In lieu of an abstract, here is a brief excerpt of the content:

Chapter 3 Prehistoric Settlement Analyses and Predictive Modeling in West-Central Louisiana Over the past 25 years considerable effort has been directed to the development of predictive models of historic and prehistoric site location in the west-central Louisiana area and, coupled with this, to developing an understanding of the nature of past settlement systems. Predictive modeling efforts have taken place primarily in two areas: on Fort Polk proper (e.g., Anderson et al. 1988, 1999b; Campbell and Weed 1986; Servello, ed. 1983; Thomas et al. 1982) and on the Kisatchie National Forest, whose districts largely surround Fort Polk (e.g., Hillman 1980; Johnson 1984a, 1984b; Johnson et al. 1986; Phillips and Willingham 1990; Willingham and Phillips 1987). These analyses have typically employed large datasets, are replicable, and have generated increasingly reliable results. Settlement analyses that have been conducted have explored questions about group technological organization and mobility, site use and site type, lithic raw material procurement and use, and how major sections of the landscape in this part of the region were used over time; these questions have been examined by a great many authors, as detailed in the pages that follow. While typically focusing on sites of all periods, the general models of site location that have been developed on Fort Polk and in the Kisatchie National Forest tend to be most useful for delimiting where prehistoric sites are located on the landscape, since these make up the overwhelming majority of the sites and isolated finds recorded to date on the installation and in the surrounding region. Predictive modeling activity directed explicitly to historic sites has occurred, however, and a review of this work is presented in Chapter 8. This chapter emphasizes research directed to prehistoric occupations in the area. Predictive models, it should be stressed, are important in both research and resource management. Having accurate information on where sites occur helps us to explore why past peoples made use of the landscape the way they did and allows cultural resource managers to plan projects to minimize impacts to important sites. Accurate predictive models also facilitate the development of appropriate field methods for the detection Prehistoric Settlement Analyses and Predictive Modeling 115 and documentation of sites.Apredictive model of site location was presented in the 1988 Historic Preservation Plan (HPP) for Fort Polk and was improved and updated in 1995 on the basis of the work described in this chapter; the Fort Polk HPP modeling efforts are detailed in Chapter 4. The Fort Polk Archaeological Survey Slope Element/Landform Predictive Model One objective of the Fort PolkArchaeological Survey (FPAS) undertaken in the late 1970s by the University of Southwestern Louisiana was to develop a testable model of site occurrence and density for the Peason Ridge area. Twenty-seven 160-acre quadrats (4,320 acres) from the Peason Ridge landscape were selected using a simple random sampling procedure (Servello and Morgan 1983:194–195) (see Figure 2.1). Four of the quadrats fell within the Peason RidgeArtillery Impact Zone and were not investigated for safety reasons. The area examined, 23 quadrats comprising 3,680 acres, represented approximately 13.2 percent of Peason Ridge. Portions of three major drainages , the Red, the Sabine, and the Calcasieu, lay within the survey area, and roughly 10 quarter section quadrats each (1,600 acres) were surveyed in the Sabine and Red River basins while three quarter section quadrats (480 acres) were examined in the Calcasieu River basin.Atotal of 115 sites and 51 isolated finds were found in these 23 sample quadrats (Servello and Morgan 1983:198) (Table 3.1). A series of environmental variables were recorded for each site and isolated find encountered, encompassing elevation, drainage, and landform characteristics (Servello and Morgan 1983:182–203; Servello 1983). While the first two variables were comparatively straightforward and entailed coding elevation and stream rank data (after Strahler 1957, 1964) from U.S. Geological Survey (USGS) quadrangles, the third was much more complex, encompassing larger subregional strata based on underlying geology and more specific landform types based on Ruhe’s (1975; Ruhe and Walker 1968) slope element classification system. The Kisatchie Wold, the major landform encompassing Fort Polk, was subdivided into five structural and erosional surfaces based on surficial and underlying geology (Servello 1983). These five strata from north to south were the Horse’s Head structural surface, the Peason Ridge structural surface, the Castor erosional surface, the Williana structural surface, and the Bentley erosional surface (Servello 1983:206). Site distributions within these strata were...

Share