Analysison Relationship Among Majoy Factors Of Several Mangroves In Dongzhai Harbour, Hainan Island

In Hainan Dongzhaigang Mangrove National Nature Reserve, six permanent plots in six different mangrove communities that locate at Hegang, Sanjiangwan and Daoxue three different places were set up with adjacent mesh method. Every tree in the plots was investigated. We mesured several indexes of these trees including height,diameter at breast height (or collar diameter),crown breadth,height blow branch, then chose their data of Bruguiera sexangula (H<2m and H≥2m), Sonneratia apetala Buch-Ham, Sonneratia caseolaris Engl., Kandelia candel, Ceriops tagal(H<2m) for relative analysis of these major characteristic factors by SAS program. The curving model of height, and relative relationship between height and diameter, crown breadth and diameter, average crown size and height, height and height below branches were studied by multivariate linear regression analysis. Comparing different modeling, we established the relative fine model. The main results are following:1.Through multivariate linear regression analysis, the multivariate linear regression equation which was used to caculate height of tree was established by use of these factors including tree diameter at breast height, average crown size and height-below-branches. After inspecting (P<0.01), the regression equations are in significant level. Mean estimated precision of the regression euqation of height of Bruguiera sexangula and Sonneratia caseolaris Engl. (all above 99%) was considered relatively fine, so we can use it to estimate their height. The estimated precision of regression equation of height of Sonneratia apetala Buch-Ham was also high to 97.86%. And the precision of multivariate linear regression equation of height of Kandelia candel was up to 92.01%.2.Research on the correlation between height and tree diameter at breast height was carried on, and then selected linear or nonlinear regression equation according to different plants species. After inspecting, when P<0.01, value of R2 or mean square stands low, that means significant level. The estimated precision of linear regression equation is above 93%, and above 94% of nonlinear regression equation. According to this result, it shows the nonlinear regression equation is better than the linear equation. In conclusion, the precision of these regression equations is so fine that we can use to forecast height of trees by means of index named tree diameter at breast height.3. It can be concluded from analysis and prediction of model of average crown size (x₂) and tree diameter at breast height (x₁) that indicated a close correlation between the two indexes. The result predicted by two model equations x₂ = ax₁ b and x₂ = ae^(-bx1) is the best, and 1/x₂=a+b/x₁ is opposite to it. When inspecting precision, the estimated precision of regression equation ranges from 84% to 95%. As is so low and wide, the model suitable to study the correlation between average crown size and height should be done further research work.4. S After research on the correlation between height (y) and average crown size (x₂), some regression equations suitable to predict the height of tree were showed. Equation model y=ax₂^b is the best, but y=a+bx₂ is not good. Laterly inspected, results of these equations are all significant or quite significant. The precisions range from 95.83% to 99.26%. As the bad precision of result, just 73.23%, further research should be done about the correlation model between the height and average crown size of Ceriops tagal. In all, result of the height regression equation established by tree diameter at breast height is better than that by crown size.5.When researching on the linear and linearized relative correlation among height(y),tree diameter at breast height (x₁) and average crown size (x₂), this paper selected some regression equation aiming to different tree species. After inspected, these equations are fine, because of their high estimated precision above 96%.6. Different regression equations suitable to different trees were set up after researching on the nonlinearized regression among height(y),tree diameter at breast height (x₁) and average crown size (x₂). Mean square of every equation is low and when carried on precision inspected, the estimated precision of the nonlinearized regression equation between height, tree diameter at breast height and average crown size of Sonneratia apetala Buch-Ham is the lowest to79.34%, and that of Bruguiera sexangula (H<2m) is 91.34%. The estimated precison of other regression equation is relative high, above 97%.7.Resarch on the correlation between height and height-below-branches resulted in some regression equation suitable to different trees. As the precision examined, estimated precision of all regression equations is relative high above 98%, except low precision of 89.70% of the linear regression equation about height and height-below-branches of Bruguiera sexangula, so they can be used to estimate trees height by height-below-branches. From this result we can see indirectly the characteristics of natural branching procession of mangrove, and its relationship with height of trees.8.Author selected regression equations suitable to different tree through research on the nonlinearized regression equation about height, height-below-branches and average crown size. After examined, when P<0.01, value of R2 or mean square is low. And through precision inspected, precision of these nonlinearized regression equation lie in between 90% and 96%. That means the result of nonlinear regression equation is better than linear.9.It indicates another ecology problem that is the significant influence of frequent typhoon on the growth as high as possible of mangrove in Hainan through the relation analysis of height and tree diameter at breast height of several species such as Sonneratia apetala Buch-Ham of Dong Zhai Gang Reserve . At the same time, it proved some scholars'viewpoints fom aspects of mangrove plants that the tropical forest trees in Hainan Island of China are not higher than those in Malasia.10.It indicates that, artificial mangrove plants grow fast and the rate of growing high is faster than that of natural branch reorganizing. In order to adapt to stern habitat, always in big winds and waves, mangrove must branch fast and its crown grow wide rapidly. That may be the cause of the function of mangrove as windproof and alleviating & preventing sea wave.11.Through the analysis of the relation between crown breadth and tree diameter at breast height of several species such as Kandelia candel in Dong Zhai Gang Reserve , we concluded that because of special habitat of mangrove, it resulted in growth differentiation of crown size between mangrove plants and land plants. According to the growth habitat differentiating, the crown breadth also is varying.