Spodoptera frugiperda is a serious agricultural pest native to tropical and subtropical areas of the Americas. It has a broad host suitability range, disperses rapidly, and has now invaded nearly 100 countries around the world by quickly establishing in the novel ecologies. Based on the native occurrence records and environmental variables, we predicted the potential geographic distribution of S. frugiperda in Central Asia using the MaxEnt model and the ArcGIS. Irrigation is considered to be the main factor for the maize crop production in the Central Asia; therefore, we sought to map the potential spread of S. frugiperda using two modeling approaches together with adjusted rainfall indices and environmental data from this region. The results showed that both approaches (MCP and Obs) could predict the potential distribution of S. frugiperda. The Observation points (Obs) approach gave predicted more conservative projections compared with the Minimum Convex Polygon (MCP) approach. Areas of potential distribution that were consistently identified by the two modeling approaches included Western Afghanistan, Southern Kazakhstan and Southern Turkmenistan. The Receiver Operating Characteristic (ROC) curve test presented herein provided reliable evidence that the MaxEnt model has a high degree of accuracy in predicting the invasion of S. frugiperda in Central Asia.
Keywords: biological invasion; ecological niche model; maize; pest; risk analysis.