Environmental gradients such as those tracking variation in elevation have been used as natural experiments for decades. Typically diversity along elevational gradients tracks abiotic variables such as temperature or productivity. But at local scales, biotic interactions, such as competition for resources, might limit diversity. In order to assess the extent to which environmental variation and local processes interact to determine the structure of local ant communities, we manipulated food availability and microclimatic conditions at 15 sites along a well-studied elevational gradient in southeastern USA. Specifically we predicted that at low elevations, where climatic conditions are mild, local processes should act to structure communities. But at high elevations, where selection should favor those species which are adapted to tolerate cold temperatures, environmental factors should limit local communities.
Results/Conclusions

We found that ant species density declined with elevation. Within one m² of leaf-litter, ant species richness varied from 1 to 8 species, and worker abundance varied from 1-259. In addition, once the effect of elevation was statistically removed, we found that experimentally reducing ground temperature decreased ant species density, whereas adding food had little effect. Ant species density was 12% lower in reduced temperature quadrats than in control quadrats and ant worker abundance was 39% lower. Our results further show that the effect of altering microclimate was independent of elevation: reducing ground temperature decreased ant species density equally at low and high elevation. We conclude that using manipulative experiments along gradients can help tease apart the relative importance of local vs. regional processes in shaping ant communities.