基于Cubist模型的天山北坡草地鼠群密度时空分布特征  

Spatiotemporal distribution of rat population density in grassland on the north slope of Tianshan Mountains based on Cubist model

作　　者：潘群 施海洋 张文强 罗格平[2,3,4] 陈春波 PAN Qun;SHI Haiyang;ZHANG Wenqiang;LUO Geping;CHEN Chunbo(College of Resources and Environmental Sciences,Xinjiang University,Urumqi 830046,Xinjiang,China;State Key Laboratory of Desert and Oasis Ecology,Xinjiang Institute of Ecology and Geography,Chinese Academy of Sciences,Urumqi 830011,Xinjiang,China;University of Chinese Academy of Sciences,Beijing 100049,China;Central Asia Center for Ecology and Environmental Research Center,Chinese Academy of Sciences,Urumqi 830011,Xinjiang,China)

机构地区：[1]新疆大学资源与环境科学学院,新疆乌鲁木齐830046 [2]中国科学院新疆生态与地理研究所荒漠与绿洲生态国家重点实验室,新疆乌鲁木齐830011 [3]中国科学院大学,北京100049 [4]中国科学院中亚生态与环境研究中心,新疆乌鲁木齐830011

出　　处：《干旱区地理》2022年第4期1200-1211,共12页Arid Land Geography

基　　金：高端外国专家项目(E0600101);2020年自治区创新环境(人才、基地)建设专项(天山青年计划)(2020Q084);2020年第一批中央财政林业草原生态保护恢复资金项目(XJCYZZ202001)资助。

摘　　要：鼠害是影响草原生态健康的重要因素,了解小型啮齿动物种群密度时空分布特征,对精准的鼠害综合防治具有重要意义。以往对小型啮齿动物时空分布的研究多局限于静态的站点分布或小范围的种群密度时间变化分析,缺乏对较大时空尺度小型啮齿动物种群密度变化的分析。从已发表的文献中收集了天山北坡草地1982—2015年小尺度的有效洞口密度实地调查信息,同时结合环境因子数据,再根据海拔将研究区划分为≤900 m和>900 m 2类,运用Cubist模型和随机森林模型,分析有效洞口密度时空分布。结果表明:(1)1982—2015年天山北坡海拔≤900 m地区的有效洞口密度总体呈增加趋势,而海拔>900 m的地区总体呈减少趋势。基于Cubist模型构建有效洞口密度与环境因子的模型拟合精度明显优于随机森林模型。(2)植被状况、气象因子和放牧强度是天山北坡有效洞口密度时空分布主要的环境驱动因素。在天山北坡内海拔≤900 m和>900 m的地区中,有效洞口密度的驱动机制存在着显著差异。(3)在海拔≤900 m地区,影响有效洞口密度时空分布主要是叶面积指数,而对于海拔>900 m地区为归一化植被指数。这可能是受到大沙鼠(Rhom⁃bomys opimus)和黄兔尾鼠(Eolagurus luteus)消耗不同类型植被的影响。Rodent damage is an important factor affecting grassland ecological health.Understanding the spatiotemporal distribution characteristics of small rodent population density is crucial for accurate comprehensive rodent control.Previous studies on the spatiotemporal distribution of small rodents were mostly limited to static site distribution or time change analysis of small-scale population density;investigation on the population density change of small rodents on a large temporal and spatial scale is lacking.In this study,the field survey information of small-scale active burrow entrance densities on the grassland on the north slope of Tianshan Mountains from 1982 to 2015 was collected from the literature and combined with environmental variable data.The study area was divided as areas with elevation of≤900 m and>900 m.The spatiotemporal distribution of active burrow entrance densities was analyzed using the Cubist and random forest(RF)models.The following results were obtained.(1)From 1982 to 2015,the active burrow entrance densities increased in the area with an elevation of≤900 m on the north slope of Tianshan Mountains,whereas that in the area with an elevation of>900 m decreased.The fitting accuracy of the Cubist model of active burrow entrance densities and environmental factors was better than that of the RF model.(2)Vegetation status,meteorological factors,and grazing intensity were found to be the primary environmental driving factors for the spatiotemporal distribution of active burrow entrance densities on the north slope of Tianshan Mountains.Significant differences in the driving mechanism of active burrow entrance densities were found between the areas with an elevation of≤900 m and those with an elevation of>900 m on the north slope of Tianshan Mountains.(3)The main factor affecting the spatiotemporal distribution of active burrow entrance densities was the leaf area index and normalized difference vegetation index for the areas with elevation≤900 m and>900 m,respectively.This difference may

关 键 词：有效洞口密度 环境因子 Cubist模型 随机森林模型 天山北坡 

分 类 号：S812.6[农业科学—草业科学]