Abstract
Gradients in cuticle lightness of ectotherms have been demonstrated across latitudes and elevations. Three key hypotheses have been used to explain these macroecological patterns: the thermal melanism hypothesis (TMH), the melanism-desiccation hypothesis (MDH) and the photo-protection hypothesis (PPH). Yet the broad abiotic measures, such as temperature, humidity and UV-B radiation, typically used to detect these ecogeographical patterns, are a poor indication of the microenvironment experienced by small, cursorial ectotherms like ants. We tested whether these macroecological hypotheses explaining cuticle lightness held at habitat and microclimatic level by using a vertical gradient within a tropical rainforest. We sampled 222 ant species in lowland, tropical rainforest across four vertical strata: subterranean, ground, understory and canopy. We recorded cuticle lightness, abundance and estimated body size for each species and calculated an assemblage-weighted mean for cuticle lightness and body size for each vertical stratum. Abiotic variables (air temperature, vapour pressure deficit and UV-B radiation) were recorded for each vertical stratum. We found that cuticle lightness of ant assemblages was vertically stratified: ant assemblages in the canopy and understory were twice as dark as assemblages in ground and subterranean strata. Cuticle lightness was not correlated with body size, and there was no support for the TMH. Rather, we attribute this cline in cuticle lightness to a combination of the MDH and the PPH. Our findings indicate that broad macroecological patterns can be detected at much smaller spatial scales and that microclimatic gradients can shape trait variation, specifically the cuticle lightness of ants. These results suggest that any changes to microclimate that occur due to land-use change or climate warming could drive selection of ants based on cuticle colour, altering assemblage structure and potentially ecosystem functioning.