Species distribution models (SDM) are amongst the most popular tools currently used in predictive ecology, yet the most common applications may be substantially limited, especially in context of invasion biology. Specifically, the basic SDMs assume equilibrium conditions, often only provide estimates of relative risk (using presence only data), typically rely on abiotic factors, are usually applied to risk of establishment only (not the other components of invasion), and to single species at a time. Despite these limitations, SDMs have been highly useful. By addressing these limitations, SDMs could be even more powerful within the context of predictive ecology.
I present novel solutions to these limitations. 1) I compare alternative methods of integrating SDMs with propagule pressure and spread models to account for non-equilibrium conditions of spreading species. 2) By integrated these models, I show how it becomes possible to estimate absolute (rather than relative) risk of establishment as well as probabilities of detection, even with presence-only data. 3) I explain how biotic interactions can be incorporated into SDMs, even with incomplete community data, and 4) how these can also be used to estimate impact of non-indigenous species. and 5) how SDMs can be integrated with propagule pressure and species trait models to predict entire pathways of invasion, rather than single species.