基于Maxent模型的葡萄皮尔斯病在中国适生性分析

葡萄皮尔斯病能够引起葡萄灾难性的毁灭,是世界性检疫性有害生物。利用Maxent模型,结合地理信息系统ArcGIS,对该病在我国的适生范围进行了首次预测研究。结果表明,葡萄皮尔斯病的潜在分布区主要集中在我国的黄淮海地区和长江中下游地区。ROC评价结果表明,Maxent生态位模型预测葡萄皮尔斯病菌潜在分布的训练数据和测试数据的AUC值分别为0.996和0.927,达到了极高的精度。各环境变量重要性的刀切法检验表明,降水量和温度因素对葡萄皮尔斯病的潜在分布影响最大。
Pierces' disease of the grape(PD), causing catastrophic destruction of the grape, is the important quarantine target in many countries around the world. In this study, the models Maxent were firstly used to predict the potential geographical distribution of Xylella fastidiosa based on geographic information system(ArcGis)in China. The results showed that the potential distribution of X. fastidiosa mainly concentrated in Huang-Huai-Hai Region and in middle and lower reaches of Yangtze River Region. The fit for the model as measured by AUC was high, with value of 0.996 for the training data and 0.927 for the test data, indicating the high level of discriminatory power for the Maxent. A jackknife test in Maxent showed that precipitation and temperature were the most important environmental elements affecting the distribution of X. fastidiosa