Abstract
Species traits may provide a short-cut to predicting generalities in species turnover in response to environmental change, particularly for poorly known taxa. We ask if morphological traits of assemblages respond predictably to macrohabitats across a large scale. Ant assemblages were collected at nine paired pasture and remnant sites from within three areas along a 300 km distance. We measured ten functional morphological traits for replicate individuals of each species. We used a fourth corner model to test associations between microhabitat variables, macrohabitats (pastures and remnants) and traits. In addition, we tested the phylogenetic independence of traits, to determine if responses were likely to be due to filtering by morphology or phylogeny. Nine of ten traits were predicted by macrohabitat and the majority of these traits were independent of phylogeny. Surprisingly, microhabitat variables were not associated with morphological traits. Traits which were associated with macrohabitats were involved in locomotion, feeding behaviour and sensory ability. Ants in remnants had more maxillary palp segments, longer scapes and wider eyes, while having shorter femurs, smaller apical mandibular teeth and shorter Weber's lengths. A clear relationship between traits and macrohabitats across a large scale suggests that species are filtered by coarse environmental differences. In contrast to the findings of previous studies, fine-scale filtering of morphological traits was not apparent. If such generalities in morphological trait responses to habitat hold across even larger scales, traits may prove critical in predicting the response of species assemblages to global change.