Abstract:

Climate change can promote the occurrence and development of diseases and insect pests, and significantly alter their distribution and damage degree. It is important to predict the occurrence risk of South American leaf blight, one of the important quarantine objects of rubber trees in China. Based on the MaxEnt model, we used the global bioclimatic variables for the baseline (1970-2000) and future (2041-2080), the topographical data and the geographic distribution information of the disease to simulate the distribution of suitable area of the disease in the baseline and future, and predicted the risk of the disease invading China. The results showed that the AUC values of the ROC curves of the model training and test data were 0.96 and 0.95, respectively. The MaxEnt model had relatively high simulation accuracy. The top three major environmental factors determining the contribution rate were temperature annual range, temperature seasonality, and mean temperature of coldest quarter. During the baseline period, the moderately and highly suitable areas of the disease were mainly concentrated in the north of South America, the south of North America, the central and western Africa, the south of Asia, and the Pacific island countries. From the baseline to the future, the moderately and highly suitable areas in America, Africa, Asia and Oceania moved toward the southeast, middle, northeast and northwest. The high risk areas of invasion into China were northeast of Hainan Island, southeast of Guangxi Province, south of Guangdong Province. The high risk areas would increase first and then decrease under climate change and move to the northeast. Our results could provide some reference for the quarantine and control of South American leaf blight in China.

Key words: South American leaf blight, MaxEnt model, climate change, environmental factor, suitable area.