Associations between environmental variables and species richness are ubiquitous. These associations have been proposed to reflect cause-effect relationships between environmental characteristics and the ability of habitats to support or generate different numbers of species; however, there are numerous different historical, ecological and stochastic mechanisms underlying species-environment relationships that are not mutually exclusive, and that probably interact to generate observed patterns of spatial variation in biological diversity. In this study, we investigated the relationship between species richness of New World bats and a group of variable sets that represent three of these possible mechanisms: energy, habitat heterogeneity and seasonality. Moreover, we studied the form of these relationships across a range of latitudinal extents and latitudinal positions. Our analyses were based on a map of the New World divided into cells of 2500km². Species richness was calculated by counting the number of species whose ranges overlapped a grid cell. The “energy” predictor set included mean values of annual precipitation, annual temperature, and annual NPP. The “habitat heterogeneity” set was comprised of spatial standard deviations of elevation, annual precipitation, annual temperature and annual NPP; finally, the “seasonality” set included measures of temporal variability in temperature and precipitation. We used regression and variation partitioning analyses to investigate relationships between hypothesized mechanisms and species richness, as well as their interactions with spatial scale and latitudinal location. 
Results/Conclusions

Our results demonstrated that both energy and seasonality explain the great majority of variation in bat species richness, while habitat heterogeneity explained about half of the total variation. However, when co-variation among predictors was accounted for, seasonality and energy were highly redundant, and there was little evidence of an important unique component associated with habitat heterogeneity. Nevertheless, explanatory power associated with several of the different hypotheses or their combinations exhibited strong patterns of variation across latitudinal extent and latitudinal position. For example, seasonality was a poor predictor at small latitudinal extents, and importance of energy increased significantly at higher latitudes. Our results stress the importance of interactions among sets of predictors in shaping large scale patterns of species richness, and suggest that influence of these predictors is scale dependent, and varies with geographic location.