文献类型：期刊文献

中文题名：广东省枫香单木干材生物量生长模型的研建

英文题名：Construction of Individual Bole Biomass Growth Models for Liquidambarformosana in Guangdong Province

作者：何潇[1] 赵嘉诚[1] 曹磊[1] 李海奎[1] 徐期瑚[2]

第一作者：何潇

机构：[1]中国林业科学研究院资源信息研究所;[2]广东省林业调查规划院

年份：2018

卷号：33

期号：6

起止页码：236-242

中文期刊名：西北林学院学报

外文期刊名：Journal of Northwest Forestry University

收录：CSTPCD;;北大核心:【北大核心2017】；CSCD:【CSCD_E2017_2018】；

基金：广东省林业科技专项"广东主要碳汇造林树种生物量模型研建"(2015-02);广东省林业科技创新平台建设项目"广东省碳汇计量监测创新平台建设"(2016CXPT03)

语种：中文

中文关键词：干材生物量生长;哑变量;树干解析;枫香

外文关键词：bole biomass growth;dummy variable;stem analysis;Liquidambar formosana

分类号：S781

摘要：枫香(Liquidambar formosana)是我国闽北林区重要的乡土树种,生长迅速、适应性强,在观赏、药用、用材等方面都有着重要作用。单木模型是林分模型的基础,通过建立单木生物量生长模型,在时间维度上分析林木生物量生长的动态变化,以期为全面评价森林的固碳能力提供科学依据。采用典型抽样的方法,从广东省9个地区的一类样地附近抽取枫香样木40株(天然起源25株,人工起源15株),破坏性实测各组分的生物量并进行树干解析得到样木的生物量生长数据。基于逻辑斯蒂(Logistic)、坎派兹(Gompertz)和理查德(Richards)3种基础生长方程,利用哑变量的方法,建立分起源和立地等级的枫香干材生物量生长模型,以决定系数R2、均方根误差RMSE、总相对误差TRE和平均预估误差MPE 4个指标来评价各模型,进而分析枫香干材生物量的生长过程。结果表明,Richards方程在基础模型中效果最佳(R2=0.61,RMSE=50kg);与Richards方程相比,哑变量模型R2提高了11%,RMSE降低了7.95kg;起源与立地等级对枫香干材生物量均产生了较大影响,在天然起源的高立地等级条件下,枫香干材理论最大连年生长量和平均生长量分别在其生长的第23年和第38年到达。
Liquidambar formosana is an important native species in the forest region of Northern Fujian.It grows rapidly with strong adaptability,and plays important roles in ornamental,medicinal and timbering values.Individual tree model is the basis of stand model.The dynamic changes in individual biomass growth were analyzed in the time dimension by establishing individual tree biomass growth model to provide a scientific basis for the comprehensive evaluation of the carbon sequestration capacity of the forest.Forty L.formosana trees were typically sampled （25 were from natural stands and 15 from plantations） around national forest inventory permanent sample plots in 9 different regions in Guangdong Province and the data of biomass growth were obtained by the destructive measurement of the biomass of each component and stem analysis.Based on the basics growth equation of Logistic,Gompertz and Richards,bole biomass growth models of distinguishing the stand origin and site quality with dummy variables of L.formosana were constructed,these models were evaluated by four indexes：coefficient of determination （ R 2）,root mean square error （RMSE）, total relative error （TRE） and mean prediction error （MPE） The growth of L.formosana was further analyzed.The results showed that the fitting of Richards equation took the lead among base models （ R 2=0.61,RMSE=50 kg）; compared with the Richards equation, R 2 raised by 11% and RMSE decreased by 7.95 kg to the model with dummy variables; the stand origin and site quality had a greater impact on the bole biomass of L.formosana .Under natural origin and high site quality,theoretical maximum of current annual increment and mean annual increment achieved at 23 a and 38 a,respectively.