Modeling
Bison Jump Dynamics at Bonfire Shelter
Modeling
Bison Jump Dynamics at Bonfire Shelter
| Results | Conclusions | References | Author Information |
| Addressing the Criteria | |||
| Final Interpretations |
The goal of the project was to determine if the site location of Bonfire Shelter and its associated topographic features are at all unique. Answering these questions may help to explain 1) why it is the only bison jump site known in the area and also 2) why it was used both during the Late Paleoindian and the Late Archaic time periods. Does Bonfire Shelter meet all the criteria necessary for a successful bison jump? And if so, do other localities in the area fail to meet those criteria, making Bonfire Shelter unique within the region? Let's address each part of the premise separately.
1. The presence of ideal grazing conditions or water sources in close proximity to the jump location.
 |
Palaeoenvironmental data for the region are not at a fine enough resolution to identify specifically where grasslands ideal for grazing would have been located. The pollen record does suggest though that there was a higher incidence of grass species in the area during Bone Bed 2 times (Bryant and Holloway 1985). The presence of bison in the region during both Bone Bed 2 and Bone Bed 3 times suggests that conditions amenable to grazing were indeed present, but the exact locations of those areas remain unknown, though it can be tentatively assumed the grasslands were located in the upland portions of the region. |
|
As for permanent water sources, given the proximity of the Rio Grande, Bonfire Shelter is obviously located close to a water source. Therefore it can be assumed that bison would have a spent a good deal of time around this area. They may have frequently traveled between the grassy uplands and the river basin. All the other canyons in the area also located in close proximity to the river, though, suggesting that Bonfire Shelter does not offer any specific advantage over any other site in the area. |
2. The herd must be large enough in number in order to ensure that enough momentum is gained by the moving herd.
|
We do not know the exact size of the herds for Bone Bed 2 or Bone Bed 3. Bone Bed 2 has a minimum number of 27 individual bison present, but is estimated to contain upwards of 120 individuals. Bone Bed 3 has minimum number of 197 bison present, but is estimated to contain upwards of 800 individuals (Dibble and Lorrain 1968) |
|
Both bone beds contain enough bison to constitute a successful bison jump (based on estimates by Frison 1978: 228) |
|
The presence of a large enough herd at Bonfire Shelter does not speak to the distinctiveness of the site, though. It shows that the bones at Bonfire Shelter could have been deposited by a bison jump. It does not show that Bonfire Shelter possesses a topographic advantage over other sites in the area. |
3. A long, flat, and relatively straight path is needed for a drive lane.
|
Based on the slope calculations, only several areas within the region contain a contiguous strip of flat land suitable for driving herds. (Click here for these locations] |
|
Of those areas, two of those do not approach any cliff edges and therefore would not serve as useful drive lanes for bison jumping. The other three all approach cliff edges. [Click here to see the remaining paths] |
|
Of the three remaining paths, two of those paths approach the main channel of the Rio Grande. It can be argued that these locations would be less ideal for bison jumping because 1) the water level in the river is higher and more permanent than would be in Mile Canyon and 2) the height of the cliff face is much higher, making accessing and removing the bison carcasses more difficult. |
|
The remaining path approaches Mile Canyon, a tributary of the Rio Grande. Mile Canyon is easier to access and is not permanently flooded with water. (Click here for map). |
|
Also, analysis of the least-cost pathways show that any bison starting from the northern or northeastern portions of the site would be funneled through this flat area in order to reach Bonfire Shelter. [Click here to see map again] |
|
Overall, it appears that the landscape leading up to Bonfire Shelter was especially effective as a bison drive lane. In fact, it appears that no other area in this 156 km2 region would have served quite as well as a drive lane. |
|
While the analysis does not eliminate all other approaches, it does suggest that Bonfire Shelter may have offered some special advantages. |
4. The bison are unable to see the cliff edge until they are directly upon it, giving them little opportunity to escape
|
Various techniques were used to approach this question. Both the viewshed analyses and the camera flybys proved useful. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Based on the viewsheds, it could be determined at what exact point the edge became visible along any of the proposed paths. (Path Line 1 and Path Line 6 were omitted early in the analysis)
|
5. The orientation of the cliff face must be in the same direction the wind is blowing.
|
The cliff face at Bonfire Shelter is oriented in a north-northwest direction. |
|
Palaeoclimatic data on site specific wind direction are unavailable. It is therefore impossible to detect if the jump-off point would have been upwind or downwind of the drive lane approach. |
|
But, this cliff face configuration matches with the configuration of other known jump sites (Polk 1979). |
|
If the prevailing westerly winds of today were also present during Late Paleoindian and Late Archaic times, then the site would not have been detectable to the bison by scent. |
6. The edge of the cliff must be steep and sharp enough and the height of the cliff must be great enough to ensure that all bison go over and are at least severely maimed by the fall.
|
At the known jump off point, the difference in elevation between the cliff overhang and the base of the canyon is over 30 m. (Click here to view a profile of the canyon edge) |
|
The fall would have been terminated short upon contact with the talus cone, which is presently ~5 m above the canyon floor (Meltzer 2004). |
|
During the Early Holocene, the talus cone would have been shallower. The drop would have been approx. 23 m from cliff edge to the top of the talus cone (Dibble and Lorrain 1968). |
|
This fall would have severely maimed or killed any animal |
|
It appears that the jump point at the notch may have been much steeper than elsewhere along the canyon edge. Other areas adjacent to the known jump point are in fact much less steep. (Click here to view map of cliff edge slope). |
|
It would have been much more difficult to attempt a bison jump at these points because the cliff edge is sloped. This would have offered the bison opportunity to escape the jump, and possibly scramble back uphill towards the hunters. This would have been very dangerous. |
The most viable route for a bison drive lane follows a north-south path parallel to Mile Canyon. The path could have begun as far as 3 km north of Bonfire Shelter. A long, flat stretch of land extends this distance, providing a ideal bison drive lane. This path connects what was probably once a grazing area in the uplands to a permanent water source, the Rio Grande. Bison herds may have frequently trekked through this area during amenable climatic episodes.
The drive lane may have been divided into two separate sections. The first and the longer of the two may have extended from the bison gathering point north of the site to a turning point immediately east of the site. Several members of the prehistoric group, known as runners, may have driven the bison along this lane, increasing herd speed as they approached the turning point. At the turning point, the bison would have been sharply turned west and prompted through a narrowing drive lane straight toward the cliff edge. Once they approached the edge, the would be either driven directly into the notch and stumble downhill or go directly over the edge without ever falling into the notch. Either way, they would probably end up dead on the canyon floor.
Of course while it is possible that a number of other routes were used to approach the site, but the propsed route includes all the necessary criteria for a successful bison jump. In fact, within this entire region there may have been only a few sites which offered this complete suite of strategic qualities. Bonfire Shelter falls at the southerly end of the best drive lane in the region. Other stretches of flat land in the area either 1) do not lead to a jump off point or 2) lead to the Rio Grande, where the jump off is so great that it may have been difficult to access the carcasses once they had been jumped. Bonfire Shelter also falls along a natural path between grazable uplands and a permanent water source. This would guarantee that bison could be found somewhere in close proximity to the site. Additionally, the cliff edge at Bonfire Shelter is fairly obscured by the topography of the surrounding features. This would have helped in successfully jumping the animals. The bison would not have been able to see the cliff edge until it was too late and the momentum of the herd had already charged them over the edge. Finally, the jump-off point at Bonfire Shelter may have been the last point along the proposed drive lane appropriate for a jump. While the resolution of the regional DEM is not high enough to accurately assess the cliff edge drop-off south of the site, it does appear that the edge becomes increasingly sloped. These conditions are not suitable for bison jumping because it offers the bison opportunities for escape. The jump-off point at Bonfire Shelter may have presented the last opportunity for a successful jump along the proposed drive lane.
Bonfire Shelter may have also been an appropriate site for bison jumping because of the configuration of the canyon floor just beneath the jump point. The position of the boulders directly beneath the jump would have provided a natural trap or barricade for the animals as they plunged into the floor. The dead animals would not have scattered about the canyon floor and the animals who had somehow survived would not have been able to escape down the canyon. The rockshelter would have provided a natural pound for the bison.
While this analysis has offered a viable explanation for how the drive lane may have been configured and how other features of the landscape may have been used by prehistoric hunters, it has failed to explain the function of the notch. Did the notch play a crucial role in the bison jump? Based on this analysis, it appears that the notch may not have been the most salient feature attracting prehistoric hunters to this spot time and again. The other characteristics discussed above may have been what made the point above Bonfire Shelter an attractive location for bison jumping. It remains unclear why hunters would have used the notch to jump the bison when a straight plunge over the edge would have just as easily (and more thoroughly) killed the animals.
On the other hand, the notch may have offered another advantage to prehistoric hunters. Historic ethnographic accounts from the 18th and 19th centuries discuss a technique sometimes used during bison jumping. Courageous herd runners would sometimes run with the herd all the way to the cliff edge in order to ensure that the animals successfully jumped. Just before the precipice, though, the runner would escape to a predetermined crevice in the cliff edge (Hornaday 2002; Verbicky-Todd 1979). This would stop the person from plunging over the cliff edge with the bison. It is possible that the notch functioned in this manner. The notch may have been used as a final hiding place for the hunter who was running the herd to the edge. The herd may have jumped the cliff edge directly adjacent to the notch, rather than through the notch. The carcasses would have ended up in the same location on the talus cone. While this technique would have been very dangerous, based on the historic accounts, it is known to have been used by Native American hunters.
Although this analysis does not prove that Bone Bed 2 and Bone Bed 3 were both the result of a bison jump, it does help to demonstrate its feasibility. It shows that a bison jump could have successfully been executed at Bonfire Shelter. The topography of the landscape surrounding Bonfire Shelter would be well suited for driving and jumping bison herds. It contains all necessary criteria.
This analysis also demonstrates that other locations in the area may have lacked the necessary criteria for a successful bison jump. Upon closer inspection of the regional topography, it appears that other locations do not contain the topographic characteristics held by most known prehistoric bison jumps. This suggests that Bonfire Shelter may have been somewhat unique to the region, and helps to explain why Bonfire Shelter was repeatedly used at such widely spaced intervals. The qualities that drew hunters to the spot during the Late Paleoindian period were probably the same qualities that drew hunters during the Late Archaic. Bonfire Shelter possesses the complete suite of topographic characteristics making it aptly suited for a bison jump locality.
In future field seasons, it may prove useful to return to Bonfire Shelter and inspect the surrounding landscape for drive line features. Although the upland area immediately surrounding the site has been closely inspected for drive line features, there has been less effort directed towards the systematic inspection of the landscape farther away from the site. Although irrefutable drive line features are absent from the immediate vicinity of the site, perhaps they are still in existence farther up the drive lane. This analysis has provided the necessary direction for pursuing such a task.
Bement, L.
(1986) “Excavation of the Late Pleistocene deposits of Bonfire Shelter Val Verde County, Texas.” Texas Archaeological Survey, Archaeology Series 1.
Binford, L.R.
(1978) Nunamiut Ethnoarchaeology. Academic Press, New York.
Bryant, V.M. and Holloway, R.G.
(1985) A Late-Quaternary paleoenvironmental record of Texas: an overview of the pollen evidence. In Pollen Records of Late-Quaternary North American Sediments, edited by V.M. Bryant and R.G. Holloway, pp. 46-66. American Association of Stratigraphic Palynologists Foundation, Dallas.
Dibble, D. and Lorrain, D.
(1968) “Bonfire Shelter: a stratified bison kill site, Val Verde County, Texas.” Texas Memorial Museum Miscellaneous Papers 1.
Frison, G.C.
(1978) Prehistoric Hunters of the High Plains. Second edition. Academic Press, New York.
Hornaday, W.T.
[2002) Extermination of the American Bison. Smithsonian Institution Press. Washington, D.C.
Llobera, M.
(2003) Extending GIS-based visual analysis: the concept of visualscapes. International Journal of Geographical Information Science 17(1): 25-48.
McDonald, J.N
(1981) North American Bison: Their Classification and Evolution. University of California Press, Berkeley.
Meltzer, D.J.
(2003) Topographic data for Bonfire Shelter. QUEST Archaeological Research Fund.
(2004) “Field Research in 2003 at the Bonfire Rockshelter site, Val Verde County, Texas.” Quest Archaeological Research Fund: Summary of field and laboratory research, 2003-2004.
Polk, M.R.
(1979) Bison Jump Sites in the Northwestern Plains of North America: a Locational Analysis. Master's Thesis, Michigan State University, East Lansing.
U.S. Geological Survey
(1999) National Elevation Dataset, First Edition. U.S. Geological Survey, Sioux Falls, SD.
Verbicky-Todd, E.
(1984) Communal Buffalo Hunting Among the Plains Indians. Archaeological Survey of Alberta, Occasional Paper No. 24.
Wheatley, D. and Gillings, M.
(2002) Spatial Technology and Archaeology: The Archaeological Applications of GIS. Taylor and Francis, New York.
Judy Cooper is a graduate student enrolled in the Archaeology PhD Program in the Anthropology Department at Southern Methodist University. As a student of Dr. David Meltzer, director of the QUEST Archaeological Research Program, she works primarily on projects pertaining to the peopling of the Americas and Paleoindian archaeology in the Great Plains of North America. Judy is interested specifically in spatial analysis of archaeological sites, namely lithic scatters, using GIS-based approaches. She is also currently enrolled in the GIS Graduate Certificate Program at the University of Texas at Dallas and will complete the program in December 2004. Judy can be contacted by e-mail at jcooper@smu.edu.
