PS 20-35
Ecology versus evolution in community trait distributions: Woodpeckers (Aves: Picidae) as a test case

Tuesday, August 6, 2013
Exhibit Hall B, Minneapolis Convention Center
Matthew J. Dufort, Ecology, Evolution, and Behavior, University of Minnesota, St Paul, MN
Background/Question/Methods

Ecologists have long been intrigued by the distributions of trait values in community members. One primary question regarding these trait distributions is whether they are driven by ecological processes, such as sorting of species from a regional species pool, or evolutionary processes, such as shifts in traits within species. In addition, these forces, whether ecological or evolutionary, have often been assumed to be repulsive (e.g. limiting similarity), but recent theoretical work indicates that attraction could result in clumps of organisms with similar trait values. I investigated these questions using traditional and geometric morphometric analysis of birds in the avian family Picidae (woodpeckers and allies). Woodpeckers have demonstrated relationships between morphology and ecology, making them ideal for such questions. To evaluate potential community assembly processes, I tested community trait distributions against predictions from several models of assembly. To determine whether any non-random patterns are based on ecological or evolutionary processes, I tested for trait shifts within species away from expectations for the species as a whole. Presence of trait shifts that contribute to non-random local patterns would support local adaptation; absence of such shifts would support species sorting.

Results/Conclusions

Preliminary results indicate that members of woodpecker communities are distributed non-randomly in trait space. I further find that both repulsion and attraction can occur, though repulsion is more frequently observed. This results in community trait distributions that are either more evenly spaced than expected, or made up clumps of similar species. These patterns appear to be driven by trait shifts away from species means, and are therefore at least partially explained by evolution. Future work will test whether such patterns scale up to differences among taxa across the entire evolutionary history of Picidae.