高级检索
    王艳君, 高泰, 石娟. 基于MaxEnt模型对舞毒蛾全球适生区的预测及分析[J]. 北京林业大学学报, 2021, 43(9): 59-69. DOI: 10.12171/j.1000-1522.20200416
    引用本文: 王艳君, 高泰, 石娟. 基于MaxEnt模型对舞毒蛾全球适生区的预测及分析[J]. 北京林业大学学报, 2021, 43(9): 59-69. DOI: 10.12171/j.1000-1522.20200416
    Wang Yanjun, Gao Tai, Shi Juan. Prediction and analysis of the global suitability of Lymantria dispar based on MaxEnt[J]. Journal of Beijing Forestry University, 2021, 43(9): 59-69. DOI: 10.12171/j.1000-1522.20200416
    Citation: Wang Yanjun, Gao Tai, Shi Juan. Prediction and analysis of the global suitability of Lymantria dispar based on MaxEnt[J]. Journal of Beijing Forestry University, 2021, 43(9): 59-69. DOI: 10.12171/j.1000-1522.20200416

    基于MaxEnt模型对舞毒蛾全球适生区的预测及分析

    Prediction and analysis of the global suitability of Lymantria dispar based on MaxEnt

    • 摘要:
        目的  舞毒蛾是一种食叶性的国际性检疫害虫,给世界上许多国家和地区造成了严重的经济损失。该研究筛选出限制舞毒蛾分布的环境变量,利用MaxEnt软件预测舞毒蛾当前及未来气候条件下的全球适生区范围,明确舞毒蛾在不同气候条件下的适生区变化。
        方法  利用ArcGIS软件设置缓冲区筛选舞毒蛾在全球的分布点数据;利用MaxEnt、SPSS和ArcGIS软件根据环境变量贡献率、刀切法和变量相关性分析对生物气候变量、月总降水量、月平均最高温度和月平均最低温度4种环境变量进行筛选;利用R软件计算调控倍频和特征组合等因子调整MaxEnt模型参数;利用MaxEnt模型预测当前和未来不同情境条件下舞毒蛾全球适生区的分布范围。
        结果  经过缓冲区筛选得到734个舞毒蛾的分布点数据;MaxEnt模型结果中,测试遗漏率与理论遗漏率吻合度高,而且模型AUC值为0.940;MaxEnt模型预测当前条件下舞毒蛾在全球的高、中度适生区主要集中在欧洲的大部分地区,北美洲中东部,亚洲的东西部,而非洲、大洋洲和南美洲分布较少。此外,舞毒蛾在未来气候条件下北半球适生区的边界向北偏移,北美洲以及欧亚大陆的高、中度适生区的面积扩增明显。
        结论  舞毒蛾的分布受多种环境变量影响,并且温度和降水与舞毒蛾的特定发育阶段相吻合。MaxEnt模型的预测结果能够反映舞毒蛾在全球的分布特征。该研究可为防治舞毒蛾和制定相关检疫措施提供理论依据。

       

      Abstract:
        Objective  Lymantria dispar (Linnaeus) is a leaf-eating international quarantine pest, which has caused serious economic losses to many countries and regions in the world. In this study, environmental variables limiting the distribution of Lymantria dispar were screened out, and MaxEnt software was used to predict the global adaptation range of Lymantria dispar under current and future climatic conditions, so as to clarify the adaptation range of Lymantria dispar under different climatic conditions.
        Method  We used ArcGIS software to set up buffer zones to screen the global distribution data of Lymantria dispar. We used MaxEnt, SPSS and ArcGIS software to screen four kinds of environmental variables, i.e. biological climate variable, monthly total precipitation, monthly mean maximum temperature and monthly mean minimum temperature, according to the contribution rate of environmental variables, knife cutting method and variable correlation analysis. We used R software to calculate the regularization multiplier and feature classes and other factors to adjust the parameters of MaxEnt model. We used MaxEnt model to predict the global distribution range of Lymantria dispar under current and future climatic conditions.
        Result  The distribution data of 734 Lymantria dispar were obtained by buffer screening. In the MaxEnt model results, the test omission rate was in high agreement with the theoretical omission rate, and the AUC value of the model was 0.940. MaxEnt model predicted that Lymantria dispar was mainly distributed in most parts of Europe, central and eastern North America, western and eastern Asia, and less in Africa, Oceania and South America under current conditions. In addition, Lymantria dispar boundaries in the northern hemisphere will shift to the north under future climatic conditions, and the areas of high and moderate adaptability in North America and Eurasia will expand significantly.
        Conclusion  The distribution of Lymantria dispar is influenced by a variety of environmental variables and temperature and precipitation are consistent with the specific development stage of Lymantria dispar. The predicted results of MaxEnt model can reflect the distribution characteristics of Lymantria dispar in the world. This study can provide theoretical basis for prevention and control of Lymantria dispar and establishment of relevant quarantine measures.

       

    /

    返回文章
    返回