This study was conducted in Gangwon-do, a mountainous region that includes extensive temperate deciduous broad-leaved forests and areas in which Q. mongolica forests are distributed.

In this study, the term “pure Q. mongolica stands” was used as an operational definition for delimiting the dataset. The term referred to plots from the 5^th National Ecosystem Survey in which Q. mongolica was recorded as the dominant species in the canopy layer in the original field survey records. It does not imply monospecific stands or the absence of other woody species in the canopy, subcanopy or shrub layers. This definition was used to distinguish the target stands from mixed forests or general forest stands in which Q. mongolica occurs only as one component of the tree layer. A total of 1,098 plots satisfying this criterion were included in the final analysis and the spatial distribution of the plots is shown in Fig. 1.

Floristic data were obtained from field investigations conducted from 2019 to 2023 as part of the 5^th National Ecosystem Survey (National Institute of Ecology 2019). From the 5^th National Ecosystem Survey, plots meeting the operational definition described above were selected and all vascular plant taxa recorded in each plot were compiled as plot-level occurrence data. For analysis, the original plot-level occurrence data were summaried to calculate taxon occurrence frequency and plot-level taxon richness, with a detailed occurrence record dataset also provided through GBIF.

The final floristic dataset consisted of 1,098 plots and 20,906 plot-level occurrence records.

The number of taxa recorded in each plot was calculated as plot-level taxon richness. Taxonomic composition was summarised at the levels of family, genus, species, variety, subspecies and form. The families with the highest numbers of taxa and the most frequently recorded taxa were identified on the basis of the number of taxa per family and the number of plots in which each taxon occurred, respectively.

Environmental variables were summarised to describe the survey scope and environmental setting of the surveyed pure Q. mongolica stands, rather than to serve as a main analytical axis of this floristic inventory study.

Environmental data were compiled from field survey records and spatial datasets. Field-based variables used in the internal compilation included plot size, topography, soil type, moisture condition, lithology and field-recorded soil texture; selected numerical and categorical variables are summarised in the Results, Table 7 and Supplementary material 3. Soil-related spatial variables, including deep soil texture, soil drainage class and effective soil depth, were obtained from spatial datasets provided by the National Academy of Agricultural Science through World Spatial Information Download (National Academy of Agricultural Science 2023a; National Academy of Agricultural Science 2023b; National Academy of Agricultural Science 2024). Elevation and slope were extracted from NASADEM 30 m data (NASA JPL 2021). Vegetation-climatic zone information followed the classification of Cho et al. (2020).

To ensure spatial consistency amongst plots, NASADEM-derived elevation and slope were used as representative variables instead of the values recorded on field survey sheets. These variables are referred to as NASADEM_GlobalDEM and Slope_NASADEM30m, respectively.

All recorded vascular plant taxa were standardised at the levels of family, genus, species, variety, subspecies and form. Each taxon was treated as an operational floristic unit for summarising taxon richness and plot-level occurrence patterns. Scientific and Korean names were based primarily on the National List of Korea dataset from the National Institute of Biological Resources and public data obtained through the Public Data Portal (data.go.kr). Some taxa absent from domestic databases or requiring taxonomic review, such as Veratrum versicolor f. viride, were supplemented through the Korea Biodiversity Information System of the Korea National Arboretum. Scientific names that could not be definitively confirmed in domestic databases were checked and verified using the international databases IPNI (International Plant Names Index) and POWO (Plants of the World Online) (National Institute of Biological Resources 2025a; Korea National Arboretum 2026; IPNI 2026; POWO 2026).

The recorded taxa were classified into conservation- and management-related categories to describe taxa of conservation or management interest recorded in pure Q. mongolica stands. These categories included Korean endemic plant taxa, endangered wild plant taxa, Red List plant taxa, floristic target plant taxa, alien plant taxa and ecosystem-disturbing plant taxa.

Korean endemic plant taxa were identified using the national endemic species reference provided by the National Institute of Biological Resources (National Institute of Biological Resources 2025b).

Endangered wild plant taxa were classified according to the national list under the Wildlife Protection and Management Act of South Korea (Ministry of Climate, Energy and Environment 2026). Korean Red List plant taxa were classified using the national Red List reference for vascular plants (National Institute of Biological Resources 2021). Floristic target plant taxa were classified into Grades I–V following national references for floristic target plant taxa (National Institute of Ecology 2018). Alien plant taxa and ecosystem-disturbing plant taxa were identified using the national alien species list and the official designation notice for ecosystem-disturbing species, respectively (National Institute of Ecology 2025; Ministry of Environment 2025). IUCN Red List plant taxa were classified using a list of 481 vascular plant species extracted from the full IUCN Red List by applying the taxonomic filter for vascular plants and the IUCN land region filter for South Korea (IUCN 2026).

Conservation- and management-related categories were not combined into a single integrated rank, but were summarised separately according to the purpose of each legal or reference list. For example, alien plant taxa and ecosystem-disturbing plant taxa were treated as independent categories because they are based on different sources and designation purposes. This classification clarifies the conservation or management context in which each taxon is relevant.

All analyses were conducted at the taxon and plot levels. First, the total number of vascular plant taxa was calculated and summarised by taxonomic rank. Second, the number of taxa in each family was calculated to identify major taxon-rich families in the flora. Third, the number of taxa recorded in each plot was counted to obtain plot-level taxon richness. Fourth, the occurrence frequency of each taxon was calculated on the basis of the number of plots in which the taxon was recorded.

For conservation-related taxa, the number of taxa in each category was calculated. Korean endemic plant taxa, endangered wild plant taxa, Red List plant taxa, floristic target plant taxa, alien plant taxa and ecosystem-disturbing plant taxa were summarised separately. Floristic target plant taxa were further summarised by grade and Red List plant taxa were summarised by threat category.

Environmental variables were summarised using descriptive statistics to characterise the ecological setting of the plots. For numerical variables such as elevation, slope, plot width and length, minimum, maximum, mean and median values were calculated. For categorical variables, such as topography, moisture condition, soil type and vegetation-climatic zone, the number of plots in each category was counted. These environmental summaries were used to describe the survey scope and plot metadata, rather than to infer causal relationships between environmental variables and species composition.

A total of 594 vascular plant taxa were recorded in 1,098 plots in pure Q. mongolica stands in Gangwon-do. These taxa comprised 103 families, 288 genera, 539 species, 45 varieties, seven subspecies and three forms. The floristic dataset included 20,906 plot-level occurrence records. The taxonomic composition is summarised in Table 1.

Species-level taxa accounted for the majority of the recorded flora, with 539 species. Infraspecific taxa consisted of 45 varieties, seven subspecies and three forms.

Asteraceae had the highest number of taxa (52), followed by Rosaceae (44), Liliaceae (37), Ranunculaceae (29), Poaceae (21) and Fabaceae (20). Other major families included Lamiaceae, Apiaceae, Betulaceae, Violaceae and Dryopteridaceae. The numbers of taxa in major families are presented in Table 2.

The number of vascular plant taxa per plot ranged from 5 to 55. Mean plot-level taxon richness was 19.04 taxa per plot and the median was 18 taxa. The first and third quartiles were 13 and 24 taxa, respectively, indicating that the middle 50% of plots contained approximately 13–24 taxa. Summary statistics are presented in Table 3 and the distribution of plot-level taxon richness is shown in Fig. 2.

Variation in plot-level taxon richness indicates that local floristic composition differed amongst plots, even though all plots were defined as pure stands dominated by Q. mongolica in the canopy layer. Such variation may reflect differences in microtopography, stand structure, disturbance history, understorey development and local environmental conditions. However, because this study was designed as a floristic inventory, the environmental drivers of plot-level taxon richness were not analysed as a main objective.

Quercus mongolica was recorded in all 1,098 plots, confirming that all plots belonged to the target forest type. The most frequently recorded associated taxa included Lindera obtusiloba, Acer pseudosieboldianum, Fraxinus rhynchophylla, Carex siderosticta, Rhododendron schlippenbachii, Lespedeza maximowiczii, Symplocos sawafutagi, Ainsliaea acerifolia and Rhododendron mucronulatum. The most frequent taxa are summarised in Table 4 and Suppl. material 1.

Woody taxa such as Lindera obtusiloba, Acer pseudosieboldianum, Fraxinus rhynchophylla, Rhododendron schlippenbachii and Rhododendron mucronulatum were repeatedly recorded across many plots and herbaceous taxa, such as Carex siderosticta and Ainsliaea acerifolia, also showed high occurrence frequencies. These results indicate that the vascular flora of pure Q. mongolica stands includes not only the canopy-dominant species, but also understorey woody and forest-floor herbaceous taxa.

A total of 32 Korean endemic taxa were recorded in plots of pure Q. mongolica stands. One taxon was classified as an endangered wild plant of Class II. A total of 21 Korean Red List plant taxa were recorded, consisting of 11 Least Concern (LC), four Endangered (EN), three Near Threatened (NT), two Vulnerable (VU) and one Data Deficient (DD) taxa. A total of 85 IUCN Red List plant taxa were recorded, consisting of two Data Deficient (DD), four Endangered (EN), 76 Least Concern (LC), two Near Threatened (NT) and one Vulnerable (VU) taxa. The conservation-related taxa and composition of Red List categories are presented in Table 5 and Suppl. material 2.

Floristic target plant taxa were particularly numerous. A total of 172 floristic target plant taxa were recorded, consisting of 42 Grade I, 44 Grade II, 57 Grade III, 24 Grade IV and five Grade V taxa. Grade III taxa were the most numerous, followed by Grade I and Grade II taxa. The grade-level composition is summarised in Table 5.

These conservation-related taxa provide occurrence evidence for conservation and phytogeographical elements within pure Q. mongolica stands in Gangwon-do. In particular, the occurrence of Grade IV and Grade V floristic target plant taxa shows that this forest type may provide occurrence sites for some taxa with restricted distributions or phytogeographical significance.

A total of seven alien plant taxa and two ecosystem-disturbing plant taxa were recorded in the plots. As this study was not designed to quantitatively assess invasion intensity, these results should be interpreted as evidence of the presence of management-relevant taxa in pure Q. mongolica stands and the need for future monitoring. Category-level counts are presented in Table 6.

The 1,098 plots covered a broad environmental range within Gangwon-do. According to NASADEM-derived terrain variables, elevation ranged from 24 m to 1,450 m, with a mean of 732.6 m and a median of 747.5 m. NASADEM-derived slope ranged from 0.3 to 45.5°, with a mean of 19.3° and a median of 19.4°. Plot width ranged from 10 m to 25 m, with a mean of 13.1 m and a median of 11 m, while plot length ranged from 5 m to 25 m, with a mean of 13.1 m and a median of 11 m. Numerical summaries of the main environmental and plot-size characteristics are presented in Table 7. Categorical summaries of environmental characteristics are provided in Suppl. material 3.

Most plots were located on upper and middle slopes. Upper slopes accounted for the largest number of plots (513 plots), followed by middle slopes (303), lower slopes (127), ridges (101), valleys (47) and summits (7). Moisture conditions were predominantly mesic, with 627 plots classified as mesic, 262 as slightly moist and 190 as slightly dry. Excessively wet and excessively dry conditions were uncommon, occurring in 18 plots and one plot, respectively.

Vegetation-climatic zone information indicated that most plots were located in the northern temperate deciduous broad-leaved forest zone. This zone included 969 plots, followed by the central temperate deciduous broad-leaved forest zone with 95 plots. The northern temperate mixed coniferous and broad-leaved forest zone, southern temperate mixed evergreen-deciduous broad-leaved forest zone and southern temperate deciduous broad-leaved forest zone included 19, 14 and one plots, respectively.

Brown forest soil was the dominant soil type, occurring in 1,024 plots, followed by red-yellow forest soil in 36 plots, dark red forest soil in 34 plots and grey-brown forest soil in four plots. Lithology was mainly igneous rock, which occurred in 562 plots, followed by metamorphic rock in 469 plots and sedimentary rock in 67 plots. Surface soil texture was dominated by loam, with 981 plots, whereas sandy and clayey/sandy soils occurred in 59 and 58 plots, respectively. Deep soil texture was mainly sandy loam, occurring in 947 plots, followed by clay loam in 135 plots, other soil textures in nine plots, clayey soil in four plots, sandy soil in two plots and gravelly soil in one plot.

Drainage was generally good. Most plots were classified as very well drained (859 plots) or well drained (228 plots), whereas the other and moderately well-drained classes occurred in nine and two plots, respectively. The most frequent effective soil depth classes were 50–100 cm and 20–50 cm, recorded in 437 and 435 plots, respectively. Shallow soils less than 20 cm deep occurred in 211 plots, whereas soils deeper than 100 cm were recorded in only six plots. Nine plots were classified as other.

These results indicate that the surveyed pure Q. mongolica stands were distributed mainly in the northern temperate deciduous broad-leaved forest zone of Gangwon-do, particularly on upper and middle mountain slopes under mesic to slightly moist conditions. The plots were also mainly associated with brown forest soils, igneous or metamorphic bedrock, loamy surface soils, sandy loam subsoils, well-drained to very well-drained conditions and effective soil depths of 20–100 cm. Therefore, the surveyed plots were mainly located on mountain slopes with relatively good drainage and intermediate effective soil depth and these site conditions represent the main environmental characteristics identified in this study.

This study provides a large-scale plot-based floristic inventory of pure Q. mongolica stands in Gangwon-do. A total of 594 vascular plant taxa were recorded in 1,098 plots, yielding 20,906 plot-level occurrence records. These results show that pure Q. mongolica stands, as operationally defined by the canopy-layer dominance of Q. mongolica in the original field survey records, contain a diverse vascular flora composed of woody plants, forest herbs, sedges, ferns and infraspecific taxa.

Previous studies of Q. mongolica forests in Korea have focused on vegetation structure, community classification, succession, diameter-class distribution or restoration models (Jang et al. 1997; Lee and Song 2011; Lee et al. 2014; Yu et al. 2024). By focusing on the floristic composition of pure Q. mongolica stands and integrating plot-level occurrence information across many plots, this study provides an occurrence-based floristic reference that complements these vegetation-structure studies.

In the international literature, studies from north-eastern China and across the broader East Asian range have examined phenological responses, climatic effects and changes in potentially suitable habitats and ecological niches of Q. mongolica under climate change (Fan et al. 2015; Liu et al. 2024; Kaviriri et al. 2025), whereas Japanese studies have addressed phytogeographical and phytosociological relationships with Northeast Asian summer-green broad-leaved forests, mainly in relation to Q. mongolica var. crispula or Q. crispula (Okaura et al. 2007; Shitara et al. 2024). Although the present dataset does not share the same analytical framework as those international studies, it can provide comparative floristic reference data on vascular plant occurrence in pure Q. mongolica stands in Gangwon-do for broader studies of Northeast Asian oak forests.

The high occurrence frequencies of Lindera obtusiloba, Acer pseudosieboldianum, Fraxinus rhynchophylla, Carex siderosticta, Rhododendron schlippenbachii, Ainsliaea acerifolia, Lespedeza maximowiczii, Symplocos sawafutagi and Rhododendron mucronulatum indicate that these taxa were repeatedly recorded in pure Q. mongolica stands in Gangwon-do. This pattern is broadly consistent with previous studies reporting Acer pseudosieboldianum, Rhododendron schlippenbachii, Tilia amurensis, Fraxinus rhynchophylla and other deciduous broad-leaved taxa as important associated taxa in Q. mongolica communities (Suh and Lee 1998; Lee and Song 2011; Lee et al. 2014; Yu et al. 2024). However, the occurrence frequencies in this study are based on plot-level presence data rather than cover or importance values and, therefore, should not be interpreted as measures of dominance or relative ecological importance within communities.

The results of this study are broadly consistent with previous studies that describe Q. mongolica communities as important deciduous broad-leaved forest types in Korea. A recent study of Q. mongolica communities in the central region of Korea reported that Q. mongolica had the highest importance value in both the tree and subtree layers and suggested that these communities may maintain Q. mongolica dominance over the long term (Yu et al. 2024). Similarly, Lee et al. (2014) indicated, on the basis of species composition and diameter-class distributions, that Q. mongolica forests may persist as stable forest communities. Song and Yun (2019) showed that Q. mongolica community groups are related to environmental gradients in the Baekdudaegan Region.

The present study differs in its main purpose from such community-structure studies. Instead of classifying vegetation units or predicting succession using TWINSPAN, CCA or importance-value analyses, this study provides a floristic inventory and an occurrence-based reference dataset for pure Q. mongolica stands. Whereas community-structure studies show how dominant and associated species are organised by vegetation layer, the floristic inventory presented here describes the range of vascular plant taxa actually recorded in a particular forest type and their plot-level occurrence patterns. Therefore, this dataset should be interpreted not as a substitute for existing vegetation-classification or vegetation-structure studies, but as a complementary reference for vascular plant occurrence in pure Q. mongolica stands.

The focus on pure stands is also important. Previous studies have often analysed Q. mongolica communities together with mixed forests or compared Q. mongolica forests with other forest types (Lee et al. 2014; Lim and Kim 2015). By restricting the dataset to stands in which Q. mongolica dominates the canopy layer, this study provides a clearer floristic reference for a specific forest type. This approach reduces interpretive ambiguity caused by mixed forests and helps identify the vascular plant assemblage recorded in pure Q. mongolica stands.

The conservation-related taxa recorded in this study provide occurrence evidence for conservation and phytogeographical elements within pure Q. mongolica stands. Floristic target plant taxa are used as indicators for identifying regionally or nationally important floristic elements and the grade-level occurrence information obtained in this study can support future interpretations of the phytogeographical characteristics of this forest type (National Institute of Ecology 2018).

The occurrence of Red List plant taxa and an endangered wild plant taxon indicates that pure Q. mongolica stands include occurrence sites for some taxa of conservation concern. However, because this study did not quantitatively assess population size or growth status, these results should be interpreted as occurrence information based on national and international reference lists, rather than as a new assessment of the conservation status of each taxon. Such occurrence information may help identify taxa that should be prioritised for confirmation in future repeated surveys and habitat management.

The records of seven alien plant taxa and two ecosystem-disturbing plant taxa show that some management-relevant taxa occur even within the surveyed pure Q. mongolica stands. Although the present data do not allow direct comparison of alien plant invasion intensity with other habitats, the occurrence of ecosystem-disturbing plant taxa suggests that continued monitoring may be needed. Previous studies have also emphasised the need for continued monitoring and management of alien plants because their distribution and spread can vary with environmental conditions and human disturbance (Jung et al. 2017; Kang et al. 2020). Monitoring of alien plant taxa and ecosystem-disturbing plant taxa can be treated as a component of the long-term management of pure Q. mongolica stands. The plot-level occurrence data provided in this study can be used to identify plots in which alien and ecosystem-disturbing plant taxa were recorded and may, therefore, provide a basis for targeted follow-up surveys and long-term monitoring.

This study contributes to the development of a reusable floristic dataset in three ways. First, it provides a taxonomic summary of vascular plants recorded in pure Q. mongolica stands. Second, it presents plot-level occurrence records that can be reused for future biodiversity assessments and monitoring. Third, it summarises conservation- and management-related taxa, including endemic plant taxa, Red List plant taxa, floristic target plant taxa, alien plant taxa and ecosystem-disturbing plant taxa, as separate categories. Internationally, plot-based vegetation data have been used as core data sources for large vegetation databases and broad-scale biodiversity analyses and the plot-level occurrence data from this study can likewise be reused for future data integration and comparative research (Dengler et al. 2011; Chytrý et al. 2016; Bruelheide et al. 2019).

In this study, environmental variables were used mainly to describe the survey scope and environmental setting of the plots, rather than to infer causal relationships between plant occurrence and environmental gradients. This was an intentional choice because the primary objective was to provide a floristic inventory based on plot-level occurrence records. If plot-level environmental metadata are made available together with the occurrence records, these data could provide input for future studies of species distributions, habitat preferences, plot-level taxon richness and floristic change.

The target habitat in this study was defined as pure Q. mongolica stands, as operationally delimited by the canopy-layer dominance of Q. mongolica in the original field survey records. This definition provides forest-type-specific occurrence information that cannot be readily obtained from general regional floras. For example, the dataset can be used to identify which vascular plant taxa are repeatedly recorded in pure Q. mongolica stands and which conservation- and management-related taxa occur within this forest type.

The plot-level structure of the dataset also increases its value. A simple species list can show which taxa occur in a region, but plot-level occurrence records indicate how widely each taxon is distributed within the target habitat. The high occurrence frequencies of species, such as Lindera obtusiloba, Acer pseudosieboldianum, Fraxinus rhynchophylla, Carex siderosticta and Rhododendron schlippenbachii, suggest that these taxa were frequently recorded in pure Q. mongolica stands in Gangwon-do. Conversely, rare or low-frequency taxa may be useful as candidates for further confirmation in future surveys, especially when they are also included in conservation- or management-related categories.

This study, therefore, contributes not only to records of forest flora in Gangwon-do, but also to the accumulation of biodiversity data for Q. mongolica-dominated temperate deciduous broad-leaved forests in Korea. By presenting the total flora together with conservation- and management-related categories, this study provides occurrence-based information that can be used in taxonomy, biodiversity monitoring, forest conservation and ecological management.

This study has several limitations. First, the analyses were based mainly on presence-absence data. Therefore, abundance, cover, dominance and vertical-layer structure of each taxon were not evaluated. For this reason, the results should be interpreted as a floristic and occurrence-based inventory rather than a complete vegetation-structure analysis.

Second, environmental variables were summarised descriptively and were not used to model species composition or plot-level taxon richness. Future studies could examine relationships between vascular plant occurrence and environmental gradients by combining plot-level occurrence matrices with environmental metadata. Such analyses would be useful for understanding how elevation, slope, moisture conditions, soil characteristics and climatic conditions are related to vascular plant richness, composition and the occurrence of conservation-related taxa (Lee et al. 2013; Woo et al. 2025; Park and Cheon 2025).

Third, this study focused on pure Q. mongolica stands in Gangwon-do. Therefore, the results should not be directly generalised to all Q. mongolica forests in Korea, especially mixed oak forests, urban Q. mongolica stands or forests in southern regions. To evaluate how floristic composition varies across the broader distribution range of Q. mongolica forests, future studies comparing different provinces, vegetation-climatic zones and pure versus mixed stands would be useful.

Finally, forest communities can be affected by climate change, pests, disturbances and management practices; therefore, repeated surveys are needed to detect temporal changes in species composition and conservation-related taxa. Studies of Q. mongolica in north-eastern China and across its broader East Asian range have analysed changes in green-up dates and potentially suitable habitats in relation to climate change (Fan et al. 2015; Liu et al. 2024; Kaviriri et al. 2025), suggesting the need to accumulate Q. mongolica forest data as reference datasets for long-term monitoring and future assessments of climate-related floristic change. However, because this study did not directly analyse climatic responses, relationships between climate change and floristic change should be examined through repeated surveys and separate analyses in the future.