Correlative and mechanistic models of species distribution provide congruent forecasts under climate change.

TLDR: It is argued that convergent lines of independent evidence provide a robust basis for predicting and managing extinctions risks under climate change.

Abstract: Good forecasts of climate change impacts on extinction risks are critical for effective conservation management responses. Species distribution models (SDMs) are central to extinction risk analyses. The reliability of predictions of SDMs has been questioned because models often lack a mechanistic underpinning and rely on assumptions that are untenable under climate change. We show how integrating predictions from fundamentally different modeling strategies produces robust forecasts of climate change impacts on habitat and population parameters. We illustrate the principle by applying mechanistic (Niche Mapper) and correlative (Maxent, Bioclim) SDMs to predict current and future distributions and fertility of an Australian gliding possum. The two approaches make congruent, accurate predictions of current distribution and similar, dire predictions about the impact of a warming scenario, supporting previous correlative-only predictions for similar species. We argue that convergent lines of independent evidence provide a robust basis for predicting and managing extinctions risks under climate change.