Chin J Plant Ecol ›› 2024, Vol. 48 ›› Issue (1): 68-79.DOI: 10.17521/cjpe.2023.0120

Special Issue: 植被生态学

• Research Articles • Previous Articles     Next Articles

Mechanisms of seedling community assembly in a monsoon evergreen broadleaf forest in Pu’er, Yunnan, China

CHEN Zhao-Quan1,2,3, WANG Ming-Hui1,3, HU Zi-Han1,3, LANG Xue-Dong1,3, HE Yun-Qiong4, LIU Wan-De1,3,*()   

  1. 1Institute of Highland Forest Science, Chinese Academy of Forestry, Kunming 650224, China
    2College of Landscape Architecture, Nanjing Forestry University, Nanjing 210037, China
    3Pu'er Forest Ecosystem Research Station, National Forestry and Grassland Administration of China; Pu’er Forest Ecosystem Observation and Research Station of Yunnan Province, Pu'er, Yunnan 665000, China
    4Yunnan Association of Scientific Research Institutes, Kunming 650228, China
  • Received:2023-05-04 Accepted:2023-11-09 Online:2024-01-20 Published:2023-11-09
  • Contact: (
  • Supported by:
    Yunnan Fundamental Research Projects(202001AS070005)


Aims Seedlings play a crucial role in the conservation of forest biodiversity. Previous studies have paid little attention to seedling communities in monsoon broadleaf evergreen forests in Pu’er, Yunnan. Our aim was to investigate the mechanisms of seedling community assembly there.

Methods We analyzed the species composition of the seedling community based on the seedling survey in a 30 hm2 forest dynamics plot. Seedling plots are divided into different groups according to the dominant species. The stepwise community assembly model (STEPCAM) was used to investigate the mechanisms of community assembly and to further analyze the functional traits of seedling communities.

<strong>Important findings</strong> Castanopsis echidnocarpa and C. calathiformis are the dominant seedling species in the monsoon broadleaf evergreen forest. Four plot types (i.e., C. echidnocarpa plot, C. calathiformis plot, mixed-dominant species plot, non-dominant species plot) were classified according to the presence or absence of dominant species. The seedling community assembly processes included: a) random dispersal assembly (with contribution rate 43.1%-61.3%); b) habitat filtering (with contribution rate 27.4%-33.9%); and c) limiting similarity (with contribution rate 5.7%-27.2%). The C. echidnocarpa and C. calathiformis plots were dominated by deterministic processes, with 56.9% and 54.6% contributions respectively, whereas the mixed-dominant and non-dominant species plots were dominated by stochastic processes, with 60.4% and 61.3% contributions respectively. Rao’s quadratic entropy (Rao’Q) was highest in the non-dominant species plot and lowest in the mixed-dominant species plot. The C. echidnocarpa plot has the lowest specific leaf area (SLA) and higher leaf thickness (LT), specific stem length (SSL), root mass fraction (RMF) and a maximum potential plant height (PPH), while the C. calathiformis plot and non-dominant species plot have a higher SLA. Among different seedling plots, the coefficients of variation for SSL and PPH were higher, while the coefficients of variation for other functional traits were lower. Multiple regression analysis showed that SLA and PPH were significantly and positively correlated with Rao’Q in different seedling plots. LT and life form were significantly and positively correlated with Rao’Q in other three types of plots, except for the non-dominant species plot. Thus, seedling community assembly in the monsoon broadleaf evergreen forest is driven by both stochastic and deterministic processes, and the contribution of the two processes varies with the type of seedling communities.

Key words: community assembly, dominant species, functional trait, functional diversity, monsoon evergreen broadleaf forest