天然落叶松林木冠幅模型与林分密度模型及其关系  

Tree Crown Width Model,Stand Density Model and Their Relationships for Natural Larch

作　　者：曾伟生 杨学云 蒲莹 Zeng Weisheng;Yang Xueyun;Pu Ying(Academy of Inventory and Planning,National Forestry and Grassland Administration Beijing 100714)

机构地区：[1]国家林业和草原局林草调查规划院,北京100714

出　　处：《林业科学》2025年第2期93-100,共8页Scientia Silvae Sinicae

基　　金：国家重点研发计划项目“典型人工林立地质量评价与生产力提升技术”(2022YFD2200501)。

摘　　要：【目的】研建林木冠幅模型、林分密度模型及其联立模型,探究林木冠幅模型与林分密度模型之间的关系,为进一步研究林分密度管理和指导森林经营提供科学依据。【方法】利用在全国范围内代表性采集的600株落叶松样木调查数据和第九次全国森林资源清查的1273个天然落叶松林样地调查数据,采用非线性回归和对数回归估计方法,分别建立了落叶松林木冠幅模型和林分密度模型;在分析二者之间内在相关性的基础上,采用含哑变量的多元非线性回归估计方法,建立了林木冠幅和林分密度联立模型,并确定了最大林分密度线。【结果】林木冠幅与胸径呈正相关,落叶松林木冠幅模型的确定系数R2在0.76以上,估计值的标准误SEE为1.03 m,平均预估误差MPE为2.16%;冠径比随胸径的增加而下降,从2 cm时的0.47快速下降至20 cm时的0.23,再缓慢下降至50 cm时的0.17。林分密度与平均胸径呈负相关,落叶松林分密度模型的R2达到0.88,SEE为147株·hm^(-2),MPE为2.58%;林分密度模型与林木冠幅模型的指数呈2倍负相关,根据联立模型得出的斜率参数为-1.384,可视为全国落叶松天然林自然稀疏线斜率的总体平均值。【结论】基于林分密度模型与林木冠幅模型之间的内在相关性,采用多元回归估计方法求解林分密度指数的斜率参数是一条可行的新途径。该方法对研究其他树种的林分密度指数及指导林分密度管理和森林科学经营也具有应用价值。【Objective】This study aims to develop models for tree crown width(CW)and stand density(SD),as well as a combined model integrating both.By exploring the correlation between CW model and SD model,this research provides a scientific basis for further studies on stand density management and guidance for forest management practices.【Method】Using the measurement data of 600 representative sample trees of larch collected throughout the whole country of China,and the data of 1273 permanent plots in natural larch forests from the 9th national forest inventory,the CW model and the SD model of larch were established by using nonlinear regression and logarithmic regression estimation methods.Based on the analysis of the relationship between CW model and SD model,a multivariate nonlinear regression method with dummy variables was used to establish the simultaneous models,and the maximum stand density line was determined.【Result】The results showed that CW was positively correlated with diameter at breast height(DBH);the determination coefficient R2 of CW model for larch was higher than 0.76,the standard error of estimate(SEE)was 1.03 m,and the mean prediction error(MPE)was 2.16%.The ratio of CW-DBH decreased with the increase of DBH,from 0.53 at 2 cm to 0.23 at 20 cm and then slowly to 0.17 at 50 cm.The SD was negatively correlated with mean DBH;the R2 of SD model for larch forests reached to 0.88,the SEE was 147 stems per hectare,and the MPE was 2.58%.There was a 2-fold negative correlation between the two power parameters of SD model and CW model,and the slope parameter from the simultaneous models is−1.384,which can be regarded as the average value of the slopes of self-thinning lines of natural larch forests in China.【Conclusion】It is a feasible new approach to use multivariate regression method for estimating the slope parameter of the stand density index(SDI)based on the correlation between SD model and CW model.This is of practical significance to study the SDI of other tree species and to guide the s

关 键 词：林木冠幅 林分密度 最大密度线 多元回归 落叶松 

分 类 号：S758.5[农业科学—森林经理学]