ASSESSING MAMMAL FUNCTIONAL DIVERSITY USING BODY MASS

Functional diversity (FD), approximated using an array of published metrics, describes the range of niches occupied by species in an assemblage. FD is correlated with various climate, biogeographic, and macroevolutionary parameters, suggesting both historical and proximate controls. Changes in the FD of communities may also reflect fundamental alteration of trophic interactions as well as the services provided by ecosystems. FD is therefore an ideal metric for understanding how communities respond to perturbation and important target for conservation. The Cenozoic fossil record, which encompasses at least eight different periods of climate change and repeated biogeographic events, is the result of a 66 Ma natural experiment that may hold the key to understanding and predicting community-scale responses to perturbation. However, the functional traits typically used in the calculation of FD for modern mammal assemblages (e.g., activity time, life history, percent dietary composition) are not known for most extinct species. The most accessible trait for estimating the functional role of species and, by extension, the FD of mammal communities, is body mass (BM), because it is highly correlated with linear measurements of the dentition and skeleton. Furthermore, differences in BM are associated with differences in ecological function, including trophic position. It remains unclear, however, whether estimates of FD from BM alone are sufficient for understanding the functional characteristics of mammal communities. Herein, we compare measures of FD for modern mammals, globally, to assess how BM-only metrics compared to more holistic measures of FD. Using phylogenetic generalized linear models, we show that BM is correlated with nearly all ecological metrics included in previous studies of extant mammal FD. We also show that macroecological patterns of mammal FD are similar whether calculated using BM only or from a range of traits, and that BM explains the majority of spatially-explicit generalized least squares model variance. We therefore suggest that the FD of fossil mammal communities can be estimated using body mass alone, allowing the ecological consequences of Cenozoic abiotic and biotic perturbations to be investigated using a similar framework as for modern species assemblages.