关键词: MaxEnt模型, 马铃薯甲虫, 潜在分布区, AUC值

Abstract: Leptinotarsa decemlineata, an alien pest, has established populations in places such as Xinjiang and Heilongjiang, destroying crops such as potatoes and tomatoes, with an annually expanding distribution in China. Predicting the suitable area of L. decemlineata is conducive to early monitoring, warning, and controlling this pest in China. According to the existing distribution records of L. decemlineata and main environmental variables, potential geographic distributions of L. decemlineata in China were predicted and analyzed using MaxEnt model and ArcGIS. Seven major environmental variables were selected by the methods of principal components analysis, correlation test, and Jackknife test. AUC value was 0.917 by the MaxEnt model, with high fitness between the predicted results and the actual values. According to the reclassification method in ArcGIS, the total area of the contemporary suitable area of L. decemlineata was 3.72×10⁶ km², accounting for 38.67% of total land area of China. Highly suitable area covered only 0.08% of land area in China, mainly located in Baotou and Hohhot in Inner Mongolia Autonomous Region. Moderately suitable area was 1.26×10⁶ km² and accounted for 13.07% of total land area of China, which included eastern parts of Heilongjiang, Liaoning and Hebei, western Sichuan, southwestern Gansu, northwestern Henan, southern Shanxi, northern Xinjiang, central parts of Jiangsu, Hubei, Guizhou and Anhui, vast areas of Shandong, and surrounding places of Baotou, Inner Mongolia. Lowly suitable area was 2.45×10⁶ km² and accounted for 25.52% of China’s land area, which mainly distributed in the vast areas of central Inner Mongolia, northern Xinjiang, eastern Northeast China, Northern China, Central China and Southern China. Mean annual temperature, the lowest temperature in the coldest month, and the precipitation of the driest quarter were the most important environmental variables affecting the distribution of suitable area of L. decemlineata.

Key words: MaxEnt model, Leptinotarsa decemlineata, potential geographic distribution, AUC value.