The role of environmental change in the late Pleistocene megafaunal extinctions remains a key question, owing in part to uncertainty about landscape changes at continental scales. We investigated the influence of environmental changes on megaherbivores using bone collagen nitrogen isotopes (n = 684, 63 new) as a proxy for moisture levels in the rangelands that sustained late Pleistocene grazers. An increase in landscape moisture in Europe, Siberia and the Americas during the Last Glacial–Interglacial Transition (LGIT; ~25–10 kyr bp) directly affected megaherbivore ecology on four continents, and was associated with a key period of population decline and extinction. In all regions, the period of greatest moisture coincided with regional deglaciation and preceded the widespread formation of wetland environments. Moisture-driven environmental changes appear to have played an important part in the late Quaternary megafaunal extinctions through alteration of environments such as rangelands, which supported a large biomass of specialist grazers. On a continental scale, LGIT moisture changes manifested differently according to regional climate and geography, and the stable presence of grasslands surrounding the central forested belt of Africa during this period helps to explain why proportionally fewer African megafauna became extinct during the late Pleistocene.
We are indebted to the following museums, curators and miners for assistance with samples, advice and encouragement: Canadian Museum of Nature (R. Harington), American Museum of Natural History (R. Tedford), Royal Alberta Museum (J. Burns), Natural History Museum London (A. Currant), Yukon Heritage Centre (J. Storer), University of Alaska, Fairbanks (D. Guthrie, P. Matheus), Institute of Plant and Animal Ecology, RAS Yekaterinburg (P. Kosintsev), Laboratory of Prehistory, St Petersburg (V. Doronichev and L. Golovanova), and a range of Yukon miners including B. and R. Johnson, the Christie family, K. Tatlow and S. and N. Schmidt. We also thank T. Faith, C. Turney, B. Shapiro, D. Froese, M. Richards, A. Sher, J. Glimmerveen, G. Larson, E. Willerslev, R. Barnett and members of ACAD (Australian Centre for Ancient DNA) for assistance with sampling and analysis. We particularly thank NRCF and the Oxford Radiocarbon Accelerator Unit (T. Higham). This work was funded by grants and fellowships from the Australian Research Council (DP140104233 and LF140100260) and UK Natural Environment Research Council to A.C. Work contributed by H.J. was supported by the Research Council of Norway through its Centres of Excellence funding scheme, project number 223272. Work contributed by M.J.W. was supported by NSF project numbers PLR 1204233 and PLR 0909527.
Dataset used in analysis.