Assistant professor McGill University Montreal, Quebec, Canada
Abstract: Extinction risk is the quantification of how likely a species is to disappear in the foreseeable future. The IUCN Red List is a widely known database that assesses the risk of extinction for animal and plant species globally. Amongst animals, the Red List has shown that there is an increased extinction rate overall for birds and mammals, and a lower estimated extinction rate for amphibians, reptiles and fish. This may reflect not only real trends but also a shortage of data for species that have not yet been evaluated, especially in less-studied taxa. One way to address the problem of biased information about species extinction risk (or conservation status) is using species attributes or traits, which are more readily obtained for most species than population dynamics data. However, predicting extinction risk from traits alone without considering the type of threats to which species are exposed can lead to confounding effects. This stems from the fact that species are more or less at risk of extinction because of their vulnerability to specific threats, which combined with their specific characteristics results in their extinction risk. A better strategy might be to find traits that confer vulnerability to a specific threat and then link that to extinction risk estimates, which can be a cost and time-efficient way to predict if a given species or taxon is of concern. Furthermore, the conservation status of a species is highly biased by how much research effort has been spent in their collection and observation, which may lead to misclassifications for listed species and data deficiency that does not allow the species to be classified as threatened. Thus, we developed a model that accounts for interactions between traits and threats, and popularity to explain extinction risk for birds. We found that larger species are more likely to be threatened by hunting and invasive species, and smaller species are more likely to be threatened by agriculture. We also found that wide habitat breadth protects species that are threatened by resource use and that migratory birds are less prone to be threatened by agriculture. We tested the correlation of the residuals of the model against popularity and found that the least popular species indeed have less accurate estimates of extinction risk, what suggests that they need to be more closely looked in extinction risk assessments.
