Geoarchaeology: An International Journal, Vol. 19, No. 5, 441–470 (2004)䉷 2004 Wiley Periodicals, Inc.Published online in Wiley Interscience (www.interscience.wiley.com). DOI:10.1002/gea.20005shortstandardGEA(Wiley) RIGHT BATCHThe Effects of Burrowing Activity onArchaeological Sites: Ndondondwane,South AfricaKent D. Fowler,1,* Haskel J. Greenfield,2and Leonard O.van Schalkwyk31Department of Archaeology, University of Calgary, Calgary, Alberta T2N 1N4Canada2Department of Anthropology, University of Manitoba, Fletcher Argue 435,Winnipeg, Manitoba R3T 5V5 Canada3eThembeni Cultural Heritage, Pietermaritzburg, KwaZulu-Natal, South AfricaBurrowing activity is a widely recognized source of site modification. Most taphonomic stud-ies of burrowers emphasize their destructive aspects on the archaeological record. Excava-tions at Ndondondwane, South Africa, suggest burrowing activity is destructive in some ways,but may also preserve cultural behavior. Drawing on both direct and indirect sources ofevidence, we discuss how burrowing activity by rodents, earthworms, and termites can informabout pedogenic and depositional processes at archaeological sites and both preserve anddestroy evidence of intra-settlement patterns and early African cultigens. Specifically, wedemonstrate the limited effect of earthworms on site stratigraphy, how the localized activityof termites have preserved casts of early African cultigens, and how the ability of archaeol-ogists to distinguish the devastating effects of rodent burrowing from remains of architecturalfeatures have permitted important inferences about social and ritual life in early Africanfarming communities. 䉷 2004 Wiley Periodicals, Inc.INTRODUCTIONIn the past 20 years, it has become widely appreciated that an understanding ofthe nature and extent of natural and cultural taphonomic postburial processes area prerequisite for interpreting past cultural behavior at archaeological sites. It isno longer assumed that there are direct links between the patterning of artefactand feature distributions on sites and human behavior. Many physical and biolog-ical processes that affect the movement and positioning of artefacts, ecofacts, andfeatures within the three-dimensional context of a site, only some of which are wellunderstood (Wood and Johnson, 1978; Butzer, 1982; Nash and Petraglia, 1987; Schif-fer, 1987; Lyman, 1994; Canti, 2003). Most often, studies of postburial processesare concerned with the destruction or distortion of archaeological data. In reality,postburial processes not only destroy and distort, but also can modify and preserveevidence of past cultural behavior.* Please direct all correspondence to this author.FOWLER, GREENFIELD, AND VAN SCHALKWYK442 VOL. 19, NO. 5shortstandardGEA(Wiley) LEFT BATCHtop of RHbase of RHtop of textbase of textOne of the most widely recognized sources of site modification is burrowingactivity. Almost all archaeological sites have some evidence of burrowing activity,and it is commonly treated as a destructive force on sites (for a recent review ofbioturbation, see Balek [2002]). Burrowers either spend most of their life under-ground (a fossorial lifestyle), or they may live above ground using burrows forshelter, nesting, hibernation, or birthing (a semifossorial lifestyle) (Schiffer, 1987;Wilkins, 1989). Alone or in combination, both lifestyles can damage the integrity ofa site’s stratigraphy and the distribution of cultural debris. However, the preciseeffects of different burrowing animal species on archaeological sites are poorlydocumented. As a result, discussions of the relationship between material culturalpatterning on sites cannot proceed without considering both the positive and neg-ative effects that burrowing animals have on the distribution and preservation ofcultural remains.Archaeologists are aware of the destructive effects caused burrowing vertebratemammals, amphibians, birds, and reptiles (Wood and Johnson, 1978; Waters, 1992).During excavation, however, most archaeologists only record the highly visibledisturbances caused by burrowing vertebrates (e.g., rodent tunnels). In contrast,burrowing invertebrates (insects, earthworms, and crustaceans) have receivedcomparably less attention. This situation cannot always be attributed to archaeo-logical negligence in terms of research design, excavation strategy, or inattentionto bioturbation processes in general. Instead, it is a function of the ability of ar-chaeologists to identify the presence of burrowing invertebrates acting on sites.In this article, we take a somewhat different approach to the issue of postburialdisturbance processes. Most taphonomic studies of burrowing activity emphasizethe destructive aspects on the archaeological record (including the works citedabove), while few document how burrowing animals may preserve the originalhorizontal and vertical relationships of cultural materials (Darwin, 1896; Johnson,1989; Michie, 1990; Kries, 1995; Van Nest, 1997). Instead, we will show how bur-rowing activity is destructive in some ways, but may also preserve cultural behav-ior. To illustrate this point, data from the Early Iron Age site of Ndondondwane inSouth Africa will be presented.THE SITERegional ContextEarly Iron Age (EIA; A.D. 420 – 1050) farmers in southeastern South Africa typi-cally established small, permanent villages on the rich alluvial soils beside lakesand rivers (Maggs, 1980a). Most EIA sites were occupied for a relatively long du-ration of time, often several hundred years (Maggs, 1984b, 1989). These reoccu-pations of the same places have created a palimpsest of flat, expansive settlements.As a result, most of the archaeological research on EIA sites in the region hasfocused on cultural disturbance processes as more immediate obstacles in datingarchaeological sites, elucidating pan-site stratigraphic relationships, and determin-ing the relationships between coterminous features and activities at settlements.BURROWING ACTIVITY AT NDONDONDWANE, SOUTH AFRICAGEOARCHAEOLOGY: AN INTERNATIONAL JOURNAL 443shortstandardGEA(Wiley) RIGHT BATCHtop of RHbase of RHtop of textbase of textTable I. Early Iron Age chronology of the eastern lowlands in South Africa. Uncalibrated ages aresummarized from Fowler (2002: Table 5.1).aCeramic Phase Uncalibrated Age Range (B.P.) Calibrated Age Range (A.D.) Mzonjani 1670 ⫾ 40– 1540 ⫾ 60 420–590Msuluzi 1460 ⫾ 50– 1275 ⫾ 60 640–780Ndondonwane 1300 ⫾ 50– 1190 ⫾ 50