The biogeography of community assembly: Latitude and predation pressure drive community structure in a guild of marine epifaunal crustaceans

Collin P Gross, Evolution and Ecology, University of California, Davis, Davis, CA, John J. Stachowicz, Department of Evolution and Ecology, University of California, Davis, Davis, CA, J. Emmett Duffy, Tennenbaum Marine Observatory Network, Smithsonian Institution, Washington, DC and Kevin Hovel, Coastal and Marine Institute, San Diego State University, San Diego, CA

Background/Question/Methods
While the last decade or so has seen a multitude of studies examining the patterns and processes of community assembly as seen in the distribution of functional traits and/or phylogenetic relationships in the community, these have typically taken place within single regions. We aimed to examine how these processes operate to produce global-scale patterns in communities at multiple spatial scales, while accounting for separate regional biogeographic histories. Using communities of peracarid crustaceans (amphipods, isopods, and tanaids) associated with eelgrass (Zostera marina L.), we modeled trait dispersion in eelgrass beds across the northern hemisphere as a function of abiotic and biotic environmental variables including predation pressure, mean annual temperature, and food (epiphyte) availability, as well as spatial variables including latitude, ocean basin, and coastline. We compared separate models for diet and microhabitat traits, and separate sets of models for plot- and site-level species pools.
Results/Conclusions
Trait dispersion strongly increased with predation and decreased with latitude at the site level. Ocean basin and food availability (epiphyte load) appeared as secondary predictors; Pacific communities were more overdispersed while Atlantic communities were more clustered, and increasing epiphyte load was associated with increased clustering in Atlantic microhabitat traits. At the plot level (within sites), bottom-up effects of microhabitat and food availability appeared to dominate, with little to no predation effect. By examining both how species interactions and environmental drivers vary within a single habitat type across a broad geographic gradient, we demonstrate an important role for latitudinal variation in species interactions in driving patterns of community assembly. Though far from definitive, we point to the potential of integrating large scale observational and experimental datasets to predict community patterns at biogeographic scales.