南京林业大学学报(自然科学版) ›› 2015, Vol. 58 ›› Issue (01): 6-10.doi: 10.3969/j.issn.1000-2006.2015.01.002

• 专题报道 • 上一篇    下一篇

基于Maxent生态位模型的松材线虫在中国的适生区预测分析

韩阳阳1,2,王 焱2,项 杨1,叶建仁1*   

  1. 1.南方现代林业协同创新中心,江苏省有害生物入侵预防与控制重点实验室,南京林业大学林学院,江苏 南京 210037;
    2.上海市林业总站,上海 200072
  • 出版日期:2015-01-31 发布日期:2015-01-31
  • 基金资助:
    收稿日期:2014-11-03 修回日期:2014-12-08
    基金项目:上海市绿化管理局2011年科研项目(JB119927)
    第一作者:韩阳阳,硕士生。*通信作者:叶建仁,教授。E-mail: jrye@njfu.edu.cn。
    引文格式:韩阳阳,王焱,项杨,等. 基于Maxent生态位模型的松材线虫在中国的适生区预测分析[J]. 南京林业大学学报:自然科学版,2015,39(1):6-10.

Prediction of potential distribution of Bursaphelenchus xylophilus in China based on Maxent ecological niche model

HAN Yangyang1,2, WANG Yan2, XIANG Yang1, YE Jianren1*   

  1. 1.Collaborative Innovation Center of Sustainable Forestry in Southern China, Jinangsu Province Key Laboratory for Prevention and Management of Invasive Species, College of Forestry, Nanjing Forestry University, Nanjing 210037, China;
    2.Shanghai Forestry Station, Shanghai 200072,China
  • Online:2015-01-31 Published:2015-01-31

摘要: 为了进一步掌握松材线虫在中国可能的扩展范围和可能造成的危害程度,利用Maxent生态位模型,结合地理信息系统ArcGIS,对松材线虫在中国的适生区范围进行了预测研究。结果表明:松材线虫在中国的适生区主要集中在华东和华南地区,华北地区南部、东北地区南部和西南地区南部则是松材线虫的中适生区或低适生区,其他地区则是松材线虫的非适生区。在适生区范围内,只要可感染的寄主松树存在,其他气候因子是可能满足松材线虫的侵染和致病的。ROC评价(receiver operating characteristic curve,受试者工作特征曲线,简称ROC曲线)结果表明,Maxent生态位模型预测松材线虫适生分布的训练数据和测试数据的AUC(areas under curve,曲线下面积)值分别为0.987和0.986,达到了极高的精度。各环境变量重要性的刀切法检验表明,温度因素对松材线虫的潜在分布影响最大。

Abstract: In order to understand the possible spread areas and harmful degree of Bursaphelenchus xylophilus in China, a prediction of potential geographical distribution of B. xylophilus was conducted by using the maximum entropy(Maxent)ecological niche models and combining with geographic information system(ArcGis). The results showed that the potential distribution of B. xylophilus mainly concentrated in eastern China and southern China, and the southern parts of north China and north-east China and southwest China were middle or low suitable areas. The other parts of China were unsuitable areas. In suitable distribution areas, as long as pine host trees appeared, other environment condition would satisfied the occurance of the disease. Results of ROC(receiver operating characteristic curve)evaluation showed that training data of predicted suitable distribution areas by Maxent ecological niche models and AUC(areas under curve)of testing data were 0.987 and 0.986, respectively. The precision of prediction was extremely high. A jackknife test in Maxent showed that temperature was the most important environmental variable affecting the distribution of this pest species.

中图分类号: