Modelling the risk of invasion by the red-swamp crayfish (Procambarus clarkii): incorporating local variables to better inform management decisions.
The correct modelling of the potential distribution of an invasive species is crucial to define effective management and monitoring strategies. Here we compared the results of models built at different spatial scales to identify the areas at risk of invasion by the red swamp crayfish (Procambarus clarkii) in the northwest of Portugal. Firstly, we surveyed crayfish at 97 locations. Secondly, we used presence-absence data and local variables to model its current distribution (local variables model) and identified slope and river width as the best explanatory factors. Thirdly, we integrated these two local variables into a former model built for the Iberian Peninsula (regional model) increasing considerably its predictive power. Finally, we compared both models focusing on the area predicted to be invaded. The local model showed a considerably narrower extent of suitable areas for crayfish in the study area than the regional model. These results show that the refinement of regional scale predictions through the incorporation of species-environment relationships at local scales may be important for supporting management decisions. By not integrating the effects of local variability, regional bioclimatic models may overlook the potential distribution of this invader at manageable extents. Results: further suggest that a wide range of native ecosystems of conservation value are probably unsuitable for this invasive species.