The Biomass Model And Distribution Pattern For Main Tree Species Of The Secondary Forest In Western Hunan

Single tree biomass model is an effective method to calculate forest biomass,and the biomass distribution pattern shows the Plant resource allocation mode.In order to measure the secondary forest biomass in Western Hunan accurately,explore the biomass distribution pattern of standing trees,improve the prediction accuracy of compatible biomass model,to provide a theoretical basis for carrying out forest resources investigation,predicting forest trees future development trend,and provide a breakthrough for understanding the growth characteristics and growth mechanism of trees.The thesis taked Cyclobalanopsis glauca,Pinus massoniana Lamb.,Quercus acutissima Carruth,Castanopsis sclerophylla(Lindl.)Schott and Castanopsis hupehensis(C.S.Caho)as the object,used harvesting method to obtain biomass of above ground components,fitted independent biomass model and compatible biomass model of 5 tree species,and selected the optimal biomass model and parameter estimation algorithm by relative ranking method,analyzed 5 tree species biomass allocation pattern,and explored the influence of diameter at breast height and age on biomass allocation pattern.The main results are as follows:(1)The biomass model with diameter at breast height has achieved a good fitting effect.After adding tree height and crown length,a few species biomass model fitting accuracy has been improved.Combining mean rank and model evaluation index to select the optimal model,the biomass models of aboveground components were as follows: Cyclobalanopsis glauca ln W=2.3112 *ln D +1.0206 *ln H-4.0331,Pinus massoniana ln W=2.0276 *ln D-1.1454,Quercus acutissima ln W=2.5337 *ln D-2.1037,Castanopsis sclerophylla ln W=2.2164 *ln D-1.8814,Castanopsis hupehensis ln W=2.3298 *ln D-1.6593.(2)The purpose of forestry is calculating the aboveground biomass.The controlling directly by proportion functions based on the ordinary least squares can calculate the above-ground biomass immediately,has high prediction accuracy and good application prospects.Using the same parameter estimation algorithm,there was little difference between the 2 compatible biomass models;Compared with ordinary least squares,two stage least squares and Three-Stage Least Squares,the fitting accuracy of ordinary least squares is highest and the calculation process is simplest.(3)Compared ordinary least squares with and seeming unrelated regressions,it is still recommended to use ordinary least squares as parameter estimation algorithm.Compared with seeming unrelated regressions,the compatibility biomass model fitted by ordinary least squares has higher fitting accuracy,and the fitting process is easier.(4)There were differences in biomass distribution patterns of 5 tree species.Total trunk ratio,total branch ratio,total leaf ratio of Pinus massoniana were 83.28%,11.51% and 5.22%,respectively;Cyclobalanopsis glauca were 81.81%,14.20% and 3.99%;Quercus acutissima were 86.94%,9.86% and 3.21%;Castanopsis hupehensis were 82.19%,15.93% and 1.89%.These 4 tree species had similar biomass allocation ratio,and their average total trunk ratio,total branch ratio,total leaf ratio were 80.67%,14.47% and 4.88%,which were quite different from Castanopsis sclerophylla(69.45%、19.95%、10.63%).With the change of tree age and diameter at breast height,there were differences between total trunk ratio,total branch ratio,total leaf ratio of Pinus massoniana and Cyclobalanopsis glauca,but on the whole,2 tree species all showed that: total trunk ratio>total branch ratio>total leaf ratio.