华北落叶松天然次生林树高曲线的混合效应模型  

Developing a height-diameter relationship model with mixed random effects for Larix principis-rupprechtii natural secondary forests

作　　者：段光爽 李学东 冯岩[4] 符利勇[1] DUAN Guangshuang1,2, LI Xuedong3, FENG Yan4 , FU Liyong1.(1. Research Institute of Forest Resource Information Techniques, Chinese Academy of Forestry, Beijing 100091, China; 2.College of Mathematics and Statistics, Xinyang Normal University, Xinyang 464000, China; 3.Experimental Centre of Forestry in North China, Chinese Academy of Forestry, Beijing 102300, China; 4. Chinese Academy of Forestry, Beijing 100091, Chin)

机构地区：[1]中国林业科学研究院资源信息研究所,北京100091 [2]信阳师范学院数学与统计学院,河南信阳464000 [3]中国林业科学研究院华北林业实验中心,北京102300 [4]中国林业科学研究院,北京100091

出　　处：《南京林业大学学报（自然科学版）》2018年第2期163-169,共7页Journal of Nanjing Forestry University：Natural Sciences Edition

基　　金：国家林业公益性行业科研专项项目(201404417);国家自然科学基金项目(31570628;31470641);河南省科技开放合作项目(172106000071)

摘　　要：【目的】建立华北落叶松天然次生林树高曲线,为森林可持续经营提供技术支撑。【方法】基于116块华北落叶松天然次生林样地单木数据,从11个具有生物学意义的备选模型中选出一个最优基础模型。考虑地域和样地对树高曲线的干扰,基于94块样地利用带哑变量的非线性混合效应模型构建了树高曲线,并对22块样地数据进行验证。【结果】利用全部数据拟合11个备选模型,其中最优模型是Logistic方程,决定系数为0.765 3,均方根误差为3.279 4。引入哑变量和随机效应后,所建模型决定系数为0.915 2,均方根误差为1.892 2,与基础模型相比拟合精度显著提高。验证数据结果表明,带哑变量的非线性混合模型对树高的预测效果良好。【结论】样地随机效应对华北落叶松树高曲线干扰较大,当模型考虑这些随机效应对树高曲线的影响时能显著提高模型预测精度。该模型可用于华北落叶松天然次生林其他样地树高的预测。【Objective】This study aimed to develop a height-diameter relationship model of Larix principis-rupprechtii natural secondary forests as technical support for sustainable forest management.【Method】Based on individual tree data for 116 plots of Larix principis-rupprechtii natural secondary forests,an optimal basic model was selected from 11 candidate models with biological significance. Taking into account the disturbance from regions and sample plots,a nonlinear mixed-effects model with dummy variables for height-diameter relationships based on 94 plots was constructed,and validation was implemented on 22 surplus plots.【Result】The optimal model of 11 alternative models was fitted to the total data with alogistic equation,and the coefficient of determination and root mean square error were 0. 765 3,3. 279 4,respectively. After adjusting the precision of the mixed-effects model with dummy variables and random effects,established with the modeling data,the coefficient of determination and root mean square error were 0. 915 2,1. 892 2,respectively,which is a distinctly improvment compared with the basic model. The prediction effects of this model were perfect with the use of the validation data.【Conclusion】The disturbance from plot level random effects significantly influenced the height-diameter relationship. The prediction accuracy of the height-diameter relationship model with these random effects was improved obviously. The prediction of tree height of other plots in Larix principis-rupprechtii natural secondary forests can be implemented with this nonlinear mixed-effects model with dummy variables.

关 键 词：华北落叶松 非线性混合效应模型 树高曲线 LOGISTIC方程 哑变量 

分 类 号：S758[农业科学—森林经理学]