Study On The Stand Density Management Diagram For Pinus Massoniana Forest And Individual-tree Basal Area Growth Model Of Close-to-nature Transformation Forest

China has the largest plantations in the world,and Pinus massoniana is one of the largest species of plantation which is widely distributed in South china.It is especially necessary to develop right management techniques for achieving their sustainable and multiple benefit.We take two measures to manage the Pinus massoniana plantation.Frist,To manage Pinus massoniana plantation,We Constructed the stand density managemengt diagram(SDMDS),and in order to maximize the forest function and guarantee the forest productivity,it is very important for the stand to maintain suitable density.Second,Close-to-Nature management for masson pine plantation,and constructing the individual tree growth model,Individual tree growth model predict the scenario of tree growth under certain site condition and thereby provide supportive information for formulating forest management regimes.We took 238 permanent sample plots as research object in the Center of Tropical Forestry.We Constructed the stand density managemengt diagram(SDMDS)for Pinus massoniana based on Reineke self-thinning equation.Due to the diagram,scenarios of variours management options were modeled to dertermine the best management option.Then we developed the denstiy stability plot for the Center of Tropical Forestry,which provided reference for judging the stability of Pinus massoniana plantation and put forward the management measures of density control.The study was based on the 20 plots of close-to-nature transformation forests in the Center of Tropical Forestry which is sited at Pingxiang County of Guangxi Province.The growth of basal area can be described as a linear function of tree size,site condition and competition.Finally,stepwise multiple linear regression were performed to produce the individual basal area growth model for Pinus Massoniana and precious broad-leaved trees.The resultant model was further refined using variance inflection factor(VIF)to investigate multicollinearity.The variables with large VIF were deleted from the model.(1)Construction and application of density management diagramWith RMA regression,self-thinning line of Pinus massoniana lanceolata plantation was developed.Whatsmore,the optimum growth interval line was determined as well as the canopy closure line.Developed dominant height model Tree volume model and site index model as the associated relationships for SDMDs,For the reasonable arrangement of operating density to provide a reference,we modeled the scenario of various management options by the diagram.(2)The stability evalua of Pinus massonianaAccording to the stand density managemengt diagram andforest resource inventory data to judgedthe stability of the stand,Between the line for up and lower density limit is tability,the other is not.different stands marked with different color,According to the state of forest stand to develop different forest management measures,to support the decision making of rational harvesting.(3)Individual accretion of close-to-nature transformation forestThe final predictive model included tree size of D2 1/D,competition indices which included the basal area in larger trees,The ratio of the square of diameter of subject trees to the average diameter of the plots and site condition.Model of R2 is over 0.7,the estimated effect is very good.The RMSE of the test model is 0.2547 and Bias is-0.00281,which indicates that the model has a certain stability and can well describe the growth of Pinus massoniana trees.(4)Individual accretion of precious broad-leaved treesThe research indicates tree size of D2.1/D,D,and site condition had a significant effect on the growth of broadleaf trees.In the early broad-leaved trees,the chest-high cross-sectional area is relatively small and influenced by Pinus massoniana.There is no competition between light and space in the broad-leaved trees.The main factors affecting the growth of trees are the trees themselves Size and soil conditions.The RMSE of the tested broad-leaved tree model was 0.3876 and Bias was 0.003108.The model has a better stability,the estimated more accurate.