South-Central Section - 47th Annual Meeting (4-5 April 2013)

Paper No. 32-7
Presentation Time: 3:50 PM

MORPHOMETRIC ANALYSIS OF SPATIOTEMPORAL VARIATION SUPPORTS REINTERPRETATION OF THE FOSSIL RECORD OF THE EASTERN BOX TURTLE (TERRAPENE CAROLINA)


VITEK, Natasha S., Jackson School of Geosciences, The University of Texas at Austin, 1 University Station C1100, Austin, TX 78712, nsvitek@utexas.edu

Poorly understood variation can obscure patterns of evolution, especially in highly variable species. The Eastern Box Turtles of the Terrapene carolina complex present an excellent opportunity to explore spatiotemporal patterns of variation because the extant species has high levels of variation in both soft-tissue and skeletal characters. It also has a published fossil record that includes shells that are suitably complete and abundant enough to be used in statistical analyses of shape. Variation within T. carolina is traditionally explained by the presence of subspecies in both the modern and fossil record. To explore patterns of spatiotemporal variation, I used geometric morphometrics to quantify the shape of 200 Recent and 44 fossilized shells of T. carolina. First, I analyzed differences in the shapes of nominative subspecies in the modern biota, both qualitatively and quantitatively. Then, I compared the results from the modern biota to that of the fossilized specimens. Although the results of pairwise comparisons indicated significant differences between extant subspecies, qualitative visual comparisons indicate that those differences may not be biologically significant. The results of assignments tests and canonical variates analyses indicated insignificant or unreliable differences. Statistical analyses could not identify individuals of a subspecies based only on skeletal features. Based on the results of these analyses, I conclude that subspecies cannot be identified in the fossil record, contrary to previous explanations. Differences between the shapes of fossilized and modern shells were greater than the differences between various nominative modern subspecies, further indicating that applying subspecific identifications to fossil specimens is inappropriate. Fossilized specimens are, on average, larger than recent specimens. In addition, fossilized shells are generally taller and have straighter marginals than Recent shells. Furthermore, fossils from two Late Pleistocene localities in Florida have patterns of size and shape variation that are unique in comparison both to other fossils and to recent specimens of T. carolina. Now that these patterns of variation are not masked by hypothesized evolution from one modern subspecies to another, better explanations can be explored.