Effects of distribution data types on climate change impact assessment using species distribution models

Abstract

Species distribution models (SDMs) are important tools to assess climate change impacts on ecosystems. The data for SDMs are derived from distribution atlases or a collection of ad hoc survey records, but little attention has been paid to the effects of applying different types of species data. Here we compared three data types, including atlas data, presence/absence and presence-only locality data, at two resolutions (1km² and 20km²) using artificial neural networks and Maxent models. Large differences in the projected impacts suggest that selection of appropriate data types are critial for SDMs.