文献类型：期刊文献

中文题名：基于多元统计分析的马尾松人工林健康评价研究——以广西热带林业实验中心为例

英文题名：Health assessment of Pinus massoniana plantation on multivariate statistical analysis: a case study of Guangxi tropical forestry experimental center

作者：赵勇钧[1] 谢阳生[2] 王建军[1] 李晗[1] 葛方兴[1] 孟京辉[1]

第一作者：赵勇钧

机构：[1]北京林业大学森林资源和环境管理国家林业局重点开放性实验室,北京100083;[2]中国林业科学研究院资源信息研究所,北京100091

年份：2019

卷号：39

期号：7

起止页码：100-107

中文期刊名：中南林业科技大学学报

外文期刊名：Journal of Central South University of Forestry & Technology

收录：CSTPCD;;北大核心:【北大核心2017】；CSCD:【CSCD2019_2020】；

基金：中央级公益性科研院所基本科研业务费专项“多功能可持续森林经营方案编制关键技术研究”(IFRIT201501);全国森林经营科技支撑科研专项

语种：中文

中文关键词：马尾松;健康评价;因子分析;聚类分析;判别分析

外文关键词：Pinus massoniana;health assessment;factor analysis;cluster analysis;discriminant analysis

分类号：S791.248

摘要：为构建一套科学实用的马尾松人工林健康评价体系,为马尾松健康经营提供理论依据。以广西热带林业实验中心的84块马尾松人工林样地为例,选取13个评价指标进行因子分析,计算森林健康指数(FHI)并对FHI值进行Ward聚类,以划分健康等级。此外,研究分析评价了马尾松样地的健康状况,并用判别分析建立Fisher判别函数对健康评价结果进行检验。结果表明:在调查的马尾松人工林中,健康林占32.1%,亚健康林占35.8%,不健康林占32.1%,且不同龄组马尾松人工林健康水平排序为:近熟林>中龄林>成熟林>幼龄林。判别分析与聚类结果具有较好的一致性,其中自身验证法正判率为97.8%,交互验证法正判率为94.3%。因此,构建的指标体系能够科学客观地反映广西热带林业实验中心马尾松人工林的健康状况。
In order to establish a scientific and practical health evaluation system for Pinus massoniana plantation and provide a theoretical basis for healthy management of Pinus massoniana, This study took 84 plots of Pinus massoniana plantations in Guangxi tropical forestry experimental center for example, then selected 13 indicators among these plots to calculate FHI index using the factor analysis method, then clustered FHI values to rank health level using the Ward cluster method. Besides, this study analyzed and assessed health state of these cluster groups, then established Fisher discriminant functional relations by discriminant analysis, finally, tested the health assessment results.The results showed that in these Pinus massoniana plantations surveyed, healthy forests accounted for 32.1%, sub-healthy forests accounted for 35.8% and unhealthy forests accounted for 32.1%, in addition, the health levels of Pinus massoniana plantations in different age groups were near-mature forest > middle age forest > mature forest > young forest.The result of discriminant analysis well agreed with clustering results, the positive verification rate of self-verification method and cross validation method were 97.8%, 94.3%.Therefore, the built indicator system can scientific and objective reflect the health status of the Pinus massoniana plantations in the Guangxi tropical forestry experimental center.