Tonsillectomy Adenoidectomy * In subjects without infection of the ear canal and/or middle ear; ** in subjects with chronic infectious middle ear disease with or without cholesteatoma or involving drainage from an infected site. Table 2 shows SAP for neonates and children undergoing dental, maxillo-facial or ENT surgeries. Antibiotic dosages are those routinely recommended [111]. Although the neonatal pharmacokinetics differs depending on gestational age, body weight and days after birth, the panel of experts did not recommend changes to the doses because of the short exposure duration of antimicrobial prophylaxis and the safety of the recommended drugs in patients of neonatal age. On the other hand, the large majority of dental, maxilla- facial and ENT surgical procedures are performed in patients with a body weight ≥2 kg, beyond the neonatal age. Table 2. Surgical antimicrobial prophylaxis (SAP) for neonates and children undergoing dental, maxillo-facial or ear-nose-throat (ENT) surgeries. Clinical Scenario Recommendation Dental surgery No peri-operative antibiotic prophylaxis is recommended. Recommendation No peri-operative antibiotic prophylaxis is recommended. Oral amoxicillin or ampicillin ev (50 mg/kg for both) should be administered during the 30 min before surgery if the operation involves the manipulation of gum tissue or the periapical region of the teeth or involves the perforation of the oral mucosa and the subject has already suffered from endocarditis, has already been operated with the application of prosthetic material, has a cyanogenic congenital heart disease which has not yet fully repaired, has a congenital heart disease already repaired with the application of prosthetic material in the 6 months following surgery or has been transplanted and has developed valvulopathy. No prophylaxis is recommended in subjects with prosthetic implants. 13 of 19 13 of 19 Antibiotics 2022, 11, 382 Table 2. Cont. Clinical Scenario Recommendation Maxillo-facial fracture surgery Pre-operative antibiotic prophylaxis with oral amoxicillin 50 mg/kg is recommended within 30 min prior to surgery when the surgery involves the mandible. Prophylaxis is not recommended in the case of maxillary or zygomatic surgery. Temporo-mandibular surgery Pre-operative antibiotic prophylaxis with cefazolin in a single dose of 30 mg/kg (maximum dose 2 g) EV is recommended in the 30 min before surgery. Cleft lip or cleft palate repair Peri-operative prophylaxis with ampicillin/sulbactam at a dose of 50 mg/kg (as ampicillin) EV is recommended to be administered within 30 min before surgery. Ear surgery Peri-operative antibiotic prophylaxis is not recommended for clean surgery, whereas it is recommended in cases of clean/contaminated or contaminated operation and for cochlear implant placement. When antibiotic prophylaxis is indicated, it is recommended to administer cefazolin as a single dose of 30 mg/kg (maximum dose 2 g) EV within 30 min before surgery. Endoscopic paranasal cavity surgery and septoplasty It is recommended to administer peri-operative antibiotic prophylaxis with cefazolin 30 mg/kg (maximum dose 2 g) EV within the 30 min before surgery. No antibiotic prophylaxis is recommended in septoplasty. Clean head and neck surgery No perioperative antibiotic prophylaxis is recommended in the case of neonatal or pediatric patients undergoing clean head and neck surgery (i.e., thyroidectomy, parathyroidectomy, salivary gland surgeries, the removal of lymphangiomas and the excision of lateral and medial neck cysts and ﬁstulas). Recommendation Clean-contaminated head and neck surgery In the case of a neonatal or pediatric patient undergoing clean-contaminated ENT surgery (i.e., oral cavity resection, laryngectomy, pharyngectomy, tracheotomy or maxillary of upper airways tumor masses), peri-operative antibiotic prophylaxis with cefazolin 30 mg/kg (maximum dose 2 g) EV administered within 30 min before surgery combined with metronidazole 15 mg/kg (max 500 mg) is recommended. tonsillectomy, adenoidectomy or both No antibiotic prophylaxis is recommended. Clinical Scenario Recommendation In the case of a neonatal or pediatric patient undergoing clean-contaminated ENT surgery (i.e., oral cavity resection, laryngectomy, pharyngectomy, tracheotomy or maxillary of upper airways tumor masses), peri-operative antibiotic prophylaxis with cefazolin 30 mg/kg (maximum dose 2 g) EV administered within 30 min before surgery combined with metronidazole 15 mg/kg (max 500 mg) is recommended. 5. Conclusions The application of uniform and shared protocols aims to improve the management of pediatric and neonatal patients with, on the one hand, the possibility of reducing SSIs and, on the other hand, containing the phenomenon of antimicrobial resistance, with the consequent rationalization of resources and costs. Our panel of experts thinks that, in the face of extremely heterogeneous prescriptions in real life characterized by the excessive and often inappropriate use of antibiotics in SAP, our document represents a balanced and shared text, derived from an extensive discussion, which can be extraordinarily beneﬁcial for patients and, more generally, for the health system. Due to there being a lack of pediatric data for the majority of dental, maxillo-facial and ENT surgeries and the fact that the recommendations for adults are currently used indicates the need for ad hoc studies to be rapidly planned for the most deﬁcient areas. This seems even more urgent for interventions such as those involving the ﬁrst airways since differences in the composition of the respiratory microbiota in children compared to Antibiotics 2022, 11, 382 14 of 19 14 of 19 adults implies the possibility that SAP schemes ideal for adults may not be equally effective in children. There is, however, once again the problem of identifying pediatric subjects as special subjects, and not as small adults, for whom the use of antibiotics should be carefully evaluated according to the speciﬁc characteristics of the various stages of development. Speciﬁc studies with new diagnostic methods on the respiratory microbiota in patients of different age ranges undergoing dental, maxillo-facial and ENT surgical procedures are needed. When our consensus document is implemented by Italian Scientiﬁc Societies, it will be interesting to analyze its clinical and economic impact in our geographical context. However, our recommendations could be generalized also to low- and middle-income countries, where the impacts of simple, cost-effective, sustainable and adaptable strategies on the reduction in morbidity risk and the associated costs have recently been highlighted. Author Contributions: E.R., E.C., F.O. and R.O. participated in the development of the methods and the literature analysis; S.B., L.N. and S.M. revised the literature review and wrote the ﬁrst draft of the manuscript; C.C. performed the evaluation of the results; D.D., M.G., A.I., L.L., A.M., G.P., N.P., S.T., E.V., A.S. and A.V. participated in the consensus and provided a substantial scientiﬁc contribution; E.S. and C.V. 1. Young, P.Y.; Khadaroo, R.G. Surgical site infections. Surg. Clin. N. Am. 2014, 94, 1245–1264. [CrossRef] [PubMed] 2. Alverdy, J.C.; Hyman, N.; Gilbert, J. Re-examining causes of surgical site infections following elective surgery in the era of asepsis. Lancet Infect. Dis. 2020, 20, e38–e43. [CrossRef] 3. Branch-Elliman, W.; O’Brien, W.; Strymish, J.; Itani, K.; Wyatt, C.; Gupta, K. Association of Duration and Type of Surgical Prophylaxis With Antimicrobial-Associated Adverse Events. JAMA Surg. 2019, 154, 590–598. [CrossRef] [PubMed] 4. Gouvêa, M.; Novaes Cde, O.; Pereira, D.M.; Iglesias, A.C. Adherence to guidelines for surgical antibiotic prophylaxis: A review. Braz. J. Infect. Dis. 2015, 19, 517–524. [CrossRef] [PubMed] 5. Conclusions supervised the consensus for their skills on dental, maxillo-facial and ENT surgery; S.E. designed the study, supervised the project and revised the ﬁrst draft of the manuscript; all the members of the Peri-Operative Prophylaxis in Neonatal and Paediatric Age (POP-NeoPed) Study Group participated in the consensus. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not required for a Delphi consensus document. Informed Consent Statement: Not required for a Delphi consensus document. Data Availability Statement: All the data are included in the manuscript. Acknowledgments: We would like to thank the other members of the Peri-Operative Prophylaxis in Neonatal and Paediatric Age (POP-NeoPed) Study Group who participated in the consensus: Alberto Argentiero, Giuseppe Maglietta, Matteo Puntoni, Cosimo Neglia, Isabella Cremonini, Mara Caramia: University Hospital of Parma, Parma, Italy; Annamaria Colombari, Agnese Corbelli, Marcella Gaffuri, Elena Gusson, Rosa Longo, Michele Piazza, Tessari Denis, Laura Venditto, Marcella Sibani, Evelina Tacconelli: Azienda Ospedaliera-Universitaria di Verona, Verona, Italy; Mario Lima, Luigia Scudeller, Federico Pea: IRCCS Azienda Ospedaliera-Universitaria di Bologna, Bologna, Italy; Matilde Ciccia, AUSL Bologna, Bologna, Italy; Andrzej Krizystoﬁak, Andrea Dotta, Sergio Picardo, Massimiliano Raponi, Paolo Rossi, Maia De Luca: IRCCS Ospedale Pediatrico Bambino Gesù, Rome, Italy; Leonardo Bussolin, Luisa Galli, Carlotta Montagnani: Meyer Hospital, University of Florence, Florence, Italy; Andrea Novelli: University of Florence, Florence, Italy; Fabio Mosca, Gloria Pelizzo, Carlo Pietrasanta: Università degli Studi di Milano, Milan, Italy; Claudia Colomba, Giovanni Corsello, Marcello Cimador: University of Palermo, Italy; Elisabetta Bignamini, Silvia Garazzino: Regina Margherita Hospital, Torino, Italy; Alfonso Papparella: Università Vanvitelli, Napoli, Italy; Giangiacomo Nicolini: Treviso Hospital, Treviso, Italy; Giorgio Conti, Rossella Garra: Catholic University, Roma, Italy; Laura Marchesini: Perugia Hospital, Perugia, Italy; Stefania Stefani: University of Catania, Catania, Italy; Valeria Caldarelli: AUSL Reggio Emilia, Reggio Emilia, Italy; Gianni Sava, Gabriele Stocco: University of Trieste, Trieste, Italy. Conﬂicts of Interest: The authors declare no conﬂict of interest. 1. Young, P.Y.; Khadaroo, R.G. Surgical site infections. Surg. Clin. N. Am. 2014, 94, 1245–1264. [CrossRef] [PubMed] 2. Alverdy, J.C.; Hyman, N.; Gilbert, J. Re-examining causes of surgical site infections following elective surgery in the era of asepsis. Lancet Infect. Dis. 2020, 20, e38–e43. [CrossRef] 3. Branch-Elliman, W.; O’Brien, W.; Strymish, J.; Itani, K.; Wyatt, C.; Gupta, K. Association of Duration and Type of Surgical Prophylaxis With Antimicrobial-Associated Adverse Events. JAMA Surg. 2019, 154, 590–598. [CrossRef] [PubMed] 4. Gouvêa, M.; Novaes Cde, O.; Pereira, D.M.; Iglesias, A.C. Adherence to guidelines for surgical antibiotic prophylaxis: A review. Braz. J. Infect. Dis. 2015, 19, 517–524. [CrossRef] [PubMed] References 15 of 19 15 of 19 Antibiotics 2022, 11, 382 5. Karamchandani, K.; Barden, K.; Prozesky, J. Adherence to surgical antimicrobial prophylaxis: “checking-the-box” is not enough. Int. J. Health Care Qual. Assur. 2019, 32, 470–473. [CrossRef] [PubMed] 6. Bratzler, D.W.; Dellinger, E.P.; Olsen, K.M.; Perl, T.M.; Auwaerter, P.G.; Bolon, M.K.; Fish, D.N.; Napolitano, L.M.; Sawyer, R.G.; Slain, D.; et al. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Surg. Infect. 2013, 14, 73–156. [CrossRef] [PubMed] [ ] 7. Ottoline, A.C.; Tomita, S.; Marques, M.; Felix, F.; Ferraiolo, P.N.; Laurindo, R.S. Antibiotic prophylaxis in otolaryngologic surgery. Int. Arch. Otorhinolaryngol. 2013, 17, 85–99. y g 8. Esposito, S.; Principi, N. Impact of nasopharyngeal microbiota on the development of respiratory tract diseases. Eur. J. Clin. Microbiol. Infect. Dis. 2018, 37, 1–7. [CrossRef] f [ ] 9. Fitch, K.; Bernstein, S.J.; Aguilar, M.D.; Burnand, B.; LaCalle, J.R.; Lazaro, P.; Loo, M.V.H.; McDonnell, J.; Vader, J.; Kahan, J.P. The RAND/UCLA Adeguateness Method User’s Manual; The RAND Corporation: Santa Monica, CA, USA, 2001. f 9. Fitch, K.; Bernstein, S.J.; Aguilar, M.D.; Burnand, B.; LaCalle, J.R.; Lazaro, P.; Loo, M.V.H.; McDonnell, J. RAND/UCLA Adeguateness Method User’s Manual; The RAND Corporation: Santa Monica CA USA 20 f Fitch, K.; Bernstein, S.J.; Aguilar, M.D.; Burnand, B.; LaCalle, J.R.; Lazaro, P.; Loo, M.V.H.; McDonnell, J.; Va 9. Fitch, K.; Bernstein, S.J.; Aguilar, M.D.; Burnand, B.; LaCalle, J.R.; Lazaro, P.; Loo, M.V.H.; McDonnell, J. RAND/UCLA Adeguateness Method User’s Manual; The RAND Corporation: Santa Monica, CA, USA, 20 10. McAlister, W.H.; Cacciarelli, A.; Shackelford, G.D. Complications associated with cystography in children. Radiology 1974, 111, 167–172. [CrossRef] [PubMed] 11. Roberts, G.J.; Jaffrey, E.C.; Spract, D.A.; Petrie, A.; Greville, C.; Wilson, M. Duration, prevalence and intensity of bacteremia after dental extractions in children. Heart 2006, 92, 1274–1277. [CrossRef] 12. Baltimore, R.S.; Gewitz, M.; Baddour, L.M.; Beerman, L.B.; Jackson, M.A.; Lockhart, P.B.; Pahl, E.; Schutze, G.E.; Shulman, S.T.; Willoughby, R., Jr. Infective Endocarditis in Childhood: 2015 Update: A Scientiﬁc Statement From the American Heart Association. Circulation 2015, 132, 1487–1515. [CrossRef] [PubMed] 13. Wilson, W.; Taubert, K.A.; Gewitz, M.; Lockhart, P.B.; Baddour, L.M.; Levison, M.; Bolger, A.; Cabell, C.H.; Takahashi, M.; Baltimore, R.S.; et al. References Prevention of infective endocarditis: Guidelines from the American Heart Association: A guideline from the American Heart Association Rheumatic Fever, Endocarditis, and Kawasaki Disease Committee, Council on Cardiovascular Disease in the Young, and the Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and the Quality of Care and Outcomes Research Interdisciplinary Working Group. Circulation 2007, 116, 1736–1754. [PubMed] g gy g y of Care and Outcomes Research Interdisciplinary Working Group. Circulation 2007, 116, 1736–1754. [P p y g p [ ] 14. Baddour, L.M.; Wilson, W.R.; Bayer, A.S.; Fowler VGJr Tleyjeh, I.M.; Rybak, M.J.; Barsic, B.; Lockhart, P.B.; Gewitz, M.H.; Levison, M.E.; Bolger, A.F.; et al. Infective endocarditis in adults: Diagnosis, antimicrobial therapy, and management of complications: A scientiﬁc statement for healthcare professionals from the American Heart Association. Circulation 2015, 132, 1435–1486. [CrossRef] [PubMed] [ ] 15. Chen, T.T.; Yeh, Y.C.; Chien, K.L.; Lai, M.S.; Tu, Y.K. Risk of infective endocarditis after invasive dental treatments. Circulation 2018, 138, 356–363. [CrossRef] [PubMed] 16. Wilson, W.R.; Gewitz, M.; Lockhart, P.B.; Bolger, A.F.; DeSimone, D.C.; Kazi, D.S.; Couper, D.J.; Beaton, A.; Kilmartin, C.; Miro, J.M.; et al. Prevention of Viridans Group Streptococcal Infective Endocarditis: A Scientiﬁc Statement From the American Heart Association. Circulation 2021, 143, e963–e978. [CrossRef] [PubMed] 17. American Academy of Pediatric Dentistry. Antibiotic Prophylaxis for Dental Patients at Risk for Infectio 386–391. 18. Alao, U.; Pydisetty, R.; Sandiford, N.A. Antibiotic prophylaxis during dental procedures in patients with in situ lower limb prosthetic joints. Eur. J. Orthop. Surg. Traumatol. 2015, 25, 217–220. [CrossRef] [PubMed] 19. Sollecito, T.P.; Abt, E.; Lockhart, P.B.; Truelove, E.; Paumier, T.M.; Tracy, S.L.; Tampi, M.; Beltrán-Aguilar, E.D.; Frantsve-Hawley, J. The use of prophylactic antibiotics prior to dental procedures in patients with prosthetic joints: Evidence-based clinical practice guideline for dental practitioners–a report of the American Dental Association Council on Scientiﬁc Affairs. J. Am. Dent. Assoc. 2015, 146, 11–16. [CrossRef] [PubMed] , , [ ] [ ] 20. Berbari, E.F.; Osmon, D.R.; Carr, A.; Hanssen, A.D.; Baddour, L.M.; Greene, D.; Kupp, L.I.; Baughan, L.W.; Harmsen, W.S.; Mandrekar, J.N.; et al. Dental procedures as risk factors for prosthetic hip or knee infection: A hospital-based prospective case-control study. Clin. Infect. Dis. 2010, 50, 8–16. [CrossRef] [PubMed] y f [ ] [ ] 21. Kao, F.C.; Hsu, Y.C.; Chen, W.H.; Lin, J.N.; Lo, Y.Y.; Tu, Y.K. References Laryngoscope 2019, 129, 82–95. [CrossRef] [PubMed] 30. McKenzie, W.S.; Louis, P.J. Temporomandibular total joint prosthesis infections: A ten-year retrospective analysis. Int. J. Oral Maxillofac. Surg. 2017, 46, 596–602. [CrossRef] [PubMed] 31. Bosco, J.; Bookman, J.; Slover, J.; Edusei, E.; Levine, B. Principles of antibiotic prophylaxis in total joint arthroplasty: Current concepts. Instr. Course Lect. 2016, 65, 467–475. [CrossRef] ing and managing temporomandibular joint total joint replacement surgical site infections. J. Oral Maxillofac –2289. [CrossRef] [PubMed] 32. Mercuri, L.G. Avoiding and managing temporomandibular joint total joint replacement surgical site inf Surg. 2012, 70, 2280–2289. [CrossRef] [PubMed] 33. Rachmiel, A.; Shilo, D. The use of distraction osteogenesis in oral and maxillofacial surgery. Ann. Maxillofac. Surg. 2015, 5, 146–147. [CrossRef] [PubMed] 34. Tan, S.K.; Lo, J.; Zwahlen, R.A. Are postoperative intravenous antibiotics necessary after bimaxillary orthognathic surgery? A prospective, randomized, double-blind, placebo-controlled clinical trial. Int. J. Oral Maxillofac. Surg. 2011, 40, 1363–1368. [CrossRef] [PubMed] [ ] [ ] 35. Zijderveld, S.A.; Smeele, L.E.; Kostense, P.J.; Tuinzing, D.B. Preoperative antibiotic prophylaxis in orthognathic surgery: A randomized, double-blind, and placebo-controlled clinical study. J. Oral Maxillofac. Surg. 1999, 57, 1403–1407. [CrossRef] 35. Zijderveld, S.A.; Smeele, L.E.; Kostense, P.J.; Tuinzing, D.B. Preoperative antibiotic prophylaxis in orthognathic surgery: A randomized, double-blind, and placebo-controlled clinical study. J. Oral Maxillofac. Surg. 1999, 57, 1403–1407. [CrossRef] 36. Smyth, A.G.; Knepil, G.J. Prophylactic antibiotics and surgery for primary clefts. Br. J. Oral Maxillofac. Surg. 2008, 46, 107–109. [CrossRef] [PubMed] p y f g 36. Smyth, A.G.; Knepil, G.J. Prophylactic antibiotics and surgery for primary clefts. Br. J. Oral Maxillofac. Surg. 2008, 46, 107–109. [CrossRef] [PubMed] 37. Rottgers, S.A.; Camison, L.; Mai, R.; Shakir, S.; Grunwaldt, L.; Nowalk, A.J.; Natali, M.; Losee, J.E. Antibiotic Use in Primary Palatoplasty: A Survey of Practice Patterns, Assessment of Efﬁcacy, and Proposed Guidelines for Use. Plast. Reconstr. Surg. 2016, 137, 574–582. [CrossRef] 38. Aznar, M.L.; Schönmeyr, B.; Echaniz, G.; Nebeker, L.; Wendby, L.; Campbell, A. Role of Postoperative Antimicrobials in Cleft Palate Surgery: Prospective, Double-Blind, Randomized, Placebo-Controlled Clinical Study in India. Plast. Reconstr. Surg. 2015, 136, 59e–66e. [CrossRef] 39. Hupkens, P.; Lauret, G.J.; Dubelaar, I.J.M.; Hartman, E.H.M.; Spauwen, P.H.M. Prevention of wound dehiscence in palatal surgery by preoperative identiﬁcation of group A Streptococcus and Staphylococcus aureus. Eur. J. Plast. Surg. 2007, 29, 321–325. [CrossRef] 40. Govaerts, P.J.; Raemaekers, J.; Verlinden, A.; Kalai, M.; Somers, T.; Offeciers, F.E. Use of antibiotic prophylaxis in ear surgery. References Prosthetic joint infection following invasive dental procedures and antibiotic prophylaxis in patients with hip or knee arthroplasty. Infect. Control Hosp. Epidemiol. 2017, 38, 154–161. [CrossRef] 22. Watters, W., 3rd; Rethman, M.P.; Hanson, N.B.; Abt, E.; Anderson, P.A.; Carroll, K.C.; Futrell, H.C.; Garvin, K.; Glenn, S.O.; Hellstein, J.; et al. American Academy of Orthopedic Surgeons; American Dental Association. Prevention of orthopaedic implant 22. Watters, W., 3rd; Rethman, M.P.; Hanson, N.B.; Abt, E.; Anderson, P.A.; Carroll, K.C.; Futrell, H.C.; Garvin, K.; Glenn, S.O.; Hellstein, J.; et al. American Academy of Orthopedic Surgeons; American Dental Association. Prevention of orthopaedic implant infection in patients undergoing dental procedures. J. Am. Acad. Orthop. Surg. 2013, 21, 180–189. [CrossRef] [PubMed] 23. Rademacher, W.M.H.; Walenkamp, G.H.I.M.; Moojen, D.J.F.; Hendriks, J.G.E.; Goedendorp, T.A.; Rozema, F.R. Antibiotic prophylaxis is not indicated prior to dental procedures for prevention of periprosthetic joint infections. Acta Orthop. 2017, 88, 568–574. [CrossRef] [PubMed] 24. Thornhill, M.H.; Crum, A.; Rex, S.; Stone, T.; Campbell, R.; Bradburn, M.; Fibisan, V.; Lockhart, P.B.; Springer, B.M.; Baddour, L.M.; et al. Analysis of Prosthetic Joint Infections Following Invasive Dental Procedures in England. JAMA Netw. Open 2022, 5, e2142987. [CrossRef] [PubMed] 25. Tomczyk, S.; Whitten, T.; Holzbauer, S.M.; Lynﬁeld, R. Combating antibiotic resistance: A survey habits of dentists. Gen. Dent. 2018, 66, 61–68. [PubMed] n, T.; Holzbauer, S.M.; Lynﬁeld, R. Combating antibiotic resistance: A survey on the antibiotic-prescribing en. Dent. 2018, 66, 61–68. [PubMed] 26. Tebano, G.; Dyar, O.J.; Beovic, B.; Béraud, G.; Thilly, N.; Pulcini, C. ESCMID Study Group for Antimicrobial Stewardship (ESGAP). Defensive medicine among antibiotic stewards: The international European Society in Clinical Microbiology and Infectious Diseases AntibioLegal Map survey. J. Antimicrob. Chemother. 2018, 73, 1989–1996. [CrossRef] [PubMed] Antibiotics 2022, 11, 382 16 of 19 27. Kreutzer, K.; Storck, K.; Weitz, J. Current evidence regarding prophylactic antibiotics in head and neck and maxillofacial surgery. BioMed Res. Int. 2014, 2014, 879437. [CrossRef] [PubMed] 27. Kreutzer, K.; Storck, K.; Weitz, J. Current evidence regarding prophylactic antibiotics in head and neck and maxillofacial surgery. BioMed Res. Int. 2014, 2014, 879437. [CrossRef] [PubMed] 28. Andreasen, J.O.; Jensen, S.S.; Schwartz, O.; Hillerup, Y. A systematic review of prophylactic antibiotics in the surgical treatment of maxillofacial fractures. J. Oral Maxillofac. Surg. 2006, 64, 1664–1668. [CrossRef] [PubMed] f g 29. Habib, A.M.; Wong, A.D.; Schreiner, G.C.; Satti, K.F.; Riblet, N.B.; Johnson, H.A.; Ossoff, J.P. Postoperative prophylactic antibiotics for facial fractures: A systematic review and meta-analysis. References Laryngoscope 1998, 108, 107–110. [CrossRef] 41. Principi, N.; Marchisio, P.; Rosazza, C.; Sciarrabba, C.S.; Esposito, S. Acute otitis media with spontane perforation. Eur. J. Clin. Microbiol. Infect. Dis. 2017, 36, 11–18. [CrossRef] pi, N.; Marchisio, P.; Rosazza, C.; Sciarrabba, C.S.; Esposito, S. Acute otitis media with spontaneous tympa ti E J Cli Mi bi l I f t Di 2017 36 11 18 [C R f] p , ; , ; , ; , ; p , p y p ration. Eur. J. Clin. Microbiol. Infect. Dis. 2017, 36, 11–18. [CrossRef] 42. Carlin, W.V.; Lesser, T.H.; John, D.G.; Fielder, C.; Carrick, D.G.; Thomas, P.L.; Hill, S. Systemic antibiotic prophylaxis and reconstructive ear surgery. Clin. Otolaryngol. Allied Sci. 1987, 12, 441–446. [CrossRef] [PubMed] ntimicrobial prophylaxis in ear surgery. Laryngoscope 1988, 98, 1116–1123. [CrossRef] [PubMed] 43. Jackson, C.G. Antimicrobial prophylaxis in ear surgery. Laryngoscope 1988, 98, 1116–1123. [CrossRef] [PubMed] 44 Winerman I ; Segal S ; Man A Effectiveness of prophylactic antibiotic treatment in mastoid surgery Am J Otol 1981 3 65–67 43. Jackson, C.G. Antimicrobial prophylaxis in ear surgery. Laryngoscope 1988, 98, 1116–1123. [CrossRef] [PubMed] 44. Winerman, I.; Segal, S.; Man, A. Effectiveness of prophylactic antibiotic treatment in mastoid surger [PubMed] 45. Hester, T.O.; Jones, R.O. Prophylactic antibiotics in surgery for chronic ear disease. Laryngoscope 1998, 108, 1334–1337. [CrossRef] [PubMed] 46. John, D.G.; Carlin, W.V.; Lesser, T.H.; Carrick, D.G.; Fielder, C. Tympanoplasty surgery and prophylactic antibiotics: Surgical results. Clin. Otolaryngol. Allied Sci. 1988, 13, 205–207. [CrossRef] [PubMed] 47. Bidkar, V.G.; Jalisatigi, R.R.; Naik, A.S.; Shanbag, R.D.; Siddappa, R.; Sharma, P.V.; Hegde, H.V. Perioperative only versus extended antimicrobial usage in tympanomastoid surgery: A randomized trial. Laryngoscope 2014, 124, 1459–1463. [CrossRef] [PubMed] 48. Verschuur, H.P.; de Wever, W.W.; van Benthem, P.P. Antibiotic prophylaxis in clean and clean-contaminated ear surgery. Cochrane Database Syst Rev 2004 3 CD003996 [CrossRef] [PubMed] 47. Bidkar, V.G.; Jalisatigi, R.R.; Naik, A.S.; Shanbag, R.D.; Siddappa, R.; Sharma, P.V.; Hegde, H.V. Perioperative only versus extended antimicrobial usage in tympanomastoid surgery: A randomized trial. Laryngoscope 2014, 124, 1459–1463. [CrossRef] [PubMed] antimicrobial usage in tympanomastoid surgery: A randomized trial. Laryngoscope 2014, 124, 1459–1463. [CrossRef] [PubMed] 48. Verschuur, H.P.; de Wever, W.W.; van Benthem, P.P. Antibiotic prophylaxis in clean and clean-contaminated ear surgery. Cochrane Database Syst. Rev. 2004, 3, CD003996. [CrossRef] [PubMed] 48. Verschuur, H.P.; de Wever, W.W.; van Benthem, P.P. Antibiotic prophylaxis in clean and clean-contaminated ear surgery. Cochrane Database Syst. Rev. References What is the bacteriology of chronic sinusitis in adults? Am. J. Otolaryngol. 1995, 16, 303–306. [CrossRef] 61. Ramadan, H.H. What is the bacteriology of chronic sinusitis in adults? Am. J. Otolaryngol. 1995, 16, gy y g 62. Rontal, M.; Bernstein, J.M.; Rontal, E.; Anon, J. Bacteriologic ﬁndings from the nose, ethmoid, and bloodstream during endoscopic surgery for chronic rhinosinusitis: Implications for antibiotic therapy. Am. J. Rhinol. 1999, 13, 91–96. [CrossRef] [PubMed] g y p py 63. Bratzler, D.W.; Dellinger, E.P.; Olsen, K.M.; Perl, T.M.; Auwaerter, P.G.; Bolon, M.K.; Fish, D.N.; Napolitano, L.M.; Sawyer, R.G.; Slain, D.; et al. American Society of Health-System Pharmacists; Infectious Disease Society of America; Surgical Infection Society; Society for Healthcare Epidemiology of America. Clinical practice guidelines for antimicrobial prophylaxis in surgery. Am. J. Health Syst. Pharm. 2013, 70, 195–283. [CrossRef] [PubMed] y 64. Annys, E.; Jorissen, M. Short term effects of antibiotics (Zinnat) after endoscopic sinus surgery. Acta Otorhinolaryngol. Belg. 2000, 54, 23–28. [PubMed] 65. Jiang, R.S.; Liang, K.L.; Yang, K.Y.; Shiao, J.Y.; Su, M.C.; Hsin, C.H.; Lin, J.F. Postoperative antibiotic care after functional endoscopic sinus surgery. Am. J. Rhinol. 2008, 22, 608–612. [CrossRef] [PubMed] 66. Liang, K.L.; Su, Y.C.; Tsai, C.C.; Lin, J.S.; Jiang, R.S.; Su, M.C. Postoperative care with Chinese herbal medicine or amoxicillin after functional endoscopic sinus surgery: A randomized, double-blind, placebo-controlled study. Am. J. Rhinol. Allergy 2011, 25, 170–175. [CrossRef] [PubMed] 67. Albu, S.; Lucaciu, R. Prophylactic antibiotics in endoscopic sinus surgery: A short follow-up study. Am. J. Rhinol. Allergy 2010, 24, 306–309. [CrossRef] [PubMed] 68. Fang, C.H.; Fastenberg, J.H.; Fried, M.P.; Jerschow, E.; Akbar, N.A.; Abuzeid, W.M. Antibiotic use patterns in endoscopic sinus surgery: A survey of the American Rhinologic Society membership. Int. Forum Allergy Rhinol. 2018, 8, 522–529. [CrossRef] [PubMed] [ ] 69. Beswick, D.M.; Messner, A.H.; Hwang, P.H. Pediatric Chronic Rhinosinusitis Management in Rhinologists and Pediatric Otolaryngologists. Ann. Otol. Rhinol. Laryngol. 2017, 126, 634–639. [CrossRef] [PubMed] 70. Karaman, E.; Alimoglu, Y.; Aygun, G.; Kilic, E.; Yagiz, C. Effect of septoplasty and per-operative antibiotic prophylaxis on nasal ﬂora. B-ENT 2012, 8, 13–19. , L. Antibiotics in septoplasty: Evidence or habit? Am. J. Rhinol. Allergy 2012, 26, 194–196. [CrossRef] 71. Ricci, G.; D’Ascanio, L. Antibiotics in septoplasty: Evidence or habit? Am. J. Rhinol. Allergy 2012, 26, 19 , ; , p p y J gy , , [ ] 72. Johnson, J.T.; Wagner, R.L. Infection following uncontaminated head and neck surgery. References 2004, 3, CD003996. [CrossRef] [PubMed] 49. Bianchini, S.; Rigotti, E.; Nicoletti, L.; Monaco, S.; Auriti, C.; Castagnola, E.; Castelli Gattinara, G.; De Luca, M.; Galli, L.; Garazzino, S.; et al. Surgical Antimicrobial Prophylaxis in Neonates and Children with Special High-Risk Conditions: A RAND/UCLA Appropriateness Method Consensus Study. Antibiotics 2022, 11, 246. [CrossRef] 50. Lander, D.P.; Durakovic, N.; Kallogjeri, D.; Jiramongkolchai, P.; Olsen, M.A.; Piccirillo, J.F.; Buchman, C.A. Incidence of Infectious Complications Following Cochlear Implantation in Children and Adults. JAMA 2020, 323, 182–183. [CrossRef] 51. Francis, H.W.; Buchman, C.A.; Visaya, J.M.; Wang, N.Y.; Zwolan, T.A.; Fink, N.E.; Niparko, J.K.; CDaCI Investigative Team. Surgical factors in pediatric cochlear implantation and their early effects on electrode activation and functional outcomes. Otol. Neurotol. 2008, 29, 502–508. [CrossRef] 52. Cohen, N.L.; Hoffman, R.A. Complications of cochlear implant surgery in adults and children. Ann. Otol. Rhinol. Laryngol. 1991, 100, 708–711. [CrossRef] [PubMed] 53. Hirsch, B.E.; Blikas, A.; Whitaker, M. Antibiotic prophylaxis in cochlear implant surgery. Laryngoscope 2007, 117, 864–867. [CrossRef] [PubMed] 17 of 19 Antibiotics 2022, 11, 382 54. Basavaraj, S.; Najaraj, S.; Shanks, M.; Wardrop, P.; Allen, A.A. Short-term versus long-term antibiotic prophylaxis in cochlear implant surgery. Otol. Neurotol. 2004, 25, 720–722. [CrossRef] [PubMed] p g y 55. Garcia-Valdecasas, J.; Jimenez-Moleon, J.J.; Sainz, M.; Fornieles, C.; Ballesteros, J.M. Prophylactic effect of clarithromycin in skin ﬂap complications in cochlear implants surgery. Laryngoscope 2009, 119, 2032–2036. [CrossRef] [PubMed] p p p g y y g p 56. Anne, S.; Ishman, S.L.; Schwartz, S. A Systematic Review of Perioperative Versus Prophylactic Antibiotics for Cochlear Implanta- tion. Ann. Otol. Rhinol. Laryngol. 2016, 125, 893–899. [CrossRef] [PubMed] y g 57. Sayed-Hassan, A.; Hermann, R.; Chidiac, F.; Truy, E.; Guevara, N.; Bailleux, S.; Deguine, O.; Baladi, B.; Gallois, Y.; Bozorg-Grayeli, A.; et al. Association of the Duration of Antibiotic Therapy With Major Surgical Site Infection in Cochlear Implantation. JAMA Otolaryngol. Head Neck Surg. 2019, 145, 14–20. [CrossRef] [PubMed] y g g , , [ ] [ ] 58. Esposito, S.; Marchisio, P.; Tenconi, R.; Tagliaferri, L.; Albertario, G.; Patria, M.F.; Principi, N. Diagnosis of acute rhinosinusitis. Pediatr. Allergy Immunol. 2012, 23 (Suppl. 22), 17–19. [CrossRef] [PubMed] 59. Doyle, P.W.; Wooham, J.D. Evaluation of the microbiology of chronic ethmoid sinusitis. J. Clin. Microbiol. 1991, 29, 2396–2400. [CrossRef] 60. Hoyt, W.H. Bacterial patterns found in surgery patients with chronic sinusitis. J. Am. Osteopath. Assoc. 1992, 92, 205–212. [CrossRef] H. 80. Simo, R.; French, G. The use of prophylactic antibiotics in head and neck oncological surgery. Curr. Opin. Otolaryngol. Head Neck Surg. 2006, 14, 55–61. [CrossRef] References Arch. Otolaryngol. Head Neck Surg. 1987, 113, 368–369. [CrossRef] [PubMed] 73. Vamvakidis, K.; Rellos, K.; Tsourma, M.; Christoforides, C.; Anastasiou, E.; Zorbas, K.A.; Arambatzi, A.; Falagas, M.E. Antibiotic prophylaxis for clean neck surgery. Ann. R. Coll. Surg. Engl. 2017, 99, 410–412. [CrossRef] [PubMed] 74. Bergenfelz, A.; Jansson, S.; Kristoffersson, A.; Mårtensson, H.; Reihnér, E.; Wallin, G.; Lausen, I. Complications to thyroid surgery: Results as reported in a database from a multicenter audit comprising 3,660 patients. Langenbecks Arch. Surg. 2008, 393, 667–673. [CrossRef] [PubMed] 75. Abboud, B.; Sleilaty, G.; Tannoury, J.; Daher, R.; Abadjian, G.; Ghorra, C. Cervical neck dissection without drains in welldifferenti- ated thyroid carcinoma. Am. Surg. 2011, 77, 1624–1628. [CrossRef] [PubMed] y g 76. Carrau, R.L.; Byzakis, J.; Wagner, R.L.; Johnson, J.T. Role of prophylactic antibiotics in uncontaminated neck dissections. Arch. Otolaryngol. Head Neck Surg. 1991, 117, 194–195. [CrossRef] [PubMed] y g 76. Carrau, R.L.; Byzakis, J.; Wagner, R.L.; Johnson, J.T. Role of prophylactic antibiotics in uncontam Otolaryngol. Head Neck Surg. 1991, 117, 194–195. [CrossRef] [PubMed] g R.L.; Byzakis, J.; Wagner, R.L.; Johnson, J.T. Role of prophylactic antibiotics in uncontaminated neck dissect 77. Seven, H.; Sayin, I.; Turgut, S. Antibiotic prophylaxis in clean neck dissections. J. Laryngol. Otol. 2004, 118, 213–216. [CrossRef] [PubMed] 78. Patel, P.N.; Jayawardena, A.D.L.; Walden, R.L.; Penn, E.B.; Francis, D.O. Evidence-Based Use of Perioperative Antibiotics in Otolaryngology. Otolaryngol. Head Neck Surg. 2018, 158, 783–800. [CrossRef] [PubMed] 79. Busch, C.J.; Knecht, R.; Münscher, A.; Matern, J.; Dalchow, C.; Lörincz, B.B. Postoperative antibiotic prophylaxis in clean- contaminated head and neck oncologic surgery: A retrospective cohort study. Eur. Arch. Otorhynolaryngol. 2016, 273, 2805–2811. [CrossRef] 80. Simo, R.; French, G. The use of prophylactic antibiotics in head and neck oncological surgery. Curr. Opin. Otolaryngol. Head Neck Surg. 2006, 14, 55–61. [CrossRef] 18 of 19 Antibiotics 2022, 11, 382 81. Velanovich, V. A meta-analysis of prophylactic antibiotics in head and neck surgery. Plast. Reconstr. Surg. 1991, 87, 429–434. [CrossRef] 82. Meccariello, G.; Montevecchi, F.; D’Agostino, G.; Zeccardo, E.; Al-Raswashdeh, M.F.H.; De Vito, A.; Vicini, C. Surgical site infections after parotidectomy: Management and beneﬁts of an antibiotic prophylaxis protocol. Acta Otorhinolaryngol. Ital. 2019, 39, 139–144. [CrossRef] [PubMed] [ ] [ ] 83. Cohen, L.E.; Finnerty, B.M.; Golas, A.R.; Ketner, J.J.; Weinstein, A.; Boyko, T.; Rohde, C.H.; Kutler, D.; Spector, J.A. Perioperative antibiotics in the setting of oropharyngeal reconstruction: Less is more. Ann. Plast. Surg. 2016, 76, 663–667. [CrossRef] [PubMed] 84. References [CrossRef] 93 Skit li N M i M M t D A tibi ti h l i i l t i t d h d d k l i l J 93. Skitarelic, N.; Morovic, M.; Manestar, D. Antibiotic prophylaxis in clean-contaminated head and nec Craniomaxillofac. Surg. 2007, 35, 15–20. [CrossRef] [PubMed] 94. Langerman, A.; Ham, S.A.; Pisano, J.; Pariser, J.; Hohmann, S.F.; Meltzer, D.O. Laryngectomy complications are associated with perioperative antibiotic choice. Otolaryngol. Head Neck Surg. 2015, 153, 60–68. [CrossRef] [PubMed] 95. Phan, M.; Van der Auwera, P.; Andry, G.; Aoun, M.; Chantrain, G.; Deraemaecker, R.; Dor, P.; Daneau, D.; Ewalenko, P.; Meunier, F. Antimicrobial prophylaxis for major head and neck surgery in cancer patients: Sulbactam-ampicillin versus clindamycin-amikacin. Antimicrob. Agents Chemother. 1992, 36, 2014–2019. [CrossRef] [PubMed] g 96. Johnson, J.T.; Kachman, K.; Wagner, R.L.; Myers, E.N. Comparison of ampicillin/sulbactam versus clindamycin in the prevention of infection in patients undergoing head and neck surgery. Head Neck 1997, 19, 367–371. [CrossRef] 97 Li S A T K C Shi J Y Chi YT P li i f i d f i i l i i f i f j h d d 97. Liu, S.A.; Tung, K.C.; Shiao, J.Y.; Chiu, Y.T. Preliminary report of associated factors in surgical site infection after major head and neck neoplasm operations—does the duration of prophylactic antibiotic matter? J. Laryngol. Otol. 2008, 122, 403–408. [CrossRef] [PubMed] 98. Sepehr, A.; Santos, B.J.; Chou, C.; Karimi, K.; Devcic, Z.; Oels, S.; Armstrong, W.B. Antibiotics in head and neck surgery in the setting of malnutrition, tracheotomy, and diabetes. Laryngoscope 2009, 119, 549–553. [CrossRef] 99. Coskun, H.; Erisen, L.; Basut, O. Factors affecting wound infection rates in head and neck surgery. Otolaryngol. Head Neck Surg. 2000, 123, 328–333. [CrossRef] 100. Krishna, P.; LaPage, M.J.; Hughes, L.F.; Lin, S.Y. Current practice patterns in tonsillectomy and perioperative care. Int. J. Pediatr. Otorhinolaryngol. 2004, 68, 779–7845. [CrossRef] 101. Dhiwakar, M.; Clement, W.A.; Supriya, M.; McKerrow, W. Antibiotics to reduce post-tonsillectomy morbidity. Cochrane Database Syst. Rev. 2012, 12, CD005607. 102. Al-Layla, A.; Mahafza, T.M. Antibiotics do not reduce posttonsillectomy morbidity in children. Eur. Arch. Otorhinolaryngol. 2013, 270, 367–370. [CrossRef] [PubMed] 103. Milder, E.A.; Rizzi, M.D.; Morales, K.H.; Ross, R.K.; Lautenbach, E.; Gerber, J.S. Impact of a new pr use with pediatric tonsillectomy. JAMA Otolaryngol. Head Neck Surg. 2015, 141, 410–416. [CrossRef] M.D.; Morales, K.H.; Ross, R.K.; Lautenbach, E.; Gerber, J.S. References Mitchell, R.M.; Mendez, E.; Schmitt, N.C.; Bhrany, A.D.; Futran, N.D. Antibiotic prophylaxis in patients undergoing head and y y p p antibiotics in the setting of oropharyngeal reconstruction: Less is more. Ann. Plast. Surg. 2016, 76, 663–667. [CrossRef] [PubMed] 84. Mitchell, R.M.; Mendez, E.; Schmitt, N.C.; Bhrany, A.D.; Futran, N.D. Antibiotic prophylaxis in patients undergoing head and neck free ﬂap reconstruction JAMA Otolaryngol Head Neck Surg 2015 141 1096–1103 [CrossRef] [PubMed] g p y g g 84. Mitchell, R.M.; Mendez, E.; Schmitt, N.C.; Bhrany, A.D.; Futran, N.D. Antibiotic prophylaxis in patients undergoing head and neck free ﬂap reconstruction. JAMA Otolaryngol. Head Neck Surg. 2015, 141, 1096–1103. [CrossRef] [PubMed] 85. Pool, C.; Kass, J.; Spivack, J.; Nahumi, N.; Khan, M.; Babus, L.; Teng, M.S.; Genden, E.M.; Miles, B.A. Increased surgical site infection rates following clindamycin use in head and neck free tissue transfer. Otolaryngol. Head Neck Surg. 2016, 154, 272–278. [CrossRef] [PubMed] 86. Johnson, J.T.; Yu, V.L.; Myers, E.N.; Wagner, R.L.; Sigler, B.A. Cefazolin vs moxalactam? A double-blind randomized trial of cephalosporins in head and neck surgery. Arch. Otolaryngol. Head Neck Surg. 1986, 112, 151–153. [CrossRef] [PubMed] 87. Robbins, K.T.; Byers, R.M.; Cole, R.; Fainstein, V.; Guillamondegui, O.M.; Schantz, S.P.; Weber, R.S.; Wolf, P.; Goepfert, H. Wound prophylaxis with metronidazole in head and neck surgical oncology. Laryngoscope 1988, 98, 803–806. [CrossRef] 88. Sawyer, R. Clinical implications of metronidazole antianaerobic prophylaxis in major head and neck surgical procedures. Ear Nose Throat J. 1988, 67, 655–656, 658, 660–662. 89. Panosetti, E.; Lehmann, W.; Smolik, J.C. Antimicrobial prophylaxis in head and neck surgery. ORL J. Otorhinolaryngol. Relat. Spec. 1987, 49, 152–156. [CrossRef] [PubMed] 90. Johnson, J.T.; Schuller, D.E.; Silver, F.; Gluckman, J.L.; Newman, R.K.; Shagets, F.W.; Snyderman, N.L.; Leipzig, B.; Wagner, R.L. Antibiotic prophylaxis in high-risk head and neck surgery: One-day vs. ﬁve-day therapy. Otolaryngol. Head Neck Surg. 1986, 95, 554–557. [CrossRef] [PubMed] 91. Gehanno, P.; Moisy, N.; Guedon, C. Cefotaxime in the prophylaxis of otorhinolaryngological cancer surgery: Long term versus short term administration, results of a multicentre study. Drugs 1988, 35 (Suppl. 2), 111–115. [CrossRef] 92. Rodrigo, J.P.; Alvarez, J.C.; Gomez, J.R.; Suarez, C.; Fernandez, J.A.; Martinez, J.A. Comparison of regimens in clean-contaminated head and neck surgery. Head Neck 1997, 19, 188–193. [CrossRef] 92. Rodrigo, J.P.; Alvarez, J.C.; Gomez, J.R.; Suarez, C.; Fernandez, J.A.; Martinez, J.A. Comparison of three prophylactic antibiotic regimens in clean-contaminated head and neck surgery. Head Neck 1997, 19, 188–193. 108. Gelardi, M.; Marchisio, P.; Caimmi, D.; Incorvaia, C.; Albertario, G.; Bianchini, S.; Caimmi, S.; Celani, C.; Esposito, S.; Fattizzo, M.; et al. Pathophysiology, favoring factors, and associated disorders in otorhinosinusology. Pediatr. Allergy Immunol. 2012, 23 (Suppl. 22), 5–16. [CrossRef] [PubMed] 107. Esposito, S.; Marchisio, P.; Capaccio, P.; Bellasio, M.; Semino, M.; Dusi, E.; Colombo, R.; Pignataro, L.; Principi, N. Risk factors for bacteremia during and after adenoidectomy and/or adenotonsillectomy. J. Infect. 2009, 58, 113–118. [CrossRef] [PubMed] y J y g 111. Committee on Infectious Diseases, American Academy of Pediatrics; Kimberlin, D.W.; Barnett, E.D.; Lynﬁeld, R.; Sawyer, M.H. Red Book: 2021–2024 Report of the Committee on Infectious Diseases, 32nd ed.; American Academy of Pediatrics Ed.: Elk Grove Village, IL, USA, 2021. [ ] 110. Sánchez-Carrión, S.; Prim, M.P.; De Diego, J.I.; Sastre, N.; Peña-García, P. Utility of prophylactic antibiotics in pediatric adenoidec- tomy Int J Pediatr Otorhinolaryngol 2006 70 1275–1281 [CrossRef] [PubMed] ( pp ), [ ] [ ] 109. Della Vecchia, L.; Passali, F.M.; Coden, E. Complications of adenotonsillectomy in pediatric age. Acta Biomed. 2020, 91, 48–53. [PubMed] 110. Sánchez-Carrión, S.; Prim, M.P.; De Diego, J.I.; Sastre, N.; Peña-García, P. Utility of prophylactic antibi tomy. Int. J. Pediatr. Otorhinolaryngol. 2006, 70, 1275–1281. [CrossRef] [PubMed] References Impact of a new practice guideline on antibiotic onsillectomy. JAMA Otolaryngol. Head Neck Surg. 2015, 141, 410–416. [CrossRef] [PubMed] 104. Or1owski, K.; Lisowska, G.; Misio1ek, H.; Paluch, Z.; Misio1ek, M. The efﬁcacy of perioperative antibiotic therapy in tonsillectomy patients. Adv. Clin. Exp. Med. 2016, 25, 493–503. [CrossRef] [PubMed] 105. Padia, R.; Olsen, G.; Henrichsen, J.; Bullock, G.; Gale, C.; Stoddard, G.; Ott, M.; Srivastava, R.; Meier, J.D. Hospital and surgeon adherence to pediatric tonsillectomy guidelines regarding perioperative dexamethasone and antibiotic administration. Otolaryngol. Head Neck Surg. 2015, 153, 275–280. [CrossRef] [PubMed] y g g [ ] [ ] 106. Mitchell, R.B.; Archer, S.M.; Ishman, S.L.; Rosenfeld, R.M.; Coles, S.; Finestone, S.A.; Friedman, N.R.; Giordano, T.; Hildrew, D.M.; Kim, T.W.; et al. Clinical Practice Guideline: Tonsillectomy in Children (Update). Otolaryngol. Head Neck Surg. 2019, 160 (Suppl. 1), S1–S42. [CrossRef] [PubMed] 19 of 19 19 of 19 Antibiotics 2022, 11, 382 g y y f 108. Gelardi, M.; Marchisio, P.; Caimmi, D.; Incorvaia, C.; Albertario, G.; Bianchini, S.; Caimmi, S.; Celani, C.; Esposito, S.; Fattizzo, M.; et al. Pathophysiology, favoring factors, and associated disorders in otorhinosinusology. Pediatr. Allergy Immunol. 2012, 23 (Suppl. 22), 5–16. [CrossRef] [PubMed] ( pp ) 109. Della Vecchia, L.; Passali, F.M.; Coden, E. Complications of adenotonsillectomy in pediatric age. Acta Biomed. 2020, 91, 48–53. [PubMed] 110. Sánchez-Carrión, S.; Prim, M.P.; De Diego, J.I.; Sastre, N.; Peña-García, P. Utility of prophylactic antibiotics in pediatric adenoidec- tomy. Int. J. Pediatr. Otorhinolaryngol. 2006, 70, 1275–1281. [CrossRef] [PubMed] 110. Sánchez-Carrión, S.; Prim, M.P.; De Diego, J.I.; Sastre, N.; Peña-García, P. Utility of prophylactic tomy. Int. J. Pediatr. Otorhinolaryngol. 2006, 70, 1275–1281. [CrossRef] [PubMed] 111. Committee on Infectious Diseases, American Academy of Pediatrics; Kimberlin, D.W.; Barnett, E.D.; Lynﬁeld, R.; Sawyer, M.H. Red Book: 2021–2024 Report of the Committee on Infectious Diseases, 32nd ed.; American Academy of Pediatrics Ed.: Elk Grove Village, IL, USA, 2021.
https://openalex.org/W4361888747	https://figshare.com/articles/journal_contribution/Supplementary_Figure_3_from_Differentiation_Syndrome_with_Ivosidenib_and_Enasidenib_Treatment_in_Patients_with_Relapsed_or_Refractory_IDH-Mutated_AML_A_U_S_Food_and_Drug_Administration_Systematic_Analysis/22476198/1/files/39927705.pdf	English	null	Supplementary Figure 1 from Differentiation Syndrome with Ivosidenib and Enasidenib Treatment in Patients with Relapsed or Refractory IDH-Mutated AML: A U.S. Food and Drug Administration Systematic Analysis	null	2,023	cc-by	126	Supplementary Figure 3. Blood count trajectory for patients with or without DS. Mean ± standard error white blood cell count, absolute neutrophil count, and platelet count over time with ivosidenib (A-C) and enasidenib (D-F) according to whether or not patients experienced DS. Supplementary Figure 3. Blood count trajectory for patients with or without DS. Mean ± standard error white blood cell count, absolute neutrophil count, and platelet count over time with ivosidenib (A-C) and enasidenib (D-F) according to whether or not patients experienced DS. Supplementary Figure 3. Blood count trajectory for patients with or without DS. Mean ± standard error white blood cell count, absolute neutrophil count, and platelet count over time with ivosidenib (A-C) and enasidenib (D-F) according to whether or not patients experienced DS.
https://openalex.org/W4242622171	https://www.qeios.com/read/89IPW2/pdf	English	null	Vasoactive	Definitions	2,020	cc-by	56	Qeios · Definition, February 7, 2020 Open Peer Review on Qeios Open Peer Review on Qeios Vasoactive National Cancer Institute National Cancer Institute Qeios ID: 89IPW2 · https://doi.org/10.32388/89IPW2 Open Peer Review on Qeios Source National Cancer Institute. Vasoactive. NCI Thesaurus. Code C94301. Able to have an effect on blood vessels. Qeios ID: 89IPW2 · https://doi.org/10.32388/89IPW2 1/1
https://openalex.org/W2626321948	https://periodicos2.uesb.br/index.php/estudosdalinguagem/article/download/1259/1085	Portuguese	null	Percepções das práticas educativas presenciais e digitais (Perceptions of interaction in digital environmental for education)	Estudos da Língua(gem)	2,014	cc-by	5,223	Danilo Rodrigues César* Universidade Federal do Triângulo Mineiro – UFTM/Brasil * Sobre os autores ver página 177. Daniervelin Renata Marques Pereira* Universidade Federal do Triângulo Mineiro – UFTM/Brasil Danilo Rodrigues César* Universidade Federal do Triângulo Mineiro – UFTM/Brasil Estudos da Língua(gem) Estudos da Língua(gem) 1 Utilizamos neste artigo algumas observações e considerações desenvolvidas em nossa pesquisa de doutorado (PEREIRA, 2013). RESUMO Apresentamos aqui um estudo de enunciados produzidos por professores e alunos num contexto de disciplina online (Educação a Distância), contexto este que vem se tornando cada vez mais um “estilo de vida” diante das práticas da sociedade atual. Situamos nosso interesse nos conflitos entre esses dois sujeitos, que percebem de maneiras diferentes a prática digital, e também na diferença essencial e afetante entre a modalidade presencial e a distância. À luz da Semiótica Tensiva, analisamos alguns enunciados, estabelecendo por horizonte as correlações entre foco e apreensão e as categorias presença/ausência no “campo de presença”, que orientam os modos de existência dos sujeitos e objetos de um ponto de vista tensivo. Nessa perspectiva, privilegiamos a subjetividade, manifestada em discurso, sobre as práticas educativas digitais, na sua relação com a modalidade presencial. Estudos da Língua(gem) Vitória da Conquista v. 12, n. 2 p. 161-177 dezembro de 2014 Estudos da Língua(gem) Vitória da Conquista v. 12, n. 2 p. 161-177 dezembro de 2014 Daniervelin Renata Marques Pereira e Danilo Rodrigues César 162 PALAVRAS-CHAVE: Percepção. Práticas pedagógicas. Discursos. Semiótica. ABSTRACT We present a study of statements produced by teachers and students in the context of online course (Distance Education). This context is becoming increasingly a “lifestyle” in the practices of contemporary society. We situate our interest in the conflicts between these two subjects (teachers and students) as they perceive differently digital practice, and also the essential difference between the face to face and distance interaction. We follow the perspective of Tensive Semiotics to analyze some statements as correlations between focus and apprehension and categories presence/absence in “field presence”, which guide modes of existence of the subjects and objects by a tensive view. From this perspective, we focus on the subjectivity manifested in discourse about digital educational practices in its relation to the face to face modality. KEYWORDS: Perception. Pedagogical practices. Discourses. Semiotics. 1 A percepção1 Problematizamos, neste texto, a noção de percepção e, apesar da complexidade do termo não nos permitir uma conclusão sobre a questão, empregamos alguns conceitos propostos pela Semiótica Francesa para encontrar marcas da percepção de práticas pedagógicas em ambiente digital. A partir dos estudos de Merleau-Ponty (1983), a Semiótica Francesa lança luz sobre a noção de percepção sensível, destaque no quadro epistemológico da Semiótica das Paixões (GREIMAS; FONTANILLE, 1993) e, mais recentemente, na Semiótica Tensiva, cujos estudiosos precursores são os franceses Jacques Fontanille e Claude Zilberberg. Esta corrente de estudos não pretende substituir a semiótica francesa “clássica”, mas acrescentar noções mais abordagens de modulações e gradações próprias do ponto de vista da complexidade, da tensividade, da afetividade, da percepção. Percepções das práticas educativas presenciais e digitais 163 Nessa visão, o sujeito compreende o mundo ao mesmo tempo em que é compreendido por ele. Dessa forma, a experiência de percepção do corpo, na concepção merleupontiana, é extensiva à percepção do mundo. “O corpo encerra uma perspectiva, uma ancoragem no instante e no espaço, e o sentido transcorre pelos canais perceptivos sem qualquer solução de continuidade entre ser-observador e ser-observado” (TATIT, 1995, p. 163). A pressuposição da noção de sentir nessa relação perceptiva orientada pelo corpo explica as flutuações entre aquilo que nos faz parar para compreender e apreender e aquilo que nos tem como resultado a continuidade. Em busca da compreensão sobre a percepção que os sujeitos têm do espaço e de seus objetos, levantamos inicialmente algumas questões: Como ter acesso a percepções de sujeitos na experiência vivida durante as práticas educativas? Essas percepções são simplesmente apreendidas no contato com o “mundo natural”/digital? Diante da dimensão discursiva do objeto selecionado aqui para análise (enunciados de professores e alunos expostos a práticas digitais de ensino e de aprendizagem), pensamos que a experiência não se dá fora da linguagem, mas é construída por ela. Nesse sentido, apoiamo-nos nas palavras de Silva (1995, p. 31- 32) para afirmarmos que a percepção do mundo pode se dar numa metamorfose, cuja forma resultante, o signo, contém, em si, marcas dessa experiência anterior: [...] a metamorfose, de um modo geral, implica destruição, ou melhor, desconstrução de uma forma anterior, que não desaparece totalmente. Dela ficam traços na forma nova, que são os elementos em que minha hipótese de trabalho se firma a fim de pensar um pouco os fundamentos da linguagem. 1 A percepção1 Exagerando um pouco, diria que na transformação de uma experiência em signo ocorre uma metamorfose fundadora. Metamorfose porque não fica tudo da experiência no signo, uma forma nova que é uma redução; fundadora porque está nas raízes da semiose. (SILVA, 1995, pp. 31-32) [...] a metamorfose, de um modo geral, implica destruição, ou melhor, desconstrução de uma forma anterior, que não desaparece totalmente. Dela ficam traços na forma nova, que são os elementos em que minha hipótese de trabalho se firma a fim de pensar um pouco os fundamentos da linguagem. Exagerando um pouco, diria que na transformação de uma experiência em signo ocorre uma metamorfose fundadora. Metamorfose porque não fica tudo da experiência no signo, uma forma nova que é uma redução; fundadora porque está nas raízes da semiose. (SILVA, 1995, pp. 31-32) Poderíamos, assim, afirmar que a metamorfose no digital ocorre a partir do processo de transformação das experiências pedagógicas tradicionais do presencial? Retomaremos essa questão nas considerações finais. Daniervelin Renata Marques Pereira e Danilo Rodrigues César 164 Do ponto de vista semiótico, segundo Fontanille (2007, p. 97), “a percepção já é uma linguagem, pois ela é significante”. Sendo assim, a partir da afirmação de Merleau-Ponty (1990, p. 92): “perceber é tornar algo presente a si com a ajuda do corpo”, Fontanille (2007, p. 97) propõe: “enunciar é tornar algo presente a si com a ajuda da linguagem”. É, então, na tomada de posição pelo ato de enunciar que a presença, “primeiro modo de existência da significação” (FONTANILLE; ZILBERBERG, 2001, p. 123), constitui-se em relação a um corpo que o sente. Para refletirmos ainda sobre a percepção, tomamos as palavras de Cassirer (2004, p. 67), que tenta expressar a posição do sujeito diante da experiência: [...] os indissolúveis conteúdos e configurações da percepção [...] não oferecem [...] nenhuma consistência ou ponto de apoio. Eles não se sujeitam a uma ordem abrangente e rígida, não suportam jamais o caráter da determinação verdadeiramente unívoca, mas, ao contrário, uma vez apreendidos em sua existência imediata, apresentam-se como algo tão somente fluido, fugidio, que resiste a toda tentativa de diferenciar nele mesmo ‘limites’ verdadeiramente nítidos e exatos. (CASSIRER, 2004, p. 67) (grifo no original). A definição não poderia ser mais coerente, no que diz respeito à posição fluida das interações em ambiente digital. 2 Consideramos, pela perspectiva estruturalista da linguística, que a modalidade presencial está relacionada à modalidade a distância, que o próximo só existe em relação ao distante, e vice-versa. 3 O docente e os discentes cujos nomes omitimos aqui permitiram a utilização dos dados por motivo da pesquisa. 4 “Campo de presença”, a partir de Merleu-Ponty (1983), é considerado como o “domínio espácio temporal em que se exerce a percepção, e, por outro, as entradas, as estadas, as saídas e os retornos que, ao mesmo tempo, a ele devem seu valor e lhe dão corpo” (FONTANILLE; ZILBERBERG, 2001, p. 125). 1 A percepção1 À luz da Semiótica Tensiva, pretendemos analisar alguns enunciados, estabelecendo por horizonte as correlações entre foco Daniervelin Renata Marques Pereira e Danilo Rodrigues César 166 e apreensão e a categoria presença/ausência, que orientam os modos de existência dos sujeitos e objetos de um ponto de vista tensivo. 1 A percepção1 Levando em conta a problemática da percepção na/da Educação a Distância (EaD), no quadro das mudanças incutidas nas relações entre a tríade “professor, aluno, conteúdo”, pesquisas (MATTE, 2009; TORI, 2010) mostram que a classificação dialética e simplista, próximo (ensino presencial) vs. distante (EaD), pode ser facilmente invertida e modulada, já que tal visão baseia-se unicamente na posição dos corpos no espaço físico, sem reconhecer a interdependência dessas categorias em cada modalidade2. Posicionando-se diferentemente quanto a essa concepção, alguns autores defendem a percepção da proximidade como aspecto alcançado pela interatividade de uma comunicação; ou seja: o potencial de dar voz, Percepções das práticas educativas presenciais e digitais 165 igualmente, a todos os atores/falantes em cena. Sobre isso, trazemos palavras do autor de Educação sem distância: “o limite da tecnologia será conseguir reproduzir a mesma sensação de proximidade oferecida pela presença física [...]” (TORI, 2010, p. 62), o que ele assume estar longe de ser alcançado. Todavia, esse autor comprova, pela análise de determinadas situações comunicativas, que a abertura à participação em certos contextos de educação online, como por videoconferência e chat, pode superar a “sensação de presença” provocada em eventos presenciais, como em uma aula magna expositiva, em que o poder da palavra é dado a apenas uma pessoa. Pesquisas como essa demonstram o quão superficiais são concepções comuns como, por exemplo: os alunos de EaD interagem apenas com a tela do computador. E não está ele mesmo permeado de traços humanos? Recortamos, para esta análise, alguns enunciados produzidos por sujeitos inseridos no contexto de ensino e de aprendizagem em ambiente digital. Trata-se de uma disciplina online para ensino da teoria Semiótica Francesa durante um semestre letivo3, oferecida a estudantes de graduação em Letras (da Universidade Federal de Minas Gerais). Interessa-nos aqui as percepções desses sujeitos, educador e educandos, da situação na qual estão inseridos por um dever e, ao mesmo tempo, por uma escolha diante de outras disciplinas igualmente optativas na grade escolar do seu curso. O que diferencia, tecnicamente e a priori, as disciplinas na modalidade de ensino a distância das ofertadas na modalidade presencial é a localização espaço-temporal dos sujeitos em comunicação. Não é à toa, então, que as noções de distanciamento e proximidade entre esses sujeitos estão presentes em seus enunciados, marcando a sua percepção do campo de presença4. 2 Campo de presença Se, por um lado, é possível observar o quadro típico dos sujeitos no contexto pedagógico como em busca dos objetos ensino e aprendizagem e dos valores a ele associados, constituindo-se essa junção sua existência para a Semiótica Francesa ou Greimasiana; por outro e de forma complementar, a percepção dos sujeitos, objeto da Semiótica Tensiva, configura-se como panorama que privilegia a complexidade na forma de se ver esse universo pedagógico. “O sujeito [para a semiótica tensiva] constrói o objeto com sua percepção e, ao fazê-lo, se constrói a si mesmo como sujeito daquela percepção” (MANCINI et al, 2007, p. 296). Por essa perspectiva, a presença sensível orienta a relação entre o sujeito e o seu entorno. “A presença é para o sujeito sensível o que o valor é para o sujeito narrativo (FONTANILLE, 2001, p. 5). Precisamos esclarecer um pouco o que é esse campo de presença em que o sujeito exerce o foco ou sua apreensão. Esse campo, para a semiótica, é o espaço-tempo discursivo percebido pelo sujeito. “[...] O campo do discurso é ao mesmo tempo campo de presença e campo posicional. Toda presença marcada no campo é dotada de uma posição de referência (o eu)” (FONTANILLE, 1999, p. 30). O sujeito pode conceber de duas formas o campo de presença: como aberto ou fechado. Se ele for sentido como aberto, a percepção será entendida como apreensão da realidade e estará fundado na extensidade, dado que a extensidade do campo de presença pode variar infinitamente fazendo com que mais e mais objetos passem a habitar o campo de presença e assim, quanto mais amplo, maior não apenas a extensidade, o que é evidente, mas também a intensidade. Por outro lado, o campo de presença pode ser fechado. Nesse caso, a percepção é guiada pelo foco: no lugar de permitir que o campo de presença varie infinitamente, ele delimita a região no interior da qual o foco, a intensidade, é Percepções das práticas educativas presenciais e digitais 167 maior (ZILBERBERG; FONTANILLE, 2001). Os dois modos de funcionamento do campo de presença são possíveis, e na verdade nós o alteramos constantemente no discurso de nossa comunicação. maior (ZILBERBERG; FONTANILLE, 2001). Os dois modos de funcionamento do campo de presença são possíveis, e na verdade nós o alteramos constantemente no discurso de nossa comunicação. Por essa lógica, podemos pensar não só em categorias, mas em graus de presença. 5A paixão a que fazemos menção é a paixão semiótica, de acordo com o conceito apresentado em Semiótica das Paixões, por Greimas e Fontanille (1993). 2 Campo de presença Se o foco e a apreensão que o sujeito tem do objeto forem tônicos, a presença é dada como plena, pois realizante. Por outro lado, se o foco for tônico, isto é, se o sujeito percebe o objeto sem, no entanto, apreendê-lo, compreendê-lo, ela será atualizante, pois instaura a falta, isto é, o desejo de apreender o objeto, de entender o que se passa, de compreender melhor a situação. Nessas duas situações o objeto em relação ao sujeito está na dêixis da presença. Entretanto, pode ocorrer de tanto o foco do sujeito quanto a apreensão que ele tem dos objetos serem átonos. Nesse caso, a presença será virtualizante, pois de fato não existe. A última posição é aquela da surpresa em que a apreensão é tônica, mas o foco é átono. Nesse caso, a presença será potencializante, pois ela poderá se constituir se o sujeito focar o objeto. Nas duas últimas posições encontramos a dêixis da ausência. No Quadro 1 a seguir, identificamos essas combinações: Presença Foco Tônico Átono Apreensão Tônica Plenitude (realizante) Inanidade (potencializante) Átona Falta (atualizante) Vacuidade (virtualizante) Quadro 1 – Correlação entre foco e apreensão da presença (baseado em Zilberberg; Fontanille, 2001, 131) As modulações da presença e da ausência mostram, dessa maneira, as relações possíveis entre o sujeito e o objeto tensivos. “A modalização existencial: a plenitude é realizante, a falta é atualizante, a vacuidade é virtualizante e a inanidade é potencializante” (FONTANILLE; ZILBERBERG, 2001, p. 131). A partir desses conceitos e dos estudos semióticos sobre o campo de presença, tentaremos vislumbrar a percepção em alguns recortes da Daniervelin Renata Marques Pereira e Danilo Rodrigues César 168 interação entre educador e educandos na referida disciplina online. Eles podem nos mostrar, sobretudo, quão afetantes são as categorias de tempo e de espaço para o foco e a apreensão na percepção dos sujeitos. 3 Percepção da diferença e diferença na percepção Torna-se interessante direcionar, então, o estudo para o campo de presença da prática educativa digital, destacando a rede de categorias percebidas pelos sujeitos como interferentes em seu ponto de vista. Encontramos pistas para essa investigação nos dados que analisaremos adiante. Citamos a seguir dois enunciados postados por duas alunas em um fórum já na fase final da disciplina online sobre Semiótica Francesa para estudantes do curso de Letras: É impressionante como os conceitos, com os exemplos explicados e explicitados, ficam mais claro. As atividades práticas são muito importantes MESMO! Essa disciplina poderia ser ofertada sem ser à distância... Apesar de os slides serem ótimos e muito didáticos, o tema é muito abstrato. Fiquei perdida sem os exemplos! (Aluna 1) Tb acho que poderia ser ofertada presencial ou semi. É muito bom este acesso que temos aqui as questoes e duvidas dos alunos. Mas acho que pessoalmente tb teríamos muito a ganhar (Aluna 2, em resposta à Aluna 1). Em um regime de concessão (manifestado pela função: embora x, y), os dois enunciados, em sintonia, permitem a seguinte interpretação: embora haja profusão de exemplos e slides didáticos, a disciplina poderia estar na modalidade de ensino presencial. Ainda que a falta, relativa ao cognitivo, seja suprida (“Fiquei perdida sem os exemplos!”), é mantida uma outra paixão5 em suspenso, a do desejo (“Essa disciplina poderia ser ofertada sem ser à distância...”). Essa paixão, que Descartes (s/d) associa ao futuro, revela a realização do sujeito encontra-se como potencialidade. Percepções das práticas educativas presenciais e digitais 169 Relaciona-se, de certa forma, a complexidade e abstração da teoria a uma dificuldade pressuposta em se cursar uma disciplina que estabelece a distância espacial entre os sujeitos. De forma contrária, os exemplos da prática tornam a apreensão da teoria (objeto buscado) mais fácil e o contato presencial (da ordem do visível, mais concreto) agregaria também o valor de segurança e afetividade que os sujeitos parecem considerar ausentes naquele ambiente. Em esquema, essas palavras podem ser sintetizadas assim: Figura 1 – Esquema a partir do discurso das alunas 1 e 2 Figura 1 – Esquema a partir do discurso das alunas 1 e 2 Ao contrário do que pode parecer indicar a Figura 1, os termos se encontram em relação de gradação, e não categórica. 3 Percepção da diferença e diferença na percepção As correlações entre abstrato/concreto e teoria/distância espacial, e entre concreto/ simples e prática/proximidade espacial se organizam em um contínuo entre tensão e relaxamento, entre ausência e presença, o que só pode ser aprendido pelos traços discursivos já sinalizados nos enunciados dos sujeitos. Em resposta às alunas, a professora afirma no mesmo fórum: De fato, o ideal é que essa disciplina fosse oferecida em 1 ano, mas pela atual falta de professores, fica muito difícil organizar entre mim e a professora B o oferecimento de semiótica 1 e semiótica 2. Em um semestre fica bem mais difícil de acompanhar. Por isso eu ainda prefiro a disciplina online, onde podemos interagir muito mais. Mesmo assim é pouco tempo. […] (Professora A) Daniervelin Renata Marques Pereira e Danilo Rodrigues César 170 Numa percepção mais aberta do campo de presença, ou “bastidores” da cena enunciativa, o sujeito professor associa a apreensão da teoria a uma extensão maior do contato com o tema e com as pessoas no tempo. Encontramos aqui um regime de implicação (pela função: se... então...): para o professor, se não há como garantir um número maior de disciplinas e um semestre letivo é pouco para a adquisição da teoria, a disciplina online seria escolhida por trazer a possibilidade de maior interação. Dessa forma, os alunos teriam o tempo necessário, passada a surpresa do surgimento do objeto, para compreendê-lo melhor e aproximá-lo do seu campo posicional (eu). Podemos representar, de acordo com o ponto de vista da professora, as modalidades a distância e presencial como dois segmentos (Figura 2). No primeiro, de tracejado curto, os pontos de comunicação ou interação são mais recorrentes, em maior número no tempo, de acordo com a visão da professora. Figura 2 – Representação dos discursos de professor e alunos em segmentos Figura 2 – Representação dos discursos de professor e alunos em segmentos No segundo, de tracejado mais longo, temos a representação de encontros em menor quantidade no tempo, mas mais notórios, pelo menos para os alunos. Enquanto os encontros no meio digital são apreendidos por uma maior extensidade, na visão do professor, a modalidade de ensino presencial é percebida como mais intensa, mais tônica pelos alunos, conforme a curva tensiva da Figura 3. 3 Percepção da diferença e diferença na percepção 171 Percepções das práticas educativas presenciais e digitais Figura 3 – Curva inversa para representar a percepção do ensino-aprendizagem no presencial e no digital Figura 3 – Curva inversa para representar a percepção do ensino-aprendizagem no presencial e no digital No gráfico (Figura 3), temos delineada a percepção dos sujeitos do presencial como lugar mais afetivo (visão dos alunos), mas de contato menos recorrente (visão do professor), contrariamente à percepção do meio digital, o que gera uma relação inversa. O que podemos destacar aí são dois tipos de concepções: 1) de que a maior interação no tempo garante a apreensão do conteúdo e 2) de que o contato presencial, mesmo que pouco extenso, é um espaço ideal para a aprendizagem de temas complexos. Precisamos, sem dúvida, de um corpus maior de dados em que essas concepções sejam melhor definidas e comprovadas, mas é fato que a introdução da Educação a Distância nas práticas pedagógicas desperta interessantes questões sobre a aprendizagem significativa, como resultado da percepção realizante pelos alunos. Por mais que esse modelo seja criticado, ajustamentos têm sido propostos a cada interação (PEREIRA, 2013), no caminho de processos de ensino-aprendizagem mais adequados ao contexto da sala de aula (enquanto prática forma ou não-formal). Não se trata de verificar se há um lugar melhor que o outro, mas de descrever e refletir sobre a percepção dos sujeitos no contato com esses meios e investigar que efeitos isso tem na sua relação com os outros Daniervelin Renata Marques Pereira e Danilo Rodrigues César 172 sujeitos, com os objetos e com o meio. Dessa forma, como resultado possível e esperado, poder-se-ia, pela pesquisa, impulsionar a criação de estratégias didáticas para tratar a questão de forma mais adequada a cada situação. Vejamos outro exemplo. Ao empregar a metalinguagem (uso da linguagem técnica da Semiótica para analisar a interação na disciplina), a professora explicita que sua relação com o objeto – uno – é direta e implicativa: É interessante pensar, por exemplo, que a participação nesses fóruns, embora seja muito importante para nossa interação mais direta e esclarecimento de dúvidas e troca de ideias, é, para alguns, simplesmente uma questão de ganhar nota e presença. Para estes, trata-se de um fazer motivado por uma manipulação por tentação (querer tirar uma nota melhor) ou até por intimidação (dever não ser reprovado). 3 Percepção da diferença e diferença na percepção Como eu tenho um olhar semiótico sobre isso, para mim realmente não importa o motivo: se a pessoa se manifesta no fórum, cumpre o contrato (merecendo a sanção positiva) e permite a interação, esclarecimento coletivo de dúvidas e troca de ideias, que é meu objetivo maior. (Professora A) Para além da discussão da localização espaço-temporal da interação, ênfase das falas analisadas anteriormente, o sujeito, nesse enunciado, recorta o seu objeto e a imagem que ele tem dos pontos de vista dos alunos. Na perspectiva do enunciador, tanto para os alunos como para o professor, o modo de eficiência, isto é, a forma de penetração do objeto no campo de presença do sujeito, estaria mais para o atingir (esperado), em oposição ao sobrevir (inesperado), uma vez que ambos agem de acordo com um objetivo a alcançar. Nessa visão, os alunos buscam a própria aprovação e o professor a aquisição do conhecimento pela interação. Ressalta-se, nesse trecho ainda, a objetividade do enunciador- professor, pela qual ele aproxima o objeto-valor buscado independente dos objetivos que movem os sujeitos-alunos. Porém, indiretamente, subjaz a esse discurso a subjetividade daquele que, ao mirar o acesso ao conhecimento e a interatividade, manifesta sua preocupação com a Percepções das práticas educativas presenciais e digitais 173 natureza formativa da educação. A percepção do que ele espera e do que os outros à sua volta esperam é também uma sensibilidade latente nesse trecho e parece resultar da interação entre os sujeitos no tempo-espaço das aulas. Zilberberg (2008, p. 7), ao tratar do espaço tensivo, aborda um conceito interessante para nossa análise: Para a extensidade, orientada para os estados de coisas, que tem a ver com a densidade do campo de presença: se as quantidades são limitadas, nós diremos que é a modalidade da concentração que é válida; se ocorre o inverso, diremos que se trata da modalidade da difusão (grifos do autor, tradução nossa). Notamos, pelo discurso da professora, que o fórum, como parte do ambiente digital em que se dão as interações (Moodle6), é visto pelo professor como local da concentração das funções de orientação quanto ao conteúdo, como podemos observar neste trecho: “importante para nossa interação mais direta e esclarecimento coletivo de dúvidas e troca de ideias”. A concentração, nesse caso, pode ser uma modalidade de controle e organização do campo de presença. 6 Plataforma muito utilizada como apoio na apresentação em ambiente digital, com diversos gêneros disponíveis: chat, fórum, questionário, entre outros. Disponível em: <https://moodle. org/?lang=pt_br>. Acesso em 25 jun. 2014. 3 Percepção da diferença e diferença na percepção Nos enunciados dos alunos, parece-nos que o espaço presencial se apresenta como mais concentrado e, por isso, mais controlável, o que poderia “contagiar” também o objeto de estudos. Por outro lado, as novidades do ambiente digital, tomado como difuso por esses mesmos sujeitos, somam-se à noção de dificuldade em dominar a teoria. Dessa forma, notamos que o valor e os efeitos de sentido não estão nos objetos a priori, mas são conferidos a eles pelos sujeitos, por suas expectativas e percepções. 4 Considerações finais Resumindo esses pontos de vista analisados pelo viés da Semiótica Tensiva, as alunas veem a sala de aula presencial como facilitadora no contato com os temas do ensino, principalmente quando Daniervelin Renata Marques Pereira e Danilo Rodrigues César 174 são complexos. Já a professora, por enfrentar problemas contextuais como a quantidade de tempo e de professores envolvidos na abordagem do tema que ensina, vê no digital uma forma de propiciar maior contato com o conhecimento e esclarecimento de dúvidas. Elas percebem o fenômeno de formas diferentes porque estão em ângulos distintos da prática educativa, e o sentem diferentemente. Dessa forma, o objeto recebe “cores” e “tons” dependentes do olhar e, na maioria das vezes, independente da aquisição dele. As formas de percepção desses sujeitos se encontram em um contínuo entre proximidade e distanciamento espaço-temporal no contato pessoal e no modo de contato com o objeto ensinado. Enquanto a sala de aula presencial é apreendida pelos alunos e professores como da ordem do habitual, do rotineiro; o espaço digital, para o ensino-aprendizagem, soa ainda novo e estranho, sendo, por isso, focado. A dificuldade dos sujeitos, especialmente dos alunos, para se inserirem na programação dada, dentro de suas regularidades, já foi verificada por nós (PEREIRA, 2010, 2013) e nos levou a cogitar que o problema se localiza essencialmente no contrato fiduciário (crença) e relação afetiva entre os sujeitos, o que pudemos comprovar em tese de doutorado (PEREIRA, 2013). A afetividade, no ponto de vista das alunas, parece explicar e justificar a preferência pelo espaço presencial, embora essa afirmação demande ainda mais estudos de uma quantidade maior de enunciados produzidos por diferentes sujeitos. A professora não manifesta no digital uma falta de afetividade, mas privilegia os resultados de acordo com o objetivo da sua prática pedagógica. Não acreditamos, no entanto, que professor e alunos se localizam em polos contrários em relação ao sensível e ao inteligível. Eles se movimentam num contínuo entre concepções que, muitas vezes, variam em relação ao objetivo. Os sintomas passionais provocados por essa variação alertam para a importância do elemento sensível para tornar os objetos e sujeitos presentes e reunidos no campo de presença, o que deve ser considerado nas reflexões entre os envolvidos e interessados no universo pedagógico. 4 Considerações finais Percepções das práticas educativas presenciais e digitais 175 Baseados nas análises feitas dos enunciados e na marcante presença das modalidades digital e presencial de ensino-aprendizagem em comparação, nos discursos, a metamorfose das práticas educativas implica uma transformação em que se perdem alguns aspectos de cada entidade (digital e presencial), como os gestos, associados em complementaridade à linguagem verbal. Essa perda dá-se no revestimento de formas novas, resultantes da metamorfose (BORDRON, 2009). As integralidades dos eventos se diluem no que, enunciado a enunciado, se constrói como projeto híbrido (COLAS-BRAISE; BEYAERT-GESLIN, 2009), o qual supõe a copresença em um conjunto misto das duas formas postas em contato em um processo contínuo de melhoração. Resta-nos, em análises futuras, acompanhar as metamorfoses dessas práticas, esperando que as práticas educativas busquem, nesse processo, e por estratégias diversas, a realização dos sujeitos envolvidos. CASSIRER, E. A filosofia das formas simbólicas. V. 2. Tradução de Cláudia Cavalcanti. São Paulo: Martins Fontes, 2004. COLAS-BLAISE, M.; BEYAERT-GESLIN, A. (Dir.). Le sens de la métamorphose. Limoges: Presses Universitaires de Limoges. 2009. REFERÊNCIAS BORDRON, J-F. Métamorphose et identités. In: COLAS-BLAISE, M.; BEYAERT-GESLIN, A. (Dir.). Le sens de la métamorphose. Limoges: Presses Universitaires de Limoges. 2009. p. 49-62. CASSIRER, E. A filosofia das formas simbólicas. V. 2. Tradução de Cláudia Cavalcanti. São Paulo: Martins Fontes, 2004. DESCARTES, R. As paixões da alma. São Paulo: Escala. S/d. DESCARTES, R. As paixões da alma. São Paulo: Escala. S/d. FONTANILLE, J. Modes du sensible et syntaxe figurative. Nouveaux Actes Sémiotiques. Limoges: PULIM, 61-62-63, 1999. FONTANILLE, J. Avant-propos. In: PARRET, H. Présences. Nouveaux actes sémiotiques. PULIM, 2001. FONTANILLE, J. Semiótica do discurso. Tradução de Jean Cristtus Portela. São Paulo: Contexto, 2007. Daniervelin Renata Marques Pereira e Danilo Rodrigues César 176 FONTANILLE, J.; ZILBERBERG, C. Tensão e Significação. Tradução de Ivã Carlos Lopes, Luiz Tatit e Waldir Beividas. São Paulo: Discurso Editorial: Humanitas/FFLCH/USP, 2001. GREIMAS, A. J.; FONTANILLE, J. Semiótica das paixões. Dos estados de coisas aos estados de alma. Tradução de Maria José Rodrigues Coracini. São Paulo: Ática, 1993. MANCINI, R.; TROTTA, M.; SOUZA, S. M. Análise semiótica da propaganda Hitler, da Folha de São Paulo. In: XIII Colóquio do CPS. 2007, p. 292-304. Disponível em: <http://www.uff.br/sedi/PDFS/2007_ Mancini_CPS2007FolhaSPHitler.pdf>. Acessado em 17 de julho de 2009. MATTE, A. C. F. Análise Semiótica da Sala de Aula no tempo da EAD. II Congresso de Tecnologias na Educação. Anais online, dezembro de 2009. Disponível em: <http://tecnologiasnaeducacao.pro.br/revista/ a1n1/pal3.pdf>. Acesso em: 24 abr. 2011. MERLEAU-PONTY, M. Phénoménologie de la perception. Paris, Gallimard, 1983. MERLEAU-PONTY, M. Phénoménologie de la perception. Paris, Gallimard, 1983. MERLEAU-PONTY, M. O primado da percepção e suas consequências filosóficas. Tradução de Constança Marcondes César. Campinas: Papirus, 1990. PEREIRA, D. R. M. Atividades didáticas para ensino de português em ambiente digital: uma análise semiótica. 2010. 319 f. Dissertação (Mestrado) – Universidade Federal de Minas Gerais, Belo Horizonte, 2010. PEREIRA, D. R. M. Semiótica e ensino: ajustamentos sensíveis em gêneros digitais da esfera educacional. 2013. 277 f. Tese (Doutorado) – Universidade de São Paulo, São Paulo, 2013. SILVA, I. A. Figurativização e metamorfose: o mito de narciso. São Paulo: Editora Unesp, 1995. TATIT, L. A semiótica e Merleau-Ponty. In: OLIVEIRA, Ana Claudia Mei Alves de; LANDOWSKI, E. Do inteligível ao sensível. São Paulo: EDUC. 1995. TORI, R. Educação sem distância. São Paulo: Editora Senac, 2010. ZILBERBERG, C. De l’événement. Tradução de M. L.Vissotto Paiva Diniz. Galáxia, n. 13, p. 13-28, jul. 2008. 177 Percepções das práticas educativas presenciais e digitais Recebido em outubro de 2013 Aceito em fevereiro de 2014 SOBRE OS AUTORES Daniervelin Renata Marques Pereira é doutora em Letras pela Universidade de São Paulo. Mestre em Linguística Aplicada pela Universidade Federal de Minas Gerais. Desenvolve pesquisa sobre o discurso da Educação a Distância mediada pelo computador, Ensino de leitura e escrita, Gêneros textuais, Cultura Livre. É professora da Universidade Federal do Triângulo Mineiro, onde atua no curso de Licenciatura em Educação do Campo. É membro do projeto Texto Livre (suporte à documentação em software livre), no âmbito do qual colabora, especialmente, com a Revista Texto Livre: linguagem e tecnologia, com o evento online Evidosol/Ciltec-online e com o projeto Português Livre. E-mail: daniervelin@gmail.com Danilo Rodrigues César é doutor em difusão do conhecimento pela Universidade Federal da Bahia, Mestre em educação pela mesma instituição. É Graduado em Tecnologia em Processamento de Dados (TPD), Licenciado em Matemática e Informática pelo CEFET-MG, Especialista em Rede de Computadores, Pós- graduado em Educação Profissional Técnica Integrada ao Ensino Médio na Modalidade Educação de Jovens e Adultos (PROEJA). É pós-doutorando em Educação na Universidade Federal de Uberlândia. É professor da Universidade de Uberaba e vinculado ao Mestrado em Educação. E-mail: danilorcesar@gmail.com
https://openalex.org/W3190082494	https://www.biorxiv.org/content/biorxiv/early/2022/02/07/2021.08.07.455508.full.pdf	English	null	The initial response of females towards congeneric males matches the propensity to hybridise in <i>Ophthalmotilapia</i>	bioRxiv (Cold Spring Harbor Laboratory)	2,021	cc-by	16,427	. CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint RESEARCH ARTICLE The initial response of females towards congeneric males matches the propensity to hybridise in Ophthalmotilapia. The initial response of females towards congeneric males matches the propensity to hybridise in Ophthalmotilapia. Van Steenberge Maarten1,2, Jublier Noémie3,4, Kéver Loïc3, Gresham Sophie1,5, Derycke Sofie1,6, Snoeks Jos2,7, Parmentier Eric3, Poncin Pascal4, Verheyen Erik1,5 Van Steenberge Maarten1,2, Jublier Noémie3,4, Kéver Loïc3, Gresham Sophie1,5, Derycke Sofie1,6, Snoeks Jos2,7, Parmentier Eric3, Poncin Pascal4, Verheyen Erik1,5 1 Operational Directorate Taxonomy and Phylogeny, Royal Belgian Institute of Natural Sciences – 1000 Brussels, Belgium 2 Laboratory of Biodiversity and Evolutionary Genomics, KU Leuven, Charles Deberiotstraat 32 – 3000 Leuven, Belgium 3 Laboratory of Functional and Evolutionary Biology, FOCUS, Liège University – 4000 Liège, Belgium 4 Laboratory of Fish and Amphibian Ethology, Behavioural Biology Unit, FOCUS, Liège University – 4000 Liège, Belgium 5 Department Biology, Evolutionary Ecology, University of Antwerp – 2610 Antwerp, Belgium 6 Flanders research institute for agriculture, fisheries and food, Animal Sciences Unit, Aquatic Environment and Quality, Ankerstraat 1 – 8400 Oostende, Belgium. 7 Section Vertebrates, Ichthyology, Royal Museum for Central Africa – 3080 Tervuren, Belgium Corresponding author Cite as: Van Steenberge M, Jublier N, Kéver L, Gresham S, Derycke S, Snoeks J, Parmentier E, Poncin P, Verheyen E (2022) The initial response of females towards congeneric males matches the propensity to hybridise in Ophthalmotilapia. bioRxiv, 2021.08.07.455508, ver. 3 peer- reviewed and recommended by Peer Community in Zoology. https://doi.org/10.1101/2021.08.07. 455508 5 Department Biology, Evolutionary Ecology, University of Antwerp – 2610 Antwerp, Belgium 6 Flanders research institute for agriculture, fisheries and food, Animal Sciences Unit, Aquatic Environment and Quality, Ankerstraat 1 – 8400 Oostende, Belgium. g Aquatic Environment and Quality, Ankerstraat 1 – 8400 Oostende, Belgium. 7 Section Vertebrates, Ichthyology, Royal Museum for Central Africa – 3080 Tervuren, Belgium 7 Section Vertebrates, Ichthyology, Royal Museum for Central Africa – 3080 Tervuren, Belgium This article has been peer-reviewed and recommended by Peer Community in Zoology https://doi.org/10.24072/pci.zool.100010 Posted: 07/02/2022 Introduction Speciation is traditionally seen as a gradual build-up of reproductive isolation between diverging populations (Mayr 1982; Coyne and Orr 2004). The classical view that this is a slow process that occurs between allopatric populations has recently been challenged by genomic findings (Marques et al. 2019) that showed how hybridisation can drive rapid speciation (Seehausen 2004; Jiggins et al. 2008). However, as unrestrained gene flow inevitably homogenizes the genomes of diverging lineages (Roux et al. 2016), the question remains what mechanisms keep incipient species separated. In scenarios of sympatric, closely related species, the ability to correctly distinguish between conspecific and heterospecific mates is probably crucial to maintain the integrity of the species (Sullivan 2009). Mating is the end point of a complex decision-making process in which several potential mates are evaluated (Luttbeg et al. 2001). An encounter with a potential mate can be seen as the first step in this process. The outcome of this initial contact can have profound implications on fitness, either through the refusal of suitable mates, or through the acceptance of suboptimal partners. Therefore, we expect intra- and interspecific differences in individual responses when confronted with a choice of partners. Additionally, the preference for a given mate can depend on environmental, social and intrinsic parameters, explaining variation in preference both between and within species, and between and within individuals (Pfennig 2008; Sommer-Trembo et al. 2017). Assortative mate selection was traditionally seen as a sequential process in which individuals first assess whether ‘the other’ is a conspecific and then assess its quality as a mate (Mayr 1982). However, empirical data and theory suggest that assessment of species and quality are not independent steps (Sullivan 2009; Mendelson and Shaw 2012) as the specific status of an individual could be judged using the same cues as its quality. Additionally, adaptive hybridisation has been observed in several taxa (overview in Mendelson and Shaw 2012), indicating that preferred mates are not necessarily always conspecific. Cichlids of the large East African Lakes form endemic, species-rich radiations (Salzburger 2018). Several suggested “key adaptations” of cichlids, such as their pharyngeal jaws, help to explain their evolution into numerous trophic niches (Kocher 2004). However, a large proportion of these closely related species coexist without apparent eco-morphological differences (Van Oppen et al. 1998). Since most cichlid assemblages are relatively young, several taxa may be classified as incipient species and they retain the potential to hybridise. ABSTRACT It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint PEER COMMUNITY IN ZOOLOGY ABSTRACT Recommender: Decaestecker Ellen Cichlid radiations often harbour closely related species with overlapping niches and distribution ranges. Such species sometimes hybridise in nature, which raises the question how can they coexist. This also holds for the Tanganyika mouthbrooders Ophthalmotilapia ventralis and O. nasuta. Earlier studies found indications of asymmetrical hybridisation with females of O. ventralis accepting males of O. nasuta, but not the other way around. We hypothesised that this was due to differences in the capacity for species recognition. Given the higher propensity of O. ventralis females towards hybridisation, we expect a reduced ability for species recognition in O. ventralis females, compared to O. nasuta females. We staged two experiments, one focusing on 22 female O. nasuta and one on 21 female O. ventralis. These fish were placed in one half of a tank and briefly exposed to a conspecific or a heterospecific male, a conspecific female, or nothing (control). Female response was evaluated by scoring six tracking parameters and by noting the occurrence of ten discrete behaviours before and during the encounter. Females always responded to the presence of another fish by approaching it. Remarkably, for both O. nasuta and O. ventralis, we did not find a different response between encounters with conspecific males and females. However, in agreement with our hypothesis, O. nasuta females behaved differently towards conspecific or heterospecific males, whereas O. ventralis females did not. When presented with a heterospecific male, O. nasuta females performed a lower number of ‘ram’ behaviours. Additionally, they never displayed the ‘flee’ behaviour, a component of the species’ mating repertoire that was seen in all but one of the presentations with a conspecific male. Our findings show that differences in species recognition at first encounter predict to a large degree the outcome of the mating process, even in the absence of mating behaviour. Reviewers: George Turner and two anonymous reviewers Correspondence: m.vansteenberge@naturalsciences.be Keywords: Species recognition, mate choice, behaviour, Lake Tanganyika, Africa . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. Introduction Hence, models that describe speciation in cichlids emphasize the importance of female mate choice in the initial stages of the speciation process (Danley and Kocher 2001). The Lake Tanganyika cichlids assemblage contains species with profoundly different mating strategies (Ronco et al. 2020). Ophthalmotilapia stands out by its extreme sexual dimorphism and female-biased reproductive investment. Although the correlation between reproductive investment, sexual selection and choosiness is well-established, it still remains debated whether choosiness is an evolutionary outcome (sensu Trivers 1972), or rather a determinant of differences between the sexes in parental investment (Thomas and Szekely 2005). Using Lake Tanganyika cichlids, Gonzales-Voyer et al. (2008) showed support for the latter hypothesis. Regardless of the evolutionary mechanism, females should be considered the choosy sex in Ophthalmotilapia (sensu Wirtz, 1999). Therefore, if a prezygotic mechanism explains the asymmetric pattern observed in nature (Nevado et al. 2011), it would be caused by differences between females of the different species in accepting matings with heterospecific males. As increased capacity for species recognition leads to increased preference in the choosier sex (Kozak & Boughman, 2009), we predict to see an interspecific difference in female response to conspecific and heterospecific males. As males of the different Ophthalmotilapia species have very similar courtship behaviours, in which the few species-specific elements are insufficient to prevent hybridisation (Kéver et al. 2018), females would mainly rely on other cues like colour patterns, body size and pheromones. Although the reproductive behaviour of Ophthalmotilapia species is well documented (Haesler et al. 2011; Immler and Taborsky 2009; Kéver et al. 2018), little is known on how Ophthalmotilapia species recognize conspecifics and select potential mates. In this study, we will investigate species recognition, which was, in spite of its shortcomings (Mendelson and Shaw 2012), defined as ‘‘a measurable difference in behavioural response toward conspecifics as compared to heterospecifics’’. We will study this by comparing the ability of female O. nasuta and O. ventralis to distinguish con- from heterospecific males. Often, dichotomous mate choice trails, in which a female can choose between two males, are used for such studies. Such a setup includes an aspect of male competition, whereas we are interested in the response of females in the absence of such behaviour. Additionally, with dichotomous trails one cannot discriminate mate preference from species recognition. Although we expect species recognition to be a good predictor of female mate preference, we did not aim to quantify preference directly. Introduction It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint large stones are available (Konings 2019). This is the only part of the lake where up to three species of Ophthalmotilapia occur in sympatry (Hanssens et al. 1999). large stones are available (Konings 2019). This is the only part of the lake where up to three species of Ophthalmotilapia occur in sympatry (Hanssens et al. 1999). Although specimens of Ophthalmotilapia can be easily assigned to one of the valid species (with the possible exception of O. heterodonta and O. ventralis, see Hanssens et al. 1999), a phylogeographic study discovered gene flow among these species. Nevado et al. (2011) observed that specimens of O. nasuta often carried mitochondrial DNA of the other species, whereas the opposite was much less often the case. They suggested that this pattern either has a postzygotic, (e.g. by cyto-nuclear incompatibilities that affects mutual crossbreedings differently) or a prezygotic (e.g. by an asymmetry in reproductive behaviour that results in a different resistance towards hybridisation) cause. The latter scenario implies that females of all species would occasionally mate with O. nasuta males, while O. nasuta females would be much less inclined to mate with heterospecific males. It also implies that the female hybrid offspring would backcross into O. nasuta. This scenario agrees with the recent description of a successful mating between a female O. ventralis and a male O. nasuta (Kéver et al. 2018). Reproductive isolation in closely related species of East African cichlids is mostly maintained through prezygotic isolation (Turner et al. 2001). Introduction The oldest East African lake, Lake Tanganyika, however, contains a mature cichlid radiation (Salzburger 2018) in which most species are well-delineated (Ronco et al. 2020). However, even between well-delineated biological species, boundaries can be permeable and molecular studies identified several instances of inter-specific hybridisation (Rüber et al. 2001; Koblmüller et al. 2007; Nevado et al. 2011). Such examples allowed us to select a case to study the importance of prezygotic, behavioural isolation after, or at the last stages, of the speciation process. Ophthalmotilapia Pellegrin, 1904 species are maternal mouth brooders that occur on the rocky and intermediate (rocky patches separated by sand) shores of Lake Tanganyika. The genus contains four currently accepted valid species: O. ventralis (Boulenger 1898), O. boops (Boulenger 1901), O. heterodonta (Poll and Matthes 1962) and O. nasuta (Poll and Matthes 1962) (Hanssens et al. 1999). They are sexually dimorphic maternal mouthbrooders with territorial males that protect a spawning site, and females that aggregate in feeding schools when they are not breeding. Ophthalmotilapia males possess egg-shaped lappets at the distal ends of their elongated pelvic fins that are unique among Great Lake cichlids (Poll 1986). These lappets function as egg dummies during the species’ mating behaviour in a similar way as the egg spots on the anal fins of the so-called ‘modern’ haplochromines (sensu Salzburger et al. 2007; Theis et al. 2012). During the mating process, the female deposits the eggs and almost immediately takes them into her mouth. By snapping at the egg dummies, which are situated close to the genital opening of the male, the intake of sperm is thought to be facilitated, increasing the fertilisation rate of the eggs within the female’s mouth (Salzburger et al. 2007). The four species of Ophthalmotilapia have different but partially overlapping, distribution ranges. Ophthalmotilapia nasuta is the sole species in the genus with a patchy but lake-wide distribution. The sister species O. heterodonta and O. ventralis have non-overlapping ranges with the former occurring in the northern half and the latter in the southern third of the Lake. The fourth species, O. boops only occurs along a rather limited stretch of Lake Tanganyika’s south-eastern shoreline. There, it prefers sites where 2 . CC-BY 4.0 International license perpetuity. Experimental setting We performed two independent experiments using females and males of O. ventralis and O. nasuta. The first focused on the behaviour of focal O. nasuta females (ON experiment), the second on that of focal O. ventralis females (OV experiment). All individuals were wild caught off the coast of Ulwile Island or nearby Kala, on the mainland (Tanzania). Females were acquired as juveniles and hence had no prior mating experience. They were kept in monospecific aquaria per species until they reached maturity. Hence, all females used had similar ages. Males were older and most had prior mating experience (Kéver et al. 2018). We verified the origin of a random selection of these fishes by sequencing the mitochondrial control region, and by comparing these sequences with the data collected by Nevado et al. (2011). The experiments were performed at the aquarium facilities of the University of Liège. Prior to the onset of the experimental trials, the sex of the specimens was checked by visually inspecting their genital papillae. Female specimens were kept jointly but isolated from males and heterospecific specimens in a separate tank for at least two weeks. This tank was devoid of hiding places, in order to prevent the development of territoriality. During that period, males were kept in monospecific tanks in which they were visually isolated from each other using opaque partitions. We kept all specimens in the same condition for at least two weeks with photoperiod: 12:12 h L:D, water temperature: 26±1°C, carbonate hardness: >8 dKH. Fishes were fed daily ad libitum with ‘Tropical Spirulina forte’ mini-granules. We used three identical experimental aquaria (88cm50cm*40cm with water level ca. 40cm), which we divided into two equal parts by a perforated transparent partition (separation wall), through which fishes could not pass, and by an opaque wall (visual barrier) that could be removed (Fig. 1A). A flower pot was placed on each side of the separation to allow the fish to take refuge. We kept the fishes in these aquaria for at least twelve hours before they were used in the experimental trials. During the trial, the visual barrier was removed. was removed. Figure 1. Experimental setup. A: aquarium setup: A focal female of either O. nasuta or O. ventralis (here O. Experimental setting nasuta) (A) was placed in one half of the experimental tank whereas no fish, a conspecific female or a hetero- or conspecific male was placed in the other half (here an O. ventralis male) (B). The tank was divided in two by a transparent wall (D) and a visual barrier (C), a hydrophone (E) was placed on the side of the non-focal specimen and an empty flowering pot was placed in both halves of the tank, allowing fishes to take refuge. Video and audio recordings were made 15 minutes prior and 45 minutes after the visual barriers were removed. B: Contrasts tested using permanova: (A): focal females presented with another fish vs. with no fish, (B): focal females presented with a conspecific female vs. a conspecific male, (C): focal females presented with a conspecific vs. a heterospecific male, (D): conspecific females and males presented to a focal female and (E): conspecific and heterospecific males presented to a focal female. Black and grey fishes represent different species, the white fish represents all possible non-focal specimens used. We recorded the behaviour of focal specimens (O. nasuta females or O. ventralis females) in four different experimental conditions. They were either exposed to (i) no other specimen (Co), (ii) a conspecific female (CF), (iii) a conspecific male (CM) or (iv) a heterospecific male (HM) (Supplement 1). For each Figure 1. Experimental setup. A: aquarium setup: A focal female of either O. nasuta or O. ventralis (here O. nasuta) (A) was placed in one half of the experimental tank whereas no fish, a conspecific female or a hetero- or conspecific male was placed in the other half (here an O. ventralis male) (B). The tank was divided in two by a transparent wall (D) and a visual barrier (C), a hydrophone (E) was placed on the side of the non-focal specimen and an empty flowering pot was placed in both halves of the tank, allowing fishes to take refuge. Video and audio recordings were made 15 minutes prior and 45 minutes after the visual barriers were removed. B: Contrasts tested using permanova: (A): focal females presented with another fish vs. with no fish, (B): focal females presented with a conspecific female vs. a conspecific male, (C): focal females presented with a conspecific vs. Introduction Instead, we choose to investigate differences in species recognition by presenting a set of O. ventralis and O. nasuta females to males of both species separately and by measuring a range of behavioural variables. This study described and compared the initial behavioural response of O. ventralis and O. nasuta females towards males of both species in an aquarium setting. In view of the pattern of asymmetric hybridisation discovered among these species (Nevado et al. 2011), we expected a difference in the capacity for species recognition between females of O. ventralis and O. nasuta. Specifically, we hypothesized that O. nasuta females would be able to differentiate between conspecific and heterospecific males at the initial stages of an encounter. For O. ventralis females, however, we expected that this capacity would be less pronounced or absent. PEER COMMUNITY IN ZOOLOGY 3 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Collection of tracking and qualitative behavioural data Video files were converted into JPG images using Adapter v2.1.6 (available at https://www.macroplant.com), capturing one frame per second and saving it as an 8-bit, gray-scale JPG file. Images taken within ten seconds before or after an experimenter was performing an action (i.e. removing the wall) were discarded from analyses. We chose to analyse the same number of frames for all trials within the ON and the OV experiment, respectively. For the ON experiment, this resulted in a minimum of 721 and 2186 frames collected before and after the removal of the separation wall, respectively. For the OV experiment, 871 and 2685 frames were available for analyses. Both focal and non- focal specimens were tracked using the ImageJ v1.49 (Schneider et al. 2012) software package. Given the presence of both light- and dark- coloured backgrounds in the aquarium setting, the set of frames was studied twice. Specimens that were present before a light coloured background were tracked by inverting black and white values whereas specimens present before a dark-coloured background were tracked using non-inverted images. For computational reasons, analyses were performed on subsets of the data containing a maximum of 1,000 frames. For each set of images, a subset of 30 frames was used to create a background using the plug-in ‘Image stack merger plus’. Backgrounds were removed using the image calculator and the resulting frames were transformed into black-and-white images using the threshold function with ‘MaxEntropy’ as the methodology. Images were adjusted using the ‘erode’ and ‘dilate’ functions to remove noise and to obtain a better representation of the fishes. The resulting image series was then used for tracking using the plug-in ‘MTrack2’, in which tracks were summarized as the x- and y- coordinates of the centroids of the tracked object. The quality of the automated tracking was checked by visually inspecting each of the frames. When the software failed to track a specimen that was clearly present in the final images, coordinates were added manually. Finally, tracks obtained from both datasets, inverted and non-inverted, were combined. When a specimen was recognized by both methods, e.g. when the fish was partially before a light- and partially before a dark-coloured background, the average of the coordinates was used. When tracking data was missing, the average value of the coordinates of the previous and the next positions were used. Experimental setting a heterospecific male, (D): conspecific females and males presented to a focal female and (E): conspecific and heterospecific males presented to a focal female. Black and grey fishes represent different species, the white fish represents all possible non-focal specimens used. We recorded the behaviour of focal specimens (O. nasuta females or O. ventralis females) in four different experimental conditions. They were either exposed to (i) no other specimen (Co), (ii) a conspecific female (CF), (iii) a conspecific male (CM) or (iv) a heterospecific male (HM) (Supplement 1). For each We recorded the behaviour of focal specimens (O. nasuta females or O. ventralis females) in four different experimental conditions. They were either exposed to (i) no other specimen (Co), (ii) a conspecific female (CF), (iii) a conspecific male (CM) or (iv) a heterospecific male (HM) (Supplement 1). For each PEER COMMUNITY IN ZOOLOGY 4 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint experiment (ON and OV) and for each experimental condition (i to iv), we conducted a minimum of five replicates. We filmed (using a CANON Legria HF R606) the entire aquarium (i.e. focal and non-focal fishes) during one hour: from 15 minutes before to 45 minutes after the visual barrier was removed. Experimenters were only briefly present in the room to remove the visual barrier. As O. ventralis males are known to produce weak-pulsed sounds during the inviting behaviour (Kéver et al. 2018), we used a HTI Min-96 hydrophone (–164.4 dB re. Experimental setting 1 V μPa−1; bandwidth 2 Hz and 30 kHz, MS, USA), connected to a Tascam DR-05 recording (TEAC, Wiesbaden, Germany) at a 44.1 kHz sampling rate to record sounds during the whole experiment. The hydrophone was positioned near the separation wall, at half the height of the water column, on the side of the non-focal specimen. At the start of each trial, we switched off the aeration of the tank so that sounds could be recorded. However, these recordings were not analysed, as we detected no communication sounds. After each trial, the focal female was euthanized. Both the focal and the non- focal specimens were weighed. Focal specimens were measured, dissected and the stage of gonad development was scored following Panfili et al. (2006). We performed a total of 28 ON and 21 OV experimental trials, with a maximum of three trials per day (Supplement 1). However, after the dissections (see below), we observed that six focal O. nasuta females from the first set of trials possessed male or ambiguous gonads (Supplement 2). These specimens were referred to as floater males and the recordings for these trials were not analysed. As we suspected that these specimens had changed sex, we photographed the genital papillae of the focal females that were to be used subsequently, two weeks before the onset of the experimental trials. A comparison between papillae of the same individuals after two weeks revealed that some could be identified as males. This either confirmed that a sex change did indeed take place in several specimens, or showed that the examination of genital papillae was insufficient for sex determination in subadult Ophthalmotilapia. These specimens were not studied. After each trial, the aquarium was cleaned and the water fully renewed. Collection of tracking and qualitative behavioural data This is justified, as missing data either corresponded to fish that remained stationary for many frames, and could hence not be distinguished from the background, or to fish that hid behind the terracotta flowerpots (Fig. 1). Frames collected before and after the removal of the 5 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint visual barrier were analysed separately. Coordinates were shifted using the lower- and anterior-most point of the separation wall as the origin, and rotated by setting the anterior water level as a reference for the horizontal plane. Finally, all coordinates were transformed from pixels to centimetres using the dimensions of the aquaria. Tracks were visualized by plotting all individual positions as well as the shift in average position of a specimen before and after the removal of the barrier (Supplement 3). visual barrier were analysed separately. Coordinates were shifted using the lower- and anterior-most point of the separation wall as the origin, and rotated by setting the anterior water level as a reference for the horizontal plane. Finally, all coordinates were transformed from pixels to centimetres using the dimensions of the aquaria. Tracks were visualized by plotting all individual positions as well as the shift in average position of a specimen before and after the removal of the barrier (Supplement 3). For each specimen, six tracking parameters were calculated from the coordinates (Table 1). Each parameter was calculated three times: once using coordinates obtained for 721/871 (OV/ON experiment respectively) seconds before (before) the removal of the visual barrier, once using coordinates obtained during 721/871 seconds (after1) after the visual barrier was removed and finally using coordinates obtained during 2186/2685 seconds (after2) after the visual barrier was removed, hence including the after1. Additionally, ten specific behaviours were defined based on Baerends and Baerends-Van Roon (1950). Collection of tracking and qualitative behavioural data These were encoded and recorded as point events in Boris v. 2.72 open source software (Friard and Gamba 2016) (Table 1). This data was collected during the same three periods: before, after1 and after2. We choose these time periods, as we aimed to study the immediate response following presentation to another fish. Teleosts respond to changing social situations at two scales, by an immediate change in behaviour, and by a modification of neural circuits (Oliveira 2012). The latter is driven by the regulation of immediate early genes (IEG). Following Oliveira (2012), we assumed that the transcription of IEG occurs roughly 15 min (after1), and their translation 45 min (after2) after the stimulus. Behaviours displayed within ten seconds before or after an experimenter was performing an action (removing the wall) were discarded from the analyses. PEER COMMUNITY IN ZOOLOGY Table 1. Tracking parameters and point events recorded for the focal and non-focal individuals du the experimental trials. For point events, interpretation of the behaviour was added. Tracking parameters Description Dist.wall Distance to the visual barrier (% of length of compartment). Dist.fish Distance to the fish on the other side of the transparent wall (cm). Sp Average speed (cm/s). SpX Horizontal speed (cm/s). SpY Vertical speed (cm/s). Height Mean height (% of height of water column). Point events Description Interpretation Chase The fish suddenly swims very fast towards the other fish and rams (or almost rams) the separation wall. Contextual: agonistic and male courtship be (Kéver et al. 2018) Flee The fish suddenly swims away from the other fish. Contextual: agonistic and female co behaviour (Kéver et al. 2018) Lateral Lateral display: The fish positions itself perpendicular to the other fish, keeping its head slightly downwards, erects its fins and bends its body. Signal movement (Baerends and Baerends-Va 1950), agonistic behaviour (Kéver et al. 2017) Frontal Frontal display: The fish faces the other fish head up and erects its fins. Signal movement (Baerends and Baerends-Va 1950), agonistic behaviour (Kéver et al. 2018) Bite Biting the wall: The fish bites the separation wall (possibly trying to bite the other fish). Signal movement (Baerends and Baerends-Va 1950), agonistic behaviour (Kéver et al. 2018) Ram Ram into the wall: The fish tries to enter the other part of the aquarium and rams (not very fast) the separation wall. Sand Sand picking: The fish takes sand in its mouth. PEER COMMUNITY IN ZOOLOGY Statistical analyses Prior to testing differences in species recognition, we visually explored the combined datasets of tracking parameters and point events. We did this by performing principal component analysis (PCA) and canonical variate analyses (CVA) in Past 3.14 (Hammer et al. 2001). The former allowed for an unbiased visualization of the variation in the data and was performed on the correlation matrices. The latter was conducted to maximize the differentiation between the different groups. Separate analyses were performed for each period of time recorded (before, after1, after2) and for each experiment (ON or OV). Point events that were not recorded during one of these periods were disregarded and missing values (i.e. for Dist.fish in the control condition Co) were treated using mean value imputation. Prior to the analyses, each of the tracking parameters and point events was normalised. This was done for each experiment (OV and ON), and for each of the time periods (‘before’, ‘after1’, ‘after2’ and ‘shift’) separately. p p y We used permanova to compare the behaviour of both focal and non-focal specimens in five different comparisons (Fig. 1B). With this non-parametric test, we could include all recorded behaviours. This allowed us to measure species recognition via differences in behaviour (sensu Mendelson and Shaw 2012) without defining a priori in what variable specimens would differ. In order to reduce the number of comparisons, we restricted ourselves to only biologically relevant contrasts. For focal specimens, we compared the behaviour between (A) females that were presented with another fish vs. with no fish, (B) focal females that were presented with a conspecific female vs. a conspecific male, and (C) females that were presented with a conspecific vs. a heterospecific male. Two additional comparisons were tested for the non-focal individuals. We tested (D) whether conspecific females and males respond differently to a focal female and (E) whether conspecific and heterospecific males respond differently to a focal female. Even though our main goal was to test whether females of O. nasuta and O. ventralis differed in behavioural response towards conspecific and heterospecific males (C), we tested the four other comparisons as well, following the recommendations of Moran (2003) and Nakagawa (2004). Tests were performed using non-parametric permanova, using the pairwise.adonis function, of the R package vegan on the combined data of tracking parameters and point events. This approach was chosen as the conditions for multivariate normality were violated. Collection of tracking and qualitative behavioural data Courtship behaviour when linked to construc bower (Kéver et al. 2018), signal movemen nipping off a substrate (Baerends and Baeren Roon 1950). Spasm A quick, strong, and unilateral contraction of the trunk musculature that results in a displacement of the head and the caudal fin in the same direction. Contextual: comfort behaviour or signal mov Observed in courtship behaviour or inter-te fights depending on the genus (Baerend Baerends-Van Roon 1950). Tail Tail-wagging: Exaggerated movements of the caudal fin (+ caudal part of the dorsal fin). At its zenith, the movement of the caudal fin is completely counterbalanced by backpedalling. Signal movement (Baerends and Baerends-Va 1950), courtship behaviour (Kéver et al. 2018 Flicker Pelvics flickering: The fish quickly and alternatively moves its right and left pelvic fins. Comfort behaviour (Baerends and Baeren Roon 1950) Contextual: agonistic and female courtship behaviour (Kéver et al. 2018) Signal movement (Baerends and Baerends-Van Roon 1950), agonistic behaviour (Kéver et al. 2017) Signal movement (Baerends and Baerends-Van Roon 1950), agonistic behaviour (Kéver et al. 2018) Signal movement (Baerends and Baerends-Van Roon 1950), agonistic behaviour (Kéver et al. 2018) Signal movement (Baerends and Baerends-Van Roon 1950), agonistic behaviour (Kéver et al. 2018) Signal movement (Baerends and Baerends-Van Roon 1950), agonistic behaviour (Kéver et al. 2018) Courtship behaviour when linked to construction of bower (Kéver et al. 2018), signal movement when nipping off a substrate (Baerends and Baerends-Van Roon 1950). Contextual: comfort behaviour or signal movement. Observed in courtship behaviour or inter-territorial fights depending on the genus (Baerends and Baerends-Van Roon 1950). Tail Tail-wagging: Exaggerated movements of the caudal fin (+ caudal part of the dorsal fin). At its zenith, the movement of the caudal fin is completely counterbalanced by backpedalling. Flicker Pelvics flickering: The fish quickly and alternatively moves its right and left pelvic fins. Signal movement (Baerends and Baerends-Van Roon 1950), courtship behaviour (Kéver et al. 2018). Comfort behaviour (Baerends and Baerends-Van Roon 1950) PEER COMMUNITY IN ZOOLOGY 6 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Visualization of the behavioural data We visually explored the data using PCA and CVA to compare the behaviour of all specimens used in each of the two experiments. In the PCAs conducted on the behavioural data collected before the removal of the barrier, values of all females as well as of conspecific males overlapped (Fig. 2A,B), suggesting a highly similar behaviour. However, heterospecific males were (somewhat) separated from all other specimens by their higher values for PC1 (ON experiment) or PC2 (OV experiment). The loadings of the main PCs (Supplement 5.1) suggest that this difference was due to a more active swimming behaviour (Sp, SpX, SpY) higher up in the water column (height) for O. ventralis males (ON experiment) and a higher number of point events (ram, sand, bite) performed at the floor of the aquarium (height) for the O. nasuta males (OV experiment), prior to their presentation to a heterospecific female. The fact that O. nasuta males build true bowers could explain the higher number of point events, especially ‘sand’. Additionally, O. nasuta might be less at ease close to the water surface as it generally occurs in deeper waters than O. ventralis (Konings 2019). The CVAs also reflected the behavioural differences of heterospecific males (Fig. 3A,B), as they had higher values for the first CVs. The behaviours that contributed strongly to the separating PCs, also contributed to the main CVs (Supplement 5.2). The PCAs performed on the data collected 15 minutes after the removal of the barrier foremost showed how the focal females that were used as controls behaved differently than the specimens that were presented with another specimen (Fig. 2C,D). For both experiments, this can be explained by control females spending less time closer to the wall (Dist.wall), and performing less agonistic behaviour (ram, lateral, flee). We did not observe any additional separation in the PCA of the OV experiment (Fig. 2D). In the ON experiment, values for (heterospecific) O. ventralis males stood out by their high values for PC1, whereas those of (conspecific) O. nasuta males had mostly low values for this axis (Fig. 2C). The tracking parameters Sp, SpX, SpY all had high, positive contributions to PC1 (Supplement 5), reflecting that O. ventralis males kept swimming actively when the barrier was removed in the ON experiment. p g y p We carried out CVAs on the same datasets (Fig. 3C,D). Statistical analyses It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Results We separately analysed two experiments, one focusing on O. nasuta and one on O. ventralis females (ON and OV experiment). In the ON experiment, two males of O. ventralis performed advanced courtship behaviours. After the encounter, these males started to swim in circles, in fast and erratic movements. This was accompanied by tail wagging, generally displayed close to the partition wall. These males often bit the hydrophone and picked up and moved around sand (49 and 45 times within 45 min vs. 0 for the other males). One of these two males (O. ventralis male presented to ON38) also tried to chase the female (79 times) and presented the egg dummies of its pelvic fins (5 times). This behaviour stopped immediately when the experimenter removed the female fish. During the encounter, these males turned dark grey, to almost black, which was swiftly reversed after the experimental trial. As we designed our experiment to study behavioural response in the absence of courtship behaviour, we removed these outliers from all analyses. All ten point events were observed in at least one of the fishes in the ON experiment, whereas ‘tail’ (i.e. tail wagging) was never observed in the OV experiment (Supplement 4). Statistical analyses When permanova revealed significant differences, we verified whether this could be due to dispersion effects (Anderson 2006). For this, in view of the size of the dataset, non-parametric Mann-Whitney U tests were performed on the within group dispersions from the mean, calculated using the betadisper function implemented in the R package vegan (Oksanen et al. 2017). A difference in dispersion across groups does, however, not contradict a difference in their average values. Hence, the data was visually explored, using PCA, to assess whether dispersion could be present together with a difference in average values. For each comparison revealed significant by permanova, Mann Whitney U non-parametrical tests were performed on each of the variables separately in order to detect which of these caused the difference between the treatments. We choose this non- parametric approach as the assumptions of normality were often not met. When significant, the effect sizes of these variables were estimated using Hedge’s g, which was calculated using the estimationstats.com web application (Ho et al. 2019). In order to test whether the observed differences in behaviour depended on the visual presence of another specimen, and whether these differences were already visible at the first stages of the encounter, permanova tests were performed on data collected before the removal of the visual separation (before) as well as on data collected over a short (after1) and a long (after2) period of time after this separation was removed. Finally, an additional test was conducted which removed individual variation between the different treatments. For this, behavioural shifts were calculated for each tracking parameter and point event by subtracting the values of the ‘before’ period before from those of the ‘after1’ period (shift). All tests were performed separately for the ON and the OV experiment. As behaviour can be influenced by gonad development and weight of the focal and non-focal specimens, Mann Whitney U tests were performed to check whether these differed between the treatments. Such tests were also performed on the amount of frames in which fishes could not be tracked. All statistical analyses were performed using Past 3.14 (Hammer et al. 2001) and R (R core team 2017). PEER COMMUNITY IN ZOOLOGY 7 7 . CC-BY 4.0 International license perpetuity. Visualization of the behavioural data For both experiments, the control females stood out by their high values for CV1. This could again be explained by their higher values for Dist.Wall. In the ON experiment, (heterospecific) O. ventralis males stood out by their low values for CV1, which would be attributed to their more active swimming behaviour (Sp, SpX, SpY). Conspecific O. nasuta males stood out by their low values for CV2, which could be due to the higher occurrence of ‘sand’ and ‘bite’ behaviour. Values for O. nasuta females that were presented with another fish had more intermediate values for CV1 and CV2. However, values of O. nasuta females that were presented to a conspecific male clustered between values of those males and of those of the other females. Similarly, females that were presented to a heterospecific male had values that were intermediate between those of these males and those of the other females (Fig. 3C). This suggests that the behaviour of focal females shares characteristics with the behaviour of the non-focal fishes presented to them. In the CVA of the OV experiment, (heterospecific) O. nasuta males stood out by their low values for CV1 and high values for CV2. This was most influenced by the higher occurrence of point behaviours (ram, spasm, sand, flicker). Values of O. ventralis males overlapped with those of female specimens that were presented with another fish (Fig. 3D). These patterns 8 PEER COMMUNITY IN ZOOLOGY 8 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint remained present when performing similar analyses on the data collected 45 minutes after the removal of the visual barrier (Supplement 6). Plotting the shift in average position before and after the removal of the barrier revealed how almost all specimens moved towards the wall when presented with another specimen. Additionally, this showed that O. Visualization of the behavioural data nasuta specimens, on average, spent more time closer to the bottom whereas O. ventralis specimens were more often found higher up in the water column (Supplement 7). Figure 2. Principal component analyses performed on the behavioural data collected 15 min before (A, B) and 15 min after (C, D) the visual barrier was removed in the ON (O. nasuta, left) and the OV (O. ventralis, right) experiments. Symbols on the scatter plots for the ON and OV experiment as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. Figure 2. Principal component analyses performed on the behavioural data collected 15 min before (A, B) and 15 min after (C, D) the visual barrier was removed in the ON (O. nasuta, left) and the OV (O. ventralis, right) experiments. Symbols on the scatter plots for the ON and OV experiment as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. 9 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Figure 3. Visualization of the behavioural data Canonical variate analyses on the behavioural data collected 15 min before and 15 min after the visual barrier was removed in the ON (O. nasuta, left) and the OV (O. ventralis, right) experiment. Symbols on the scatter plots for the ON and OV experiments as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. Figure 3. Canonical variate analyses on the behavioural data collected 15 min before and 15 min after the visual barrier was removed in the ON (O. nasuta, left) and the OV (O. ventralis, right) experiment. Symbols on the scatter plots for the ON and OV experiments as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. Figure 3. Canonical variate analyses on the behavioural data collected 15 min before and 15 min after the visual barrier was removed in the ON (O. nasuta, left) and the OV (O. ventralis, right) experiment. Symbols on the scatter plots for the ON and OV experiments as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. Behaviour of focal females However, these dispersion effects can be explained by the (much) lower variation in values for control females compared to those of other focal females. As visualisation of the data showed a clear separation of both groups, we can assume that a difference in dispersion is present together with a difference in average values (Fig. 2C). Mann-Whitney U tests revealed that controls differed from other focal females by their higher values for the variable Dist.wall (ON: 46.3 +-18.2 vs. 20.1 +-8.8, p=0.005, g=1.32; OV: 55.2 +-13.3 vs. 20.8 +-9.7, p=0.002, g=1.13) and their lower values for ‘ram’ (ON: 8.6+-12.1 vs. 74.5+-54.0, p=0.005, g=-2.14; OV: 8.0 +-11.9 vs. 48.8 +-38.4, p=0.014, g=-3.12). Table 2. PERMANOVA performed on the behavioural parameters of the ON (O. nasuta) and OV (O. ventralis) experiment. Tests were performed on the data collected during 15 minutes before (B), 15 minutes after (A1) and 45 minutes after (A2) the removal of the opaque wall as well as on behavioural shifts (S) calculated by subtracting data recorded during 15 minutes before from that of 15 min after the removal of the wall (A1-B). We tested five comparisons (Fig. 1B): by comparing the behaviour of focal females that were presented with (A) another fish vs. with nothing, (B) with a conspecific female vs. a conspecific male, and (C) with a conspecific vs. a heterospecific male. We further compared (D) the behaviour of non-focal conspecific males vs. females, and (E) conspecific vs. heterospecific males. Behaviours that were neither recorded before or after the removal of the wall, ‘tail’ and ‘sand’ for the ON, and ‘flicker’ and ‘sand’ for the OV experiment, were excluded. For the first comparison, Dist.fish was excluded, as it could not be calculated. Values in bold are significant at the 0.05 level. For these comparisons, ° and † denote that the assumption of equal dispersion was violated at the 0.05 and 0.01 le els levels. Behaviour of focal females We tested whether the behavioural responses of females of O. nasuta and O. ventralis differed in three different comparisons (Fig. 1B), i.e. when they were presented to (A) another fish vs. no fish, (B) a conspecific female vs. a conspecific male, and (C) a conspecific vs. a heterospecific male. Within each of the comparisons in the ON experiment, we detected no significant difference in the gonad development of the focal females, in their weights, in the weights of the non-focal fishes, and in the percentage of missing frames. The same applies for the OV experiment, although here, non-focal females had lower body weights than conspecific males. Prior to the removal of the visual barrier, no significant difference in behaviour was PEER COMMUNITY IN ZOOLOGY 10 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint recorded for focal females from the different treatments in both experiments, and for all three comparisons (Table 2). recorded for focal females from the different treatments in both experiments, and for all three comparisons (Table 2). For the three comparisons, and for both experiments (ON and OV), we performed permanova on the behavioural data recorded during the 15 minutes after the removal of the visual barrier (Table 2). For both the OV and the ON experiment, this revealed a significant difference in the behaviour of focal females that were not presented to another fish (controls) and focal females that were presented with another fish (comparison A). For the ON experiment, we could not exclude that these differences stem from dispersion effects. Behaviour of focal females A B C D E ON B F:0.31; p: 0.789 F:0.95; p: 0.423 F:0.91; p: 0.474 F:1.17; p: 0.339 F:1.15; p: 0.24 A1 F:7.28; p: 0.012 † F:0.42; p: 0.768 F:11.33; p: 0.016 ° F:0.45; p: 0.651 F:0.55; p: 0.73 A2 F:4.71; p: 0.027 † F:0.49; p: 0.553 F:8.13; p: 0.040 F:0.66; p: 0.545 F:0.85; p: 0.47 S F:8.40; p: 0.006 † F:1.03; p: 0.416 F:8.75; p: 0.018 † F:1.18; p: 0.332 F:0.25; p: 0.93 OV B F:0.23; p: 0.762 F:1.62; p: 0.203 F:0.85; p: 0.387 F:3.19; p: 0.061 F:5.13; p: 0.013 A1 F:8.00; p: 0.005 F:0.91; p: 0.386 F:0.40; p: 0.546 F:0.28; p: 0.633 F:0.48; p: 0.524 A2 F:4.28; p: 0.026 F:0.78; p: 0.438 F:0.48; p: 0.610 F:0.09; p: 0.837 F:0.49; p: 0.525 S F:7.61; p: 0.004 F:0.89; p: 0.418 F:0.87; p: 0.407 F:0.42; p: 0.679 F:0.73; p: 0.447 Unexpectedly, in both experiments, we did not observe a difference in behaviour between focal females that were presented with a conspecific female or a conspecific male (comparison B). However, when comparing the behaviour of focal females presented with conspecific and heterospecific males (comparison C), a difference became evident between the ON and the OV experiment. In support of our hypothesis, females of O. nasuta responded differently towards conspecific and heterospecific males, whereas females of O. ventralis did not (Table 2). Mann-Whitney U tests revealed that this was due to the lower number of observed ‘ram’ behaviours (43.8+-17.9 vs. 138+-52.9; p=0.04, g=2.25) in O. nasuta females that were presented to O. nasuta males compared to those presented to O. ventralis males. Additionally, O. nasuta females never performed a ‘flee’ behaviour when presented to an O. ventralis male, whereas this was observed in all but one of the O. nasuta females presented to a O. nasuta male (0 vs. 2.6+-2.3; p=0.04, g=-1.33). We obtained highly similar results when analysing the data collected 45 minutes after the visual barrier was removed, or when analysing the shift data (Table 2). Responses of Ophthalmotilapia females Nevado et al. (2011) discovered signatures of unidirectional hybridisation in Ophthalmotilapia, which could either be explained by cyto-nuclear incompatibilities, or by asymmetric mate choice. The latter explanation implies that O. nasuta females would discriminate more strongly against heterospecific males, than females of O. ventralis. This is supported by our experiments. As we did not present focal females with heterospecific females, we cannot say that the observed species recognition in O. nasuta females was due to a different response towards males, or towards all specimens of the other species. However, as females in Ophthalmotilapia are non-territorial and therefore often encounter heterospecific congeners, we expect that the female response is specific to heterospecific males. Unexpectedly, females of both species behaved similarly towards conspecific males and females. Hence, we cannot say whether we observed sexually motivated behaviour, or just the routine behaviour of a (isolated) female that encounters a conspecific individual. In the wild, non-breeding females of both species aggregate in large feeding groups (Konings 2019). Hence, being isolated for 12 hours, as was the case prior to the start of the experiment, represents an unnatural situation for Ophthalmotilapia females. It would, therefore, not be unlikely if Ophthalmotilapia females are behaviourally hardwired to reunite immediately with conspecifics, regardless of whether these are female or male. Females of O. nasuta only performed the ‘flee’ behaviour towards conspecific males and displayed the ‘ram’ behaviour less frequently. Although we found no evidence that this behaviour is sexually motivated, male chasing and female fleeing (i.e. ‘flee’) form the first steps in the mating process of Ophthalmotilapia (Kéver et al. 2018). The ram behaviour, on the other hand, was seen in all experimental trials in which a focal female was presented to another fish. Indeed, the main difference in behaviour between focal females of both species that were, or were not, presented with another fish was the amount of time spent close to the wall, and in the display of the ‘ram’ behaviour. We discovered that several of the presumed females that we planned to use in the experiments had been, or changed into, males. This was also observed for one of the non-focal females in the ON experiment, which was kept isolated from males after the experiment (Supplement 2). Although there are several reports of sex changes occurring in cichlids (Peters 1975; Naish and Ribbink 1990), evidence hereof remained, until now, limited. Behaviour of non-focal specimens We tested whether the behavioural response of conspecific females and males differed when presented to a focal female (D) and whether the behavioural response of conspecific and heterospecific males differed when presented to a focal female (E). Unexpectedly, permanova revealed a difference in the PEER COMMUNITY IN ZOOLOGY 11 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint behaviour of O. nasuta and O. ventralis males in the OV experiment, prior to the removal of the barrier. Mann-Whitney U tests revealed that this was due to the higher average vertical swimming speed of O. ventralis males compared to O. nasuta males (SpY 1.7+-0.5 vs. 0.7+-0.4, p= 0.014, g=2.05). None of the other comparisons was shown to be significantly different, neither before, nor after the removal of the barrier (Table 2). Summary We tested the initial response of O. nasuta and O. ventralis females towards conspecific and heterospecific males. In support of our hypothesis, O. nasuta females differentiated between conspecific and heterospecific males, whereas O. ventralis females did not. Visualisation of the data revealed that O. nasuta females mirrored the behaviour of the males to which they were presented. We also presented females of both species with a conspecific female or with nothing (control). Although females always responded to the presence of another fish, their behaviour did not differ when presented with conspecific males and females. Comparisons of non-focal specimens did not reveal any differences in behaviour after presentation to a focal female. However, before the removal of the wall, males of O. ventralis and O. nasuta behaved differently in the OV experiment. The role of males Asymmetric propensities towards hybridisation are expected in the intermediate stages of reproductive isolation (Arnold et al. 1996). This has been observed in a variety of animal taxa including lungless salamanders (Verrell 1990), spadefoot toads (Pfennig 2007), swordtails (Crapon de Caprona and Ryan 1990), pupfishes (Strecker and Kodric-Brown 1999), and several cichlids (Egger et al. 2008; Nevado et al. 2011). Although most examples are related to female mate choice, these patterns can also be caused by asymmetries in male choosiness (Svensson et al. 2007). Although male mate choice is common in fishes (Wong and Jennions 2003; Werner and Lotem 2003), we choose to focus on the role of females (see introduction, Seehausen et al. 2008; Sefc et al. 2017). However, the mode of fertilization in Ophthalmotilapia could also have an influence on male choosiness. Haesler et al. (2011) studied the reproductive behaviour of O. ventralis, but it can be assumed that the behaviour of its congeners is highly similar. In O. ventralis, a ripe female will visit the territories of several males, either to spawn, or just to collect additional ejaculates. Subsequently, sperm competition will take place within her mouth, resulting in clutches with multiple sires (Haesler et al. 2011). Given that this dilutes the effect of a ‘wrong’ choice, a female can afford to be less choosy. Differences in both male and female courtship effort towards genetically distant or similar mates have been documented in another mouth brooding cichlid: Tropheus Boulenger, 1898 (Zoppoth et al. 2013). However, Tropheus species are sexually monomorphic and both sexes are territorial. Additionally, Tropheus males invest significantly more in raising the clutch, by providing the female access to their feeding territories. As males of Ophthalmotilapia do not share their resources, we can expect these males to be less choosy than those of Tropheus. Additionally, a substantial role of male mate choice is not supported by our data, as we did not observe a difference in behaviour between non-focal males of O. nasuta and O. ventralis when presented with females of the two species. It should be noted, however, that two O. ventralis males that displayed mating behaviour towards O. nasuta females were excluded from the analyses. Whereas our experiments only revealed the capacity for species recognition in females of O. nasuta, we cannot conclude that males cannot distinguish between females of the two species. Whereas the males of O. nasuta and O. Interpretation Although we uncovered a significant difference in the behaviour of O. nasuta females that were presented with a conspecific and a heterospecific male, permanova did not reveal a significant interspecific difference in the behaviour of the males. This could indicate that O. nasuta females interpreted behaviour differently when displayed by O. nasuta or by O. ventralis males. Such species- or sex-dependent interpretation of behaviour is known for several cichlid species, in which territorial males present themselves identically towards both visiting females and intruding males (Baerends and Baerends-Van Roon 1950). Visual exploration of the data, however (Fig. 3C, Supplement 6), revealed a potential difference in male behaviour, which was mirrored by female response. As female O. ventralis did not appear to differentiate between conspecific and heterospecific males, one could ask why hybridisation is not even more prevalent. However, we only examined the very first stage in a potential mating process, so other differences that are present in the mating behaviour between both species could be responsible for this. For example, O. ventralis males display a specific late mating behaviour, called 'invite’, which O. nasuta males never display (Kéver et al. 2018). Additionally, hybrids might have a lower fitness. In order to reach mitochrondrial introgression, female hybrids would also need to mate with O. nasuta males. This is not unlikely given that, in cichlids, female mate choice is influenced, via imprinting, by the maternal phenotype (Verzijden and ten Cate 2007; Verzijden et al. 2008). Responses of Ophthalmotilapia females As no reproductions were recorded when these specimens still had a female morphology, we cannot rule out that they were males that delayed the development of the conspicuous male colouration and the elongated pelvic fins. Hence, we did not provide solid evidence of sex change in Ophthalmotilapia. 12 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Interpretation The importance of visual cues Although animals can use multiple kinds of cues to assess the quality of a potential mate, their final assessment depends on the overall information available. This is exemplified by female mate choice in the allopatric swordtail species Xiphophorus nigrensis Rosen 1960 and X. pygmaeus Hubbs and Gordon 1943. Here, mating preferences differed depending on whether visual, olfactory or a combination of both cues were available (Crapon de Caprona and Ryan 1990). Different responses to visual and olfactory cues were also shown for females of sympatric Cyprinodon Lacipède, 1803 pupfish species from Lake Chichancanab (Mexico). Here, different degrees of asymmetric discrimination of males were observed depending on whether females had access to visual or olfactory information (Strecker and Kodric-Brown 1999). In species-rich systems and in species that form leks, such as Ophthalmotilapia spp., females must be able to rapidly assess the quality of a potential mate (Barlow 2002). Ophthalmotilapia females could use the morphological characteristics, conspicuous bowers and/or stereotypical displays of males to distinguish them from sympatric congeners. However, even though multiple cues can be involved, mate choice decisions in radiations are often based on just a small amount of (combinations of) these traits (Hohenlohe and Arnold 2010). The separation wall used in our experiments contained holes that allowed for the exchange of water between both compartments. Hence, besides visual clues, the fishes most likely also received olfactory and acoustic information. Although visual cues were suggested to be the primary factor in species-isolating, female mate choice in other cichlids (Jordan et al. 2003; Kidd et al. 2006), we cannot rule out the importance of other types of information. Studies have shown that olfactory (Blais et al. 2009; Plenderleith et al. 2005), acoustic (Nelissen 1978; Amorim et al. 2004; Kéver et al. 2018) and behavioural (Barlow 2002) information can also influence the mating process. Although Seehausen and van Alphen (1998) showed a certain hierarchy of information, where other cues are taken into account when visual information is absent or masked, other experiments showed that female cichlids are more likely to select the right male when both olfactory and visual cues are available (Plenderleith et al. 2005; Blais et al. 2009). When visual information suffices for mate recognition, the behaviour throughout the mating process, i.e. potentially leading toward spawning, does not need to diverge between closely-related species (Barlow 2002). The role of males ventralis behaved differently when a visual barrier was present, no significant difference was found after its removal. This could imply that males of the two species behave in a similar way when presented with a conspecific or a heterospecific female. However, an alternative explanation would be that males recognise conspecific and heterospecifics, and use this knowledge to court females using a repertoire appropriate to the species. Although this was found in sister species pairs of freshwater sticklebacks (Gasterosteus spp. L. 1758) (Kozak et al. 2009), our experimental design did not allow us to test this in Ophthalmotilapia. We cannot exclude that morphological, physiological and behavioural features that distinguish O. nasuta males from males of congeners could have caused the asymmetric pattern of introgression. 13 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Foremost, O. nasuta males become larger and possess longer pelvic fins. This feature could render them more attractive as O. ventralis females have a preference towards males with strongly elongated pelvic fins (Haesler et al. 2011). As a change in the feature associated with attractiveness can alter species recognition in the mating process (Phelps et al. 2006), the extra-long pelvic fins of O. nasuta males could serve as a super-natural stimulus (sensu Tinbergen 1948). Additionally, even though Haesler et al. (2011) found no correlation between female choice and male body length in O. ventralis, they did observe that larger males outcompeted their rivals in sperm competition within the females’ mouth. Additionally, sperm of O. The role of males nasuta remains viable for a significantly longer amount of time than that of O. ventralis (Morita et al. 2014). Lastly, O. nasuta males construct true bowers (elaborate, crater-shaped sand mounts), whereas the nests of males of the other species of Ophthalmotilapia only consist of a small area of cleaned rock, or of a small pit in the sand (Konings 2019). The importance of visual cues This may explain why spawning behaviour of Ophthalmotilapia is remarkably similar across the genus (Kéver et al. 2018) and why differently-coloured, sympatric mbuna cichlids from Lake Malawi have identical courtship behaviours (McElroy and Kornfeld 1990). Supplementary material is available on line under: https://doi.org/10.1101/2021.08.07.455508. Supplementary material is available on line under: https://doi.org/10.1101/2021.08.07.45550 Ethical note All specimens were obtained from a certified commercial supplier (Cichlidenstadl, Alerheim, Germany). Specimens were housed in the aquarium facilities of the University of Liège. Experimental procedures were performed in accordance with Belgian law, and approved by the University of Liège Institutional Animal Care and Use Committee (protocol #1759) in accordance with the regulations of the ethical committee of the University of Liège. All manipulations were performed by a FELASA-certified experimenter. Ecological reasons for asymmetric hybridisation Although they can be found in sympatry, O. ventralis is more associated with the rocky shores of Lake Tanganyika, whereas O. nasuta has a wider ecological range. At rocky shores, O. ventralis can be one of the most abundant cichlid species (Sturmbauer et al. 2008). Hence, for an O. ventralis female, a random encounter with another Ophthalmotilapia male is much more likely to result in a conspecific than a heterospecific encounter. In contrast, for an O. nasuta female venturing into the preferred O. ventralis habitat, a conspecific encounter would be less often the case. Therefore, the ability to discriminate between conspecific and heterospecifics would be less important for females of O. ventralis than for those of O. nasuta. A similar interpretation was given to explain asymmetries in female discrimination of sympatric Cyprinodon species, where the species with the highest abundance had the lowest choosiness PEER COMMUNITY IN ZOOLOGY 14 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint (Strecker and Kodric-Brown 1999). Although a species’ propensity for discrimination could be a consequence of its distribution range, the opposite could also hold. Species that are better in recognizing conspecifics are more likely to maintain the integrity of their gene pool. Hence, they could be better in colonising habitats that have already been occupied by related species. Finally, we showed that substantial behavioural differences can be observed between closely-related species. This should be a warning to be cautious when assuming similarities in the behaviour of certain (model) organisms and related taxa. Acknowledgements We are grateful to Siegfried Loose for his advice on the rearing of Ophthalmotilapia and to Nedim Tüzün for his advice on the tracking software. We thank Ad Konings (Cichlid press) to provide us with pictures of Ophthalmotilapia. This study is part of the GENBAS project, which was funded by the BELSPO (Belgian Science Policy) BRAIN-be program. Version 3 of this preprint has been peer-reviewed and recommended by Peer Community In Zoology (https://doi.org/10.24072/pci.zool.100010). PEER COMMUNITY IN ZOOLOGY Data, script and code availability The data and script are deposited on Zenodo under: https://doi.org/10.5281/zenodo.5534485. Conflict of interest disclosure The authors of this preprint declare that they have no financial conflict of interest with the content of this article. MVS is one of the PCI Zool recommenders. PEER COMMUNITY IN ZOOLOGY 15 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint References Amorim MCP, Knight ME, Stratoudakis Y, Turner GF (2004) Differences in sounds made by courting males of three closely related Lake Malawi cichlids. J Fish Biol. 65(5):1358-1371. https://doi.org/10.1111/j.0022-1112.2004.00535.x Amorim MCP, Knight ME, Stratoudakis Y, Turner GF (2004) Differences in sounds made by courting males of three closely related Lake Malawi cichlids. J Fish Biol. 65(5):1358-1371. https://doi.org/10.1111/j.0022-1112.2004.00535.x Anderson MJ (2006) Distance-based tests for homogeneity of multivariate dispersions. Biometrics. 62(1):245–253. https://doi.org/10.1111/j.1541-0420.2005.00440.x Anderson MJ (2006) Distance-based tests for homogeneity of multivariate dispersions. Biometrics. 62(1):245–253. https://doi.org/10.1111/j.1541-0420.2005.00440.x Arnold SJ, Verrell PA, Tilley SG (1996) The evolution of asymmetry in sexual isolation: A model and a test case. Evolution 50:1024-1033. https://doi.org/10.1111/j.1558-5646.1996.tb02343.x Arnold SJ, Verrell PA, Tilley SG (1996) The evolution of asymmetry in sexual isolation: A model and a test case. Evolution 50:1024-1033. https://doi.org/10.1111/j.1558-5646.1996.tb02343.x Barlow GW (2002) How behavioural studies contribute to the species problem: a piscine perspective. Fish Fish. 3(3):197-212. https://doi.org/10.1046/j.1467-2979.2002.00083.x Barlow GW (2002) How behavioural studies contribute to the species problem: a piscine perspective. Fish Fish. 3(3):197-212. https://doi.org/10.1046/j.1467-2979.2002.00083.x Baerends GP, Baerends-Van Roon JM (1950) An introduction to the study of the ethology of the cichlid fishes, Leiden, The Netherlands, E. J. Brill. Baerends GP, Baerends-Van Roon JM (1950) An introduction to the study of the ethology of the cichlid fishes, Leiden, The Netherlands, E. J. Brill. Blais J, Plenderleith M, Rico C, Taylor MI, Seehausen O, van Oosterhout C, Turner GF (2009) Assortative mating among Lake Malawi cichlid fish populations is not simply predictable from male nuptial colour. BMC Evol Biol. 9(53):53. https://doi.org/10.1186/1471-2148-9-53 Coyne JA, Orr HA (2004) Speciation. Sunderland, MA: Sinauer Associates. Crapon de Caprona MD, Ryan MJ (1990) Conspecific mate recognition in swordtails, Xiphophorus nigrensis and X. pygmaeus (Poeciliidae): olfactory and visual cues. Anim Behav. 39(2):290-296. https://doi.org/10.1016/S0003-3472(05)80873-5 Danley PD, Kocher TD (2001) Speciation in rapidly diverging systems: lessons from Lake Malawi. Mol Ecol. 10:1075-1086. https://doi.org/10.1046/j.1365-294X.2001.01283.x Egger B, Obermüller B, Eigner E, Sturmbauer C, Sefc KM (2008) Assortative mating preferences between colour morphs of the endemic Lake Tanganyika cichlid genus Tropheus. Hydrobiologia. 615:37-48. https://doi.org/10.1007/978-1-4020-9582-5_3 Friard O, Gamba M (2016) BORIS: A Free, Versatile Open-Source Event-Logging Software for Video / Audio Coding and Live Observations. Methods Ecol Evol. 7(11):1–6. https://doi.org/10.1111/2041-210X.12584 Gonzalez-Voyer A, Fitzpatrick JL, Kolm N (2008) Sexual selection determines parental care pat cichlid fishes. Evolution 62(8): 2015-2026. https://doi.org/10.1111/j.1558-5646.2008.00426.x Haesler MP, Lindeyer CM, Otti O, Bonfils D, Heg D, Taborsky M (2011) Female mouthbrooders in control of pre- and postmating sexual selection. Behav Ecol. References 22(5):1033–1041. https://doi.org/10.1093/beheco/arr087 Hammer Ø, Harper DA, Ryan PD (2001) PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica. 4:9pp. Hanssens M, Snoeks J, Verheyen E (1999) A Morphometric Revision of the Genus Ophthalmotilapia (Teleostei, Cichlidae) from Lake Tanganyika (East Africa). Zool J Linn Soc-Lon. 125(4):487–512. https://doi.org/10.1111/j.1096-3642.1999.tb00602.x Ho J, Tumkaya T, Aryal S, Choi H, Claridge-Chang A (2019) Moving beyond P values: data analy estimation graphics. Nat Met. 16(7):565-566. https://doi.org/10.1038/s41592-019-0470-3 Ho J, Tumkaya T, Aryal S, Choi H, Claridge-Chang A (2019) Moving beyond P values: data analysis with estimation graphics. Nat Met. 16(7):565-566. https://doi.org/10.1038/s41592-019-0470-3 Hohenlohe PA, Arnold SJ (2010) Dimensionality of mate choice, sexual isolation, and speciation. PNAS. Hohenlohe PA, Arnold SJ (2010) Dimensionality of mate choice, sexual isolation, and speciatio . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint 107(38):16583-16588. https://doi.org/10.1073/pnas.1003537107 107(38):16583-16588. https://doi.org/10.1073/pnas.1003537107 ( ) p // g/ /p Immler S, Taborsky M (2009) Sequential polyandry affords post-mating sexual selection in the mouths of cichlid females. Behav Ecol Sociobiol. 63(8):1219–1230. https://doi.org/10.1007/s00265-009-0744-3 Jiggins CD, Salazar C, Linares M, Mavarez J (2008) Hybrid trait speciation and Heliconius butterflies. Philos T R Soc B. 363(1506):3047-3054. https://doi.org/10.1098/rstb.2008.0065 Jordan R, Kellogg K, Juanes F, Stauffer J (2003) Evaluation of female mate choice cues in a group of Lake Malawi mbuna (Cichlidae). Copeia. 2003(1):181-186. https://doi.org/10.1643/0045- 8511(2003)003[0181:EOFMCC]2.0.CO;2 Kéver L, Parmentier E, Derycke S, Verheyen E, Snoeks J, Van Steenberge M, Poncin P (2018) Limited possibilities for prezygotic barriers in the reproductive behaviour of sympatric Ophthalmotilapia species (Teleostei, Cichlidae). Zoology. 126:71-81. https://doi.org/10.1016/j.zool.2017.12.001 16 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Kidd MR, Danley PD, Kocher TD (2006) A direct assay of female choice in cichlids: all the eggs in one basket. J Fish Biol. 68(2):373–384. https://doi.org/10.1111/j.0022-1112.2006.00896.x Kidd MR, Danley PD, Kocher TD (2006) A direct assay of female choice in cichlids: all the eggs in one basket. J Fish Biol. 68(2):373–384. https://doi.org/10.1111/j.0022-1112.2006.00896.x p g j Kocher TD (2004) Adaptive evolution and explosive speciation: the cichlid fish model. Nat Rev Genet. 5(4):288-298. . https://doi.org/10.1038/nrg1316 Koblmüller S, Duftner N, Sefc KM, Aibara M, Stipacek M, Blanc M, Egger B, Sturmbauer C (2007) Reticulate phylogeny of gastropod-shell-breeding cichlids from Lake Tanganyika - the result of repeated introgressive hybridisation. BMC Evol Biol. 7:7. https://doi.org/10.1186/1471-2148-7-7 g y p g Konings A (2019) Tanganyika cichlids in their natural habitat. 4th edition. El Paso, TX: Cichlid press. Kozak GM, Boughman JW (2009) Learned conspecific mate preference Behav Ecol. 20:1282–1288. https://doi.org/10.1093/beheco/arp134 Kozak GM, Reisland M, Boughmann JW (2009) Sex differences in mate recognition and conspecific preference in species with mutual mate choice. Evolution. 63:353-365. https://doi.org/10.1111/j.1558- 5646.2008.00564.x Luttbeg B, Towner MC, Wandesforde-Smith A, Mangel M, Foster SA (2001) State-dependent mate- assessment and mate-selection behavior in female threespine sticklebacks (Gasterosteus aculeatus, Gasterosteiformes: Gasterosteidae). Ethology. 107(6):545-558. https://doi.org/10.1046/j.1439- 0310.2001.00694.x Mayr E (1982) The growth of biological thought: Diversity, evolution and inheritance. Cambridge, MA: Harvard University Belknap Press. Marques DA, Meier JI, Seehausen O (2019) A combinatorial view on speciation and adaptive radiation. Trends Ecol Evol. 34(6):531-544. https://doi.org/10.1016/j.tree.2019.02.008 Mendelson TC, Shaw KL (2012) The (mis)conce https://doi.org/10.1016/j.tree.2012.04.001 Mendelson TC, Shaw KL (2012) The (mis)concept of species recognition. Trends Ecol Evol. 27(8):421-427. https://doi.org/10.1016/j.tree.2012.04.001 McElroy DM, Kornfeld I (1990) Sexual selection, reproductive behaviour, and speciation in the species fock of Lake Nyasa (Pisces: Cichlidae). Environ Biol Fish. 28:273-284. p y ( ) Moran MD (2003) Arguments for rejecting the sequential Bonferroni in ecological studies. Oicos. 100(2):403 405 https://doi org/10 1034/j 1600 0706 2003 12010 x Moran MD (2003) Arguments for rejecting the sequential Bonferroni in ecological studies 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x Adaptive sequence divergence in a colour gene involved in the formation of the characteristic egg-dummies of male Haplochromis cichlid fishes. BMC Biol. 5(51):13pp. https://doi.org/10.1186/1741-7007-5-51 https://doi.org/10.1016/j.tree.2004.01.003 Seehausen O, van Alphen JJM (1998) The effect of male coloration on female mate choice in closely related Lake Victoria cichlids (Haplochromis nyererei complex). Behav Ecol Sociobiol. 42:1–8. https://doi.org/ en O, van Alphen JJM (1998) The effect of male coloration on female mate choice in closely related Victoria cichlids (Haplochromis nyererei complex). Behav Ecol Sociobiol. 42:1–8. https://doi.org/ Seehausen O, Terai Y, Magalhaes IS, Carleton KL, Mrosso HDJ, Miyagi R, van der Sluijs I, Schneider MV, Maan ME, Tachida H, Imai H, Okada N (2008) Speciation through sensory drive in cichlid fish. Nature. 455:620-626. https://doi.org/10.1038/nature07285 Seehausen O, Terai Y, Magalhaes IS, Carleton KL, Mrosso HDJ, Miyagi R, van der Sluijs I, Schneider MV, Maan ME, Tachida H, Imai H, Okada N (2008) Speciation through sensory drive in cichlid fish. Nature. 455:620-626. https://doi.org/10.1038/nature07285 Sefc KM, Mattersdorfer K, Ziegelbecker A, Neuhüttler N, Steiner O, Goessler W, Koblmüller S (2017) Shifting barriers and phenotypic diversification by hybridisation. Ecol Lett. 20(5):651-662. https://doi.org/10.1111/ele.12766 Sefc KM, Mattersdorfer K, Ziegelbecker A, Neuhüttler N, Steiner O, Goessler W, Koblmüller S (2017) Shifting barriers and phenotypic diversification by hybridisation. Ecol Lett. 20(5):651-662. https://doi.org/10.1111/ele.12766 Sommer-Trembo C, Plath M, Gismann J, Helfrich C, Bierbach D (2017) Context-dependent female mate choice maintains variation in male sexual activity. R Soc Open Sci. 4:170303. https://doi.org/10.1098/rsos.170303 Sommer-Trembo C, Plath M, Gismann J, Helfrich C, Bierbach D (2017) Context-dependent female mate choice maintains variation in male sexual activity. R Soc Open Sci. 4:170303. https://doi.org/10.1098/rsos.170303 Strecker U, Kodric-Brown A (1999) Mate recognition systems in a species flock of Mexican p Evolution Biol. 12(5):927-935. https://doi.org/10.1046/j.1420-9101.1999.00096.x Sturmbauer C, Fuchs C, Harb G, Damm E, Duftner N, Maderbacher M, Koch M, Koblmüller S (2008) Abundance, distribution, and territory areas of rock-dwelling Lake Tanganyika cichlid fish species. Hydrobiologia. 615(1):57-68. https://doi.org/10.1007/978-1-4020-9582-5_5 Sturmbauer C, Fuchs C, Harb G, Damm E, Duftner N, Maderbacher M, Koch M, Koblmüller S (2008) Abundance, distribution, and territory areas of rock-dwelling Lake Tanganyika cichlid fish species. Hydrobiologia. 615(1):57-68. https://doi.org/10.1007/978-1-4020-9582-5_5 Sullivan BK (2009) Mate recognition, species boundaries and the fallacy of “species recognition”. Open Zool. 2:86-90. https://doi.org/10.2174/1874336601002009086 Sullivan BK (2009) Mate recognition, species boundaries and the fallacy of “species recognition”. Open Zool. 2:86-90. https://doi.org/10.2174/1874336601002009086 Sullivan BK (2009) Mate recognition, species boundaries and th 2:86-90. Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x https://doi.org/10.1098/rsbl.2005.0355 Plenderleith M, van Oosterhout C, Robinson RL, Turner GF (2005) Female preference for conspecific males based on olfactory cues in a Lake Malawi cichlid fish. Biol Lett. 1(4):411-414. https://doi.org/10.1098/rsbl.2005.0355 Poll M (1986) Classification des cichlidae du Lac Tanganyika: Tribus, genres et espèces. Bull Acad R Belg. 8:1-163. R core team (2017) R: A Language and environment for statistical computing. R foundation for statistical computing, Vienna, Austria. R core team (2017) R: A Language and environment for statistical computing. R foundation for statistical computing, Vienna, Austria. Ronco F, Büscher HH, Indermauer A, Salzburger W (2020). The taxonomic diversity of the cichlid fish fauna of ancient Lake Tanganyika, East Africa. J. Great Lakes Res. 46:1067-1078. https://doi.org/10.1016/j.jglr.2019.05.009 Ronco F, Büscher HH, Indermauer A, Salzburger W (2020). The taxonomic diversity of the cichlid fish fauna of ancient Lake Tanganyika, East Africa. J. Great Lakes Res. 46:1067-1078. https://doi.org/10.1016/j.jglr.2019.05.009 Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Met. 9(7):671-675. https://doi.org/10.1038/nmeth.2089 Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Met. 9(7):671-675. https://doi.org/10.1038/nmeth.2089 Roux C, Fraïsse C, Romiguier J, Anciaux Y, Galtier N, Bierne N (2016) Shedding light on the grey zone of speciation along a continuum of genomic divergence. PLOS Biol. 14(12):e2000234. https://doi.org/10.1371/journal.pbio.2000234 Roux C, Fraïsse C, Romiguier J, Anciaux Y, Galtier N, Bierne N (2016) Shedding light on the grey zone of speciation along a continuum of genomic divergence. PLOS Biol. 14(12):e2000234. https://doi.org/10.1371/journal.pbio.2000234 Rüber L, Meyer A, Sturmbauer C, Verheyen E (2001). Population structure in two sympatric species of the Lake Tanganyika cichlid tribe Eretmodini: Evidence for introgression. Mol Ecol. 10(5):1207–1225. https://doi.org/10.1046/j.1365-294x.2001.01259.x Rüber L, Meyer A, Sturmbauer C, Verheyen E (2001). Population structure in two sympatric species of the Lake Tanganyika cichlid tribe Eretmodini: Evidence for introgression. Mol Ecol. 10(5):1207–1225. https://doi.org/10.1046/j.1365-294x.2001.01259.x Salzburger W (2018). Understanding explosive diversification through cichlid fish genomics. Nat Rev Genet. 19:705-717. https://doi.org/10.1038/s41576-018-0043-9 Salzburger W (2018). Understanding explosive diversification through cichlid fish genomics. Nat Rev Genet. 19:705-717. https://doi.org/10.1038/s41576-018-0043-9 Salzburger W (2018). Understanding explosive diversification t 19:705-717. https://doi.org/10.1038/s41576-018-0043-9 Salzburger W, Braasch I, Meyer A (2007). Adaptive sequence divergence in a colour gene involved in the formation of the characteristic egg-dummies of male Haplochromis cichlid fishes. BMC Biol. 5(51):13pp. https://doi.org/10.1186/1741-7007-5-51 Salzburger W, Braasch I, Meyer A (2007). Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x Morita M, Awata S, Yorifuji M, Ota K, Kohda M, Ochi H (2014) Bower-building behaviour is associated with increased sperm longevity in Tanganyikan cichlids. J Evolution Biol. 27(12):2629–2643. https://doi.org/10.1111/jeb.12522 Naish KA, Ribbink AJ (1990) A preliminary investigation of sex change in Pseudotropheus lombardoi (Pisces: Cichlidae). Environ Biol Fish. 28:285–294. Nakagawa S (2004) A farewell to Bonferroni: the problem if low statistical power and publication bias. Behav. Ecol. 15:1044-1045. https://doi.org/10.1093/beheco/arh107 Nelissen MHJ (1978) Sound production by some Tanganyikan cichlid fishes and a hypothesis for the evolution of their communication mechanisms. Behaviour. 64(1-2):137–147. https://doi.org/10.1163/156853978X00477 Nevado B, Fazalova V, Backeljau T, Hanssens M, Verheyen E (2011) Repeated unidirectional introgression of nuclear and mitochondrial DNA between four congeneric Tanganyikan cichlids. Mol Biol Evol. 28(8):2253-2267. https://doi.org/10.1093/molbev/msr043 Oksanen JF, Blanchet G, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin PR, O'Hara RB, Simpson GL, Solymos PM, Stevens HH, Szoecs E, Wagner H (2017) vegan: Community Ecology Package. R package version 2.4-5. Panfili J, Thior D, Ecoutin J-M, Ndiaye P, Albaret J-J (2006) Influence of salinity on the size at maturity for fish species reproducing in contrasting West African estuaries. J Fish Biol. 69(1):95-113. https://doi.org/10.1111/j.1095-8649.2006.01069.x Peters HM (1975) Hermaphroditism in cichlid fishes. pp. 228–235.In: R. Reinboth (ed.) Intersexuality in the Animal Kingdom. Springer-Verlag, Berlin. https://doi.org/10.1007/978-3-642-66069-6_22 Pfennig KS (2007) Facultative mate choice drives adaptive hybridization. Science, 318(5852):965. https://doi.org/10.1126/science.1146035 Phelps SM, Rand AS, Ryan MJ (2006) A cognitive framework for mate choice and species recognition. Am Nat. 167(1):28–42. https://doi.org/10.1086/498538 17 17 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Plenderleith M, van Oosterhout C, Robinson RL, Turner GF (2005) Female preference for conspecific males based on olfactory cues in a Lake Malawi cichlid fish. Biol Lett. 1(4):411-414. Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x Animal Behav. 58:1–12. https://doi.org/10.1006/anbe.1999.1144. Wong BMB, Jennions MD (2003) Costs influence male mate choice in 270(Suppl 1):S36-S38. https://doi.org/10.1098/rsbl.2003.0003 Wong BMB, Jennions MD (2003) Costs influence male mate choice in a freshwater fish. P Roy Soc B-Biol Sci. 270(Suppl 1):S36-S38. https://doi.org/10.1098/rsbl.2003.0003 Wong BMB, Jennions MD (2003) Costs influence male mate choice in a freshwater fish. P Roy Soc B-Biol Sci. 270(Suppl 1):S36-S38. https://doi.org/10.1098/rsbl.2003.0003 Zoppoth P, Koblmüller S, Sefc KM (2013) Male courtship preferences demonstrate discrimination against allopatric colour morphs in a cichlid fish. J Evolution Biol. 26(3):577-586. https://doi.org/10.1111/jeb.12074 Zoppoth P, Koblmüller S, Sefc KM (2013) Male courtship preferences demonstrate discrimination against allopatric colour morphs in a cichlid fish. J Evolution Biol. 26(3):577-586. https://doi.org/10.1111/jeb.12074 Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x Van Oppen MJH, Turner GF, Rico C, Robinson RL, Deutsch JC, Genner MJ, Hewitt GM (1998) Assortative mating among rock-dwelling cichlid fishes supports high estimates of species richness from Lake Malawi. Mol Ecol. 7(8):991-1001. https://doi.org/10.1046/j.1365-294x.1998.00417.x Van Oppen MJH, Turner GF, Rico C, Robinson RL, Deutsch JC, Genner MJ, Hewitt GM (1998) Assortative mating among rock-dwelling cichlid fishes supports high estimates of species richness from Lake Malawi. Mol Ecol. 7(8):991-1001. https://doi.org/10.1046/j.1365-294x.1998.00417.x Verrell PA (1990) Frequency of interspecific mating in salamanders of the plethodontid genus Desmognathus: different experimental designs may yield different results. J Zool. 221(3):441-451. https://doi.org/10.1111/j.1469-7998.1990.tb04012.x Verrell PA (1990) Frequency of interspecific mating in salamanders of the plethodontid genus Desmognathus: different experimental designs may yield different results. J Zool. 221(3):441-451. https://doi.org/10.1111/j.1469-7998.1990.tb04012.x Verzijden MN, Korthof REM, ten Cate C (2008) Females learn from mothers and males learn from others. The effect of mother and siblings on the development of female mate preferences and male aggression biases in Lake Victoria cichlids, genus Mbipia. Behav Ecol Sociobiol. 62: 1359–1368. https://doi.org/10.1007/s00265-008-0564-x Verzijden MN, Korthof REM, ten Cate C (2008) Females learn from mothers and males learn from others. The effect of mother and siblings on the development of female mate preferences and male aggression biases in Lake Victoria cichlids, genus Mbipia. Behav Ecol Sociobiol. 62: 1359–1368. https://doi.org/10.1007/s00265-008-0564-x Verzijden MN, ten Cate, C (2007) Early learning influences species assortative mating preferences in Lake Victoria cichlid fish. Biol lett. 3(2): 134-136. https://doi.org/10.1098/rsbl.2006.0601 Verzijden MN, ten Cate, C (2007) Early learning influences species assortative mating preferences in Lake Victoria cichlid fish. Biol lett. 3(2): 134-136. https://doi.org/10.1098/rsbl.2006.0601 Werner NY, Lotem A (2003) Choosy males in a haplochromine cichlid: first experimental evidence for male mate choice in a lekking species. Anim Behav. 66(2):293-298. https://doi.org/10.1006/anbe.2003.2208 Werner NY, Lotem A (2003) Choosy males in a haplochromine cichlid: first experimental evidence for male mate choice in a lekking species. Anim Behav. 66(2):293-298. https://doi.org/10.1006/anbe.2003.2208 Wirtz, P (1999) Mother species-father species: unidirectional hybridization in animals with female choice. Animal Behav 58:1–12 https://doi org/10 1006/anbe 1999 1144 Werner NY, Lotem A (2003) Choosy males in a haplochromine cichlid: first experimental evidence for male mate choice in a lekking species. Anim Behav. 66(2):293-298. https://doi.org/10.1006/anbe.2003.2208 Wirtz, P (1999) Mother species-father species: unidirectional hybridization in animals with female choice. Animal Behav. 58:1–12. https://doi.org/10.1006/anbe.1999.1144. Wirtz, P (1999) Mother species-father species: unidirectional hybridiz Animal Behav. 58:1–12. https://doi.org/10.1006/anbe.1999.1144. Wirtz, P (1999) Mother species-father species: unidirectional hybridization in animals with female choice. Supplementary material Supplementary material is available on line under: https://doi.org/10.1101/2021.08.07.455508. Moran MD (2003) Arguments for rejecting the sequential Bonferron 100(2):403-405. https://doi.org/10.1034/j.1600-0706.2003.12010.x https://doi.org/10.2174/1874336601002009086 Svensson EI, Karlsson K, Friberg M, Eroukhmanoff F (2007) Gender differences in species recognition and the evolution of asymmetric sexual isolation. Curr Biol. 17(22):1943–1947. https://doi.org/10.1016/j.cub.2007.09.038 Svensson EI, Karlsson K, Friberg M, Eroukhmanoff F (2007) Gender differences in species recognition and the evolution of asymmetric sexual isolation. Curr Biol. 17(22):1943–1947. https://doi.org/10.1016/j.cub.2007.09.038 Tinbergen N (1948) Social releasers and the experimental method required for their study. Wilson Bull. 60(1):6-51. Tinbergen N (1948) Social releasers and the experimental method required for their study. Wilson Bull. 60(1):6-51. Turner GF, Seehausen O, Knight ME, Allender CJ, Robinson RL (2001) How many species of cichlid fishes are there in African lakes. Mol Ecol. 10(3):793-806. https://doi.org/10.1046/j.1365-294x.2001.01200.x Turner GF, Seehausen O, Knight ME, Allender CJ, Robinson RL (2001) How many species of cichlid fishes are there in African lakes. Mol Ecol. 10(3):793-806. https://doi.org/10.1046/j.1365-294x.2001.01200.x Theis A, Salzburger W, Egger B (2012) The function of anal fin egg-spots in the cichlid fish Astatotilapia burtoni. PLoS ONE 7(1):e29878. https://doi.org/10.1371/journal.pone.0029878 Theis A, Salzburger W, Egger B (2012) The function of anal fin egg-spots in the cichlid fish Astatotilapia burtoni. PLoS ONE 7(1):e29878. https://doi.org/10.1371/journal.pone.0029878 18 PEER COMMUNITY IN ZOOLOGY . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint Thomas GH, Székely T (2005) Evolutionary pathways in shorebird breeding systems: sexual conflict, parental care, and chick development. Evolution 59(10):2222–2230. https://doi.org/10.1111/j.0014- 3820.2005.tb00930.x Thomas GH, Székely T (2005) Evolutionary pathways in shorebird breeding systems: sexual conflict, parental care, and chick development. Evolution 59(10):2222–2230. https://doi.org/10.1111/j.0014- 3820.2005.tb00930.x Trivers RL (1972) Parental investment and sexual selection. in Sexual selection and the descent of man, EBG. Campbell, ed. (136-179). Aldine, Chicago. Trivers RL (1972) Parental investment and sexual selection. in Sexual selection and the descent of man, EBG. Campbell, ed. (136-179). Aldine, Chicago. Supplementary material is available on line under: https://doi.org/10.1101/2021.08.07.455508. Supplement 1. Summary of the experiments. Supplement 2. Indications for sex change observed in O. nasuta females. Top row: gonads of some of the specimens that were female in external phenotype but had male or ambiguous gonads (ON33, 35, 26, 43). Vertical left: female gonads in developmental stage 3 (ON34), 4 (ON24) and 5 (ON37) respectively (Panfili et al. 2006). Bottom: two presumed non-focal O. nasuta females used in the ON experiment, one of which could have undergone transition after the experiments, and horizontal right, the ventral area of the same specimens, with A: anal pore and UG: urogentital pore. Supplement 3. Visualization of tracking data. The position of each fish is plotted for each second in which specimens were tracked with the positions recorded before and after the removal of the opaque wall (grey) coloured differently. Ellipses denote the area in which 90% of tracks are situated, large dots denote the average positions before and after the removal of the barrier and arrows shows the change in mean position. The separation wall is visualized as a meshed partition. The average speed before (v0) and after (v1) the removal of the barrier is plotted for each tracked specimen with data for focal specimens given in bold. Abbreviations (ON: O. nasuta, OV: O. ventralis, F: female, M; male) Supplement 4. Summary of the data. pp y Supplement 5. Loadings and variance of the main axes of the PCAs and CVAs conducted in the study. Supplement 5. Loadings and variance of the main axes of the PCAs and CVAs conducted in the study. Supplement 6. Principal component analyses and Canonical variate analyses performed on the behaviours recorded 45 min after the visual barrier was removed of the ON (O. nasuta, left) and the OV (O. ventralis, right) experiment. Symbols on the scatter plots for the ON and OV experiment as in E and F, respectively, with full circles denoting focal females presented with no fish (red), a conspecific female (blue) an O. ventralis male (purple), and an O. nasuta male (green), empty circles denote non-focal conspecific PEER COMMUNITY IN ZOOLOGY 19 . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . PEER COMMUNITY IN ZOOLOGY Supplementary material is available on line under: https://doi.org/10.1101/2021.08.07.455508. CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint . CC-BY 4.0 International license perpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for this this version posted February 7, 2022. ; https://doi.org/10.1101/2021.08.07.455508 doi: bioRxiv preprint females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. females and full squares O. ventralis (purple) and O. nasuta (green) males. Explained variances are added to the axes. Supplement 7. Shift in average position of the specimens analysed 15min before and 15min after the removal of the separation wall for the ON (above) and OV (below) experiment. Focal specimens are all visualised on the left, and non-focal specimens on the right. Dimensions in cm, with the vertical bar representing the separation wall. Dashed arrows represent individual fishes, bold arrows the average per treatment. Colours, for focal females (ON and OV) presented with no fish (red), a conspecific female (turquoise) an O. ventralis male (pink), and an O. nasuta male (light green), and non-focal specimens (right): conspecific females (blue), O. ventralis males (purple) and O. nasuta males (green). 20 PEER COMMUNITY IN ZOOLOGY 20
https://openalex.org/W4393387616	https://interactive-plus.ru/e-articles/903/Action903-561991.pdf	Russian	null	Printsip neopredelennosti dlia raznykh sistem koordinat	Relevant lines of scientific research: theory and practice	2,024	cc-by	2,041	Scientific Cooperation Center "Interactive plus" ПРИНЦИП НЕОПРЕДЕЛЕННОСТИ ДЛЯ РАЗНЫХ СИСТЕМ КООРДИНАТ Аннотация: при рассмотрении нескольких систем координат появляется периодичность как следствие аналитических выражений уравнений сдвинутых функции в разных системах координат. Доказательство теорем и опирается только на известные факты математического анализа. Во второй части статьи приводятся теоремы, описывающие непрерывно развивающиеся во времени процессы, позволяющие восстанавливать произвольные значения функций по значениям в целых точках 0, 1, 2, …, связанные с преобразованиями Лапласа. Данные теоремы аналогичны тождествам Шеннона-Котельникова для преобразований Лапласа, и приводят к доказательству проблемы Гильбер- та о существовании дискретного гильбертова пространства на всей дей- ствительной положительной полуоси. Ключевые слова: периодичность аналитической функции, неоднознач- ность представления функций, разные системы координат, тождества типа Шеннона-Котельникова, преобразование Лапласа. Введение. Статья посвящена двум темам: периодичности произвольных действитель- ных и комплексных функций, возникающих при рассмотрении разных систем координат, [1–3], и проблеме Гильберта о существовании гильбертова про- странства для функций, определенных на бесконечных носителях типа ( , ),[0, ) − , вытекающей из результатов автора, посвященных тождествам Шеннона-Котельникова для преобразований Лапласа в работе [4]. 1 Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0) Центр научного сотрудничества «Интерактив плюс» В первой части работы, являющейся продолжением статьи [3] автора, мы используем обозначения: уравнение произвольной функции ( ) y f x = ( ( ) z f p = для комплексного случая) в новой системе координат с центром в точке ( 0) A совпадает с уравнением. ( ) ( ) , z f r A g r p x r X G = + =  = =  p r A r = +  – комплексные или действительные переменные в исходной и новой системе координат при произвольном действительном 0 A , G – некоторая откры- тая область комплексной плоскости, (интервал для действительного случая). В тео- реме 1 с точки зрения обыкновенных фактов математического анализа возможно возникновение периодичности для практически произвольного класса функций при использовании новых систем координат при разных A. Во второй части данной статьи доказывается тождество типа тождества Шеннона-Котельникова, из которого следует явное представление значений та- ких функций при ( , ), x− по значениям в целых точках 0, 1, 2, …, (тожде- ство (1)). Приводится явная форма такого представления в виде ряда с извест- ными коэффициентами одними и теми же для разных функций, (равенство (2)). 2 https://interactive-plus.ru 2 https://interactive-plus.ru Содержимое доступно по лицензии Creative Commons Attribution 4.0 license (CC-BY 4.0) ПРИНЦИП НЕОПРЕДЕЛЕННОСТИ ДЛЯ РАЗНЫХ СИСТЕМ КООРДИНАТ 1 Периодичность произвольных функций Как и в статье [3] определим понятие аналитического отображения точек плоскости для функции ( ) z f p = как отображения точек (не векторов) концов радиус-векторов p плоскости с помощью уравнения ( ) z f p p G =   . Для простоты изложения в первой части статьи в области значений функ- ции ( ) z f p = не существует двух одинаковых значений: 1 2 ( ) ( ) f p f p  при всех 1 2 1 2 p p p p G     . Доказательство теоремы 1 для случая действительных или комплексных переменных дословно одно и то же. Теорема 1. Для произвольной функции ( ) z f p = , определенной в некоторой открытой области G (интервале) исходное аналитическое многообразие точек (многооб- 2 https://interactive-plus.ru Содержимое доступно по лицензии Creative Commons Attribution 4.0 license (CC-BY 4.0) Scientific Cooperation Center "Interactive plus" разие для уравнений ( ), ( ) z f p z g r = = ) имеет два решения уравнения ( ) z g r = при любом z из области значений z . Доказательство. Первое решение определяется из исходного равенства 0 ( ) z g r = . Для доказательства существования второго решения отметим, что в равен- стве ( ) ( ) z g r A f P = − = при 0 P r = дополнительную переменную P с концом P в точке конца радиус-вектора r A − можно считать переобозначенным комплекс- ным аргументом p , (радиус-вектор P проведен из центра координат в точке (0,0), радиус-вектор r A − проведен из центра координат в точке (A,0)). Данный факт вытекает из совпадения многообразия {( , ): ( )}, z P z f P P G =  и {( , ): ( ), } z r P z g P A P G = = −  ввиду равенства ( ) ( ) g p A f p − = , [3]. Так как 1 1 0 ( ) ( ), z f P g r r r A = = = − то теорема 1 доказана. (Можно также было заметить, что ( ), ( ) z f P z f p A = = − при всех P p G =  – два уравнения одного ана- литического исходного многообразия точек, для уравнения ( ) z g r = ). Замечание. ПРИНЦИП НЕОПРЕДЕЛЕННОСТИ ДЛЯ РАЗНЫХ СИСТЕМ КООРДИНАТ , t t   выполнено: 0 0 0 0 0 0 ( ) [1 ( 1)] ( ) ( )( 1) ( ) ( ) ( 1) ( ) . k u t u k m k m mu k k k m r t e Z u du s k e Z u du s k C e Z u du      − − − − = = = = + − = − = −       (2) (2) так как ряд под интегралом сходится равномерно [1–6], и следовательно сумма может быть вынесена за знак интеграла. При [0, ), 1.2,..., t K K  −  остаток ряда под интегралом в (1.2) ввиду / !, t  k \|s(k) t мажорируется: ([ ]) [ ] [ ] ([ ]) \| ( )( 1) \| \| ( )( 1) \| ( 1) 1 (( 1) ) . ( ) t u k u k t u u k k N k N k N t u u N t u s k e s k e e e e e e    − − − = = = − − −  −  −   −    (3) (3) Здесь [t] – целая часть числа \|\|t\|, [ ] \| \|. t t  После [t] интегрирований по частям в интеграле: ( ) ([ ]) [ ] \| ( ) \| ( 1) ( ) t t u u N u R e Z u e e du A R  −− =  (4) (4) получаем из условия леммы 1 с учетом ( ) ( 1) \| ( ) \| 0, , 0,1,..., , l K x Z x e x l K + → → = что ( ) 0, . A R R → → получаем из условия леммы 1 с учетом ( ) ( 1) \| ( ) \| 0, , 0,1,..., , l K x Z x e x l K + → → = что ( ) 0, . A R R → → Так как в интеграле до R возможность вынести бесконечную сумму за знак интеграла очевидна ( \| 1\| \| 1\| u R e e − − − − ), то теорема 2 доказана. ПРИНЦИП НЕОПРЕДЕЛЕННОСТИ ДЛЯ РАЗНЫХ СИСТЕМ КООРДИНАТ Отметим также следующий факт: результаты отражения функции ( ) z f P = относительно точек (0,0,) и (A,0) совпадают с функциями ( ), (2 ) z f p z f A p = − = − . В случае двойной четности относительно данных точек эти равенства эквива- лентны периодичности аналитической функции ( ) z f p = , [0,2 ] A G  . 2. Тождество типа Шеннона-Котельникова для спектральных разложе- ний и преобразований Лапласа. 2. Тождество типа Шеннона-Котельникова для спектральных разложе- ний и преобразований Лапласа. Теорема 2. Пусть 0 ( ) ( ) . ut r t Z u e du  − =  Пусть 0 ( ) ( ) . ut r t Z u e du  − =  Пусть 0 ( ) ( ) . ut r t Z u e du  − =  1. Если Z(x) произвольная функция с к не более чем конечным числом разры- вов на [0, )  такая, что 0 \| ( ) \| . Z u du    то при всех [0, ) t   имеет место тождество: 0 0 ( ) ( ) ( 1) ( ), ( ) ( 1)...( 1)/ !, 0; (0) 1. k m k m k k m r t s k C r m s k t t t k k t s  − = = = − − = − − +  =   (1) Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0) 3 Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0) 3 Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0) Центр научного сотрудничества «Интерактив плюс» 2. Если ( 1) 0 \| ( ) \| . K X Z X e dX  +   , и функция Z(x) имеет непрерывную на [0, )  производную К -ого порядка, то при [0, ) t K  , K=1, 2, …, тождество (1) тоже выполнено. (При целых положительных 0,1,... t N =  , первая сумма в (1) заменяется на соответствующую конечную сумму по k от нуля до N и равна r(N)). Доказательство. При 0,1,... t  теорема очевидна. 0,1,...  теорема очевидна. При 0,1,... t  теорема очевидна. При всех 0, 1,2,... , t t   выполнено: При всех 0, 1,2,... 4 https://interactive-plus.ru Содержимое доступно по лицензии Creative Commons Attribution 4.0 license (CC-BY 4.0) 4 https://interactive-plus.ru Содержимое доступн ПРИНЦИП НЕОПРЕДЕЛЕННОСТИ ДЛЯ РАЗНЫХ СИСТЕМ КООРДИНАТ Так как в интеграле до R возможность вынести бесконечную сумму за знак интеграла очевидна ( \| 1\| \| 1\| u R e e − − − − ), то теорема 2 доказана. Scientific Cooperation Center "Interactive plus" 6. Чубариков В.Н. Об асимптотических формулах для интеграла И.М. Ви- ноградова / В.Н. Чубариков // Математический институт АН СССР: сборник трудов. – 1981. – Т. 157. – С. 214 -232. Список литературы 1. Pavlov, A.V. Different coordinate systems and periodicity. Vol. State Univer. Mathematical Physics and Computer Simulation. 2023. 26. No. 3. Pp. 114–118. DOI 10.15688/mpcm.jvolsu.2023.3.9. EDN ZMJMLT 2. Павлов А.В. Отраженные функции и периодичность / А.В. Павлов // International Journal of Open Information Technologies. – 2022. – №6. – С. 33–39. – EDN RREOAZ 2. Павлов А.В. Отраженные функции и периодичность / А.В. Павлов // International Journal of Open Information Technologies. – 2022. – №6. – С. 33–39. – EDN RREOAZ 2. Павлов А.В. Отраженные функции и периодичность / А.В. Павлов // International Journal of Open Information Technologies. – 2022. – №6. – С. 33–39. – EDN RREOAZ EDN RREOAZ 3. Павлов А.В. Периодичность для разных систем координат / А.В. Павлов // Вопросы науки и образования: новые подходы и актуальные исследования: сбор- ник материалов Всеросс. науч.-практ. конф. (Чебоксары, 29 февраля 2024 г.). – Чебоксары: Интерактив плюс, 2024. DOI 10.21661/r-561906. 3. Павлов А.В. Периодичность для разных систем координат / А.В. Павлов // Вопросы науки и образования: новые подходы и актуальные исследования: сбор- ник материалов Всеросс. науч.-практ. конф. (Чебоксары, 29 февраля 2024 г.). – Чебоксары: Интерактив плюс, 2024. DOI 10.21661/r-561906. 4. Павлов А.В. Случайные ряды Фурье и их применение к теории фильтра- ции-прогноза / А.В. Павлов. – М.: МГУ им. М.В. Ломоносова, 2000. – 64 с. ISBN 5–93839–002–8. 4. Павлов А.В. Случайные ряды Фурье и их применение к теории фильтра- ции-прогноза / А.В. Павлов. – М.: МГУ им. М.В. Ломоносова, 2000. – 64 с. ISBN 5–93839–002–8. 5. Лаврентьев М.А. Методы теории функций комплексного переменного / М.А. Лаврентьев, Б.В. Шабат. – M.: Наука, 1987. – 688 с. 5. Лаврентьев М.А. Методы теории функций комплексного переменного / М.А. Лаврентьев, Б.В. Шабат. – M.: Наука, 1987. – 688 с. 6. Чубариков В.Н. Об асимптотических формулах для интеграла И.М. Ви- ноградова / В.Н. Чубариков // Математический институт АН СССР: сборник трудов. – 1981. – Т. 157. – С. 214 -232. 6. Чубариков В.Н. Об асимптотических формулах для интеграла И.М. Ви- ноградова / В.Н. Чубариков // Математический институт АН СССР: сборник трудов. – 1981. – Т. 157. – С. 214 -232. 5 5 Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0) Content is licensed under the Creative Commons Attribution 4.0 license (CC-BY 4.0)
https://openalex.org/W1545987713	http://www.scielo.br/pdf/rbme/v20n4/1517-8692-rbme-20-04-00315.pdf	Portuguese	null	Foot strike patterns in veteran runners with respect to the increase of the speed and fatigue	null	2,014	cc-by	5,368	ABSTRACT Introduction: In the last few decades the number of participants in the popular endurance races has increased,especially for athletes in the category of veterans (over 35 years), as well as the number of organized events. Objective: To analyze the foot strike patterns in veteran runners with respect to an increase of speed and fatigue. Methods: Participation of 43 athletes, 40 men (mean age: 41.3±1.10 years old) and 3 women (mean age: 44.3±3.17 years old). After warm-up, three running protocols were performed: comfortable pace, competition pace and competicion pace until the increase two values of the Borg scale, freely chosen by each athlete. Foot strike pat- terns (contact surface, lateral rotation and tilting of the foot) and a time sequence were analyzed by high-speed video. Results: 84.5% of the athletes had rearfoot, increasing with the speed and fatigue. Among 52.7% of the athletes had mild external rotation and 69% had mild lateral tilt. No relationship was found between athletes, either with or without lesions in the foot strike patterns on the three race protocols. The contact time was higher in rearfoot regarding the rest of the types of foot strike patterns in comfortable running and competition pace. Conclusion: The findings of this study may be relevant for the design of athletic shoes and contribute information on the relationship between the foot strike patterns of the veteran runner and the risk of injuries. Keywords: athletes, running, athletic injuries. DINÂMICA DO APOIO EM CORREDORES VETERANOS COM RELAÇÃO AO INCREMENTO DA VELOCIDADE E DA FADIGA FOOT STRIKE PATTERNS IN VETERAN RUNNERS WITH RESPECT TO THE INCREASE OF THE SPEED AND FATIGUE Artigo Original DINÁMICA DE APOYO EN CORREDORES VETERANOS CON RELACIÓN AL INCREMENTO DE LA VELOCIDAD Y DE LA FATIGA DINÁMICA DE APOYO EN CORREDORES VETERANOS CON RELACIÓN AL INCREMENTO DE LA VELOCIDAD Y DE LA FATIGA Jesús Salas Sánchez1 (Educador Físico) Pedro Ángel Latorre Román2 (Educador Físico) María Aparecida Santos e Campos2 (Educadora Física) Víctor Manuel Soto Hermoso3 (Educador Físico) Correspondência: Palavras-chave: atletas, corrida, lesões em atletas. RESUMO Jesús Salas Sánchez1 (Educador Físico) Pedro Ángel Latorre Román2 (Educador Físico) María Aparecida Santos e Campos2 (Educadora Física) Víctor Manuel Soto Hermoso3 (Educador Físico) 1. Universidad Autónoma de Chile. 2. Universidad de Jaén, Espanha. 3. Universidad de Granada, Espanha. Correspondência: Maria Aparecida Santos e Campos. Calle Fuente Del Alamillo 3 puerta 2 1º D, Jaén, 23006, España. salas644@hotmail.com Introdução: Nas últimas décadas, o número de participantes nas corridas populares de resistência tem aumentado, especialmente, dos atletas da categoria de veteranos (acima de 35 anos), assim como o número de provas organizadas. Objetivo: Analisar a dinâmica de apoio na corrida do atleta veterano com relação ao incremento da velocidade e da fadiga. Métodos: Participaram 43 atletas, 40 homens (média de idade: 41,3 ± 1,10 anos) e 3 mulheres (média de idade 44,3 ± 3,17 anos). Depois do aquecimento, foram realizados três protocolos de corrida: confortável, ritmo de competição e ritmo de competição até o incremento dos valores da escala de Borg, escolhidos livremente por cada atleta. A dinâmica do apoio (superfície de contato, rotação lateral e inclinação do pé) e sua sequência temporal foram analisadas em vídeo de alta velocidade. Resultados: Do total, 84,5% dos atletas apresentaram apoio atrasado, com incre- mento correspondente ao aumento da velocidade e da fadiga na corrida. Dos atletas, 52,7% apresentaram rotação externa leve e 69% apresentaram inclinação lateral leve. Não foi encontrada relação entre atletas lesionados ou não nem entre a dinâmica de apoio nos três tipos de corrida. O tempo de contato na fase de amortecimento foi superior no apoio atrasado com relação ao resto dos tipos de passadas na corrida confortável e no ritmo de competição. Conclusão: As descobertas deste estudo podem ser importantes para o desenho de tênis esportivos e para aportar informações sobre a relação entre o padrão da passada do corredor veterano e o risco de lesões. Correspondência: Maria Aparecida Santos e Campos. Calle Fuente Del Alamillo 3 puerta 2 1º D, Jaén, 23006, España. salas644@hotmail.com MATERIAIS E MÉTODOS Participaram do estudo 43 atletas, sendo 40 homens (idade = 41,3 ± 1,10 anos) e 3 mulheres (idade = 44,3±3,17 anos). Esta é uma amostra não probabilística de conveniência dos clubes esportivos pertencentes às províncias de Jaén e Granada (Espanha). Depois de receber detalhes do estudo, cada participante assinou um consenti- mento declarando que cumpre com as normas éticas da Declaração de Helsinque da Associação Médica Mundial. Entre os critérios de inclusão levados em conta estavam: pertencer à categoria veterana, de acordo com os critérios da Real Federação Espanhola de Atletismo (a partir dos 35 anos); ter 2 ou mais anos de experiência na prática de corrida de fundo; não ter nenhuma doença cognitiva nem incapaci- dade intelectual; não haver padecido de nenhuma lesão nos últimos 3 meses; estar em treinamento. Os atletas que haviam se afastado da prática esportiva por um período superior a 7 dias foram excluídos. Aproximadamente 69% das lesões crônicas nos corredores estão relacionadas à alta quilometragem do treinamento, ao aumento rápido da distância, da intensidade, a treinamentos em superfícies irregulares, à instabilidade do terreno de treino3, e a certas variações anatômicas como: pronação e inversão do pé, que relacionam-se com lesões es- pecíficas. No entanto, as modificações nos calçados esportivos, não reduziram as incidências destas lesões nos corredores4. Deste modo, apesar da alta tecnologia aplicada aos tênis esportivos para corredores, eles continuam sofrendo ferimentos. Neste sentido, Fields et al.5 enfatizam que nas carreiras de corredores de resistência a taxa de lesões se mantém relativamente constante. Durante os úl- timos 30 anos, tanto os corredores daquela época quanto os atuais, continuam apresentando lesões. Para realizar a análise de observação e tempo de apoio, foram em- pregadas quatro câmaras Dalsa Genier HM 1024. Estas câmaras incorpo- ram um sensor CMOS desenhado por DALSA que permite trabalhar a 100 imagens por segundo com uma resolução de 1024 x 768 píxels. A dinâmica de apoio foi analisada em fotogramétrica 2D com o programa virtual-Dub-1.9.11 e videoSpeed v1.37 (Universidad de Granada). As cor- ridas foram realizadas em uma esteira rolante BH Prisma M80. A altura (cm) e o peso (kg) foram medidos por um estadiômetro e uma balança Seca 222 e 634 respectivamente, Hamburgo, Alemanha. A frequência cardíaca (Fc) foi analisada mediante um pulsômetro Garmin Forerunner 305. RESUMEN Introducción: En las últimas décadas, el número de participantes en las carreras populares de resistencia ha aumentado, especialmente, de los atletas de la categoría de veteranos (más de 35 años), así como el número de pruebas organizadas. Objetivo: Analizar la dinámica de apoyo en la carrera del atleta veterano con relación al incremento de la velocidad y de la fatiga. Métodos: Participaron 43 atletas, 40 hombres (promedio de edad: 41,3 ± 1,10 años) y 3 mujeres (promedio de edad 44,3 ± 3,17 años). Después del calentamiento, fueron realizados tres protocolos de carrera: confortable, ritmo de competición y ritmo de competición hasta el incremento de los valores de la escala de Borg, escogidos libremente por cada atleta. La dinámica de apoyo (superficie de contacto, rotación lateral e inclinación del pie) y su secuencia temporal fueron analizados en video de alta velocidad. Resultados: Del total, 84,5% de los atletas presentaron apoyo atrasado, con incremento correspondiente al aumento de la velocidad y de la fatiga en la carrera. De los atletas, 52,7% presentaron rotación externa leve y 69% presentaron inclinación lateral leve. No fue encontrada relación entre atletas lesionados o no ni entre la dinámica de apoyo en los tres tipos 315 Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 de carrera. El tiempo de contacto en la fase de amortiguación fue superior en el apoyo atrasado con relación al resto de los tipos de pasos en la carrera confortable y en el ritmo de competición. Conclusión: Los descubrimientos de este estudio pueden ser importantes para el diseño de zapatos deportivos y para aportar informaciones sobre la relación entre el estándar del paso del corredor veterano y el riesgo de lesiones. Palabras clave: atletas, carrera, lesiones en atletas. Artigo recebido em 04/04/2013, aprovado em 15/01/2014 Artigo recebido em 04/04/2013, aprovado em 15/01/2014 DOI: http://dx.doi.org/10.1590/1517-86922014200401693 INTRODUÇÃO impulso de freado e de impulso, além de uma maior pré-ativação dos músculos do tríceps sural do que quando se está calçado14. De qualquer forma, faltam evidências científicas que afirmem que as taxas de lesões e de rendimento na corrida se incrementam quando a corrida é praticada descalça ou com calçados minimalistas3. Por outro lado , diversos estudos documentaram a marcha assimétrica dos corredores, sobre tudo, no que se relaciona ao risco de lesões7,15. Nas últimas décadas o número de participantes nas corridas po- pulares de resistência tem aumentado, especialmente na categoria de atletas veteranos (acima de 35 anos), bem comoo número de provas organizadas. A corrida Internacional de San Antón, em Jaén, Espanha, teve, durante aedição de 2011, um total de 3.900 atletas corredores, 50,5% eram veteranos, sendo 43,15% homens e 6,89% mulheres. O objetivo deste estudo foi analisar a dinâmica da SC e sequência temporal na corrida de atletas veteranos e sua evolução em relação a velocidade e a fadiga. No atleta veterano, à medida que a idade avança, uma das con- sequências mais indesejadas é o risco de sofrer uma lesão musculo- esquelética não intencional do exercício físico. Os atletas, jovens ou veteranos, parecem ter um maior risco ao aparecimento de tais lesões do que o restante da população1, as quais podem ser determinadas por fatores internos (intrínsecos) relacionado com o atleta, ou externos (extrínsecos) relacionados com o ambiente2. MATERIAIS E MÉTODOS A percepção do esforço (RPE) foi avaliada com a escala de Borg (1982)16 na qual se estabelece subíndice 16, itens de 6 a 20 graus de menor à máxima intensidade. continuam apresentando lesões. Esse quadro pode estar relacionado à dinâmica do contato do pé com o solo. Geralmente, são definidas três categorias de superfícies de contato (SC) mais frequente nos corredores de fundo4,6-8: (1) retropé no qual o contato inicial se realiza no calcanhar ou parte posterior do pé; (2) médio, no qual o calcanhar e a planta do pé fazem o contato quase que simultaneamente (3) adiantado, em que o contato inicial é feito na parte da frente do pé. O padrão da SC depende, de certa forma, das seguintes variáveis: velocidade (V), superfície, calçado e fadiga6. A corri- da com apoio do metatarso parece ser uma característica da evolução humana6. Nas altas velocidades e entre os corredores descalços ou com calçados minimalistas, o apoio adiantado parece ser o mais comum4,6,9. Aproximadamente 80% dos corredores calçados apresentam uma SC atrasada10. Lieberman et al.9 observam que os corredores descalços com apoio adiantado apresentam uma força de impacto três vezes menor do que os corredores calçados com apoio atrasado. É plausível que os corredores com apoio atrasado, com o aumento das cargas de impacto, possam ter um maior risco de desenvolver lesões. A modificação de apoio atrasado para adiantado pode reduzir a dor patelo-femural11, e a dor associada à síndrome compartimental crônica do esforço12 com a redução de impactos é provável que provoque uma diminuição global do risco de lesões e neste sentido, seria interessante a adoção de uma SC plana o adiantada13. Aliás, correr descalço com apoio adiantado reduz o tempo de apoio e o tempo de voo, há um menor pico de força, maior Esse quadro pode estar relacionado à dinâmica do contato do pé com o solo. Geralmente, são definidas três categorias de superfícies de contato (SC) mais frequente nos corredores de fundo4,6-8: (1) retropé no qual o contato inicial se realiza no calcanhar ou parte posterior do pé; (2) médio, no qual o calcanhar e a planta do pé fazem o contato quase que simultaneamente (3) adiantado, em que o contato inicial é feito na parte da frente do pé. O padrão da SC depende, de certa forma, das seguintes variáveis: velocidade (V), superfície, calçado e fadiga6. Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 Palabras clave: atletas, carrera, lesiones en atletas. MATERIAIS E MÉTODOS A corri- da com apoio do metatarso parece ser uma característica da evolução humana6. Nas altas velocidades e entre os corredores descalços ou com calçados minimalistas, o apoio adiantado parece ser o mais comum4,6,9. Aproximadamente 80% dos corredores calçados apresentam uma SC atrasada10. Lieberman et al.9 observam que os corredores descalços com apoio adiantado apresentam uma força de impacto três vezes menor do que os corredores calçados com apoio atrasado. É plausível que os corredores com apoio atrasado, com o aumento das cargas de impacto, possam ter um maior risco de desenvolver lesões. A modificação de apoio atrasado para adiantado pode reduzir a dor patelo-femural11, e a dor associada à síndrome compartimental crônica do esforço12 com a redução de impactos é provável que provoque uma diminuição global do risco de lesões e neste sentido, seria interessante a adoção de uma SC plana o adiantada13. Aliás, correr descalço com apoio adiantado reduz o tempo de apoio e o tempo de voo, há um menor pico de força, maior A fim de facilitar a observação sistemática, alguns instrumentos foram concebidos ad hoc com observação representada por imagens, exemplos da dinâmica de apoio dos corredores em relação com a SC, a rotação vertical (ROT) e a rotação ou inclinação anteroposterior (BA) do pé. Foi diferenciada a SC em atrasada, plana e adiantada. A ROT em rotação interna, externa muito intensa, externa média, externa leve e alinhada. A BA se categorizou em inclinação medial, lateral intensa, lateral leve e centrada (figura 1). Também foram analisados os distintos tempos nas fases de apoio e voo. Tempo de amortização: desde o início do apoio até a eleva- ção do tornozelo; tempo de impulso: desde a fase de elevação do 316 Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 cícios extenuantes 72 horas antes do protocolo de análise da corrida. Inicialmente, foram registrados o peso, altura e as variáveis sociodemo- gráficas. Posteriormente, os atletas realizaram um aquecimento de 5 minutos de alongamento e técnicas de corrida. MATERIAIS E MÉTODOS Logo em uma esteira rolante foi solicitado aos atletas que realizassem 3 protocolos de cor- rida eleitos livremente, ao qual se ajustou a velocidade da esteira de acordo com cada atleta: (1)1.5 minutos de corrida confortável (CC); (2) 2.5 minutos de corrida em ritmo de competição (CRC); (3) Foi mantida a velocidade de corrida em ritmo competição até que se produziu um incremento na fadiga em duas unidades RPE (CRCFA - corrida em ritmo competição com fadiga acumulada). A execução das corridas foi gravada com quatro câmaras dispostas em forma de círculo. Para o calculo da confiabilidade intraobservador, dois meses depois (pós) da primeira observação (pré), as observações com uma amostra aleatória de 25 sujeitos foram realizadas novamente. calcanhar até o final do apoio, e tempo de voo: tempo sem contato dos pés com o solo. Em um questionário feito ad hoc, foram utilizados os dados sociode- mográficos nos quais houve a inclusão do número de lesões padecidas nos últimos três anos (tabela 1). Os atletas foram citados individualmente no laboratório de ergo- nomia da empresa “Ergonomía Solei” e foram instruídos a evitar exer- Figura 1. Dinâmica do apoio em corredores. Rotação Interna Rotação externa muito intensa Rotação externa média Rotação externa leve Perfeitamente alinhado “Dinâmica da rotação” do pé (sobre eixo vertical) “Superfície de contato” no impacto inicial Adiantado Atrasado Plano Centrada “Inclinação” “ Inclinação lateral” da parte posterior do pé no impacto inicial Lateral muito intensa Lateral leve Medial “Superfície de contato” no impacto inicial Rotação Interna Rotação externa muito intensa Rotação externa média Rotação externa leve Perfeitamente alinhado “Dinâmica da rotação” do pé (sobre eixo vertical) Adiantado Atrasado Plano Análise estatística Os dados deste estudo foram encontrados usando o programa SPSS. V.19.0 para Windows (SPSS Inc, Chicago, EUA). O nível de sig- nificância foi estabelecido em p <0,05. Os resultados são mostrados nas estatísticas descritivas de frequência, porcentagens, médias e desvios-padrão. Foi utilizado o teste de Shapiro-Wilk para verificar a distribuição normal dos dados contínuos. A análise temporal foi realizada por meio de análise de variância (ANOVA). A comparação de dados entre as repetições foi realizada mediante a análise ANOVA de medidas repetidas para dados contínuos,onão paramétrico de Friedman e Wilcoxon contraste para esses dados ordinais, e Q de Cochran para variáveis dicotômicasPara a análise da confiabilidade intraobserbador foi utilizado índice Kappa de Cohen. “Dinâmica da rotação” do pé (sobre eixo vertical) “Dinâmica da rotação” do pé (sobre eixo vertical) Figura 1. Dinâmica do apoio em corredores. Centrada “Inclinação” “ Inclinação lateral” da parte posterior do pé no impacto inicial Lateral muito intensa Lateral leve Medial “Inclinação” “ Inclinação lateral” da parte posterior do pé no impacto inicial “Inclinação” “ Inclinação lateral” da parte posterior do pé no impacto inicial Figura 1. Dinâmica do apoio em corredores. Centrada “Inclinação” “ Inclinação lateral” da parte posterior do pé no impacto inicial Lateral muito intensa Lateral leve Medial Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 RESULTADOS Em relação às variáveis sociodemográficas dos participantes, se destacou que 74,4% dos atletas não estão federados e 62,8% não tem treinador. Os anos de treinamento são 7,20±6,31. O índice de massa corporal (IMC) = 23,83±1,97 Kg/m2 e qualifica estes atletas na categoria de normopeso. Registrou-se uma média de 1,64±0,98 lesões nos últi- mos três anos, 37,2% dos atletas não haviam sofrido nenhuma lesão e 62,8% haviam sofrido lesão. Figura 1. Dinâmica do apoio em corredores. Os índices de concordância Kappa de Cohen para as observações pré e pós das categorias de observação foram: 0,685 p<0,001 para a SC, 0,653 (p<0,001) para ROT e 0,754 ( p<0,001) para BA. Tabela 1. Variáveis sociodemográficas. Estudos (%) Estudos Primários 18,6 Estudos Secundários 37,2 Estudios Universitarios 44,2 Estão federados (%) Sim 25,6 Não 74,4 Tem trenador (%) Sim 37,2 Não 62,8 Sessões de entrenamiento semanais (%) Hasta 4 sessões semanais 48,8 Mais de 4 sessões semanales 51,2 Anos de treinamento Média (DT) 7,20 (6,31) IMC(kg/m²) Média (DT) 23,83 (1,97) Lesiones Média (DT) 1,64 (0,98) Tabela 1. Variáveis sociodemográficas. Há um aumento significativo (p<0,001) do RPE e a Fc a medida que se incrementa a velocidade e a duração da corrida (maior fadiga) (tabela 2). Há um aumento significativo (p<0,001) do RPE e a Fc a medida que se incrementa a velocidade e a duração da corrida (maior fadiga) (tabela 2). Há um aumento significativo (p<0,001) do RPE e a Fc a medida que se incrementa a velocidade e a duração da corrida (maior fadiga) (tabela 2). Em relação a SC existe uma tendência (p=0,027) ao apoio atra- sado com o incremento da velocidade e a fadiga (entre CC y CRCFA) (figura 2 e tabela 3). Não foram observadas alterações significativas na ROT Y BA entre os três tipos de corridas (figura 3 e 4). No total de todas as corridas, 84,5% dos atletas deste estudo apresentaram apoios atra- sados, 15,5% realizam apoios planos e adiantados e, 52,7% dos atletas apresentam um apoio com rotação externa leve e 69% inclinação lateral leve nas três corridas. Em relação a assimetria dos pés, podemos comprovar que existe uma tendência ao incremento da assimetria em relação a ROT (p=0,084) com o acréscimo da velocidade e fadiga (figura 5). Não foram apresentadas diferenças significativas (p<0,05) entre atletas lesionados ou não lesionados com relação à dinâmica de apoio (SC,ROT,BA) e assimetria de pés em três tipos de corridas. A tabela 4 mostra a sequência temporal dos três tipos de corridas executados em relação SC. Se destaca um incremento do tempo de amortização (TA) na CC e CRC conforme o apoio se torna atrasado. 317 Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 Tabela 2. Velocidade, RPE e Fc dos tres tipos de corridas. CC CRC CRCFA p Velocidade (km/h) 10,23 (0,95) 13,73 (1,34) 13,73 (1,34) <0,001 RPE (6-20) 10,55 (1,00) 13,69 (0,70) 16,11 (0,79) <0,001 Fc (pulxmin) 139,58 (12,94) 159,32 (10,87) 167,13 (10,40) <0,001 CC (corrida confortável). CRC (corrida ritmo de competição). CRCFA (carrera ritmo competição com fadiga acumulada). Figura 2. Superfície de contato em relação aos três tipos de corridas. 120,0 100,0 80,0 60,0 40,0 20,0 0 Porcentagem (%) CC CRC CRCFA (SC= superfície de contato) SC Atrasado SC Plano SC Adiantado Figura 2. Superfície de contato em relação aos três tipos de corridas. 120,0 100,0 80,0 60,0 40,0 20,0 0 Porcentagem (%) CC CRC CRCFA (SC= superfície de contato) SC Atrasado SC Plano SC Adiantado Tabela 3. Evolução dos parámetros observacionais nos tres tipos de corridas. CC CRC CRCFA p SC 1,26 (0,44) 1,18 (0,45) 1,10 (0,25) 0,027 ROT 3,76 (0,56) 3,81 (0,65) 3,82 (0,67) NS BA 3,23 (0,52) 3,27 (0,52) 3,23 (0,48) NS NS: não significativo. CC (corrida confortável). CRC (corrida ritmo de competição). Há um aumento significativo (p<0,001) do RPE e a Fc a medida que se incrementa a velocidade e a duração da corrida (maior fadiga) (tabela 2). CRCFA (carrera ritmo com- petição com fadiga acumulada). Figura 3. Rotação relacionada aos três tipos de corridas. 60 50 40 30 20 10 0 Porcentagem (%) CC CRC CRCFA (RE= rotação externa) Figura 2. Superfície de contato em relação aos três tipos de corridas. CC CRC CRCFA (SC= superfície de contato) Figura 4. Inclinação relacionada aos três tipos de corridas. (BL= inclinação. AC= apoio centrado) 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 0 Porcentagem (%) CC CRC CRCFA BL Muito intensa BL Leve AC Alinhado RE Leve RE Media RE Muito intensa (SC= superfície de contato) Figura 2. Superfície de contato em relação aos três tipos de corridas. Figura 3. Rotação relacionada aos três tipos de corridas. 60 50 40 30 20 10 0 Porcentagem (%) CC CRC CRCFA (RE= rotação externa) Alinhado RE Leve RE Media RE Muito intensa Figura 3. Rotação relacionada aos três tipos de corridas. 60 50 40 30 20 10 0 Porcentagem (%) CC CRC CRCFA (RE= rotação externa) Alinhado RE Leve RE Media RE Muito intensa NS: não significativo. CC (corrida confortável). CRC (corrida ritmo de competição). CRCFA (carrera ritmo com- petição com fadiga acumulada). DISCUSSÃO Os tênis para corridas podem debilitar os músculos do pé e a força do arco através da dependência de suportes este e solas rígidas e, que esta mesma debilidade pode contribuir para a pronação excessiva e a facits plantar. Enquanto isso, corredores que correm descalços ou com calçados minimalistas evitam aterrissagens posteriores do pé e os impactos fortalecem os músculos do pé e com isso podem evitar lesões9. Neste sentido Daoud et al.6, indicam que os corredores que ha- bitualmente apoiam atrasados tem aproximadamente o dobro da taxa de lesões por esforços repetitivos em relação aos indivíduos que apoiam com a parte dianteira do pé. E a causa destas diferenças se deve a ausência de um pico forte de impacto na força de reação solo durante o apoio do antepé. Os modernos calçados de corrida podem ser perigosos porque pro- piciam uma aterrissagem de calcanhar no solo que produz um impacto muito maior que a aterrissagem com a parte dianteira dos pés. Além disso, a magnitude da força vertical máxima durante o período de impacto é aproximadamente três vezes menor nos corredores descalços que ater- rissam de antepé do que nos corredores que habitualmente aterrissam de calcanhar9. Os tênis para corridas podem debilitar os músculos do pé e a força do arco através da dependência de suportes este e solas rígidas e, que esta mesma debilidade pode contribuir para a pronação excessiva e a facits plantar. Enquanto isso, corredores que correm descalços ou com calçados minimalistas evitam aterrissagens posteriores do pé e os impactos fortalecem os músculos do pé e com isso podem evitar lesões9. Figura 4. Inclinação relacionada aos três tipos de corridas. Figura 4. Inclinação relacionada aos três tipos de corridas. Figura 5. Assimetria da dinâmica de apoio em relação aos três tipos de corridas. 20 18 16 14 12 10 8 6 4 2 0 Frequencia SC ROT BA CC CRC CRCFA Neste sentido Daoud et al.6, indicam que os corredores que ha- bitualmente apoiam atrasados tem aproximadamente o dobro da taxa de lesões por esforços repetitivos em relação aos indivíduos que apoiam com a parte dianteira do pé. E a causa destas diferenças se deve a ausência de um pico forte de impacto na força de reação solo durante o apoio do antepé. Figura 5. Assimetria da dinâmica de apoio em relação aos três tipos de corridas. Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 DISCUSSÃO Os resultados deste estudo indicam que o apoio atrasado, com rotação externa leve e ligeira inclinação lateral leve, é a execução di- nâmica mais comum do atleta veterano, independentemente da raça. Apresentado uma tendência de SC atrasada e a assimetria dos pés em relação ao ROT a medida que aumenta a velocidade crescente e a fadiga da corrida. A tendência da SC muito atrasada com a fadiga está em consonân- cia com o estudo de Lason et al.7 entretanto, este aumento associado à velocidade de corrida não corresponde aos dados de Hasegawa et al.10 que informam um aumento de apoio plano nos atletas mais velozes. Figura 3. Rotação relacionada aos três tipos de corridas. Dos atletas deste estudo, 84,5% tem apoio atrasado na soma das três corridas, 15,5% tem apoios planos e adiantados. Dados seme- lhantes aos de Hamil17 que projeta valores de 75% a 80% de apoios atrasados e de outros estudos7,10, que destacam que este tipo de SC é a mais habitual em atletas de corrida de fundo recreativas aportando porcentagem de apoio atrasado entre 93% - 74.9% e 87,8 % respectiva- mente. Ainda que inferiores aos dados de Daoud et al.6 que informada presença deste tipo de apoio em 59% dos corredores. Alguns estudos5 assinalam que os atletas universitários que possuem apoio atrasado apresentam o dobro de lesões por esforço repetitivo em comparação aos que se apoiam no metatarso. Figura 4. Inclinação relacionada aos três tipos de corridas. (BL= inclinação. AC= apoio centrado) 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 0 Porcentagem (%) CC CRC CRCFA BL Muito intensa BL Leve AC Figura 4. Inclinação relacionada aos três tipos de corridas. (BL= inclinação. AC= apoio centrado) 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 0 Porcentagem (%) CC CRC CRCFA BL Muito intensa BL Leve AC q p Os modernos calçados de corrida podem ser perigosos porque pro- piciam uma aterrissagem de calcanhar no solo que produz um impacto muito maior que a aterrissagem com a parte dianteira dos pés. Além disso, a magnitude da força vertical máxima durante o período de impacto é aproximadamente três vezes menor nos corredores descalços que ater- rissam de antepé do que nos corredores que habitualmente aterrissam de calcanhar9. CONCLUSÃO O estudo da dinâmica de apoio na corrida do atleta veterano pode ser um aspecto interessante no desenho de calçados especializados para esta população. Ainda que futuras investigações com amostras amplas devessem estabelecer de maneira precisa a relação entre a dinâmica de apoio na corrida de resistência do atleta veterano e a incidência de lesões, assim como determinar se o emprego do calçado esportivo atual pode ser um elemento de risco. Por outro lado, Hasegawa et al.10 indicam que o tempo de contato é maior nos apoios atrasados que nos planos e adiantados. Neste sentido este estudo destaca que o tempo de contato na fase de amortização é superior no apoio atrasado em relação ao resto de tipos de pisada na CC e CRC. Em recente revisão Lorenz e Pontillo18 observaram que há dados suficientes para afirmar que os corredores devem utilizar um padrão de SC adiantada no lugar de um atrasada para reduzir as forças de reação do solo o tempo de contato com a superfície e a duração do passo. Se há um efeito positivo ou negativo sobre a lesão ainda não se determinou. Sem lugar a dúvidas se necessita mais investigação antes de tirar conclusões definitivas a respeito. Todos os autores declararam não haver qualquer potencial conflito de interesses referente a este artigo. REFERÊNCIAS 10. Hasegawa H, Yamauchi T, Kraemer WJ. Foot strike patterns of runners at the 15-km point during an elite-level half marathon. J Strength Cond Res. 2007;21(3):888-93. 1. Alvero JR. Lesiones deportivas en competición en atletas veteranos. Apunts Medicina de l’esport. 2008;43:113-7. 1. Alvero JR. Lesiones deportivas en competición en atletas veteranos. Apunts Medicina de l’esport. 2008;43:113-7. 2. Osorio JA, Clavijo MP, Arango E, Patiño S, Gallego IC. Lesiones deportivas. Latreia. 2007;20:167-77. 2. Osorio JA, Clavijo MP, Arango E, Patiño S, Gallego IC. Lesiones deportivas. Latreia. 2007;20:167-77. 11. Cheung RT, Davis IS. Landing pattern modification to improve patellofemoral pain in runners: a case series. J Orthop Sports Phys Ther. 2011;41(12):914-9. 3. Lohman EB, Balan Sackiriyas KS, Swen RW. A comparison of the spatiotemporal parameters, kine- matics and biomechanics between shod, unshod and minimally supported running as compared to walking. Phys Ther Sport. 2011; 12(4):151-63. 3. Lohman EB, Balan Sackiriyas KS, Swen RW. A comparison of the spatiotemporal parameters, kine- matics and biomechanics between shod, unshod and minimally supported running as compared to walking. Phys Ther Sport. 2011; 12(4):151-63. 12. Diebal AR, Gregory R, Alitz C, Gerber JP. Forefoot running improves pain and disability associated with chronic exertional compartment syndrome. Am J Sports Med. 2012;40(5):1060-7. 4. Rixe JA, Gallo RA, Silvis ML. The Barefoot Debate: Can Minimalist Shoes Reduce Running-Related Injuries?. Curr Sports Med Rep. 2012;11(3):160-5. 13. Davis I, Bowser B, Mullineau D. Do Impacts Cause Running Injuries? A Prospective Investigation. Pre- sented at the American Society of Biomechanics Mtg., Providence, RI, August, 2010. 5. Fields KB, Sykes JC, Walker KM, Jackson JC. Prevention of running injuries. Curr Sports Med Rep. 2010;9(3):176-82. 14. Divert C, Mornieux G, Baur H, Mayer F, Belli A. Mechanical comparison of barefoot and shod running. Int J Sports Med. 2005;26:593-8. 6. Daoud AI, Geissler GJ, Wang F, Saretsky J, Daoud YA, Lieberman DE. Foot strike and injury rates in endurance runners: a retrospective study. Med Sci Sports Exerc. 2012;44(7):1325-34. 15. Zifchock RA, Davis I, Hamill J. Kinetic asymmetry in female runners with and without retrospective tibial stress fractures. J Biomech. 2006;39:2792-7. 7. Larson P, Higgins E, Kaminski J, Decker T, Preble J, Lyons D, et al. Foot strike patterns of recreational and sub-elite runners in a long-distance road race. J Sports Sci. 2011;29(15):1665-73. and sub-elite runners in a long-distance road race. J Sports Sci. 2011;29(15):1665 16. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sport Exer. DISCUSSÃO O 52,7% dos atletas nas três corridas apresentam um apoio com 318 Tabela 4. Sequência de tempo nos três tipos de corrida em relação ao SC. CC CRC CRCFA TA TI TV TA TI TV TA TI TV SC retrasado (s) 0,167 (0,01) 0,122 (0,01) 0,072 (0,03) 0,142 (0,02) 0,105 (0,01) 0,090 (0,02) 0,142 (0,02) 0,106 (0,01) 0,088 (0,02) SC plano (s) 0,140 (0,01) 0,130 (0,01) 0,076 (0,03) 0,126 (0,01) 0,107 (0,01) 0,103 (0,03) 0,127 (0,00) 0,105 (0,00) 0,122 (0,02) SC adiantado (s) - - - 0,107 0,102 0,110 - - - p <0,001 NS NS 0,043 NS NS NS NS NS NS: não significativo. TA: tempo de amortização. TI: tempo de impulso. TV: tempo de voo. SC= superficie de contato. CC (corrida confortável). CRC (corrida ritmo de competição). CRCFA (carrera ritmo competição com fadiga acumulada).. rotação externa leve e 69% uma inclinação lateral leve, não sendo fatores graves de desalinhamento. Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014 REFERÊNCIAS 1982; 14: 377-381. 8. Latorre PA. Análisis de los parámetros biomecánicos de la economía de carrera en atletas de fondo a velocidades competitivas. Tesis Doctoral.Jaén: Universidad de Granada. Servicio de publicaciones de la Universidad de Jaén; 2003. 17. Hamill J. Rearfoot and forefoot footfall patterns: implications for barefoot running. J Foot Ankle Res. 2012;5(Suppl 1):K1. 9. Lieberman DE, Vankadesan M, Werbel WA, Daoud AI, D’Andrea S, Davis IS, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners. Nature. 2010;463:531-5. 18. Lorenz DS, Pontillo M. Is there evidence to support a forefoot strike pattern in barefoot runners? A review. Sports Health. 2012;4:480. 9. Lieberman DE, Vankadesan M, Werbel WA, Daoud AI, D’Andrea S, Davis IS, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners. Nature. 2010;463:531-5. 319 Rev Bras Med Esporte – Vol. 20, No 4 – Jul/Ago, 2014
https://openalex.org/W3119491455	https://www.researchsquare.com/article/rs-144842/latest.pdf	English	null	Tropical drivers of interannual vegetation variability in eastern Africa	null	2,021	cc-by	6,690	Tropical Drivers of Interannual Vegetation Variability in Eastern Africa In-Won Kim (  iwkimi@pusan.ac.kr ) Center for Climate Physics, Institute for Basic Science Malte Stuecker University of Hawai฀i Sea Grant Axel Timmermann Center for Climate Physics, Institute for Basic Science Jong-Seong Kug Pohang University of Science and Technology So-Won Park Pohang University of Science and Technology Jin-Soo Kim University of Zurich Research Article Keywords: tropical Paci¦c, sea surface temperature (SST), Tropical Indian Ocean (TIO), Indian Ocean warming Posted Date: January 19th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-144842/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Research Article Posted Date: January 19th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-144842/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Tropical drivers of interannual vegetation 1 variability in eastern Africa 2 3 In-Won Kim1,2, Malte F. Stuecker3, Axel Timmermann1,2, Jong-Seong Kug4, 4 So-Won Park4, and Jin-Soo Kim5 5 6 1 Center for Climate Physics, Institute for Basic Science, Busan, Republic of Korea 7 2 Pusan National University, Busan, Republic of Korea 8 3 Department of Oceanography and International Pacific Research Center, School of 9 Ocean and Earth Science and Technology, University of Hawaiʻi at Mānoa, Honolulu, HI, 10 USA 11 4 Division of Environmental Science and Engineering, Pohang University of Science and 12 Technology, Pohang, Republic of Korea 13 5 Department of Evolutionary Biology and Environmental Studies, University of Zurich, 14 Zurich, Switzerland 15 Corresponding author: In-Won Kim (iwkimi@pusan.ac.kr) 16 17 Tropical drivers of interannual vegetation variability in eastern Africa 1 1 ABSTRACT 18 Mechanisms by which tropical Pacific and Indian Ocean sea surface temperatures influence vegetation in 19 Eastern Africa and which role drought-induced fires play have not been fully explored. Here, we use a suite of 20 idealized Earth system model simulations to elucidate the governing processes for eastern African interannual 21 vegetation changes. Our analysis focuses on Tanzania. In the absence of ENSO-induced sea surface temperature 22 (SST) anomalies in the Tropical Indian Ocean (TIO), El Niño causes during its peak phase negative precipitation 23 anomalies over Tanzania due to a weakening of the tropical-wide Walker circulation and anomalous descending 24 motion over the Indian Ocean and southeastern Africa. Resulting drought conditions increase the occurrence of 25 wildfires, which leads to a marked decrease in vegetation cover. Subsequent wetter La Niña conditions in boreal 26 winter reverse the trend in vegetation, causing a gradual 1-year-long recovery phase. The 2-year-long vegetation 27 response in Tanzania can be explained as a double-integration of the local rainfall anomalies, which originate 28 from the seasonally-modulated ENSO Pacific-SST forcing (Combination mode). In the presence of interannual 29 TIO SST forcing, the southeast African precipitation and vegetation responses to ENSO are muted due to Indian 30 Ocean warming and the resulting anomalous upward motion in the atmosphere. 31 ABSTRACT Mechanisms by which tropical Pacific and Indian Ocean sea surface temperatures influence vegetation in 19 Eastern Africa and which role drought-induced fires play have not been fully explored. Here, we use a suite of 20 idealized Earth system model simulations to elucidate the governing processes for eastern African interannual 21 vegetation changes. Our analysis focuses on Tanzania. In the absence of ENSO-induced sea surface temperature 22 (SST) anomalies in the Tropical Indian Ocean (TIO), El Niño causes during its peak phase negative precipitation 23 anomalies over Tanzania due to a weakening of the tropical-wide Walker circulation and anomalous descending 24 motion over the Indian Ocean and southeastern Africa. Resulting drought conditions increase the occurrence of 25 wildfires, which leads to a marked decrease in vegetation cover. Subsequent wetter La Niña conditions in boreal 26 winter reverse the trend in vegetation, causing a gradual 1-year-long recovery phase. The 2-year-long vegetation 27 response in Tanzania can be explained as a double-integration of the local rainfall anomalies, which originate 28 from the seasonally-modulated ENSO Pacific-SST forcing (Combination mode). In the presence of interannual 29 TIO SST forcing, the southeast African precipitation and vegetation responses to ENSO are muted due to Indian 30 Ocean warming and the resulting anomalous upward motion in the atmosphere. 31 2 2 Introduction Natural fluctuations in Africa’s vegetation are affected by rainfall and wildfire variability1-7. Especially the Sahel, eastern Africa, and southern Africa show large interannual variations in terrestrial productivity, which can be attributed to year-to-year changes in water stress8. The El Niño-Southern Oscillation (ENSO) has been considered a primary climate driver for rainfall variability in parts of Africa. Evidence from observations shows that El Niño events can cause drought in southern Africa, and enhanced precipitation and corresponding floods in eastern Africa9. Earlier studies documented a strong relationship between ENSO and the Normalized Difference Vegetation Index (NDVI) over eastern and southern Africa10-13. In contrast, over the Sahel, the relationship between ENSO and NDVI is weak14. Observations and model experiments show an asymmetric atmospheric response over Africa between El 42 Niño and La Niña15. In addition, nonlinear ENSO teleconnections over Africa might be also affected by 43 ENSO-induced asymmetric sea surface temperature (SST) responses over the Atlantic and Indian Oceans. 44 SST variability over the south Atlantic Ocean influences rainfall over the Sahel in the opposite sense of 45 ENSO10. Indian Ocean Dipole (IOD) events, typically accompanied by ENSO, positively correlate with 46 eastern African rainfall during the short rainy season16,17. Apart from ENSO, the vegetation response to climate 47 factors is also modulated by nonlinear land processes. Globally, a nonlinear relationship between net primary 48 production and rainfall is observed for grasslands18. Interannual vegetation changes over eastern Africa show 49 a nonlinear relationship with rainfall variability and a strong dependency on land cover type is observed2. 50 Although the aforementioned studies have demonstrated the impacts of ENSO on African vegetation based on observations, we still lack a deeper understanding of how interannual SST changes in the Indian and Pacific Ocean influence vegetation anomalies and which role wildfires play. In this study, we investigate the vegetation response over sub-Saharan Africa to ENSO through a series of model experiments and compare them to the observations. The Walker circulation response to ENSO However, the Periodic and the Pacific experiments show a 10-20 % increase in burned 105 area over Tanzania in September-October-November in year 1 after ENSO event peak time [SON(1)] (Fig. 106 3a-d), whereas the observations and the Tropics experiment show statistically insignificant differences (Fig. 107 3e-h). 108 109 The response of rainfall, vegetation, and wildfire over Africa to ENSO Here, to focus on the symmetric (i.e., linear) response to El Niño and La Niña events, we show El Niño minus La Niña composites. Observed composite differences between El Niño and La Niña events display pronounced positive precipitation anomalies over the Horn of Africa and negative anomalies over Southern Africa in D(0)JF(1) (Fig. 2j). The three experiments (Periodic, Pacific, and Tropics) reproduce these anomalies reasonably well (Fig. 2a, d, g). However, a small positive precipitation anomaly simulated by the Tropics experiment in the northeastern part of South Africa is not captured in the observations. Interestingly, the Periodic and the Pacific experiments exhibit a 50 % D(0)JF(1) rainfall reduction over 84 Tanzania for the El Niño minus La Niña composite (Fig. 2a, d) and an accompanying negative Net Primary 85 Production (NPP) anomaly during January-February-March of the decaying ENSO year [JFM(1)] (Fig. 2b, e). 86 In contrast, the observations and the more realistic Tropics experiment show only very weak rainfall, NPP, 87 and LAI anomalies over Tanzania (Fig. 2g, 2h, 2i, 2j, 2l), in agreement with Latif et al.22. This suggests that 88 tropical Indian or Atlantic Ocean SST anomalies might play an important role in muting the direct Pacific 89 response over this region. We hypothesize specifically that the negligible observed rainfall response over 90 Tanzania in the observations can be attributed to a compensation between the direct Pacific effect and the El 91 Niño-related Indian Ocean warming effect on the Walker circulation (Fig. 1b, c). 92 This hypothesis is further supported by the lead-lag relationship between ENSO and LAI anomalies in the 93 Periodic and the Pacific experiments (Fig. 2k). According to this analysis ENSO is leading LAI anomalies in 94 Tanzania by about one year in these two experiments, whereas no statistically significant correlation can be 95 found in the Tropics experiment. ENSO negatively correlates with LAI over Tanzania at a maximum lag of 96 16-months (R = 0.49, p < 0.00001) in the Pacific and 18-months (R = 0.58, p < 0.00001) in the Periodic 97 experiments (Fig. 2k). In contrast, for the Tropics experiment, the correlation is not significant (R = 0.06, 98 p=0.11) (Fig. 2k). The response of rainfall, vegetation, and wildfire over Africa to ENSO Regarding the LAI response to ENSO at this 16-18 months lag (that is, in May-June-July 99 in year 2 after the ENSO event peak time: MJJ(2)), we find larger negative anomalies over Tanzania in the 100 Pacific and the Periodic experiments (Fig. 2c, f), while they are much weaker anomalies in the Tropics 101 experiment and the observations (Fig. 2i, l). Moreover, the delayed response over Tanzania to ENSO is also 102 found in wildfire activity (Fig. 3). The periodic experiment shows negative anomalies in burned area over 103 Tanzania in D(0)JF(1) and statistically insignificant differences in the Pacific and Tropics experiments, as well 104 as in the observations. However, the Periodic and the Pacific experiments show a 10-20 % increase in burned 105 area over Tanzania in September-October-November in year 1 after ENSO event peak time [SON(1)] (Fig. 106 3a-d), whereas the observations and the Tropics experiment show statistically insignificant differences (Fig. 107 3e-h). 108 Combination mode-driven rainfall response over Tanzania The Walker circulation response to ENSO To investigate the drivers of the vegetation response over Africa, we first focus on the tropical large-scale 59 atmospheric circulation and its interannual variations. The position and strength of the Walker circulation are 60 closely coupled to SST anomalies in the tropical Pacific. Both the Periodic and the Pacific experiments (SST 61 anomalies are only prescribed in the tropical Pacific) show pronounced Walker circulation changes between 62 El Niño and La Niña events with anomalous ascending motion over the eastern Pacific region (and 63 corresponding upper-level divergence) and anomalous descending motion (and corresponding upper-level 64 convergence) during the peak ENSO phase of December-January-February [D(0)JF(1)] (Fig. 1a, b). 65 Importantly, the edge of the descending motion extends to the African continent in the two experiments. In 66 contrast, the Tropics experiment shows that the center of the descending motion shifts toward the Maritime 67 Continent, inducing weaker subsidence around the Indian and Atlantic Ocean, accompanying tropical Indian 68 Ocean (TIO) warming (Fig. 1c). This large-scale circulation response is similar to what is seen for the 69 observations (Fig. 1d). The TIO warming pattern seen in Figure 1c, d is largely forced by El Niño and then is 70 prolonged for several months after the El Niño event due to the so-called capacitor effect19,20. The pattern of 71 3 3 large-scale atmospheric anomalies in the Tropics experiment (Fig. 1c) is more consistent with the observations 72 (Fig. 1d) than the Periodic and Pacific experiments (Fig. 1a, b). This suggests that TIO warming affects the 73 change of the large-scale atmospheric circulation around the African continent related to ENSO, as suggested 74 by Liu et al.21. 75 76 The response of rainfall, vegetation, and wildfire over Africa to ENSO 77 Here, to focus on the symmetric (i.e., linear) response to El Niño and La Niña events, we show El Niño 78 minus La Niña composites. Observed composite differences between El Niño and La Niña events display 79 pronounced positive precipitation anomalies over the Horn of Africa and negative anomalies over Southern 80 Africa in D(0)JF(1) (Fig. 2j). The three experiments (Periodic, Pacific, and Tropics) reproduce these anomalies 81 reasonably well (Fig. 2a, d, g). However, a small positive precipitation anomaly simulated by the Tropics 82 experiment in the northeastern part of South Africa is not captured in the observations. The Walker circulation response to ENSO 83 Interestingly, the Periodic and the Pacific experiments exhibit a 50 % D(0)JF(1) rainfall reduction over 84 Tanzania for the El Niño minus La Niña composite (Fig. 2a, d) and an accompanying negative Net Primary 85 Production (NPP) anomaly during January-February-March of the decaying ENSO year [JFM(1)] (Fig. 2b, e). 86 In contrast, the observations and the more realistic Tropics experiment show only very weak rainfall, NPP, 87 and LAI anomalies over Tanzania (Fig. 2g, 2h, 2i, 2j, 2l), in agreement with Latif et al.22. This suggests that 88 tropical Indian or Atlantic Ocean SST anomalies might play an important role in muting the direct Pacific 89 response over this region. We hypothesize specifically that the negligible observed rainfall response over 90 Tanzania in the observations can be attributed to a compensation between the direct Pacific effect and the El 91 Niño-related Indian Ocean warming effect on the Walker circulation (Fig. 1b, c). 92 This hypothesis is further supported by the lead-lag relationship between ENSO and LAI anomalies in the 93 Periodic and the Pacific experiments (Fig. 2k). According to this analysis ENSO is leading LAI anomalies in 94 Tanzania by about one year in these two experiments, whereas no statistically significant correlation can be 95 found in the Tropics experiment. ENSO negatively correlates with LAI over Tanzania at a maximum lag of 96 16-months (R = 0.49, p < 0.00001) in the Pacific and 18-months (R = 0.58, p < 0.00001) in the Periodic 97 experiments (Fig. 2k). In contrast, for the Tropics experiment, the correlation is not significant (R = 0.06, 98 p=0.11) (Fig. 2k). Regarding the LAI response to ENSO at this 16-18 months lag (that is, in May-June-July 99 in year 2 after the ENSO event peak time: MJJ(2)), we find larger negative anomalies over Tanzania in the 100 Pacific and the Periodic experiments (Fig. 2c, f), while they are much weaker anomalies in the Tropics 101 experiment and the observations (Fig. 2i, l). Moreover, the delayed response over Tanzania to ENSO is also 102 found in wildfire activity (Fig. 3). The periodic experiment shows negative anomalies in burned area over 103 Tanzania in D(0)JF(1) and statistically insignificant differences in the Pacific and Tropics experiments, as well 104 as in the observations. 𝑷∗(𝒕) = 𝛂 𝐄𝐍𝐒𝐎 (𝒕) + 𝛃 𝐄𝐍𝐒𝐎(𝒕) ∙𝐜𝐨𝐬(𝝎𝒂𝒕)(𝟏), where α and β are the regression coefficients on the ENSO and theoretical C-mode predictors, and 𝜔𝑎 the frequency of the annual cycle. One can also include a white noise precipitation forcing, but since we consider ensemble mean properties in a linear model, the noise forcing is not necessary to understand the temporal evolution. The time-series in the Periodic and Pacific experiments show that the reconstructions of precipitation anomalies over Tanzania via the C-mode equation reproduce the seasonally varying simulated rainfall response to ENSO well (Periodic: R=0.64, p < 0.00001; Pacific: R= 0.65, p < 0.00001) (Fig. 4). The simulated La Niña response is somewhat reduced as compared to the El Niño rainfall anomaly. This is reminiscent of an atmospheric nonlinearity to otherwise symmetric SST forcing. Combination mode-driven rainfall response over Tanzania 4 4 The temporal evolution of the rainfall response over Tanzania to ENSO shows a rapid transition during the peak phase of both El Niño and La Niña in both the Periodic and the Pacific experiments, but not in the Tropics experiment (Fig. 4). The rainfall response is particularly pronounced in the former during the peak phase of ENSO in D(0)JF(1), which is also the climatological wet season (Fig. 4, Fig. S2). This illustrates the tight coupling between climatological conditions and the imposed ENSO signal. To further understand the distinct atmospheric response to ENSO in the absence of TIO SST anomalies, we hypothesize that the precipitation response over Tanzania to ENSO is driven by the seasonally modulated interannual ENSO variability, which is referred to as a Combination mode (C-mode)23. According to this simple model the precipitation anomalies can be written as Role of wildfires in the vegetation response to ENSO Wildfires can play a potential role in vegetation change through climate-fire-vegetation interactions24,25. In the absence of TIO warming in the Periodic and Pacific experiments, El Niño induced drying increases the occurrence of fires, which is manifest in the prolonged positive anomalies in burned area lasting for about one year after the peak of El Niño (Fig. 4 a, b). For wet savannas in Africa, an increase in fuel moisture can lead to a decrease in the burned area7, while for dry savannas, an increase in moisture facilitates more fires26. To investigate the causal linkages between precipitation and wildfire responses to ENSO, we hypothesize that changes in burned area B, are driven by precipitation variability P. Here we choose P as the ENSO- reconstructed precipitation anomaly from equation (1). We assume in its simplest linearized form that the burned area does not depend on the available vegetation which allows us to introduce a fixed mean recovery timescale (), in which the burned area can regrow. The simplified linearized model then reads: Wildfires can play a potential role in vegetation change through climate-fire-vegetation interactions24,25. In 131 the absence of TIO warming in the Periodic and Pacific experiments, El Niño induced drying increases the 132 occurrence of fires, which is manifest in the prolonged positive anomalies in burned area lasting for about 133 one year after the peak of El Niño (Fig. 4 a, b). For wet savannas in Africa, an increase in fuel moisture can 134 lead to a decrease in the burned area7, while for dry savannas, an increase in moisture facilitates more fires26. 135 To investigate the causal linkages between precipitation and wildfire responses to ENSO, we hypothesize 136 that changes in burned area B, are driven by precipitation variability P. Here we choose P as the ENSO- 137 reconstructed precipitation anomaly from equation (1). We assume in its simplest linearized form that the 138 burned area does not depend on the available vegetation which allows us to introduce a fixed mean recovery 139 timescale (), in which the burned area can regrow. The simplified linearized model then reads: 140 𝒅𝑩(𝒕) 𝒅𝒕 = −𝝁𝟏𝑩−𝜽𝟏𝑷∗(𝒕)(𝟐). 141 Appropriate parameters values are given in Table S1. Role of wildfires in the vegetation response to ENSO The reconstruction of burned area response over 142 Tanzania captures the simulated temporal evolution reasonably well (R=0.82, p < 0.00001), suggesting that 143 the burned area response can be determined essentially by the time integral of the direct ENSO effect and the 144 C-mode term. Previous studies support the notion that the lagged response of wildfire activity in some areas 145 can be linked to the integrated effect of antecedent precipitation anomalies7,27. In the Periodic experiment, less 146 rainfall over Tanzania during the wet season [D(0)JF(1)] and successive dry season promote a lagged response 147 in burned area in SON(1) (Fig. S2). Subsequently, LAI anomalies over Tanzania slowly develop after the peak 148 of El Niño and are prolonged until the following La Niña event. Especially, the peak of negative anomalies in 149 Wildfires can play a potential role in vegetation change through climate fire vegetation interactions . In 131 the absence of TIO warming in the Periodic and Pacific experiments, El Niño induced drying increases the 132 occurrence of fires, which is manifest in the prolonged positive anomalies in burned area lasting for about 133 one year after the peak of El Niño (Fig. 4 a, b). For wet savannas in Africa, an increase in fuel moisture can 134 lead to a decrease in the burned area7, while for dry savannas, an increase in moisture facilitates more fires26. 135 To investigate the causal linkages between precipitation and wildfire responses to ENSO, we hypothesize 136 that changes in burned area B, are driven by precipitation variability P. Here we choose P as the ENSO- 137 reconstructed precipitation anomaly from equation (1). We assume in its simplest linearized form that the 138 burned area does not depend on the available vegetation which allows us to introduce a fixed mean recovery 139 timescale (), in which the burned area can regrow. The simplified linearized model then reads: 140 𝒅𝑩(𝒕) 𝒅𝒕 = −𝝁𝟏𝑩−𝜽𝟏𝑷∗(𝒕)(𝟐). 141 Appropriate parameters values are given in Table S1. The reconstruction of burned area response over 142 Tanzania captures the simulated temporal evolution reasonably well (R=0.82, p < 0.00001), suggesting that 143 the burned area response can be determined essentially by the time integral of the direct ENSO effect and the 144 C-mode term. Role of wildfires in the vegetation response to ENSO Accordingly, we hypothesize that the LAI response can be 152 largely explained by the integrated effect of burned area (equation 2), where L represents temporal variation 153 of LAI, and λ is 8 month-1 as an inverse damping time scale (characterizing vegetation resilience): 154 𝒅𝑳(𝒕) 𝒅𝒕 = − 𝝁𝟐𝑳−𝜽𝟐𝑩(𝒕) (𝟑). 𝒅𝑳(𝒕) 𝒅𝒕 = − 𝝁𝟐𝑳−𝜽𝟐𝑩(𝒕) (𝟑). According to this simplified double-integration model (equations 1-3) the LAI response over Tanzania correlates highly with simulated LAI anomalies (R=0.72, p < 0.00001), indicating that the lagged and prolonged vegetation response to ENSO is explained by vegetation resilience and the integrated effect of wildfire activity. Similar double-integration models have been introduced to explain also the emergence of low-frequency marine biogeochemical variability29. Role of wildfires in the vegetation response to ENSO Previous studies support the notion that the lagged response of wildfire activity in some areas 145 can be linked to the integrated effect of antecedent precipitation anomalies7,27. In the Periodic experiment, less 146 rainfall over Tanzania during the wet season [D(0)JF(1)] and successive dry season promote a lagged response 147 in burned area in SON(1) (Fig. S2). Subsequently, LAI anomalies over Tanzania slowly develop after the peak 148 of El Niño and are prolonged until the following La Niña event. Especially, the peak of negative anomalies in 149 𝒅𝑩(𝒕) 𝒅𝒕 = −𝝁𝟏𝑩−𝜽𝟏𝑷∗(𝒕)(𝟐). 𝒅𝑩(𝒕) 𝒅𝒕 = −𝝁𝟏𝑩−𝜽𝟏𝑷∗(𝒕)(𝟐). 𝒅𝑩(𝒕) 𝒅𝒕 = −𝝁𝟏𝑩−𝜽𝟏𝑷∗(𝒕)(𝟐). Appropriate parameters values are given in Table S1. The reconstruction of burned area response over 142 Tanzania captures the simulated temporal evolution reasonably well (R=0.82, p < 0.00001), suggesting that 143 the burned area response can be determined essentially by the time integral of the direct ENSO effect and the 144 C-mode term. Previous studies support the notion that the lagged response of wildfire activity in some areas 145 can be linked to the integrated effect of antecedent precipitation anomalies7,27. In the Periodic experiment, less 146 rainfall over Tanzania during the wet season [D(0)JF(1)] and successive dry season promote a lagged response 147 in burned area in SON(1) (Fig. S2). Subsequently, LAI anomalies over Tanzania slowly develop after the peak 148 of El Niño and are prolonged until the following La Niña event. Especially, the peak of negative anomalies in 149 5 5 LAI occurs during the mature La Niña phase in December-January-February in year 2 [DJF(2)], in spite of the 50 maximum rainfall anomalies during this time (Fig. 4, Fig. S2). The vegetation response to climate factors also 51 depends on vegetation resistance and resilience28. Accordingly, we hypothesize that the LAI response can be 52 largely explained by the integrated effect of burned area (equation 2), where L represents temporal variation 53 of LAI, and λ is 8 month-1 as an inverse damping time scale (characterizing vegetation resilience): 54 LAI occurs during the mature La Niña phase in December-January-February in year 2 [DJF(2)], in spite of the 150 maximum rainfall anomalies during this time (Fig. 4, Fig. S2). The vegetation response to climate factors also 151 depends on vegetation resistance and resilience28. Discussion and Conclusions In this study, we explored how vegetation in the southeastern part of Africa changes in response to interannual ENSO variability through a series of model experiments. Focusing on Tanzania, we found that, in the absence of TIO variability, the rapid transition of precipitation anomalies during ENSO events are determined by the interaction between ENSO and the annual cycle (the so-called C-mode). After the occurrence of El Niño, the pronounced decrease in rainfall over Tanzania leads to enhancement in burned area with a time delay, thereby prolonging a marked vegetation decrease for 2 years. This response can be explained by the integrated effect of wildfire (double integrated effect of precipitation) and vegetation resilience through an idealized dynamical model, which explains the AGCM results reasonably well. However, in the real world, there is no evidence for robust changes in precipitation, wildfire, and vegetation over Tanzania, in relationship to ENSO. This is because TIO warming during El Niño events compensates the rainfall response to ENSO over Tanzania (Fig. 4c) by weakening the anomalous atmospheric subsidence (Fig. 1b, c). This offset response is consistent with the opposite impact between Indian Ocean Basin-wide mode (IOBM) and ENSO on seasonal rainfall variability over Africa discussed in Preethi et al.30. The IOD is another primary climate factor which can affect rainfall and vegetation variability over East Africa2,30,31, but the IOD impact to eastern Africa peaks in September-November [SON(0)] (Fig. S3). This is too early to cause major precipitation and vegetation anomalies in Tanzania (Fig. 4, Fig. S3). Furthermore, we emphasize the necessity to understand African vegetation variations driven by the interaction between ENSO and TIO warming under a warmer climate through further studies. Future projections show that TIO warming related to ENSO will likely be intensified32-34. Thus, our results provide a framework to assess future coupled changes in rainfall, wildfires, and vegetation induced by the relationship between ENSO and TIO warming in response to global warming. Model and Experiments We conducted a suite of atmospheric general circulation model (AGCM) experiments with the Community Earth System Model (CESM 1.2.2) using the Community Atmosphere Model version 4.0 (CAM4)40 and Community Land Model version 4.0 (CLM4)41,42 with active Carbon-Nitrogen (CN) biogeochemistry. The model, which uses a horizontal of approximately 1-degree, was spun up until the carbon and nitrogen pools were equilibrated to a 1957-2016 SST climatology boundary forcing and present-day greenhouse gas concentrations. We then performed four different types of AGCM experiment ensembles to investigate the vegetation response over sub-Saharan Africa to interannual tropical SST variability starting from these equilibrated initial conditions. First, a control experiment (CTRL) was carried out with a repeating global climatological SST forcing for the period 1957-2016 using a 3-member ensemble. The CTRL largely reproduces the observed precipitation (PRCP), LAI, and burned area climatological patterns. Climatological mean precipitation over central Africa and southeastern Africa and burned area over some parts of Ethiopia, Tanzania, Angola, and South Africa are somewhat overestimated in the model (Figure S1). To illustrate the impact of observed ENSO variability, a “Pacific” experiment was conducted by adding the observed SST anomalies over the tropical eastern Pacific (15°S-15°N, 180°-90°W) for the period 1957-2016 to the climatology with a 10-member ensemble. A “Tropics” experiment was forced with SST anomalies over the whole tropics (15°S-15°N) for the period 1957-2016 to investigate the response to other modes of pantropical SST variability in addition to ENSO with a 3-member ensemble. An idealized “Periodic” experiment was designed to investigate the response to symmetric ENSO variability (see for instance Stuecker et al.43). The regressed ENSO SST anomaly pattern over the tropical eastern Pacific with an idealized sinusoidal 2.5 years periodicity was added to the observed SST climatology (1957-2016) and the experiment was run for 100 years with a 3-member ensemble. The climate response in all perturbation experiments is defined relative to the control experiment climate. Outside the tropical SST perturbation regions, the SST is the same as in the CTRL simulation. Observations We used precipitation data from Global Precipitation Analysis Products of the Global Precipitation Climatology Centre (GPCC)35, 200 hPa wind from European Centre for Medium-Range Weather Forecasts 6 6 (ECMWF) reanalysis generation 5 (ERA5)36, and SST from the Hadley Centre Sea Ice and Sea Surface 189 Temperature data set version 1 (HadISST1)37. To characterize observed vegetation changes, we utilized leaf 190 area index (LAI) data derived from the Global Inventory Modeling and Mapping Studies (GIMMS) 191 Normalized Difference Vegetation Index (NDVI3g) for the period 1982 to 201138. The monthly Global Fire 192 Emissions Database version 4 (GFEDv4)39 was used to characterize the 1994-2014 wildfire activity. 193 194 Acknowledgments I.-W. Kim and A. Timmermann were supported by the Institute for Basic Science (project code IBS-R028- D1). I.-W. Kim thanks K. Rodgers for comments and editing of the manuscript. M. F. Stuecker thanks Abigail Swann, Marysa Laguë, and Keith Oleson for discussions on CLM. J.-S. Kug and S.-W. Park were supported by NRF-2018R1A5A1024958. J.-S. Kim was supported by University of Zurich Research Priority Programme “Global Change and Biodiversity” (URPP GCB). This is SOEST publication X and IPRC contribution Y. 7 I.-W.K. performed the analysis and wrote the initial draft of the manuscript. A.T. and M.F.S. designed the 229 study. M.F.S. conducted the idealized model experiments. All authors discussed the results and reviewed the 230 manuscript. 231 8 8 1 Camberlin, P., Martiny, N., Philippon, N. & Richard, Y. Determinants of the interannual 233 relationships between remote sensed photosynthetic activity and rainfall in tropical Africa. 234 Remote Sensing of Environment 106, 199-216 (2007). 235 2 Hawinkel, P. et al. Vegetation response to precipitation variability in East Africa controlled by 236 biogeographical factors. Journal of Geophysical Research: Biogeosciences 121, 2422-2444 237 (2016). 238 3 Musau, J., Patil, S., Sheffield, J. & Marshall, M. Spatio-temporal vegetation dynamics and 239 relationshipwithclimate over East Africa. (2016). 240 4 Archibald, S., Nickless, A., Govender, N., Scholes, R. J. & Lehsten, V. Climate and the inter‐ 241 annual variability of fire in southern Africa: a meta‐analysis using long‐term field data and 242 satellite‐derived burnt area data. Global Ecology and Biogeography 19, 794-809 (2010). 243 5 Andela, N. & Van Der Werf, G. R. Recent trends in African fires driven by cropland expansion 244 and El Nino to La Nina transition. Nature Climate Change 4, 791-795 (2014). 245 6 Papagiannopoulou, C. et al. Vegetation anomalies caused by antecedent precipitation in most 246 of the world. Environmental Research Letters 12, 074016 (2017). 247 and El Nino to La Nina transition. Nature Climate Change 4, 791-795 (2014). 245 6 Papagiannopoulou, C. et al. Vegetation anomalies caused by antecedent precipitation in most 246 of the world. Environmental Research Letters 12, 074016 (2017). 247 7 Zubkova, M., Boschetti, L., Abatzoglou, J. T. & Giglio, L. Changes in fire activity in Africa 248 from 2002 to 2016 and their potential drivers. Geophysical research letters 46, 7643-7653 249 (2019). 250 Williams, C. A. et al. Interannual variability of photosynthesis across Africa and its attributio Journal of Geophysical Research: Biogeosciences 113 (2008). 9 Nicholson, S. Acknowledgments E. & Kim, J. The relationship of the El Niño–Southern oscillation to African 253 rainfall. International Journal of Climatology: A Journal of the Royal Meteorological Society 254 17, 117-135 (1997). 255 10 Camberlin, P., Janicot, S. & Poccard, I. Seasonality and atmospheric dynamics of the 256 teleconnection between African rainfall and tropical sea‐surface temperature: Atlantic vs. 257 ENSO. International Journal of Climatology: A Journal of the Royal Meteorological Society 258 21, 973-1005 (2001). 259 11 Anyamba, A., Tucker, C. J. & Mahoney, R. From El Niño to La Niña: Vegetation response 260 patterns over east and southern Africa during the 1997–2000 period. Journal of climate 15, 261 3096-3103 (2002). 262 12 Philippon, N., Martiny, N., Camberlin, P., Hoffman, M. & Gond, V. Timing and patterns of the 263 ENSO signal in Africa over the last 30 years: Insights from normalized difference vegetation 264 index data. Journal of Climate 27, 2509-2532 (2014). 265 f , ( ) 13 Anyamba, A., Glennie, E. & Small, J. Teleconnections and interannual transitions as observed 266 in African Vegetation: 2015–2017. Remote Sensing 10, 1038 (2018). 267 14 Philippon, N., Martiny, N. & Camberlin, P. Forecasting the vegetation photosynthetic activity 268 over the Sahel: a model output statistics approach. International Journal of Climatology: A 269 Journal of the Royal Meteorological Society 29, 1463-1477 (2009). 270 Frauen, C., Dommenget, D., Tyrrell, N., Rezny, M. & Wales, S. Analysis of the nonlinearity of El Niño–Southern Oscillation teleconnections. Journal of Climate 27, 6225-6244 (2014). Wenhaji Ndomeni, C., Cattani, E., Merino, A. & Levizzani, V. An observational study of the variability of East African rainfall with respect to sea surface temperature and soil moisture. Quarterly Journal of the Royal Meteorological Society 144, 384-404 (2018). 17 Wolff, C. et al. Reduced interannual rainfall variability in East Africa during the last ice age. 276 Science 333, 743-747 (2011). 277 18 Yang, Y., Fang, J., Ma, W. & Wang, W. Relationship between variability in aboveground net 278 primary production and precipitation in global grasslands. Geophysical Research Letters 35 279 (2008). 280 ( ) 19 Xie, S.-P. et al. Indian Ocean capacitor effect on Indo–western Pacific climate during the 281 summer following El Niño. Journal of Climate 22, 730-747 (2009). 282 g f ( ) 20 Cai, W. et al. Pantropical climate interactions. Science 363 (2019). 283 p ( ) 21 Liu, W., Cook, K. H. & Vizy, E. K. Acknowledgments Influence of Indian Ocean SST regionality on the East 284 African short rains. Climate Dynamics 54, 4991-5011 (2020). 285 9 22 Latif, M., Dommenget, D., Dima, M. & Grötzner, A. The role of Indian Ocean sea surface 286 temperature in forcing East African rainfall anomalies during December–January 1997/98. 287 Journal of Climate 12, 3497-3504 (1999). 288 23 Stuecker, M. F., Timmermann, A., Jin, F.-F., McGregor, S. & Ren, H.-L. A combination mode 289 of the annual cycle and the El Niño/Southern Oscillation. Nature Geoscience 6, 540-544 (2013). 290 24 Ryan, K. C. Vegetation and wildland fire: implications of global climate change. Environment 291 International 17, 169-178 (1991). 292 25 Chikamoto, Y., Timmermann, A., Stevenson, S., DiNezio, P. & Langford, S. Decadal 293 predictability of soil water, vegetation, and wildfire frequency over North America. Climate 294 Dynamics 45, 2213-2235 (2015). 295 y 26 Archibald, S., Roy, D. P., van Wilgen, B. W. & Scholes, R. J. What limits fire? An examination 296 of drivers of burnt area in Southern Africa. Global Change Biology 15, 613-630 (2009). 297 298 27 Westerling, A. L., Gershunov, A., Brown, T. J., Cayan, D. R. & Dettinger, M. D. Climate and 298 wildfire in the western United States. Bulletin of the American Meteorological Society 84, 595- 299 604 (2003). 300 28 De Keersmaecker, W. et al. A model quantifying global vegetation resistance and resilience to 301 short‐term climate anomalies and their relationship with vegetation cover. Global Ecology and 302 Biogeography 24, 539-548 (2015). 303 29 Di Lorenzo, E. & Ohman, M. D. A double-integration hypothesis to explain ocean ecosystem 304 response to climate forcing. Proceedings of the National Academy of Sciences 110, 2496-2499 305 (2013). 306 ( ) 30 Preethi, B., Sabin, T., Adedoyin, J. & Ashok, K. Impacts of the ENSO Modoki and other 307 tropical Indo-Pacific climate-drivers on African rainfall. Scientific reports 5, 16653 (2015). 308 31 Williams, C. & Hanan, N. ENSO and IOD teleconnections for African ecosystems: evidence 309 of destructive interference between climate oscillations. Biogeosciences 8, 27 (2011). 310 g ( ) 32 Zheng, X.-T., Xie, S.-P. & Liu, Q. Response of the Indian Ocean basin mode and its capacitor 311 effect to global warming. Journal of climate 24, 6146-6164 (2011). 312 33 Chu, J.-E. et al. Future change of the Indian Ocean basin-wide and dipole modes in the CMIP5. 313 Climate dynamics 43, 535-551 (2014). 314 34 Tao, W. Acknowledgments et al. Interdecadal modulation of ENSO teleconnections to the Indian Ocean Basin 315 Mode and their relationship under global warming in CMIP5 models. International Journal of 316 Climatology 35, 391-407 (2015). 317 gy 35 Schneider, U. et al. GPCC's new land surface precipitation climatology based on quality- 318 controlled in situ data and its role in quantifying the global water cycle. Theoretical and Applied 319 Climatology 115, 15-40 (2014). 320 gy 36 Malardel, S. et al. A new grid for the IFS. ECMWF Newsletter 146, 23-28 (2016) 321 37 Rayner, N. et al. Global analyses of sea surface temperature, sea ice, and night marine air 322 temperature since the late nineteenth century. Journal of Geophysical Research: Atmospheres 323 108 (2003). 324 38 Zhu, Z. et al. Global data sets of vegetation leaf area index (LAI) 3g and fraction of 325 photosynthetically active radiation (FPAR) 3g derived from global inventory modeling and 326 mapping studies (GIMMS) normalized difference vegetation index (NDVI3g) for the period 327 1981 to 2011. Remote sensing 5, 927-948 (2013). 328 g 39 Randerson, J., Van der Werf, G., Giglio, L., Collatz, G. & Kasibhatla, P. Global Fire Emissions 329 Database, Version 4.1 (GFEDv4). ORNL DAAC (2015). 330 40 Neale, R. B. et al. The mean climate of the Community Atmosphere Model (CAM4) in forced 331 SST and fully coupled experiments. Journal of Climate 26, 5150-5168 (2013). 332 41 Oleson, K. W., Bonan, G. B., Feddema, J., Vertenstein, M. & Kluzek, E. Technical description 333 of an urban parameterization for the Community Land Model (CLMU). NCAR, Boulder (2010). 334 42 Lawrence, D. M. et al. Parameterization improvements and functional and structural advances 335 i i 4 f th C it L d M d l J l f Ad i M d li E h S 336 41 Oleson, K. W., Bonan, G. B., Feddema, J., Vertenstein, M. & Kluzek, E. Technical description 333 of an urban parameterization for the Community Land Model (CLMU). NCAR, Boulder (2010). 334 5 p y ( ) ( ) 42 Lawrence, D. M. et al. Parameterization improvements and functional and structural advances 335 in version 4 of the Community Land Model. Journal of Advances in Modeling Earth Systems 336 3 (2011). 337 43 Stuecker, M. F., Jin, F.-F., Timmermann, A. & McGregor, S. Combination mode dynamics of 338 the anomalous northwest Pacific anticyclone. Journal of Climate 28, 1093-1111 (2015). Acknowledgments 339 10 340 11 11 Figures Figure 1 The Periodic and the Paci¦c experiments (SST anomalies are only prescribed in the tropical Paci¦c) show pronounced Walker circulation changes between El Niño and La Niña events with anomalous ascending motion over the eastern Paci¦c region (and corresponding upper-level divergence) and anomalous descending motion (and corresponding upper-level convergence) during the peak ENSO phase of December-January-February [D(0)JF(1)] Figure 1 The Periodic and the Paci¦c experiments (SST anomalies are only prescribed in the tropical Paci¦c) show pronounced Walker circulation changes between El Niño and La Niña events with anomalous ascending motion over the eastern Paci¦c region (and corresponding upper-level divergence) and anomalous descending motion (and corresponding upper-level convergence) during the peak ENSO phase of December-January-February [D(0)JF(1)] Figure 2 Figure 2 Figure 4 The temporal evolution of the rainfall response over Tanzania to ENSO shows a rapid transition during the peak phase of both El Niño and La Niña in both the Periodic and the Paci¦c experiments, but not in the Tropics experiment (Fig. 4). Figure 2 Observed composite differences between El Niño and La Niña events display pronounced positive precipitation anomalies over the Horn of Africa and negative anomalies over Southern Africa in D(0)JF(1) Observed composite differences between El Niño and La Niña events display pronounced positive precipitation anomalies over the Horn of Africa and negative anomalies over Southern Africa in D(0)JF(1) Figure 3 The periodic experiment shows negative anomalies in burned area over Tanzania in D(0)JF(1) and statistically insigni¦cant differences in the Paci¦c and Tropics experiments, as well as in the observations. However, the Periodic and the Paci¦c experiments show a 10-20 % increase in burned area over Tanzania in September-October-November in year 1 after ENSO event peak time [SON(1)] The periodic experiment shows negative anomalies in burned area over Tanzania in D(0)JF(1) and statistically insigni¦cant differences in the Paci¦c and Tropics experiments, as well as in the observations. However, the Periodic and the Paci¦c experiments show a 10-20 % increase in burned ar over Tanzania in September-October-November in year 1 after ENSO event peak time [SON(1)] Figure 4 The temporal evolution of the rainfall response over Tanzania to ENSO shows a rapid transition during the peak phase of both El Niño and La Niña in both the Periodic and the Paci¦c experiments, but not in the Tropics experiment (Fig. 4). SupportingInformation.docx Supplementary Files This is a list of supplementary ¦les associated with this preprint. Click to download. SupportingInformation.docx
https://openalex.org/W2795413964	https://jmedicalcasereports.biomedcentral.com/track/pdf/10.1186/s13256-017-1552-4	English	null	A novel single variant in the MEFV gene causing Mediterranean fever and Behçet’s disease: a case report	Journal of medical case reports	2,018	cc-by	3,475	© 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. CASE REPORT Open Access Abstract Background: Familial Mediterranean fever is an autoinflammatory disease of unknown etiology, characterized clinically by recurrent attacks of sudden-onset fever with arthralgia and/or thoracoabdominal pain and pathogenetically by autosomal recessive inheritance due to a mutation in the MEFV gene. Behçet’s disease is an inflammatory disease characterized by recurrent oral and genital aphthous ulcerations, uveitis, and skin lesions. Preliminarily, our literature review suggested that patients with familial Mediterranean fever who also have Behçet’s disease have only a single mutated familial Mediterranean fever gene. The MEFV gene mutation responsible for familial Mediterranean fever is probably a susceptibility factor for Behçet’s disease, particularly for patients with vascular involvement, and both disorders can occur concurrently in a patient, as in the present case. Case presentation: A 10-year-old girl of Moroccan origin presented to our institution for genetic consultation for genetic testing of the MEFV gene. She had fever associated with abdominal and diffuse joint pain in addition to headache. These symptoms have oriented pediatricians to familial Mediterranean fever. The evolution was marked by Behçet’s syndrome symptoms. Sanger sequencing followed by complete exome sequencing analysis of the MEFV gene for the proband mutation revealed a novel variant. We conclude that the novel single variant c.2078 T > A (p.Met693Lys) could be responsible for the association of familial Mediterranean fever and Behçet’s disease. Conclusion: To the best of our knowledge, this is the first report of a new variant in exon 10 of the MEFV gene in a Moroccan family. This novel variant should be listed in the MEFV sequence variant databases. rds: MEFV gene, Familial Mediterranean fever, Behçet’s disease, FMF-BD coexistence, Case report amyloidosis, which can be prevented by the daily and lifelong administration of colchicine therapy [3]. Behçet’s disease (BD) is a complex chronic and relapsing inflam- matory disorder extensively described in young adults from East Asian and Mediterranean countries. Its essential manifestations are oral and genital ulcerations, folliculitis, erythema nodosum, and uveitis. The presence of vasculitis worsens the prognosis, with the potential to introduce life-threatening complications such as thrombophlebitis, arterial aneurysms, and occlusion [4]. The etiology of BD is unknown. A higher frequency of MEFV muta- tions in patients with BD with respect to their ethni- city has previously been demonstrated [5]. Other groups have confirmed these data and found an asso- ciation between the presence of MEFV mutations and the severity of vasculitis. A novel single variant in the MEFV gene causing Mediterranean fever and Behçet’s disease: a case report Maria Zerkaoui1,2, Fatima Zahra Laarabi2, Yousra Ajhoun3, Bouchra Chkirate4 and Abdelaziz Sefiani1,2 Maria Zerkaoui1,2, Fatima Zahra Laarabi2, Yousra Ajhoun3, Bouchra Chkirate4 and Abdelaziz Se Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 DOI 10.1186/s13256-017-1552-4 Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 DOI 10.1186/s13256-017-1552-4 * Correspondence: maria.zerkaoui@gmail.com 1Human Genomic Centre, Faculty of Medicine and Pharmacy, Mohammed V University, Rabat, Morocco 2Department of Medical Genetics, National Institute of Health, Rabat, Morocco Full list of author information is available at the end of the article Case presentation In 2015, we received in genetic consultation a consanguin- eous 10-year-old Moroccan girl, the second of three sib- lings, for genetic testing of the MEFV gene (Fig. 1). At admission, the patient had a periodic fever with abdominal pain, mucocutaneous symptoms (pseudo erysipelas), vom- itus, and joint pain (monoarthritis) in addition to head- ache. Pediatricians had excluded acute rheumatic fever on the basis of normal levels of antistreptolysin O and normal echocardiographic results. During the treatment, the pa- tient demonstrated an intolerance to colchicine with diar- rhea and abdominal pain, so the response to colchicine was not determined. We performed deoxyribonucleic acid (DNA) analysis. Initially, we excluded recurrent mutations in exons 2 and 10. Molecular testing of the entire MEFV gene revealed the novel variant c.2078 T > A, p.M693K, predicted as damaging and deleterious by the SIFT and PROVEAN tools (Fig. 2). This result pointed to a diagno- sis of FMF. Fig. 2 Electrophoregrams of the novel heterozygous missense mutation at the 2078 nucleotide position of the MEFV gene. a (red arrow) Normal sequence of unaffected mother and heterozygous mutation c.2078 T > A (p.Met693Lys) of unaffected father. b (red arrow) Heterozygous mutation c.2078 T > A (p.Met693Lys) of our patient, a Moroccan girl presenting with coexistence of familial Mediterranean fever and Behçet’s disease junction and the sigmoid sinus on the left and right sides. The patient’s clinical history had been taken over. For a few months, the patient was showing one episode of aphthosis in the buccal mucosa as well as urogenital ulcers. She was given a bolus of intravenous methylpred- nisolone for 3 consecutive days, followed by oral pred- nisone; simultaneously, we administered intravenous anticoagulation (enoxaparin sodium) followed by oral acenocoumarol to maintain an international normalized ratio between 2 and 3. Resolution of the thrombus was achieved by the time of the 6-month follow-up visit. Thrombophlebitis of the left peroneal vein visualized by Doppler echography appeared 2 weeks after starting medical treatment. An ophthalmologic examination re- vealed papillary edema degree 2 with scleromalacia. We reconsidered the diagnosis according to the established Tel-Hashomer clinical criteria and the International Study Group for Behçet Disease classification criteria for diagnosis of BD [8]. According to these criteria, the young girl had FMF-BD. We completed the genetic ana- lysis of her parents. Her mother was normal, and her father was heterozygous for the same mutation but remained asymptomatic. Case presentation Meanwhile, the child’s condition had worsened 2 weeks after her admission; high fever had persisted through the prior week and was associated with intense headache and abdominal pain that were relieved with opioids. This severity is atypical in FMF. The laboratory data revealed an elevated level of C-reactive protein, although the levels of white blood cells were within normal range and antiphospholipid antibodies were negative. Cerebral angiography showed cerebral venous thrombosis of the Fig. 1 Pedigree of the patient’s family Background Familial Mediterranean fever (FMF) is an ethnically re- stricted autosomal recessive disease, usually caused by loss-of-function mutations in the MEFV gene on the short arm of chromosome 16 [1]. It is the most common genetic disease in the Eastern Mediterranean Basin, af- fecting in particular Armenians, Turks, North Africans, Ashkenazi Jews, and Arabs. FMF is characterized by re- current episodes of fever accompanied by peritonitis, pleuritis, arthritis, or erysipelas-like erythema [2]. The most severe complication is the development of renal * Correspondence: maria.zerkaoui@gmail.com 1Human Genomic Centre, Faculty of Medicine and Pharmacy, Mohammed V University, Rabat, Morocco 2Department of Medical Genetics, National Institute of Health, Rabat, Morocco Full list of author information is available at the end of the article * Correspondence: maria.zerkaoui@gmail.com 1Human Genomic Centre, Faculty of Medicine and Pharmacy, Mohammed V University, Rabat, Morocco 2Department of Medical Genetics, National Institute of Health, Rabat, Morocco Full list of author information is available at the end of the article Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 Page 2 of 4 Fig. 2 Electrophoregrams of the novel heterozygous missense mutation at the 2078 nucleotide position of the MEFV gene. a (red arrow) Normal sequence of unaffected mother and heterozygous mutation c.2078 T > A (p.Met693Lys) of unaffected father. b (red arrow) Heterozygous mutation c.2078 T > A (p.Met693Lys) of our patient, a Moroccan girl presenting with coexistence of familial Mediterranean fever and Behçet’s disease The MEFV gene encodes the pyrin/marenostrin expressed primarily in the myeloid cell lineage, which modulates cell susceptibility to apoptosis through the regulation of inflammation [6]. More than 300 variants throughout the MEFV gene have been identified, with M694V, M680I, and E148Q being the most common [7]. We report a case of a Moroccan patient with FMF who also had BD (FMF-BD) who had been regarded as having FMF alone 3 years from the onset of disease. Mo- lecular testing revealed a novel single variant that could be responsible for the association of FMF and BD. Discussion FMF is a genetic disease characterized by attacks of fever, arthritis, serositis, and abdominal pain. BD is an inflam- matory disorder with a genetic basis that is characterized by oral and genital ulcers, uveitis, pustular erythematous cutaneous lesions, arthritis, central nervous system in- volvement, and possible vascular manifestations such as Fig. 1 Pedigree of the patient’s family Page 3 of 4 Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 Page 3 of 4 venous thrombosis, arteritis, and aneurysms. The diagno- sis of BD is independently associated with higher inci- dence of FMF, especially in females, among people of Arab descent. Watad et al.’s data implied that understating the differentiation between FMF and BD was not evident and clear in a real-life population of patients with BD [9]. The clinical picture of FMF is often similar to that of BD with regard to inflammatory reactions with activation of neutrophils and therapeutic efficacy of colchicine, and the possible coexistence of the former should be considered, particularly in cases with periodic fever as in our patient. Because MEFV encodes the neutrophil protein pyrin, mu- tations in the MEFV gene may play an important role in the pathogenesis of BD [10]. One hypothesis is that a re- duced level of functional pyrin on the one hand, presum- ably characterizing heterozygosity, and increased demand for pyrin on the other hand, by activated granulocytes of BD, may ultimately lead to FMF attacks [11]. simple loss of function of the recessive model of FMF in- heritance. In some cases, FMF is a dominant condition with low penetrance and variable disease expression, presenting not only in homozygous subjects but also in heterozygous subjects [2]. According to their phenotype classification in the pub- lication by Balci et al. [14], patients with a classic pheno- type who were genetically confirmed to have MEFV mutations are defined as phenotype I. Phenotype III was defined as those who have no clinical signs but present with the required genotype, whereas phenotype II refers to patients who develop amyloid A amyloidosis without any previous attacks typical of FMF [14]. Therefore, our patient’s father was asymptomatic but carried the same novel variant c.2078 T > A in the MEFV gene. FMF is typically inherited in an autosomal recessive fashion, although there have been rare reports of domin- ant inheritance with specific heterozygous mutations and complex alleles [15]. Conclusions We present what is, to the best of our knowledge, the first identification of an MEFV mutation in a Moroccan family with associated FMF and BD. To the best of our know- ledge, this mutation has not been reported before. There- fore, we think that the c.2078 T > A (p.M693K) mutation should be listed in the MEFV sequence variants of FMF- BD databases and may help to increase public awareness of this rare association in the Moroccan population. To identify specific phenotype/genotype correlations exactly, the novel mutation and its relationship should be confirmed by studying homozygous individuals for this mutation. In 2001, Livneh et al. [11] suggested that about 60% of patients with FMF-BD have only a single mutated FMF gene (MEFV). Moreover, in three families in that study, mutational analysis revealed at least one heterozygous family member who shared the noncarrier chromosome with the FMF-BD proband but remained unaffected. This may be explained by the presence of a low- penetrance mutation, not detected by that team, in non- coding regions (for example, intron, promoter, enhan- cer), which causes FMF expression only in a fraction of carriers [11]. A significant number of patients diagnosed with FMF show only a single mutation despite sequen- cing of the entire MEFV genome region or other autoin- flammatory genes, and this has led to reconsideration of Abbreviations BD: Behçet’s disease; FMF: Familial Mediterranean fever; FMF-BD: Familial Mediterranean fever with associated Behçet’s disease Discussion The parents of our proband were clinically normal. Genetic analysis of the mother’s MEFV gene was normal. The father, even if he was un- affected, carried the proband mutation. y y Diagnosis of FMF remains clinical and requires informa- tion about family history and response to colchicine. Our patient presented with FMF symptoms of fever, abdominal pain, joint pain, and erysipelas-like erythema. In the literature, severe abdominal and chest pain are reported to be the most typical symptoms, occurring in more than 90% and more than 40% of patients, respectively. Suggest- ive of an FMF diagnosis is the presence of a recurrent erysipelas-like rash on the skin of the inferior limbs [12]. Our patient presented with a recurrent headache, which is an atypical symptom of FMF. She kept having headaches, but her pediatricians considered this symptom a conse- quence of fever. She presented with aphthosis in the buccal and genital regions, and after investigation, we ob- jectified cerebral and peroneal vein thrombophlebitis, symptoms related to BD. It has been reported that the MEFV genetic mutation responsible for FMF is probably a susceptibility factor for BD, particularly in cases with vas- cular involvement, and that both disorders can coexist in a patient, as was the case in our patient [10]. MEFV muta- tions might be a more crucial genetic factor in female pa- tients with vascular involvement [13]. There is good evidence that position M694 of the pyrin molecule is critically important for its normal function. The p.M694 is located in the putative binding site of caspase-1, and the substitution or deletion of this residue may interfere with the inhibitory interaction between pyrin and caspase-1 and thus promote interleukin- 1β generation [16]. This reasoning could be applied to the p.M693K mutation of the pyrin molecule found in our pa- tient and thus highly suggests its pathogenic effect. Our pa- tient carried a single mutated allele of the MEFV gene that could be responsible for the coexistence of FMF and BD. Competing interests Competing interests The authors declare that they have no competing interests. The authors declare that they have no competing interests. 16. Rowczenio DM, Iancu DS, Trojer H, Gilbertson JA, Gillmore JD, Wechalekar AD, Tekman M, Stanescu HC, Kleta R, Lane T, et al. Autosomal dominant familial Mediterranean fever in Northern European Caucasians associated with deletion of p.M694 residue—a case series and genetic exploration. Rheumatology (Oxford). 2017;56:209–13. Availability of data and materials Not applicable. 13. Cattan D. MEFV mutation carriers and diseases other than familial Mediterranean fever: proved and non-proved associations; putative biological advantage. Curr Drug Targets Inflamm Allergy. 2005;4(1):105–12. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author details 1 1Human Genomic Centre, Faculty of Medicine and Pharmacy, Mohammed V University, Rabat, Morocco. 2Department of Medical Genetics, National Institute of Health, Rabat, Morocco. 3Department of Ophthalmology, Mohammed V Military Teaching Hospital, Mohammed V University, Rabat, Morocco. 4Pediatric Department IV, Children’s Hospital, University M h d V R b t M Mohammed V, Rabat, Morocco. Received: 24 March 2017 Accepted: 23 December 2017 Consent for publication Written informed consent was obtained from the patient’s legal guardian(s) for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal. 15. Stoffels M, Szperl A, Simon A, Netea MG, Plantinga TS, van Deuren M, Kamphuis S, Lachmann HJ, Cuppen E, Kloosterman WP, et al. MEFV mutations affecting pyrin amino acid 577 cause autosomal dominant autoinflammatory disease. Ann Rheum Dis. 2014;73(2):455–61. Israeli patients suffering from familial Mediterranean fever and Behcet’s disease (FMF-BD). Eur J Hum Genet. 2001;9(3):191–6. 12. Rigante D, Vitale A, Lucherini OM, Cantarini L. The hereditary autoinflammatory disorders uncovered. Autoimmun Rev. 2014;13(9):892–900. 13. Cattan D. MEFV mutation carriers and diseases other than familial Mediterranean fever: proved and non-proved associations; putative biological advantage. Curr Drug Targets Inflamm Allergy. 2005;4(1):105–12. 14. Balci B, Tinaztepe K, Yilmaz E, Gucer S, Ozen S, Topaloglu R, Besbas N, Ozguc M, Bakkaloglu A. MEFV gene mutations in familial Mediterranean fever phenotype II patients with renal amyloidosis in childhood: a retrospective clinicopathological and molecular study. Nephrol Dial Transplant. 2002;17(11):1921–3. 15. Stoffels M, Szperl A, Simon A, Netea MG, Plantinga TS, van Deuren M, Kamphuis S, Lachmann HJ, Cuppen E, Kloosterman WP, et al. MEFV mutations affecting pyrin amino acid 577 cause autosomal dominant autoinflammatory disease. Ann Rheum Dis. 2014;73(2):455–61. 16. Rowczenio DM, Iancu DS, Trojer H, Gilbertson JA, Gillmore JD, Wechalekar AD, Tekman M, Stanescu HC, Kleta R, Lane T, et al. Autosomal dominant familial Mediterranean fever in Northern European Caucasians associated with deletion of p.M694 residue—a case series and genetic exploration. Rheumatology (Oxford). 2017;56:209–13. Authors’ contributions 14. Balci B, Tinaztepe K, Yilmaz E, Gucer S, Ozen S, Topaloglu R, Besbas N, Ozguc M, Bakkaloglu A. MEFV gene mutations in familial Mediterranean fever phenotype II patients with renal amyloidosis in childhood: a retrospective clinicopathological and molecular study. Nephrol Dial Transplant. 2002;17(11):1921–3. MZ was responsible for data analysis, literature review, and manuscript preparation. FZL was responsible for data analysis and carried out the molecular genetic studies. YA provided medical care. BC provided medical care. AS reviewed the manuscript. All authors read and approved the final manuscript. Ethics approval and consent to participate Not applicable. Funding 12. Rigante D, Vitale A, Lucherini OM, Cantarini L. The hereditary autoinflamma disorders uncovered. Autoimmun Rev. 2014;13(9):892–900. 12. Rigante D, Vitale A, Lucherini OM, Cantarini L. The hereditary autoinflammator disorders uncovered. Autoimmun Rev. 2014;13(9):892–900. Availability of data and materials Not applicable. Acknowledgements W h k h Acknowledgements We thank the patient and his family for their kind cooperation. Page 4 of 4 Page 4 of 4 Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 Zerkaoui et al. Journal of Medical Case Reports (2018) 12:53 References 1. Nakamura A, Yazaki M, Tokuda T, Hattori T, Ikeda S. A Japanese patient with familial Mediterranean fever associated with compound heterozygosity for pyrin variant E148Q/M694I. Intern Med. 2005;44(3):261–5. 1. Nakamura A, Yazaki M, Tokuda T, Hattori T, Ikeda S. A Japanese patient with familial Mediterranean fever associated with compound heterozygosity for pyrin variant E148Q/M694I. Intern Med. 2005;44(3):261–5. 2. Joo K, Park W, Chung MH, Lim MJ, Jung KH, Heo Y, Kwon SR. Extensive thrombosis in a patient with familial Mediterranean fever, despite hyperimmunoglobulin D state in serum. J Korean Med Sci. 2013;28(2):328–30. 3. Öztürk A, Elsobky E, Elsayed SM, Alhodhod M, Akar N. Mutational analysis of the MEFV gene in Egyptian patients with familial Mediterranean fever. Turk J Med Sci. 2009;39(2):229–34. 4. Kone-Paut I, Sanchez E, Le Quellec A, Manna R, Touitou I. Autoinflammatory gene mutations in Behcet’s disease. Ann Rheum Dis. 2007;66(6):832–4. 5. Touitou I, Magne X, Molinari N, Navarro A, Quellec AL, Picco P, Seri M, Ozen S, Bakkaloglu A, Karaduman A, et al. MEFV mutations in Behcet’s disease. Hum Mutat. 2000;16(3):271–2. 6. Centola M, Wood G, Frucht DM, Galon J, Aringer M, Farrell C, Kingma DW, Horwitz ME, Mansfield E, Holland SM, et al. The gene for familial Mediterranean fever, MEFV, is expressed in early leukocyte development and is regulated in response to inflammatory mediators. Blood. 2000;95(10):3223–31. 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: 7. Rigante D, Frediani B, Cantarini L. A comprehensive overview of the hereditary periodic fever syndromes. Clin Rev Allergy Immunol. 2016. https://doi.org/10.1007/s12016-016-8537-8. • 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: 9. Watad A, Tiosano S, Yahav D, Comaneshter D, Shoenfeld Y, Cohen AD, Amital H. Behcet’s disease and FMF: two sides of the same coin or just an association? A cross-sectional study. Eur J Intern Med. 2017;39:75–8. 10. Matsuda M, Nakamura A, Tsuchiya S, Yoshida T, Horie S, Ikeda S. Coexistence of familial Mediterranean fever and Behcet’s disease in a Japanese patient. Intern Med. 2006;45(12):799–800. 10. Matsuda M, Nakamura A, Tsuchiya S, Yoshida T, Horie S, Ikeda S. Coexistence of familial Mediterranean fever and Behcet’s disease in a Japanese patient. Intern Med. 2006;45(12):799–800. 11. Livneh A, Aksentijevich I, Langevitz P, Torosyan Y, G-Shoham N, Shinar Y, Pras E, Zaks N, Padeh S, Kastner DL, et al. A single mutated MEFV allele in 11. Livneh A, Aksentijevich I, Langevitz P, Torosyan Y, G-Shoham N, Shinar Y, Pras E, Zaks N, Padeh S, Kastner DL, et al. A single mutated MEFV allele in
https://openalex.org/W3196788968	https://europepmc.org/articles/pmc8464704?pdf=render	English	null	COVID-19 and Inherited Metabolic Disorders: One-Year Experience of a Referral Center	Children	2,021	cc-by	3,535	Citation: Tummolo, A.; Paterno, G.; Dicintio, A.; Stefanizzi, P.; Melpignano, L.; Aricò, M. COVID-19 and Inherited Metabolic Disorders: One-Year Experience of a Referral Center. Children 2021, 8, 781. https://doi.org/10.3390/ children8090781 Academic Editor: Kong Y. Chen Keywords: COVID-19; inherited metabolic disorders; rare diseases; SARS-CoV-2 Received: 4 August 2021 Accepted: 4 September 2021 Published: 6 September 2021 children children children Article COVID-19 and Inherited Metabolic Disorders: One-Year Experience of a Referral Center molo 1,,†, Giulia Paterno 1, Annamaria Dicintio 1 , Pasquale Stefanizzi 2, Livio Melpignano 3 A i ò 4 Albina Tummolo 1,,†, Giulia Paterno 1, Annamaria Dicintio 1 , Pasquale Stefanizzi 2, Livio Melpignano 3 and Maurizio Aricò 4 1 Department of Metabolic Diseases and Clinical Genetics, Giovanni XXIII Children Hospital, Azienda Ospedaliero-Universitaria Consorziale Policlinico, 70126 Bari, Italy; giupatvi@gmail.com (G.P.); annamaria.dicintio@gmail.com (A.D.) 1 Department of Metabolic Diseases and Clinical Genetics, Giovanni XXIII Children Hospital, Azienda Ospedaliero-Universitaria Consorziale Policlinico, 70126 Bari, Italy; giupatvi@gmail.com (G.P.); annamaria.dicintio@gmail.com (A.D.) 2 Department of Biomedical Science and Human Oncology, Post Graduate School of Hygiene and Preventive Medicine, University of Bari Aldo Moro, 70124 Bari, Italy; pasquale.stefanizzi@uniba.it 3 Medical Direction, Giovanni XXIII Children Hospital, Azienda Ospedaliero-Universitaria Consorziale Policlinico, 70126 Bari, Italy; livio.melpignano@policlinico.ba.it 4 Strategic Control Azienda Ospedaliero-Universitaria Consorziale Policlinico, Piazza Giulio Cesare, 70124 Bari, Italy; maurizio.arico@policlinico.ba.it y p * Correspondence: albina.tummolo@policlinico.ba.it † Metabolic Diseases and Clinical Genetics Children Hospital Giovanni XXIII, Azienda Abstract: Understanding the potential risks of patients with inherited metabolic disorder (IMD) exposed to the COVID-19 pandemic is an unmet need for those involved in their management. Here, we report on the incidence of COVID-19 in a cohort of patients with IMD treated at a children’s hospital and compare them with a matched control group. Among the total number of 272 patients actively followed at a referral center, 19 (7%) tested positive for SARS-CoV-2 between March 2020 and March 2021. Their median age was 16.2 years (range 1.4–32.8 years). In two-thirds of the cases, the source of infection was a family member; 12/19 patients (63%) were asymptomatic, only one required hospitalization, and none of them died. In our single-center experience, COVID-19 had a moderate impact on a relatively large cohort of patients with IMD, including children and young adults. The clinical course was very mild in all but one case. The proportion of symptomatic cases and the clinical course were comparable in patients with IMD and in a group of matched, non-IMD COVID-19 controls from the general population. Citation: Tummolo, A.; Paterno, G.; Dicintio, A.; Stefanizzi, P.; Melpignano, L.; Aricò, M. COVID-19 and Inherited Metabolic Disorders: One-Year Experience of a Referral Center. Children 2021, 8, 781. https://doi.org/10.3390/ children8090781 Academic Editor: Kong Y. Chen Received: 4 August 2021 Accepted: 4 September 2021 Published: 6 September 2021 1. Introduction Rare diseases are diseases that affect a small number of people compared to the general population. Despite their rarity, these conditions collectively affect ~30 million people across Europe [1] and come with particular challenges related to a higher risk of serious illness like COVID-19-related disease. Small numbers and limited experience may hamper the conﬁdence of the attending specialist when facing this situation. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Inherited metabolic disorders (IMD) represent one of the most common groups of Mendelian diseases [2]. Indeed, despite being individually rare or ultra-rare (<1:2,000,000), they are increasingly recognized all over the world thanks to neonatal screening programs and improved diagnostic methods [3]. Copyright: © 2021 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/). p g In the course of events such as the COVID-19 pandemic, patients with IMD may represent a group of fragile subjects with an increased risk or at least special needs. In Italy, the spread of the pandemic occurred earlier than in other European countries. Since then, the frequency of the virus variants has changed over time, with the alpha variant of the virus [4] becoming progressively prevalent in the country, reaching an 88.1% prevalence by March 2021 [5]. https://www.mdpi.com/journal/children Children 2021, 8, 781. https://doi.org/10.3390/children8090781 Children 2021, 8, 781 2 of 6 The available reports on the impact of the pandemic on the management of IMD patients [6,7] have emphasized the effects on clinical planning, access to therapies, and management of emergencies. Few studies have so far focused on the clinical impact of SARS-CoV-2 infection on this group of patients. Here, we report on the incidence of COVID-19 in a cohort of children and young adult patients with IMDs cared for at a children’s hospital. We aim for characterizing the frequency, the course, and the severity of COVID-19; its impact on the course of the underlying disease; and its management. 2. Materials and Methods All patients with IMD followed at our unit who tested positive for SARS-CoV-2 infection between March 2020 and March 2021 were included. No local ethics committee approval was needed, as this was an observational study. The diagnosis was based on the positivity of the nasopharyngeal swab analyzed by real-time reverse-transcription polymerase chain reaction (rRT-PCR). Swabs were performed either under the direction of the general practitioner, as part of the tracing of social/familial contacts, or as part of the screening program upon patient admission. A communication of positivity to external testing was given by email. Information on the clinical status and follow-up was collected by telephone interview or teleconsulting. During the national lockdown period, biochemical monitoring was performed at home, the results were reported to the clinical center, and the treatment was modiﬁed accordingly. Then, to evaluate the possible differences compared to the general population, we retrieved data on the subjects who tested positive for COVID19 in the same time interval from the health surveillance database of the Apulian region. Subjects with COVID-19 were matched with IMD patients by propensity score matching. The variables used to compute the propensity scores were age and sex. p p y g By the institutional policy, given the retrospective and noninterventional design of the study, formal approval by our institutional review board was not considered necessary. 3. Results Among a total of 272 patients with IMD treated and followed-up on over the last two years at our referral center, 19 patients (7%): 12 females and 7 males, with a median age of 16.2 years (range 1.4–32.8 years), tested positive for SARS-CoV-2. In 12 out of the 19 cases (63%), the source of infection was a family member (Table 1). All the cases occurred between October 2020 and March 2021. All of them were documented as infected by the alpha variant of SARS-CoV-2. None of them were vaccinated against SARS-CoV-2. p g In 5/19 cases, the scheduled visit at the referral center was postponed because of COVID-19. Two patients were affected by lysosomal storage disorders (LSD) and nine patients by altered phenylalanine metabolism: three by hyperphenylalaninemia (PHA) and six by phenylketonuria (PKU). Five patients were affected by urea cycle disorders (UCD), and the other three patients had cobalamin C deﬁciency, type 2 glutaric aciduria, and maple syrup urine disease (MSUD) (Table 1). Of the two patients with lysosomal storage disease—namely, Gaucher type 1 and juvenile Pompe disease, the ﬁrst one developed only rhinitis, and the enzyme replacement therapy (ERT) was not modiﬁed. The second one, with long-lasting restrictive respiratory syndrome requiring nocturnal noninvasive ventilation, developed a fever as the only COVID-related manifestation and could be treated at home. The ERT was modiﬁed with the omission of one scheduled injection. Children 2021, 8, 781 3 of 6 Table 1. Patients and infection characteristics. Table 1. Patients and infection characteristics. 3. Results Case Age (y)/Gender Disease Traced Contact Time of Infection COVID Manifestations; Duration Impact on Therapy/Diet 1 2.8/F Hyperphenylalaninemia Family October 2020 1-day fever and diarrhea; 6-day diffused eritemato-papular rash free diet 2 2.5/F Hyperphenylalaninemia Family November 2020 None free diet 3 15.3/M Ornitin Transcarbamilase deﬁciency Family November 2020 none no changes 4 16.5/F Lysinuric protein intolerance Family December 2020 None no changes 5 12/M Type 2 glutaric aciduria Family December 2020 None no change 6 4.6/F Hyperphenylalaninemia Family December 2020 None free diet 7 20.3/F Cobalamin C deﬁciency Family December 2020 1-day fever, cough, and rhinitis no changes 8 20.4/M Phenylketonuria Work December 2020 1-day fever and headache no changes 9 11.9/F Phenylketonuria Family December 2020 none no changes 10 7.8/F Phenylketonuria Family December 2020 none no changes 11 24.3/M Juvenile Pompe disease Social December 2020 3-day fever 1 ERT infusion missed 12 29.9/M Phenylketonuria Work December 2020 1-day osteoarticular pain, 7 days loss of taste no changes 13 17.6/M Arginine succinic aciduria Family January 2021 none no changes 14 32.8/F Gaucher Disease type 1 work February 2021 3-day rhinitis no ERT infusion missed 15 1.4/M Arginine succinic aciduria Family February 2021 none no changes 16 26.8/F Maple Syrup Urine Disease Family February 2021 2-day fever, lack of appetite emergency diet regimen 17 28.1/F Phenylketonuria Family February 2021 none no changes 18 1.6/F Citrullinemia type 1 Family March 2021 none no changes 19 16.2/F Phenylketonuria Family March 2021 none no changes Among the nine patients (47%) affected by altered phenylalanine metabolism, none of them showed any increase in Phe plasma values; therefore, a reduction in Phe intake was not necessary in those with protein-restricted diets or in the case of a fever. Among the other patients with disorders of intermediate metabolisms with decompen- sation risks, only the patient with MSUD (OMIM # 248600), aged 26.8 years, experienced acute metabolic decompensation. She presented with a fever (>38 ◦C) and inappetence, which resulted in a progressive reduction of her daily protein intake. Ten days after the diagnosis of COVID-19, she developed altered consciousness, hallucinations, and vomit- ing; her leucine levels peaked at 708 µmol/L, conﬁrming an acute metabolic imbalance. Among the other patients with disorders of intermediate metabolisms with decompen- sation risks, only the patient with MSUD (OMIM # 248600), aged 26.8 years, experienced acute metabolic decompensation. 4. Discussion Understanding the special needs and potential risks of patients with IMD exposed to the COVID-19 pandemic has been an unmet need for the pediatricians involved in their treatment. Thus, we revised our experience on patients followed-up over the last two years at our referral center. Of them, only 19, i.e., 7%, developed COVID-19. The source of infection was familial/social exposure in two-thirds of them. This ﬁnding is widely in keeping with the trend observed in our hospital for children with COVID-19 [8] but, also, for healthcare workers, suggesting that the application of a stringent policy for DPI use and distancing turned out to be effective in preventing the intrahospital spread of SARS-CoV-2 infection. [9] The manifestation of SARS-CoV-2 infection is much less severe in children and ado- lescents than in adults and the elderly population [10]. In this series of patients with IMD, the manifestation of COVID-19 was very mild, with only one case requiring hospital admission, and none of them requiring admission to the ICU. This may be explained by the young ages of our population, in which the clinical manifestations mirrored that of the general population of the same age group. Furthermore, the fact that all the cases were concentrated in the second and third pandemic waves of our country is in line with the general trend of the incidences in the pediatric and young adult individuals in the Apulian population [8], which has seen an increase in the index of contagion and hospitalization starting from October 2020 [11]. Two patients with lysosomal storage disease on ERT contracted SARS-CoV-2 infection in a social environment and had a very favorable clinical course. In a recent study, 88 out of 181 patients with Gaucher disease were tested for SARS-CoV-2; of them, 16 (18%) tested positive. None of them required COVID-19-directed treatment or died [12]. Our juvenile-onset Pompe patient, with pre-existent respiratory impairment and noc- turnal ventilation support, did not show any further respiratory deterioration. The clinical picture led to a very short ERT discontinuation, in keeping with the recommendation of Wenninger et al., who, in a prospective cohort study of 12 patients with late-onset Pompe disease, showed that the interruption of ERT was associated with a deterioration in the core clinical outcome measures [13]. 3. Results She presented with a fever (>38 ◦C) and inappetence, which resulted in a progressive reduction of her daily protein intake. Ten days after the diagnosis of COVID-19, she developed altered consciousness, hallucinations, and vomit- ing; her leucine levels peaked at 708 µmol/L, conﬁrming an acute metabolic imbalance. Children 2021, 8, 781 4 of 6 Hospitalization was required for seven days, with an enteral emergency diet intravenous infusion of ﬂuids, as well as nonprotein calories. The comparison of this group of 19 patients with IMD with 34 COVID-19-positive control subjects matched by propensity scores revealed that the proportion of asymptomatic subjects was 76% in the control group vs. 63% in IMD patients (p-value = 0.32); no hospitalizations occurred in the control group (p-value = 0.97); no deaths were reported in both groups (Table 2). Table 2. Comparison of the main parameters between the IMD and control groups. IMD/COVID+ Control Group COVID+ p-Value Number 19 34 − Sex female, number (%) 12 (63%) 22 (64%) 0.92 Age, median (range) 16.2 (1.4–32.8) 16 (1.2–33.5) 0.32 Asymptomatic, number (%) 12 (63%) 26 (76%) 0.32 Hospitalization, number (%) 1 (5%) 0 (0%) 0.97 Deaths, number (%) 0 (0%) 0 (0%) − Table 2. Comparison of the main parameters between the IMD and control groups. Discussion 4. Discussion Among the patients with IMD with decompensation risk, MSUD represented the only disorder with decompensation requiring hospitalization because of a reduction in protein intake and fever, well-known causes of decompensation in these patients. Interestingly, the patient with type 2 glutaric aciduria and the one with arginine succinic aciduria aged 17.6 years in 2019 developed, respectively, two and four decompensations due to intercur- Children 2021, 8, 781 5 of 6 rent illnesses, requiring hospitalization. Other patients with IMD with decompensation risk, such as mitochondrial diseases, did not develop COVID19 over the study period. The overall number of admissions/referrals to the center due to intercurrent infections diminished compared to the previous year, likely due to the lockdown social restrictions. In our center, we followed a large cohort of 183 subjects, children and adults, with Phe metabolism disorders, of which nine (4.9%) were diagnosed with COVID-19. It is interesting to note that a fever was not associated with increased plasma Phe values during regular monitoring; thus, an adjustment of the Phenylalanine intake was not required. Within the small world of children (and adults) with rare diseases, the role of the referral center has been one of mainly supporting the spokes represented by local hospitals— on one hand, conﬁrming the minimal need for treatment adjustment, but in the meanwhile, offering support for the FAQs and worries of the parents and the adolescents facing an unknown event perceived as potentially catastrophic. Conﬁrming that the usual diet and long-lasting therapies were adequate in this condition turned out to be of help and relief in most cases. In keeping with this approach, the European Society of Endocrinology (ESE) Rare Disease Committee, alongside the Endocrinology European Reference Network (ENDO-ERN), have engaged in an initiative to collect essential data concerning speciﬁc groups of patients with rare endocrine conditions who develop COVID-19 [14]. g p p p This study has some limitations. Since we did not perform a universal screening of all our patients by nasopharyngeal swab, we cannot rule out that other patients may have developed asymptomatic SARS-CoV-2 infections. Thus, the incidences we report of symptomatic COVID-19 in patients with IMD have to be taken as the minimal amount. Second, not all the patients were directly cared for in our hospital during the manifestation of COVID-19; yet, these were so few that a reporting bias should not represent a major issue. 4. Discussion A further study on the effects of other variants of COVID-19 on these patient populations is recommended. In conclusion, our experience suggests that the COVID-19 circulation among young patients with IMD was minimal, followed by the epidemiologic distribution of the in- fection and the severity of the manifestations in the general population of the same age. The accurate monitoring and close interactions between the referral centers, the local spokes and patient charities may be instrumental in reassuring the patients and improving the outcomes. Author Contributions: A.T. interpreted data and wrote the manuscript, G.P. and A.D. made contri- butions to the study sample follow-up, P.S. and L.M. made the statistical analysis and interpreted the data; and M.A. critically revised the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: The authors declare no sources of study funding, including sponsorship, for the described work. Institutional Review Board Statement: This study was conducted according to the guidelines of the “Declaration of Helsinki” and did not require approval by the local ethics committee. Informed Consent Statement: Written informed consent was obtained from the participants of the study/their parents. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Conﬂicts of Interest: No ﬁnancial beneﬁts were received or will be received from any party related directly or indirectly to the subject of this article. Conﬂicts of Interest: No ﬁnancial beneﬁts were received or will be received from any party related directly or indirectly to the subject of this article. 3. Mak, C.M.; Lee, H.C.; Chan, A.Y.; Lam, C.W. Inborn errors of metabolism and expanded newborn screening: Review and update. Crit. Rev. Front. Med. 2013, 50, 142–162. Available online: www.frontiersin.org (accessed on 8 February 2021). e ce e es oc o ogy Spo g o e se ses , , 5 [C oss e ] 2. Morava, E.; Rahman, S.; Peters, V.; Baumgartner, M.R.; Patterson, M.; Zschocke, J. Quo vadis: The redeﬁnition of ‘inborn metabolic diseases’. J. Inherit. Metab. Dis. 2015, 38, 1003–1006. [CrossRef] [PubMed] References 1. The Lancet Diabetes Endocrinology. Spotlight on rare diseases. Lancet Diabetes Endocrinol. 2019, 7, 75. [CrossRef] gy p g [ ] 2. Morava, E.; Rahman, S.; Peters, V.; Baumgartner, M.R.; Patterson, M.; Zschocke, J. Quo vadis: The redeﬁnition of ‘inborn metabolic diseases’. J. Inherit. Metab. Dis. 2015, 38, 1003–1006. [CrossRef] [PubMed] 3. Mak, C.M.; Lee, H.C.; Chan, A.Y.; Lam, C.W. Inborn errors of metabolism and expanded newborn screening: Review and update. Crit. Rev. Front. Med. 2013, 50, 142–162. Available online: www.frontiersin.org (accessed on 8 February 2021). Children 2021, 8, 781 6 of 6 4. O’Toole, Á.; Hill, V.; Pybus, O.G.; Watts, A.; Bogoch, I.I.; Khan, K.; Messina, J.P.; The COVID-19 Genomics UK (COG-UK) consortium; Network for Genomic Surveillance in South Africa (NGS-SA); Brazil-UK CADDE Genomic Network; et al. Tracking the international spread of SARS-CoV-2 lineages B.1.1.7 and B.1.351/501Y-V2 [version 1; peer review: 3 approved]. Wellcome p 5. COVID-19 Data Portal–Italy. Available online: https://www.covid19dataportal.it/highlights/highlight23/ (accessed on 2 September 2021). p ) 6. Elmonem, M.A.; Belanger-Quintana, A.; Bordugo, A.; Boruah, R.; Cortès-Saladelafont, E.; Endrakanti, M.; Giraldo, P.; Grünert, S.C.; Gupta, N.; Kabra, M.; et al. The impact of COVID-19 pandemic on the diagnosis and management of inborn errors of metabolism: A global perspective. Mol. Genet. Metab. 2020, 131, 285–288. [CrossRef] [PubMed] 7. Lampe, C.; Dionisi-Vici, C.; Bellettato, C.M.; Paneghetti, L.; van Lingen, C.; Bond, S.; Brown, C.; Finglas, A.; Francisco, R.; Sestini, S.; et al. The impact of COVID-19 on rare metabolic patients and healthcare providers: Results from two MetabERN surveys. Orphanet J. Rare Dis. 2020, 15, 341. [CrossRef] [PubMed] p 8. Loconsole, D.; Caselli, D.; Centrone, F.; Morcavallo, C.; Campanella, S.; Aricò, M.; Chironna, M. SARS-CoV-2 Infection in Children in Southern Italy: A Descriptive Case Series. Int. J. Environ. Res. Public Health 2020, 17, 6080. [CrossRef] [PubMed] C ll l Ch A ò ff f l f 9. Caselli, D.; Loconsole, D.; Dario, R.; Chironna, M.; Aricò, M. Effectiveness of Preventive Measures in Keeping Low Prevalence of SARS-CoV-2 Infection in Health Care Workers in a Referral Children’s Hospital in Southern Italy. Pediatr. Rep. 2021, 13, 118–124. [CrossRef] [PubMed] 10. Caselli, D.; Aricò, M. 2019-nCoV: Polite with children! Pediatr. Rep. 2020, 12, 8495. [CrossRef] [PubMed 10. Caselli, D.; Aricò, M. 2019-nCoV: Polite with children! Pediatr. Rep. 2020, 12, 8495. [CrossRef] [PubMed] 10. Caselli, D.; Aricò, M. 2019-nCoV: Polite with children! Pediatr. Rep. 2020, 12, 8495. [CrossRef] [PubMed] 11. Bollettino Epidemiologico Regione Puglia. References Available online: https://press.regione.puglia.it/-/bollettino-epidemiologico-regione- puglia-26-aprile-2021 (accessed on 26 April 2021). 11. Bollettino Epidemiologico Regione Puglia. Available online: https://press.regione.puglia.it/-/bollettino-epidemiologico-regione- puglia-26-aprile-2021 (accessed on 26 April 2021). p wat, N.; Naik, H.; Narayanan, P.; Mistry, P.K.; Balwani, M. Gaucher disease and SARS-CoV-2 infection 1 patients in New York. Mol. Genet. Metab. 2021, 132, 44–48. [CrossRef] [PubMed] 12. Fierro, L.; Nesheiwat, N.; Naik, H.; Narayanan, P.; Mistry, P.K.; Balwani, M. Gaucher disease an Experience from 181 patients in New York. Mol. Genet. Metab. 2021, 132, 44–48. [CrossRef] [PubMed] p p 13. Wenninger, S.; Gutschmidt, K.; Wirner, C.; Einvag, K.; Montagnese, F.; Schoser, B. The impact of interrupting enzyme replacement therapy in late-onset Pompe disease. J. Neurol. 2021, 268, 2943–2950. [CrossRef] [PubMed] py p 14. ESE and Endo-ERN Launch an Initiative to Collect Data on Patients with Rare Endocrine Conditions and COVID-19. Available online: https://endo-ern.eu/ese (accessed on 28 April 2020).
https://openalex.org/W2283593732	https://europepmc.org/articles/pmc4759729?pdf=render	English	null	Comparative speed of kill of sarolaner (Simparica™) and afoxolaner (NexGard®) against induced infestations of Ctenocephalides felis on dogs	Parasites & vectors	2,016	cc-by	4,963	* Correspondence: robert.six@Zoetis.com 1Zoetis, Veterinary Medicine Research and Development, 333 Portage St., Kalamazoo, MI 49007, USA Full list of author information is available at the end of the article © 2016 Six et al. 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. Comparative speed of kill of sarolaner (Simparica™) and afoxolaner (NexGard®) against induced infestations of Ctenocephalides felis on dogs Robert H. Six1, Julian Liebenberg2, Nicole A. Honsberger1 and Sean P. Mahabir1 Robert H. Six1, Julian Liebenberg2, Nicole A. Honsberger1 and Sean P. Mahabir1 Robert H. Six1, Julian Liebenberg2, Nicole A. Honsberger1 and Sean P. Mahabir1 Six et al. Parasites & Vectors (2016) 9:90 DOI 10.1186/s13071-016-1372-1 Six et al. Parasites & Vectors (2016) 9:90 DOI 10.1186/s13071-016-1372-1 RESEARCH Open Access Comparative speed of kill of sarolaner (Simparica™) and afoxolaner (NexGard®) against induced infestations of Ctenocephalides felis on dogs Robert H Six1 Julian Liebenberg2 Nicole A Honsberger1 and Sean P Mahabir1 Animals g py y p pp [ ] Effective control of fleas and their associated condi- tions and diseases is dependent upon the rapid removal of live fleas from the host. Fleas feed almost immediately on attaining a host [10], and direct irritation and allergic reactions are dependent upon the frequency and dur- ation of feeding [11]. Similarly, the longer that a host is exposed to feeding fleas the greater the chance of the transmission of flea-borne pathogens. The use of highly effective, topical and systemic parasiticides has allowed the primary means of flea control to be via direct treatment of the pet. These treatments have largely eliminated the need to treat indoor and outdoor envi- ronments, and their use as host targeted therapies markedly reduces the severity and prevalence of FAD [2]. Orally administered compounds, whether adminis- tered daily like nitenpyram [12] or monthly like spino- sad [13] have been introduced that provide rapid systemic control of fleas for immediate daily treatment, nitenpyram [12], or for up to a month following a sin- gle dose, spinosad [13]. These applications have been readily accepted by veterinarians and dog owners for their efficacy, ease of use and the fact that these sys- temic agents are not topically applied with the resultant lower exposure of the owner or children to potentially toxic residues. Recently, a new class of compounds, the isoxazolines, have shown excellent efficacy against both fleas and ticks for a month or longer following a single orally administered dose [14, 15]. Twenty-four, male and female, purpose-bred mongrel dogs ranging in age from 15 to 105 months and weigh- ing from 7.8 to 21.0 kg were used in the study. All dogs had undergone an adequate wash-out period to ensure that no residual ectoparasiticide efficacy remained from any previously administered compound. Each dog was individually identified by electronic transponder. The dogs were acclimatized to study conditions for a mini- mum of 14 days before treatment on Day 0. Dogs were in- dividually housed in indoor runs such that no physical contact was possible between dogs. Dogs were fed an ap- propriate maintenance ration of a commercial dry canine feed, and water was available ad libitum. All dogs were given a physical examination to evaluate general health and suitability for inclusion into the study. Animals General health observations were performed twice daily from the start of the acclimation period to the end of the study. Design The study followed a randomized complete block design. Dogs were ranked according to decreasing flea counts into blocks of three and within each block a dog was randomly allocated to treatment with placebo, sarolaner, or afoxolaner. There were eight dogs per treatment group. Dogs were infested with fleas prior to treatment and then weekly following treatment for 5 weeks. Flea counts were conducted at 8, 12 and 24 h after treatment and each subsequent weekly re-infestation. Sarolaner (Simparica™, Zoetis) is a new isoxazoline effective against fleas and ticks for at least one month following a single dose. A laboratory study was con- ducted to compare the speed of kill of a single dose of sarolaner (Simparica™, Zoetis) and afoxolaner (NexGard®, Merial) against an existing flea (C. felis) infestation and subsequent re-infestations for a period of 5 weeks after treatment. Abstract Background: Fleas are the most common ectoparasite infesting dogs globally. The many possible sequellae of infestation include: direct discomfort; allergic reactions; and the transmission of pathogens. Rapid speed of kill is an important characteristic for a parasiticide in order to alleviate the direct deleterious effects of fleas, reduce the impact of allergic responses, and break the flea infestation cycle. In this study, the speed of kill of a novel orally administered isoxazoline parasiticide, sarolaner (Simparica™) against fleas on dogs was evaluated and compared with afoxolaner (NexGard®) for 5 weeks after a single oral dose. Methods: Twenty-four dogs were randomly allocated to treatment with a single oral dose at label rate of either sarolaner (2 to 4 mg/kg) or afoxolaner (2.5 to 6.8 mg/kg) or placebo, based on pretreatment flea counts. Dogs were combed and live fleas counted at 8, 12 and 24 h after treatment and subsequent re-infestations on Days 7, 14, 21, 28 and 35. Efficacy was determined at each time point relative to counts for placebo dogs. Results: There were no adverse reactions to treatment. A single oral dose of sarolaner provided ≥98.8 % efficacy (based on geometric means) within 8 h of treatment or subsequent weekly re-infestations of fleas to Day 35. By 12 h, fleas were virtually eradicated from all dogs, with only two fleas recovered from a single sarolaner-treated dog on Day 7; efficacy was 100 % at all other time points. Significantly greater numbers of live fleas were recovered from afoxolaner-treated dogs at 8 h on all days and at 12 h on Days 28 and 35 (P < 0.05). Conclusions: In this controlled laboratory evaluation, sarolaner had a significantly faster speed of kill against fleas than afoxolaner. This was noticeably more evident towards the end of the treatment period. The rapid and consistent kill of fleas within 8 to 12 h after a single oral dose of sarolaner over 35 days indicates that this treatment will provide highly effective control of flea infestations, relief for dogs afflicted with flea allergy dermatitis, and should reduce the risk of flea-borne pathogen transmission. Keywords: Ctenocephalides felis, Sarolaner, Simparica™, Afoxolaner, Speed of kill, Flea, Dog, Oral Page 2 of 6 Six et al. Parasites & Vectors (2016) 9:90 Page 2 of 6 Page 2 of 6 Page 2 of 6 Ethical approval The study was a masked, negative controlled, randomized, laboratory efficacy design conducted in the Republic of South Africa. Procedures were in accordance with the World Association for the Advancement of Veterinary Parasitology (WAAVP) guidelines for evaluating the effi- cacy of parasiticides for the treatment, prevention and control of flea and tick infestation on dogs and cats [16] and complied with the principals of Good Clinical Practices [17]. The protocol was reviewed and approved by the local Institutional Animal Care and Use Committee. Masking of the study was assured through the separation of functions. All personnel conducting observations or animal care or performing infestations and counts were masked to treatment allocation. Methods g Flea infestation of pets and the home are a common oc- currence and their control and elimination can become an expensive and time consuming challenge [1]. The cat flea, Ctenocephalides felis felis, is the most abundant ectoparasite of dogs and cats worldwide [2], and they cause flea allergic dermatitis (FAD), serve as the vector of various bacterial pathogens, and are the intermediate host for filarid and cestode parasites. FAD has been recognized as the most common dermatologic disease of dogs [3]. Ctenocephalides felis has been implicated in the transmis- sion of Rickettsia typhi, Rickettsia felis, Bartonella hense- lae, Mycoplasma haemofelis and, in rare cases, Yersinia pestis [4–7]. Ctenocephalides felis also serves as the inter- mediate host for the nonpathogenic subcutaneous filarid nematode of dogs, Acanthocheilonema (Dipetalonema) reconditum, as well as several species of cestodes includ- ing Dipylidium caninum and Hymenoleptis spp. [1, 8, 9]. Results There were no treatment-related adverse events ob- served during the study. Placebo-treated dogs main- tained good flea infestations throughout the study and these counts were maintained even following the comb- ing and re-infestation procedures at 8 and 12 h (Tables 1, 2 and 3). The C. felis used in the study were from a locally main- tained laboratory colony of European origin. The colony was initiated in 2010 with fleas from Germany; fleas ob- tained from Ireland were introduced into the colony in 2012. Flea infestations were performed on Days -7 (host suitability and allocation), -1, 7, 14, 21, 28 and 35. At each infestation a precounted aliquot of 100 (±5) adult, unfed C. felis were directly applied to the animal which was gently restrained for a few minutes to allow the fleas to penetrate and disperse into the hair coat. Each dog was examined and combed to remove and count fleas at 24 h after the initial host suitability infestation, and at 8 (±1), 12 (±1) and 24 (±2) hours after treatment and each subsequent weekly re-infestation. Fleas were replaced on the dogs immediately after each 8 and 12 h evaluation, and discarded after the 24 h counts. At the 8-h time point, both treatments resulted in sig- nificantly lower flea counts than placebo-treated dogs (P ≤0.0005) throughout the study (Table 1). Treatment with sarolaner also resulted in significantly lower flea counts than afoxolaner at 8 h on all assessment days (P ≤0.0499) and the sarolaner treatment provided greater and more consistent efficacy at 8 h (≥98.8 %, ≥97.9 % geo- metric, arithmetic mean) for the entire study. Efficacy based on geometric means for afoxolaner ranged from 76.3 to 99.4 %, and on arithmetic means from 65.2 to 99.1 % (Table 1). Notably, the efficacy for the afoxolaner- treated dogs declined as the study progressed, and was <90 % from Day 21 onwards (Fig. 1). Flea counts were performed by personnel trained in the standard procedures in use at the test facility. Commercial fine-toothed flea combs were used. Dogs were combed using repeated strokes initially while standing starting from the head, then proceeding caudally along the dor- sum. The dog was then placed on each side and then on its back for combing of the sides and ventral surfaces. Results After a few combing strokes were completed, the comb was examined and hair and fleas were removed from the comb and all live fleas were counted. Each animal was examined for a minimum of 10 min; if any fleas were re- covered in the last minute, combing was continued in one-minute increments until no fleas were detected. At the 12-h timepoint, both treatments resulted in sig- nificantly lower flea counts than placebo-treated dogs (P ≤0.0001) throughout the study (Table 2) and flea counts were significantly higher for afoxolaner-treated dogs on Days 28 and 35 (P ≤0.0161). Efficacy was very high for sarolaner with only two live fleas found on one dog on Day 7 (efficacy of 99.6 %); on all other days efficacy for sarolaner was 100 %. For afoxolaner, 100 % efficacy was only achieved on Day 0, and efficacy tended to decline as the study progressed (Fig. 1). On Days 28 and 35, effi- cacy for afoxolaner-treated dogs was <95 % (arithmetic mean) (Table 2). Statistical analysis The individual dog was the experimental unit and the primary endpoint was the live flea count. Data for post- treatment flea counts were summarized with arithmetic and geometric means by treatment group and timepoint. Flea counts were transformed (loge(count + 1)) prior to analysis in order to stabilize the variance and normalize the data. Using the PROC MIXED procedure (SAS 9.2, Cary NC), transformed counts were analyzed using a mixed linear model. The fixed effects were treatment, timepoint and the interaction between timepoint and treatment by timepoint. The random effects included room, block within room, block by treatment interaction within room, and error. Testing was two-sided at the significance level α = 0.05. At the 24-h time point, both treatments resulted in significantly lower flea counts than placebo-treated dogs (P ≤0.0001) throughout the study (Table 3). Both prod- ucts provided ≥99.8 % reduction in mean live flea counts through Day 35 with 100 % efficacy at most evaluations (Table 3); there was no difference between the counts for the two products on any assessment day (P ≥0.5043). flea counts relative to placebo calculated using Abbott’s formula: flea counts relative to placebo calculated using Abbott’s formula: doses were administered by hand pilling to ensure ac- curate and complete dosing. Each dog was observed for several minutes after dosing for evidence that the dose was swallowed, and for general health at 1, 4 and 24 h after treatment administration. % reduction ¼ 100 mean count placebo ð Þ–mean count treated ð Þ mean count placebo ð Þ % reduction ¼ 100 mean count placebo ð Þ–mean count treated ð Þ mean count placebo ð Þ Treatment Day -1 bodyweights were used to determine the appro- priate dose to be administered. On Day 0, dogs received either a placebo tablet, a sarolaner chewable tablet (Simparica™) to deliver sarolaner at the minimum label dose of 2 mg/kg (range 2 to 4 mg/kg), or NexGard® per label directions (afoxolaner at 2.5 to 6.8 mg/kg). All Six et al. Parasites & Vectors (2016) 9:90 Page 3 of 6 Page 3 of 6 Discussion For any parasiticide product providing flea control, the product’s initial speed of kill is important to provide the pet with rapid relief from the existing infestation. The residual speed of kill is critical to decrease the likelihood of transmission of vector borne disease, ensure that any newly acquired fleas are rapidly eliminated before they can reproduce, and assist in the management of flea The assessment of efficacy was based on the percent reduction in the arithmetic and geometric mean live Page 4 of 6 Six et al. Parasites & Vectors (2016) 9:90 Page 4 of 6 Table 1 Mean live flea counts and efficacy relative to placebo at 8 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 1 Mean live flea counts and efficacy relative to placebo at 8 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 1 Mean live flea counts and efficacy relative to placebo at 8 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 treated with a single oral dose of sarolaner or afoxolaner on Day 0 Treatment Day of treatment or re-infestation 0 7 14 21 28 35 Placebo Range 44–99 51–91 69–89 55–92 62–97 75–100 A. mean 69.9 74.9 77.0 79.8 79.1 87.5 G. mean1 68.1a 73.3a 76.6a 78.8a 78.4a 87.1a Sarolaner Range 0–0 0–3 0–0 0–0 0–8 0–5 A. mean 0.0 0.4 0.0 0.0 1.6 0.6 Efficacy (%) 100 99.5 100 100 97.9 99.3 G. mean1 0.0c 0.2c 0.0c 0.0c 1.0c 0.3c Efficacy (%) 100 99.7 100 100 98.8 99.7 P-value vs. placebo <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Afoxolaner Range 0–2 0–35 0–31 0–44 3–66 0–71 A. mean 0.6 6.1 10.1 15.3 27.5 25.0 Efficacy (%) 99.1 91.8 86.9 80.9 65.2 71.4 G. mean1 0.4b 2.4 b 6.8b 8.5b 18.6b 11.3b Efficacy (%) 99.4 96.7 91.1 89.2 76.3 87.0 P-value vs. placebo <0.0001 <0.0001 <0.0001 <0.0001 0.0005 0.0005 P-value vs. 1Geometric means within columns with the same superscript are not significantly different (P > 0.05) Discussion Parasites & Vectors (2016) 9:90 Table 3 Mean live flea counts and efficacy relative to placebo at 24 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 3 Mean live flea counts and efficacy relative to placebo at 24 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 3 Mean live flea counts and efficacy relative to placebo at 24 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 3 Mean live flea counts and efficacy relative to placebo at 24 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Treatment Day of treatment or re-infestation 0 7 14 21 28 35 Placebo Range 36–87 28–88 48–87 53–73 43–73 58–83 A. mean 56.5 59.3 64.3 66.3 59.8 71.4 G. mean1 54.6a 56.6a 62.7a 66.0a 59.0a 70.8a Sarolaner Range 0–0 0–1 0–0 0–0 0–0 0–0 A. mean 0.0 0.1 0.0 0.0 0.0 0.0 Efficacy (%) 100 99.8 100 100 100 100 G. mean1 0.0b 0.1b 0.0b 0.0b 0.0b 0.0b Efficacy (%) 100 99.8 100 100 100 100 P-value vs.placebo <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Afoxolaner Range 0–0 0–0 0–0 0–0 0–0 0–0 A. mean 0.0 0.0 0.0 0.0 0.0 0.0 Efficacy (%) 100 100 100 100 100 100 G. mean1 0.0b 0.0b 0.0b 0.0b 0.0b 0.0b Efficacy (%) 100 100 100 100 100 100 P-value vs. placebo <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 P-value vs. sarolaner 1.000 0.5043 1.000 1.000 1.000 1.000 1Geometric means within columns with the same superscript are not significantly different (P > 0.05) control of greater than 97 % within 8 h after treatment and weekly re-infestations for 35 days, and fleas were vir- tually eliminated from all dogs within 12 h over the entire 35 days. This rapid efficacy will disrupt the flea life cycle by killing adult fleas before they can lay eggs, thus redu- cing the environmental infestation level [18, 19]. allergic dermatitis. The added benefit for products pre- scribed by a veterinarian is instilling pet owner confidence in the outcome of the treatment and the treatment recom- mendation which can enhance compliance. Discussion sarolaner 0.0499 0.0094 <0.0001 <0.0001 <0.0001 <0.0001 1Geometric means within columns with the same superscript are not significantly different (P > 0.05) Table 2 Mean live flea counts and efficacy relative to placebo at 12 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 2 Mean live flea counts and efficacy relative to placebo at 12 h after treatment and post treatment re-infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Table 2 Mean live flea counts and efficacy relative to placebo at 12 h after treatment and post treatment re infestations for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Treatment Day of treatment or re-infestation 0 7 14 21 28 35 Placebo Range 40–95 33–79 60–89 48–88 56–85 62–95 A. mean 62.3 59.5 74.6 75.8 70.0 80.4 G. mean1 60.4a 57.0a 74.1a 74.7a 69.3a 79.8a Sarolaner Range 0–0 0–2 0–0 0–0 0–0 0–0 A. mean 0.0 0.3 0.0 0.0 0.0 0.0 Efficacy (%) 100 99.6 100 100 100 100 G. mean1 0.0b 0.1b 0.0b 0.0b 0.0c 0.0c Efficacy (%) 100 99.7 100 100 100 100 P-value vs.placebo <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 Afoxolaner Range 0–0 0–2 0–2 0–4 0–28 0–25 A. mean 0.0 0.3 0.3 0.9 5.5 4.3 Efficacy (%) 100 99.6 99.7 98.8 92.1 94.7 G. mean1 0.0b 0.1b 0.1b 0.5b 2.5b 1.4b Efficacy (%) 100 99.7 99.8 99.4 96.4 98.2 P-value vs. placebo <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 P-value vs. sarolaner 1.000 1.000 0.2455 0.0713 0.0008 0.0161 1Geometric means within columns with the same superscript are not significantly different (P > 0.05) Page 5 of 6 Page 5 of 6 Six et al. Discussion A single oral dose of sarolaner at the proposed commercial dose of 2 to 4 mg/kg resulted in the rapid reduction of an existing flea infestation as well as rapid kill of newly infested fleas for 35 days. Treatment with sarolaner resulted in effective The efficacy provided by sarolaner at 8 h was signifi- cantly greater than that provided by afoxolaner at all Fig. 1 Percent efficacy based on geometric mean counts relative to placebo at 8 and 12 h after treatment and weekly post treatment re-infestations of fleas for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Fig. 1 Percent efficacy based on geometric mean counts relative to placebo at 8 and 12 h after treatment and weekly post treatment re-infestations of fleas for dogs treated with a single oral dose of sarolaner or afoxolaner on Day 0 Six et al. Parasites & Vectors (2016) 9:90 Page 6 of 6 evaluations but especially when performance is com- pared towards the end of the monthly treatment period when the efficacy of afoxolaner declined below 90 % from Day 21 onwards. Efficacy of afoxolaner was im- proved at 12 h and again sarolaner had consistently bet- ter kill of fleas and there were significantly more live fleas recovered from the afoxolaner-treated dogs on Days 28 and 35; sarolaner was 100 % effective on these days. There was no difference between sarolaner and afoxolaner at the 24 h assessments. The results for afox- olaner in this study are in agreement with those reported by Hunter et al. [20] who reported 99.6 % efficacy at 12 h and 100 % efficacy at 24 h against an existing in- festation and efficacy ranging from 89.7 to 99.5 % at 12 h and ≥99.9 % efficacy at 24 h for 35 days. Received: 4 December 2015 Accepted: 16 January 2016 Received: 4 December 2015 Accepted: 16 January 2016 Competing interests Th d f d d 18. Jacobs DE, Hutchinson MJ, Ryan WG. Control of flea populations in a simulated home environment model using lufenuron, imidacloprid or fipronil. Med Vet Entomol. 2001;25:73–7. The study was funded by Zoetis. RHS, NAH and SPM are current employees of Zoetis. JL is a current employee of ClinVet which was contracted by Zoetis to conduct the study. 19. Dryden MW, Payne PA, Lowe A, Mailen S, Smith V, Rugg D. Efficacy of a topically applied formulation of metaflumizone on cats against the adult cat flea, flea egg production and hatch, and adult flea emergence. Vet Parasitol. 2007;150:263–7. References l b Effects on the development of Dipylidium caninum and on the host reaction to this parasite in the adult flea (Ctenocephatides felis felis). Parasitol Res. 1987;73:171–7. 9. Marshall AG. The cat flea, Ctenocephalides felis felis (Bouche, 1835) as an intermediate host for cestodes. Parasitology. 1967;57:419–30. Acknowledgements The authors would like to thank Douglas Rugg for his assistance in preparing this manuscript. Conclusions A randomized, blinded, controlled and multi-centered field study comparing the efficacy and safety of Bravecto® (fluralaner) against Frontline® (fipronil) in flea- and tick-infested dogs. Parasitol Vectors. 2014;7:83. 15. Shoop WL, Harline EJ, Gould BR, Waddell ME, McDowell RG, Kinney JB, et al. Discovery and mode of action of afoxolaner, a newisoxazoline parasiticide for dogs. Vet Parasitol. 2014;201:179–89. 15. Shoop WL, Harline EJ, Gould BR, Waddell ME, McDowell RG, Kinney JB, et al. Discovery and mode of action of afoxolaner, a newisoxazoline parasiticide for dogs. Vet Parasitol. 2014;201:179–89. 16. Marchiondo AA, Holdsworth PA, Fourie LJ, Rugg D, Hellmann K, Snyder DE, et al. World Association for the Advancement of Veterinary Parasitology (W. A.A.V.P.) 2nd. ed.: Guidelines for evaluating the efficacy of parasiticides for the treatment, prevention and control of flea and tick infestations on dogs and cats. Vet Parasitol. 2013;194:84–97. 17. EMEA. Guideline on good clinical practices. VICH Topic GL9. http://www. ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/10/ WC500004343.pdf. Accessed 23 Aug 2015. Authors’ contributions All authors participated in study and protocol design and reviewed and approved the manuscript. JL conducted the study which was monitored by NH. SPM conducted the statistical analyses. 20. Hunter JS, Dumont P, Chester TS, Young DR, Fourie JJ, Larsen DL. Evaluation of the curative and preventive efficacy of a single oral administration of afoxolaner against cat flea Ctenocephalides felis infestations on dogs. Vet Parasitol. 2014;201:207–11. References l b 1. Blagburn BL, Dryden MW. Biology, treatment and control of flea and tick infestations. Vet Clin Small Anim. 2009;39:1173–200. 1. Blagburn BL, Dryden MW. Biology, treatment and control of flea and tick infestations. Vet Clin Small Anim. 2009;39:1173–200. 2. Rust MK. Advances in the control of Ctenocephalides felis felis (cat flea) on cats and dogs. Trends Parasitol. 2005;21:232–6. 2. Rust MK. Advances in the control of Ctenocephalides felis felis (cat flea) on cats and dogs. Trends Parasitol. 2005;21:232–6. 3. Scott DW, Miller WH, Griffin CE. Muller and Kirk’s small animal dermatology. Philadelphia: Saunders; 2001. p. 626. 3. Scott DW, Miller WH, Griffin CE. Muller and Kirk’s small animal dermatology. Philadelphia: Saunders; 2001. p. 626. 4. Breitschwerdt EB. Feline bartonellosis and cat scratch disease. Vet Immunol Immunopathol. 2008;123:167–71. 5. Kamrani A, Parreira VR, Greenwood J, Prescott J. The prevalence of Bartonella, hemoplasma, and Rickettsia felis infections in domestic cats and in cat fleas in Ontario. Can J Vet Res. 2008;72:411–9. 5. Kamrani A, Parreira VR, Greenwood J, Prescott J. The prevalence of Bartonella, hemoplasma, and Rickettsia felis infections in domestic cats and in cat fleas in Ontario. Can J Vet Res. 2008;72:411–9. 6. Brewer MM, Woods JE, Hawley JR, Wisnewski N, Lappin M. Evaluation of experimental transmission of Candidatus Mycoplasma haemominutum and Mycoplasma haemofelis by Ctenocephalides felis to cats. Am J Vet Res. 2005;66:1008–12. 6. Brewer MM, Woods JE, Hawley JR, Wisnewski N, Lappin M. Evaluation of experimental transmission of Candidatus Mycoplasma haemominutum and Mycoplasma haemofelis by Ctenocephalides felis to cats. Am J Vet Res. 2005;66:1008–12. 7. Eisen RJ, Borchert JN, Holmes JL, Amatre G, Van Wyk K, Enscore RE, et al. Early-phase transmission of Yersinia pestis by cat fleas (Ctenocephalides felis) and their potential role as vectors in a plague-endemic region of Uganda. Am J Trop Med Hyg. 2008;78:949–56. 7. Eisen RJ, Borchert JN, Holmes JL, Amatre G, Van Wyk K, Enscore RE, et al. Early-phase transmission of Yersinia pestis by cat fleas (Ctenocephalides felis) and their potential role as vectors in a plague-endemic region of Uganda. Am J Trop Med Hyg. 2008;78:949–56. The rapid and consistent speed of kill of fleas over a period of 35 days makes sarolaner an excellent option for monthly flea control that will reduce the direct irrita- tion caused by flea infestation, assist in the prevention of FAD, and should reduce the risk of flea-borne diseases. 8. Pugh RE. Conclusions 10. Dryden MW, Gaafar SM. Blood consumption by the cat flea, Ctenocephalides felis (Siphonaptera: Pulicidae). J Med Entomol. 1991;28:394–400. 10. Dryden MW, Gaafar SM. Blood consumption by the cat flea, Ctenocephalides felis (Siphonaptera: Pulicidae). J Med Entomol. 1991;28:394–400. A single oral dose of sarolaner at 2 to 4 mg/kg resulted in more rapid control of an existing flea infestation than afoxolaner with >97 % efficacy within 8 h of both treat- ment and subsequent re-infestations for 35 days. Fleas were virtually eliminated from all dogs within 12 h. Effi- cacy of sarolaner at 8 h was significantly superior to afoxolaner on all days, and at 12 h on Days 28 and 35. Afoxolaner’s efficacy at 12 h decreased towards the end of the month long treatment interval. The rapid and consistent speed of kill provided by sarolaner over the 5 weeks following a single dose can provide peace of mind for veterinarians and pet owners that there will be no gap in protection against fleas when sarolaner is dosed monthly, will reduce the deleterious effects of flea infest- ation and the clinical signs of flea allergy dermatitis, and should reduce the risk of flea-borne pathogen transmission. 11. Dryden MW. Flea and tick control in the 21st century: challenges and opportunities. Vet Dermatol. 2009;20:435–40. 11. Dryden MW. Flea and tick control in the 21st century: challenges and opportunities. Vet Dermatol. 2009;20:435–40. 12. Rust MK, Waggoner MM, Hinkle NC, Stansfield D, Barnett S. Efficacy and longevity of nitenpyram against adult cat fleas (Siphonaptera: Pulicidae). J Med Entomol. 2003;40:678–81. 12. Rust MK, Waggoner MM, Hinkle NC, Stansfield D, Barnett S. Efficacy and longevity of nitenpyram against adult cat fleas (Siphonaptera: Pulicidae). J Med Entomol. 2003;40:678–81. 13. Snyder DE, Meyer J, Zimmermann AG, Qiao M, Gissendanner SJ, Cruthers LR, et al. Preliminary studies on the effectiveness of the novel pulicide, spinosad, for the treatment and control of fleas on dogs. Vet Parasitol. 2007;150:345–51. 13. Snyder DE, Meyer J, Zimmermann AG, Qiao M, Gissendanner SJ, Cruthers LR, et al. Preliminary studies on the effectiveness of the novel pulicide, spinosad, for the treatment and control of fleas on dogs. Vet Parasitol. 2007;150:345–51. 14. Rohdich N, Roepke RKA, Zschiesche E. A randomized, blinded, controlled and multi-centered field study comparing the efficacy and safety of Bravecto® (fluralaner) against Frontline® (fipronil) in flea- and tick-infested dogs. Parasitol Vectors. 2014;7:83. 14. Rohdich N, Roepke RKA, Zschiesche E. 18. Jacobs DE, Hutchinson MJ, Ryan WG. Control of flea populations in a simulated home environment model using lufenuron, imidacloprid or fipronil. Med Vet Entomol. 2001;25:73–7. 17. EMEA. Guideline on good clinical practices. VICH Topic GL9. http://www. ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/10/ WC500004343.pdf. Accessed 23 Aug 2015. Author details 1 1Zoetis, Veterinary Medicine Research and Development, 333 Portage St., Kalamazoo, MI 49007, USA. 2ClinVet International (pty) Ltd, Uitsigweg, Bainsvlei 9338, Bloemfontein, Republic of South Africa. 1Zoetis, Veterinary Medicine Research and Development, 333 Portage St., Kalamazoo, MI 49007, USA. 2ClinVet International (pty) Ltd, Uitsigweg, Bainsvlei 9338, Bloemfontein, Republic of South Africa.
https://openalex.org/W3110522641	https://iris.unipa.it/bitstream/10447/564592/2/10-1055-s-0040-1719119.pdf	English	null	The Impact of Narrow-band Imaging on the Pre- and Intra- operative Assessments of Neoplastic and Preneoplastic Laryngeal Lesions. A Systematic Review	International Archives of Otorhinolaryngology	2,020	cc-by	5,926	DOI https://doi.org/ 10.1055/s-0040-1719119. ISSN 1809-9777. Published online: 2020-11-30 Published online: 2020-11-30 Published online: 2020-11-30 THIEME Systematic Review 4 471 Carmelo Saraniti1 Enzo Chianetta1 Giuseppe Greco1 Norhaﬁza Mat Lazim2 Barbara Verro1 Enzo Chianetta1 Giuseppe Greco1 Norhaﬁza Mat Lazim2 Barbara Verro1 Carmelo Saraniti1 Enzo Chianetta1 Giuseppe Greco1 Norhaﬁza Mat Lazim Address for correspondence Carmelo Saraniti, MD, PhD, Department of Biomedicine, Neurosciences and Advanced Diagnostics, ENT Clinic, Università degli Studi di Palermo, Via del Vespro, 129, Palermo, 90127, Italy (e-mail: carmelo.saraniti@unipa.it). 1Department of Biomedicine, Neurosciences and Advanced Diagnostics, ENT Clinic, Università degli Studi di Palermo, Palermo, Italy 2Department of Otorhinolaryngology-Head and Neck Surgery, School of Medical Sciences, Universiti Sains Malaysia, Kubang Kerian, Kelantan, Malaysia 1Department of Biomedicine, Neurosciences and Advanced Diagnostics, ENT Clinic, Università degli Studi di Palermo, Palermo, Italy 2Department of Otorhinolaryngology-Head and Neck Surgery, School of Medical Sciences, Universiti Sains Malaysia, Kubang Kerian, Kelantan, Malaysia Int Arch Otorhinolaryngol 2021;25(3):e471–e478. Thieme Revinter Publicações Ltda., Rua do Matoso 170, Rio de Janeiro, RJ, CEP 20270-135, Brazil © 2020. Fundação Otorrinolaringologia. All rights reserved. This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial-License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/ licenses/by-nc-nd/4.0/) Key Points y Narrow-band imaging is useful in detecting suspicious laryngeal lesions and proper resection margins showing intraepithelial papillary capillary loops (IPCLs) that are considered a main cancer feature. Narrow-band imaging is used both pre- and intraoperatively, but it provides more precise information if used during Narrow-band imaging is useful in detecting suspicious laryngeal lesions and proper resection margins showing intraepithelial papillary capillary loops (IPCLs) that are considered a main cancer feature. Narrow-band imaging is used both pre- and intraoperatively, but it provides more precise information if used during surgery. Compared with white-light endoscopy, narrow-band imaging enables a better assessment of the lesions covered by a thick white plaque (such as in cases of leukoplakia) The classiﬁcation of the European Laryngological Society is the simplest and the most practical for the identiﬁcation of various laryngeal lesions compared with other classiﬁcations. on sensitivity (SE), speciﬁcity (SP), positive and negative predictive values (PPVs and NPVs) and accuracy (ACC). Search Methodology Using the Preferred Reporting Items for Systematic Reviewand Meta-Analysis (PRISMA) guidelines, two independent authors (CS, BV) analyzed the literature published on the PubMed, Scopus and Web of Science databases (up to March 1st, 2020). The keywords “narrow-band imaging” or “NBI” and “larynx cancer” or “glottic” or “larynx” were used to detect the relevant articles. The two researchers independently selected articles by reading titles and abstracts; then,the included articles were read entirely. Moreover, some articles were extracted from references mentioned in the previously selected articles. Eligibility Criteria The inclusioncriteriawere originalarticlesdescribing research that performed NBI in patients with larynx lesions; studies that reported accuracy, sensitivity, speciﬁcity, PPVs and NPVs; and studies published in English. The exclusion criteria were studies that reported lesions involving sites other than larynx; reviews, editorials, com- mentaries, thesis and conference abstracts; studies that calculated statistical data of WL combined with NBI; studies that included less than 40 lesions. Statistical Data Analysis In the case of lack of statistical data in some articles,3–7 we have calculated them by using the Prism GraphPad software (GraphPad Software, San Diego, CA, US). Study Selection and Characteristics The process of selection of articles is shown in ►Fig. 1. In total, 638 articles were identiﬁed through the search in the databases and references. After removing the duplicates and reading titles and abstracts, 48 articles were selected. Then, according to eligibility criteria, only 14 of the full-text articles were included in the present review. Abstract Introduction Narrow-band imaging is an endoscopic diagnostic tool that, focusing on superﬁcial vascular changes, is useful to detect suspicious laryngeal lesions, enabling their complete excision with safe and tailored resection margins. Objectives To analyze the applications and beneﬁts of narrow-band imaging in detecting premalignant and malignant laryngeal lesions through a comparison with white-light endoscopy. Data Synthesis A literature search was performed in the PubMed, Scopus and Web of Science databases using strict keywords. Then, two authors independently analyzed the articles, read the titles and abstracts, and read completely only the relevant studies according to certain eligibility criteria. In total, 14 articles have been included in the present review; the sensitivity, speciﬁcity, positive and negative predictive values, and accuracy of pre- and/or intraoperative narrow-band imaging were analyzed. The analysis showed that narrow-band imaging is better than white-light endoscopy in terms of sensitivity, speciﬁcity, positive and negative predictive values, and accuracy regarding the ability to identify cancer and/or precancerous laryngeal lesions. More- over, the intraoperative performance of narrow-band imaging resulted more effective than the in-ofﬁce performance. Conclusion Narrow-band imaging is an effective diagnostic tool to detect premalig- nant and malignant laryngeal lesions and to deﬁne proper resection margins. Moreover, narrow-band imaging is useful in cases of leukoplakia that may cover a possible malignant lesion and that cannot be easily assessed with white-light endosco- py. Finally, a shared, simple and practical classiﬁcation of laryngeal lesions, such as that of the European Laryngological Society, is required to identify a shared lesion management strategy. Keywords ►narrow-band imaging ►early diagnosis ►larynx cancer received June 5, 2020 accepted September 27, 2020 published online November 30, 2020 received June 5, 2020 accepted September 27, 2020 published online November 30, 2020 Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 472 Key Points Narrow-band imaging is useful in detecting suspicious laryngeal lesions and proper resection margins showing intraepithelial papillary capillary loops (IPCLs) that are considered a main cancer feature. Narrow-band imaging is used both pre- and intraoperatively, but it provides more precise information if used during surgery. Compared with white-light endoscopy, narrow-band imaging enables a better assessment of the lesions covered by a thick white plaque (such as in cases of leukoplakia) The classiﬁcation of the European Laryngological Society is the simplest and the most practical for the identiﬁcation of various laryngeal lesions compared with other classiﬁcations. Introduction Laryngeal squamous cell carcinoma (LSCC) is one of the most frequent head and neck cancers. Early diagnosis and radical excision represent the main goals. Indeed, the challenge for laryngeal cancer is to perform its tailored resection with proper safe margins. Until now, transoral laser microsurgery (TLM) has been the gold standard for the treatment of early glottic cancer, since it enables a tailored and customized resection, ensuring functional preservation as well. Moreover, otolaryngologists can use a quite recent diagnostic tool to achieve this target: narrow-band imaging (NBI), which is an optical technique ﬁrst used in gastroen- terology and then introduced in laryngology in 2004.1 This endoscopic tool is based on narrow-band optical ﬁlters with two ranges of wavelength that are signiﬁcantly absorbed by hemoglobin. In particular, the ﬁrst peak wavelength is of 415 nm (range: 400 nm to 430 nm) with blue light that enhances the visualization of capillary vessels on the su- perﬁcial layer of the mucosa. The second narrow-band light has a peak wavelength of 540 nm (range: 525 nm to 555 nm), and it is a green light that ensures a better visualization of the submucosal vascular pattern. Actually, vascular changes and neoangiogenesis are the main cancer features that can be better detected with NBI endoscopy than with white-light (WL) endoscopy. In particular, accord- ing to the current literature, any well-demarcated brownish areas and/or thick dark spots and/or winding and hypertro- phic vascular patterns can be considered suspicious NBI lesions.2 However, a consensus classiﬁcation based on intraepithelial papillary capillary loop (IPCL) patterns is required to properly assess these lesions. Therefore, NBI endoscopy represents the so-called “optical biopsy”,1 be- cause it provides a better distinction between benign and malignant laryngeal lesions, and, therefore, a more appro- priate management of the patient, even before the biopsy and histological assessment, which are the gold standard for tumor diagnosis. Thus, NBI is a useful tool not only in the preoperative phase for early diagnosis, but also during surgery, for a proper and effective tumor resection, and during the follow-up for the early detection of a possible recurrence. Therefore the aim of the present review is to analyze the International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. Preoperative Performance of NBI Therefore, the aim of the present review is to analyze the applications and advantages of NBI compared with WL endoscopy in cases of malignant laryngeal lesions, focusing Six of the selected articles reported the preoperative perfor- mance of NBI (►Table 1). These studies included patients affected by premalignant and malignant laryngeal lesions, Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 473 Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 474 Table 1 Characteristics of the studies that analyzed the preoperative performance of NBI Authors Rzepakowska et al.8 Rzepakowska et al.10 Šifrer et al.9 Volgger et al.3 Staníková et al.6 Hosono et al.11 Number of assessed lesions 91 105 49 41 63 166 Type of lesion Leukoplakia Laryngeal lesions Laryngeal lesions Laryngeal lesions Leukoplakia Laryngeal lesions IPCL classiﬁcation Ni et al.15 Ni et al.15 ELS Ni et al.15 Ni et al.15 None SE of NBI (%) 100% 98.8% 100% 100% 88% 84.4% SP of NBI (%) 97.4% 73.9% 95% 79.4% 89.5% 88.6% PPV of NBI (%) 87.5% 93.1% 88% 50% 85% 91% NPV of NBI (%) 100% 94.4% 100% 100% 92% 80.5% ACC of NBI (%) 97.8% 93.3% 96% 82.9% 88.9% 86.1% Table 1 Characteristics of the studies that analyzed the preoperative performance of NBI Abbreviations: ACC, accuracy; ELS, European Laryngological Society; IPCL, intraepithelial papillary capillary loop; NBI, narrow-band imaging; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, speciﬁcity. Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI (the data calculated by the reviewers are reported in italics). Table 2 Characteristics of the studies that analyzed the intra- operative performance of NBI Authors Šifrer et al.12 Lukes et al.7 Klimza et al.5 Number of assessed lesions 191 78 90 Type of lesion Laryngeal lesions Laryngeal lesions T1-T2 glottic cancer IPCL classiﬁcation ELS None None SE of NBI (%) 98% 100% 100% SP of NBI (%) 95% 82% 0% PPV of NBI (%) 88% 85% 85.7% NPV of NBI (%) 99% 100% Not able to evaluate ACC of NBI (%) 95% 91% 85.7% Abbreviations: ACC, accuracy; ELS, European Laryngological Society; IPCL, intraepithelial papillary capillary loop; NBI, narrow-band imaging; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, speciﬁcity. Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI (the data calculated by the reviewers are reported in italics). Preoperative NBI versus Intraoperative NBI p p Two of the selected articles13,14 compared the results of in- ofﬁce and intraoperative NBI in the evaluation of laryngeal lesions (►Table 3). De Vito e al.13 performed a prospective study on 73 vocal-cord lesions. I Regarding the detection of premalignant and malignant lesions, the preoperative per- formance of NBI showed an SE of 97%, an SP of 92.5%, a PPV of 91.4%, an NPV of 97.4%, and an ACC of 94.5%. Meanwhile, the intraoperative NBI evaluation enabled the authors to achieve higher speciﬁcity (95%), PPV (94.1%) and accuracy (95.9%) compared with in-ofﬁce NBI. On the other hand, the SE and NPV were similar for the pre- and intra- operative NBI. Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI (the data calculated by the reviewers are reported in italics). Piazza et al.14 performed a similar study protocol on 279 lesions highly suspicious for carcinoma. Therefore, they performed NBI endoscopy before and during surgery, getting better statistical results in the latter case. another article, Rzepakowska et al.10 analyzed 105 suspected laryngeal lesions with NBI to detect malignancy, and they found a high ability of the NBI to identify malignant lesions, with an SE of 98.8%, an SP of 73.9%, a PPV of 93.1%, an NPV of 94.4%, and an ACC of 93.3%. NBI versus WL Five of the included studies4,5,14–16 compared the statistical results of WL and NBI (►Table 4). Klimza et al.5 analyzed 90 T1-T2 glottic cancers performing intraoperative WL and NBI. They found that NBI proved to have higher SE (100% versus 79.5%), PPV (85.7% versus 83.3%) and ACC (85.7% versus 71.1%) than WL regarding the endoscopic role to detect proper resection margins during TLM. The same trend of statistical results was observed in the article by Piazza et al.,14 in which they also made a comparison between intraoperative WL and NBI. Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. Table 2 Characteristics of the studies that analyzed the intra- operative performance of NBI Table 2 Characteristics of the studies that analyzed the intra- operative performance of NBI PPV of 85%, an NPV of 100%, and an ACC of 91%. Only seven lesions were found to be false positive. Šifrer et al.12performed rigid NBI during direct laryngoscopy in191laryngeallesionsandachievedstatisticalvaluessimilar to those of Lukes et al.:7 SE of 98%, SP of 95%, PPV of 88%, NPV of 99%, and ACC of 95%. Actually, performing NBI during surgery, they detected 78 malignant lesions, 100 benign lesions, and 13 discrepancies between the NBI and the histology. Authors Šifrer et al.12 Lukes et al.7 Klimza et al.5 Number of assessed lesions 191 78 90 Type of lesion Laryngeal lesions Laryngeal lesions T1-T2 glottic cancer IPCL classiﬁcation ELS None None SE of NBI (%) 98% 100% 100% SP of NBI (%) 95% 82% 0% PPV of NBI (%) 88% 85% 85.7% NPV of NBI (%) 99% 100% Not able to evaluate ACC of NBI (%) 95% 91% 85.7% RISMA 2009 ﬂow diagram of the selection process of articles from the current literature. Fig. 1 PRISMA 2009 ﬂow diagram of the selection process of articles from the current literature. Staníková et al.6 included 63 cases of leukoplakia on the vocal cords in their article. Comparing the preoperative NBI assessments with the histological results, they achieved rates of SE, SP, ACC, PPVs and NPVs of NBI of 88%, 89.5%, 88.9%, 85% and 92% respectively. such as papilloma, leukoplakia and SCC, which were treated with TLM. In these cases, the preoperative performance of NBI has ensured high sensitivity, speciﬁcity, accuracy, PPVs and NPVs in the diagnosis of malignancy. Furthermore, other authors3,9–11 studied the preopera- tive performance of NBI in all laryngeal lesions, including benign ones. In particular, Šifrer et al.9 performed NBI endoscopy on 49 laryngeal lesions, 26 of which were benign, and 23 of which were malignant after the histological evaluation. In this study, the NBI showed an SE of 100%, an SP of 95%, a PPV of 88%, an NPV of 88% and an ACC of 96%. In In particular, Rzepakowska et al.8 reported 91 cases of vocal cordleukoplakia,andtheydemonstratedthat thepreoperative performance of NBI enabled the detection of 75/77 cases of benign lesions and all cases of malignant ones, with a PPV of 87.5% and an NPVof 100%. Moreover, in this study, NBI showed a sensitivity of 100%, a speciﬁcity of 97.4%, and an accuracy of 97.8% in predicting malignancy in cases of leukoplakia. Intraoperative Performance of NBI Only a few authors5,7,12 analyzed the results of the perfor- mance of NBI during direct laryngoscopies (►Table 2). In particular, Lukes et al.7 studied 78 laryngeal lesions, clearly excluding the benign ones. The intraoperative perfor- mance of NBI enabled the identiﬁcation of 39 LSCCs and 32 papillomas which were then conﬁrmed by the histological assessment. Thus, they found an SE of 100%, an SP of 82%, a International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 475 Table 3 Characteristics of the studies that compared the pre- and intraoperative performances of NBI Authors De Vito et al.13 Piazza et al.14 Number of assessed lesions 73 279 Type of lesion Laryngeal lesions Laryngeal lesions IPCL classiﬁcation Ni et al.15 None SE of preoperative NBI (%) 97% 61% SE of intraoperative NBI (%) 97% 98% SE of WL (%) 33% (intraoperative) SP of preoperative NBI (%) 92.5% 87% SP of intraoperative NBI (%) 95% 90% SP of WL (%) 95% (intraoperative) PPV of preoperative NBI (%) 91.4% 83% PPV of intraoperative NBI (%) 94.1% 86% PPV of WL (%) 85% (intraoperative) NPV of preoperative NBI (%) 97.4% 88% NPV of intraoperative NBI (%) 97.4% 88% NPV of WL (%) 100% (intraoperative) ACC of preoperative NBI (%) 94.5% 76% ACC of intraoperative NBI (%) 95.9% 92% ACC of WL (%) 64% (intraoperative) cial vascular changes (IPCLs). It is often described as “optical biopsy”,1 since it is useful to differentiate various laryngeal lesions and to provide a very reliable diagnosis – almost like the histological assessment –, which enables a proper management of the lesion. Analyzing all the literature since its ﬁrst use, dwelling on statistical results, we found the usefulness and beneﬁts of NBI in the detection of precan- cerous and cancerous lesions, as well as in identifying the proper resection margins during surgery,13,14 as proven by its high SE, SP and ACC rates. In fact, due to its ability to highlight alterations in the mucosal vascularization, it enables physicians to deﬁne more precisely the limit be- tween healthy and pathological tissue and perform a com- plete and tailored resection, which is more difﬁcult to obtain with WL.3 In addition, based on this analysis, the intraoperative NBI results were better than the preoperative ones. Preoperative NBI versus Intraoperative NBI Preoperative NBI versus Intraoperative NBI As stated before, NBI is used not only in-ofﬁce but also during surgery. In particular, the performance of preopera- tive NBI is useful in the screening and follow-up of suspi- cious laryngeal lesions, to reduce the number of patients who undergo surgery. In contrast, the intraoperative NBI results were useful especially in the detection of proper resection margins.5,14 Indeed, complete tumor excision in TLM represents one of the main issues that may impair the local control of the disease. Thus, intraoperative NBI could be useful because it enables the detection of the superﬁcial spread of the cancer, as well as the performance of a proper and customized tumor resection.17 Thereby, as reported also by Klimza et al.,5 the need for a second-look procedure may be reduced by the systematic performance of intra- operative NBI. Abbreviations: ACC, accuracy; ELS, European Laryngological Society; IPCL, intraepithelial papillary capillary loop; NBI, narrow-band imaging; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, speciﬁcity; WL, white-light endoscopy. Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI performed pre-operatively and intra-operatively. Concerning this, referring to the selected articles, De Vito et al.13 reported a slightly better efﬁciency of intraoperative NBI compared with in-ofﬁce NBI, although the results were not statistically signiﬁcant. It is possible to interpret the result obtained by De Vito et al.13 by analyzing the differ- ences between the pre- and intraoperative performance of NBI. In fact, in the ﬁrst case, it is an in-ofﬁce procedure that is repeatable, with minimal execution time and costs, but burdened by the persistence of oropharyngeal reﬂexes that can reduce the quality of the images. In contrast, intra- operative NBI has higher costs and execution time, but, at the same time, it enables a closer and magniﬁed observation of the superﬁcial spread thanks to general anesthesia and to the use of rigid telescopes with different angles of visualiza- tion. Furthermore, it is necessary to underline that all authors of the studies included in the review used a rigid endoscope in the intraoperative phase,5,7,12–14 while almost all of them used a ﬂexible endoscope in the preoperative phase.3–6,8,10,13–16 However, in none of the included studies, Ni et al.15,16 performed two studies in which they com- pared the two endoscopic tools in the preoperative phase. Preoperative NBI versus Intraoperative NBI In 2011, they studied premalignant and malignant laryngeal lesions,15 and, in 2018, they focused just on vocal-cord leukoplakia.16 In both studies, the authors reported better results of NBI with regards to SE, SP, PPV, NPV and ACC. In the article written by Popek et al.,4 333 laryngeal lesions, including benign ones, were analyzed in-ofﬁce with both WL and NBI. They achieved higher values of SE (98.5% versus 95.4%), SP (98.5% versus 84.2%), PPV (97.7% versus 79.6%), NPV (99% versus 96.6%) and ACC (98.5% versus 88.3%) for the NBI in comparison with WL. Intraoperative Performance of NBI Furthermore, some studies4,5,14–16 stressed not only the overall beneﬁts of NBI, but also its remarkable advan- tages compared with WL. International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. NBI versus WL sa. However, to provide a consensus assessment of various vascular patterns, a shared classiﬁcation is required. Until now, the classiﬁcation of the morphologic endoscopic char- acteristics of IPCLs developed by Ni et al.15 is the most widely used. This classiﬁcation identiﬁes ﬁve different types of IPCL changes. Thus, type I is characterized by thin, oblique and arborescent vessels without IPCLs. In type II, the IPCLs are not visible and, the oblique and arborescent vessels have larger diameters. Type III consists of white mucosa that covers the IPCLs preventing their visualization. In type IV, the IPCLs are visible as dark brown spots. Type V, in turn, is divided into three subtypes. In type Va, the IPCLs appear signiﬁcantly enlarged, with brownish speckles and various shapes. In type Vb, the IPCLs have irregular and snake-like shapes. Type Vc is characterized by necrotic tissue on the surface of the lesion and by tortuous, irregular and line-like or brownish speckled IPCLs. Accord- ing to this classiﬁcation, type V corresponds to high-grade dysplasia and cancer, type IV suggests a suspicious lesion, whereas types I, II and III are related to benign lesions. To date, WL endoscopy is the most widely used diagnostic tool in the assessment of premalignant lesions or cancer of the larynx. However, WL sometimes fails in identifying leukoplakia or early and slight mucosal changes in the vocal cords. Indeed, a thick white patch may cover an underlying malignant lesion and impair the diagnosis, leading to a false negative result.13,15 Moreover, neoangiogenesis and vascular changes represent one of the ﬁrst alterations in carcinogen- esis that WL laryngoscopy cannot identify.18 Thereby, in this scenario, NBI makes the difference in better identifying benign and malignant lesions. Actually, NBI enables the study of mucosal and submucosal vascular patterns and the clas- siﬁcation of any changes in the IPCLs.10,16 Thus, regarding the issue of white mucosal thickening, the surrounding mucosa should be analyzed for any vascular abnormalities with NBI.5,15,19 Indeed, in the included studies4,5,14–16 reported that the SE and NPV of NBI were signiﬁcantly greater than those of WL. In another study, Ni et al.16 reported that NBI was better than WL, especially concerning SP and PPV. These results are probably due to the fact that, in the latter study, the authors analyzed only one type of lesion, which is leukoplakia, proving mainly the ability of NBI to detect any malignant lesion covered by a white patch. Summary of the Main Results Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI and WL, by specifying if they were performed preoperatively and intraoperatively [the data calculated by the reviewers are reported in italics]. NBI versus WL On the contrary, the other studies examined various laryngeal lesions, and the authors demonstrated the usefulness of NBI in distinguishing benign and malignant lesions. However, in 2015, the European Laryngological Society (ELS) proposed a new classiﬁcation.20 The main objective of the ELS was to provide a simpler and practical strategy. This new classiﬁcation consists of only two vascular patterns: the longitudinal and perpendicular patterns. The former is characterized by vascular changes that remain in two dimensions (length and width) of the vocal folds, such as ectasia, increased vessel density or increased vessel branches, and it is related to benign lesions. Meanwhile, in the perpendicular pattern, the vessels grow in the third dimension, and represent the IPCLs. This pattern could correspond to recurrent respiratory papillomatosis (RRP) or premalignant or malignant lesions. Therefore, to Therefore, NBI is a more effective diagnostic tool than WL, mainly in cases of detection of early cancer or leukoplakia. nternational Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. Summary of the Main Results the effectiveness of NBI in terms of SE, SP, PPV, NPVand ACC is analyzed according to the type of endoscope used (rigid or ﬂexible). the effectiveness of NBI in terms of SE, SP, PPV, NPVand ACC is analyzed according to the type of endoscope used (rigid or ﬂexible). Narrow-band imaging is a recent endoscopic tool used in laryngeal oncology to detect any suspicious lesion and to provide a customized tumor resection, focusing on superﬁ- Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 476 Table 4 Characteristics of the studies that compared NBI and WL Authors Klimza et al.5 Piazza et al.14 Ni et al.16 Ni et al.15 Popek et al.4 Number of assessed lesions 90 279 120 104 333 Type of lesion T1-T2 glottic cancer Laryngeal lesions Leukoplakia Laryngeal lesions Laryngeal lesions Pre- or intraoperative NBI Preoperative Intraoperative Preoperative Preoperative Preoperative IPCL classiﬁcation None None Ni et al.15 Ni et al.15 Ni et al.15 SE of NBI (%) 100% 98% 82.6% 88.9% 98.5% SE of WL (%) 79.5% 33% 60.8% 68.9% 95.4% SP of NBI (%) 0% 90% 92.8% 93.2% 98.5% SP of WL (%) 20% 95% 72.2% 89.8% 84.2% PPV of NBI (%) 85.7% 86% 73.1% 90.9% 97.7% PPV of WL (%) 83.3% 85% 34.1% 83.8% 79.6% NPV of NBI (%) Not able to evaluate 88% 95.7% 91.7% 99% NPV of WL (%) 16.7% 100% 88.6% 79.1% 96.6% ACC of NBI (%) 85.7% 92% 90.8% 90.4% 98.5% ACC of WL (%) 71.1% 64% 70% 76.9% 88.3% Abbreviations: ACC, accuracy; ELS, European Laryngological Society; IPCL, intraepithelial papillary capillary loop; NBI, narrow-band imaging; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, speciﬁcity; WL, white-light endoscopy. Note: the number and type of assessed lesions were classiﬁed according to the IPCLs, the SE, the SP, the PPV, the NPV, and the ACC of NBI and WL, by specifying if they were performed preoperatively and intraoperatively [the data calculated by the reviewers are reported in italics]. Table 4 Characteristics of the studies that compared NBI and WL Abbreviations: ACC, accuracy; ELS, European Laryngological Society; IPCL, intraepithelial papillary capillary loop; NBI, narrow-band imaging; NPV, negative predictive value; PPV, positive predictive value; SE, sensitivity; SP, speciﬁcity; WL, white-light endoscopy. Study Limitations The present review found some limitations in the analysis of the current literature. Even though we used narrow inclusion criteria, some selected studies showed heteroge- neity regarding the number of enrolled patients, types of lesion assessed, study design, and number and level of experience of the endoscopists who performed NBI. At the beginning of the learning curve for NBI, the cases of positive lesions may be overestimated with a consequent increase in the number of false positives. This is mostly observed in the ﬁrst six months of use.14,21 Indeed, regard- ing this issue, Zurek et al.22 showed that after 65 to 70 NBI examinations, the plateau of the learning process is reached. However, the included studies reported the ability of NBI to detect premalignant and malignant lesions as far as SE, SP, PPV, NPV, and ACC. Moreover, the SP of NBI is based only on outpatient evaluation, without a histological conﬁrmation, since it is unfeasible and unethical to perform either random or systematic biopsies in “non-suspicious” NBI lesions. 2 Lin YC, Watanabe A, Chen WC, Lee KF, Lee IL, Wang WH. Narrow- band imaging for early detection of malignant tumors and radia- tion effect after treatment of head and neck cancer. Arch Otolaryngol Head Neck Surg 2010;136(03):234–239 3 Volgger V, Felicio A, Lohscheller J, et al. Evaluation of the com- bined use of narrow band imaging and high-speed imaging to discriminate laryngeal lesions. Lasers Surg Med 2017;49(06): 609–618 4 Popek B, Bojanowska-Poźniak K, Tomasik B, Fendler W, Jeruzal- Świątecka J, Pietruszewska W. Clinical experience of narrow band imaging (NBI) usage in diagnosis of laryngeal lesions. Otolaryngol Pol 2019;73(06):18–23 5 Klimza H, Jackowska J, Piazza C, Banaszewski J, Wierzbicka M. The role of intraoperative narrow-band imaging in transoral laser microsurgery for early and moderately advanced glottic cancer. Rev Bras Otorrinolaringol (Engl Ed) 2019;85(02):228–236 6 Staníková L, Šatanková J, Kučová H, Walderová R, Zeleník K, Komínek P. The role of narrow-band imaging (NBI) endoscopy in optical biopsy of vocal cord leukoplakia. Eur Arch Otorhinolar- yngol 2017;274(01):355–359 7 Lukes P, Zabrodsky M, Lukesova E, et al. The role of NBI HDTV magnifying endoscopy in the prehistologic diagnosis of laryngeal papillomatosis and spinocellular cancer. BioMed Res Int 2014; 2014:285486 Another limitation of the present review refers to the ACC of NBI and WL. NBI Classiﬁcation of Laryngeal Lesions As aforementioned, the efﬁciency of NBI is related to its ability to reveal IPCL changes on the mucosa and submuco- Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 477 distinguish these lesions, the angle of the turning point of the IPCL should be evaluated. Indeed, turning points with wide angles are suggestive of RRP, whereas narrowangles are indicators of cancer. tic error. Finally, we would like to emphasize the need to use a shared, simple and practical classiﬁcation for laryngeal lesions, such as the ELS classiﬁcation, to identify a consensus lesion management strategy. Analyzing the included articles, only two of them9,12 eval- uated vascular changes according to the ELS classiﬁcation, and the other eight used the classiﬁcation made by Ni et al.15 The remaining studies did not specify the classiﬁcation used. Due to this gap in the number of articles which used one classiﬁca- tion rather than the other one, we cannot include this topic as an evaluation criterion in the present review. Further studies could investigate and evaluate the effec- tiveness of NBI in another important aspect of laryngeal cancer, namely during the follow-up care, for the early detection of potential recurrences. Study Limitations Narrow-band imaging is usually performed after WL endoscopy; therefore, the former necessarily shows a higher ACC than the latter in detecting precancerous and cancerous lesions. 8 Rzepakowska A, Sielska-Badurek E,Żurek M,Osuch-WójcikiewiczE, Niemczyk K. Narrow band imaging for risk stratiﬁcation of glottic cancer within leukoplakia. Head Neck 2018;40(10):2149–2154 9 Šifrer R, Rijken JA, Leemans CR, et al. Evaluation of vascular features of vocal cords proposed by the European Laryngological Society. Eur Arch Otorhinolaryngol 2018;275(01):147–151 Furthermore, NBI is also used and recommended in the follow-up of patients after surgery or radiotherapy, as it enables an earlier detection of relapse than the evaluation with WL.14,23 Unfortunately, due to the limited bibliography on the efﬁcacy of NBI in the follow-up, it has not been possible to study in details its beneﬁts and advantages compared with WL. 10 Rzepakowska A, Sielska-Badurek E, Cruz R, Sobol M, Osuch- Wójcikiewicz E, Niemczyk K. Narrow band imaging versus lar- yngovideostroboscopy in precancerous and malignant vocal fold lesions. Head Neck 2018;40(05):927–936 11 Hosono H, Katada C, Okamoto T, et al. Usefulness of narrow band imaging with magnifying endoscopy for the differential diagnosis of cancerous and noncancerous laryngeal lesions. Head Neck 2019;41(08):2555–2560 References 1 Piazza C, Del Bon F, Peretti G, Nicolai P. Narrow band imaging in endoscopic evaluation of the larynx. Curr Opin Otolaryngol Head Neck Surg 2012;20(06):472–476 International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. Final Comments 12 Šifrer R, Šereg-Bahar M, Gale N, Hočevar-Boltežar I. The diagnostic value of perpendicular vascular patterns of vocal cords deﬁned by narrow-band imaging. Eur Arch Otorhinolaryngol 2020;277(06): 1715–1723 The objective of the present review was to assess the applications and beneﬁts of NBI by comparing them to those of WL. Therefore, based on the analysis of the current literature, the present study conﬁrms that NBI is an effective diagnostic tool in detecting precancerous and cancerous lesions and in identifying proper surgical margins. Moreover, NBI is useful when laryngeal lesions cannot be easily assessed with WL, such as in the presence of a thick white patch. Therefore, it’s important to develop a deep experience in the performance of NBI to interpret correctly and accu- rately laryngeal lesions, thus signiﬁcantly reducing diagnos- 13 De Vito A, Meccariello G, Vicini C. Narrow band imaging as screening test for early detection of laryngeal cancer: a prospec- tive study. Clin Otolaryngol 2017;42(02):347–353 14 Piazza C, Cocco D, De Benedetto L, Del Bon F, Nicolai P, Peretti G. Narrow band imaging and high deﬁnition television in the assessment of laryngeal cancer: a prospective study on 279 patients. Eur Arch Otorhinolaryngol 2010;267(03):409–414 15 Ni XG, He S, Xu ZG, et al. Endoscopic diagnosis of laryngeal cancer and precancerous lesions by narrow band imaging. J Laryngol Otol 2011;125(03):288–296 15 Ni XG, He S, Xu ZG, et al. Endoscopic diagnosis of laryngeal cancer and precancerous lesions by narrow band imaging. J Laryngol Otol 2011;125(03):288–296 International Archives of Otorhinolaryngology Vol. 25 No. 3/2021 © 2020. Fundação Otorrinolaringologia. All rights reserved. Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. 478 Impact of NBI in the Assessment of Neoplastic and Preneoplastic Laryngeal Lesions Saraniti et al. committee on endoscopic laryngeal imaging of the European Laryngological Society. Eur Arch Otorhinolaryngol 2016;273 (05):1207–1214 16 Ni XG, Zhu JQ, Zhang QQ, Zhang BG, Wang GQ. Diagnosis of vocal cord leukoplakia: The role of a novel narrow band imaging endoscopic classiﬁcation. Laryngoscope 2019;129(02):429–434 21 Piazza C, Cocco D, Del Bon F, Mangili S, Nicolai P, Peretti G. Narrow band imaging and high deﬁnition television in the endoscopic evaluation of upper aero-digestive tract cancer. Acta Otorhinolar- yngol Ital 2011;31(02):70–75 17 Campo F, D’Aguanno V, Greco A, Ralli M, de Vincentiis M. The prognostic value of adding Narrow-Band Imaging in transoral laser microsurgery for early glottic cancer: a review. Lasers Surg Med 2019;52(04):301–306 22 Żurek M, Rzepakowska A, Osuch-Wójcikiewicz E, Niemczyk K. Learning curve for endoscopic evaluation of vocal folds lesions with narrow band imaging. Rev Bras Otorrinolaringol (Engl Ed) 2019;85(06):753–759 18 Yang Y, Fang J, Zhong Q, et al. The value of narrow band imaging combined with stroboscopy for the detection of applanate indis- cernible early-stage vocal cord cancer. Acta Otolaryngol 2018;138 (04):400–406 23 Lukes P, Zabrodsky M, Syba J, Lukesova E, Votava M, Plzak J. Efﬁcacy of transnasal ﬂexible videoendoscopy with narrow band imaging for follow-up of patients after transoral laser cordectomy. Lasers Surg Med 2020;52(04): 333–340 19 Kraft M, Fostiropoulos K, Gürtler N, Arnoux A, Davaris N, Arens C. Value of narrow band imaging in the early diagnosis of laryngeal cancer. Head Neck 2016;38(01):15–20 20 Arens C, Piazza C, Andrea M, et al. Proposal for a descriptive guideline of vascular changes in lesions of the vocal folds by the
https://openalex.org/W2091841729	http://www.isez.pan.krakow.pl/journals/folia/pdf/62(2)/62(2)_08.pdf	English	null	Sexual Maturation in Common Vole (<I>Microtus arvalis</I>) Males Raised under Laboratory Conditions	Folia Biologica	2,014	cc-by	6,156	Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA Accepted February 19, 2014 Accepted February 19, 2014 MISKA A., KRUCZEK M., KAPUSTA J. 2014. Sexual maturation in Common vole (Microtus arvalis) males raised under laboratory conditions. Folia Biologica (Kraków) 62: 135-142. The common vole is one of the most numerous rodents in Europe and Asia but its reproductive biology is not fully described. It is thought that females reach reproductive abilities at a very young age, however, there is no data concerning male sexual maturation. The aim of the present study was to determine the rate of sexual maturation of males of the common vole. Research was carried out on 4, 6, 8 and 10 week-old animals. Body, testes and accessory sex glands weights were compared to evaluate morphological development. Epididymal sperm quality was assessed by a motility test, sperm tail membrane integrity, viability, maturity and sperm head morphology. Moreover, the number of sperm cells in 1 mm3 of semen was evaluated. The largest body weights were observed in 8 and 10 week-old males, which also possessed the highest relative weights of gonads and accessory sex glands. These groups of males produce about 2 times more gametes than 6 week-old individuals. The highest seamen quality was noted in 8 and 10 week-old males. Based on the obtained results, it is concluded that among the investigated age groups the most appropriate male age for fertile copulation is between 8 and 10 weeks. Key words: Male sexual maturation, spermatozoa, reproductive abilities, common vole, Microtus arvalis. Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA, Institute of Environmental Sci- ences, Jagiellonian University, Gronostajowa 7, 30-387 Kraków, Poland. E-mail: agata.miska@uj.edu.pl Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA, Institute of Environmental Sci- ences, Jagiellonian University, Gronostajowa 7, 30-387 Kraków, Poland. E-mail: agata.miska@uj.edu.pl The common vole, Microtus arvalis (PALLAS 1779) is widespread all over Europe and Asia (WILSON & REEDER 2005). Originally, it inhabited grass- lands and rural areas. Today, common voles as well as other small mammals are common in culti- vated lands (FREEMARK 1995; SINGLETON & BROWN 1999). It is estimated that populations in cultivated areas constitute 60-90% of the total number of small rodent populations (STEIN 1955; MACKIN-ROGALSKA 1981; JACOB & TKADLEC 2010). PL-ISSN 0015-5497 (print), ISSN 1734-9168 (online) Ó Institute of Systematics and Evolution of Animals, PAS, Kraków, 2014 PL-ISSN 0015-5497 (print), ISSN 1734-9168 (online) Ó Institute of Systematics and Evolution of Animals, PAS, Kraków, 2014 Folia Biologica (Kraków), vol. 62 (2014), No 2 doi:10.3409/fb62_2.135 *Supported by grants from the Polish State Committee for Scientific Research (DS/V/WBiNoZ/UJ/757). The authors are grate- ful to J. Styrna for acrosome visualization and E. Pochroñ for her technical assistance. Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA The common vole is considered to be one of the most populous small mammals in Europe and every 3 to 5 years a population outbreak is ob- served (TKADLEC & STENSETH 2001; LAMBIN & BRETAGNOLE 2006; JACOB & TKADLEC 2010; LUQUE-LARENA et al. 2013). the potential crop yield (TERTIL 1977; TRUSZ- KOWSKI 1982), especially if its diet is considered: mostly plant material, mainly Monocotyledone, and seeds (HOOGENBOOM et al. 1984). Moreover, like M. oeconomus, usually males have larger home ranges than adult females (TAST 1966). Bet- ter understanding of the reproduction biology of M. arvalis may improve methods of species con- trol. Presently, research concerning sexual matura- tion of the common vole is limited. The rate of fe- male sexual maturation is a controversial subject. TKADLEC and ZEJDA (1995) claim that at the age of 12-13 days the vagina opening is developed, and BOYCE and BOYCE (1988) report a mean age of 13.7 days as early sexual maturation for daugh- ters of solitary breeding females. Daughters from grouped breeding females reach sexual maturity at a mean age of 27.4 days (BOYCE & BOYCE 1988), nevertheless at the age of 14 days precocious From the economic point of view, the common vole is considered a pest. When the total number of individuals reaches 3000 per 1 ha, it is estimated that the possible damage may reach up to 90% of 136 A. MISKA et al. breeding occurs (TKADLEC & ZEJDA 1995). There is no data concerning the age of sexual maturation in males. chips were provided as bedding material, which was changed once a week. Standard pelleted chow for laboratory rodents (Labofeed H, Kcynia) and tap water were available ad libitum. The aim of this study was to determine the age of sexual maturation in males and verify if it is com- parable to the early sexual maturation of females. The described experiment was conducted in 2008 and 2009. Tested individuals came from the 3rd and 4th generation of animals bred in the Institute of Environmental Sciences, Jagiellonian Univer- sity. To examine the rate of male sexual matura- tion, 4, 6, 8 and 10 weeks-old males were used in the experiment. Adult common vole males and fe- males were mated; 1 female with 1 male in a cage. Following 19 to 21 days of pregnancy, the number of pups was noted and males were randomly as- signed to four experimental groups. Material and Methods After applying gentle pressure to each cauda epididymis with forceps, allowing epididymal sperm to press to the vasa deferentia, the content of the latter was suspended in 100 Fl of M2 medium (Sigma-Aldrich, Germany), and allowed to dis- perse for 2 minutes. Body, testes and accessory sex glands weights After cervical dislocation, males were weighed, next paired testes as well as seminal vesicles and coagulation glands were dissected and weighed (the latter two jointly). Animals Common voles (Microtus arvalis) came from the laboratory colony of the Institute of Environ- mental Sciences, Jagiellonian University in Kraków. The original stock was obtained in 2006 from Ma- gura National Park, and was maintained as an out- bred stock colony according to the system de- scribed by GREEN (1966). Each generation con- sisted of at least 18 breeding pairs; the male and female in each mating pair did not have common parents or grandparents. This breeding system en- sures the heterogeneity of the colony. Animals were housedinpolyethylenecages(40cm×25cm×15cm) under 14h of light (7am-9pm) and 10h of darkness (9pm-7am), in 21 ± 1ºC, and 60% humidity. Wood Epididymal sperm concentration Experimental procedures Body, testes and accessory sex glands weights Agata MISKA, Ma³gorzata KRUCZEK, and Joanna KAPUSTA For the 4-week old age group, three week-old males were sepa- rated from their parents and placed into individual cages. For the 6, 8 and 10-week old groups, the males were kept in groups of 3 to 5 individuals per cage. Two weeks before the experiment, each indi- vidual assigned to 6-, 8- or 10-weeks old group had been placed into a separate cage. Various fields of research use the common vole asamodelspecies(KRÓLetal.2012;LANTOVÁetal. 2012; FISCHER et al. 2013). Its main attribute is that it is considered to be a wild living rodent. Only a well-known biology of this rodent will permit an explicit understanding of selected physiological mechanisms. Up till now, the age of sexual matu- ration in common vole males has been defined based on knowledge derived from other rodent species. For example, golden hamster males reach puberty at the age of 42.5 days (VANDENBERGH 1971), laboratory mice at 40-60 days – depending on the strain (BARTKE et al. 1974), bank voles at the age of 6-12 weeks (KRUCZEK 1986). Moreo- ver, it has been shown that the odor of sexually ma- ture males may accelerate female sexual maturation in mice (VANDENBERGH 1967, 1969) and also block pregnancy in newly conceived females (BRUCE 1959). Each experimental group consisted of 10 males. Inordertoavoidlitterimpact,amaximumof2males from 1 litter took part in the experiment. In addi- tion, siblings were separated. Male reproductive success and sperm quality and quantity are usually correlated with body and testes weights (KRUCZEK 1986; KLEMME et al. 2007; KRUCZEK & STYRNA 2009). To determine male sexual maturation in the common vole, the body, testes and accessory sex gland weights were compared in various age groups. Additionally, we examined epididymal sperm quantity and quality obtained from males of different ages. This method is widely considered a sufficient proce- dure to discriminate male reproductive abilities in rodents (KRUCZEK & STYRNA 2009). Experimental procedures Epididymal sperm viability – eosin-Y test The test reflects the structural and morphological integrity of the sperm membrane (WALCZAK et al. 1994; STYRNA & KRZANOWSKA 1995). To assess sperm viability, 20 Fl of epididymal sperm sus- pension (as described above) was mixed with 20 Fl of 0,2% eosin Y, incubated for 10 min at 37°C and smeared on a slide. The percentage of spermatozoa with unstained sperm heads (viable spermatozoa) among 200 counted spermatozoa from each male is reported. Malebody,testesandaccessorysexglandsweights Malebody,testesandaccessorysexglandsweights The results summarized in Table 1 show that 10 week-old male body weights, testes relative weights and accessory sex gland relative weights were higher than in 4, 6, and 8 week-old males. There were no significant differences in body weights between 4, 6 and 8 week-old males (Table 1), whereas testes relative weight in 4 week-old males was significantly lower than in 6 week-old males as well as in 8 week-old males. Moreover, acces- sory sex gland weight was significantly lower in 4 week-old males as compared to older individu- als. As reported in Table 1, there were no statisti- cally important differences between 6, 8 and 10 week-old males in testes relative weight, nor in accessory sex gland weight. As shown in Table 1, relative weight differences amongst 6, 8 and 10 week-old males were significant in testes and accessory sex glands. Epididymal sperm concentration A 1:20 dilution of epididymal sperm suspension with M2 medium was prepared, and the number of live spermatozoa in 100 squares of a hemocytome- ter (Bûrker chamber) was counted under a light microscope at 400x magnification. A coverslip was placed on the sample to restrict spermatozoa movements. The average of 2 sperm counts was used to estimate the sperm concentration as de- scribed by STYRNA and KRZANOWSKA (1995) and KRUCZEK and STYRNA (2009). 137 Sexual Maturation in Common Vole Males Epididymal sperm motility Individual counts and assessment of sperm were blind to the males age. Spermatozoa motility was assessed in a hemocy- tometer. The percentage of motile sperm, i.e. sperm showing progressive movement, among 200 counted spermatozoa from each male is reported (SEED et al. 1996). Epididymal sperm morphology For morphological examination, a small drop of the epididymal sperm suspension was smeared on a slide, air-dried, fixed in acetic-alcohol (absolute alcohol:glacial acetic acid, 3:1), and stained with Papanicolau to assess the proportions of different sperm head abnormalities. Epididymal sperm tail membrane integrity – wa- ter test Epididymal sperm tail membrane integrity – wa- ter test The integrity of the epididymal sperm tail mem- brane was determined by the hypoosmotic swel- ling test. The procedure was the same as used for mice and bank voles (STYRNA & KRZANOWSKA 1995; STYRNA et al. 2003; KRUCZEK & STYRNA 2009): 20 Fl of epididymal sperm suspension (as described above) was mixed with 120 Fl distilled water on a clean glass slide, then the mixture was gently covered with a coverslip and incubated for 5 min at 37°C before it was examined. The percent- age of spermatozoa showing swelling among 200 counted spermatozoa from each male is reported. Epididymal spermatozoa without a cytoplasmic droplet Twenty Fl of epididymal sperm suspension (as described above) was transferred to a slide and gently covered with a coverslip. The percentage of spermatozoa without a cytoplasmic droplet among 200 counted spermatozoa with progressive move- ments from each male is reported (KRUCZEK & STYRNA 2009) Ethical standards The experimental procedures for this study were approved by the Regional Committee on Animal Experimentation in Kraków (Protocol No. 38/2008) acting in compliance with the European Commu- nitiesCouncilDirectiveof24Nov1986(86/609/EEC). Statistical analysis In order to obtain linearity, percentage outcomes were subjected to angular transformation. To com- pare means, a one-way ANOVA test was used. To show significant differences, a post hoc test (Tukey test) was used. All results were presented as means ± S.E. and P<0.05 was considered sig- nificant. To eliminate the impact of body mass on testes weights, the relative testes weights were compared. All calculations were made using Sta- tistica PL ver. 10.0. Epididymal sperm cells properties Only in 1 of the 4 week-old males a single, non- motile spermatozoa was observed and that is why this data was excluded from further consideration. In 7 out of 10, 6 week-old individuals and in all 8 and 10 week-old males, epididymis sperm cells in semen were observed. Furthermore, sperm evalua- tion was made on 6, 8 and 10 week-old males. As demonstrated in Table 1, the highest number of spermatozoa occurred in 8 and 10 week-old males, This study discusses the following abnormalities (Fig. 1): Normal – sperm with correct morphology; Class 2 – lack of top part of hook and abnormalities in bottom part of the head; Class 3 – lack of hook, serious abnormalities in the proximal part of sperm heads, possible changes in the bottom part of the head. Table 1 Body weight, relative testes and accessory sex glands (seminal vesicles and coagulation glands) weights, and spermatozoa concentration in 4, 6, 8, 10 week-old common vole males; (Mean ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01 and a, b – P<0.05 Males F P 4 week-old 6 week-old 8 week-old 10 week-old body weight (g) 19.21A ± 1.98 23.01a ± 1.64 22.87b ± 1.18 30.18Aba ± 1.61 F3.36 = 7.98 <0.01 testes relative weight (mg/10g body weight) 50.26ABa ± 9.69 97.84a ±16.21 114.75A ± 5.63 124.80B ± 8.26 F3.36 = 9.56 <0.01 accessory sex glands relative weight (mg/10g body weight) 9.54ABa ± 4.25 41.34a ± 12.88 68.25A ± 9.98 92.14B ± 4.34 F3.36= 16.71<0.01 spermatozoa concentration (×10-4/ml) 24.07 ± 8.05 42.78 ± 8.26 46.84 ± 8.13 F2.36 = 2.22 N.S. Fig. 1. Photography (A) and diagram (B) presenting the morphology of common vole spermatozoa: Normal-sperm with correct morphology, class 2 and 3- sperm with head alterations. A. MISKA et al. 138 138 A. MISKA et al. Discussion Our results have shown that common vole males reach sexual maturation at the age of 8-10 weeks. Maturing males in these age groups have the heaviest body, gonad and accessory sex gland rela- tive weights and produce sperm cells of the best quality and quantity. This age is rather delayed when taking into consideration female age at sex- ual maturation. TKADLEC & ZEJDA (1995) indi- cate that for common vole females, the possible age for precocious breeding is 13-14 days, while BOYCE & BOYCE (1988) indicate a mean age of 13.7 days for female sexual maturation in daugh- ters of solitary bred females, and 24.7 days as the mean age of sexual maturation for daughters of group-bred females. Our results contribute to the biology of this species, and at the same time facili- tate further research especially in the area of neuro-hormonal activity regulation. Fig. 3. Proportion of motile spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letter differ significantly at A – P<0.01. Fig. 4. Proportion of not swollen spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01. Fig. 4. Proportion of not swollen spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01. Morphological development is correlated with body weight. Research has demonstrated that bank vole (Myodes glareolus) males with higher body weight have heavier testes and accessory sex glands than males with lower body weight (KRU- CZEK1986; KLEMME et al. 2007). Moreover, body mass is one of the factors which influences male reproductive success. The bigger and heavier the males, the more they are preferred by bank vole fe- males (KRUCZEK & STYRNA 2009) and the higher the number of offspring they sire (KLEMME et al. 2007). In our experiment, the lowest body mass was observed in 4 week-old males, while the high- est in 10 week-old males; 6 and 8 week-old male body weights were comparable. We assume that adult common vole body weight will not change withtime,asinothermaturerodents(HOFFMAN etal. 2002). Furthermore, in mammals, the quality and Fig. 5. Epididymal sperm cells properties The highest proportion of spermatozoa without a cytoplasmic droplet was 139 Sexual Maturation in Common Vole Males Sexual Maturation in Common Vole Males Fig. 2. Proportion of viable spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01. noted in 10 week-old males, while the lowest pro- portion was in 6 week-old males (P<0.01). There were significant differences between 6 week-old males and both 8 and 10 week-old animals as well as between 8 and 10 week-old males. Sperm head abnormalities are presented in Table 2. There were significant differences in abnormal sperm head proportions between 6 week-old males and both 8 and 10 week-old males. Six week-old animals had the highest proportion of abnormal sperm heads. There were no statistically important differences between 8 and 10 week-old males. The highest number of sperm heads with normal mor- phology was observed in those 2 age groups. Nor- mal sperm head morphology was divided into 2 classes (Table 2 and Fig. 1). In each class, the same dependence as in the combined normal sperm head morphology group appeared. Fig. 2. Proportion of viable spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01. Fig. 3. Proportion of motile spermatozoa in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letter differ significantly at A – P<0.01. Epididymal sperm cells properties Body weight, relative testes and accessory sex glands (seminal vesicles and coagulation glands) weights, and spermatozoa concentration in 4, 6, 8, 10 week-old common vole males; (Mean ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01 and a, b – P<0.05 Males F P 4 week-old 6 week-old 8 week-old 10 week-old body weight (g) 19.21A ± 1.98 23.01a ± 1.64 22.87b ± 1.18 30.18Aba ± 1.61 F3.36 = 7.98 <0.01 testes relative weight (mg/10g body weight) 50.26ABa ± 9.69 97.84a ±16.21 114.75A ± 5.63 124.80B ± 8.26 F3.36 = 9.56 <0.01 accessory sex glands relative weight (mg/10g body weight) 9.54ABa ± 4.25 41.34a ± 12.88 68.25A ± 9.98 92.14B ± 4.34 F3.36= 16.71<0.01 spermatozoa concentration (×10-4/ml) 24.07 ± 8.05 42.78 ± 8.26 46.84 ± 8.13 F2.36 = 2.22 N.S. Fig. 1. Photography (A) and diagram (B) presenting the morphology of common vole spermatozoa: Normal-sperm with correct morphology, class 2 and 3- sperm with head alterations. Fig. 1. Photography (A) and diagram (B) presenting the morphology of common vole spermatozoa: Normal-sperm with correct morphology, class 2 and 3- sperm with head alterations. while the lowest occurred in 6 week-old males. However, there were no significant differences among these 3 age groups. and 10 week-old males reached a significant level (P<0.01). The results obtained from the water test which presents the proportion of spermatozoa with cor- rect form of sperm tail membrane (not swollen), are shown in Fig. 4. There were no significant dif- ferences between 8 and 10 week-old animals. In 6 week-old males, the proportion of sperm with the correct form of the tail membrane was signifi- cantly lower than in both 8 and 10 week-old males (P<0.01). Proportions of viable spermatozoa (Fig. 2) were almost the same in 8 and 10 week-old males. There were significant differences in the proportion of viable spermatozoa between the above 2 groups and 6 week-old males (P<0.01).The lowest propor- tion of viable sperm was observed in the 6 week-old group. Similarly, the proportion of motile spermato- zoa (Fig. 3) was similar in the 8 and 10 week-old males and higher than in the 6 week-old males, al- though only differences between 6 week-old males Data concerning sperm without a cytoplasmic droplet is shown in Fig. 5. Discussion Moreover, recent research has shown strong evi- dence for multiple paternities in common voles: 50% of litters are sired by 2 or 3 males (BORKOW- SKA & RATKIEWICZ 2010). In these situations, only sperm with the best properties: sperm tail membrane integrity and high number of mature sperm, can “win” by reaching the ovum (SNOOK 2005; STOCKLEY 2004). Our results show that the highest proportion of mature sperm is observed in 10 week-old males, while the best sperm tail mem- brane properties occur in the 8 and 10 week-old male sperm cells. Spermatozoa in these age groups probably have the best chance to “win the race” to- wards the ovum quantity of sperm depends on gonad size; bigger testes produce more spermatozoa that are of better quality (GOMENDIO et al. 1998; MONTOTO et al. 2011). Testosterone is the main male androgen which controls male reproductive system development. Production of this hormone takes place in testes (SOLOMON et al. 2011). After conversion into di- hydrotestosterone, testosterone controls secon- dary sexual trait development, accessory sex gland development and spermatogenesis (FRANÇA et al. 2006). Sperm production and sperm maturation is also regulated by testosterone (RACHMAN & CHRISTIAN 2007; SUN et al. 1990) and the number of sperm is an indicator of spermatogenesis effec- tiveness (RUIZ-PESINI et al. 2000). Considering these factors, we expected different quantities and qualities of spermatozoa in different male age groups. In our experiment, the youngest, i.e. the 4 week-old males, have the lowest testes weight and only in 1 out of 10 individuals, sperm cells were present in the visual field. Because there were only a few non-motile sperm cells in one in- dividual, this age group was considered unable to reproduce. This data was excluded from further consideration. Therefore, only data concerning sperm cell quantity and quality obtained from 6, 8, and 10 week-old males was used in the statistical analysis. Although there were no statistically im- portant differences in 6-, 8- and 10 week old ani- mal body and testes weights, a tendency to increase with age was observed. In general, higher testes and accessory sex glands were possessed by older males. The probable interaction of accessory sex gland in Microtus is copulatory plug formation (TAMARIN 1985). Testes weight indicates acces- sory sex gland development in Yorkshire pigs (RAESIDE et al. 1997). Discussion The best sperm viability is ob- served in 8 and 10 week-old males. Multiple copu- lations in rodents (for example in bank voles – M. glareouls) are possible, which in consequence may cause sperm competition and hidden female choice (KLEMME et al. 2006; EBERHARD 1996). Moreover, recent research has shown strong evi- dence for multiple paternities in common voles: 50% of litters are sired by 2 or 3 males (BORKOW- SKA & RATKIEWICZ 2010). In these situations, only sperm with the best properties: sperm tail membrane integrity and high number of mature sperm, can “win” by reaching the ovum (SNOOK 2005; STOCKLEY 2004). Our results show that the highest proportion of mature sperm is observed in 10 week-old males, while the best sperm tail mem- brane properties occur in the 8 and 10 week-old male sperm cells. Spermatozoa in these age groups probably have the best chance to “win the race” to- wards the ovum. ences in the number of sperm cells amongst 3 male age groups, 6 week-old males produce about 2 times less spermatozoa than 10 week-old males. With the increase in the number of spermatozoa, the possibility of fertilization becomes higher (SNOOK 2005). However, the effectiveness of fertilization also depends on sperm quality. Only sperm with highest parameters of motility, viability, sperm tail membrane integrity, sperm maturity, and sperm head morphology may survive in the female geni- taltrackandreachtheovum(LAMEO&GIAMBERSIO 1991; SOMFARI et al. 2002). In Microtine species (e.g. montane voles – M. montanus; GRAY et al. 1974, bank vole – M. glareolus; CLARK 1970, prairie voles – M. ochrogaster, RICHMOND & CONAWAY 1969) provoked estrus is presented and we assume that in the common vole the situation is similar. Generally, copulation takes place before ovulation and that is why sperm cells have to sur- vive in the female genital track (SNOOK 2005). Only viable sperm may survive the capacitation process, acrosome reaction and finally may be capable of fertilizing the mammalian ovum (YA- NAGIMACHI 1981). The best sperm viability is ob- served in 8 and 10 week-old males. Multiple copu- lations in rodents (for example in bank voles – M. glareouls) are possible, which in consequence may cause sperm competition and hidden female choice (KLEMME et al. 2006; EBERHARD 1996). Discussion Proportion of spermatozoa without a cytoplasmic droplet in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P < 0.01 and a – P < 0.05. Fig. 5. Proportion of spermatozoa without a cytoplasmic droplet in 6, 8, 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P < 0.01 and a – P < 0.05. 140 A. MISKA et al. Table 2 Proportion of spermatozoa head abnormalities in 6, 8, and 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01 Spermatozoa number Males F P 6 week-old 8 week-old 10 week-old normal 0.600AB ± 0.03 1.142A ± 0.02 1.172B ± 0.03 F2.26 = 43.64 0.01 abnormal – total 0.76AB ± 0.1 0.18A ± 0.03 0.16B ± 0.03 F2.26 = 42.48 0.01 abnormal – class 1 0.488AB ± 0.08 0.242A ± 0.03 0.245B ± 0.05 F2.26 = 32.07 0.01 abnormal – class 2 0.729A.B ± 0.03 0.342A ± 0.02 0.304B ± 0.03 F2.26 = 38.82 0.01 Proportion of spermatozoa head abnormalities in 6, 8, and 10 week-old common vole males; (Means ± S.E.); Means marked by the same letters differ significantly at A, B – P<0.01 ences in the number of sperm cells amongst 3 male age groups, 6 week-old males produce about 2 times less spermatozoa than 10 week-old males. With the increase in the number of spermatozoa, the possibility of fertilization becomes higher (SNOOK 2005). However, the effectiveness of fertilization also depends on sperm quality. Only sperm with highest parameters of motility, viability, sperm tail membrane integrity, sperm maturity, and sperm head morphology may survive in the female geni- taltrackandreachtheovum(LAMEO&GIAMBERSIO 1991; SOMFARI et al. 2002). In Microtine species (e.g. montane voles – M. montanus; GRAY et al. 1974, bank vole – M. glareolus; CLARK 1970, prairie voles – M. ochrogaster, RICHMOND & CONAWAY 1969) provoked estrus is presented and we assume that in the common vole the situation is similar. Generally, copulation takes place before ovulation and that is why sperm cells have to sur- vive in the female genital track (SNOOK 2005). Only viable sperm may survive the capacitation process, acrosome reaction and finally may be capable of fertilizing the mammalian ovum (YA- NAGIMACHI 1981). Discussion Our data has shown that with an increase of testes weight, the accessory sex gland weight is increased in rodents as well. This relationship was clearly evident in the case of 8 and 10 week-old males. In addition, those 2 age groups produce the largest number of sperm cells. Al- though there were no important statistical differ- 141 Sexual Maturation in Common Vole Males Sperm cell progressive movement is caused by the tail and is necessary to move towards the ovum for fertilization to take place (MORTIMER 1997). Sperm motility depends mainly on tail membrane properties (VETTER et al. 1998). In our experi- ment, the best semen quality was observed in 8 to 10 week old males. It can be therefore concluded that males of this age are considered best for mat- ing, however 6-week old individuals are also capa- ble of fertile copulations. GOMENDIO M., HARCOURT A.H., ROLDAN E.R.S. 1998. Sperm competition in mammals. (In: Sperm Competition and Sexual Selection. Academic, T.B. Birkhead, A.P. Mol- ler ed., New York): 667-751. GRAY G.D., DAVIS H.N., ZERYLNICK M., DEWSBURY D.A. 1974. Oestrus and induced ovulation in montane voles. J. Reprod. Fert. 38: 193-196. GREEN E.L. 1966. Breeding systems. Biology of the labora- torymouse.McGraw-HillBookCompany,NewYork:13-14. HOFFMAN W.P., NESS D.K., VAN LIER R.B.L. 2002. Analysis of roden growth data in toxicology studies. Toxicol. Sci. 66: 313-319. When spermatozoa reach the ovum, the correct form of sperm head is necessary. Hydrolytic en- zymes present in the acrosome partake in an acro- somal reaction which allows sperm cells to fertilize the ovum (KRZANOWSKA et al. 1995). There is a positive relationship between a high amount of spermatozoa with head abnormalities and limited fertilization abilities in hamsters (WEISSENBERG et al. 1987) and mice (KRZANOW- SKA et al. 1995). Our results show that the largest amount of abnormal sperm head morphology is present in the sperm of 6 week-old males, while the largest amount of normal morphology is prevalent in 8 and 10 week-old males. Results concerning sperm quality and quantity indicate that both 8 and 10 week-old males produce the best spermatozoa which may contribute to more successful repro- duction. However, this conclusion should be con- firmed with further research that would assess the number of successful breedings in different male age groups. HOOGENBOOM I., DAAN S., DALIINGA J.H., SCHOENMAK- ERS M. 1984. References KRZANOWSKA H., STYRNA J., WABIK-ŒLIZ B. 1995. Analy- sis of sperm quality in recombinant inbred mouse strains: correlation of sperm head shape with sperm abnormalities and with the incidence of supplementary in the perivitelline space. J. Reprod. Fertil. 104: 347-354. BARTKE A., WEIR J.A., MATHISON P., ROBERTSON C., DEL- TORIO S. 1974. Testicular function in mouse strains with dif- ferent age of sexual maturation. J. Hered. 65: 204-208. LAMBIN X., BRETAGNOLE N.G. 2006. Vole population cycles in northern and southern Europe: is there a need for different explanations for singlepattern? J. Animal Sci. 75:340-349. BORKOWSKA A., RATKIEWICZ M. 2010. Promiscuity, male reproductive success and mate relatedness in natural popula- tion of the common vole. J. Zool. 280: 195-201. LAMEO A.M., GIAMBERSIO A.M. 1991. Water-test: a Simple method to assess sperm membrane integrity. Int. J. Androl. 14: 278-282. BOYCE C.C.K., BOYCE J.L. 1988. Population of Microtus ar- valis. I. Lifetime reproductive success of solitary and grouped breeding females. J. Anim. Ecol. 57: 711-722. LANTOVÁ P., BRIXOVÁ L., LANTA V. 2012. Urine marking in male common vole: does behavioural activity matter? Be- hav. Process 90: 174-179. BRUCE H.M. 1959. An exteroceptive block to pregnancy in the mouse. Nature 184: 105. CLARK J.R. 1970. Ovulation in the bank vole, Clethrionomys glareolus. J. Reprod. Fertil. 23: 351. LUQUE-LARENA J.J., MOUGEOT F., VIÑUELA J., JAREÑO D., ARROYO L., LAMBIN X., ARRAOYO B. 2013. Recent large- scale range expansion and outbreaks of the common vole (Microtus arvalis) in NW Spain. BAAE. http://dx.doi.org/10.1016/j.baae.2013.04.006 EBERHARD W. 1996. Female Control: Sexual Selection by Cryptic Female Choice. Princton University Press, New Jer- sey. FISCHER D., IMHOLT C., PROKOP A., JACOB J. 2013. Efficacy of methyl nonyl ketone as an in-soil repellent for common voles (Microtus arvalis). Pest. Menag. Sci. 69: 431-436. MACKIN-ROGALSKA R. 1981. Spatial structure of rodent populations co-occurring in different crop Fields. Pol. Ecol. Stud. 7: 213-227. FRANÇA L.R., SUESCUN M.O., MIRANDA J.R., GIOVAMBAT- TISTA A., PERELLO M., SPENEDI E., CALANDRA R.S. 2006. Testis structure and function in a nongenetic hyperadipose rat model at prepubertal and adult ages. Endocrinology 22: 343-351. MONTOTO L.G., MAGAÑA C., TOURMENTE M., MAR- TIN-COELLO J., CRESPO C., LAQUE-LARENA J.J., GOMEN- DIO M., ROLDAN E.R.S. 2011. Sperm competition, sperm number and sperm quality in Muroid Rodents. Plos One 6: e18173. FREEMARK K. 1995. Assessing effects of agriculture on ter- restrial wildlife: developing a hierarchical approach for the US EPA. Landscape Urban Plan. 31: 99-115. Discussion Seasonal changes in the daily timing of behav- ior of the common vole, Microtus arvalis. Oecologia 61: 18-31. JACOB J., TKADLEC E. 2010. Rodent outbreaks in Europe: dy- namics and damage in rodent outbreaks. (In: Rodent Out- breaks: Ecology and Impact. G.R. Singleton, S. Belmain, P. Brown, B. Hardy eds. IRRI, Philippines): 218. KLEMME I., ECCARD J., YL_NEN H. 2006. Do female bank vole (Clethrionomys glareolus) mate multiply to improve on previous mates? Beh. Ecol. Sociobiol. 60: 415-421. KLEMME I., ECCARD J.A., YLÖNEN H. 2007. Reproductive success of male bank voles (Clethrionomys glareolus): the effect of operational sex ratio and body size. Behav. Ecol. Sociobiol. 6: 1911-1918. KRÓL E., DOUGLAS A., DARDENTE H., BIRNIE M.J., VINNE V.V.D., EIJER W.G., GERKEMA M.P., HAZLERIGG D.G., HUT R.A. 2012. Strong pituitary and hypothalamic response to photoperiod but not to 6-methuloxy-2-benzoxazolinone in female common voles (Microtus arvalis). Gen. Comp. Endocr. 179: 289-295. KRUCZEK M. 1986. Seasonal effects on sexual maturation of male bank voles (Clethionomys glareolus). J. Reprod. Fertil. 76: 83-89. KRUCZEK M., STYRNA J. 2009. Semen quantity and quality correlated with bank vole males’ social status. Behav. Proc- ess 82: 279-285. References MORTIMER S.T. 1997. A critical review of the physiological importance and analysis of sperm movement in mammals. Hum. Reprod. Update 3: 403-439. 142 A. MISKA et al. SUN Y.T., WREFORD N.G., ROBERTSON D.M., KRETSER D.M. 1990. Quantitative cytological studies of spermatogenesis in intact and hypophysectomized rats: identification of androgen-dependent stages. Endocrinology 127: 1215-1223. RAESIDE J.I., FRIENDSHIP R.M., VRABLIC O.E. 1997. Effects of castration on early postnatal development of male acces- sory sex glands in the domestic pig. Eur. J. Edocrinol. 137: 287-292. TAMARIN R.H. 1985. Biology of New World Microtus. ASM Special Publication 8, Shippensburg. Pp. 155-159. RAHMAN F., CHRISTIAN H.C. 2007. Non-classical actions of testosterone: an update. Trends Endocrinol. Metab. 18: 371-378. TAST J. 1966. The root vole, Microtus oeconomus (Pallas), as an inhabitant of seasonally flooded land. Ann. Zool. Fenn. 3: 127-171. RICHMOND M., CONAWAY C.H. 1969. Induced ovulation and oestrus in Microtus ochrogaster. J. Reprod. Fert. Suppl. 6: 357-376. TERTIL R. 1977. Impact of the common vole Microtus arvalis (Pallas), on the winter wheat and alfalfa crops. EPPO Bull. 7: 317-339. RUIZ-PESINI E., LAPEÑA A.C., DIEZ C., ÃLVEREZ E., ENRI- QUEZ J.A., LÓPEZ-PÌREZ M.J. 2000. Seminal quality corre- lates with mitochondrial functionality. Clin. Chim. Acta 300: 97-105. TKADLEC E., STENSETH N.C. 2001. A new geographical gra- dient in vole population dynamics. Proc. R. Soc. Land. B. 268: 1547-1552. SEED J.,CHAPIN R.E.,CLEGG D.E.,DOSTAL L.A.,FOOTE R.H., HURTT M.E., KLINEFELTER G.R., MAKRIS S.L., PER- REAULT S.D., SCHRANER S., SEYLER D., SPRADNO R., TREINEN K.A., VEERAMACHANENI D.N. WISE L.D. 1996. Methods for assessing sperm mobility, morphology, and counts in the rat, rabbit, and dog: consensus report. Reprod. Toxicol. 3: 237-244. TKADLEC E., ZEJDA J. 1995. Precocious breeding in female common voles and its relevance to rodent fluctuation. Oikos 73: 231-236. TRUSZKOWSKI J. 1982. The impact of the common vole on the vegetation of agroecosystems. Acta Theriol. 27: 305-345. WALCZAK R., SRTUMILLO E., KULA K. 1994. Eosin and wa- ter tests and results of conventional semen analysis. Ginekol. Pol. 65: 99-102. SINGLETON G.R., BROWN P.R. 1999. Occurrence and distri- bution of the common vole, Microtus arvalis (Pallas), in leg- umes and seed grasses in Poland between 1977 and 1992. Polish Ecol. Stud. 20: 503-508. WEISSENBERG R., BELLA R., LUNENFELD B. 1987. The fer- tilization capacity of golden hamster epididymal spermato- zoa in relations to age, number, motility and morphology. Andrologia 19: 47-53. SNOOK R.R. 2005. References Sperm in competition: not playing by the numbers. Trends Ecol. Evol. 20: 46-53. WILSON D., REEDER D. 2005. Mammal species of the World, a taxonomy and Geography references. (The Johns Hopkins University Press, Baltimore). Pp. 991-992. SOLOMON E., BERG L., MARTIN D.M. 2011. Biology. Life sciences Belmont. Pp. 1083-1084. SOMFARI T., BODÓ S., NAGY S., GÓCZA E., IVÀNCICS J., KOVÀCS A. 2002. Simultaneous evaluation of viability and acrosome integrity of mouse spermatozoa using light mi- croscopy. Biotech. Histochem. 77: 117-120. VANDENBERGH J.G. 1967. Effect of the presence of male on the sexual maturation of female mice. Endocrinology 81: 345-349. VANDENBERGH J.G. 1969. Male odor accelerates females sexual maturation in mice. Endocrinology 84: 658-660. STEIN G.W.H. 1955. Die Kleinsäuger ostdeutcher Acker- flächen. Z. saugetierkd. 20: 89-113. VANDENBERGH J.G. 1971. The penile smear: an index of sex- ual maturity in male golden hamsters. Biol. Reprod. 4: 234-237. STOCKLEY P. 2004. Sperm competition in mammals. Hum. Fert. 7: 91-97. STYRNA J., KRZANOWSKA H. 1995. Sperm select penetration test reveals differences in sperm quality in strains with dif- ferent Y chromosome genotype in mice. Arch. Andrology 35: 111-118. VETTER C.M., MILLERJ.E., CRAWFORD L.M., ARMSTRONG M.J., CLAIr J.H., CONNER M.W., WISE L.D., SKOPEK T.R. 1998. Comparison of motility and membrane integrity to assess rat sperm viability. Reprod. Toxicol. 12: 105-114. STYRNA J., KILARSKI W., KRZANOWSKA H. 2003. Infuence of the CBA genetic back-ground on sperm morphology and fertilization efficiency in mice with partial Y chromosome detection. Reprod. 126: 579-588. YANAGIMACHI R. 1981. Mechanisms of fertilization in mam- mals. (In: Fertilization and Embryonic Development in vi- tro. L. Mastroianni, J. Biggers eds, Plenum Press, New York): 81-182.

Subsets and Splits

No community queries yet

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