摘要
利用油茶象的分布点数据和环境因子数据,通过MaxEnt生态位模型预测油茶象在中国的潜在分布区域。结果表明:油茶象的潜在分布区主要集中于我国南方省份;在ArcGIS中进行显示与风险等级划分,按栅格数值的大小将油茶象的适生范围分为4级,分级标准为:高适生区、中适生区、低适生区、非适生区,经ROC曲线分析法验证,MaxEnt生态位模型的AUC值为0.979(非常接近1),表明预测获得了较好的效果。
Maximum Entropy（MaxEnt） was used to predict potential distribution of Curculio chinensis Cheveolat in China based on the Curculio chinensis distribution points datas and environmental factors datas.The results showed that the suitable areas for Curculio chinensis Cheveolat infestations were south areas of China.With the revelation in ArcGIS,the grades were made and the areas for Curculio chinensis Cheveolat infestations were divided into four grades： the highly suitable areas;the moderate suitable areas;the low suitable areas and the unsuitable areas.The results were proved by ROC analytical method and the AUC value of MaxEnt was 0.979 which close to 1 and it meaned that the prediction was highly dependable.
出处
《福建林业科技》
2011年第4期55-58,共4页
Journal of Fujian Forestry Science and Technology
基金
云南省重点学科建设项目(XKZ200905)
