基于MaxEnt模型的大果圆柏生境适宜性评价  

Prediction of potentially suitable distribution areas for Juniperus tibetica based on MaxEnt model

作　　者：王义贵 何学高 胡云云 王含予 文妙霞 刘超 丁鹏 WANG Yigui;HE Xuegao;HU Yunyun;WANG Hanyu;WEN Miaoxia;LIU Chao;DING Peng(Northwest Surveying and Planning Institute of National Forestry and Grassland Administration,Key Laboratory National Forestry and Grassland Administration on Ecological Hydrology and Disaster Prevention in Arid Regions,Xi’an 710048,Shaanxi,China;Xi’an green ring forestry technical service co,Xi’an 710048,Shaanxi,China)

机构地区：[1]国家林业和草原局西北调查规划院旱区生态水文与灾害防治国家林业和草原局重点实验室,陕西西安710048 [2]西安绿环林业技术服务公司,陕西西安710048

出　　处：《中南林业科技大学学报》2023年第7期41-51,共11页Journal of Central South University of Forestry & Technology

基　　金：国家林业和草原局研发计划项目“青海柏树生态系统有机碳空间分异特征研究”(LC-2-12);国家林业和草原局科技项目“黄河流域林草生态资源专题监测技术研究”(LC-2-02)。

摘　　要：【目的】大果圆柏是我国特有树种,探讨限制大果圆柏地理分布的主导环境变量,模拟大果圆柏潜在适宜分布区,可为大果圆柏资源保护恢复和人工林培育提供理论参考。【方法】利用ENMTools筛选得到大果圆柏122个分布点数据,应用MaxEnt模型比较采用单一气候变量(C),气候变量和地形变量(C+T),气候、地形和水文变量(C+T+H)及气候、地形、水文、土壤4类综合环境变量(C+T+H+S)模拟结果的准确性与可靠性,探讨限制大果圆柏地理分布的环境变量。【结果】基于单一气候变量、2类、3类和4类综合环境变量的MaxEnt模型AUC值训练集分别为0.976±0.001、0.976±0.001、0.979±0.001、0.978±0.001,验证集分别为0.972±0.012、0.968±0.010、0.969±0.010、0.967±0.008,表明4类综合环境变量模拟模型稳定性最好,可信度高。大果圆柏地理分布主要受气温(昼夜温差与年温差比值、近地表气温和温度季节变化)、降水(年平均降水量)、地形(海拔、坡向)、土壤(营养保留能力)等4类综合环境变量影响;基于单一气候变量(C)模拟的大果圆柏适宜分布面积到4类综合环境变量(C+T+H+S)模拟结果呈递减趋势;基于MaxEnt模型模拟的大果圆柏适生区主要集中在我国在西藏(南部和东部)、四川(北部、西北部及西部)、青海南部、甘肃甘南地区、云南北部。【结论】地形、温度、降水、土壤多种环境因素综合影响着大果圆柏的地理分布格局。4类综合环境变量(C+T+H+S)模拟可提高模型预测的准确性和反映大果圆柏的地理分布。【Objective】Juniperus tibetica is an indigenous tree in China.Exploring the dominant environmental variables that limit the distribution of J.tibetica and simulating the potentially suitable distribution areas of J.tibetica can provide guidelines for the conservation and restoration of J.tibetica resources and plantation cultivation.【Method】ENMTools was used to screen the data of 122 distribution points of J.tibetica.The Maximum Entropy(MaxEnt)model was used to compare the accuracy and reliability of the simulation results of 4 comprehensive environmental variables:single climate variable(C),climate variable and terrain variable(C+T),climate,topographic and hydrological variables(C+T+H),as well as climate,topography,hydrology and soil(C+T+H+S).The dominant environmental variables limiting the geographical distribution of J.tibetica were discussed.【Result】The area under the receiveroperating characteristic curve(AUC)showed that the training data of the 4 comprehensive environmental variables were 0.976±0.001,0.976±0.001,0.979±0.001 and 0.978±0.001,respectively.In comparison,the test data was 0.972±0.012,0.968±0.010,0.969±0.010 and 0.967±0.008,respectively.These results indicated that the simulation model of the 4 types of comprehensive environmental variables had satisfactory stability and high reliability.The geographical distribution of the species was affected by Climate(Isothermality(Bio2/Bio7)(×100)),Near-surface air temperature,Temperature Seasonality(standard deviation×100)and Annual Precipitation.Topography(Elevation and Slope)and Soil(Nutrient retention capacity)also affect the geographical distribution.The decreasing trend of the suitable distribution area of J.tibetica was found based on the simulation of the single climate variable(C)and the simulation results of 4 types of comprehensive environmental variables(C+T+H+S).Based on the MaxEnt model simulation,The suitable distribution area of J.tibetica was mainly concentrated in Tibet(south and east),Sichuan(north,northwest and west)

关 键 词：大果圆柏 MAXENT 环境变量 适宜分布区 

分 类 号：S791.44[农业科学—林木遗传育种]