LAUSR

Predictions of suitable habitat areas within a specific region can provide important information to assist in the management of invasive plants. Here, we predict the current and future potential distribution of Solidago altissima (tall goldenrod) in South Korea using climatic and topographic variables and anthropogenic activities. We adopt four single models (the generalized linear model, generalized additive model, random forest, and an artificial neural network) and a weighted ensemble model for the projection based on 515 field survey points. The results showed that suitable areas for S. altissima were mainly concentrated in the southwest regions of South Korea, where temperatures are higher than in other regions, especially in the winter season. Solar radiation and Topographic Wetness Index (TWI) were also positively associated with the occurrence of S. altissima . Anthropogenic effects and distances from rivers were found to be relatively less important variables. Based on six selected explanatory variables, suitable habitat areas for S. altissima have expanded remarkably with climate changes. This range expansion is likely to be stronger northward in west coastal areas. For the SSP585 scenario, our model predicted that suitable habitat areas increased from 16,255 km 2 (16.2% of South Korea) to 44,551 km 2 (44.4%) approximately over the past thirty years. Our results show that S. altissima is highly likely to expand into non-forest areas such as roadsides, waterfront areas, and abandoned urban areas. We propose that, based on our projection maps, S. altissima should be removed from its current margin areas first rather than from old central population areas.