Using empirical and simulated data to study the influence of environmental heterogeneity on fish species richness in two biogeographic provinces

Loss of species richness in aquatic ecosystems is occurring rapidly and many factors, including habitat heterogeneity, have been suggested to affect the diversity of aquatic communities. We used fish community data (> 200 species) from extensive surveys conducted in two biogeographic provinces (extent > 1000 km) in North America to test the hypothesis that fish species richness is greater in more heterogeneous habitats (grain < 10 km2). Our tests are based on samples collected at nearly 800 stations over a period of five years. Using a set of environmental variables routinely measured by monitoring programs and a random placement model of community assembly, we demonstrate that fish species richness in coastal ecosystems is associated locally with the spatial heterogeneity of environmental variables but not with their magnitude. The observed effect of heterogeneity on species richness was substantially greater than that generated by simulations. Our modeling framework opens avenues for targeted conservation of habitat heterogeneity at broader temporal and spatial scales.