How organisms assemble to form a community is a central question in ecology, with commonly inferred mechanisms including environmental filtering, competitive exclusion, and random processes. Community phylogenetic and functional trait analyses have been used to identify the relative importance of these processes, most often in terrestrial angiosperm systems consisting of a single taxon (e.g., class). Here, community phylogenetic and functional trait analyses are applied to multiphyletic subtidal epibenthic invertebrate assemblages at eight sites in the Gulf of Maine, to investigate the relative importance of deterministic or stochastic forces in structuring these communities. At local sites, some communities were phylogenetically overdispersed, suggesting community assembly via competitive exclusion. Five traits exhibited phylogenetic signal and for these traits, phylogeny can be used as a proxy for ecological similarity. Functional trait diversity was overdispersed for one site and clustered for another, indicative of assembly via competitive exclusion and environmental filtering respectively. Regionally, taxonomic beta diversity was less than expected by chance but phylogenetic and functional trait beta diversity were greater than expected, indicating that species are not dispersal limited and that different environmental conditions among sites select for different species. These results suggest that subtidal epibenthic invertebrate assemblages in the Gulf of Maine are structured by deterministic forces and that different assembly processes operate at different spatial scales: locally, competitive exclusion appears to be important whereas regionally, environmental filtering plays a role.

To conduct a community phylogenetic analysis for subtidal epibenthic invertebrate species at 8 sites in the Gulf of Maine, 3 data sets were used: a phylogenetic tree, a species by site presence-absence matrix, and functional trait information. The site presence-absence matrix was created from data found in Table 1 from Miller and Etter (2011) using the species on vertical surfaces only, and the other two datasets are described below.

Dataset 1—Bayesian newick tree

DNA was extracted and amplified from 50 invertebrate species or downloaded from GenBank and 18S sequences aligned, checked by eye, and trimmed at the beginning and end to minimize gaps. The general time-reversible (GTR) model of evolution was selected, with an estimated proportion of DNA sites invariant (I), and mutation rates among sites following a gamma distribution (G). The GTR+I+G model was then used to estimate a Bayesian gene tree with a ‘Bootstrapping procedure’ (100 times) to assess support for tree nodes.

Dataset 2—Trait data

Functional traits relevant to the study species were selected to cover a range of life history, morphological, and behavioral characteristics relating to space and food acquisition. All traits were divided into discrete categories and species were placed in categories based on their characteristics. For the one continuous trait, categories were assigned based on the species average. Trait information was obtained from a variety of sources including personal knowledge, on-line databases, natural history texts, and primary literature. To summarize, the trait data comprises five nominal traits (coloniality, reproduction, growth form, food capture, and defense), and one continuous trait (body size) divided into nominal categories.

Miller, R. J., and R. J. Etter. 2011. “Rock walls: small–scale diversity hotspots in the subtidal Gulf of Maine.” Marine Ecology Progress Series 425: 153–165.