Forecasting the Risk of Reinvasion by the Giant African Snail in Ogasawara, Japan

ABSTRACT Species distribution models (SDMs) are widely used to predict the potential distribution of invasive species, even in areas with limited occurrence data. However, their accuracy may be compromised when input data are spatially biased or geographically restricted. This study addresses such challenges by focusing on the giant African snail ( Achatina fulica (Ferussac, 1821) (Gastropoda: Stylommatophora: Achatinidae)), a globally invasive species present on Chichijima island, Ogasawara Archipelago, Japan. Although the species has declined in range and abundance on the island since the 1990s, a time‐lagged population resurgence remains possible. To support proactive management, we developed SDMs using MaxEnt with three current observation datasets: two from Chichijima and Hahajima (both within the archipelago), and one from Hualien County, Taiwan. Notably, the latter two areas host high‐density populations. We projected these models onto Chichijima and assessed their predictive performance using historical monitoring data collected during peak abundance. Among the models, the one trained on Hahajima data exhibited the highest predictive accuracy, likely due to similar environmental conditions with Chichijima. This model identified several high‐risk areas on Chichijima that have not been recently surveyed but may serve as reinvasion hotspots. Our results demonstrate the utility of transferring SDMs across regions to enhance risk assessment in data‐limited contexts. By leveraging information from ecologically analogous populations, management strategies for invasive species can be more efficiently targeted, supporting early intervention and resource prioritization.