Christine Dupuy

The wetlands of Marais breton (MB) and Marais poitevin (MP) on the French Atlantic coast are commonly used by several duck species, especially as stopover sites during the prenuptial migration. Understanding the ecological requirements of Anatidae at spring stopover sites is important to define appropriate management actions that might have a carry over effect on the subsequent reproduction success. This study focused on the Northern Shoveler (Spatula clypeata), a species that regularly visits the two marshes during spring and fall migrations and is highly dependent on freshwater invertebrates as the food resource. Fifteen Northern Shovelers were equipped with GPS/GSM tags and monitored during their stopover in both marshes in 2020 and 2021. The aims of the study were to understand the habitat use on stopover sites and relate home range (HR) size with characteristics of the feeding habitats (such as freshwater invertebrates’ density and diversity). The HR area of the studied individuals was mainly constituted of ponds in MB (83% of the HR) and wet meadows in MP (71% of the HR). The Northern Shovelers equipped with tag spent more than 72 consecutive hours in 31 wetlands, using them during the day, at night or all day. The diurnal visited sites were deep ponds that were sparsely vegetated and dominated by microcrustaceans, whereas the nocturnal visited sites were wet meadows or ponds with high aquatic vegetation cover and high invertebrate taxonomic diversity. The 31 described sites appeared to be rich in freshwater invertebrates, with no significant difference in invertebrate densities between the diurnal and nocturnal sites. HR sizes were highly homogenous between the two study sites (MB and MP), between sexes or between age classes. In conclusion, according to this study, an appropriate HR for the Northern Shoveler at spring stopover is 8.49 ± 5.95 km² (mean ± standard error).

Trophic interaction networks are notoriously difficult to understand and to diagnose (i.e., to identify contrasted network functioning regimes). Such ecological networks have many direct and indirect connections between species, and these connections are not static but often vary over time. These topological changes, as opposed to a dynamic on a static (frozen) network, can be triggered by natural forcings (e.g., seasons) and/or by human influences (e.g., nutrient or pollution inputs). Aquatic trophic networks are especially dynamic and versatile, thus suggesting new approaches for identifying network structures and functioning in a comprehensive manner. In this study, a qualitative model was devised for this purpose. Applying discrete-event models from theoretical computer science, a mechanistic and qualitative model was developed that allowed computation of the exhaustive dynamics of a given trophic network and its environment. Once the model definition is assumed, it provides all possible trajectories of the network from a chosen initial state. In a rigorous and analytical approach, for the first time, we validated the model on one theoretical and two observed trajectories recorded at freshwater stations in the La Rochelle region (Western France). The model appears to be easy to build and intuitive, and it provides additional relevant trajectories to the expert community. We hope this formal approach will open a new avenue in identifying and predicting trophic (and non-trophic) ecological networks.

A bstract: Trophic interaction networks are notoriously difficult to understand and to diagnose (i.e., to identify contrasted network functioning regimes). Such ecological networks have many direct and indirect connections between species, and these connections are not static but often vary over time. These topological changes, as opposed to a dynamic on a static (frozen) network, can be triggered by natural forcings (e.g., seasons) and/or by human influences (e.g., nutrient or pollution inputs). Aquatic trophic networks are especially dynamic and versatile, thus suggesting new approaches for identifying network structures and functioning in a comprehensive manner. In this study, a qualitative model was devised for this purpose. Applying discrete-event models from theoretical computer science, a mechanistic and qualitative model was developed that allowed computation of the exhaustive dynamics of a given trophic network and its environment. Once the model definition is assumed, it provides all possible trajectories of the network from a chosen initial state. In a rigorous and analytical approach, for the first time, we validated the model on one theoretical and two observed trajectories recorded at freshwater stations in the La Rochelle region (Western France). The model appears to be easy to build and intuitive, and it provides additional relevant trajectories to the expert community. We hope this formal approach will open a new avenue in identifying and predicting trophic (and non-trophic) ecological networks.

The Northern shoveler, Spatula clypeata, makes several migratory stopovers to reach its breeding site in the best conditions. These stopovers allow the species to replenish their reserves. Therefore, feeding efficiency at such sites is essential. Despite its importance, few studies have been conducted on the spring ecology of the shoveler, particularly on its diet at stopover sites. Therefore, this study focused on the Northern shoveler's feeding habit during its spring migratory stopover in the Marais breton (MB), a wetland situated in Vendée (France, Atlantic coast). The shoveler's plasma and potential food resources were studied using a stable carbon and nitrogen isotope analysis. The study showed that the shoveler feeds primarily on microcrustaceans, especially Cladocera and Copepoda; Chironomidae larvae, Corixidae, Hydrophilidae larvae, and particulate organic matter. This last food source, the POM, had never been highlighted before.