Abstract: [Objective] To provide scientific basis for optimizing the management models of natural secondary forest in northeast China and improving forest quality.[Method] Take natural secondary forest of Quercus mongolica of Danqinghe Experimental Forest Farm of Harbin Forestry Bureau in Heilongjiang Province was studied, and a combination of random sampling and typical sampling was used to survey the sample-plots to study the influences of target tree management model, comprehensive tending model and non-interference model on the structure and diversity of plant community.[Results]After 17 years of management practice, the extent of dominance of Quercus mongolica in the target tree management model was significantly reduced, while those of Pinus koraiensis and Picea koraiensis were increased greatly, and they became dominant species. The dominance of Quercus mongolica in the comprehensive tending management and non-interference models remained the same, while that of other tree species was low but the composition of the tree species changed slightly. The new precious tree species (Phellodendron amurense and, Juglans mandshurica) appeared in the target tree management and comprehensive tending management models. The absolute values of basal area at breast height of Quercus mongolica were very different among the three models, showing the target tree management model (5.3 m²·hm^-2) < comprehensive tending model (11.0 m²·hm^-2) < non-interference model (21.6 m²·hm^-2). Quercus mongolica had the greatest advantage in all models, and was significantly less advantageous in the target tree management model (55.1%) than in the comprehensive tending model (78.1%) and the non-interference model (79.6%). The basal area (2.8 m²·hm^-2) and the extent of dominance (32.3%) of Pinus koraiensis in the target tree management model were significantly higher than those in other two models. The average density and average tree height of the forest stand in the non-interference model were significantly larger than those of the other two models (P<0.05), and there was no significant difference between the target tree management model and the comprehensive tending model. There was no significant difference among the average diameters at breast height of the forest in the three models. From the diameter class structure, except the two diameter classes of 1-5 cm and 35-40 cm, the number of other diameter classes in the non-interference model was significantly larger than that in the target tree management and the comprehensive tending models. From the vertical structure, the non-interference model was obviously larger than the target tree management model and comprehensive tending model at all layers > 10 m. There was no significant difference (P<0.05) in tree species diversity of arborous layer in the target tree management, comprehensive tending and non-interference models. The diversity of species (P<0.05) of shrub layer in the comprehensive tending model was significantly greater than that in the target tree management and non-interference models. The richness index of species in herbaceous layer in the target tree management and comprehensive tending models was significantly greater than that in the non-interference model. There was no significant difference in other types of species diversity indices in the tree management models (P<0.05). The similarity of forest stands between the target tree management model and the comprehensive tending model was the highest, while that between the target tree management model the non-interference model was the lowest.[Conclusion]The target tree management model successfully helped introduce the later-period tree species into the forest stand by replanting species such as Pinus koraiensis, and promoted the growth of some precious trees. In the short term, the productivity of the forest stand without interference was higher than that of the forest stand in the target tree management and the comprehensive tending models. In terms of the natural secondary forest of Quercus mongolica, how to improve the growth of forest stand through forest management in a long time requires long-term monitoring and in-depth study.

Key words: natural secondary Quercus mongolica forests, target tree forest management, comprehensive tending management, vegetation structure, plant biodiversity, species similarity