Modelling and mapping the distribution, diversity and abundance of freshwater mussels (Family Unionidae) in wadeable streams of Illinois, USA

Freshwater mussels are one of the most imperilled animal groups in the world. Their effective conservation and restoration require a better understanding of their spatial distributions at a relevant scale and of their relationships with natural environmental factors and human disturbances. In this study, we sampled over 900 sites on wadeable streams throughout Illinois, U.S.A., and compiled environmental data for a wide range of natural and anthropogenic factors related to climate, geology, land use, and connections to large rivers, dams and ponds. Using random forest classification and regression, we modelled the presence-absence of mussels as a group (87% accuracy), the abundances of 29 individual mussel species (R-2=0.2-0.51), species richness (R-2=0.52) and total mussel abundance in a standard sample (R-2=0.41). The abundances of most species increased with stream size, the proportion of agricultural land in the catchment and the distance to the nearest dam or pond, but decreased with increasing catchment or channel slope and the proportion of forest in the catchment. Species varied in their relationships with climate variables, suggesting that they respond differently to climate change. Geology, particularly bedrock depth, was important for many species. Species richness and total mussel abundance responded positively to stream size and negatively to the slope of streams or catchments. The models were applied to unsampled wadeable stream reaches to generate mussel distribution maps at the reach scale, useful tools for resource managers to effectively protect and restore mussel biodiversity. The models also improve our understanding of how mussel populations and assemblages are structured by natural factors and human disturbances at a broad scale.