竞争对天然针阔混交林树木直径生长的影响  

Effect of Competition on Tree Growth in Natural Coniferous and Broad-leaved Mixed Forest

作　　者：刘耀凤 董利虎[1] 郝元朔 李凤日[1] Liu Yaofeng;Dong Lihu;Hao Yuanshuo;Li Fengri(Northeast Forestry University,Harbin 150040,P.R.China)

机构地区：[1]东北林业大学,哈尔滨150040

出　　处：《东北林业大学学报》2023年第9期1-7,共7页Journal of Northeast Forestry University

基　　金：国家重点研发计划项目(2022YFD2201000)。

摘　　要：为探究不同竞争水平条件时天然针阔混交林中主要组成树种单木直径生长的变化规律,将竞争因子作为重要影响因素来构建天然针阔混交林单木直径生长模型。根据穆棱林业局44块天然针阔混交林固定样地的3期调查数据,利用方差分析分别比较了不同竞争强度时红松(Picea koraiensis)、冷杉(Abies nephrolepis)、云杉(Picea koraiensis)、桦木(白桦(Betula platyphylla),枫桦(Betula costata))、椴树(糠椴(Tilia mandshurica),紫椴(Tilia amurensis))、色木槭(Acer mono)、其他树种间直径生长量的差异,利用最小二乘法建立了与距离无关的单木直径生长基础模型,然后构建树种哑变量单木直径生长模型,并以此为基础引入样地随机效应因子构建样地水平混合效应模型。使用独立检验数据对基础模型、哑变量模型、混合模型进行检验。结果表明:天然针阔混交林各树种间直径生长量不同,随竞争强度的增大,各树种直径生长量均显著降低。不同树种在同一竞争强度下直径生长量不同,所有林木断面积之和(B_(AL))≤9 m^(2)时,其他树种的直径生长量最大,色木槭的最小;9 m^(2)<B_(AL)≤36 m^(2)时,云杉的直径生长量最大,色木槭的最小;36 m^(2)<B_(AL)≤45 m^(2)时,色木槭的直径生长量最大,桦木的最小。与基础模型相比,哑变量模型在保证预估精度的同时能够体现树种间直径生长量的差异,其决定系数(R^(2))从0.6115提高到0.6425,均方根误差(ERMS)从0.7669 cm下降到0.7360 cm;混合效应模型的R^(2)从0.6425提升到0.6627,ERMS降低到0.7149 cm,提高了模型的预测精度。在天然针阔混交林中,所有树种直径生长量均随竞争强度的增大而减小,但不同树种间存在差异。包含哑变量的线性混合效应模型可以提高天然针阔混交林的单木直径生长模型预测精度,能较好地描述天然针阔混交林树木直径生长过程。To explore the changes in the diameter growth of the main constituent tree species in natural coniferous broad-leaved mixed forests under different competition levels,competition factors were used as important influencing factors to construct a single tree diameter growth model for natural coniferous broad-leaved mixed forests.According to the data of three surveys in 44 fixed sample plots of natural coniferous and broad-leaved mixed forests in Muling Forestry Bureau,the analysis of variance was used to compare Picea koraiensis,Abies nephrolepis,Picea koraiensis,Betula platyphylla,B.costata,Tilia mandshurica,T.amurensis,and Acer mono.From the differences in diameter growth among other tree species,a distance independent single tree diameter growth basic model was established using the least squares method.Then,a tree species dummy variable single tree diameter growth model was constructed,and a sample level mixed effect model was constructed by introducing random effect factors from the sample plot.The basic model,dummy variable model and mixture model were tested with independent test data.The diameter growth of various tree species in the natural coniferous broad-leaved mixed forest was different,and with the increase of competition intensity,the diameter growth of each tree species significantly decreased.The growth of different tree species under the same competition intensity is different.When BAL is less than 9,the growth of other tree species is the largest,and that of Acer mono is the smallest.When 9<B_(AL)≤36,the growth of Picea koraiensis is the largest and that of Acer mono is the smallest.When 36<B_(AL)≤45,the growth of Acer mono is the largest and that of B.platyphylla is the smallest.Compared with the basic model,the dummy variable model can not only guarantee the prediction accuracy,but also reflect the differences in growth among tree species.Its coefficient of determination(R^(2))increased from 0.6115 to 0.6425,and the root-mean-square deviation(E_(RMS))decreased from 0.7669 to 0.7360 cm.R^(2

关 键 词：天然针阔混交林 林木生长 竞争 哑变量模型 混合效应模型 

分 类 号：S757[农业科学—森林经理学]