| Forest biomass is a basic quantity character of the forest ecological system. Biomass data are foundation of researching many forestry and ecology problems, thus, accurate quantification of tree biomass is critical and essential for calculating carbon storage, as well as for studying climate change, forest health, forest productivity, nutrient cycling, and etc. Although directly measuring the actual weight of each component (i.e., stem, branch, foliage and root) is undoubtedly the most accurate method, it is destructive, time consuming, and costly. Thus, developing biomass models is considered a better approach to estimating forest biomass.Based on the biomass data of 1049 sample trees for 17 tree species and 20996 permanent plots for 27 forest types in northeast China forest area, we made a comprehensive analysis to the changing laws of root to shoot ratios for the main tree species and forest types. The additive system of biomass equations for the main tree species and forest types were developed. The model error structure (additive vs. multiplicative) of the biomass allometric equation was evaluated using the likelihood analysis, while nonlinear seemly unrelated regression (NSUR) was used to estimate the parameters in the additive system of biomass equations. The biomass model validation was accomplished by Jackknifing technique. In general, we studied assessing the model error structure, the model structures of the additive systems of biomass equations, the changing laws of biomass partitioning and root to shoot ratios deeply. These provided technical and theoretical support for accounting and monitoring the Chinese forest biomass and carbon stocking.The detail contributions and conclusions were as followed:(1) The choice between linear regression on log-transformed data or nonlinear regression on original data depends on the model error structure. The likelihood analysis is considered to be consistent with core statistical principles, and more suitable in determining model error structures.(2) 3SPW and SUM (3) not only ensure the additivity property of biomass equations but also have the good prediction accuracy. However, the 3SPW is more complicated in estimating the parameters and model application, SUM (3) is relatively easy toimplement.(3) There are differences in the biomass partitioning of different diameters at breast height, and the changing laws are inconsistent and unstable. The biomass partitioning of some components increased or increased trend with the increase of diameter at br east height, the biomass partitioning of some components decreased or decreaed trend with the increase of diameter at breast height, and the biomass partitioning of the other components didn’t change or change significantly with diameter at breast height. In addition, the results of this study suggest a significant effect of age on tree biomass partitioning of different components.(4) Stem biomass had the largest relative contribution to total biomass (62.6%), followed by root biomass (22.0%), while brancha and foliage biomass had the smallest relative proportion (11.2% and 4.4%, respectively) for the seventeen species in northeast China forest area. In addition, there were three significant linear relationships for all sample trees of the seventeen species, and the individual root to shoot ratios of all sample trees ranged from 0.196 to 0.378.(5) For the aggregation additive system of biomass equations, the biomass equations of total, aboveground and stem were better than biomass equations of branch and foliage. The precision of each biomass equations in the additive system was above 80% that would be suitable for predicting the biomass of the seventeen species in northeast China forest area.(6) For different forest types, there are significant differences in the biomass partitioning of stem, branch, foliage and root, and the biomass partitioning of them changed with quadratic mean diameter, average tree height, and stand density, and the changing laws were different. For Da xing’an mountains of Heilongjiang Province, the proportion of different forest types is 64.1% for stem,23.0% for root,10.2% for branch and 2.7% for foliage, respectively; for Chang bai mountains of Jilin Province, the proportion of different forest types is 63.1% for stem,21.4% for root,12.5% for branch and 3.0% for foliage, respectively; for Xiao xing’an mountains and Chang bai mountains of Heilongjiang Province, the proportion of different forest types is 62.8% for stem,21.4% for root,11.9% for branch and 3.9% for foliage, respectively.(7) For the root to shoot ratios of different forest types, there are some differences, and for most forest types, the root to shoot ratios changed significantly with quadratic mean diameter, average tree height, and stand density. The average root to shoot ratio for different forest types of different regions ranged from 0.210 to 0.370, with an average value of 0.280. Among them, the average root to shoot ratio is 0.300 for Da xing’an mountains of Heilongjiang Province, 0.275 for Chang bai mountains of Jilin Province and 0.274 for Xiao xing’an mountains and Chang bai mountains of Heilongjiang Province, respectively.(8) For the aggregation additive system of stand-level biomass equations, the biomass equations of total, aboveground and stem were better, and the biomass equations of belowground (root), branch and foliage were worse. Overall, the three kinds of aggregation additive system of stand-level biomass equations had a good model fitting performance. The precision of each stand-level biomass equations in the aggregation additive system was above 90% that would be suitable for predicting the biomass of the main forest types in northeast China forest area.(9)The result of selecting the most suitablemethod method to estimate the stand-level biomass is uncertain. Stand biomass equations including stand volume (SBEV) and stand biomass equations including BEF and stand volume (SBEVBEF) belong to volume-derived biomass. They are different from the stand biomass equations including stand variables (SBESV) in essence. The similar precision was obtained for SBEV, SBEVBEF and SBESV, but the precision was relatively low for a constant biomass expansion factor. |
PDF Full Text Request