基于MaxEnt和ArcGIS模型的山海螺生长适宜区预测分析  

Prediction and Analysis of Suitable Growth Area of Codonopsis lanceolata Based on MaxEnt and ArcGIS

作　　者：庄怡雪 李敏[1] 陆一南 徐水宇 陈懿冲 陈艳柠 杨成梓[1,2] ZHUANG Yi-xue;LI Min;LU Yi-nan;XU Shui-yu;CHEN Yi-chong;CHEN Yan-ning;YANG Cheng-zi(College of Pharmacy,Fujian University of Traditional Chinese Medicine,Fuzhou 350122,China;Fujian Quality Monitoring and Technology Service Center for Chinese Medicine Raw Materials,Fuzhou 350122,China)

机构地区：[1]福建中医药大学药学院,福建福州350122 [2]福建省级中药原料质量监测技术服务中心,福建福州350122

出　　处：《中药材》2022年第10期2335-2339,共5页Journal of Chinese Medicinal Materials

基　　金：2019年中医药公共卫生服务补助专项“全国中药资源普查项目”(财社[2019]39号)

摘　　要：目的:利用山海螺中国分布点的205条地理分布信息及55个环境因子,采用最大熵(MaxEnt)模型结合地理信息系统(ArcGIS)对山海螺的全国适宜性分布区进行预测。方法:采用刀切法(Jackknife)与贡献率筛选主导环境因子,选用受试者工作特征曲线(ROC曲线)验证模型的精确度,最后采用ArcGIS软件运算预测山海螺的全国适宜性分布区。结果:MaxEnt模型预测精度较高(AUC>0.9),符合预测要求。综合分析得出影响山海螺生长的主导因子共5个,均为降水量相关因子;山海螺的全国适宜性分布区域与中国植物志的记载稍有不同,最适宜生长地区主要为分布于东北地区的辽宁省,华东地区的福建省、江西省、浙江省、安徽省以及中南地区的湖南省、广东省和广西壮族自治区。中国植物志未记载的西南地区亦可有分布,为低适宜生长区。结论:该结果对山海螺扩大种植面积和提高栽培技术等实践工作具有指导意义。Objective: To predict the national suitability distribution area of Codonopsis lanceolata by using the 205 geographic distribution information and 55 environmental factors of its distribution points in China through the maximum entropy(MaxEnt) model combined with geographic information system(ArcGIS).Methods: The Jackknife and contribution rate were used to screen the dominant environmental factors, the receiver operating characteristic curve(ROC curve) was selected to verify the accuracy of the model, the ArcGIS software was used to calculate the national suitability distribution area of Codonopsis lanceolata.Results: The prediction accuracy of MaxEnt model was high(AUC >0.9),which met the prediction requirements.The results of comprehensive analysis showed that there were five dominant factors affecting the growth of Codonopsis lanceolata,all of which were precipitation related factors.The national suitable distribution area of Codonopsis lanceolata was slightly different with the recorded in Flora of China.The most suitable growth area was mainly distributed in Liaoning province in northeast China, Fujian province, Jiangxi province, Zhejiang province and Anhui province in east China, Hunan province, Guangdong province and Guangxi Zhuang autonomous region in central south China.It could also be found in southwest China where the Flora of China was not recorded, it was a low suitable growing area.Conclusion: The results can provide guidance for the practice of expanding the planting area and improving the cultivation technology of Codonopsis lanceolata.

关 键 词：山海螺 最大熵模型 ARCGIS 适宜性分布区 生态因子 

分 类 号：R282.71[医药卫生—中药学]