Data from: Individual-level trait diversity concepts and indices to comprehensively describe community change in multidimensional trait space
Fontana, Simone; Petchey, Owen L.; Pomati, Francesco (2016), Data from: Individual-level trait diversity concepts and indices to comprehensively describe community change in multidimensional trait space, Dryad, Dataset, https://doi.org/10.5061/dryad.b28p8
Global environmental change can influence ecosystem processes directly or through changes in the trait composition of natural communities. Traits are individual-level features of organisms, and theory predicts that diversity in traits should relate to ecosystem processes. Validated indices that account for both intra- and interspecific trait variation in multidimensional trait space are lacking. In this article, we highlight how an individual-level perspective requires new concepts for trait diversity (TD) and we validate a set of measures suitable to study trait richness, evenness and divergence at the individual scale. First, we tested a selection of multivariate indices for trait richness, evenness and divergence from the literature (FRic, FEve, FDis and the Rao coefficient) using simulated and real individual-level data. We compared the observed changes in the tested indices with those predicted from their expected/required behaviour (i.e. increase or decrease under specific manipulation of community trait structure) and found unsatisfactory results only for FRic and FEve, whereas FDis and the Rao coefficient showed the expected changes. Therefore, we propose two novel concepts and related indices for individual-level trait richness (TOP = trait onion peeling) and evenness (TED = trait even distribution). TOP represents the sum of all successive convex hull areas touching all individuals (points) within a multidimensional trait distribution. TED is a measure of how evenly distributed are individuals within the multidimensional trait space. It is calculated comparing the probability distributions of pairwise distances between individuals and between points of a perfectly even reference distribution. We tested TOP and TED on the same simulated and real data as above, and results indicated appropriate behaviour for TOP (trait richness) and TED (trait evenness). By validating TD indices in an individual-level context, this study contributes to the expansion of functional ecology towards individual-level dynamics. Future comprehensive investigations of individual trait differences in natural communities may improve our understanding of the pathways by which environmental changes affect ecosystem functioning through biodiversity change.