Scale dependence in phylogenetic and trait-based community structure of woodpeckers

Background/Question/Methods

Phylogenetic and trait-based analyses of co-occurring species provide complementary information on the processes structuring biological communities.  However, detection of patterns in both types of data is highly dependent on the spatial and taxonomic scale of the study.  I tested for phylogenetic and trait-based community structure at varying spatial and taxonomic scales in woodpeckers (Aves: Picidae) of North America and Europe.  I derived distribution and abundance data from taxon lists and the citizen science project eBird, and focused on regions with sufficient data to conduct tests at multiple scales.  I assessed phylogenetic community structure using a species-level molecular phylogeny, and quantified trait-based community structure using measurements of skeletal characters that correlate with differences in diet and foraging mode.  For both phylogenetic and trait-based analyses, I varied the spatial scale from local (<1 km²) to regional (10,000 km²), and varied the taxonomic scope from species found at the regional scale up to the entire global diversity of woodpeckers.  In addition, I examined the evolution of characters used in the trait-based analyses, in order to clarify how trait evolution may affect agreement between community patterns of phylogenetic relatedness and trait similarity.

Results/Conclusions

I detected non-random patterns in both phylogenetic and trait-based structure, though as predicted, these patterns varied strongly with spatial and taxonomic scale.  Increasing spatial scale results in a shift from phylogenetic and trait-based evenness at local scales to clustering at regional scales.  Increasing taxonomic inclusion similarly results in a shift from phylogenetic and trait-based evenness under minimum taxonomic inclusion, to clustering with maximum taxonomic inclusion.  Analyses of trait evolution show strong phylogenetic signal and a general lack of convergence, such that phylogenetic relatedness is a good proxy for trait similarity; this explains the correspondence between phylogenetic and trait-based results.  These results are generally in line with predictions and previous results.  However, the inclusion of phylogenetic structure, trait-based structure, and trait evolution allowed me to more clearly discriminate interactions among the multiple processes affecting patterns of community structure.