Allometric equations for estimating biomass of natural shrubs and young trees of subtropical forests

Shrubbery and young plantations, with a large number of tree species, take up a broad area in subtropical Chinese forests and contribute a significant part to forest biomass and carbon (C) stocks. However, the biomass and C stocks of shrubbery and young plantations were generally underestimated or excluded in forest biomass calculations due to the scarce of standard equations for estimating biomass. The aim of this study was to develop appropriate regression equations for biomass estimation of shrubbery and young plantations. A total of 108 individuals of 15 most widespread shrub and young tree species were sampled by destructive harvesting, and the dry weight of each component of trees, i.e., foliage, branch, stem and root were obtained. The dry biomass of each component was correlated with plant height (H), basal diameter of stem (D), crown area (CA) and their composite variables (DH)-H-2 and CV (CA x H) by using seemingly uncorrelated regression, and the best fitted model was chosen according to the determination coefficient (R-2), root mean squared error (RMSE), Akaike's information criterion (AIC) and percent relative standard errors (PRSE) less than 25%. In the species-specific equations, H, D or (DH)-H-2 were used as the appropriate independent variables in most of the equations, and only a few of them were CA or CV. In the multiple species equations, H was an important variable to predict the biomass, but the predictors of biomass equation for different functional groups or life forms were different due to diversity of external morphology. The species-specific equations had low biases, while the general equations for functional groups and life forms showed comparable biases, and the general equations for all species had the highest prediction biases. Therefore, general equations for functional groups or life forms are recommended to estimate biomass for species without species-specific allometric equations.