Aboveground Biomass Of Three Secondary Evergreen Forests At The Xiaokeng Of Nanling Mountain, Southern China

Aboveground biomass (AGB) of three typical evergreen broad-leaved forest communities which were Schima superba community, Castanopsis carlesii+Schima superba community and Castanopsis fissa community at Xiaokeng of Nanling Mountain, Soutnern China were measured using clean felling method in this study. 722 trees were cut down, and components of each tree were measured. AGB of each forest community and its distribution among different layers, tree species and organs of tree layer were analyzed, respectively. The regression models of biomass were established based on these AGB data. Two new methods were put forward to establishing AGB regression model, use basal area of each tree as variable to establish regression model and use virtual standard tree to establish regression model.The main results are as follows:(1) Total aboveground biomass of natural secondary evergreen forest Schima superba community, Castanopsis carlesii+Schima superba community and Castanopsis fissa community were estimated to be 118.472 t.hm^-2, 120.034t.hm^-2 and 134.149t.hm^-2 respectively. Biomass of tree layer accounted for most of community AGB, tree layer biomass of three communities are more than 96.0%. Understory layer and liana biomass of all three forest communities only accounted for less than 1.5% of total AGB. Litter fall biomass only accounted for less than 2.5% of total AGB in three communities. Biomass distribution characteristics of each layer reflected the evergreen broadleaved forest biomass distribution characteristics.(2) Stem biomass of the Castanopsis fissa community accounted for 85% of tree layer biomass, stem biomass of the Castanopsis carlesii+Schima superba community accounted for 79.1% and stem biomass of the Schima superba community accounted for 78.2%, in contrast, proportion of stem biomass in tree layer of Castanopsis fissa community was the biggest. Proportion of branch biomass in tree layer of Castanopsis carlesii+Schima superba community was biggest, the second one was Schima superba community, proportion in Castanopsis fissa community was smallest; Proportion of leaf biomass in tree layer of Schima superba community Was biggest, secondary was Castanopsis carlesii+Schima superba community, proportion in Castanopsis fissa community was smallest. Tree density peak all occurred at 5-15 cm DBH class range in the three communities, however, stand biomass peak lag the tree density peak slightly, all occurred at 10-20 cm DBH class range.(3) By comparing different biomass regression models, it is showed that the power regression models with relative higher R² values and relative lower deviation, especially in the fitting stem biomass and total aboveground biomass.When using D²H as variable to fit stem biomass and total aboveground biomass power regression models were better than using D as variable to fit models. However, when fitting branch biomass and leaf biomass power regression models, using D as variable were better than using D²H as variable in most tree species. It is suggested that in the secondary evergreen broadleaved forest under the same DBH, tree height and stem biomass had a good relationship, however, relationship between tree height and branch-leaf biomass was not obvious.(4) In this study, two improved methods were developed, one method was using basal area of each tree species as variable to fit biomass power regression models, the other method was using virtual standard tree to fit biomass power regression models. Two methods all showed higher R² values and lower deviation, which can improve biomass estimation.