Objective The development process of forest stand is reasonably divided into multiple stages, and corresponding forest management strategies can be formulated for forests at different stages of development in a more targeted manner. However, the division of natural forest development stages is difficult, and there is no best method, therefore, this study takes the Changbaishan spruce conifer mixed forest as an example, aiming to compare the methods of the development stages of five forest stands, and clarify the most suitable method for the Changbaishan spruce coniferous mixed forest.

Method A total of 2589 sample plot samples were divided by interspecies linkage-optimal segmentation method (A), forest facies characteristic discrimination method (B), TWINSPAN bidirectional indicator species analysis method (C), MRT multiple regression tree method (D), and TWINSPAN-based discriminant analysis method (E) of Jingouling Forest of Wangqing Forestry Bureau of Jilin Province. Multiple comparisons were used to verify whether there were significant differences between different stages of stand development, and the anastomosis coefficient was used to compare the anastomosis between the classification results of different methods, and finally the classification results of the five methods were compared from the aspects of tree species composition and stand characteristics.

Result The results of multiple comparisons show that the nine numerical indicators used to describe the characteristics of forest stands in this paper basically have significant differences between different stages of stand development, indicating that the results are more reliable. Method D and Method E had the highest anastomosis coefficient, indicating that the division results were more consistent, while Method B had a low agreement with other methods. On the whole, the classification results of each method were generally that with the gradual development of forest stands, the characteristic indicators of forest stands increased, the diversity index decreased, and the composition of tree species tended to be simple. The differences are mainly reflected in the number of hectares and tree species composition indicators, as well as the degree of difference between different indicators and the sample size of each stage. From all aspects, the division results of Method E are more in line with the growth and development law of forest stands, The first stage is the formation of spruce-fir-coniferous mixed forest, the rapid growth of forest stands, the second stage is the strengthening of inter-forest competition and significant individual differentiation, and the third stage is forest division into near-natural forest, which is the target state of forest management.

Conclusion The discriminant analysis method based on TWINSPAN can better divide the development stage of forest stands, and divide the development process of Changbai Mountain spruce-fir-coniferous mixed forest into group establishment stage, competition stage and near-natural forest stage, which provides an important theoretical basis for forest scientific management.