Evaluation Of Site Quality In The Natural Spruce-fir Mixed Forests Of Changbai Mountains

Sixty plots of natural mixed spruce-fir coniferous and broadleaved forest located in Jingouling forest farm of Wangqing Forestry Bureau,Jilin Province were investigated to evaluate their site quality by usng the relationship between environment factors(topographical factors and soil factors)and site form.To solve the mixed and unevenaged problems in the natural forest,we divided the forest into 3 tree species(groups)including spruce,fir,and deciduous group(ribbed birch,amur linden,white birch,elm,other broadleaved trees).The dominant trees were selected in each tree species(groups).The guide curves were fitted with the relationship between the dialeter at breast height(instead of age)and dominant mean height.The site form tables of 3 tree species(groups)were constructed by using standard deviation adjustment method.Three BP neural network models were built with topographical factors,soil factors as input data and dominant mean height and average diameter at breast height as output data.We evaluated the site quality of 60 plots on the level of each tree species(groups)and overall forest in the hope of providing a theoretical basis for matching species with the site and forest management.It could also be used to provide reference for other researches of natural forest The main results were as follows:(1)Fitting the guide curves and constructing the site form tables by using the relationship between the diameter at breast height(instead of age)and dominant mean height were feasible and the results showed that the fitting curve seemed good,and the site form table was reasonable and accurate.(2)The model successfully added topographic factors and soil factors to site quality evaluation using BP neural network,which makes the evaluation results more objective and accurate.The model does not depend on the stand factors and thus can be used for site quality evaluation of non-forest land.(3)The site quality of 60 plots were evaluated and graded.The plots were divided into 5 levels:the rating level of Ⅰ(excellent)has 3 plots,Ⅱ(good)has 20 plots,and they account for 38.3%of the total.There is no V grade(very poor)plots,which means the site quality is good as a whole,but there are still 34 plots on the level of Ⅲ(medium)and 3 plots on the level of IV(poor)which indicate the site quality of many plots need to be improved.