Abstract:

Objective: The objective was to stratify the arbor trees of a mid-subtropical forest community with a single dominant species of Castanopsis kawakamii, in order to provide a precondition for further study of the characteristics and dynamics of each layer. Method: Two sample plots in a size of 50 m×50 m of C. kawakamii forest community located in the provincial nature reserve of C. kawakamii were investigated respectively using the methods of maximum light receiving plane (MLRP), profile diagram, and model-based clustering (MCLUST) to stratify the forest community of the two sample plots. Result: The two sample plots were all closed forest stands and the canopy density were all more than 0.95. In sample plot 1, the stand density and volume were 424 trees·hm² and 588.04 m³·hm², and the average DBH was 39.2 cm, the average tree height was 26.69 m. In sample plot 2, the stand density and volume were 542 tree·hm² and 417.19 m³·hm², and the average DBH was 29.1 cm, the average tree height was 26.69 m. All the three methods can solve the problem of stratifying the arbor layer, but MLRP method was the optimal because it is relatively simple and convenient with less field work and having explicit biological interpretability. According to the MLRP, the arbor layer of the C. kawakamii forest community should be divided into two sub-layers which were consistent with the results of the profile diagram and the MCLUST. The upper layer was light receiving layer (LRL) and the lower layer was none-light receiving layer (NLRL). The thresholds between LRL and NLRL of the two sample plots were 16.00 and 16.90 m. In sample plot 1, the ratio was 6:4 for counts of trees of LRL and NLRL, and 99:1 for volume. In sample plot 2, the ratio was about 3:7 for counts of trees of LRL and NLRL, and 97:3 for volume. In sample plot 1, C. kawakamii distributed in LRL accounted for 40.6% of the total number of trees in the arbor layer and 76.9% of the total stand volume; in NLRL that were 3.8% and 0.4%. As to sample plot 2, C. kawakamii in LRL accounted for 27.0% of the total number of trees, and 84.1% of total stand volume; in NLRL that was 5.7% and 0.2%. Conclusion: The MLRP method is more suitable than the other two methods for stratifying the arbor layer of the C. kawakamii community. Both communities can be divided into LRL and NLRL. C. kawakamii is absolutely dominant in terms of number of trees and the volume in LRL but not NLRL. This vertical distribution pattern implies in sufficient regeneration and recruitments under forest, which can be a limitation for the sustainable development of C. kawakamii population. Further studies are needed to characterize the layers and their relationships, as well as the dynamics of formation and development of the layers, providing scientific basis for the protection of natural C. kawakamii forests.

Key words: mono-dominant community, vertical structure, forest strata, stratification of arbor layer, maximum light receiving plane method, mid-subtropical zone