基于MaxEnt模型和ArcGIS的杠板归适宜分布区预测

Prediction of Suitable Distribution Area of Polygonum perfoliatum L.Based on MaxEnt Model and ArcGIS

目的采用最大熵(MaxEnt)模型和地理信息系统对杠板归全国适宜分布区进行预测。方法将杠板归476条地理分布信息及55个环境因子导入MaxEnt模型运算,采用刀切法和贡献率筛选主导环境因子,选择受试者工作特征曲线验证模型精确度,通过ArcGIS软件加载模型预测杠板归适宜分布区划图。结果本研究模型预测结果可信度高(AUC>0.9),杠板归的主导环境因子为最湿月降水量、年均降水量、6月份降水量、11月份降水量、最湿季降水量、9月份平均气温和植被亚类,杠板归适宜分布区较广,高度适生区主要为福建、浙江、江西、重庆、广东(除南部)、湖南(除东部)、广西北部、贵州北部、湖北西部、安徽南部、陕西南部、江苏南部。结论本研究预测杠板归的分布区与实际情况基本一致,预测适宜分布区可为开展杠板归野生抚育及合理规划栽培用地提供依据。

Objective To use the maximum entropy(MaxEnt)model and geographic information system to predict the suitable distribution area of Polygonum perfoliatum L.in China.Methods Totally 476 pieces of geographic distribution information and 55 environmental factors were introduced into the MaxEnt model operation.Jackknife test and contribution rate were used to screen the dominant environmental factors,and the receiver operating characteristic curve was selected to verify the accuracy of the model.Finally,the ArcGIS software was used to load the model-predicted suitable distribution area of Polygonum perfoliatum L..Results The prediction results of this model were reliable(AUC>0.9),and the dominant environmental factor was the wetest monthly precipitation quantity,average annual precipitation,monthly precipitation in June,monthly precipitation in November,wetest season precipitation,monthly mean temperature in September and vegetation subclasses.Polygonum perfoliatum L.belongs to a wide range of suitable distribution areas,and the highly suitable areas were mainly Fujian,Zhejiang,Jiangxi,Chongqing,Guangdong(except south),Hunan(except east),northwest Guangxi,northern Guizhou,western Hubei,southern Anhui,southern Shannxi,and southern Jiangsu.Conclusion The predicion of the study for the distribution area of Polygonum perfoliatum L.is basically consistent with the actual situation.The prediction of suitable distribution area can provide a basis for the development of wild breeding and rational planning of cultivated land of Polygonum perfoliatum L.