Background & Summary

Mountains cover about 25% of the world’s land surface, host more than 85% of the world’s amphibians, birds and mammals and play a crucial role in maintaining biodiversity as refuges, biodiversity hotspots and ecological corridors that support ecosystem function and stability1,2,3. Due to steep environmental gradients, diverse microhabitats, and complex topographical features, mountain plants exhibit high diversity, distinct distribution patterns, and specialized adaptations to varying climatic and soil conditions4,5. Meanwhile, as a result of geographical isolation and adaptive evolution, many mountain plants are highly endemic, resulting in a wealth of endemic species and unique vegetation types6. Documenting plant community composition across mountainous regions is essential for mapping species distributions and detecting spatial shifts in response to environmental changes. High-quality datasets on plant community-level monitoring could support the identification of biodiversity hotspots and inform conservation strategies under ongoing climate change.

Elevational gradients have long been recognized as natural laboratories for studying patterns of species composition, diversity, and ecosystem processes across environmental gradients7. Along these gradients, abiotic conditions such as temperature, precipitation, solar radiation, and soil nutrients change systematically and often steeply over short spatial scales. These environmental shifts lead to pronounced variations in species composition, functional traits, and community structure, offering valuable opportunities to disentangle the ecological and evolutionary drivers of biodiversity. Importantly, elevational gradients can be used to infer species climatic niches and potential distribution limits, making them especially relevant for assessing biological responses to global climate change. For instance, warming may lead to range shifts or increased extinction risks for high-elevation specialists8. These dynamics highlight the value of elevational studies in predicting distributional shifts and offering essential insights for biodiversity conservation in the Anthropocene.

Plant community data provide the basic information in species composition, community structure, and ecosystem functions9. By combining with environmental variables, plant evolutionary history and functional traits, these data help us understand biodiversity patterns, species interactions, the effects of environmental changes on vegetation dynamics, and evolutionary and ecological mechanisms10,11. Detailed plant community data, including species richness and abundance, are essential for revealing species interactions and other forces that determine species coexistence within and across plant communities12,13.

China harbors over 35,000 vascular plant species, with half of them being endemic14. Mountainous regions in China with varied climates and geological histories, covering over 50% of the land, host most of the country’s plant diversity15 and endangered species16,17. We initiated a long-term research network BEST (Biodiversity along Elevational gradients: Shifts and Transitions; https://BEST-mountains.org) in 2017, to monitor biodiversity dynamics of multiple taxa (e.g., vascular plants, mosses, soil microbes, soil fauna, insects and birds) under the changes of climate and land uses. This paper reports a comprehensive plant community dataset collected from 17 mountains along elevational gradients in China. Across 370 permanent sampling plots along 17 elevational transects, we used standard filed protocols18 to measure the diameter at breast height (DBH ≥ 1 cm) of each individual, tag one unique number and identify to species. For each plot, we calculated species-level basal area (BA), relative abundance (RA), relative dominance (RD), and relative importance value (RIV). These metrics were also calculated separately for the tree layer and the shrub and sapling layer. Following the rough DBH-based classification of vegetation layers commonly used in textbooks and published studies, individuals with DBH ≥ 5 cm were classified as the tree layer, representing mature trees, while trees and shrubs with DBH < 5 cm were classified as the shrub and sapling layer. These data provide a baseline for understanding how plant community composition and species diversity vary along elevational gradients in these mountains. Additionally, this dataset is valuable for global comparisons among mountain ecosystems, especially in the context of global changes in climate and land uses.

These 370 plots across 17 mountains span elevations from 166 m to 3,835 m (Fig. 1), and extend across 46° in longitude and 24° in latitude, encompassing diverse climatic zones, including temperate continental, temperate monsoon, subtropical monsoon, and tropical monsoon climates. Correspondingly, the vegetation types represented range from temperate coniferous forests, temperate coniferous broad-leaved mixed forests, and warm-temperate deciduous broadleaved forests, to subtropical evergreen broad-leaved forests and tropical seasonal rainforests (Table 1). Totally, there are 1,493 woody plant species spanning 121 families and 449 genera (Fig. 2). The most species-rich families are Rosaceae, Lauraceae, and Fagaceae, contributing 123, 117, and 76 species (21% of total species), respectively. At the genus level, Rhododendron, Ilex, and Acer are the most diverse, comprising 49, 47, and 40 species (9% of total species), respectively. Among these, 995 species (67% of total species) occur only in a single mountain, and only 9 species are shared by half or more of the mountains. Plant diversity varies across mountain regions. The Bawangling on Hainan Island, south China, contains one of China’s largest tropical rain forests and supports the highest species richness, with 330 woody plants in the current dataset. Species richness in subtropical mountains ranges from 88 species in Daiyunshan to 261 species in Tianmushan. Qinling, situated in the subtropical-temperate transition zone and served as a natural boundary between south and north China, harbors 207 woody plants in our study sites. Temperate regions show relatively low richness, with West Tianshan in western China having only 23 species, and Changbaishan in northeastern China having 36 species. Overall, this dataset offers a baseline data for advancing biodiversity conservation of China’s forest flora.

Fig. 1
figure 1

Locations of the 17 studied mountains in China and representative plant species from each mountain: (a) Vatica mangachapoi from Bawangling; (b) Schima argentea from Damingshan; (c) Machilus chinensis from Chebaling (photo provided by Yongcai Zhao); (d) Fokienia hodginsii from Nanling; (e) Rhododendron sinogrande and (f) Euonymus frigidus from Gaoligong Mountains; (g) Pinus taiwanensis from Daiyunshan; (h) Rhododendron oreotrephes from Yulong Mountain; (i) Castanopsis eyrei from Guanshan (photo provided by Lin Chen); (j) Quercus engleriana from Jinfoshan; (k) Philadelphus sericanthus var. kulingensis from Lushan; (l) Daphniphyllum macropodum from Tianmushan; (m) Prunus obtusata from Dabieshan; (n) Abies chensiensis from Qinling (photo provided by Peiliang Liu); (o) Rhododendron concinnum from Baiyunshan; (p) Quercus variabilis from Lishan; (q) Pinus koraiensis from Changbaishan; (r) Malus sieversii from West Tianshan. The map of China is from: http://bzdt.ch.mnr.gov.cn.

Full size image
Table 1 The summary information of elevational sites in 17 mountains.
Full size table
Fig. 2
figure 2

The family-level phylogeny showing the evolutionary history of 1,473 plant species across 121 families. The inner circle shows the distribution of genus richness, and the outer circle shows species richness across families. Genus richness ranged from 1 to 27, and species richness ranged from 1 to 123. Both richness metrics were log10-transformed and then standardized to a range of 0 to 1 across the 121 families.

Full size image

Methods

Field survey

A total of 370 permanent sampling plots were established during growing seasons from 2013 to 2025 along elevational gradients across 17 mountains, spanning elevations from 166 to 3,835 m a.s.l. and covering a total area of 31.95 ha (Table 1). The study sites range in longitude from 82°E to 128°E and in latitude from 19°N to 43°N (Fig. 1). According to the CHELSA dataset19, the 17 mountains show wide climatic variation, with annual mean temperature from −6.2 °C to 24 °C and precipitation from 351 to 2,925 mm (Table 1). For each mountain and elevation, typical vegetation types were selected. Plots of appropriate size were established along the gradient, with intervals of approximately 100 m in elevation. At each interval, one or two plots were established, depending on the complexity of the vegetation composition. Basic information such as latitude, longitude, and elevation was recorded for each plot. All woody plants with a diameter at breast height (DBH) ≥ 1 cm were tagged in the field, and their species names and DBH were recorded. For specimens that could not be confidently identified in the field, voucher specimens were collected and subsequently verified by local taxonomic experts. Scientific species names were standardized using World Flora Online20 and the Flora of China21, with the R package22,23. We used V.PhyloMaker224 with the GBOTB.extended.WP.tre backbone, which incorporates the megatrees of Smith & Brown25 and Zanne et al.26, to build a family-level phylogeny. A total of 121 families were represented in the mega-tree (Fig. 2). All these plots are scheduled to resurvey every five years. Currently, nine mountains (Bawangling, Chebaling, Gaoligong Mountains, Daiyunshan, Yulong Mountain, Guanshan, Tianmushan, Dabieshan, West Tianshan) have finished the first recensus of plant communities. Meanwhile, microclimatic variables, soil properties, and canopy cover have also been recorded. The microclimatic data have been partially shared through the SoilTemp network27, but are still incomplete for some recently established sites. Therefore, we did not include these data in the current release and will incorporate them in future updates.

To assess species dominance within woody plant communities, we calculated the importance value as a comprehensive measure of relative species significance. We used a standard filed protocol18 to record species identity, abundance, and diameter at breast height (DBH ≥ 1 cm) of all woody plants. Vegetation layers were classified as tree layer (DBH ≥ 5 cm) and shrub and sapling layer (DBH < 5 cm). We calculated species-level basal area (BA), abundance, relative abundance (RA), relative dominance (RD), and relative importance value (RIV) for each plot, as well as for tree layer and shrub and sapling layer separately. To calculate the RIV, only the main stems were considered, excluding root suckers and branches. The RIV for each species was calculated using the formula: RIV = (RA + RD) / 2, where Relative Dominance (RD) = (species’ basal area / total basal area of all species) × 100%; Relative Abundance (RA) = (number of individuals of the species / total number of individuals of all species) × 100%.

Data Records

The BEST plant community dataset described above is available at Figshare (https://doi.org/10.6084/m9.figshare.30877097)28. It includes four main data files in CSV format: a stand table summarizing stand-level variables for all BEST sites across 17 mountains (Table 2), a species list of all recorded plant species (Table 3), a community table containing detailed plant community composition data (Table 4), and an occurrence table documenting plant species occurrences across mountains (Table 5). Additionally, a family-level phylogenetic tree is provided in Newick format (“family_phylogeny.tre” on Figshare), which can be used for phylogeny-related studies.

Table 2 Data structure of the header file of the stand-level data for all BEST sites in 17 mountains.
Full size table
Table 3 Data structure of the header file for the species list of all recorded plant species across all BEST sites in 17 mountains.
Full size table
Table 4 Data structure of the header file of the BEST plant community data.
Full size table
Table 5 Data structure of the header file for plant species occurrence across mountains.
Full size table

Technical Validation

All plant surveys followed a standard field protocol to ensure consistency and comparability across these 17 mountains. Geographic coordinates of all 370 plots were precisely recorded using GPS devices. Overall, these rigorous technical validation steps ensure the dataset is robust, reliable, and valuable for a wide range of ecological and conservation research.

Usage Notes

These data enable the exploration of elevational patterns in plant species richness, community composition, and phylogenetic diversity across montane forest ecosystems in China. To ensure consistency across mountains, data cleaning was performed by excluding records related to branches and retaining only those associated with the main stem. Users should consider differences in sampling years among mountains when conducting temporal analyses. Additionally, some records could not be identified to the species level and remain unknown. Such records should be handled cautiously depending on the specific research question.