文献类型：期刊文献

中文题名：基于距离相关Hegyi指数的云冷杉天然林单木胸径生长模型

英文题名：Single tree DBH growth model of spruce-fir natural forest based on distance related Hegyi index

作者：周泽宇[1] 周超凡[2] 胡兴国[3] 陈科屹[4] 杜满义[1] 张会儒[1] 符利勇[5]

第一作者：周泽宇

机构：[1]中国林业科学研究院华北林业实验中心,北京102300;[2]中国林业科学研究院森林生态环境与保护研究所,北京100091;[3]吉林省汪清林业局,吉林汪清133200;[4]中国林业科学研究院林业科技信息研究所,北京100091;[5]中国林业科学研究院资源信息研究所,国家林业和草原局森林经营与生长模拟重点实验室,北京100091

年份：2023

卷号：45

期号：10

起止页码：59-69

中文期刊名：北京林业大学学报

外文期刊名：Journal of Beijing Forestry University

收录：CSTPCD;;北大核心:【北大核心2020】；CSCD:【CSCD2023_2024】；

基金：国家重点研发计划课题(2022YFD2200503)。

语种：中文

中文关键词：Hegyi竞争指数;期初胸径;混合效应模型;胸径生长

外文关键词：Hegyi competition index;initial diameter at breast height;mixed effect model;DBH increment

分类号：S791.14;S791.18;S750

摘要：【目的】基于与距离有关的Hegyi种内和种间竞争指数,构建以期初胸径及Hegyi竞争指数为预测变量的云冷杉天然林单木胸径生长预测模型。【方法】使用2013年金沟岭云冷杉林3块1 hm2固定样地内实测数据,以非线性Logistic模型为基础模型,逐步引入期初胸径、种内和种间竞争指数,以探究竞争和生长对于单木胸径变化的影响,并利用非线性混合效应模型进行模型精度的提升。【结果】模型拟合结果显示:以期初胸径、种内Hegyi竞争指数、种间Hegyi竞争指数为预测变量,树种水平下随机效应作用在参数a_(0)、a_(2)、a_(3)上时,模型具有最优的拟合效果,且未出现异方差现象。建模数据的调整决定系数(R^(2)_(adj))、均方根误差(RMSE)、总相对误差(TRE)分别为0.5126、0.6071、3.6519%,利用检验数据对模型进行独立样本检验,检验数据的R^(2)_(adj)为0.5098,检验RMSE均为0.6242,检验TRE均为3.8831%,检验数据的残差分布未出现明显的异方差现象。【结论】云冷杉天然林中影响对象木胸径生长的因素包括单木大小因子与竞争因子。期初胸径对胸径生长的影响较大,为正向作用。竞争因子中,种内竞争和种间竞争均对单木胸径的生长具有明显的调控作用,为负向抑制作用。本文所构建的基于树种水平的非线性混合效应模型能够为研究区天然云冷杉林中对象木的胸径生长预测提供一定的理论基础以及技术参考。
[Objective]Based on the Hegyi intraspecific and interspecific competition indexes related to distance,the prediction model of DBH increment of natural spruce-fir forest was constructed based on initial DBH and Hegyi competition index.[Method]In order to explore the effects of competition and growth on DBH growth of individual trees,the measured data of three permanent sample plots(each of 1 ha)of sprucefir in Jingouling Forest Farm,Jilin Province of northeastern China in 2013 were utilized.Based on the nonlinear Logistics model,initial DBH,intraspecific and interspecific competition indexes were gradually added.The nonlinear mixed-effect model was utilized to improve the model accuracy.[Result]The results of model fitting showed that the model had the best fitting effect when initial DBH,intraspecific and interspecific Hegyi competition indexes were used as predictive variables,the best fitting efficiency occurred based on the species-level of random effect worked on parameters a_(0),a_(2),a_(3),and without heteroscedasticity.The R^(2)_(adj),root mean squared error(RMSE)and total relative error(TRE)of modeling data were 0.5126,0.6071,3.6519%,respectively.The R^(2)_(adj),RMSE and TRE of validation data were 0.5098,0.6242,3.8831%,respectively.The residual distribution of validation data did not show obvious heteroscedasticity.[Conclusion]The factors affect the diameter growth of target trees in natural spruce-fir forest,including self-growth factors and competition factors.Among the self-growth factors,initial DBH is the main factor and plays a positive role to promote DBH increment.Among the competition factors,interspecific competition and intraspecific competition have obvious inhibition effects on the growth of individual tree DBH increment.The nonlinear mixed effect model based on species-level can provide a theoretical basis and technical reference for the DBH growth of target trees in natural spruce-fir forest in the study area.