基于气候和寄主因素的栗山天牛中国成灾和扩散风险评估  

Risk Analysis of Dispersal and Outbreak of Massicus raddei(Coleoptera:Cerambycidae)in China Based on Climate and Host Distribution

作　　者：张宇凡 党英侨 王小艺[1] Zhang Yufan;Dang Yingqiao;Wang Xiaoyi(Key Laboratory of Forest Protection of National Forestry and Grassland Administration Ecology and Nature Conservation Institute,Chinese Academy of Forestry,Beijing 100091;Yunfu Forestry Science and Technology Promotion Center,Guangdong Province,Yunfu 527300)

机构地区：[1]中国林业科学研究院森林生态环境与自然保护研究所国家林业和草原局森林保护学重点实验室,北京100091 [2]广东省云浮市林业科学和技术推广中心,云浮527300

出　　处：《林业科学》2022年第6期95-109,共15页Scientia Silvae Sinicae

基　　金：十三五重点研发计划项目(2018YFC1200400);中国林业科学研究院基础研究项目(CAFYBB2018ZB001)。

摘　　要：【目的】明确不同喜食程度的寄主植物对栗山天牛局部成灾和未分布区扩散定殖的潜在影响,为局部成灾区生态防控和未发生区早期预警提供理论指导。【方法】通过文献整理、数据库检索与实地调查,获得我国137个栗山天牛分布点。利用ArcGIS 10.0从WorldClim下载1970—2000年的19个环境变量,从中筛选出相关系数<∣0.9∣且对栗山天牛潜在分布概率贡献率较大的8个气候变量,将其转换为Maxent需要的ASCII格式数据。通过数据库检索和进一步筛选获得我国2548个栗山天牛寄主分布点,基于该虫对我国各县分布的寄主树种的喜食程度,对栗山天牛在每个县内的适生程度进行赋值,并采用ArcGIS获得寄主树种风险图的ASCII格式数据。在Maxent模型中依次导入栗山天牛分布点和环境变量文件、栗山天牛分布点和寄主树种风险图文件,分别将重复运行10次的平均结果作为仅基于气候因素或寄主分布的栗山天牛潜在适生区。采用ArcGIS将以上2个潜在适生区进行叠加获得同时考虑寄主和气候因素的栗山天牛潜在适生区,并通过实际调查点检验所有预测结果的准确性。3个潜在适生区均根据适宜指数分为4类,即非适生区、低适生区、中适生区和高适生区,计算3个适生区预测结果中的4个适生等级的面积并作图。【结果】在当前气候条件下,栗山天牛的高适生区集中在吉林、辽宁、北京、河北、天津和浙江北部,适生区范围未包含在云南省实际调查中所获得的分布点。在当前寄主分布情况下,栗山天牛对我国南北地区的寄主种类在喜食程度上存在显著差异,北部地区分布易感树种蒙古栎和辽东栎,该虫表现为高度适生性,而在南部地区分布其他次要偏嗜危害的寄主种类,该虫表现为中度适生性和低度适生性。同时考虑气候和寄主因素预测出的适生区更符合栗山天牛现有分布情况,即适生区�【Objective】Massicus raddei is widely distributed in eastern Asia,but the insects only outbreak in regional oak forests in Jilin and Liaoning provinces of China.This study aims to ascertain potential effects of host plants with different preference levels on the spread and outbreak of M.raddei in China,so as to provide theoretical guidance for ecological management in outbreak regions and early warning in distribution-free regions.【Method】A total of 137 distributional points of M.raddei in China were obtained based on literature,database search,and field investigation.Eight climate variables with correlation coefficients<|0.9|and relatively larger contribution rate to the potential distribution of M.raddei were screened out from 19 environmental variables downloaded from WorldClim between 1970 and 2000 by ArcGIS 10.0.Then the data were converted into the ASCII format data required for Maxent.Through database search and further artificial screening,2548 host distributional points of M.raddei were obtained in China.Based on the feeding preference degree of host species distributed in every counties of China,the suitability degrees of M.raddei in each county were assigned as specific values,and then the ASCII format data of risk map of hosts was obtained by using ArcGIS.In the MaxEnt model,two files of distribution points of M.raddei and environmental variables and two files of distribution points of M.raddei and the risk map of hosts were successively imported,and the average predictive results of repeated 10 runs by MaxEnt were regarded as the potential suitable areas of M.raddei based on only climate factors or host distribution,respectively.By applying ArcGIS,aforementioned two layers were overlapped to obtain a new potential distribution combining climate with host distribution,and then the precision of the three predicted models was all evaluated through actual distribution points obtained from field surveys.In the three predicted results,the habitat of M.raddei in China was divided into four categories

关 键 词：栗山天牛 Maxent模型 易感树种 局部成灾 适生区 

分 类 号：S718.8[农业科学—林学]