Objective  Leptocybe invasa is an invasive wasp common to China, which was first introduced into the Guangxi Province of southern China in 2007. It had endangered the eucalyptus plantations in southern China within just a few years, causing serious economic and ecological losses. This study employed the MaxEnt latest prediction model, a novel maximum entropy model to firstly analyze this invasive wasp distribution in China under current and future climate conditions. The results will be helpful to understand the range dynamic change of L. invasa under the influence of temperature change.

  Method  Data were collected on the current locations of this wasp, along with the damage incurred to eucalyptus. These data were used to create a forecast model for the present and future trends of the favorable habitat under the RCP 8.5 climate scenario. And model will test the accuracy of simulation results.

  Result  The test omission rate of simulation results was basically consistent with that of theory. The value of AUC was 0.898 while a standard deviation of 0.022. The verification results showed that the test and training data were not independent and the constructed model was highly reliable and accurate. MaxEnt model predicted that the potential distribution for L. invasa was mainly concentrated at the south of the Yangtze River in the provinces of Fujian, Guangdong, Guangxi and Hainan. The RCP 8.5 scenario predicted a small-scale potential reduction of moderately favorable habitat, but the whole favorable habitat increased.

  Conclusion  Based on the analysis of the predicted scope of this invasive wasp distribution, this study has established the grounds for the effective formulation of preventive and control measures for the monitoring of L. invasa and the prediction of its possible colonies, which is a significant theoretical reference on its own.