The Geography of Altithermal Subsistence Adaptations on the North American Great Plains

On the Great Plains, the Middle Holocene (8200–5000 yr BP) was a warm period where foraging peoples expanded into new food resources and developed new food processing technologies. It is unclear precisely what drove these cultural changes, and environmental explanations tend to assume that climatically mediated causes occurred along latitudinal gradients, with the southern Plains experiencing severe water shortages and declining abundances of important resources relative to the northern Plains. A decline in one such resource, bison, has been repeatedly proposed as the mechanism that drove the adoption of new resources and technologies, with the most pronounced consequences at southern latitudes. To evaluate whether the availability of bison was responsible for subsistence variability in the Middle Holocene, this dissertation examines bison remains from sites dating to the Middle and preceding Early Holocene. This entailed establishing first assessing spatio-temporal variability in bison population sizes, then evaluating subsistence responses to that variability. A reliable proxy measure of bison population sizes is lacking, and I therefore assumed a link between bison population sizes and habitat suitability. I first analyzed stable oxygen isotope ratios and rates of linear enamel hypoplasia (LEH) in bison teeth to understand geographic and temporal variability in habitat suitability. The former variable was used as a proxy measure of atmospheric temperature, where warmer temperatures exerted more stress on these animals. The latter variable records cases of physiological stress, where the incidence rate of hypoplasia increases as grassland conditions deteriorate and populations decline. I expected that suitability declined from the Early to the Middle Holocene, and that the magnitude of this decline was negatively associated with latitude. Next, I used the prey-as-patch model to predict that declines in bison population sizes prompted people to more completely process each hunted animal. To evaluate this prediction, this dissertation examined spatial variability in bison bone fragmentation patterns in the Middle Holocene. I predicted that levels of fragmentation would increase moving south, tracking declines in bison populations. The oxygen isotope data indicate that Middle Holocene warming was most pronounced at southern latitudes, and that the Canadian prairies and northernmost U.S. grasslands experienced little to no change in average temperature compared to the Early Holocene. However, isotope data only extend as far south as Kansas, leaving the true severity of Middle Holocene warming on the southern Plains unknown. Warming at southern latitudes was a winter phenomenon, which would have adversely affected grazing conditions. The incidence of LEH is too low to detect spatio-temporal patterns in the dataset, and I could not use these data to evaluate predictions about habitat suitability. Long bone fragmentation is consistent with higher processing intensity at northern latitudes, contradicting expectations. These data support the hypothesis that Middle Holocene warming was primarily a southern phenomenon, and that the decline of bison in this period was likely a restricted to the central and southern areas of the region. However, the motivations for increased carcass processing intensity appear unrelated to bison population levels within the Middle Holocene. These results suggest increasingly divergent subsistence pathways between the northern and southern Plains during the Middle Holocene. Much like this period, North America’s grasslands are warming today, imperiling the social and ecological systems that support bison. A better grasp on the relationship between bison and climate change in the Middle Holocene holds potential lessons for the persistence of these animals.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.