* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Defining drivers of the trophic niche width in reef fish communities
Survey
Document related concepts
Storage effect wikipedia , lookup
Unified neutral theory of biodiversity wikipedia , lookup
Introduced species wikipedia , lookup
Island restoration wikipedia , lookup
Biodiversity action plan wikipedia , lookup
Molecular ecology wikipedia , lookup
Habitat conservation wikipedia , lookup
Ecological fitting wikipedia , lookup
Lake ecosystem wikipedia , lookup
Latitudinal gradients in species diversity wikipedia , lookup
Reconciliation ecology wikipedia , lookup
Transcript
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olet à renseigner par l’enseignant responsable de l’option/spécialisation* Ou son représentant Date : !./!/! Signature Bon pour dépôt (version définitive) ! Autorisation de diffusion : Oui ! Non! ! !"#$%&'("')*+,',!! #/!,!^L!%/1&/G<'/!OM^L! !"#$%&'%()*$+,-./-%,-%,!! 7(_&'/!4/!%&(>/!,!&.9G0:*N(O6=%N=[!4)'/2&/$'!4/!'/2./'2./%[!"EF!Q*E/5%!O[!9(+]$#%:%$':G/'! 6+%/)>+(+&!'K0K'/+&!,!!$:./.01*O+*'!'+P*+:B5.0*#+Q+=OO!%*?U8#/!\(#)/$&)Z$/!:!=>'*2(G1$%!@$/%&[!E/++/%A! =$&'/%!G/G<'/%!4$!`$']!?P*G[!a$(#)&KA!,!R07-?&719*#$())+OP!"PE=!E/++/%! ! ! ! « Les analyses et les conclusions de ce travail d’étudiant n’engagent que la responsabilité de son auteur et non celle d’AGROCAMPUS OUEST. » ! ! Avec la participation de l’Ecole Pratique des Hautes-Etudes et de l’Université de la Nouvelle-Calédonie ! Fiche de diffusion du mémoire A remplir par l’auteur !"# avec le maître de stage. Aucune confidentialité ne sera prise en compte si la durée n’en est pas précisée. Préciser les limites de la confidentialité (2) : Confidentialité absolue : " # oui $ non (ni consultation, ni prêt) % Si oui #1 an #5 ans #10 ans A l’issue de la période de confidentialité ou si le mémoire n’est pas confidentiel, " merci de renseigner les éléments suivants : Référence bibliographique diffusable(3) : $ oui # non Résumé diffusable : $ oui # non Mémoire consultable sur place : $ oui # non Reproduction autorisée du mémoire : # oui # non Prêt autorisé du mémoire : $ oui # non !!!!!!!!!!!!!!!!!. Diffusion de la version numérique : % Si oui, l’auteur !"# $ oui # non complète l’autorisation suivante : Je soussigné(e) , propriétaire des droits de reproduction dudit résumé, autorise toutes les sources bibliographiques à le signaler et le publier. Date : Signature : Rennes/Angers, le !"# Le maître de stage(4), L’auteur , L’enseignant référent, (1) auteur = étudiant qui réalise son mémoire de fin d’études (2) L’administration, les enseignants et les différents services de documentation d’AGROCAMPUS OUEST s’engagent à respecter cette confidentialité. (3) La référence bibliographique (= Nom de l’auteur, titre du mémoire, année de soutenance, diplôme, spécialité et spécialisation/Option)) sera signalée dans les bases de données documentaires sans le résumé. (4) Signature et cachet de l’organisme. ! Remerciements En cette fin de période de stage, je tiens tout particulièrement à remercier l’équipe de l’IRD CoReUs du Laboratoire Arago de Banyuls-sur-mer, qui m’a accueillie et encadrée tout au long de ces six mois : un grand merci à mon encadrant, Michel Kulbicki, pour sa grande humanité, ses nombreuses corrections du brouillon de ce rapport mais aussi toutes ses connaissances sur l’univers des poissons de récifs, qui pourtant n’a décidément pas fini de nous cacher ses secrets. Merci également à René Galzin, de l’EPHE, d’avoir participé à la mise en place de ce stage et de nous avoir rapporté du bout du monde de nouveaux échantillons des Marquises, qui je l’espère dissiperont un peu les mystères isotopiques qui entourent ces îles pour que nous puissions les étudier &. Merci Valeriano, pour ton bel accent italien mais surtout pour ta présence, ton soutien et ton aide précieuse pour l’utilisation de R et des tests statistiques. Un grand merci également à tous ceux qui, de près ou de loin, se sont penchés sur mon travail pour y apporter leur éclairage et leurs suggestions : David Mouillot et Sébastien Villéger, de l’UMR Ecosym à Montpellier ainsi que Sarah Nahon, du CRIOBE (EPHE) à Perpignan. Les pistes que vous nous avez suggérées ne tarderont pas, je l’espère, à être explorées ! Merci à tous ceux qui depuis plusieurs années ont agrémenté la base de données isotopes de nombreuses nouvelles espèces et individus. Par ailleurs, je tiens également à exprimer mes remerciements à l’équipe administrative du Laboratoire Arago : merci à Philippe Lebaron, directeur de la station marine, de m’avoir accueillie dans un cadre si idyllique. Merci au personnel de tous horizons que j’ai pu croiser durant mon séjour, et qui ont rendu celui-ci agréable tout du long. Et bien sûr, merci à tous les étudiants de passage, stagiaires et thésards de la station pour leur gentillesse et leur bonne humeur &. Vous êtes nombreux et je vous exprime à tous ma profonde reconnaissance. Merci notamment à Sandrine, le trio d’Amandines, Sara, Mathieu, Antoine, Karine, Sabrina, Margot, Marlène, Cynthia, Elsa, Marine, Marion, Romain ainsi qu’à tous les M1 qui ont mis le bazar et l’ambiance dans le foyer des étudiants &. Merci à Maria-Jesus from Chile, à Sven from Ireland et à Tim from Germany de m’avoir permis d’améliorer mon anglais et d’en savoir un peu plus sur leurs contrées respectives. Les soirées plage et les difficiles nuit d’été passées à tenter de dormir malgré les concerts de la place centrale resteront gravées dans ma mémoire, dans le bon sens du terme ! Merci finalement à mes professeurs de l’Agrocampus Ouest à Rennes, toujours disposés à nous apporter leur aide et leur soutien, et à ma famille pour m’avoir encouragée durant ces deux années de master à l’autre bout de la France. J’espère avoir été à la hauteur de cet investissement. Merci à tous ! ! Résumé étendu Les enjeux actuels de gestion de la biodiversité nécessitent une compréhension fine et précise des processus écologiques qui gouvernent le fonctionnement des assemblages complexes, tels que les communautés de poissons de récifs coralliens. Un suivi rigoureux et des décisions efficaces s’appuient notamment sur la connaissance de la vulnérabilité de ces systèmes, afin de prévoir leur capacité de résistance ou résilience suite à des contraintes ou des perturbations (pêche, pollutions, etc.) Dans le Pacifique Sud, la diversité des assemblages de poissons de récifs coralliens et les bénéfices humains qui en découlent sont d’une importance considérable. Ces assemblages sont également d’une très grande complexité, tant par leur composition taxonomique ou fonctionnelle que par leur dynamique et leur fonctionnement. Leur étude nécessite donc plusieurs approches complémentaires dont les caractéristiques de la niche trophique font partie. Au cours de ce travail, les caractéristiques de la niche trophique d’une espèce donnée seront résumées par la surface de niche trophique qu’elle occupe. Nous supposons que certaines espèces adoptent un comportement « spécialiste » (faible diversité de proies, faible variabilité intra-spécifique du niveau trophique) tandis que d’autres ont un comportement plutôt opportuniste, donc « généraliste » (forte diversité des proies, possibilité de forte variabilité de niveau trophique entre deux individus de la même espèce). L’hypothèse principale de notre travail est que la combinaison de certaines caractéristiques intrinsèques à l’espèce (classe alimentaire, taille, etc.) avec différents critères environnementaux (biotiques ou abiotiques) vont favoriser la dominance, en termes de fréquence ou d’abondance, d’espèces spécialistes, ou au contraire d’espèces généralistes. L'importance relative d'espèces spécialistes ou généralistes peut influer sur les caractéristiques des communautés, telles que résilience, résistance, stabilité. L'étendue de la niche trophique de plusieurs espèces de poissons de récifs a été mesurée via l’utilisation des ratios des isotopes stables 13C et 12C du carbone (!13C) et 14N et 15 N de l’azote (!15N). Ces isotopes sont issus du prélèvement de muscle blanc dorsal de plusieurs milliers d’individus regroupant en moyenne 300 espèces, échantillonnées au niveau de trois régions contrastées du Pacifique Sud : Nouvelle-Calédonie, Archipel des Gambier et Atoll de Mururoa. Des prélèvements de sources de carbone (algues, plancton, matière organique en suspension) ont également été effectués sur le site de Mururoa. Pour chaque espèce, nous disposons également d’informations sur leurs traits de vie issus de la base de données FishBase et des campagnes d’observations de l’IRD et de l’EPHE. Enfin, des informations en termes de biomasse (g/m"), fréquence (présence/absence par site) et abondance (poissons/m") des espèces sont disponibles via une base de données regroupant les observations de transects UVC (Underwater Visual Census) effectués au niveau des régions étudiées. Nous avons défini une "niche isotopique" pour chaque espèce, basée sur la variabilité intra-spécifique de !13C (représentative de la diversité de sources consommées) couplée à la variabilité intra-spécifique de !15N (représentative de la diversité des niveaux trophiques occupés). Cette surface de niche isotopique est calculée à l’aide de la méthodologie « Stable Isotope Bayesian Ellipses in R » développée par Jackson et al. 2011 (SIBER - Package SIAR in R Development Core Team 2012). Cette niche isotopique sera donc utilisée comme proxy de la niche trophique, en tenant compte de certaines contraintes propres à l’utilisation des isotopes stables afin d’éviter au maximum l’introduction de facteurs confondants. Une première série d'analyses a examiné les variations de !13C et !15N en fonction de la taille des poissons, de leur espèce, du groupe fonctionnel, du site ou de la région. Il y est ! notamment montré un effet de la taille qui se caractérise le plus souvent par une augmentation du ratio isotopique de l’azote avec la taille individuelle. Toutefois, des tendances opposées sont également observées de façon moins fréquente. De plus, ces relations entre taille et signature isotopique L’estimation de la niche isotopique des espèces communes à au moins deux des trois régions étudiées suggère une forte variabilité de celle-ci, que l’on peut imputer à (1) la variabilité inter-régionale de la signature isotopique des sources consommées, (2) à une plasticité réelle du comportement alimentaire d’une même espèce d’une région à une autre. Le rang des espèces le long du gradient spécialistes-généralistes reste cependant stable d’une région à l’autre ; ce qui suggère un maintien du comportement alimentaire moyen des espèces les unes par rapport aux autres entre deux régions contrastées. Mis à part l'alimentation, les effets sur la niche isotopique des autres traits de vie tels que la grégarité, le comportement nycthéméral, la mobilité ou encore le niveau dans la colonne d’eau où évolue une espèce sont très variables et ont difficilement pu être reliés à des facteurs environnementaux. De même, aucune relation significative n’a été mise en évidence entre la surface de la niche isotopique et : • • La répartition géographique (indicatrice du degré d’endémisme) d’une espèce L’indice de Singularité Fonctionnelle qui permet de classer les espèces en fonction de l'originalité de leurs traits de vie Comme précisé précédemment, seul le régime alimentaire indique une différenciation en termes de largeur de niche trophique occupée, les espèces herbivores et omnivores présentant en moyenne une niche trophique plus étendue, donc une alimentation relativement plus diversifiée que les piscivores stricts, carnivores, planctonophages ou coralivores. Toutefois, ce pattern n'est pas stable d’une région à l’autre, la surface de la niche isotopique n'étant que faiblement liée aux traits de vie des espèces. Connaître le mode de vie précis d’une espèce ne suffit donc pas à prévoir la niche trophique qu’elle occupera par rapport à une autre espèce ayant le même comportement alimentaire. Seul ce dernier facteur sera donc retenu pour analyser l'influence des facteurs environnementaux sur la niche isotopique. En revanche, une relation significativement positive a été mise en évidence entre la surface moyenne de niche isotopique et la fréquence, biomasse ou abondance des espèces herbivores. Le phénomène inverse été observé chez quatre espèces coralivores. Cela suggère que les interactions spécifiques au sein d’un groupe fonctionnel influencent le potentiel d’occupation de la niche trophique, et qu’elles entraînent des réponses différentes en termes de diversification du régime alimentaire selon le groupe fonctionnel considéré. La seconde étape de ce travail se concentre sur l’atoll de Mururoa, où l’échantillonnage est plus exhaustif, afin d’analyser l’influence d’un gradient environnemental. Dans les écosystèmes récifaux, le degré d’exposition à l’océan est connu pour être un facteur impactant la composition des assemblages ichtyologiques. L’atoll de Mururoa a été subdivisé en quatre zones de superficies comparables selon un gradient d'exposition à l’influence océanique. Les niches isotopiques des espèces disponibles sur au moins deux zones de l’atoll ont alors été estimées à l’échelle de la zone. Les espèces étudiées présentent une surface de niche isotopique très souvent significativement différente d’une zone à l’autre mais sans qu’un pattern stable ait pu être mis en évidence, même en tenant compte des traits de vie en particulier le régime alimentaire. Dans la zone lagonaire (zone centrale de l’atoll, caractérisée par l’absence de récif mais la présence de pinacles) la surface isotopique de la niche est cependant plus restreinte pour la majeure partie des espèces étudiées, y suggérant une plus grande homogénéité des ressources. Cette seconde étape ! confirme la capacité d’une forte versatilité alimentaire de la plupart des espèces, le comportement alimentaire étant contraint par les conditions locales. La dernière étape a consisté à analyser la structuration de la niche isotopique globale, c.à.d. celle occupée par l’ensemble des individus regroupés par groupes fonctionnels, et ce en tenant compte de la zonation de l’atoll de Mururoa. Une méthodologie récemment dérivée des métriques de Layman et al. 2007 a permis d’estimer la surface de niche isotopique occupée par les individus d'un régime alimentaire tout en tenant compte de l’abondance ou de la biomasse des différentes espèces qui le composent. Ceci a permis de mettre en évidence certaines spécificités de groupes fonctionnels, notamment les herbivores, ces derniers montrant une combinaison de plasticité au niveau de l'individu et de l'espèce qui leur confère probablement une grande adaptabilité. Ce travail, qui porte sur la plus importante base de données sur les isotopes stables de C et N chez les poissons, suggère donc: • • • • • une forte adaptabilité alimentaire, dans les limites de chaque comportement trophique, qui devrait favoriser la résistance et la stabilité des communautés de poissons de récif face aux perturbations confirmation que les communautés très diversifiés sont probablement plus résilientes les stratégies opposées des herbivores et coralivores en termes de relation entre largeur de niche trophique et abondance sont l'indication que les effets de ces poissons sur l'équilibre algue-corail est sans doute bien plus complexe que ce qui est actuellement décrit l'absence de relation entre la largeur de la niche trophique et l'endémisme ou la singularité fonctionnelle est un indice que le rôle écologique, en particulier en termes de résistance et résilience, de telles espèces est probablement non lié à leurs caractéristiques trophiques l'absence de relation entre la largeur de la niche trophique et la plupart des traits de vie est signe d'un grande flexibilité dans les relations trophiques au sein des groupes fonctionnels, ce qui pourrait favoriser résilience et stabilité. Les apports et les limites de l’utilisation des isotopes stables pour l’étude de la structure fonctionnelle des communautés de poissons seront discutées. Enfin, quelques perspectives seront alors présentées. ! Table of contents Introduction & Review ................................................................................... ^! ! A) Reef Fish Communities in the South Pacific ........................................................ I. The ecological and trophic niche in ecology : specialists VS generalists .......... II. Trophic specialization in reef fishes : a few examples from the literature ....... III. Biodiversity and community structure of reef fish across scales .................... ^! ^! L! L! ! ! ! ! B) On the use of stable isotopes to explore the trophic niche in ecology .................. C! I. Principle & fields of application ........................................................................ C! II. “A niche for isotopic ecology” (Newsome et al. 2007) .................................... C! III. Limitations and constrains of the use of C and N stable isotopes in ecology .. B! ! ! ! ! In a Nutshell (1) .......................................................................................................................b! ! Materials and Methods ................................................................................... T! ! A) Three complementary databases ............................................................................ T! I. Stable carbon and nitrogen isotopes sampling ................................................... T! II. Field observations data : Underwater Visual Censuses (UVC) ........................ S! III. Specific information : Biological and Behavioral Traits .............................. ^M! ! ! ! B) The population and community niche metrics ....................................................... ^^! I. The SIBER ellipses methodology ....................................................................... ^^! II. The Layman metrics modified by Cucherousset & Villéger (submitted) ......... ^O! ! ! ! In a Nutshell (2) .......................................................................................................................^L! ! Results ............................................................................................................... ^N! ! Part 1 : Preliminary analysis ....................................................................................... ^N! (1) The differences between sources in a given Region ........................................... ^N! (2) Effect of individual size on the isotopic signature ….......................................... ^N! ! ! ! ! Part 2 : Effect of inherent characteristics on the isotopic niche width ........................ ^B! (1) The influence of biological and behavioral traits .............................................. ^B! (2) The influence of the Geographical Range and Functional Distinctiveness ....... ^B! (3) The influence of the frequency, biomass or density ............................................ ^T! ! Part 3 : Effect of environmental gradient in the isotopic niche width ........................ OM! (1) The evolution of the isotopic niche width with habitat variations : ! feeding plasticity ?.................................................................................................... OM! (2) The community scale : insights from the methodology of modified ! Layman’s metric .......................................................................................................... OM! ! ! ! ! ! Discussion ......................................................................................................... OO! ! Specialization in reef fish communities ...................................................................... OO! Effect of individual size on isotopic signatures ............................................... OO! Specialist to generalist gradient : is a typology at hand ? ............................... ON! Effect of the biology of species on isotopic niche width ................................ ON! Effect of endemism and functional distinctiveness on isotopic niche width .. OC! Effect of frequency, biomass and density on isotopic niche width ................. OB! Effect of an oceanic gradient ........................................................................... Ob! ! ! ! ! ! ! ! Stability versus Versatility .......................................................................................... OT! ! Comments & Critics of the methods : which perspectives ? ....................................... OS! ! Conclusion ........................................................................................................ LM! ! Literature Cited .................................................................................................. LO! ! Table of appendices ! !EE0-5.S*=*J*Geographical repartition of coral reef .................................................................... 8% % !EE0-5.S*==*J Major application of stable isotopes in ecology.................................................... * 9% % !EE0-5.S*===*J*Sampling design for isotopic analysis ................................................................. * :% % !EE0-5.S*=Q*J*Underwater Visual Censuses in Mururoa ............................................................ * 88% % !EE0-5.S*Q*J*R routine for the estimation of Functional Distinctiveness Index......................... * 89% % !EE0-5.S*Q=*J*Origin of the Layman’s derived metrics............................................................... 8:% % !EE0-5.S*Q==*J*Isotopic signatures of diet sources in New-Caledonia ........................................ *8;% % !EE0-5.S*Q===*J*Details about effect of size on isotopic signature .............................................. * 8<% 15 % !EE0-5.S*=T*J*Example of the relationship between size and ! N for six species ..................... *8=% % !EE0-5.S*T*J*Main information about SIBER ellipses by specie within a region ......................* >?% % !EE0-5.S*T=*J*Table of significancy of the biological factors effect on isotopic niche width .... >9% % !EE0-5.S*T==*J*Residuals of the model (1) and (2) performed on Appendix XI .......................... >:% % !EE0-5.S*T===*J*Plots of isotopic signatures in zones of Mururoa ............................................... >@% 2 !EE0-5.S*T=Q*J*Table of density (individuals/m ) for each diet (regardless to specie) % within zones of Mururoa ........................................................................................................... *>;% % !EE0-5.S*TQ*J*Layman’s derived metrics graphical outputs ...................................................... *><% % !EE0-5.S*TQ=*J*Divergences between FishBase and isotopic data .............................................. 9=% ! List of figures * C.8D10* U*?1cSA* J* 7(1%! 3)&.! /G1#(2/G/+&%! *0! <)*#*>)2(#! %(G1#)+>! 0*'! d^LQed^CP! (+(#]%/%! .(I/! <//+! 1/'0*'G/4c! ! 6(2.! 0)##/4! 2)'2#/! '/1'/%/+&%! (! %(G1#)+>! %&(&)*+c! f'/]! ('/(%! ('/! 2*'(#! '//0c! D$<4)I)%)*+! g*+/%!*0!7$'$'*(!('/!<*3/4!)+!4*&&/4!#)+/%!?;!,!;(>**+[!D!,!D./#&/'/4[!D@!,!D*$&.:h/%&[!@!,!@2/(+Ac! * C.8D10*@*?1c^^A!,!J)>$'/!O!,!D&/1!0*##*3/4!<]!&./!D"96E!/##)1%/!'*$&)+/!&*!/%&)G(&/!&./!)%*&*1)2!+)2./! 3)4&.!*0!(!>'*$1!*0!)+4)I)4$(#%!)+!&./!d^LQed^CP!)%*&*1)2!%1(2/c! * C.8D10* V* ?1c^CA* J* E/2$'%)I/:1*'&)*+)+>! 1#*&%! *0! &./! /00/2&! *0! F)/&! *+! &./! '/#(&)*+%.)1! </&3//+! )+4)I)4$(#! %)g/! ?-;[! GGA! (+4! ?(A! d^LQ! *'! ?<A! d^CP! )+! ?^A! P/3:Q(#/4*+)(! (+4! ?OA! 7$'$'*(c! F/>'//! *0! %)>+)0)2(+2/!)%!>)I/+!0*'!/(2.!0)+(#!<*R!?i!1:I(#$/Ac! * ^C ^L C.8D10*W*?1c^bA!J!U#*&%!*0!d P!(>()+%&!d Q!%)>+(&$'/%!*0!0)%.!%1/2)/%!*0!4)00/'/+&!4)/&!2#(%%!)+!?(A!P/3: Q(#/4*+)([!?<A!f(G<)/'!"%#(+4%!(+4!?2A!7$'$'*(!'/>)*+%c!-*!)G1'*I/!#)%)<)#)&][!4/&()#!*0!%1/2)/%!.(I/! +*&!<//+!>)I/+!(+4!*+#]!&./!G/(+!I(#$/!*0!/(2.!F)/&!2#(%%!3)&.!/''*'%!<('%!('/!'/1'/%/+&/4c! ! C.8D10* X!?1c^bA!J!7/(+!(+4!2*+0)4/+2/!)+&/'I(#%!?SCjA!I(#$/%!*0!D6=c9!/%&)G(&/4!4$')+>!&./!7Q7Q! 1'*2/4$'/!*0!2(#2$#(&)*+!*0!&./!)%*&*1)2!2*'/!+)2./!0*'!/(2.!%1/2)/!)+!?(A!P/3:Q(#/4*+)(!kD!f(G<)/'! "%#(+4%! (+4! ?<A! 7$'$'*(! kD! f(G<)/'! "%#(+4%c! D1/2)/%! ('/! '(+l/4! <]! F)/&! 2#(%%! &./+! )+! (#1.(</&)2! *'4/'c ! C.8D10*Y!?1c^SA!J*E/#(&)*+%.)1!</&3//+!D6=A%I(#$/%!(+4!&./!f/*>'(1.)2(#!E(+>/!*0!%1/2)/%!)+!?(A!P/3: Q(#/4*+)([!?<A!f(G<)/'!"%#(+4%!(+4!?2A!7$'$'*(!'/>)*+%c! ! C.8D10* A*?1c^SA! J* E/#(&)*+%.)1! </&3//+! D6=A! I(#$/%! (+4! &./! J$+2&)*+(#! F)%&)+2&)I/+/%%! "+4/R! *0! %1/2)/%!)+!?(A!P/3:Q(#/4*+)([!?<A!f(G<)/'!"%#(+4%!(+4!?2A!7$'$'*(!'/>)*+%c! ! C.8D10*Z!?1c^SA!J*E/#(&)*+%.)1!</&3//+!D6=c9!I(#$/%!(+4!&./!?(A!0'/Z$/+2][!?<A!'(+l!)+!<)*G(%%!(+4!?2A! '(+l!)+!4/+%)&]!*0!%1/2)/%!)+!7$'$'*(c!-./!#)+/%!2*''/%1*+4!&*!#)+/('!'/>'/%%)*+%c!D$2.!(!'/>'/%%)*+! G*4/#! )+! f(G<)/'! "%#(+4%! 2*$#4! +*&! </! 1/'0*'G/4! (%! &./! +$G</'! *0! (I()#(<#/! %1/2)/%! 3(%! +*&! %$00)2)/+&c ! C.8D10*[!?1cOOA*J!6I*#$&)*+!*0!?(A!&./!-'*1.)2!E)2.+/%%!?-E)2A[!?<A!&./!-'*1.)2!F)I/'>/+2/!?-F)IA[!?2A!&./! -'*1.)2! F)%1/'%)*+! ?-F)%A[! ?4A! &./! -'*1.)2! 6I/+/%%! ?-6I/+A! (+4! ?/A! &./! -'*1.)2! 5+)Z$/+/%%! ?-6I/+A! )+4)2/%!0*'!/(2.!F)/&!2#(%%!(22*'4)+>!&*!&./!g*+/!*0!7$'$'*(c! * ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! List of Tables ,7M:0*U*?1cBA*J*7()+!#)G)&(&)*+%!(+4!+//4/4!2($&)*+!3)&.!%&(<#/!)%*&*1/%!(+(#]%)%!)+!*$'!0'(G/3*'lc* * ,7M:0* @*?1cSA* J* F/&()#/4! +$G</'! *0! %1/2)/%e)+4)I)4$(#%! %(G1#/4! )+%)4/! /(2.! '/>)*+! (+4! &./! %$<4)I)%)*+%!G(4/!3)&.)+!&./!'/>)*+!*0!7$'$'*(c! * ,7M:0*V*?1c^MA*J*F/%2')1&)*+!*0!&./!9)*#*>)2(#!(+4!9/.(I)*'(#!&'()&%!$%/4!)+!&.)%!'/1*'&c!! ! ,7M:0* W* ?1c^OA* J! F/%2')1&)*+! (+4! $%/! *0! &./! %)R! 2*GG$+)&]! G/&')2%! 1'*1*%/4! <]! Q$2./'*$%%/&! m! k)##K>/'!0'*G!;(]G(+!/&!(#c!?OMMbAc!! ! ,7M:0*X*?1c^TA!J!E/%$#&%!*0!-$l/]!\DF!1*%&:.*2!&/%&!1/'0*'G/4!*+!&./!G/(+!D6=c9!I(#$/%!*0!%1/2)/%!&*! /R(G)+/!&./!4/&()#/4!)+0#$/+2/!*0!/(2.!<)*#*>)2(#!0(2&*'!&(l/+!)+4)I)4$(##]! ! ,7M:0* Y*?1cO^A* J* k(')(&)*+! *0! ?(A! )%*&*1)2! %)>+(&$'/! (+4! ?<A! D6=c9! G/(+! I(#$/! *0! (! >)I/+! %1/2)/! )+! 4)00/'/+&!g*+/%!*0!7$'$'*(c!-./!%1/2)/%!#)%&/4!('/!&.*%/!(11/(')+>!)+!(&!#/(%&!&3*!*0!&./!0*$'!g*+/%! (+4!0*'!3.)2.!3/!2*$#4!G(l/!4)'/2&!2*G1(')%*+%!(%!&./)'!)+4)I)4$(#!%)g/!?-;A!'(+>/!('/!%)G)#('!0'*G! *+/!g*+/!&*!(+*&./'c% % ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Introduction & Review ! INTRODUCTION & REVIEW I ! Explaining the origin and the future of biodiversity patterns around the world is at present a major challenge. Many systems have been facing diversity loss during the last decades. Global warming and anthropogenic disturbances are causing ecological disequilibrium which consequences have to be evaluated in order to improve resource management. Even though they occupy a very small proportion of the ocean surface (less than 0,1% of the total area of world oceans) and are restricted to subtropical regions (See Appendix I), coral reef systems shelter nearly 25% of the total fish biodiversity. Thus they are one of the most complex and diversified systems in the world. Because of the distribution of these species along strong environmental gradients (Parravicini et al. 2013) these fish assemblages show a great potential for ecological studies and understanding how communities are structured. The major traits of the ecology of reef fish communities will be briefly reviewed along with some of the major hypotheses on the processes underlying the organization of these communities in order to set the outline of this report. A) Reef Fish Communities in the South Pacific I. The ecological and trophic niche in ecology : specialists VS generalists The study of niche width represents a key point in ecology, however it covers different concepts and there is no general consensus. The first ideas about ecological niche were developed mainly by Grinnell (1917), Elton (1927) and Gause (1936) (See Chase & Leibold, 2003 for a review). For a better understanding of our framework, the major different approaches used to define the niche of a given species are briefly described hereafter. Hutchinson (1957) defined the ecological niche as an abstract hypervolume, i.e a ndimensional set of points in a space which axes represent varied environmental and/or biotic variables. The environmental variables are subdivided in two categories that are the bionomic axes (which represents the pool of resources used by a species) and the scenopoetic axes (the set of habitat, physical and biogeochemical variables where the species could live) (Hutchinson, 1978). Afterwards more precise distinctions were proposed : !!!!"#$!%&'(($))'*(!+,!-#$!.)-/('*(!('0#$!!" An ecological niche may be defined in terms of the subset of the n-dimensional space taken by a species from its environment, but also in terms of the impact of that species. Grinnell (1917) describes the ecological niche as « the response of species to a given set of variables » while Elton (1927) is more specifically focused on « the impact of species in the environment », for instance in terms of foraging strategies. Thus the trophic niche is usually linked to the Eltonian approach of the ecological niche. !!!!"#$!12(3*4$(-*)!+,!-#$!&$*)'5$3!('0#$!!" The whole range of environmental constrains (both biotic and abiotic) rarely enable species to be as efficient as they theoretically could. Indeed, the biotic relationships with other species and the environmental variability of both available resources and conditions may « shape » the volume of the occupied niche of a species (which represents the realized niche), even though the inherent characteristics of individuals (in particular phenotypic features) would provide them a larger range of possibilities (designed as the fundamental niche). Afterwards more precise distinctions need to be given accordingly to the objectives of the study (Devictor et al. 2010). As an example, the study of foraging behavior is related to eltonian realized niche. However the width of the ecological niche is concretely difficult to measure since this notion is strongly dependent upon the considered spatial, temporal or taxonomic scale. Devictor et al. (2010) defined four major types of factors that influence the width of the ecological niche : ! !" • • • • Spatial and temporal variability of environmental conditions (for instance resource availability or temperature range that optimize metabolic pathways, etc.) Species interactions (predation, competition, increasing usually with species richness) Local genetic speciation (that provides different patterns of performance in a given habitat) and phenotypic plasticity of individuals (i.e. their potential of adaptability) The considered level of organization (individual < population < community) Our perception of a niche is also dependant on the methodology and sampling design. Currently, the most common theory states that, due to trade-off constrains, a species should either be adaptable (i.e with a large fundamental niche ; so it can fill in a diversified set of narrower realized niches with respect to the environmental instability) but with low performance, or specialized (i.e with a small range of usable environmental variables, for instance a very specific kind of foraging resource) but with a high performance within the range it belongs to. The degree of specialization of species in a given community or ecosystem will shape its functioning (Devictor et al. 2008). Indeed the respective ecological functions of these species will be different and provide different features against perturbations or constrains (Burkepile & Hay, 2008). A system with a high degree of heterogeneity (in terms of topography or substrat for instance, but also possibly a high species richness) could either shelter generalist species that highly overlap or specialist species with a high degree of resource partitioning, which in each case finally confers stability to the whole system as it ends up into a certain dynamical equilibrium linked to interspecific competition. One may propose the same and opposite hypotheses in an homogeneous habitat (for example in a very small and/or isolated island when the amount of reef corals diversity is low because of larvae dispersion constrains) with either specialized species which may be very competitive and exclude generalists, or oppositely generalists which may be less competitive but more adapted to changes in this homogeneous habitat. However specialized species are more expected to be found in very temporally stable and predictable environment, where the needed subset of environmental resource will be available at all times ; while a changing and constrained habitat would be more adapted for generalist species which could face the changes and therefore be more resilient in case of major perturbation (Munday, 2004). Resilience describes the ability for a system to keep its structure and functioning around a particular set of mutually reinforcing processes, known as a regime (Notion introduced for the first time by Holling, 1973). Thus, a stable ecosystem where most of species are specialists is likely to be less resilient than a system dominated by generalists ; as species will be more sensitive to habitat or resources variation (Pratchett et al. 2012). As a consequence perturbations in stable environments are likely to generate larger risks than in less stable environments supporting a similar but more generalized diversity. In the case of very diversified and threatened systems like coral reef communities, the study of specialization is thus essential to evaluate the vulnerability and the risks these communities are facing. Defining the degree of specialization of these fish could enable better management measures to protect them and more accurate predictions regarding biodiversity erosion. !!" INTRODUCTION & REVIEW I ! II. Trophic specialization in reef fishes : a few examples from the literature Several hypotheses have been proposed to explain the high diversity of coral reef fishes (see review of Hixon, 2011). One hypothesis states that high intra- and interspecific competition prevents species from becoming so dominant as to exclude others species. Diversity is maintained thanks to resource or spatial partitioning. Therefore habitat or trophic specializations are expected to be high when competition is important, i.e. when species richness and abundance are high relatively to the available resources and shelters. Literature confirms some of these statements, for instance on the Holocentridae family in the Atlantic (Gladfelter & Johnson, 1983), Pomacentridae family in the Great Barrier Reef (Berumen & Pratchett, 2008) but a second hypothesis could be proposed. A generalist behavior could also enable a high diversity as the competitive exclusion would be less important. For instance in the Toliara lagoon (Madagascar), Dascyllus aruanus is a highly habitat-specialized species but within this population there is a partitioning of the resource both spatially and with fish size. As a result the population of D. aruanus can be considered as foraging generalist even if at the individual level these fish are specialists (Frédérich et al. 2010). Similarly, a study conducted on two others damselfish species showed no distinction between their trophic niche but only between their spatial niche (Dromard et al. 2013). Slatyer et al. (2013) in a review suggested that endemism should be associated with a small niche, as the distribution of the species is often restricted to quite specific habitats while widespread species use more common resources (Hanski, 1993). However at present the life-history of endemic species has not been much studied in reef fish communities. III. Biodiversity and community structure of reef fish across scales The dynamics of biodiversity and the community structure in reef fish strongly depends on the spatial, temporal or even taxonomic scale of the study. Most of the existing literature focuses on local drivers of biodiversity but regional drivers also need to be accounted for (Tuya et al. 2011). According to Pandolfi (2002) for instance, determining the temporal patterns of communities dynamic is essential to determine the time range in which anthropic disturbance could modify the functioning of these systems. Besides both biotic and abiotic drivers could shape the community structure in a reef system and they are expected to be strongly linked, even though the biotic interaction factors may apply their influence mainly at low geographical scales. Finally, the pool of species that we find at a given location depends on the constraints applied by that we could call “ecological filters”. Biogeographical factors, as reef area and reef isolation are important drivers of reef fish communities structure (Mellin et al. 2010). For example, the islands theory developed by MacArthur and Wilson according to which a large and not isolated area should shelter a higher biodiversity has been showed to apply to coral reef (Galzin et al. 1994 ; McNeil et al., 2009 ; Huntington & Lirman, 2012). Species richness, density and biomass may change according to regional drivers like sea surface temperature or nutrient input (Mellin et al. 2010b) whereas locally, drivers like habitat complexity (i.e the topography of the reef), coral or algal covering, or even wave exposure, distance to the ocean, reef type (Grimaud & Kulbicki, 1998 ; Fulton et al. 2005) as well as biological features like predation (Hixon & Beets, 1993), competition, recruitement and immigration influence the specific composition of communities (Cornell & Karlson, 2000 ; Bellwood & Hughes, 2001 ; Bellwood et al. 2002). Stability, resilience, resistance are a matter of scale, reef fish assemblages being more stable at large scale than they are at a local scale (Pandolfi, 2002 ; Tuya et al. 2011 ; Huntington & Lirman, 2012). A comparison of three major reef systems (Great Barrier Reef of Australia, French Polynesia and Caribbean) shows that the structures of Labridae fish communities are similar despite of very different geological and geographical histories, and a ! !" very low species overlap between two systems like Caribbean and Polynesia (Bellwood et al. 2002). This emphasizes the need to take functional groups in account when studying community dynamics across different locations (Sale & Guy, 1992), so that species turnover does not obscure our understanding of the influence of regional or local drivers. A functional group is usually defined as a group of individuals or species that share a comparable role or impact on their ecosystem, due to their intrinsic phenotypic or behavioral features (i.e grazing, filtering, chasing, etc). The functional role is often linked to trophic requirements, for example within reef fish herbivores there are several feeding behaviors (grazing, scrapping, browsing and so on) which all target the same resource but end up in different actions on the resource dynamics. Some functions may be central to the ecological stability of a system and therefore their disappearance could induce major changes, the extreme case being the transition between a stable and productive ecosystem toward a degraded one. Yet all functional groups are not represented equally. In particular many rare functions are filled by rare species (rare because of their endemism and/or their low abundance) (Mouillot et al., 2013). Therefore some unique functions are characterized by a low redundancy1 which increases their vulnerability and their disappearance may reduce the potential of resilience of the system (Hoey & Bellwood, 2009 ; Mouillot et al. 2013). Moreover, increasing species richness usually favors the redundancy of functional groups that are already well represented (Halpern & Floeter, 2008 ; Guillemot et al. 2011). Even though the same functional groups are found on most reefs, redundancy of each functional groups is often unstable from one geographical area to another, attributing to each system unique characteristics (Bellwood et al. 2004). Thus, in order to preserve these systems, it is essential to understand their vulnerability. This notion of vulnerability seems quite elusive as it also depends on actual existing threats and other ecological or biological characteristics, but it has often been linked to the functional redundancy, considering that a system represented by functional groups with low redundancy will be more vulnerable than a system in which redundancy is high (McCann, 2000 ; Bellwood et al. 2004). In conclusion, it is important to better understand functions in a community as they insure the stability and the dynamic of the whole community. However, species redundancy in functional groups may not be the only parameter that matters. As already mentioned, the degree of specialization of each species in a given functional group within a habitat is also important, as a given functional group is likely to be more vulnerable if only composed by specialist species. However this approach has not been much used in ecological studies of reef fish communities. Our knowledge on trophic specialization remains very patchy, with most studies focusing on a single species or genus within narrow spatial limits. Patterns of reef fish specialization have never been measured at a large scale or for a large pool of species within different functional groups, and generally researchers focus on spatial specialization more than on trophic specialization. The latter could be a particularly informative approach. Foraging behavior in reef fish is mainly measured by two methods: - Study of behavior in the field with direct observations - Study of stomach contents However, these methods rapidly reach their limitations when large scale studies are needed. For instance, stomach content does only represent a snapshot of the diet of a given individual at a given time, but does not provide a temporal integrated measurement. Similarly stomach contents and field observations cannot provide information about what is really assimilated by fishes. Thus these methods enable to observe some of the impact fish may have on their environment, but they yield part of the information on the true width of their trophic niche. This is where stable isotopes may constitute as a useful tool. 1 !!" : Redundancy : In this case, the number of species within a functional group INTRODUCTION & REVIEW I ! B) On the use of stable isotopes to explore the trophic niche in ecology I. Principle & fields of application By their long-term stability in natural environment and the development of recent effective technologic tools such as mass spectrometry, specific non-radiogenic isotopes have become well used tools in biology (Post, 2002 ; Wolf et al. 2009). Stable isotopes are used in many fields. In marine biology and ecology, they may be classified in three main approaches presented in Appendix II. In ecology, the most common used isotopes are the ratios 12C/13C and 14N/15N. These two elements are predominant in metabolic pathways, biogeochemistry and biomass flows across trophic webs, but even though these ratios are relatively similar among biological compartments, slight differences are detectable as a result of fractionation processes (the proportion of heavy isotope conserved from source to product) and source characteristics (Fry, 2006). This phenomenon is called “enrichment” and leads to each species or individual to have a quite unique isotopic signature with enables to reconstruct some of its history (Fry, 2006). Nitrogen isotope ratio !15N enables the assessment of a species trophic level thanks to enrichment processes (DeNiro & Epstein, 1981). Indeed each trophic level shows an increase of the !15N value which is a result of bioaccumulation of heavy isotopes 15 N from diet whereas 14N are easier to eliminate than 15N by metabolism and excretion. This tracer enables to determine the position of a species relatively to the others in a trophic web and is thus described as a predictor of the bionomic axis of a trophic niche (Hutchinson, 1978). Oppositely, carbon isotope ratios !13C are rather stable through trophic levels and therefore enable to trace back the origin of the carbon sources in a given environment (DeNiro & Epstein, 1978). It is thus described as a predictor of the range of scenopoetic axes of a trophic niche (Hutchinson, 1978). II. “A niche for isotopic ecology” (Newsome et al. 2007) Newsome et al. (2007) provided a well documented review of the perspectives opened by stable isotopes analysis in the field of trophic niche ecology. In order to rank species or individuals within a species along a specialist / generalist gradient, they stated that plotting !13C versus !15N values is somewhat analog to looking at two axes within the n-dimensional niche described by Hutchinson. Indeed the isotopic signature is both influenced by what the individuals consume and assimilate (bionomic axis given by !13C and !15N) and the habitat and range of resources they use (scenopoetic axis given by !13C). Thus a different diet behavior between two individuals of a given species will lead to a different isotopic signature for each of them, and finally a higher variability in the !13C/!15N biplot than two individuals with a similar diet. The !-space given by the combination of !13C and !15N in a biplot thus constitutes an isotopic niche that could represent an accurate description of the trophic niche, as both resources range and inter-individual variability are observable at the same time. In addition, one, two or more species may be represented together in the same !-space in order to compare their degree of specialization. This approach, first described by Bearhop et al. (2004) and developed by Layman et al. (2007) and Jackson et al. (2011) with the use of different statistical metrics, has often been applied to the study of a set of two or three species in a narrow habitat but has never been tested at large scale on a wide pool of species. Our objective is therefore to take advantage of these tools to bring some new insights in the field of specialist to generalist species gradients within reef fish communities and test for the potential influence of several factors (See “In a nutshell (1)” below). However, there are a number of constrains to this approach which we will need to take into account. ! !" III. Limitations and constrains of the use of C and N stable isotopes in ecology The estimation of isotopic niche requires to take into account many biological and technical constrains (Table I). All of these contrains will be taken in account when possible. #$%&'"(")"*$+,"&+-+.$.+/,0"$,1",''1'1"2$3.+/,"4+.5"0.$%&'"+0/./6'0"$,$&70+0"+,"/38"98$-'4/8:")" " ;$-6&+,<"+,1+=+13$&0" ;$-6&+,<"+0/./6'0" >'8+/1"/9"0$-6&+,<" ?+0.+,<3+05"0/382'0" #//"-$,7"0/382'0" @-,+=/87" A8$2.+/,$.+/," B00+-+&$.+/,"8$.'" $,+-)&.+#!) #+."/0%&!#)123) "!!&#&%".&0')4!)&'(+!.&0') *$.'8+$&"%+$0" ?'&+6+1$.+/, %"/0,".0,-)"'"%-!&!) • Choose the most representative tissue : the white muscle integrates patterns of feeding behavior on several weeks. • Standardization could be done to make comparisons easier. • Temporal contrasts have been avoided (i.e sampling is done in a narrow time scale, if possible within a month). • Spatial variability of isotopic baseline need to be assessed. However feeding sources were sampled for one region only, which limits inter regional comparisons. • Sampling sources was limited because of the high fish species diversity and the need for a large spectrum of sources. This is why comparing locations with different isotope ratios for the sources has to be conducted with much caution. • Population VS Individual specialization is not easy to distinguish but should not strongly interfere with conclusions as the global pattern should be conserved. (See also “In a nutshell (2)”) • Coefficient of fractionation may differ according to species, life stage, sex, sea surface temperature, etc. However it is very difficult to assess. We chose to use the mean value currently accepted in literature : 3,4 ‰ for nitrogen and 0,1 ‰ for carbon. • Same conditions as previous paragraph for assimilation rate, but this aspect could not be taken into account here. However this should not alter the generalist-to-specialist pattern if it does exist. Choice and justification • Samples were analyzed in different laboratories. A test on 100 samples indicated no significant differences amongst labs. • Lipid content may influence isotopic signatures, but preliminary tests found no significant difference. Consequently no delipidation was performed. * Related to either the issue, the chosen solutions or both. !!" Post et al. 2002 Boecklen et al. 2011 #".+,&"%)123) • Visualize intraspecific variability : we will account for the individual size by sampling the wider size range as possible. Sex effect could not be tested as it could not be determined in all cases. References * Bearhop et al. 2004 Boecklen et al. 2011 Matthews & Mazumder, 2004 Newsome et al. 2007 Technical cautions Choice and justification Boecklen et al. 2011 Newsome et al. 2007 !"#$%&'()*+!&(') Biological cautions References * INTRODUCTION & REVIEW I ! 6:!;!:<=2>?@@!ABC! ! This review brings up several questions : ! Are the usual ecological statements about specialists / generalists dominance true for reef fish communities? Do some environmental factors favor specialization ? ! Do some functional groups contains more specialized species than others ? Is the trophic niche width conditioned by certain biological traits (Diet, Schooling, etc) ? ! Is the proportion of trophic specialist (or generalist) species an important component of the structure of a given community ? What could we conclude in terms of redundancy and potential resilience ? ! How accurate is stable isotope analysis to study community functioning and stability at a large geographic and taxonomic scale ? Our hypothesis : ! A specialized behavior is an inherent characteristic and might therefore be maintained wherever the species lives. ! Species may however present a potential feeding plasticity across environmental gradients. "#!$%&&'$#!()*)+,#!(#-.##+!$-)(/*/-0!12!3#4$)-/*/-05! ! A singular combination of biological and behavioral traits will favor a specialized feeding behavior: the more original a species happens to be, the more specialized it is. In the same way, a high degree of endemism should favor trophic specialization. ! When competition is expected to be high, most of species will be either specialist due to resource-partitioning or generalist with a high resource overlap. ! In a given functional group, relationship between species may interact with specialization / generalism. 67&*/,)-/'+$!8'4!&'-#+-/)*!'8!4#$/*/#+,#!9! Our framework : The use of an important and large set of data will be used. This will enable to combine the available knowledge about biology of hundreds of reef fish species and their stable isotopes signatures, in a large geographic area of the South Pacific. Data on frequency, biomass and density will also be available for most of these species. ! First we will attempt to classify species from specialist to generalist and see if this classification remains stable across regions. ! Then for each region we will test for the influence of some biological traits and dominance for each species on the isotopic niche width. A sheltered-to-exposed gradient effect on niche width will also be tested in an atoll of French Polynesia. ! Finally a change in organizational scale (from species to functional group, given by diet of species) will be performed in order to analyse the relationship amongst species within a community. ! At last, the use of isotopes to assess for the specialist - generalist structure among reef fishes communities will be discussed and some perspectives will be presented. ! !" Material & Methods ! MATERIALS AND METHODS II ! A) Three complementary databases This work uses three databases which are the result of a large sampling effort in several contrasted regions of the South Pacific Ocean : (i) The isotopes database contains the carbon and nitrogen stable isotopic ratios of an important number of species sampled mainly in South of New-Caledonia and French Polynesia (Gambier Islands and the atoll of Mururoa). (ii) The field observations on reef fish (Underwater Visual Censuses) are compiled in another database, which report biomass, frequency and abundance estimates. (iii) The third database contains the observed characteristics of the sampled individuals, and the life-history traits of each species derived from Kulbicki et al. (2011). I. Stable carbon and nitrogen isotopes sampling This study took place in three contrasted regions covering a large geographical area of the South of Pacific Ocean, from New-Caledonia to French Polynesia as presented in Fig.1. These three regions are distinct mainly by their geomorphology and the different types of reef they present. o South of New-Caledonia (NC) is a continental region that supports a large panel of contrasted coastal environments. There is an important gradient in terrestrial influence from the coast to the barrier reef. NC also shows the most diversified fish assemblages, due to its history and its proximity to the Coral Triangle center of biodiversity (Kulbicki et al., 2011 ; Parravicini et al. 2013). o Gambier Islands (GB) is an archipelago with four major islands surrounded by a barrier reef, at the extreme south-east of French Polynesia. o Mururoa (MU) is an atoll of the Tuamotu archipelago, 400 km west from the Gambier Islands. As in any atoll terrestrial influence is very limited. The reef belt surrounding the lagoon supports a thin emerged sandy land and several motus. Along the East of the atoll, the exposition to ocean is higher due to the presence of a large pass. The atoll of Mururoa, where half of the sampling was performed, could be subdivided into four different areas along an environmental gradients as described in Figure 1 (c). The number of species and individuals captured for each region and area are summarized in Table 2. The detailed list of the families, genera and species sampled, as well as their diet are presented in Appendix III. All the samples used for our analyses were collected at the same season for all areas in a region, in order to reduce the temporal variability of isotopic signature and specific composition of the assemblages. However, the period of sampling could vary between two regions (Table 2), besides the isotopic baseline (main sources isotopic signatures) is presently unavailable for two of the three regions, therefore no direct comparison between regions will be presented. Fishes were stunned with rotenone then picked up with scoop nets. Although it contains carbon, any potential effect of rotenone on the carbon isotope signatures of fish can be considered as negligible since the residues are rapidly cleared from fish organism (Gingerich et al. 1986). The fish were fork-length measured (TL in mm) and placed on ice immediately after their removal from water. For each individual, a fillet of white muscle tissue taken from the dorsal region was dissected for isotopic analyses. Isotopic signatures were accessed after the samples were freeze-dried, and stable isotopes values were converted in ! notation as !13C or !15N = [(Rsample/Rstandard) - 1] x 1000, where R is 13C/12C or 15N/14N. The standard reference was Pee Dee Belemnite carbonate for CO2 and atmospheric nitrogen for N2. " " ! !" " " " " NEW-CALEDONIA (Island) " " GAMBIER (Archipelago) " " " " " " " " MURUROA " (Atoll) " " " " #$%&'(" )"*" +,-." /$01" (2-3,4(2(50." 67" 8$636%$4,3" .,2-3$5%" 76'" 9):;<9)=>" ,5,3?.(." 1,@(" 8((5" -('76'2(AB" " C,41" 7$33(A" 4$'43(" '(-'(.(50." ," .,2-3$5%" .0,0$65B" D'(?" ,'(,." ,'(" 46',3" '((7B" E&8A$@$.$65" F65(."67"+&'&'6,",'("86/(A"$5"A600(A"3$5(."GH"*"H,%665I"E"*"E1(30('(AI"EJ"*"E6&01KL(.0I"J"*"J4(,5MB" " " " N,83(" O"*" P(0,$3(A" 5&28('" 67" .-(4$(.<$5A$@$A&,3." .,2-3(A" /$01$5" (,41" '(%$65" ,5A" 01(" .&8A$@$.$65." 2,A("/$01$5"01("'(%$65"67"+&'&'6,"*"" DQ+RSCT"SEHQ>PE" G#,33"OU)UM" Number of species & fishes sampled )VW".-(4$(."K"X)Y"7$.1" >CLK";QHCPJ>SQ" GE&22('"OU)U"Z"OU)OM" )V:".-(4$(."K")U!!"7$.1" !""$"#'('!%&&,&-' QNJHH"J#"+[T[TJQ" GE-'$5%"OUUXM" O)V".-(4$(."K"OU:!"7$.1" !"%$).'('!%#)$*+' " ):V".-(4$(."K"WW="7$.1" !!".-(4$(."K"V=V"7$.1" )U!".-(4$(."K"V)V"7$.1" )UV".-(4$(."K":!X"7$.1" " Location !"#$%&'()$*+!,-" !%(.$(/(0*+!-" ,1(23*+,-" 425""3*+4-" Period of sampling Average coordinates (Latitude ; Longitude) !"#$%&'('!%#)$*+' II. Field observations data : Underwater Visual Censuses (UVC) The species composition of the sampled sites are available in an ecological database containing the frequency, biomass and abundance for most of the sampled species of MU and GB (such data are not directly available for NC, as UVC data collected there were not simultaneous to isotope sampling). These data were collected by underwater transects performed by IRD and EPHE. The sampling methods were different, IRD data was collected with distance sampling (Labrosse et al. 2002), EPHE data were collected using fixed width transects (Planes et al. 2005). UVC samples are non destructive, repeatable and represent a large proportion of the reef fish communities (Kulbicki, 1990). For the region of Mururoa, the counting strategy is stratified according to the same subdivision as presented in Figure 1. The corresponding details (mainly the location of transects and the number of species collected) are presented in Appendix IV. !" ! MATERIALS AND METHODS II ! III. Specific informations : Biological and Behavioral Traits Biological traits are useful to classify species into functional groups (or functional entities). Traits are also important to understand some of the differences in isotopic ratios observed amongst and even within species. The traits used in this work were extracted from Kulbicki et al. (2011), with a large part of the information obtained from FishBase1. The traits retained for our analyses are indicated in Table 3. A species could not be listed in more than one category by trait. On the basis of all these traits, an index of singularity was estimated for each species. This index is derived from the Evolutional Distinctiveness (ED) described by Isaac et al. (2007) and enables to classify species by a degree of « Functional singularity ». From now this index will be designed as Functional Distinctiveness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he Functional Distinctiveness value for a species will be higher when a species presents a singular combination of traits relatively to the other species. FD value have been estimated for each species within each region separately. As FDs are dependent upon the ! 1 : www.fishbase.org !"# species pool considered, we retained all the species available from checklists within each region (Kulbicki et al. 2011). The R script used to derive Functional Distinctiveness from Evolutional Distinctiveness is presented in Appendix V. B) The population and community niche metrics I. The SIBER ellipses methodology As initially defined by Hutchinson (1957) and more recently by Newsome et al. (2007), ecological niche could be approximated as a surface or volume between two or more axes representing environmental and biotic variables. Stable isotope analysis provides quantitative information on both bionomic (!15N) and scenopoetic factors (!13C). Thus the core isotopic niche area could be! revealed by fitting standard ellipses to the isotopic data in the bi-dimensional plot of !13C/!15N, as described in Jackson et al. (2011). The standard ellipse area of a set of bivariate data is calculated from the variance and covariance of x and y data and is expected to be less sensitive to sample size than former methods, which enable robust estimations of the isotopic niche provided we have a minimum of ten samples for one species. The steps providing the final mean value of the isotopic niche width of a species in a given region are described in Figure 2 : $%&''"&(")*+*+&,"-.",/&%01+"+123&/"-.","4&/1"" Biplot of !13C/!15N for Species A (filled circles) and Species B (Opened circles) Step 1 : Standardization between 0 and 100. Each value is transformed as : Xstd = 100 # ((Xi - Xmin) / (Xmax - Xmin)) as proposed by Cucherousset & Villéger (2008). Step 2 : Fitting of a standard ellipse on each species, using the covariance table of the bivariate data x and y to define the shape and size of the ellipse : The resultant standard ellipse area (SEA) is given by the formula "ab, a being the semi-major axis of the ellipse and b the semi-minor axis, obtained via the eigenvalues of the matrix. SEA can be corrected in order to account for small sample size : SEAc = SEA x [(n-1)/(n-2)]. Step 3 : The covariance matrix could also be estimated by a bayesian approach, which enable an open number of estimates of the ellipse area (SEAB) and thus the constitution of a confidence interval for SEA in order to perform statistical analyses. 532*+1"6"7"8%19"(&''&:1;"<="%01"8>?@."1''39A1"+&*%3/1"%&"1A%3B,%1"%01"3A&%&93C"/3C01":3;%0"&("," 2+&*9"&("3/;3D3;*,'A"3/"%01"E!FGHE!IJ"3A&%&93C"A9,C1K" !!" ! MATERIALS AND METHODS II ! For our study, 500 simulations were performed for each species within a region or a zone of Mururoa, using a Monte-Carlo Markov-Chain with the following priors as proposed by Jackson et al. (2011) : The priors : µx ~ dnorm(0, !2 = 103) µy ~ dnorm(0, !2 = 103) ! ~ wishart-1(! = 2, V = ) The Likelihood : Yi ~ MVN([µx , µy] , !) ) See Jackson et al. (2011) for detailed methodology on calculation of SIBER metric, and Jackson et al. (2012) for an example of the use of these standard and bayesian ellipses in a biological case of invasion ecology. As indicated in the mentioned steps, sampling locations were treated separately to provide insight into the feeding ecology of species across our study sites and avoid the issues of the missing isotopic baseline. II. The Layman metrics modified by Cucherousset & Villéger (submitted) Layman metrics, based on the elaboration of Convex Hulls in the bi-dimensional " C/" N plot, were developed with the purpose to describe with precision the isotopic niche of a species or assemblage of several species (Layman et al. 2007, Hoeinghaus & Zeug 2008, Layman & Post 2008). These metrics are characterized by a set of seven indices ("13C Range CR, "15N Range NR, Total area of Convex Hull TA, Mean distance to centroid CD, Mean nearest neighbor distance NND and Standard deviation of nearest neighbor distance SDNND) that, beyond the simple description of the area of a niche, also enable to describe the relationship between individuals of a species or between several species of a group (a group being defined, for us, by diet as it represent a functional group), in order to analyze their functional structure and the potential effect of external factors (for example the habitat). However, Cucherousset & Villéger proposed to weight the information brought by these metrics with the abundance, biomass or other representative data of the importance of each « individual » of the group (actual individual or a given species). Thus they derived the seven Layman metrics into a set of six new metrics as described in Table 4 and Appendix VI. 13 15 $%&'(#)#*#+(,-./01/23#%34#5,(#26#17(#,/8#-29953/1:#9(1./-,#0.202,(4#&:#;5-7(.25,,(1#<#=/''>?(.#6.29# # ##@%:9%3#(1#%'A#B"CCDE#*# Metric Abreviation Use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he trophic niche of a species may be represented by its isotopic niche width : • • Mean isotopic signature of !13C and !15N Mean value of the isotopic niche core, given by SIBER ellipses Assumptions : There is a gradient from specialist to generalist species : Narrow niche #! "! $! Large niche A = Specialist species with low degree of individual specialization B = Generalist species with low degree of individual specialization C = Generalist species with high degree of individual specialization !"#$%&'!'(%)$*%)"#$+&#,!-$%&$./%0#*$12$: (Hypothetized factors with statistical test that were used to test them) 1345463785$987:4;<$ #=>3;4=?@=:85$987:4;<$ Individuals size : . Simple linear regression . Model-based recursive partitioning (See Zeileis et al. 2008) Sheltered-to-exposed gradient in the atoll of Mururoa : Specific size Diet Home-Range Schooling Activity Level General linear model runned on !13C, !15N and SEAB. Geographical Range Functional Distinctiveness Biomass/Frequency/Density Large niche, high or low functional redundancy !"# ! Effect of zone tested on !13C, !15N and SEAB with ANOVA for species taken individually. !"#$%!& Narrow niche, high or low functional redundancy Results ! RESULTS III ! Part 1 : Preliminary analyses (1) The differences between sources in a given Region The isotopic signatures of some foraging sources are available but only in Mururoa (Turf algae, Particulate Organic Matter POM and Zooplankton). Other sources have been studied in New-Caledonia in 1996 (See Carassou et al. 2008 and Appendix VII) but have not been recently re-sampled for the current study. No information on the isotopic values of sources is available for the Gambier Islands either. The differences between zones of Mururoa for each foraging sources were tested with a one-sided ANOVA. No significant differences were found for POM across zones within Mururoa (p = 0.19 for !13C and p = 0.57 for !15N), whereas Zooplankton was significantly different for !13C (p = 0.0007) and Turf was also significantly different for !15N (p = 0.034). The Tukey HSD post-hoc test shows that these differences are due to the values of the Ocean zone, that are higher than in the other zones for !13C (i.e closer to zero) and lower for !15N (see also Appendix XIII for graphical representation). As the number of species and individuals in New-Caledonia and Gambier is quite low to work at a local scale, the estimation of the isotopic niche width by the SIBER method will be performed at a regional scale, i.e by pooling the individuals sampled at different locations even though the isotopic baseline remains unknow for each one of them. This fact implies a hypothesis whereby either the isotopic baseline does not change substantially across zones, or if it changes, the ecological pattern given by the isotopic niche core study is still stable from one zone to another and will therefore be observable even when pooling species from different locations. In Mururoa, the particular signature of Ocean zone will be taken into account when studying environmental gradient effects on isotopic niche. (2) Effect of individual size on the isotopic signature The values of !13C or !15N are known to vary with the size of the individuals. As the magnitude of these variations may influence our main hypotheses, they were evaluated for each species within each region. For most species, a significant effect of the individual size exists, at times combined with a significant interaction with the region (see Appendix VIII for details). • The dominant pattern was an absence of relationship between size and !13C but a significant positive slope for !15N with size in a given species ; • However, for 13 species among the 83 with a significant relationship between size and at least one of the two isotopes values, the slope between size and !15N was negative. These 13 species belong to various diets ; • Twenty four species showed no significant relationship between size and !15N ; • The slope between size and !13C was positive for 37 species, not significant for 33 and negative for 19 species. The relationship between !13C or !15N and size for a species may change with the region. The slope of the regression may either increase or decrease or even reverse, from positive to negative for instance, between two regions for the same species, indicating that the relationship is not spatially stable. This could be verified for 33 species and show that 15 species kept a similar slope, 4 species presented a significant change in slope but no slope reversal and 10 species reversed their slope across regions for at least one of the two isotopes (Appendix VIII for details). When looking at the characteristics of these species, no general common biological trait could explain the observed patterns. However the effect of the diet the species belong to has been tested in order to work at a larger functional scale (Figure 3). Indeed diet is expected to be an important factor to discriminate the ontogenetic size effect. ! !"# >9?# # # # # >:?# >!?# >C?# # # # # # # # # # # # # # # # # # # # # # # # $%&'()#*#+#,)-'(.%/)012(3%24%4()).#25#36)#)55)-3#25#7%)3#24#36)#()893%24.6%1#:)3;))4#%4<%/%<'98#.%=)# 94<#>9?#@!*A#2(#>:?#@!"B#%4#>!?#B);0A98)<24%9#94<#>C?#D'('(29E#7)&())#25#.%&4%5%-94-)#%.#&%/)4#52(#)9-6# 5%498#:2F#>G#10/98')?E # Model-based recursive portioning trees (Zeileis et al., 2008) compare the level of importance of each diet in predicting the relationship between size and !13C or !15N. Three steps are followed for each recursive cycle during in developing the tree. A parametric model is first fitted to the data, then parametric instability is tested with generalized M-"uctuation tests (Zeileis and Hornik, 2007). If instability is detected, the data is split into two child data sets, or nodes. The steps are repeated until all data sets are stable. This results in a tree diagram with multiple terminal nodes that hold distinct sets of parametrically stable data, each of which are represented by different models. When included as a factor of portioning in a model-based recursive portioning (Figure 3), it may be noticed that if no pattern were detected for !13C, diet influenced the relationship between size and !15N. In both Mururoa and New-Caledonia, the pattern of portioning is equivalent, low trophic level diets (HM, HS, OM) presenting a negative relationship between individual size and !15N while high trophic level diets (FC, IM, IS, PK) display a positive relationship between size and !15N. Each circle containing a p-value indicates that Diet influences the regression slope between size and !13C or !15N, leading to consequently different regressions shape, which are summarized below. For each node the number of individuals implicated in the analysis is given. Specific examples are given on Appendix IX for one species per diet class. The chosen species are among the most represented in our database and in the field. Both examples for HM and HD show a decrease of the values of !15N with size. For instance, N. lituratus shows a high decrease of !15N values for the smallest size range, then the relationship is more stable if not increasing. Oppositely, A. nigrofuscus shows a stable decrease of !15N values when the size becomes more important. The four last species presented in Appendix IX have an !"# ! RESULTS III ! increasing !15N signature when they grow. However these examples must not be taken as a generality since there is a high variability across species even within the same diet. Part 2 : Effect of inherent characteristics on the isotopic niche width (Regional scale) (1) The influence of biological and behavioral traits As the level of sources may be significantly different amongst regions it is difficult to make direct comparisons. Therefore, the three regions will be mainly compared on a relative scale. The distribution of the mean isotopic signature for each species and for each diet class (Figure 4) follows a similar pattern from one region to another. Piscivore and planktivore species show the highest !15N values followed by corallivore and consumers of mobile invertebrates. The diet with the lowest trophic levels, represented by omnivores and herbivores present the lowest values of !15N. The characteristics of SIBER ellipse for each species within a region are given in Appendix X. The following analysis is performed with general linear model (family = gamma with link function = inverse, as it enables the best fit and better residuals, See also Appendix XII). Each biological and behavioral traits had significant effects on isotopic niche width (Appendix XI), diet being the most important for both New-Caledonia and Mururoa (such a model could not be performed in Gambier Islands as the number of species with more than ten individuals was too low to accurately represent each biological trait). Before examining the effect of the traits in details, the relative gradient from generalist to specialists species for each region was compared. To realize a typology of species according to their degree of isotopic specialization, species were ranked within a region from the largest isotopic niche width (i.e the highest SEAB mean value) to the narrowest isotopic niche score (i.e the smallest SEAB mean value). A graphical comparison of the ellipse areas for species in common across two regions is given on Figure 5. In most instances the order of each species was kept amongst regions within a trophic group. Thus, G. chilospilus presents a larger isotopic niche than G. eurostus in both regions of Mururoa and Gambier Islands for instance. The same pattern is observed for other diets except for “IM” as the rank from large to small isotopic niche core vary between Mururoa and Gambier Islands. As said before, all Biological and Behavioral Traits (Diet, Home-Range, Size of species, Level, Schooling, Activity) had a significant effect on the mean SEAB value (p < 0.001). The details are presented in Appendix XI. A post-hoc test on the outputs of the general linear model presented in Appendix XI was planed (instead of taking each factor individually like a one-sided ANOVA and then using Tukey HSD), using the glht function of the “multcomp” R package. However, a technical problem was encountered which could not be fixed, which is the reason a ANOVA was used instead (Table 5). In Mururoa, the species from Diet class (HM) have a mean SEAB value that is higher than the other functional groups. The effect of the other traits is unstable as it could change across diet. Moreover, both of general patterns and patterns of a given diet could evolve between Mururoa and NewCaledonia : the performed post-hoc test did not rank modalities in the same order. (2) The influence of the Geographical Range and Functional Distinctiveness No significant relationship between the Geographical Range of a given species and its mean SEAB value could be found either in New-Caledonia, Gambier Islands or Mururoa (Figure 6). Thus a narrower trophic niche is not linked to the degree of endemism, at least when using isotopic variability as a proxy. In the same way, no relationship was found between the mean SEAB value of species and their degree of Functional Distinctiveness (Figure 7). Thus even a species that presents a singular combination of biological or ! !"# behavioral traits will not necessarily present a narrower isotopic niche width when using isotopic variability as a proxy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iological traits tested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odalities that need to be considered cautiously because only represented by one species. (3) The influence of the frequency, biomass or density Using the Underwater Visual Census database, the frequency (number of transects where species is present), biomass (g/m2) and density (individuals/m2) of each species were calculated for each zone of the regions of Mururoa and Gambier Islands. The relationship between frequency or the rank in biomass/density of each species, and their mean SEAB value is given in Figure 8. A weak but significantly positive relationship exists between each of these measures and the width of the isotopic niche. When looking at the effect of the Diet (i.e does the slope of the regression change according to the functional group) reveals that the global pattern is mainly linked to herbivores (HM) for all three measures Frequency (ANCOVA : p = 0.0107 for herbivores), Biomass (ANCOVA : p = 0.0139 for herbivores) or Density (ANCOVA : p = 0.0403 for herbivores). For other diets, the regression is never significant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he case of coralivore : A previous SIBER analysis, run with a minimum of eight individuals by species instead of ten, enabled the inclusion of two supplementary species from Chaetodon genus, which are all coralivore to some extent. Even if the minimum number individual is lower than recommended, it is interesting to report that the isotopic niche width of these species decreased with increasing frequency (R2 = 0.92), biomass (R2 = 0.97) or density (R2 = 0.87). Thus these four coralivores species present an opposite pattern relatively to the strict herbivores. The ecological metrics and the community functioning appears to be a better factor for shaping the isotopic trophic niche than biological traits (except for diet) at least for herbivores and coralivores. The next step will be to test the effect of a sheltered-to-exposed environmental gradient on the local isotopic niche width of species, in order to see if ecological features may influence the diversity of their diet in addition to the competition. Part 3 : Effect of environmental gradient in the isotopic niche width # (1) The evolution of the isotopic niche width with habitat variations : feeding plasticity ? The mean SEAB values per diet and zone are presented in Appendix XIII. The zone effect on the isotopic signatures !13C and !15N are summarized in Table 6 ; as well as the zone effect on the isotopic niche width given by the SIBER ellipses for species taken individually. A post-hoc test (glht, R package multicomp) performed on the entire pool of species shows that the Lagoon zone (L) is separate from the others as it presents higher global !15N value and lower !13C value (i.e more negative). The values of the available sources were different in the Ocean zone. This suggests that the change of the global isotopic signature of species between the zones of Mururoa are not - at least not only - the results of the variability of sampled sources. For each species, a zone effect on the isotopic niche width was detected, the only exception being Myripristis berndti. Testing together for the effect of zone on SIBER ellipses and on the isotopic values for !13C and !15N allows to discriminate which species just change their niche width from those which modify both their isotopic niche width and their diet range. Most species present no significant difference in isotopic signature across zones. However for several species, the change in the isotopic niche width is also accompanied by a shift in the consumed sources, but without consistent pattern. (2) The community scale : insights from the modified Layman’s metrics The community scale analysis was performed by pooling species in functional groups (i.e their diet class). Species were added to a functional group when both the isotopic signature and the density metric were available (See Appendix XIV for the global density for each diet). For each zone of the atoll of Mururoa, the six modified Layman’s metrics were estimated using the script by the Ecosym laboratory (IRD, Montpellier) for all functional groups. Graphical outputs of the script, which give a visual representation of the structure of the community and the total area of the biplot occupied, are presented in Appendix XV. The larger this area is, the wider the subsets of both bionomic and scenopoetic axes, where the community could evolve, may be. The Trophic Divergence (TDiv), Trophic Eveness (TEven) and Trophic Uniqueness (TUni) bring a representation of the relationship between species in a given community. The major findings, when pooling all of these results for each functional groups to compare zones of Mururoa (Figure 9) are : - The Sheltered and the Lagoon zones present the narrowest Trophic Richness (< 0.15) for all functional groups except IM in Lagoon zone (0.25). The general pattern for TRich is quite maintained across zones for differents diet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one effect on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a) !13C (a) !15N (b) SIBER Ellipses area K#I#:K# :K#I#:# :K#I#:# :#I#:K#I#K# E#I#K#I#:K# :K#I#:#I#E# K#I#:K#I#:# # K#O#:K# K#I#:K# K#I#:K#I#:#P#E# :#O#:K# :#O#:K# E#O#K#I#:K# # :#I#:K# :K#I#:# :K#I#:# E#I#:K# :#I#:K#P#E# :K#P#:#P#E# # K#P#E#I#:K# :K#I#:#I#E# # E#O#K#I#:K#I#:# :#I#:K#I#K# :#I#:K#I#K# K#I#:K#I#:# # :#I#:K#I#K# E#I#:K# :K#I#K#I#:# K#I#:K# :K#I#:# :K#I#:# :#I#:K#I#K# E#I#K#I#:K# :K#I#:#I#E# K#I#:K#I#:# # K#P#:K# K#I#:K# K#I#:K#I#E#P#:# :#P#:K# :#I#:K# E#I#K#I#:K# # :K#I#:# :K#I#:# :#I#:K# E#I#:K# :#I#:K#O#E# :K#I#:#I#E# # K#I#E#I#:K# :K#I#:#I#E# # E#P#K#I#:K#I#:# :#I#:K#P#K# :#I#:K#I#K# K#I#:K#I#:# # :#I#:K#I#K# E#I#:K# :K#I#K#I#:# K#P#:K# :K#P#:# :K#P#:# :#P#:K#P#K# E#P#K#P#:K# :K#P#:#P#E# K#P#:K#P#:# # K#P#:K# K#I#:K# K#P#:K#P#:#P#E# :#P#:K# :#P#:K# E#P#K#P#:K# # :K#P#:# :K#P#:# :#P#:K# E#P#:K# :#P#:K#P#E# :K#P#:#P#E# # K#P#E#P#:K# :K#P#:#P#E# # E#P#K#P#:K#P#:# :#P#:K#P#K# :#P#:K#P#K# K#P#:K#P#:# # :#P#:K#P#K# :K#P#E# :K#P#K#P#:# - Trophic Divergence (TDiv) is stable across functional groups and zones (around 0.8). The low values for IM in Lagoon zone and PK in Ocean zone are due to the presence of a very abundant species that influence the index (Amblycirrhitus bimacula for IM and Dascyllus flavicaudus for PK). TDiv being relatively high (> 0.5 whatever the considered zone), we may suppose the feeding behavior or/and diet of species to be quite singular for each species within functional groups. - The global pattern of Trophic Diversity (TDis) generally increases along the gradient from sheltered to oceanic zone, except for the sessile invertebrates carnivores (IS) for which the values decrease. For instance omnivore individuals (OM) present a consistently increasing TDis values when going from Sheltered zone to Ocean. Piscivores (FC), Planktivores (PK) and Herbivores (H) present a similar pattern with higher TDis in Lagoon zone. Mobile invertebrates consumers (IM), oppositely, present lower TDis value in Lagoon zone. Low values correspond to a community that presents a lower variability in the various isotopic signatures (and by extension the trophic behavior) in species of this assemblage. This !"# ! RESULTS III ! means that a small TDiv corresponds to communities that occupy a narrow isotopic niche, or present a small diversity between species or both. - The Trophic Eveness (TEven) increases across the environmental Sheltered-toOcean gradient (except for the low value observed for Planktivores in Ocean zone). The similarity between values increases as the oceanic influence increases. For instance, all values for all diets cover the range from 0.1 to 0.9 in Sheltered zone while around 0.7 to 0.9 near to the Ocean. - Finally, the Trophic Uniqueness (TUni) gives a representation of the « trophic singularity » of each species in a given functional group and zone. A similar pattern is observed for Piscivores (FC), Herbivores (H) and Planktivores (PK) with TUni being higher close to the ocean (Lagoon and Ocean zones). The three other diets (IM, IS and OM) display more stable values across zones (except for Omnivores that show a slightly increasing value when getting closer to the Ocean, but more slowly than H, PK and FC). IS and OM present quite high values (around 0.6 to 0.7) whereas IM present the lowest values of all (around 0.2 to 0.4). #$%&'(" )"*" +,'$,-$./" .0" 1,2" -3(" 4'.53$6" 7$63/(88" 147$629" 1:2" -3(" 4'.53$6" ;$<('%(/6(" 14;$<29" 162" -3(" 4'.53$6";$85('8$./"14;$829"1=2"-3("4'.53$6"><(/(88"14><(/2",/="1(2"-3("4'.53$6"?/$@&(/(88"14><(/2" $/=$6(8"0.'"(,63"=$(-",66.'=$/%"-."-3("A./("$/"B&'&'.,C"43("A./(8",'(",::'(<$,-(=",8"0.DD.E8"*"F"*" F3(D-('(=9"FG"*"F.&-3HI(8-9"J"*"J,%../9"G"*"G6(,/C! ! !!" Discussion & Conclusion ! DISCUSSION & CONCLUSION IV ! ! SPECIALIZATION IN REEF FISH COMMUNITIES Effect of individual size on isotopic signatures Results relative to the effect of size on !13C and !15N values showed that in more than half of the studied species a significant relationship existed between size and the value of at least one of the two isotope ratios. This ontogenic pattern has been reported for many fish species (Galván et al. 2010) and could be explained by several factors. First, the modification of isotopic signatures could be due to variations in individual metabolism, for example from juvenile to sexually mature stage when gametogenesis requires more energy investment (Kolasinski et al. 2009). Maturation could also enable a better assimilation of the ingested items and thus modify the isotopic fractionation rate (Sweetings et al, 2007 (a) and (b)). The isotopic signature can be modified by a change in foraging location (e.g. shift of habitat between juveniles and adults) if the isotopic signature of the targeted source is different from one foraging location to another (Gladfelter & Johnson 1983, Brischoux et al. 2011). Finally, there may be a diet shift between early and late life stages (Cocheret de la Morinière et al. 2003, Graham et al. 2007, Frédérich et al. 2010). This diet shift could either avoid intraspecific competition between individuals of different size, or be only due to the fact that larger individuals possess a better capture ability or could ingest prey from higher trophic levels (Coates 1980, Luczkovich et al. 1995, Cocheret de la Morinière et al. 2003, Graham et al. 2007). In our study, diet shift is the most probable explanation for species which amplitude between the smallest and the largest individuals is important, based upon a mean fractionation coefficient of 3,4‰ for !15N (Post et al. 2002). A diet shift may have important consequences on trophic niche width as depending upon the proportion of juveniles and adults in a population, the global isotopic niche width could change (size-linked intra-population variability). Moreover, in the literature the slope between size and !15N is positive whilst for !13C there is usually no relationship (Galván et al. 2010 ; Greenwood et al. 2010). Cases when the isotopic value of !15N decreases with size are not frequent in the literature. Our study is by far the most comprehensive, covering almost 150 species, whereas so far the largest study available looked at around ten species (Cocheret de la Morinière, 2003; Pasquaud et al. 2008), and a we recorded a negative slope between size and !15N in more than ten cases, three of which with a sufficient amplitude to induce a gradual change in trophic level (Acanthurus nigrofuscus, Naso lituratus and Naso unicornis).These three herbivores contradict suggestions of Cocheret de la Morinière et al. (2003) who found no significant relationship between size and isotopic signatures for herbivores. Indeed for a given species, the comparison of our results with the literature (e.g. Carassou et al. 2008, Frédérich et al. 2010, Greenwood et al. 2010) does not systematically lead to similar conclusions. Similarly our results show some variability across regions. This suggests that some environmental, biotic or abiotic constrains may influence the relationship of isotopic ratios with fish size (Reid et al. 2013, Nagelkerken et al. 2008). By contrast, when scaling up to the next degree of organization, i.e. functional groups, patterns are stronger as the relationship between size and isotopic ratios is conserved between New-Caledonia and Mururoa despite major environmental differences. This shows a convergence with the biodiversity patterns mentioned in the Review of this report : functional groups show a stronger stability at large scale than species. However, within a given functional group, the variability of the responses of the species is important. This suggests that fish are very versatile / adaptable : beyond their biological traits (and by extension their “evolutionary background”) there is a potential of adaptability that enables them to fit in various habitat constrains, otherwise a conserved pattern across regions for each species would have been the rule. ! !"# Specialist to generalist gradient : is a typology at hand ? It was noticed that the rank of species along a generalist-to-specialist gradient remains stable within a functional group across regions, the only exception being mobile invertebrates consumers species. Despite a number limitations due to sampling and the absence of confirmation from stomach content analyses and despite the difficulty to check for all the underlying hypotheses (i.e fractionation rates by species, exact signatures of sources, etc) the present results still constitute a strong argument to support the use of stable isotopes variability as a proxy of trophic niche width. Up to now, specialization in coral reef fish has been analyzed in a narrow range of species which are phylogenetically close, and the specialization was often considered in its spatial aspect. Such a trophic specialist-to-generalist gradient had never been analyzed, to our knowledge, at the scale of the present study which shows that there is on one hand some constancy in the classification of species along this specialist-generalist gradient but that the trophic niche width is highly variable across regions, locations and even within species as discussed in the following paragraphs. Effect of the biology of species on isotopic niche width All the tested biological or behavioral traits have a significant effect on the mean isotopic niche width of a species, implying that the degree of specialization is at least in part linked to the evolutive traits of species (See Futuyma & Moreno, 1988 for a Review and Litsios et al. 2012 for an example in the Pomacentridae family). In both New-Caledonia and the atoll of Mururoa the most influential traits were diet and mean species size. This is relevant insofar as the diet of species does not lead to the same opportunities in terms of prey diversity. For instance, strict coralivore fish often present feeding preferences or space partitioning in order to avoid habitat interspecific competition or aggressiveness (BouchonNavarro et al. 1985 ; Cox, 1994 ; Zekeria et al. 2002). Coral feeders have a narrower niche width than piscivore or herbivore species, which are usually far more mobile and able to feed on a wider range of prey (e.g for herbivores Roughgarden 1974). This is actually observed in the atoll of Mururoa where sessile invertebrate feeders (most of which are coralivore) present the narrowest isotopic niche width. In the same way, omnivore fish are expected to present the widest isotopic niche, which was however not the case in Mururoa in contrast to NewCaledonia where this expectation was confirmed. In Mururoa the largest isotopic niche were presented by the two herbivore diets (strict herbivore HM and mixed herbivore-detritivore HD) followed by omnivore species. The simplest explanation is that omnivore species still present the most eclectic diet, but in the atoll of Mururoa, herbivore species present a high degree of inter-individual variability that results in a wide individual specialization (Green & Bellwood, 2009 ; Araujo et al. 2011 ; Elmqvist et al. 2013), and thus a larger global specific niche than omnivores. The absence of such patterns in New-Caledonia could be linked to statistical power, as isotopic niche width could be estimated for only two herbivore and two omnivore species in this region, whereas six to sixteen species of each diet were available in Mururoa. Behavioral traits like schooling or diurnal activity, for instance, are often linked to trophic requirements, and therefore interact with diet. For example herbivore reef fish are always diurnal species (the reason is uncertain but could be linked to either predation avoidance or nutrient content), many carnivore species forage mostly at night because invertebrates come out of their shelter at that time (Nagelkerken et al. 2000). In the same way, large schools are more likely to shelter species of low trophic level (Foster, 1985 ; Welsh & Bellwood, 2012) as they protect individuals from predators or enable them to reach their resource more easily in case of direct competition. Oppositely predators are more often !"# ! DISCUSSION & CONCLUSION IV ! solitary species. Furthermore, behavioral traits are expected to have an influence on the trophic niche width within a given functional group. For instance Lawson et al. (1999) showed that an herbivore species forming large schools had a broader foraging influence than another herbivore which does not aggregate. This very likely results into a broader diet range. However, within our study the effect of a biological or behavioral trait could change across regions (New-Caledonia vs the atoll of Mururoa). For example, when considering the interaction between diurnal activity and piscivore species (FC), a larger isotopic niche was observed for nocturnal species in Mururoa while the exact opposite was observed in NewCaledonia. Such interactions were frequent (observed for almost all biological traits across regions) and show the versatility or adaptability of fish to food availability. However, there was no predictable pattern, with the exception of the Level factor : the isotopic niche width being almost always wider for species that live on the bottom than for species living higher in the water column. This may be explained by either a higher prey diversity and availability, or by a more generalist trophic behavior of these species in order to avoid competition over ressource. This absence of predictability in the effect of biological traits on trophic niche width has to be associated with similar results found for : (1) individual size ; (2) within a species across locations and; (3) the species effect within functional groups (the two latter points are discussed further). All these results converge towards a high versatility or adaptability of these reef fish to food resources. Very little literature is presently available on this topic. Futuyma & Moreno (1988) suggested that the trophic niche width and specialization were the result of a highly complex interaction between inherent (biological traits) and local constrains that balance each other, and consequently it is very difficult to disentangle the respective effect of each force. Our results strongly suggest that there is an interaction between food availability and trophic niche width. Before investigating the consequences of this adaptability/versatility of reef fish diet, we wish to examine two other species features: endemism and functional distinctiveness. Effect of endemism and functional distinctiveness on isotopic niche width Narrow geographical ranges (endemism) and functional singularity may confer more vulnerability to species. Endemism is expected to be linked to a narrow potential of habitat adaptability and thus a relatively high degree of trophic specialization (Munday, 2004 ; Slatyer et al. 2013). The combination of a narrow geographical range and low abundance implies a double jeopardy (Hobbs el al. 2010). Double jeopardy combined with specialization would be even more critical to species vulnerability. However we found no significant effect of geographical range upon mean isotopic niche width of species, and neither did Hobbs et al. (2010) with ecological specialization of angelfishes in the Indian Ocean. Endemism may therefore not be linked to a better utilization of trophic resources or to the presence of specific trophic resources. However the level of endemism was low in our data set. Data from Marquesas Islands, which are characterized by a high level of endemism, are currently being analyzed and initial results suggest also that trophic niche width is not related to endemism. In the same way, there was no significant relationship between singularity of a species and the mean value of its isotopic niche width, even inside a functional group, opposite to expectations. Indeed the rarest species have been shown to often fill key roles in reef ecosystems by supporting rare functional features ; and rare species often have a narrow ecological niche (Mouillot et al. 2013). In the present study niche width could not be associated to species singularity neither in New Caledonia nor Mururoa or the Gambier archipelago. Thus these “singular” species are neither linked to a very high generalist behavior nor oppositely to a very high specialization that could enable their persistence. As it was hypothesized for degree of endemism, maybe singular species have developed some ! !"# evolutionary advantages in terms of spatial competition, or a better resistance towards predation for instance. Hobbs et al. (2010) did not find any relationship between abundance and ecological niche width while they hypothesized that the least abundant species would be the most singular. Endemism and functional singularity yield therefore a similar signal than the other analyses (size, location, functional groups), the trophic niche width of reef fishes is highly variable and not linked to specific biological features. This merges with the considerations of Pavoine et al. (2005) on the lack of predictability of species singularity. Effect of frequency, biomass and density on isotopic niche width Data from the atoll of Mururoa enabled to analyze the effects of species frequency, biomass and density on trophic niche width. A significantly positive relationship with the mean isotopic niche width was found. This suggests that the most common species, in terms of occurrence or abundance, have a more eclectic feeding behavior than the species that are less common. A high frequency could be the cause or consequence of a high degree of opportunist feeding behavior, that enable the species to survive in a wide variety of locations despite of varying prey availability or inter- (but more likely intraspecific) competition (Semmens et al. 2009). Many studies have reported a positive relationship between abundance and individual specialization that could lead to a larger global isotopic niche (Bean et al. 2002 ; Araujo et al. 2011). A high biomass may originate either from a high density, or from a low density but large individuals. Large individuals may have access to wider prey spectra and therefore display larger mean values of isotopic niche width. When looking at the effect of the species size class (Table 5), however, the largest species do not actually have necessarily a larger niche width. So the relationship between biomass and mean isotopic niche width is probably linked to the density effect. When looking at the effect of the diet (i.e does the slope of the regression change according to the functional group) it appears that the global pattern is mainly linked to herbivores for all three measures frequency, biomass or density. This is in opposition to Hobbs et al. (2010) who found no such relationship for Centropyge angelfishes which are herbivore. However they took into account the habitat niche also which could explain this dilution. The positive regression between abundance and isotopic niche width leads to quite important considerations in terms of resilience, less abundant species being likely to become more vulnerable to resource depletion. Oppositely, when considering the four corallivorous species Chaetodon ephippium, C. flavirostris, C. lunulatus and C. trifascialis, the less frequent or abundant (in terms of frequency, biomass and density) had the largest mean isotopic niche width. This could be linked to a process of trophic niche-partitioning amongst coralivore species. A frequent or abundant species may get more specialized in order to avoid interspecific competition (Cox, 1994 ; Zekeria et al. 2002 ; Pratchett & Berumen, 2008) or has a better access to its favorite resource (Berumen & Pratchett, 2008). Thus each species focuses on a narrower range of items. However, as two of these four species were only represented by eight individuals instead of the recommended minimum of ten (Jackson et al. 2011 ; Syväranta et al. 2013), this result may not be considered as plainly robust despite the high level of significance. The opposition between herbivore and coralivore species, in terms of the response of specialization to increasing competition may be considered to support our previous results showing the diet versatility of species. Herbivore species may react to intraspecific competition, presenting a higher degree of individual specialization resulting into a wide global isotopic niche which may overlap (or not; Bolnick et al. 2010) with other species ; whereas coralivores are more specialized in a common pool of resources which suggest dietpartitioning amongst other coralivore species. Balance between these two forces leads to different responses as described by Bolnick et al. (2010) and thus to different vulnerability to !"# ! DISCUSSION & CONCLUSION IV ! resource depletion. For the other diets, the absence of any trend between frequency / abundance and the mean isotopic niche width may reflect an adaptability to maintain a similar diet for a wide range of resource conditions. Omnivore species could feed upon a wide range of prey and are less likely to be constrained by competition. Planktivore species may not be strongly selective and feed on the entire range of plankton organisms with low or no intra or interspecific interactions, as the resource availability is high (however, intraspecific diet variability has been shown to increase with density for Dascyllus aruanus, Frédérich et al. 2010). Carnivores and piscivores have usually wide home ranges which allow them to find their favorite prey items, and therefore decrease their need to adapt to local competition pressure. Effect of an oceanic gradient At the species scale As the effect of distance from the ocean on community structure is well documented (e.g. Grimaud & Kulbicki, 1998) we wanted to test if isotopic niche width could explain some of this variation. In particular the effect of wave exposure, turbidity and currents condition resource availability and habitat features thus influencing the functional structure of the associated ecosystems (Grimaud & Kulbicki, 1998 ; Friedlander et al. 2003). For example resource regularity is expected to be higher in sheltered than in exposed zones for herbivores and planktivores which may induce wider isotopic niches. In Mururoa the mean isotopic niche width of a given species varied across zones while the isotopic signatures of !13C and !15N generally remained stable. This suggests a variation in prey selection or availability but without a shift of sources. These variations could not be linked to any consistent pattern within a functional or taxonomic group, in opposition to the patterns observed within functional groups at the regional scale. This suggests that factors driving the isotopic niche width at a local scale cannot be predicted from biological or ecological traits alone. In the absence of local data on environmental variables such as nutrient levels, chlorophyll a, topography of the substratum, and so on, it is not possible to know the influence of the local environment on trophic niche width for a given species. However, shifting from the species level to the entire available species pool within a functional group may bring stability, as observed for the effect of individual size. At the community scale : support from Layman’s modified metrics In Mururoa, the environmental gradient from sheltered-to-exposed zones had a more stable effect on the isotopic space at the functional group level. The narrower isotopic space (TRic) in Sheltered and Lagoon zones for almost all diets, as the high mean Trophic Evenness and Trophic Uniqueness values in the Lagoon zone suggest that species may share a similar range of prey, but with a need for partitioning (1) because of low prey availability in the Lagoon zone, the complexity of habitat being low or (2) because of high species richness in the Sheltered zone, leading to competition. An example from planktivores may illustrate our point : - According to UVC data the density of planktivores is higher in the Sheltered and Lagoon zones (Given in Appendix XIV), which could be linked to the proximity to coral communities that provide feeding sources. This converges with literature (Friedlander et al. 2003). - Trophic Richness is weak in the Lagoon and Ocean zones, suggesting that fish probably feed upon a similar range of zooplankton organisms. TRich is intermediate in the Sheltered zone and high in South-West zone suggesting a trophic diversification. - Plankton availability is thus supposedly higher in the Sheltered and South-West zones, maybe because of a higher rate of nutrient input from the atoll ring. ! !"# - Trophic Eveness is maximum in the South-West and Lagoon zones, Trophic Uniqueness is maximum in the Lagoon zone. These results suggest that in the Lagoon zone density is high but prey availability is low, resulting in a high degree of partitioning. In the Ocean zone there is a low density but also a low prey availability, with a balance between these two constraints. Oppositely in the Sheltered zone density is high and so is prey availability with a potential balance also between these two constraints. In the South-West a low density goes with a high prey availability which may explain the high Trophic Evenness and Trophic Richness : each species may feed on its favorite prey as they are diversified with little competition expected to occur. However this remains hypothetic as it is difficult to describe such an ecological pattern for all diets. There are however some converging features, e.g.: • • Trophic Richness is maximum for herbivores (with exception of mobile invertebrates consumers) across all four zones, while Trophic Evenness and Uniqueness remain relatively low, supporting that herbivores show both an intra-specific and interspecific diversity in their diet. This pleads for a globally generalist behavior of herbivore species. Oppositely, sessile invertebrates feeders (most of which are corallivores) present a low Trophic Richness and a high Trophic Uniqueness, suggesting a high behavioral specialization in conjunction with strict resource partitioning The community approach is complementary to the study of the relationship at the species level between abundance and isotopic niche width, as it accounts for species interrelationships with potential hypotheses on competition and resource. However, this approach remains exploratory because (1) the species for which both abundance and isotopic data were available may not represent the typical structure of a community in a given habitat ; (2) UVC data are an instantaneous representation whereas isotopes integrate variations over several months in diet ; (3) this methodological framework still needs to be fully validated. However, at this stage the interest of such an approach is undeniable as the combination of isotopic niche values and abundance data enables a precise representation of the trophic structure and functioning of a community (Cucherousset & Villéger, submitted). Mouillot et al. (2013b) suggested a similar tool to analyze the functional structure of communities while accounting for phenological, morphological or any other measurable trait, susceptible to fill a potential role in resilience of the system as expected for the degree of diet specialization. ! Stability versus Versatility So far we found that the isotopic niche width may be mainly predicted from species diet and abundance. There is also an effect of individual size and within a functional group species tend to keep the same rank. However, in our opinion two major considerations highlight this work : (a) as already often described in the literature, within a functional group the species composition may change across space and time. However, the isotopic niche width or the size-based response of these species may change as well, in other words species may present a high degree of diet versatility which amplitude was unexpected ; in particular it is observed for almost all studied levels. (b) all functional groups do not respond in a similar manner, suggesting different answers to changes in environmental conditions and therefore resilience (Nyström, 2006). For example coralivores may be much more vulnerable and less resilient than herbivores (Wilson et al. 2008). Their susceptibility to habitat loss has been demonstrated many times (Bouchon-Navarro et al. 1985 ; Pratchett et al. 2012). Oppositely herbivore species may be well adapted in preventing shifts to macro-algae dominance in reef ecosystems, as they display a high degree of diet versatility/adaptability both at the individual and the species levels which corresponds with their generalist diet behavior (Roughgarden, !"# ! DISCUSSION & CONCLUSION IV ! 1974) with these individual and species levels acting in a complementary way (Elmqvist et al. 2003). This versatility/adaptability of species may also be essential in the resilience potential of a community. We found out that species could be way more adaptable than previously supposed, as their isotopic niche width depends on both deterministic factors like traits or abundance, but also stochastic factors as environmental constraints and/or prey availability. Moreover, for some species, isotopic signatures showed a significant divergence compared to the mean trophic levels given by Fishbase, which are based upon stomach contents or feeding behavior (See Appendix XV). The trophic versatility of reef fishes is not an unknown pattern, in particular Curtis-Quick et al. (2012) showed very similar processes on reef fish of the Great Barrier in Australia, with at times a complete shift in diet according to environmental features. Similarly Albouy et al. (2011) could not use ecomorphological traits to predict the level of diet-partitioning or diet overlap, and Bellwood et al. (2006) stated that a specialized morphology was not systematically linked to a specialist diet. Yet, the broad range of species and geographical area we used for the current study highlights that this trophic versatility/adaptability occurs at all scales. Besides looking at which guilds are the most resilient one may ask to which degree of perturbation may a guild be resilient ? What are the specific potential of adaptability and to what extent species may resist to factors like habitat and environmental changes or variations in resource availability before they collapse ? The trophic versatility of reef fishes may be considered in the light of the theory of the diversity response, that states that the resilience of a guild does not depend only on its specific redundancy and overall niche adaptability, but also on the diversity of responses within this guild (Nyström et al. 2006 ; Mori et al. 2013). Versatility may act as a security to insure a better resilience of coral reef fish assemblages despite habitat loss and other threats. This may also explain in part the origin of the high biodiversity among coral reef fish : the instability of recruitment (Hixon, 2011) combined with versatility could enable the coexistence of both specialist and generalist species, preventing any competitive exclusion (even though there still are ubiquitous species found over a wide range of conditions). Many other types of niche specialization, generalization and many biological traits, besides trophic niche, could contribute to the versatility or the diversity of responses inside a reef ecosystem (Curtis-Quick et al. 2012). These may interact with the versatility / adaptability patterns we highlighted in this work. However high resilience may not necessarily means a high ability for recovery. These two notions have to be cautiously distinguished as even a pool of generalist and/or adaptable species, like herbivores, may not easily recover after a phase shift (Elmqvist et al. 2003). For instance Bellwood et al. (2006b) showed that after a major collapse, the recovery to the previous coral-dominated system occurred thanks to a single species, thus pointing to the importance of keystone species. Therefore, even though the future of reef fish depends on the health of coral reefs, the removal of species by overharvesting or pollution (Chabanet et al. 2010) for instance may have unpredictable impact on the potential of resilience (Folke et al. 2004 ; Bellwood et al. 2012 for an example with parrotfish) ! COMMENTS & CRITICS OF THE METHODS : WHICH PERSPECTIVES ? The use of stable isotopes for a large scale study is at present uncommon. Most studies that illustrate the use of SIBER ellipses focus on a range of two or three species that are directly related (one being for example a direct competitor of the other - Jackson et al. 2012 ; Guzzo et al. 2013 ; Hayden et al. 2013) in a very restricted geographical area. The use of such methods for large scale considerations has much potential but brings also several important issues. The Review section of this report indicated many technical and methodological constrains which need to be considered when analyzing large scale patterns. ! !"# First, the absence of measures on potential sources could lead to some criticisms as differences in isotope ratios may in part originate from differences in the sources and therefore interfere in the analysis of the rank of species on a specialist-generalist gradient (Bearhop et al. 2004 ; Araujo et al. 2011), particularly when patterns were slightly pronounced. Solutions could be the analysis of a third isotope in order to obtain a tridimentional niche and add an axis to the typology of generalist-to-specialist. Another solution could have been to analyze the isotopic signature of blood, that integrate the diet range at a narrower time scale (Boecklen et al. 2011). This could bring information on temporal variability or inversely stability of diet range. However blood sampling on reef fish, which are often very small would be a major challenge. The lack of sources is also problematic when comparing contrasted zones or regions. A determination of a robust isotopic baseline at each sampled location could enable a better standardization across areas or regions which would allow direct comparisons, something which could not be done in the present work. A precise knowledge about sources would also enable to get a very tight estimation of the niche via SIAR mixing models for example, that enable to estimate the relative proportion of each type of prey in the whole diet range of a species. However this is probably out of reach in reef systems as sources are too numerous and may not be isotopically sufficiently distinct. This could be a perspective for a smaller-scaled study in coral reef fishes though, for instance to deepen some of our results. Another issue with SIBER ellipses analysis is the difficulty to estimate overlap between a large set of species. Overlap can easily be estimated between two species but this quickly becomes very complex when having more than ten species as we did. The Layman’s derived metrics may bring some information related to overlap but will not enable a precise analysis of competition. There are numerous other sources of concern in this large scale analysis. The major ones are ; (1) the difficulty to get samples for the same species over a wide array of environmental conditions which is bound to the very high beta diversity of reef fish assemblages ; (2) the need to sample a range of sizes for most species, as isotope ratios are size dependant ; (3) get sufficient replicates in order to reduce variance ; (4) collect more environmental information in order to analyze the effects of environmental factors such as coral cover, habitat complexity, depth or fishing pressure. However the complexity and diversity of reef fish assemblages and of their habitat makes such concerns very difficult to take into account. The present study, despite its shortcomings, is by far the largest study to date on the isotope ratios of fish and it is precisely this sample size which enabled us to discover how versatile and adaptable reef fish may be regarding their diet. CONCLUSION This study proceeds to decline the use of the isotopic niche width at a large scale and over a large set of species, which to our knowledge has never been done before. However, such a large scale approach is confronted to the difficulty to account for all the very precise considerations that stable isotopes biology usually requires. Despite the shortcomings that were discussed, this work brings confirmations of previous findings as well as some new insights and suggests some new venues that were before scarcely studied and documented. The most interesting result may be the combination of (1) the versatility of species found at nearly all studied levels : the effect of size on the signature of !13C and !15N but also the effect of biological traits/environmental gradient on the mean isotopic niche width inside a functional group, with (2) the stability observed when considering the functional group instead of the species. The latter phenomenon was known when considering reef fish species distribution across biogeographical regions in the world, !"# ! DISCUSSION & CONCLUSION IV ! but finding the same pattern with regards of trophic specialization leads to several interesting conclusions in term of resilience and functional groups. • • • • • A high trophic versatility/adaptability may favor the resistance and stability of reef fish assemblages to perturbations. This confirms that assemblages with a higher diversity will probably be more resilient The opposite strategies of coralivores and herbivores in the relationship between abundance and trophic niche width suggest that the effects of these fish on coral-algae phase shifts may be far more complex than presently thought. The lack of relationship between trophic niche width with endemism and singularity suggests that the ecological role, in particular resilience or resistance, of such species is probably not related to their trophic characteristics. The lack of relationships between trophic niche width and most life-history traits suggest a higher flexibility in the trophic relationships amongst functional groups, which may further favor resilience and stability. This work also highlights several perspectives to explore : Investigate the effects of environmental factor on the isotopic niche width by focusing on subsets of the databases. Complete the data on sources in order to improve our understanding of how the generelist-to-specialist ranking may change across locations and regions. If the ranking of species across this gradient, remains stable this would constitute a good index to make comparative analyses of ecological processes across locations or regions. Add information from other elements such as H, S, or P to distinguish different types of specialization. trophic overlap amongst species, amongst functional groups or even within species across locations could be very informative on competition / exclusion processes. This overlap can be easily be assessed using the SIBER ellipses, but in highly complex assemblages the number of combinations rapidly becomes very important. • • • • This study has shown both the potential of very large scale C and N isotope ratios studies and their limitations. It is likely that technological progress may be decisive in this field with smaller and smaller samples needed and more and more elements being available for analysis. Ecology experts should not be worried, the journey is just beginning ! ! ! !"# Litterature Cited Albouy C., Guilhaumon F., Villéger S., Mouchet M., Mercier L., Culioli J.M., Tomasini J.A., Le Loc’h F., Mouillot D. (2011). Predicting trophic guild and diet overlap from functional traits: statistics, opportunities and limitations for marine ecology. Marine Ecology Progress Series, 436, 17-28. Araujo M.S., Bolnick D.I., Layman C.A. (2011). The ecological causes of individual specialisation. Ecology Letters, 14, 948-958. Bean K., Jones G.P., Caley M.J.(2002). Relationships among distribution, abundance and microhabitat specialisation in a guild of coral reef triggerfish (family Balistidae). Marine Ecology Progress Series, 233, 263-272. Bearhop S., Adams C.E., Waldron S., Fuller R.A., MacLeod H. (2004). Determining Trophic Niche Width : A Novel Approach Using Stable Isotope Analysis. Journal of Animal Ecology, 73, 1007-1012. Bellwood D.R. & Hughes T.P. (2001). Regionalassembly rules and biodiversity of coral reefs. Science, 292, 1532–1535. Bellwood D.R., Wainwright P.C., Fulton C.J., Hoey A. (2002). Assembly rules and functional groups at global biogeographical scales. Functional Ecology, 16, 557-562. Bellwood D.R., Hughes T.P., Folke C., Myström M. (2004). Confronting the coral reef crisis. Nature, 429, 827-833. Bellwood D.R., Wainwright P.C., Fulton C.J., Hoey A.S. (2006). Functional versatility supports coral reef biodiversity. Proceedings of the Royal Society, 273, 101-107. Bellwood D.R., Hughes T.P., Hoey A.S. (2006b). Sleeping functional group drives coral-reef recovery. Current Biology, 16, 2434-2439. Bellwood, D.R., Hoey A.S., Hughes T.P. (2012). Human activity selectively impacts the ecosystem roles of parrotfishes on coral reefs. Proceedings of the Royal Society B, 279, 1621-1629. Berumen M.L. & Pratchett M.S. (2008). Trade-offs associated with dietary specialization in corallivorous butterflyfishes (Chaetodontidae: Chaetodon). Behaviour, Ecology and Sociobiology, 62, 989-994. interspecific competition leads to decoupled changes in population and individual niche width. Proceedings of the Royal Society B, 277, 1789-1797. Bouchon-Navarro Y., Bouchon C., Harmelin-Vivien M.L. (1985). Impact of coral degradation on a chaetodontid fish assemblage (Moorea, French Polynesia). Proceedings of 5th International Coral Reef Symposium, 5, 427-432. (résilience chez les corallivores). Brischoux F., Bonnet X., Cherel Y., Shine R. (2011). Isotopic signatures, foraging habitats and trophic relationships between fish and seasnakes on the coral reefs of New Caledonia. Coral Reefs, 30, 155-165. Burkepile D.E. & Hay M.E. (2008). Herbivore species richness and feeding complementarity affect community structure and function on a coral reef. PNAS, 105 (42), 16201-16206. Carassou L., Kulbicki M., Nicola T.J.R., Polunin N.V.C. (2008). Assessment of fish trophic status and relationships by stable isotope data in the coral reef lagoon of New Caledonia, southwest Pacific. Aquatic Living Ressources, 21, 1-12. Chabanet P., Guillemot N., Kulbicki M., Vigliola L., Sarramegna S. (2010). Baseline study of the spatiotemporal patterns of reef fish assemblages prior to a major mining project in New Caledonia (South Pacific). Marine Pollution Bulletin, 61, 598-611. Chase J.M. & Leibold M.A. (2003). Ecological niches : linking classical and contemporary approaches. University of Chicago Press. 221 pp. Coates D. (1980). Prey-size intake in humbug damselfish,Dascyllus aruanus (Pisces, Pomacentridae) living within social groups. J Anim Ecol, 49, 335-340. Cocheret de la Morinière E., Pollux B.J.A., Nagelkerken I., Hemminga M.A., Huiskes A.H.L., Van der Velde G. (2003). Ontogenetic dietary changes of coral reef fishes in the mangrove-seagrass-reef continuum: stable isotopes and gut-content analysis. Marine Ecology Progress Series, 246, 279-289. Cornell, H. V. & Karlson, R. H. (2000). Coral species richness : Ecological vs. biogeographical influences. Coral Reefs, 19, 37-49. Cox E.F. (1994). Resource use by corallivorous butterfly fishes (family Chaetodontidae) in hawaii. Bulletin of Marine Science, 54 (2), 535-545. Boecklen W.J., Yarnes C.T., Cook B.A., James A.C. (2011). On the Use of stable isotopes in trophic ecology. Annu. Rev. Ecol. Evol. Syst. 42, 411–440. Cucherousset J. & Villéger S. (Submitted). Quantifying the multiple facets of trophic diversity: new integrative metrics for stable isotope ecology. Methods in Ecology and Evolution, x, xxx-xxx. Bolnick D.I., Ingram T., Stutz W.E., Snowberg L.K., Lau Lee O., Paull J.S. (2010). Ecological release from Curtis-Quick J.A., Ahmadia G.N., Smith D.J. (2012). Feeding plasticity of reef fish. Proceedings of the 12th ! !"# International Coral Reef Symposium, Cairns, Australia, 9-13 July 2012. (Pour les effets de la compétition) DeNiro M.J. & Epstein S. (1978). Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta, 42, 495-506. DeNiro M.J. & Epstein S. (1981). Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta, 45, 341-351. Devictor V., Julliard R., Jiguet F. (2008). Distribution of specialist and generalist species along spatial gradients of habitat disturbance and fragmentation. Oikos, 117, 507-514. Devictor V., Clavel J., Julliard R., Lavergne S., Mouillot D., Thuiller W., Venail P., Villéger S., Mouquet N. (2010). Defining and measuring ecological specialization. Journal of Applied Ecology, 47, 15-25. Dromard C.R., Bouchon-Navaro Y., Cordonnier S., Fontaine M-F., Verlaque M., Harmelin-Vivien M., Bouchon C. (2013). Resource use of two damselfishes, Stegastes planifrons and Stegastes adustus, on Guadeloupean reefs (Lesser Antilles): Inference from stomach content and stable isotope analysis. Journal of Experimental Marine Biology and Ecology, 440, 116–125. Elmqvist T., Folke C., Nystrom M., Peterson G., Bengtsson J. (2003). Response Diversity, Ecosystem Change, and Resilience. Frontiers in Ecology and the Environment, 1 (9), 488-494. Elton C. (1927). Animal Ecology. London : Sidgwick and Jackson. Fulton C.J., Bellwood D.R., Wainwright P.C. (2005). Wave energy and swimming performance shape coral reef fish assemblages. Proceedings of the Royal Society B, 272, 827-832. Futuyma D.J. & Moreno G. (1988). The Evolution of Ecological Specialization. Annual Reviews of Ecological Systems, 19, 207-233. Galván D.E., Sweeting C.J., Reid W.D.K. (2010). Power of stable isotope techniques to detect size-based feeding in marine fishes. Marine Ecology Progress Series, 407, 271-278. Galzin R., Planes S., Dufour V., Salvat B. (1994). Variation in diversity of coral reef fish between French Polynesian atolls. Coral Reefs, 13, 175-180. Gause G.F. (1936). The struggle for existence. Baltimore : Williams and Wilkins. Gingerich W.H. (1986). Tissue distribution and elimination of rotenone in rainbow trout. Aquat. Toxicol. 8, 27-40. Gladfelter W.B. & Johnson W.S. (1983). Feeding niche separation in a guild of tropical reef fishes (Holocentridae). Ecology, 64 (3), 552-563. Graham B.S., Grubbs D., Holland K., Popp B.N. (2007). A rapid ontogenetic shift in the diet of juvenile yellowfin tuna from Hawaii. Marine Biology, 150, 647-658. Green A.L. & Bellwood D.R. (2009). Monitoring functional groups of herbivorous reef fishes as indicators of coral reef resilience – A practical guide for coral reef managers in the Asia Pacific region. IUCN working group on Climate Change and Coral Reefs. IUCN, Gland, Switzerland. 70 pages. Folke C., Carpenter S., Walker B., Scheffer M., Elmqvist T., Gunderson L., Holling C.S. (2004). Regime shifts, resilience, and biodiversity in ecosystem management. Annual Review in Ecology, Evolution and Systematics, 35, 557-581. Greenwood N.D.W., Sweeting C.J., Polunin N.V.C. (2010). Elucidating the trophodynamics of four coral reef fishes of the Solomon Islands using !15N and !13C. Coral Reefs 29, 785-792. Foster S.A. (1985). Group foraging by a coral reef fish: a mechanism for gaining access to defended resources. Animal behaviour, 33 (3), 782-792. Grimaud J. & Kulbicki M. (1998). Influence de la distance à l’océan sur les peuplements ichtyologiques des réifs frangeants de Nouvelle-Calédonie. Life Sciences 321, 923-931. Frédérich B., Lehanse O., Vandewalle P., Lepoint G. (2010). Trophic Niche Width, Shift, and Specialization of Dascyllus aruanus in Toliara Lagoon, Madagascar. Copeia, 2, 218-226. Grinnell J. (1917). The niche-relationship of the California thrasher. Auk. 34, 427-433. Friedlander A.M., Brown E.K., Jokiel P.L., Smith W.R., Rodgers K.S. (2003). Effects of habitat, wave exposure, and marine protected area status on coral reef fish assemblages in the Hawaiian archipelago. Coral Reefs, 22, 291-305. Fry B. (2006) Stable Isotope Ecology. Springer. 308 pp. !!" ! Guillemot N., Kulbicki M., Chabanet P., Vigliola L. (2011). Functional Redundancy Patterns Reveal NonRandom Assembly Rules in a Species-Rich Marine Assemblage. Plos One 6 (10), e26735. Guzzo M.M., Haffner G.D., Legler N.D., Rush S.A., Fisk A.T. (2013). Fifty years later: trophic ecology and niche overlap of a native and non-indigenous fish species in the western basin of Lake Erie. Biological Invasions. DOI 10.1007/s10530-012-0401-z. Halpern B.S. & Floeter S.R. (2008). Functional diversity responses to changing species richness in reef fish communities. Marine Ecology Progress Series, 364, 147-156. Hanski, I. (1993). Three explanations of the positive relationship between distribution and abundance of species. In : Historical and Geographical Determinants of Community Diversity (eds Ricklefs, R. & Schluter, D.). University of Chicago Press, Chicago, pp. 108– 116. Hayden B., Holopainen T., Amundsen P-A., Eloranta A.P., Knudsen R., Praebel K., Kahilainen K.K. (2013). Interactions between invading benthivorous fish and native whitefish in subarctic lakes. Freshwater Biology, 58, 1234-1250. Hixon M.A. & Beets J.P. (1993). Predation, Prey Refuges, and the Structure of Coral-Reef Fish Assemblages. Ecological Monographs, 63 (1), 77-101. Isaac N.J.B., Turvey S.T., Collen B., Waterman C., Baillie J.E.M. (2007). Mammals on the EDGE : Conservation Priorities Based on Threat and Phylogeny. Plos One (3), e296. Jackson A.L., Inger R., Parnell A.C., Bearhop S. (2011). Comparing isotopic niche widths among and within communities : SIBER – Stable Isotope Bayesian Ellipses in R. Journal of Animal ecology, 80, 595-602. Jackson M.C., Donohue I., Jackson A.L., Britton J.R., Harper D.M., Grey J. (2012). Population-level metrics of trophic structure based on stable isotopes and their application to invasion ecology. Plos One, 7 (2), e31757. Kassen R. (2002). The experimental evolution of specialists, generalists, and the maintenance of diversity. Journal of Evolutionary Biology, 15, 173190. Hixon M.A. (2011). 60 years of coral reef fish ecology: past, present, future. Bulletin of Marine Science, 87 (4), 727-765. Kolasinski J., Frouin P., Sallon A., Rogers K., Bruggemann H.J., Potier M. (2009). Feeding ecology and ontogenetic dietary shift of yellowstripe goatfish Mulloidichthys flavolineatus (Mullidae) at Reunion Island, SW Indian Ocean. Marine Ecology Progress Series, 386, 181-185. Hobbs J-P.A., Jones G.P., Munday P.L. (2010). Rarity and extinction risk in coral reef angelfishes on isolated islands : interrelationships among abundance, geographic range size and specialisation. Coral Reefs 29, 1-11. Kulbicki M. (1990). Comparisons between rotenone poisoning and visual counts for density and biomass estimates of coral reef fish populations. Proc. Cong. 1990 Int. Soc. Reef Studies, Noumea 14-18 Nov. 105112. Hoeinghaus D.J. & Zeug S.C. (2008). Can stable isotope ratios provide for community-wide measures of trophic structure ? Comment. Ecology, 89 (8), 2353-2357. KULBICKI Michel, MOU-THAM Gérard, VIGLIOLA Laurent, WANTIEZ Laurent, MANALDO Esther, LABROSSE Pierre, LETOURNEUR Yves. Major coral reef fish species of the South Pacific with basic information on their biology and ecology. CRISP-IRD report. Noumea SPC, 2011. 107 pp. Hoey A.S. & Bellwood D.R. (2009). Limited Functional Redundancy in a High Diversity System : Single Species Dominates Key Ecological Process on Coral Reefs. Ecosystems, 12, 1316-1328. Holling C.S. (1973). Resilience and stability of ecological systems. Annual Review of Ecology and Systematics, 4, 1-23. Huntington B.E. & Lirman D. (2012). Species-area relationships in coral communities: evaluating mechanisms for a commonly observed pattern. Coral Reefs, 31, 929-938. Hutchinson G.E. (1957). Concluding remarks : Cold Spring Harbour symposium. Quant. Biol., 22, 415427. Hutchinson G.E. (1978). An Introduction to Population Biology. New Haven, CT : Yale University Press. ! Labrosse P., Kulbicki M., Ferraris J. (2002). Underwater visual fish census surveys. Proper use and implementation. Reef resource assessment tools. Noumea, New Caledonia. Secretariat of the Pacific Community. 54 pp. Lawson G.L., Kramer D.L., Hunte W. (1999). Sizerelated habitat use and schooling behavior in two species of surgeonfish (Acanthurus bahianus and A. coeruleus) on a fringing reef in Barbados, West Indies. Environmental Biology of fish, 54 (1), 19-33. Layman C.A., Arrington D.A., Montana C.G., Post D.M. (2007) Can stable isotope ratios provide for community-wide measures of trophic structure ? Ecology, 88 (1), 42-48. Layman C.A. & Post D.M. (2008). Can stable isotope ratios provide for community-wide measures of trophic structure ? Reply. Ecology, 89 (8), 2358-2359. !"# Litsios G., Pellissier L., Forest F., Lexer C., Pearman P.B., Zimmermann N.E., Salamin N. (2012). Trophic specialization influences the rate of environmental niche evolution in damselfishes (Pomacentridae). Proceedings of the Royal Society, 279, 3662-3669. Luczkovich J. J., Norton S.F. and Gilmore, R.G. Jr. (1995). The influence of oral anatomy on prey selection during the ontogeny of two percoid fishes, Lagodon rhomboides and Centropomus undecimalis. Environmental Biology of Fishes, 44, 79–95. MacNeil M.A., Graham N.A., Polunin N.V.C., Kulbicki M., Galzin R., Harmelin-Vivien M., Rushton S.P. (2009). Hierarchical drivers of reef-fish metacommunity structure. Ecology, 90 (1), 252-264. Matthews B. & Mazumder A. (2004). A critical evaluation of intrapopulation variation of !13C and isotopic evidence of individual specialization. Oecologia 140, 361-371. McCann K.S. (2000). The diversity-stability debate. Nature, 405, 228-233. Mellin C., Huchery C., Caley J., Meekan M.G., Bradshaw C.J.A. (2010). Reef size and isolation determine the temporal stability of coral reef fish populations. Ecology, 91 (11), 3138-3145. Mellin C., Bradshaw C.J.A., Meekan M.G., Caley M.J. (2010b). Environmental and spatial predictors of species richness and abundance in coral reef fishes. Global Ecology and Biogeography, 19, 212-222. Mori A.S., Furukawa T., Sasaki T. (2013). Response diversity determines the resilience of ecosystems to environmental change. Biological Reviews of the Cambridge Philosophical Society, 88 (2), 349-364. Mouillot D., Bellwood D.R., Baraloto C., Chave J., Galzin R., Harmelin-Vivien M., Kulbicki M., Lavergne S., Lavorel S., Mouquet N., Paine C.E.T., Renaud J., Thuiller W. (2013). Rare Species Support Vulnerable Functions in High-Diversity Ecosystems. Plos Biology, 11 (5), e1001569. Mouillot D., Graham N.A.J., Villéger S., Mason N.W.H., Bellwood D.R. (2013b). A functional approach reveals community responses to disturbances. Trends in Ecology and Evolution, 28 (3), 167-177. Munday P.L. (2004). Habitat loss, resource specialisation, and extinction on coral reefs. Global Change Biology, 10, 1642–1647. Nagelkerken I., Dorenbosch M., Verberk W.C.E.P., Cocheret de la Morinière E., Van der Velde G. (2000). Day-night shifts of fishes between shallow-water biotopes of a Caribbean bay, with emphasis on the !"# ! nocturnal feeding of Haemulidae and Lutjanidae. Marine Ecology Progress Series, 194, 55-64. Nagelkerken I, Bothwell J, Nemeth RS, Pitt JM, van der Velde G (2008) Interlinkage between Caribbean coral reefs and seagrass beds through feeding migrations by grunts (Haemulidae) depends on habitat accessibility. Marine Ecology Progress Series 368. Newsome S.D., Martinez del Rio C., Bearhop S., Phillips D.L. (2007). A niche for isotopic ecology. Frontiers in Ecology and the Environment, 5, 429436. Nyström M. (2006). Redundancy and Response Diversity of Functional Groups: Implications for the Resilience of Coral Reefs. A Journal of the Human Environment, 35 (1), 30-35. Pandolfi J.M. (2002). Coral community dynamics at multiple scales. Coral Reefs, 21, 13-23. Parravicini V., Kulbicki M., Bellwood D.R., Friedlander A.M., Arias-Gonzales E., Chabanet P., Floeter S., Vigliola L., D'Agata S., Myers R., Mouillot D. (2013). Global patterns and predictors of tropical reef fish species richness. Ecography, 36, 001–009. Pasquaud S., Elie P., Jeantet C., Billy I., Martinez P., Girardin M. (2008). A preliminary investigation of the fish food web in the Gironde estuary, France, using dietary and stable isotope analyses. Estuar Coast Shelf Sci, 78, 267–279. Pavoine S., Ollier S., Dufour A.B. (2005). Is the originality of a species measurable ? Ecology Letters, 8, 579-586. Planes S., Galzin R., Bablet J-P., Sale P.F. (2005). Stability of coral reef fish assemblages impacted by nuclear tests. Ecology, 86 (10), 2578-2585. Post D.M. (2002). Using Stable Isotopes to Estimate Trophic Position : Models, Methods, and Assumptions. Ecology, 83 (3), 703-718. Pratchett M.S. & Berumen M.L. (2008). Interspecific variation in distributions and diets of coral reef butterflyfishes (Teleostei: Chaetodontidae). Journal of Fish Biology, 73, 1730-1747. Pratchett M.S., Coker D.J., Jones G., Munday (2012). Specialization in habitat use by coral damselfishes and their susceptibility to habitat Ecology and Evolution - Open Access, 10.1002/ece3.321, 2168-2180. P.L. reef loss. doi: Reid W.D.K., Sweeting C.J., Wigham B.D., McGill R.A.R. (In Press). High variability in spatial and temporal size-based trophodynamics of deep-sea fishes from the Mid-Atlantic Ridge elucidated by stable isotopes. Deep-Sea Research II. Roughgarden J. (1974). Species packing and the competition function with illustrations from coral reef fish. Theoretical Population Biology, 5 (2), 163-186. Sale P.F. & Guy J.A. (1992). Persistence of community structure: what happens when you change taxonomic scale ? Coral Reefs, 11, 147-154. Semmens B.X., Ward E.J., Moore J.W., Darimont C.T. (2009). Quantifying inter- and intra-population niche variability using hierarchical bayesian stable isotope mixing models. Plos One, 4 (7), e6187. Slatyer R.A., Hirst M., Sexton J.P. (2013). Niche breadth predicts geographical range size : a general ecological pattern. Ecology Letters, doi: 10.1111/ele.12140. Sweeting C.J., Barry J., Barnes C., Polunin N.V.C., Jennings S. (2007). Effects of body size and environment on diet-tissue !15N fractionation in fishes. Journal of Experimental Marine Biology and Ecology, 340, 1-10. Sweeting C.J., Barry J., Barnes C., Polunin N.V.C., Jennings S. (2007b). Effects of body size and environment on diet-tissue !13C fractionation in fishes. Journal of Experimental Marine Biology and Ecology, 352, 165-176. Syväranta J., Lensu A., Marjomäki T.J., Oksanen S., Jones R.I. (2013). An empirical evaluation of the utility of Convex Hull and Standard Ellipse Areas for assessing population niche widths from stable isotope data. Plos One, 8 (2), e56094. ! Tuya F., Wernberg T., Thomsen M.S. (2011). The relative influence of local to regional drivers of variation in reef fishes. Journal of Fish Biology, 79, 217-234. Welsh J.Q. & Bellwood D.R. (2012). How far do schools of roving herbivores rove? A case study using Scarus rivulatus. Coral Reefs, 31, 991-1003. Wilson S.K., Burgess S.C., Cheal A.J., Emslie M., Fisher R., Miller I., Polunin N.V.C., Sweatman H.P.A. (2008). Habitat utilization by coral reef fish: implications for specialists vs. generalists in a changing environment. Journal of Animal Ecology, 77, 220-228. Wolf N., Carleton S.A., Martinez del Rio C. (2009). Ten years of experimental animal isotopic ecology. Functional Ecology, 23, 17-26. Zeileis A. & Hornik K. (2007) Generalized MFluctuation Tests for Parameter Instability. Statistica Neerlandica, 61 (4), 488-508. Zeileis A., Hothorn T., Hornik K. (2008). ModelBased Recursive Partitioning. Journal of Computational and Graphical Statistics, 17 (2), 492514. Zekeria Z.A., Dawit Y., Ghebremedhin S., Naser M., Videler J.J. (2002). Resource Partitioning among Four Butterflyfish Species. Marine and Freshwater Research, 53, 1-6. !"# Appendices ! Appendix I : Geographical repartition of coral reef Figure I-1: Map of geographical distribution of the major coral reefs around the world. Note that the whole Indo-Pacific area concentrates the most important proportion of these reefs. ! 1 ! 2 Appendix II : Major applications of stable isotopes in ecology Table II-1 : Three major fields of application of stable isotope analysis in ecology, with a few examples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ppendix III : Sampling design for isotopic analyses Table III-1 : List of species captured within each region for isotopic analysis and their diet (main characteristic) (Only species with a minimum of five individuals were included) : SOUTH OF NEW-CALEDONIA (NC) : !"#$%&' 454/6789:-4;' 4CBDB/:-4;' HI;//::-4;' 54;(:B/:-4;' 574;6B-B/6:-4;' 7BIB5;/69:-4;' I4H9:-4;' I;679:/:-4;' I86K4/:-4;' @8II:-4;' @894;/:-4;' /;@:C6;9:-4;' ! ()*+$*,' !"#$%&'(')*+,-"&..* !"#$%&'(')*/'))'0.1(.* 21+(#)-0#*)"-3#)* 4#)-*'$."-($.)* 5%1$-"%')*)%(.#%')* 5&1.,-/.3%1(')*#(%')* 5&1.,-/.3%1(')*6'.$6'1,.$1#%')* !3-7-$*18-)%.70#* !3-7-$*"-03(1))')* !3-7-$*/-1/1(,1.$.* !("�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ppendix IV : Underwater Visual Census in Mururoa Location of Underwater Visual Censuses transects in Mururoa : Figure IV-1 : UVC Transects performed by EPHE (Ecole Pratique des Hautes Etudes) Figure IV-2 : UVC Transects performed by IRD (Institut de Recherche pour le Développement) ! 11 ! 12 Appendix V : R routine for the estimation of Functional Distinctiveness Index R script and functions used to estimate Functional Distinctiveness Index (FD) from Evolutional Distinveness (ED) of Isaac et al. (2007) : # Required libraries : library(MASS) library(ape) library(cluster) library(picante) library(ecodist) library(vegan) library(ade4) # Upload the Biological & Behavioral Traits database : THV <- read.csv2("THV_Region.csv", head = T, sep=";", dec=",") # Upload the needed function « Phylo2phylog » : phylo2phylog <function(phy, ...) { if(!require(ade4)) {stop("This function requires the ade4 package")} newick2phylog(write.tree(phy),...) } ##### Example for Gambier islands ##### # Calcul of the interspecies distances : dist.fish <- daisy(GB_THV, metric = "gower", stand = TRUE) # Construction of the cluster on the species dissimilarity : FD <- hclust(dist.fish, method="average") # Transformation of the cluster in a phylogenetic format : fish.func.tree=as.phylo(FD) plot(fish.func.tree) # Application of the ED function performed by Isaac et al. (2007) to obtain FD : fish.func.phylog <- phylo2phylog(fish.func.tree) fish.func.FD <- originality(fish.func.phylog, method=6) # 6 is equivalent to the ED function. dotchart.phylog(fish.func.phylog, originality(fish.func.phylog, 6), scaling=FALSE, yjoining=0, cleaves=0, ceti=0.5, csub=0.7, cdot=0.5) # # Storage of the results : write.csv(fish.func.ED) ##### End of script ##### ! 13 Appendix VI : Origin of the Layman’s derived metric Table VI-I : Litterature used by Cucherousset & Villéger* to construct their Layman’s derived metrics indexes : * for review only : the article has not been accepted yet, so the interpretation of these indexes has to be taken with caution. ! 14 Literature cited in Appendix VI : ! ! Cornwell W.K., Schwilk D.W. & Ackerly D.D. (2006) A trait-based test for habitat filtering : convex hull volume. Ecology, 87, 1465-1471. Garnier E., Cortez J., Billès G., Navas M.-L., Roumet C., Debussche M., Laurent G., Blanchard A., Aubry D. & Bellmann A. (2004) Plant functional markers capture ecosystem properties during secondary succession. Ecology, 85, 2630-2637. Laliberté E. & Legendre P. (2010) A distance-based framework for measuring functional diversity from multiple traits. Ecology, 91, 299-305. Layman C.A., Arrington D.A., Montaña C.G. & Post D.M. (2007a) Can stable isotope ratios provide for for community-wide measures of trophic structure ? Ecology, 88, 42-48. Mouillot D., Graham N.A.J., Villéger S., Mason N.W.H. & Bellwood D.R. (2013) A functional approach reveals community responses to disturbances. Trends in Ecology & Evolution, 28, 167-177. Villéger S., Grenouillet G. & Brosse S. (2013) Decomposing functional !-diversity reveals that low functional !-diversity is driven by low functional turnover in European fish assemblages. Global Ecology and Biogeography, in press. Villéger S., Mason N.W.H. & Mouillot D. (2008) New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology, 89, 2290-2301. Villéger S., Novack-Gottshall P.M. & Mouillot D. (2011) The multidimensionality of the niche reveals functional diversity changes in benthic marine biotas across geological time. Ecology Letters, 14, 561-568. ! 15 Appendix VII : Isotopic signatures of diet sources in New-Caledonia, from Carassou et al. 2008 : Table VII-1: Extract from Table 3 of Carassou et al. (2008) publication (Realized on a subset of the dataset used in the current work, data of New-Caledonia 1996). ! 16 Appendix VIII : Details about effect of size on isotopic signature Table VIII-1: Detailed Table of Results of the Size or the Size x Region effect on the isotopic signature of !13C and !15N for each species. Only species for which significant linear regressions between size and isotopic signature (p < 0.05) were found (for at least one isotope) are presented. For this analysis, a minimum of five individuals per Region was required. The boxed species are those for which the amplitude of the isotopic shift reach 3,4 ‰ for !15N in the sampled range of individual size, in the region that are underlined. The question marks indicate a relationship between size and isotopic signature which may be best fit by a non lienar relationship, therefore the linear relationship has to be considered cautiously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�,,',% 84#$%"/+7+,% 84#$%1"-+% 84#$%)4#?,'#"% 84#$%:+#+2+,% 84#$%/"71"% B",$%"#'"(',% B",$%97":+)"'2',% C?#$%10#(2*+% C?#$%6'(*00% =0.$%"#/0(*0',% =0.$%,"--"#"% @.-$%&":.% !"# $%# !"# !"# $%&#!"&#'(# !"# $%# $%&#!"# !"# '(# !"# $%&#!"# '(# '(# !"&#'(# !"&#'(# '(# $%&#!"# !"# $%&#!"# !"&#'(# $%&#!"&#'(# !"&#'(# !"# $%&#!"&#'(# !"# $%# $%# !"# $%&#!"# $%&#!"# $%&#!"# !"# !"# !"# '(# $%&#!"# $%# '(# !"# $%# $%&#!"# !"# !"# !"&#'(# !"# $%&#!"# $%&#!"# 18 )# )# *# +# *# )# +# *# *# *# )# )# )# )# *# *#&#+#2# )# *#&#)#2# )# )# )#&#+# *# +# )# )# )# +# )# +# +#&#)# )# +#&#)# *# )# )# *# )# )# )# )# *# *# *# +# *# )# *# )# +# )# )# )# +#&#)#&#)# )# )# )# )# )# *# +# *# *# )# )# )# )# *# +#&#)# *# )# *# +# +#&#+#&#*# *# *# +# *# )# +#2# *# )# *# )# )# +# *# )# *# +# )# )# )# )#&#+#2# *# )# *# # # # # ,#-#'## # # '#-#'# # # '#-#'# # # '#-#'# 2#-#,# # 2#-#'# # '#-#,# ,#-#'# '#-#'# '#-#'# # '#-#,# # # # ,#-#'# '#-#'# ,#-#'# # ,#-#'# # # # # '#-#'# # # '#-#2# # '# '# !15N! Appendix IX : Example of relationship between size and !15N for six species (one per diet) Size (mm)! Figure IX-1: Example of the relationship between size and !15N for six species, one per diet, in the region of Mururoa. The model for each species is represented in dotted lines with pvalue and adjusted R-squared. ! 19 Appendix X : Main information about SIBER ellipses by species within a region Table X-1: Main information about isotopic niche width based upon SIBER ellipses for each species represented by at least ten individuals in each of the three regions of Mururoa, NewCaledonia and Gambier Islands. The Functional Distinctiveness Index (FD) is indicated for each species also. GAMBIER ISLANDS !"#$%#!& "#$%&'%$'()&*+&,'-)./-+$)( "#$%&'%$'()01%/%$)( ".-2+3$1$)(2/410'$).$)( ".-2+3$1$)(+1/0)+&4$)( 53&/6/2$)(.36010$1$)( 5310,/)(-4/6/)( 5310,/)(#-,/( 5+&20.3-&+$)('6-7/.-$%-( 5+&20.3-&+$)()+1/-+$)( 8-).966$)(-1$-2$)( :*/2&*3&6$)(3&;-402-+$)( :*/2&*3&6$)(,&11-( <-6/.30&1&)(+1/,-.$6-+$)( =-140.&2+102(%/-%&,-( =.-1$)(*)/++-.$)( >3-6-))0,-(6$+&).&2)( !#'(& ./.01/21./.& 8.04676/358& 85/03431367& .5704//5/73& .2102368756& 44023346/44& .7048812381& 6103858258& 157042828.5& 4420/167884& 438085.78.1& 8/30.66.687& 15.0557.8/.& 4640831.2.7& 415034.2578& 18506733867& !#'!&)*+& .330/44144.& 3/03411./85& 3730475677.& .160542785& 4620126268& 42037/44771& .40873/1183& 54083843233& 13.0/78458& 17706//.833& 1/50...8614& 8330//.2324& 1330358656.& 4470/6/21./& 1720535/328& 1.507836/37& !#'!&!,& //05416.3.7& 130188267/7& 1.50127558/& 7.081/.2674& 5/073/4442.& 506///548.4& 120/11181/3& 160/551/.8.& 3.0681/4/3/& 85037447423& 5.014615824& 11.02712824& 32041/46278& 55054418..4& 8801.14331& .6075452138& -,& 20116723481& 20116723481& 2026/7/47.6& 202/6/45751& 20265255438& 2026652/4.7& 2026685833.& 202674/.5/4& 202/222//86& 20275481773& 20267557574& 20267433715& 2026871418/& 202728/8671& 2027344.127& 20268313226& !#'(& .2.02.3/3.& 44305//87/3& 4360//5.741& 148088///25& 848078288.3& .2202224.6/& 45701.4.571& .1404171837& .2706./.84/& 1/502433721& 4210441647& 174086543.1& 4580477747.& 180/477/2/8& !#'!&)*+& 4.801.4824.& 1/.033.//78& 41207866416& 1240544518.& 3.20/235.27& 4/8063/4/55& 4410488/255& 4/80.57.184& 48603218.58& 18/04/55288& 17/0.314/88& 18501245268& 4420235828& 1308/746756& !#'!&!,& 7.02384358/& 810/5833.7.& 87041237575& .10.5223.11& 1.80//76258& 45055873/74& 610.486/87.& 82075485864& 7.04//1538/& 850.1/6576.& .502.1/4653& 33025512377& 560388...67& 30542145115& -,& 20263577168& 20268.3213.& 2027846/437& 2026333436& 20263531615& 20263838264& 2026358152/& 202633/6251& 2026.16862.& 2026.24./56& 202682.4574& 2027774656/& 2027641.16.& 202657/4/5& NEW-CALEDONIA !"#$%#!& "*0402(%0&%&16&/2/( 5&2+10*94&(#/)*/20)-( 53&/60%/*+&1$)(-1+$)( 5/11/*&.+&)()+/4,-+/.$)( ?9,20+301-;(-6#/,-14/2-+$)( ?9,20+301-;(.3/60)*/6$)( ?9,20+301-;(&$10)+$)( ?9,20+301-;('/,#1/-+$)( ?9,20+301-;(,06$..&2)/)( ?9,20+301-;(*/2%-&( ?9,20+301-;($2%$6-+$)( @&+31/2$)(4&2/7/++-+$)( @$+A-2$)(B$/2B$&6/2&-+$)( C91/*1/)+/)(D$2+&&( ! 20 !"#$%#$&'$&()$*+,-.,( /,&*(01$-*#1$&( /.2$%'.#0&(30#-*&0&( 4+.-'#*%*20&(+.*%,#50&( 4*2,-.1'#0&(2*+0--.1&$&( 6,#7*-.1'#*1(5$,5.2,( 6-,#0&(#$)0+,'0&( 6-0'$-,#$,('$7#$1,( !"#!$"%&'!$( $&)#&+'!$"'( '"#!,%"%*&%( ,'!#$%)"&%,( +*$#+*"*''&( ,%'#+$+)"''( +,!#&$%%""( %%#,%)")+&)( $%#)*+!'&!$( $+&#&&+""%( ')#*"",+,%)( ,%,#&$%%,( '+#,$*)""''( ,,'#'+*)$+"( +*&#&%,*","( "*#"%*!'))!( +!#",%+%)*+( %&#$,$!'+%!( +"#*'$!+"$'( !&#)$+!,'+,( ),#&'!"!!$)( %&#$)%*)*,+( ,$#'','"&*!( ))#+%$!+&'$( *#*"&"&,%$,( *#*%+"+*++$( *#*",,)+$&,( *#*%)%%'%!&( *#*""*$$*"!( *#*"&,!&!*%( *#*%+*&!"&$( *#*"!&++!+( MURUROA -./01/-( (((-/23( 9:05.3503(&.;3,&-$,'0&( %&#,&%,$*)$( 9-,1'<0#0&(1$7#*30&-0&( ,,*#,*!$,+&( 9-,1'<0#0&('#$*&'.70&( +)$#&$$&!),( 9%*7*1(5*#"&&,( +,'#'&$&*%"( 9%*7*1(',.1$*%'.#0&( $%#$+)+$!&)( 90+*&'*20&(-<$1.1&$&( $"#"+)+"*",( =,1'<$7,&'.#(&*+,15#$( +,"#%"!!)),( =.1'#*%"7.(3+,)$&&$2,( +'+#,%',$*%( =.%<,+*%<*+$&(,#70&( +)*#&"%)$"+( =<,.'*5*1(,0#$7,( +!$#)+++)"( =<,.'*5*1(+010+,'0&( &*#,$)%)$&!( =<,.'*5*1('#$3,&-$,+$&( ++,#+",%"!"( =<.$+$10&(-<+*#*0#0&( +"$#)"$)!!+( =<.$+*5$%'.#0&(>0$1>0.+$1.,'0&( +"*#"%!!*&,( =<+*#0#0&(2$-#*#<$1*&( +$$#%*+,$'+( =<+*#0#0&(&*#5$50&( !%#&$+$*"!!( =<#*2$&(,'#$%.-'*#,+$&( !+#")'&&$&$( =<#*2$&()$#$5$&( ++'#""%$*'&( =#.1$207$+(-#.1$+,:$&( !"#+"*'+,)&( ='.1*-<,.'0&(&'#$,'0&( )"$#&'+$*&,( ?,&-"++0&(,#0,10&( +%'#""*&$+( ?,&-"++0&(3+,)$-,050&( ''#)&$*%$*&( @%$:0+0&($1&$5$,'*#( &+#!&'+&'&'( @%$1.%<.+0&(<.;,7*1,'0&( +',#)$&+%)"( @%$1.%<.+0&(2.##,( )$*#+"%*'&'( A1,'<*5.1'.;(,0#.*+$1.,'0&( )!$#**)&',&( A*2%<*&0&(),#$0&( "&#'$!",$,+( A"21*'<*#,;(-<$+*&%$+0&( +")#"&&*$*"( A"21*'<*#,;(.0#*&'0&( '!#+**+'!))( A"21*'<*#,;(B,),1$-0&( %"#'',)+'),( A"21*'<*#,;(2,#7,#$'*%<*#0&( ++)#!!,&!+( A"21*'<*#,;(0150+,'0&( %'#&!)!*"$,( C,+$-<*.#.&('#$2,-0+,'0&( +,*#!*+)+%'( C.'.#*%#$,-,1'<0&(-#0.1','0&( &*#)!*%"%""( ! 21 -/28(456( ""#)%'!+%%&( ,++#%$'$*!%( +*&#%''&"&%( ++!#+*+)%'"( $*#&%%)),''( $)#,*$&+!!%( +,*#!'$++*%( +"&#%)"*+"'( +*"#%%+,)''( +$)#$)!!*"+( !,#+)),!*!!( ''#$,*',,%'( +&&#&'+),$&( +$%#$$$'"$"( +)&#+,$$)%%( !,#%)!*!$)"( $"#)')'!)*+( +*%#'))!%),( !)#$,&$$$&+( )&)#$++',,$( +")#$"))$',( ',#+!"%!'*+( !+#$%"*$&,&( +"'#,'$+$%$( ))'#$),&*!( )$+#*"+&"+( ")#,%$)'&$"( +$+#*'!%)')( %*#%,+)+$"( ""#"$'$'+',( ',#))!!"&+,( %)#,$$$'&!'( ++"#$$)%%*,( !*#,*")$&",( -/28(-7( +%#'%&!+$,,( !)#!*&$!!$"( ,,#"!,*&*!%( $*#*,$*&,"'( +,#,"&%**$'( +)#!!!'"*))( )+#&)"!!,&)( ,%#')!+,,+%( )&#,!$)+%$+( )"#%)+'*"&"( +!#&&&$'&$%( )&#+&**)$!+( )&#+*")',&,( $*#,,*)!+*$( ,$#**)+"!')( +)#!'!$++)( +*#&$++*&*!( ))#&")''+',( +)#%!&+,+)"( ,'#$""%"%&)( ,'#',!!!&%+( +"#*+)')+$( +'#%'!!*$+( ,"#%!"*&%%+( ,,#$++)&)%+( ,"#)$%"),&'( +,#+"""+&")( $!#')''+!)+( )!#*,"%)**)( ))#')%&$%''( ,+#$)*$*&%'( +%#,!$"*'')( )%#',,)!)%!( +!#%+,*&$),( 87( *#*%,$%,)&!( *#*%,+$!'&'( *#+)"$,$%,&( *#++*'!,'++( *#*%'*+$$,$( *#+)')")"'"( *#*"&&"'%'"( *#*%,'&&)!( *#+*!*',"&+( *#*"!'%%+!"( *#*"'$&','+( *#*"'!)"'+( *#*""*!$"%'( *#*''$$&)+$( *#*%"&'&+++( *#*%!&%&"( *#*'*"+)%',( *#*'*"+)%',( *#++%)'+"%)( *#*%'*')+&%( *#*',+!'",!( *#*',,,',%$( *#*'&"*)!$+( *#*%+,+")"$( *#*%*"'+&,$( *#+$$"+&*$( *#*"&!"",%$( *#*%+$&'))%( *#*%*,,+$$'( *#+,$$$",+&( *#*%+*&+&'( *#*%'%)*$'!( *#*"&$!$",)( *#+*+'*','"( !"#$%&'(#&)*+',)-)*+ ./-&%01)*+'-201*$10+ ./-&%01)*+$304',/)*+ .)-5'1)*+2'*(0%'+ 6)33$070,&-&"*+83'4$301/'-)*+ 6)33$070,&-&"*+4'10,$3/1*0*+ 6"%0#%0*-0*+9/%17-0+ 6"%0#%0*-0*+2)1-//+ 6"%0#%0*-0*+40$3',/'+ :'*$+30-)%'-)*+ :'*$+)10,$%10*+ :/,-'(0'+*'4'"/1*0*+ :/$10#&$1+'%;/1-/)*+ :/$10#&$1+*'(('%'+ <'%'#/%,0*+(033/#)1,-'-'+ <'%)#/1/)*+()3-08'*,0'-)*+ <'%)#/1/)*+#3/)%$*-0;('+ <3/,-%$;3"#&07$7$1+70,200+ ='%;$,/1-%$1+70'7/('+ ='%;$,/1-%$1+(0,%$*-$('+ ='%;$,/1-%$1+#)1,-'-0**0()*+ ='%;$,/1-%$1+*#0108/%)(+ ='%;$,/1-%$1+-0/%/+ =')%07'+;%',030*+ =,'%)*+;&$99'1+ =,'%)*+;3$90,/#*+ =,'%)*+*,&3/;/30+ =0;'1)*+'%;/1-/)*+ =-/;'*-/*+8'*,0$3'-)*+ =-/;'*-/*+10;%0,'1*+ =-/-&$5)30*+9'17'1/1*0*+ ="1$7)*+4'%0/;'-)*+ >&'3'**$('+&'%7?0,2/+ >&'3'**$('+3)-/*,/1*+ >&'3'**$('+@)01@)/40--'-)(+ A/9%'*$('+*,$#'*+ A/9%'*$('+4/308/%)(+ ) ! !"#$"!%&'$() (,#%,,,"%"&) '$!#$,+&%&') ++#&%!*',&%) *,"#*(**%,") '$'#%!%&!!") '&(#(!&$&+") ''&#$!(,+'") ''+#&!'%&%") &%'#!((++'') +!$#'%&($"+) '&+#((&!"&) %!#,+,!(*&*) '*%#,,'',$,) &,#("+%%*,() (*#"&!%"&$%) '+&#%'""((*) ''$#',(!"$') ",#++"$(*'*) '*!#*!"$!(%) '&$#*!$+$(%) '*$#'*+"'(+) %$#!*!*"+&") '!$#++&&*&%) '"'#"'*"$(!) *%&#&!'%&"") *&!#$&($&"*) '(&#%$'*!() *'$#,+,$&(') '!*#+(*$"!,) '(!#'"(*!"%) *,$#!&*"') '+'#&*****) %&#+*"+$&"&) ',,#$+"&**,) '!'#&",&&*&) '*'#+,%+&!') ) 22 !'#%*&*$""") !!#,(*&,!(+) %(#&""'",%&) +'#+'('%!$+) *+$#%"&&'%() %!#!+*+%(&') '*"#$$!&(%&) '$*#(&$",,!) ''$#$$&'"&*) &(&#,"$!!&,) &!!#%***!$() '*"#&$%+!*&) ",#",(*$"$%) '*'#,$+(''() &$#('$(&*%&) !%#++,&'(+) ''(#*%*'*,!) %*#,+&',',,) "'#*(!%%*&&) '*$#&&+"*$&) ''%#"&$%*"") '',#&,*(!*") ",#(,+$$+*%) ',$#!(%''"+) '!'#%&$%(%') *,$#"%$'&'') *$,#+++,!'%) ',%#$%*,""!) '%,#+&$,$"() '&"#$'!("**) ',"#'&,'%,,) *',#"!$(*,!) '*'#%(!&$%) "%#(&'('*+*) '+$#&%'($"*) '+!#+&**'%&) ''&#**%+("+) '+#",%'%+",) *$#*+!,%*(!) *&#*$,$!%$%) %#%*,"+"&+") +%#$(!(&&') '(#!,!'*"*%) *!#%&&*"++') **#!$!&%&'%) '(#!*$&,&,+) !'#,+"+%,$+) '*'#(%**"&') **#!!"+(&'%) *&#!%'!$""&) **#!'"++'") %#%"',+!$%() '*#&(%%*%!,) &(#!!(+(%'() *"#*&'%**$() '*#"$%,(((!) '"#(,(&'("() &$#!,$&"%!%) '(#&'&,*"$") '+#*&&((+"*) *%#,""!$!"() +*#+($"(*'%) (*#&&+(%"&!) ,&#%%!+,+"!) &,#$%%(+,!%) &"#$,,''''() &%#!(+"!+,') &%#(!,!!!*,) !$#,'",+'+*) &(#'%"(%(*%) '&#,'(!('*) &'#,!!(&!&() &&#$%*%*%!') *'#(%'%"&(") $#''$!&((*&) $#'!&'(*$'+) $#'*!*&!*(() $#'++('!$+) $#$",($%+&) $#$",($%+&) $#$"!"$"%'!) $#$"!%",*&") $#$"!,!%&"&) $#$"&(!!&*%) $#$""(*&&'%) $#$""+',!$*) $#$"!%",*&") $#$"!"$"%'!) $#$(!$""",*) $#$(!+&%(!') $#$(!+&%(!') $#$(!!*$$'+) $#$"*&($*%() $#$"*!$%"*%) $#$"*!$%"*%) $#$",!$"++) $#$"&$!"'(+) $#$"$""'+,") $#$"(!%!''') $#$"!+$%"+%) $#$"+**!',') $#'*(+&+"&!) $#$"*!&&!&,) $#$"(",&%(") $#$(,%!!"%*) $#$%%&*+&%*) $#$(!*!%!$+) $#$(,%(,!!() $#$(!*$!+*+) $#$"+&$,*(() $#'$(&'+("!) ) ) ) Appendix XI : Table of significancy of the biological factors effect on isotopic niche width Table XI-1: significance of the biological factors, tested on (a) !13C and !15N, (b) SEAB (i.e isotopic niche width) : Despite the experimental design was unbalanced, the interactions between biological traits (Table 5) have been tested even though they should be taken cautiously, as they strongly improve the model (See residuals in Appendix XII). However the effect of each trait will be tested within diet with ANOVA and Tukey post-hoc tests when possible (See Table 5 in Results for details) in order to account for functional group without statistical bias. Factor effect ! ! ! (a) !13C (a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b) SIBER Ellipses area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ppendix XII : Residuals of the model (1) and (2) performed on Appendix XI Residuals of model (1) on !13C Residuals of model (1) on !15N Residuals of model (2) on !13C Residuals of model (1) on ellipses Appendix XI Tableau des SIBER par specie et région, avec data SEA.B, eccentricity, etc. Residuals of model (2) on !15N ! Residuals of model (2) on ellipses 24 Appendix XIII : Plots of isotopic signatures in zones of Mururoa (a) Sheltered zone (b) South-West zone (c) Lagoon zone (d) Ocean zone Figure XIII-I : Plots of !15N against !13C signatures of fish species and different diets in Mururoa for the (a) Sheltered, (b) South-West, (c) Lagoon and (d) Ocean zones. The grey points represent the mean value per species, the mean value for each diet are indicated on the graph in black points with errors bars. ! 25 Appendix XIV : Table of density (individuals/m2) for each diet (regardless to species) within zones of Mururoa : Table XIV-1 : Density (individuals/m2) for each diet within zones of Mururoa "#$%&$'$(! %-.))/! )0$-/! &)&-%! !!"#$"%&'(#) 12134! 12155! 12163! 12167! 18159! *('+"%&'(#) 12:7! 1256! 12;;! 1247! 18;6! ,-'."%&'(#)/0) 12;6! 4214! 1297! 12;4! 1896! ,-'."%&'(#)/1) 121;:! 12195! 12176! 121:<! 1819! 23."%&'(#) 12;9! 1297! 1259! 12:5! 1835! !4-.56"%&'(#) 4261! 1274! 423<! 1251! 4849! ! ! ! ! ! ! ")*&#+,$"&! 26 ! Appendix XV : Layman’s derived metrics graphical outputs Legend of following figures : a. Mean isotopic position, b. Trophic Richness, c. Trophic Divergence, d. Trophic Dispersion, e. Trophic Eveness, f. Trophic Uniqueness. I. Piscivores (FC) Figure XV-I (a) : Layman’s derived metrics values for FC in Sheltered zone of Mururoa - Figure XV-I (b) : Layman’s derived metrics values for FC in South-West zone of Mururoa ! 27 Figure XV-I (c) : Layman’s derived metrics values for FC in Lagoon zone of Mururoa - Figure XV-I (d) : Layman’s derived metrics values for FC in Ocean zone of Mururoa ! 28 Appendix XV : Layman’s derived metrics graphical outputs II. Herbivores (HD + HM = H) Figure XV-II (a) : Layman’s derived metrics values for H in Sheltered zone of Mururoa - Figure XV-II (b) : Layman’s derived metrics values for H in South-West zone of Mururoa ! 29 Figure XV-II (c) : Layman’s derived metrics values for H in Lagoon zone of Mururoa - Figure XV-II (d) : Layman’s derived metrics values for H in Ocean zone of Mururoa ! 30 Appendix XV : Layman’s derived metrics graphical outputs III. Mobile invertebrate consumers (IM) Figure XV-III (a) : Layman’s derived metrics values for IM in Sheltered zone of Mururoa - Figure XV-III (b) : Layman’s derived metrics values for IM in South-West zone of Mururoa ! 31 Figure XV-III (c) : Layman’s derived metrics values for IM in Lagoon zone of Mururoa - Figure XV-III (d) : Layman’s derived metrics values for IM in Ocean zone of Mururoa ! 32 Appendix XV : Layman’s derived metrics graphical outputs IV. Sessile invertebrate consumers (IS) Figure XV-IV (a) : Layman’s derived metrics values for IS in Sheltered zone of Mururoa - Figure XV-IV (b) : Layman’s derived metrics values for IS in South-West zone of Mururoa ! 33 Figure XV-IV (c) : Layman’s derived metrics values for IS in Ocean zone of Mururoa (No sessile invertebrate consumer (IS) species were available for Lagoon zone in our data). ! 34 Appendix XV : Layman’s derived metrics graphical outputs V. Omnivores (OM) Figure XV-V (a) : Layman’s derived metrics values for OM in Sheltered zone of Mururoa - Figure XV-V (b) : Layman’s derived metrics values for OM in South-West zone of Mururoa ! 35 Figure XV-V (c) : Layman’s derived metrics values for OM in Lagoon zone of Mururoa - Figure XV-V (d) : Layman’s derived metrics values for OM in Ocean zone of Mururoa ! 36 Appendix XV : Layman’s derived metrics graphical outputs VI. Planktivores (PK) Figure XV-VI (a) : Layman’s derived metrics values for PK in Sheltered zone of Mururoa - Figure XV-VI (b) : Layman’s derived metrics values for PK in South-West zone of Mururoa ! 37 Figure XV-VI (c) : Layman’s derived metrics values for IS in Lagoon zone of Mururoa - Figure XV-VI (d) : Layman’s derived metrics values for IS in Ocean zone of Mururoa ! 38 Appendix XVI : Divergences between FishBase and isotopic data Table XVI-I: Examples of divergences between isotopic mean value (!15N) and the Trophic Level of a species as provided by Fishbase : SPECIES Trophic Level (Fishbase) Mean !15N Suggestion Acanthurus blochii Apogon doryssa Apogon exostigma Bothus pantherinus Lutjanus kasmira P. johnstonianus Pomacentrus adelus Sargocentron rubrum S. spiniferum Sargocentron tiere Chl. frontalis Chl. microrhinos Chl. sordidus Scarus altipinnis Scarus globiceps Scarus niger Scarus psittacus Scarus schlegeli 2.27 3.24 3.45 3.96 3.48 2.26 2.26 3.48 3.48 3.58 2.59 2.59 2.52 2.59 2.52 2.52 2.52 2.52 8.88 9.81 6.91 6.66 11.36 8.95 9.05 11.15 10.45 10.60 5.14 5.82 5.77 5.01 5.96 5.84 5.22 5.58 Could eat small invertebrates ? Could eat fish larvae Could be planktivore May not be piscivore Mainly piscivore ? Could be omnivore Could be omnivore Could be piscivores Could be piscivores Could be piscivores May be herbivores May be herbivores May be herbivores May be herbivores May be herbivores May be herbivores May be herbivores May be herbivores Other examples could be given. These species are those for which a high divergence was found with our data. ! 39 Diplôme : Master de l’Institut Supérieur des Sciences Agronomiques, Agroalimentaires, Horticoles et du Paysage Spécialité : Approche Ecosystémique de l’Halieutique Spécialisation / option : Enseignant référent : Etienne RIVOT Auteur(s) : Chrystelle DELORD Organisme d'accueil : IRD CoReUs 2 Date de naissance* : 08 avril 1991 Adresse : Laboratoire Arago Nb pages : 36 Annexe(s) : 16 (39 pp.) Année de soutenance : 2013 Avenue du Fontaulé 66650 BANYULS-SUR-MER Maître de stage : Michel KULBICKI Titre français : Analyse des facteurs influençant la niche trophique dans les communautés de poissons récifaux : apport des isotopes stables de C et N. Titre anglais : Defining drivers of the trophic niche width in reef fish communities : support from C & N stable isotopes. Résumé (1600 caractères maximum) : Si la composition spécifique des communautés de poissons de récifs a été fréquemment mise en relation avec les caractéristiques du milieu, il en est autrement de la composition fonctionnelle. A l’aide des isotopes stables du carbone et de l’azote, considérés comme représentatifs du régime alimentaire la largeur de la niche trophique de plusieurs espèces représentatives des groupes fonctionnels majeurs des poissons de récifs de Nouvelle-Calédonie et Polynésie Française a été analysée. La taille des individus influe sur les compositions isotopiques mais sans pattern régional. Aucune relation stable n’a pu être mise en évidence entre la largeur de la niche trophique des espèces et leurs traits de vie (excepté l’alimentation), ni avec le degré d’endémisme ou la singularité fonctionnelle. En revanche une relation significative existe entre la taille de niche trophique d’espèces herbivores ou coralivores et leur fréquence/biomasse/abondance, suggérant un effet de la compétition. Aucune tendance stable n’a été mise en évidence quant à l’effet d'un gradient d'influence océanique sur la surface de la niche trophique. Le regroupement des observations par groupe fonctionnel, par contre, engendre davantage de stabilité quel que soit le facteur considéré. Les résultats suggèrent que les poissons de récifs coralliens présentent une versatilité trophique inattendue, leur conférerant un fort potentiel d’adaptabilité et donc de résilience face aux perturbations avec cependant des variations suivant les groupes fonctionnels considérés. Abstract (1600 caractères maximum) : While the species composition of reef fish communities has frequently been linked to habitat characteristics, the functional structure of these communities still remains poorly known. With the support of carbon and nitrogen stable isotopes, which are considered to be representative of diet, we attempt to understand how the trophic niche width evolves for a representative number of species from the major functional groups of reef fish communities from New-Caledonia and French Polynesia. Individual size influenced the isotopic composition, but no regional pattern could be demonstrated. No stable relationship could be found between the trophic niche width and the life-history traits of species (except for diet), nor with endemism or functional distinctiveness. However, a significant relationship exists between niche trophic width of herbivores or corallivore species and their frequency/biomass/abundance, suggesting an effect of competition. No major trend was highlighted when considering a gradient of oceanic influence on trophic niche width. Grouping data according to functional groups generated more stable trophic niche widths, for anyof the analyzed factors . These results suggest that reef fish species are unexpectedly versatile in respect of their trophic niche, giving them a high potential of adaptability and thus enhance their resilience against perturbations. This potential may however vary from one functional group to another. Mots-clés : isotopes stables, poissons de récifs, spécialisation, niche trophique, pacifique sud Key words : stable isotopes, reef fishes, specialization, trophic niche, south pacific