Global biogeographical patterns of ants and their abiotic determinants

Abiotic determinants of global biogeographical patterns are well known for a range of taxa, particularly plants and vertebrates. However, they are largely undescribed for most arthropods which constitute the bulk of the world biodiversity. Based on recently updated ant databases, we estimate the relative strength of abiotic and topographic determinants of ant species richness at the global scale, and investigate how such determinants vary across zoogeographic realms. We used ant species distribution to estimate ant species richness, the number of sampling points, and estimated species richness per cell of 2° worldwide. We fit simultaneous autoregressive models (SAR) to test the relative importance of six explanatory variables: annual rainfall, rainfall seasonality, mean temperature, actual evapotranspiration, mean altitude, and topographic heterogeneity, on estimated species richness. At the global scale, estimated ant species richness was best explained by annual rainfall and mean temperature. However, while annual rainfall affected positively estimated species richness for most realms, the sign of the mean temperature varied between realms. Furthermore, rainfall seasonality, actual evapotranspiration, mean altitude and topographic heterogeneity affected negatively estimated ant species richness in a number of realms. The variation in ant diversity determinants between realms challenges the generality of the global pattern and demonstrates the importance of regional ecological and historical settings. This provides useful information for conservation planning and helps predict how ants will respond to global and regional climate change.