Characteristics Of Carbon Density And Its Influencing Factors Of Pinus Massoniana Forest In Southern Jiangxi Province

As the main body of terrestrial ecosystem,forest ecosystem has huge carbon storage capacity and has an important impact on the global carbon cycle.Carbon density is one of the important indicators for evaluating forest carbon sequestration capacity of forest ecosystem.How to accurately estimate forest carbon density has received extensive attention.Pinus massoniana is one of the most widely distributed native tree species in southern China,which plays an important role in regional carbon cycle and climate change.This study takes Southern Jiangxi Province as the research area,using the data of Pinus massoniana sample plots of forest management inventory in 2019,calculation of its carbon intensity based on the Carbon Budget Model of the Canadian Forest Sector Model(CBM-CFS3),and the carbon density distribution characteristics were analyzed;the spatial distribution characteristics of carbon density were analyzed by geostatistical methods,and the main factors influencing the total carbon density of forest stands were screened using multiple stepwise regression and other methods;based on the comparison of different spatial regression models,the scale effect of the main influencing factors on the total carbon density of the stand was analyzed by the optimal spatial regression model,and the spatial distribution law was analyzed.The research results can provide reference basis for the development of carbon sink management strategies and the construction of carbon density estimation models at the regional scale for the Pinus massoniana forest in the study area.The main results and conclusions are as follows:(1)The Richards equation,Gompertz equation,Korf equation and Logistic equation were selected to construct the age-volume growth equation of Pinus massoniana forest with different origins of two forest categories,namely,ecological forest and commercial forest,respectively.Their optimal model was used as the driving equation of CBM-CFS3 model to calculate stand carbon density.After model evaluation and test,the optimal models of natural forest and plantation of ecological forest were Richards model and Gompertz model,respectively,while the optimal models of natural forest and plantation of commercial forest were Gompertz model and Richards model,respectively.(2)Based on the CBM-CFS3 model,the total carbon density of the stand was 135.08Mg C/hm~2,of which the carbon density of vegetation layer was 41.51 Mg C/hm~2 and the carbon density of the dead organic matter(DOM)was 93.57 Mg C/hm~2,accounting for 30.73%and 69.27%of the total carbon density of the stand,respectively.The order of each component of the vegetation layer was trunk(22.89 Mg C/hm~2)>branch(9.55 Mg C/hm~2)>root(6.41 Mg C/hm~2)>leaf(2.66 Mg C/hm~2).DOM carbon pool showed soil layer(75.94Mg C/hm~2)>litter(14.89 Mg C/hm~2)>dead wood(3.19 Mg C/hm~2).The global Moran’s I of the total carbon density of the stand is greater than zero,and it gradually decreases as the distance increases.At the same time,when the distance exceeded 180 km,the total carbon density of the stand has almost no spatial correlation.(3)Through multiple linear stepwise regression method,the main factors affecting the total carbon density of stand were age group,average DBH,canopy density,annual average temperature.At the same time,the least squares model,spatial error model,spatial lag model,geographically weighted regression model and multi-scale geographically weighted regression model were used to fit the relationship between the total carbon density of the stand and its influencing factors.The results showed that the multi-scale geographically weighted regression model had the best fitting effect,and its R~2,RMSE and AIC were 0.655,0.587 and 1232.265,respectively.And this model has obvious advantages in reducing the spatial autocorrelation of model residuals.(4)Scale effect and degree of different influencing factors on stand total carbon density based on multi-scale geographically weighted regression model.Among them,the age group are small-scale variables,canopy density is large-scale variables,while the average DBH and average temperature are global scale variables;the age group is the most important factor affecting the total carbon density of stands,followed by the average DBH,canopy density,and the smallest is the annual average temperature.The coefficient values of each influencing factor are quite different in different locations,and there are obvious spatial trends.At the same time,their spatial distribution patterns were also different.