Janaína de Andrade Serrano, Lars Iversen, Laura J. Pollock

As the biodiversity crisis deepens, understanding extinction risk is essential for conserving at-risk species and triaging those potentially overlooked. Extinction risk is often estimated with traits (e.g. larger species are more vulnerable) without considering the context of threats or human bias in the listing process (e.g. more popular species are more or less likely to be listed). On the other hand, current global assessments of threats do not include the context of the biological variation of species (e.g. hunting mainly impacts large species). Here, we show that biological traits, threats, and popularity all interact to influence extinction risk for birds globally. We find particularly strong interactions between body mass and hunting (large body mass increases extinction risk for species threatened by hunting), habitat breadth and agriculture (narrow habitat breadth increases extinction risk for species threatened by agriculture), body mass and agriculture (small bodied species have increased extinction risk when threatened by agriculture) and range size and agriculture (for mid range-sized species, agriculture increases extinction risk). Further, we find that extinction risk increases with popularity, likely reflecting the increased chance of popular species having been listed given the same traits and threats. Overall, our results show the importance and necessity of including both biological and human biases, as well as human-posed threats when estimating extinction risk and identifying regions harbouring disproportionally high numbers of vulnerable species.