A new method to estimate habitat potential for Atlantic salmon (Salmo salar): predicting the influence of dam removal on the Selune River (France) as a case study

Maintaining connectivity in aquatic ecosystems is important to ensure adequate ecological functioning. A large dam removal project in the Selune River (Normandy, France) would reconnect 827 km(2) of catchment area to the sea. Only the downstream section of the Selune is currently available to diadromous fish, which migrate between freshwater and the marine environment. In particular, managers focus on the future potential abundance of Atlantic salmon, Salmo salar, for conservation and fishery purposes. As in stream channel habitat drives carrying capacity of juvenile salmon, salmon abundance is usually inferred from intensive and linear habitat surveys. However, this approach is neither cost-effective for large-scale surveys nor feasible for riverbed sections with low accessibility for measurement with traditional methods, e.g. dam lakes. We used well-defined relationships between gradient, hydrology and channel habitat structure to construct a simple model to estimate potential suitable habitat for juvenile salmon. Using fine-scale habitat data from nearby rivers, we parameterized a linear mixed model to estimate the area of suitable habitat based on simple physical descriptors of river characteristics. We compared our predictions to fine-scale habitat surveys on the upper Selune. Using only slope and width, our model was able to explain 80% of the variance in suitable habitat. Estimates indicated that dam removal on the Selune River would generate a threefold increase in suitable habitat for juveniles. This could increase the mean number of adult salmon returning to the river by 1420.9 (s.e. = 1015.5). More generally, this model provides an alternative and cost-effective tool to help better manage salmon populations in rivers impacted by dams.