Radiocarbon dating Late Pleistocene megafauna in southern Europe

Thesis

In the last 100,000 years, mainland Europe lost over a dozen large mammal species. These were part of a global extinction event that occurred during heightened climatic variability and the expansion of modern humans. The role of these two factors in driving the extinctions has been debated for decades. To move the debate forward, precise and accurate chronological estimates of fossils are needed to link species to climatic events and human activities. Unfortunately many fossil chronology datasets are small, contain unreliable dates, and are not spatially representative.

This poor temporal constraint is in part due to the challenges of radiocarbon dating Pleistocene bone. Radiocarbon dating requires sufficient collagen to be extracted from a bone sample. Heat and aridity both degrade collagen, so faunal specimens from southern European sites in particular often fail to produce sufficient collagen. This has resulted in a dearth of reliable dates from the Mediterranean and an incomplete picture of when and where fauna disappeared in Europe. Resolving southern European faunal histories is particularly important as southern Europe has been posited as both the region species retracted to prior to extinction and where the earliest modern humans arrived in Europe.

This thesis tested approaches to improving sampling and pretreatment for radiocarbon dating poorly-preserved bone. It then produced new radiocarbon dates from three carnivore species in southern Europe: cave lions, cave hyenas, and leopards. These dates were combined with critically evaluated radiocarbon dates to model extinctions and interactions between fauna, climate, and modern humans using both established and novel modelling tools.

Key methodological results found sample preservation could vary dramatically within sites and even individual bones. Proposed alternatives to the standard ABA pretreatment produced lower-quality collagen but sometimes produced yields where ABA did not, presenting an opportunity for further purification. Critical evaluation of radiocarbon datasets was important, as inclusion of less-robust dates altered some model results.

Models were inconclusive on the still-small leopard dataset but suggested both cave lions and hyenas went extinct in southern Europe earlier than northern Europe. Time series analysis of OxCal kernel density estimate models suggested observational densities of hyenas lagged human densities and cave lions lagged climate changes, while Bayesian spatial extinction models estimated that both species went extinct in mid-latitude Europe at the last glacial maximum and terminal Pleistocene, respectively. A ‘pincer effect’ is suggested to interpret these results, where modern humans limited species’ ability to shift ranges during glacial expansions. These results are compared to other Pleistocene extinctions and discussed in relation to ecological theories.

Authors: Jacob, EM

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Subjects: Paleoecology--Pleistocene; Radiocarbon dating