IPRRG webinar: Validating standardized Mahalanobis distance as a species distribution model for invasive alien species
Species distribution models (SDMs) can be used to identify areas of climatic suitability for invasive alien species under present and future climate conditions. However these models face issues including correlations between climate variables and results that are not often validated. We have implemented a standardized multivariate distance metric as an SDM, allowing for the rapid production of climate suitability maps with reliability metrics. This approach characterizes the observed multivariate climate envelope by taking advantage of the correlation between climate variables, and identifies areas where the climate is outside the range of the observed climate envelope. We include built-in metrics of reliability, including measures of variation across climate models. Model outputs include climate suitability maps, measures of variability, and most important covariates, to assess model reliability and underlying environmental drivers of the results. Using a list of 9 species from CFIA’s regulated pest list, we demonstrate that this method achieves a high level of accuracy (> 80%) for determining climate suitability for North American plant pest invasions. This suggests that the model is suitable for identifying areas of North America that are susceptible to future invasions. We show that many of the errors occur at the edge of climate suitable areas, where we would expect greater uncertainty in model predictions due to potential over-constraining and geospatial averaging. This standardized multivariate distance model can be used to assess the potential area for invasion by a plant pest in minimal time (< 1 hr) with minimal data (standard bioclimatic variables and species presence data).