Biodiv Sci ›› 2023, Vol. 31 ›› Issue (11): 23251.  DOI: 10.17520/biods.2023251

• Special Feature: Semi-humid Evergreen Broad-leaved Forests • Previous Articles     Next Articles

Population structure and regeneration characteristics of dominant tree species in a semi-humid evergreen broad-leaved forest in the Jizu Mountains, Yunnan

Wencong Liu1, Xi Tian1, Tao Yang1, Jiesheng Rao1, Xiaofeng Wang1, Hengjun Qian1, Mengling Tu1, Ziming Shan1, Xiaokun Ou1, Zehao Shen2,1,*()   

  1. 1 School of Ecology and Environmental Science, Yunnan University, Kunming 650091
    2 Kay Laboratory of MOE for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, Beijing 100871
  • Received:2023-07-11 Accepted:2023-11-19 Online:2023-11-20 Published:2023-12-29
  • Contact: * E-mail:


Aim: Subtropical semi-humid evergreen broad-leaved forest is a unique forest type in Southwest China. We expect to enhance our understanding of the characteristics and drivers of community dynamics in this ecosystem by studying the spatial and temporal patterns and regeneration strategies of the dominant tree species.

Methods: Using the first community survey data of the forest dynamics plot in a semi-humid evergreen broad-leaved forest in the Jizhu Mountains, Yunnan Province, we explored the patterns of seedlings, saplings, and adults of five dominant tree species and compared their population structure in DBH classes. The multi-scale spatial patterns of regeneration and adult populations were modeled using a probability density function g(r). A comparison of ordinary least squared and multi-scale geographical weighting regression models were used to assess the factors influencing the seedling and sapling density of the dominant tree species.

Result: (1) Castanopsis orthacantha, C. delavayi, Schima argentea, and Lithocarpus craibianus exhibited a growing population DBH class structure. The Pinus yunnanensis had a declining population structure. (2) The patterns of regeneration and adult populations of all five species were clustered at small scales (0-50 m), but the intensity of clustering decreased with increasing scales, shifting to random or uniform distributions. The distributions of juvenile and adult populations of Pinus yunnanensis showed a significant negative correlation at small scales, while the distribution of juvenile and adult populations of the other four broad-leaved species showed a significant positive correlation. (3) The multi-scale geographical weighting regression model explained 58.7% and 26.9% more than the ordinary least squares model for the change of population regeneration densities of seedlings and saplings, respectively, showing a significant effect of spatial autocorrelation in population patterns. The model results indicated that the seedling density pattern was constrained more by stochasticity in seed dispersal than by habitat factors, while the sapling density pattern was dominated by that of habitat factors.

Conclusions: Except for Pinus yunnanensis, the other four dominant tree species in the sample plot of the semi-humid evergreen broad-leaved forest in the Jizhu Mountains regenerate well. The spatial patterns of regeneration of all five species are constrained by both stochastic dispersal limitation and deterministic environmental filtration, but the dominant constraint shifts from the former to the later as regeneration progressesfrom the seedling phase to the sapling phase.

Key words: semi-humid evergreen broad-leaved forest, DBH class structure, spatial structure, natural regeneration, influencing factors