Validating the use of unique trait combinations for measuring multivariate functional richness

Abstract

Summary Quantifying functional trait diversity has provided important new insights for understanding ecosystem processes and functioning. Functional diversity is often partitioned into three components richness, evenness and divergence. Currently, a convex hull is used as a measure of multivariate richness, but this approach has some serious limitations. Consequently, we propose using the number of unique trait combinations (UTCs) as an approach to measure the filled trait space (the hypervolume containing all possible trait combinations) and propose a new index, sUTC, as the amount of filled trait space divided by the trait space range. Like convex hull approaches, UTC can be partitioned into alpha and beta components when used across sites, and the beta component can be further partitioned into turnover and nestedness components. Unlike convex hull approaches, UTC can be used more intuitively with existing diversity measures as it can be used in conjunction with abundance information. We present the concepts these indices are based on and give examples of their use. The new index, sUTC, is compared to the existing indices on the basis of criteria specified in the literature and one novel criterion. We test and evaluate the approach using simulated data and field data. We found that the UTC approach provided a more accurate assessment of functional richness than did the existing indices. The UTC approach is a multivariate approach to measuring functional richness that can accommodate continuous and categorical traits and can account for holes in the trait space.