Les relations habitat – poissons, de l’éthohydraulique à la dynamique des populations et des communautés

Fish-Habitat relationships, from ethohydraulics to population & community dynamics

Capra, H.

Type de document
HDR
Langue
Français
Affiliation de l'auteur
IRSTEA LYON UR MALY FRA
Année
2014
Résumé / Abstract
L'objectif de mes recherches est d’acquérir des connaissances sur l’écologie des poissons en cours d’eau, et plus particulièrement sur les stratégies d’utilisation de l’habitat par les poissons. Mes travaux sont réalisés principalement sur le terrain, à différentes échelles d'espace et de temps avec la volonté de se rapprocher à la fois des mécanismes qui agissent à l'échelle des individus (locale) mais également à l’échelle des populations (globale). Le premier axe de mes recherches vise à comprendre les réponses individuelles des poissons à des changements de conditions d’habitat physique dans les cours d’eau. Ces travaux se concrétisent par le développement de modèles de préférences d’habitat à partir d’échantillonnage ponctuels par pêche électrique sur de nombreux cours d’eau, par l’étude des stratégies de sélection de l’habitat grâce à des études comportementales ou des expérimentations en milieu contrôlé pour tester des hypothèses avec le moins de facteurs confondants possibles. Le second axe de mes recherches repose sur la modélisation de la dynamique des populations ou des peuplements en intégrant les effets significatifs des facteurs physiques sur la démographie des populations, établis à l’échelle individuelle ou à partir de méta-analyses. Je privilégie l'approche multi-échelles pour mieux cerner les relations de cause à effet des facteurs physiques de l’environnement aquatique concomitants des effets d’autres facteurs environnementaux souvent nombreux (ex : température, qualité d'eau, biogéographie). Les connaissances acquises (e.g. courbes de préférences hydrauliques, détermination de phases limitantes du cycle de développement) sont utilisées dans des outils tels que les modèles d’habitat pour la gestion des aménagements industriels sur les cours d’eau (barrages, usines hydroélectriques, centrales nucléaires) et la restauration des écosystèmes d'eau courante. Les modèles d’habitat sont des outils qui permettent de prédire les conséquences probables d’un changement de débit ou d’une modification morphologique en termes de structure des communautés piscicoles (e.g. proportion d’espèces rhéophiles vs lenitophiles). Ils permettent de définir un compromis entre les exigences d’habitat des biocénoses et les usages anthropiques. J’ai maintenant comme objectif d’approfondir les recherches sur les relations habitat – poissons à l’échelle du cycle vital, non plus stade par stade, mais dans la continuité spatiale et temporelle du cycle. Quelques résultats importants Echelle locale – individu – court terme Les modèles régionaux de préférences hydrauliques établis pour 24 espèces échantillonnées dans 6 cours d’eau sont très robustes puisqu’ils permettent d’expliquer dans deux tiers des cas les préférences locales de ces espèces. Les tests de transférabilité des courbes de préférences font de ces modèles une référence en France, pour toutes les études utilisant des outils de simulations d’habitat. L’expérimentation en milieu contrôlé menée avec M. Daufresne (doctorant 2000 – 2004) confirmait que les alevins émergents de truites étaient sensibles à une augmentation de vitesse de courant les 6 premiers jours après l’émergence, mais que cela ne faisait qu’accélérer la cinétique d’un processus de dispersion qui concerne 80% des alevins émergents. Les études comportementales dans le Rhône (télémétrie) nous ont permis de connaitre les stratégies comportementales mises en œuvre par les poissons pour répondre aux fluctuations des conditions environnementales. Les adultes de cyprinidés et de silure utilisaient principalement les zones de berge (non échauffées pour les cyprinidés) où la variabilité artificielle (éclusées) des conditions hydrodynamiques était la moins contraignante. Avec J. Bergé (doctorant 2008 – 2012) nous avons mis en évidence l’importance de la variabilité interindividuelle des réponses. Associé à l’étude des comportements un travail conséquent de terrain, d’étalonnage et de validation a été réalisé pour établir un modèle hydrodynamique en deux dimensions (2D). Ce modèle était nécessaire pour connaitre en tout point du site d’étude et pour tous les débits observés dans le Rhône à Bugey les conditions d’habitat disponibles (hydraulique et température) pour les poissons. Echelle globale – metapopulations – long terme Ma participation active à la « Cellule des débits réservés » m’a permis de réaliser une étude de validation des modèles d’habitat à l’échelle des populations de truites communes. Les habitats jugés classiquement structurant pour les populations de truite, tels que l’habitat favorable disponible pour les adultes lors des étiages estivaux (notion de capacité d’accueil) sont finalement rarement limitant. D’autres stades de développement sont plus sensibles à certaines phases de la variabilité saisonnière de l’habitat (e.g. effets limitants des crues printanières sur le recrutement) et d’autres facteurs tels que les déplacements de juvéniles (1+) peuvent aussi influencer la dynamique des populations de truites. Les suivis à long terme sur les rivières salmonicoles et sur le Rhône ne mettent en évidence des réponses significatives des communautés de poissons aux modifications des conditions d’habitat que lorsque ces dernières sont importantes (e.g. vitesse moyenne du courant multipliée par 2). Les études comportementales à l’échelle des saisons et de plusieurs dizaines de kilomètres de cours d’eau renseignent sur la taille des domaines vitaux individuels et par extension, de l’emprise spatiale de la dynamique des populations. Les domaines vitaux annuels des cyprinidés dans le Rhône varient de quelques centaines de mètres à plusieurs dizaines de kilomètres. Ces valeurs sont une première approche des exigences de ces espèces holobiotiques en termes de continuité physique dans un grand fleuve, mais sont fortement liés à la taille des tronçons de cours d’eau entre deux barrages infranchissables. Perspectives Après avoir identifié quelques éléments clefs du rôle de l’habitat physique dans l’écologie des poissons en cours d’eau (dispersion de 80% des très jeunes alevins, déplacements des adultes, influence de la capacité d’accueil) je souhaite m’orienter vers des modèles de dynamique des populations et des communautés plus « fonctionnels », i.e. en lien avec la disponibilité spatiotemporelle des habitats vitaux (reproduction, développement des très jeunes alevins, alimentation, repos). D’autre part les flux d’individus au sein d’une mosaïque d’habitat décrite à grande échelle spatiale, celle des communautés de poissons, tels que la dispersion des très jeunes alevins, les flux de juvéniles et les migrations d’adultes, doivent être quantifiés pour mieux modéliser la dynamique des communautés.
Position at Irstea My research at Irstea (Lyon-Villeurbanne) is within the research group (TR) QUASARE (Quality of Aquatic Systems and Ecological Restoration) and of the team Dynam of the research unit AEEP (Aquatic Ecosystems, Ecology and Pollution). TR QUASARE includes 7 teams at 4 Irstea centers to conduct research on the responses of aquatic anthroposystems (rivers, lakes, estuaries) and their biological communities to local and global anthropogenic pressures (risk of hydro- morphological alterations, biological and chemical risks). It aims to develop methodologies to maintain or restore the functional state of these systems. The team Dynam, led by Nicolas Lamouroux, develops a comparative and quantitative approach to interactions between hydrology, ecology and substrate in a variety of ecosystems (mountain streams, large rivers, intermittent streams). My career path: from ecohydraulics to ethohydraulics The aim of my research is to gain knowledge on the ecology of fish in rivers, by studying the strategies developed by fish for habitat use. This knowledge (e.g. hydraulic preference curves, determining limiting phases of the development cycle) is used in habitat models for management and restoration of rivers. My research is field oriented and carried out at different scales of space and time to identify mechanisms acting at the level of individuals (local) but also at the level of populations (global). This multiscale approach is important to better understand the cause and effect of physical factors in the aquatic environment versus effects of other environmental factors (e.g. temperature, water quality, biogeography of species). The topic ethohydraulics well summarizes my current research directions: towards a behavioral analysis (ethology) of the use of physical habitat (hydraulics) by fish in the rivers. The development of a sampling strategy of the fish communities by direct observation (snorkeling) adds a new and unique perspective to an ecological-based approach. Management context: towards new habitat models My research fits into the operational framework for the management of industrial development on rivers (dams, hydroelectric plants, nuclear power plants) and more recently for the physical restoration. Habitat models are tools that predict the likely consequences of a change in instream flow, or in river morphology, on structure of fish communities (e.g. proportion of rheophilic species vs lenitic species). Fish habitat modeling is an essential part of quantitative flow management and physical restoration of streams and rivers (Tharme 2003). Within the regulatory framework established by the French Water Law of December 2006 and the European Framework Directive (Acreman and Ferguson, 2010), habitat models contribute to the definition of green regimes in rivers, i.e. looking for a compromise between habitat requirements of biotic communities and human uses (e.g. water supply, irrigation, hydropower, cooling plants). A strong desire to restore ecological continuity (blue frame) at a global scale has become indispensable. It is important to perform habitat modeling and sampling at the scale of river networks rather than the conventional river section scale. The number of studies aimed at defining minimum flow (or regime) at the watershed scale is increasing (Acreman in press ; Wilding et al. in press). We have to leave the river section to better understand the determinisms of the fish population dynamics at a global scale. I now aim for further research on the habitat relationships of fish across the spatial and temporal continuity of their life cycle. Theoretical context: from ecohydraulics to ecological continuity Changes in fish community structure are related to their evolutionary history and current environment (Blanck et al. 2007). Over time, environmental constraints have often contributed to the emergence of distinct communities between regions and continents (Moyle & Herbold 1987). However, comparable environmental characteristics in different regions can promote the same morphologies or demographic strategies in a predictable manner (Schluter 1986, Townsend & Hildrew 1994, Lamouroux & Cattanéo 2006). My work focuses on understanding the role of physical habitat in structuring fish communities in rivers. The first direction of my research is to understand individual responses of fish to changes in conditions of physical habitat in streams. I am developing models of habitat preferences from point sampling with electrofishing on many rivers. I also study strategies of habitat selection through behavioral experiments in situ or in a controlled environment to test hypotheses with the least possible confounding factors. The second direction of my research is focused on modeling population dynamics (or community) by incorporating the significant effects of physical factors on population demographics, established at the individual level or from meta-analysis. Main results Local scale – individual – short term Regional models of hydraulic preferences developed for 24 species sampled in six rivers are very robust since they can explain two-thirds of local preferences of these species. The transferability of these models have been tested and they are now used as a reference for all habitat simulations in France (Lamouroux et al. 1999). Experiments conducted in a controlled environment (with my first PhD. Student M. Daufresne ; Daufresne et al. 2005) confirmed that the emerging fry trout were sensitive to an increase in flow velocity the first 6 days after emergence, but this is just an acceleration of the kinetics of a dispersion process which involves 80 % of the emerging fry. Behavioral studies in the Rhône River (with acoustic telemetry) identified different behavioral strategies used by fish in response to changes in environmental conditions (discharge, temperature). The adults of cyprinids (barbel and chub) and of catfish primarily used habitat along the banks where artificial variability of hydrodynamic conditions (during peaking flows) were less stringent. Barbel and chub selected rarely habitat with water temperature higher than 25°C. With my second PhD. Student, J. Bergé, we highlighted the importance of interindividual variability in responses. A 2D hydrodynamic model of the study site was calibrated and validated based on a substantial fieldwork. This model was developed to simulate the habitat conditions (hydraulics and temperature) available for fish in , in any point of the study site and for all the observed discharges. Global scale – metapopulations – long term To validate IFIM in biological terms on reaches with power plants, experiments have been carried out on eight bypassed river reaches downstream of hydropower stations following an increase in the minimum instream flow. They are being conducted by a France-wide working group on “Guaranteed Flow” that groups experts from different research bodies, government agencies and Electricité de France (Merle and Eon 1996 ; Sabaton et al. 2004). My active participation in the Guaranteed Flow Working Group has allowed me to carry out a validation study of habitat models at the population scale (brown trout populations). Habitat conditions which are classically considered as structuring for trout populations, such as availability of suitable habitat for adults during the summer low flows, are actually only limiting factors for a few populations. Indeed, other developmental stages may be more sensitive to certain phases of the seasonal variability of the habitat conditions (e.g. limiting effects of spring flooding on the recruitment; Cattanéo et al. 2002). Other factors, such as immigration of juveniles (1 year +), can also influence the trout population dynamics. The long-term monitoring of fish assemblages in the Rhone River (e.g. monitoring of the structure of the fish community in the Rhone River at Bugey or in the bypass section of Pierre- Benite) highlight responses of fish communities to significant changes in habitat conditions (Sabaton et al. 2008 ; Daufresne et al.submitted). The long-term monitoring of trout populations and fish communities of the Rhone River, made at the scale of the river section, is needed to detect major changes in the evolution of these populations and of these communities. Expanding the study scale in fish surveys is needed to better understand the influence of physical habitat on community structure. Behavioral studies across seasons and over several kilometers of waterways provide information on the size of individual home ranges and by extension, the spatial extent of population dynamics. The annual range of the cyprinid movements in the Rhône vary from a few hundred meters to several kilometers. These values enable to define more precisely requirements of these holobiotic species in terms of physical continuity in a large river, but are strongly related to the size of the sections of river between two impassable dams. Prospects Having identified some key elements relating physical habitat to the ecology of fish in rivers (see before: 80% dispersion of very young fry, the movements of adults, the role of the carrying capacity) I wish to develop more functional models of population dynamics and of community dynamics. Such approaches will take account of the processes which take place at different levels of space and time, to connect the real time for fish (local scale) to the long term which determine population structures (global scale). Research at the individual level (i.e. behavioral approaches) are needed to better understand the strategies of fish in response to changes in habitat availability, also including the temperature (Junge et al. 2011; Tonolla et al. 2012; Wawrzyniak 2012; Armstrong et al. 2013), in connection with vital / functional habitats (reproduction, development of very young fry, feeding, resting). On the other hand, the movements of fish within a mosaic of habitat, such as the dispersal of very young fry, must be quantified at a large spatial scale (Fausch 2002 ; Torgersen et al. 2006), not only in terms of spatial extent, but also in terms of impact on the population dynamics and genetics (Massa- Gallucci et al. 2010 ; Junge et al. 2011 ; Blum et al. 2012 ; Shipham et al. 2013; Hughes et al. 2014). Local scale – individual – short term * Dynamics of functional habitats During the 63° Mission of the Permanent Commission of the “Franco –Québécoise” Cooperation, I started an international collaboration which aims to develop a dynamic link between the available habitat (availability varying in time and space) and the habitat required by fish to maintain their vital functions (Le Pichon et al. 2006). The habitat characterization (estimation of a habitat value) will be based on knowledge about the behavioral use of the functional habitats. * Translocation project This project aims to complete the behavioral study conducted in 2009 and 2010 on the Rhône River (collaboration with M. Ovidio and H. Pella). The objective would be to study the movements of individuals collected from a highly variable environment and released in a much more stable environment (e.g. in a natural or in a heavily regulated system) and vice versa, if possible on several river sections. This study will determine how genetic (i.e. species) or environmental (i.e. adaptation) factors influence individual responses to changes in habitat variability. * Post-Doc for fish trajectories I will select one post-doctoral researcher (collaborations with H. Pella, H. Drouineau, P. Lambert and M. Ovidio) to describe more accurately the behavior of fish in their environment, especially during important changes in habitat availability. Data from the telemetry experiment carried out in 2009 (PhD thesis of J. Bergé), with a recording of the positions of fish every 3 seconds, may allow us to estimate the distance, the speed, the frequency and the spatial extent of trajectories performed by the marked individuals. We already showed fish habitat selection and the spatial repartition of fish in the study site. But further research may be conducted to know the shape of the trajectories of fish, the variations of these forms under certain environmental influences (changes of water level, of temperature, or presence of intra- and inter-species relationships) and depending on the habitat availability in the river. Global scale – metapopulations – long term * Direct observation (snorkeling) sampling project The results of long-term monitoring of fish communities have demonstrated the need to quantify the movements of fish at a large spatial scale (dispersal of fry, annual movements of juveniles and migrations of breeding adults). This metapopulation context (Massa- Gallucci et al. 2010; Benjamin et al. 2014; Mari et al. 2014; Yeakel et al. 2014) leads us to sample fish populations at a large spatial scale. But electro-fish sampling (i.e. conventional sampling technique) make it difficult to sample long sections of river habitat. Direct observation sampling is being used within DYNAM to (1) assess the spatial distribution of species and of species associations and (2) to detect the effects of the confluences or of the fragmentation (e.g. dams, dry river sections) on these distributions. * Eels project It is important to understand the ecology of eels because of the significant decrease the stock. However, habitat preference, migration, behavior, and population dynamics of the European eel is rarely assessed in large rivers. I am developing a project (1) to estimate the stock of silver eels, (2) to study their behavior during their downstream migration, and particularly in front of the dams and (3) to track downstream migration routes. A first experiment has just been completed (April 2014) which confirms the effectiveness of acoustic telemetry equipment available at Irstea (HTI telemetry system) to study the downstream migration of silver eels (collaboration with CNR, Onema and MRM). * Dispersal project (trout and/or cyprinids) The dispersal of fry just after their emergence or their hatching is poorly known for fish. This is, however, a critical phase of the development cycle. I am developing a project to monitor fry behavior in situ (ideally individually marked) from the spawning redds to the habitat used before the first winter. I will compare these movements (voluntary or due to habitat changes) to a dynamic description of the available habitat (spatial and temporal) to assess the role of the physical environment in the dispersal of young larvae / fry. Questions addressed by M. Novak in her PhD thesis (supervised by J.-M. Olivier and P. Sagnes, Lehna, Lyon 1 - 2014 to 2017) may identify spatial origins of the individuals sampled in the Rhône (isotopic labels) and track their dispersal. I also intend to quantify the movements of fish (dispersal of fry and migrations of juveniles) in streams.
Diplôme
Habilitation à Diriger des Recherches, Université Lyon I

puce  Accés à la notice sur le site Irstea Publications / Display bibliographic record on Irstea Publications website

  Liste complète des notices de CemOA