UNIVERSITE DE GENEVE FACULTE DES SCIENCES Institut Forel Prof. Anthony Lehmann HESSO//GE HAUTE ECOLE DU PAYSAGE, D’INGENIERIE ET D’ARCHITECTURE Institut Terre-Nature-Environnement Prof. Beat Oertli Biodiversité des mares et étangs: impact du réchauffement climatique et de l’eutrophisation THESE Présentée à la Faculté des sciences de l’Université de Genève pour obtenir le grade de Docteur ès sciences, mention sciences de l’environnement par Véronique ROSSET de Corsier (GE) Thèse n°4396 GENEVE Atelier d’impression ReproMail 2012 Rosset, V.: Biodiversité des mares et étangs: impact du réchauffement climatique et de l’eutrophisation. Terre & Environnement, vol. 108, 245 pp. (2012) ISBN 978-2-940472-08-6 Section des sciences de la Terre et de l'environnement, Université de Genève, 13 rue des Maraîchers, CH-1205 Genève, Suisse Téléphone ++41-22-379.66.28 - Fax ++41-22-379.32.11 http://www.unige.ch/sciences/terre/ « Ne désespérez jamais. Faites infuser davantage. » Henri Michaux Remerciements Je tiens tout d’abord à remercier Beat Oertli pour m’avoir fait découvrir le monde des mares et étangs et pour m’avoir donné l’opportunité de me lancer dans cette grande aventure. Un grand merci pour avoir passé du temps à relire et encore relire mes articles et pour m’avoir poussée à me dépasser. Je remercie également Anthony Lehmann de m’avoir initiée à GRASP et d’avoir toujours été très disponible pour répondre à mes questions statistiques ainsi que de m’avoir accueillie à Battelle pour finir cette thèse. Je remercie également Dominique Vallod pour m’avoir fait découvrir les étangs de la Dombes scientifiquement lors de discussions sur leur fonctionnement ou concrètement en participant à une pêche ainsi que pour sa grande disponibilité. Je tiens également à remercier Michael Samways (Université de Stellenbosch) et Emmanuel Castella (Université de Genève) d’avoir accepté d’évaluer le présent travail. Ce travail n’aurait pas pu aboutir sans l’aide de nombreuses personnes qui ont participé aux campagnes de terrain, au travail de laboratoire et à la gestion des bases de données. Je remercie toute l’équipe PLOCH du LEBA, toute l’équipe d’hepia, de l’ISARA et de l’Université de Lyon. Je remercie vivement Pascale Nicolet et Jane Park qui ont toutes les deux corrigés avec brio l’anglais de mes publications. Ce travail a également bénéficié de ma collaboration développée en Afrique du Sud avec le Département d’Ecologie de la Conservation et d’Entomologie. Je remercie vivement Michael Samways pour m’avoir si bien accueillie et pour les discussions constructives et John Simaika pour tous les moments partagés ainsi que toutes nos discussions animées. Je remercie les différentes institutions, fondations et collectivités qui ont participé au financement de ce travail : l’Office fédéral suisse de l’environnement (OFEV), de nombreux cantons suisses, l’Université de Genève (Laboratoire d’Ecologie et de Biologie Aquatique), la HES-SO (RCSO RealTech), la commission de recherche du Parc National Suisse, l’Agence de l’Eau Rhône Méditerranée Corse, le Département français de l’écologie, de l’énergie, du développement durable et de la mer, la PEP Aquaculture Rhône-Alpes et la fondation Vérots. Ce travail a également bénéficié de la participation des propriétaires d’étangs dombistes ayant participé à l’étude. Je remercie également pour leur soutien financier lors de congrès et de déplacements en Afrique du Sud : le Programme de recherche commune entre Suisse et Afrique du Sud, les relations internationales HES-SO, la Société Suisse 1 d’Hydrologie et de Limnologie SGHL, et pour leur soutien lors des déplacements dans la Dombes, la bourse Germaine de Staël de l’Académie Suisse des Sciences Techniques. Un grand merci également à tous les collègues de hepia avec qui j’ai partagé le goût de la nature pendant plus de 3 ans et en particulier, l’axe aquatique et mes collègues de bureau: Sandrine Angélibert, Nicola Indermuehle, Michaël Delaharpe, David Leclerc et Eliane Demierre, pour tous les moments de joie et de panique partagés. Un merci particulier à Jane O’Rourke pour sa relecture du résumé en anglais. Merci également à tous les stagiaires et étudiants avec qui j’ai travaillé durant ma thèse. Je remercie tous les collègues de Battelle pour leur accueil, pour toutes les discussions ainsi que pour les readings groups qui étaient fort stimulants. Je remercie ma famille et mes amis qui ont suivi mes aventures d’assistante et de doctorante, m’ont soutenu lors de ce parcours et se sont réjouis à chaque étape franchie, et tout particulièrement ma maman pour sa relecture de certains passages de ma thèse. Enfin, un grand merci à mon mari pour m’avoir soutenue et supportée tout au long de ce travail, et pour avoir accepté de donner son avis sur des figures incompréhensibles, de s’occuper de notre fils Clément pendant que je filais sur le terrain, en congrès ou en Afrique du Sud. Merci également à Clément pour son intérêt pour le travail de maman « attraper les grenouilles » et pour sa patience lorsque je m’éclipsais au travail. Merci à tous. 2 Résumé Les mares et étangs et les réseaux qu’ils constituent collectivement, sont des milieux remarquables ayant une forte valeur écologique, sociale et économique. Ils abritent une biodiversité unique qui est souvent plus riche que celle présente dans les eaux courantes ou les lacs. Malgré cette biodiversité exceptionnelle, les mares et étangs ont été et sont encore fortement altérés par les activités humaines. Deux des menaces principales pesant sur les mares et étangs sont le réchauffement climatique et l’eutrophisation. Le réchauffement climatique est manifeste et observé dans le monde entier et a de multiples impacts sur les espèces et les communautés. Il provoque notamment des déplacements d’aires de répartition des espèces qui induisent des colonisations et des extinctions dans une région ou un écosystème donné et modifient la richesse locale et régionale. Il est observé actuellement que la richesse spécifique régionale augmente dans les milieux terrestres comme aquatiques. Bien que l’échelle locale soit moins connue, les études actuelles suggèrent également une augmentation de la richesse spécifique. L’augmentation de la charge en nutriments, eutrophisation, est un problème important pour la biodiversité des écosystèmes d’eau douce, causant notamment une augmentation de la productivité primaire, des blooms algaux et une forte désoxygénation. La relation entre productivité et richesse spécifique est peu connue dans les milieux aquatiques, mais les études actuelles suggèrent qu’elle consiste en une courbe en cloche, présentant une richesse maximale dans des conditions mésotrophes et une richesse plus faible dans des conditions oligotrophes ou hypertrophes. Les objectifs du présent travail sont (1) de déterminer l’impact du réchauffement climatique sur la biodiversité locale (étang) et régionale (pays ou région), (2) de déterminer la forme de la relation entre productivité et biodiversité à l’échelle locale et finalement (3) d’essayer d’évaluer l’impact combiné du réchauffement climatique et de l’eutrophisation. Une centaine d’étangs de Suisse ainsi qu’une centaine d’étangs de France (Dombes, Ain) ont été étudiés. Le jeu de données suisse couvre un large spectre de conditions de température et de niveaux trophiques. Le jeu de données français comporte essentiellement des étangs hypertrophes. La biodiversité de quatre différents groupes est étudiée : les macrophytes, les macroinvertébrés aquatiques (le peuplement entier, mais avec une attention particulière sur les gastéropodes et les coléoptères), les odonates adultes et les amphibiens. Deux métriques descriptrices de la biodiversité sont utilisées : la richesse taxonomique et la valeur de conservation des peuplements, cette dernière exprimant le degré de menace des espèces. 3 A l’échelle régionale (ici en Suisse), le présent travail a montré que, en réponse au réchauffement climatique, la proportion d’espèces colonisatrices (« gagnantes »), donc d’espèces nouvelles pour le pays, serait plus importante que celle d’espèces à risque d’extinction (« perdantes »), suggérant une augmentation de la richesse spécifique régionale. De plus, les espèces identifiées comme à risque d’extinction à cause du réchauffement climatique ne sont pas toutes sur Liste Rouge ; il serait alors important pour la conservation de la biodiversité de mettre en place un indice de sensibilité face au réchauffement climatique ou d’intégrer cet indice dans les critères de sélection des espèces prioritaires. Concernant l’impact de l’eutrophisation à l’échelle régionale (ici en Dombes), une analyse préliminaire suggère que, pour des étangs de plaine riches en nutriments, la richesse taxonomique diminue pour tous les groupes taxonomiques sauf les amphibiens, mais que cette diminution n’est marquée que pour les macrophytes. A l’échelle locale, la richesse spécifique des étangs va potentiellement augmenter sous l’effet du réchauffement climatique et cette augmentation sera particulièrement importante en altitude. Cette augmentation est prédite pour tous les groupes taxonomiques étudiés (macrophytes, gastéropodes, coléoptères, odonates adultes et amphibiens) mais avec des différences d’amplitude. La forte sensibilité des étangs au réchauffement climatique en fait des écosystèmes idéaux pour la surveillance de l’impact du réchauffement climatique sur la biodiversité, et ce en particulier en milieu alpin. Concernant l’eutrophisation, une forte hétérogénéité de réponse selon les groupes taxonomiques considérés a été mise en évidence pour des étangs de plaine riches en nutriments : la richesse taxonomique locale en macrophytes et macroinvertébrés diminue avec l’augmentation de leur charge trophique alors que la richesse en gastéropodes et en amphibiens ne montre aucune tendance significative. La valeur de conservation des peuplements s’est avérée ne pas varier avec l’augmentation de la charge trophique, sauf pour les plantes aquatiques où elle diminue. Ces différences entre richesse taxonomique et valeur de conservation démontrent que la richesse ne peut pas être un indicateur de la valeur de conservation. De plus, l’absence de relation entre valeur de conservation et niveau trophique suggère qu’une même valeur de conservation peut potentiellement être observée dans un écosystème riche en nutriments comme dans un écosystème moins chargé en nutriments. Des réseaux incluant des étangs avec des niveaux trophiques différents pourraient donc favoriser la biodiversité au niveau régional. Concernant l’impact combiné du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs, le présent travail a montré que, à l’échelle locale, l’eutrophisation atténuerait l’augmentation de la richesse spécifique due au réchauffement climatique en plaine, mais 4 uniquement pour les macrophytes et les gastéropodes. En altitude, pour tous les groupes taxonomiques excepté les amphibiens, une eutrophisation modérée accentuerait l’impact du réchauffement climatique, alors qu’une eutrophisation élevée l’atténuerait. A l’échelle régionale, l’exemple des macrophytes suggère que le réchauffement climatique et l’eutrophisation ne menacent pas toujours les mêmes espèces et que la prise en compte des deux perturbations ensemble pourrait augmenter la proportion d’extinction prédites. Au final, les étangs doivent être protégés et des moyens humains devraient être mis en place afin d’aider les organismes à s’adapter aux nouvelles conditions créées par l’eutrophisation et le réchauffement climatique. En plus des stratégies de conservation habituelles (conservation des habitats et des espèces), des stratégies ciblées sur ces deux perturbations seraient particulièrement utiles, telles qu’intensifier le contrôle des sources ponctuelles et diffuses de nutriments ou favoriser les migrations des espèces en augmentant la connectivité. De plus, un des objectifs de la conservation de la biodiversité devrait être de conserver des réseaux d’étangs différant en charge trophique, ainsi qu’en un maximum de paramètres abiotiques afin de maximiser la biodiversité régionale. Mots-clés : mares et étangs, biodiversité, eutrophisation, réchauffement climatique, valeur de conservation, macrophytes, macroinvertébrés, odonates adultes, amphibiens. 5 Abstract Ponds and the networks they collectively support are exceptional ecosystems with a high ecological, social and economic value. They shelter a very diverse and sometimes unique biodiversity, often richer than those found in running waters or larger lakes. Despite this outstanding biodiversity, ponds have been and are threatened by many human activities. Two of the main threats hanging over pond biodiversity are climate warming and eutrophication. Climate warming is obvious worldwide and has many consequences on communities and species. It leads in particular to species geographic distribution shifts leading to colonization and extinction events in a given area or ecosystem and to changes in local and regional richness. An increase in regional species richness has been described in both terrestrial and aquatic systems. The local scale is less known, but up-to-present studies suggest also an increase in species richness. The increase in the nutrient load, eutrophication, is a major threat for freshwater biodiversity. It leads to an increase in primary productivity, algal blooms, and high oxygen depletion. The relationship between productivity and species richness is not well-known in aquatic systems, but studies suggest a humpshaped pattern, with maximal species richness in intermediate conditions and a lower richness in oligotrophic and hypertrophic conditions. The objectives of the thesis are: (1) to ascertain the impact of climate warming on local (pond) and regional biodiversity (country or region), (2) to determine the shape of the relationship between productivity and biodiversity at the local scale and finally (3) to try evaluating the joint impact of climate warming and eutrophication. About one hundred ponds of Switzerland and about one hundred ponds of France (Dombes, Ain) have been studied. The Swiss data set consists of a large range of thermal conditions and nutrient loads. The French data set consists mainly of hypertrophic ponds. The biodiversity of four different biological groups was studied: macrophytes, macroinvertebrates (the whole assemblage, but with special focus on gastropods and water beetles), adult dragonflies and amphibians. Two metrics describing biodiversity were selected: taxonomic richness and conservation value, the latter expressing the threat level of species. At the regional scale (here in Switzerland), this thesis evidenced that, in answer to the warming climate, the proportion of colonization events (“winners”), therefore of new species in the country, is higher than the proportion of extinction events (“losers”), suggesting an increase in regional species richness. Moreover, as the potential threat from climate warming is not reflected by the current Red Lists, conservation management could gain from a complementary label indicating the degree of 6 sensitivity to warming or from an integration of this label among the criteria used to select priority species. Concerning the effects of eutrophication at the regional scale (here the Dombes area), a preliminary analysis suggests that, for nutrient-rich lowland ponds, taxonomic richness will decrease for all taxonomic groups apart from amphibians, but that a marked decrease will occur only for macrophytes. At the local scale, pond species richness will potentially increase due to climate warming. This increase will be particularly high in mountain areas. An increase is predicted for all taxonomic groups (macrophytes, gastropods, water beetles, adult dragonflies and amphibians), but with different magnitudes. The high sensitivity of ponds to climate warming confirms their role as ideal early warning systems, in particular in alpine areas. With regards to eutrophication, responses of the different taxonomic groups differed largely in nutrient-rich lowland ponds: local taxonomic richness of plants and macroinvertebrates decreased in response to eutrophication, whereas richness of gastropods and amphibians showed no significant trend. The conservation of assemblages showed no variation with increasing nutrient load, except for plants where it decreases. These discrepancies between taxonomic richness and conservation value showed that taxonomic richness cannot act as a surrogate for conservation value. Moreover, the absence of relationship between the conservation value and the nutrient load suggests that a same conservation value can potentially occur in a nutrient rich system as in a nutrient poor system. Networks of ponds encompassing different nutrient loads could therefore support a higher regional biodiversity. Concerning the joint impact of climate warming and eutrophication on pond biodiversity, this thesis showed that, at the local scale, eutrophication could hinder the increase in species richness due to climate warming in lowland areas for macrophytes and gastropods only. In high altitude areas, for all taxonomic groups apart from amphibians, a moderate eutrophication will intensify the effects of climate warming, whereas a high eutrophication will hinder it. At the regional scale, the example of macrophytes suggests that climate warming and eutrophication do not always threaten the same species and that considering both perturbations could increase the proportion of species at risk of extinction. Globally, measures should be taken to help organisms adapt to the new conditions produced by eutrophication and climate warming. Besides usual conservation strategies (habitat and species conservation), strategies targeting these two perturbations would be useful, such as for example increasing the control of nutrient point and non-point sources or promoting species migrations by an increase in connectivity. 7 Key words: pond, biodiversity, eutrophication, climate warming, conservation value, macrophytes, macroinvertebrates, adult dragonflies, amphibians 8 Contenu Remerciements ....................................................................................................................................... 1 Résumé.................................................................................................................................................... 3 Abstract ................................................................................................................................................... 6 Contenu ................................................................................................................................................... 9 1. Introduction générale ................................................................................................................... 13 1.1. Notion de mare et d’étang .................................................................................................... 13 1.2. Valeurs écologique, sociale et économique des étangs........................................................ 14 1.3. Abondance des étangs .......................................................................................................... 15 1.4. Notion de biodiversité ........................................................................................................... 15 1.5. Variables environnementales influençant la biodiversité des étangs................................... 16 1.6. Menaces pesant sur les étangs et leur biodiversité .............................................................. 21 1.6.1. Menaces principales ...................................................................................................... 21 1.6.2. Réchauffement climatique ............................................................................................ 22 1.6.3. Eutrophisation ............................................................................................................... 27 1.7. 2. Hypothèses de travail ............................................................................................................ 30 Méthodologie ............................................................................................................................... 33 2.1. Sites d’étude .......................................................................................................................... 33 2.2. Approche méthodologique globale ....................................................................................... 35 2.3. Mesure de la biodiversité ...................................................................................................... 36 2.3.1. Mesure de la richesse taxonomique ..................................................................................... 37 2.3.1.1. Article 1 « The pond biodiversity index “IBEM” : a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 2. Method description and examples of application » .................................................................................................................................. 39 2.3.2. Mesure de la valeur de conservation des peuplements ....................................................... 61 9 2.3.2.1. De la problématique aux hypothèses ............................................................................ 61 2.3.2.2. Article 2 « Comparative assessment of scoring methods to evaluate the conservation value of pond and small lake biodiversity » .................................................................................. 63 2.3.2.3. Synthèse concernant l’évaluation comparative de méthodes mesurant la valeur de conservation de la biodiversité des étangs ................................................................................... 89 3. Impact du réchauffement climatique sur la biodiversité des étangs ......................................... 91 3.1. De la problématique aux hypothèses.................................................................................... 91 3.2. Article 3 « The local diversity of macroinvertebrates in alpine ponds as an indicator of global change: a Gastropod case-study ».......................................................................................... 93 3.3. Article 4 « Warmer and richer? Predicting the impact of climate warming on species richness in small temperate waterbodies » .................................................................................... 101 3.4. Article 5 « Freshwater biodiversity under climate warming pressure: identifying the winners and losers in temperate stagnant waterbodies ». .......................................................................... 125 3.5. Synthèse concernant l’impact du réchauffement climatique sur la biodiversité des étangs ... ............................................................................................................................................. 155 3.6. 4. Discussion complémentaire ................................................................................................ 156 Impact de l’eutrophisation sur la biodiversité des étangs ........................................................ 161 4.1. De la problématique aux hypothèses.................................................................................. 161 4.2. Article 6 « The relationship between nutrient load and biodiversity in lowland ponds and small lakes: implications in a changing climate context »............................................................... 163 5. 4.3. Synthèse concernant l’impact de l’eutrophisation sur la biodiversité des étangs ............. 181 4.4. Discussion complémentaire ................................................................................................ 182 Synthèse et discussion générale ................................................................................................ 185 5.1. Vers une compréhension globale de l’impact combiné du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs ............................................................................... 185 5.1.1. Echelle régionale (diversité γ) ..................................................................................... 185 10 5.1.2. 5.2. Echelle locale (diversité α)........................................................................................... 188 Le changement global : quelles conséquences supplémentaires sur la biodiversité des étangs ? ........................................................................................................................................... 193 5.3. 6. Implications pour la conservation de la biodiversité .......................................................... 195 Conclusions et perspectives ....................................................................................................... 201 6.1. Conclusions.......................................................................................................................... 201 6.2. Perspectives......................................................................................................................... 203 7. Références bibliographiques ...................................................................................................... 205 8. Annexes ....................................................................................................................................... 221 8.1. Article « The pond biodiversity index "IBEM": a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 1. Index development. » .......................................... 223 8.2. Article « Les libellules (Odonates) des étangs piscicoles de la Dombes. ».......................... 235 11 12 Chapitre 1 Introduction générale 1. Introduction générale Cette thèse traite de l’état présent, et de l’évolution potentielle future, de la biodiversité d’écosystèmes aquatiques stagnants, les mares et étangs, qui malgré leur valeur écologique, socioéconomique et esthétique exceptionnelle, sont fortement altérés par les activités humaines. L’objectif principal est de déterminer quels sont et/ou seront les effets de deux perturbations anthropiques majeures, l’eutrophisation et le réchauffement climatique, sur la biodiversité de ces écosystèmes. 1.1. Notion de mare et d’étang Les mares et les étangs sont des écosystèmes aquatiques d’origine naturelle ou anthropique très diverse. Parmi les étangs d’origine naturelle, on peut citer par exemple les dépressions topographiques créées lors du retrait de glaciers (il y a 4’000 à 10'000 ans), les bras morts de plaines alluviales ou les étangs créés par l’action des animaux (castor, sanglier par exemple). Au cours des derniers siècles, un grand nombre de mares et étangs ont également été créés par l’Homme, d’abord pour avoir une source d’eau à proximité puis pour des besoins industriels, agricoles, piscicoles ainsi que pour embellir le paysage. Aujourd’hui, leur création se poursuit à des fins également écologiques ou de loisir (baignade, chasse par exemple). Les origines très diverses des mares et étangs en font des écosystèmes très variés. En conséquence, la majorité des définitions existantes se basent sur des critères de surface et de profondeur, plutôt que sur des différences d’origine. Une première définition basée principalement sur la surface a été développée dans les années 1990 par une ONG anglaise se nommant Pond Conservation : un étang (« pond » en anglais) est un « plan d’eau entre 1 m2 et 2 hectares de surface qui peut être permanent ou temporaire, naturel ou artificiel » (Biggs et al. 2005). De Meester et al. (2005) ont une définition très proche de l’étang (« pond » en anglais) : « plan d’eau peu profond, artificiel ou naturel, de petite taille (1 m2 à environ 5 hectares), qui contient de l’eau en permanence ou temporairement ». Une autre définition, basée principalement sur la profondeur, a été proposée par Oertli et al. (2000). Un étang est « une pièce d’eau dont la profondeur maximale est inférieure à 8 mètres et qui offre la possibilité aux plantes aquatiques supérieures de se développer sur la plus grande partie de la surface des fonds ». Cette faible profondeur (< 8 mètres) empêche la formation de couches thermoclines (couches de transition thermique entre des eaux de surface et des eaux plus profondes de température différente) comme dans les espaces lacustres plus profonds. Une définition de la 13 mare est utilisée par le Pôle Relais « mares, zones humides intérieures, vallées alluviales » de France : « étendue d’eau de formation naturelle ou anthropique, à renouvellement généralement limité, de taille variable et de 5000 m² au maximum, dont la faible profondeur (maximum 2 mètres) permet à toutes les couches d’eau d’être sous l’action du rayonnement solaire, ainsi qu’aux plantes de s’enraciner sur tout le fond ». Plus récemment, le Réseau Européen de Conservation des Mares et Etangs (EPCN) a opté pour une définition des mares et étangs (« ponds » en anglais) basée également sur la profondeur et la surface : « petits plans d’eau de formation naturelle ou anthropique, ayant une surface d’un mètre carré à quelques hectares et une profondeur allant de quelques centimètres à plusieurs mètres, qui sont généralement en eau toute l’année, bien que certains puissent passer par des phases d’assèchement » (EPCN 2011). Au vu des définitions existantes, nous considérerons ici comme mares et étangs : « les masses d’eau douce stagnantes, d’origine naturelle ou anthropique dont la profondeur maximale est inférieure à 8 mètres et la surface maximale est de quelques hectares ». Quelques mares et étangs ayant une surface supérieure à celle définie ici seront néanmoins considérés dans le cadre de cette étude au vu de leur fonctionnement identique à celui des autres mares et étangs étudiés. Dans le texte qui suit, pour faciliter la lecture, nous utiliserons uniquement le terme « étang » (ou « pond » en anglais) pour nommer indifféremment les mares et les étangs. 1.2. Valeurs écologique, sociale et économique des étangs Les étangs sont des habitats indispensables pour de nombreuses espèces rares et en danger. Les réseaux qu’ils constituent collectivement, hébergent les métapopulations de nombreuses espèces d’amphibiens, de plantes aquatiques et d’invertébrés (EPCN 2008). A l’échelle régionale, ils abritent collectivement une biodiversité unique et souvent plus riche que celle présente dans les eaux courantes ou les lacs (e.g. Williams et al. 2004, Angelibert et al. 2006). Les étangs jouent également un rôle essentiel dans l’amélioration de la connectivité entre les habitats d’eau douce en tant qu’écosystèmes-relais ou « stepping-stone » (De Meester et al. 2005, EPCN 2008). Les étangs font partie de notre histoire et de notre culture. Ils jouent un rôle essentiel dans le développement et le maintien de liens entre la population et la nature (EPCN 2008). En tant que zones humides, ils offrent également des sources d’inspiration religieuse ou philosophique (Sajaloli et al. 2002), des opportunités de loisir et ont une valeur esthétique indéniable (Millennium Ecosystem Assessment 2005). De plus, les étangs constituent des supports pédagogiques remarquables (EPCN 2008) et des excellents écosystèmes-modèles pour la recherche scientifique (De Meester et al. 2005). 14 En plus de leur rôle écologique et social, les étangs, comme toute zone humide, jouent également un rôle économique important, pour la production piscicole et comme bassins d’abreuvement du bétail ou d’irrigation (Millennium Ecosystem Assessment 2005). En Europe, ces utilisations sont en déclin, mais les étangs jouent toujours un rôle économique essentiel par le biais des nombreux services écologiques qu’ils fournissent : amélioration de la gestion hydrique dans les bassins versants, atténuation des effets du réchauffement climatique par fixation du carbone et réduction des pollutions diffuses (EPCN 2008). 1.3. Abondance des étangs Malgré la forte valeur écologique, économique et sociale des étangs, beaucoup d’entre eux ont disparu ou ont été dégradés par les activités humaines. Plus de 50% de certains types de zones humides ont été détruits durant le 20ème siècle dans certaines régions d’Amérique du Nord, d’Europe, d’Australie et de Nouvelle-Zélande (Millennium Ecosystem Assessment 2005). En France, une des zones d’étude de cette thèse, 50% des zones humides ont disparu entre 1960 et 1990 (Bernard 1994). En Suisse, région également étudiée dans cette thèse, la situation est tout aussi préoccupante avec une disparition de près de 90% des zones humides depuis 1800 (Imboden 1976). Les étangs restent néanmoins relativement abondants. En ne considérant que les plus grands d’entre eux (1’000 – 10’000 m2), ils sont estimés à 277 millions à travers le monde (d'après Downing et al. 2006). En Suisse par exemple, environ 32’000 étangs compris entre 0,01 et 5 hectares ont été comptabilisés contre 365 lacs (Oertli et al. 2005a). En France, on estime qu’il y a au moins un million d’étangs de moins de 0,5 hectares (EPCN 2008). 1.4. Notion de biodiversité La biodiversité (ou diversité biologique) peut être définie comme la variabilité de la vie à tous les niveaux d’organisation biologique (Gaston and Spicer 2004). Elle se subdivise en trois niveaux principaux : la diversité génétique qui correspond à la variabilité des gènes au sein d’une même espèce ou d’une même population (diversité intra-spécifique), la diversité spécifique qui correspond à la diversité des espèces (diversité inter-spécifique) et la diversité écosystémique qui correspond à la diversité des écosystèmes et biomes présents sur Terre. 15 Dans le cadre de cette thèse, nous entendrons par biodiversité, diversité spécifique. Cette biodiversité peut être mesurée à trois différentes échelles spatiales (Figure 1). La première échelle est celle de l’écosystème (ici l’étang) et est nommée biodiversité α. La deuxième échelle est celle de la variabilité entre écosystèmes ou biodiversité β. Finalement, l’échelle spatiale la plus large est l’échelle régionale (pays ou région) ou biodiversité γ. Figure 1 : Illustration des trois différentes échelles spatiales de la biodiversité spécifique (Source : hepia). 1.5. Variables environnementales influençant la biodiversité des étangs Différentes variables environnementales ont un impact sur la biodiversité des étangs, et ce à différentes échelles spatiales. En effet, les variables environnementales présentes à ces différentes échelles, allant du global à l’habitat, agissent comme des filtres de sélection sur le pool d’espèces global pour définir les peuplements locales (e.g. Gaston and Blackburn 2000, Heino et al. 2009) (Figure 2). Deux échelles spatiales en particulier seront considérées dans cette thèse : l’échelle locale et l’échelle régionale. Par échelle locale, nous entendons ici l’écosystème entier (l’étang), échelle à laquelle se mesure la biodiversité α (Figure 2). Par échelle régionale, nous entendons un pays ou une région, échelle à laquelle se mesurent les biodiversités β et γ (Figure 2). 16 Figure 2 : Modèle schématique des filtres environnementaux influençant les peuplements régionaux et locaux, du bassin versant au microhabitat (redessiné à partir de Heino et al. 2009) et positionnement le long de ce schéma des deux échelles spatiales considérées dans cette thèse (en vert). Les processus abiotiques influençant la biodiversité des étangs sont extrêmement complexes comme permet de le constater la Figure 3. A l’échelle la plus globale, la position à la surface de la Terre, c’està-dire la latitude et l’altitude, a une influence sur la biodiversité des étangs. En effet, la latitude et l’altitude sont parmi les facteurs indirects les plus importants responsables de l’hétérogénéité de la biodiversité sur Terre (Rahbek 1995, Gaston and Spicer 2004). De nombreux facteurs climatiques y sont associés, dont la température de l’air, les précipitations, la radiation et le vent. A chaque facteur climatique s’ajoute sa variabilité et son imprévisibilité : les événements climatiques aléatoires ou extrêmes tels que gels ou sécheresses ont un impact marqué sur la biodiversité d’un étang. 17 Figure 3 : Principaux paramètres abiotiques influençant un étang et sa biodiversité animale et végétale. Les flèches en rouge et en gras correspondent à deux relations majeures étudiées dans le cadre de cette thèse. Certaines flèches ont volontairement été omises, car moins importantes pour la compréhension de ces deux relations majeures. 18 Cette échelle globale influence l’environnement proximal et le bassin versant de l’étang. A ce niveau spatial plus proche de l’étang, des facteurs naturels (pédologie, géologie) et anthropiques (occupation du sol) influencent la qualité de l’eau (Brönmark and Hansson 2000). L’occupation du sol influence trois paramètres de la qualité de l’eau : la teneur en nutriments, la teneur en polluants et la transparence, mais par des mécanismes différents. Par exemple, la présence d’agriculture intensive dans le bassin versant implique (i) l’épandage d’engrais qui ont un impact sur la teneur en nutriments, (ii) l’usage de pesticides qui ont un impact sur la teneur en polluants et (iii) le lessivage des sols nus qui a un impact sur la transparence de l’eau. Le degré de connectivité entre les étangs a également une influence, cette fois directe, sur les peuplements abrités par les étangs. En effet, la distance entre un étang donné et les étangs alentours influence fortement les peuplements présents dans cet étang (Briers and Biggs 2005). L’environnement proximal a une importance particulière pour les peuplements d’un étang, car il fait souvent partie intégrante des habitats nécessaires aux espèces aquatiques (ou amphibiotiques) pour parfaire leur cycle de vie (stades adultes terrestres des amphibiens et de nombreux insectes aquatiques). L’ombrage par exemple a une influence directe sur les peuplements abrités par les étangs. La richesse taxonomique en Odonates adultes est par exemple sensible au degré d’ombrage de l’étang (Angelibert et al. 2010). Ces deux niveaux spatiaux (échelle globale et échelle proximale) ont des impacts sur la biodiversité des étangs soit directs (par exemple la connectivité), soit indirects en modifiant les conditions abiotiques au sein des étangs qui elles auront un impact sur la biodiversité (par exemple les précipitations) (Figure 3). Au sein d’un étang, la diversité d’habitats, qui dépend principalement de la surface et de la profondeur de l’étang, de la nature des rives (naturelle ou modifiée) et de la structure des rives (degré de découpage des rives), a une influence sur la biodiversité d’un étang (Brönmark and Hansson 2000, Oertli and Frossard in prep). Deux des paramètres-clés pour la biodiversité sont la température de l’eau et la concentration en nutriments (Brönmark and Hansson 2000, Bornette and Puijalon 2011). La température de l’eau a de nombreux effets directs sur les organismes d’eau douce. Tout d’abord, elle affecte de nombreux processus physiologiques, comme le métabolisme, la respiration, la photosynthèse et le taux de développement (Macan 1975, Brönmark and Hansson 2000). Par exemple, pour le coléoptère aquatique Dytiscus marginalis, le nombre d’œufs qui éclosent et le nombre de jours de développement nécessaire à leur éclosion varient en fonction de la température de l’eau (Blunck 1914). Pour deux autres espèces de macroinvertébrés, Planaria alpina et Planaria gonocephala, des différences de consommation d’oxygène en fonction de la température de l’eau 19 ont été mises en évidence : Planaria alpina est plus adaptée aux eaux froides car elle consomme moins d’oxygène à basse température alors que Planaria gonocephala est adaptée aux eaux plus chaudes, car elle consomme moins d’oxygène à des températures dépassant les 10°C (Figure 4) (Schlieper 1952). Figure 4 : Consommation d’oxygène à différentes températures de deux espèces de Planaria (tiré de Schlieper 1952). La température a également une influence sur le comportement des individus et leurs rythmes biologiques ; la somme de degrés-jours détermine de nombreux processus (Begon et al. 2006). La date d’éclosion des œufs du gardon, Rutilus rutilus est par exemple dépendante de la somme de degrés-jours (Brönmark and Hansson 2000). La date de pontes des œufs ainsi que le succès de leur développement dépend, chez d’autres poissons comme l’achigan à petite bouche, Micropterus dolomieu, de la température de l’eau (Rawson 1945). L’activité peut dépendre également de la température, comme par exemple le coléoptère, Ilybius montanus : une augmentation de la température induit une augmentation de la fréquence de ses passages à la surface de l’étang et un temps plus court en profondeur (Calosi et al. 2007). De plus, chaque espèce ayant un spectre thermique propre, la température est un déterminant-clé de la distribution géographique des espèces dans le monde (Begon et al. 2006). La température de l’eau a également des effets indirects sur les organismes. En effet, elle a un impact majeur sur les processus physico-chimiques, comme par exemple la teneur en oxygène dissous, ainsi que sur les processus biologiques, comme par exemple le taux de photosynthèse chez les plantes (Brönmark and Hansson 2000, Bornette and Puijalon 2011). La température influence également le régime hydrique de l’étang et a ainsi un impact majeur sur le biotope. Quant aux nutriments, ils sont des éléments-clés pour la vie. En effet, ce sont des constituants élémentaires des cellules, le phosphore pour le stockage de l’information génétique et le 20 métabolisme de la cellule, l’azote pour la fabrication des protéines (Brönmark and Hansson 2000). De plus, les nutriments sont indispensables à la croissance des plantes et ont ainsi un impact majeur sur la composition des peuplements de plantes aquatiques (Bornette and Puijalon 2011), qui ellesmêmes influencent les peuplements animaux. Un exemple d’impact des nutriments sur les peuplements animaux est la succession d’espèces de macroinvertébrés du genre Sigara qui a été observée dans des lacs d’Angleterre de niveau trophique croissant (Macan 1970). La teneur en nutriments a également une influence sur la transparence de l’eau en favorisant la croissance du phytoplancton (Scheffer 2004). L’usage et l’histoire des étangs ont également un impact sur leur biodiversité (Figure 3). L’âge d’un étang influence les peuplements qu’il abrite (Macan 1975, Jeffries 2008). Les pratiques de gestion, comme par exemple la fertilisation et le chaulage pratiqués dans la Dombes (France), vont modifier la physico-chimie de l’eau, ce qui influencera les peuplements occupant un étang. Dans le cadre de cette thèse, deux variables environnementales ayant un impact majeur sur la biodiversité des étangs, seront étudiées plus en détail : la température de l’air et la teneur en nutriments (flèches rouges en gras sur la Figure 3). Ces deux variables environnementales sont au cœur de deux perturbations anthropiques : la température du réchauffement climatique et la teneur en nutriments de l’eutrophisation. 1.6. Menaces pesant sur les étangs et leur biodiversité 1.6.1. Menaces principales Les étangs et les espèces qui leur sont associées sont confrontés à de nombreuses menaces telles que l’intensification agricole, la pollution, le captage excessif de l’eau, le drainage des terres, une gestion inappropriée ou inexistante, l’invasion par les espèces exotiques ou encore le changement climatique (Brönmark and Hansson 2002, EPCN 2007). Toutes ces pressions de par leur action séparée et synergétique vont avoir des conséquences négatives sur la qualité et la quantité d’eau disponible pour la consommation humaine ainsi que sur la biodiversité (Brönmark and Hansson 2002). Les étangs sont néanmoins peu protégés dans le cadre des législations nationales et européennes (EPCN 2007). 21 1.6.2. Réchauffement climatique Effets physiques En raison des activités humaines, la concentration de gaz carbonique (CO2) et d’autres gaz à effet de serre augmente dans l’atmosphère, conduisant à un réchauffement croissant (IPCC 2007a). Le réchauffement climatique est manifeste et observé dans le monde entier ; durant le siècle dernier, la température globale a augmenté de 0.76°C (IPCC 2007a). Pour la même période, la température a augmenté 2,05 fois plus en Suisse, une des zones d’études de cette thèse, que dans l’hémisphère Nord en moyenne (Rebetez and Reinhard 2008). Le réchauffement est également plus important dans les régions montagneuses (les Alpes par exemple) que dans le reste du monde (Beniston et al. 1997, Diaz and Bradley 1997, Noges et al. 2007). Des changements du régime des précipitations ont été observés en maintes régions au cours du siècle dernier (IPCC 2007a). Une diminution de la couverture neigeuse et un recul des glaciers ont également été observés (IPCC 2007a). Les changements climatiques se traduisent aussi par une augmentation des vagues de chaleur, des sécheresses et des fortes précipitations (IPCC 2007a). Une augmentation de la température des lacs et rivières ainsi qu’une plus longue période sans glace pour les lacs ont déjà été observées sous l’effet du réchauffement climatique (Kernan et al. 2010). Au niveau continental, comme régional et local, ces changements climatiques vont continuer dans le futur (IPCC 2007a). Pour la dernière décennie du 21e siècle (2090 à 2099), les scientifiques prédisent une augmentation de la température moyenne globale de 1,8°C pour le scénario B1 (faibles émissions) à 4,0°C pour le scénario A2 (émissions élevées) (IPCC 2007a). L’étude des changements climatiques ayant eu lieu durant la fin du Quaternaire indique que le rythme et l’amplitude des changements prédits pour la fin du siècle sont sans précédent (Jackson and Overpeck 2000). Concernant les précipitations, dans les régions européennes, une augmentation des précipitations est prédite en hiver, mais une diminution est prédite en été (IPCC 2007a). En Suisse, un des sites d’étude de cette thèse, il faut compter avec un réchauffement plus important en été qu’en hiver (C2SM et al. 2011). D’ici 2085, le réchauffement devrait être de 3,2 à 4,8°C selon la région de Suisse et selon la saison (Figure 5). Quant aux précipitations moyennes, il faut compter qu’elles changeront en hiver uniquement dans le Sud de la Suisse, avec une augmentation d’environ 23%, et diminueront en été de 21-28% dans toute la Suisse (Figure 5). Concernant les événements extrêmes, les vagues de froid et périodes de gel seront plus rares en Suisse, alors que les vagues de chaleur et les sécheresses estivales plus fréquentes (OcCC 2008). 22 Figure 5 : Changements attendus des températures moyennes et des précipitations moyennes d’été en Suisse en 2085 par rapport à 1980-2009 selon le scénario A1B (source: C2SM et al. 2011). L’ombrage de l’arrière-plan correspond à la fiabilité de la prédiction ; plus l’ombrage est foncé plus la prédiction est fiable. Bien que la fiabilité des projections de changements climatiques soit élevée, il reste des incertitudes. En effet, parmi les différents scénarios d’émissions de gaz à effet de serre existants (A1B, A1T, A1FI, A2, B1, B2), il n’est pas possible à l’heure actuelle de savoir lequel se réalisera dans le futur. Selon le scénario utilisé, les projections régionales peuvent différer fortement (partie droite de la Figure 6), créant une incertitude relativement élevée. Figure 6 : Moyennes globales multi-modèles du réchauffement en surface (relatif à 1980-1999) pour les scénarios A2, A1B e et B1, indiqués comme le prolongement des simulations du XX siècle (partie de gauche). Les barres grises sur la droite représentent les meilleures estimations (ligne solide à l’intérieur de chaque barre) et l’étendue probable évaluée pour les six scénarios. Projections de températures en surface pour l’horizon 2020-2029 et pour l’horizon 2090-2100 (relatif à 1980-1999) pour les scénarios B1, A1B et A2 (partie de droite) Source : IPCC 2007a. 23 Effets sur la biodiversité Malgré l’incertitude des prévisions concernant les changements physiques à venir, l’évidence scientifique du réchauffement climatique n’est plus à prouver. Par contre, ses effets sur la biodiversité ne sont pas encore totalement compris (Parmesan 1996, Noges et al. 2007). Il ne fait néanmoins aucun doute que le réchauffement climatique a déjà, et aura encore plus dans le futur, de forts impacts sur la biodiversité dans le monde (Hughes 2000, Theurillat and Guisan 2001, Walther et al. 2002, Parmesan 2006, Rosenzweig et al. 2008). De plus, une étude à l’échelle globale a montré que le climat, et donc la température, a un impact majeur sur les changements de la biodiversité dans les régions alpines (Sala et al. 2000) (Figure 7). Figure 7 : Effet relatif de cinq différents paramètres déterminant les changements de la biodiversité pour différents types d’écosystèmes (tiré de Sala et al. 2000). Barres : 1, utilisation du sol ; 2, climat ; 3, déposition d’azote ; 4, interactions biotiques ; 5, CO2 atmosphérique. Les effets du réchauffement climatique sur les espèces et les peuplements peuvent être résumés en 4 catégories principales. Les changements de température couplés avec les changements de concentration en CO2 et de précipitations, ont et vont avoir des conséquences directes sur la physiologie des organismes, comme par exemple le taux métabolique, la photosynthèse, la respiration et la croissance (Hughes 2000, McCarty 2001). Il a été observé par exemple que la taille moyenne des poissons diminue sous l’effet du réchauffement de ces 25 dernières années dans trois rivières françaises (Daufresne and Boet 2007). Les espèces, en particulier celles ayant des temps de génération courts et des taux de croissance élevés, pourraient subir des changements micro-évolutionnaires rapides afin de s’adapter aux conditions changeantes (Hughes 2000, Walther 2010). Le réchauffement climatique provoque et provoquera dans le futur un avancement de la phénologie des espèces marines, terrestres et d’eau douce (McCarty 2001, Walther et al. 2002, Parmesan 2006). Une méta-analyse globale a montré un avancement de 2,3 24 jours/décennie pour 677 espèces réparties sur la Terre entière pendant les 40 dernières années (Parmesan and Yohe 2003). Les Odonates par exemple ont avancé leurs dates de vol aux Pays-Bas (Dingemanse and Kalman 2008) et en Angleterre (Hassall et al. 2007). En Suisse, la floraison des cerisiers est plus précoce de 15-20 jours qu’en 1950 (North et al. 2007). En plus de ces changements de phénologie, le réchauffement climatique permet et permettra dans le futur aux individus de réaliser des générations supplémentaires durant la saison favorable, passant de cycles univoltins à des cycles multivoltins (Musolin 2007). A ces impacts directs du réchauffement climatique, s’ajoutent la problématique de certaines espèces qui risquent d’être « piégées » par leurs réponses évolutives à d’autres espèces dont la phénologie a avancé (Schlaepfer et al. 2002). Par exemple, les liens entre consommateur et ressources pourraient être brisés par des décalages de phénologie (Durant et al. 2007). Des déplacements d’aires de répartition des espèces vers les pôles en latitude et vers les sommets en altitude sous l’effet du changement climatique ont été documentés pour un grand nombre d’espèces terrestres (Parmesan 1996, Parmesan et al. 1999, McCarty 2001, Walther et al. 2002, Parmesan and Yohe 2003, Lenoir et al. 2008, Chen et al. 2011). En ce qui concerne les écosystèmes d’eau douce, les changements de distribution des espèces en lien avec le réchauffement ont été moins étudiés. Une étude a cependant montré que les espèces d’odonates, de coléoptères aquatiques et de poissons d’Angleterre subissent en moyenne des déplacements de leurs aires de distribution aussi importants que les groupes terrestres (Hickling et al. 2006). Ces déplacements d’aires de répartition des espèces sont la conséquence de colonisations (espèces favorisées par le réchauffement climatique ou « gagnants ») et d’extinctions (espèces à risque d’extinction ou « perdants »). Dans une région ou un écosystème donné, si la proportion de colonisations est supérieure à la proportion d’extinctions, la richesse spécifique augmentera. Il est observé actuellement que la richesse spécifique régionale augmente sous l’influence du réchauffement climatique pour les milieux terrestres comme aquatiques (e.g. Iverson and Prasad 2001, Buisson et al. 2008). A l’échelle locale (écosystème entier), les changements de richesse spécifique associés avec le réchauffement climatique sont moins connus. Néanmoins, une augmentation de la richesse spécifique devrait se produire dans les régions alpines, si l’on se base sur la diminution connue de la richesse spécifique en fonction de l’altitude largement décrite (Rahbek 1995). De plus, plusieurs études prédisent une augmentation de la richesse spécifique locale, par exemple dans les forêts montagnardes et subalpines (Kienast et al. 1998). Certaines séries temporelles confirment également cette tendance. Par exemple, la richesse spécifique des plantes 25 terrestres a augmenté durant le siècle dernier dans des placettes situées sur des sommets alpins-nivaux (Grabherr et al. 1994, Vittoz et al. 2009). Pour les écosystèmes d’eau douce, les connaissances actuelles restent éparses et les études existantes décrivent plutôt les changements de composition que de richesse. Néanmoins, la diminution de la richesse spécifique avec l’altitude décrite dans les rivières (Lods-Crozet et al. 2001, Jacobsen 2008) suggère que la richesse spécifique locale d’eau douce augmente également sous l’effet du réchauffement climatique. Derrière ces changements de richesse se cachent des extinctions. Les estimations du nombre d’espèces à risque d’extinction diffèrent fortement entre les régions et entre les espèces : de seulement quelques pourcents (Levinsky et al. 2007) à presque 80% (Thomas et al. 2004) du pool régional. La biodiversité d’eau douce en particulier est extrêmement vulnérable au changement climatique, avec des taux d’extinction similaires ou dépassant ceux des espèces terrestres mieux connues (Heino et al. 2009). Si les taux d’extinction ont été largement étudiés, les espèces à risque ne sont que rarement identifiées individuellement, bien que cela soit essentiel pour mettre en place des plans d’action pour la conservation de la biodiversité. L’étude de critères liés à l’écologie, l’histoire de vie ou la distribution géographique des espèces afin de détecter les espèces les plus menacées par le réchauffement climatique se développe néanmoins (e.g. Ott 2001, Heikkinen et al. 2010, Williams et al. 2010, Graham et al. 2011). Les modèles de distribution des espèces (« species distribution models » en anglais) peuvent également être utiles pour identifier les espèces les plus menacées (e.g. Domisch et al. 2011). La plupart des études concernant l’impact du réchauffement climatique sur la biodiversité se focalisent sur une sélection restreinte de groupes taxonomiques, tels que la végétation terrestre, les papillons ou les oiseaux (Root et al. 2003). Cependant, étant donné que des espèces ayant des temps de génération ou des capacités de dispersion différents pourraient répondre différemment au réchauffement climatique, les réponses des taxons les plus étudiés pourraient ne pas être représentatives des changements de la biodiversité dans sa totalité (Thomas et al. 2004). 26 1.6.3. Eutrophisation L’eutrophisation est provoquée par une augmentation de la charge en nutriments, principalement en phosphore et azote (Carpenter et al. 1998). Il s’agit d’un problème majeur pour la biodiversité des écosystèmes d’eau douce, et des étangs en particulier (Carpenter et al. 1998, Brönmark and Hansson 2002, EPCN 2007). C’est un phénomène naturel qui s’est accéléré dès les années 1950-1960 en Europe sous l’effet d’apports excessifs en nutriments, en phosphore principalement, en provenance des ménages, de l’industrie et de l’agriculture (Carpenter et al. 1998, Brönmark and Hansson 2002, Dodds 2002). Une étude à l’échelle globale a montré que l’utilisation du sol, et donc l’eutrophisation et la quantité de nutriments, est un des principaux paramètres déterminant les changements de biodiversité des lacs et rivières (Figure 7) (Sala et al. 2000). L’eutrophisation cause notamment des blooms algaux de cyanobactéries qui sont problématiques en particulier pour les plans d’eau utilisés pour le loisir ou pour l’eau de boisson (Brönmark and Hansson 2002, Dodds 2002, Smith 2003). L’eutrophisation peut provoquer également une forte désoxygénation qui peut avoir des conséquences négatives importantes notamment pour les poissons (Dodds 2002, Khan and Ansari 2005). L’eutrophisation mène également à une diminution de la transparence qui peut avoir des conséquences majeures sur les macrophytes, allant jusqu’à leur disparition (Brönmark and Hansson 2002, Smith 2003). Si l’on se réfère à la théorie de l’équilibre des lacs peu profonds de Scheffer et al. (1993), l’eutrophisation des milieux lentiques provoque un passage d’un état clair dominé par les macrophytes à un état turbide, dominé par le phytoplancton. Néanmoins, des études récentes suggèrent qu’il n’y a pas un passage d’un état à un autre, mais un gradient de configurations le long du gradient trophique (Scheffer and van Nes 2007). Avant la disparition totale des macrophytes, les peuplements changent, comme observé par exemple dans le lac Léman (Suisse-France) (Lachavanne 1985, Lehmann and Lachavanne 1999). Dans des étangs de France (Dombes), il a été observé que les fortes productivités favorisent les espèces de macrophytes flottantes qui possèdent des organes de réserves et pratiquant une reproduction végétative (Arthaud 2011). Les étangs de plaine sont en général eutrophes et hypertrophes alors que les étangs d’altitude sont en général oligotrophes et mésotrophes, comme par exemple en Suisse qui est un des sites d’étude de cette thèse (Figure 8). En effet, les milieux sont naturellement plus riches en nutriments en plaine du fait des conditions climatiques favorables, mais le changement global, et notamment les pressions humaines croissantes et le réchauffement climatique, s’y ajoute et augmente encore la charge trophique. Dans certains cas, comme par exemple dans la Dombes (France) qui est un des sites 27 d’étude de cette thèse, les étangs de plaine peuvent être particulièrement hypertrophes (Figure 8), sous l’effet principalement d’apports allochtones visant à améliorer leur rendement piscicole. Figure 8 : Proportion d’étangs dans chaque classe trophique selon les classes de phosphore et de nitrogène définies par OCDE (1982) et Wetzel (1983). En haut : étangs suisses de plaine (< 1000m d’altitude, n = 89), au milieu : étangs suisses d’altitude (>1800m d’altitude, n = 19) et en bas : étangs de la Dombes, France (< 350m d’altitude, n = 83). (d'après les données de Oertli et al. 2005b, Indermuehle et al. 2010, Arthaud et al. in prep). Impact de l’eutrophisation sur la richesse et la valeur de conservation L’augmentation de la charge trophique d’un étang entraîne une augmentation de la productivité primaire (notamment du phytoplancton). L’impact de l’augmentation de la productivité primaire sur la biodiversité a été largement démontré pour un grand nombre d’écosystèmes, dont les écosystèmes aquatiques. Néanmoins, la relation productivité primaire-richesse spécifique montre de grandes différences de force et de forme selon l’échelle spatiale, le groupe taxonomique ou le type d’écosystème considéré (Waide et al. 1999, Dodson et al. 2000, Mittelbach et al. 2001). En comparaison avec les écosystèmes terrestres, les études conduites sur les écosystèmes aquatiques sont sous-représentées (Waide et al. 1999) ; les études existantes à l’heure actuelle suggèrent 28 néanmoins que la relation entre productivité et richesse spécifique la plus fréquente soit une courbe en cloche, soit une richesse maximale à des taux de productivité intermédiaires, avec une richesse plus faible à des taux faibles ou forts (Waide et al. 1999, Mittelbach et al. 2001). Concernant les étangs en particulier, cette courbe en cloche a été décrite à l’échelle locale pour différents groupes taxonomiques, comme par exemple les poissons, le phytoplancton, les macrophytes et les invertébrés benthiques (Jeppesen et al. 2000, Chase and Leibold 2002). Comme la plupart des études concernant les écosystèmes aquatiques se sont néanmoins focalisées sur un set restreint de groupes taxonomiques, tels que les organismes planctoniques ou les plantes, et n’ont généralement pas considéré les niveaux élevés dans la chaîne trophique, il n’est pas encore prouvé que la courbe en cloche soit une réponse générale valable pour tous les groupes taxonomiques. Relevons également que, jusqu’à présent, les études se sont surtout concentrées sur la richesse spécifique qui constitue en soi un aspect incomplet de la biodiversité. Aucune étude n’a inclus la valeur de conservation des espèces, qui est pourtant particulièrement importante car elle considère la composition spécifique, et donc le niveau de menace ou d’endémisme des espèces présentes dans un peuplement. Cette question est cruciale car l’eutrophisation menace la biodiversité dans sa totalité (Brönmark and Hansson 2000, Heino et al. 2009). Bien que la valeur de conservation soit sensible aux conditions physico-chimiques de l’eau (e.g. Simaika and Samways 2011), la preuve n’a pas été faite que la richesse spécifique et la valeur de conservation varient dans le même sens lorsque l’eutrophisation augmente. 29 1.7. Hypothèses de travail Etudier l’impact du réchauffement climatique sur la biodiversité des étangs est important, car les écosystèmes d’eau douce y sont particulièrement sensibles. Comme décrit dans les chapitres précédents, des déplacements d’aires de répartition des espèces ont été fréquemment observés pour les espèces terrestres, mais moins pour les espèces d’eau douce. Ces déplacements sont la conséquence de changements de la composition des écosystèmes, c’est-à-dire de colonisations (espèces favorisées par le réchauffement climatique ou « gagnants ») et d’extinctions (espèces à risque d’extinction ou « perdants »). Si les taux d’extinction ont été largement étudiés (changements quantitatifs), les espèces à risque n’ont été que rarement identifiées individuellement (changements qualitatifs). Une augmentation de la richesse spécifique des espèces terrestres comme aquatiques a été observée à l’échelle régionale (pays ou région). Une telle augmentation correspond à une proportion de colonisations supérieure à la proportion d’extinctions. L’échelle locale (écosystème) est moins étudiée que l’échelle régionale. Néanmoins, la diminution de la richesse spécifique locale largement observée avec l’augmentation en altitude dans les milieux terrestres comme aquatiques suggère une augmentation de la richesse spécifique avec le réchauffement climatique. De plus, certaines séries temporelles ont confirmé cette tendance dans les écosystèmes terrestres. Il est également à noter que les réponses des espèces les plus étudiées pourraient ne pas être représentatives des changements de la biodiversité dans sa totalité. En conséquence, il est important d’étudier, dans les étangs, l’impact du réchauffement climatique sur la richesse spécifique ainsi que sur les changements de composition spécifique pour plusieurs groupes taxonomiques. Deux hypothèses seront testées : H1 A l’échelle locale (étang), la richesse spécifique augmente sous l’effet du réchauffement climatique. H2 A l’échelle régionale (pays ou région), la proportion d’espèces à risque d’extinction est plus faible que la proportion d’espèces favorisées par le réchauffement climatique. Si cette hypothèse se vérifie, la richesse spécifique régionale augmentera. 30 Etudier l’impact de l’eutrophisation sur la biodiversité des étangs est important car il s’agit d’un problème majeur pour les écosystèmes d’eau douce. Comme décrit dans les chapitres précédents, il a été montré que l’augmentation de la charge trophique entraîne une augmentation de la productivité primaire. La relation entre productivité primaire et richesse spécifique est largement décrite dans les écosystèmes terrestres et montre de larges différences de force et de forme. Cette relation est moins connue dans les écosystèmes aquatiques. Néanmoins, les études existantes à l’heure actuelle suggèrent que la relation la plus fréquente soit une courbe en cloche. Concernant les étangs en particulier, quelques études ont pu également mettre en évidence cette courbe en cloche, mais il n’est pas prouvé que cette réponse soit valable pour tous les groupes taxonomiques. Toutes ces études se sont concentrées sur la richesse spécifique, mais on ne sait pas si la valeur de conservation des espèces, autre aspect important de la biodiversité prenant en compte le niveau de menace des espèces, suit la même relation de courbe en cloche que la richesse spécifique. En conséquence, il est important d’étudier l’impact de l’eutrophisation sur deux facettes de la biodiversité, la richesse taxonomique et la valeur de conservation, pour différents groupes taxonomiques. Cette question est étudiée uniquement pour les étangs de plaine, riches en nutriments, dont la forte charge trophique les place à la fin de la courbe en cloche. Une troisième hypothèse sera donc testée : H3 A l’échelle locale (étang), en plaine, la biodiversité (richesse taxonomique et valeur de conservation) diminue linéairement sous l’effet d’une augmentation de la charge trophique. Ce travail tentera ensuite d’évaluer l’impact combiné du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs en se basant notamment sur les réponses apportées à ces trois hypothèses. 31 32 Chapitre 2 Méthodologie 2. Méthodologie 2.1. Sites d’étude Le présent travail se base sur deux sites d’étude principaux : l’un en Suisse et l’autre en France (Figure 9 et Figure 10). Le jeu de données suisse consiste en une centaine d’étangs étudiés dans le contexte de différents projets antérieurs à cette thèse ayant notamment pour objectif de mettre au point des instruments standardisés pour l’évaluation des étangs (e.g. Oertli et al. 2000, Indermuehle et al. 2010, Menetrey et al. 2010). Il s’agit d’étangs répartis dans toute la Suisse et couvrant une large gamme d’altitudes et donc de température (moyennes annuelles de l’air de -2.2°C à 12.1°C) ainsi que de niveaux trophiques (oligotrophe à hypertrophe, Figure 8). Leur répartition plus précise dans la Suisse est donnée au Chapitre 3.3. Ils ont une surface médiane de 1240 m2 (6 – 58'000 m2) et une profondeur médiane de 1.3 mètres. Ils sont soit d’origine naturelle, soit artificiels et liés aux activités humaines passées ou actuelles, mais aucun n’est utilisé pour la pisciculture. Le jeu de données français (Dombes, Ain) regroupe une centaine d’étangs étudiés dans le contexte d’un projet qui s’est déroulé pendant cette thèse dont l’objectif était d’évaluer la valeur écologique des étangs piscicoles (e.g. Arthaud et al. in prep). Il s’agit d’étangs répartis dans une région du SudEst de la France à environ 300 mètres d’altitude (Figure 9), qui ont une surface médiane de 91'000 m2 (22'900 – 795'000 m2) et une profondeur médiane de 0.65 mètres. Cette région comprend un réseau très dense d’étangs artificiels créés au Moyen-Âge pour la pisciculture extensive et toujours utilisés pour cette activité. Les étangs étudiés ont un niveau trophique élevé (eutrophe à hypertrophe, Figure 8) qui est lié notamment à leur usage piscicole. Ces étangs sont vidangés quelques mois par année lors de leur pêche. Ils sont également mis en assec, puis cultivés ou mis en jachère pendant une année tous les 4-5 ans afin de permettre une utilisation piscicole durable. Ces périodes d’assec rendent l’approvisionnement en eau vital pour le fonctionnement de ces étangs. A ces deux sites d’études s’ajoute un troisième jeu de données qui a été utilisé pour un approfondissement méthodologique (voir Chapitre 2.2). Il s’agit d’un jeu de données d’Afrique du Sud basé sur une centaine d’étangs dont les données proviennent d’observations privées et de collections de musées (Figure 9 et Figure 10). Ces étangs et réservoirs ont une surface médiane de 640 m2 (4 – 110’000 m2) et une profondeur médiane de 0.8 mètre. Ils varient considérablement en niveau trophique et sont disséminés dans tout le pays à des altitudes allant de 0 à 2300 mètres. Ils sont artificiels et utilisés pour l’irrigation, la pisciculture, la pêche de loisir et comme des points d’eau pour le bétail ou le gibier. 33 Suisse © Samways © Samways Figure 9 : Localisation des étangs étudiés dans les deux sites d’études principaux : A. la Dombes (France) et B. la Suisse, ainsi que dans le site d’étude utilisé pour un approfondissement méthodologique : C. l’Afrique du Sud. Figure 10 : Photos de deux étangs typiques de Suisse : à gauche en plaine (canton de Genève), à droite en altitude (cirque de Macun), d’un étang typique de la Dombes (France) à gauche en eau, à droite en assec et de deux étangs typiques d’Afrique du Sud : à gauche dans la savane du KwaZulu-Natal et à droite dans la Province du Limpopo Nord. 34 2.2. Approche méthodologique globale Les deux questions principales étudiées dans le cadre de cette thèse, impact du réchauffement climatique et impact de l’eutrophisation, sont examinées à l’aide de jeux de données différents et sur des facettes différentes de la biodiversité (Tableau 1 et Tableau 2). Tableau 1 : Jeux de données utilisés pour chaque question traitée dans le cadre de cette thèse. Suisse Approfondissement méthodologique, richesse x Approfondissement méthodologique, valeur de conservation x Réchauffement climatique x x (plaine) Eutrophisation Dombes (France) Afrique du Sud x x x Tableau 2 : Composants de la biodiversité évalués pour chaque question traitée dans le cadre de cette thèse. Richesse taxonomique Approfondissement méthodologique, richesse x Approfondissement méthodologique, valeur de conservation x Réchauffement climatique x Eutrophisation x Valeur de conservation Composition spécifique x x x Afin de répondre à ces deux questions principales, la méthodologie de mesure de deux composants de la biodiversité a été évaluée : la richesse taxonomique et la valeur de conservation. La question de la mesure de la richesse taxonomique a été évaluée à l’aide du jeu de données suisse et la question de la mesure de la valeur de conservation à l’aide des jeux de données de Suisse, de la Dombes et d’Afrique du Sud. La première question, impact du réchauffement climatique sur la biodiversité des étangs, qui sera traitée au Chapitre 3, est étudiée à l’aide du jeu de données suisse. Ce site d’étude couvrant une large gamme de températures, il est un excellent modèle pour répondre à cette question. Deux facettes de la biodiversité, la richesse taxonomique et la composition spécifique, sont utilisées pour répondre à cette question. La deuxième question, impact de l’eutrophisation sur la biodiversité des étangs, qui sera traitée au Chapitre 4, est étudiée à l’aide d’une partie du jeu de données suisse (étangs de plaine uniquement) 35 ainsi qu’à l’aide du jeu de données de la Dombes. Deux facettes de la biodiversité, la richesse taxonomique et la valeur de conservation, sont utilisées pour répondre à cette question. Les éléments de réponse apportés à ces deux questions permettront ensuite avec l’appui de la littérature d’effectuer une tentative de synthèse concernant l’impact combiné de l’eutrophisation et du réchauffement climatique sur la biodiversité des étangs. 2.3. Mesure de la biodiversité La biodiversité spécifique à une échelle spatiale donnée peut être mesurée par différents paramètres. Le plus communément utilisé en écologie est le nombre d’espèces ou richesse spécifique (Magurran 2004). La valeur de conservation considère la composition spécifique, et donc le niveau de menace ou le degré d’endémisme. Cet aspect de la biodiversité est complémentaire à la richesse spécifique et particulièrement important. La Convention mondiale sur la biodiversité de Rio (1992) mentionne justement qu’il faut considérer les écosystèmes ayant une forte richesse, mais également ceux abritant des espèces endémiques ou menacées. Bien que la rareté n’ait pas de valeur en soi, elle permet aux écosystèmes avec leur capacité d’accueil restreinte, d’accueillir un plus grand nombre d’espèces ; la rareté rend la diversité possible et prend ainsi une valeur particulière (Rolston 1994). 36 2.3.1. Mesure de la richesse taxonomique La richesse taxonomique d’un étang peut être évaluée soit de façon exhaustive, soit à l’aide d’un indice standardisé. L’utilisation d’un indice est généralement moins coûteuse et plus reproductible. En Suisse, une méthode d’évaluation de la biodiversité a été développée par le Laboratoire d’Ecologie et de Biologie Aquatique de l’Université de Genève : la méthode PLOCH (Oertli et al. 2005b). Cette méthode se base sur la richesse spécifique de 5 groupes taxonomiques : la végétation aquatique, les gastéropodes, les coléoptères, les libellules adultes et les amphibiens. Elle permet d’obtenir une note standardisée allant de 0 à 1 en comparant les richesses observées avec des richesses prédites dans des conditions de référence. Nous avons adapté cette méthode au sein du Groupe Ecologie et Ingénierie des Systèmes Aquatiques de la Haute école du paysage, de l’ingénierie et de l’architecture, en une méthode moins coûteuse et plus accessible à des gestionnaires non-spécialistes. Cette adaptation, l’Indice Biologique des Etangs et Mares IBEM est décrite en détail dans le chapitre qui suit (Article 1) ainsi que par Angélibert et al. (2010) (Annexe 1). L’IBEM se base sur le même principe et utilise les mêmes groupes taxonomiques que la méthode PLOCH, mais avec une détermination au genre, sauf pour les amphibiens pour lesquels la détermination à l’espèce est maintenue. Les données de richesse pour les étangs suisses utilisées dans le cadre de cette thèse ont été obtenues avec la méthode PLOCH (détermination à l’espèce pour tous les groupes). Pour les étangs de la Dombes, les méthodes PLOCH et IBEM ont été adaptées : détermination à l’espèce de tous les groupes sauf pour les gastéropodes et les coléoptères (genre). 37 38 Article 1 The pond biodiversity index “IBEM”: a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 2. Method description and examples of application. Indermuehle Nicola, Angélibert Sandrine, Rosset Véronique & Oertli Beat This manuscript is published in Limnetica 29 (2010). University of Applied Sciences Western Switzerland, hepia Geneva technology, architecture and landscape, 1254 Jussy-Geneva, Switzerland. 39 40 KEYWORDS Bioassessment, monitoring, small waterbodies, nature conservation, case study, practitioners, macroinvertebrates, aquatic plants, amphibians ABSTRACT Ponds are now widely recognized to contribute significantly to regional freshwater biodiversity. Therefore, tools to easily and rapidly assess biological quality specifically for these aquatic habitats have been increasingly requested by conservation planners and nature managers. In close association with practitioners, we developed such a method for Switzerland; the pond biodiversity index “IBEM”. The IBEM-Index is based on the assessment of the taxonomic richness of 5 groups: aquatic vegetation, Gastropoda, Coleoptera, adult Odonata and Amphibia. No abundance data are necessary and genus level identification is required for all groups except Amphibia (species level). The sampling methodology is a stratified random strategy and allows the use of richness estimators to transform the observed taxonomic richness (Sobs) into true taxonomic richness (Strue). As the IBEM assessment follows the methodology presented in the Water Framework Directive, it is based on the calculation of the ratio of true taxonomic richness (Strue) to reference-based predicted richness (Sref). Each of the five taxonomic groups is assessed separately and the overall biological quality of any given pond (i.e. the IBEM-Index) is the average of the five ratios. This score is later converted into one of five quality classes for each pond: bad (0 to 0.2), poor (>0.2 to 0.4), moderate (>0.4 to 0.6), good (>0.6 to 0.8), and high (>0.8 to 1). In this paper, the implementation of the IBEM-Index is described in detail. The sampling methodologies are developed (for the biodiversity and the environmental variables) as well as the assessment methodology. Finally, two examples are presented in detail, for a “good” quality pond and for a “bad” quality pond. The method implementation also includes a website (http://campus.hesge.ch/ibem) which allows the online calculation of the index, and provides support for both sampling and assessment methodologies to users. The IBEM-Index is a rapid assessment method which gives an overall value of pond biodiversity in terms of taxa richness and can be used, for example, in regional screenings or site monitoring in Switzerland. Moreover, as biodiversity is generally recognized as a good indicator of global ecological quality, the IBEM-Index can also be used to investigate ecosystem quality. 41 INTRODUCTION Pond ecosystems contribute significantly to regional freshwater biodiversity (Nicolet et al. 2004, Oertli et al. 2004, Williams et al. 2004, Angelibert et al. 2006). In the last 15 years, this has consistently been shown in many parts of Europe. As a result, tools to easily and rapidly assess the biological quality of these aquatic habitats have been increasingly requested by conservation planners and nature managers. Method have previously been developed (e.g. Biggs et al. 2000; Boix et al. 2005; Chovanec et al. 2005; Oertli et al. 2005; Menetrey et al. 2008; Solimini et al. 2008), but the characteristics of many of these (e.g. special type of ecosystems, restricted geographical use, high cost) prevent their use by practitioners from Switzerland. To address this, we developed such a method specifically for, and in association with Swiss practitioners: the pond biodiversity index “IBEM”. Throughout the process, a selection of representative end users such as nature conservation managers, consultants, governmental organizations and taxonomic experts were consulted on the theoretical and practical aspect of the method in order to fulfill their requirements. The IBEM-Index is simple, standardized, cheap, adjustable and consistent with the relevant legislative framework (Angélibert et al. 2009). The new method, IBEM, is based on the biodiversity of five taxonomic groups, four of which are identified at genus (aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata), and one at species level (Amphibia). The sampling methodology is a stratified random strategy. The assessment follows the methodology adopted by the European Water Framework Directive, and the ratio of the observed richness to a reference-based predicted richness is converted into one of five quality classes for each pond. The final index is the mean of the five assessment scores. To facilitate the method implementation, a website (http://campus.hesge.ch/ibem) enables the calculation of the index online, and provides support on both sampling and assessment methodologies to users. Full details of the development of IBEM have been presented in another paper (part 1, see Angelibert et al. 2009). In this paper (part 2), we present the implementation of the IBEM-Index, including the sampling (for biodiversity and environmental variables) and assessment methodologies. Finally, two detailed examples are presented, one for a “good” quality pond and one for a “bad” quality pond. RANGE OF APPLICATION (TYPE OF POND – GEOGRAPHICAL AREAS) The IBEM-Index can be used to assess ponds with a surface area of 50 m2 to 60 000 m2, and a mean depth between 30 cm and 900 cm. 42 The method is valid (i) in Switzerland and the close border regions of neighbour countries (i.e. with a 100 km-wide belt), and (ii) for waterbodies situated in the colline or montane altitudinal belt (300 – 1000 m.a.s.l.). For other geographical regions (with different species pools), the sampling strategy can be adopted as it stands or easily adapted. However, a different reference system based on biological data or expert knowledge would have to be developed in order to predict the reference richness (Sref) used in the IBEM-Index assessment. METHOD FOR SAMPLING BIODIVERSITY AND MEASURING ENVIRONMENTAL VARIABLES The IBEM-Index for sampling biodiversity has been specifically adapted (see Angélibert et al. 2009) from the PLOCH assessment method (Oertli et al. 2005). The IBEM-Index is based on the assessment of the taxonomic richness of five groups: aquatic vegetation, Gastropoda, Coleoptera, adult Odonata and Amphibia. The choice of these indicator groups has been largely discussed by Oertli et al. (2005) and supported by further studies (Auderset Joye et al. 2004; Menetrey et al. 2005, 2008; see also Cordoba-Aguilar 2008 for Odonata). For Odonata, the adult stage was selected because identification and sampling are easier and less expensive than for larvae or exuviae. Moreover, even though allochthonous taxa can generate background noise when recording an adult assemblage, this noise can be coped with because its magnitude does not prevent identification of the main trends (Oertli 2008). Presence of adults is also a metric indicating the quality of the environment of a pond (shoreline, helophytic vegetation, buffer area) and has therefore to be considered. No abundance data is required and taxa identification is to genus level for all groups except Amphibia (species level). Exotic taxa are not taken into consideration to evaluate the biodiversity with the IBEM-Index as they are not representative of the autochtonous biodiversity of a pond. The IBEMIndex measures the “quality” (and not the functioning of the ecosystem) and cannot therefore include exotic species. The sampling methodology allows the use of richness estimators (Jackknife-1, Burnham & Overton 1979) to transform the observed taxonomic richness (Sobs) into true taxonomic richness (Strue). Finally, this true richness is compared to the reference richness (Sref) that would be expected for reference conditions. Aquatic vegetation Aquatic plants are sampled once in early July, with square plots (0.5 x 0.5 m) equally distributed along transects which are regularly spaced perpendicular to the longest axis of the pond (see 43 examples in Fig. 4). Areas deeper than 3 m are not sampled. The two square plots located at both ends of each transect must be placed directly against the shoreline, covering only the water (and not the shore). In case of fluctuating water level, shoreline square plots must be placed at the usual pond limit. The position of aquatic or terrestrial plants can help to locate this limit. For example, Mentha aquatica or Veronica beccabunga are usually located at the shoreline but with their stems reaching out of the water. If the pond has a dense reedbed or sedges that are impossible to penetrate, the square plots are located in front of this area, on the open water. The number of sample plots (n) in relation to pond area is calculated as follows: n = 30 – 29.1 * log10(area) + 8.6 * (log10(area))2 (see part 1, Angelibert et al. 2009). This number allows enough data to be gathered for each pond so that observed aquatic plant richness (Sobs) reaches on average 70% of true richness (Strue). In each plot, the presence or absence of aquatic plant genera is recorded, with the help a grapnel or an aquascope as necessary. The depth is recorded in each square plot, and is later used to calculate the mean pond depth (see environmental variables section). Only aquatic plants are recorded and these are defined as one of 254 species listed in the highest humidity class (= 5) by Landolt (1977). This includes true hydrophytes (species submerged or with floating leaves) and many emergent plants. To this ‘aquatic’ species pool were added 22 species listed by Landolt (1977) under humidity class 4: Juncus effusus, Carex canescens, Carex flava, Carex lepidocarpa, Carex nigra, Eleocharis acicularis, Eleocharis quinqueflora, Equisetum palustre, Galium palustre, Agrostis stolonifera, Juncus conglomeratus, Scirpus sylvaticus, Juncus filiformis, Juncus inflexus, Lysimachia nummularia, Lythrum salicaria, Lysimachia vulgaris, Mentha longifolia, Myosotis scorpioides, Ranunculus repens, Rorippa palustris, Juncus articulatus. The Characeae are considered as a single taxon. In the case of “mixed” genera which include both aquatic and non-aquatic species (such as Carex sp.), the genus is considered only if the observed specimen belongs to an aquatic species. Exotic species are not taken into account for the IBEM-Index (for example Elodea nuttallii). In both these instances, a specimen may require identification to species level to be either included or discarded in the results. A table with aquatic genera and species taken into account in the IBEM-Index can be downloaded from the IBEM website (http://campus.hesge.ch/ibem/flore.asp). Gastropoda and Coleoptera Aquatic Gastropoda and aquatic Coleoptera (larvae and adults) are sampled once in early July, with a small-framed hand-net (rectangular frame 14x10 cm, mesh size 0.5 mm). This sampling date was chosen as the best compromise between acceptable cost of the method and satisfactory results. Indeed, the sampling of aquatic invertebrates can be coupled with the sampling of aquatic 44 vegetation. Gastropoda and Coleoptera are present in the pond all along the year (with the exception of a few Coleoptera families). Furthermore, both adults and larvae of Coleoptera are sampled, increasing the chance to sample the Coleoptera taxa. The number of required samples (n) in relation to pond area is calculated as follows: n = 15.5 - 10.5 * log10(area) + 2.7 * (log10(area))2 (see part 1, Angelibert et al. 2009). This number allows enough data to be gathered for each pond so that observed richness (Sobs) reaches on average 90% of true Gastropoda richness (Strue) and 70% of true Coleoptera richness (Strue). Sampling is stratified across the dominant mesohabitats. Sediments and algae (except Characeae) are not sampled because of their low taxonomic richness for the selected taxa. Mesohabitats are divided into two main categories: (i) shoreline aquatic mesohabitats, and (ii) those occurring between the shoreline (excluding the shoreline itself) to a depth of 2 m (deeper zones are not sampled). Only mesohabitats covering more than 1% of the total mesohabitat area are taken into account and only the pond area comprising the mesohabitats listed in Table 1 is considered (this list is also available on the IBEM website, http://campus.hesge.ch/ibem/ coleopteres.asp). Two thirds of the samples are then allocated to the first mesohabitat category and the remaining samples are allocated to the second. The samples are distributed between the mesohabitats in proportion to the coverage of each, with a minimum of one sample per mesohabitat. One unit sample consists of the intensive sweeping of the net through the habitat for 30 seconds. If one mesohabitat is composed of scattered patches, the sampling time (30 s) is divided into shorter periods and distributed between patches (= one composite sample). If the number of mesohabitats is larger than the number of samples, the surveyor groups together the mesohabitats situated in the lowest position in Table 1 (for example: group together mesohabitats 3.2.1. and 3.1. (Table 1)) and then samples each habitat for 15 s (= one composite sample). If there is one sample to distribute and two habitats have the same coverage, the user has to choose the habitat listed in the highest position in Table 1 (for example: hydrophytes (1) are preferred to Helophytes (2); submerged plants (1.1.) are preferred to floating leaves (1.2.); etc). Finally, Gastropoda and Coleoptera are sorted in the field and presence/absence of genera in each sample is recorded in the laboratory. Empty shells of Gastropoda are not sorted. For inexperienced staff, additional sorting in the laboratory is recommended. Identification can be made either in the field or in the lab on preserved material. Exotic species are not taken into account for the IBEMIndex; consequently it can be necessary to identify the species of a given specimen in order to discard an exotic taxon (for example Gyraulus parvus). The list of Gastropoda and Coleoptera genera used for the IBEM-Index is available on the IBEM website (http://campus.hesge.ch/ibem/ coleopteres.asp). 45 Table 1: List of the mesohabitats taken into account for the IBEM-Index sampling method. Two thirds of the samples are allocated to the habitats occurring at the shoreline (land-water interface) (A); one third of the samples are allocated to the habitats occurring between the shoreline and a depth of 2 m (B). Mesohabitats A. Habitats occurring at the shoreline (land-water interface) 1. Small-sized helophytes (Carex sp., Eleocharis sp., …) 2. Roots 3. Bare ground 4. Mineral substrate 5. Accumulations of CPOM (Coarse Particulate Organic Matter) (Leaf litter) 6. Large-sized helophytes (Phragmites sp., Phalaris sp., Typha sp., …) 7. Other B. Habitats occurring between the shoreline and 2 m depth (excluding the land-water interface and the sediments) 1. Hydrophytes 1.1.1.1. Submerged with strongly dissected leaves (Myriophyllum sp., Utricularia sp., Ceratophyllum sp., Ranunculus sp. …) 1.1.1.2. Submerged with thread-like leaves (Potamogeton pusillus, P. pectinatus, Zanichellia palustris) 1.1.2.1. Submerged with large entire leaves (Sagittaria sp., Potamogeton crispus, P. lucens, P. perfoliatus) 1.1.2.2. Submerged with small entire leaves (Elodea sp.) 1.1.3. Characeae 1.2.1. Floating large leaves (Water lilies, Trappa natans, Hydrocharis sp., Potamogeton natans, Polygonum amphibium, …) 1.2.2. Floating small leaves (Lemna sp.) 1.3. Moss 1.4. Other hydrophytes (Menyanthes trifoliate, …) 2. Helophytes 2.1. Reedbed (Glyceria maxima, Phragmites australis, Phalaris sp., Typha sp.) 2.2. Large-sized Scirpus (Scirpus lacustris, …) 2.3. Flooded sedge formations 2.4.1. Alisma sp., Equisetum sp., … 2.4.2. Eleocharis sp., small Scirpus sp., Juncus sp. 2.5. Other helophytes 3. Other habitats 3.1. Leaf litter 3.2.1. Loose mineral substrate (sand, gravel) 3.2.2. Consolidated mineral substrate (rock, stones) 3.3. Other 46 Odonata Adult Odonata are sampled twice; at the end of spring and in mid-summer (Fig. 1). The sampling dates depend on the altitude of the studied pond. Observations are made in plots (10m x 30 m) distributed along one third of the shore length, including all the occurring habitats (Fig. 2). Figure 1. Late-spring (1) and mid-summer (2) sampling periods for adult Odonata in relation to altitude. These periods were identified by means of phenological data on adult Odonata provided by the Swiss Biological Records Centre (number of observations per species, pooled in function of altitude and date). Figure 2: Example of distribution of Odonata plots around a pond with route used by the surveyor. 47 At least 3 plots must be distributed along the shoreline (i.e. ponds with a shoreline length < 270 m are sampled along more than a third of the shoreline). Each plot is sampled for 10 minutes. Sampling day conditions are: (i) air temperature between 20° and 30°C (approximately between 11h30 and 16h00), (ii) sunshine and (iii) no wind. Presence of Odonata genera is recorded in each plot using binoculars. If identification is not possible with binoculars, Odonata can be captured using a butterfly net. Strictly lotic taxa, such as Calopteryx and Cordulegaster, are not recorded. The list of Odonata genera used for the IBEM-Index is available on the IBEM website (http://campus.hesge.ch/ibem/ odonates.asp). Amphibia The field protocol follows the method by Schmidt (2004), used for the red list update in Switzerland. Presence of amphibian species is recorded during four visits (March, April, May and June). Each visit lasts 1 hour. The first visit is made during the night, the other three at dusk. Standardised sampling conditions are mild nights, with no wind or rain. Sampling after a long period of drought must be avoided. The amphibians (adults, subadults, larvae) are surveyed by means of (i) search by flashlight, (ii) identification of calls, and (iii) dip netting. The two species Rana esculenta and R. lessonae are considered as one single taxon (green frog complex). The taxonomic reference list, used for the IBEM-Index, is available on the IBEM website (http://campus.hesge.ch/ibem/amphibiens.asp). Amphibians are a flagship group, often with a central importance for managers. As there is a low number of species, this is the only group where an exhaustive inventory (or nearly so) is possible. Such exhaustive inventory is particularly important for detection of rare species (also often threatened). This is for example the case in Switzerland and the gathered species list, even not useful for the IBEM index, is forwarded to the national managers of the Swiss Amphibian breading sites (the KARCH, Swiss Amphibian and Reptile Conservation Programme). Environmental variables Six environmental variables are measured for the IBEM-Index assessment (see next section): pond surface area (m2), mean depth (cm), shoreline index, pond shade (4 classes), percentage of woodland in a 50 m radius surrounding the pond, and altitude (m.a.s.l.). Methods are summarized in Table 2. 48 Table 2: Methods to measure the 6 environmental variables used for the assessment of a given pond by the IBEM-Index. Variables Units Pond surface area m Calculated using GIS, aerial photography or graph paper Mean depth cm Mean of the depths recorded in each vegetation square plot using a 2 Methods 1 ruler or a handheld depth sounder 2 D L /( 2 * ( * S ) with L = shoreline length (m), S = pond area (m ), Shoreline index D = 3.141 Pond shade Class Vertical projection of the shadow of woody vegetation expressed in four classes: (1) 0%, (2) >0-5%, (3) >5-25%, (4) >25-100% Woodland (within 50 m) % Altitude m 1 Forest coverage in a radius of 50 m around the pond If the pond is deeper than 3 m, additional depth measurements must be carried out. METHOD FOR ASSESSING BIOLOGICAL QUALITY The IBEM assessment follows the methodology presented in the Water Framework Directive, and is based on the calculation of the ratio between true taxonomic richness (Strue) and reference-based predicted richness (Sref). This score is translated into one of five quality classes for each pond: bad (0 to 0.2), poor (>0.2 to 0.4), moderate (>0.4 to 0.6), good (>0.6 to 0.8), and high (>0.8 to 1). Each of the five taxonomic groups is assessed separately and the overall biological quality of any given pond (i.e. the IBEM-Index) is calculated by the average of the five ratios. True taxonomic richness (Strue) To compensate for the bias of a non-exhaustive sampling, observed taxonomic richness (Sobs) is transformed into true taxonomic richness (Strue) by a statistical estimator (Jackknife-1, Burnham & Overton 1979). Strue is calculated for aquatic vegetation, Gastropoda, Coleoptera and Odonata either with specific software (for example EstimateS (Colwell 2005)) or by means of our downloadable Microsoft EXCEL file (“calcul_richesse_Strue”), available at http://campus.hesge.ch/ibem/calcul.asp. 49 The sampling of amphibian species is considered to be exhaustive (or nearly so); therefore the observed Amphibian richness equals Strue. Predicted taxonomic richness (Sref) The predicted taxonomic richness for reference conditions (Sref) is calculated for each taxonomic group using GAM models, based on a subset of 12 predicting variables (see Angelibert et al. 2009 for details). Six of these variables (trophic state, transparency, conductivity, percentage of floatingleaved and submerged vegetation, and fish presence) potentially describe pond degradation; they are therefore used to model reference conditions for each site. Indeed, the reference condition of a taxonomic group of a given pond is simulated by setting these 6 indicators of degradation to their “non-degraded” value, i.e. allowing the highest possible taxonomic richness. The other 6 predictors (surface, mean depth, shoreline development, pond shading, percentage of woodland in a 50 m radius, and altitude) are not sensitive to pond degradation and are therefore set to the fieldmeasured values. A downloadable tool calculates Sref automatically (see next section). Calculating the IBEM-Index The IBEM-Index is calculated by a user-friendly tool, either directly online on the IBEM website (http://campus.hesge.ch/ibem/calcul_de_l_indice/initialisation.asp) or by means of a downloadable Microsoft EXCEL file (“calcul_IBEM_v1.0”), available from the same website. The following elements are required to process the index: (i) true genus richness (Strue) of aquatic vegetation, Gastropoda, Coleoptera and Odonata, (ii) observed species richness of Amphibia, (iii) 6 field-measured environmental variables. The user-friendly tool produces the predicted richness for each taxonomic group (Sref), calculates the ratio Strue/Sref and finally computes the IBEM-Index (see example in Fig. 6). APPLIED EXAMPLES As a demonstration, two ponds were assessed by the IBEM-Index and the whole process described here. The two ponds, ZH0002 and ZG0023, are located in lowland Switzerland (Fig. 3). These ponds are located in Adlikon (canton of Zurich) and Menzingen (canton of Zoug), respectively. Both waterbodies are relatively small (640 m2 for ZH0002 and 1608 m2 for ZG0023). Other physical pond characteristics are presented in Table 3. 50 Figure 3: Geographical location of the ponds ZH0002 and ZG0023 in Switzerland. Table 3: Values of the six environmental variables measured in the two ponds (ZH0002 and ZG0023) and required for the IBEM assessment. Variables Ponds ZH0002 ZG0023 435 720 Surface area (m ) 640 1608 Mean depth (cm) 107 108 Forested surrounding (%) 0 8 Shoreline development 1.29 1.22 Shade (% of the pond shaded) 1 1 Altitude (m a.s.l.) 2 51 Sampling According to the pond surface area and using the mathematical formula presented in the method section, aquatic plants were sampled in 16 and 25 square plots in ZH0002 and ZG0023, respectively. These square plots were equally distributed along transects (Fig. 4). Figure 4: Distribution of square vegetation sampling plots along transects in the two ponds ZH0002 (a) and ZG0023 (b). The mathematical formula presented in the method section was used to calculate the number of samples needed to survey for Gastropoda and Coleoptera: 7 and 10 samples in ZH0002 and ZG0023, respectively. The samples were stratified across the dominant mesohabitats (two mesohabitats in ZH0002 and 3 in ZG0023) (Fig. 5). Two thirds of the samples (5 and 7 respectively) were distributed along the shoreline aquatic habitats. The other third was distributed between the shoreline (excluding the shoreline itself) to a depth of 2 m. 52 Adult Odonata were sampled in 3 plots distributed along the shoreline (Fig. 5). As these two ponds have a shoreline length < 270 m (e.g. 116 m and 124 m for ZH0002 and ZG0023 respectively), they were sampled along more than a third of the shoreline. Amphibian species were recorded as described in the methods section. The six environmental variables required for the assessment by the IBEM-Index were also recorded (Table 3). Figure 5: Example of distribution of the sweep-net samples for Gastropoda and Coleoptera and plots for adult Odonata in the two ponds ZH0002 (a) and ZG0023 (b). 53 Calculation of the IBEM-Index The observed taxonomic richness (Sobs) was transformed into true taxonomic richness (Strue) by the statistical estimator Jackknife-1 (Burnham & Overton 1979) (Table 4). These values of true richness varied between 5 (Amphibians) and 12.3 (Coleoptera) for pond ZH0002, and between 0 (Gastropoda) and 8 (Odonata) for pond ZG0023. A list of the taxa recorded in both ponds is given in Appendix 1. Table 4: Values of the observed taxonomic richness (Sobs) and true taxonomic richness (Strue) for the two ponds ZH0002 and ZG0023. V: aquatic vegetation, G: Gastropoda, C: Coleoptera, O: Odonata, A: Amphibia. Ponds ZH0002 ZG0023 Taxonomic group V G C O A Sobs 10 7 8 10 5 Strue 11 7.9 12.3 12 5 Sobs 3 0 3 8 2 Strue 4 0 5.8 8 2 To calculate the IBEM-Index, we used the Microsoft EXCEL file “calcul_IBEM_v1.0” (Fig. 6). The user entered values in the grey cells (six environmental variables, five observed richness), and the results were automatically produced (Fig. 6, cells Ratio and Quality class). Note that the five taxonomic groups had to be used for a reliable assessment with the IBEM-Index. However, the user can exclude one or more groups (Fig. 6, cells Group retained yes/no) in order to get a rough estimate of the biodiversity value of a pond. ZH0002 has a good overall biological quality (Fig. 6a, IBEM-index = 0.79). In this pond, there was a high diversity of Odonata and Gastropoda, but aquatic vegetation was moderately diverse. ZG0023 has a poor overall biological quality (Fig. 6b, IBEM-Index = 0.28) mainly due to the poor aquatic vegetation, Gastropoda and Coleoptera diversity. 54 Figure 6: Calculation of the IBEM-Index for the two ponds ZH0002 (a) and ZG0023 (b) using the EXCEL file “calcul_IBEM_v1.0” (available in French and translated into English for this example) downloadable at http://campus.hesge.ch/ibem. This calculation can also be done online at: http://campus.hesge.ch/ibem/ calcul_de_l_indice/initialisation.asp. 55 DISCUSSION The IBEM-Index is a rapid assessment method which gives an indication of the value of a pond for biodiversity based on the number of taxa. It enables the identification of taxon-rich pond ecosystems, a task required by the 1992 Convention on Biodiversity. The IBEM method can be used in Switzerland for rapid biodiversity assessment, for example in regional surveys or for site monitoring. It is a reliable indicator of site quality, adapted for the assessment or monitoring of ponds belonging to natural sites of national importance (national inventory of marshes, moorlands, river backwaters, amphibian breeding sites). Besides producing the IBEM-Index, the datasets collected by the IBEM sampling method can later be used to study patterns of taxon richness and similarity between sites. Overall, the IBEM-Index is one of the tools available for nature conservation. For strictly species-related conservation issues, other tools which are also part of the “nature conservation toolbox” should be used, for example exhaustive inventories or red lists. Each tool has its specific objective and should be used appropriately. As biodiversity is generally recognized as a good indicator of global ecological quality, the IBEM-Index can also be used to investigate the question of ecosystem quality, a central objective of the WFD. For example, the IBEM-Index was calculated for 63 Swiss lowland ponds, revealing a high proportion of ponds with poor or moderate biological quality (49%, Fig. 7). Good quality was assigned to 38% of the ponds, and only 13% achieved the High quality class, and none of the assessed ponds were ranked in the lower quality class (i.e. Bad). This highlights that about one pond out of two is actually degraded in terms of biodiversity, and this is likely to reflect global ecological quality. In the UK, the Countryside Survey 2007 shows that only 8% of ponds are currently in good condition and that the biological quality of lowland ponds decreased between 1996 and 2007 (Carey et al. 2008). The main objective of the WFD is to restore the quality of all waterbodies in Europe by 2015. However, in all European countries the implementation of the directive covers only waterbodies with a surface area greater than 50 ha, therefore excluding ponds. Despite this, some European regions are also applying WFD-type evaluation and monitoring programmes to ponds (for example some Spanish states e.g. Catalonia, Aragon). If Switzerland followed the WFD for small waterbodies, according to our results half of Swiss lowland ponds would have to be restored to good quality. Although this is not realistic because of the limited funding available for nature conservation and water quality management, our assessment shows that it is crucial to raise awareness of the importance of the conservation of ponds in Switzerland. Currently, ponds are mainly seen as a habitat for flagship species on the Red List. In the future, they should also be considered as an important element of a global landscape where all freshwater systems should have good ecological quality. Moreover, the consideration of the whole pond network is also very important at the regional scale. Although the IBEM-Index will not give a 56 quality value to a regional richness, the taxa list gathered through the IBEM sampling can be useful to address the question of the pond network richness. For global ecological quality assessments, the IBEM-Index can be combined with metrics recently developed specifically for the assessment of the ecological quality of ponds in Switzerland (see Menetrey et al. 2005, 2008; Sager & Lachavanne 2009). Figure 7: Biological quality of Swiss lowland ponds (n = 63), evaluated by the IBEM-Index. The IBEM-Index is valid in Switzerland and the close border regions of neighbour countries (i.e. with a 100 km-wide belt). In others European regions, the sampling strategy and methodology can nevertheless be used directly. Conversely, the assessment of the biological quality (i.e. the calculation of the IBEM-Index) has to be adapted for each region: a reference condition must be assessed for each pond. The assessment of this reference value (i.e. for good ecological condition) can be done in four different ways (as specified in the WFD): i) using historical data (from a few years ago to paleoecological data) on similar ecosystems (same surface area, depth, altitude, shoreline) relatively naturals (i.e. unimpacted by human activities) at the time of sampling; ii) using current data on similar ecosystems, relatively naturals and located in the same region; iii) by consulting taxonomic experts to define the reference value of richness or iv) through prediction (i.e. using mathematical model of the relationship between diversity and the driving variables). 57 ACKNOWLEDGEMENTS The IBEM-Index was developed with support from: Groupe d’Etude et de Gestion de la GrandeCariçaie (GEG), Fondation des Grangettes, Musée Cantonal de Zoologie de Lausanne, Swiss Amphibian and Reptile Conservation Programme (KARCH), University of Geneva - Laboratoire d’Ecologie et Biologie Aquatique (LEBA), Laboratoire des technologies de l'Information (Haute Ecole de Gestion de Genève), Consulting offices AMaibach Sàrl, Aquabug, Aquarius, GREN, and Natura. The study of the Swiss ponds, which made the development of the IBEM-index possible, was supported by many partners: Swiss Federal Office for the Environment (FOEN), Cantons of Geneva, Jura, Vaud and Lucerne, Research commission of the Swiss National Park and HES-SO // University of Applied Sciences Western Switzerland (RCSO RealTech). Moreover we are grateful for the data provided by the Swiss Biological Records Center (CSCF) and the Swiss Floristic Database (CRSF). Many thanks to the following persons for their various contributions: Céline Antoine, Dominique Auderset Joye, Diana Cambin, Gilles Carron, Emmanuel Castella, Jessica Castella, Michaël de la Harpe, Raphaelle Juge, Jean-Bernard Lachavanne, Anthony Lehmann, Simon Lézat, Nathalie Menetrey, Jane O’Rourke, Patrice Prunier, Corinne Pulfer, Nathalie Rimann, Mirko Saam, Lionel Sager, Emilie Sandoz. 58 APPENDIX 1: Taxa recorded in the two test ponds. +: presence Taxonomic group Genus or species Ponds Aquatic vegetation Alisma sp. + Carex sp. + Ceratophyllum sp. + Juncus sp. + Lemna sp. + Lycopus sp. + Lythrum sp. + Mentha sp. + Phragmites sp. + ZH0002 Potamogeton sp. Gastropoda Coleoptera Typha sp. + Ferrissia sp. + Gyraulus sp. + Hippeutis sp. + Physella sp. + Planorbarius sp. + Planorbis sp. + Radix sp. + Agabus sp. + + + Enochrus sp. + Haliplus sp. + Helochares sp. + Hydrophylus sp. + Hydroporus sp. + Hyphydrus sp. + Ilybius sp. + Noterus sp. + Oulimnius sp. + Aeshna sp. + + Anax sp. + + Coenagrion sp. + + Cordulia sp. + Enallagma sp. + Erythromma sp. + Ischnura sp. + + Libellula sp. + + Pyrrhosoma sp. Amphibia + + Dytiscus sp. Odonata ZG0023 + + Sympecma sp. + Sympetrum sp. + + Bufo bufo + + Hyla arborea + Green frog complex (Rana esculenta and R. lessonae) + Rana temporaria + Triturus alpestris + 59 + 60 2.3.2. Mesure de la valeur de conservation des peuplements 2.3.2.1. De la problématique aux hypothèses La conservation de la biodiversité fait face à des ressources limitées en temps, en fonds et en personnel et a donc besoin d’outils pratiques efficaces pour mesurer la valeur de conservation des écosystèmes. En écologie, la mesure de la biodiversité la plus utilisée est la richesse spécifique. Cependant, les mesures de la biodiversité basées uniquement sur la richesse spécifique ont le désavantage de ne pas prendre en considération la composition spécifique et donc le niveau de menace ou d’endémisme des espèces. Des indices standardisés mesurant la valeur de conservation des écosystèmes aquatiques ont été développés dans le monde entier. Ces indices diffèrent selon les régions et aucun consensus n’existe à l’heure actuelle sur une méthode unique. Il est donc nécessaire d’entreprendre une évaluation comparative mettant en évidence les différences entre types de méthodes afin de permettre aux scientifiques comme aux gestionnaires de choisir le type de méthode le plus adapté à leur situation. En conséquence, l’objectif principal du chapitre suivant (Article 2) est d’évaluer comparativement différents types de méthode se distinguant de par le poids donné à la Liste Rouge et de par leur expression, comme une moyenne par espèce ou pour le peuplement en entier. Cette analyse est réalisée pour des étangs de plaine comme d’altitude en Suisse, en France (Dombes) et en Afrique du Sud. L’hypothèse centrale est que l’évaluation de la valeur de conservation de la biodiversité des étangs diffère selon le type de méthode. Cette hypothèse est testée (i) en analysant les différences entre les valeurs de conservation obtenues avec les différents types de méthodes afin d’identifier d’éventuelles redondances ou complémentarités et (ii) en identifiant le potentiel de chaque type de méthode à donner une information supplémentaire à la richesse spécifique. Il s’agit aussi de déterminer si ces méthodes d’évaluation de la valeur de conservation peuvent être utilisées en plus de la richesse spécifique pour mesurer la qualité du biotope. Un objectif complémentaire est de tester l’applicabilité générale dans le monde de chaque méthode, indépendamment de la zone géographique où elle a été développée. Les conclusions principales de l’Article 2 sont rappelées au chapitre 2.3.2.3. 61 62 Article 2 Comparative assessment of scoring methods to evaluate the conservation value of pond and small lake biodiversity Rosset Véronique1, Simaika John2, Arthaud Florent3,4, Bornette Gudrun4, Vallod Dominique3, Samways Michael 2 & Oertli Beat1 This manuscript has been submitted the 9th December 2011. 1 University of Applied Sciences Western Switzerland, hepia Geneva technology, architecture and landscape, 1254 Jussy-Geneva, Switzerland; 2 Stellenbosch University, Department of Conservation Ecology and Entomology, P Bag X1, Matieland 7602, South Africa; 3 ISARA-Lyon, Lyon, F-69364, France; University of Lyon, Lyon, F‐69003, France ; University Lyon 1, Villeurbanne, F‐69622, France ; ENTPE, Vaulx‐en‐ Velin, F‐69518, France 4 CNRS, UMR5023 “Ecology of natural and anthropised hydrosystems”, Villeurbanne, F-69622, France. 63 64 KEYWORDS Freshwater ecosystems, Red List, species richness, dragonflies, macrophytes, biotope quality ABSTRACT 1. Freshwaters are among the most endangered ecosystems in the world. Practical tools to measure their biodiversity value are needed for their effective conservation. Besides species richness, other aspects of biodiversity, including the threat level of species also need to be considered. Currently, existing scoring methods for assessing the conservation value of freshwaters are very varied, and guidelines to select an appropriate method are lacking. 2. In this paper, we hypothesize that scores to assess the conservation value of a given waterbody can vary markedly according to the type of method used. To test this, we applied four types of scoring methods differing in the weight given to the Red List categories and in the expression of the score, i.e. either using mean per species or the assemblage as a whole, on sets of dragonfly and macrophyte data collected from varied types of small lakes and ponds in three different countries (France, Switzerland and South Africa). 3. The comparison of the different types of methods showed that the type of method used had a marked impact on the assessment of the conservation value of a waterbody. 4. Overall, our results also confirmed that the different types of methods could be applicable in different geographical areas and types of freshwater ecosystems, independently of the original area where the method was developed. 5. Our results illustrated that, besides the species richness assessment commonly used, calculating conservation value as a mean per species is useful because it provides additional information. In contrast, those scores based on the assemblage as a whole showed high redundancy with species richness. Overall, using methods expressed as a mean per species and coupling the Red List with other criteria (e.g. the Dragonfly Biotic Index) gave the best performance. 65 INTRODUCTION Freshwaters are among the most threatened ecosystems worldwide (Ricciardi and Rasmussen 1999, Millennium Ecosystem Assessment 2005). This is particularly true for ponds and small lakes which are threatened by habitat loss, excessive nutrient load, chemical pollution, climate change and invasion by alien species (Brönmark and Hansson 2002, EPCN 2007). Ponds and small lakes are numerous across many landscapes (Downing et al. 2006), and provide important ecological, social and economic services such as wildlife habitat, livestock watering, fish production or recreational activities (EPCN 2007). At the regional scale, they collectively support diverse, and in some cases unique biodiversity, often richer than those in running waters or large lakes (e.g. Williams et al. 2004, Angelibert et al. 2006). The conservation of biodiversity faces limited resources in time, funding and personnel (e.g. Kati et al. 2004) and needs effective, practical tools for measuring the conservation value of sites. In ecology, the most commonly used measure of biodiversity is species richness (Magurran 2004, Fleishman et al. 2006). However, biodiversity measures based on species richness alone have the disadvantage of not taking into account species composition and therefore the level of threat to, or endemism among, the species present in a community. Standardized scoring methods have been developed for the assessment of the conservation value of freshwaters worldwide. These use different combinations of physical and/or biological criteria (Boon and Pringle 2009). The most used biological criterion (see among the selection of representative scoring methods in Table 1) is the IUCN Red List Categories (IUCN 2001), hereafter referred to as “Red List”. Methods to assess the conservation value of the biodiversity, hereafter referred to as “conservation value”, are regionally diverse, with no consensus currently existing on a unified strategy. For example, all the existing scoring methods described in Table 1 have been developed for specific countries and have rarely been tested in other geographical areas. Differences among conservation value scoring methods are generally related to the weight they give to the Red List. Some methods are based exclusively on the Red List. For example the Species Quality Score (SQS), developed by Foster et al. (1989), is based solely on the Red List for water beetles in the United Kingdom. The SQS concept has since been applied to other taxa in a variety of freshwater systems, from lentic to lotic, in the United Kingdom (Painter 1999, Nicolet et al. 2004, Williams et al. 2004), France (Oertli 1995, Godreau et al. 1999), and Switzerland (Oertli et al. 2002). Other methods which couple the Red List with other criteria include, for example, the Community Conservation Index (CCI), the Swedish System Aqua (System Aqua), the Dragonfly Biotic Index (DBI), the Lake Assessment for Conservation system (LACON) and the System for Evaluating Rivers for Conservation 66 (SERCON). The CCI is used in the United Kingdom to assess the conservation value of freshwater invertebrate communities (Chadd and Extence 2004). The System Aqua uses the Red List threat level, naturalness of the catchment and species richness to assess the conservation value of seven freshwater and terrestrial groups (Willen 2009). The DBI, developed for South African freshwater systems, also uses the Red List in combination with other criteria, in this case the geographical extent of species and the sensitivity of species to habitat disturbance (Samways 2008, Simaika and Samways 2009a). The LACON system and the SERCON system, developed in the United Kingdom, also use the Red List in combination with other biological criteria, as well as with physical criteria, such as the naturalness of the flow regime (Boon and Howell 1997, Boon 2000, Duker and Palmer 2009). Finally, some scoring methods do not make use of the Red List at all, and focus rather on the functioning of the focal ecosystems using a combination of geomorphological, hydrological and ecological criteria of their health (e.g. Amoros et al. 2000), or biological indicators as for example, the Index of Centres of Density (ICD) developed in the USA by Angermeier and Winston (1997), which uses the number of source populations in an area to assess the conservation value of fish assemblages. Another method, independent of the Red List, has been developed in France for terrestrial plants, assessing the conservation value on the basis of a combination of local rarity, regional responsibility and habitat vulnerability criteria (Gauthier et al. 2010). Scoring methods to assess the conservation value of freshwater systems differ also in the fact that they are expressed at different levels either (i) the assemblage (‘per assemblage’), or (ii) the species (‘per species’, often a statistical mean of the species belonging to the assemblage, see Table 1). For example the SQS is expressed per assemblage, as it consists of the sum of the threat levels of all species belonging to the assemblage. An adaptation of the SQS, the Species Rarity Index (SRI), is, in contrast, expressed per species, and consists of the sum of the threat levels of all species belonging to the assemblage (i.e. the SQS) averaged by the number of species. These many different types of methods can be confusing for practitioners, and could potentially lead to different management recommendations. Moreover, their interpretation, e.g. in terms of biotope quality, is not evident. Our main aim here is therefore to clarify the differences between the types of methods in order to help practitioners choose the type of method best tailored to their particular situation. 67 Table 1 : Main characteristics of a selection of scoring methods used for the assessment of the conservation value of freshwater biodiversity. Scoring method Biological groups Ecosystem Spatial scale Macrophytes Species Quality Score (SQS) and Macroinvertebrates Standing or Ecosystem other similar methods (as C value) Odonata running waters Amphibia Macrophytes Species Rarity Index (SRI) and Macroinvertebrates Standing or other similar methods (as Csp Ecosystem Odonata running waters value) Amphibia Community Conservation Index Invertebrates (CCI) AQUA Standing or Ecosystem running waters Macrophytes Macroinvertebrates Ecosystem Crayfish & Fish Standing or or Amphibia running waters catchment Birds Mammals United-Kingdom Switzerland - Threat category on Sum of species threat levels the Red List United-Kingdom Switzerland Sum of species threat levels - Threat category on averaged by the number of Per species the Red List species Davies et al. 2007, Nicolet et al. 2004, Oertli et al. 2002 Great-Britain For the threat level, sum of Threat level species threat levels averaged Per species - Species richness by the number of species multiplied by a community score Chadd & Extence 2004 Sweden For the threat level, sum of (i) a Naturalness score depending on the maximal - Threat category on species threat level with (ii) the Per assemblage the Red List weighted sum of the number of - Species richness species from other threat levels Willen 2009 Dragonfly Biotic Index (DBI) Odonata LACON Macrophytes Standing waters Ecosystem United-Kingdom SERCON Macrophytes Macroinvertebrates Running waters Fish Breeding birds Ecosystem United-Kingdom Fish Running waters per per Authors Criteria Standing or Ecosystem running waters Index of Centres of Density (ICD) Expressed species or assemblage Geographic location South Africa Site of an USA ecosystem - Species distribution - Threat category on the Red List - Species sensitivity to habitat change Species richness - Rarity (incl. threat category on the Red List) Naturalness - Population size Species richness - Rarity (incl. threat category on the Red List) Naturalness - Population size - Population density 68 Principle of calculation Per assemblage Foster et al. 1990 then Davies et al. 2007, Nicolet et al. 2004, Oertli et al. 2002, Painter 1999 Sum of the total per species of the three sub-indices averaged Per species by the number of species Samways 2008, Simaika & Samways 2009a None Per assemblage Duker & Palmer 2009 None Per assemblage Boon et al. 1997, 2002 Sum of the ratio of the density of each species at each site to the sum of densities of the Per species species over all sites averaged by the number of species Angermeier & Winston 1997 Our central hypothesis is that the assessment of the conservation value of pond and small lake biodiversity differs markedly depending on the type of method, i.e. according to the weight given to the Red List and whether expressed per species or per assemblage. This hypothesis is tested by (i) analysing the differences between the conservation values given by different types of methods for the same assemblages (macrophytes or dragonflies) in order to identify potential redundancy or complementarity, and (ii) identifying the potential of different types of methods to provide additional information over species richness. In addition, we aim to determine if these conservation value scoring methods could be used in addition to species richness to measure biotope quality. Another objective of this study is to test the applicability of a given method in different geographical areas, independently of the original area where the method was developed, to assess each method’s general applicability across the world. 69 MATERIAL AND METHODS Biodiversity data sets Waterbodies from three different geographical areas were selected, two areas in Europe (France and Switzerland) and one in Africa (South Africa) (Figure 1). In France (Figure 1A), a total of 78 ponds were studied in the context of a project aiming at assessing the ecological value of fish ponds (e.g. Arthaud et al. submitted, Robin et al. submitted). These ponds have a median area of 9580 m2 (1,840-86,500 m2) and a median depth of 0.65 m. They are located in the Dombes region (North-East from Lyon), an area of 1,000 km2 at about 300 m. a.s.l. This region comprises a very dense network of artificial eutrophic to hypertrophic ponds (about one thousand), used for fish farming. In Switzerland (Figure 1B), a total of 90 ponds and small lakes were studied in the context of various projects (e.g. Oertli et al. 2002, Indermuehle et al. 2010, Menetrey et al. 2010). They have a median area of 2270 m2 (6-96,000 m2) and a median depth of 1.15 m. They are oligotrophic to hypertrophic and are scattered throughout Switzerland at elevations ranging from 210 to 2757 m. a.s.l. They are either of natural origin (glacial retreat), or artificial, and linked to past or present human activities (e.g. nature conservation, recreation, gravel or clay extraction, fish production). In South Africa (Figure 1C), a total of 116 ponds and reservoirs were studied. The data consisted of a synthesis of museum and private collections and sightings (from 1901 to present) (see e.g. Simaika and Samways 2009b). These ponds and reservoirs have a median area of 640 m2 (4-110,000 m2) and a median depth of 0.8 m. These waterbodies vary considerably in nutrient level, and are scattered in the Eastern part of the country at elevations ranging from 0 to 2300 m a.s.l. These waterbodies are mostly artificial in origin and used as watering points for game or domestic livestock, or for irrigation, fish farming or recreational fishing. 70 Figure 1: Location of the three study areas in Europe and Africa and location of the studied sites A. in Dombes region from France B. in Switzerland and C. in South Africa. Two biological groups, adult dragonflies (Odonata) and macrophytes, were studied using presence/absence data. The two groups were chosen because data on adult dragonflies were available for all waterbodies in the three countries, and data were available on macrophytes for all waterbodies in Switzerland and for about two thirds of the waterbodies in France (55 out of the 78 ponds). Types of conservation value scoring methods We assessed the conservation value of dragonfly and macrophyte assemblages of small waterbodies (ponds and small lakes) with four different types of methods distinguished by the weight given to the Red List and their expression, whether per species or per assemblage (see Introduction section). 71 Scoring methods based exclusively on the Red List and expressed per assemblage The first type of method is based exclusively on the Red List and expressed per assemblage. Two examples are used here: the C value (Oertli et al. 2002) and the rarity component of the Swedish System Aqua (Willen 2009). C value The C conservation value is an application (Oertli et al. 2002) equivalent to the SQS (Foster et al. 1989, Foster et al. 1992, Painter 1999, Williams et al. 2004, Copp et al. 2010). Species are ranked according to their degree of rarity on the national Red List in geometric progression, successively doubling from 1 (commonest species) to 32 (rarest): species of Least Concern (LC) or Data Deficient (DD) status were given the rank of 1, species of Near Threatened (NT) status 4, species of Vulnerable (VU) status 8, species of Endangered (EN) status 16 and species of Critically Endangered (CR) or Regionally Extinct (RE) status 32. The conservation value per site of the species assemblage (C value) is the sum of the scores of all species present at the site. Aqua method The rarity component of the System Aqua, hereafter named the Aqua method, consists of a weighted value ranging from 0 to 5 based on species’ threat status on the national Red List (Willen 2009). A total score per site is calculated according to the formula below, where the score of the highest ranked category is assigned according to a specific table (available in Willen 2009), X3 corresponds to the number of species within the Red List category DD, X4 to the number of species within the category VU and X5 the number of species within the category NT : ( ) ( ) ( ) Scoring method based exclusively on the Red List and expressed per species: the Csp value The second type of method is based also exclusively on the Red List, but expressed per species. The example of the Csp value is used here. The Csp conservation value is an application (Oertli et al. 2002) of the Species Rarity Index developed in the United Kingdom. As for the C value, species are ranked according to their degree of rarity on the national Red List following the same geometric 72 progression as the C value. The mean conservation value per site per species (Csp value) is the C value divided by the number of species present in the site. Scoring method coupling the Red List with other criteria and expressed per species: the Dragonfly Biotic Index The third type of method couples the Red List with other criteria and is expressed per species. The example of the South African Dragonfly Biotic Index (DBI) (Samways 2008, Simaika and Samways 2009a) is used here. The DBI is a composite index that consists of three sub-indices: species geographical distribution in the investigated area, threat status based on the national and global Red List, and species sensitivity to habitat disturbance (Simaika and Samways 2009a). Disturbance, in the sense of the DBI sub-index, refers here to anthropogenic disturbance, whether direct or indirect, such as for example, habitat degradation by invasion of alien species, cattle trampling, over-abstraction, and agricultural run-off. Each sub-index consists of 4 weightings ranging from 0 to 3, where zero corresponds to the lowest conservation value. The sum of the three sub-indices for any one species is the standard DBI score, which has a defined range from 0 to 9 in South Africa. In Switzerland and France, the DBI was slightly modified. First, the mean of the three sub-indices was used instead of the sum, because of lack of data for one of the three sub-indices for some species. Then, the weightings of the sub-indices were replaced by a continuous gradient ranging from 0 to 1, so that the DBI score ranged from 0 to 1, rather than 0 to 3. To calculate the DBI score per site, the total of all the species’ DBIs is divided by the total number of species. The need for information on species geographical distribution in the investigated area and species sensitivity to habitat disturbance makes the DBI longer to compute in comparison with the methods based exclusively on the Red List. However, once the DBI is calculated, it is, for practical purposes, virtually permanent (Samways 2008). Scoring method independent of the Red List and expressed per species: the nested ranking method The fourth type of method includes those methods that are independent of the Red List, and expressed per species. The example of the French nested ranking system, developed by Gauthier et al. (2010), is used here. 73 The scoring method of Gauthier et al. (2010), the ‘nested ranking method’, is based on three criteria: regional responsibility, local rarity and habitat vulnerability. The first criterion, regional responsibility, corresponds to the extent of the species’ geographical distribution outside the study area, while the second criterion, local rarity, corresponds to the extent of the geographical distribution inside the study area. The third criterion provides information on the likelihood of habitat loss for a given species in the study region. Each criterion consists of five weightings ranging from 1 to 5, where 1 corresponds to the lowest conservation value. We slightly modified this scoring method from weightings to a continuous gradient of values rounded off to the first decimal place, and ranging from 0 to 1, where zero corresponds to the lowest conservation value. The three criteria were combined to classify all n species of a studied area among themselves, from one to n, where species n has the highest conservation value. This final classification depends on a hierarchical approach, where the regional responsibility criterion is a first order, the local rarity a nested criterion of a second order and habitat vulnerability a criterion of a third order. To arrive at a nested ranking per site, we applied the same standardization procedure as in the DBI, in which the total of all species ranks is divided by the total number of species. As with the DBI, the need for information on species geographical distribution inside and outside the investigated area and species habitat vulnerability makes the nested ranking method longer to compute in comparison with the methods based exclusively on the Red List. The nested ranking method does not consider changes in species’ population sizes at a global scale, whereas the other methods indirectly incorporate this criterion through the Red List. Data sources for the calculation of conservation values The conservation values given by all scoring methods were calculated for adult dragonflies of South Africa and for adult dragonflies and macrophytes of France and Switzerland. Due to a lack of information concerning the species’ geographical distribution of macrophytes in Europe, the nested ranking method has not been calculated for macrophytes of Switzerland. All scoring methods, except the nested ranking method, required the national Red List. Red Lists were available for both dragonflies and macrophytes in Switzerland (Gonseth and Monnerat 2002, Moser et al. 2002, Auderset Joye et al. 2010) and for dragonflies in South Africa (Samways 2006). In France, no Red List is currently available for dragonflies and macrophytes, so the Swiss Red Lists were used as a surrogate, because of geographical proximity (see Figure 1A) and species pool similarity (90% of the recorded French macrophyte and dragonfly species are present in Switzerland). 74 The extent of species’ geographical distribution within the study area was required for both DBI and nested ranking method. In Switzerland, it was quantified for both biological groups by the number of grid cells of 20x20 km where a species is currently present (data from the Swiss Biological Records Center (CSCF), and the Swiss Floristic Records Center (CRSF)). In France, it was quantified using the frequency of occurrence of each species among the studied shallow lakes. In South Africa, categories of extent of the distribution in South Africa and Africa as a whole were distinguished according to occurrence records collated for the South African dragonfly database (Samways and Simaika, unpublished detailed database). The nested ranking method requires the quantification of the species geographical distribution beyond the study area, i.e. in Europe for the Swiss data set, in Africa for the South African data set, and in France for the French data set, which consists only of a small portion of the country (see Figure 1A). For the dragonflies of Switzerland, it was quantified based on the presence of species in 500 km x 500 km grid cells in Europe (data from Dijkstra and Lewington 2006). For the macrophytes of Switzerland, it was not quantified because of lack of information. For both dragonflies and macrophytes of France, the geographical distribution outside the study area was quantified by the number of counties of France currently occupied by a species (Grand and Boudot 2006, Le réseau des Botanistes Francophones 2010). For the dragonflies of South Africa, it was quantified by the number of African countries where a species occurred (Dijkstra et al. 2011b). Information about habitat vulnerability or species sensitivity to habitat disturbance was required for both the DBI and the nested ranking method. In France and Switzerland, it was quantified for dragonflies on the basis of the affinity of each species for 20 types of freshwater habitats (Dommanget (1998) adapted by C. Deliry, available in Rosset and Oertli (2011)) and for macrophytes on the basis of the presence of each species in 52 types of aquatic communities (Rodwell 2000). For dragonflies of South Africa, categories of species sensitivity to habitat disturbance were distinguished on the basis of the South African dragonfly database (Samways 2008; Samways and Simaika, unpublished detailed database). Overall, the way of calculating the different sub-indices of each scoring method may vary depending on the data available in a particular study area for each biological group. Such variability does not strongly affect the conclusions of the present paper, because it compares the conservation values given by the different scoring methods for a particular group in a particular study area and not among the biological groups nor among the study areas. 75 Statistical analyses Differences and similarities among the conservation values calculated with the four types of scoring methods, as well as with species richness, were explored through Spearman-rank correlations. Ability of the scoring methods for measuring biotope quality The ability of the scoring methods to measure biotope quality was assessed according to the method used by Barbour et al. (1996), U.S.EPA (1998) and Hering et al. (2006) for evaluating the freshwaters of North America and Europe in the context of the Clean Water Act and of the Water Framework Directive respectively. The principle of this method is to compare the number of sites classified as high/low biotope quality by each type of scoring method to the number of sites classified as high/low biotope quality according to an independent assessment. This method was applied to the Swiss and French data sets. The independent assessment of the quality of 18 of the Swiss ponds was based on seven biological and environmental criteria confirmed by expert opinion (Menetrey et al. 2010). The independent assessment of the quality of 25 of the French ponds was based on expert opinion and on two criteria fitting to the specificities of these ponds, i.e. fish-farming practices and ecosystem equilibrium (Vallod et al. unpublished data). Ponds with conservation values above the 25th percentile of reference high quality sites were classified as a ’high biotope quality’ sites, and ponds with conservation values under the 75th percentile of low quality sites were defined as ’low biotope quality’ sites. 76 RESULTS Comparison of the conservation values indicated by the different types of scoring methods Adult dragonflies Most of the conservation values of adult dragonfly assemblages were significantly correlated among each other (exceptions described below), but at different strengths (minimum: 0.29; maximum: 0.99) (Table 2). Overall, the correlations between the conservation values were moderate (mean Spearman ρ of 0.54). As expected, when considering the scoring methods based exclusively on the Red List (the C value, the Csp value and the Aqua method), the correlations between the scores were all high (0.59 < Spearman ρ < 0. 99) (Table 2). The correlations between the conservation values calculated with the Aqua method and the Csp value were particularly strong in the three study areas (Spearman ρ > 0.92). The scoring method independent of the Red List, the nested ranking method, produced scores that were the most weakly correlated to the others (-0.02 < Spearman ρ < 0.83). In the Swiss and South Africa data sets, the correlations were particularly weak, with half of the correlations nonsignificant. The scoring method coupling the Red List with other criteria, the DBI, produced scores showing intermediate correlations with the other conservation values (0.21 < Spearman ρ < 0.83). In the Swiss data set, the conservation value obtained with this method was not significantly correlated with the C value. Considering the way the conservation values are expressed (by assemblage or by species), the values expressed per species, the Csp value, the DBI and the nested ranking method were, on average, not more correlated among each other than with the other values expressed per assemblage, the C value and the Aqua (mean Spearman ρ of 0.50 versus 0.53). 77 others, sp. RL + others, sp. nested ranking DBI RL, sp. Csp RL, assembl. Aqua Table 2: Correlations (Spearman’s rank) between the conservation values of dragonfly assemblages from Switzerland (upper value), France (middle value) and South Africa (bottom value) indicated by the different types of methods (in bold). Significant correlations: ** p < 0.01, **** p < 0.0001. “RL” corresponds to methods based exclusively on the Red List, “RL + others” to methods coupling the Red List with other criteria, and “others” to methods independent of the Red List. “assembl.” corresponds to methods expressed per assemblage and “sp.” to methods expressed per species. 0.665**** 0.905**** 0.619**** 0.731**** 0.915**** 0.831**** 0.963**** 0.593**** 0.991**** 0.205 0.442**** 0.538**** 0.794**** 0.761**** 0.731**** 0.369**** 0.582**** 0.585**** -0.023 0.187 0.294** 0.723**** 0.616**** 0.513**** 0.444**** 0.832**** -0.073 0.053 0.066 0.305** RL, assembl. C RL, assembl. Aqua RL, sp. Csp RL + others, sp. DBI Macrophytes All the conservation values of macrophyte biodiversity were significantly correlated with each other, but at different strengths (minimum: 0.47; maximum: 0.95) (Table 3). Overall, the correlations between the conservation values were moderate (mean Spearman ρ of 0.75). As expected, the three scoring methods based exclusively on the Red List (the C value, the Csp value and the Aqua method) produced highly correlated conservation values (0.71 < Spearman ρ < 0. 95) in the Swiss data set, but not in the French one (Table 3). In the French data set, the strongest correlation (Spearman ρ = 0.91) occurred between the conservation value indicated by the method coupling the Red List with other criteria (DBI) and the one obtained independently of the Red List (nested ranking method). The correlations between the conservation values based exclusively on the Red List, the C value, the Csp value and the Aqua method, were weaker (0.71 < Spearman ρ < 0.79). 78 The correlations between the conservation values indicated by the three methods expressed per species (the Csp value, the DBI and the nested ranking method) were not, on average, much stronger than those with the two methods expressed per assemblage, i.e. the C value and the Aqua (mean Spearman ρ of 0.71 versus 0.82). RL, assembl. Aqua Table 3: Correlations (Spearman’s rank) between the conservation values of macrophyte assemblages from Switzerland (upper value) and France (bottom value) indicated by the different types of methods (in bold). Significant correlations: **** p < 0.0001. “RL” corresponds to methods based exclusively on the Red List, “RL + others” to methods coupling the Red List with other criteria, and “others” to methods independent of the Red List. “assembl.” corresponds to methods expressed per assemblage and “sp.” to methods expressed per species. 0.949**** others, sp. RL + others, sp. nested ranking DBI RL, sp. Csp 0.769**** 0.714**** 0.865**** 0.799**** 0.714**** 0.470**** 0.575**** 0.708**** 0.773**** 0.641**** 0.839**** NA NA NA 0.789**** 0.703**** 0.805**** RL, assembl. C RL, assembl. Aqua RL, sp. Csp NA 0.913**** RL + others, sp. DBI Synthesis Considering all data sets and both biological groups, the correlations of the conservation values indicated by the three scoring methods based exclusively on the Red List (the C value, the Csp value and the Aqua method) were, on average, higher between each other than with the other types of methods (mean Spearman ρ between 0.73 and 0.89 versus between 0.36 and 0.70). The method independent of the Red List (the nested ranking method) produced conservation values only weakly 79 correlated with all other sets of conservation values (mean Spearman ρ between 0.36 and 0.70). The strongest correlation was with the scores produced by the method coupling the Red List with other criteria (the DBI) (mean Spearman ρ = 0.70). When considering the way the scores are expressed (by assemblage or by species), the scores expressed per species (the Csp value, the DBI and the nested ranking method) were not more correlated between each other than with the other scoring methods expressed per assemblage (the C value and the Aqua) (mean Spearman ρ of 0.56 versus 0.59). Relationship between the different types of conservation values and species richness The different types of conservation values were mostly significantly correlated to species richness in the three study areas, but with different levels of strength (minimum: 0.20; maximum: 0.95) (Table 4). Overall, the strength of the correlations between the values given by the different types of methods and species richness was not particularly strong (mean Spearman ρ of 0.50). Two conservation values expressed per species were not significantly correlated to species richness in the Swiss and South African data set. These were (i) the DBI method which couples the Red List with other criteria and (ii) the nested ranking method, which is independent of the Red List. The conservation values based exclusively on the Red List (the C value, the Csp value and the Aqua method) were correlated to species richness in a similar way to the other types of methods (the DBI and the nested ranking method). The conservation values obtained through the two methods expressed per assemblage (the C value and the Aqua system) showed the highest correlations to species richness for both dragonflies and macrophytes in all study areas (0.86 < Spearman ρ < 0. 95 and 0.26 < Spearman ρ < 0. 75). All other significant correlations between conservation values and species richness were weaker (0.20 < Spearman ρ < 0.66). 80 Table 2: Correlations (Spearman’s rank) between the species richness and the conservation values indicated by the different scoring methods for dragonfly and macrophyte assemblages from France, Switzerland and South Africa. Significant correlations: ** p < 0.01, **** p < 0.0001. “RL” corresponds to methods based exclusively on the Red List, “RL + others” to methods coupling the Red List with other criteria, and “others” to methods independent of the Red List. “assembl.” corresponds to methods expressed per assemblage and “sp.” to methods expressed per species. France Switzerland South Africa dragonflies macrophytes dragonflies macrophytes dragonflies Type of method Example RL, assembl. C 0.863**** 0.949**** 0.938**** 0.871**** 0.880**** RL, assembl. Aqua 0.623**** 0.745**** 0.423**** 0.697**** 0.258** RL, sp. Csp 0.475**** 0.616**** 0.508**** 0.320** 0.201** RL + others, sp. DBI 0.609**** 0.611**** 0.041 0.171 0.139 others, sp. nested ranking 0.588**** 0.660**** -0.080 - -0.131 Ability of the different types of conservation values for measuring biotope quality For both the dragonflies and macrophytes datasets from Switzerland and France, the conservation values indicated by the different scoring methods were able to detect biotope quality in 11-85% of cases (Table 5). Species richness was able to detect biotope quality in a percentage of cases higher than any conservation value, in 53% to 84% of cases (average 71%). There were no differences in ability to measure biotope quality according to the weight given to the Red List. The conservation values based exclusively on the Red List (the C value, the Csp value and the Aqua method) did not perform better or worse than the conservation value coupling the Red List with other criteria (the DBI), or the conservation value independent of the Red List (the nested ranking method). When considering the way the conservation values are expressed (by assemblage or by species), the conservation values expressed per assemblage and previously demonstrated to be highly correlated to species richness (the C value and the system Aqua) were the most sensitive in detecting biotope quality (on average 57% and 55% cases respectively). Table 5: Percentage of sites for which the biotope quality was correctly detected by the different conservation values and by the species richness for dragonflies and macrophytes from France and Switzerland (18 sites for France and 25 sites for Switzerland). “RL” corresponds to methods based exclusively on the Red List, “RL + others” to methods coupling the Red List with other criteria, and “others” to methods independent of the Red List. “assembl.” corresponds to methods expressed per assemblage and “sp.” to methods expressed per species. France Switzerland Type of method Example dragonflies macrophytes dragonflies macrophytes RL, assembl. C 49 60 49 72 RL, assembl. Aqua 45 85 14 77 RL, sp. Csp 30 55 14 56 RL + others, sp. DBI 35 45 32 44 others, sp. nested ranking 50 60 11 - - 53 75 72 84 species richness 81 Methodological differences between the types of methods for assessing conservation value of waterbodies In order to help practitioners (nature conservation managers and environmental consultants) choosing among the different types of methods for assessing the conservation value of waterbodies, there are two important methodological differences. Firstly, two types of methods: i) those that couple the Red List with other criteria (here the DBI), and ii) those that are independent of the Red List (here the nested ranking system), encompass a great range of values, even when no or only few species are threatened on the Red List. Indeed, these two types of methods allow the classification of species of Least Concern (LC), and so a more precise classification of sites with only LC species (see the example of the dragonfly assemblages from South Africa, Figure 2A). Figure 2: Scatter plots A. between scores produced by a conservation value independent of the Red List (here the nested ranking method) and scores produced by a conservation value based exclusively on the Red List (here the Csp value) for the dragonfly assemblages from South Africa, showing the large range of values taken by the nested ranking when no species is Red Listed, i.e. when the Csp value consists of 1, and B. between scores produced by a conservation value expressed per species (here the nested ranking) and species richness from dragonfly assemblages of Switzerland, showing the high variability of the conservation values expressed per species when species richness is low. The three conservation values expressed per species have a high variability when species richness is low (overall less than five species), which produced, in some cases, extremely high values. These extremely high values may overestimate the conservation value of the assemblage (see the example of the dragonfly assemblages from Switzerland, Figure 2B). 82 DISCUSSION Approaches to the assessment of conservation value of ponds The differences in conservation value according to the method used to calculate it (i.e. to the weight given to the Red List and to the expression of the score per assemblage or per species) was investigated by a two-step process. Firstly, the differences between the conservation values given by four types of methods for the same assemblages (macrophytes or dragonflies) were analysed in order to identify potential redundancy or complementarity. Then, the potential of the four types of methods to provide additional information over species richness was identified. Comparison of the different types of scoring methods The weight given to the Red List by the different types of scoring methods had a marked impact on the assessment of the conservation value of a particular pond. The scoring methods based exclusively on the Red List gave strongly correlated conservation values. This situation was to be expected because these methods rely entirely on the same data source, the Red List, to assess the conservation value. The relationship of these scoring methods with other methods, either coupling the Red List with other criteria, or independent of the Red List, was distinctly weaker. The way the conservation values are expressed (per assemblage or per species), in contrast, did not have any impact on the strength of the correlations between conservation values. The conservation values expressed per species were not more highly correlated to each other than to the conservation values expressed per assemblage. Do the different types of scoring methods provide additional information over species richness? The different types of scoring methods provided conservation values more or less correlated with species richness. The weight given to the Red List by the different types of scoring methods did not impact the strength of the correlation with species richness, but the expression of the score either per species or per assemblage did. The scores obtained through methods expressed per assemblage were most strongly correlated with species richness, and this was mostly the case in all study areas, and for both macrophytes and dragonflies. This high correlation can be explained by the fact that, as with methods expressed per assemblage, each species, whether Red Listed or not, increases the conservation value. Therefore, 83 conservation values expressed per assemblage provide very little extra information than does solely species richness. Calculation of this type of conservation value appears to be an unnecessary step in the assessment of the conservation value of waterbodies. In contrast, the conservation values expressed per species do provide additional information over species richness assessments. Indeed, conservation values expressed per species were weakly or not correlated at all to species richness. Such weak correlations have already been demonstrated for the DBI in South African rivers (Simaika and Samways 2011) and for the Csp value in Switzerland (Oertli et al. 2002). These large differences between the conservation value and species richness confirm that measuring the conservation value of a site per species could provide additional information over measurement of species richness, while also revealing perspectives on species composition. In summary, the way the conservation values are expressed (per assemblage or per species) has a marked impact on the strength of the correlation with species richness, and the conservation values expressed per assemblage brought no additional information over species richness. This high redundancy suggests that there is no need to use conservation values expressed per assemblage, but that using conservation values expressed per species is useful in assessing waterbodies. The added usefulness of using scoring methods assessing conservation values over species richness for measuring biotope quality In our study, species richness alone was a better metric than any measure of conservation value for describing biotope quality. Although not as powerful as species richness, conservation value scores expressed per assemblage were also powerful for assessing biotope quality. Evidently, the good performance of methods expressed per assemblage is directly related to their strong correlation with species richness. Our results suggest that conservation values are only weakly related to biotope quality, and therefore, that a biotope of poor quality can, surprisingly, host communities of high value. This may be related to the potential large differences in the autecology of rare species (e.g. contrasting habitat or water requirements). This is why the conservation value is a weak indicator of biotope quality, in contrast to species richness. This means that determination of indigenous species richness is sufficient for measuring biotope quality. This is consistent with the previously demonstrated ecological significance of species richness, which is frequently highly related to abiotic stresses that affect freshwater ecosystems (e.g. Bornette et al. 1998, Bornette et al. 2001, Riis and Sand-Jensen 2001, Hinden et al. 2005). 84 Geographical and ecological limits to the applicability of the different types of scoring methods The different scoring methods of conservation value investigated here were developed in specific countries (Switzerland, France, Sweden, South Africa and UK) and have not previously been tested in other geographical areas. The mechanics of the different types of scoring methods suggests that they can be readily transferred from the specific context of one country to another, as well as from one particular taxonomic group to another; this transferability is confirmed by the present study. The different methods were also tolerant of the particularities and constraints of each geographical species pool. The only limitation to worldwide applicability is the availability of information concerning the regional species pool of a particular area (e.g. Red Lists, geographical distribution, ecological information). The present investigation also showed that the different scoring methods can be easily transferred among the different types of waterbodies (e.g. small lakes, fish ponds, reservoirs). The high transferability of the different methods suggests that the conclusions of this study are valid for freshwater systems other than the ponds and small lakes studied here. Recommendations on the use of a method for assessing the conservation value of a waterbody Among the four types of methods tested, the conservation values expressed per assemblage need not be used, not because of any lack of power, but because of redundancy with species richness. The three remaining types of scoring methods, all the ones expressed per species, produced sets of conservation values (for dragonfly and macrophyte assemblages) that were overall moderately correlated with each other. This suggests that they provide different information about the conservation value of an ecosystem, reinforcing the need for recommendations concerning their use. We therefore propose here, based on our results as well as on the mechanics of the methods, some methodological differences which could help practitioners tailor the choice of a type of conservation value(s) (Table 6). First, the scoring methods based only on the Red List are faster to calculate than the other methods that require additional criteria. Then, two types of conservation values (combining the Red List with other criteria and independent from the Red List) have a larger spread of values, when a large number of species is of LC status. This could have the advantage of classifying more precisely sites that have only LC species, and that have the same conservation value when using methods based exclusively on the Red List. Finally, the conservation values independent of the Red List have the disadvantage of not taking into consideration ongoing changes in species’ population size at the regional scale. It assigns the same conservation value to a species currently rare but stable or in 85 decline, as one currently rare but increasing in abundance (due to climate change for example). All the other types of scoring methods indirectly incorporate this aspect, which is taken into account in the Red List. Table 6: Criteria which could help practitioners in choosing among three different types of conservation value scoring methods. The methods expressed per assemblage have been discarded because of their high redundancy with species richness (see Discussion for more details). “RL” corresponds to methods based exclusively on the Red List, “RL + others” to methods coupling the Red List with other criteria, and “others” to methods independent of the Red List. “assembl.” corresponds to methods expressed per assemblage and “sp.” to methods expressed per species. Example (method tested in the present study) Time to compute RL RL + others others Species Species Species Csp value DBI nested ranking short long long Values spread on a large gradient (even when many species without RL status) no yes yes Calculation using knowledge on the trend in species’ population size yes yes no In conclusion, the methods expressed per species and coupling the Red List with other criteria, for example the DBI, give the best performance. Where there are financial limitations for an assessment, the methods expressed per species and based exclusively on the Red List, even if performing less well, could be used because of their ease of calculation. The selection of a type of method would evidently also change according to the information on species available (i.e. Red List status, geographical distribution, ecology). Where there is a lack of information on species geographical distribution and ecology, the methods expressed per species and based exclusively on the Red List would be the only option. In case of imprecise Red List assessments or of absence of Red List assessments, the methods expressed per species and independent of the Red List would be preferred. 86 CONCLUSIONS The comparative assessment of four types of scoring methods of conservation value on dragonfly and macrophyte assemblages of small lakes and ponds from different countries (France, Switzerland and South Africa) confirmed our hypothesis that the assessment of the conservation value of a given assemblage varies markedly according to the type of scoring method used. Indeed, the weight given to the Red List categories by the different types of scoring methods, and their expression, either per species or per assemblage, had a marked effect on the assessment of the conservation value of a pond. Overall, our results confirmed that the different types of scoring methods of conservation value would generally be applicable in different geographical areas and types of freshwater ecosystems, independently of the original area where the method was developed. Our results indicated that, when evaluating an ecosystem, two indices should be used. First, species richness should be used because, at “equal species interest”, sites with high species richness have a higher priority in terms of conservation than those with low species richness. In addition, species richness gave the best performance for measuring biotope quality. Secondly, a scoring method of conservation value should also be used so that, in cases of equal species richness, priority species can be highlighted. Among the four different types of methods for assessing conservation value, we found that the methods expressed per assemblage do not provide any extra information, and need not be used, because of their high redundancy with species richness. The remaining types of methods, i.e. all those expressed per species, provided additional information over that of species richness. On the basis of methodological differences, we conclude that the methods expressed per species and coupling the Red List with other criteria, such as the DBI, gave the best performance. However, where there are financial limitations for undertaking assessments, or where there is a lack of information available on the focal species (e.g. Red List status, geographical distribution, ecology), it may be necessary to use other methods expressed per species which do not use both Red List and other criteria. 87 ACKNOWLEDGMENTS We thank the numerous people from the University of Geneva (Laboratory of Ecology and Aquatic Biology), the University of Applied Sciences Western Switzerland (hepia, Geneva), ISARA-Lyon (Engineering school in agriculture, alimentation, rural development and environment), the Lyon 1 University (UMR CNRS 5023, laboratory of ecology of natural and anthropized hydrosystems), Stellenbosch University (Department of Conservation Ecology and Entomology) for the fieldwork, the laboratory work and database management. Various supports contributed for gathering the database for the 90 Swiss ponds studied: the Federal Office for the Environment (OFEV), many Swiss Cantons, the University of Geneva (Laboratory of Ecology and Aquatic Biology), the University of Applied Sciences of Western Switzerland (RCSO RealTech), and the Research Commission of the Swiss National Park. Various supports contributed to gathering the database for the 78 French ponds: the owners of the studied ponds, the Rhone-Mediterranean and Corsica Water Agency, the French Department of Ecology, energy, sustainable development and sea (DIVA2 research grant), the PEP Aquaculture Rhône-Alpes, the French Vérots foundation, and the French-Swiss PHC Germaine de Staël grant. Véronique Rosset travels were supported by the Swiss – South African Joint Research Programme, the international relations HES-SO, the Swiss Society of Hydrology and Limnology SGHL, and the grant from the Swiss Academy of Technical Sciences. John Simaika was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) PGS-D3 scholarship. We also would like to thank the anonymous reviewers for their insightful comments which helped to improve the manuscript, P. Nicolet and J. Park for editing the English and D. Leclerc for scanning raw data. 88 2.3.2.3. Synthèse concernant l’évaluation comparative de méthodes mesurant la valeur de conservation de la biodiversité des étangs La conclusion générale apportée par l’article « Comparative assessment of scoring methods to evaluate the conservation value of pond and small lake biodiversity » (Article 2) au sujet de l’utilisation d’indices mesurant la valeur de conservation des peuplements est que : L’évaluation de la valeur de conservation du peuplement d’un étang donné varie fortement selon le poids donné à la Liste Rouge et selon son expression, comme une moyenne par espèce (par espèce) ou pour le peuplement en entier (par peuplement). Les différents types de méthodes produisent donc des scores hétérogènes. Cet article a également permis de montrer que, en plus de l’évaluation habituelle de la richesse spécifique, mesurer la valeur de conservation est utile, mais seulement quand elle est mesurée d’une certaine façon : Mesurer la valeur de conservation en l’exprimant par peuplement n’est pas utile, car redondant avec la mesure de la richesse spécifique. La mesure de la valeur de conservation par espèce apporte au contraire une information supplémentaire à la mesure de la richesse spécifique. Parmi les types de méthodes exprimées par espèce, celles associant la Liste Rouge avec d’autres critères biogéographiques ou écologiques, sont les plus performantes. Cet article démontre également que les différents types d’indices ont une applicabilité géographique très large : ils pourraient être utilisés à travers le monde, indépendamment de la région où ils ont été développés. 89 90 Chapitre 3 Impact du réchauffement climatique sur la biodiversité des étangs 3. Impact du réchauffement climatique sur la biodiversité des étangs 3.1. De la problématique aux hypothèses Comme déjà décrit en Introduction, la biodiversité d’eau douce, et notamment des étangs, est particulièrement sensible au réchauffement climatique. Une augmentation de la richesse spécifique sous l’effet du réchauffement climatique a été décrite à l’échelle régionale dans les milieux terrestres comme aquatiques. L’échelle locale (écosystème) est moins connue que l’échelle régionale. Cependant, la diminution de la richesse locale avec l’augmentation en altitude décrite dans les écosystèmes terrestres et aquatiques suggère une augmentation de la richesse locale avec le réchauffement climatique. La richesse spécifique augmente si la proportion de colonisations (espèces favorisées par le réchauffement climatique ou « gagnants ») est supérieure à celle des extinctions (espèces à risque d’extinction ou « perdants »). Si beaucoup d’études prédisent la proportion d’espèces à risque d’extinction (changements quantitatifs), peu identifient individuellement les espèces à risque (changements qualitatifs). Relevons également que toutes les espèces pourraient ne pas répondre de façon identique au réchauffement climatique. En conséquence, l’objectif principal de ce chapitre est de déterminer l’impact du réchauffement climatique sur la biodiversité des étangs à l’échelle locale (étang) et régionale (pays ou région). Cette analyse est réalisée pour des étangs de plaine et d’altitude de Suisse. Les réponses de 5 groupes taxonomiques différents sont évaluées. La première partie de ce chapitre (Articles 3 et 4) étudie les conséquences du réchauffement climatique à l’échelle locale. L’hypothèse testée est l’hypothèse H1 : à l’échelle locale (étang), la richesse spécifique augmente sous l’effet du réchauffement climatique. Deux sous-hypothèses sont également testées : l’augmentation de la richesse spécifique locale varie (i) en fonction des groupes taxonomiques considérés et (ii) en fonction de l’altitude des étangs. La deuxième partie de ce chapitre (Article 5) étudie les conséquences du réchauffement climatique à l’échelle régionale. L’hypothèse testée est l’hypothèse H2 : à l’échelle régionale (pays ou région), la proportion d’espèces à risque d’extinction est plus faible que la proportion d’espèces favorisées par le réchauffement climatique. Les conclusions principales des Articles 3, 4 et 5 sont rappelées au chapitre 3.5. 91 92 Article 3 The local diversity of macroinvertebrates in alpine ponds as an indicator of global change: a Gastropod case-study. Rosset Véronique, Oertli Beat, Angélibert Sandrine & Indermuehle Nicola This manuscript is published in Verh. Internat. Verein. Limnol. 30 (2008). University of Applied Sciences Western Switzerland, hepia technology, architecture and landscape - CH-1254 Jussy-Geneva, Switzerland. 93 94 KEYWORDS alpine ponds, freshwater snails, global change, predictive model, species richness INTRODUCTION Anthropogenic global change is a potential threat for biodiversity worldwide. Increasing temperatures are responsible for numerous shifts in species distributions and abundances which may lead to species-level extinctions (Thomas et al. 2004). Regional diversity can be particularly threatened as demonstrated in the Alps (Theurillat and Guisan 2001). Furthermore, the local diversity, e.g. ecosystem species richness, is also likely to be significantly impacted. Such consequences at the local scale are insufficiently documented, especially for areas where temperature increases are expected to be important. This is the case for high altitude regions, like the Swiss Alps, where temperature increases are predicted to be particularly high (BRADLEY et al. 2004). Ponds may play a central role in such local assessments not only because of their small size and their simple community structure (De Meester et al. 2005), but also because of their sensitivity to physical environment changes. Aquatic macroinvertebrates are good bioindicators of temperature changes (Hodkinson and Jackson 2005). Among them, the passive dispersers, like Gastropods, are good sentinels of pond biodiversity changes because of their complete dependency on aquatic habitats and their weak colonization ability. In this paper, we evaluated the potential response of the local biodiversity to global change in alpine and subalpine ponds from the Swiss Alps by modelling the changes in gastropod species richness according to seven different global change scenarios. MATERIAL AND METHODS Between 1996 and 2005, 119 ponds were studied in Switzerland. They belong to 4 different altitudinal belts (collinear, montane, subalpine and alpine; from 210 to 2760 m. a.s.l.), and therefore cover a large range of thermal conditions. Gastropods were sampled and their species richness was estimated (Jackknife estimation) using the PLOCH method (Oertli et al. 2005b). Fourteen environmental variables (Table 1) were recorded, based on their potential importance for gastropod diversity and their role on invertebrate richness evidenced in previous studies (Oertli et al. 2002, 95 2005b). Annual mean air temperature was the best variable to characterise pond thermal conditions. It was calculated on the basis of monthly values from 115 climate stations and from a digital terrain model with a 25-m grid (data from the Swiss Federal Institute for Forest, Snow and Landscape Research). Temperatures ranged between -2.2 to 12.1°C (median = 7.9). The relationships between species richness and environmental variables were first explored using Pearson “r” or Spearman “ρ” correlation coefficients. We then used a stepwise linear regression procedure (LR) to model the relationship between estimated gastropod richness and the environmental variables (including annual mean air temperature), by setting the probability to enter to 0.05. The model was used to predict the changes in species richness for two “virtual” alpine and subalpine ponds in response to global change. These two virtual ponds were attributed median values of the selected variables, for their corresponding altitudinal belt: the typical alpine pond had a mean annual air temperature of 0.6, a connectivity of 4, a proportion of forests in the environment of 0 and a water transparency of 60; the typical subalpine pond had respectively values of 3.4, 5, 21 and 60. The temperature increases were based on the IPCC emissions scenarios for 2090-2100 (IPCC 2007a) and the CH2070 scenario for Switzerland in 2070 (OcCC & ProClim– 2007), and ranged between +1.8 °C to +4 °C. Additional analyses using Generalized Additive Models (Oertli et al. 2005b) lead to similar results, so they are not presented here. RESULTS AND DISCUSSION Gastropod richness (Jackknife estimation) in the 119 ponds ranked from 0 to 14.9 with a mean of 3.5 (±0.3) species per pond. Annual mean air temperature best correlated with species richness (r = 0.504, p < 0.0001). Five other variables correlated with the species richness: the proportion of agriculture in the catchment area, conductivity, the proportion of pond area covered by floating vegetation and by submerged vegetation, and the connectivity in a radius of 1000 m (r = 0.357, 0.296, 0.232, 0.220 and 0.188, respectively with p < 0.0001 for all variables). 96 Table 1: Environmental characteristics of the studied ponds (n = 119). Variable Unit Mean Minimum Maximum Median Annual mean air temperature °C 6.42 -2.23 12.13 7.92 - 3.61 0 8.25 4.61 Proportion of agriculture in the catchment area % 29 0 100 15 Proportion of forest in the surroundings (at 50 m) % 34 0 100 28 Proportion of pond area covered by floating vegetation % 29 0 100 16 % 34 0 100 22 class - 1 4 3 Water transparency in summer (Snellen tube) cm 42 1 60 50 Conductivity in winter µS/cm 388.6 2.9 1367 397 class - 1 2 2 class - 1 2 2 class - 1 5 4 class - 1 2 1 Mean depth cm 166 26 910 110 Log10(area) - 3.46 0 9.09 3.27 Connectivity (in a radius of 1000 m) a Proportion of pond area covered by submerged vegetation Trophic class based on P, N and conductivity Fish c Extent of shade cover by trees Age b d d Shoreline development e a Measure of pond isolation. This measure takes into account the number and size of ponds within a radius of 1000m and their distance from the studied pond. Small values indicate high degree of isolation. b Maximum class of total P, total N and Conductivity classes according to Wetzel, 1983 (between 1 and 5). Large values indicate high degree of trophy. c Measure the presence of fish: 1 = presence, 2 = probably absent. d Classes (between 1 and 6 for the shaded proportion and between 1 and 5 for the age). Large values indicate high proportion shaded or old ponds. e Ratio of the length of the shoreline to the circumference of a circle area equal to that of the pond (Wetzel, 1983) (1 < 1.40, 2 >= 1.40) 97 The model predicting species richness included four variables (annual mean air temperature, 1000-m connectivity, proportion of forest in the environment and water transparency), and explained 38.4 % of the variability (Figure). Temperature was the most important variable (standardised contribution coefficient = 7.9). There was a significant correlation between observed and predicted richness (r = 0.620 with p < 0.0001, slope = 1). This model allowed the testing of seven scenarios of annual mean air temperature increase for an alpine or a subalpine typical pond (Figure 1). All scenarios clearly lead to an increase in gastropod richness, by 1 to 3 taxa. The predicted changes would involve a shift towards a higher biodiversity, and suggest that species-poor alpine ponds would evolve toward richer subalpine systems. In our case, the species richness of alpine ponds would rise from 1 to about 3 species, while the richness of subalpine ponds would increase from 2.6 to about 5 species. Figure 1: Potential change in the gastropod species richness (S) for an alpine or a subalpine pond as predicted by the stepwise LR equation according to seven climate change scenarios (six IPCC emission scenarios and the CH2070 scenario). “T°” is the annual mean air temperature, “conn” is 1000-meter connectivity, “forest” is forested environment and “trans” is water transparency. The two virtual ponds are typical for subalpine or alpine altitudinal belts. For the alpine one, “T°” is 0.6, “conn” is 4, “forest” is 0 and “trans” is 60, whereas for the subalpine “T°” is 3.4, “conn” is 5, “forest” is 21 and “trans” is 60. 98 These predicted increases in the local richness of alpine/subalpine ponds will be partly due to the upward expansion of the geographical range of those species for which low temperatures act as a limiting factor. Therefore, colonisation events from species present at lower altitudes would be a rule. In addition, the extinction of cold stenothermal species might also happen (Oertli et al. 2008), although this is unlikely in our study sites, since none of these gastropods are cold stenothermal taxa. The richness predicted in this study are potential values. For passive dispersers with slow dispersion rates, like freshwater gastropods (Brönmark 1985), the time necessary to colonise upward habitats can be long. Moreover, further research is needed to understand better the consequences of global change on additional variables important to subalpine/alpine pond diversity, such as the hydraulic functioning, the duration of snow cover and natural hazards. ACKNOWLEDGMENTS We thank the numerous people from the Laboratory of Ecology and Aquatic Biology (LEBA, University of Geneva) and the University of Applied Sciences of Western Switzerland (EIL, Geneva) for the fieldwork, the laboratory-work and the data base management. We also thank the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL) for providing temperature data. 99 100 Article 4 Warmer and richer? Predicting the impact of climate warming on species richness in small temperate waterbodies. Rosset Véronique1, Lehmann Anthony2, Oertli Beat1 This manuscript is published in Global Change Biology 16 (2010). 1 University of Applied Sciences Western Switzerland, hepia technology, architecture and landscape - CH-1254 Jussy-Geneva, Switzerland; 2 Institute for Environmental Sciences, University of Geneva, - CH-1227 Carouge, Switzerland. 101 102 KEYWORDS ponds, biodiversity, climate warming, predictive model, aquatic macrophytes, macroinvertebrates, Coleoptera, Gastropoda, amphibians, Odonata. ABSTRACT Climate change is expected to affect communities worldwide. Many studies focus on responses at the regional level and show an increase in species richness. However, less is known about the consequences of climate change at the local scale (in ecosystems). Small waterbodies, such as ponds, could play an important role for the assessment of the impact of future changes in climate at the local level. We evaluated here the potential changes due to climate warming in the species richness for various groups (plants, snails, beetles, dragonflies, amphibians) across 113 lowland and high altitude ponds in Switzerland. We modelled the relationships between species richness and environmental variables (including temperature) and predicted species richness changes for the end of the century (20902100; using the A2 IPCC scenario). Temperature rise could significantly increase pond species richness. For the five taxonomic groups pooled, species richness would potentially increase from 41 to 75 (+83%) in lowland ponds. In presently species-poor high altitude ponds, the potential increase would be particularly marked, with a proportional increase (+150%; from 14 to 35 species) almost double that in lowland areas. A strong increase in species richness also resulted from models including changes in additional variables, such as landuse or water quality. Future reductions in water quality (e.g. increase in nutrients) may limit the predicted increase in lowland species richness or, conversely, result in a greater increase in species richness in high altitude areas. Nutrient enrichment is shown to affect the taxonomic groups differentially, with plant species richness the most negatively influenced. Climate warming could therefore affect species richness of temperate ponds not only regionally, but also at the local, within ecosystems - scale; species richness could increase markedly in temperate regions, and especially so at higher altitude. 103 INTRODUCTION Global average temperature has increased by approximately 0.74°C since the early 20th century (IPCC 2007a) and is expected to continue rising. Some parts of the world are more impacted than the average. For example, temperature increases are predicted to be particularly high in mountain regions (Nogues-Bravo et al. 2007). Climate change is expected to affect biological systems worldwide (Rosenzweig et al. 2008). Most research is involved in the description and prediction of changes affecting species or taxonomic groups (Walther et al. 2002; Parmesan & Yohe 2003; Parmesan 2006) and describe, for example, species distribution shifts upwards in elevation and northwards in latitude (Hickling et al. 2006; Lenoir et al. 2008), or changes in phenology and voltinism (Menzel et al. 2006; Morin et al. 2009). All these processes have an important impact on species composition at the regional or local scale. Changes in composition of species assemblages have often been described at the regional level (biogeographic regions or political entities) or at the local scale (e.g. site or ecosystem diversity). Such changes have been detected in Europe and North America for various taxa like diatoms (Ruhland et al. 2008), invertebrates (Burgmer et al. 2007), birds (Lemoine & Bohning-Gaese 2003), and fish (Fodrie et al. 2009). For species richness, values at the regional scale have been shown to increase under the influence of climate warming in Europe and North America (e.g. Iverson & Prasad 2001; Daufresne & Boet 2007; Buisson et al. 2008). It is also well-known and well-described in almost every ecology textbook, that terrestrial and freshwater species richness tends to be lower in colder areas, i.e. at higher altitude or latitude (e.g. Gaston & Spicer 2004; Nagy & Grabherr 2009). This trend has also been well-described at the local scale for plants, invertebrates and vertebrates (review in Rahbek 1995). In contrast, the resultant changes in patterns of species richness associated specifically with climate warming have seldom been investigated, especially at the ecosystem level. Among the few existing studies, Henderson (2007) and Hiddink & ter Hofstede (2008), using time series, report an increase in fish species richness in marine ecosystems in response to climate warming. The local scale has generally been more often investigated through plots rather than for the whole ecosystem. For example, longterm monitoring of vegetation plots in terrestrial environments indicates an increase in local species richness (Pauli et al. 2007; Vittoz et al. 2009). For freshwater ecosystems, whose biodiversity is highly endangered (Dudgeon et al. 2006), knowledge remains patchy about future trends under climate change (Heino et al. 2009). Freshwater studies are scarcer than those covering the terrestrial environment and existing work tends to describe species composition of assemblages rather than local species richness. However, the 104 elevational gradient of local biodiversity (a decline of species richness with altitude) has been described by a number of studies (e.g. Lods-Crozet et al. 2001; Jacobsen 2008). Understanding the effects of climate change at the local scale is currently one of the main challenges of ecology, particularly in freshwater ecosystems. Ponds are defined here as ‘waterbodies between 1 m2 and 2 ha in surface area, which may be permanent or seasonal, and include both man-made and natural waterbodies’ (Biggs et al. 2005). Ponds are extremely numerous worldwide, and taking into account only the largest ponds (1,000 to 10,000 m2) there are an estimated 277 million across the globe (from Downing et al. 2006). Furthermore, at the regional scale, they collectively support a very diverse, and in some cases unique biodiversity, often richer than those found in running waters or lakes (e.g. Williams et al. 2004; Angélibert et al. 2006). Ponds are therefore central to the conservation and management of freshwater biodiversity (Oertli et al. 2009). Ponds have numerous characteristics which enable them to play an essential role in the assessment and monitoring of the impacts of climate changes at the local level. Firstly, they are small and their community structure is relatively simple (De Meester et al. 2005). Secondly, freshwater systems respond strongly to physical environmental changes, such as climate change (Heino et al. 2009). Thirdly, as ponds collectively host a large part of regional freshwater species richness (e.g. Williams et al. 2004), their monitoring in relation to anthropogenic pressure (such as climate change) is of critical importance. Finally, it is also evident that because ponds have relatively well-defined boundaries, their species richness at the ecosystem level is easier to measure than for systems with less well-defined limits. Most studies of the effects of climate change on biodiversity have focused on a restricted selection of taxonomic groups, such as terrestrial vegetation, butterflies, or birds (Root et al. 2003), making these taxa of particular interest as bioindicators. However, as species with different generation times or dispersal capacities might show different responses to changes in climate, responses of well-studied taxa could be non-representative of biodiversity changes as a whole (Thomas et al. 2004). It is therefore important to conduct investigations across a range of taxonomic groups to ascertain the impact of climate change on species richness. The main aim of this paper is to evaluate the potential effects of climate warming (independently of other physical changes) on biodiversity at the local scale (ecosystem); we therefore hypothesize that local richness, as does regional richness, would increase in the future. We focus on species richness, one important component of biodiversity according to the Rio Convention (1992), and the most commonly used biodiversity measure in ecology (Magurran 2004) and in bioassessments. Freshwater communities from ponds are taken here as a model, with five representative groups which are 105 taxonomically and ecologically distinct: vascular plants, invertebrates (beetles, snails and dragonflies) and vertebrates (amphibians). These taxa are ecologically complementary with respect to their life cycles (amphibiotic vs aquatic) and colonization abilities (passives vs actives). The questions investigated are: i) will pond species richness increase in response to future warming? ii) what will be the impacts of other variables expected to change with climate warming (e.g. landuse and water quality)? iii) are there differences in response between taxonomic groups? iv) will the response be of similar magnitude in lowland and high altitude areas? To conduct the study, species richness was assessed in 113 ponds across Switzerland, a country characterised by different altitudes (from lowland to high altitude) and therefore heterogeneous thermal conditions. The relationship between species richness and environmental variables (including temperature and other variables e.g. landuse and water chemistry) was used to build predictive models (Generalized Additive Models, GAM). To begin, we predicted the future changes in species richness for the end of the next century according to the temperature increase from the A2 IPCC emission scenario. We then took into consideration changes in other related environmental variables. Both predictive approaches will demonstrate that climate warming could significantly increase the local species richness of temperate waterbodies, especially in high altitude areas. 106 METHODS Study area A total of 113 ponds were studied in Switzerland, a country situated in both the continental and alpine biogeographic regions of Europe. Study sites were situated at a range of altitudes (210 2760m above sea level), thus covering a large range of thermal conditions (Figure 1). Because of this large range of thermal conditions, the ponds were separated into four thermal groups according to the altitudinal belt in which they were located: colline, montane, subalpine and alpine ponds. Colline ponds (n = 55) have a mean annual air temperature of more than 8°C, montane ponds (n = 27) of 5 to 8°C, subalpine (n = 15) of 2.5 to 5°C, and alpine (n = 16) of less than 2.5°C. Figure 1: Location of the 113 study ponds in Switzerland. The greyscale corresponds to mean annual air temperature (data from the Swiss Federal Institute for Forest, Snow and Landscape Research). Local species richness A selection of 113 ponds was studied. Only ponds with relatively stable biological communities were included in the study: other ponds, such as temporary ponds (according to managers’ knowledge), pioneer ponds (less than 3 years) and ponds under active management (e.g. shoreline restoration, removing of sediments, planting of helophytes) were discarded. Local species richness was assessed 107 once in each pond between 1996 and 2005. Quality tests were conducted to ensure that these data covering a 6 year timescale (even 10 for some of them) could be used as a single dataset, i.e. that species richness had a low variability over such a time lag. In a first test, removing from the dataset the ponds studied outside a shorter timescale (5 years), did not lead to significant differences in predictions of species richness (mean Pearson correlation = 0.852). In further steps, investigations on five ponds of high altitude (Robinson & Oertli 2009) and eleven lowland ponds (Auderset Joye et al. 1993; Auderset Joye et al. 1994, and unpublished data) confirmed that ponds in stable conditions, as defined above, presented a stable species richness over a time lag from 6 to 10 years (differences not significant; Wilcoxon test p<0.05). Furthermore, only small modifications in the species composition were evidenced (through PCA analyses). Pond species richness was assessed using the PLOCH standardized method (Oertli et al. 2005b), which estimates the real species richness (thereafter “species richness”) for five taxonomic groups: aquatic vascular plants, aquatic Gastropoda, aquatic Coleoptera (larvae and adults), Odonata adults and Amphibia. The five groups represent heterogeneous taxonomic groups according to their systematic classification (vascular plants, invertebrates and vertebrates), but also in the ecological characteristics of their life cycle, amphibiotic (Odonata, Amphibia) vs aquatic (plants, Gastropoda, Coleoptera), and of their colonization abilities (passive: plants, Gastropoda. active: Coleoptera, Odonata, Amphibia). For each group, sampling effort was proportional to pond area: the aim was to sample at least 70% of the species for later estimating the real richness (100%) through species richness estimators (Chao1 or Jackknife1). Semi-quantitative sampling techniques were adapted to each taxonomic group: quadrats along transects for vascular plants, net sweeping samples for Gastropoda and Coleoptera, plot inventories for Odonata, and field observations for Amphibia. Inventories were carried out in both open water and interface between water and land. Each species presence was quantified through an abundance class, but this information was not used in the present study that relies solely on presence/absence data. Environmental data Approximately 100 local and regional environmental variables were recorded. A subset of fifteen variables was selected for the analysis (Table 1). The rationale for variable selection included: (i) the potential relevance of the variable in explaining patterns of species richness, as demonstrated in previous studies (Gee et al. 1997; Oertli et al. 2002; Jeffries 2003; Biggs et al. 2005; Hinden et al. 108 2005), and/or (ii) the absence of multicolinearity among variables (Pearson’s correlation coefficient ≤ 0.600; Oertli et al. 2000). Annual mean air temperature was the best climatic variable to characterise pond thermal conditions. Indeed, this variable was significantly correlated with several other climatic variables such as daily mean degree days (> 3°) (Pearson r = 0.986, p < 0.05) and solar radiation in July (Pearson r = -0.770, p < 0.05). Annual mean air temperature was also significantly correlated with mean annual precipitations, the potential evapotranspiration and the water budget in July, but with lower coefficients (respectively Pearson r = -0.412; -0.241; -0.225 for p < 0.05). Annual mean air temperature was calculated on the basis of monthly values from 115 weather stations and from a digital topographic model with a 25-m grid (data from the Swiss Federal Institute for Forest, Snow and Landscape Research). Temperatures for the 113 study ponds ranged from -2.2 to +10°C (Table 1). Table 3 : Local and regional environmental variables characterising the 113 study ponds. Variable Unit Median Mean Minimum Maximum °C 6.5 8.0 -2.2 10.0 - 4.6 3.7 0.0 8.2 Proportion of catchment area occupied by agriculture % 23 30 0 100 Proportion of forest in the surroundings (within 50 m radius) % 28 34 0 100 Proportion of pond area covered by floating vegetation % 16 29 0 100 Proportion of pond area covered by submerged vegetation % 22 35 0 100 class 3.0 3.1 1.0 4.0 cm 50 42 1 60 µS/cm 403.0 387.1 2.9 945.0 class 2 2 1 2 class 2 2 1 4 class 4 3 1 5 - 1.3 1.4 1.0 2.0 - 4 5 1 10 - 3.4 3.4 1.2 5.0 Annual mean air temperature Connectivity (in a radius of 1000 m) a Trophic class based on P, N and conductivity b Water transparency in summer (Snell tube) Water conductivity (winter measures) Fish presence c Extent of pond area shaded (trees coverage) Age d d Shoreline development Mean depth e d Log10(area) a Measure of pond isolation. This measure considers the number and size of ponds within a radius of 1000m and their distance from the study pond. Small values indicate high degree of isolation. b Maximum class of total phosphorus, total nitrogen and conductivity classes according to Wetzel, 1983 (between 1 and 4). Large values indicate high level of trophy. c Measure of the pressure of fish: 1 = low pressure, 2 = high pressure. d Classes (between 1 and 4 for the proportion of the pond surface area shaded, between 1 and 5 for age, and between 1 and 10 for mean water depth). Large values indicate: a high proportion of shade, old ponds, or deep ponds. e Ratio of the length of the shoreline to the circumference of a circle area equal to that of the pond (Wetzel 1983). 109 Predictive models Generalized Additive Models (GAM) were used to model the relationship between environmental variables (including mean annual air temperature) and species richness for each of the five taxonomic groups. GAM (Hastie & Tibshirani 1990) were used because they can fit any shape of response curve without having to assume particular relationships between species richness (response) and environmental variables (predictors). Before starting the modelling process, 6 outlier ponds (either for area, conductivity or temperature) were removed to homogenize the dataset. GAM were performed using S-PLUS and a set of functions developed for generalised regression analyses and spatial predictions (Lehmann et al. 2002; Maggini et al. 2006). Predictors were selected using a stepwise method based on an F-test with a p-value of 0.05. This was used for all groups except Amphibia, for which an Akaike’s Information Criteria (AIC) was used because the response curves of the F-model could not be explained ecologically. Cross-selection procedure and BIC-test were also tested but were not retained for any of the groups because the resulting models were poorly relevant from an ecological viewpoint. Three methods were used to assess the performance of the models: a simple correlation (COR), a ten-fold cross-validation correlation (cvCOR) (Lehmann et al. 2002), and the explained deviance (D2). Moreover, the ratio between mean pointwise standard errors and predictions for present conditions was calculated for each model (Hastie & Tibshirani 1990). Basic assumptions for prediction of species richness changes To predict the effects of climate warming on local pond richness, the future global temperature increase predicted by the A2 IPCC emission scenario for 2090-2100 was used: +3.4°C (IPCC 2007a). The “A2” scenario is the most commonly used in studies on the impacts of climate warming on biodiversity. The other five most confident scenarios (B1, B2, A1T, A1B, and A1FI) were also tested; they produced the same trends and therefore will not be presented. Three different types of predictions were made, in which values for a variable (for example temperature) were arbitrarily chosen in order to simulate a given condition. In the first type of prediction (“prediction 1”), only temperature was considered to increase during the century (+3.4 °C). This is the simplest prediction, giving a baseline for further investigations. As other environmental variables are also expected to be impacted indirectly by climate warming, we conducted two other types of predictions. In addition to changes in temperature, these consider 110 changes in six variables: proportion of the catchment area occupied by agriculture, proportions of pond area covered by floating and submerged vegetation, trophic level, water transparency and conductivity. These two types of predictions differed in the simulated magnitude of the impact of climate warming on the six variables: “prediction 2” is conservative (small changes) and “prediction 3” is pessimistic (large changes). Predictions 2 and 3 therefore represent minimal and maximal conditions respectively. The magnitudes of these expected changes were fixed according to the present distribution of the six variables in the study dataset (n=113). This level was based on the difference between two statistical parameters (median and quartiles). For the conservative scenario, the magnitude of change considered is the difference between the median and the first quartile, and for the pessimistic scenario, the difference between the median and the third quartile. The direction of the simulated changes was fixed according to previous studies and correlations in the dataset. Trophic level (and consequently conductivity) is expected to increase with climate warming, as evidenced by models and experiments on the relationship between primary productivity and warming (Rustad et al. 2001; Park et al. 2004). Furthermore, positive correlations were observed in the study dataset between temperature and trophic level (Spearman ρ = 0.478), and between temperature and conductivity (Pearson r = 0.616). Changes in proportions of pond surface area covered by aquatic vegetation are based on the hypothesis that an augmentation of the trophic level is linked to an increase in pond macrophyte cover - a hypothesis supported for example by Sondergaard et al. (2009). Water transparency is expected to decrease because of the negative relationship between trophic level and transparency (Spearman ρ = -0.432). A rise in temperature should increase the extent of land under agricultural management, as it has been predicted for Switzerland under moderate warming scenarios (Rebetez 2006; OcCC & ProClim– 2007). The area of agricultural landuse is therefore assumed to increase in the pond catchments. This hypothesis is supported by a positive correlation in the dataset between temperature and proportion of agricultural land around the pond (Pearson r = 0.506). Pond depth and pond surface area could possibly decrease with warming because of an increase in evaporation. Nevertheless, these variables were not considered here; indeed, predictions factoring in these two variables have been made in this study, but appear to have no significant influence on the predicted species richness. The seven variables considered for the three types of predictions and the magnitude of their expected changes are presented in Table 2. 111 Table 4: Environmental variables used for predicting species richness in 2090-2100. The magnitude of expected changes to variables with climate warming is indicated for each of three different scenarios: prediction 1 with only temperature increase, prediction 2 (conservative) and 3 (pessimistic) with changes in the other additional variables (see text for explanation of the derivation of percentage changes). Prediction 1 Prediction 2 Prediction 3 + 3.4 °C + 3.4 °C + 3.4 °C Proportion of catchment area occupied by agriculture no change + 23 % + 50 % Proportion of pond area covered by floating vegetation no change + 16 % + 42 % Proportion of pond area covered by submerged vegetation no change + 22 % + 72 % Trophic class based on nutrients charge (P, N) and conductivity no change +1 +2 Water transparency in summer (Snell tube) no change - 10 cm - 35 cm Conductivity no change + 230 μS.cm Annual mean air temperature -1 + 370 μS.cm The magnitude of expected changes to variables with climate warming is indicated for each of three different scenarios: prediction 1 with only temperature increase, prediction 2 (conservative), and 3 (pessimistic) with changes in the other additional variables (see text for explanation of the derivation of percentage changes). 112 -1 RESULTS Present values of pond species richness Species richness assessed for each of the five taxonomic groups in the 113 ponds ranged widely: from zero to 32 species per pond according to the taxa (i.e. “present” richness presented in Figure 2). This high variability reflects the different ecological conditions in the ponds surveyed. Indeed, ponds are pseudo-replicates: they can be large or small, deep or shallow, shaded or exposed, etc. Nevertheless, this variability is explained to a high extent by the large range of thermal conditions in these ponds. Indeed, species richness was positively correlated to the mean annual air temperature, with high Pearson r-correlation coefficients: 0.45 (vascular plants), 0.50 (Gastropoda), 0.53 (Coleoptera), 0.66 (Odonata), 0.67 (Amphibia). Predictive models The five Generalized Additive Models produced here (that will further be used as predictive models) linking species richness to environmental variables performed well for each of the five taxonomic groups (Table 3). This is indicated by their high correlation coefficients (mean COR = 0.75) and crossvalidation correlation coefficients (mean cvCOR = 0.68). They also explained more than fifty percent of the richness variability (mean D2= 0.57). The ratio between pointwise standard errors and predictions for present conditions correspond to values of 10% to 24% according to the model (mean values, see Table 3). Temperature was the environmental variable which made the highest contribution in the five models: from 56 to 60 % (from Table 3), with a positive relationship with species richness. Apart from temperature; shade, pond area, water conductivity, and trophic state were the most significant in contributing to the models (in at least 2/5 models). 113 Table 5 : Model contribution (expressed in the scale of the linear predictor) of the 15 variables selected by a stepwise procedure in the five predictive models (GAM) and validation diagnostic. Validation coefficients are correlation coefficients (COR), cross-validation correlation coefficients (cvCOR) and percentage of variability explained (D2). MODELS Vascular plants Gastropoda Coleoptera Odonata Amphibia Variables Annual mean air temperature 2.40 Extent of pond area shaded 0.85 Log10(area) 4.30 2.57 1.07 Conductivity 0.68 Trophic class 0.30 Water transparency 3.41 2.21 0.87 0.55 0.87 0.57 0.54 1.02 Age 0.48 Mean depth 0.44 Shoreline development 0.55 Proportion of pond area covered by submerged vegetation 0.48 Proportion of pond area covered by floating vegetation 0.59 Fish presence 0.34 Connectivity (in a radius of 1000 m) 0.53 Proportion of catchment area occupied by agriculture 0.34 Validation coefficients 2 48 54 52 71 62 COR 0.69 0.69 0.75 0.86 0.77 cvCOR 0.59 0.60 0.70 0.79 0.70 Ratio pointwise standard errors/predictions (mean) 0.10 0.24 0.15 0.16 0.16 Ratio pointwise standard errors/predictions (max) 0.26 0.87 0.29 0.40 0.26 D Validation coefficients are percentage of variability explained (D2) correlation coefficients (COR) and cross-validation correlation coefficients (cvCOR). 114 Potential increase in pond species richness in response to climate warming Pooled species richness (all five taxonomic groups) Because of the high contribution of temperature in the five models, these can be used for predicting future climate warming effects on pond species richness. A temperature increase (+ 3.4°C for 20902100) clearly enhanced pond richness for each of the taxonomic groups (prediction 1 in Figure 2). Species richness of the five taxonomic groups pooled increased from 34 to 70 (+106%). Factoring in the other environmental variables indirectly impacted by climate (Table 2), both conservative (prediction 2) and pessimistic scenarios (prediction 3) led to a clear increase in species richness (Figure 2), raising the species richness to 64 and 62 (i.e. +88% and +82% respectively). All three types of predictions show a high variability (Figure 2), as already observed with the present values. Despite this high variability, the richness increases were all statistically significant (Wilcoxon p-values < 0.05). 115 Figure 2: Present and predicted pond species richness changes in response to climate warming for 2090-2100 for five taxonomic groups (n= 113 ponds). Prediction 1 takes into account only temperature rises (A2 IPCC emission scenario: +3.4°C). Predictions 2 and 3 consider six additional variables expected to be indirectly influenced by climate warming (same as Table 2). Prediction 2 is the ‘conservative’ scenario and prediction 3 the ‘pessimistic’ scenario for climate warming. Vertical boxes represent the interquartile range (Q3-Q1), within which the line represents the median. Whiskers represent the largest non-outlier values and dots represent outlier values. Magnitude of increased richness according to taxonomic groups The species enrichment predicted in response to warming differed in magnitude according to taxonomic groups. Indeed, pond species richness clearly increased for each taxonomic group, but at different rates (Figure 2). Considering a change only in temperature (prediction 1), species richness of vascular plant increased from 11 to 18 (+68%), from 3 to 5 for Gastropoda (+84%), from 7 to 15 for Coleoptera (+119%), from 4 to 9 for Amphibia (+119%), and from 8 to 18 for Odonata (+128%). Both 116 conservative and pessimistic scenarios (predictions 2 and 3) led to the same magnitude of increase for all taxonomic groups, excepting vascular plants. For vascular plants, a particularly lower increase (1 new species, +9%) was predicted with the pessimistic scenario (prediction 3). This low increase resulted from two variables - trophic level and conductivity - which strongly buffered the species enrichment predicted by a temperature increase alone. The smallest changes occurred for vascular plants and Gastropoda when considering only ponds from the colline altitudinal belt: less than one species, i.e. -1% and +18% respectively with the conservative scenario, +2% and +6% respectively with the pessimistic one. The variables responsible were two indicators of water quality (conductivity and trophic level) for vascular plants and the proportion of pond covered by floating vegetation for Gastropoda. Magnitude of increased richness according to altitude The predicted increases in pond species richness associated with climate warming differed in magnitude between altitudinal levels, for the five taxonomic groups pooled, as illustrated in Figure 3. According to all three types of predictions, the sites situated in high altitude areas (subalpine and alpine) were predicted to gain fewer species than lowland areas (colline and montane). When considering only temperature increase (prediction 1), the species richness increased from 41 to 75 (+83%) in lowland ponds (from 51 to 89 and from 30 to 61 in colline and montane ponds respectively, i.e. from +75% and +103%). In high altitude ponds, the species richness was predicted to increase from 41 to 75 (+83%) (from 18 to 23 and from 10 to 18 in subalpine and alpine ponds respectively, i.e. from +128% and +180%). Nevertheless, the increase in species richness was proportionally almost twice higher in high altitude than lowland ponds. Environmental variables other than temperature can have an antagonistic influence on species richness depending on altitude. In the conservative scenario (prediction 2), the inclusion of these variables in the models produced a lower increase in species richness for colline ponds than for temperature alone. On the other hand, these variables produced a greater increase in species richness in ponds at higher altitude (Figure 3). Furthermore, there could also be contrasting effects between the two predictive scenarios (minor or major changes in environmental variables). For example, in alpine ponds, the conservative scenario resulted in a greater increase in species richness, whereas the pessimistic scenario showed a smaller increase (Figure 3). The variables most involved in these antagonist patterns were linked to nutrient availability (conductivity, trophic level, and proportion of floating vegetation). 117 When the five taxonomic groups are considered separately, species richness tends to increase according to the three types of predictions in both lowland and high altitude areas (data not presented here). Figure 3: Present and predicted pond species changes in response to climate warming for 2090-2100 for four thermal levels: colline, montane, subalpine and alpine (n= 55 for colline ponds, 27 for montane ponds, 15 for subalpine ponds, and 16 for alpine ponds). Predictive scenarios same as in Figure 2. Vertical boxes and whiskers represent the same parameters as in Figure 2. 118 DISCUSSION Potential increase in pond species richness in response to climate warming A clear increase in pond species richness was predicted for 2090-2100 for vascular plants, Gastropoda, Coleoptera, Odonata and Amphibia. This increase took place for three types of predictions, one considering only temperature increase (prediction 1), and two factoring in additional variables (e.g. landuse, water quality) in conservative (prediction 2) or pessimistic scenarios (prediction 3). The magnitude of the increase in species richness was high and when all taxonomic groups were pooled, the increase was 88% and 82%, raising the species richness to 64 and 62 for the conservative and pessimistic scenarios respectively. The predicted local increase in species richness agrees with the most commonly described trend at the regional scale – increased richness – (Iverson & Prasad 2001; Daufresne & Boet 2007; Buisson et al. 2008), even though a few studies reported possible trends of richness impoverishment (Walker et al. 2006; Wilson et al. 2007). Most studies considering species richness at the site level have shown the same increasing trend. Monitoring of terrestrial vascular plants in permanent plots in the Alps showed an increase in richness (+86%) during the 20th century (Vittoz et al. 2009); such an increase (+12%) has already been observed recently over a shorter timescale of 10 years (Pauli et al. 2007). Magnitude of increased richness according to taxonomic groups The predicted increase in species richness varied slightly according to the taxonomic groups for all three types of predictions, with the exception of vascular plants. Vascular plants were the only taxonomic group to respond markedly to changes in environmental variables other than temperature. Water quality degradation (conductivity and trophic level) significantly reduced the increase in plant richness predicted by a temperature increase alone. This is clearly due to the high sensitivity of many aquatic plant species to eutrophication (e.g. Melzer 1976). Magnitude of increased richness according to altitude The increase in species richness (all taxonomic groups pooled) was predicted to be higher in lowland ponds than in high altitude ponds, but the proportion of that increase was almost double for high altitude (150%) over lowland ponds (83%). Notwithstanding the clear and positive effect of 119 temperature, other environmental variables also had a marked impact on richness, which could be antagonistic for the conservative or pessimistic scenarios depending on whether the pond was located in high altitude or lowland areas. These different and sometimes antagonistic responses are linked to the unimodal relationship between species richness and productivity frequently described in aquatic systems (review in Mittelbach et al. 2001). As lowland ponds are already relatively nutrient-rich (e.g. Menetrey et al. 2008), a potential future nutrient loading is likely to lower species richness. Conversely, in relatively nutrient-poor alpine ponds (Menetrey et al. op.cit.), an increase in nutrient loading is likely to raise species richness. This confirms the role of ponds as ideal early warning systems. Indeed, the high sensitivity of high altitude ecosystems to climate warming highlighted in this study has frequently been reported in the Alps (Theurillat & Guisan 2001; Beniston 2003), and in other mountainous areas (Nogues-Bravo et al. 2007). Moreover, as local richness is relatively easy to measure in ponds because of their small size, simple community structure (De Meester et al. 2005) and well-defined boundaries, ponds are ideal tools for monitoring the long-term impacts of climate change at the local scale. In the Swiss National Park for example, ponds are already used as a sensor to assess the long-term impacts of climate change (Robinson & Oertli 2009). Which species would be affected by climate warming? The predicted increase in pond species richness will be the result of a positive balance between numerous colonization events and sparse extinction events. Cold thermal specialists of ponds (i.e. species reproducing solely in ponds characterised by low water temperature) are at risk of extinction. In Switzerland, they are absent (Gastropoda or Amphibia) or represent 22 species among 287 (8%, vascular plants), 13 species among 122 (11%, Coleoptera) or 7 species among 54 (13 %,Odonata), and are mostly restricted to high altitude areas (unpublished data, Swiss Centre for Fauna Cartography and Swiss Centre of Floristic Network). All other species from the study area are thermal generalists which occupy principally lowland areas and are likely to colonise high altitude ponds under climate warming. These changes are likely to occur very quickly in Switzerland because lowland and high altitude regions are geographically interlinked. For lowland ponds, new colonists will come mostly from southern regions (including the Mediterranean basin): some are currently expanding their geographic range, e.g. the Odonate Aeshna affinis (Grand & Boudot 2006). This is likely to be a slow process overall. However, it may also occur more rapidly due to the introduction and expansion of exotic species (Rahel & Olden 2008). In the studied area, there are already 4 exotic Gastropoda species (9% of the species pool) and 2 exotic submerged plant species (5% of the species pool). 120 Uncertainty in the magnitude of the increase in species richness The magnitude of the predicted increase in species richness depends primarily on the magnitude of the temperature change. The temperature changes predicted by the IPCC emission scenarios (used in this study) have underlying uncertainties and are on a worldwide scale. The future warming of the study area could be higher than predicted by these global scenarios, as evidenced by the fact that the current observed warming in the study area is higher than the trend in the Northern Hemisphere (Rebetez 2006). The uncertainties of the predictive models were reduced to depict most adequately the changes in pond species richness. All key potential drivers of species richness contributed significantly to the predictive models. Furthermore, the high percentage of variability of the dataset explained by the models demonstrates that these variables are good predictors. Moreover, cross-validation coefficients show that the models have a high stability because they are based on a relatively large number of ponds (n=113). The magnitude of the predictions relies also on the ability of the ecosystems to support new species, i.e. they should not be “species-saturated”. Because of their low species richness and abundances (Oertli et al. 2008), coupled to their high potential of increase in resources (nutrients and primary production), high altitude ponds have the potential to host a large number of new species. For lowland ponds, it is less obvious. However, regressions between regional (200-meters altitudinal classes) and local species richness for the study ponds suggested that lowland ponds are likely to be currently unsaturated (unpublished data). Additional parameters that could reduce the uncertainty of the predicted increase in species richness should be investigated in future research. Physical changes in wind speed, precipitation (IPCC 2007a) and evaporation should be considered because they may have a greater impact than warming alone (Körner & Walther 2001). Strong hydrological perturbations (e.g. changes in the hydroperiod) are also likely to affect small waterbodies. The variability of physical changes would also have effects on small waterbodies, as ecological responses do not only depend on global averages (Walther et al. 2002). Uncertainty in the rate of increase in species richness Predictions were made for 2090-2100. However, the colonization and dispersal abilities of species could influence the rate of increase in species richness. Species have different dispersal strategies 121 (Bilton et al. 2001), and those with higher colonization abilities are likely to colonise new areas faster than passive dispersers. Furthermore, once a species reached a new ecosystem, it should persist and this persistence potential will depend on (i) ecological interactions, such as trophic or habitat interdependence and local competition, (ii) species niche breadth (Kotiaho et al. 2005), and (iii) habitat diversity (MacArthur & Wilson 1967). As an example, rapid colonization by active dispersers could be impaired by the absence of favourable habitats created by passive dispersers, e.g. aquatic plant stands. Therefore, habitat availability could make colonization events more the exception than the rule (Hill et al. 2001), and delay the predicted increase in species richness. Some recent studies have addressed this issue by including the species colonization rate in dynamic simulations of plant migration under climate change (Randin 2007). 122 CONCLUSION Earlier investigations have shown that species richness is likely to increase in many parts of the world by the end of the century, in particular in high altitude areas. This has mainly been investigated at the regional scale (γ diversity) and rarely at the local scale (α diversity). Our predictions for pond ecosystems, based on expected temperature increase and its indirect effects on a set of related environmental variables, showed that species richness would increase markedly at the local scale. The species-poor ecosystems characterising high altitude regions would proportionally enrich markedly. This overall increase in species richness should not mask the fact that species extinctions will also occur, and even if in smaller numbers than colonisation events, this will affect the pool of cold thermal specialists. This enhances the pressure on freshwater ecosystems, already suffering from many other anthropogenic alterations. The next steps in investigating the increase in species richness of ecosystems associated with climate change should now focus on: (i) model improvements taking into account other important factors, such as species distribution changes, species dispersal abilities and physical changes in habitats and (ii) field evidence recorded by long-term monitoring programs. Further developments should also investigate components of biodiversity other than species richness, such as: (i) richness in rare and endemic species, or conservation value linked to degree of threat of species, (ii) diversity indices (e.g. Shannon index) taking into account species abundances, (iii) biomass, productivity, and energy flow measures, and (iv) species traits linked to the functioning of the ecosystem. The use of pond ecosystems seems particularly appropriate for such investigations, and this study has successfully demonstrated that they constitute an excellent sentinel ecosystem for detecting climate change effects at the local scale. 123 ACKNOWLEDGMENTS We thank the numerous people from the University of Geneva (Laboratory of Ecology and Aquatic Biology) and the University of Applied Sciences of Western Switzerland (EIL, Geneva) for the fieldwork, the laboratory-work and the database management. Various supports contributed for gathering the database for the 113 ponds studied: the Federal Office for the Environment (OFEV), many Swiss Cantons, the University of Geneva (Laboratory of Ecology and Aquatic Biology), the University of Applied Sciences of Western Switzerland (RCSO RealTech), and the Research Commission of the Swiss National Park. We also thank the Swiss Federal Institute for Forest, Snow and Landscape Research (WSL) for providing temperature data, and the Swiss Centre for fauna cartography (CSCF) and the Swiss Centre of floristic network (CRSF) for providing biological data at the regional scale. We also would like to thank four anonymous reviewers for their insightful comments which greatly helped to improve the manuscript, and P. Nicolet from the Pond Conservation, J. Park and F. Cattaneo for editing the English. 124 Article 5 Freshwater biodiversity under climate warming pressure: identifying the winners and losers in temperate stagnant waterbodies. Rosset Véronique & Oertli Beat This manuscript is published in Biological Conservation 144 (2011). University of Applied Sciences Western Switzerland, hepia technology, architecture and landscape - CH-1254 Jussy-Geneva, Switzerland. 125 126 KEYWORDS Ponds and wetlands, sensitivity to climate warming, thermal preferences, species resilience to perturbations, extinction, colonisation. ABSTRACT Climate warming is affecting the biodiversity all around the world, resulting in the expansion or contraction of the geographical range of species, and leading to colonisation (winners) and extinction (losers) events in ecosystems. It is crucial for the conservation of biodiversity to identify these potential winners and losers. We focus here on small standing waterbodies in Switzerland and on five taxonomic groups: vascular plants, snails, beetles, dragonflies and amphibians. We first assessed the sensitivity of each species to climate warming through their thermal preferences, using current altitudinal and latitudinal distribution, as a surrogate for temperature. We then evaluated the resilience of species to perturbations through five ecological and biogeographical criteria applicable to the perturbation “warming”: dispersal ability, degree of habitat specialisation, geographical extent in the study area, future trend in geographical extent, and future trend of habitat availability for species. Potential winners and losers of a warming climate could be quantified through their thermal preferences. The proportion of potential losers ranged from zero species for snails to 33% of the regional species pool for dragonflies. The set of potential winners was much larger, ranging from 53% for amphibians to 61% for dragonflies. A multimetric index combining the five resilience criteria enabled the further prioritisation of the species along a gradient of extinction risk. This potential threat from climate warming is not reflected by the current Red Lists of dragonflies and amphibians, suggesting that conservation management could gain from a complementary label indicating the degree of sensitivity to warming. 127 1. INTRODUCTION The Earth’s climate is predicted to become warmer during the 21th century with global average increases in temperature of 1.8 - 4°C, and it has already warmed by 0.74°C over the 20th century (IPCC 2007a). In addition, there is no longer any doubt that climate warming has drastic effects on biodiversity. Indeed, it is affecting the geographical range of species and the composition of communities all around the world (reviews in McCarty 2001, Walther et al. 2002). An increase in species richness under the influence of climate warming is expected in alpine areas according to the well-known decline in species richness with increasing altitude (see review from Rahbek (1995)), for both terrestrial (e.g. Herzog et al. 2005, McCain 2005) and aquatic taxa (e.g. Lods-Crozet et al. 2001, Jones et al. 2003, Jacobsen 2004). Moreover, several studies have predicted that climate warming will induce an increase in local species richness in different ecosystems, from montane and subalpine forests (Kienast et al. 1998) to small temperate waterbodies (Rosset et al. 2010). Some long-term datasets have already confirmed this predicted trend. For example, the species richness of terrestrial plants has been reported to increase during the last century in permanent plots situated on alpinenival summits (e.g. Grabherr et al. 1994, Vittoz et al. 2009). These observed or predicted increases in taxonomic richness are the consequence of changes in species composition. Indeed, as it has occurred over geological time scales in response to a changing climate, many taxa will shift their geographical range and move into regions where climate is more suitable, while others will fail to do so and will suffer extinction. Therefore, in a given region (and in a given ecosystem), taxa can become extinct (“the losers”), while at the same time colonisations by new taxa can occur (“the winners”). An increase in taxonomic richness is observed when the colonization events are more frequent than the extinction events. The ability of taxa to adapt to the changing climatic conditions (e.g. Ashley et al. 2003) would make this balance even more complex. The geographic range of many taxa has already been observed to shift towards regions where their climatic suitability is increasing, and this is likely to be the response of most species (e.g. Parmesan et al. 1999, Parmesan and Yohe 2003, Hickling et al. 2006). However, some taxa will not be able to track the changing climate and are at risk of extinction. Estimates of the number of these taxa differ greatly between regions and between taxa: from only a few percent (e.g. Levinsky et al. 2007) to almost eighty percent (e.g. Thomas et al. 2004) of the regional species pool. Freshwater biodiversity, in particular, is highly vulnerable to climate change, with extinction rates matching or exceeding those suggested for better-known terrestrial taxa (Heino et al. 2009). Besides these quantitative estimates, it is essential to identify the species which are at risk of extinction in order to design targeted action plans for conservation. The use of species-specific 128 criteria related to ecology, life-history or geographical distribution is frequent in extinction risk assessments independent of climate changes (e.g. Keith 1998, O'Grady et al. 2004, Kotiaho et al. 2005, Mattila et al. 2008). Investigation of such criteria to detect taxa which are the most extinctionprone because of climate warming is more recently emerging (e.g. Ott 2001, Heikkinen et al. 2010, Williams et al. 2010, Graham et al. 2011): several species-specific criteria have been related to the sensitivity of species to global changes in climate, such as thermal preferences, dispersal ability, geographical extent, population trend and degree of habitat specialisation (Ott 2001, Beaumont and Hughes 2002, Thuiller et al. 2005a, Broennimann et al. 2006, Jiguet et al. 2007, Calosi et al. 2008, Vittoz et al. 2009). Furthermore, the combination of these criteria could enhance the effects of climate change on species. Travis (2003) demonstrated that habitat specialists with poor dispersal abilities or suffering from habitat loss will be more affected by climate change than other habitat specialists. The present study assess the sensitivity to climate warming of species from five taxonomic groups (vascular plants, beetles, snails, dragonflies and amphibians) living in small standing waterbodies (ponds or wetlands) in Switzerland. Ponds were used as an ecosystem model because: (i) their small size and relatively simple communities make them attractive model systems for research (De Meester et al. 2005), (ii) freshwater systems respond strongly to physical environmental changes, such as climate change (Heino et al. 2009), (iii) ponds are predicted to increase in species richness in response to warming (Rosset et al. 2010). The present investigation aims to detect the potential winners and losers involved in this local species enrichment, with the final aim of providing baseline information to prioritise conservation measures towards the species most at risk of extinction. Specifically, we will evaluate the sensitivity of each species to warming through two successive and complementary approaches. The first approach investigates the thermal preferences of species using altitudinal and latitudinal distributions as a surrogate for temperature. This will allow to quantify the number of species in four distinct groups of winners and losers: two groups of potential losers at risk of extinction (cold and cool thermal specialists), and two groups of potential winners (warm thermal specialists and thermal generalists). A second approach takes into consideration the resilience of species to perturbations, and therefore their ability to track the climate warming; it is used for groups for which sufficient ecological and biogeographical knowledge is available, i.e. for Odonata and Amphibia. This second approach aims to prioritise further the species from each of the four groups along a gradient of extinction risk. The method uses a multimetric index, with five criteria describing the resilience of species to a “warming” perturbation: the dispersal ability, the degree of habitat specialisation, the geographic extent in the study area, the future trend in geographic extent and the future trend of habitat availability for species. 129 2. METHODS 2.1 Study area Switzerland is an outstanding model for investigating climate issues and particularly the sensitivity of species to climate warming. Indeed, this country covers a large range of altitudes (210 – 2760 meters above sea level), and therefore a large range of thermal conditions (mean annual air temperatures from -2°C to 12°C). Despite its small size (41,000 km2), Switzerland includes many standing waterbodies with an estimated 32,000 ponds (defined here as waterbodies between 100 m2 and 5 ha in surface area) and 360 lakes. 2.2 Study taxonomic groups Five taxonomic groups were investigated: aquatic vascular plants, aquatic Gastropoda, aquatic Coleoptera (larvae and adults), Odonata adults and Amphibia. These represent heterogeneous groups in terms of their systematic classification (vascular plants, invertebrates and vertebrates), but also in terms of their life cycle, i.e. amphibiotic (Odonata, Amphibia) vs aquatic (plants, Gastropoda, Coleoptera), and of their colonization abilities, i.e. passive (plants, Gastropoda) vs active (Coleoptera, Odonata, Amphibia). For each taxonomic group, only species dependent on pond habitats for an important part of their life cycle were investigated; therefore running water species were discarded. For vascular plants in particular only aquatic species, as defined by their humidity index scores, were investigated (for details and full species list see Oertli et al. 2002). Information concerning these five taxonomic groups came mainly from large national databases available in Switzerland for the period 1990-2007 (databases from the Swiss Centre for fauna cartography CSCF, the Swiss Centre of floristic network CRSF). More than 60’000 species records were available. These species records came from different sources: research projects, volunteer records, species inventories or museum records. The scientific validity of all species records was rigorously verified by experts before their inclusion in the databases. 2.3 Methodological approach for the identification of potential winners and losers Species-specific criteria related to ecology or geographical distribution were used to assess the sensitivity of species to warming through two successive and complementary approaches. The first 130 approach investigated the thermal preferences of species, using knowledge of their altitudinal and latitudinal distribution as a surrogate for temperature in order to classify species into distinct groups of winners and losers. The second approach takes into consideration the resilience of species to perturbations, and therefore their ability to track the climate warming, in order to prioritise further the species from each of the groups identified through the first step along a gradient of extinction risk. 2.4 Thermal preferences of species Thermal preferences of species were investigated with the final aim to quantify the number of species in two groups of losers: the cold thermal specialists and the cool thermal specialists, and in two groups of winners: the warm thermal generalists and the warm generalists. The underlying assumption is that species with restricted thermal ranges, in particular in cold or cool temperatures are potential losers and that species with ranges restricted to warm temperatures or with large thermal ranges are potential winners. Thermal preferences of species were investigated with (i) the altitudinal distribution of species in the study country (Switzerland), which allowed the identification of the cold thermal specialists, and (ii) with the latitudinal distribution of species at the European scale, which allowed the identification of the cool thermal specialists, the warm thermal specialists and the thermal generalists. 2.4.1 Altitudinal distribution Altitudinal distribution was investigated for the whole of the Swiss territory, either through occurrence data in altitudinal classes or through literature review and expert advice. Occurrence data were available for Gastropoda, Odonata and Amphibia. Data came from the large national databases described above. Altitudinal classes consisted of 200-meters intervals. Lower altitudinal classes were combined because of the small amount of data available, and so class 1 represented all altitudinal classes lower than 500 meters (mean annual air temperature of above 9°C). The same procedure was applied to the highest altitudinal classes: class 11 represented all altitudinal classes over 2500 meters (mean annual temperature under 0°C). The altitudinal distributions of species were explored through boxplots representing the extent of their thermal preferences. Because of the large species pools, species having similar thermal preferences were grouped (more details can be obtained on request). 131 There was a lack of information on altitudinal distribution for both Coleoptera and vascular plants, and therefore literature review and/or expert advice were used. For Coleoptera (Hydradephaga only), altitudinal distributions were investigated through both literature review (Schmedtje & Colling 1996; Carron 2005, 2008) and expert advice (G. Carron, personal communication, 2008). For vascular plants, the thermal value from Landolt (1977) was used: from 1 (cold, high altitude) to 5 (warm, lowland). Occurrence data in altitudinal classes together with literature review and expert advice enabled the identification of one group of potential losers: the cold thermal specialists. In our dataset, these boreo-alpine species live exclusively in regions characterised by cold temperature (mean annual air temperatures about lower than 6°C). For Gastropods, Odonata and Amphibia, they were defined as species whose median occurrence was at 900 meters or higher. For vascular plants, cold thermal specialists were those species with a Landolt thermal value lower or equal to 2. For Coleoptera, they were defined as species inhabiting exclusively eukrenal, hypokrenal and epirhitral areas (Schmedtje and Colling 1996); in addition, a set of supplementary species was added according to unpublished data about their geographical range in the study country (G. Carron, personal communication, 2009). 2.4.2 Latitudinal distribution The latitudinal distribution was investigated at the European scale for Odonata and Amphibia, using the size of the entire geographical range of each species and the ratio between the maximal latitudinal width of each species range in Southern (at a lower latitude than Switzerland) and Northern Europe. The latitudinal distribution was described by the occurrence of species in grid cells of 500x500 km (data from Dijkstra & Lewington 2006 for Odonata, and Meyer et al. 2009 for Amphibia). Investigation of latitudinal data allowed first the identification of one group of potential winners: the thermal generalists. They were defined as species having a geographical distribution covering more than 75% of Europe. Analysis of the latitudinal distribution of the remaining species allowed the identification of one other group of potential winners: the warm thermal specialists, which were defined as species having a latitudinal extent two times larger in Southern than in Northern Europe. It also identified one group of potential losers: the cool thermal specialists. These were defined as species having a latitudinal extent two times smaller in Southern than in Northern Europe and not previously identified as cold thermal specialists. 132 2.5 Resilience to perturbations The resilience of species to perturbations was assessed with the final aim of prioritising the species from each of the four groups of winners and losers along a gradient of resilience to warming. This was achieved through the use of five ecological and biogeographical criteria applicable to the perturbation “warming”: (1) dispersal ability, (2) degree of habitat specialisation, (3) geographic extent in the study area, (4) future trend in geographical extent in the study area, and (5) future trend of habitat availability for species. These resilience criteria could be investigated for Odonata and Amphibia, because good quality data on ecology and biogeography was available. The first criterion of resilience, dispersal ability, is directly related to the ability of a species to track change, in temperature for example. It was estimated for Odonata by the minimal size of posterior wings (data from Dijkstra & Lewington(2006)), and for Amphibia by the maximal dispersal distance (ACEMAVcoll. et al. 2003, Smith and Green 2005, Meyer et al. 2009). For the second criterion of resilience, the degree of habitat specialisation of each species, the underlying assumption was that species with restricted habitat requirements (habitat specialists) would be more sensitive to large-scale changes such as climate warming than species with broad habitat requirements (habitat generalists). The degree of habitat specialisation was estimated by adding up scores from 1 to 3 (1 high affinity, 3 low affinity) describing the affinity of each species towards different types of freshwater habitats, 20 habitat types for Odonata (Appendix A) and 8 habitat types for Amphibians (Appendix B from Meyer et al. 2009). Small values of the criterion correspond to a high degree of habitat specialisation. Affinity of Odonata species towards different types of freshwater habitats consists of information provided by national and local experts (Appendix A). The third criterion of resilience was the present geographical extent of each species in the study area. Indeed, species with restricted distribution could suffer local extinction which could lead to regional extinction. This criterion was assessed through the number of grid cells of 20x20 km where a species is currently present in Switzerland (national databases described above). The fourth criterion of resilience was the future expected trend in the geographical extent of each species in Switzerland. Past and present temporal trends were used as a surrogate for future trends. The underlying assumption was that species already undergoing a diminution of their geographical range will be less resilient to a perturbation (such as climate warming) than species expanding their range. This criterion was assessed by comparing the occurrence of species between two periods (1970-2002 to 2003-2004 for Amphibia; 1970-1988 to 1999-2000 for Odonata). The trend was 133 assessed for each species according to the method developed by Gonseth & Monnerat (2001). For Odonata, trends were obtained directly from Y. Gonseth (personal communication) and for Amphibia they were calculated according to data from the Swiss Red List (Schmidt and Zumbach 2005). Trends were only determined for species with more than 25 records. The last criterion of resilience was the future trend of habitat availability for species. The underlying assumption was that the potential for freshwater habitats to host a species within the study area would influence to the ability of a species to track change - for example, in temperature. This criterion was estimated by combining the affinity of each species towards different types of freshwater habitats (degree of habitat specialisation) and the present and future abundance of these habitats in Switzerland. The abundance of each freshwater habitat within the study area consists of an expert assessment informed by national hydrological inventories. A score was given for each habitat, from 0 (absent) to 5 (extremely frequent) (see Appendix C). A weighting factor (w) was then applied to this score (see Appendix C), in order to take into account the impairment or improvement of the habitats (present and future). Indeed, increasing or decreasing human pressure can considerably alter or enhance the future capacity of freshwater habitats to host Odonata and Amphibia populations. This final criterion was calculated according to the following equation, where “si,h” is the affinity of each species i towards the habitats h , “aph” and “afh” are the present and future abundance of each habitat h in the study area, and “wph” and “wfh” are the present and future proportion of each type of habitat h in a unimpaired state (able to host Odonata or Amphibia): ∑( ) ∑( ∑( ) ) 2.5.1 The multimetric index of resilience An index of resilience to warming was calculated for each Odonata and Amphibia species by combining the five ecological and biogeographical criteria. In order to give the same weight to the five criteria, each criterion was transformed into a metric ranging from 0 to 1. As the five metrics (= the five transformed criteria) showed no multicolinearity (Pearson r coefficients lower than 0.70), they were combined into a multimetric resilience index, which consists of the mean of the five metrics. 134 3. RESULTS 3.1 Thermal preferences of species The species of the regional species pool of the study country were classified according to their thermal preferences into one to four groups of winners and losers; for all taxonomic groups cold thermal specialists were identified and for both Odonata and Amphibia three supplementary groups of winners and losers were identified. From the regional pool of 287 aquatic vascular plants inhabiting ponds, 22 species (8%) were identified as cold thermal specialists (potential losers) on the basis of their “Landolt” thermal value (Figure 1 and Appendix D).The other plant species could not been classified as potential winners or losers based on their thermal preferences (unknown status). Figure 1: Percentages of species inhabiting ponds in Switzerland identified as potential winners or losers on the basis of the altitudinal and latitudinal distribution as a surrogate for temperature. (A) Percentages of species identified as cold thermal specialists (potential losers in a context of climate warming) for vascular plants, Gastropoda, Coleoptera, Odonata and Amphibia. (B) Zoom on two taxonomic groups, Odonata and Amphibia, for which the percentages of cool thermal specialists (potential losers), warm thermal specialists (potential winners) and thermal generalists (potential winners) were also identified because of sufficient ecological and biogeographical knowledge. For Gastropods (44 pond species), there were no cold thermal specialists (Figure 1, Figure 2 and Appendix D). The 44 Gastropod species were of unknown status. 135 Figure 2: Altitudinal distribution of pond species in Switzerland, for Odonata (A), Gastropoda (B) and Amphibia (C). Horizontal axes are altitude classes of 200 meters or more (see Methods) as a surrogate for temperature. N (vertical axes) is the number of occurrences in the regional data set from the CSCF (Swiss Centre for fauna cartography). More details about the species included in the general patterns (47 for Odonata, 44 for Gastropoda and 17 for Amphibia) can be obtained on request. Of the regional pool of 122 Coleoptera pond species, 13 (11%) were identified as cold thermal specialists on the basis of literature or expert knowledge (Figure 1 and Appendix D). The other Coleoptera species could not be classified as potential winners or losers based on their thermal preferences (unknown status). For Odonata, 7 out of 54 pond species (13%) were identified as cold thermal specialists based on their present altitudinal distributions in Switzerland (Figure 1, Figure 2 and Appendix E). Among the other Odonata species, 11 (20%) were cool thermal specialists (potential losers) based on their latitudinal distribution in Europe. Five species (9%) were identified as warm thermal specialists (potential winners) and 28 species (52%) were identified as thermal generalists (potential winners). The other 3 Odonata species (6%) could not been classified as potential winners or losers based on their thermal preferences (unknown status). 136 For Amphibia (17 pond species), there were no cold thermal specialists (Figure 1, Figure 2 and Appendix F). According to their latitudinal distribution in Europe, 3 Amphibia species (18%) were identified as cool thermal specialists, 8 species (47%) as warm thermal specialists, and one species (6%) as a thermal generalist (Figure 1 and Appendix F). The 5 remaining Amphibia species (29%) could not be classified as potential winners or losers according to their thermal preferences (unknown status). 3.2 Resilience to warming The resilience of species to warming could be investigated within two ecologically and biogeographically well-known groups: Odonata and Amphibia. The main patterns will be presented here with a few examples; the detailed classification of all species along the gradient of resilience to perturbations is available in Appendices E and F. 3.2.1 Odonata The values of resilience obtained for Odonata differed largely among the species, allowing to prioritise further the species of each group of winners or losers (Figure 3). Among the ten Odonata species predicted to be the least resilient eight were already detected as potential losers based on their thermal preferences, but two were identified as potential winners: Coenagrion mercuriale and Erythromma lindenii. The ten Odonata species predicted as the most resilient were all potential winners according to their thermal preferences. Some Odonata species showed large differences in scores according to the five criteria that made up the resilience index (Appendix E, maximum standard deviation = 0.42). Different degrees of resilience were also found between the groups of winners and losers (Figure 3). The cold thermal specialists (potential losers) had a lower resilience than other groups (MannWhitney p-values < 0.05), ranging from 0.19 to 0.41. The two most threatened cold thermal specialists are Coenagrion hastulatum and Leucorrhinia dubia. The cool thermal specialists detected as potential losers had a slightly higher resilience than cold thermal specialists (Mann-Whitney pvalues = 0.005), ranging from 0.32 to 0.56. The two most threatened cool thermal specialists are Erythromma najas and Leucorrhinia caudalis. The thermal generalists (potential winners) had the highest resilience (Mann-Whitney p-values < 0.05), ranging from 0.4 to 0.73. The warm thermal specialists also identified as potential winners had a resilience not significantly different from the 137 cold and cool thermal specialists (potential losers) (Mann-Whitney p-values = 0.14 and 0.43 respectively). The species which were not classified in a particular group according to their thermal preferences (unknown status) showed an intermediate degree of resilience. Figure 3: Values of the resilience index for Odonata species inhabiting ponds in Switzerland, based on five ecological and biogeographical criteria (dispersal ability, degree of habitat specialisation, geographic extent in the study area, future trend in geographical extent and future trend of habitat availability for species). The species are separated according to their thermal preferences into four groups of potential winners and losers and one group of unknown status (see main text and Figure 1): the cold thermal specialists (potential losers), the cool thermal specialists (potential losers), the warm thermal specialists (potential winners) and the thermal generalists (potential winners). 3.2.2 Amphibia As with Odonata, the values of resilience obtained for Amphibia differed largely among the species, allowing to prioritise the species of each group of winners or losers (Figure 4). Among the five least resilient Amphibia species, only one of them, Triturus cristatus, was already detected as at risk of extinction according to its thermal preferences. Among the five most resilient Amphibia two were already detected as potential winners according to their thermal preferences: Bufo bufo and Bombina variegata. Some Amphibia species showed large differences in scores according to the five criteria that made up the resilience index (Appendix F, maximum standard deviation = 0.52). For Amphibia, as for Odonata, differences in resilience could be observed between the different groups of winners and losers identified on the basis of their thermal preferences, although none 138 were significant (Figure 4). The cool thermal specialists detected as potential losers had a resilience ranging from 0.43 to 0.74.The two most threatened cool thermal specialists are Triturus cristatus and Pelophylax esculentus/lessonae. There was only one warm thermal specialist (potential winner) among Amphibia species, but it had the highest resilience, from 0.7. The warm thermal specialists identified as potential winners had the lowest resilience, ranging from about 0.1 to about 0.6. As with Odonata, the species which were not classified in a particular group according to their thermal preferences (unknown status) showed an intermediate degree of resilience. Figure 4: Values of the resilience index for Amphibia species inhabiting ponds in Switzerland, based on five ecological and biogeographical criteria (dispersal ability, degree of habitat specialisation, geographic extent in the study area, future trend in geographical extent and future trend of habitat availability for species). The species are separated according to their thermal preferences into four groups of potential winners and losers and one group of unknown status (see main text and Figure 1): the cold thermal specialists (potential losers), the cool thermal specialists (potential losers), the warm thermal specialists (potential winners) and the thermal generalists (potential winners). 3.3 Does the species sensitivity to warming express a different threat than the Red List status? Among the ten least resilient Odonata species, and therefore the most at risk under the warming pressure, only four were classified on the Red List of the study area (Appendix E); six were not considered as threatened. Moreover, the resilience index for the 54 Odonata species was significantly different from the threat classification from the Red List (Wilcoxon signed-ranks test, p < 0.0001), evidencing that the current Red List of Odonata does not reflect the potential threat from climate warming. 139 For Amphibia, the five least resilient species were all threatened according to the Red List of the study area, but they do not have the highest degree of threat (Appendix F). Moreover, the resilience index for the 17 Amphibia species was significantly different from the threat classification from the Red List (Wilcoxon signed-ranks test, p < 0.0001), evidencing, as for Odonata, the current Red List of Amphibia does not reflect the potential threat from climate warming. 140 4. DISCUSSION 4.1 The potential winners and losers in Swiss ponds A set of potential losers linked directly to their thermal preferences – the cold and the cool thermal specialists - were identified in the study ponds. 8% of aquatic vascular plants (22 species), 11% of Coleoptera (13 species), 18% of Amphibia (3 species) and 33% of Odonata species (18 species) are at risk of extinction according to their thermal preferences. Concerning Gastropoda, none of the species is a potential loser according to its thermal preferences. The predicted extinction of cold thermal specialists is of particular concern for mountainous areas of Switzerland (more than above 900 m a.s.l.). The predicted extinction of cool thermal specialists would also affect mountainous areas, but over a longer timescale than the extinction of cold thermal specialists. Indeed, cool thermal specialists are currently found in areas of low and middle altitude in Switzerland (less than above 900 m a.s.l.). So these species would in a first time move upwards and northwards to colonise higher altitude or latitude habitats, and only then become extinct from these habitats. The proportion of Odonata and Coleoptera species at risk of extinction according to their thermal preferences is in the mid-range of other estimates for different groups of freshwater and terrestrial insects (e.g. Thomas et al. 2004, Durance and Ormerod 2007, Settele et al. 2008, Hering et al. 2009), although direct comparisons with other studies are difficult because of methodological differences. The proportion of vascular plant species at risk of extinction according to their thermal preferences is within the range given by Thomas et al. (2004) who predicted that, by 2050, between 3 and 16% of plant species in Europe would undergo an extinction event according to mid-range climate change scenarios. Thuiller et al. (2005b) reported that as much as 27 to 63% of plant species are at risk from climate-driven extinctions by 2080 in Europe, depending on the scenarios used. Warm thermal specialists and thermal generalists (potential winners) were identified in the study country. A proportion of 53% of Amphibian species (9 species) and of 63% of dragonfly species (33 species) are potential winners according to their thermal preferences. These species today occur in lowland areas, and under climate warming, these species would move upwards and northwards to colonise new habitats. In summary, the regional and local richness of the cold and cool thermal specialists (up to 33% of the regional species pools) will undergo a significant impoverishment in the future. In contrast, the warm thermal specialists and thermal generalists which are much more numerous will be a source of colonization events. The positive balance between these two categories (winners and losers) will result in a shift in community composition at the local scale which explains, for example, the 141 predicted increase in the species richness of Swiss ponds by the end of the century (Rosset et al. 2010). Thermal preferences are not the only criterion indicating the species’ sensitivity to climate warming. Taking into consideration five other ecological and biogeographic criteria linked to the resilience of a species to perturbations (dispersal ability, degree of habitat specialisation, geographic extent in the study area, future trend in geographic extent and future trend of habitat availability for species), it was possible to prioritise further the species of each group of winners and losers along a gradient of extinction risk. Indeed, the values of resilience obtained for Odonata and Amphibia differed largely among the species. Among the five least resilient Amphibia species, only one was also identified as being at risk of extinction on the basis of its thermal preferences. This corroborates that the main threats hanging over Amphibia species in Switzerland are not directly dependent on rising temperatures, but on other perturbations such as habitat loss and urbanization. These threats are as important in Switzerland (75% of investigated species are on Red Lists) as in other parts of the world (Kiesecker et al. 2001, Pounds et al. 2006). Eight of the ten least resilient Odonata species were identified as being at risk of extinction because of its low resilience to perturbations and not because of its thermal preferences. This suggests that the extinction risk for Odonata, in contrast with Amphibia, is mostly dependent on rising temperatures. Species restricted to small temperature ranges had the lowest resilience and thermal generalists had a higher resilience. This reveals that species restricted to cold or cool temperatures (cold and cool thermal specialists) will doubly suffer from climate warming, on the one hand because of their restricted thermal preferences and on the other hand because of their low resilience to perturbations such as warming. 4.2 Benefits and limitations of the methodological approach There is evidence that pond communities within a country covering a large range of temperatures, such as Switzerland, could act as a good model to identify potential winners and losers. Our results demonstrate that, in the context of climate warming, some of the potential winners and losers could be detected through an assessment of the thermal preferences of species. In addition, we showed here that an approach based solely on thermal preferences should be coupled with an analysis of criteria linked to the resilience of species, such as dispersal ability, degree of habitat specialisation, or geographic extent. The high standard deviations between the different criteria used in the resilience 142 index developed here illustrate the importance of using a combination of different criteria to properly assess the sensitivity of individual species to climate warming. The resilience index could also be improved by weighting the different criteria included according to their ecological importance for each taxonomic group. 4.3 Additional winners and losers The species included in the present study were those in the current Swiss species pool. However, southern species associated with warmer climates and currently recorded in neighbouring countries (e.g. Mediterranean species) are another group of potential winners. Indeed, because of the high dispersal abilities of many Mediterranean species (e.g. macroinvertebrates in Bonada et al. (2007)), several are likely to move northwards and successfully colonise lowland Swiss ponds. For example, five southern Odonata species have already colonized Switzerland in the last 20 years: Sympetrum meridionale, Sympetrum fonscolombii, Crocothemis erythrea, Coenagrion scitulum and Aeshna affinis. Other European Odonata species with a distribution range close to Switzerland could potentially migrate to Switzerland in the future, such as Trithemis annulata and Coenagrion caerulescens (see European distribution maps in Dijkstra & Lewington (2006)). Among Amphibia, one species, Pelodytes punctatus, could also migrate from its present natural populations located near the western boundary of Switzerland (B. Schmidt, personal communication, 2009). There are similar examples for plant species: as an example Ludwigia peploides is moving northwards along the Rhone River (G. Bornette, personal communication, 2009) and could potentially colonise Swiss ponds in the future. Among water beetles, there are also potential candidates for northwards expansions among the western Mediterranean species (D. Bilton, personal communication, 2009). In contrast, there is still no evidence of an expansion of aquatic snails from the South (E. Castella, personal communication, 2009). It is important to identify which species are expanding their range because these could potentially have a negative effect on existing pond communities, as it is already the case with the aquatic plant L. peploides. Some of the colonising species could even have public health impacts: for example the mosquito species responsible for the transmission of West Nile virus, Culex modestus, is currently expanding in French ponds (Pradel et al. 2008) and has already colonised part of Switzerland. Even though this is not necessarily directly related to climate warming, exotic species, such as the vascular plant Ludwigia grandiflora and the aquatic snail Gyraulus parvus, are also expanding their range in Swiss ponds. Biological interactions could provoke cascade events which could lead to “secondary losers”: extinctions due to the extinction of other species (Gaston and Spicer 2004), disruptions due to the 143 invasion of exotic species (Rahel and Olden 2008), “trapping” of organisms because of their evolutionary responses to formerly reliable cues which shifted in phenology (Schlaepfer et al. 2002), or increases in disease vectors (Pounds et al. 2006). It is important to consider these biotic interactions, especially for habitat specialists (Preston et al. 2008). Such cascades have been investigated in several case-studies (e.g. Pounds et al. 2006), but further research is needed to identify all the species likely to be affected by such cascade events at a national or regional scale. 4.4 Implications for conservation The current Red List categories do not reflect the potential threat from climate warming. Therefore, the Red List could benefit from including the likely effects of changes in climate in their assessment. In order to promote the conservation of species which are not already threatened, but which will highly suffer from future changes in climate, it would be important to implement a complementary label indicating the degree of sensitivity to warming based for example on the index developed here. None of the species identified as potential losers are endemic to Switzerland. At the European scale, most of these species have large populations inhabiting boreal and arctic regions, and could persist there even if they would undergo a southern contraction of their distribution. The study country (Switzerland) has, however, a responsibility for their conservation, since for a large proportion of these species the study country represents one of their last locations in central and southern Europe. The use of basic ecological and biogeographical knowledge is valuable for conservation management because it allows one to indicate the degree to which each species will be affected by climate warming. This should facilitate the prioritisation of conservation efforts, for example by putting in place action plans for the most sensitive species. Moreover, management measures should also consider species requirements for migration and colonisation of new habitats. Indeed, they should target the habitats, aiming to improve habitat connectivity (corridors and networks), as suggested by the reviews of Heller and Zavaleta (2009) and Lawler (2009). 144 5. CONCLUSIONS The present study showed that some of the winners and losers of a warming climate could be quantified through their thermal preferences. For pond ecosystems in Switzerland, a large set of potential losers were identified, ranging from zero species to 33% of the regional species pool according to taxonomic group and represented by the cold and the cool thermal specialists. A large set of potential winners was also identified, comprising from 53% to 61% of the regional species pool depending on the taxonomic group and represented by the thermal generalists and the warm thermal specialists. This approach was greatly enhanced by an analysis of ecological and biogeographical criteria linked to the resilience of individual species to perturbations (and therefore to warming), such as dispersal ability, degree of habitat specialisation or geographical extent, which allowed to prioritise further the species of each group of potential winners and losers along a gradient of extinction risk. This potential threat from climate warming is not reflected by the current Red Lists. Therefore, conservation efforts could gain from a complementary label indicating the degree of sensitivity to warming. 145 ACKNOWLEDGMENTS We would like to thank the Swiss Centre for fauna cartography (CSCF) and the Swiss Centre of floristic network (CRSF) for providing occurrence data on the altitudinal distributions of species in Switzerland. We thank S. Angélibert, A.Lehmann and R. Maggini for their valuable comments. We thank D. Auderset, D. Bilton, G. Bornette, G. Carron, E. Castella, P. Prunier and B. Schmidt for their expertise on specific taxonomic groups. We would like to thank C. Deliry for providing expertise concerning the habitat affinity of Odonata (Appendix A) and Y. Gonseth for providing temporal trends on the geographic extent of Odonata in Switzerland. We also would like to thank P. Nicolet from the Pond Conservation and J. Park for editing the English. 146 APPENDIX A Affinity scores for 54 Odonata Swiss pond species with 20 types of freshwater habitats (“3” is low and “1” is high), 3 1 1 3 3 3 2 2 3 2 2 3 3 3 1 3 3 1 3 3 3 3 3 3 3 3 3 3 3 3 3 2 2 2 3 2 2 3 3 3 3 3 3 3 2 3 2 3 3 3 3 2 2 3 2 3 3 3 2 3 3 3 1 3 3 3 3 3 3 3 2 2 3 3 3 large "natural" open ponds large "natural" forested ponds lowland marshland, alkaline peat bogs lowland acidic peatbogs 2 3 3 2 3 1 2 3 3 2 3 2 2 2 3 2 3 3 3 2 3 2 3 3 2 3 2 2 1 3 3 2 3 3 2 3 1 3 3 1 1 3 3 3 3 3 3 2 3 3 3 2 2 2 3 2 2 3 3 3 2 3 3 3 2 2 3 3 2 3 2 3 3 3 3 2 2 1 1 3 3 2 3 3 3 3 3 3 2 3 3 3 3 3 3 3 3 2 3 3 2 2 2 3 3 2 2 2 3 3 3 2 2 3 3 2 3 3 3 3 3 2 2 2 3 3 3 147 3 3 1 1 1 3 3 3 1 3 3 3 3 3 2 3 2 3 3 3 2 2 3 3 1 2 2 3 2 3 3 3 2 2 3 2 2 3 3 2 3 3 2 3 2 3 3 3 3 3 2 3 2 2 2 3 2 3 3 3 2 3 1 2 2 3 3 3 2 1 2 3 3 3 3 3 1 2 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 3 3 3 2 3 1 1 3 3 2 3 3 3 3 3 3 2 3 3 2 3 2 2 3 2 3 2 2 1 3 3 2 2 3 3 3 2 2 3 3 3 2 3 3 3 2 2 2 3 3 3 3 3 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 1 1 3 3 3 2 2 3 3 2 2 3 3 3 3 2 2 2 1 3 2 2 3 3 3 3 2 2 2 3 1 1 3 3 1 1 3 3 3 2 3 3 3 3 3 3 3 3 3 3 2 3 3 3 2 3 3 3 3 3 3 3 3 3 3 2 3 3 3 2 3 3 3 3 3 3 3 2 3 3 3 2 2 1 lowland and middle altitude lakes artificial ponds salty-water systems 3 3 1 3 3 3 2 2 2 2 3 2 3 2 2 2 3 2 3 3 2 2 2 3 ponds and marshland of altitude 3 3 3 3 3 acidic peat bogs of altitude 3 temporary systems 2 3 3 seepages, resurgences large streams (calm parts) large streams (flowing) flowing streams 3 3 3 3 small forested ponds caerulea cyanea grandis isosceles juncea mixta subarctica imperator parthenope pratense lindenii tenellum hastulatum mercuriale puella pulchellum aenea erythraea cyathigerum bimaculata najas viridulum pulchellus elegans pumilio dryas sponsa virens viridis caudalis dubia pectoralis depressa fulva quadrimaculata albistylum brunneum cancellatum coerulescens pennipes nymphula alpestris arctica flavomaculata metallica fusca paedisca danae depressiusculum flaveolum pedemontanum sanguineum striolatum vulgatum small open ponds Aeshna Aeshna Aeshna Aeshna Aeshna Aeshna Aeshna Anax Anax Brachytron Cercion Ceriagrion Coenagrion Coenagrion Coenagrion Coenagrion Cordulia Crocothemis Enallagma Epitheca Erythromma Erythromma Gomphus Ischnura Ischnura Lestes Lestes Lestes Lestes Leucorrhinia Leucorrhinia Leucorrhinia Libellula Libellula Libellula Orthetrum Orthetrum Orthetrum Orthetrum Platycnemis Pyrrhosoma Somatochlora Somatochlora Somatochlora Somatochlora Sympecma Sympecma Sympetrum Sympetrum Sympetrum Sympetrum Sympetrum Sympetrum Sympetrum flowing ditches, irrigation channels Species waterway channels Genus rivulets and brooks spring areas adapted from Dommanget (1998) through local expert advice (C. Deliry, personal communication, 2009). 3 3 3 3 2 2 2 3 3 3 3 3 2 2 2 3 3 2 3 3 2 3 3 APPENDIX B Affinity scores for 17 Amphibia Swiss pond species with 8 types of freshwater habitats (“3” is low and “1” is high), data 2 2 3 3 1 3 3 1 3 2 3 artificial ponds peatbogs marhsland and wet meadows streams (calm parts) brooks, flowing streams spring areas, ditches, rivulets lakes small and large ponds from Meyer et al. (2009). Alytes obstetricans 1 1 Bombina variegata 1 Bufo bufo 1 1 Bufo calamita 1 3 Bufo viridis 1 Hyla arborea 1 1 2 1 Hyla intermedia 1 1 2 3 Lissotriton helveticus 1 2 1 3 3 Lissotriton vulgaris 1 3 1 2 3 Mesotriton alpestris 1 2 1 2 3 1 Pelophylax esculenta/lessonae 1 1 1 2 3 2 Pelophylax ridibunda 2 1 2 2 Rana dalmatina 1 1 2 Rana latastei 3 3 1 3 Rana temporaria 1 3 2 2 Triturus carnifex 1 1 2 3 Triturus cristatus 1 1 2 2 1 3 2 1 1 3 3 1 2 148 3 3 2 APPENDIX C Present and future abundance of 20 freshwater habitat types in Switzerland (0 corresponds to absence and 5 to high abundance), and their present and future expected proportion to remain in an unimpaired state (able to host Odonata and Amphibia). Scoring based on expert assessment (B. Oertli and V. Rosset) informed by national hydrological atlas and 149 2 5 1 0.8 3 5 1.5 0.8 lowland and middle altitude lakes ponds and marshland of altitude * lowland acidic peatbogs acidic peat bogs of altitude large "natural" forested ponds 4 1 4 1 lowland marshland, alkaline peat bogs 0 0 0 0 artificial ponds 4 1 4 1 large "natural" open ponds small forested ponds 5 1 1 5 0.2 0.5 0.5 0.8 5 1 2 5 0.4 0.4 2 0.8 salty-water systems * small open ponds temporary systems * 1 1 1 1 seepages, resurgences waterway channels * 2 1 2 1 flowing ditches, irrigation channels large streams (flowing) 1 5 4 3 0.5 0.6 0.5 0.3 1 5 4 3 0.4 0.8 0.8 0.5 large streams (calm parts) flowing streams spring areas Present habitat abundance aph Present proportion in a unimpaired state wph Future expected habitat abundance afh Future proportion in a unimpaired state wf h rivulets and brooks inventories. Habitats with a star were used only for Odonata. 2 0.5 1 5 5 0.8 1 1 0.9 0.4 2 0.5 1 5 5 0.8 1 0.5 0.9 0.4 APPENDIX D Species of vascular plants, Gastropoda and Coleoptera inhabiting Swiss ponds identified as cold thermal specialists (potential losers in a context of climate warming) using altitudinal distribution as a surrogate for temperature. Cold thermal specialists Vascular plants Arabis subcoriacea, Cardamine amara, Carex frigida, Carex juncella, Carex nigra, Carex norvegica, Carex paupercula, Cochlearia pyrenaica, Dactylorhiza cruenta, Epilobium alsinfolium, Epilobium nutans, Eriophorum angustifolium, Eriophorum scheuchzeri, Isoëtes echinospora, Isoëtes lacustris, Juncus filiformis, Montia fontana, Ranunculus reptans, Rorippa islandica, Saxifraga stellaris, Sparganium angustifolium and Vaccinium microcarpum Gastropoda None Coleoptera Agabus laponicus, Agabus melanarius, Hydroporus foveolatus, Hydroporus incognitus, Hydroporus marginatus, Hydroporus melanarius, Hydroporus memnonius, Hydroporus nigellus, Hydroporus nigrita, Hydroporus obscurus, Hydroporus sabaudus, Ilybius erichsoni and Stictotarsus griseostriatus 150 APPENDIX E Thermal preferences, threat levels according to the Swiss Red List, values of five ecological and biogeographical criteria (dispersal ability, degree of habitat specialisation, geographic extent in the study area, future trend in geographical extent and future trend of habitat availability for species) and multimetric index of resilience based on the five criteria for the 54 Odonata species inhabiting ponds in Switzerland. Species are regrouped by thermal preferences and then classified along a resilience gradient to perturbations, from less resilient (low values) to most resilient (high values). For details on the calculation of the five criteria, see Methods. NA corresponds to values not available because of lack of Degree of habitat specialisation Geographic extent in the studied area Future expected trend in the geographic extent Resilience index (mean) Standard deviation of the resilience index Expected changes in the hosting potential of the studied area Species Coenagrion hastulatum cold 0,12 0,10 0,29 0,36 0,06 0,19 0,13 Leucorrhinia dubia cold 0,33 0,00 0,37 0,41 0,11 0,24 0,18 Aeshna subartica cold VU 0,82 0,00 0,07 NA 0,11 0,25 0,38 Aeshna caerulea cold VU 0,76 0,03 0,24 0,29 0,00 0,26 0,30 Somatochlora artica cold 0,48 0,08 0,30 0,43 0,12 0,28 0,18 Somatochlora alpestris cold 0,55 0,00 0,38 0,67 0,11 0,34 0,28 Aeshna juncea cold 0,85 0,03 0,64 0,44 0,08 0,41 0,36 Erythromma najas cool 0,21 0,46 0,35 0,26 0,29 0,32 0,09 Leucorrhinia caudalis cool CR 0,52 0,31 0,05 NA 0,51 0,35 0,22 Epitheca bimaculata cool CR 0,73 0,10 0,09 NA 0,56 0,37 0,32 Sympetrum danae cool 0,24 0,33 0,53 0,13 0,64 0,37 0,21 Sympecma paedisca cool CR 0,18 0,28 0,06 NA 0,99 0,38 0,42 Leucorrhinia pectoralis cool CR 0,55 0,38 0,12 NA 0,49 0,39 0,19 Sympetrum vulgatum cool 0,36 0,46 0,51 0,21 0,67 0,44 0,17 Somatochlora flavomaculata cool 0,61 0,46 0,32 0,30 0,63 0,46 0,15 Somatochlora metallica cool 0,67 0,51 0,52 0,38 0,50 0,52 0,10 Cordulia aenea cool 0,52 0,64 0,52 0,56 0,52 0,55 0,05 Aeshna grandis cool 0,88 0,72 0,51 0,26 0,45 0,56 0,24 Orthetrum albistylum - EN 0,64 0,59 0,21 0,02 0,66 0,42 0,29 Sympetrum pedemontanum - CR 0,27 0,62 0,28 NA 0,72 0,47 0,23 Thermal preferences Genus Dispersal ability Species threatenend according to the Swiss Red List knowledge. 151 Sympetrum depressiusculum - Coenagrion pulchellum generalist Lestes virens generalist Sympecma fusca Lestes VU 0,36 0,79 0,33 0,28 0,86 0,53 0,28 0,12 0,85 0,39 0,19 0,44 0,40 0,28 0,21 0,49 0,13 NA 0,78 0,40 0,30 generalist 0,18 0,44 0,40 0,19 0,84 0,41 0,27 sponsa generalist 0,15 0,67 0,49 0,00 0,78 0,42 0,33 Erythromma viridulum generalist 0,12 0,46 0,35 0,21 1,00 0,43 0,35 Ischnura pumillo generalist 0,06 0,77 0,55 0,31 0,60 0,46 0,28 Sympetrum flaveolum generalist 0,33 0,41 0,34 NA 0,78 0,47 0,21 Orthetrum brunneum generalist 0,64 0,56 0,43 0,29 0,54 0,49 0,13 Lestes viridis generalist 0,33 0,85 0,48 0,16 0,65 0,49 0,27 Lestes dryas generalist 0,24 0,44 0,56 NA 0,75 0,50 0,21 Platycnemis pennipes generalist 0,21 0,85 0,44 0,42 0,58 0,50 0,23 Enallagma cyathigerum generalist 0,09 0,67 0,66 0,29 0,84 0,51 0,31 Brachytron pratense generalist 0,67 0,49 0,23 0,50 0,70 0,51 0,19 Coenagrion puella generalist 0,09 0,69 0,69 0,36 0,75 0,52 0,28 Ischnura elegans generalist 0,06 1,00 0,56 0,33 0,73 0,54 0,36 Aeshna mixta generalist 0,76 0,46 0,41 0,25 0,85 0,55 0,25 Sympetrum sanguineum generalist 0,33 0,85 0,48 0,33 0,79 0,56 0,25 Pyrrhosoma nymphula generalist 0,21 0,77 0,57 0,47 0,76 0,56 0,23 Aeshna isosceles generalist 0,82 0,41 0,28 0,74 0,59 0,57 0,22 Orthetrum coerulescens generalist 0,48 0,85 0,34 0,41 0,82 0,58 0,24 Sympetrum striolatum generalist 0,36 0,97 0,60 0,47 0,64 0,61 0,23 Anax parthenope generalist 0,97 0,38 0,36 NA 0,73 0,61 0,29 Libellula depressa generalist 0,61 0,95 0,64 0,34 0,61 0,63 0,22 Libellula quadrimaculata generalist 0,61 0,69 0,62 0,55 0,77 0,65 0,09 Orthetrum cancellatum generalist 0,70 0,85 0,53 0,41 0,77 0,65 0,18 Aeshna cyanea generalist 0,94 0,59 0,69 0,29 0,78 0,66 0,24 Libellula fulva generalist 0,61 0,77 0,32 1,00 0,61 0,66 0,25 Anax imperator generalist 1,00 0,77 0,62 0,50 0,76 0,73 0,19 Coenagrion mercuriale warm 0,00 0,18 0,15 NA 0,65 0,24 0,28 Erythromma lindenii warm 0,21 0,36 0,25 0,24 0,64 0,34 0,17 Ceriagrion tenellum warm EN 0,09 0,56 0,12 NA 0,65 0,35 0,29 Gomphus pulchellus warm VU 0,45 0,49 0,27 0,20 0,66 0,41 0,18 Crocothemis erythrea warm 0,33 0,72 0,37 0,93 0,66 0,60 0,25 CR EN CR CR Abbreviations: IUCN species threat status: VU vulnerable, EN endangered, CR critically endangered. * Species of Least Concern (LC) and Nearly Threatened (NT) are not considered as threatened. 152 APPENDIX F Thermal preferences, threat levels according to the Swiss Red List, values of five ecological and biogeographical criteria (dispersal ability, habitat specialisation, geographic extent in the study area, future trend in geographical extent and future trend of habitat availability for species) and a multimetric index of resilience based on the five criteria for the 17 Amphibia species inhabiting ponds in Switzerland. Species are regrouped by thermal preferences and then classified along a resilience gradient to perturbations, from less resilient (top of table, low values) to most resilient (bottom of table, high values). For details on the calculation of the five criteria, see Methods. NA corresponds to values not available Habitat specialisation Geographic extent in the studied area Future expected trend in the geographic extent Expected changes in the hosting potential of the studied area Resilience index (mean) Standard deviation of the resilience index 0,22 0,41 0,47 0,77 0,43 0,22 cool 1,00 0,39 0,61 0,66 0,36 0,60 0,26 cool 0,39 0,61 1,00 1,00 0,69 0,74 0,26 EN 0,02 0,39 0,44 0,13 0,71 0,34 0,27 - EN 0,00 0,35 0,46 0,61 0,92 0,47 0,34 calamita - EN 0,82 0,39 0,37 0,60 0,24 0,48 0,23 Mesotriton alpestris - 0,32 0,52 0,91 0,76 0,27 0,56 0,27 Lissotriton helveticus - 0,69 0,39 0,50 0,55 0,73 0,57 0,14 Pelophylax ridibundus - 0,73 0,52 0,39 NA 0,95 0,65 0,24 Bufo bufo Rana Genus Species Triturus cristatus cool Pelophylax esculentus/lessonae Rana temporaria Lissotriton vulgaris - Alytes obstetricans Bufo Species threatenend according to the Swiss Red List 0,28 Thermal preferences Dispersal ability because of a lack of knowledge. EN generalist VU 0,72 0,57 0,66 0,80 0,76 0,70 0,09 dalmatina warm EN 0,28 0,13 0,15 0,00 0,00 0,11 0,12 Rana latastei warm VU 0,18 0,43 0,02 NA 0,55 0,30 0,24 Triturus carnifex warm EN 0,28 0,26 0,08 0,51 1,00 0,43 0,36 Bufo viridis warm RE 1,00 0,04 0,01 NA 0,83 0,47 0,52 Hyla arborea warm EN 1,00 0,17 0,39 0,60 0,45 0,52 0,31 Hyla intermedia warm EN 0,59 0,26 0,06 0,71 1,00 0,52 0,37 Bombina variegata warm EN 0,29 0,35 0,56 0,74 0,95 0,58 0,27 Abbreviations: VU vulnerable, EN endangered, RE regionallyextinct. * Species of Least Concern (LC) and Nearly Threatened (NT) are not considered as threatened. 153 154 3.5. Synthèse concernant l’impact du réchauffement climatique sur la biodiversité des étangs Les conclusions principales apportées dans ce chapitre concernant l’impact du réchauffement climatique sur la biodiversité des étangs de Suisse sont : L’augmentation de température devrait entraîner une augmentation significative de la richesse spécifique locale des étangs (Articles 3 et 4), ce qui confirme l’hypothèse H1. Cette augmentation de richesse serait particulièrement marquée dans les étangs d’altitude (actuellement pauvres en espèces) en comparaison avec les étangs de plaine (Article 4). L’amplitude de l’augmentation de richesse serait hétérogène, variant selon le groupe taxonomique considéré. Elle sera plus grande pour les odonates adultes, les amphibiens et les coléoptères que les gastéropodes et les macrophytes (Article 4). Dans le pool régional suisse, la proportion d’espèces à risque d’extinction est plus faible que la proportion d’espèces qui vont bénéficier du réchauffement climatique (Article 5), ce qui confirme l’hypothèse H2. Cette situation devrait entraîner une augmentation de la richesse régionale. De plus, ceci corrobore l’augmentation de richesse locale prédite dans les Articles 3 et 4. Les espèces menacées par le réchauffement climatique mises en évidence dans l’Article 5 ne sont actuellement pas toutes sur Liste Rouge. Un indice de sensibilité au réchauffement climatique serait donc un outil de conservation de la biodiversité utile. 155 3.6. Discussion complémentaire Quelques éléments de discussion complémentaire à ceux développés jusqu’à présent dans ce chapitre peuvent être apportés, principalement concernant les méthodes de modélisation utilisées, l’incertitude sur l’ampleur des augmentations de richesse spécifique locale prédites et l’impact du réchauffement climatique sur la richesse spécifique régionale. Méthodes de modélisation Bien que les incertitudes quant aux prédictions effectuées dans ce Chapitre (Article 4) aient été réduites au maximum, d’autres méthodes de modélisation que les Modèles Additifs Généralisés (GAM) pourraient encore affiner les tendances prédites dans l’Article 4. Par exemple, la technique de modélisation des « multivariate adaptive regression splines » (MARS) pourrait permettre de prendre en compte les interactions entre prédicteurs (Moisen and Frescino 2002). L’incorporation d’autres paramètres tels que la vitesse de colonisation des espèces dans les modèles, comme cela a été fait par exemple pour simuler la migration des plantes sous l’effet du réchauffement climatique (Randin 2007), pourrait également améliorer le réalisme des prédictions. Incertitude sur l’ampleur des augmentations de richesse spécifique locale prédites Si la tendance d’une augmentation de la richesse spécifique sous l’effet du réchauffement climatique est certaine, le réalisme de l’ampleur de ces augmentations mérite d’être discuté. La question de savoir si des augmentations aussi importantes (+83% en plaine et +150% en altitude) pourront se produire dépend de la capacité des écosystèmes à accueillir de nouvelles espèces, c’est-à-dire à ne pas être « saturés ». En effet, la richesse d’un écosystème ne pourra pas augmenter si ses peuplements sont « saturés », c’est-à-dire si l’écosystème ne peut pas accueillir d’espèces supplémentaires. Cette question a déjà été discutée brièvement (Article 4) mais sans entrer dans le détail des résultats sur lesquels cette discussion se basait. Le test de la saturation des étangs a été effectué à l’aide des relations entre richesse spécifique locale et richesse spécifique régionale (Gaston and Blackburn 2000). La richesse spécifique régionale a été calculée en regroupant les étangs par tranche d’altitude de 200 m, exception faite des altitudes très basses et très élevées à cause de la faible quantité de données disponible. En effet, toutes les altitudes inférieures à 500m ont été regroupées en une région et toutes les altitudes supérieures à 2500m ont été regroupées en une région. Les écosystèmes non-saturés sont définis comme ceux pour lesquels l’augmentation de 156 richesse locale est proportionnelle à la richesse de la région dans laquelle ils se trouvent (relation linéaire) et les écosystèmes saturés sont définis comme ceux dont la richesse locale augmente avec la richesse régionale jusqu’à un niveau maximum (relation logarithmique). D’après Gaston et Blackburn (2000), la richesse des peuplements d’un écosystème saturé serait limitée par les interactions locales, qui forceraient la richesse locale à rester à un niveau plus bas que possible d’après le pool d’espèces régional, alors que la richesse des peuplements d’un écosystème nonsaturé ne serait pas limitée par les interactions locales. La saturation des étangs a été évaluée ici en comparant les paramètres R2 des deux différentes régressions (linéaire et logarithmique). Ce test a montré que les régressions linéaires sont plus fiables que les régressions logarithmiques pour tous les groupes taxonomiques étudiés, suggérant que les étangs étudiés sont non-saturés (Figure 11). Néanmoins, les différences entre les R2 de ces deux types de régressions étant faibles pour la plupart des groupes (Figure 11), on ne peut pas conclure avec certitude que les étangs étudiés sont non-saturés et que l’augmentation de richesse régionale peut potentiellement provoquer une augmentation de la richesse à l’échelle locale de l’amplitude de celle prédite dans l’Article 3 et l’Article 4. 157 Figure 11 : Régressions linéaires et logarithmiques entre la richesse spécifique locale et la richesse spécifique régionale de 113 étangs de Suisse pour cinq groupes taxonomiques différents. 158 Impact du réchauffement climatique sur la richesse spécifique régionale Une question importante n’a pas encore été discutée à la fin de ce Chapitre 3: les changements de composition (extinctions et colonisations) décrits dans l’Article 5 vont provoquer une augmentation de la richesse régionale suisse, mais de quelle amplitude ? La proportion d’espèces à risque d’extinction plus faible que celle des espèces potentiellement colonisatrices suggère que la richesse régionale augmente. Cependant, l’amplitude de cette augmentation ne dépendra pas seulement des proportions de perdants et de gagnants identifiés dans l’Article 5, mais également du nombre d’espèces colonisant la Suisse depuis le Sud. Bien qu’on ait déjà observé certaines espèces du Sud migrer en Suisse (exemples dans l’Article 5), leur nombre exact n’est pas connu à l’heure actuelle. L’amplitude finale de l’augmentation de la richesse régionale suisse ne peut donc pas être prédite avec certitude. 159 160 Chapitre 4 Impact de l’eutrophisation sur la biodiversité des étangs 4. Impact de l’eutrophisation sur la biodiversité des étangs 4.1. De la problématique aux hypothèses Comme déjà décrit en Introduction, l’eutrophisation est un problème majeur pour les écosystèmes d’eau douce, dont notamment les étangs. Ce phénomène entraîne une augmentation de la productivité primaire. La relation entre productivité primaire et richesse spécifique est largement connue dans les écosystèmes terrestres et montre des différences de forme et de force selon l’échelle spatiale, le groupe taxonomique ou le type d’écosystème considéré. La question est moins connue dans les écosystèmes aquatiques. L’état des connaissances actuelles suggère néanmoins que cette relation consiste en une courbe en cloche. Concernant les étangs en particulier, une même réponse a été mise en évidence pour quelques groupes taxonomiques, mais on ne sait pas si cette réponse est valable pour tous les groupes taxonomiques. De plus, aucune étude n’a déterminé à l’heure actuelle si la valeur de conservation, facette de la biodiversité importante car elle considère le degré de menace des espèces, suit la même courbe en cloche que la richesse spécifique. En conséquence, l’objectif principal de ce chapitre (Article 6) est d’évaluer, à l’échelle locale, la relation entre charge trophique et biodiversité (richesse et valeur de conservation) dans les étangs de plaine de Suisse et du plateau de la Dombes en France pour différents groupes taxonomiques. La forte charge trophique de ces étangs les place à la fin de la courbe en cloche. L’hypothèse testée est l’hypothèse H3 : à l’échelle locale, en plaine, la biodiversité (richesse taxonomique et valeur de conservation) diminue linéairement sous l’effet de l’augmentation de la charge trophique. Un objectif complémentaire de ce chapitre est de simuler l’impact d’une augmentation de la charge trophique due au réchauffement climatique sur la richesse et la valeur de conservation des étangs. Les conclusions principales de l’Article 6 sont rappelées au Chapitre 4.3. 161 162 Article 6 The relationship between nutrient load and biodiversity in lowland ponds and small lakes: implications in a changing climate context. Rosset V.1, Angélibert S.1, Arthaud F.2,3, Bornette, G.3, Vallod D.2, Oertli B.1 This manuscript will be submitted. 1 University of Applied Sciences Western Switzerland, hepia Geneva technology, architecture and landscape, 1254 Jussy-Geneva, Switzerland; 2 ISARA-Lyon, Ecosystems and Aquatic Resources, 69364 Lyon, France; 3 UMR CNRS 5023 LEHNA “Ecology of natural and anthropized hydrosystems”, Lyon 1, Lyon University, 69622 Villeurbanne, France. 163 164 KEYWORDS Freshwaters, nutrient load, species richness, macrophytes, dragonflies, macroinvertebrates, amphibians ABSTRACT 1. The global changes occurring worldwide are leading to an increase in nutrient load and primary productivity in freshwater ecosystems. In particular, the warming climate and the land-use changes will increase the eutrophication and its impact on living communities. Our knowledge on the relationships between the eutrophication of freshwater systems and the biodiversity (species richness and threatened species) suggests a hump-shaped pattern between primary productivity and species richness. Nevertheless, large variations in the strength and pattern of the relationship are commonly reported, and the relevance of the hump-shaped pattern remains to be confirmed for small waterbodies as ponds and small lakes. Furthermore, it is still not demonstrated if the threatened species, measured here through the conservation value of assemblages, respond in the same way than the species richness to nutrient load increases. 2. The present study investigated, at the local scale (ecosystem), the relationship between primary productivity and both taxonomic richness and conservation value from temperate lowland ponds and small lakes. We hypothesized that in these nutrient-rich waterbodies, richness and conservation value will decrease with increasing nutrient load. Then, based on these results, we forecasted the potential impact on biodiversity of an increase in nutrient load due to climate change. 3. For this purpose, the relationship between nutrient load and both richness and conservation value was investigated for four different taxonomic groups: macrophytes, aquatic macroinvertebrates, adult dragonflies and amphibians, in two data sets from Switzerland and Eastern France (55 and 82 waterbodies), covering a nutrient gradient mainly from eutrophic to hypereutrophic conditions. Richness was assessed at species level for macrophytes, dragonflies and amphibians, at genus level for two orders of macroinvertebrates (gastropods and beetles), and at family level for the whole macroinvertebrate assemblage. The conservation value of assemblages was assessed on the basis of species’ rarity level on the 165 national Red List. The relationship of both taxonomic richness and conservation value with nutrient load was investigated using robust linear regressions. The potential impact on biodiversity of an increase in nutrient load due to climate change was simulated through the increase of the nutrient values in these regressions. 4. The response of biodiversity to increasing nutrient load varied according to the taxonomic group considered, underlining that different groups can show different trends. The richness of vascular plants and macroinvertebrates decreased with the increase in nutrients, but this was not the case for gastropods and amphibians. Dragonflies and water beetles presented an intermediate situation with a decrease in richness in only one of both data sets. Moreover, discrepancies between richness and conservation value were evidenced in their response to an increase in nutrient load; this demonstrates that richness alone cannot act as a surrogate for conservation value and therefore for biodiversity. 5. The patterns revealed by the present study have strong implications in the context of climate change, underlining a potential decrease of local richness for some groups, as macrophytes and macroinvertebrates. Nevertheless all taxonomic groups will not be impacted, as for example gastropods and amphibians. 166 INTRODUCTION The global changes happening worldwide are leading to an increase in nutrient load and primary productivity in freshwater ecosystems. In particular, the warming climate and the land-use changes will increase the eutrophication and its impact on living communities (Vitousek 1994, e.g. McKee et al. 2003, Park et al. 2004, review in Hering et al. 2010). Relationships between primary productivity and taxonomic richness have been described for a wide array of ecosystems, including aquatic ones. Nevertheless, large variations in the pattern and strength of this productivity-richness relationship have been reported, depending on the spatial scale, the taxonomic group or the type of ecosystems considered (Waide et al. 1999, Dodson et al. 2000, Mittelbach et al. 2001). Compared to terrestrial systems, studies conducted in aquatic systems are underrepresented (Waide et al. 1999); the existing investigations suggest nevertheless that the hump-shaped pattern is particularly common (Waide et al. 1999, Mittelbach et al. 2001). This hump-shaped pattern was described at the local scale, for example in shallow lakes or ponds for several taxonomic groups: fish, phytoplankton, and floatingleaves macrophytes and benthic invertebrates, macrophytes and macroalgae (Jeppesen et al. 2000, Chase and Leibold 2002, Menetrey et al. 2005). Feuchtmayr et al. (2009) showed major effects of increased temperatures and nutrient additions in experimental mesocosms, such as a decrease in plant species richness, although some other mesocosm experiments found only a slight effect of these changes on the taxonomic richness of macrophytes, macroinvertebrates and zooplankton (McKee et al. 2002, McKee et al. 2003, Moss et al. 2003, Feuchtmayr et al. 2007). Globally, most investigations of aquatic ecosystems have focused on a restricted set of taxonomic groups, such as planktonic organisms and plants, and did not considered high trophic levels. Furthermore, these studies failed to include the conservation value of assemblages, a component of biodiversity of particular interest because it takes into account species composition and therefore the level of threat to, or endemism among, the species present in a community. This issue is however of particular importance, as eutrophication is a major threat of freshwater biodiversity as a whole (Brönmark and Hansson 2002, Dudgeon et al. 2006, Heino et al. 2009). Even if conservation value of freshwaters has been demonstrated to be sensitive to physico-chemical conditions of water (e.g. Simaika and Samways 2011), suggesting that changes in nutrient load would affect this conservation value, it is still not demonstrated that species richness and conservation values vary in the same way when nutrient load increases. Among freshwater ecosystems, small lakes and ponds are seriously threatened by eutrophication (Brönmark and Hansson 2002, EPCN 2007), despite the fact that they collectively support a very diverse, and sometimes unique biodiversity, often richer than those found in running waters or larger 167 lakes (e.g. Williams et al. 2004, Angelibert et al. 2006). This makes crucial to reveal the relationships between eutrophication and biodiversity in these ecosystems. Moreover, as these ecosystems are usually characterized by various levels of trophic complexity along the nutrient gradient (Jeppesen et al. 2000, Scheffer et al. 2006), they are adequate ecosystems for studying the relationships between nutrient load and biodiversity for several groups dispatched along the trophic chain. Consequently, the present study aimed to investigate, at the local scale (ecosystem), the relationship between nutrient load and both taxonomic richness and conservation value of temperate lowland ponds and small lakes, for four different taxonomic groups: macrophytes, aquatic macroinvertebrates, adult dragonflies, and amphibians. As lowland waterbodies usually support high level of nutrients (mainly from eutrophic to hypertrophic), we hypothesized that taxonomic richness and conservation value of all the taxonomic groups investigated may decrease with increasing nutrient load. This hypothesis will be tested using two sets of waterbodies: ponds and small lakes in Switzerland, ranging mainly from eutrophic to hypertrophic, and ponds in Eastern France, mainly hypertrophic. Additionally, on the basis of the relationships evidenced, the impact on biodiversity of an increase in nutrient load due to climate change will be simulated. 168 MATERIAL AND METHODS Study sites Two sets of temperate lowland waterbodies encompassing different ranges of nutrient loads were investigated: a Swiss data set ranging mainly from eutrophic to hypertrophic conditions and a French data set of mainly hypertrophic conditions. The Swiss data set consists in 55 ponds and small lakes which were studied in the context of various projects (e.g. Oertli et al. 2002, Indermuehle et al. 2010, Menetrey et al. 2010). They have a median area of 1240 m2 (6 – 58’000 m2) and a median depth of 1.3 m. They are scattered throughout lowland Switzerland at elevations ranging from 210 to 700 m. a.s.l. They are either of natural origin (glacial retreat), or artificial and linked to past or present human activities (e.g. nature conservation, recreation, gravel or clay extraction). None of them is used for fish-farming. The French data set consists of 82 ponds from South-Eastern France, studied in the context of a project aiming at assessing the ecological value of fish-ponds (e.g. Arthaud et al. in prep, Robin et al. submitted). They have a median area of 91’000 m2 (22’900 – 795’600 m2) and a median depth of 0.65 m. They are located in the Dombes region (North-East from Lyon) at about 300 m of elevation. This region comprises a very dense network of artificial ponds (about one thousand), built in the Middle Age for extensive fish farming, and still used for this activity. As the two data sets differed in several key parameters potentially ruling biodiversity (range of nutrient concentrations, geographical location, morphometry, anthropic use), we decided to keep the analyses in parallel for both data sets. Measures of biodiversity Four taxonomic groups were chosen, differing in their ecology and biology (e.g. life cycles, habitats, feeding strategies): macrophytes, aquatic macroinvertebrates (larvae and adults), adult dragonflies and amphibians (adults, subadults, larvae). In both data sets (Switzerland and France), they were sampled on the basis of the PLOCH-IBEM standardized method (Oertli et al. 2005b, Angelibert et al. 2010, Indermuehle et al. 2010). Inventories were carried out in both open water and interface between water and land. The sampling effort was proportional to pond area. Macrophytes were sampled once during summer on square plots equally distributed along transects perpendicular to the longest axis of each pond. Macroinvertebrates were sampled once during the spring or summer months with a small-framed hand-net (rectangular frame 14 × 10 cm, mesh size 0.5 mm) following a 169 stratified strategy across the dominant mesohabitats. Adult dragonflies were inventoried twice, at the end of spring and in mid-summer, in plots (10 m × 30 m) distributed in all the habitats occurring along the shore. Amphibians were inventoried two to four times in the year by means of (i) search by flashlight, (ii) identification of calls, and (iii) net dipping. Two measures of biodiversity, taxonomic richness and conservation value, were calculated at the local scale (for the whole ecosystem). Richness was assessed at species level for macrophytes, dragonflies and amphibians, at genus level for two orders of macroinvertebrates (gastropods and beetles), and at family level for the whole macroinvertebrate assemblage. When the sampling was not exhaustive according to species accumulation curves, the observed taxonomic richness was transformed by a statistical estimator (Jackknife-1, Burnham and Overton 1979) in order to estimate the true richness of the waterbody. The conservation value of assemblages was assessed for each waterbody. It was calculated for the three taxonomic groups identified at the species level: macrophytes, dragonflies and amphibians using the Csp value. The Csp value is a Swiss application (Oertli et al. 2002) of the Species Quality Score developed in the United Kingdom (e.g. Foster et al. 1989, Painter 1999, Williams et al. 2004). Species are ranked according to their degree of rarity on the national Red List in geometric progression, successively doubling from 1 (commonest species) to 32 (rarest): species of LC or DD status were given the rank of 1, species of NT status 4, species of VU status 8, species of EN status 16 and species of CR or RE status 32. The Csp value is the sum of the scores of all species present at the site divided by the number of species present in the site. Swiss Red Lists were used in Switzerland. In France, no Red Lists are currently available, so the Swiss Red Lists were used as a surrogate, because of species pool similarity. The Csp value was chosen according to the recent recommendations of Rosset et al. (submitted) and because of its potential to be grounded on the same information about species, here the same Red Lists, in the two study data sets. Measures of nutrient load An integrative index of nutrient load was calculated by combining total nitrogen and total phosphorus concentrations. Total nitrogen and total phosphorus concentrations were measured in samples of water at the point of maximal depth of each waterbody. Water sampling timing was adapted to the characteristics of each data set. Water was sampled during the whole period when 170 the pond was full of water, i.e. from the end of the winter period to the autumn period in the French data set, and during the end of the winter period in the Swiss data set. The determination of the index of nutrient status of each pond was based on classes defined by OCDE (1982) and Wetzel (1983): oligotrophic, mesotrophic, eutrophic and hypertrophic. A fifth class following the same progression was added for very high nutrient concentrations in order to fit the highest values occurring in the data sets. Each of the five classes was subdivided equally in ten subclasses. Each pond was thus classified in one of these fifty classes according to its total nitrogen concentrations and also according to its total phosphorus concentrations. The highest class of the two was retained as the nutrient status of the pond. Statistical analysis The relationship of both taxonomic richness and conservation value with nutrient load was investigated using robust linear regressions (lmrob function in the robustbase package of R (Rousseeuw et al. 2011)). When no significant robust linear regressions were evidenced, polynomial regressions were conducted to test for a potential hump-shaped pattern. Differences in the slope of the regressions were tested through an Analysis of covariance (ANCOVA). For modeling the effects of an increase in nutrient load (expected in the climate change context), two projections were made by changing the values of nutrient load in the “biodiversity-nutrient load” regressions. The two projections differed in the magnitude of the simulated increase in nutrient load: the optimistic projection consisted in a slight nutrient increase and the pessimistic projection in a high nutrient increase. The magnitude of these changes was fixed according to the present distribution of the index of nutrient status in each of the study datasets, i.e. according to the difference between two statistical parameters (median and quartiles). For the “slight nutrient increase” scenario, the magnitude of change was the difference between the median and the first quartile, and for the “high nutrient increase” scenario, it was the difference between the first and the third quartile. For each taxonomic group separately, the differences between the current biodiversity and the projected biodiversity under the two scenarios were tested through Mann-Whitney nonparametric tests. 171 RESULTS Response of taxonomic richness to nutrient load The response of taxonomic richness to an increase in nutrient load varied according to the taxonomic group considered (Figure 1). Macrophyte species richness and invertebrate family richness were negatively related to nutrient load in both data sets. Macrophyte richness decreased linearly from 18 to 9 species for the Swiss data set (p = 0.005) (Figure 1, column A), and from 18 to 7 species for the French one (p = 0.002) (Figure 1, column B). The effect of nutrient load on vascular plant richness was twice stronger in the French data set than in the Swiss one (slope of -5.1 versus -2.3, ANCOVA p-value < 0.0001). Macroinvertebrate family number linearly decreased from 33 to 18 for the Swiss data set (p = 0.0008) and from 19 to 12 (p = 0.005) for the French one (Figure 1). The effect of nutrient load on macroinvertebrate richness was slightly stronger in the Swiss data set than in the French one (slope of -4.0 versus -3.4, ANCOVA p-value of 0.002). Water beetle and dragonfly richness were negatively linearly related to nutrient load, but in only one data set. Dragonfly richness was negatively related to nutrient load in the French data set, from 15 to 11 dragonfly species (p = 0.03), but not in the Swiss one (Figure 1). Water beetle richness was negatively related to nutrient load in the Swiss data set, from 16 genus to 4 genus (p = 0.006), but not in the French one (Figure 1). No significant relationship between nutrient load and taxonomic richness was revealed for gastropods and amphibians, in both data sets (Figure 1). 172 Figure 1: Scatter plots of macrophyte species richness, gastropod genus richness, water beetle genus richness, dragonfly species richness and amphibian species richness along a trophic gradient (O: oligotrophic, M: mesotrophic, E: eutrophic, H: hypertrophic, HH: highly hypertrophic) for A. the Swiss data set, and B. the French data set. Robust linear regressions and polynomial regressions are represented when significant (p < 0.05). 173 Response of conservation value to nutrient load The conservation values of dragonfly and amphibian assemblages were not correlated to nutrient values and showed scattered values for all trophic status (Figure 2). The conservation value of macrophyte assemblages decreased linearly with increasing nutrient load but only in the Swiss data set. This decrease ranged from an index value of 3 to a value of 2 (p = 0.01) (Figure 2). Figure 12: Scatter plots of macrophyte, dragonfly and amphibian Csp conservation value along a trophic gradient (O: oligotrophic, M: mesotrophic, E: eutrophic, H: hypertrophic, HH: highly hypertrophic) for A. the Swiss data set and B. the French data set. Robust linear regressions and polynomial regressions are represented when significant (p < 0.05). NA corresponds to values not available due to statistical limits of the robust linear regression method. 174 Forecasted loss in local biodiversity due to a future nutrient increase The response of pond biodiversity (taxonomic richness and conservation value) to a future increase in nutrient load (expected to happen in a climate change context) varied according to the taxonomic group considered (Figure 3). On one side, the taxonomic richness of gastropods and both taxonomic richness and conservation value of amphibians are not predicted to change in ponds with an increase in nutrient load. On another side, the taxonomic richness of macrophytes, macroinvertebrates, water beetles and dragonflies is predicted to be negatively affected in ponds by the increase in nutrient load (Figure 3); their conservation value is nevertheless negatively impacted only for macrophytes and only in one of the data sets. A decrease in macrophyte and macroinvertebrate pond richness was projected in both data sets (Mann-Whitney p-values < 0.04, Figure 3). Considering macrophytes, the increase in nutrient load would lead to a loss of 1.4 to 2.6 species (10 to 19%) in a Swiss pond (median) and of 3.9 to 5.5 species (32 to 44%) in a French one. Considering macroinvertebrates, the increase in nutrient load would lead to a loss of 3.3 to 5.3 families (13 to 20%) in a Swiss pond (median) and of 1.6 to 2.6 families (11 to 18%) in a French one. A decline of the conservation value of macrophyte assemblages was also projected, but only in the Swiss data set, with a loss of a value of 0.3 to 0.5 (11 to 19%) per pond (median) (Mann-Whitney p-values < 0.04). A decrease in genus richness of water beetles was projected only in ponds from the Swiss data set under the pessimistic projection; the projected median loss would be of 2.4 genus (27%) (MannWhitney p-value < 0.003). Dragonfly species richness was projected to decrease in ponds for the French data set only, with a median loss of 0.9 to 1.4 species (7 to 11%) (Mann-Whitney p-values < 0.006). 175 Figure 13: Taxonomic richness (value for a median pond) observed today (“current”) and forecasted with an optimistic projection (“slight nutrient increase”) and with a pessimistic projection (“high nutrient increase”). A. corresponds to the Swiss data set, and B. to the French data set. The non-significant changes appear in grey (Mann-Whitney p-value > 0.05). 176 DISCUSSION Response of taxonomic richness to nutrient load Contrarily to our hypotheses, a decrease in taxonomic richness with increasing nutrient load was not observed for all taxonomic groups. If such a decrease was evidenced here for the vascular plants and the macroinvertebrates, this was not the case for the gastropods and the amphibians. Dragonflies and water beetles presented an intermediate situation with a decrease in richness in only one of both data sets. These discrepancies underline that different taxonomic groups can show different trends. The waterbodies investigated here are nutrient-rich; therefore, the decrease in richness revealed by the present study for vascular plants and macroinvertebrates, and partially for dragonflies and water beetles, is consistent with the decreasing part of the well-known hump-shaped pattern reported by the reviews of Mittelbach et al. (2001) and Waide et al. (1999). It is also consistent with more specific studies conducted in freshwaters (Dodson et al. 2000, Jeppesen et al. 2000, Chase and Leibold 2002, Feuchtmayr et al. 2009). The decrease of richness, however, contrasts with some experimental studies which found only a slight effect of nutrient charges on taxonomic richness of freshwater communities (McKee et al. 2002, McKee et al. 2003, Moss et al. 2003, Feuchtmayr et al. 2007). For vascular plants and macroinvertebrates, a decrease in taxonomic richness with increasing nutrient load was observed in both Swiss and French waterbodies. The high sensitivity of vascular plants to eutrophication has been widely recognized in freshwaters (e.g. Forsberg 1964, Melzer 1976, Haslam 1982, Haury et al. 2006, Bornette and Puijalon 2011). Moreover, the indirect negative impact of eutrophication on macrophytes due to phytoplankton-induced turbidity has been largely reported (e.g. Scheffer 2004, Walker and Preston 2006, Scheffer and van Nes 2007). The negative response of macroinvertebrates to increasing nutrient load has also been reported elsewhere (e.g. Simpson et al. 1986, Menetrey et al. 2005), and confirms their well-described indicator potential of the water quality (e.g. Rosenberg and Resh 1993, Smith et al. 2007). For two other taxonomic groups, dragonflies and water beetles, a decrease in richness with increasing nutrient load occurred for only one of the two data sets. Contrasting responses to productivity among types of ecosystems have already been reported by Mittelbach et al. (2001). The linear decrease of water beetle genus richness with increasing nutrient load occurred only in the Swiss less anthropised ponds. The absence of relationship in the French data set may relate to some other environmental strains which could be of greater importance for water beetles, such as shoreline configuration. The sensitivity of water beetles to eutrophication revealed in the Swiss data 177 set is in accord with previous studies in Irish wetlands (Foster et al. 1992, Cooper et al. 2005). Dragonfly species richness only decreased with increasing nutrient load in French ponds. The absence of relationship in the Swiss data set may be explained by some other environmental strains, such as pond area. Indeed, as pond area was found to explain 31% of the variability of dragonfly species richness in the Swiss data set (Oertli et al. 2002), area could be of greater importance for dragonflies than nutrient load. Finally, in both data sets, gastropod and amphibian taxonomic richness was not related to nutrient load. This tolerance of eutrophication is not surprising for gastropods. Indeed, gastropods have a greater affinity for eutrophic conditions than other macroinvertebrate orders. As an example, 20% of all Swiss gastropod genera prefer eutrophic conditions, contrasting with other freshwater macroinvertebrates such as mayflies and caddisflies those only 7% and 2% genera prefer eutrophic conditions (unpublished data coupling ecological traits from Tachet et al. 2000 with Swiss data bases). The tolerance of amphibians to eutrophication may be related to their high plasticity and high dispersal abilities which enable them to occupy a large range of ecological conditions (e.g. Feder and Burggren 1992, Smith and Green 2005). Response of conservation value to nutrient load The conservation value of assemblages did respond to an increase in nutrient load only for one of the investigated groups: the macrophytes. As already discussed, vascular plants are highly sensitive to eutrophication. Differences in the response of the richness and the conservation value, two main facets of biodiversity, demonstrate that richness cannot act as a surrogate for conservation value or for biodiversity. Such discrepancies between richness and conservation value have already been evidenced in ponds from United Kingdom (Biggs et al. 1998). The weak relationship between conservation value and nutrient load may be related to potential large differences in the autecology of rare species, with for example some being associated to nutrient-poor systems, when others are associated to nutrient-rich systems. This means that a same conservation value can potentially occur in a nutrient rich system as in a nutrient poor system. Implications of the “nutrient load-biodiversity” relationship in a climate change context Among all global changes, the rise in temperature due to climate change will increase the eutrophication and its impact on living communities from freshwater ecosystems, (e.g. McKee et al. 178 2003, Park et al. 2004, review in Hering et al. 2010). The projections made in the present study revealed that this eutrophication may lead to a decrease of species richness in temperate lowland ponds and small lakes, reaching 7% to 27%. Nevertheless, the loss of species does not impact all taxonomic groups and for example, amphibians and gastropods, may not suffer directly from eutrophication. 179 ACKNOWLEDGMENTS We thank the numerous people from the University of Geneva (Laboratory of Ecology and Aquatic Biology), the University of Applied Sciences Western Switzerland (hepia, Geneva), ISARA-Lyon (Engineering school in agriculture, alimentation, rural development and environment), and the Lyon 1 University (UMR CNRS 5023, laboratory of ecology of natural and anthropized hydrosystems) for the field-work, the laboratory-work and the database management. Various supports contributed for gathering the database for the Swiss ponds studied: the Federal Office for the Environment (OFEV), several Swiss Cantons, the University of Geneva (Laboratory of Ecology and Aquatic Biology) and the University of Applied Sciences Western Switzerland (RCSO RealTech). Various supports contributed for gathering the database for the French ponds: the owners of the studied ponds, the RhoneMediterranean and Corsica Water Agency, the French Department of Ecology, energy, sustainable development and sea (DIVA2 research grant), the PEP Aquaculture Rhône-Alpes, the French Vérots foundation and the French-Swiss PHC Germaine de Staël grant. We also would like to thank the anonymous reviewers for their insightful comments which greatly helped to improve the manuscript and P. Nicolet for editing the English. 180 4.3. Synthèse concernant l’impact de l’eutrophisation sur la biodiversité des étangs de plaine La conclusion de l’article « The relationship between nutrient load and biodiversity in lowland ponds and small lakes: implications in a changing climate context » (Article 6) est que : La réponse de la biodiversité à une augmentation de la charge trophique varie selon les groupes taxonomiques. Ce résultat souligne que différents groupes taxonomiques ne suivent pas nécessairement les mêmes tendances. Plus précisément, cet article met en évidence que : L’eutrophisation a un impact linéaire négatif sur la richesse taxonomique des macrophytes et des macroinvertébrés, confirmant l’hypothèse H3, alors que la richesse taxonomique des gastéropodes et des amphibiens ne répond pas à l’eutrophisation. La richesse en coléoptères et en odonates adultes montre une situation intermédiaire avec une diminution de la richesse seulement dans un des deux jeux de données étudiés. La réponse de la valeur de conservation des peuplements est faible en comparaison à la richesse taxonomique, voire nulle. La valeur de conservation des peuplements ne diminue pas avec l’eutrophisation, sauf pour les macrophytes. Cette différence de réponse à l’eutrophisation entre richesse et valeur de conservation démontre que la richesse ne peut pas être un indicateur de la valeur de conservation. Les réponses de la biodiversité mises en évidences dans ce chapitre ont des implications majeures dans un contexte de réchauffement climatique. En effet, la biodiversité locale des quatre groupes taxonomiques sensibles à l’eutrophisation pourrait diminuer fortement. 181 4.4. Discussion complémentaire Quelques éléments de discussion complémentaire à ceux développés jusqu’à présent dans ce chapitre peuvent être apportés, principalement concernant l’impact de l’eutrophisation sur la composition spécifique à l’échelle locale et l’impact de l’eutrophisation sur la richesse taxonomique à l’échelle régionale. Impact de l’eutrophisation sur la composition spécifique à l’échelle locale La diminution de la richesse locale due à l’eutrophisation mise en évidence certains des groupes taxonomiques étudiés dans cette thèse (macrophytes et macroinvertébrés) est le bilan d’extinctions et de colonisations à l’échelle de chaque étang, donc de changements de composition spécifique. Des changements de la composition des peuplements ont déjà été observés sous l’effet de l’eutrophisation. Dans le lac Léman, par exemple, l’abondance du macrophyte Potamogeton pusillus a augmenté sous l’effet de l’eutrophisation alors que les macrophytes Zannichellia palustris et Potamogeton crispus ont presque disparu (Lachavanne 1985, Lehmann and Lachavanne 1999). Dans des rivières danoises, toutes les espèces de Potamogeton ont soit diminué en abondance soit disparu sous l’effet de l’eutrophisation (Riis and Sand-Jensen 2001). Certaines espèces d’Ephéméroptères sont présentes préférentiellement dans des étangs de Suisse à faible (Caenis horaria par exemple) ou fort niveau trophique (Cloeon dipterum) (Menetrey et al. 2008). La diminution de richesse locale due à l’eutrophisation n’empêche toutefois pas certaines espèces menacées de développer des populations stables dans des milieux hypertrophes, comme par exemple la libellule Leucorrhinia pectoralis dans les étangs de la Dombes (Leclerc et al. 2010, en Annexe 2). Impact de l’eutrophisation sur la richesse taxonomique à l’échelle régionale (pays ou région) En plus des conséquences de l’eutrophisation à l’échelle locale (étang), l’eutrophisation a des conséquences sur la richesse à l’échelle régionale (région, pays). A l’échelle régionale, la relation entre productivité et richesse spécifique consiste généralement en une courbe en cloche (Rosenzweig 1995). Ainsi, dans une gamme de productivités élevées, la richesse diminuerait en réponse à l’augmentation de productivité. L’impact de l’eutrophisation à l’échelle régionale n’a pas été étudié spécifiquement dans le cadre de cette thèse. Néanmoins, une simulation de l’effet d’une augmentation de la productivité sur la richesse régionale peut être faite pour des étangs de plaine riches en nutriments (exemple du 182 plateau des Dombes). Cette simulation consiste en un remplacement des 50% des étangs les moins riches en nutriments par les 50% les plus riches, soit à un passage d’une région avec 40% d’étangs hypertrophes et 60% d’étangs hautement hypertrophes à une région avec 100% d’étangs hautement hypertrophes. Cette analyse montre que l’eutrophisation pourrait provoquer une diminution de la richesse régionale pour tous les groupes taxonomiques sauf les amphibiens (Figure 14). L’amplitude de cette diminution est néanmoins très différente selon le groupe taxonomique considéré. La richesse régionale en macrophytes serait la plus affectée passant de 45 espèces à 37. Des espèces telles que Rorippa amphibia et Potamogeton berchtoldi, qui sont par ailleurs respectivement vulnérables ou quasi-menacées selon la Liste Rouge Suisse, pourraient disparaître. La richesse régionale en odonates, coléoptères et gastéropodes diminuerait au contraire très faiblement. La forte diminution de la richesse régionale des macrophytes sous l’effet de l’eutrophisation pourrait augmenter l’impact de l’eutrophisation à l’échelle locale. En effet, pour un étang donné, si la taille des populations d’une espèce diminue dans les étangs environnants, la probabilité d’extinction des populations de cette espèce va augmenter comme décrit par la théorie des métapopulations (Levins 1969, 1970, Hanski 1991). Cette espèce sera donc probablement moins résiliente aux perturbations se produisant dans l’étang en question (Lawler 2009), comme par exemple une augmentation de la charge trophique. Figure 14 : Simulation de l’effet d’une augmentation de la productivité sur la richesse taxonomique régionale de la Dombes pour cinq groupes taxonomiques. Les 50% d’étangs les moins riches en nutriments sont remplacés par les 50% les plus riches, ce qui consiste en un passage d’une région avec 40% d’étangs hypertrophes (H) et 60% d’étangs hautement hypertrophes (HH) à une région avec 100% d’étangs hautement hypertrophes (HH). 183 184 Chapitre 5 Synthèse et discussion générale 5. Synthèse et discussion générale 5.1. Vers une compréhension globale de l’impact combiné du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs L’impact de deux perturbations majeures, le réchauffement climatique (Chapitre 3) et l’eutrophisation (Chapitre 4) sur la biodiversité des étangs a été étudié dans le cadre de cette thèse. Les résultats obtenus et discutés aux Chapitres 3 et 4 permettent d’avancer dans la compréhension de l’importance relative de chacune des deux perturbations ainsi que de leur impact combiné. Comme exposé en introduction, la compréhension de l’impact combiné de l’eutrophisation et du réchauffement climatique requiert la prise en considération de différentes échelles spatiales (Figure 2). Cette discussion est donc divisée entre échelle régionale (pays ou région) et échelle locale (écosystème). 5.1.1. Echelle régionale (diversité γ) Comme déjà décrit, la proportion d’espèces à risque d’extinction (perdants) est importante (jusqu’à 33%) en Suisse, mais elle est moindre que la proportion d’espèces favorisées par le réchauffement (gagnants, 53-61%) (Chapitre 3). En ce qui concerne l’eutrophisation, les analyses effectuées pour des étangs de plaine riches en nutriments suggèrent des réponses différentes selon le groupe taxonomique considéré avec par exemple une forte diminution de la richesse régionale pour les macrophytes, mais une légère diminution de la richesse régionale pour les gastéropodes, odonates et coléoptères et une absence de changement pour les amphibiens (Chapitre 4). Mais, quel sera l’impact combiné du réchauffement climatique et de l’eutrophisation à l’échelle régionale ? Cette question est très complexe, car étant donné l’échelle spatiale large, les processus en jeu sont très nombreux. Deux éléments principaux de réponse peuvent néanmoins être apportés. Premièrement, comme brièvement présenté en introduction, une étude à l’échelle globale apporte quelques éléments de réponse sur l’importance relative du réchauffement climatique et de l’eutrophisation. Sala et al. (2000) constatent que l’importance relative du climat (notamment la température) et de l’occupation du sol (notamment l’eutrophisation) sur la biodiversité varie selon les écosystèmes et les zones géographiques. Dans les régions alpines, le climat a plus d’importance que l’occupation des sols (Figure 7), suggérant un impact prépondérant de la température. Lorsque l’on considère les lacs en particulier, l’utilisation du sol a plus d’impact sur la biodiversité que le 185 climat (Figure 7), suggérant un impact prépondérant de l’eutrophisation. La composante eutrophisation pourrait donc s’avérer importante, voire majeure par rapport à la composante température pour les étangs dans les régions de plaine, mais au contraire mineure pour les étangs d’altitude. Néanmoins, cette prépondérance de la température dans les régions d’altitude pourrait changer : l’implantation de cultures là où il n’y en a pas aujourd’hui pourrait augmenter l’impact de l’eutrophisation en altitude. Deuxièmement, les prédictions des taux d’extinction d’espèces sous l’effet du réchauffement climatique sont nombreuses (e.g. Thomas et al. 2004, Heikkinen et al. 2010), alors que celles sous l’effet de l’eutrophisation sont plus rares (e.g. Penning et al. 2008). Toutes ces prédictions considèrent soit l’effet du réchauffement climatique, soit celui de l’eutrophisation. Est-ce que la prise en compte des deux perturbations ensemble augmenterait la proportion d’extinctions prédites ou est-ce que les espèces à risque sous l’effet du réchauffement sont en grande partie les mêmes que celles à risque sous l’effet de l’eutrophisation ? Bien que cette question n’ait pas été traitée spécifiquement dans cette thèse, un test sur un des groupes taxonomiques étudiés, les macrophytes, a néanmoins pu être effectué. Il révèle que les espèces menacées par le réchauffement climatique (espèces sténothermes froides identifiées au Chapitre 3) sont également sensibles à l’eutrophisation. En effet, 90% d’entre elles ont une valeur N de Landolt entre 1 et 2 (Landolt 1977), ce qui correspond à une préférence écologique pour les milieux très pauvres à maigres, comme par exemple Carex norvegica et Isoëtes echinospora (Tableau 3). Au contraire, parmi les espèces à risque sous l’effet de l’eutrophisation (liste des espèces européennes de Penning et al. (2008) présentes en Suisse), seulement un quart d’entre elles sont sensibles au réchauffement climatique, comme par exemple Sparganium angustifolium, Isoëtes lacustris et I. echinospora (Tableau 4). Au final, ces résultats préliminaires suggèrent que le réchauffement climatique et l’eutrophisation ne menacent pas toujours les mêmes espèces et que la prise en compte des deux perturbations ensemble pourrait bien augmenter la proportion d’extinction prédites. 186 Tableau 3 : Valeurs N de Landolt (Landolt 1977) pour les espèces sténothermes froides menacées par le réchauffement climatique identifiées au Chapitre 3. Une valeur N de Landolt de 1 correspond à très pauvre, 2 à maigre, 3 à ni maigre-ni fumé, 4 à riche et 5 à surfumé. Espèce sténotherme froide menacée par le réchauffement climatique Valeur N de Landolt Arabis subcoriacea 2 Cardamine amara 3 Carex frigida 2 Carex juncella 2 Carex nigra 2 Carex norvegica 1 Carex paupercula 2 Cochlearia pyrenaica 2 Dactylorhiza cruenta NA Epilobium alsinfolium 2 Epilobium nutans 2 Eriophorum angustifolium 2 Eriophorum scheuchzeri 2 Isoëtes echinospora 1 Isoëtes lacustris NA Juncus filiformis 2 Montia fontana 2 Ranunculus reptans 2 Rorippa islandica NA Saxifraga stellaris 2 Sparganium angustifolium 3 Vaccinium microcarpum 1 Tableau 4 : Sensibilité au réchauffement climatique (espèces sténothermes froides) des espèces européennes présentes en Suisse identifiées comme à risque d’extinction sous l’effet de l’eutrophisation par Penning et al. (2008). Espèces à risque sous l'effet de l'eutrophisation Sensibilité au réchauffement climatique Eleocharis acicularis Isoetes echinospora oui I. lacustris oui Ranunculus reptans oui Callitriche hamulata Myriophyllum alterniflorum Potamogeton filiformis P. polygonifolius P. x nitens P. x zizii Ranunculus peltatus Utricularia australis U. intermedia U. minor U. ochroleuca Nuphar lutea x pumila Sparganium angustifolium oui 187 5.1.2. Echelle locale (diversité α) Comme déjà décrit, sous l’effet du bilan positif entre gagnants et perdants à l’échelle régionale, le réchauffement climatique devrait significativement augmenter la richesse spécifique locale et cette augmentation de richesse serait particulièrement marquée dans les étangs d’altitude pauvres en espèces en comparaison aux étangs de plaine (Chapitre 3). En ce qui concerne l’impact de l’eutrophisation sur les étangs de plaine riches en nutriments, la réponse de la biodiversité varie selon les groupes taxonomiques : l’eutrophisation a un impact négatif sur la richesse taxonomique des macrophytes et des macroinvertébrés, mais pas sur celle des gastéropodes et des amphibiens (Chapitre 4). La richesse en coléoptères et en odonates adultes montre une situation intermédiaire avec une diminution de la richesse taxonomique seulement dans l’un des deux jeux de données étudiés (Chapitre 4). Quel sera alors l’impact combiné du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs à l’échelle locale ? Une des conclusions concernant les effets synergétiques de ces deux perturbations à l’échelle régionale apportée par cette thèse - le réchauffement climatique et l’eutrophisation ne menacent pas toujours les mêmes espèces – suggère que leur impact combiné soit supérieur à l’impact de chacun des deux perturbations individuellement. De plus, il est largement reconnu que le réchauffement climatique va exacerber l’eutrophisation à l’échelle de l’écosystème (Flanagan et al. 2003, Park et al. 2004, Schindler 2006, Heino et al. 2009). Les effets seront par contre complexes et varieront selon les conditions initiales et la localisation (Kernan et al. 2010). L’augmentation de la température favorise en soi la croissance de cyanobactéries et augmente la durée des périodes d’anoxie sédimentaire qui provoquent une hausse du relargage de phosphore dans les lacs (Ulen and Weyhenmeyer 2007, European Environment Agency 2008). Certaines études suggèrent également qu’une dégradation de la qualité de l’eau (en termes de teneur en nutriments notamment) pourrait augmenter ou accélérer les impacts du réchauffement climatique sur les macroinvertébrés ou les poissons des rivières (Daufresne et al. 2007, Durance and Ormerod 2009). De plus, le réchauffement pourrait augmenter la sensibilité à l’eutrophisation et menacer la « phase claire » (clear state en anglais) riche en biodiversité (e.g. Meerhoff et al. 2007, Mooij et al. 2007), bien que cette possibilité ne soit mise en évidence par aucune des études expérimentales effectuées à l’heure actuelle (Jeppesen et al. 2010). D’autres études, bien que n’ayant pas spécifiquement pour objectif de déterminer qui du réchauffement climatique ou de l’eutrophisation a un impact prépondérant, apportent également des éléments de réponse. Daufresne and Boet (2007) suggèrent que les effets du réchauffement pourraient dépasser ceux des autres perturbations anthropiques non-climatiques pour les 188 peuplements piscicoles des grandes rivières de France. Au contraire, certaines expériences en mésocosmes (e.g. Moss et al. 2003) et l’étude de larges jeux de données (e.g. Gyllstrom et al. 2005) suggèrent que le réchauffement a et/ou aura un impact faible en comparaison avec l’eutrophisation. La question de l’impact prépondérant du réchauffement climatique ou de l’eutrophisation sur la biodiversité d’eau douce à l’échelle locale reste donc ouverte. Les modélisations effectuées dans le cadre de cette thèse suggèrent que l’eutrophisation soit prépondérante pour deux des cinq groupes taxonomiques étudiés : les macrophytes et les gastéropodes (Figure 15 et Figure 16). Pour les autres groupes taxonomiques, l’eutrophisation aurait un impact important sur la réponse de la richesse spécifique au réchauffement climatique uniquement en altitude (Figure 17 et Figure 18 et Figure 19). Pour les macrophytes et les gastéropodes, l’augmentation de richesse prédite sous l’effet du réchauffement climatique (scénario 1, Figure 15 et Figure 16) serait diminuée voire éliminée par la prise en compte de l’augmentation de la charge trophique dans les étangs collinéens (scénarios 2 et 3, Figure 15 et Figure 16). Pour les macrophytes, cette atténuation de l’augmentation de richesse prédite se produirait également dans les étangs montagnards et subalpins (Figure 15). En haute altitude (étangs alpins), une eutrophisation modérée accentuerait l’augmentation de richesse des macrophytes et des gastéropodes, alors qu’une eutrophisation élevée l’atténuerait (Figure 15 et Figure 16), tendances cohérentes avec la relation la plus commune entre productivité et richesse, une courbe en cloche. Figure 15 : Richesse actuelle en macrophytes dans les étangs de Suisse et changements prédits en réponse au réchauffement climatique pour 2090-2100 à différents étages : collinéen, montagnard, subalpin et alpin. Le scénario 1 considère uniquement l’augmentation de température (scénario A2 du GIEC : + 3.4°C). Les scénarios 2 et 3 considèrent également des changements de niveau trophique et de conductivité sous l’effet du réchauffement climatique. Le scénario 2 est « conservatif » et le scénario 3 est « pessimiste ». (Résultats basés sur les modélisations effectuées dans le Chapitre 3.3). 189 Figure 16 : Richesse actuelle en gastéropodes dans les étangs de Suisse et changements prédits en réponse au réchauffement climatique pour 2090-2100 à différents étages : collinéen, montagnard, subalpin et alpin. Le scénario 1 considère uniquement l’augmentation de température (scénario A2 du GIEC : + 3.4°C). Les scénarios 2 et 3 considèrent également des changements de niveau trophique et de conductivité sous l’effet du réchauffement climatique. Le scénario 2 est « conservatif » et le scénario 3 est « pessimiste ». (Résultats basés sur les modélisations effectuées dans le Chapitre 3.3). Pour les coléoptères et les odonates adultes, l’augmentation de richesse prédite sous l’effet du réchauffement climatique (scénario 1, Figure 17 et Figure 18) n’est pas influencée par la prise en compte de l’augmentation de la charge trophique dans les étangs collinéens, montagnards et subalpins (scénarios 2 et 3, Figure 17 et Figure 18). Dans les étangs alpins par contre, une eutrophisation modérée accentuerait l’augmentation de richesse des coléoptères et des odonates adultes, alors qu’une eutrophisation élevée l’atténuerait (Figure 17 et Figure 18), tendances cohérentes avec la relation de courbe en cloche entre productivité et richesse largement décrite. 190 Figure 17 : Richesse actuelle en coléoptères dans les étangs de Suisse et changements prédits en réponse au réchauffement climatique pour 2090-2100 à différents étages : collinéen, montagnard, subalpin et alpin. Le scénario 1 considère uniquement l’augmentation de température (scénario A2 du GIEC : + 3.4°C). Les scénarios 2 et 3 considèrent également des changements de niveau trophique et de conductivité sous l’effet du réchauffement climatique. Le scénario 2 est « conservatif » et le scénario 3 est « pessimiste ». (Résultats basés sur les modélisations effectuées dans le Chapitre 3.3). Figure 18 : Richesse actuelle en odonates adultes dans les étangs de Suisse et changements prédits en réponse au réchauffement climatique pour 2090-2100 à différents étages : collinéen, montagnard, subalpin et alpin. Le scénario 1 considère uniquement l’augmentation de température (scénario A2 du GIEC : + 3.4°C). Les scénarios 2 et 3 considèrent également des changements de niveau trophique et de conductivité sous l’effet du réchauffement climatique. Le scénario 2 est « conservatif » et le scénario 3 est « pessimiste ». (Résultats basés sur les modélisations effectuées dans le Chapitre 3.3). 191 Pour les Amphibiens, l’augmentation de richesse prédite sous l’effet du réchauffement climatique n’est influencée par la prise en compte de l’augmentation de la charge trophique à aucun étage altitudinal (Figure 19). Figure 19 : Richesse actuelle en amphibiens dans les étangs de Suisse et changements prédits en réponse au réchauffement climatique pour 2090-2100 à différents étages : collinéen, montagnard, subalpin et alpin. Le scénario 1 considère uniquement l’augmentation de température (scénario A2 du GIEC : + 3.4°C). Les scénarios 2 et 3 considèrent également des changements de niveau trophique et de conductivité sous l’effet du réchauffement climatique. Le scénario 2 est « conservatif » et le scénario 3 est « pessimiste ». (Résultats basés sur les modélisations effectuées dans le Chapitre 3.3). L’impact prépondérant de l’eutrophisation en plaine mis en évidence pour les macrophytes et les gastéropodes est cohérent avec les résultats de l’étude global de Sala et al. (2000) décrits au chapitre 5.1.1 qui montrent un impact prépondérant de l’eutrophisation en plaine. Les modélisations effectuées pour les coléoptères, les odonates adultes et les amphibiens semblent par contre en contradiction avec les résultats de Sala et al. (2000). Néanmoins, l’absence de réponse à l’eutrophisation en plaine mise en évidence pour ces trois groupes taxonomiques pourrait s’expliquer par leur faible sensibilité à l’eutrophisation décrite au Chapitre 4. 192 5.2. Le changement global : quelles conséquences supplémentaires sur la biodiversité des étangs ? Comme montré dans la Figure 3, bien que la température et la teneur en nutriments soient deux facteurs abiotiques ayant un impact majeur sur la biodiversité des étangs, de nombreux autres paramètres ont un impact sur la biodiversité des étangs. Ainsi, en plus du réchauffement climatique et de l’eutrophisation, d’autres changements ont et vont avoir un impact sur la biodiversité des étangs, participant à ce qu’on nomme changement global. Changements abiotiques Tout d’abord, le changement climatique consiste en d’autres changements physiques directs que le réchauffement. Les changements de vitesse du vent, de précipitations (IPCC 2007a) et d’évaporation pourraient avoir un impact majeur sur la biodiversité (Körner and Walther 2001). Le changement climatique va également provoquer des modifications de l’hydropériode qui auront des effets sur la biodiversité d’eau douce (e.g. Xenopoulos et al. 2005). De plus, la variabilité des changements climatiques aura des conséquences, étant donné que les réponses écologiques ne dépendent pas uniquement de moyennes globales (Walther et al. 2002). Cette question de la variabilité est très importante en Suisse, où les changements futurs de température et de précipitations seront différents entre saisons (C2SM et al. 2011). Elle le sera également dans la Dombes où les précipitations de fin d’hiver sont déterminantes pour le remplissage des étangs ayant été mis en assec (Goubier-Martin 1991). De plus, l’augmentation de la fréquence et la force des événements climatiques extrêmes aléatoires - inondations, sécheresses, gel, etc - (IPCC 2007a), va avoir des conséquences majeures sur la structure et le fonctionnement des peuplements aquatiques (e.g. Ledger et al. 2011). Ensuite, le changement climatique va induire des changements indirects de la physico-chimie de l’eau. Comme déjà largement décrit, l’augmentation de température va augmenter la productivité primaire et exacerber l’eutrophisation (e.g. Park et al. 2004, Heino et al. 2009). De plus, les phases de désoxygénation devraient être plus fréquentes et plus marquées (Portnoy 1991), ce qui va affecter les cycles de vie, les aires de distribution, le comportement et les interactions avec d’autres organismes de la majorité des espèces d’eau douce (Brönmark and Hansson 2000). Le changement climatique va également avoir un impact sur le pH des étangs. Néanmoins, les observations actuelles montrent dans certains cas une acidification et dans d’autres une alcalinisation et ne permettent pas encore de prédire avec certitude les conséquences futures (Heino et al. 2009). 193 En plus des changements climatiques et physico-chimiques, les activités humaines provoquent des modifications à l’échelle du bassin versant des étangs qui vont également avoir des conséquences sur leur biodiversité. Par exemple, la fragmentation des habitats terrestres et la perte de connectivité en découlant vont avoir des conséquences dramatiques pour les amphibiens qui utilisent l’habitat terrestre pour migrer et se déplacer (Joly et al. 2001, Dodd 2010). De plus, le réchauffement climatique va provoquer des changements d’occupation du sol. Dans les Alpes suisses, par exemple, un réchauffement modéré devrait augmenter la proportion de terres cultivées (Rebetez 2006, OcCC and ProClim– 2007). Interactions biologiques Les changements de la biodiversité dus à tous ces changements abiotiques (climatiques, physicochimiques et anthropiques) vont influencer en retour la biodiversité (Chapin et al. 2000) du fait de modifications des interactions biologiques. Les interactions biologiques sont connues depuis longtemps comme d’importance majeure dans le fonctionnement des écosystèmes en général (Begon et al. 2006), mais également pour la compréhension de l’impact du réchauffement climatique (e.g. Klanderud and Totland 2007) et de l’eutrophisation (e.g. Micheli 1999). De nombreuses études récentes ont mis en évidence la nécessité d’inclure ces interactions biologiques dans les modélisations (e.g. Preston et al. 2008, Van der Putten et al. 2010) car (i) les distributions géographiques des espèces peuvent être limitées par les conditions biotiques (Moore et al. 2007) et (ii) les perturbations anthropiques, le réchauffement climatique par exemple, peuvent découpler les histoires de vie des espèces (Schlaepfer et al. 2002, Parmesan 2006). De plus, les changements globaux peuvent favoriser des maladies et accélérer ainsi la disparition de certaines espèces (Kiesecker et al. 2001, Pounds et al. 2006). Espèces invasives La progression des espèces invasives, sous l’effet notamment du réchauffement climatique (Tockner and Stanford 2002, Hellmann et al. 2008, Rahel and Olden 2008), peut également avoir des conséquences majeures sur la biodiversité aquatique. De nombreux exemples incluent des espèces liées aux eaux chaudes qui sont apparues dans la mer Méditerranée, la mer du Nord et la mer Rouge (Walther et al. 2002). Même dans des environnements relativement isolés d’altitude ou arctiques (Smith 1996, Pauchard et al. 2009), des invasions se produisent. La majorité de ces invasions résulte 194 en une perte de la biodiversité ainsi qu’en des changements dans la structure des peuplements et dans le fonctionnement de l’écosystème (Mooney and Hobbs 2000). 5.3. Implications pour la conservation de la biodiversité Si de nombreuses recherches sont encore nécessaires pour comprendre avec précision les réponses multiples des organismes au réchauffement climatique et à l’eutrophisation, il ne fait aucun doute que ces deux perturbations ont et vont avoir dans le futur des conséquences majeures sur la biodiversité des étangs. Les organismes vont donc devoir s’adapter à ces changements ou migrer. Lorsque les adaptations des organismes eux-mêmes ne sont pas suffisantes ou suffisamment rapides ou lorsque la migration n’est pas possible, des moyens humains peuvent être nécessaires pour aider les organismes à survivre. Bien que le réchauffement climatique et l’eutrophisation posent des défis supplémentaires, les stratégies de conservation habituelles - établissement de réseaux de réserves et maintien de la connectivité entre populations ou processus - peuvent être extrêmement efficaces (e.g. Hunter et al. 2010). De plus, une des approches les plus intuitives pour améliorer la capacité des espèces à supporter ces deux perturbations est de minimiser toutes les autres perturbations anthropiques (Poff et al. 2002, Heller and Zavaleta 2009, Lawler 2009). En effet, protéger les écosystèmes existants en diminuant l’impact des espèces invasives, la fragmentation et la perte des habitats ou la surexploitation résulte généralement en de plus larges populations qui auront de meilleures chances de supporter l’eutrophisation et/ou le réchauffement climatique (Lawler 2009). En ce qui concerne l’eutrophisation en particulier, la meilleure stratégie est de traiter la cause, c’està-dire de contrôler les sources ponctuelles et diffuses de nutriments (Dodds 2002). L’introduction du traitement des eaux usées et l’interdiction du phosphate dans les lessives a permis par exemple la diminution de l’eutrophisation de plusieurs lacs suisses (Lachavanne et al. 1991, Lazzarotto and Rapin 2010). Afin de limiter les apports de nutriments en provenance du bassin versant, il faut chercher également à maîtriser les pollutions organiques d’origine agricole en favorisant l’utilisation rationnelle des engrais ou en modifiant les modes d’exploitation. Les mesures de protection des étangs face à l’eutrophisation doivent également avoir pour objectif de délimiter une zone tampon suffisante dans l’environnement immédiat de façon à limiter les atteintes dues aux exploitations agricoles environnantes. Une telle zone devrait avoir une largeur de plusieurs dizaines de mètres. En ce qui concerne les annexes fluviales par exemple, différents types de zones tampon sont recommandés dont une « zone tampon trophique ». Cette zone tampon trophique est une bande de 195 terres agricoles cultivées soumise à des restrictions d’exploitation dont l’objectif est de réduire ou prévenir l’engraissement indirect (Service conseil Zones alluviales Berne et Yverdon-les-Bains 20012008). L’Office Fédéral suisse de l’Environnement, des Forêts et du Paysage (OFEFP) recommande pour les marais une largeur comprise entre 20 et 40 mètres (chiffres en cours de révision) qui doit être adaptée en fonction des conditions locales (topographie, couverture végétale, types d’agriculture environnante) (Marti et al. 1997). Cette stratégie de zone tampon ne s’applique néanmoins pas à toutes les situations. Dans le cas des étangs de la Dombes par exemple, la présence de zones tampon n’est pas nécessaire, car l’eutrophisation est gérée par la production de poissons et par la mise en assec. De plus, il se pourrait que, dans ce cas précis, les ceintures de végétation entourant les étangs ainsi que la végétation recouvrant les fossés jouent déjà un rôle de zone tampon. Il est à noter également que, comme mis en évidence au Chapitre 4, la valeur de conservation des peuplements n’est pas nécessairement liée à la charge trophique et qu’ainsi des milieux riches en nutriments peuvent héberger des espèces rares. Dans ces cas, on cherchera à maintenir les conditions propices à ces espèces, plutôt qu’à lutter directement contre l’eutrophisation. En ce qui concerne le réchauffement climatique en particulier, les deux stratégies principales sont la mitigation qui vise à réduire l’intensité du réchauffement climatique et l’adaptation qui vise à atténuer les impacts du réchauffement climatique qui ne peuvent pas être évités. La mitigation passe par des appels politiques pour la diminution des émissions de gaz à effet de serre, par l’adoption de mesures de compensation (séquestration du carbone) et par la promotion d’un développement durable (Hannah et al. 2002, IPCC 2007b). L’adaptation passe par les stratégies de conservation habituelles comme par exemple protéger et restaurer les habitats existants, mais dans une perspective plus modulable et proactive (Heino et al. 2009, Lawler 2009), qui tient compte des changements prédits dans le futur. Une étape préliminaire à la mise en place de stratégies d’adaptation est d’identifier les espèces menacées. La Figure 20 rappelle ici les espèces principales menacées par le réchauffement climatique en Suisse. Ces espèces à risque d’extinction sous l’effet du réchauffement climatique ne sont pas identiques à celles menacées selon la Liste Rouge (Chapitre 3). Un indice de sensibilité au réchauffement climatique devrait donc être établi à des fins de conservation de la biodiversité. Cet indice pourrait être un outil de conservation complémentaire aux Listes Rouges nationales actuelles. Dans une deuxième étape, il serait utile d’ajouter les critères utilisés pour déterminer le risque d’extinction dû au réchauffement climatique dans les critères de sélection des espèces prioritaires en Suisse (OFEV 2011). Le risque d’extinction dû au réchauffement climatique pourrait être ajouté 196 comme filtre supplémentaire s’ajoutant aux deux critères utilisés à l’heure actuelle, degré de menace et responsabilité de la Suisse. Dans l’exemple des Odonates, deux espèces sténothermes froides menacées par le réchauffement climatique s’ajouteraient à cette Liste prioritaire : Somatochlora alpestris et Aeshna juncea. De plus, le degré de priorité des cinq autres espèces sténothermes froides (Coenagrion hastulatum, Leucorrhinia dubia, Aeshna subartica, Aeshna caerulea, Somatochlora artica), actuellement entre moyen et modéré, serait vraisemblablement augmenté si l’on prenait en compte les changements de température prévus dans le futur. 197 P. Prunier E. Demierre Copyright CSCF, 22.09.2011, Base cartographique : OFS, OFT © Biopix: N Sloth C. Kerihuel Figure 20 : Liste des espèces à risque d’extinction sous l’effet du réchauffement climatique en Suisse (espèces sténothermes froides) classées en fonction de leur résilience au réchauffement quand disponible (flèche avec dégradé) ainsi que photos et éventuellement cartes de distribution en Suisse de certaines de ces espèces. 198 Un des axes majeurs des stratégies d’adaptation au réchauffement climatique est l’augmentation de la connectivité, notamment aux moyens de corridors de migration et de réseaux de zones protégées (Heino et al. 2009, Heller and Zavaleta 2009, Kernan et al. 2010). Concernant en particulier les espèces menacées par le réchauffement climatique en Suisse (Chapitre 3), une stratégie intéressante à développer serait de promouvoir la restauration des habitats de ces espèces et de leurs réseaux par la création de nouveaux étangs, en tenant compte des migrations actuelles des espèces mais aussi des migrations futures induites par le réchauffement. Cette stratégie d’adaptation passe par (i) l’identification des corridors écologiques qu’utiliseront les espèces devant migrer sous la pression du réchauffement, (ii) l’identification des contraintes et potentialités liées aux usages économiques et sociaux, puis (iii) la localisation optimale de nouveaux habitats (relais et/ou permanents) pour ces espèces et (iv) le développement des populations de ces espèces en altitude (grâce à la création de milieux d’accueil, Figure 21). Cette stratégie est justement en cours d’application en Suisse pour certaines des espèces identifiées au Chapitre 3 dans le cadre d’un projet expérimental dans le Valais (Oertli et al. 2011). L’identification des corridors écologiques peut se faire grâce à l’étude de leur structure et/ou de leur fonctionnalité (Fischer and Lindenmayer 2007). Le lien entre la structure et la fonctionnalité des corridors peut être fait grâce aux outils issus d’une discipline récente : la génétique du paysage (« Landscape Genetics ») (Manel et al. 2003) : les ressemblances ou différences génétiques entre les populations sont corrélées avec la structure du paysage afin de déterminer quels éléments paysagers améliorent la connectivité (et donc favorisent les mouvements des individus) et quels éléments agissent comme des barrières aux déplacements. 199 Figure 21 : Devenir des espèces sensibles au réchauffement climatique d’ici à 2070 en l’absence de création de nouveaux plans d’eau (à gauche) et avec création de nouveaux plans d’eau (à droite). Les augmentations de température jusqu’à 2070 sont tirées de OcCC (2007). (Source : projet de recherche « Restauration des habitats alpins pour conserver la biodiversité menacée par le réchauffement climatique » (RestorAlps) de la HES-SO//GE (Oertli et al. 2011)). Une autre stratégie d’adaptation au réchauffement climatique est la colonisation assistée qui consiste à déplacer une espèce dans une nouvelle région qui devrait être plus favorable pour la persistance de l’espèce dans le futur. Cette stratégie est sujette à de nombreuses controverses, dues à sa nature très perturbatrice ainsi qu’aux incertitudes associées avec la probabilité que les espèces déplacées puissent devenir invasives (Loss et al. 2011). La colonisation assistée pourrait néanmoins être bénéfique si utilisée sur la base de connaissances scientifiques détaillées et accompagnée de stratégies de management concernant les autres menaces (Hoegh-Guldberg et al. 2008, Loss et al. 2011). 200 Chapitre 6 Conclusions et perspectives 6. Conclusions et perspectives Comprendre l’impact de deux perturbations anthropiques majeures, l’eutrophisation et le réchauffement climatique, sur la biodiversité des étangs est complexe : cela implique de nombreux processus imbriqués à des échelles spatiales multiples. Cette question est néanmoins cruciale, car les étangs abritent une biodiversité remarquable et fournissent de nombreux services environnementaux. Cette question a été étudiée ici à deux échelles spatiales : à l’échelle locale (l’écosystème : ici l’étang) et à l’échelle régionale (le pays ou la région : ici la Suisse ou la Dombes). 6.1. Conclusions Cette thèse contribue de manière originale à l’étude de l’impact du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs. A l’échelle régionale (ici en Suisse), le présent travail a montré que, sous l’influence du réchauffement climatique, la proportion d’espèces colonisatrices (« gagnantes »), donc nouvelles pour le pays, serait plus importante que celle d’espèces à risque d’extinction (« perdantes »), suggérant une augmentation de la richesse spécifique régionale. De plus, les espèces identifiées comme à risque d’extinction à cause du réchauffement climatique ne sont pas toutes sur Liste Rouge : il serait donc important pour la conservation de la biodiversité de mettre en place un indice de sensibilité au réchauffement climatique ou d’intégrer cet indice dans les critères de sélection des espèces prioritaires existants. Concernant l’eutrophisation, l’exploration effectuée à l’échelle régionale (ici la Dombes) suggère que, pour des étangs de plaine riches en nutriments, la richesse taxonomique diminue pour tous les groupes taxonomiques sauf les amphibiens, mais que cette diminution n’est marquée que pour les macrophytes. Cette forte diminution de la richesse régionale des macrophytes pourrait augmenter l’impact de l’eutrophisation sur les macrophytes à l’échelle locale. A l’échelle locale (étang), le présent travail a permis de montrer que la richesse spécifique devrait augmenter sous l’effet du réchauffement climatique. Cette augmentation est prédite pour tous les groupes taxonomiques étudiés (macrophytes, gastéropodes, coléoptères, odonates et amphibiens) mais avec des différences d’amplitude. Une augmentation de richesse particulièrement élevée en altitude a été mise en évidence. Cette forte sensibilité des étangs au réchauffement climatique fait des étangs des écosystèmes idéaux (« sentinelles ») pour la surveillance de l’impact du réchauffement climatique, et ce en particulier en milieu alpin. 201 Concernant l’impact de l’eutrophisation, une forte hétérogénéité de réponse entre les différents groupes taxonomiques a été mise en évidence pour les étangs de plaine riches en nutriments. La richesse en macrophytes et en macroinvertébrés d’un étang diminue avec l’augmentation de la charge trophique, alors que la richesse en gastéropodes et en amphibiens ne suit aucune tendance significative. La richesse en coléoptères et en odonates montre une situation intermédiaire avec une diminution de la richesse seulement dans un des deux jeux de données étudiés. Ces résultats révèlent que plusieurs groupes taxonomiques peuvent suivre des tendances différentes, ce qui renforce la nécessité d’étudier différents groupes taxonomiques et de les prendre en compte dans les stratégies de conservation de la biodiversité. La valeur de conservation d’un étang, facette de la biodiversité complémentaire à la richesse taxonomique, s’est avérée ne pas varier avec l’augmentation de la charge trophique, sauf pour les macrophytes où elle diminue. Ces différences entre richesse et valeur de conservation démontrent que la richesse taxonomique ne peut pas être un indicateur de la valeur de conservation. De plus, l’absence de relation entre valeur de conservation et niveau trophique suggère qu’une même valeur de conservation peut potentiellement être observée dans un écosystème riche en nutriments comme dans un écosystème ayant une faible concentration de nutriments. Des réseaux d’étangs ayant des niveaux trophiques différents pourraient donc favoriser la biodiversité β (variabilité entre étangs) et en conséquence favoriser la biodiversité γ (biodiversité régionale). Au-delà de la compréhension de l’impact du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs, comprendre l’impact combiné de ces deux perturbations anthropiques est crucial, bien que délicat. A l’échelle locale, il est avéré que le réchauffement climatique va exacerber les effets de l’eutrophisation. Les résultats apportés par cette thèse suggèrent que l’eutrophisation atténuerait l’augmentation de richesse prédite en plaine uniquement pour les macrophytes et les gastéropodes. En altitude, pour tous les groupes taxonomiques excepté les amphibiens, une eutrophisation modérée accentuerait l’impact du réchauffement climatique, alors qu’une eutrophisation élevée l’atténuerait. A l’échelle régionale, la littérature actuelle suggère une prépondérance de la composante eutrophisation pour les étangs dans les régions de plaine, alors que la composante température est prépondérante pour les étangs d’altitude. Les résultats apportés par cette thèse pour le cas des macrophytes suggèrent que le réchauffement climatique et l’eutrophisation ne menacent pas toujours les mêmes espèces et que la prise en compte des deux perturbations ensemble pourrait augmenter la proportion d’extinctions prédites. Des études expérimentales à différentes températures (altitudes) et trophies ou des suivis temporels de l’évolution d’étangs permettraient de confirmer ou non ces résultats. 202 Au final, les étangs sont des écosystèmes fortement menacés par le réchauffement climatique et l’eutrophisation. Il est important de conserver leur biodiversité car ils sont des éléments-clés dans la mosaïque d’écosystèmes d’eau douce qui composent notre paysage. Les réseaux d’étangs abritent une biodiversité d’autant plus riche (diversité γ) que les étangs qui les composent sont différents (diversité β). Un des objectifs majeurs de la conservation de la biodiversité devrait donc être de conserver des réseaux d’étangs différant en charge trophique, ainsi qu’en un maximum de paramètres abiotiques tels que la morphométrie ou l’ombrage. 6.2. Perspectives Bien que cette thèse apporte des éléments de réponse originaux concernant l’impact du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs, il reste de nombreuses questions à résoudre. Tout d’abord, comme discuté au Chapitre 5.1, l’impact combiné de l’eutrophisation et du réchauffement climatique reste mal connu. Les études futures devraient avoir pour objectif de traiter ces deux problématiques en même temps afin de déterminer leur impact combiné sur la biodiversité des étangs. En plus de l’impact combiné de l’eutrophisation et du réchauffement, l’influence d’autres paramètres décrits au Chapitre 5.2, comme par exemple l’hydropériode et les interactions biologiques, pourrait améliorer notre compréhension de la situation. L’hydropériode, qui dépend de la température et des précipitations mais également des usages du plan d’eau, apparaît en particulier comme un aspect crucial à explorer dans le futur. En effet, le manque d’eau futur pourrait provoquer des extinctions locales non-négligeables (e.g. Xenopoulos et al. 2005). Les interactions biologiques devraient également être considérées avec attention dans les investigations futures (Hulme 2005), et en particulier la progression des espèces invasives (Tockner and Stanford 2002, Rahel and Olden 2008). L’échelle régionale est importante, car la composition spécifique, la richesse et l’abondance locales sont influencées par les processus régionaux (Gaston and Blackburn 2000). Cette échelle régionale mériterait d’être étudiée plus en détails, en particulier en ce qui concerne la problématique de l’eutrophisation. La biodiversité β, dont les réponses au réchauffement climatique sont contradictoires (Brown et al. 2007, Dijkstra et al. 2011a), devrait également être approfondie afin de saisir les changements de la biodiversité dans leur globalité. 203 Aux différentes échelles spatiales s’ajoutent la dimension temporelle. Il est crucial que cette dimension soit prise en compte pour évaluer les changements de la biodiversité et ses conséquences (Chapin et al. 2000). De nombreuses études paléolimnologiques ont permis de comprendre les changements passés de la biodiversité dans les lacs (e.g. Smol et al. 2005, Ruhland et al. 2008, Battarbee 2010) mais également dans les étangs (Antoniades et al. 2005, Keatley et al. 2006). La question mérite néanmoins une attention supplémentaire et des réseaux de surveillance devraient être mis en place afin d’évaluer les futurs changements temporels de la biodiversité (Heino et al. 2009) et de confirmer ou non les tendances prédites sous l’effet du réchauffement climatique. Il serait bénéfique d’étendre à d’autres régions les réseaux de surveillance mis en place sur les étangs et rivières du Parc National Suisse (Robinson and Oertli 2009) ou sur les lacs alpins européens (Livingstone 2000, Magnuson et al. 2000). En ce qui concerne les étangs de la Dombes qui sont périodiquement mis en assec, l’influence de l’âge depuis le dernier assec sur la biodiversité devrait être considérée avec attention. La biodiversité inclut de nombreuses facettes. En plus des deux facettes abordées dans le cadre de cette thèse, la richesse taxonomique et la valeur de conservation des peuplements, la diversité fonctionnelle (traits fonctionnels des espèces) pourrait apporter des éléments de réponse complémentaires, notamment pour les macroinvertébrés et les macrophytes. Les macroinvertébrés présentent une large gamme de caractéristiques biologiques et écologiques (ou traits) qui sont influencés par les conditions environnementales et renseignent sur les fonctions de l’écosystème ou sa résilience à une perturbation (e.g. Poff 1997, Usseglio-Polatera et al. 2000, Verberk et al. 2008). Les macrophytes présentent également un grand nombre de traits d’histoire de vie liés aux conditions environnementales et notamment à la productivité (e.g. Southwood 1977, Grimen 2002). L’étude de la diversité fonctionnelle pourrait améliorer notre compréhension de la réponse de la biodiversité au réchauffement climatique (Daufresne et al. 2009), comme à l’eutrophisation (Hulot et al. 2000). Finalement, les conclusions concernant l’impact du réchauffement climatique et de l’eutrophisation sur la biodiversité des étangs apportées par cette thèse pourraient s’avérer valides pour d’autres types d’écosystèmes d’eau douce. Les méthodes d’analyse utilisées dans le présent travail pourraient être appliquées à d’autres milieux d’eau douce, comme les rivières alpines, afin de confirmer ou non les tendances mises en évidence ici. Les analyses effectuées dans le cadre de cette thèse pourraient également être transposées et utilisées pour modéliser l’impact du réchauffement et de l’eutrophisation à un niveau plus global, celui des services environnementaux rendus par les écosystèmes. 204 Chapitre 7 Références bibliographiques 7. Références bibliographiques ACEMAV coll., R. Duguet, and F. Melki. 2003. Les Amphibiens de France, Belgique et Luxembourg. éditions Biotope, Mèze, 480p. Amoros, C., G. Bornette, and C. P. Henry. 2000. Environmental Auditing, A Vegetation-Based Method for Ecological Diagnosis of Riverine Wetlands. Environmental Management 25:211-227. Angelibert, S., N. Indermuehle, D. Luchier, B. Oertli, and J. Perfetta. 2006. Where hides the aquatic biodiversity of macroinvertebrates in the Canton of Geneva (Switzerland)? Archives Des Sciences 59:225-234. Angelibert, S., V. Rosset, N. Indermuehle, and B. Oertli. 2010. The pond biodiversity index "IBEM": a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 1. Index development. Limnetica 29:93-104. Angermeier, P. L. and M. R. Winston. 1997. Assessing conservation value of stream communities: A comparison of approaches based on centres of density and species richness. Freshwater Biology 37:699-710. Antoniades, D., M. S. V. Douglas, and J. P. Smol. 2005. Quantitative estimates of recent environmental changes in the Canadian High Arctic inferred from diatoms in lake and pond sediments. Journal of Paleolimnology 33:349-360. Arthaud, F. 2011. Fonctionnement des étangs en réponse aux stress et perturbations d’origine anthropique : diversité, structure et dynamique des communautés végétales. Université Claude Bernard - Lyon I, Lyon, 205p. Arthaud, F., D. Vallod, J. Robin, A. Wezel, and G. Bornette. in prep. Phytoplankton diversity and productivity in eutrophic shallow lakes. Auderset Joye, D., D. Cambin, J. Détraz-Méroz, P. Durand, R. Juge, J.-B. Lachavanne, B. Lods-Crozet, A. Noetzlin, B. Oertli, C. Oïhénart, and O. Rossier. 1993. Les plans d'eau du Canton de Genève. Partie II. Caractérisation et qualification écologiques de 13 étangs., LEBA / Université de Genève & Département de l'intérieur et de l'Agriculture, Genève, Switzerland, 165p. Auderset Joye, D., B. Oertli, D. Cambin, and J.-B. Lachavanne. 1994. Etude des étangs du canton de Genève: Etang des Douves, Etang Hainard, Marais du Château., LEBA /Université de Genève & Service des Forêts, Faune et Protection de la Nature/Département de l'Intérieur et de l'Agriculture, Geneva, 43p. Auderset Joye, D., B. Oertli, R. Juge, and J.-B. Lachavanne. 2004. Evaluation écologique des étangs du Bois de Jussy., Département de l'Intérieur, de l'Agriculture et de l'Environnement du Canton de Genève (Service des forêts, de la protection de la nature et du paysage); Université de Genève, Genève, 24p. Auderset Joye, D., A. Schwarzer, J.-B. Lachavanne, and R. Juge. 2010. Liste rouge des Characées – Liste rouge des espèces menacées en Suisse. Office fédéral de l’environnement, Berne et Laboratoire d’écologie et biologie aquatique, Université de Genève., Berne, 84p. Barbour, M. T., J. Gerritsen, G. E. Griffith, R. Frydenborg, E. McCarron, J. S. White, and M. L. Bastian. 1996. A framework for biological criteria for Florida streams using benthic macroinvertebrates. Journal of the North American Benthological Society 15:185-211. Battarbee, R. W. 2010. Aquatic Ecosystem Variability and Climate Change - A Paleoecological Perspective. Pages 119-151 in M. Kernan, R. Battarbee, and B. Moss, editors. Climate change impacts on freshwater ecosystems. Blackwell, Oxford. Beaumont, L. J. and L. Hughes. 2002. Potential changes in the distributions of latitudinally restricted Australian butterfly species in response to climate change. Global Change Biology 8:954-971. Begon, M., C. R. Townsend, and J. L. Harper. 2006. Ecology : from individuals to ecosystems. 4th edition edition. Blackwell publishing, Malden (USA), Oxford (UK) and Carlton (Australia), 738p. Beniston, M., H. F. Diaz, and R. S. Bradley. 1997. Climatic change at high elevation sites: An overview. Climatic Change 36:233-251. 205 Bernard, P. 1994. Les zones humides., Comité interministériel de l’évaluation des politiques publiques. Commissariat au Plan. La Documentation française, 391p. Biggs, J., A. Corfield, P. Gron, H. O. Hansen, D. Walker, M. Whitfield, and P. Williams. 1998. Restoration of the rivers Brede, Cole and Skerne: a joint Danish and British EU-LIFE demonstration project, V-Short-term impacts on the conservation value of aquatic macroinvertebrate and macrophyte assemblages. Aquatic Conservation : Marine and Freshwater Ecosystems 8:241-255. Biggs, J., P. Williams, M. Whitfield, P. Nicolet, and A. Weatherby. 2005. 15 years of pond assessment in Britain: results and lessons learned from the work of Pond Conservation. Aquatic Conservation-Marine and Freshwater Ecosystems 15:693-714. Biggs, J., P. J. Williams, M. Whitfield, G. Fox, P. Nicolet, and S. Howard. 2000. A new biological method for assessing the ecological quality of lentic waterbodies. Pages 235-250 in S. Wicherek, editor. L'eau, de la cellule au paysage. Elsevier, Paris. Bilton, D. T., J. R. Freeland, and B. Okamura. 2001. Dispersal in freshwater invertebrates. Annual Review of Ecology and Systematics 32:159-181. Blunck, H. 1914. Die Entwicklung des Dytiscus marginalis L. vom Ei bis zur Imago 1. Teil. Das Embryonalleben. Z. wiss. Zool. 111:76-151. Boix, D., S. Gascón, J. Sala, M. Martinoy, J. Gifre, and X. D. Quintana. 2005. A new index of water quality assessment in Mediterranean wetlands based on crustacean and insect assemblages: the case of Catalunya (NE Iberian peninsula). Aquatic Conservation-Marine and Freshwater Ecosystems 15:635-651. Bonada, N., S. Doledec, and B. Statzner. 2007. Taxonomic and biological trait differences of stream macroinvertebrate communities between mediterranean and temperate regions: implications for future climatic scenarios. Global Change Biology 13:1658-1671. Boon, P. J. 2000. The development of integrated methods for assessing river conservation value. Pages 413-428. Kluwer Academic Publ. Boon, P. J. and D. L. Howell. 1997. Freshwater Quality: Defining the Indefinable? The Stationery Office, Edinburgh, 552p. Boon, P. J. and C. M. Pringle. 2009. Assessing the conservation value of fresh waters. an international perspective. Cambridge University Press, Cambridge, 293p. Bornette, G., C. Amoros, and N. L. Lamouroux. 1998. Aquatic plant diversity in riverine wetlands: The role of connectivity. Freshwater Biology 39:267-283. Bornette, G., H. Piegay, A. Citterio, C. Amoros, and V. Godreau. 2001. Aquatic plant diversity in four river floodplains: a comparison at two hierarchical levels. Biodiversity and Conservation 10:1683-1701. Bornette, G. and S. Puijalon. 2011. Response of aquatic plants to abiotic factors: a review. Aquatic Sciences 73:1-14. Bradley, R. S., F. T. Keimig, and H. F. Diaz. 2004. Projected temperature changes along the American cordillera and the planned GCOS network. Geophysical Research Letters 31:L16210. Briers, R. A. and J. Biggs. 2005. Spatial patterns in pond invertebrate communities: separating environmental and distance effects. Aquatic Conservation-Marine and Freshwater Ecosystems 15:549-557. Broennimann, O., W. Thuiller, G. Hughes, G. F. Midgley, J. M. R. Alkemade, and A. Guisan. 2006. Do geographic distribution, niche property and life form explain plants' vulnerability to global change? Global Change Biology 12:1079-1093. Brönmark, C. 1985. Fresh-Water Snail Diversity - Effects of Pond Area, Habitat Heterogeneity and Isolation. Oecologia 67:127-131. Brönmark, C. and L. A. Hansson. 2000. The Biology of Lakes and Ponds. Oxford University Press, New York, 211p. Brönmark, C. and L. A. Hansson. 2002. Environmental issues in lakes and ponds: current state and perspectives. Environmental Conservation 29:290-307. 206 Brown, L. E., D. M. Hannah, and A. M. Milner. 2007. Vulnerability of alpine stream biodiversity to shrinking glaciers and snowpacks. Global Change Biology 13:958-966. Buisson, L., W. Thuiller, S. Lek, P. Lim, and G. Grenouillet. 2008. Climate change hastens the turnover of stream fish assemblages. Global Change Biology 14:2232-2248. Burgmer, T., H. Hillebrand, and M. Pfenninger. 2007. Effects of climate-driven temperature changes on the diversity of freshwater macroinvertebrates. Oecologia 151:93-103. Burnham, K. P. and W. S. Overton. 1979. Robust estimation of population size when capture probabilities vary among animals. Ecology 60:927-936. C2SM, MeteoSwiss, ETH, NCCR Climate, and OcCC. 2011. Swiss Climate Change Scenarios CH2011. Zurich, Switzerland, 88p. Calosi, P., D. T. Bilton, and J. I. Spicer. 2007. The diving response of a diving beetle: effects of temperature and acidification. Journal of Zoology 273:289-297. Calosi, P., D. T. Bilton, and J. I. Spicer. 2008. Thermal tolerance, acclimatory capacity and vulnerability to global climate change. Biology Letters 4:99-102. Carey, P. D., S. Wallis, P. M. Chamberlain, A. Cooper, A. Emmett, L. C. Maskell, T. McCann, J. Murphy, L. R. Norton, B. Reynolds, W. A. Scott, I. C. Smart, and J. M. Ullyett. 2008. Countryside Survey: UK Results from 2007., Centre for Ecology & Hydrobiology, Lancaster, 105p. Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N. Sharpley, and V. H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8:559-568. Carron, G. 2005. Kommentierte Checkliste der Dytiscidae und Noteridae (Coleoptera) der Schweiz. Mitteilungen der Entomologischen Gesellschaft 55:93-114. Carron, G. 2008. Checklist des coléoptères aquatiques de Suisse. Deuxième partie: Gyrinidae, Haliplidae, Paelobiidae, Sphaeriusidae. Mitteilungen der Entomologischen Gesellschaft 81:53-60. Chadd, R. and C. Extence. 2004. The conservation of freshwater macroinvertebrate populations: a community-based classification scheme. Aquatic Conservation-Marine and Freshwater Ecosystems 14:597-624. Chapin, F. S., E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L. Reynolds, D. U. Hooper, S. Lavorel, O. E. Sala, S. E. Hobbie, M. C. Mack, and S. Diaz. 2000. Consequences of changing biodiversity. Nature 405:234-242. Chase, J. M. and M. A. Leibold. 2002. Spatial scale dictates the productivity-biodiversity relationship. Nature 416:427-430. Chen, I.-C., J. K. Hill, R. Ohlemüller, D. B. Roy, and C. D. Thomas. 2011. Rapid Range Shifts of Species Associated with High Levels of Climate Warming. Science 333:1024-1026. Chovanec, A., J. Waringer, M. Straif, W. Graf, W. Reckendorfer, A. Waringer-Löschenkohl, H. Waidbacher, and H. Schultz. 2005. The Floodplain Index - a new approach for assessing the ecological status of river/floodplain-systems according to the EU Water Framework Directive. Archiv Fur Hydrobiologie Suppl. 155/1:169-185. Colwell, R. K. 2005. EstimateS: Statistical estimation of species richness and shared species from samples. Version 7.5. . User's guide published at: hhp://purl.oclc.org/estimates. Cooper, A., T. McCann, R. Davidson, and G. N. Foster. 2005. Vegetation, water beetles and habitat isolation in abandoned lowland bog drains and peat pits. Aquatic Conservation-Marine and Freshwater Ecosystems 15:175-188. Copp, G. H., A. S. Serhan Tarkan, M. J. Godard, N. J. Edmonds, and K. J. Wesley. 2010. Preliminary assessment of feral goldfish impacts on ponds, with particular reference to native crucian carp. Aquatic Invasions 5:413–422. Corbet, P. S. 1999. Dragonflies. Behaviour and ecology of Odonata. Harley Books, Colchester, 829p. Cordoba-Aguilar, A. 2008. Dragonflies: Model Organisms for Ecological and Evolutionary Research. Oxford University Press, Oxford, 304p. 207 Daufresne, M., P. Bady, and J. F. Fruget. 2007. Impacts of global changes and extreme hydroclimatic events on macroinvertebrate community structures in the French Rhone River. Oecologia 151:544-559. Daufresne, M. and P. Boet. 2007. Climate change impacts on structure and diversity of fish communities in rivers. Global Change Biology 13:2467-2478. Daufresne, M., K. Lengfellner, and U. Sommer. 2009. Global warming benefits the small in aquatic ecosystems. Proceedings of the National Academy of Sciences of the United States of America 106:12788-12793. De Meester, L., S. Declerck, R. Stoks, G. Louette, F. Van de Meutter, T. De Bie, E. Michels, and L. Brendonck. 2005. Ponds and pools as model systems in conservation biology, ecology and evolutionary biology. Aquatic Conservation-Marine and Freshwater Ecosystems 15:715-725. Diaz, H. F. and R. S. Bradley. 1997. Temperature variations during the last century at high elevation sites. Climatic Change 36:253-279. Dijkstra, J. A., E. L. Westerman, and L. G. Harris. 2011a. The effects of climate change on species composition, succession and phenology: a case study. Global Change Biology 17:2360-2369. Dijkstra, K.-D. B. and R. Lewington. 2006. Field Guide to the Dragonflies of Britain and Europe. British Wildlife Publishing, Gillingham, 320p. Dijkstra, K. D. B., J.-P. Boudot, V. Clausnitzer, J. Kipping, J. J. Kisakye, S. S. Ogbogu, B. Samraoui, M. J. Samways, K. Schutte, J. P. Simaika, F. Suhling, and S. L. Tchibozo. 2011b. Dragonflies and damselflies of Africa (Odonata): history, diversity, distribution, and conservation. . Pages 126173 in W. Darwall, K. Smith, D. Allen, R. Holland, I. Harrison, and E. Brooks, editors. The Diversity of Life in African Freshwaters: Underwater, Under Threat. . IUCN, Gland, Switzerland and Cambridge , UK. Dingemanse, N. J. and V. J. Kalman. 2008. Changing temperature regimes have advanced the phenology of Odonata in the Netherlands. Ecological Entomology 33:394-402. Dodd, C. K. 2010. Amphibian Ecology and Conservation. A handbook of techniques. Oxford University Press, Oxford, 556p. Dodds, W. K. 2002. Freshwater ecology. Concepts and environmental applications. Academic Press, San Diego, 569p. Dodson, S. I., S. E. Arnott, and K. L. Cottingham. 2000. The relationship in lake communities between primary productivity and species richness. Ecology 81:2662-2679. Domisch, S., S. C. Jähnig, and P. Haase. 2011. Climate-change winners and losers: stream macroinvertebrates of a submontane region in Central Europe. Freshwater Biology 56:2009– 2020. Dommanget, J.-L. 1998. Les Libellules et leurs habitats. Société Française d'Odonatologie, Bois-d'Arcy, 20p. Downing, J. A., Y. T. Prairie, J. J. Cole, C. M. Duarte, L. J. Tranvik, R. G. Striegl, W. H. McDowell, P. Kortelainen, N. F. Caraco, J. M. Melack, and J. J. Middelburg. 2006. The global abundance and size distribution of lakes, ponds, and impoundments. Limnology and Oceanography 51:23882397. Dudgeon, D., A. H. Arthington, M. O. Gessner, Z. I. Kawabata, D. J. Knowler, C. Leveque, R. J. Naiman, A. H. Prieur-Richard, D. Soto, M. L. J. Stiassny, and C. A. Sullivan. 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biological Reviews 81:163-182. Duker, L. and M. Palmer. 2009. Methods for assessing the conservation value of lakes. Pages 166-199 in P. J. Boon and C. M. Pringle, editors. Assessing the conservation value of fresh waters. an international perspective. Cambridge University Press, Cambridge. Durance, I. and S. J. Ormerod. 2007. Climate change effects on upland stream macroinvertebrates over a 25-year period. Global Change Biology 13:942-957. Durance, I. and S. J. Ormerod. 2009. Trends in water quality and discharge confound long-term warming effects on river macroinvertebrates. Freshwater Biology 54:388-405. 208 Durant, J. M., D. O. Hjermann, G. Ottersen, and N. C. Stenseth. 2007. Climate and the match or mismatch between predator requirements and resource availability. Climate Research 33:271-283. EPCN. 2007. Developping the Pond Manifesto. Annales De Limnologie-International Journal of Limnology 43:221-232. EPCN. 2008. The Pond Manifesto. 20p. EPCN. 2011. Rationale for Pond Conservation, http://campus.hesge.ch/epcn/history-rationale.asp. European Environment Agency. 2008. Impacts of Europe's changing climate — 2008 indicator-based assessment. European Environment Agency, Copenhagen, 247p. Feder, M. E. and W. W. Burggren. 1992. Environmental Physiology of the Amphibians. University of Chicago Press, Chigago, IL., 634p. Feuchtmayr, H., D. McKee, I. F. Harvey, D. Atkinson, and B. Moss. 2007. Response of macroinvertebrates to warming, nutrient addition and predation in large-scale mesocosm tanks. Hydrobiologia 584:425-432. Feuchtmayr, H., R. Moran, K. Hatton, L. Connor, T. Heyes, B. Moss, I. Harvey, and D. Atkinson. 2009. Global warming and eutrophication: effects on water chemistry and autotrophic communities in experimental hypertrophic shallow lake mesocosms. Journal of Applied Ecology 46:713-723. Fischer, J. and D. B. Lindenmayer. 2007. Landscape modification and habitat fragmentation: a synthesis Global Ecology and Biogeography 16:265-280. Flanagan, K. M., E. McCauley, F. Wrona, and T. Prowse. 2003. Climate change: the potential for latitudinal effects on algal biomass in aquatic ecosystems. Canadian Journal of Fisheries and Aquatic Sciences 60:635-639. Fleishman, E., R. F. Noss, and B. R. Noon. 2006. Utility and limitations of species richness metrics for conservation planning. Ecological Indicators 6:543-553. Forsberg, C. 1964. Phosphorus, a maximum factor in the growth of Characeae. Nature 201:517-518. Foster, G. N., A. P. Foster, M. D. Eyre, and D. T. Bilton. 1989. Classification of water beetle assemblages in arable fenland and ranking of sites in relation to conservation value. Freshwater Biology 22:343-354. Foster, G. N., B. H. Nelson, D. T. Bilton, D. A. Lott, R. Merritt, R. S. Weyl, and M. D. Eyre. 1992. A classification and evaluation of Irish water beetle assemblages. Aquatic Conservation 2:185208. Gaston, K. J. and T. M. Blackburn. 2000. Pattern and Process in Macroecology. Blackwell Science Ltd, Oxford, 377p. Gaston, K. J. and J. I. Spicer. 2004. Biodiversity. An Introduction. 2nd edition. Blackwell Science Ltd, Malden, Oxford & Victoria, 191p. Gauthier, P., M. Debussche, and J. D. Thompson. 2010. Regional priority setting for rare species based on a method combining three criteria. Biological Conservation 143:1501-1509. Gee, J. H. R., B. D. Smith, K. M. Lee, and S. W. Griffiths. 1997. The ecological basis of freshwater pond management for biodiversity. Aquatic Conservation-Marine and Freshwater Ecosystems 7:91-104. Godreau, V., G. Bornette, B. Frochot, C. Amoros, E. Castella, B. Oertli, F. Chambaud, D. Oberti, and E. Craney. 1999. Biodiversity in the floodplain of Saône: a global approach. Biodiversity and Conservation 8:839-864. Gonseth, Y. and C. Monnerat. 2001. Recent changes in distribution of dragonflies in Switzerland (Odonata). Proc. 13th Int. Coll. EIS:23-31. Gonseth, Y. and C. Monnerat. 2002. Liste rouge des espèces menacées en Suisse. Libellules. OFEFP & CSCF, Berne & Neuchâtel, 46p. Goubier-Martin, M. 1991. Climat et aquaculture application aux étangs de la Dombes. Université Jean Moulin - Lyon III, Lyon, 483p. Grabherr, G., M. Gottfried, and H. Pauli. 1994. Climate Effects on Mountain Plants. Nature 369:448448. 209 Graham, N. A. J., P. Chabanet, R. D. Evans, S. Jennings, Y. Letourneur, M. A. MacNeil, T. R. McClanahan, M. C. Ohman, N. V. C. Polunin, and S. K. Wilson. 2011. Extinction vulnerability of coral reef fishes. Ecology Letters 14:341–348. Grand, D. and J.-P. Boudot. 2006. Les libellules de France, Belgique et Luxembourg. Biotope, Mèze, (Collection Parthénope), 480p. Grimen, J. P. 2002. Plant strategies, vegetation processes, and ecosystem properties. Second edition. John Wiley and Sons, Chichester, UK, 456p. Gyllstrom, M., L. A. Hansson, E. Jeppesen, F. Garcia-Criado, E. Gross, K. Irvine, T. Kairesalo, R. Kornijow, M. R. Miracle, M. Nykanen, T. Noges, S. Romo, D. Stephen, E. Van Donk, and B. Moss. 2005. The role of climate in shaping zooplankton communities of shallow lakes. Limnology and Oceanography 50:2008-2021. Hannah, L., G. F. Midgley, T. Lovejoy, W. J. Bond, M. Bush, J. C. Lovett, D. Scott, and F. I. Woodward. 2002. Conservation of biodiversity in a changing climate. Conservation Biology 16:264-268. Hanski, I. 1991. Single-species metapopulation dynamics: concepts, models and observations. Pages 17-38 in M. E. Gilpin and I. Hanski, editors. Metapopulation Dynamics. Academic Press, London. Haslam, S. M. 1982. Major factors determining the distribution of macrophytic vegetation in the watercourses of the european economic community. Pages 105-111 in EWRSGH Symposium on Aquatic Weeds. Hassall, C., D. J. Thompson, G. C. French, and I. F. Harvey. 2007. Historical changes in the phenology of British Odonata are related to climate. Global Change Biology 13:933-941. Hastie, T. J. and R. J. Tibshirani. 1990. Generalized Additive Models. Chapman & Hall, London, 352p. Haury, J., M.-C. Peltre, M. Tremolieres, J. Barbe, G. Thiebaut, I. Bernez, H. Daniel, P. Chatenet, G. Haan-Archipof, S. Muller, A. Dutartre, C. Laplace-Treyture, A. Cazaubon, and E. LambertServien. 2006. A new method to assess water trophy and organic pollution - the Macrophyte Biological Index for Rivers (IBMR): its application to different types of river and pollution. Hydrobiologia 1:153-158. Heikkinen, R. K., M. Luoto, N. Leikola, J. Poyry, J. Settele, O. Kudrna, M. Marmion, S. Fronzek, and W. Thuiller. 2010. Assessing the vulnerability of European butterflies to climate change using multiple criteria. Biodiversity and Conservation 19:695-723. Heino, J., R. Virkkala, and H. Toivonen. 2009. Climate change and freshwater biodiversity: detected patterns, future trends and adaptations in northern regions. Biological Reviews 84:39-54. Heller, N. E. and E. S. Zavaleta. 2009. Biodiversity management in the face of climate change: A review of 22 years of recommendations. Biological Conservation 142:14-32. Hellmann, J. J., J. E. Byers, B. G. Bierwagen, and J. S. Dukes. 2008. Five potential consequences of climate change for invasive species. Conservation Biology 22:534-543. Henderson, P. A. 2007. Discrete and continuous change in the fish community of the Bristol Channel in response to climate change. Journal of the Marine Biological Association of the United Kingdom 87:589-598. Hering, D., C. K. Feld, O. Moog, and T. Ofenbock. 2006. Cook book for the development of a Multimetric Index for biological condition of aquatic ecosystems: Experiences from the European AQEM and STAR projects and related initiatives. Hydrobiologia 566:311-324. Hering, D., A. Haidekker, A. Schmidt-Kloiber, T. Barker, L. Buisson, W. Graf, G. Grenouillet, A. Lorenz, L. Sandin, and S. Stendera. 2010. Monitoring the responses of freshwater ecosystems to climate change. Pages 84-118 in M. Kernan, R. Battarbee, and B. Moss, editors. Climate change impacts on freshwater ecosystems. Blackwell, Oxford. Hering, D., A. Schmidt-Kloiber, J. Murphy, S. Lucke, C. Zamora-Munoz, M. J. Lopez-Rodriguez, T. Huber, and W. Graf. 2009. Potential impact of climate change on aquatic insects: A sensitivity analysis for European caddisflies (Trichoptera) based on distribution patterns and ecological preferences. Aquatic Sciences 71:3-14. Hickling, R., D. B. Roy, J. K. Hill, R. Fox, and C. D. Thomas. 2006. The distributions of a wide range of taxonomic groups are expanding polewards. Global Change Biology 12:450-455. 210 Hiddink, J. G. and R. ter Hofstede. 2008. Climate induced increases in species richness of marine fishes. Global Change Biology 14:453-460. Hill, J. K., C. D. Thomas, and B. Huntley. 2001. Climate and recent changes in butterflies. Pages 77-88 in G. R. Walther, C. A. Burga, and P. J. Edwards, editors. "Fingerprints" of Climate Change Adapted Behaviour and Shifting Species Ranges. Kluwer Academic/Plenum Publishers, New York. Hinden, H., B. Oertli, N. Menetrey, L. Sager, and J. B. Lachavanne. 2005. Alpine pond biodiversity: what are the related environmental variables? Aquatic Conservation-Marine and Freshwater Ecosystems 15:613-624. Hodkinson, I. D. 2005. Terrestrial and aquatic invertebrates as bioindicators for environmental monitoring, with particular reference to mountain ecosystems. Environmental Management 35:649-666. Hoegh-Guldberg, O., L. Hughes, S. McIntyre, D. B. Lindenmayer, C. Parmesan, H. P. Possingham, and C. D. Thomas. 2008. Assisted colonization and rapid climate change. Science 321:345-346. Hughes, L. 2000. Biological consequences of global warming: is the signal already apparent? Trends in Ecology & Evolution 15:56-61. Hulme, P. E. 2005. Adapting to climate change: is there scope for ecological management in the face of a global threat? Journal of Applied Ecology 42:784-794. Hulot, F. D., G. Lacroix, F. O. Lescher-Moutoue, and M. Loreau. 2000. Functional diversity governs ecosystem response to nutrient enrichment. Nature 405:340-344. Hunter, M., E. Dinerstein, J. Hoekstra, and D. Lindenmayer. 2010. A Call to Action for Conserving Biological Diversity in the Face of Climate Change. Conservation Biology 24:1169-1171. Imboden, C. 1976. Eaux vivantes - Initiation à la biologie des zones humides. Ligue Suisse pour la Protection de la Nature, Bâle, 240p. Indermuehle, N., S. Angélibert, V. Rosset, and B. Oertli. 2010. The pond biodiversity index "IBEM": a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 2. Method description and examples of application. Limnetica 29:105-119. IPCC. 2007a. Climate Change 2007 - The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the IPCC. 996p. IPCC. 2007b. Fourth Assessment Report. Climate Change 2007: Climate Change Impacts, Adaptation and Vulnerability. Cambridge University Press, Cambridge, New-York, 976p. IUCN. 2001. IUCN Red List Categories and Criteria: version 3.1. . IUCN Species Survival Comission, Gland, Switzerland and Cambridge, UK, 30p. Iverson, L. R. and A. M. Prasad. 2001. Potential changes in tree species richness and forest community types following climate change. Ecosystems 4:186-199. Jackson, S. T. and J. T. Overpeck. 2000. Responses of plant populations and communities to environmental changes of the late Quaternary. Paleobiology 26:194-220. Jacobsen, D. 2008. Low oxygen pressure as a driving factor for the altitudinal decline in taxon richness of stream macroinvertebrates. Oecologia 154:795-807. Jeffries, M. 2008. The spatial and temporal heterogeneity of macrophyte communities in thirty small, temporary ponds over a period of ten years. Ecography 31:765-775. Jeffries, M. J. 2003. Idiosyncratic relationships between pond invertebrates and environmental, temporal and patch-specific predictors of incidence. Ecography 26:311-324. Jeppesen, E., J. P. Jensen, M. Sondergaard, T. Lauridsen, and F. Landkildehus. 2000. Trophic structure, species richness and biodiversity in Danish lakes: changes along a phosphorus gradient. Freshwater Biology 45:201-218. Jeppesen, E., B. Moss, H. Bennion, L. Carvalho, L. De Meester, H. Feuchtmayr, N. Friberg, M. O. Gessner, M. Hefting, T. Lauridsen, L. Liboriussen, H. L. Malmquist, L. May, M. Meerhoff, J. S. Olafsson, M. B. Soons, and J. T. A. Verhoeven. 2010. Interaction of climate change and eutrophication. Pages 119-151 in M. Kernan, R. Battarbee, and B. Moss, editors. Climate change impacts on freshwater ecosystems. Blackwell, Oxford. 211 Jiguet, F., A. S. Gadot, R. Julliard, S. E. Newson, and D. Couvet. 2007. Climate envelope, life history traits and the resilience of birds facing global change. Global Change Biology 13:1672-1684. Joly, P., C. Miaud, A. Lehmann, and O. Grolet. 2001. Habitat matrix effects on pond occupancy in newts. Conservation Biology 15:239-248. Jones, J. I., W. Li, and S. C. Maberly. 2003. Area, altitude and aquatic plant diversity. Ecography 26:411-420. Kati, V., P. Devillers, M. Dufrene, A. Legakis, D. Vokou, and P. Lebrun. 2004. Testing the value of six taxonomic groups as biodiversity indicators at a local scale. Conservation Biology 18:667-675. Keatley, B. E., M. S. V. Douglas, and J. P. Smol. 2006. Early-20th century environmental changes inferred using subfossil diatoms from a small pond on Melville Island, NWT, Canadian high Arctic. Hydrobiologia 553:15-26. Keith, D. A. 1998. An evaluation and modification of World Conservation Union Red List criteria for classification of extinction risk in vascular plants. Conservation Biology 12:1076-1090. Kernan, M., R. Battarbee, and B. Moss. 2010. Climate change impacts on freshwater ecosystems. Blackwell Publishing, Oxford, 314p. Khan, F. A. and A. A. Ansari. 2005. Eutrophication: An ecological vision. The Botanical Review 71:449482. Kienast, F., O. Wildi, and B. Brzeziecki. 1998. Potential impacts of climate change on species richness in mountain forests - an ecological risk assessment. Biological Conservation 83:291-305. Kiesecker, J. M., A. R. Blaustein, and L. K. Belden. 2001. Complex causes of amphibian population declines. Nature 410:681-684. Klanderud, K. and O. Totland. 2007. The relative role of dispersal and local interactions for alpine plant community diversity under simulated climate warming. Oikos 116:1279-1288. Körner, C. and G. R. Walther. 2001. Fingerprints of climate change - concluding remarks. Pages 305316 in G. R. Walther, C. A. Burga, and P. J. Edwards, editors. "Fingerprints" of Climate Change - Adapted Behaviour and Shifting Species Ranges. Kluwer Academic/Plenum Publishers, New York. Kotiaho, J. S., V. Kaitala, A. Komonen, and J. Paivinen. 2005. Predicting the risk of extinction from shared ecological characteristics. Proceedings of the National Academy of Sciences of the United States of America 102:1963-1967. Lachavanne, J.-B. 1985. The influence of accelerated eutrophication on the macrophytes of Swiss lakes: Abundance and distribution. Verh.Internat.Verein.Limnol. 22:2950-2955. Lachavanne, J. B., R. Juge, and J. Perfetta. 1991. The consequences of water oligotrophication on macrophytic vegetation of Swiss lakes. Pages 943-948 in International Association of Theoretical and Applied Limnology Proceedings. E Schweizerbart'sche Verlagsbuchhandlung, Stuttgart. Landolt, E. 1977. Ökologische Zeigerwerte zur Schweizer Flora, Zürich, 208p. Lawler, J. J. 2009. Climate Change Adaptation Strategies for Resource Management and Conservation Planning. Pages 79-98 Year in Ecology and Conservation Biology 2009. Blackwell Publishing, Oxford. Lazzarotto, J. and F. Rapin. 2010. Evolution physico-chimique des eaux du Léman. Pages 31-56 Rapport de la Comission internationale de protection des eaux du Léman. Campagne 2009. CIPEL. Le réseau des Botanistes Francophones. 2010. http://www.tela-botanica.org/page:eflore. Leclerc, D., S. Angélibert, V. Rosset, and B. Oertli. 2010. Les libellules (Odonates) des étangs piscicoles de la Dombes. Martinia 26:98-108. Ledger, M. E., F. K. Edwards, L. E. Brown, A. M. Milner, and G. Woodward. 2011. Impact of simulated drought on ecosystem biomass production: an experimental test in stream mesocosms. Global Change Biology 17:2288-2297. Lehmann, A. and J. B. Lachavanne. 1999. Changes in the water quality of Lake Geneva indicated by submerged macrophytes. Freshwater Biology 42:457-466. 212 Lehmann, A., J. M. Overton, and J. R. Leathwick. 2002. GRASP: generalized regression analysis and spatial prediction. Ecological Modelling 157:189-207. Lemoine, N. and K. Bohning-Gaese. 2003. Potential impact of global climate change on species richness of long-distance migrants. Conservation Biology 17:577-586. Lenoir, J., J. Gégout, P. Marquet, P. de Ruffray, and H. Brisse. 2008. A significant upward shift in plant species optimum elevation during the 20th century. Science 320:1768-1771. Levins, R. 1969. Some demographic and genetic consequences of environmental heterogeneity for biological control. Bulletin of the Entomological Society of America 15:237-240. Levins, R. 1970. Extinction. Pages 77-107 in M. Gerstenhaber, editor. Some mathematical questions in biology. Lectures on Mathematics in Life Sciences. American Mathematical Society, Providence. Levinsky, I., F. Skov, J. C. Svenning, and C. Rahbek. 2007. Potential impacts of climate change on the distributions and diversity patterns of European mammals. Biodiversity and Conservation 16:3803-3816. Livingstone, D. M. 2000. Large-scale climatic forcinf detected in historical observations of lake ice break-up. Verh. Internat. Verein. Limnol. 27:2775-2783. Lods-Crozet, B., E. Castella, D. Cambin, C. Ilg, S. Knispel, and H. Mayor-Simeant. 2001. Macroinvertebrate community structure in relation to environmental variables in a Swiss glacial stream. Freshwater Biology 46:1641-1661. Loss, S. R., L. A. Terwilliger, and A. C. Peterson. 2011. Assisted colonization: Integrating conservation strategies in the face of climate change. Biological Conservation 144:92-100. Macan, T. T. 1970. Biological studies of the English lakes. Longman, London, 260p. Macan, T. T. 1975. Freshwater Ecology. Longman, London, 343p. MacArthur, R. H. and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey, 203p. Maggini, R., A. Lehmann, N. E. Zimmermann, and A. Guisan. 2006. Improving generalized regression analysis for the spatial prediction of forest communities. Journal of Biogeography 33:17291749. Magnuson, J. J., D. M. Robertson, B. J. Benson, R. H. Wynne, D. M. Livingstone, T. Arai, R. A. Assel, R. G. Barry, V. Card, E. Kuusisto, N. G. Granin, T. D. Prowse, K. M. Stewart, and V. S. Vuglinski. 2000. Historical trends in lake and river ice cover in the Northern hemisphere. Science 289:1743-1746. Magurran, A. E. 2004. Measuring biological diversity. Blackwell Publishing, Malden, Oxford & Carlton, 256p. Manel, S. M., M. Schwartz, G. Luikart, and P. Taberlet. 2003. Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology and Evolution 18:189-197. Marti, K., O. K. Krüsi, J. Heeb, and E. Theis. 1997. Clé de détermination des zones tampon. Guide pour déterminer des zones-tampon suffisantes du point de vue écologique pour les marais. 2e edition. Office fédéral de l'environnement, des forêts et du paysage, Berne, 50p. Mattila, M., J. S. Kotiaho, V. Kaitala, and A. Komonen. 2008. The use of ecological traits in extinction risk assessments: A case study on geometrid moths. Biological Conservation 141:2322-2328. McCarty, J. P. 2001. Ecological consequences of recent climate change. Conservation Biology 15:320331. McKee, D., D. Atkinson, S. Collings, J. Eaton, I. Harvey, T. Heyes, K. Hatton, D. Wilson, and B. Moss. 2002. Macro-zooplankter responses to simulated climate warming in experimental freshwater microcosms. Freshwater Biology 47:1557-1570. McKee, D., D. Atkinson, S. E. Collings, J. W. Eaton, A. B. Gill, I. Harvey, K. Hatton, T. Heyes, D. Wilson, and B. Moss. 2003. Response of freshwater microcosm communities to nutrients, fish, and elevated temperature during winter and summer. Limnology and Oceanography 48:707-722. Meerhoff, M., J. M. Clemente, F. T. De Mello, C. Iglesias, A. R. Pedersen, and E. Jeppesen. 2007. Can warm climate-related structure of littoral predator assemblies weaken the clear water state in shallow lakes? Global Change Biology 13:1888-1897. 213 Melzer, A. 1976. Makrophytische Wasserpflanzen als Indikatoren des Gewässerzustandes oberbayerischer Seen, dargestellt im Rahmen limnologischer Untersuchungen an den Osterseen und den Eggstätt-Hemhofer Seen (Oberbayern). Vaduz, 195p. Menetrey, N., B. Oertli, and J. B. Lachavanne. 2010. The CIEPT: A macroinvertebrate-based multimetric index for assessing the ecological quality of Swiss lowland ponds. Ecological Indicators 11:590-600. Menetrey, N., B. Oertli, M. Sartori, and J. B. Lachavanne. 2008. Eutrophication: are mayflies (Ephemeroptera) good bioindicators for ponds? Hydrobiologia 597:125-135. Menetrey, N., L. Sager, B. Oertli, and J. B. Lachavanne. 2005. Looking for metrics to assess the trophic state of ponds. Macroinvertebrates and amphibians. Aquatic Conservation-Marine and Freshwater Ecosystems 15:653-664. Menzel, A., T. H. Sparks, N. Estrella, E. Koch, A. Aasa, R. Ahas, K. Alm-Kubler, P. Bissolli, O. Braslavska, A. Briede, F. M. Chmielewski, Z. Crepinsek, Y. Curnel, A. Dahl, C. Defila, A. Donnelly, Y. Filella, K. Jatcza, F. Mage, A. Mestre, O. Nordli, J. Penuelas, P. Pirinen, V. Remisova, H. Scheifinger, M. Striz, A. Susnik, A. J. H. Van Vliet, F. E. Wielgolaski, S. Zach, and A. Zust. 2006. European phenological response to climate change matches the warming pattern. Global Change Biology 12:1969-1976. Meyer, A., S. Zumbach, B. Schmidt, and J.-C. Monney. 2009. Les amphibiens et les reptiles de Suisse. Haupt, Bern, 336p. Micheli, F. 1999. Eutrophication, fisheries, and consumer-resource dynamics in marine pelagic ecosystems. Science 285:1396-1398. Millennium Ecosystem Assessment. 2005. Ecosystems and human well-being: wetlands and water. Synthesis. Washington, 68p. Mittelbach, G. G., C. F. Steiner, S. M. Scheiner, K. L. Gross, H. L. Reynolds, R. B. Waide, M. R. Willig, S. I. Dodson, and L. Gough. 2001. What is the observed relationship between species richness and productivity? Ecology 82:2381-2396. Moisen, G. G. and T. S. Frescino. 2002. Comparing five modelling techniques for predicting forest characteristics. Ecological Modelling 157:209-225. Mooij, W. M., J. H. Janse, L. Domis, S. Hulsmann, and B. W. Ibelings. 2007. Predicting the effect of climate change on temperate shallow lakes with the ecosystem model PCLake. Hydrobiologia 584:443-454. Mooney, H. A. and R. J. Hobbs. 2000. Invasive Species in a Changing World. Island Press, Washington, DC, USA, 461p. Moore, P., S. J. Hawkins, and R. C. Thompson. 2007. Role of biological habitat amelioration in altering the relative responses of congeneric species to climate change. Marine Ecology Progress Series 334:11-19. Morin, X., M. J. Lechowicz, C. Augspurger, J. O' Keefe, D. Viner, and I. Chuine. 2009. Leaf phenology in 22 North American tree species during the 21st century. Global Change Biology 15:961-975. Moser, D. M., A. Gygax, B. Bäumler, N. Wyler, and R. Palese. 2002. Liste rouge des espèces menacées de Suisse. Fougères et plantes à fleurs. Ed. Office fédéral de l’environnement, des forêts et du paysage, Berne; Centre du Réseau Suisse de Floristique, Chambésy; Conservatoire et Jardin botaniques de la Ville de Genève, Chambésy, 118p. Moss, B., D. McKee, D. Atkinson, S. E. Collings, J. W. Eaton, A. B. Gill, I. Harvey, K. Hatton, T. Heyes, and D. Wilson. 2003. How important is climate? Effects of warming, nutrient addition and fish on phytoplankton in shallow lake microcosms. Journal of Applied Ecology 40:782-792. Musolin, D. L. 2007. Insects in a warmer world: ecological, physiological and life-history responses of true bugs (Heteroptera) to climate change. Global Change Biology 13:1565-1585. Nagy, L. and G. Grabherr. 2009. The Biology of Alpine Habitats. Oxford Unversity Press, Oxford, 376p. Nicolet, P., J. Biggs, G. Fox, M. J. Hodson, C. Reynolds, M. Whitfield, and P. Williams. 2004. The wetland plant and macroinvertebrate assemblages of temporary ponds in England and Wales. Biological Conservation 120:261-278. 214 Noges, P., W. Van de Bund, A. C. Cardoso, and A. S. Heiskanen. 2007. Impact of climatic variability on parameters used in typology and ecological quality assessment of surface waters implications on the Water Framework Directive. Hydrobiologia 584:373-379. North, N., N. Kljun, F. Kasser, J. Heldstab, M. Maibach, J. Reutimann, and M. Guyer. 2007. Changements climatiques en Suisse - Indicateurs des causes, des effets et des mesures., Berne, 77p. O'Grady, J. J., D. H. Reed, B. W. Brook, and R. Frankham. 2004. What are the best correlates of predicted extinction risk? Biological Conservation 118:513-520. OcCC. 2008. Le climate change - que faire? Le nouveau rapport des Nations Unies sur le climat (GIEC 2007) et ses principaux résultats dans l'optique de la Suisse., Berne, 47p. OcCC and ProClim–. 2007. Les changements climatiques et la Suisse en 2050. Impacts attendus sur l’environnement, la société et l’économie. Bern, 168p. OCDE. 1982. Eutrophisation des eaux. Méthodes de surveillance, d'évaluation et de lutte., OCDE, Paris, 174p. Oertli, B. 1995. Odonates de la vallée de la Saône. Martinia 11:35-42. Oertli, B. 2008. The use of dragonflies in the assessment and monitoring of aquatic habitats. Pages 79-95 in A. Córdoba-Aguilar, editor. Dragonflies and Damselflies : Model Organisms for Ecological and Evolutionary Research. Oxford University Press, Oxford, New York. Oertli, B., D. Auderset Joye, E. Castella, R. Juge, and J. B. Lachavanne. 2000. Diversité biologique et typologie écologique des étangs et petits lacs de Suisse. University of Geneva & Swiss Agency for the Environment, Forests and Landscape, Geneva, 434p. Oertli, B., J. Biggs, R. Cereghino, P. Grillas, A. Jolly, and J.-B. Lachavanne. 2005a. Conservation and monitoring of pond biodiversity : introduction. Aquatic Conservation-Marine and Freshwater Ecosystems 15:535–540. Oertli, B., R. Cereghino, A. Hull, and M. R. Miracle. 2009. Pond conservation: from science to practice. Hydrobiologia 634:1-9. Oertli, B. and P. A. Frossard. in prep. Mares et étangs : écologie, conservation, gestion, valorisation. . Presses Polytechniques Universitaires Romandes, Lausanne, XXp. Oertli, B., N. Indermuehle, S. Angelibert, H. Hinden, and A. Stoll. 2008. Macroinvertebrate assemblages in 25 high alpine ponds of the Swiss National Park (Cirque of Macun) and relation to environmental variables. Hydrobiologia 597:29-41. Oertli, B., D. A. Joye, E. Castella, R. Juge, A. Lehmann, and J. B. Lachavanne. 2005b. PLOCH: a standardized method for sampling and assessing the biodiversity in ponds. Aquatic Conservation-Marine and Freshwater Ecosystems 15:665-679. Oertli, B., D. A. Joye, N. Indermuehle, R. Juge, and J. B. Lachavanne. 2004. 1st European Pond Workshop - "Conservation and monitoring of pond biodiversity". Archives Des Sciences 57:69-71. Oertli, B., S. Schmid, and J. Bolliger. 2011. Projet RestorAlps HES-SO RealTech rapport soumis, 23p. OFEV, O. f. d. l. e. 2011. Liste der National Prioritären Arten. Arten mit nationaler Priorität für die Erhaltung und Förderung, Stand 2010., Bern, 132p. Ott, J. 2001. Expansion of Mediterranean Odonata in Germany and Europe - consequences of climatic changes. Pages 89-111 in G. R. Walther, C. A. Burga, and P. J. Edwards, editors. "Fingerprints" of Climate Change - Adapted Behaviour and Shifting Species Ranges. Kluwer Academic/Plenum Publishers, New York. Painter, D. 1999. Macroinvertebrate distributions and the conservation value of aquatic Coleoptera, Mollusca and Odonata in the ditches of traditionally managed and grazing fen at Wicken Fen, UK. Journal of Applied Ecology 36:33-48. Park, S., M. T. Brett, A. Muller-Solger, and C. R. Goldman. 2004. Climatic forcing and primary productivity in a subalpine lake: Interannual variability as a natural experiment. Limnology and Oceanography 49:614-619. Parmesan, C. 1996. Climate and species' range. Nature 382:765-766. 215 Parmesan, C. 2006. Ecological and evolutionary responses to recent climate change. Annual Review of Ecology Evolution and Systematics 37:637-669. Parmesan, C., N. Ryrholm, C. Stefanescu, J. K. Hill, C. D. Thomas, H. Descimon, B. Huntley, L. Kaila, J. Kullberg, T. Tammaru, W. J. Tennent, J. A. Thomas, and M. Warren. 1999. Poleward shifts in geographical ranges of butterfly species associated with regional warming. Nature 399:579583. Parmesan, C. and G. Yohe. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421:37-42. Pauchard, A., C. Kueffer, H. Dietz, C. C. Daehler, J. Alexander, P. J. Edwards, J. R. Arevalo, L. A. Cavieres, A. Guisan, S. Haider, G. Jakobs, K. McDougall, C. I. Millar, B. J. Naylor, C. G. Parks, L. J. Rew, and T. Seipel. 2009. Ain't no mountain high enough: plant invasions reaching new elevations. Frontiers in Ecology and the Environment 7:479-486. Pauli, H., M. Gottfried, K. Reier, C. Klettner, and G. Grabherr. 2007. Signals of range expansions and contractions of vascular plants in the high Alps: observations (1994-2004) at the GLORIA*master site Schrankogel, Tyrol, Austria. Global Change Biology 13:147-156. Penning, W. E., M. Mjelde, B. Dudley, S. Hellsten, J. Hanganu, A. Kolada, M. van den Berg, S. Poikane, G. Phillips, N. Willby, and F. Ecke. 2008. Classifying aquatic macrophytes as indicators of eutrophication in European lakes. Aquatic Ecology 42:237-251. Poff, N. L. 1997. Landscape filters and species traits: Towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391409. Poff, N. L., M. M. Brinson, and J. W. Day. 2002. Aquatic Ecosystems and Global Climate Change. Potential Impacts on Inland Freshwater and Coastal Wetland Ecosystems in the United States. Pew Center on Global Climate Change, Arlington, 45p. Portnoy, J. W. 1991. Summer oxygen depletion in a diked New-England estuary. Estuaries 14:122129. Pounds, J. A., M. R. Bustamante, L. A. Coloma, J. A. Consuegra, M. P. L. Fogden, P. N. Foster, E. La Marca, K. L. Masters, A. Merino-Viteri, R. Puschendorf, S. R. Ron, G. A. Sanchez-Azofeifa, C. J. Still, and B. E. Young. 2006. Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439:161-167. Pradel, J., T. Martin, D. Rey, R. Foussadier, and D. Bicout. 2008. Culex modestus (Ficalbi), vecteur de virus West Nile en Camargue: une espèce en augmentation dans la Dombe? Epidémiologie et santé animale 53:105-119. Preston, K., J. T. Rotenberry, R. A. Redak, and M. F. Allen. 2008. Habitat shifts of endangered species under altered climate conditions: importance of biotic interactions. Global Change Biology 14:2501-2515. Rahbek, C. 1995. The Elevational Gradient of Species Richness - a Uniform Pattern. Ecography 18:200-205. Rahel, F. J. and J. D. Olden. 2008. Assessing the effects of climate change on aquatic invasive species. Conservation Biology 22:521-533. Randin, C. F. 2007. Providing more informative projections of climate change impact on plant distribution in a mountain environment. Université de Lausanne, Lausanne, 321p. Rawson, D. S. 1945. The experimental introduction of small mouth black bass into lakes of the Prince Albert National Park, Saskatchewan. Trans. Amer. Fish Soc. 73:19-31. Rebetez, M. 2006. La Suisse se réchauffe Effet de serre et changement climatique. Presses ploytechniques et universitaires romandes, Lausanne, 142p. Rebetez, M. and M. Reinhard. 2008. Monthly air temperature trends in Switzerland 1901-2000 and 1975-2004. Theoretical and Applied Climatology 91:27-34. Ricciardi, A. and J. B. Rasmussen. 1999. Extinction rates of North American freshwater fauna. Conservation Biology 13:1220-1222. 216 Riis, T. and K. Sand-Jensen. 2001. Historical changes in species composition and richness accompanying perturbation and eutrophication of Danish lowland streams over 100 years. Freshwater Biology 46:269-280. Robinson, C. T. and B. Oertli. 2009. Long-term Biomonitoring of Alpine Waters in the Swiss National Park. Eco.mont 1:23-34. Rodwell, J. S. 2000. British Plant Communities: Aquatic Communities, Swamps and Tall-Herb Fens Cambridge University Press, Cambridge, 282p. Rolston, H. I. 1994. Conserving natural value. Columbia University Press, New York, 259p. Root, T. L., J. T. Price, K. R. Hall, S. H. Schneider, C. Rosenzweig, and J. A. Pounds. 2003. Fingerprints of global warming on wild animals and plants. Nature 421:57-60. Rosenberg, D. M. and V. H. Resh. 1993. Freshwater biomonitoring and benthic macroinvertebrates, New York, London, 488p. Rosenzweig, C., D. Karoly, M. Vicarelli, P. Neofotis, Q. G. Wu, G. Casassa, A. Menzel, T. L. Root, N. Estrella, B. Seguin, P. Tryjanowski, C. Z. Liu, S. Rawlins, and A. Imeson. 2008. Attributing physical and biological impacts to anthropogenic climate change. Nature 453:353-U320. Rosenzweig, M. L. 1995. Species diversity in space and time. Cambridge University Press., Cambridge, 436p. Rousseeuw, P., C. Croux, V. Todorov, A. Ruckstuhl, M. Salibian-Barrera, T. Verbeke, M. Koller, and M. Maechler. 2011. robustbase: Basic Robust Statistics. R package version 0.7-3. Ruhland, K., A. M. Paterson, and J. P. Smol. 2008. Hemispheric-scale patterns of climate-related shifts in planktonic diatoms from North American and European lakes. Global Change Biology 14:2740-2754. Rustad, L. E., J. L. Campbell, G. M. Marion, R. J. Norby, M. J. Mitchell, A. E. Hartley, J. H. C. Cornelissen, and J. Gurevitch. 2001. A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia 126:543-562. Sager, L. and J. B. Lachavanne. 2009. The M-NIP: a macrophyte-based Nutrient Index for Ponds. Hydrobiologia 634:43-63. Sajaloli, B., O. Limoges, V. Girod, F. Pirot, and C. Dutilleul. 2002. La mare, oeil et miroir du paysage. Contribution à l’étude des perceptions et représentations des petites zones humides et à la définition d’une stratégie d’adhésion sociale à leur préservation. Page 15 in Actes du Colloque "Hydrosystèmes, Paysages, Territoires", Université des Sciences et Technologies de Lille. Sala, O. E., F. S. Chapin, J. J. Armesto, E. Berlow, J. Bloomfield, R. Dirzo, E. Huber-Sanwald, L. F. Huenneke, R. B. Jackson, A. Kinzig, R. Leemans, D. M. Lodge, H. A. Mooney, M. Oesterheld, N. L. Poff, M. T. Sykes, B. H. Walker, M. Walker, and D. H. Wall. 2000. Biodiversity - Global biodiversity scenarios for the year 2100. Science 287:1770-1774. Samways, M. J. 2006. National Red List of South African dragonflies (Odonata). Odonatologica 35:341-368. Samways, M. J. 2008. Dragonflies and Damselflies of South Africa. Pensoft Publishers, Sofia, Bulgaria, 297p. Scheffer, M. 2004. Ecology of shallow lakes. Kluwer Academic Publishers, Dordrecht, Boston, London, 357p. Scheffer, M., S. H. Hosper, M. L. Meijer, B. Moss, and E. Jeppesen. 1993. Alternative Equilibria in Shallow Lakes. Trends in Ecology & Evolution 8:275-279. Scheffer, M., G. J. van Geest, K. Zimmer, E. Jeppesen, M. Sondergaard, M. G. Butler, M. A. Hanson, S. Declerck, and L. De Meester. 2006. Small habitat size and isolation can promote species richness: second-order effects on biodiversity in shallow lakes and ponds. Oikos 112:227-231. Scheffer, M. and E. H. van Nes. 2007. Shallow lakes theory revisited: various alternative regimes driven by climate, nutrients, depth and lake size. Hydrobiologia 584:455-466. Schindler, D. W. 2006. Recent advances in the understanding and management of eutrophication. Limnology and Oceanography 51:356-363. 217 Schlaepfer, M. A., M. C. Runge, and P. W. Sherman. 2002. Ecological and evolutionary traps. Trends in Ecology & Evolution 17:474-480. Schlieper, C. 1952. Uber die Temperatur-Stoffwechsel-Relation einiger eurythermer Wassertiere. Biol. Zbl. 71:449-461. Schmedtje, U. and M. Colling. 1996. Ökologische Typisierung der aquatischen Makrofauna. Bayersiches Landesamt für Wasserwirtschaft, München, 543p. Schmidt, B. 2004. Actualisation de la Liste Rouge des amphibiens 2002-2005, Directives pour le travail de terrain, Directives révisées pour 2004., Bern, 8p. Schmidt, B. and S. Zumbach. 2005. Liste Rouge des espèces menacées en Suisse. Amphibiens., Berne, 46p. Service conseil Zones alluviales Berne et Yverdon-les-Bains. 2001-2008. Dossier Zones alluviales: fiches. Office fédéral de l'environnement OFEV Berne, 174p. Settele, J., O. Kudrna, A. Harpke, I. Kuhn, C. van Swaay, R. Verovnik, M. Warren, M. Wiemers, J. Hanspach, T. Hickler, E. Kuhn, I. van Halder, K. Veling, A. Vliegenthart, I. Wynhoff, and O. Schweiger. 2008. Climatic Risk Atlas of European Butterflies. Pensoft, Sofia-Moscow, 710p. Simaika, J. P. and M. J. Samways. 2009a. An easy-to-use index of ecological integrity for prioritizing freshwater sites and for assessing habitat quality. Biodiversity and Conservation 18:11711185. Simaika, J. P. and M. J. Samways. 2009b. Reserve selection using Red Listed taxa in three global biodiversity hotspots: Dragonflies in South Africa. Biological Conservation 142:638-651. Simaika, J. P. and M. J. Samways. 2011. Comparative assessment of indices of freshwater habitat conditions using different invertebrate taxon sets. Ecological Indicators 11:370-378. Simpson, K. W., J. P. Fagnani, R. W. Bode, D. M. DeNicola, and L. E. Abele. 1986. Organism-substrate relationships in the main channel of the Lower Hudson River. Journal of the North American Benthological Society 5:41-57. Smith, A. J., R. W. Bode, and G. S. Kleppel. 2007. A nutrient biotic index (NBI) for use with benthic macroinvertebrate communities. Ecological Indicators 7:371-386. Smith, M. A. and D. M. Green. 2005. Dispersal and the metapopulation paradigm in amphibian ecology and conservation: are all amphibian populations metapopulations? Ecography 28:110-128. Smith, R. I. L. 1996. Introduced plants in Antarctica: Potential impacts and conservation issues. Biological Conservation 76:135-146. Smith, V. H. 2003. Eutrophication of freshwater and coastal marine ecosystems - A global problem. Environmental Science and Pollution Research 10:126-139. Smol, J. P., A. P. Wolfe, H. J. B. Birks, M. S. V. Douglas, V. J. Jones, A. Korhola, R. Pienitz, K. Ruhland, S. Sorvari, D. Antoniades, S. J. Brooks, M. A. Fallu, M. Hughes, B. E. Keatley, T. E. Laing, N. Michelutti, L. Nazarova, M. Nyman, A. M. Paterson, B. Perren, R. Quinlan, M. Rautio, E. Saulnier-Talbot, S. Siitoneni, N. Solovieva, and J. Weckstrom. 2005. Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America 102:4397-4402. Solimini, A., M. Bazzanti, A. Ruggiero, and G. Carchini. 2008. Developing a multimetric index of ecological integrity based on macroinvertebrates of mountain ponds in central Italy. Hydrobiologia 597:109-123. Sondergaard, M., L. S. Johansson, T. L. Lauridsen, T. B. Jorgensen, L. Liboriussen, and E. Jeppesen. 2009. Submerged macrophytes as indicators of the ecological quality of lakes. Freshwater Biology 55:893–908. Southwood, T. R. E. 1977. Habitat, the templet for ecological strategies ? Journal of Animal Ecology 46:337-365. Tachet, H., P. Richoux, M. Bournaud, and P. Usseglio-Poletra. 2000. Invertébrés d'eau douce: systématique, biologie, écologie. CNRS édition edition, 588p. Theurillat, J. P. and A. Guisan. 2001. Potential impact of climate change on vegetation in the European Alps: A review. Climatic Change 50:77-109. 218 Thomas, C. D., A. Cameron, R. E. Green, M. Bakkenes, L. J. Beaumont, Y. C. Collingham, B. F. N. Erasmus, M. F. de Siqueira, A. Grainger, L. Hannah, L. Hughes, B. Huntley, A. S. van Jaarsveld, G. F. Midgley, L. Miles, M. A. Ortega-Huerta, A. T. Peterson, O. L. Phillips, and S. E. Williams. 2004. Extinction risk from climate change. Nature 427:145-148. Thuiller, W., S. Lavorel, and M. B. Araujo. 2005a. Niche properties and geographical extent as predictors of species sensitivity to climate change. Global Ecology and Biogeography 14:347357. Thuiller, W., S. Lavorel, M. B. Araujo, M. T. Sykes, and I. C. Prentice. 2005b. Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America 102:8245-8250. Tockner, K. and J. A. Stanford. 2002. Riverine flood plains: present state and future trends. Environmental Conservation 29:308-330. Travis, J. M. J. 2003. Climate change and habitat destruction: a deadly anthropogenic cocktail. Proceedings of the Royal Society of London Series B-Biological Sciences 270:467-473. U.S.EPA. 1998. Lake and Reservoir Bioassessment and Biocriteria: Technical Guidance Document, EPA841-B-98-007., United States, 204p. Ulen, B. M. and G. A. Weyhenmeyer. 2007. Adapting regional eutrophication targets for surface waters - influence of the EU water framework directive national policy and climate change. Environmental Science & Policy 10:734-742. Usseglio-Polatera, P., M. Bournaud, P. Richoux, and H. Tachet. 2000. Biological and ecological traits of benthic freshwater macroinvertebrates: relationships and definition of groups with similar traits. Freshwater Biology 43:175-205. Van der Putten, W. H., M. Macel, and M. E. Visser. 2010. Predicting species distribution and abundance responses to climate change: why it is essential to include biotic interactions across trophic levels. Philosophical Transactions of the Royal Society B-Biological Sciences 365:2025-2034. Verberk, W. C. E. P., H. Siepel, and H. Esselink. 2008. Applying life-history strategies for freshwater macroinvertebrates to lentic waters. Freshwater Biology 53:1739–1753. Vittoz, P., N. Dussex, J. Wassef, and G. A. 2009. Diaspore traits discriminate good from weak colonisers on high-elevation summits. Basic and Applied Ecology 10:508-515. Waide, R. B., M. R. Willig, C. F. Steiner, G. Mittelbach, L. Gough, S. I. Dodson, G. P. Juday, and R. Parmenter. 1999. The relationship between productivity and species richness. Annual Review of Ecology and Systematics 30:257-300. Walker, K. J. and C. D. Preston. 2006. Ecological predictors of extinction risk in the Flora of Lowland England, UK. Biodiversity and Conservation 15:1913-1942. Walker, M. D., C. H. Wahren, R. D. Hollister, G. H. R. Henry, L. E. Ahlquist, J. M. Alatalo, M. S. BretHarte, M. P. Calef, T. V. Callaghan, A. B. Carroll, H. E. Epstein, I. S. Jonsdottir, J. A. Klein, B. Magnusson, U. Molau, S. F. Oberbauer, S. P. Rewa, C. H. Robinson, G. R. Shaver, K. N. Suding, C. C. Thompson, A. Tolvanen, O. Totland, P. L. Turner, C. E. Tweedie, P. J. Webber, and P. A. Wookey. 2006. Plant community responses to experimental warming across the tundra biome. Proceedings of the National Academy of Sciences of the United States of America 103:1342-1346. Walther, G. R. 2010. Community and ecosystem responses to recent climate change. Philosophical Transactions of the Royal Society B-Biological Sciences 365:2019-2024. Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M. Fromentin, O. HoeghGuldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389-395. Wetzel, R. G. 1983. Limnology. 2nd edition. Saunders College Publishing, Philadelphia, 760p. Willen, E. 2009. System Aqua - a Swedish system for assessing nature conservation values of freshwaters. Page 293 in P. J. B. C. M. Pringle, editor. Assessing the conservation value of fresh waters. An international perspective. Cambridge University Press, Cambridge. 219 Williams, N. M., E. E. Crone, T. H. Roulston, R. L. Minckley, L. Packer, and S. G. Potts. 2010. Ecological and life-history traits predict bee species responses to environmental disturbances. Biological Conservation 143:2280-2291. Williams, P., M. Whitfield, J. Biggs, S. Bray, G. Fox, P. Nicolet, and D. Sear. 2004. Comparative biodiversity of rivers, streams, ditches and ponds in an agricultural landscape in Southern England. Biological Conservation 115:329-341. Wilson, R. J., D. Gutierrez, J. Gutierrez, and V. J. Monserrat. 2007. An elevational shift in butterfly species richness and composition accompanying recent climate change. Global Change Biology 13:1873-1887. Xenopoulos, M. A., D. M. Lodge, J. Alcamo, M. Marker, K. Schulze, and D. P. Van Vuuren. 2005. Scenarios of freshwater fish extinctions from climate change and water withdrawal. Global Change Biology 11:1557-1564. 220 Chapitre 8 Annexes Liste des Annexes 1. Angelibert, S., V. Rosset, N. Indermuehle, and B. Oertli. 2010. The pond biodiversity index "IBEM": a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 1. Index development. Limnetica 29:93-104. 2. Leclerc, D., S. Angélibert, V. Rosset, and B. Oertli. 2010. Les libellules (Odonates) des étangs piscicoles de la Dombes. Martinia 26:98-108. 221 222 Limnetica, 29 (1): x-xx (2008) Limnetica, 29 (1): 93-104 (2010) c Asociación Ibérica de Limnolog´a, Madrid. Spain. ISSN: 0213-8409 The pond biodiversity index “IBEM”: a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 1. Index development Sandrine Angélibert, Véronique Rosset, Nicola Indermuehle & Beat Oertli∗ hepia Geneva, University of Applied Sciences Western Switzerland, technology, architecture, landscape. CH-1254 Jussy-Geneva, Switzerland. 2 ∗ Corresponding author: [email protected] 2 Received: 12/12/08 Accepted: 30/6/09 ABSTRACT The pond biodiversity index “IBEM”: a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 1. Index development Due to legal requirements, nature managers increasingly have to carry out assessments of biodiversity for conservation purposes. For ponds, a type of waterbody now widely recognized as an important reservoir for freshwater biodiversity, standardized bioassessment methods are needed, but still rare. We produced such a tool for small lowland waterbodies in Switzerland: the Pond Biodiversity Index (“IBEM”). This Index is the adaptation of a method used by researchers for assessing the biodiversity in ponds, PLOCH, which does not currently meet the requirements for routine use by nature managers because it is too expensive and requires a high skill level in taxonomic identication. A method intended for practitioners has to be simple, standardized, cheap, adjustable, and consistent with the legislative framework. In order to fulll these requirements, the theoretical and practical aspects of IBEM were developed with a group of representative end users including nature conservation managers, consultants, governmental organizations and taxonomic experts. To develop the method, we used a species dataset from 63 Swiss lowland ponds which included ve taxonomic groups: aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata and Amphibia. The following topics were addressed: (i) the number and type of taxonomic groups which should be used for producing the index (is it possible to use surrogates?) (ii) the level of identication for each taxonomic group (species? genus? family?) (iii) the sampling strategy (sampling technique, number of replicates), (iv) the calculation of a unique index and the strategy for assessing its score, and (v) the transfer of this new method to end users. The new method IBEM uses all ve taxonomic groups, because a subset of groups did not produce reliable assessments of pond biodiversity. Identication to genus level is required for four groups (aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata) and species level for Amphibia. The sampling methodology is based on the stratied random strategy used in the PLOCH method, but with a slight modication in the number of samples per pond. The assessment follows the methodology adopted by the European Water Framework Directive, and the ratio of the observed richness to a reference-based predicted richness is translated into one of ve quality categories for each pond. The nal index is the mean of the ve assessment scores. To facilitate the implementation of the IBEM method, a website (http://campus.hesge.ch/ibem) enables online calculation of the index, and provides instructions on both sampling and assessment methodologies. Furthermore, training courses are organized by the authors of the method for end users. Key words: Bioassessment, monitoring, small waterbodies, nature conservation, practitioners, macroinvertebrates, aquatic plants, amphibians. RESUMEN El ´ndice de biodiversidad “IBEM”: una nueva herramienta para evaluar la biodiversidad de charcas en Suiza. Parte I. Desarrollo del ´ndice Debido a requerimientos legales, es cada vez más necesario que los gestores del medio ambiente lleven a cabo evaluaciones de la biodiversidad dirigidas a la conservación de la naturaleza. Para las charcas, pequeñas masas de agua ampliamente reconocidas como importantes reservorios de diversidad biológica acuática, los métodos normalizados de bio-evaluación son 223 94 Angélibert et al. necesarios, pero aún escasos. Para esta tipolog´a de pequeñas masas de agua situadas a baja altitud en Suiza se ha elaborado el ´ndice de Biodiversidad de charcas (“IBEM”). Este ´ndice es la adaptación de un método utilizado por los investigadores para evaluar la diversidad biológica en charcas, PLOCH, que no cumpl´a los requisitos para un uso rutinario por parte de los gestores del medio natural por ser demasiado caro y requerir un alto nivel de experiencia en la identicación taxonómica. Un método destinado a estos profesionales tiene que ser sencillo, estandarizado, económico, ajustable y en consonancia con el marco legislativo. Con el n de cumplir estos requisitos, los aspectos teóricos y prácticos de IBEM se han desarrollado con un grupo representativo de posibles usuarios, incluyendo gestores conservadores, consultores, organizaciones gubernamentales y expertos en taxonom´a. Para desarrollar el método, se ha utilizado una base de datos de 63 charcas Suizas, situadas en altitudes bajas, que incluye cinco grupos taxonómicos: plantas acuáticas, gasterópodos acuáticos, coleópteros acuáticos, odonatos adultos y anbios. Se han estudiado los siguientes aspectos: (i) el número y tipo de grupos taxonómicos que se deben utilizar (es posible el uso de sustitutos?) (ii) nivel de identicación para cada grupo taxonómico (¿especie, género, familia?) (iii) estrategia de muestreo (técnica, número de réplicas), (iv) cálculo de un ´ndice único y procedimiento para la asignación de valores y (v) la transferencia de este método a los posibles usuarios. El nuevo método IBEM utiliza los cinco grupos taxonómicos, ya que un subconjunto de ellos no producir´a evaluaciones ables de la diversidad biológica de la charca. La identicación a nivel de género es necesaria para cuatro de estos grupos (plantas acuáticas, gasterópodos acuáticos, coleópteros acuáticos, y odonatos adultos) y para los anbios es necesario el nivel de especie. El muestreo sigue un diseño aleatorio estraticado, utilizado en el método PLOCH, pero con una ligera modicación en el número de muestras por charca. La evaluación sigue la metodolog´a adoptada por la Directiva Marco de Aguas, y la relación entre la riqueza observada y la del estado de referencia se traduce en una de las cinco categor´as de calidad para cada charca. El ´ndice nal es la media de las cinco puntuaciones de la evaluación. Para facilitar la aplicación del método IBEM, un sitio web (http://campus.hesge.ch/ibem) permite cálculo del ´ndice a través de la red y proporciona instrucciones tanto de las metodolog´as de muestreo como de la valoración. Además, los autores han organizado cursos de formación sobre el método para los usuarios. Palabras clave: Índices bióticos, indicadores biológicos, pequeñas masas de agua, conservación de la naturaleza, medioambientalistas, macroinvertebrados, plantas acuáticas, anbios. INTRODUCTION Ponds contribute in a unique way to aquatic biodiversity, supporting as many species as rivers or lakes, including many that are rare or threatened (Williams et al. 2004, Grillas et al. 2004, Nicolet et al. 2004, Oertli et al. 2004, Angélibert et al. 2006). In order to assess and monitor these freshwater ecosystems, conservation planners and nature managers need to have tools to easily and rapidly evaluate the biological quality of these aquatic habitats. These rapid biodiversity assessment tools should be standardized, cheap and consistent with the legislative framework. However, such tools are still rare for ponds. Existing methods (e.g. Biggs et al., 2000; Gernes & Helgen, 2002; Boix et al., 2005; Chovanec et al., 2005; Oertli et al., 2005; Solimini et al., 2008; Trigal et al., 2009; Menetrey Perrotet, 2009) all have features hindering their use by practitioners from Switzerland. For example, some methods apply only to a restricted geographical region, others are too expensive, and many re- quire a high level of skills in taxonomic identication (Indermuehle et al., 2004; Sandoz, 2006). In addition, in the absence of simple methods to assess still waters, managers tend to misuse methods designed for running waters. The Swiss-based pond biodiversity index IBEM (from the French Indice de Biodiversité des Etangs et Mares) was developed to ll this gap. Following suggestions made by Green et al. (2005) to improve biodiversity monitoring, the development process relied strongly on consultations with stakeholders and took into account the needs of end users. According to these requirements, the new tool had to be: (i) simple in terms of sampling and data processing, (ii) standardized, (iii) adjustable, (iv) cheap and (v) eurocompatible. IBEM is based on a method for assessing the biodiversity in ponds originally used by researchers: the PLOCH method (Oertli et al., 2005). PLOCH relies on the species richness of ve taxonomic groups: aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata and Amphibia. The choice of these indicator 224 The IBEM-Index: index development groups has been discussed by Oertli et al. (2005) and supported by further studies (Auderset Joye et al., 2004; Menetrey et al., 2005, 2008). To summarize, the ve taxonomic groups (aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata and Amphibia) fulll most indicator-criteria stated by New (1995) and are ecologically complementary with respect to their life cycle, their position in the food web, their habitat preferences and their ways of dispersal (for further reading on the use of Odonata in biodiversity assessments, see also Cordoba-Aguilar (2008)). The PLOCH method is relatively expensive to use (sampling, sorting and identication time) and requires species level taxonomic identication skills, and is therefore not suitable for use by pond conservation practitioners. A group of experts were consulted throughout the development of the IBEM-Index. This group was composed of ve future end users and seven taxonomic specialists who were involved in all major decision making. In parallel, ve teams of nature conservation managers (three environmental consultant teams and two nature reserve management groups) tested both the practical and theoretical aspects of the IBEM method. They assessed the method’s strengths and weaknesses, and identied the key issues to be resolved before successful implementation. Three academic theses (Lezat 2006; Sandoz 2006; Frey 2007) were Figure 1. Distribution of the 63 sampled lowland ponds (circles) in Switzerland with location of the four 4 test ponds (black circles). Distribución de las 63 charcas muestreadas en Suiza (c´rculos) con la localización de las 4 charcas de prueba (c´rculos negros). 225 95 furthermore carried out within the framework of the IBEM development. Cross-taxon and withintaxon surrogacies for the ve taxonomic groups were also explored using an existing, compatible dataset of 63 ponds. The aim was to determine (i) whether all or a subset of those groups were mandatory for a reliable biodiversity assessment, and (ii) whether a higher taxa approach could be implemented, i.e. if species level identication could be replaced by genus or even family level identication. The sampling and assessment methodologies were then adapted with respect to the chosen taxonomic level. Finally, strategies were drawn up to implement this new method and make it easily available to end users. METHODS Study sites and practitioner teams Testing of the method by practitioners was carried out by ve teams of nature managers: the environmental consultants GREN (Geneva, GE), AMaibach Sàrl (Oron-la-Ville, VD), NATURA (Les Reussilles, JU) and two nature reserve management groups (“Groupe d’Etude et de Gestion de la Grande-Cariçaie” GEG (Yverdonles-Bains, VD), and “Fondation des Grangettes /Musée Cantonal de Zoologie de Lausanne” (Lausanne, VD)). They applied the PLOCH method (detailed methodology described in Oertli et al., 2005) to assess the biological quality of four ponds located in different regions of Western Switzerland (La Grande Cariçaie FR, Les Grangettes VD, Rouelbeau GE, La Combe Tabeillon JU, Fig. 1). These ponds were sampled during 2005 or 2006. Experts in the taxonomy and ecology of the selected taxonomic groups took part in workshops to provide additional support for the development of the method: P. Prunier, R. Juge and J.-B. Lachavanne (aquatic plants), P. Stucki (Gastropoda), G. Carron (Coleoptera), A. Maibach (Odonata) and S. Zumbach/KARCH (Amphibia). For the development of the IBEM-Index, a dataset of 63 Swiss lowland ponds (Fig. 1) with an altitudinal range of 305 to 967 m.a.s.l. was used, constituting a subset of the data collected during 96 Angélibert et al. the PLOCH project (Oertli et al., 2000; 2002) by the Laboratory of Aquatic Ecology and Biology (LEBA) of the University of Geneva. The main pond characteristics are given in Appendix 1. Sampling of biodiversity (aquatic plants, aquatic Gastropoda, aquatic Coleoptera, adult Odonata and Amphibia) and measurements of around 100 environmental variables were carried out following standardized procedures (detailed information in Oertli et al., 2005). Developing the IBEM-Index How many taxonomic groups are required for an accurate assessment? In order to investigate if one or more taxonomic groups can be discarded from the ve sampled groups without losing accuracy in the global assessment (cross-taxon surrogacy), we measured the PLOCH quality class (bad, poor, moderate, good and high) for 63 lowland ponds, based on species level data (i) for all ve taxonomic groups, and (ii) for all the possible combinations using less than ve groups (n = 30 combinations). The performance of these 30 combinations was then assessed by the percentage of ponds remaining in the same quality class as that produced by considering all ve taxonomic groups (e.g. % of correctly classied ponds). leoptera. True Odonata richness was estimated by abundance-based Chao-I (Chao, 1984), as the minimal number of replicates (samples) requested by Jackknife-1 was not available for this group. Jackknife-1 and Chao-1 are both non-parametric estimators, which assess true species richness relying on the observed richness measured in the eld; the use of such true richness estimators reduces bias linked to heterogeneous sampling effort due to non-exhaustive sampling. The true richness was estimated at all the taxonomic levels (i.e. species, genus and family). A good surrogate (genus or family richness) should have a good correlation with species richness. The identication levels presenting a low correlation (r2 values below 0.75) were therefore discarded from further analysis. Secondly, the accuracy of the remaining potential surrogates was evaluated by their ability to correctly assess pond biodiversity. This was done by comparing the real quality class of 63 lowland ponds (PLOCH species level assessment) with the quality classes obtained with combinations of the different identication level (species, genus, family). The performance of the combinations was evaluated by the percentage of the 63 ponds remaining in the same quality class as that produced by considering identication at the species level for the 5 taxonomic groups (e.g. % of correctly classied ponds). Number of samples Choice of taxonomic resolution: species, genus or family? Species level identication is a time consuming and hence expensive task that requires high taxonomic skills often lacking in end users (environmental consultants and other nature managers). For this reason we investigated if species richness could be replaced by genus or even family richness without losing the relevance of the index for 63 lowland ponds. This within-taxon investigation on surrogacy was carried out in two steps. Firstly, we tested within-taxon correlations, between species, genus or family richness. True richness was calculated by samplebased Jackknife-1 (Burnham & Overton, 1979) estimation for vegetation, Gastropoda and Co- The aim of sampling is to gather the observed taxonomic richness (Sobs ) reaching at least 70 % of true pond richness (Strue ). This level is sufcient for subsequently estimating the true richness with richness estimators. The number of samples (vegetation plots or macroinvertebrate sweep net samples) to be collected was assessed with data from 63 Swiss lowland ponds. By means of EstimateS software (Colwell, 2005), 63 accumulation curves of Sobs were drawn and Strue was computed by thenon-parametric Jackknife-1 estimator (Burnham & Overton, 1979) to compensate for the bias of a non-exhaustive sampling. This data was then used to estimate the mean number of samples necessary to gather at least 70 % of Strue (i.e. PLOCH method, Oertli et al., 2005). 226 The IBEM-Index: index development 97 Biodiversity was assessed by calculating the ratio between the observed condition and an unimpaired reference condition. This ratio allowed the classication of the pond into one of ve quality classes: bad, poor, moderate, good and high (e.g. the methodology presented in the WFD (EC, 2000)). As the -Index is based on taxonomic richness, reference conditions stand for conditions enabling high potential richness. We predicted these reference conditions with Generalized Additive Models (GAMs; Hastie & Tibshirani, 1990; Lehmann et al., 2002) built on the relationship between environmental variables and taxonomic richness of the ve indicator groups. Statistical details on the GAM-procedure used are described by Oertli et al. (2005). thod’s cost-effectiveness without affecting the quality of the results? An effort was therefore made to reduce the time necessary for a complete pond biological assessment. Training opportunities were another concern of the practitioners. Above all, they wanted to improve eldwork standardisation (sampling technique and methodology), but also develop their taxonomic identication skills. It was therefore decided to implement an online support system, with the objective of improving the use of the index. This interactive website (http://campus.hesge.ch/ibem) contains documents, illustrations and video tutorials, as well as an online index calculator. Training courses, targeted at nature reserve managers and consultants are also part of the strategy to facilitate implementation of the method in Switzerland. RESULTS Developing the index Test of the method by practitioners Cross-taxon investigation: how many taxonomic groups? Prediction of reference conditions The ve teams of practitioners (environmental consultants and nature reserve management groups) all endorsed the concept of a standardized sampling approach. They highlighted the usefulness of the rapid assessment index and its euro-compatibility (according to the WFD methodology). However, two specic questions were raised concerning the proposed taxonomic identication level (species level) and the number of taxonomic groups to be sampled (ve). Is species identication compulsory for all the biological groups or could genus or even family level data do? Could one group (or several groups) be left aside, depending on the skills of the staff involved in the assessment of a given pond? These issues were taken into account and tested during the further development of the index (see below). Additional questions concerned the fieldwork methodology, for example the sampling periods to be chosen or the strategy for sample distribution. These remarks led to methodological changes in the new method (see Indermuehle et al., 2009). Furthermore, cost reduction was a central issue raised by practitioners during this preliminary test stage. Is it possible to enhance the me227 The cross-taxon surrogacy test (with species level data) (Fig. 2) showed that the four taxa combination VGCA performed best when compared to the reference combination (VGCOA, for: Vegetation, Gastropoda, Coleoptera, Odonata, Amphibia) with 83 % of the ponds correctly classied and 17 % with only a one-class shift. GCOA performed second best (80 % of the ponds correctly classied), followed by VGCO (73 %) and VCA (70 %). All single taxa performed badly, with less than 45 % of the ponds correctly classied. In conclusion, at least four taxonomic groups would have to be retained for a reliable assessment, either with the combination VGCA (i.e. without Odonata) or GCOA (i.e. without aquatic Vegetation). Within-taxon investigation: species, genus or family level? A total of 243 (= 35 ) potential combinations were available for this within-taxon investigation, depending on the identication level (species, genus or family) of the ve taxonomic groups. The rst step was to test correlations between species, genus or family richness (Table 1). Spe- 98 Angélibert et al. Figure 2. Percentage of correctly classied (“no change”) and misclassied ponds (shifts from one to four classes) obtained by the assessment with different taxa combinations. V: Vegetation, G: Gastropoda, C: Coleoptera, O: Odonata, A: Amphibia. (n = 63 ponds). Porcentaje charcas clasicadas correctamente (“sin cambio”) y mal clasicadas (cambios de entre una a cuatro categor´as) obtenido a partir de la evaluación con diferentes combinaciones de los taxones. V: Vegetación, G: Gastropoda, C: Coleoptera, O: Odonata, A: Amphibia. (n = 63 charcas). cies richness and genus richness showed strong correlations for aquatic Vegetation (r2 = 0.80), Gastropoda (r2 = 0.87), Coleoptera (r2 = 0.90) and Odonata (r2 = 0.88). These results showed that for these four groups, genus richness could potentially be used as a surrogate for species richness. This was not the case for Amphibia (r2 = 0.72), which should therefore be identified to species level. After discarding 162 combinations involving genus and family richness of Amphibia, only 81 (= 34 ) remained from the initial 243 combinations. For correlations between family richness and species richness, only Vegetation presented a high value (r2 = 0.78); the values for the other taxonomic groups were low (r2 from 0.47 to 0.67). Thus, family level cannot be used as surrogate for species richness, except possibly for Vegetation. From the 81 original combinations, only 24 re- mained, involving species level for all 5 groups, genus level for Vegetation, Odonata, Gastropoda and Coleoptera, and family level for Vegetation. Finally, as species level identication requires high taxonomic skills and is likely to hinder the implementation of a new rapid index, all 22 combinations involving species level data for Vegetation, Gastropoda, Coleoptera and Odonata were discarded. Consequently, two combinations remained: “(VGCO)genus-(A)species” and “(V) family-(GCO)genus-(A)species”. These two surrogate combinations differed only in terms of the identication level of aquatic Vegetation (V): either family or genus. The accuracy of these two combinations was evaluated for their ability to correctly assess the biodiversity of pond dataset. Compared to the reference combination (“VGCOA species”), both Table 1. Correlations between species richness (S) and genus and family richness of the ve indicator groups. Correlaciones entre la riqueza de especies (S) y la riqueza de géneros y la de familias de los cinco grupos indicadores. Genus richness Family richness n (ponds) Vegetation S Gastropoda S Coleoptera S Odonata S Amphibia S 0.80 0.78 57 0.87 0.47 42 0.90 0.55 62 0.88 0.52 58 0.72 0.67 102 228 The IBEM-Index: index development 99 Based on the two previous tests, which included discarding some groups and changing the taxonomic identification level, seven combinations were considered for the most relevant index (Table 2): “(VGCO)genus-(A)species” and “(V)family-(GCO)genus-(A)species” (i.e. the two best combinations based on all ve indicator groups), and 5 combinations involving only four indicator groups at different taxonomic levels (see previous sections). The combination “(VGCO)genus-(A)species” performed better than the other indices, with respect to the percentage of correctly classied ponds (88 %, Table 2). The second best option was “(VGCA) species”, but this combination was discarded because it was based on species level data and was therefore less suitable for a rapid index. The combination “(V)family-(GCO)genus-(A)species” was second equal in effectiveness, but was discounted as it relied on family level data for Vegetation. Family level identication for plants is likely to be less intuitive and therefore more time consuming for generalists used to genus level identication. It was deemed important for the development of the Index to nd a reasonable trade off between ease of use (e.g. genus level identication) and relevance for biological assessment; and consequently combinations which classied less than 80 % of sites correctly were considered inadequate as indices. For these reasons, the combination (a) (b) Table 2. Percentage of correctly classied ponds for seven different indices. V: Vegetation, G: Gastropoda, C: Coleoptera, O: Odonata, A: Amphibia. Porcentaje de charcas correctamente clasicadas para siete ´ndices diferentes. V: Vegetación, G: Gasterópodos, C: Coleópteros, O: Odonatos, A : Anbios Index (VGCO)genus-(A)species (VGCA)species (V)family-(GCO)genus-(A)species (GCOA) species (VGC)genus-(A)species (GCO)genus-(A)species (V)family-(GC)genus-(A)species % of correctly classied ponds 88 % 83 % 82 % 79 % 72 % 72 % 65 % combinations produced satisfying results. The “(VGCO)genus-(A)species” combination performed best, with 88 % of the ponds correctly classied compared to 82 % for the “(V)family(GCO)genus-(A)species” combination. In both cases, ponds misclassied only shifted one category. Taking into account both the cross-taxon and the within-taxon investigations Figure 3. Mean number of samples necessary to gather at least 70 % of Strue as a function of pond area. (a) Aquatic vegetation. Equation of the relationship: n = 30 – 29.1 ∗ log10 (area) + 8.6 * (log10 (area))2 . (b) Macroinvertebrates (Coleoptera and Gastropoda). Equation of the relationship: n = 15.5 – 10.5 ∗ log10 (area) + 2.7 * (log10 (area))2 . Número medio de muestras necesarias para obtener al menos el 70 % del Strue en función del área de la charca. (a) Vegetación acuática, ecuación de la función: n = 30 – 29.1 ∗ log10 (área) + 8.6 * (log10 (área))2 . (b) Macroinvertebrados (Coleoptera y Gastropoda), ecuación de la función: n = 15.5 – 10.5 ∗ log10 (área) + 2.7 * (log10 (área))2 . 229 100 Angélibert et al. set of 15 was selected as potential predictors for the stepwise selection within the GAM procedure. GAMs integrated 12 of these variables, with four to ve predictors for each model (Table 3). Area represented the most important contribution to all models, except for Coleoptera, with a contribution between 0.63 and 0.93. The other predictors were mean depth, shoreline development, percentage of pond surface shaded by trees, percentage of woodland in the pond’s surrounding (in a 50-m buffer zone), altitude, sh presence, proportion of pond area covered by oatingleaved or submerged vegetation, water conductivity, turbidity, and nutrient concentration (trophic state). Three variables were not integrated in the ve GAMs: pond connectivity (a measure of isolation from other waterbodies), percentage of agriculture in the catchment area, and pond age. These ve models were used to predict reference conditions, i.e. highest possible richness for each type of pond. For predicting these ve richness values for a given pond, 6 of the 12 variables, describing the pond typology, have to be measured in the eld: pond area, mean depth, “(VGCO)genus-(A)species” was ultimately chosen for the IBEM-Index. Number of samples The genus accumulation curves of vegetation and macroinvertebrates (Gastropoda and Coleoptera) (Sobs ) and the associated curves of Strue were computed for 63 ponds. This was then used to estimate the mean number of samples required to reach 70 % of Strue , in relation to the surface area of each of the 63 ponds. These results were used to produce the relationship between pond area and the number of samples to be collected (Fig. 3). Prediction of reference conditions In order to dene the reference conditions and assess the taxonomic richness of the ve indicator groups, ve predictive models were produced. The relation between environmental variables and the richness of the ve taxonomic groups was modelled with GAMs. Out of more than 100 local and regional environmental variables, a sub- 0.32 0.33 0.68 r2 0.36 0.73 r1 0.40 0.67 0.39 0.30 0.22 %D 0.59 PNC 0.82 0.66 Subm. veget. Floating veget. 0.35 0.65 0.77 0.78 Fish Altitude # Woodland # Shade # SI # 0.57 Transparency 0.63 0.93 Conductivity Vegetation Gastropoda Coleoptera Odonata Amphibia Mean depth # Area # Table 3. Selected predictors and validation diagnostic of the ve GAM models for aquatic Vegetation, Gastropoda, Coleoptera, Odonata and Amphibia. The range of measured values is presented in Appendix 1. The models were evaluated using percentage of explained deviance ( %D), simple variation coefcient (r1 ), and cross-validation coefcient (r2 ). All models were selected with threshold p < 0.05. Predictores seleccionados y diagnóstico de validación de los cinco modelos GAM para vegetación acuática, gasterópodos, coleópteros, odonatos y anbios. El rango de valores medidos se presenta en el apéndice 1. Los modelos fueron evaluados utilizando el porcentaje de desviación explicada ( % D), el coeciente de variación (r1 ), y el coeciente de validación cruzada (r2 ). Todos los modelos fueron seleccionados con p < 0.05. 0.42 0.29 0.37 0.32 0.62 0.20 0.53 0.61 0.61 0.80 0.46 0.37 0.43 0.51 0.73 0.29 0.34 Area: log10 (area); SI: shoreline index (dened in Appendix 1); shade: percentage of pond surface area shaded; woodland: percentage of woodland in a 50m radius from the pond edge; sh: sh presence; oating veget: proportion of pond surface area covered by oating-leaved vegetation; subm. veget.: proportion of pond surface area covered by submerged vegetation; PNC: trophic state (dened in Appendix 1). # Variables to be measured for the IBEM pond assessment. Área: log10 (área); SI: desarrollo del per´metro (denido en el Apéndice 1); shade: porcentaje de área de la charca sombreada; woodland: porcentaje de terreno forestal en un radio de 50 m desde el borde de la charca; sh: presencia de peces; oating veget.: proporción de supercie de la charca cubierta por vegetación de hojas otantes; subm. veget.: proporción de supercie de la charca cubierta por vegetación sumergida; PNC: estado tróco (denido en el Apéndice 1). # Variables que deben ser medidas para la evaluación de las charcas con el IBEM. 230 The IBEM-Index: index development shoreline index, percentage of pond surface shaded, percentage of woodland in a 50 m radius from the pond edge, and altitude. The other 6 variables are potential indicators of pond degradation and are consequently not to be measured on the field: they are set to their “optimal” value, i.e. allowing the highest possible taxonomic richness for each taxonomic group (see Indermuehle et al., 2009). Cost of the implementation of the IBEM method The investigations and tests carried out always kept in mind that one of the major requests of practitioner was low cost. Every effort was therefore made to reduce the time necessary for a complete pond biological assessment. Time reduction was achieved mainly by allowing a higher taxonomic identication level for four taxonomic groups (i.e. genus instead of species). Another noticeable gain was obtained by replacing macroinvertebrate sorting in the laboratory (Gastropoda and Coleoptera) with eld sorting. For one sample, the reduction in time is about 60 % (from 120 minutes to 45 minutes). Overall, the time needed to calculate the IBEM-Index was reduced by 50 % compared to the PLOCH method (50 hours for a 5000 m2 waterbody, instead of 100 hours). DISCUSSION The IBEM-Index was developed in close collaboration with future end users in order to meet their needs. The overall aim was to create a simple, standardized, rapid index to routinely assess pond biodiversity. By pursuing this aim, an important issue arose in dening reasonable trade off between ease of use (e.g. avoiding species level identication), low cost, and relevance to biological assessment. During the development of the index, each trade off was weighted-up in order to optimize the nal assessment tool. For example, the combination “(VGCO)genus(A)species” was chosen over “(VGCOA)species” even though its performance was slightly worse. This was because it required lower taxonomic skills (often lacking in end users) and was less time consuming. As time is money, and funding for bio231 101 diversity assessments is generally lacking, addressing the cost issue was essential for a new index. Cost reduction was one of the most important concerns raised by practitioners during the preliminary test stage. Therefore, this was the focus of effort to reduce the time necessary for a complete pond biological assessment. Approximately 50 hours are necessary to calculate the Index for a 5000 m2 waterbody, including sampling and data processing. Routine monitoring of biological quality for running water is in the same range of costs. For example, a half-yearly assessment of a stream section with the IBGN Index (AFNOR, 1992) is estimated to require the same amount of time (i.e. 50 hrs) for one year. Another important new feature of the IBEMIndex is its interactive online tutorial website (http://campus.hesge.ch/ibem) with online index calculation, developed to enhance the use of the index. Training courses, targeted at nature reserve managers and consultants, are also part of the strategy to facilitate implementation of the method in Switzerland. To summarize, the IBEM method is a tool for the rapid assessment of the biological quality of Swiss lowland ponds developed for practitioners (see Indermuehle et al., 2009). It produces an index by assessing the taxonomic richness of a given pond as an indicator of its overall biodiversity, and is therefore particularly useful for comparing ponds in local or regional scale assessments. The index may also, in time, be used for monitoring conservation actions and policy issues. The IBEM-Index has been designed to meet the specic needs of practitioners, and, as an index, constitutes a new tool for nature conservation. ACKNOWLEDGEMENTS The IBEM-Index was developed with support from: Groupe d’Etude et de Gestion de la Grande-Cariçaie (GEG), Fondation des Grangettes, Musée Cantonal de Zoologie de Lausanne, Swiss Amphibian and Reptile Conservation Programme (KARCH), University of GenevaLaboratoire d’Ecologie et Biologie Aquatique (LEBA), Laboratoire des technologies de l’In- 102 Angélibert et al. formation (Haute Ecole de Gestion de Genève), Consulting ofces AMaibach Sarl, Aquabug, Aquarius, GREN, and Natura. The study of the Swiss ponds, which made the development of the IBEM-Index possible, was supported by many partners: The Swiss Federal Ofce for the Environment (FOEN), Cantons of Geneva, Jura, Vaud and Lucerne, Research commission of the Swiss National Park and HES-SO // University of Applied Sciences Western Switzerland (RCSO RealTech). Moreover we are grateful for the data provided by the Swiss Biological Records Center (CSCF) and the Swiss Floristic Database (CRSF). Many thanks to the following persons for their various contributions: Céline Antoine, Dominique Auderset Joye, Diana Cambin, Gilles Carron, Emmanuel Castella, Jessica Castella, Michaël de la Harpe, Raphaelle Juge, Jean-Bernard Lachavanne, Anthony Lehmann, Simon Lézat, Nathalie Menetrey, Jane O’Rourke, Patrice Prunier, Corinne Pulfer, Nathalie Rimann, Mirko Saam, Lionel Sager, Emilie Sandoz. Furthermore, we are grateful to Helen Keeble and Pascale Nicolet (Pond Conservation) for improving the English style of the manuscript, as well as to two anonymous reviewers for their helpful comments. REFERENCES AFNOR. 1992. Détermination de l’Indice Biologique Global Normalisé (IBGN). NF T90-350, 9 pp. ANGÉLIBERT, S., N. INDERMUEHLE, D. LUCHIER, B. OERTLI & J. PERFETTA. 2006. Where hides the aquatic biodiversity in the Canton of Geneva (Switzerland)? Arch. Sci., 59: 225-234. AUDERSET JOYE, D., B. OERTLI, R. JUGE & J.B. LACHAVANNE. 2004. Évaluation écologique des étangs du Bois de Jussy, Département de l’Intérieur, de l’Agriculture et de l’Environnement du Canton de Genève (Service des forêts, de la protection de la nature et du paysage), University of Geneva, Geneva. 24 pp + appendix. BIGGS, J., P. WILLIAMS, M. WHITFIELD, G. FOX & P. NICOLET. 2000. Biological techniques of still water quality assessment: phase 3. Method development. Environment Agency, Bristol. 147 pp. BOIX, D., GASCON, S., SALA, J., MARTINOY, M., GIFRE, J. & X.D. QUINTANA. 2005. A new index of water quality assessment in Mediterranean wetlands based on crustacean and insect assemblages: the case of Catalunya (NE Iberian peninsula). Aquatic Conserv: Mar. Freshw. Ecosyst., 15: 635-651. BURNHAM, K. P. & W. S. OVERTON. 1979. Robust estimation of population size when capture probabilities vary among animals. Ecology, 60: 927-936. CHAO, A. 1984. Nonparametric estimation of the number of classes in a population. Scandinavian Journal of Statistics 11: 265-270. CHOVANEC, A., J. WARINGER, M. STRAIF, W. GRAF, W. RECKENDORFER, A.WARINGERLÖSCHENKOHL, H. WAIDBACHER & H. SCHULTZ. 2005. The Floodplain Index – a new approach for assessing the ecological status of river/oodplain-systems according to the EU Water Framework Directive. Large Rivers, 15: 169185. COLWELL, R. K. 2005. EstimateS: Statistical estimation of species richness and shared species from samples. Version 7.5. User’s Guide and application published at: http://purl.oclc.org/estimates. CORDOBA-AGUILAR, A. (ed) 2008. DragoniesDamselies: Model Organisms for Ecological and Evolutionary Research. Oxford University Press. Oxford, UK. 304 pp. [EC EUROPEAN COMMISSION.] 2000. Directive No 2000/60/EC of the European Parliament and of the Council. Ofcial Journal of the European Community L 327. 72 pp. FREI, M. 2007. Erfassung von Libellen (O. Odonata) an stehenden Kleingewässern-ein Methodenvergleich. Diploma thesis, University of Basel, Basel. 58 pp. GERNES, M. C. & J. C. HELGEN. 2002. Index of Biotic Integrity (IBI) for large depressional wetlands in Minnesota. Report USEPA, CD995525-01. 59 pp + appendix. GREEN, R. E., A. BALMFORD, P. R. CRANE, G. M. MACE, J. D. REYNOLDS & R. K. TURNER. 2005. A framework for improved monitoring of biodiversity: Responses to the World Summit on Sustainable Development. Conserv. Biol., 19: 56-65. GRILLAS, P., P. GAUTHIER, N. YAVERCOVSKI & C. PERENNOU. 2004. Mediterranean temporary pools. Vol. 1, Issues relating to conservation, 232 The IBEM-Index: index development functioning and management. Station Biologique de La Tour du Valat, France. 119 pp. HASTIE, T. J. & R. J. TIBSHIRANI. 1990. Generalized Additive Models. Chapman & Hall: London. 335 pp. INDERMUEHLE, N., B. OERTLI, N. MENETREY & L. SAGER. 2004. An overview of methods potentially suitable for pond biodiversity assessment. Arch. Sci., 57: 131-140. INDERMUEHLE, N., S. ANGELIBERT, V. ROSSET & B. OERTLI. 2010. The pond biodiversity index IBEM: a new tool for the rapid assessment of biodiversity in ponds from Switzerland. Part 2. Methods and examples of application. Limnetica, 29: 105-120. LEHMANN, A., J. M. OVERTON & J. R. LEATHWICK. 2002. GRASP: generalized regression analysis and spatial prediction. Ecol. Model., 157: 189-207. LEZAT, S. 2006. Biodiversité des étangs: Test de qualité de la méthode PLOCH-2, Volet Invertébrés. Bachelor Thesis, University of Applied Sciences of Western Switzerland, Jussy, Geneva. 69 pp. MENETREY, N., L. SAGER, B. OERTLI & J. B. LACHAVANNE. 2005. Looking for metrics to assess the trophic state of ponds. Macroinvertebrates and amphibians. Aquatic Conserv: Mar. Freshw. Ecosyst., 15: 653-664. MENETREY N., B. OERTLI, M. SARTORI, A. WAGNER & J. B. LACHAVANNE. 2008. Eutrophication: are mayies (Ephemeroptera) good bioindicators for ponds? Hydrobiologia, 597: 125135. MENETREY PERROTET, N. 2009. Elaboration d’une méthode d’évaluation de la qualité écologique des petits plans d’eau basée sur les communautés de macroinvertébrés. Terre & Environnement, 83: 170 pp. NEW, T. R. 1995. Introduction to Invertebrate Conservation Biology. Oxford University Press. New York. 194 pp. NICOLET, P., J. BIGGS, G. FOX, M. J. HODSON, C. REYNOLDS, M. WHITFIELD & P. WIL- 233 103 LIAMS. 2004. The wetland plant and macroinvertebrate assemblages of temporary ponds in England and Wales. Biol. Conserv., 120: 261-278. OERTLI, B., D. AUDERSET JOYE, E. CASTELLA, R. JUGE & J.-B. LACHAVANNE. 2000. Diversité biologique et typologie écologique des étangs et petits lacs de Suisse. Laboratoire d’Ecologie et de Biologie aquatique de l’Université de Genève et OFEFP. 434 pp. OERTLI, B., D. AUDERSET JOYE, E. CASTELLA, R. JUGE, D. CAMBIN & J.-B. LACHAVANNE. 2002. Does size matter? The relationship between pond area and biodiversity. Biol. Conserv., 104: 59-70. OERTLI, B., D. AUDERSET JOYE, N. INDERMUEHLE, R. JUGE, J.-B. LACHAVANNE. 2004. First European Pond Workshop. Arch. Sci., 57: 69-72. OERTLI, B., D. AUDERSET JOYE, E. CASTELLA, R. JUGE, A. LEHMANN & J.-B. LACHAVANNE. 2005. PLOCH: a standardized method for sampling and assessing the biodiversity in ponds.Aquatic Conserv: Mar. Freshw. Ecosyst., 15: 665-679. SOLIMINI, A., M. BAZZANTI, A. RUGGIERO & G. CARCHINI. 2008. Developing a multimetric index of ecological integrity based on macroinvertebrates of mountain ponds in central Italy. Hydrobiologia, 597: 109-123. SANDOZ, E. 2006. Comparaison de deux méthodes d’évaluation de la qualité biologique des étangs (PLOCH et NPS). Application à l’étang de Font (Grande Cariçaie, FR), Bachelor Thesis, University of Applied Sciences of Western Switzerland, Jussy, Geneva. 130 pp. TRIGAL, C., GARCÍA-CRIADO, F., FERNÁNDEZALÁEZ, C., 2009. Towards a multimetric index for ecological assessment of Mediterranean atland ponds: the use of macroinvertebrates as bioindicators. Hydrobiologia 618: 109-123. WILLIAMS, P., M. WHITFIELD, J. BIGGS, S. BRAY, G. FOX, P. NICOLET & D. SEAR. 2004. Comparative biodiversity of rivers, streams, ditches and ponds in an agricultural landscape in Southern England. Biol. Conserv., 115: 329-341. 104 Angélibert et al. Appendix 1. Mean values and ranges of 12 variables characterizing 63 ponds. Valores medios y rangos de las 12 variables utilizadas para la caracterización de las 63 charcas. Variable Unit Mean Minimum Maximum Median area m2 7939 66 58064 3100 mean depth cm 154 32 850 109 1.5 1.0 2.6 2.0 shoreline index (D)a conductivity μS cm−1 446 61 856 254 transparency cm 39 4 60 50 class 3.33 2 4 3.67 trophic class (PNC)b oating-leaved vegetation % 35 0 100 49 submerged vegetation % 41 0 100 52 altitude m.a.s.l. 542 305 967 423 class 2.2 1 4 3.1 pond shadec woodland (50 m environment) % 37 0 100 50 sh (1: absence; 2: presence) class 1.65 1 2 1.83 √ a Shoreline index: D = L/(2 ∗ (π ∗ S), with L = shoreline length (m), S = pond area (m2 ), π= 3.141 b Trophic class PNC: trophic class indicated by total phosphorus, total nitrogen and conductivity: (1) oligotrophic, (2) mesotrophic, (3) eutrophic, (4) hypertrophic c Pond shade: percentage of pond surface area shaded. Four classes: (1) 0 %, (2) > 0-5 %, (3) > 5-25 %, (4) > 25-100 % √ a Desarrollo del per´metro: D = L/(2 ∗ (π ∗ S), dónde L = per´metro (m), S = área de la charca (m2 ), π= 3.141 b Categor´as trócas PNC: categor´a tróca indicada por el fósforo total, nitrógeno total y conductividad (1) oligotróco, (2) mesotróco, (3) eutróco, (4) hipertróco. c Sombreado de la charca: porcentaje de supercie de la charca sombreada. Cuatro categor´as: (1) 0 %, (2) > 0-5 %, (3) > 5-25 %, (4) > 25-100 % 234 235 236 237 238 239 240 241 242 243 244 245