Variation in niche and distribution model performance: The need for a priori assessment of key causal factors

Ecological niche models and species distribution models are becoming important elements in the toolkit of biogeographers and ecologists. Although burgeoning in use, much variation exists in implementation of these techniques, leading to considerable diversity of methodology and discussion of what is the ‘best’ approach. In this analysis, we explore implications of different configurations of major factors that constrain species’ distributions—abiotic factors and dispersal limitation—for the success or failure of these models. We analyze variation in performance among modeling approaches as a function of the relative configuration of these two factors and the spatial extent of training region, with the result that a clear understanding of the abiotic-dispersal configuration is a prerequisite to effective model implementations; the effects of spatial extent of the training region are less consistent and clear. Model development will be powerful only when set in an appropriate and explicit biogeographic and population ecological context.

► Ecological niche models are now very popular in the ecological literature.
► Many studies using these approaches lack a proper conceptual framework.
► We analyze effects of abiotic niche and accessibility on model performance.
► Understanding the abiotic-dispersal configuration is crucial.