Paper No. 63-22
Presentation Time: 9:00 AM-5:30 PM
PREDICTIVE SITE MODEL FOR PRE-CLOVIS BURIES MATERIALS IN CENTRAL TEXAS
In the mid-20th century, Clovis (~12,800 – 13,100 yr BP) was identified as the oldest archaeological horizon in North America (Waters et al., 2011). In the decades to follow, evidence for a pre-Clovis (~13,000 – 16,000 yr BP) component surfaced (Goebel et al., 2008). The discovery of pre-Clovis at the Debra L. Friedkin site near Buttermilk Creek in central Texas strongly indicates that there is a pre-Clovis component in this region (Waters et al., 2011). To help find more pre-Clovis sites in central Texas, a predictive buried site potential model was created using attributes, such as age and landform type, described for the Debra L. Friedkin site using a geographic information system (ArcGIS). Because of the way in which GIS integrates layers of information, it is one of the best methods to accurately determine buried site potential (Mehrer and Wescott, 2006). GIS is integrated where archaeology incorporates the hard sciences to spatially identify the stratigraphic units defined by geomorphologists, soil scientists, and sedimentologists (Green, 1990). Predictive modeling is made up of coincidences in the present that must be explained in context of the past (Ebert, 2000). This model has been derived from niche and habitat suitability models employed by ecologists, and was first applied to human habitation by Hudson (1969). Buried site potential includes a combination of depositional environment, landform stability, and landform age. The basis for site prediction is the theory that non-cultural aspects of the environment will correlate with site locations (Savage, 1990). Landform age and stability were determined with the stratigraphy records from other sites in the area. Soil maps provided information about the age of the surfaces as well as their geomorphic landform, erodibility, drainage, and slope. LiDAR provided fine scale topography. We were able to make recommendations for excavation based on landform attributes. The next stage of this project will involve ground truthing using continuous soil cores extracted using the Geoprobe MC5 direct push method along transects perpendicular to the Buttermilk Creek. Results will determine the accuracy of the model and be used to improve the one presented here.