1Department of Anthropology, University of Arkansas, 2Environmental Dynamics Program, University of Arkansas, 3Department of Anthropology, University of Minnesota, 4Department of Anthropology, Lehman College CUNY, 5Graduate Center, CUNY, 6Divison of Paleontology, American Museum of Natural History, 7Department of Palaeoanthropology and Messel Research, Senckenberg Research Institute and Museum of Natural History, 8Department of Sociology and Anthropology, University of Rhode Island
Thursday 1:45-2:00, Broadway III/IV
Miocene deposits on Rusinga Island and at Songhor, Kenya, have yielded some of the most important specimens in the primate fossil record. However, there has been little consensus concerning paleohabitats at these sites. Reconstructions of Rusinga habitats have ranged from semi-arid/open to rainforest settings; and Songhor is generally regarded as “forested” but the specific type(s) of forest are not known. The lack of widespread early Miocene C4 grasses, detectable by δ13C approaches, exacerbates the problem. Here we present dental microwear textures of tragulids from Rusinga Island (n=22; with the majority of specimens coming from localities R1 and R106, which date to between 18-20 Ma) and Songhor (n=14; with specimens dating to between 19.5-20 Ma) as a proxy for paleoenvironmental reconstruction of these early Neogene primate sites. Extant ruminants evince strong associations between microwear and habitat. Open-setting grazers tend to have low texture complexity and high anisotropy, whereas closed-forest browsers have high complexity and low anisotropy. The texture data for the tragulids from both sites fall within the range of extant bovid generalists, though Rusinga specimens have significantly higher complexity than do those from Songhor. These results are in accord with mixed habitat settings at these sites, with Rusinga exhibiting perhaps slightly more closed habitats than Songhor. Texture data distributions do not suggest marked changes in graze-browse ratios in seasonal environments.
This study was funded by grants and support from the US National Science Foundation (0852609), Leakey Foundation, University of Minnesota, NYCEP, and Baylor University.