森林工程2025，Vol.41Issue(5)：883-895,13.DOI:10.7525/j.issn.1006-8023.2025.05.002
基于树种分类的阔叶混交林单木枝下高预测模型
Prediction Model of Individual Height to Crown Base in Mixed Broad-leaved Forest Based on Tree Species Classification
摘要
Abstract
The mixed broad-leaved forest at Maoer Mountain is characterized by its complex stand structure,rich species diversity,and intricate interactions between tree species,all of which influence height to crown base.Therefore,accu-rately constructing a height to crown base model can provide important reference value for guiding forest management and improving the accuracy of growth prediction.In this study,28 representative sample plots of mixed broad-leaved forests in Maoer Mountain area were selected,and 18 tree species widely distributed in the sample plots were divided into groups according to the variability coefficient of tree species diameter,average tree species diameter,average height-di-ameter ratio,proportion of tree species,classification of soft broad-leaved and hard broad-leaved species,and shade tol-erance of tree species,thereby solving the problem of tree species complexity and the insufficiency of sample size of a single tree species in model construction.In this study,Wykoff model was adopted as the basic model.At the same time,factors such as stand,competition and species diversity were considered to express the mixed degree and competi-tion of stand,and a generalized model of height to crown base was established.At the same time,the difference of height to crown base among plots and tree species groups was considered to build a model of subbranch height mixed effect.The influence of sampling on the result was analyzed by using the method of'remaining one'.The results showed that the common tree species in Maoer Mountain could be divided into four tree species groups.Besides DBH and tree height,the average height of dominant trees,the sum of the section area of intra-species larger than that of the target trees and the Shannon index had significant effects on height to crown base.Considering the mixed effects of plots and tree species groups,the basic and generalized subbranch height models had high fitting accuracy R2 of 0.638 and 0.627,and RMSE of 1.880 and 1.909,respectively.In addition,when one sample was randomly selected from each tree group in each plot to correct the mixed model,compared with the fixed effect model,the MAE and MAPE of the generalized height to crown base model decreased by 7.51％and 13.51％,respectively,showing better prediction effect without wasting man-power,material and financial resources.This study analyzed the effects of various tree species and ecological functions on the growth of height to crown base in Maoer Mountain,and provided some reference for predicting the height to crown base of different tree species in broad-leaved mixed forest in Maoer Mountain.关键词
阔叶混交林/枝下高/树种分组/混合效应模型/模型校正Key words
Mixed broad-leaved forest/height to crown base/grouping of tree species/mixed effect model/calibration of model分类
农业科技引用本文复制引用
李晨,董利虎,苗铮..基于树种分类的阔叶混交林单木枝下高预测模型[J].森林工程,2025,41(5):883-895,13.基金项目
"十四五"国家重点研发计划项目(2022YFD2201000) （2022YFD2201000）
国家自然科学基金面上项目(31971649) （31971649）
黑龙江省博士后科学基金项目(LBH-Z23050). （LBH-Z23050）
