Data Recovery at 9GE2084 the 19 Rock Piles Site, Greene County, Georgia

Site 9GE2084 is a somewhat large (~186 m by 80 m) concentration of rock piles in Greene County, Georgia, approximately 14 kilometers south-south-of-west from the town of Greensboro. The site is located on top of a ridge toe knoll (USGS Liberty Topographic Map, UTM 17: 294013E and 3703557N (NAD 27)) and extends over the toe's northwestern slope down to within 10 meters of a tributary creek of the former Richland Creek. The site is almost 30 meters above Richland Creek's original channel, which is roughly 550 meters to the northeast. Lake Oconee's eastern arm currently inundates the Richland Creek valley up to the 350 feet contour (Figure 1.1). During a Phase I archaeological survey of the area, Ledbetter (2004) identified 19 distinct rock piles (5 of which were labeled A-E) at Site 9GE2084 (Figure 1.2). Based on a single horseshoe recovered from one of the four shovel test pits and acknowledging the prevalence of historic period rock piles in the eastern United States (Gresham 1990), including the general project area (e.g.,Pluckhahn et al. 1995), Ledbetter initially proposed that the site represents historic period field clearing piles. But rock pile complexes are also known to contain prehistoric piles (e.g., Fish et al. 1978; Loubser and Greiner 2002), so "several insightful review comments" (Ledbetter 2005:5) led Ledbetter to conduct Phase II test excavations looking for evidence of a likely prehistoric presence. An additional 37 shovel test pits were excavated across Site 9GE2084 during Phase II testing (Ledbetter 2005). The shovel test pits were arranged in a 20-meter interval grid across the site. Eight of the shovel test pits yielded artifacts, including 18 quartz flakes, a ceramic sherd, and an iron rod fragment. Also, Ledbetter excavated a trench into the most prominent rock pile on the site, labeled Rock Pile C. The trench, measuring four meters long by one meter wide, yielded 20 quartz flakes, 39 ceramic sherds (which included pieces from a Lamar faintly complicated stamped bowl), and 74 fragments of human bone. Three additional test units, each measuring one by one-meter, were excavated within three of the other stone piles (i.e., Rock Piles A, E, and F) at the site. The test unit within Rock Pile A produced six quartz flakes and the test unit within Rock Pile E produced nine quartz flakes and two ceramic sherds. In spite of an average artifact density of 5.7 artifacts per square meter, the test excavations exposed no postholes, pits, middenlike layers, or hearth-like lenses to suggest that the site might have been a camp or village. The recovery of quartz flakes and one unifacial tool suggested to Ledbetter (2005:30-33) that Site 9GE2084 might have been an ephemeral activity locale, possibly dating back to the Late Archaic period. The Lamar complicated stamped sherds, mostly from a single bowl associated with human bone remains from below Rock Pile C, hinted at mortuary activities during Late Mississippian times. Ledbetter furthermore suggested that at least some of the quartz debitage could date to the Late Mississippian too. The horseshoe and iron rod fragment moreover indicated a historic period presence, probably in the form of field clearing and building stockpiling activities. Sandy loams, ranging in thickness from 10 to 20 centimeters at the site (Ledbetter 2005:6), suggest loss of topsoil due to intensive agriculture during the late nineteenth and early twentieth centuries. Logging later in the twentieth century disturbed much of the site as well. Considering evidence for multi-component use and disturbance of the ridge toe knoll, the question remains how the thin scatter of prehistoric artifacts at Site 9GE2084 stratigraphically relates to the majority of rock piles. Whereas Rock Pile C was clearly prehistoric and Rock Pile E was a good prehistoric candidate, Ledbetter could not determine if Rock Pile A was prehistoric or historic. Moreover, according to Ledbetter's positive shovel test pits, three concentrations of artifacts seem to be present at Site 9GE2084; a southeastern one centered on Rock Pile C, a central one centered on Rock Pile B, and another northern one centered on Rock Pile E (Figure 1.2). Because of the human remains beneath Rock Pile C and the likelihood of additional human remains beneath some of the other un-excavated rock piles on Site 9GE2084, Ledbetter recommended the site eligible for nomination to the National Register of Historic Places. The US Army Corps of Engineers (USACE), Savannah District, concurred with this recommendation, as well as the Georgia Department of Natural Resources, Historic Preservation Division (HPD) in 2005. Site 9GE2084 was recommended for preservation in place, and set aside so as not to be impacted by development. More recently, however, it was determined that residential development was preferred and that additional excavations at the site should be employed to confirm the age of the remaining rock piles. It was recommended that any impact by development was to be mitigated through data recovery excavations, and to relocate any human remains in accordance with the "General Plan for Preserving and Protecting Human Burials at Reynolds Plantation" dated May 14, 1999.