Forecasting the Invasion of Eurasian Watermilfoil ( Myriophyllum spicatum ) at the Northern Edge of Its North American Range

Eurasian watermilfoil ( Myriophyllum spicatum ) is an invasive aquatic plant posing increasing ecological and management challenges in North America. In Quebec, Canada, the species is rapidly expanding northward into uncolonised catchments. Thus, accurate forecasting of invasion risk is essential to guide proactive monitoring and intervention. This study aims to (i) compare the predictive performance of species distribution models (SDMs) trained at different geographic scales; (ii) identify environmental and anthropogenic factors influencing Eurasian watermilfoil presence at its northern range expansion; and (iii) project future invasion risk under scenarios of climate warming and population growth. We developed random forest SDMs using Eurasian watermilfoil occurrence data from Quebec, North America and Europe, coupled with 24 environmental and anthropogenic variables for 11,823 sub‐basins in Quebec. Model performance was assessed across regional, continental and global training datasets. Future projections were based on six scenarios combining population and climate change predictions through 2070 and 2100. Models trained on regional data from Quebec yielded significantly higher prediction accuracy than models based on their broader North American invasive range, their native range in Europe, or combined data. Maximum temperature, mean temperature, precipitation, lake volume and population density emerged as the most important predictors. Under high‐warming scenarios, the area predicted to be suitable for Eurasian watermilfoil increased nearly fivefold, with invasion hotspots expanding notably in western Quebec and along the St. Lawrence and Saguenay rivers. Regionally calibrated models more accurately reflect the realised niche of Eurasian watermilfoil at its current invasion front. Projected climate and demographic changes are likely to dramatically expand suitable habitat, increasing invasion risk across Quebec's freshwater landscapes. This study highlights the importance of local context in invasion forecasting and provides a scalable, spatially explicit framework to inform monitoring and management strategies in a globally relevant conservation issue.