The role of habitat fragmentation in Pleistocene megafauna extinction in Eurasia

The idea that several small, rather than a single large, habitat areas should hold the highest total species richness (the so-called SLOSS debate) brings into question the importance of habitat fragmentation to extinction risk. SLOSS studies are generally addressed over a short time scale, potentially ignoring the long-term dimension of extinction risk. Here, we provide the first long-term evaluation of the role of habitat fragmentation in species extinction, focusing on 22 large mammal species that lived in Eurasia during the last 200 000 years. By combining species distribution models and landscape pattern analysis, we compared temporal dynamics of habitat spatial structure between extinct and extant species, estimating the size, number and degree of the geographical isolation of their suitable habitat patches. Our results evidenced that extinct mammals went through considerable habitat fragmentation during the last glacial period and started to fare worse than extant species from about 50 ka. In particular, our modelling effort constrains the fragmentation of habitats into a narrow time window, from 46 to 36 ka ago, surprisingly coinciding with known extinction dates of several megafauna species. Landscape spatial structure was the second most important driver affecting megafauna extinction risk (ca 38% importance), after body mass (ca 39%) and followed by dietary preferences (ca 20%). Our results indicate a major role played by landscape fragmentation on extinction. Such evidence provides insights on what might likely happen in the future, with climate change, habitat loss and fragmentation acting as the main forces exerting their negative effects on biodiversity. © 2021 The Authors. Ecography published by John Wiley & Sons Ltd on behalf of Nordic Society Oikos