|
- 2015
贵州杉木人工林蓄积相关因子的主成分分析DOI: 10.13360/j.issn.1000-8101.2015.06.033 Keywords: 杉木, 人工林, 林分蓄积, 主成分分析, 多元线性回归Cunninghamia lanceolata, artificial forest, stand volume, Principal component analysis, multiple liner regression Abstract: 利用第8次国家森林资源连续清查贵州省第6次复查(2010年)的数据资料,从中提取杉木人工林样地321块,通过主成分分析和多元回归分析研究林分蓄积的相关因子。结果显示:中心区特征值大于1的有3个公因子,累积贡献率68.012%; 第1公因子包括平均胸径、平均树高、平均年龄、郁闭度和林分密度,定义为林分因子,贡献率为37.353%; 第2公因子包括枯枝落叶、土壤厚度、腐殖质层,定义为土壤因子,贡献率17.606%; 第3公因子为海拔,定义为环境因子,其方差贡献率为13.052%。一般区特征值大于1的有2个公因子,累积贡献率66.151%; 第1公因子的定义与中心区一致,其方差贡献率分别为47.772%; 第2公因子包括海拔、土壤厚度,定义为环境因子,其方差贡献率分别为18.378%。多元回归建模并对其进行t检验发现,中心区、一般区的回归方程与其因子分析结果均有差异。综合表明,中心区和一般区林分蓄积的相关因子均是平均胸径、平均树高、平均年龄、郁闭度和林分密度; 此外,中心区林分蓄积相关因子还包括海拔。Based on the 8th forest resources continuous investigation data of Guizhou Province in 2010,this study took Cunninghamia lanceolata artificial forest as the research object and extracted 321 pieces of sample plots,and the factors related to stand volume were analyzed by the method of principal component analysis and multiple liner regression. There had some results in this study:In central region,three common factors with the eigenvalue more than 1 were selected,and the cumulative variance contribution was 68.012%.The first common factor was composed of age,height,DBH,canopy density and stand density,which was defined as the basic stand factor with variance contribution of 37.353%. The second common factor was composed of thickness of soil,humus soil and litter,which was defined as the soil factor with variance contribution of 17.606%.The third common factor was altitude,which was defined as the environment factor with variance contribution of 13.052%. In ordinary region,two common factors with the eigenvalue more than 1 were selected,and the cumulative va-riance contribution was 66.151%.The first common factor were defined as same as them in central region,and the variance contribution were 47.772%.The second common factor was soil thickness and altitude,which was defined as environment factor with variance contribution of 18.378%.The multiple linear regression analysis of the modeling samples were conducted,and the analysis results were verified and different with the results of factor analysis in central region or ordinary region. The findings showed that the related factors of stand volume were age,height,DBH,canopy density and stand density in central region or ordinary region,and the altitude was an important related factor on central region,while the other factors exerted less influence on stand volume growth. The conclusion provides a foundation for conservation and management about C.lanceolata artificial forest in Guizhou Province
|