Abstract
General principles that shape community structure can be described based on a functional trait approach grounded on predictive models; increased attention has been paid to factors accounting for the functional diversity of species assemblages and its association with species richness along environmental gradients. We analyze here the interaction between leaf-litter ant species richness, the local communities' morphological structure and fundamental niche within the context of a northeast-southeast latitudinal gradient in one of the world's most species-rich ecosystems, the Atlantic Forest, representing 2,700 km of tropical rainforest along almost 20° of latitude in eastern Brazil. Our results are consistent with an ecosystem-wide pattern in communities' structure, with relatively high species turnover but functionally analogous leaf-litter ant communities' organization. Our results suggest directional shifts in the morphological space along the environmental gradient from overdispersed to aggregated (from North to South), suggesting that primary productivity and environmental heterogeneity (altitude, temperature and precipitation in the case) determine the distribution of traits and regulate the assembly rules, shaping local leaf-litter ant communities. Contrary to the expected and most common pattern along latitudinal gradients, the Atlantic Forest leaf litter ant communities show an inverse pattern in richness, that is, richer communities in higher than in lower latitudes. The morphological specialization of communities showed more morphologically distinct communities at low latitudes and species redundancy at high latitudes. We claim that an inverse latitudinal gradient in primary productivity and environmental heterogeneity across the Atlantic forest may affect morphological diversity and species richness, enhancing species coexistence mechanisms, and producing thus the observed patterns. We suggest that a functional framework based on flexible enough traits should be pursued to allow comparisons at local, regional and global levels.