基于MaxEnt模型预测胸窗萤在北京市的潜在适生区  

Prediction of the potential habitats of terrestrial firefly Pyrocoelia pectoralis in Beijing based on MaxEnt model

作　　者：卢林 张志伟[2] 林美英[3] 赵梓伊 肖能文[1] 王琦[1] Lu Lin;Zhang Zhiwei;Lin Meiying;Zhao Ziyi;Xiao Nengwen;Wang Qi(Key Laboratory of National Environmental Protection for Assessment of Regional Ecological Processes and Functions,Chinese Research Academy of Environmental Sciences,Beijing 100012,China;College of Forestry,Shanxi Agricultural University,Taigu 030800,Shanxi Province,China;Institute of Zoology,Chinese Academy of Sciences,Beijing 100101,China;School of Life Sciences,Lanzhou University,Lanzhou 730000,Gansu Province,China)

机构地区：[1]中国环境科学研究院生态研究所,国家环境保护区域生态过程与功能评估重点实验室,北京100012 [2]山西农业大学林学院,太谷030800 [3]中国科学院动物研究所,北京100101 [4]兰州大学生命科学学院,兰州730000

出　　处：《植物保护学报》2022年第4期1217-1224,共8页Journal of Plant Protection

基　　金：生态环境部生物多样性调查评估项目(2019HJ2096001006)。

摘　　要：为明确胸窗萤Pyrocoelia pectoralis在北京市的潜在适生分布情况,利用MaxEnt模型结合25个与胸窗萤生存相关性强的环境因子对其潜在适生区进行分析预测。结果表明:北京市胸窗萤的高适生区面积(适生值>0.5)为2444.42 km^(2),占全市总面积的14.90%。高适生区主要在昌平区的北部和西部,海淀区的西南部,石景山区、门头沟区的东部和南部,房山区的西南部以及平谷区的中部。测试集AUC值为0.977,表明预测结果准确可靠。刀切法分析结果显示最湿季度降雨量、最热季度均温和最湿季度均温对胸窗萤的分布状况影响最大,贡献率分别为29.21%、25.54%和13.56%;夜间灯光和植被类型也有一定的影响,贡献率分别为9.00%和8.16%,累计贡献率为85.45%,通过核密度分析可识别出胸窗萤保护的热点区和空缺区。In order to identify the potential habitats of terrestrial firefly Pyrocoelia pectoralis in Beijing,the potential habitats are predicted by MaxEnt model combined with 25 environmental factors which are strongly related to the survival of P.pectoralis.The results showed that the highly suitable area for growth of P.pectoralis in Beijing(suitable growth value>0.5)was 2444.42 km^(2),accounting for 14.90% of the total area of the city.The highly suitable areas were mainly in the north and west of Changping District,the southwest of Haidian District,the east and south of Shijingshan District and Mentougou District,the southwest of Fangshan District and the middle of Pinggu District.The area under curve value of test set was 0.977,which indicated that the prediction results were accurate and reliable.Jackknife analysis results showed that the precipitation of wettest quarter,mean temperature of wettest quarter and mean temperature warmest quarter had the greatest influence on the distribution of P.pectoralis,with a contribution rates of 29.21%,25.54% and 13.56%,respectively.Night light data and vegetation type also had an influence to some degree,with a contribution rate of 9.00% and 8.16%,respectively,and the cumulative contribution rate was 85.45%.The hot spots and gaps for conservation of P.pectoralis were identified by kernel density analysis.

关 键 词：胸窗萤 MaxEnt模型 环境指示生物 潜在适生区 

分 类 号：Q968[生物学—昆虫学]