北亚热带高山区日本落叶松林相容性生物量模型研究  

Research on compatible model for biomass of Larix kaempferi forest in northern sub-tropical alpine area

作　　者：王宏全[1] 孙晓梅[1] 张灿明[2] 李子敬[1] 李忠国[1] 

机构地区：[1]国家林业局林木培育重点实验室,中国林业科学研究院林业研究所,北京100091 [2]湖南省林业科学研究院,长沙410004

出　　处：《东北农业大学学报》2011年第2期131-138,共8页Journal of Northeast Agricultural University

基　　金：国家“十一五”科技支撑课题(2006BAD24B06,2006BAD01A14)

摘　　要：研究在北亚热带高山区日本落叶松林内均匀设置51块标准地,对单木、灌草和林分的生物量进行测定并建模,再通过独立检测样地验证模型的精度。结果表明,日本落叶松单木生物量模型以D2H为参数的拟合精度高于仅以D为参数的模型,其中幂函数B=4.6×(D2H)0.827对预测北亚热带高山区日本落叶松单木生物量效果最佳,γ值和F值分别为0.979和926.1,达极显著水平。以代数和关系模型为基础建立了单木各器官的相容性生物量模型,树干(96.3%)和树皮(92.1%)的预报精度高于树叶和树根的精度(86.3%和88.4%)。灌草层生物量的估计误差与样品干、鲜质量的测量精度以及抽样误差密切相关。采用逐步回归法对郁闭度C、平均林龄A、平均胸径D、平均树高H、林分蓄积M、林分密度N、胸高断面积B这7个自变量进行有效选优,构建了北亚热带高山区日本落叶松生物量预测模型B=-20.01+36.87C+0.36M-0.62D+2.51H+0.34A(精度达94.4%),通过联立方程组方式解决了生物量模型中总量与各分量的相容性估计问题。在建立的相容性模型系统中,乔木层的生物量预估精度为95.4%,高于林下灌草层(83.2%和80.09%),乔木层中以树干生物量预估精度最高(96.89%),其次为树冠生物量(95.47%),树根生物量的预报精度较低(89.91%)。Total 51 standard plots with five quadrats of shrub and grass for each were designed during the experiment in northern sub-tropical alpine area and the biomass of individual tree,shrub,grass and forest stand were measured and modeled. The examined plots were used to check all the models. The results showed that the predicting accuracy of models with D2H as parameter was higher than that of the models with only D as parameter;the fitting result of power function B=4.6×（D2H） 0.827 was the best,R value,F value of which were 0.979 and 926.1,respectively and regression performance was significant;the estimating error of the biomass of shrub and herb was closely related with sampling accuracy rather than the sample size;in the model for forest,the information contained in C,A,D,H,V,N and B were overlapped but the invalid variables were eliminated by the method of correlation analysis and stepwise regression;the precision of the model for trunk was the highest with 96.89% but that of the model for shrub and grass were the lowest with 83.20% and 80.09%,respectively;forest biomass was calculated by the jointed estimation method,the predicting accuracy of which was 94.44%,and 95% of residual for the model was reasonable;and the seried models in this research solved the compatibility problem for Larix kaempferi in this region.

关 键 词：日本落叶松 单木生物量模型 逐步回归 联合估计 相容性生物量模型 

分 类 号：S758[农业科学—森林经理学]