北亚热带常绿阔叶林林下灌木生物量模型的建立

Biomass models for shrubs under evergreen broad-leaved forest in north subtropics

【目的】灌木是森林生产力的重要组成部分,探索北亚热带地区常绿阔叶林林下灌木生长模型,为森林生物量及其碳储量估算奠定基础。【方法】在安徽南部查湾自然保护区,选取4种常见林下灌木树种(老鼠矢、乌药、朱砂根和香桂),通过野外实测获得地径(D)、树高(H)、地径平方乘树高(D2 H)、冠径树高乘积(CH)、植冠面积(AC)和植冠体积(VC)等模型参数,拟合生物量模型,基于独立检验数据对模型进行验证,获得生物量最优模型。【结果】各灌木树种单器官及全株生物量模型以D2 H和CH为自变量都具有较高的拟合优度(0.815~0.983)和较小的标准误(SEE)。不同灌木树种、不同器官之间的生物量最优模型选用方程均存在一定差异,以幂函数、二项式方程为主,且模型检验精度均较高(总相对误差(RS)<30%,平均相对误差绝对值(RMA)<20%)。模型的普适性研究表明,叶、枝和根生物量最优通用模型为W=a+bX+cX2(X为D2 H(叶、枝)或CH(根)),拟合效果较优;而全株生物量最优模型为W=1.423 2(D2 H)0.832 4,拟合指数(FI)=0.960,适用于4种灌木叶、枝、根和全株生物量的估算,但根系通用模型的估算精度低于叶、枝与全株的最优生物量通用模型。【结论】基于生物量模型可以精准地估算亚热带地区的灌木生物量。

[Objective] The shrub layer is an important part of stand productivity in forest ecosystems. This study investigated allometric models for biomass estimation of shrubs to provide basis for accurate es- timation of forest biomass and related carbon stock. [Method] Four common shrub trees （Symplocos stel- laris ,Lindera aggregate ,Ardisia crenata, and Cinnamomum subavenium） at Zhawan Natural Reserve in southern Anhui were selected and descriptive parameters of D （basal diameter）, H （total plant height）, D2H （square of basal diameter X height） ,CH （mean of crown diameter X H） ,Ac（means crown area） ,Vc （means plant crown volume）,and Ac X H were obtained through filed investigation to fit biomass models. Models were also validated based on independent test data. [Result] The biomass models of both single or- gan and whole plant with D2H and CH as independent variables had higher goodness of fit （0. 815-0. 983）and lower standard error of estimation （SEE）. different shrub species, and the optimal model The biomass equations differed between different or s mostly followed power function and binomial with gans and high ex- and RMA〈20%. The universality analysis of the models for these four aal universal biomass model for leaf,branch and root was W=a-F-bX-t-cX2 lile the best model for entire plant biomass was binomial （W = 1. 423 2%. universal model can apply to biomass estimation of leaf, branch, root and trees,but the estimate accuracy of universal model was lower for root blo- whole biomass. [Conclusion] The allometric equations obtained is a valid accurate estimation of shrub biomass in subtropical forest ecosystems. whole plant for the mass than for leaf, and nondestructive method to