Disentangling the importance of body-size, shape and phylogenetic variation in community assembly: The case of Neotropical treefrogs communities

Background/Question/Methods

Current approaches in community ecology have been considering phylogenetic history and trait variation among species to infer processes driving community assembly. There is evidence showing that harsh environments select species that either similar or closely related species whereas stable environments select distinct or distantly related species to compose local communities. One current limitation of recent studies is that they either concentrate on trait variation or phylogenetic structure, despite the fact they are often uncorrelated (i.e., trait without phylogenetic signal). To test if phylogeny and traits reveal consistent patterns in community assembly, we quantified the phylogenetic and trait structure of South American treefrog assemblages and related their structure to multiple environmental variables. We used distributions range maps of 187 species and divided the study area into 602 grids to delimit assemblages. Trait data was based on body size and shape represented by multiple morphometric variables that were size adjusted. We assessed whether treefrog assemblages show non-random phylogenetic and trait structure using the standardized effect size (SES) of mean phylogenetic and trait distances within communities. Additionally, we evaluated the relative contribution of spatial and environmental variables in explaining phylogenetic structure and trait dispersion using a variation partitioning in a multiscale framework. 

Results/Conclusions

We found that the majority of assemblages were clustered: 87% of the assemblages showed negative phylogenetic SES and 71% presented negative shape SES. Regarding body size, 98% of the assemblages showed positive SES, indicating size evenness. Phylogenetic and shape clustering were positively associated with humidity levels and negatively with temperature. Conversely, size evenness was negatively associated with variables related to total precipitation. Variation partitioning showed that spatial variables were more important in explaining large scale distribution and environmental variables were more important at medium to fine scales. In our study we found a different pattern from previous one involving phylogenetic or trait based frameworks, since phylogenetic and body shape clustering was related to less harsh environmental conditions whereas size evenness was related to somewhat harsher conditions. Additionally we showed that body size and shape variation show opposite contributions to community assembly in relation to phylogenetic variation.  However, the relative importance of spatial and environmental variables in explaining this opposed patterns were similar, indicating that body and shape variation are similarly important to community assembly but in opposite directions.