Phylogenetic and Trait-Based Prediction of Extinction Risk for Data-Deficient Amphibians.

Amphibians are among the most highly threatened lineages, with at least 2,000 species estimated to be in danger of extinction [1, 2]. Alarmingly, another ∼2,200 species (∼25% of all ∼7,900 known species) are data deficient or not evaluated (hereinafter termed data deficient) by the International Union for Conservation of Nature (IUCN) [1]. Without an estimate of their status, data-deficient species are usually overlooked in conservation planning and resource allocation [3]. Amphibians have the highest proportion of data-deficient species of any vertebrate group [1, 4], which highlights the need to estimate their threat status considering potentially imminent extinctions. We apply a trait-based spatio-phylogenetic statistical framework [5] to predict threat status for data-deficient species. Because ecological, geographical, and evolutionary attributes increase extinction risk [6, 7], we used geographic distribution data [1, 8], phylogenetically imputed ecological traits, and an amphibian phylogeny [9] to provide initial baseline predictions. We estimate that half of the ∼2,200 data-deficient species are threatened with extinction (vulnerable, endangered, or critically endangered), primarily in the Neotropics and Southeast Asia. This increases the number of amphibian species estimated to be threatened with extinction by ∼50%. Of these, we predict that ∼500 species are endangered or critically endangered, and three may be extinct already. We highlight families that are most at risk and suggest where urgent conservation is needed to avert their loss. We show that some of the most vulnerable species may also be the most poorly known and offer an analytical framework for preliminary analysis of their threat status in the face of deficient empirical data.