The European Green Crab (EGC) is a high-risk global invader that can devastate coastal marine ecosystems by displacing native species, degrading and disturbing native habitats (including eelgrass), and altering food webs. EGC has recently been detected in the Canadian portion of the Salish Sea. As EGC continue to establish in the region, identifying locations on which to focus limited monitoring resources is an ongoing problem given the vast amount of coastal habitat that could be occupied by the species. A variety of methods can be used to identify highly suitable habitats for EGC at a range of spatial scales. However, none have been evaluated in the context of informing EGC management, nor for the Canadian portion of the Salish Sea. Here we evaluate five individual methods developed to assess habitat suitability for EGC (i.e., MaxEnt, stochastic gradient boosted linear and logistic regression models, a rapid site selection tool, and a qualitative site assessment and ranking tool) and five derived models generated by multiplying the outputs of these individual models. Each model relied on slightly different environmental and habitat input variables affecting EGC invasion success. Thus, rather than identifying a single preferred model, we used a multi-model ensemble approach to identify sites that are expected to be most suitable for the species. The ensemble approach likely increases predictive power by including both environmental and habitat characteristics when identifying priority sites for early detection/monitoring for EGC in the Canadian waters of the Salish Sea. Finally, we describe how the models evaluated here, alone or in combination, could be used to identify additional sites either within the Salish Sea or into new areas.
This dataset contains predicted habitat suitability from five models for European Green Crab at beaches in the Salish Sea (British Columbia, Pacific Region).