The geographic variation in species richness across continents remains one of the most intriguing patterns in ecology, biogeography and evolution. Correlative approaches have identified important environmental variables describing such pattern, but no causal mechanisms are incorporated into these approaches. Alternatively, new approaches focus on the mechanisms that determine the distribution of individual species to understand patterns of species richness. Overlapping species’ geographic ranges has become standard to assess spatial variation in species richness and describe broad-scale patterns of diversity. The overlap of geographic ranges, in turn, is determined by the size, shape and location of ranges. Consequently, considering the internal structure of geographic ranges can enhance the understanding of large-scale patterns of species richness. In this study, we examined the geographic variation in species richness for New World leaf-nosed bats focusing in geographic associations among species and introducing the “diversity field”. Our goals were to describe how species richness is distributed inside individual species’ ranges, identify patterns related to species’ characteristics and determine the potential factors shaping patterns of species associations at geographic scales. We constructed a database of the continental distribution of phyllostomid bats and generated a presence-absence matrix. Taking advantage of recent analytical developments, we extracted information from such matrix to measure the overlap of all species within the range of each individual species. We described and examined the diversity field, defined as the set of species richness values of sites where a given species occurs, with three complementary approaches: 1) the analysis of species richness frequency distributions, 2) the examination of maps showing the range of species with the corresponding richness values, and 3) the construction of range-diversity plots. We model the shape and location of species’ geographic ranges to generate statistical null hypothesis.

Results/Conclusions

Our results show a high degree of association among these bats, most species associate positively with the rest of species. More interesting, the majority of species doesn’t have a homogeneous pattern of range overlaps, hence experiencing different coexistence scenarios throughout their ranges, describing a mosaic of potential interactions. Furthermore, observed patterns are significantly different from null expectations; co-occurrence is much higher than expected. Altogether, our analyses highlight the importance of mechanistic approaches to evaluate geographic patterns of diversity and support a critical role for historical factors in the structure of geographic ranges and large-scale biodiversity patterns.