Bonfire Shelter (41VV218) is located in Val Verde County, TX, just outside of the town of Langtry (see map)
	Located at the southern edge of the Great Plains region on the Edwards Plateau
	The site is located in Mile Canyon, a small tributary of the Rio Grande, running northward, perpendicular to the river
	The region is presently characterized by desert, scrubland vegetation, though prehistorically, the area may have been marked by grassland vegetation
	Though the climate is extremely arid and there is little standing water in Mile Canyon, the occasional thunderstorm can create dangerous flash floods within the canyon

	Located within a rockshelter carved into the eroding limestone cliff face (~26m high)
	The rockshelter is approx. 80-90 m in length and a maximum of 15m in depth (Dibble and Lorrain 1968)
	The shelter is approx. 18m above the floor of Mile Canyon (Meltzer 2004)
	The front opening of the site is naturally barricaded by large, limestone boulders which detached from the cliff overhang above.  This collapse event occurred prior to the accumulation of the archaeological deposits (Dibble and Lorrain 1968)
	A large talus cone has formed at the southern edge of the rockshelter.  It was formed when eroding soils from the surrounding uplands were funneled down slope through a deep crevice in the overhanging cliff face known as "the notch" (Click here to view to notch, photo by D. Dibble from Texas Beyond History website)
	Click here to view a plan map of the rock shelter (Figure from Bement 1986)

	The archaeological deposits at Bonfire Shelter were first discovered in 1958 by a young Dr. Michael Collins and were brought to the attention of state archaeologists
	A team of archaeologists from the University of Texas lead by Dr. David Dibble conducted excavations at Bonfire Shelter in 1963-1964.
	Bonfire Shelter was investigated again in 1983-1984 by a group of archaeologists including Dr. Solveig Turpin, Dr. David Dibble, and Dr. Leland Bement
	In 2003, the QUEST Archaeological Research Program of Southern Methodist University conducted brief fieldwork at the site, focusing mostly on topographic mapping of the site
	For a full discussion of the archaeological investigations at Bonfire Shelter, visit http://www.texasbeyondhistory.net/bonfire/investigations.html

Bonfire Shelter contains three stratigraphic deposits: (Click here for profile, from Bement 1986) Bone Bed 1: 12,430 ±490 years BP (Early Paleoindian), contains bones of extinct Pleistocene mammals (including Equus, Camelops, Mammuthus) Bone Bed 2: 10,080 ±100 years BP (Late Paleoindian), contains bones of extinct bison species, Bison antiquus Bone Bed 3: 2,645 ±75 years BP (Late Archaic), contains bones of modern bison species, Bison bison

 

Bone Bed 1:  contains possible bone expediency tools, charcoal flecks, but no stone tools (Bement 1986).  May be a palaeontological deposit rather than an archaeological deposit (Meltzer 2004).

 

Bone Bed 2:  contains at least 27 individual bison, though it has been estimated that the entire deposit could contain 120 individuals.  Has been interpreted as three separate bison jump events.  Contains a very small lithic assemblage, including 5 nearly complete projectile points (Plainview, Folsom, and possible Midland).  Deposit concentrated in talus cone (Dibble and Lorrain 1968).    Click here to see projectile points (photo by Milton Bell, from Texas Beyond History website)

 

Bone Bed 3:  contains at least 197 individuals, though it has been estimated that the entire deposit could contain 800 individuals.  Based on the sheer quantity of animals, it is interpreted as multiple jump events (though there is no evidence for stratigraphic separation).  Much of the bone is heavily burned.  Two hearths have been associated with Bone Bed 3.  The lithic assemblage contains 88 artifacts, 25 of which are identifiable projectile points (Castroville and Montell).  Deposits concentrated in the talus cone (Dibble and Lorrain 1968).       Click here to see example of projectile points (photo by Milton Bell, from Texas Beyond History website)

[back to Table of Contents]

Problems

Bonfire Shelter has generated a fair amount of controversy amongst archaeologists.  Not only is Bonfire Shelter the most southern bison jump recorded in North America, but Bone Bed 2 also represents the oldest.  This deposit predates all other known bison jump sites by nearly 5,000 years.  For this reason, many have questioned the interpretation of Bone Bed 2 as a bison jump locality. 

Some have suggested that Bone Bed 2 of Bonfire Shelter may represent a processing locality, rather than than the primary kill locality (Binford 1978).  The assemblage lacks low-utility elements, suggesting that the actual kill occurred elsewhere and select meat packages were butchered and then transported to Bonfire Shelter for further processing.  The answers to these questions lie predominantly in the fauna itself, which are presently being reanalyzed by Ryan Byerly, member of the QUEST research team.

Bonfire Shelter is the only known bison jump site in the area.  While this may suggest that Bonfire Shelter possessed unique qualities making it an especially attractive location for bison jumping, it may also simply be a consequence of the excellent preservation offered by the rockshelter.  It is completely possible that people were jumping bison over cliffs all over the region throughout prehistory, but these sites have not been preserved.  Bonfire Shelter has been protected from destructive elements, such as wind and water--the same elements which may have destroyed other bison jump localities in the region. 

[back to Table of Contents]

Prehistoric Bison Hunting

 

	Hunting tactic frequently used during the Late Prehistoric and historic periods in North America.
	Bison jumping was probably a communal hunting tactic.  It requires the cooperation and planning of multiple people in order to successfully execute.
	Its success and frequency probably increased after the invention of the bow and arrow and the later introduction of the horse, though it was practiced for several thousand years before.
	Earliest known bison jump site recorded is Head-Smashed-In Buffalo Jump.  It's earliest levels date to 5,660 ±100 years BP (date from Frison 1978)
	Bison jumps involved the following series of events (summarized from Verbicky-Todd 1984): Members of the hunting group, known as runners, would scout out bison herds in the area and notify the rest of the group When the group was ready to jump the herd, the runners would round-up the bison, sometimes several kilometers from the jump point.  The runners would sometimes dress in bison skins in order to act as members of the herd and trick the bison into following. The herd would be directed down predetermined drive lanes, often marked with piles of rocks, brush, or dried bison dung known as drive lines.  Other members of the group would hide behind these piles, jumping out to scare the bison and make sure they dido not deviate from the set path. The speed of the herd was increased as the jump point approached.  The goal was to increase the anxiety and panic amongst the bison. The drive lane would narrow as it approached the edge.  Greater numbers of hunters would appear from the sides, exciting the bison even more.  The bison would then be forced over the edge and plunge to their death.   For the animals who survived the fall, hunters at the base of the drop would emerge and finish them off with spears or other weapons. Because bison travel in herds and follow the direction of the bison immediately in front of them, it is possible to force a herd of bison to jump a cliff edge.  Even if the front members of the herd are able to avoid the jump, the remaining members, compelled by the momentum of the herd behind them, will be forced over the edge before they have time to react.

[back to Table of Contents]