Multi-taxa trait and functional responses to physical disturbance.

Examining assemblage trait responses to environmental stressors extends our understanding beyond patterns of taxonomic diversity and composition, with results potentially transferable among bioregions. But the degree to which trait responses may be generalized across taxonomic groups remains incompletely understood. We compared trait responses among carabids, spiders and plants to an experimentally manipulated gradient of physical disturbance, replicated in open habitats within a forested landscape. Recolonization of recently disturbed habitats is expected to favour species with traits that promote greater dispersal ability, independent of taxa. We specifically predicted that physical disturbance would increase the representation of carabids with smaller body size, wings or wing dimorphism, spiders able to disperse aerially, and plants with therophyte life-history and wind-dispersed seed. We sampled 197 arthropod species (14,738 individuals) and 164 species of plant. The strength of association between each trait and the disturbance intensity was quantified by correlating matrices of species by traits, species abundance by sites and sites by environment, with significance assessed by comparison with a null model. Responses of biological traits varied among taxa but could be consistently interpreted in terms of dispersal ability. Trait shifts for carabid and plant assemblages were as predicted and correspond to those observed in other disturbance regimes. Assemblages after disturbance comprised smaller and winged carabids, and smaller plants with wind-dispersed seed, consistent with selection for species with better dispersal ability. In contrast, aerial dispersal did not appear important in spider recolonization, instead terrestrial dispersal ability was suggested by the increased abundance of larger-bodied and cursorial species. However, larger spider body size was also associated with an active-hunting strategy, also favoured in the post-disturbance environment. Trait-function linkage differed among taxa and was sometimes diffuse, with covariance among biological traits and the mapping of individual traits to multiple ecological functions. In particular, body size responses reflected correlations with life history, susceptibility to perturbation and dispersal ability that were inconsistent between the two arthropod groups. Selection of traits for assessment should therefore be taxa specific. Generalizations of trait responses across taxa should only be conducted where functional or ecological significance of assembly-level changes can be understood.