基于最大熵模型(Max Ent)和地理信息系统(GIS)的四川省眉山市东坡区晚熟柑橘适宜性评价  

Suitability Evaluation of Late Maturing Citrus in Dongpo District,Meishan City,Sichuan Province Based on Maximum Entropy Model(MaxEnt)and Geographic Information System(GIS)

作　　者：姚莉[1] 邓春秀 唐彪 王宏[1] 吴月颖 张奇[1] 李源洪[2] 刘忠友[2] 卢昌艾[4] 林超文[1] YAO Li;DENG ChunXiu;TANG Biao;WANG Hong;WU YueYing;ZHANG Qi;LI YuanHong;LIU ZhongYou;LU ChangAi;LIN ChaoWen(Agricultural Resources and Environment Institute,Sichuan Academy of Agricultural Sciences,Chengdu 610066;Institute of Remote Sensing and Digital Agriculture,Sichuan Academy of Agricultural Sciences,Chengdu 610066;General Station of Arable Soil Quality and Fertilizer of Sichuan Province,Chengdu 610041;Institute of Agricultural Resources and Regional Planning,Chinese Academy of Agricultural Sciences,Beijing 100081)

机构地区：[1]四川省农业科学院农业资源与环境研究所,成都610066 [2]四川省农业科学院遥感与数字农业研究所,成都610066 [3]四川省耕地质量与肥料工作总站,成都610041 [4]中国农业科学院农业资源与农业区划所,北京100081

出　　处：《中国农业科学》2025年第4期748-758,共11页Scientia Agricultura Sinica

基　　金：国家重点研发计划(2024YFD1700200,2022YFD1901405);四川省财政自主创新专项项目(2023ZZCX013);“1+9”揭榜挂帅科技攻关项目(1+9KJGG005)。

摘　　要：【目的】开展四川省眉山市东坡区晚熟柑橘适宜性评价,可为当地晚熟柑橘合理种植、优化生产空间提供科学参考。【方法】基于14个气候指标、3个地形指标和22个土壤理化性状指标,以及372个晚熟柑橘种植区采样点数据,利用最大熵模型(MaxEnt)和地理信息系统(GIS)结合相关性分析筛选影响四川省眉山市东坡区晚熟柑橘分布的关键环境因子,构建评价晚熟柑橘种植适宜性模型。【结果】MaxEnt模型计算效果非常好,受试者特征工作曲线(ROC曲线)中的下面积值(AUC)达0.843,具有较高的准确性与可靠性。MaxEnt分析表明,年平均气温、海拔、全磷、pH、年平均相对湿度、粉粒量、年降水量、年日照时数、秋季降水量、土壤交换性钙、全钾和有机质是影响晚熟柑橘分布的主要环境因子。影响柑橘分布的气候变量累计贡献率为43.0%,其中,年平均气温(22.9%)、年平均相对湿度(7.0%)和年降水量(4.6%)是影响晚熟柑橘分布的3个最主要的气候因子;影响晚熟柑橘分布的土壤变量累计贡献率为41.3%,其中,全磷(13.8%)、pH(10.4%)和粉粒量(5.7%)是影响晚熟柑橘分布的3个最主要的土壤因子;海拔是影响柑橘分布的地形因子,贡献率为15.7%。通过统计分析各主导环境因子在不同等级分布区的参数值,结合各生态因子响应曲线、实地调查情况以及查阅文献,综合分析得出晚熟柑橘适宜分布的基本气候和地形环境要求为:年平均气温17℃左右,年平均相对湿度80%,年降雨量1300 mm左右,年日照时数1350 h,秋季降雨量390 mm,海拔400—700 m;土壤pH为4.5—7.5,土壤养分状况受立地条件和栽培管理措施的综合影响。全区34673.29 hm~2的宜园土地中,优势区、适宜区和不适宜区分别占23.51%、61.37%和15.12%。【结论】客观地反映了在气候、地形和土壤因素影响下,东坡区晚熟柑橘种植适宜性变化特征、适宜性面积及分布。【Objective】Carrying out the suitability evaluation of late-maturing citrus in Dongpo District,Meishan City,can provide scientific references for the reasonable cultivation and optimization of production space of local late-maturing citrus.【Method】Based on 14 climatic indicators,3 topographic indicators and 22 soil physicochemical indicators,as well as data from 372 sampling points of late-maturing citrus planting areas,the article utilizes the Maximum Entropy Model(MaxEnt)and Geographic Information System(GIS)in combination with correlation analysis to screen key environmental factors affecting the distribution of late-maturing citrus in Dongpo District,Meishan City,Sichuan Province,and constructs a model for evaluating the suitability of late-maturing citrus planting.The model was constructed to evaluate the suitability of late-maturing citrus planting.【Result】The MaxEnt model was very effective,and the area under the curve(AUC)in the characteristic working curve(ROC curve)of the subjects reached 0.843,which was highly accurate and reliable.The MaxEnt analysis showed that the average annual air temperature,elevation,total phosphorus,pH,average annual relative humidity,pulverization,annual precipitation,annual sunshine hours,autumn precipitation,soil exchangeable calcium,total potassium and organic matter were the main environmental factors affecting the distribution of late-maturing citrus;the cumulative contribution of climatic variables affecting the distribution of citrus was 43.0%,of which mean annual temperature(22.9%),mean annual relative humidity(7.0%)and annual precipitation(4.6%)were the three most important climatic factors affecting the distribution of late-maturing citrus;the cumulative contribution of soil variables affecting the distribution of late-maturing citrus was 41.3%,of which total phosphorus(13.8%),pH(10.4%)and chalkiness(5.7%)were the three most important soil factors affecting the distribution of late-maturing citrus;elevation was a topographic factor affecting the distributi

关 键 词：晚熟柑橘 最大熵模型 地理信息系统 适宜性评价 

分 类 号：S666[农业科学—果树学] S127[农业科学—园艺学]