Abstract: The large forest dynamics monitoring plot is an important research platform for answering key questions in ecology, such as species coexistence and community succession. According to the protocol set forth by CTFS-ForestGEO, a 20-ha tropical seasonal rainforest dynamics monitoring plot was established by the Administration Bureau of the Nabanhe River National Nature Reserve and the Xishuangbanna Tropical Botanical Garden (Chinese Academy of Sciences) between 2016–2017. Here, we analyzed the species composition and classified community types for this large forest dynamics monitoring plot, where all free-standing woody plants with diameter at breast height (DBH) ≥ 1 cm were tagged, measured, and identified to the species level. There were 59,498 individuals with DBH ≥ 1cm belonging to 296 species, 197 genera, and 63 families. There were 15 species designated as common with individuals numbering more than 1,000 and 126 rare species with fewer than 20 individuals in the plot. At the family level, tropical-associated families were dominant, accounting for 65.08% of the total family; temperate-associated families only accounted for 17.46% of the total family. At the genus level, the tropical-associated genera were dominant, accounting for 88.32% of the total genus. temperate genera only accounted for 10.66%. Elevation and convexity are the key topographical factors underlying the formation of community types in this plot. There were four distinct associations identified among 500 quadrates (20 m × 20 m) as follows (1) Cipadessa baccifera + Mallotus philippensis – Harpullia cupanioides + Pometia pinnata Association; (2) Trigonostemon thyrsoideus + Boehmeria glomerulifera – Acrocarpus fraxinifolius + Pometia pinnata Association; (3) Aporusa yunnanensis + Turpinia pomifera – Castanopsis mekongensis + Phoebe puwenensis Association; (4) Ficus cyrtophylla + Leea compactiflora – Sladenia celastrifolia + Pometia pinnata Association.

Key words: large forest dynamics monitoring plot, floristic characteristics, multivariate regression trees, association