文献类型：期刊文献

中文题名：八种亚热带森林类型乔木层地上生物量分配模型

英文题名：Above-Ground Biomass Distribution Models for Arbor Layer of Eight Subtropical Forest Types

作者：付甜[1] 朱建华[1] 肖文发[1] 曾立雄[1]

第一作者：付甜

机构：[1]中国林业科学研究院森林生态环境与保护研究所

年份：2014

卷号：50

期号：9

起止页码：1-9

中文期刊名：林业科学

外文期刊名：Scientia Silvae Sinicae

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：国家林业局948引进项目"林业碳收支模型CBM-CFS3引进(2012-4-78);国家‘十二五’科技支撑项目"长江流域防护林体系整体优化与调控技术研究"(2011BAD38B04)

语种：中文

中文关键词：幂函数;多项;logit模型;材积;-生物量转换;生物量比例

外文关键词：power function;multinomial logit model;volume-to-biomass conversion;biomass proportion

分类号：S758.4;S758.8

摘要：以我国亚热带的8种森林类型的乔木层为研究对象，采用幂函数模型和多项 logit模型分别构建各森林类型的材积-生物量转换模型和林木器官生物量比例模型。结果表明：材积-生物量转换模型的拟合结果表现为除常绿阔叶林外，其他森林类型的材积与生物量均显著相关（R2＞0.5，P ＜0.001），预测残差随自变量的增大而升高，均方根误差为6.520～23.123 t·hm -2，抽样精度检验结果显示，材积-生物量转换模型的预测精度为31.14%～91.79%，除常绿阔叶林以外，其他森林类型的模型预测精度均达到70%以上；林木器官生物量比例模型的拟合结果表现为除常绿阔叶林外实测值与预测值的相关性均显著（P ＜0.05），预测残差随自变量的增大而减小；干、皮、枝和叶生物量比例模型的均方根误差均不超过0.1，分别为0.031～0.085，0.005～0.041，0.029～0.103和0.016～0.083，各林木器官比例的预测精度分别达到91.04%～96.14%，62.71%～94.48%，63.70%～94.47%和-8.86%～83.92%；各森林类型（除常绿阔叶林外）的乔木层地上材积-生物量模型和林木器官生物量比例模型的可行性较高，所得模型参数也可为亚热带森林生物量分配研究提供参考。
In this study,the volume-to-biomass conversion models and the proportion models for biomass components of 8 subtropical forest types in China were established by using the power function model and multinomial logit model, respectively. The results show that the fitting results of the volume-to-biomass conversion models for every forest type except the evergreen broad-leaved forest exhibit significant relations （R2 〉0. 5,P〈0. 001） between volume and biomass,and the predicting residuals increase with the increment of the independent variable value. The root mean square errors are all controlled within 6. 520-23. 123 t·hm -2 . The testing accuracy is between 31. 14% -91. 79%,which are all above 70%except for evergreen broad-leaved forest. On the other hand,the fitting results of proportion models for the biomass components reach significant level （P 〈0. 05） except for the evergreen broad-leaved forest,and the predicting residuals decrease with the increased independent variable value. The root mean square errors of stem,bark,branch and leaf biomass proportion of total tree biomass are all less than 0. 1,that is 0. 031 -0. 085,0. 005 -0. 041,0. 029 -0. 103 and 0. 016 -0. 083. The prediction accuracy reaches 91. 04% -96. 14%,62. 71% -94. 48%,63. 70% -94. 47%,and -8. 86% -83. 92%,respectively. The volume-to-biomass conversion models and the proportion models for biomass components in this paper are reasonable models for the 8 subtropical forest types except for the evergreen broad-leaved forest,and the obtained model parameters also can provide a reference for studying biomass distribution of the other subtropical forests.