COS 39-5
Partitioning the effects of spatial and environmental variation on phylogenetic structure of a bat metacommunity in a human-modified landscape

Tuesday, August 6, 2013: 2:50 PM
L100C, Minneapolis Convention Center
Laura M. Cisneros, Ecology and Evolutionary Biology, University of Connecticut, Storrs, CT
Michael R. Willig, Center for Environmental Sciences and Engineering, University of Connecticut, Storrs, CT
Background/Question/Methods

The past decade has witnessed a revolution in biodiversity science in that functional and phylogenetic dimensions, rather than simply the taxonomic dimension, have increasingly been used to understand ecological and biogeographic processes governing species assembly at various spatial scales. For the phylogenetic dimension, two critical questions remain unanswered. The first concerns the extent to which variation in phylogenetic structure arises from differences in phylogenetic composition (central tendencies) versus differences in phylogenetic diversity (dispersion). The second concerns the role of spatial versus environmental variation in molding these patterns. As such, we evaluated phylogenetic composition and phylogenetic diversity of a bat metacommunity within a human-modified landscape in Costa Rica. Via variation partitioning, we estimated the contributions of space-free environmental filtering, spatially structured environmental filtering, and spatial influences of unknown origin on phylogenetic metacommunity structure. Spatial characteristics were based on site coordinates. Eight environmental characteristics estimated compositional (metrics sensitive to proportions of land-cover) and configurational (metrics sensitive to spatial arrangement of land-cover) attributes of the landscape and were quantified at each of three spatial scales (1, 3, 5 km radius circles). Because seasonal resource variation may affect the importance of predictors, we conducted analyses separately for each combination of landscape scale and season. 

Results/Conclusions

Phylogenetic diversity, rather than phylogenetic composition, accounted for most of the phylogenetic variation at the metacommunity level (i.e., variance ratio of diversity = 0.95; variance ratio of composition = 0.05). The mean phylogenetic characteristics of communities did not differ greatly among sites, whereas the phylogenetic diversity did differ greatly among sites within the metacommunity. Regardless of season or scale, contributions of landscape and spatial predictors to variation in phylogenetic diversity were similar. Spatially structured environmental characteristics accounted for ~ 98% of the variation in phylogenetic diversity. When considering only pure spatial and only pure environmental characteristics, each accounted for less than 0.3% and less than 1.9% of the variation in phylogenetic diversity, respectively. In northeastern Costa Rica, space and landscape characteristics are highly correlated due to the non-random implementation of anthropogenic land modification. Given the considerable vagility of bats and the intermediate spatial scale of this study, we suspect that the influence of most of the spatially structured environmental variation operates via influences of compositional and configurational aspects of the landscape.