| Forest ecosystem is one of the most important components of terrestrial ecosystem,which plays an important role in mitigating global climate change and global carbon balance.The accurate estimation of carbon storage of the forest ecosystem is the basis for the precise assessment of carbon sink at country or the global scales.Based on the data of 97 analytical trees from 10 ecological stations,and enumeration survey data of 45 plots and forest resource inventory data of Shanxi Province,we compared the accuracy of different biomass estimation methods.Spatial autocorrelation and hot spot analysis were used to reveal the spatial distribution characteristics of the carbon density in tree layer in Shanxi Province.In addition,the main driving factors of the spatial distribution of the carbon density of the tree layer were analyzed by using correlation analysis and geographic detector.The main conclusions are as follows:(1)Among trunk biomass models,the model including both diameter at breast height(D)and height(H)variables fitted well for all the species.The determination coefficient R~2of the model for deciduous broadleaved forest(DBF),temperate coniferous forest(TCF)and cold temperate coniferous forest(CCF)species,when both D and H was used as a combining variable(D~2×H),was 0.818,0.969 and 0.961,respectively.The fitted R~2 of the model,when D and H used as additive variable(D+H),was 0.843,0.969 and 0.983,respectively,for DBF,TCF and CCF species.For canopy and root biomass the performance of the fitted models was similar to those models for the trunks,and both D and H could be well used to estimate their biomass.The biomass prediction deviation from the current models was significantly less than those from the currently used Chinese forest model(CFM).(2)The biomass estimated by the local tree height(H)model was lower than that estimated by the national tree height(CFM)model.This indicated that the model established in this paper was more suitable for estimating the biomass of different forest types in Shanxi Province.For the same tree species and plot,the biomass estimated by different methods was of big difference,and the coefficient of variation of the biomass estimated by the stock method was less than that of estimated by the weighted biomass regression model.The accumulation method was suitable for the plot scale,and the weighted biomass regression model was suitable for the sample wood scale.(3)The carbon density of the tree layer in Shanxi Province was 23.47 MgC/hm~2 and the carbon storage was 82.63 TgC.The highest carbon storage was 38.43 TgC which was located in the eastern soil-rock mountainous area,and the lowest was 6.17 TgC located in the central and southern basins.The carbon reserves of the ecological regions was in the order of the eastern soil and rock mountain,Luliang soil and rock mountain,western loess hilly region,northern aeolian sand source region and central and southern basin region.The carbon density of the natural forest and plantation was maximum,occuring in mature forest(28.97 MgC/hm~2)and near mature forest(24.25 MgC/hm~2).There was no significant difference in carbon density among slope directions,and the carbon storage on shaddy slope was higher than that on sunny slope.The highest carbon density of both natural forest and artificial forest occurred at the elevation of 1500～2000 m.The spatial autocorrelation analysis showed that the carbon density of Shanxi Province and among the five ecological regions had strong spatial autocorrelation.Precipitation and canopy cover were the two main driving factors affecting the spatial distribution of tree carbon density in Shanxi Province,accounting for 20.78%and 17.15%of the spatial variation,respectively.The interaction of any two factors was greater than that of a single factor on the spatial distribution of carbon density,showing double synergies and nonlinear synergies,among them the interaction between precipitation and canopy cover is the strongest. |
PDF Full Text Request