The Licking Bison Site (39HN570):
An Early Archaic Bison Kill in Northwest South Dakota

The 2000 field season drew to a close the current phase of investigations at the Licking Bison site (39HN570), an Early Archaic Bison kill, located in Harding County, South Dakota (Figure 1). The land is owned by Gary Licking of Buffalo, South Dakota. During the 1994-2000 field seasons volunteers from across the plains consisting of professional, avocational, and amateur archaeologists helped to excavate the site. The goal of the program was to preserve by excavation, that portion of the site that would be lost to erosion in the upcoming years.

Figure 1

The investigations began during a period of time when the State Historical Society, Archaeological Research Center (SARC) was conducting some exploratory investigations in the northwest portion of the state. SDAS member Joe Nonnast showed the author a couple of points that he had collected from a location in Harding County. The points appeared to be Early Archaic side-notched (roughly 7000-5000 years before present) variety or possibly Late Prehistoric/Late Archaic Besant side/corner-notched points (2000-1500 years before present). Both styles are very similar in size and shape. While agreement was not necessarily reached as to whether the points were from the Early Archaic period or Late Prehistoric period, we decided to investigate the site more fully with a field visit.

Upon arriving at the site, we found a light scatter of eroded bone and tooth enamel on a hard packed sandy surface with small amounts of bone extending from an eroded terrace scarp. A little trowelling exposed bison bone in a very good state of preservation in the terrace. The following year we returned with SDAS members to begin excavations at the site. We quickly found that the site was the remains of a bison kill site buried in an alluvial sand deposit. Based upon the geomorphic setting, we assumed that the site was a Late Prehistoric Besant Complex kill (Albanese 1997, Fosha 1995, 1996). Later with radiocarbon dates we found this to be in error and that the site does indeed date to the Early Archaic Period.

The site is situated in a terrace of Graves Creek, an ephemeral stream that empties into the South Fork of the Grand River, at a point located 5.3 km southeast of the site area. In the Licking Bison site area, the floodplain of Graves Creek is 122 m wide. It is bordered by a 2-3 m high scarp that forms the riser (step) of the T1 Holocene terrace that borders the floodplain (Figure 2). The Early Archaic bison bone bed, that constitutes the bulk of the prehistoric site lies on the southwest edge of the modern floodplain of Graves Creek, adjacent to and buried by approximately 1.5 m of the T1 terrace (Albanese 1997:2). Additional cultural activity can be observed on the opposite side of the stream in the form of bison bone and hearths which may be activity areas related to the kill event. However temporal data is lacking at these locations at the present time. What may be a separate site is a number of bison elements eroding from what clearly appears to be older terrace fill approximately 40 meters upstream from the site.

Figure 2

Excavations began in the spring of 1995 and continued into 2000. The excavations were conducted intermittently on weekends and 3-4 day visits which concluded with the excavation of 26 units (Figure 3). The 2000 field season marked the end of these investigations until the bone analysis is completed and funding can be arranged to properly continue excavations.

Figure 3

Though we have not excavated a large portion of the site, we have gleaned data that indicates butchering practices and inferential data as to how these animals may have been procured. Currently, our model is based upon the excavation of the above mentioned units. In the southwest corner of the excavations, we have uncovered complete, or nearly complete articulated animals (one on top of the other) suggesting that these individuals were never processed (Figure 4). Moving east in the grid, large, articulated portions of the animals were located. These included largely halved animals such as rear or frontal sections with intact vertebral columns. To the north of this area, bison remains consist of articulated quarters, and some individual elements (Figure 5). North of this area are smaller sections of articulated bison and many individual elements (Figure 6). Also in this area, a small unprepared surface hearth with small burned sections of bison bone and a light charcoal scatter was investigated.

Figure 4 Figure 5 Figure 6

A geologic feature is located in the southeast corner of the excavations. This consists of a small gully buried by Holocene fill that was present at the time of the kill event. The gully is located at the base of a high Cretaceous-aged terrace, approximately one meter in width, and 60 cm in depth (Figure 7). This is also the location where the fully articulated bison are located. This gully appears to be of natural origin and may represent the stream channel at the time of the kill and was possibly used to aid in trapping or impeding the bison.

Figure 7

Bone from the site awaits a detailed study. The bone is generally in excellent condition due in part to what appears to be an immediate burrial in sand. One exception is the uppermost bone in the area near the relict gully which had been buried with an increase in clay content. Much of this bone was to friable to effectively remove intact. Seasonality has not yet been established for the site, nor has the identification of species. Bison from other bison kill sites of this period such as the Simonsen site in Iowa (Agogino and Frankforter 1960) and Hawken site from the western edge of the Black Hills in Wyoming (Frison 1991) with the same type of point style are interpeted as B. bison occidentalis. The range of this extinct subspecies during this time period suggests that the Licking Bison site Bison be of a similar nature.

The Licking Bison points (Figure 8) are side-notched lanceolate forms with straight to concave base. The majority of the specimens have concave bases. Breakage and modification makes some of the examples appear to have straight bases while others suggest that they were intentionally made with a nearly straight base. Bases and notches were generally ground and complete points exhibit extremely sharp tips. This is consistent with the points from the Hawken site. Frison (1991:191) makes the reasonable suggestion that grinding was conducted to withstand thrusting pressure without splitting the shaft or cutting a sinew binding and that the tips were made "needle sharp" to penetrate the hide. As with the Hawken site, one of the points from Licking Bison was recovered from the vertebral column that had cut through two vertebrate before severing the spinal cord supporting the penetrating advantage of extra-sharp tips.

Figure 8

The Licking Bison points are made of relatively local materials. Siliceous wood is the most common material type with local acquisition possible, with lesser amounts of porcelinite from northwest South Dakota-southeast Montana. One projectile point is made of a local Bijou Hills like quartzite. One projectile point is made from Minnelusa or Paha sapa chert from the Northern Black Hills and represents the most distant of the lithic materials recovered to date.

The Early Archaic side-notched points are seen in the northern plains as the beginning of the Early Archaic Tradition. One obvious change from the preceding lanceolate points of the Paleoindian period are the addition of side notches. A greater use of local materials has also been suggested for this period. While it is certainly evident in this, and other assemblages dating to this period in the Northern Plains (e.g. Frison 1983; Harrell and McKern 1985; Reher and Frison 1980:134), locally it is a continuation of a pattern beginning with Foot Hills Mountain Tradition hunters and gatherers in the Black Hills (Sellet and Fosha 2001).

The points go by many regional names such as Blackwater and Pahaska Side-Notched (Husted n.d.) and excavated sites such as Hawken (Frison 1991), Simonsen (Agogino and Frankforter 1960) and Logan Creek (Kivett 1962). In well stratified sites such as Mummy Cave in Wyoming (Husted n.d.), the points change over time during the Early Archaic period. Also of importance from sites such as Mummy Cave, there is no progression of change from the Late Paleoindian points to the side-notched lanceolate forms. This prompted Husted (1969:88) to suggest that these forms are from eastern Early Archaic populations moving onto the Northern Plains into the foothills and mountains of the Rocky and Bighorn, a view that recent researchers are now addressing (Kornfeld et al. 2000; Kornfeld and Frison 2000:147).

One AMS sample (funded by the Northern Hills Chapter, SDAS) was submitted on charcoal from the bone bed. The results were approximately 50,000 years to early and it was suggested that an unassociated section of charcoal had been introduced into the bone bed. A second sample (funded by the author) was run on bone (LE-5478). The two-sigma calibrated results at the 95 percent confidence limit expressed an age of cal B.C. 4780-4460 or 6730-6410 years before present (B.P.). The intercept of radiocarbon age with calibration curve is cal B.C. 4600 or 6560 B.P. This is very close to the Hawken I site date of 6470 B.P. (Frison et al. 1974), the approximate 6000 B.P. date for Hawken III as well as the 6633 B.P. date from Logan Creek. Comparing the points to those that were recovered from Mummy Cave, Wil Husted found the closest similarities in Layers 17 and 18 and estimated their age at around 6500-6000 B.P. but suggested the older end of this scale might be the closest (Wil Husted personal communication 2002).

While the current on-site investigations have concluded, there is still much to do on the analysis of the bone from Licking Bison.

Figure 9

Portions of this have been completed but the bulk of the bone analysis remains. Once this is completed, it is hoped that with money secured through grants, excavations may continue.

The site has great potential in addressing many of the research questions concerning the Early Archaic. These relate to such topics as the cause and effect of the Altithermal on indigenous populations, mobility and range of period hunter gathering groups, content models for Early Archaic bison kills, local subsistence procurement systems and economy, and gain a better understanding of the Paleoindian-Early Archaic continuum (or lack there-of) to name just a few. The last five years of the project have been a great adventure. Those who participated gained an understanding not just of the site but of each other. Late nights around the campfire, the brilliance of the Harding County night sky, the floods and explosions made it was an experience I am certainly glad I was a part of and I thank everyone who shared in the project.