To visualize and analyse the spatial distribution of peat cores, a shapefile (SHP) layer was created for WSPC database. All sites were georeferenced and compiled into a unified GIS layer using QGIS 3.16.2 software (QGIS Development Team, 2025), with the database attribute structure described in Table 1. The extent of the study region, as well as all subsequent spatial analyses, followed the WSL boundary as described by Sheng et al. (2004). Given the vast extent of the study region and clustering of coring sites, peat core density was assessed by dividing the area into 50×50 km grid cells. The number of cores within each cell was calculated using the join attributes by location (summary) tool (geometric predicate: contains) in QGIS. Besides the attributes obtained from the original publications, each peat core record was supplemented with information on its location within specific vegetation, permafrost, and peatland zones using the same tool in QGIS. Spatial data about vegetation zones were obtained from the map of terrestrial ecoregions by Olson et al. (2001), while permafrost extent was derived from the circum-Arctic map of permafrost and ground-ice conditions by Brown et al. (2002). The spatial extent of peatland zones was based on the map of major peatland zones originally published by Ivanov and Novikov (1976) and reproduced in Kremenetski et al. (2003). This map was georeferenced and digitized for use in this study.

To analyse the percentage of peat cores covering each time interval, we assumed that the age of the record represents the period from the time of sampling back to the oldest dated level, although in some cases (particularly in the northern part of the study area) this may not have been the case. The same assumption was used to analyse the temporal coverage of individual proxies across WSL. Given the substantial influence of permafrost on sampling and data availability, peat cores were classified into three groups corresponding to different permafrost zones: (1) non-permafrost and isolated permafrost, (2) sporadic and discontinuous permafrost, and (3) continuous permafrost.

To evaluate peat core datasets, we developed a quantitative scoring system to assess the quality and informational value of each peat core record. Proxies were grouped into three categories (biological, physical, and chemical) and each proxy was assigned an individual weight reflecting its relative contribution to environmental reconstructions (Table A3). In addition to proxy-specific weights, we also applied an additional weighting step when integrating proxy information into the final scores. For each proxy group (biological, physical, and chemical), the total sum of all proxy weights from a given group was multiplied by a factor of 4 to emphasize the contribution of group-level proxy richness. When calculating the overall proxy score, the total sum of all proxies across groups was instead multiplied by 5, giving greater influence to records that include a broad suite of proxy types. Each record was also evaluated based on (1) the presence of a chronology, (2) total record length, and (3) the number of proxy types available. These components were scaled to 0–1 and combined with the weighted proxy information to generate four aggregated scores (combined_bio, combined_phys, combined_chem, and combined_all) (Table A4). Higher scores indicate records with greater chronological control, longer temporal coverage, and richer multiproxy data. All calculations were performed in R Statistical Software (R Core Team, 2022) using the dplyr package (Wickham et al., 2023) for data manipulation. Spatial patterns of combined scores were visualized using QGIS. We applied the Heatmap (Kernel Density Estimation, KDE) tool to each dataset, using a 30 km radius and a pixel size of 100×100 m. The KDE was weighted by the combined scores of all proxies as well as separately for the biological, chemical, and physical proxy groups. The resulting density maps provide a spatially smoothed representation of proxy distribution and highlight both well-sampled regions and areas with insufficient proxy coverage.

For the literature analysis, if the full text of publication was unavailable, information on the language and proxies used was inferred from the available title, abstract, or bibliographic data accessible to the authors. All plots and diagrams were prepared in R Statistical Software (R Core Team, 2022) using ggplot2 (Wickham, 2016) package, and final figures were assembled in CorelDRAW Graphic Suite 2021.

The database was compiled from a total of 156 publications published between 1953 and 2025, which were classified into two categories. The first group primary sources included 99 publications identified through an extensive literature review, from which data were directly obtained and used to build the database. The second group secondary sources included 57 original publications that were not directly accessible, but were cited within the primary publications used for data extraction. For each core, references are provided both to the publication from which the data were obtained (primary source) and, where available, to the original source publication cited therein (secondary source). The WSPC database is based on 82 English-language, 73 Russian-language, and one German-language publications. The authors had direct access to 79 of the English-language and 20 of the Russian-language sources.

The publications used in the compilation of the database contained information on varying numbers of peat cores. The majority of studies (82 %) presented data from one to five cores, typically providing reconstructions at the local or regional scale. Another 13 % of the publications included results from six to fifty cores. These studies analysed various regions within the Western Siberian Lowland (WSL), often along transects across peatlands, or incorporated WSL data as part of larger datasets. Only eight publications contained datasets with more than fifty peat cores. These studies generally focused on large-scale syntheses addressing specific research questions at the global scale (such as carbon accumulation) or provided a more detailed analysis of the WSL, and include Kremenetski et al. (2003), Sheng et al. (2004), Smith et al. (2004, 2012), MacDonald et al. (2006), Loisel et al. (2014), Hugelius et al. (2020), and Lapshina and Zarov (2023).

Based on conducted literature review, it was observed that between 1950 and 1990, the number of peat core studies in WSL increased notably, with most results published in Russian-language journals or books and only a limited number appearing in English (Fig. 3a). This pattern reflects the early growth of peat research during the Soviet period and the dominance of domestic Soviet research activity, as the region remained largely inaccessible to the international scientific community. After 1990, this trend shifted, as the number of English-language publications increased rapidly, reflecting the onset of international collaboration after the collapse of the USSR. The decade 2010–2020 saw the highest number of publications, reflecting not only regional research dynamics but also the broader global growth in scientific output. Although the decade 2020–2030 is only halfway through, 28 papers on this topic have already been published. While this trend may suggest that the current decade will produce the highest number of publications, ongoing geopolitical tensions, limited accessibility of some areas, and difficulties in collaborating with Russian scientists may slow down further research activity.

Based on the use of proxies, two distinct periods can be identified in WSL studies: an early stage (1950–2000), characterized by single-proxy studies, and the modern period (after 2000), marked by increasing integration of biological, physical, and chemical indicators (Fig. 3b). A major step forward in peatland research was the implementation of radiocarbon dating, developed in the 1940s (Olsson, 2009). Since the 1960s, it has become increasingly important in paleoenvironmental studies and remains the most widely applied chronological tool in peatland research in WSL (Kremenetski et al., 2003). Whereas earlier studies often reported results solely based on depth, radiocarbon dating is now used almost universally. Basal depth was an important physical parameter not only from a scientific perspective but also from an economic one. The discovery of oil in the 1950s initiated intensive studies of WSL peatlands, as information on peat thickness was essential for constructing roads and infrastructure required for oil exploration (Kremenetski et al., 2003). Early decades were dominated by classic palynological studies. Pollen was the first biological proxy applied in this region and has maintained its importance to the present day. However, although pollen analysis is widely used, it primarily provides information about the vegetation and environment surrounding the peatland rather than the peatland itself. Growing interest in peatland development led to the increasing use of plant macrofossil analysis, which began to be applied from the 1980s onward. After 2000, more proxies started to be implemented in research on WSL. New biological proxies, such as testate amoebae and micro- and macrocharcoal, were introduced, providing valuable tools for reconstructing changes in peatland hydrology and fire activity at different spatial scales. More laboratory analyses aimed at obtaining more precise data on the physical and chemical properties of peat began to be applied (e.g., bulk density, pH, ash content), along with newly developed technologies introduced over the past decade (e.g., XRF (X-ray fluorescence), FTIR (Fourier-transform infrared spectroscopy)). Peat and carbon accumulation were already studied in research conducted before 2000; however, their frequency had increased noticeably since then. The increased variety of applied proxies over the past 25 years, on one hand reflects the development of new research methods, and on the other, a growing awareness of the importance of peatland ecosystems and the desire to better understand their history.

The Western Siberian Peat Core Database (WSPC) includes information on the locations and applied palaeoecological proxies of 654 peat cores or peat cross-sections collected from the WSL (Fig. 4a). The spatial distribution of cores highlights the considerable challenges associated with conducting research in the WSL peatlands. Their vast extent, the widespread presence of permafrost, and limited transport accessibility largely determine the areas where sediment sampling is logistically feasible. Notably, 47 % of all peat cores (306 out of 654) included in the WSPC database were collected within a 100 km buffer zone of the major rivers: Ob, Yenisey, Irtysh, and Tobol, with high density of cores in the Middle Ob' Lowland (Fig. 4b). A distinct concentration of collected cores is observed in the Great Vasyugan Mire (west of Tomsk), which represents the largest contiguous peatland in the world (Fig. 4c) (Kirpotin et al., 2009). Various parts of this extensive complex have been the focus of numerous previous studies (e.g. Blyakharchuk et al., 2019; Lapshina and Zarov, 2023; Rudmin et al., 2018; Syso and Peregon, 2009) due to the absence of permafrost and the proximity to major cities facilitating field work organisation. Another distinct concentration of peat cores is observed in the central part of the study region, between the Ob and Yenisey rivers, where numerous cores were collected during extensive Russian-American field campaigns conducted between 1999 and 2001 (Sheng et al., 2004; Smith et al., 2000). Although the railway network in WSL is relatively sparse, a major line crosses this area, connecting Tyumen with Novy Urengoy, Nadym, and Yamburg, thereby facilitating logistical access to this part of the region. Peatlands in the northern part of the WSL have been studied mostly on the Yamal Peninsula, with a few coring sites on the Gyda Peninsula, mainly along the shore of the Gydan Bay. The northernmost core in the database was located on Bely Island (73.34190° N, 70.12760° E) and southernmost on Dolgon'koye swamp (53.21250° N, 84.40720° E).

The WSPC database includes peat cores, representing a broad range of environmental settings across the region. Peatlands in WSL occupy all major permafrost zones and vegetation types, although their spatial extent and sampling intensity vary (Fig. 5a, c). Peatlands are most extensive in the non-permafrost zone, which covers 41.5 % of the total peatland area, followed by the discontinuous (18.6 %), sporadic (14.1 %), isolated (13.9 %), and continuous (11.9 %) permafrost zones. In general, the number of peat cores reflects the relative area of peatlands in each zone. The two permafrost zones with the largest peatland areas also have the highest numbers of peat cores: the non-permafrost zone (278 cores, 42.5 % of the total) and the discontinuous permafrost zone (132 cores, 20.2 %). The number of peat cores in the isolated (94 cores, 14.4 %) and continuous (101 cores, 15.4 %) permafrost zones also corresponds closely to the proportion of peatland area in these zones. The main exception is the sporadic permafrost zone, which contains the third-largest share of peatland area (14.1 %) but is relatively underrepresented, accounting for only 7.5 % of all peat cores in the database.

Across vegetation types, the majority of WSL peatlands are associated with boreal forests/taiga (78.9 %), followed by tundra (13.7 %), temperate broadleaf and mixed forests (6.1 %), and temperate grasslands and shrublands (1.3 %). The distribution of peat cores reflects this pattern: cores from peatlands in boreal forests dominate the dataset (405 cores, 61.9 %), followed by tundra sites (118 cores, 18.0 %), while mixed forests (92 cores, 14.1 %) and shrubland peatlands (39 cores, 6.0 %) are considerably less represented by peat cores in the database.

Based on major peatland zones, polygonal peatlands, which dominate north of 67° N and are associated with continuous permafrost and tundra vegetation, are represented by 90 cores (13.8 %) in the database (Fig. 5b, c). Flat hill and large hill peatlands, which occur predominantly within the discontinuous permafrost zone and are associated with taiga vegetation between 61–67° N, account for 107 (16.4 %) and 49 (7.5 %) cores, respectively. Oligotrophic (Sphagnum-dominated) peatlands, concentrated between 56–61° N and typically associated with boreal forest/taiga areas exhibiting either non-permafrost, isolated or sporadic permafrost, represent the most frequent peatland type in WSL and are represented by the largest group of 337 cores (51.5 %). The two southernmost peatland zones, where permafrost is absent and temperate vegetation types predominate, are the least represented in the dataset. Flat eutrophic and mesotrophic peatlands, occurring in a narrow belt from 55–58° N in the east and up to 60° N in the west, include 34 cores (5.2 %), while eutrophic and brackish peatlands, located south of 55° N, form the smallest group, with 30 cores (4.6 %). Additionally, 7 cores (1.0 %) are outside the extent of the defined peatland zones.

The WSPC database documents the application of 26 palaeoecological proxies in peat cores in the WSL, grouped into four categories: biological, physical, chemical and chronological (Fig. 6b, Table A1). Peat cores in the database differ substantially in the number of applied proxies, ranging from only one to a maximum of 12 per core (Figs. 6a, 8a). The most common in the database were cores containing one or two palaeoecological proxies, representing 26.3 % and 28.4 % of all records, respectively. Peat cores that include only a single proxy most commonly contain information on basal depth or sediment age; and in some cases, pollen or plant macrofossil data. For cores containing only biological proxies, this may be explained by the extraction of data from more recent, method-focused publications (e.g. palynology), whereas broader and potentially more comprehensive information may have been reported in earlier sources that were inaccessible for the authors of this database. In cores containing two proxies, the most frequent combinations include basal depth or age paired with either pollen or carbon accumulation data. Peat cores containing data from three to six proxies typically include several biological proxies combined with age or basal depth information, and in some cases supplemented with peat property measurements (e.g. peat or carbon accumulation, or organic matter content). The second most common proxy count is seven (13.0 %). Within this group, two distinct core types can be distinguished: (1) cores that resemble the “3–6 proxy” category but incorporate more biological proxies, and (2) cores that lack biological proxies but provide more extensive data on the physical and chemical properties of peat. Cores containing eight or more proxies represent multi-proxy analyses, integrating biological, physical, and chemical data along with sediment dating, and thus provide the most comprehensive information on the history of the studied peatlands. Only 9.8 % of all cores include this number of proxies, while just 1.2 % contain ten or more. The spatial distribution of peat cores with a given number of proxies is also evident (Fig. 8a). High concentrations of well-studied peatlands occur along major rivers, particularly the Ob (Mukhrino, Sredne Vasyuganskoe, Samara, and Bakchar) and the Yenisey (Igarka). Given its large size and ecological significance, the Great Vasyugan Mire also contains numerous peat cores with a high number of proxies, reflecting its long-term use as a key site for multi-proxy palaeoecological research. Another cluster of cores analysed with multiple proxies is observed in the central part of the study region, between the Ob and Yenisey rivers. These cores were mostly collected during Russian-American field campaigns and include detailed physical and chemical analyses of peat, as well as cores described in single-core studies, such as the Khanymei peatland (Halaś et al., 2025).

In the WSPC database, a total of 26 palaeoecological proxies has been applied to peat cores from WSL; however, their use varies significantly, both in quantitative frequency and in spatial distribution across the study area (Figs. 6b, 7, B1).

Chronological proxies are essential for reconstructing peatland development over time and were applied in 58.0 % of all peat cores. Radiocarbon dating is by far the most frequently used method in the region, enabling age-depth modelling and temporal correlation of palaeoecological data (Chambers and Charman, 2004). The resolution of dating varies considerably between cores: some cores were dated only at the basal level, whereas others include several dated horizons. This variability often reflects financial constraints, as multipoint dating remains relatively expensive. Nevertheless, in recent decades researchers have increasingly adopted multi-level dating within individual cores to achieve more robust reconstructions and to facilitate comparison with paleoclimate data (e.g. Feurdean et al., 2019; Lamentowicz et al., 2015; Leonova et al., 2022; Novenko et al., 2023). Due to its widespread application, the spatial distribution of this proxy is relatively uniform, and it is represented in peat cores across most of WSL (Fig. B1a).

Physical proxies, which provide information on peat properties and stratigraphy, are frequently applied in the peat cores included in the database in the database. Basal depth is by far the most common proxy, present in 84.3 % of cores (Fig. 6b). Such a large number of basal depth measurements is primarily due to the fact that this parameter is relatively easy to obtain in the field and usually does not require specialist expertise (Parry et al., 2014). Many basal depth data were collected during geological prospecting campaigns or as part of detailed investigations of the development of individual peatland complexes (Kremenetski et al., 2003; Lapshina and Zarov, 2023). This widespread use reflects both its fundamental importance for peatland development studies and its relevance as an economically significant parameter (Fig. B1b). Another important physical proxy useful e.g. for carbon content calculations is bulk density, reported in 24.6 % of cores, mostly from the central part between the Ob and Yenisey rivers and from the Vasyugan Plain (Fig. B1c). Degree of decomposition and peat humification are both analyses that indicate the intensity of decay processes; however, the methods differ (Biester et al., 2014). Peat humification is an older technique and is rarely used (1.5 % of cores), whereas the degree of decomposition, usually assessed via the von Post scale – a traditional Russian field method – is much more common in WSL cores (17.4 %). Their occurrence within the study area is characterized by a clear concentration in the south, on the Vasyugan Plain, while in the remaining regions the proxy is represented only in a small number of sites (Fig. B1d, g). Moisture or water content, a proxy useful for interpreting decomposition processes or nutrient mobility (Rydin et al., 2006), was applied in 16.5 % of all cores, and its spatial distribution is similar to that of bulk density (Fig. B1e). Peat accumulation is an important proxy that not only provides information about peatland development but can also indirectly inform about past climate variability (Swindles et al., 2025). It was used in 13.3 % of the cores, which are distributed across the study region, with more sites located on the Vasyugan Plain (Fig. B1f). The form in which this information is reported varies depending on analytical resolution, sometimes it is a single value for the entire core, and its accuracy increases with increasing resolution. Nevertheless, accumulation rates in permafrost-affected peatlands should be interpreted with caution, as cryogenic structures and variable ice content may substantially influence peat accumulation and preservation processes (Treat et al., 2016). However, such information was not consistently reported in the analysed studies and therefore could not be systematically included in the database.

Chemical properties of accumulated peat in WSL have also been frequently analysed in peat cores. The most commonly used chemical proxy is carbon accumulation, applied in 34.3 % of all studied cores (Fig. 6b). This proxy provides valuable information about the total amount of stored carbon in the peat and about long-term changes in peatland carbon sink strength (Charman et al., 2015). WSL is recognized as a global carbon hotspot; therefore, many studies focus on carbon budgets and accumulation patterns, which explains the widespread use of this proxy. This proxy is predominantly reported in sites from the central part of the study region and the peatlands of the Vasyugan Plain; however, cores containing this information are also frequently found in river valleys and on the Yamal Peninsula (Fig. 7a). Ash content and organic matter, typically measured via loss-on-ignition (LOI) analysis, are also commonly analysed in WSL peat cores, appearing in 24.2 % and 23.9 % of peat cores, respectively. These simple and low-cost methods are widely used in paleoecology and provide information about peat composition and the input of inorganic material (Chambers and Charman, 2004; Tolonen, 1984). Their spatial distribution is very similar and resembles the distribution pattern of bulk density (Figs. 7b, B1t). Other chemical proxies included in this database: geochemistry, pH, stable isotopes, XRF or XRD and FTIR have been rarely applied in WSL peat cores. Data on pH, which reflects peat acidity, are more characteristic of modern ecological surveys than of palaeoecological studies. Consequently, this proxy was used in only 1.1 % of all cores, exclusively in non-permafrost peatlands in the southern part of the region (Fig. B1w). Traditional geochemical analyses (2.4 %), stable isotopes (0.9 %), XRF/XRD (0.9 %), and FTIR (0.2 %) all require specialised expertise and laboratory infrastructure, which until recently were not widely accessible. Geochemical analyses and XRF/XRD are different methods for determining the elemental composition of peat. Geochemistry has become increasingly popular in recent years, whereas XRF and XRD, traditionally used in lake sediment studies, have only recently begun to be applied to peat (Longman et al., 2019; Shotyk, 1988). Geochemical analyses have been conducted in several peatlands south of 64° N, while XRF/XRD has been used primarily in peatlands near Tomsk in the south-eastern part of the study area (Figs. 7f, B1v). Stable isotopes in peat provide valuable insights into past environmental and climatic conditions; however, they are costly and require specialized isotope facilities (McClymont et al., 2010). Such data are scattered across WSL and are typically collected from cores analysed in extensive studies that integrate multiple other proxies (Fig. 7e) (Feurdean et al., 2019; Startsev et al., 2022).

The group of palaeoecological proxies that has been the least frequently applied in WSL peat cores are biological proxies (Fig. 6b). They rely on the identification of remains of organisms preserved in peat and are therefore highly dependent on preservation conditions. Additionally, they are usually time-consuming and require specialist expertise. The most frequently used biological proxy, applied in 21.4 % of all cores, is plant macrofossils (referred to as “botanical composition” in older publications), which reflect local peatland vegetation and are essential for reconstructing peatland development and hydrology (Birks and Birks, 2000). However, the level of analytical detail varies between studies. In some cores, particularly in older works, the data are limited to reporting dominant Sphagnum species or other major plant groups, often at a low sampling resolution of 10 or 25 cm. In contrast, more recent studies tend to use higher-resolution sampling and more detailed identifications, frequently including subfossil vegetative fragments and carpological remains (e.g. Halaś et al., 2025; Novenko et al., 2022). Peat cores with plant macrofossil data are spread across WSL, with a higher concentration on Vasyugan Plain and in the area near Tomsk (Fig. B1h). Pollen is an excellent proxy for reconstructing past regional vegetation and landscape-scale ecological changes (Mander and Punyasena, 2018). Owing to its long-established use in palaeoecological studies, it was introduced very early in WSL research. To date, 16.1 % of all cores in the database contain pollen data. These cores cover the majority of WSL, with their highest frequency occurring in areas where core density is greatest (Fig. 7c). Non-pollen palynomorphs (NPPs) are additional microfossils that can be identified during standard palynological analysis (Shumilovskikh et al., 2021). Because Quaternary palynologists historically did not focus on these other palynomorphs, their identification is challenging, and their systematic documentation only began in the late 1970s, their use in WSL cores has remained limited. Although the scientific community has increasingly recognized the potential of this proxy to reflect local ecological conditions and it is now being applied at more sites (e.g. Feurdean et al., 2022; Halaś et al., 2025), only 1.1 % of all cores currently include NPP data, and these come from only a few locations in the WSL (Fig. B1m). Peatlands are also valuable archives of past fire activity, with charcoal commonly used as a proxy for this purpose. Depending on the intended spatial scale of fire reconstruction, different charcoal size fractions are analysed: macrocharcoal for local fires and microcharcoal for regional fire activity (Conedera et al., 2009). Both types are present in the WSL, although at low frequency. Macrocharcoal has been used more often, occurring in 4.0 % of all cores, whereas microcharcoal appears in only 1.5 %. There is no clear spatial pattern in their use; instead, they are found sporadically in isolated cores across the study area (Fig. B1j, l). A lack of fire-history research in WSL (and in Russia more broadly) was also noted by Pupysheva and Blyakharchuk (2023). They emphasized that fire proxies are relatively new to this region and highlighted the importance of expanding fire-history studies across this territory. Hydrology is one of the most important environmental variables controlling the condition of peatland (Rydin et al., 2013), and testate amoebae are a well-established proxy for reconstructing past hydrology of these ecosystems (Charman, 2001). Although their widespread use in palaeoecological research began roughly 40 years ago (Mitchell, 2025), their application in WSL peat cores remains very limited, appearing in only about 2.6 % of all studied cores. Testate amoebae have been used primarily in non-permafrost peatlands in the southern part of the WSL, while in other permafrost zones they have been applied only sporadically (Fig. 7d). Less frequently used biological proxies include arthropods, molluscs, oribatid mites, diatoms, and zoogenic remains, each recorded in less than 0.3 % of cores. Although these proxies can offer valuable insights into local environmental conditions, their preservation in peat is generally poor, and they are more characteristic of lake sediment studies. As a result, they are rarely applied in peatland research. This pattern is also evident in WSL, where their use in peat cores is limited to only a few isolated cases (Fig. B1n–r).

Overall, the WSPC database demonstrates a strong focus on fundamental physical and chemical proxies, particularly basal depth, carbon accumulation, and organic matter content, reflecting the importance of these variables in peatland palaeoecology. Biological proxies are applied more selectively, often depending on the research question, while more specialized proxies such as geochemistry, stable isotopes, or FTIR are rarely used. This distribution highlights both the breadth of multi-proxy approaches in WSL peatlands and the gaps in underrepresented proxy types, particularly in biological and specialized chemical analyses.

The WSPC database includes three types of peat cores: (1) full-length sequences that extend from the basal mineral substrate to the sampling date and thus capture the complete peat accumulation history; (2) upper-only sequences that begin at the modern peat surface but do not reach the basal layer, typically resulting from targeted sampling strategies such as upper-core ecological reconstructions or from logistical constraints, particularly in permafrost regions; and (3) cores that may or may not reach the basal mineral substrate but are distinguished by the absence of the youngest peat layer due to climate-driven cessation of peat accumulation or degradation of previously accumulated peat, a pattern most commonly observed in polygonal peatlands in the northern part of the region. The temporal distribution of peat core ages, summarised in 300-year intervals, shows a highly uneven pattern, revealing several distinct phases across the time span up to ∼ 16 000 cal. yr BP (Fig. 9b). Since the majority of cores in the WSCP database were published after 1950 and type-3 cores are relatively rare, we assumed for this analysis that each peat core represents continuous data from 1950 to the deepest dated layer.

Peat cores covering only the near-modern period (0–2000 cal. yr BP) are quite rare in the database, usually less than 5 cores per 300-year interval (Fig. 9b). There is an increase in the number of cores that cover longer periods of the Late Holocene up to 4000 cal. yr BP, with 6–12 cores per interval. Cores covering the Middle Holocene (∼ 4000–8000 cal. yr BP) form the second most frequent group, averaging 10 cores per interval. The highest representation is observed for cores spanning ∼ 8000–11 000 cal. yr BP, with nearly half of all chronologically established cores reaching this period. Beyond ∼ 11 500 cal. yr BP, the number of cores declines sharply, with only individual cores reaching into the Late Glacial. The oldest peat core in the WSPC database dates to 15 623 cal. yr BP (core WSL_246) and is located near the Yenisey River within a present-day non-permafrost region (Fig. 9a). Peat cores older than 11 500 cal. yr BP are exclusively situated south of 62° N, where current ground conditions correspond to non-permafrost or isolated permafrost zones, and all of these oldest records are located within 100 km of major river channels. The lack of records from this period is consistent with limited peatland development in WSL during the Late Glacial, when conditions were generally unfavourable for peat accumulation (Alexandrov et al., 2016; Velichko et al., 2011). In terms of coverage, ∼ 80 % of cores extend over the last 4000 cal. yr BP, ∼ 50 % reach 8000 cal. yr BP, and only ∼ 5 % cover 11 000 cal. yr BP.

Because the database includes both basal-depth and upper-only cores, these data cannot be interpreted as direct records of peatland initiation and serve as a basis for analysis spatial peatland development in WSL. Instead, it reflects the age extent of available peat profiles. Despite this limitation, the broad temporal pattern – many cores spanning the early Holocene and very few extending into the Late Glacial – broadly aligns with regional peatland development phases described for this region (Kremenetski et al., 2003).

Temporal coverage of selected proxies varies across permafrost zones, reflecting differences in peat core density, maximum peat age in each zone, and the overall frequency of proxy application (Fig. 8b). The longest peat records occur in isolated and non-permafrost zones, where the majority of studied proxies have been applied, and where most of them provide the greatest temporal coverage, with some extending to almost 15 000 cal. yr BP. In other permafrost zones less proxies have been used, in sporadic and discontinuous permafrost zones proxies' coverage span up to 11 500 cal. yr BP, whereas in continuous permafrost zone, maximum temporal coverage reaches ∼ 11 000 cal. yr BP. Temporal coverage of proxies also depends on the number of cores available for each time interval, which directly affects the completeness and spatial representativeness of the data. When only a few cores provide information for a given variable, such as peat accumulation, the insights are limited and largely local, whereas larger numbers of cores allow for more comprehensive and regionally representative interpretations.

The peat cores from isolated and non-permafrost zone provide the earliest and most continuous proxy coverage of all three regions, with several proxies extend into the Late Glacial, providing broader and more detailed insights into the regional peatland history (Fig. 8b). Pollen data for this region are present from ∼ 15 000 cal. yr BP, making it the oldest palaeoecological proxy across all regions and providing a continuous record of vegetation composition and broader regional ecological transitions from the Late Glacial through the Holocene. Although the earliest period is represented by only a few sites, the number of cores grows rapidly after 11 000 cal. yr BP, reaching more than 60 cores in the recent period. Thus, this region offers the highest resolution, most spatially extensive vegetation reconstructions from peatlands in WSL. Peat-property proxies, including organic matter and bulk density, as well as peat and carbon accumulation extend to ∼ 14 000 and 12 000 cal. yr BP, earlier than in any other region. Their steady increase to dozens of cores through the Holocene offers robust reconstruction of substrate evolution, decomposition dynamics, carbon accumulation rates, and peat accumulation phases. Plant macrofossils data are available until ∼ 12 000 cal. yr BP and reach over 60 cores in the Late Holocene, enabling high-resolution reconstructions of local plant assemblages, peatland microtopography, and fen-bog transitions. Hydrological reconstructions through testate amoebae extend to ∼ 11 000 cal. yr BP and increase to 14 cores, supporting long-term and spatially extensive water-table reconstructions. Geochemical and XRF/XRD records are present from ∼ 10 000–8500 cal. yr BP, capturing changes in mineral influx, ash content, and geochemical environment. These datasets are largely only available in this part of WSL; although representation increases through time, they still allow only spatially restricted interpretations. Fire proxies vary in temporal coverage. Macrocharcoal is recorded from ∼ 12 500 cal. yr BP, with moderate representation after 8500 cal. yr BP, providing the deepest temporal perspective on local fire history across WSL. Microcharcoal, which can inform on regional fire dynamics, is traceable to ∼ 8500 cal. yr BP but remains very sparse.

The temporal coverage in the discontinuous and sporadic permafrost zones differs markedly from the previous region. Here, a few proxies dominate, providing long-term data across the Holocene, while others are absent or offer temporally and spatially restricted information (Fig. 8b). Organic matter, carbon accumulation, and bulk density data are available until ∼ 11 500 cal. yr BP, and their representation increases sharply through the Early Holocene, remaining exceptionally high (> 60 cores) throughout the Mid- and Late Holocene. This suggests that this zone provides the strongest quantitative foundation for Holocene carbon-accumulation reconstructions among all permafrost regions. In contrast, peat accumulation data extend only to ∼ 9500 cal. yr BP and are available from a very limited number of cores. While these data add a quantitative dimension to vegetation productivity, low representation restricts their use to localized reconstructions rather than regional-scale trends. Pollen data, similar to bulk density, are present from ∼ 11 500 cal. yr BP but remain sparse until the Mid-Holocene. Only from ∼ 7000 cal. yr BP onward does pollen coverage expand to 7–9 cores per interval, allowing spatially broader vegetation reconstructions. Plant macrofossils and microcharcoal data extend to ∼ 8500 cal. yr BP but remain rare, providing long-term yet localized records of plant communities, peatland-surface conditions, and regional fire history. Geochemical information obtained from traditional geochemical analyses is limited to a single site but extends back to ∼ 6000 cal. yr BP, whereas XRF/XRD data are unavailable for peatlands located in these permafrost zones. Hydrological records through testate amoebae are largely absent, appearing only within the last 500 years, which restricts water-table reconstructions to the very recent past. A similar pattern is observed in macrocharcoal data; information on local fire activity, in contrast to the southern region of WSL, is almost entirely absent for peatlands in this region.

Proxy data in the continuous permafrost zone are available up to ∼ 11 000 cal. yr BP, with pollen providing the longest and most continuous environmental record (Fig. 8b). Most other proxies are present until ∼ 10 500–10 000 cal. yr BP, and in all cases, representation peaks around 6500 cal. yr BP, though it remains relatively low compared to the other regions. These datasets allow reconstruction of peatland vegetation changes, substrate development, and peat and carbon accumulation dynamics across almost the entire Holocene; however, spatial coverage is limited due to the low number of cores. Hydrological information from testate-amoebae extends to ∼ 7000 cal. yr BP but is available from only a single site, making water table reconstructions largely site-specific. Fire proxies (micro- and macrocharcoal) are present through the Holocene but remain sparse, restricting fire history interpretations to localized events and preventing robust regional-scale reconstructions. Geochemical information, derived from traditional analyses or XRF/XRD, is absent for peatlands in the continuous permafrost zone.

In summary, there is a clear regional variation in the temporal coverage of selected proxies across different permafrost zones. The longest and most continuous records are found in isolated and non-permafrost zones, reflecting both the early initiation of peatlands in this region and the relatively easy accessibility for research due to the absence or spare presence of permafrost. In discontinuous and sporadic permafrost zones, only a few proxies provide long-term, spatially extensive coverage, primarily for basic peat and carbon properties, while other biological and geochemical proxies remain sparse. In continuous permafrost zones, many proxies extend into the early Holocene, but the low number of cores limits spatial representativeness, meaning that long-term reconstructions of ecological dynamics are largely site-specific.