Comparative functional community assembly: Scale outweighs context
Functional community assembly and its manifestation in functional biodiversity (FD) are foundational cornerstones in modern ecological theory because they address the nature of ecological community structure and the influence of structure on ecosystem functioning. Principal challenges in developing a general understanding of both community assembly and FD include the problems of the myriad of metrics, the ever-present issue of scaling, and the delightful quagmire of (for the most part anyway) idiosyncratic biotic species pools. The problem of measurement/metrics can solved technically through comparison and thoroughness. The problem of scaling can be solved technically at least in part, but it remains also a non-technical biotic issue of the degree of clumping and mixing of individuals in space. The problem of idiosyncratic natural history has no simple technical solution. We have undertaken a comparative study in which we sampled five disparate vegetation types (grassland, mesic forest understory, pine barren, wet forest understory, wetland; within a single climatic region) at multiple spatial scales (alpha, within-patch beta, patch alpha, among-patch beta, and gamma) in order quantify standardized effect sizes (SES) of multiple FD metrics for multiple traits (height, leaf size, leaf economics) with multiple species pool null models (local, regional) in order to begin to build a more comparative understanding of functional community assembly.
Effects associated with scaling, either in terms of the alpha, beta, and gamma diversity partition, or in terms of scale of the species pool outweighed effects of habitat, trait, or diversity metric. SES of FD metrics were largely consistent regardless of their sophistication, so we needn’t worry too much about which indices are best. Quadrat alpha FD was usually underdispersed relative to regional pools, while it was only sometimes overdispersed relative to local species pools. Beta FD within patches was consistently underdispersed. Patch-scale alpha FD was more strongly underdispersed than quadrat alpha, while among-patch beta was strongly overdispersed. Together, these results suggest a high degree of deterministic functional assembly across the vegetation types. Alternatively, the results suggest a high degree of mixing at the smallest scales.
The vegetation types were taxonomically distinct with jaccard distances > 0.85, but were functionally less distinct with functional distances (1 – niche overlap) of nearly 0 to 0.6. In addition, the slope of the SAR was remarkably consistent across the vegetation types.