Natural communities are often assembled through an interaction between species traits and habitat conditions, i.e., niche assembly. Consequently, community composition should be strongly influenced by interactions between the filtering properties of different biogeographic regions and the evolved traits of taxa within regional species pools. We investigated how biogeographic differences in the variability (both within and between years) of environmental stress gradients alters local and regional community assembly. We predicted that species pools in regions with predictable environmental gradients across years should consist of species with narrow habitat ranges, leading to high between-habitat β-diversity. In contrast, regions where stress gradients vary among or within years are likely dominated by habitat generalists, and have low β-diversity. We tested this hypothesis with a meta-analysis of published studies of wetland invertebrate communities across regions that vary in the seasonal predictability of environmental stress from drying disturbances. Within regional clusters of wetlands, we calculated indices of species turnover including β-diversity, nestedness and species habitat breadth. We also quantified climatic and hydrological variables that determine the predictability of drying and filling regimes.
Results/Conclusions
We found that wetland communities in alpine regions with predictable hydrology from snowmelt were dominated by taxa with narrow distributions across the drying gradient, resulting in high β-diversity among habitats. In contrast, species pools in regions with variable hydrology from unpredictable rainfall (either inter-annual or timing within a year) were dominated by habitat generalists, leading to high overlap in species composition among wetlands. The presence and identity of vertebrate predators in a regional species pool also influenced species turnover across the drying gradient. These results indicate that climatic properties and evolutionary constraints acting at biogeographic scales can modify patterns of community assembly across environmental gradients. Integrating these mechanisms across biogeographic, regional and local scales is likely to be necessary to predict patterns of community assembly.