French naiad (Bivalvia: Margaritiferidae, Unionidae) species distribution models: prediction maps as tools for conservation

Freshwater mussels are amongst the most threatened invertebrate species worldwide. For some species, current distribution has contracted to a point whereby original extent and ecological requirements are difficult to assess. As a consequence, extirpation rate is poorly evaluated, surveys are not targeted towards suitable areas and species are not taken into account in development impact studies. In this paper, we developed species distribution models (SDM) and produced maps of suitable habitat for conservation purposes. We focused on Margaritifera and Unio species for which enough reliable data could be collated for whole continental France. For the first time at this scale, SDM were based on a river stretch framework which proved to be very efficient. Models performances were highly discriminative, allowing drawing predictive maps of habitat suitability. As no human-induced variables were included, our models approximate species’ ideal original range. Range contraction could then be quantified by comparing the extent of occurrence (EOO) predicted by the models to the currently known EOO of the species. For most species, known distribution matches the predicted suitable habitat suggesting that geographic barriers between main drainages do not impact naiads’ distribution.