2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 194-6
Presentation Time: 9:15 AM


SMITH, Jansen A., Earth and Atmospheric Sciences, Cornell University, Ithaca, NY 14850, GOODWIN, David H., Department of Geosciences, Denison University, 100 Sunset Hill Drive, Granville, OH 43023 and DIETL, Gregory P., Paleontological Research Institution, 1259 Trumansburg Road, Ithaca, NY 14850, jas933@cornell.edu

Species abundance distributions (SADs) transform complex community data into an easily understood format. This simplification makes comparisons of structure between and within communities readily attainable. SADs can also be deconstructed into rare and common components, allowing for elucidation of the processes driving community structure. Using this deconstruction approach, we tested the hypothesis that distributions of rare species tend to be structured by neutral mechanisms (i.e., stochastic dispersal), whereas common species are governed by niche-based dynamics (i.e., ecological interactions).

Molluscan assemblages in the Colorado River delta were bulk sampled at three localities along the delta’s former salinity gradient. Samples were sieved with a 5-mm mesh and all larger individuals were identified to the species level and counted (n>40,000). Abundance data were used to create SADs and the best-fit model (e.g., log-normal, log-series) was determined using AIC, BIC, and maximum likelihood in the statistical program R. Species were classified as rare or common based on the percentage of samples in which they occurred at each locality (>50% = common; <50% = rare). SADs were created for individual rare and common components and the best-fit models were determined, as above. Rare and common species were categorized as marine or estuarine species and correlation of their occurrence along the salinity gradient was assessed using a chi-square test.

Molluscan species abundance distributions along the past salinity gradient in the Colorado River delta showed overlapping bimodal distributions of rare and common species. Assigning species to habitat groups revealed core species were predictably distributed along the gradient according to habitat preference. Rare species were, however, commonly found outside of their preferred habitat. All three rare SADs best-fit a log-series model, suggesting neutral processes. On the contrary, common SADs best-fit log-normal models, suggesting niche-based processes. These preliminary data suggest the core species in the molluscan community are distributed based on ecological preference while rare, or transient species, are more randomly distributed. We suggest neutral and niche-based processes function complimentarily to produce community structure.