Study On The Strata Characteristics Of Natural Broad-leaved Forests In Middle-subtropical Zone

Natural broad-leaved forest in mid-subtropical zone in China is a rare zonal vegetation in the world and one of the regions with the most abundant plant resources in China. It is meaningful to study their vertical structure for forests protection, utilization and management of that region. To reveal the natural stratification rules, the typical and sub-typical of natural broad-leaved forests in subtropical zone were selected as the experimental objects, and a new quantitative method was developed to divide the experimental objects into different forest forest stratum. Diameter distribution, tree height-diameter relation and stand description factors of each forest stratum were analyzed on the basis of stratum division.(1)Stratification defined. A new quantitative stratification method—maximum light receiving stratum was developed on the study to disclose the stratification pattern of typical middle-subtropical Natural broad-leaved forests According to the natural stratification pattern of natural broad-leaved forests in mid-subtropical zone, after profile diagram method, TSTRAT method, LMS method, and cluster method were made attempts on. To the typical forests, the method defined 3 strata, the stratumⅡ lower limited values were 17.0m,16.5m,17.0m,17.0m,16.0m, the stratumⅠ lower limited values were 25.0m,27.0m,25.0m,22.9m,25.0m.To the sub-typical forests,the method divided into 2 strata, the stratumⅠ lower limited values were 17.0m and 16.5m. The stratification results of the 7 sample plots were closed to profile diagram, and also complied with the various standards indicators of GBT 26424-2010.Compared with profile diagram method, TSTRAT method, LMS method, and cluster method,the new method reflects the natural stratification pattern of middle-subtropical natural broad-leaved forests better, stratification results essentially in agreement with the situation of trees whether accepted the direct light in the spot, and conform to the correlative regulations in the nation standard. The new method defined the strata basis on the trees whether can acceptthe direct light or not and height difference of direct light. Reflect the competition between the trees for light and space resources, has a certain biological significance.(2)Diameter distribution of typical forests. After the strata defined, the Shapiro-Wilk test was used to test normality of the diameter, the SK and KT were used to analyses graph shape features, Meyer negative exponential function and Weibull distribution function were used to discuss the strata(including stand and strata) diameter distribution, and the Chi-square test was set to inspect whether the diameter distribution obeyed or not. In typical experimental forests,the value of S-W test of the whole stand of all sample plots, all stratum Ⅱ and all stratumⅢ were less than 0.05, so all of them disobeyed the normal distribution; to StratumⅠ, the normal distribution were obeyed because the value of S-W of sample plot 1,2 and 3 were more than 0.05, but other plots did not expect; the W value of S-W increase with the increase of the mean height of the stratum, indicated that the diameter class of stratum tended to normal distribution with increase of the stratum mean height. The value of SK and KT were decreased with the increase of mean height of stratum, and the Stratum Ⅰ had the least abs-value of SK and KT meant the chat feature moving to the nomal distribution, were verified the rusults of S-W test. The results of Chi-test showed, sample plot 2 obeyed the Meyer negative exponential function, other plots did not; sample plot 1 did not obey the Weibull distribution function, and other plots did, the shape parameters c were less than 1, implied all the diameter distribution were reverse “J” shape. To stratum Ⅲ, all plots did not obey the normal distribution; sample plot 3-5 obeyed the Meyer negative exponential function; stratum Ⅲ of all the plots obeyed the Weibull distribution function. The diameter class of stratum Ⅲ were similar to the whole stand diameter class, except for the less diameter range, seemed as a censored reverse “J”shape. To stratum Ⅱ, all plots did not obey the normal distribution; sample plot 2 and 4obeyed the Meyer negative exponential, other plots did not; all the plots obeyed the Weibull distribution function, the shape parameter c between 1 and 3.6, suggested the diameter class of all plots were right partial mountain curve. Stratum Ⅰof sample plot 1,2 and 3 obeyed the normal distribution, but other plots did not; all plots obeyed both the Meyer negative exponential function and the Weibull distribution function. In general, the Weibull distributionfucution had the better flexibility in fitting the diameter distribution of strata of natural broad-leaved forest in mid-subtropical zone.(3) In the sub-typical forests(sample plot 6 and 7), the whole stand of sample plot 6 had a complicated diameter distribution, it liked as reverse “J” shape bounded up with right partial mountain curve, and it did not obey the normal distribution, the Meyer negative exponential function and the Weibull distribution function. The whole stand of sample plot 7 was a typical reverse “J” shape, and it obeyed the Meyer negative exponential function and the Weibull distribution function, did not obey the normal distribution. stratum Ⅰ of two sample plots did not obey the normal distribution and the Meyer negative exponential function, obeyed the Weibull distribution function, and the shape parameters between 1 and 3.6, suggested the diameter class of stratum Ⅰ were right partial mountain curve. stratum Ⅱ of sample plot 6did not obey the normal distribution and the Meyer negative exponential function, obeyed the Weibull distribution function; sample plot 7 obeyed the Meyer negative exponential function and the Weibull distribution function, did not obey the normal distribution. The W value of S-W test were similar to the typical forest, increase with the mean DBH of stratum increase.(4) Height-diameter relation of typical and sub-typical forests. Schumacher(S) function and Curtis(C) function were fitted for tree height-diameter relation of each forest stratum. C function was selected for analysis tree height-diameter relation because the results of C function has bigger R2 and smaller MASE and AMR. Models of tree height-diameter of whole forest stand were fitted and applied for the height-diameter relation analysis of stratum I and stratum II with scatter diagram. It was not showed a significant relation between height and DBH of stratumI and stratumII, and common models couldn’t fit that relation well. Bigger errors could appear when models of tree height-diameter of whole forest stand were used to extrapolate heights of trees in l stratum I and stratum II, the value of AMR was bigger than models fitted according each stratum. So that the typical forest stands with 3 strata could not described well by model of whole stand because of the bigger error.But to sub-typical experimental forests, the model of whole stand fitted the height-diameter in each stratum with asamller error, it could use the model of whole stand to fit the height-diameter in each stratum essentially.(5) Main stand description factors of typical and sub-typical experimental forests. The standard deviation and coefficient of variation were calculated for mean DBH, mean height,proportion of trees and their volume in each stratum. In typical stands, the total mean of DBH was approximately equal to that of stratum II. The coefficient of variation of total mean DBH and total mean height were bigger than that of stratums. The coefficient of variation of strata mean DBH became smaller with stratum height decrease. The coefficient of variation of stratum I and II mean heights were similar and less than that of stratum III. Proportions of trees in stratum I and stratum II just owned 20%-30% of whole typical forest stands, but their proportion volume were 90% of the whole. The mean height or median of stratum were substitute for tree height when calculating stand volume. All these three kind of height for estimating stand volume were tested by relative error analysis and variance analysis, and results showed that the relative error was less than 5% when using mean stratum height. If using the median, the relative error is less than 5% for stratum I and II, and 10% for stratum III.Since there were no significant different among the three methods for measuring stand volume,it was feasible by using mean or median of stratum tree heights to measure whole stand volume.In sub-typical forest, The standard deviation and coefficient of variation of mean DBH, mean height, proportion of trees and their volume in each stratum were similar to typica experimental forests; the different including total mean DBH were different to each stratum, the stems proportion of each stratum, sample plot 6 have more than 50% proportion in stratum I, that is shapely different from typical forests, proportion of stratum I in sample plot 7was 25%, liked to the typical forests.