Abstract

Across the last ~50,000 years (the late Quaternary) terrestrial vertebrate faunas have experienced severe losses of large species (megafauna), with most extinctions occurring in the Late Pleistocene and Early to Middle Holocene. Debate on the causes has been ongoing for over 200 years, intensifying from the 1960s onward. Here, we outline criteria that any causal hypothesis needs to account for. Importantly, this extinction event is unique relative to other Cenozoic (the last 66 million years) extinctions in its strong size bias. For example, only 11 out of 57 species of megaherbivores (body mass ≥1,000 kg) survived to the present. In addition to mammalian megafauna, certain other groups also experienced substantial extinctions, mainly large non-mammalian vertebrates and smaller but megafauna-associated taxa. Further, extinction severity and dates varied among continents, but severely affected all biomes, from the Arctic to the tropics. We synthesise the evidence for and against climatic or modern human (Homo sapiens) causation, the only existing tenable hypotheses. Our review shows that there is little support for any major influence of climate, neither in global extinction patterns nor in fine-scale spatiotemporal and mechanistic evidence. Conversely, there is strong and increasing support for human pressures as the key driver of these extinctions, with emerging evidence for an initial onset linked to pre-sapiens hominins prior to the Late Pleistocene. Subsequently, we synthesize the evidence for ecosystem consequences of megafauna extinctions and discuss the implications for conservation and restoration. A broad range of evidence indicates that the megafauna extinctions have elicited profound changes to ecosystem structure and functioning. The late-Quaternary megafauna extinctions thereby represent an early, large-scale human-driven environmental transformation, constituting a progenitor of the Anthropocene, where humans are now a major player in planetary functioning. Finally, we conclude that megafauna restoration via trophic rewilding can be expected to have positive effects on biodiversity across varied Anthropocene settings.