Catchment‐scale applications of hydraulic habitat models: Climate change effects on fish
Résumé
Approaches available for estimating the ecological impacts of climate change on
aquatic communities in river networks range from detailed mechanistic models applicable
locally to correlative approaches applicable globally. Among them, hydraulic
habitat models (HABMs) link hydraulic models of streams with biological models that
reflect how organisms select microhabitat hydraulics. Coarser but more general species
distribution models (SDMs) predict changes in geographic distributions; they
generally involve coarse predictors such as air temperature or distance to source but
neglect proximate habitat descriptors such as microhabitat hydraulics. We propose
an original application of HABM for predicting the ecological impacts of climate
change at large scales, a comparison of their predictions with those of SDM and a
linkage of the two modelling approaches. We showcase our approach in a large
catchment (Rhône River) where an available distributed hydrological model estimates
present and future unregulated daily flows over the whole river network. Despite
large local uncertainties, simulations showed that climate change may strongly reduce
low flow percentiles (e.g., a median reduction of 38.6% for a pessimistic climate scenario),
inducing important alteration of fish hydraulic habitat suitability (e.g., a median
loss of 3.9%–18.7% for three modelled fish species with contrasting habitat use:
brown trout, barbel and sculpin). The HABM and SDM individually predicted consistent
or opposite fish responses to climate change, depending on the species and their
habitat requirements. Our results illustrate that accounting for ecological responses
to proximate habitat variables such as hydraulics can strongly modify projections
related to climate change.