| The canopy and its growth are restricted by stand factors(such as site conditions and stand density),At the same time,the canopy is also an important indicator for reasonable regulation of stand density.The prediction of the shape and size of the canopy can directly reflect the change of the spatial distribution structure of the forest stand,and has important guiding significance for the regulation of reasonable density in the process of tree growth.In this study,the Larix principis-rupprechtii plantation forest in Saihanba mechanical forest farm was taken as the research object,and the crown profile model and crown width model were constructed based on 21 sample plots data of different forest ages.Using the analytical data of 10 sample plots of different ages,58 average trees and 1789 branches,using pearson correlation analysis method,the stand factors and tree factors were analyzed with the maximum crown radius(MOR)and the maximum radius relative crown depth(MRDINC).Based on the power function,modified Kozak equation,Weibull equation,the basic model was selected,and the outer contour of the canopy of Larix principis-rupprechtii plantation was predicted by constructing a two-level mixed effect model of sample plots and samples and quantile regression model,and the crown profiles of trees of different ages were simulated..Using the crown width data of 5553 trees in 21 standard plots,the relationship between stand factors and tree factors and crown width was analyzed using pearson correlation analysis method.A basic model was selected from 10 common crown width models,and a two-level mixed-effect model based on density level and plot level was constructed to analyze the impact of stand factors or tree factors on crown width.The main conclusions are as follows:(1)Correlation coefficients of maximum canopy radius(MOR)and stand age(Age),breast diameter(DBH),crown length(CL),tree height(HT),crown length ratio(CR),height-to-diameter ratio(HDR)were 0.5618,0.7135,0.6264,0.5656,0.7395,-0.5647 and P<0.0001.MOR is directly proportional to stand age(Age),breast diameter(DBH),crown length(CL),tree height(HT),crown length rate(CR),and inversely proportional to height-to-diameter ratio(HDR).Maximum crown radius corresponds to crown depth(MRDINC),stand age(Age),breast diameter(DBH),crown length(CL),tree height(HT),crown length ratio(CR),and height-to-diameter ratio(HDR)are inversely proportional.Correlation coefficients of crown width and density(Sd),crown length(CL),dominant tree average breast diameter(DD),tree height(HT),dominant tree average height(DH),height to diameter ratio(HDR)were-0.6217,0.3341,0.5351,0.4177,0.4005,-0.5094.Crown width is greatly affected by density(Sd),crown length(CL),dominant tree average breast diameter(DD),tree height(HT),dominant tree average height(DH),and height-to-diameter ratio(HDR).The mean height(DH)and tree height(HT)are positively correlated,but negatively correlated with density(Sd)and height-to-diameter ratio(HDR).(2)Compared with the modified Kozak equation and the modified Weibull equation,the power function model determines that the coefficient R2 is 0.712,the root mean square error(RMSE)is 0.495,and the average relative error(ME)is 15.388.The mixed effect model of single wood level AIC is 408.1755,BIC is 450.3415.Considering the two-level mixed effect model of plot and sample wood,the AIC is 350.0507 and the BIC is 396.4334.The two-level mixed effect model significantly improves the accuracy of the model.In the two-level mixed effect model,the parameter a6 considers the plot effect,the parameters a4 and a5 consider the sample effect,and the model determination coefficient(R2)It is 0.873,the root mean square error(RMSE)is 0.319m,and the average relative error(ME)is 6.642m.In the quantile regression model,R2 reaches the maximum value at the quantile of 0.65 to 0.75.In the quantile regression model,R2 reaches the maximum value at the quantile point of 0.65 to 0.75.The quantile regression model can better describe the average change trend of the outer contour of the largest canopy,but it does not well represent the largest outer contour of the canopy.With the increase of q,the curve of the quantile model is infinitely close to the maximum outer contour of the canopy,and the quantile regression model of 0.9 quantile relative to other quantiles is closest to the maximum outer contour of the canopy.(3)The logistic equation model determines that the coefficient R2 is 0.593,the root mean square error(RMSE)is 0.735,and the average relative error(ME)is 14.963.It is the optimal basic model for describing the crown amplitude model of Larix principis-rupprechtii.The fitting evaluation indexes of the mixed effect model are better than the corresponding basic model.The mixed effect model R2 is 18.89%higher than the basic model,RMSE is 17.87%lower than the basic model,and ME is 32.03%lower than the basic model.In the simulation of the crown width model,the crown width is inversely proportional to the average tree height(DH)and height-to-diameter ratio(HDR)of the dominant tree,and inversely proportional to the tree height(HT).The average height of dominant wood(DH)and tree height(HT)affect the crown width with the increase of DBH,and the height of the first live branch height(HCB)does not follow the(DBH)change and change. |
PDF Full Text Request