Bison remains from deeply stratified deposits at the Castle Canyon Site, an area of southwestern Texas where bison remains are outstandingly rare, date about 1500 B.C.-A.D. 1000. The last bison known in the area died about 1885. It is... more
Bison remains from deeply stratified deposits at the Castle Canyon Site, an area of southwestern Texas where bison remains are outstandingly rare, date about 1500 B.C.-A.D. 1000. The last bison known in the area died about 1885. It is suggested that environmental conditions in the area may not have changed, and that presence-absence periods might rather reflect changing conditions on the Central or Northern Plains. These data do not alter Dillehay's conclusions, but present information not available to him.
The Wold Bison Jump in Johnson County, Wyoming, is one of many prehistoric, mass kill sites scattered across the Plains. At Wold, a foraging basin of prime ungulate grazing habitat abuts the gently sloping backside of a bluff.... more
The Wold Bison Jump in Johnson County, Wyoming, is one of many prehistoric, mass kill sites scattered across the Plains. At Wold, a foraging basin of prime ungulate grazing habitat abuts the gently sloping backside of a bluff. Funnel-shaped drive lines of cairns extend across the top of the bluff towards a treacherous cliff. The drive was configured to constrain stampeding bison (Bison sp.) as prehistoric hunters communally drove them from the foraging basin to the precipice. Previous GIS analyses of bison jumps inductively analyze surrounding landscapes by classifying jump locations as known, unvarying focal points of analysis. While this approach can be informative, at Wold we attempt to obtain a more general understanding of how bison jumps operate. Using iterative models of least cost paths, topographic cross-sections, and visibility analysis, we test which landscape-embedded variables are optimized at Wold as compared to other potential localities across the study area. We find that this site’s placement is primarily explained by minimizing the distance at which the cliff face is visible and secondarily by minimizing the cost of slope and curvature routes ascending into the drive lines. Our procedure could hypothetically be used to predict optimal jump locations on similar landscapes.
