American megafaunal extinctions and human arrival: Improved evaluation using a meta-analytical approach

Abstract

The Late Quaternary megafaunal extinctions have been discussed for a long time, and human impacts have been proposed as a potential driver, especially in North America. However, coexistence among first humans and now-extinct megafauna has been assessed only by narrative reviews and local evaluations. This paper involves a systematic review of the literature and analyzed data using a formal meta-analytic approach to assess the chronological overlap among earliest humans and now-extinct mammals in the New World under an explicit biogeographical context. Reliable radiocarbon dates indicating First-Appearance of Humans (FAH ages) and Last-Appearance of Megafauna (LAM ages) from South and North American archaeological sites were used to compute the summary effect size. Positive effect sizes indicate the time of human–megafauna coexistence, whereas negative ones indicate lack of coexistence. The findings probabilistically indicate that humans did not coexist with megafauna throughout the New World. Therefore, humans were not the unique driver (if they had any influence) of end-Pleistocene extinctions. Subgroup meta-analyses indicate that non-Sloths, and Proboscideans and non-Proboscideans, already had been extinct when humans arrived in northern South America and Alaska/Canada. Consequently, non-human causes must have been the main driver of Pleistocene extinction in these regions without evidence of human–megafauna coexistence. Thus, assuming that humans had any influence on megafaunal extinction in the other regions of the continent (i.e., where humans and megafauna possibly coexisted), there must be multiple causes, geographically structured, for the late Quaternary extinctions.