Catchment‐scale modeling of riverine species diversity using hydrological simulation: application to tests of species‐genetic diversity correlation

Abstract Species distribution models were developed to predict the spatial patterns of the species diversity and the genetic diversity of stream organisms using a distributed hydrological model. We derived annual metrics of water depth and flow velocity in rivers using a hydrological model covering an entire catchment in northeastern Japan. We acquired geospatial data throughout the catchment and the presence records of six taxa within the part of the catchment. Subsequently we derived habitat suitability indices (HSIs) for these taxa using a frequency analysis or the maximum‐entropy approach (MaxEnt) to predict three metrics of species diversity. The Shannon‐Wiener's diversity index based on MaxEnt (Shannon_MaxE) most effectively represented empirical taxon richness. Subsequently, by comparing Shannon_MaxE and empirical genetic diversity for the four species of stream insects, we evaluated species‐genetic diversity correlations (SGDCs). Of the four species, only one caddisfly species ( Hydropsyche orientalis ) displayed significant positive SGDCs. The result reflects the broad habitat range of this taxon throughout the catchment and its poor dispersal ability, whereas the other three species lacked SGDCs and displayed either a strong dispersal potential ( Stenopsyche marmorata and Ephemera japonica ) or a narrower habitat range within upstream domains ( Hydropsyche albicephala ). Finally, we estimated the spatial distribution of genetic diversity of H. orientalis populations based on the calculated Shannon_MaxE using the positive SGDC. This framework is promising for projecting future biodiversity in the context of anthropogenic perturbations such as climate change.