Summary
As conservation increasingly recognizes the importance of species’ functional roles in ecosystem processes, studies are shifting away from measuring species richness towards measures that account for the functional differences between species in a community. These functional diversity (FD) indices have received much recent attention and refinement, but their greatest limitation remains their inability to incorporate information about intraspecific trait variation (ITV).
We use an individual‐based model to account for ITV when calculating the functional diversity of two avian communities in Borneo; one in primary (unlogged) forest and one in selectively logged forest. We deal with the scarcity of trait data for individual species by developing a simulation approach, taking data from the literature where necessary. Using a bootstrapping procedure, we produce a range of ecologically feasible FD values taking account of ITV for five commonly used FD indices, and we quantify the confidence that can be placed in these values, using a newly developed bootstrapping method: btFD.
We find that incorporating ITV significantly alters the FD values of all indices used in our models. The rank order of FD for the two communities, indicating whether diversity was higher in primary or selectively logged forest, is largely unchanged by the inclusion of ITV. However, by accounting for ITV, we are able to reveal previously unrecognized impacts of selective logging on avian functional diversity through a narrower dispersion of individuals in functional trait space in logged forest.
Our results highlight the importance of incorporating ITV into measures of functional diversity, whilst our simulation approach addresses the frequently encountered difficulty of working with sparse trait data and quantifies the confidence that should be placed in such findings.