Biodiversity oases in the southern tundra: the role of shrub-herbaceous communities (Western Siberia)SheferNikita V.vchifz@mail.ruKritskovIvan V.vchifz@mail.ruIstigechevGeorgy I.vchifz@mail.ruLoikoSergey V.vchifz@mail.ruRussian FederationRussian FederationRussian FederationRussian Federation24122025
Biodiversity oases in the southern tundra: the role of shrub-herbaceous communities (Western Siberia)
Nikita V. Shefer1, Ivan V. Kritskov1, Georgyi I. Istigechev1, Sergey V. Loiko1
1National Research Tomsk State University, 36 Lenin Ave., Tomsk, 634050, Russia
Corresponding author: Nikita V. Shefer (vchifz@mail.ru)
Academic editor: A. Matsyura | Received 15 November 2025 |Accepted 17 December 2025 | Published 24 December 2025 |
Citation: Shefer NV, Kritskov IV, Istigechev GI, Loiko SV (2025) Biodiversity Oases in the Southern Tundra: The Role of shrub-herbaceous communities (Western Siberia). Acta Biologica Sibirica 11: 1563–1587. https://doi.org/10.5281/zenodo.18051200
Abstract
The southwestern Gydan Peninsula remains a significant data gap in understanding tundra ecosystem responses to climate change and shrub expansion. Detailed studies linking plant community composition to soil properties in this region are critically lacking. This study provides the integrated analysis of shrub-herbaceous communities and their underlying soils in this region. Based on comprehensive geobotanical and soil surveys at 52 sites, we identified four distinct community types using multivariate analysis. Our results reveal a clear hierarchy of environmental drivers. Primary vegetation patterns and the formation of biodiversity hotspots, including the discovery of several plant species new to the local flora, are predominantly shaped by landscape heterogeneity, such as the mosaic of slopes, valleys, and snow hollows. Soil properties, notably those related to texture, active layer dynamics, and hydromorphy, act as significant secondary filters, structuring communities along a key soil gradient. However, the overall correlation between soil and vegetation matrices was limited, indicating a substantial role for other unmeasured factors. We conclude that plant community assembly in the southern Gydan tundra is governed by an interplay where the geomorphic template sets the stage for diversity, while localized soil conditions fine-tune composition. These findings establish an essential baseline for monitoring and underscore that conserving landscape heterogeneity is crucial for the resilience of these rapidly changing Arctic ecosystems.
Ongoing climatic changes in the Subarctic and Arctic (Anisimov and Kokorev 2017; Vasiliev and Drozdov 2020; Kaverin et al. 2021; Fewster et al. 2022) are increasing the length of the growing season, precipitation, and nutrient availability, thereby enhancing vegetation productivity (Wu et al. 2011; Huang et al. 2017; Reichle et al. 2018). This "greening" effect, confirmed by satellite monitoring of the Normalized Difference Vegetation Index (NDVI) (Tishkov and Krenke 2015; Frost et al. 2019; Berner et al. 2020; Myers-Smith et al. 2020), stems not only from increased productivity but also from successional processes. These include the colonization of drained lake basins (Lara et al. 2018; Loiko et al. 2020), landslides, thermokarst depressions (Walker et al. 2009), and most notably, tundra shrubification (Tape et al. 2006; Frost et al. 2013; Tishkov and Krenke 2015). As the most extensive of these successional processes, shrubification is crucial for modeling future landscapes, as it alters tundra ecosystem structure, regional climate, soil-atmosphere exchange, and species interactions (Myers-Smith et al. 2011).
Although shrubification driven by alder, willow, and dwarf birch expansion (Tape et al. 2012) has been extensively studied globally (Vowles and Björk 2019; Chen et al. 2021; Mekonnen et al. 2021), significant gaps remain at the regional scale. The soils and vegetation of the Yamalo-Gydan Ecoregion are poorly documented, even from a primary inventory standpoint (Loiko et al. 2024). While the northern and central Gydan Peninsula are relatively well-studied botanically (Khitun and Rebristaya 1998; Khitun 2002; Telyatnikov et al. 2019; Telyatnikov et al. 2021), data for the south are scarce (Khitun 2003; Loiko et al. 2024). Studies of soils for the peninsula are exceptionally limited (Loiko et al. 2024), contrasting with the better-studied southern boundary of the Yamalo-Gydan Ecoregion (Titov and Potokin 2001; Pismarkina and Byalt 2016; Glazunov and Nikolaenko 2018; Zarov et al. 2022; Loiko et al. 2023).
This regional knowledge gap is critical because these ecosystems are facing dual pressures from climate-driven vegetation change (Frost et al. 2023), alterations in hydrological regimes (Mann et al. 2010), and increasing industrial development that can rapidly alter landscapes (Loiko et al. 2024). Detailed ground-based studies with broad spatial coverage are needed to provide baseline data for vulnerable Arctic ecosystems (Syso et al. 2022) and to interpret trends in tundra productivity from remote sensing like NDVI (Tassone et al. 2024).
In this context, obtaining baseline data on the soils and vegetation of the poorly studied southwestern Gydan Peninsula is a priority. Particular interest lies in ecosystems that act as local biodiversity hotspots and environmental indicators. These include species-rich slopes, ravine and hollow complexes, and streamside communities, where favorable microclimate, drainage, and moisture support highly productive and structurally complex shrub-dominated ecosystems. However, the structure and composition of these shrub-herbaceous communities, as well as their intricate relationships with soil properties, remain poorly studied for this region.
Existing literature provides strong evidence for the profound effects of shrub encroachment on tundra ecosystems (Bonfils et al. 2012; Loranty and Goetz 2012; Frost et al. 2018; Kropp et al. 2020; Liu et al. 2022; Villani et al. 2022). Yet, these studies often treat "shrubs" as a uniform functional group. A critical, understudied aspect is the internal biodiversity of shrub-dominated communities. The distinct composition of older versus newly established thickets, and the specific correlations between this floristic diversity and a suite of soil characteristics, have not been integrally addressed, representing a key knowledge gap in predicting ecosystem trajectories.
Our study continues a series of hierarchical morphogenetic analyses of Western Siberian soils (Loiko et al. 2011; Loiko et al. 2013; Loiko et al. 2015; Konstantinov et al. 2019) and aims to fill the critical gap in understanding the coevolution of plant communities and soils under shrub expansion in the southern tundra of the Gydan Peninsula. The primary objectives of this work are: 1) to identify and classify the main types of shrub-herbaceous communities in the studied streamsides and slopes, 2) to assess the changes in soil morphological properties and taxonomy resulting from shrub expansion, and 3) to establish the relationships between the identified vegetation clusters and soil characteristics. As shrubs expand and form consortium cores, understanding this plant-soil interplay is essential for predicting the transformation of regional phytodiversity and the trajectory of future landscape changes.
Materials and methodsStudy Area
The studies were conducted in the southern part of the Gydan Peninsula (Fig. 1). The research area is situated within the Messoyakha subprovince, forming part of the southern tundra subzone in the Gydan province of the Yamal-Gydan tundra region (Atlas... 2004). From a pedological perspective, the territory falls within the subzone of tundra Cryosols and Histic Cryosols characteristic of the subarctic tundra in the West Siberian tundra province, as classified by the soil-ecological zoning of Russia (Map... 2020). The Tazovsky-Gydansky district, where the study was conducted, exhibits diverse soil textures ranging from sandy to clayey, with binomial soil profiles and marine sediments being common features of the local pedogenesis (Loiko et al. 2024).
The relief of the Gydan Peninsula, similar to the entire northern part of the West Siberian Plain, is characterized by a terraced surface. The Yamal-Gydan region features five geomorphological levels (marine terraces) of different ages (Atlas … 1971). The surveyed areas are predominantly composed of the lowest and youngest surfaces of the first and second marine terraces, with absolute elevations ranging from 10 to 25 m above sea level, as well as even lower floodplain areas (Telyatnikov et al. 2021). The highest elevations of the interfluves areas reach up to 40 m above sea level. The upper part of the river catchments is a flat, peaty plains with elevations exceeding 40 m, into which the river valleys and its tributaries are deeply incised (Loiko et al. 2024). The study area constitutes a gently rolling plain, dissected by a network of ravines and numerous lacustrine basins. The lithology of the investigated areas is homogeneous, with sandy deposits prevailing everywhere, while loamy deposits are exceedingly rare (Telyatnikov et al. 2021).
<bold id="bold-7edd420d37a9aa1b79a5032604ebd20f"/>Maps and locations of the study sites.
Figure 1. Maps and locations of the study sites.
In the southern part of the Gydan Peninsula, the climate is continental, with long and severe winters and relatively brief and cool summers. During coldest month (January), the average air temperature ranges from -27.5 °C to -28.5 °C, while at warmest (July) – it is around 12–14.5 °C. However, on some summer days, temperatures can reach up to 27–30.5 °C. The average annual temperature is between -9.5 and -10.5 °C (Trofimova and Balybina, 2014). Precipitation in this area is approximately 350 mm per year. A stable snow cover usually forms by the end of October and melts by the beginning of June. The snow reaches its maximum depth at the end of April, and there are frequent snowstorms that can cause uneven snow depths, ranging from almost none to 1–2 m, and even more in areas with alder bushes and at the bottom of tundra ravines (Geocryology... 1989).
According to the zonal classification of vegetation cover by I.S. Ilyina (1985), the south of Gydan Peninsula is associated to the subarctic tundra subzone (subzone of southern tundra). This subzone is characterized by the presence of hypoarctic shrubs, including dwarf birch (Betula nana L.) and several species of willow (Salix glauca L., S. lanata L., and S. pulchra Cham.). In the southern part of the subarctic tundra, diverse shrub-moss-lichen and shrub-moss tundra with dwarf birches and willows develop on watershed plains. These tundra communities can be found on watershed plains. When waterlogged, these tundra communities on watershed plains transition into shrub-moss, shrub-lichen-moss, or shrub-cottongrass-moss tundra. These tundras have a hummocky or mound-hollow microtopography, and sphagnum moss dominates the vegetation (Ilyina et al. 1985; Pisarenko et al. 2017). In the southern tundra, there are transitional sedge communities with Carex aquatilis Wahlenb. and sedge-sphagnum communities that develop in waterlogged lake basins (khasyreys), flow-through depressions, extensive mires of hummock-swamp complexes, and various depressions within waterlogged tundras (Loiko et al. 2020). The moss layer is typically dominated by Sphagnum obtusum Warnst. and S. riparium Ångstr. (Pisarenko et al. 2017). On elevated terrain, communities with multiple layers of shrubs are widespread. These include alder (Alnus fruticosa Rupr.), various types of willows (Salix glauca and S. pulchra), and dwarf birches (Betula nana). Under these shrubs, cloudberries, various herbs and sedges are found, and occasionally larch trees (Larix sibirica Ledeb.). Along the banks of larger rivers, alder and willow thickets (Alnus fruticosa, Salix gmelini Pall.) can be found. Eroded slopes are occupied by communities at different stages of succession as they revegetate after sand blowouts or landslides (Telyatnikov et al. 2021; Loiko et al. 2024).
Fieldwork and laboratory analysis
Fieldwork was conducted during the summer of 2024. We surveyed 52 herbaceous-shrub sites characterized by high floristic diversity relative to the surrounding areas. Site selection was based on shrub cover exceeding 40%, with each site measuring 10 × 10 meters. Plant community descriptions were recorded using the presence-absence method. The surveyed communities were classified into 3 distinct habitat types: alder thickets (Suppl. material 1: Table S1), streamside shrub-herbaceous communities (Suppl. material 2: Table S2) and slope/hollow shrub-herbaceous communities (Suppl. material 3: Table S3). In addition to botanical surveys, we described soil profiles at each site (Suppl. material 4: Table S4). We studied soil profiles by digging soil pits, cutting and photographing their front walls, identifying soil horizons and describing them morphologically, followed by sampling.
To identify the plants, they were herbarised. Collected herbarium specimens were transferred to the P.N. Krylov Herbarium (TK) of Tomsk State University for storage. Herbarisation was carried out according to the rules of herbarium (Gureeva 2013). The collected herbarium was identified according to the "Flora of Siberia" (1987–2003) and collection of TK.
Under laboratory conditions, using the collected photographic materials and soil samples, we conducted measurements of the following quantitative parameters (Suppl. material 4: Table S4). Active Layer Thickness (the depth to the permafrost table from the soil surface). Permafrost was not recorded if it was beyond the reach of a 150-cm probe. Thickness of the peat horizon covering the mineral soil surface. Thickness of the AO (moder humus) horizon, where present. Depth from the soil surface to the onset of morphological features diagnostic of gleyization. Thickness of the gleyed layer, measured from the mineral soil surface to a depth of 40 cm. The areal percentage of gley patches within the upper 40 cm of the mineral soil horizons. The percentage of gley patces relative to the total area of the gley horizon, where present. Physical clay content.
Multivariate analysis
All statistical analyses were performed in R version 4.5.2, using the following packages: “vegan” version 2.7-2 (Oksanen et al. 2025) for multivariate analysis and permutation tests, “cluster” version 2.1.8.1 (Maechler et al. 2025) for clustering algorithms, “ggplot2” version 4.0.1 (Wickham 2016) for data visualization and patchwork version 1.3.2.9000 (Pedersen 2025) for plot arrangement at clusters. Data handling was facilitated by «dplyr» version 1.1.4 (Wickham et al. 2025) and “tidyr” version 1.3.1 (Wickham et al. 2025b).
The original dataset comprised 52 geobotanical descriptions with 117 species recorded in presence-absence format. To reduce the influence of rare species on the analysis, we filtered the dataset to include only species that occurred in more than one sample, resulting in a final dataset of 52 geobotanical descriptions and 82 species for subsequent analysis.
Analysis of vegetation structure. During preliminary data analysis, we evaluated multiple approaches to optimize clustering performance by testing two distance matrix measures (Bray-Curtis dissimilarity and Jaccard index, as this measure effectively handles presence-absence data while accounting for joint absences) and several hierarchical clustering methods including average linkage, complete linkage, and flexible beta method with varying beta parameters (β = - 0.25, - 0.30, - 0.35, 0.40, - 0.45, - 0.50). The final analytical workflow proceeded through five sequential steps:
1. Distance matrix calculation: We computed Bray-Curtis dissimilarity matrices using the vegdist() function from the vegan package.
2. Hierarchical clustering: Cluster analysis was performed using the flexible beta method (β = - 0.25) with the AGNES (Agglomerative Nesting) algorithm implemented in the cluster package. This method was selected for its ability to handle ecological data with varying cluster structures.
3. Cluster determination: The number of clusters was determined through expert assessment of the dendrogram structure, identifying natural groupings based on visual inspection of branch lengths and clustering patterns.
4. Ordination analysis: Principal Coordinates Analysis (PCoA) was conducted using the cmdscale() function to visualize the multivariate data in two-dimensional space.
5. Cluster validation: The quality of cluster assignments was evaluated using silhouette analysis, which measures how well each site fits within its assigned cluster compared to other clusters, and the AGNES coefficient, which assesses the overall strength of the clustering structure.
Analysis of soil-vegetation relationships.
Prior to the analysis, we assessed multicollinearity among the soil predictors using Spearman's correlation coefficient. Variables E (proportion of gleyed area 40 cm) and F (proportion of gleyed area within the gleyed at 0 to 40 cm) showed a high correlation (r = 0.88), indicating potential redundancy. To ensure model stability and interpretability, we retained only variable E in the final RDA model, as it is ecologically more representative of the gradient. The core analysis consisted of several complementary multivariate techniques to quantify and characterize the influence of the eight measured soil variables on plant community composition:
1. Redundancy Analysis (RDA): We used RDA, a constrained ordination method, to model the species presence-absence data as a function of the soil variables. The global significance of the model and the significance of each canonical axis were tested using permutation tests (999 permutations). The relative importance of individual soil variables was assessed by their sum of squared loadings on the first two RDA axes.
2. Mantel Test: To evaluate the overall correlation between compositional dissimilarity and environmental dissimilarity, we performed a Mantel test. A Jaccard dissimilarity matrix was calculated from the species data and compared to a Euclidean distance matrix based on standardized soil variables, using Pearson's correlation and 999 permutations.
3. Non-metric Multidimensional Scaling (NMDS) with Environmental Fitting: We performed an unconstrained NMDS ordination based on Jaccard dissimilarity to visualize the main gradients in species composition. The relationship between the ordination space and individual soil variables was then assessed by fitting environmental vectors using the envfit() function (999 permutations).
4. Indicator Species Analysis: To identify plant taxa significantly associated with high or low values of specific soil parameters, we conducted an indicator species analysis. For each soil variable, samples were split into two groups based on the median value. Associations between species occurrence and these groups were tested using Fisher's exact test.
ResultsFloristic Diversity and Taxonomic Composition of the Studied Plant Communities
A total of 117 species from 32 families were recorded in the study (Table 1). The vascular flora comprised two families of Lycopodiophyta (Lycopodiaceae, Huperziaceae), one family of Equisetophyta (Equisetaceae), two families of Pinophyta (Cupressaceae and Pinaceae), and 28 families of Magnoliophyta. Among the angiosperms, 25 families belonged to the Magnoliopsida class, and four to the Liliopsida class (Table 1).
Table 1. Summary of floristic richness: numbers of families, genera, and species
Taxonomic Group
Number of Families
Number of Genus
Number of Species
Vascular Cryptogams (Lycopodiophyta and Equisetophyta)
3
3
5
Gymnosperm (Pinophyta)
2
2
2
Dicotyledon Dicots (Magnoliopsida)
25
55
80
Dicotyledon Monocots (Liliopsida)
4
13
30
Total
34
73
117
The families with the highest species richness were Poaceae (13 species from 8 genus), Cyperaceae (11 species from 2 genus), Salicaceae (11 species from 1 genus), Asteraceae (11 species from 11 genus), and Ericaceae (8 species from 6 genus). Several families were represented by a moderate number of species: Caryophyllaceae (7 species from 4 genus), Rosaceae (5 species from 4 genus) and Ranunculaceae (5 species from 3 genus). Four families included four species each: Lycopodiaceae (2 genus), Polygonaceae (3 genus), Scrophulariaceae (3 genus), and Juncaceae (Luzulagenus). Two families were represented by three species each: Rubiaceae (Galiumgenus) and Saxifragaceae (2 genus). Six families were represented by two species: Equisetaceae (Equisetum genus), Betulaceae (2 genus), Pyrolaceae (Pyrola genus), Caprifoliaceae (2 genus), Fabaceae (2 genus) and Melanthiaceae (2 genus). Thirteen families were represented by a single species in the study area: Pinaceae, Cupressaceae, Boraginaceae, Polemoniaceae, Geraniaceae, Violaceae, Apiaceae, Onagraceae, Primulaceae, Empetraceae, Campanulaceae, Grossulariaceae and Brassicaceae.
Classification of Shrub-Herbaceous Communities
Silhouette analysis revealed relatively weak clustering structure across all tested parameters. The highest silhouette value (0.2) was achieved using Bray-Curtis dissimilarity with the flexible beta method (β = - 0.25) when the dataset was restricted to taxa occurring in 6 or more sites and specifying 6 clusters. However, such extensive data filtration was considered ecologically inappropriate as it eliminated meaningful biological information. We therefore use a more conservative approach, including all taxa occurring in more than one geobotanical description. Given the minimal differences in silhouette values across cluster solutions (k = 2–10) (Fig. 2C), we employed expert-based cluster determination guided by botanical composition and ecological interpretation rather than lean only on statistical optimization.
<bold id="bold-fd01fb99adf96e9004e9bfdc35b96ca9"/>Analysis of shrub-herbaceous community based on presence-absence data. <bold id="bold-40680149b6c87a12d964a7c55d1430dd">A </bold>– Dendrogram showing hierarchical clustering using flexible beta method (β = – 0.25) with manually identified clusters (k = 4). <bold id="bold-c82faf1cb5b64a56adc18aaece58278e">B </bold>– Principal Coordinates Analysis (PCoA) ordination. <bold id="bold-5f7053f1406d6084f3980bd9ec42b2f3">C </bold>– Plot of average silhouette width versus number of clusters. <bold id="bold-2b404eeef0dcc058276de1f8f2db33d3">D </bold>– Bar plot displaying the size distribution of identified clusters.
Figure 2. Analysis of shrub-herbaceous community based on presence-absence data. A – Dendrogram showing hierarchical clustering using flexible beta method (β = – 0.25) with manually identified clusters (k = 4). B – Principal Coordinates Analysis (PCoA) ordination. C – Plot of average silhouette width versus number of clusters. D – Bar plot displaying the size distribution of identified clusters.
Expert assessment of the dendrogram constructed using the flexible beta method (β = - 0.25) revealed four distinct clusters of shrub-herbaceous communities (Fig. 2A). The AGNES coefficient (0.73) confirmed the presence of a reasonably strong clustering structure in the data. However, the average silhouette width (0.15, Fig. 2C) suggested that individual cluster assignments were not well defined, which may reflect the continuous nature of shrub-herbaceous community gradients in the study area. This pattern suggests that while distinct shrub community types can be identified, they form along environmental gradients where intermediate compositions are common.
The Principal Coordinates Analysis (PCoA, Fig. 2B), based on the Bray-Curtis dissimilarity, revealed the underlying structure of the shrub-herbaceous community data. The first two axes explained 19.4% and 13.7%, respectively, of the total variance. A moderate cumulative variance of 33.1% explained by the first two axes is typical for ecological community data, as multiple environmental gradients shape vegetation patterns collectively. Ordination pattern confirm the results of the cluster analysis, showing a partial separation of the four community types in a reduced- dimensional space.
The four identified clusters varied in size and composition (Fig. 2D):
Cluster 1 (red, 5.8%) represents alder thicket communities. This cluster comprises a single, distinct community characterized by the lowest species richness among all studied alder thickets, with a mean of 8 species. The most frequently occurring species were Alnus fruticosa, Betula nana, Solidago lapponica With., Polemonium caeruleum L. and Carex globularis L.
Cluster 2 (blue, 19.2%) consists of the most species-poor streamside communities, with minor representation from hollow and slope habitats (7 streamside and 3 slope communities). The mean species richness for the cluster is 10, driven by the notably low average of about 7 species in the streamside communities that dominate it. The most frequent species were Betula nana, Salix glauca, Comarum palustre L., Rubus chamaemorus L., and Carex aquatilis. Cluster 2 is likely associated with the most severe waterlogging, as indicated by the constant presence of Comaru mpalustre and Carex aquatilis.
Cluster 3 (green, 30.8%) includes both alder thickets and slope communities (7 alder thickets and 8 slope communities). The mean species richness was 16. The most common species included Equisetum arvense L., Betula nana, Vaccinium uliginosum L., Ledum palustre L., Vaccinium vitis-idaea subsp. minus (Lodd.) Hultén, Empetrum nigrum L. and Stellaria peduncularis Bunge. This cluster represents communities with a significant presence of tundra species, typical of well-drained shrub tundra ecosystems with high taxonomic diversity.
Cluster 4 (violet, 44.2%) contains a mixture of slope shrub-herbaceous communities and streamside communities (14 slope and 10 streamside communities). This cluster had the highest mean species richness (17), which indicates more favorable and heterogeneous conditions. The most common species were Betula nana, Salix glauca, Salix pulchra, Polemonium caeruleum, Equisetum arvense, Veratrum lobelianum Bernh., Rubus arcticus L., Vaccinium uliginosum, Viola epipsiloides Á. Löve & D. Löve and Pedicularis labradorica Wirsing.
Soil-vegetation relationship
The constrained ordination model, Redundancy Analysis (RDA) (Fig. 3A), provided a detailed view of this relationship and was statistically significant (F = 1.43, p = 0.001). It explained 18.6% of the total variance in species composition (Constrained variance = 1.78; Total inertia = 9.60). A permutation test on individual axes confirmed that only the first canonical axis (RDA1) was significant (p < 0.001). A Mantel test, assessing the overall correlation between the soil and vegetation distance matrices, confirmed a significant but weak positive relationship (r = 0.157, p = 0.003).
The relative contribution of individual soil variables to the RDA model, assessed by the sum of squared loadings on the first two axes, showed a clear hierarchy. The most influential parameters were physical clay content (H, 0.59), active layer thickness (A, 0.59), proportion of gleyed patches (G, 0.57) and proportion of gleyed patches 40 cm (E, 0.43). The remaining variables showed substantially lower values were peat thickness (B, 0.22), moder humus horizon (AO) thickness (C, 0.20), and gleying depth (D, 0.17).
<bold id="bold-cf7fbba7e420ea57662a5744ec5869b6"/>The relationship between soil variables and plant community composition. <bold id="bold-368cdd29745b23c3d670718cbe2b89f5">A </bold>– Redundancy Analysis (RDA) biplot. <bold id="bold-050ddd50ed1b8f103b3f942d72e59ad9">B </bold>– Non-metric Multidimensional Scaling (NMDS) ordination based on Jaccard dissimilarity. Soil variable codes: <bold id="bold-2c2c81f32f7959102c5a340113172008">A </bold>– active layer thickness, <bold id="bold-2c0c344cd08384d932c6ee7760c77c6c">B </bold>– peat thickness, <bold id="bold-38f52af296a162b872cf6dc236efcd84">C </bold>– moder humus horizon (AO) thickness, <bold id="bold-5566a25ae3a218fce2dd358162da55c3">D </bold>– gleying depth, <bold id="bold-242ae0b45b340257b09c200283b01f84">E </bold>– proportion of gleied patches at 40 cm, <bold id="bold-819c0a3d1cb5c2479c94a8ed643459b1">G </bold>– proportion of gleyed patches within the gleyed horizon, <bold id="bold-a6e509b7c4b08da9db97ced08f5aa69b">H </bold>– physical clay content.
Figure 3. The relationship between soil variables and plant community composition. A – Redundancy Analysis (RDA) biplot. B – Non-metric Multidimensional Scaling (NMDS) ordination based on Jaccard dissimilarity. Soil variable codes: A – active layer thickness, B – peat thickness, C – moder humus horizon (AO) thickness, D – gleying depth, E – proportion of gleied patches at 40 cm, G – proportion of gleyed patches within the gleyed horizon, H – physical clay content.
To visualize the main patterns in species composition independently of a predefined environmental model, we performed a Non-metric Multidimensional Scaling (NMDS) (Fig. 3B). This unconstrained ordination (stress = 0.238) represents the overall dissimilarity between samples based solely on species data (Jaccard distance). When soil variables were fitted post-hoc to this ordination space, active layer thickness (A) emerged as the strongest and only statistically significant vector (loadings: NMDS1 = 0.87, NMDS2 = 0.49). This result refines the interpretation of the RDA: while RDA identified a suite of three co-influential factors (A, G, H,) within the constrained model, NMDS suggests that the primary, unconstrained floristic gradient across all samples is most strongly and directly aligned with the cryogenic factor – active layer thickness. The other key variables from the RDA (physical clay proportion and gleyed area) likely contribute to more complex or secondary patterns of community variation.
An indicator species analysis based on Fisher's exact test identified ten statistically significant associations (p < 0.05) between plant species occurrences and soil variables (Table 2). The most frequently occurring indicator was Petasites frigidus (L.) Fr., which showed significant associations with three soil variables: proportion of gleyed area within the gleyed horizon (G) (p = 0.0006), proportion of gleyed area at 40 cm depth (E) (p = 0.0016), and physical clay content (H) (p = 0.0041).
Table 2. Plant species significantly associated with specific soil conditions (Fisher's exact test)
Plant species frequency
Soil variable
Species
P value
High soil parameter value
Low soil parameter value
Active layer thickness
Equisetum arvense
0.0047
0.7692
0.3462
Stellaria peduncularis
0.0034
0.5769
0.1538
Raw humus horizon thickness
Empetrum nigrum
0.0012
0.5625
0.1111
Tanacetum bipinnatum(L.) Sch. Bip.
0.0021
0.4375
0.0556
Gleying depth
Trollius asiaticus L.
0.0048
0.3846
0.03846
Proportion of gleyed area 40 cm
Petasites frigidus
0.0016
0.5500
0.1250
Proportion of gleyed area within the gleyed horizon
Petasites frigidus
0.0006
0.5200
0.0741
Comarum palustre
0.0013
0.1200
0.5556
Physical clay proportion
Petasites frigidus
0.0041
0.5238
0.1290
Chamaenerion angustifolium (L.) Scop .
0.0027
0.2857
0
Diversity and distribution of soil in key biotopes
The most common soils among those studied are Reductaquic Turbic cryosols. These soils were examined at 6 sites. The next most common soils are Reductaquic Turbic Criosols (Arenic) and Histic Reductaquic Turbic Cryosols, studied at 5 and 4 points. Below we provide descriptions of the characteristics of the most common soils under shrubs (Table 3).
Reductaquic Turbic Cryosols are more often formed on gentle slopes with a steepness of 3 to 5°. The upper part of the profile of these soils is characterized by a cryoturbated suprapermafrost part and abundant gleying. Gleyization appears as grey-blue and rusty-ochre patches and swirls, indicating periodic changes in oxidation-reduction conditions. Cryoturbations in the active layer are found almost everywhere. Cryoturbation results in the burial of significant amounts of carbon through the burial of organogenic particles and charcoal. The granulometric composition varies across the profile from light to heavy loam and clay. Permafrost occurs at a depth of 30–70 cm and often contains ice veins.
Table 3. Soil classification (WRB 2022) of the study sites
Soil name
Site number
Reductaquic Turbic Cryosols
25, 26, 8, 35, 36, 21
Reductaquic Turbic Cryosols (Arenic)
22, 23, 24, 30, 34
Reductaquic Cryosols
19, 20
Histic Reductaquic Turbic Cryosols
18, 29, 7, 32, 28
Fibric Histosols
5
Pantofluvic Fluvisols (Arenic)
15
Fluvisols (Arenic)
1
Gleyic Epifluvic Fluvisols (Arenic)
31
Stagnic Pantofluvic Fluvisols (Ochric)
6
Reductaquic Turbic Criosols (Arenic) are less common and are confined to the lower parts of slopes, stream valleys, and ancient lake terraces. The upper part of the profile has been significantly influenced by cryoturbation processes. The soil-forming substrate consists of ancient alluvial deposits with a sandy texture. Gleyization is mostly confined to the suprapermafrost part of the profile. The alluvial genesis of the parent rock is also identified in the upper part of the profile, but it has been heavily transformed by cryoturbation. The depth of permafrost varies from 25 to 100 cm.
Reductaquic Cryosols are found in the valleys of ephemeral (temporary) watercourses. The upper part of the profile (0-10 cm) consists of weakly decomposed peat. This is underlain by a thin layer (less than 5 cm) of raw-humus material. Beneath this raw-humus interlayer, a mineral horizon with signs of humus illuviation is present. A thixotropic horizon with gleyic features lies below. The texture class is light loam. The permafrost table is found at a depth of 25-30 cm.
Histic Reductaquic Turbic Cryosolsare characterized by a Histic horizon (15-20 cm thick) of highly decomposed peat at the surface. It is underlain by a mineral horizon impregnated with illuvial organic matter. The texture varies from sandy to light loam. The permafrost table is found at 30-80 cm.
Fibric Histosols are formed at the confluence of two streams. The soil is represented by a well-decomposed peat horizon with a thickness of 60 cm.
Pantofluvic Fluvisols (Arenic) are confined to the floodplains of small and medium-sized rivers, stream valleys, where sandy alluvium predominates. The soils are distinguished by their layered structure with layers of varying thickness. The thickness of the layers is determined by the intensity of sedimentation and the composition of the suspended matter. Burial of surface horizons is often found in the layers. For a significant period of time during the year, soils are affected by water-logging, which is manifested in the presence of gleying of the profile, which is more characteristic of the lower part of the profile.
Fluvisols (Arenic) have formed on the bed of an ephemeral watercourse. The soil profile is composed of layered, sandy alluvial deposits from the surface and exhibits signs of gleyic properties throughout.
Gleyic Epifluvic Fluvisols (Arenic) The upper part of the profile (down to 50 cm depth) consists of layered alluvial sandy deposits. These are underlain by gleyed, cryoturbated light loamy deposits. The permafrost table is found at a depth of 55 cm.
Stagnic Pantofluvic Fluvisols (Orhric) are characterized by an organic horizon in the upper part of the profile, interlayered with sandy alluvial material. The mineral part of the profile consists of alternating sandy and light loamy layers. The permafrost table occurs at a depth of 110 cm. Gleyic features are confined to the upper 20–30 cm of the soil profile.
DiscussionLandscape heterogeneity as a driver of diversity and floristic discoveries
The formation of high plant diversity zones in Arctic environments is often driven by local environmental heterogeneity (Tape et al. 2012; Loiko et al. 2024). Our results demonstrate how specific geomorphic features, such as snow accumulation hollows, well-drained slopes, and stream valleys, create a mosaic of contrasting ecological conditions that support distinct plant assemblages and enhance regional taxonomic diversity in the southern Gydan tundra.
Analysis of species richness across the researched habitats identified several consistent biodiversity hotspots. The high number of co-existing species is attributed to the proximity to contrasting landforms such as southern slopes and stream floodplains, which support the unique microclimatic regime of these habitats.
The most prominent hotspots were ecosystems associated with drained southern slopes (20–25 species, sites 7, 9, 41, 47), stream valleys (20–23 species, sites 16, 17), and hollow ecosystems (20–25 species, sites 28, 49, 50). The latter represent topographical depressions, sheltered by shrubs and experiencing snowmelt two to four weeks later than the surrounding plains, functioning as micro-refugia.
High floristic diversity was also a consistent feature of other contrasting habitat types – particularly in late-successional slope shrub communities recovering from anthropogenic impact (29 species, site 10) and in vegetation complexes adjacent to lake terraces (26 species, site 42). This further demonstrates that floristic richness in southern tundra is primarily associated with areas of high environmental heterogeneity.
However, the relationship between habitat type and community structure is not unidimensional. We observed that within the same broad categories of slopes, hollows, and valleys, certain plots exhibited an alternative ecological strategy: low to moderate species richness (3–9 species, e.g., sites 3, 6, 12–14, 19–22, 30, 34) coupled with apparent high productivity and often strong dominance of one or a few species. This pattern suggests that environmental heterogeneity drives not only peaks in taxonomic diversity but also the emergence of highly productive, low-diversity patches. These areas may represent more stable, mature, or nutrient-enriched states within the landscape mosaic, where competitive exclusion reduces species number, or particularly favorable microsites that allow for vigorous growth of a limited set of well-adapted species (Walker et al. 2009; Liljedahl et al. 2020).
Despite this variation in diversity patterns, the capacity of these heterogeneous habitats to act as biodiversity reservoirs is unequivocally demonstrated by our floristic discoveries. The discovery of several vascular plant species previously undocumented for this part of the Gydan Peninsula provides direct, taxon-specific evidence for the refugial capacity of these heterogeneous landscapes. According to the collections of the P.N. Krylov Herbarium of Tomsk State University and the National Depository Bank of Live Systems (https://plant.depo.msu.ru/?d=P), our research documents the first local records of the following species, many of which were found precisely within the identified hotspot habitats (Fig. 4).
Carex rostrata Stokes (Fig. 4A), found in slope communities (Suppl. material 3: Table S3, site 38), previously known from the Pur-Taz interfluve and the Yenisei River floodplain.
Stellaria palustris Retz. (Fig. 4B), recorded in streamside habitats (Suppl. material 2: Table S2, sites 15, 16, 27), with prior findings across the Taz Peninsula, northern Gydan, and the Yenisei floodplain.
Bistorta elliptica (Willd. ex Spreng.) Kom. (Fig. 4C), located on a southern slope near a stream (Suppl. material 2: Table S2, site 16), previously observed in the Pur-Taz interfluve and on the Taz Peninsula.
Arnica iljinii (Maguire) Iljin (Fig. 4D), documented on an upper southern slope of a river valley (Suppl. material 2: Table S2, site 26), known from the Taz Peninsula and northern Gydan.
Salix reticulata L. (Fig. 4E), found in snow accumulation hollows near a stream surrounded by shrubs (Suppl. material 2: Table S2, sites 26, 27), previously recorded on the Taz Peninsula and in the Yenisei floodplain.
Vaccinium myrtillus L. (Fig. 4F), found in a slope ecosystem (Suppl. material 3: Table S3, site 42), previously recorded in the Pur-Taz interfluve, on the Taz Peninsula, and in the Yenisei River floodplain.
Dianthus repens Willd. (Fig. 4G) found at the eroded sandy margin of an alder thicket (Suppl. material 1: Table S1, site 7), previously known from the Taz Peninsula, the central part of the Gydan Peninsula, the Arctic part of the Yenisei floodplain, and the south near the village of Tolka on the Taz River.
Figure 4. Newly reported plant species for the study area. A – Carex rostrata, B – Stellaria palustris, C – Bistorta elliptica, D – Arnica iljinii, E – Salix reticulata, F – Vaccinium myrtillus, G – Dianthus repens, H – Micranthes hieraciifolia, I – Pedicularis sudetica, J – Cardamine nymanii.
Micranthes hieraciifolia (Waldst. & Kit. ex Willd.) Haw. (Fig. 4H), located in a hollow within snow accumulation ecosystems (site 49; Suppl. material 3: Table S3), previously known from the Taz Peninsula and northern Gydan.
Pedicularis sudetica Willd. (Fig. 4I), found in thermokarst stream valleys (Suppl. material 2: Table S2, site 17) and shrub-dominated slopes (Table S3, sites 44, 45, 47, 50), documented earlier on the Taz Peninsula and in central Gydan.
Cardamine nymanii Gand. (Fig. 4J), recorded in waterlogged stream valleys (site 5; Suppl. material 1: Table S1) and near-lake shrub systems (Suppl. material 2: Table S2, site 15), previously known from the northern Gydan Peninsula and Norilsk areas.
In summary, the southern Gydan tundra shrub-herbaceous communities supports a complex mosaic driven by geomorphic heterogeneity. This mosaic includes both high-diversity microrefugia evidenced by species-rich sites and the discovery of new floristic records and high-productivity, lower-diversity patches. The cooccurrence of these distinct community types within the same landforms strongly supports the hypothesis that landscape heterogeneity is the primary architect of both taxonomic and functional diversity. The identified hotspots are not merely areas of high species count but function as essential microrefugia and productive nodes, providing the stable, buffered niches necessary for species coexistence, the persistence of rare taxa, and the optimization of biomass production under different local conditions. This underscores the paramount importance of conserving the full spectrum of landscape heterogeneity to maintain the ecological integrity and resilience of Arctic tundra ecosystems.
The role of soil drivers at vegetation structure
Our integrated analysis confirms that specific soil properties are significant, yet secondary, factors for plant community assembly in the studied tundra landscapes. The significant RDA model (p=0.001) and Mantel correlation (r=0.157, p=0.003) unequivocally demonstrate an soil signal. However, the modest proportion of explained variance (18.6%) and the weak Mantel correlation indicate that the measured soil parameters are not the primary determinant of vegetation structure. Within this constrained variance, a clear hierarchy emerged: soil texture (physical clay proportion, H), cryogenic conditions (active layer thickness, A), and hydromorphic features (proportion of gleyed area within the gleyed horizon, G) were the principal drivers. This hierarchy aligns with the known mechanisms of shrub-soil feedbacks in the Arctic, where shrub expansion is frequently linked to altered soil texture, deepened active layers, and modified hydrology (Frost et al. 2018; Kropp et al. 2020) These factors govern base conditions like moisture regime and rooting depth.
The significance of only the first RDA axis (p < 0.001) indicates that the primary variation in plant communities is primarily driven by a single dominant factor as a complex interplay of soil development and hydrology, shaped by these key factors.
The indicator species analysis provides concrete evidence for these relationships, with species like Petasites frigidus showing significant associations with multiple variables from this core group (proportion of gleyed area within the gleyed horizon, G; proportion of gleyed area at 40 cm, E; and physical clay proportion, H). However, the modest overall explanatory power of our soil model is itself an insightful result. It underscores a key principle in contemporary tundra ecology: while soil properties provide an important filter, vegetation patterns at the landscape scale are emergent properties of a more complex interplay (Myers-Smith et al. 2011; Heijmans et al. 2022). Thus, the obtained results suggest that, despite the identification of specific significant soil parameters, they control only a part of the observed vegetation structure. A substantial proportion of the variation (81.4%) is likely due to other factors not accounted for in this model. This explains the overall weak relationship identified by the Mantel test and points to the complex nature of plant community formation in the studied ecosystems of Gyda Peninsula.
Conclusions
This integrated study provides novel baseline data on the co-evolution of plant communities and soils in the poorly documented southwestern Gydan Peninsula. Our findings offer several key insights into the drivers of tundra ecosystem diversity:
1. Landscape-scale geomorphic heterogeneity – the mosaic of slopes, valleys, and hollows – emerges as the primary architect of both taxonomic and functional diversity. These features create essential microrefugia that support high-diversity hotspots and are directly evidenced by the discovery of several plant species previously unreported for this region. This underscores the critical importance of conserving the full spectrum of landscape heterogeneity to maintain ecological resilience.
2. While a significant soil signal was detected, its measured influence was quantitatively limited compared to the overarching landscape template. A hierarchy of key soil drivers was identified, with properties related to texture, cryogenic conditions, and hydromorphy being the most influential. However, the substantial portion of vegetation variation unexplained by the soil model points to the dominant role of other factors.
3. Vegetation structure itself exhibits a continuous gradient, suggesting that shrub-herbaceous communities assemble along environmental continua rather than as discrete units. The strong alignment of the main floristic gradient with active layer thickness highlights the fundamental role of this cryogenic factor.
In summary, our research establishes a critical baseline for monitoring change. It demonstrates that the transformation of the southern Gydan tundra is governed by a hierarchy where the landscape template sets the stage, while localized soil and cryogenic processes fine-tune community composition. Future studies should prioritize quantifying the microclimatic and historical drivers behind the unexplained variation to improve predictive models of ecosystem trajectory under ongoing climate change.
Acknowledgements
The study was performed within the framework of a state assignment of the Ministry of Science and Higher Education of the Russian Federation (No FSWM-2024- 0006).
The authors would like to thank M.V. Olonova and I.I. Gureeva for their consultations during the work with the herbarium material.
References
Anisimov OA, Kokorev VA (2017) Russian permafrost in the 21st century: model-based projections and analysis of uncertainties. Earth`s Cryosphere 21(2): 3–9. [In Russian]
Atlas Tyumenskoy oblasti (1971) Issue 1. GUGK, Tyumen, 198 pp. [In Russian]
Atlas of the Yamalo-Nenets Autonomous Okrug (2004) Omsk Cartographic Factory, Sale- khard, 303 pp. [In Russian]
Berner LT, Massey R, Jantz P, Forbes BC, Macias-Fauria M, Myers-Smith I, Kumpula T, Gauthier G, Andreu-Hayles L, Gaglioti BV, Burns P, Zetterberg P, D’Arrigo R, Goetz SJ (2020) Summer warming explains widespread but not uniform greening in the Arctic tundra biome. Nature Communications 11(1): 1–12. https://doi.org/10.1038/s41467-020-18479-5
Bonfils CJW, Phillips TJ, Lawrence DM, Cameron-Smith P, Riley WJ, Subin ZM (2012) On the influence of shrub height and expansion on northern high latitude climate. Environmental Research Letters 7(1): 015503. https://doi.org/10.1088/1748-9326/7/1/015503
Chen Y, Hu FS, Lara MJ (2021) Divergent shrub-cover responses driven by climate, wildfire, and permafrost interactions in Arctic tundra ecosystems: Global change biology 27(3): 652–663. https://doi.org/10.1111/gcb.15451
Ilyina IS, Lapshina EI, Lavrenko NN, Meltzer LI, Romanova EA, Bogoyavlensky BA, Makhno VD (1985) Vegetation cover of the West Siberian Plain. Nauka, Novosibirsk, 251 pp. [In Russian]
Fewster RE, Morris PJ, Ivanovic RF, Swindles GT, Peregon AM, Smith CJ (2022) Imminent loss of climate space for permafrost peatlands in Europe and Western Siberia. Nature Climate Change 12(4): 373–379. https://doi.org/10.1038/s41558-022-01296-7
Frost GV, Epstein HE, Walker DA, Matyshak G, Ermokhina K (2013) Patterned-ground facilitates shrub expansion in Low Arctic tundra. Environmental Research Letters 8(1): 015035. https://doi.org/10.1088/1748-9326/8/1/015035
Frost GV, Epstein HE, Walker DA, Matyshak G, Ermokhina K (2018) Seasonal and Long-Term Changes to Active-Layer Temperatures after Tall Shrubland Expansion and Succession in Arctic Tundra. Ecosystems 21(3): 507–520. https://doi.org/10.1007/s10021-017-0165-5
Frost GV, Roland CA, Schmidt JH (2023) Dynamic disequilibrium: Recent widespread increases in vegetation cover on subarctic floodplains of Interior Alaska. Ecosphere 14: e4344. https://doi.org/10.1002/ecs2.4344
Geocryology of the USSR (1989) Western Siberia. Nedra, Moscow, 453 pp. [In Russian]
Glazunov VA, Nikolaenko SA (2018) Materials on the flora of the vicinity of the villages of Tazovsky and Gaz-Sale (Yamalo-Nenets Autonomous Okrug). In: Man and North: Anthropology, Archaeology, Ecology. Materials of All-Russian Scientific Conference, Tyumen, April 2–6. Tyumen Scientific Center SB RAS Press, Tyumen, 501–504. [In Russian]
Gureeva II (2013) Herbarium Science: Guide to Organizing a Herbarium and Working with Herbarium Collections. Study Guide. 2nd edition, corrected and supplemented. Tomsk State University Publishing House, Tomsk, 194 pp. [In Russian]
Heijmans MMPD, Magnusson RI, Lara MJ, Frost GV, Myers-Smith IH, van Huissteden J, Jorgenson MT, Fedorov AN, Epstein HE, Lawrence DM, Limpens J (2022) Tundra vegetation change and impacts on permafrost. Nature Reviews Earth and Environment 3(1): 68–84. https://doi.org/10.1038/s43017-021-00233-0
Huang M, Piao S, Janssens IA, Zhu Z, Wang T, Wu D, Ciais P, Myneni RB, Peaucelle M, Peng S (2017) Velocity of change in vegetation productivity over northern high latitudes. Nature ecology and evolution 1(11): 1649–1654. https://doi.org/10.1038/s41559-017-0328-y
Kaverin D, Malkova G, Zamolodchikov D, Shiklomanov N, Pastukhov A, Novakovskiy A, Sadurtdinov M, Skvortsov A, Tsarev A, Pochikalov A, Malitsky S, Kraev G (2021) Long-term active layer monitoring at CALM sites in the Russian European North. Polar Geography 44(3): 203–216. https://doi.org/10.1080/1088937X.2021.1981476
Khitun OV, Rebristaya OV (1998) Vegetation and ecotopical structure of the flora of cape Khonorasale surroundings (Gydansky peninsula, the arctic tundra subzone). Botanicheskii Zhurnal 83 (12): 21–37. [In Russian]
Khitun OV (2002) Intralandscape structure of the flora of the Tinikyakha river lower reaches (northern hypoarctic tundra, Gydansy peninsula). Botanicheskii Zhurnal 87 (8): 1–24. [In Russian]
Khitun OV (2003) Analysis of the intralandscape structure of the flora of the Khalmeryakha river middle reaches (West Sberian Arctic). Botanicheskii Zhurnal 88 (10): 9–30. [In Russian]
Konstantinov A, Loiko S, Kurasova A, Konstantinova E, Novoselov A, Istigechev G, Kulizhskiy S (2019) First findings of buried late-glacial paleosols within the dune fields of the Tomsk Priobye region (SE western Siberia, Russia). Geosciences 9(2): 82. https://doi.org/10.3390/geosciences9020082
Kropp H, Loranty MM, Natali SM, Kholodov AL, Rocha AV, Myers-Smith I, Abbot BW, Abermann J, Blanc-Betes E, Blok D (2020) Shallow soils are warmer under trees and tall shrubs across Arctic and Boreal ecosystems. Environmental research letters 16(1): 015001. https://doi.org/10.1088/1748-9326/abc994
Lara MJ, Nitze I, Grosse G, Martin P, McGuire AD (2018) Reduced Arctic tundra productivity linked with landform and climate change interactions. Scientific reports 8(1): 1–10. https://doi.org/10.1038/s41598-018-20692-8
Liljedahl AK, Timling I, Frost GV, Daanen RP 2020 Arctic riparian shrub expansion indicates a shift from streams gaining water to those that lose flow. Communications Earth and Environment 1(1): 1–9. https://doi.org/10.1038/s43247-020-00050-1
Liu Y, Riley WJ, Keenan TF, Mekonnen ZA, Holm JA, Zhu Q, Torn MS (2022) Dispersal and fire limit Arctic shrub expansion. Nature communications 13(1): 1–10. https://doi.org/10.1038/s41467-022-31597-6
Loiko SV, Gerasko LI, Kulizhskiy SP (2011) Grouping of Soil Memory Carriers (Case Study of the Northern Part of the Chernevaya Taiga Range). Tomsk State University Journal of Biology 3: 38–49.
Loiko SV, Bobrovsky MV, Novokreshchennykh TA (2013) Signs of Windthrow Morphogenesis in Background Soils of the Chernevaya Taiga (Case Study of the Tom-Yaya Interfluve). Tomsk State University Journal of Biology 4: 20–35.
Loiko SV, Gerasko LI, Kulizhskiy SP, Amelin II, Istigechev GI (2015) Soil Cover Structure in the Northern Part of the Chernevaya Taiga Range in the Southeast of Western Siberia. Eurasian Soil Science 48(4): 410–423. https://doi.org/10.7868/S0032180X15040061
Loiko S, Klimova N, Kuzmina D, Pokrovsky O (2020) Lake drainage in permafrost regions produces variable plant communities of high biomass and productivity. Plants 9: 867. https://doi.org/10.3390/plants9070867
Loiko SV, Klimova NV, Kritckov IV, Kuzmina DM, Kulizhsky SP (2023) Soils and vegetation of the riverside floodplain in the hydrological continuum of the southern tundra within the Pur-Taz interfluve (Western Siberia). Acta Biologica Sibirica 9: 293–315. https://doi.org/10.5281/zenodo.7879848
Loiko SV, Kriсkov IV, Shefer NV, Nedyak I, Manasypov RM, Kulizhskiy SP (2024) Soils and vegetation of the permafrost floodplain of the small river Tenyakha (Messoyakha basin, Western Siberia). Acta Biologica Sibirica. 10: 1779–1803. https://doi.org/10.5281/zenodo.12668698
Loranty MM, Goetz SJ (2012) Shrub expansion and climate feedbacks in Arctic tundra. Environmental Research Letters 7(1): 011005. https://doi.org/10.1088/1748-9326/7/1/011005
Maechler M, Rousseeuw P, Struyf A, Hubert M, Hornik K (2025) cluster: Cluster Analysis Basics and Extensions. R package version 2.1.8.1. https://CRAN.R-project.org/package=cluster
Mann DH, Groves P, Reanier RE, Kunz ML (2010) Floodplains, permafrost, cottonwood trees, and peat: What happened the last time climate warmed suddenly in arctic Alaska? Quaternary Science Reviews 29: 3812–3830. https://doi.org/10.1016/j.quascirev.2010.09.002
Map of soil-ecological zoning of the Russian Federation: scale 1:8 000 000 (2020) Explanatory text and legend to the map. MAKS Press, Moscow, 98 pp. [In Russian]
Mekonnen Z, Riley W, Berner L, Bouskill N, Torn M, Iwahana G, Breen A, Myers-Smith I, Criado M, Liu Y (2021) Arctic tundra shrubification: A review of mechanisms and impacts on ecosystem carbon balance. Environmental Research Letters 16(5): 053001. https://doi.org/10.1088/1748-9326/abf28b
Myers-Smith IH, Forbes BC, Wilmking M, Hallinger M, Lantz T, Blok D, Tape KD, Macias-Fauria M, Sass-Klaassen U, Levesque E, Boudreau S, Ropars P, Hermanutz L, Trant A, Collier LS, Weijers S, Rozema J, Rayback SA, Schmidt NM, Schaepman-Strub G, Wipf S, Rixen C, Menard CB, Venn S, Goetz S, Andreu-Hayles L, Elmendorf S, Ravolainen V, Welker J, Grogan P, Epstein HE, Hik DS (2011) Shrub expansion in tundra ecosystems: dynamics, impacts and research priorities. Environmental Research Letters 6(4): 045509. https://doi.org/10.1088/1748-9326/6/4/045509
Myers-Smith IH, Kerby JT, Phoenix GK, Bjerke JW, Epstein HE, Assmann JJ, John C, Andreu-Hayles L, Angers-Blondin S, Beck PSA, Berner LT, Bhatt US, Bjorkman AD, Blok D, Bryn A, Christiansen CT, Cornelissen JHC, Cunliffe AM, Elmendorf SC, Forbes BC, Goetz SJ, Hollister RD, de Jong R, Loranty MM, Macias-Fauria M, Maseyk K, Normand S, Olofsson J, Parker TC, Parmentier F-JW, Post E, Schaepman-Strub G, Stordal F, Sullivan PF, Thomas HJD, Tømmervik H, Treharne R, Tweedie CE, Walker DA, Wilmking M, Wipf S (2020) Complexity revealed in the greening of the Arctic. Nature Climate Change 10(2): 106–117. https://doi.org/10.1038/s41558-019-0688-1
Oksanen J, Simpson GL, Blanchet FG, Kindt R, Legendre P, Minchin PR, O'Hara RB, Solymos P, Stevens MHH, Szoecs E, Wagner H, Barbour M, Bedward M, Bolker B, Borcard D, Borman T, Carvalho G, Chirico M, De Caceres M, Durand S, Evangelista HBA, Fitz-John R, Friendly M, Furneaux B, Hannigan G, Hill MO, Lahti L, Martino C, McGlinn D, Ouellette M-H, Ribeiro Cunha E, Smith T, Stier A, Ter Braak CJF, Weedon J (2025) vegan: Community Ecology Package. R package version 2.7-2. https://CRAN.R-project.org/package=vegan
Pedersen T (2025) patchwork: The Composer of Plots. R package version 1.3.2.9000. https://patchwork.data-imaginist.com
Pisarenko OU, Lapshina ED, Bezgodov AG (2017) On the bryoflora of the Yamalo-Nenets Autonomous Okrug. Turczaninowia 20(1): 35–51. https://doi.org/10.14258/turczaninowia.20.1.3 [In Russian]
Pismarkina EV, Byalt VV (2016) Materials for the study of biodiversity in the Yamalo-Nenets Autonomous District: vascular plants of the Nuny-Yaha River basin. Bulletin of the Orenburg State Pedagogical University 17: 49–69. [In Russian]
Reichle LM, Epstein HE, Bhatt US, Raynolds MK, Walker DA (2018) Spatial heterogeneity of the temporal dynamics of Arctic Tundra vegetation. Geophysical Research Letters 45(17): 9206–9215. https://doi.org/10.1029/2018GL078820
Syso AI, Sokolov DA, Siromlya TI, Ermolov YV, Makhatkov ID (2022) Anthropogenic transformation of soil properties in Taimyr landscapes. Eurasian Soil Science 55: 541–555. https://doi.org/10.1134/S1064229322050088
Tape K, Sturm M, Racine C (2006) The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology 12(4): 686–702. https://doi.org/10.1111/j.1365-2486.2006.01128.x
Tape KD, Hallinger M, Welker JM, Ruess RW (2012) Landscape Heterogeneity of Shrub Expansion in Arctic Alaska. Ecosystems 15(5): 711–724. https://doi.org/10.1007/s10021-012-9540-4
Tassone MS, Epstein HE, Armstrong AH, Bhatt US, Frost GV, Heim B, Raynolds MK, Walker DA (2024) Drivers of heterogeneity in tundra vegetation productivity on the Yamal Peninsula, Siberia, Russia. Environmental Research: Ecology 3: 015003. https://doi.org/10.1088/2752-664X/ad220f
Telyatnikov MYu, Troeva EI, Ermokhina KA, Pristyazhnyuk SA (2019) Vegetation of the two regions of the northern part of the Gydan Peninsula (the subzone of typical tundras). Turczaninowia 22(4): 128–144. https://doi.org/10.14258/turczaninowia.22.4.14 [In Russian]
Telyatnikov MYu, Khitun OV, Chernyadyeva IV, Kuzmina EYu, Ermokhina KA (2021) New data on vegetation in two areas of the southern part of the typical tundra subzone of the Gydan Peninsula. Turczaninowia 24(3): 5–23. https://doi.org/10.14258/turczaninowia.24.3.1 [In Russian]
Tishkov AA, Krenke AN (2015) "Greening" of the Arctic in the 21st century as a synergistic effect of global warming and economic development. Arctic: Ecology and Economics 4: 28–38. [In Russian]
Titov YuV, Potokin AF (2001) Vegetation of the Taz River floodplain. Surgut State University Publishing House, Surgut, 140 pp. [In Russian]
Trofimova IE, Balybina AS (2014) Classification of climates and climatic regionalization of the West-Siberian plain. Geography and Natural Resources 35: 114–122. [In Russian]
Vasiliev AA, Drozdov DS, Gravis AG, Malkova GV, Nyland KE, Streletskiy DA (2020) Permafrost degradation in the western Russian arctic. Environmental Research Letters 15(4): 045001. https://doi.org/10.1088/1748-9326/ab6f12
Villani M, Mauclet E, Agnan Y, Druel A, Jasinski B, Taylor M, Schuur EAG, Opfergelt S (2022) Mineral element recycling in topsoil following permafrost degradation and a vegetation shift in sub-Arctic tundra. Geoderma 421: 115915. https://doi.org/10.1016/j.geoderma.2022.115915
Vowles T, Björk RG (2019) Implications of evergreen shrub expansion in the Arctic. Journal of Ecology 107(2): 650–655. https://doi.org/10.1111/1365-2745.13081
Walker DA, Leibman MO, Epstein HE, Forbes BC, Bhatt US, Raynolds MK, Comiso JC, Gubarkov AA, Khomutov AV, Jia GJ (2009) Spatial and temporal patterns of greenness on the Yamal Peninsula, Russia: interactions of ecological and social factors affecting the Arctic normalized difference vegetation index. Environmental Research Letters 4(4): 045004. https://doi.org/10.1088/1748-9326/4/4/045004
Wickham H (2016) ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. https://ggplot2.tidyverse.org
Wickham H, François R, Henry L, Müller K, Vaughan D (2025) dplyr: A Grammar of Data Manipulation. R package version 1.1.4. https://dplyr.tidyverse.org
Wickham H, Vaughan D, Girlich M (2025b) tidyr: Tidy Messy Data. R package version 1.3.1. https://tidyr.tidyverse.org
Wu Z, Dijkstra P, Koch GW, Peñuelas J, Hungate BA (2011) Responses of terrestrial ecosystems to temperature and precipitation change: A meta-analysis of experimental manipulation. Global Change Biology 17(2): 927–942. https://doi.org/10.1111/j.1365-2486.2010.02302
Zarov EA, Golubyatnikov LL, Lapshina ED, Loiko SV (2021) Vegetation and Soils of Tundra Landscapes in the Pur–Taz Interfluvial Region. Biology Bulletin of the Russian Academy of Sciences 48: S118–S127. https://doi.org/10.1134/S1062359022010186
Supplementary material 1Table S1. Taxonomical composition of shrub-herbaceous alder thickets communities
Authors: Nikita V. Shefer, Ivan V. Kritskov, Georgyi I. Istigechev, Sergey V. Loiko
Data type: table
Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Supplementary material 2Table S2. Taxonomical composition of shrub-herbaceous streamside communities
Authors: Nikita V. Shefer, Ivan V. Kritskov, Georgyi I. Istigechev, Sergey V. Loiko
Data type: table
Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Supplementary material 3Table S3. Taxonomical composition of shrub-herbaceous communities of slopes and hollows
Authors: Nikita V. Shefer, Ivan V. Kritskov, Georgyi I. Istigechev, Sergey V. Loiko
Data type: table
Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Supplementary material 4Table S4. Morphological parameters of soil profiles
Authors: Nikita V. Shefer, Ivan V. Kritskov, Georgyi I. Istigechev, Sergey V. Loiko
Data type: table
Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.