基于空间矩阵模型及0~1测度的美国白蛾风险格局分析  

Risk Pattern Analysis of Hyphantria cunea Based on Spatial Matrix Model and 0-1 Measure

作　　者：叶江霞[1,2] 王敬文 张明莎 周汝良[2] 石雷[1] Ye Jiangxia;Wang Jingwen;Zhang Mingsha;Zhou Ruliang;Shi Lei(Institute of Resource Insects,Chinese Academy of Forestry,Kunming 650224;Southwest Forestry University,Kunming 650233)

机构地区：[1]中国林业科学研究院资源昆虫研究所,昆明650224 [2]西南林业大学,昆明650233

出　　处：《林业科学》2021年第1期140-152,共13页Scientia Silvae Sinicae

基　　金：国家自然科学基金项目(31760212);国家自然科学基金项目(31860213)。

摘　　要：【目的】分析美国白蛾在中国的扩散风险格局,为其防治和风险管理提供精准的空间位置和技术支持。【方法】收集全国2011年至2016年美国白蛾的乡镇检疫数据,进行寄主、气象要素、地理环境、人为影响变量的空间模拟。将影响有害生物发生发展的因子及风险描述为空间栅格点上的连续变量,利用GIS(地理信息系统)的空间矩阵模型表达变量,并借助SPSS筛选和建立0~1测度的多变量Logistic概率模型,以250 m为最小空间栅格,描述全国任意一个空间地理单元发生扩散的风险概率。【结果】除生物气候变量外,人流物流变量的影响显著,高风险位于东部农区、建设用地及人工植被区域,主要集中在辽宁、北京、天津、上海、河北、山东、河南、安徽、湖北、江苏、陕西等省市,并有扩散到吉林、内蒙、湖南、江西、新疆、宁夏等省区的趋势。【结论】利用矩阵模型及0~1概率化测度描述有害生物发生及扩散风险,对基层开展检疫和防治具明显的指导作用。引入人流物流扩散影响模拟变量,可有效提高测报预警精度。人为活动密集区既是疫区,又是传播通道。山地森林系统的自然植被对传播扩散具有阻隔作用,加强高风险区的近自然林修复和建设,加大重要通道的检疫,对防控具有重要意义。【Objective】This study aimed at providing spatial guidance for prevention and control of risk management,and the factors and risks affecting the occurrence and spread of pests were described as continuous variables on the spatial grid points to improve the precision of measurement and warning.The raster model of GIS(geographic information system)was used to express the variables in digital matrix and spatial unit,and the multivariate logistic probability model with 0-1 measure was established by SPSS to present the risk of pest occurrence.The flowchart was illustrated by taking Hyphantria cunea as an example.【Method】Based on the national quarantine data in township of Hyphantria cunea during 2011 and 2016,simulation of host,meteorological factors,biogeographic environment,and main factors impact variables was carried out to analyze the risk pattern of occurrence and spread of Hyphantria cunea.Taking 250 m as the minimum spatial grid,the risk probability of any spatial geographic unit is described.【Result】Besides bioclimates,the impact of human mobility and logistics variables is significant in risk assessment.The high risk is located in the eastern farmland and construction land of man-made vegetation regions,mainly in Liaoning,Beijing,Tianjin,Shanghai,Hebei,Shandong,Henan,Anhui,Hubei,Jiangsu,Shanxi and other provinces and cities,and has the trend of spreading to Jilin,Inner Mongolia,Hunan,Jiangxi,Xinjiang,Ningxia and other provinces.【Conclusion】The matrix model and 0-1 probabilistic measure are used to describe the risk of occurrence and diffusion of pests.The simulation of human impacts spread can effectively improve the accuracy of measurement and warning.The human activity intensive region is not only the epidemic area,but also the transmission path.The natural vegetation of the mountain forest system has a blocking effect on the propagation and diffusion.It is of great significance for the prevention and control to strengthen the restoration and construction of near-natural forests in high-ris

关 键 词：空间矩阵 风险分析 美国白蛾 人流物流 概率化测报 

分 类 号：Q958.15[生物学—动物学] S154.5[农业科学—土壤学]