基于无人机遥感的天然草原鼠害发生面积调查方法研究  

Monitoring of rodent damage areas in grassland using unmanned aerial vehicle remote sensing technology

作　　者：花蕊 包达尔罕 董瑞 唐庄生 楚彬[1] 郝媛媛 花立民[1] HUA Rui;BAO Daerhan;DONG Rui;TANG Zhuang-sheng;CHU Bin;HAO Yuan-yuan;HUA Li-min(College of Prataculture Science,Gansu Agricultural University,Key Laboratory of Grassland Ecosystem of the Ministry of Education,Engineering and Technology Research Center for Alpine Rodent Pest Control,National Forestry and Grassland Administration,Lanzhou 730070,China)

机构地区：[1]甘肃农业大学草业学院,草业生态系统教育部重点实验室,国家林业草原高寒草地鼠害防控工程技术研究中心,甘肃兰州730070

出　　处：《草业学报》2023年第5期71-82,共12页Acta Prataculturae Sinica

基　　金：甘肃高寒草地鼠害非药物环境友好型防控技术与产品研发示范项目(2021CYZC-05);甘肃省教育厅优秀研究生“创新之星”项目(2021CXZX-343);甘肃省高等学校青年博士基金项目(2021QB-028);甘肃农业大学公招博士项目(GAU-KYQD-2020-19);甘肃农业大学伏羲青年英才项目“草地鼠害无人机监测技术研发”(GAUfx-04Y06)资助。

摘　　要：草原鼠害是目前我国为害面积最大、影响最为广泛的草原生物灾害。掌握草原鼠害发生面积是评估其为害程度和制定防治策略的重中之重。当前草原鼠害发生面积调查采用填图法或勾绘法,这两种方法均存在效率低、准确性差的问题,难以满足草原鼠害精准管理的要求。本研究以生态学和统计学理论为基础,采用无人机低空遥感技术,以青藏高原优势害鼠高原鼠兔为测试对象,以甘肃玛曲县4个乡镇为测试区域展开天然草地鼠害发生面积调查。1)首先确定被调查鼠种的一级调查区,即害鼠适生区;2)在一级调查区依据不同栖息地类型作为二级调查样地并划分为网格;3)在二级样地中随机抽样一定数量的样方,作为三级样方用于无人机调查;4)依据多个样方平均鼠害发生面积统计出二级样地的草原鼠害发生面积,汇总各二级样地发生面积获得调查区的鼠害发生总面积。结果显示,测试区域草原总面积4.6×10^(5)hm^(2),其中高原鼠兔适生区面积2.5×10^(5)hm^(2),经抽样调查得到研究区高原鼠兔为害发生面积为1.05×10^(4)hm^(2)。本测试区域共有50个1 hm^(2)样地,以调查1个样地为例,1人需20 min左右完成内外业工作。相比填图法或勾绘法,本方法高效便捷、结果精度高,可为大尺度草原鼠害发生面积调查提供技术支持。Rodent pests are the greatest and most widespread biological problem affecting pastoral land in China and their impact can reach‘disaster’level.Obtaining accurate data for rodent-damaged areas in grassland is a priority for evaluating the degree of pest damage and for devising pest management strategies.At present,survey grid and mapping methods are the main methods used to identify rodent-damaged areas in grassland.However,both of those methods have problems in terms of low efficiency and poor accuracy,so they do not meet the requirements for precise management of grassland pests.According to the principles of ecology and statistics,we used unmanned aerial vehicle(UAV)remote sensing technology and a stratified sampling method to detect damage areas from plateau pika,a dominant rodent pest on the Qinghai-Tibetan Plateau.These trials were conducted at four townships in Maqu County in Gansu Province.First,we selected the primary investigation area according to its suitability for rodents.Then,the first survey area was divided according to different habitat types into secondary survey plots,and each plot was divided into a grid.A certain number of quadrats in the secondary sample plots was randomly sampled,each as a three-level sample with a UAV survey.According to the average area of rodent pest damage across multiple quadrats,the damaged area in each secondary plot was calculated.The areas of damage in all the secondary plots were added to obtain the total pest-damaged area in the primary investigation area.The total area of grassland in the test area was 4.6×10^(5) ha,the area suitable for plateau pika was 2.5×10^(5)ha,and the area of plateau pika damage in the test area was 1.05×10^(4) ha.There were 50 sampling sites in total,and each site had an area of 1 ha.Taking one site as a sample,it takes one person 20 min to complete the work in terms of UAV flying and image interpretation.Compared with the grid survey method and mapping method,this new method can detect areas of rodent damage in grassland on a la

关 键 词：草原鼠害 发生面积 无人机 分层抽样 

分 类 号：S812.6[农业科学—草业科学] TP751[农业科学—畜牧学]