首页 | 本学科首页   官方微博 | 高级检索  
     检索      

上海青菜田黄曲条跳甲发生环境因子的随机森林建模分析
引用本文:任向辉,崔建新,杨萌,李鹏,牛新月,李文杰.上海青菜田黄曲条跳甲发生环境因子的随机森林建模分析[J].中国农业科技导报,2017,19(7):117-123.
作者姓名:任向辉  崔建新  杨萌  李鹏  牛新月  李文杰
作者单位:河南科技学院资源与环境学院, 新乡市农作物重大有害生物防控重点实验室, 河南 新乡 453003
基金项目:国家现代农业产业技术体系建设专项(CARS-45-SYZ10);河南科技学院重点资助基金项目(0500136);国家级大学生创新创业项目(20130515403)资助。
摘    要:应用LAI-2000植物冠层仪对一2 000 m2上海青菜田10个样方的节肢动物群落和菜心小气候、植株冠层数据进行了调查,在对相关影响因子进行初步的典范对应分析基础上应用R软件建立黄曲条跳甲发生量与环境因子的随机森林模型,然后利用该模型对上海青植株黄曲条跳甲分布的环境因子进行了排序分析。随机森林模型分析结果表明黄曲条跳甲的发生与植株冠层信息、小气候信息与各种节肢动物的分布关联性明显,9个自变量均方残差仅为19.698 22。叶心温湿度、叶面积指数与平均叶倾角是影响该虫发生的最重要环境因子,这些因子加上天敌龟纹瓢虫因子进行主成分分析的特征值累计百分率达91.474 5%。因此,黄曲条跳甲在上海青植株的分布受外界风速气温和作物长势所引起的叶心温湿度和叶片遮阴程度影响极大,在该害虫综合治理中要重视气象因素与作物生长指标对其分布的影响。

关 键 词:上海青  黄曲条跳甲  随机森林模型  

Radom Forest Model Analysis on Environment Factors of Phyllotreta striolata Occurrence in Brassica chinensis Field#br#
REN Xianghui,CUI Jianxin,YANG Meng,LI Peng,NIU Xinyue,LI Wenjie.Radom Forest Model Analysis on Environment Factors of Phyllotreta striolata Occurrence in Brassica chinensis Field#br#[J].Journal of Agricultural Science and Technology,2017,19(7):117-123.
Authors:REN Xianghui  CUI Jianxin  YANG Meng  LI Peng  NIU Xinyue  LI Wenjie
Institution:Key Laboratory for Crops Pest Management of Xinxiang, School of Resource and Environment Science, ; Henan Institute of Science and Technology,  Henan Xinxiang 453003, China
Abstract:The data of arthropod communities, plant center microclimate and plant canopy were surveyed on 10 samples in a 2 000 m2 field of Brassica chinensis with LAI-2000 plant canopy analyzer. On the basis of a primary CCA analysis to the related factors of this survey, a random forest model between the Phyllotreta striolata occurrence data and environment factors was established using R software. Moreover the ordinary analysis of these environment factors about Phyllotreta striolata occurring was carried out by this model. The result of random forest model analysis showed that there was an evident relationship between Phyllotreta striolata occurrence and microclimate information and corresponding distribution information of various arthropod. Furthermore MSR of this model’s 9 independent variables was only 19.698 22. In addition the most important environment factors affecting this pest occurrence were temperature and RH% of leaf center, LAI and MTA of every leaf blade. The cumulative percentage of PCA Eigen value about these environment factors plus the natural enemy factor of Propylea japonica was 91.474 5%. In conclusion the distribution of Phyllotreta striolata at Brassica chinensis plant was restricted by the values of temperature and RH% of leaf center, the shading degree of leaf blades brought by outdoor wind velocity, temperature and plant growth vigour. Therefore the influence of weather factors and crop growth indexes to Phyllotreta striolata distribution should be paid more attention in the comprehensive control and management of this insect.
Keywords:Brassica chinensis  Phyllotreta striolata  random forest model  
本文献已被 CNKI 等数据库收录!
点击此处可从《中国农业科技导报》浏览原始摘要信息
点击此处可从《中国农业科技导报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号