Discerning patterns of diversity from biogeographical distributions: testing models of metacommunity dynamics using non-marine ostracodes from San Salvador Island, Bahamas

Empirical examples of natural metacommunities lag behind theoretical developments and therefore are needed to understand how the relative contributions of dispersal and environmental filtering varies taxonomically and in different environments. Here, we use the geographic distributions of ostracode species and their morphological traits in lakes on San Salvador Island, Bahamas to test the hypothesis that their metacommunity dynamics are dominated by species sorting. We sampled thirty-two lakes for ostracode abundance, morphological traits, and limnological variables. The abiotic environment of lakes was found to vary mostly independently of space, allowing for the evaluation of the distinct roles of spatial and environmental variables. Differences between assemblages were not found to be spatially auto-correlated, indicating that dispersal limitation of species is not an important factor influencing community assembly. Ostracode assemblages and community-weighted aggregated species’ traits were found to vary systematically with conductivity, alkalinity, and dissolved oxygen, showing that the abiotic environment is a strong filter. While conductivity did display low, but significant, spatial structure, halophilic species followed this spatial pattern. This environmental filter combined with no effects of dispersal is consistent with only a species sorting model. This study thus provides a valuable example of metacommunity models applied to a natural system by specifying the relevant ecological factors that govern community assembly.