ADVANCES IN SPECIES-ENVIRONMENTAL MAPPING MODELING

Species-environmental matching models have surged in use to map rare species and harmful invasive species around the globe. We tested several models to identify suitable habitat for invasive plants, animals, and diseases. The models have proved useful in mapping suitable habitat as a hypothesis of potential species distributions to be refined as new data become available. Recent Advances in data assimilation, remote sensing integration, and innovative modeling algorithms now make it possible to address a variety of questions important to conservation biologists, resource managers, and policymakers. We provide a brief overview of species-environmental matching models, and we discuss the recent advances in the field. We discuss caveats, shortcomings, and misuses of the models before demonstrating new applications of the models to address metapopulation issues, the effects of multiple tree pathogens, and changes in species distributions over time. Then, we show how five individual models (logistic regression, boosted regression trees, random forest, multivariate adaptive regression splines, and Maxent) can be combined for an “ensemble model” for selected nonnative plant species in National Parks. We conclude with a look to the future, where the models integrate climate change, land use change, and remote sensing to map and model the leading edges of harmful invasive species. We show how citizen scientists can contribute to species mapping capabilities.