Study On Stand Volume And Biomass Growth Model Of Main Tree Species In Guangdong Province

The calculation of forest volume and biomass is related to many forestry and ecological problems,and its is an important data source to study forest carbon storage,forest productivity,material flow and energy cycle of forest ecosystem.The growth model is one of ways to express the volume and biomass,and it can predict the future forest growth without repeated investigation,and it has the characteristics of fast,economical and reliable.In order to establish the growth equation of forest stock volume and biomass in Guangdong Province.Data were collected from 358 natural forest plots,including Pinus massoniana,Pinus elliottii,Cunninghamia lanceolata,Quercus spp.,Phoebe zhennan,Schima superba,other hard broad leaf species,Acacia confusa and other soft broad leaf species.The forest resources inventory data were collected in 1997,2002,2007,2012 and 2017.According to the relationship between stand average DBH and stand age,there were eight stand growth models of different parameters(growth potential classification and growth rate classification),which considered the competition or not,stepwise modeling(unitary nonlinear regression)or joint modeling(nonlinear error variable simultaneous equations).In the stepwise model,after the initial age and other parameters were estimated by the Mitscherlich growth model,the ordered sample clustering was performed on parameter a or b,and the volume and biomass growth models were constructed based on the Richards growth model.In the joint model,after the ordered sample clustering of parameter a or b,the stand average DBH was constructed based on the Mitscherlich growth model,and the nonlinear error variable simultaneous equations of volume and biomass were constructed based on the Richards growth model.In order to verify and compare the feasibility and stability of volume and biomass models based on multi-temporal averag DBH data,eight growth equations were used to fit the stand average DBH of the four forest resources inventory data,and the accuracy of volume and biomass estimation of different species was compared and evaluated by different models.The main results are as follows:(1)For the 8 models based on the growth equation of Mitscherlich and Richards,the evaluation indexes of the models are all very good,the determination coefficient and the modified determination coefficient are mostly above 0.95,the average prediction error is mostly less than 2%,and the overall relative error is less than±1%,and most of them tend to be 0.(2)Considering the competition or not,the volume model is superior to the biomass model.Among the eight model systems,the joint modeling based on parameter b classification was the best.The determination coefficients of the volume model and biomass model were consistent with the modified determination coefficients regardless of competition.(3)The evaluation results show that the model fitting effect is good,and the estimated precision parameters of the model are all less than 10%,indicating that the estimated accuracy is above 90%.Among them,the joint estimation with competition based on parameter a classification,the joint estimation without competition based on parameter b classification and the joint estimation with competition based on parameter b classification have the highest precision and the model fitting effect is the best.The accuracy of the joint model is higher than that of the stepwise model,and the accuracy of the model with competition is higher than that without competition.Except Pinus massoniana,the accuracy of the other eight species’volume model is higher than that of the biomass model.(4)Among the dominant tree species in Guangdong Province,the order of volume growth potential from high to low was:Pinus massoniana>other hard broad leaf species>Schima superba>Quercus spp.>other soft broad leaf species>Phoebe zhennan>Acacia confusa>Pinus elliottii>Cunninghamia lanceolata.The tree with the highest growth potential was Pinus massoniana with the maximum value of 2131.90 dm~3,and the tree with the highest growth potential was Cunninghamia lanceolata with the minimum value of 838.74 dm~3.The order of biomass growth potential from high to low was:Pinus massoniana>other hard broad leaf species>Quercus spp.>Schima superba>other soft broad leaf species>Acacia confusa>Phoebe zhennan>Pinus elliottii>Cunninghamia lanceolata,the tree with the highest growth potential in stock volume was Pinus massoniana,with the maximum value of 2191.60 kg,and the tree with the lowest growth potential was Cunninghamia lanceolata,with the minimum value of 544.65 kg.In general,the growth potential of broad-leaved species is higher than that of coniferous species.(5)Geographical factors,topographic factors,climate factors and soil factors all had effects on the volume and biomass,and geography,climate and soil factors had greater effects,while topographic factors had less effects.In this paper,a stand growth model was established based on the relationship between DBH,stock volume and biomass based on the continuous forest inventory data of five periods.The combined model based on parameter growth rate classification was the best one.The results of model validation show that with competition is better than without competition.Joint modeling is better than stepwise modeling.According to the modeling analysis of different tree species,the growth potential of broadleaf species is greater than coniferous species.The stand growth model established in this paper has realized the accurate estimation of the volume and biomass in Guangdong Province.The research can provide scientific reference for the data of stand stock volume and biomass and in many fields,and it can provide scientific reference for the management and sustainable management of forest resources.