GEOGRAPHIC AND TEMPORAL COMPARISONS OF BIVALVE DIVERSITY IN THE PLIO-PLEISTOCENE OF VIRGINIA AND THE CAROLINAS

Diversity of bivalves from the late Pliocene of South Carolina (Duplin Fm) and Virginia (Yorktown Fm) was analyzed by an undergraduate Invertebrate Paleontology class at Univ. North Carolina Wilmington. We bulk sampled the Tearcoat Branch (Sumter Co., SC) and Kirby Pond (Timmonsville, SC) sites and studied samples from the Morgarts Beach and Moore House members of the Yorktown (Fort Boykin and Deep Creek, VA). Samples were sieved through a 5 mm mesh and bivalves with beaks were picked, sorted and identified to genus. Sample sizes ranged from ~1200 – 1500 specimens. Diversity was analyzed using Holland’s Analytic Rarefaction (http://www.uga.edu/~strata/software/) and PAST (http://folk.uio.no/ohammer/past/). We compared diversity among the South Carolina and Virginia samples and with previously collected data from the Duplin of North Carolina. We expected the southern samples to be more diverse than the northern samples. We also compared our results to those from previous studies of the Pleistocene lower Waccamaw and James City formations of the Carolinas. We hypothesized that the Duplin samples, which predate the Plio-Pleistocene regional extinctions, would be richer than the Pleistocene material.

Tearcoat Branch and Morgarts Beach samples were dominated by Mulinia; dominance was ~0.60 in contrast to 0.26 and 0.22 for Kirby Pond and Moore House samples. As predicted, when rarefied to the same size (1180 specimens), the South Carolina samples were much richer (24 and 28 genera) than the Virginia samples (13 and 16 genera). The disparity in richness between north and south remained when rarefaction omitted the dominant genus. Compared to the North Carolina Duplin, South Carolina samples were significantly less rich than at the Lumber River and Robeson Farm but similar to Natural Well, especially when Mulinia was omitted. We did not find the expected drop in diversity from the Pliocene to Pleistocene. South Carolina Duplin samples were less rich than the Waccamaw samples but richness overlapped (95% confidence intervals) with the James City at Lee Creek when rarefied without the dominant genus. Confidence intervals overlapped for most lower Waccamaw and Duplin sites studied. Yorktown samples were less rich than all others. Lack of richness decline may be related to taphonomy, environment, or taxonomic level of study.