Natural and anthropogenic factors influencing changes in peatland management in Poland

The main aim of this study is to identify trends in peatland management from the end of the nineteenth century to late twentieth century in the Tuchola Pinewoods (TP) located in NW Poland (young glacial zone, temperate climate). The analyses were performed based on the 1:100,000 Prussian topographic maps from the years 1876–1879 and the 1:50,000 Polish topographic maps from the years 1966–1986. A total of 744 peatlands were identified in TP (total area — 10,762 ha, 3.03% of the studied region). Smaller peatlands of up to 1–2 ha were found to be most numerous (15.32% of total number), whereas those exceeding 15 ha cover the largest area (75.42% of total area). The analysis revealed a tendency regarding land management of peatlands in the study period — large peatlands were adapted for agricultural purposes, and the smaller ones were mostly transformed into woodlands. The most important factors influencing the directions of changes in the use of peatlands include as follows: the type of use of the adjacent areas, implementation of anthropogenic drainage systems and location in a lake catchment or a catchment drained by rivers or streams. Considering that 85.7% of total peatland area had been drained by 1986, the risk arising from greenhouse gas emission from peatlands should be regarded as high.


Introduction
Wetlands, including peatlands, are some of the most valuable natural ecosystems on account of their particular communities of fauna and flora, which include many protected species (Lowe and Walker 1997;Ilnicki 2002;Joosten and Clarke 2002). Their unique role in the environment: ecological (Joosten and Clarke 2002;Thorslund et al. 2017), hydrological (Šimanauskienė et al. 2019;Ghajarnia et al. 2020), biogeochemical: (Rydin et al. 2013;Grzybowski and Glińska-Lewczuk 2020) has been emphasised, inter alia, by the Ramsar Convention of 1975(UNESCO 1982Dinesen et al. 2021), which proposed several methods/activities to protect these unique habitats. Peatlands, where biogenic accumulation has been ongoing for millennia, act as a natural archive of the environment and as such provide ample information on climate change (Barber et al. 2003;Heijmans et al. 2008;Booth et al. 2010;Gałka et al. 2014), vegetation cover (Feurdean et al. 2014;Krąpiec et al. 2016;Grzybowski and Glińska-Lewczuk 2020) and human activity (Lamentowicz et al. 2008(Lamentowicz et al. , 2019aGlina et al. 2019a;Słowiński et al. 2019).
The main groups of factors responsible for the transformation of peatlands highlighted in literature are global climate change (Ojala and Louekari 2002;Barber et al. 2003;Charman et al. 2013;Lamentowicz et al. 2019a;Lunt et al. 2019) and human impacts: melioration (Glina et al. 2017;Kiryluk 2020), development of settlements and agriculture (Ojala and Louekari 2002;Page and Baird 2016;Glina et al. 2017), changes in land use (Lamentowicz et al. 2007;Glina et al. 2019a), drop in water level (Glina et al. 2019b(Glina et al. , 2022. As far as transformation of the existing peatlands is concerned, natural determinants of this process include global climate change (Joosten and Clarke 2002;Harenda et al. 2018), topoclimatic changes (Kucharzyk and Szary 2012) and lowering of the water Communicated by Jasper van Vliet.  (Ilnicki 2002;Joosten and Clarke 2002;Gumbricht et al. 2017).
Before the Anthropocene (Waters et al. 2016), climatic conditions had been the primary factor affecting the water balance of peatlands. They determined the potential for peat-forming vegetation communities to occur, and were also a factor limiting mineralisation of organic matter, which allows for peat layers to accumulate over time (Tong et al. 2014;Laamrani et al. 2020). Since the second half of the twentieth century, peatlands/wetlands have been under strong stress factor associated with climate warming (Clark et al. 2005;Dise 2009;Davidson 2014;Hu et al. 2017;Sirin et al. 2020). Climate-driven changes in peatlands have coincided with those induced by humans (inter aliaOjala and Louekari 2002; Lamentowicz et al. 2007Lamentowicz et al. , 2009Booth et al. 2010). The anthropogenic degradation of peatlands and their habitats stems from direct transformation of the peatlands themselves: drainage, peat harvesting, changes in land use (Joosten and Clarke 2002;Clark et al. 2005;Dawson et al. 2010;Adesiji et al. 2015;Urák et al. 2017;Šimanauskienė et al. 2019), as well as from transformation of their catchment areas -especially over the last few hundred years (de Jong et al. 2010;Gałka et al. 2014;Lamentowicz et al. 2015Lamentowicz et al. , 2019aDyderski et al. 2016;Mieczan et al. 2018).
The aim of the presented research is to analyse the human and environmental impact (including climate) on changes in the peatland use between the end of the nineteenth century and the end of the twentieth century in selected region of north-western Poland (Tuchola Pinewoods). The peatlands found in the research area have been relatively well investigated, but mainly as part of paleogeographic and paleobotanical studies (Tobolski 2006(Tobolski , 2007Grover and Baldock 2013;Błaszkiewicz et al. 2015;Słowiński et al. 2015Słowiński et al. , 2016, and the efforts focused on individual objects or small clusters of peatland areas (Żurek 1987;Tobolski 2003;Cedro and Lamentowicz 2011;Gałka et al. 2014;Słowiński et al. 2015;Kołaczek et al. 2015). There is a lack of regional analyses involving larger datasets of peatlands that would address the issue of peatland transformation.
The main hypothesis tested in this study is that human pressure constitutes a crucial factor shaping changes in peatland use in the last 150 years in the Tuchola Pinewoods, a region located in the young glacial geomorphological zone of north central Europe. To verify this, spatial variability pertaining to changes in peatland management was analysed in relation to several drivers of these changes. Research encompassing the entire region (as opposed to individual peatlands) may significantly improve our understanding of peat-forming processes in recent centuries, i.e. a period of increasing anthropopression. Furthermore, comprehensive understanding of the past and present landscape layout is essential for the effective conservation and restoration of these vulnerable ecosystems (Klimkowska et al. 2010) at regional and global scales in view of threats arising from the observed and predicted climate warming.

Study area
The Tuchola Pinewoods mesoregion (A = 3550 km 2 , share of forests = 70.3%) (Fig. 1) is part of the West Pomeranian Lakeland (north-western Poland) and represents young glacial geomorphological zone, typical of northern Europe (Giętkowski 2008(Giętkowski , 2009Łuców et al. 2021). It is located within two peatlands zones: 'raised bogs and fens' in the northern and central part of the study area and 'fens' in the southern part (Fig. 1). The analysed area covers parts of two catchments (Brda and Wda), with a hydrographic network layout connected with glacial and postglacial (Holocene) land transformation (Błaszkiewicz 2010;Lamentowicz et al. 2015;Błaszkiewicz et al. 2015;Słowiński et al. 2015) (Suppl. Mat., Fig. S1). The young glacier landscape of the study area is characterised by high share of lakes and peatlands, especially in the north, west and south-east of the region. Major peatland drainage works in the region were carried out until the mid-twentieth century (Szumińska 2014). The first works were conducted in the nineteenth century by the Prussian government (Poland was under partitions in 1772(Poland was under partitions in -1913 and were continued until the 1940s. However, up until the Second World War, around 60% of all drainage works encompassed the 1st-order ditch only (Szumińska and Absalon 2012). After the Second World War, drainage has continued, and in the 1980s, the main works related to the construction of the 2nd-order ditches were completed. In 2010, a biosphere reserve was established within the study area, and almost the entirety of the area is currently protected under Natura 2000. Furthermore, the Tuchola Pinewoods encompass one national park, five landscape parks and numerous nature reserves and protected landscape areas (Suppl. Mat., Fig. S2) (The General Directorate for Environmental Protection of Poland (GDOŚ) 2016; Łuców et al. 2021).
The region is located in the temperate climate zone, with characteristics of a transitional oceanic-to-continental climate (Kottek et al. 2006). Mean annual precipitation (MAP) in the years 1965-2003 increased from 575-600 mm in the south east to 800-825 mm in the north west (Szumińska 2014). Mean annual air temperatures (MAAT) in 1951-70 increased from 6.5 °C in the north of the Tuchola Pinewoods to 6.9 °C in the south (Wójcik and Marciniak 1987). MAAT in the period of 1952-2018 calculated for the Chojnice station, which is located in the central part of the study area, was higher and amounted to 7.4 °C (Okoniewska and Szumińska 2020).

Spatial analysis
Changes in peatland areas were analysed on the basis of 1:100,000 Prussian maps -Karte des Deutschen Reiches from the years 1876-1879 (11 map sheets) and 1:50,000 Polish topographic maps showing the state of land cover in the years 1966-1986 1 (22 map sheets) ( Table 1). The study period demonstrating changes is approximately 100 years. Based on the Prussian maps, the boundaries of peatlands were delimited. Due to the scale of the analysed maps, the analysis accounted for features larger than 0.10 ha.
The boundaries of peatlands were initially delimited using Karte des Deutschen Reiches, which distinguishes contours described as bruch, morr, moss (Ger. bruchwetland, marsh, quagmire; Ger. morr, moss -peat (Ilnicki 2002). In this study, we assumed that features with these designations included areas that were active peatlands at the time of the mapping surveys for the 1879 editions. Areas of wetlands in which the peat-forming processes that accumulate organic deposits were not active were marked on these maps using a separate contournasse wiese (wet meadows). There is a possibility that some of the wet meadows were established on previously drained peatlands. However, this study aims to analyse peatlands that in 1879 were natural or semi-natural, and that had not been subject to significant changes in land use.
The maps were rectified and digitised in ArcGIS 9.0, and the results were then double-checked to eliminate  Partial change into water bodies Partially changed areas; rise in groundwater level (waterlogging) 7 Complete change into water bodies Complete change into water bodies errors. Information on land amelioration and land use in each individual wetland and in its surroundings was entered into the database. Because peatlands lack clear boundaries on late twentieth-century Polish maps and because of possible differences in base maps resulting from technological and methodological differences in how they were prepared, the qualitative data in the vector layer attribute table was updated. Such a method was proposed by Urbański (2011). A new geodatabase was not elaborated; rather, the existing database was simply augmented with data on the directions of changes in peatlands and their surroundings. The hierarchy of changes presented in Table 1 was adopted for the analysis. The database also contains information on peat extraction and the extent of peatland drainage. Peatlands contours were included into the grid analysis (1 × 1 km) of peatlands percentage in the Tuchola Pinewoods region. The grid calculations were carried out with regard to the peatland areas delimitated in 1879. Furthermore, grids were calculated for the peatlands demonstrating no distinct change, complete change and partial change in 1986.
In addition to the data pertaining to peatlands, data on selected features of the environment were collected in a GIS database ( Table 2). The environmental variables were selected taking into account the criterion of availability of data for the entire study area and the criterion of potential effect of a given factor on changes within peatlands. Additionally, the distance of a peatland from the nearest human settlement was also taken into account, due to the potentially greater risk of peatlands being converted into meadows. Another factor that was taken into account was distance from groundwater intake station. Intensive groundwater intake in the selected sites of the study area caused the development of a cone of depression within ground water aquifer (Szumińska 2014). This can potentially affect water table stability within peatlands. Among the natural factors, the permeability of surface geological formations was taken into account (permeability classes from 0 to 4) ( Table 1). Permeability was assigned on the basis of the particular permeability class that accounted for the largest area of a given peatland. It should be noted, however, that not all peatlands marked on the 1:100,000 topographic map of 1879 were shown as peat deposits on the 1:50,000 geological map. This may have been caused by the difference in scale. At the same time, there were cases in which there was no active peatland contour on the 1879 map, while there was a peat deposit on the geological map, which suggests a lack of an active peat-forming process in this area in 1879. Because precipitation in the study area increases significantly from south-east to northwest, the authors elected to account for the mean annual sum of atmospheric precipitation within individual peatlands as an environmental variable.

Statistical methods
Detrended correspondence analysis (DCA) was used to examine the structure of the data set. The structure was revealed to be linear, and consequently, as recommended in such cases (Jongman et al. 1987;ter Braak and Šmilauer 2002), the relationship between changes in peatland use and selected control variables was analysed by means of redundancy analysis (RDA) (ter Braak and Šmilauer 2002). The method uses regression procedures to examine relations between variables. Additionally, RDA enables visual representation of said relations in ordination diagrams, which are considered reader-friendly. Hence, the method is commonly employed in environmental studies (e.g. Bossolani et al. 2020;Sewerniak and Puchałka 2020;Raduła et al. 2022). RDA was conducted in the CANOCO package v. 4.5 (ter Braak and Šmilauer 2002). The change in peatland use was treated in CANOCO calculations as 'species data', and described on a three-level scale (0 -no change, 1partial change, 2 -complete change) with regard to the studied peatland use change (into woodlands, into agricultural lands and into water bodies; Table 1). In the ordination analysis, on the other hand, all examined explanatory factors (related to either environmental or anthropogenic agents) were treated as 'environmental' CANOCO variables. Subsequently, in the CANOCO package, forward selection and a Monte Carlo permutation test were used to assess the statistical significance of particular factors for the change of peatland use in the studied period (ter Braak and Šmilauer 2002). The Monte Carlo permutation test is a test of statistical significance obtained by repeatedly shuffling (permuting) the samples. The basis of the test lies in the observation that under the null hypothesis the 'species data' can be randomly linked with the samples in the 'environmental data'. Each permutation in the Monte Carlo procedure leads to a new data set, from which the test statistic can be calculated (F 1 , F 2 …, F K ). Finally, the Monte Carlo significance level is obtained from the place of F 0 among F 1 , F 2 …, F K ; from the proportion of values greater than or equal to F 0 . Thus, the Monte Carlo significance level is the rank of F 0 among all values F 0 , F 1 , F 2 …, F K divided by K + 1 (ter Braak and Šmilauer 2002).

Peatland distribution in the Tuchola Pinewoods
The analyses resulted in a map of density of peatland distribution in the Tuchola Pinewoods in 1879 ( Fig. 2A). A total of 744 peatlands were identified in the Tuchola Pinewoods in the late nineteenth century, with a total area of 10,762 ha, or 3.03% of the study area. The smallest marked peatland covered 0.12 ha, while the largest was 792.8 ha.
There were 595 items of less than 15 ha, combined area of which amounted to 2645 ha. Only 149 peatlands were found to exceed 15 ha; however, they encompassed an area of 8117 ha ( Fig. 3; Suppl. Mat., Table S1, Table S2). The most numerous were peatlands of 1 to 2 ha (114 items) and those of less than 1 ha (91 items) ( Fig. 4; Suppl. Mat., Table S2). The group of peatlands larger than 15 ha shows the prevalence of objects in the class of 15-50 ha (Fig. 3). However, despite the low number (14) of very large peatlands (> 100 ha), their total area is larger than in the case of peatlands representing any other size class ( Fig. 3; Suppl. Mat., Table S1). Peatlands are not regularly distributed, but rather occur in patches, which may be attributed to the differences in lithology (Suppl. Mat., Fig. S1) and hydrological conditions (Suppl. Mat., Fig. S2). The high share of peatlands is visible in the watersheds zones of the Brda, Wda and Mątawa catchments (Fig. 1A). Furthermore, peatlands are well represented in the vicinity of the main lakes in the studied region. The areas with a high peatland percentage are characterised by complex lithological conditions with the mosaic of sediments of glacial, outwash plain, fluvial and lake origin (Suppl. Mat., Fig. S1). One should note that the selected groups of peatland patches are located outside of landscape or national parks, inter alia peatlands located in the watershed of the Wda and Mątawa catchments, and the larger peatland in the Niechwaszcz river valley (Suppl. Mat., Fig. S2). In the case of the latter peatland complexes, they are located in a forested area as well as in agriculture zones.

Changes in the peatland management
The analysis reveals that only 37 objects (5.0% of the total number of peatlands studied) showed no indication of changes in use, whereas the majority of peatlands had been transformed into agricultural lands and woodlands ( Fig. 5A and Table 3). In the case of agricultural land, these changes included 53.5% of the total number of items, amounting to 73.4% of the total area of all studied peatlands (partial and complete changes combined). Of the total number of peatlands, 38.8% (23.5% by area) have been changed into woodland (partial and complete changes combined). Analysis of changes in the use of the peatland surroundings in the Tuchola Pinewoods shows that the change in land use affected around 342 objects (of a total of 744). Peatlands surroundings were mostly turned into agricultural lands and woodlands ( Fig. 5B and 6). Such changes were observed in the case of 27 and 315 peatlands, respectively. The conducted analyses indicate that a shift in land use in the surroundings of peatlands triggers a change in how the peatlands themselves are used (Fig. 5A, B). However, there are peatlands whose surroundings have not changed (402 peatlands), which are located across the entire study area (Fig. 5B), and most of these sites are located in non-woodland areas (Suppl. Mat., Fig. S2).
The summary of directions of peatland transformations by size class (Fig. 7) shows that peatlands exceeding 15 ha were mostly changed into agricultural land. By contrast, smaller peatlands were more often transformed into woodlands. There is an exception to this, however, as 218 small peatlands (< 15 ha) were completely converted to agricultural land. These items are mainly small meadows adjacent to or within woodlands. However, peatlands that were changed to agricultural land represent the largest mean area (Suppl. Mat., Table S3). The largest items in this group are located in the north of the studied region (the northern Wda basin, and the Wieckie, Smolińskie and Kruszyńskie lake basins) and in the south of the Tuchola Pinewoods (the Kałębie Lake basin and the immediate surroundings of the Wda and Mątawa rivers) (Fig. 5A). The largest peatland (792 ha) is located in the Niechwaszcz river valley.
The number of peatlands that became wooded (completely or partially) amounted to 289. In this group, the number of drained peatlands increased from 44 to 103 between 1879 and 1986 (Fig. 8). Peatlands that were completely wooded numbered 201 (Table 3). They are located mainly in the south-east and north of the study area (Fig. 5A) and are considered small (Suppl. Mat., Table S3). The 88 items that were only partially wooded ( Fig. 7 and Table 3) are scattered throughout the Tuchola Pinewoods, but their largest concentration is found in the south-east of the study area (Fig. 5A). Many of the peatlands in this group are large. Peatlands that were changed into agriculture lands numbered 398 and are found all across the study area (Fig. 5A). At the same time, in the north and south of the Tuchola Pinewoods, there is a concentration of peatlands that were completely converted into agricultural land (275 peatlands). The largest peatland, located in the Niechwaszcz catchment area, was partially changed to agricultural land. Many areas that were partially transformed to agricultural land also show a shift in land use in the surrounding area -mainly from agricultural land to woodland (Fig. 6). In 1879, as many as 126 objects repurposed for agriculture had been drained, and the number increased to 265 by 1986 (Fig. 8). An interesting direction of changes involves the transformation of peatlands into water bodies (Fig. 5A), which pertains to 20 sites with a total area of 57.9 ha (partial and complete change, combined). These peatlands were affected by a rise in water table, effectively causing them to become waterlogged. They are found in the north of the study area and are relatively small. Of this group, three peatlands had been drained by 1879, and drainage systems were still visible at one site in 1986.
As shown in the analysis of the number and area of drained features (Fig. 8), the scope of land amelioration in the Tuchola Pinewoods in the late nineteenth century was certainly lesser than in the late twentieth century. In the years 1879-1986, the number of ameliorated peatlands increased from 182 to 381 and the total area from   Table S4). Vast majority of these were drainage works. Ameliorations mainly concerned peatlands changed into agricultural land, but amelioration works were also conducted in peatlands that were partially or completely transformed into woodland areas (Fig. 8). The main subjects of land amelioration were large peatlands of economic importance because, once drained, they could be used for agriculture or as additional acreage for woodland cultivation.

Factors affecting changes in peatland management
The Monte Carlo permutation test calculated for the datasets indicated that the followed variables related to land use is the surrounding areas were statistically significant for the change of peatland use in the examined period: S: woodlands 1879 and S: woodlands 1986, S: wood/agr. 1986, S: agriculture 1986 (Table 4). Furthermore, peatland drainages in 1879 and 1986 (Table 4) were also found to be a significant factor. Redundancy analysis (Fig. 9) shows that factor 1, being related mainly to the changes occurring in peatland surroundings (afforestation vs. transformation into agricultural land) as well as to the occurrence of drainage, explains 37% of total variance in the peatland dataset. In turn, factor 2 was found to be irrelevant to the variability studied. The conducted ordination analysis also shows that the change of peatlands into woodlands (negative correlation with factor 1) was prevalent in peatlands not connected to the surface runoff system (endorheic basin), and in the surroundings of peatlands used for forestry both in 1879 and at the end of the twentieth century (S: woodlands). Moreover, the transformation of peatlands into woodland was also favoured by remoteness from a population centre (dist. to settlement, Fig. 9). In turn, the location of peatlands in a drainage basin or an endorheic lake basin favoured the occurrence of peatlands converted into agriculture land. This change was also influenced by land reclamation works carried out in the twentieth century and by agricultural use of the surroundings of peatlands in 1879 and 1986 (S: agriculture, Fig. 9). Redundancy analysis also shows that the peatland area correlates positively with the drainage works conducted in 1879 (Fig. 9), which in turn indicates that the size of peatland was important for the decision of drainage in the nineteenth century, and peatlands of large or medium area were drained in that period. Moreover, the change of peatlands into water bodies is also positively correlated with factor 2. Although the relation is weak (Fig. 9), it can be associated with the flooding of peatlands located in the vicinity of lakes (S: water).

Spatial distribution of peatlands
According to estimates, peatlands constitute about 7.7% of northern Poland, 4.0% of the whole country and about 5.5% of Europe (Okruszko 1996;Ilnicki 2002). The present research results indicate that the percentage share of areas identified as peatlands on the 1879 source map was 3.03% of the entire area of the Tuchola Pinewoods, which is less than the average concentration in northern Poland. The difference in peatland percentage presented in this study and cited works may arise from different data sources. Map of peatland distribution put forward by Ilnicki (2002) was compiled based on peat source documentations and field research conducted in the second half of the twentieth century, and included active peatlands and peat deposits. The Prussian map used in this study shows active peatlands in the last decades of the nineteenth century and may be less precise compared to later field documentation. However, it is the only source material indicating peatlands in the nineteenth century and offers a possibility to analyse historical changes related to peatlands in the study region. Based on this map, it can be seen that, spatially, the peatlands of the Tuchola Pinewoods are not evenly distributed (Fig. 2). Large complexes of peatlands are located in the vicinity of watershed zones, where they form distinct clusters. The share of peatlands in these zones on a 1 × 1-km grid exceeds 15% of the area -and in some grid units up to 60%. These zones are characterised by a thinner outwash cover and the presence of moraine islands (Szumińska 2014), and by considerable distance from the main drainage axes  Fig S1). The Tuchola Pinewoods mesoregion is located within a zone of 'raised bogs and fens' according to Żurek (1987). In this zone, not only is the occurrence of peatlands related to temperature and water regime (Montanarella et al. 2006), but also largely depends on local morphological and hydrological conditions (Dembek et al. 2000;Ilnicki 2002;Joosten and Clarke 2002). Therefore, peatlands may occur not only within river valleys and glacial channels. However, the comparison of peatland distribution in this study and the map presented by Ilnicki (2002) shows that several peatlands found in these geomorphological locations appear to be missing from the Prussian map. It is possible that selected parts of river valleys and glacial channels in the Tuchola Pinewoods were subject to intense farming at the time of drafting field documentation for the Prussian map, and thus, peat-forming process may have been impeded there until the second half of the twentieth century. The share of peatlands in the area does not appear to be related to local climate conditions. Due to its longitudinal extent, the Tuchola Pinewoods covers areas with different MAAT (6.5-6.9 °C) and MAP (575-825 mm). At the same time, it should be noted that MAAT decrease in a north-westwards direction, as MAP increase (Szumińska 2014). Higher sums of precipitation and lower air temperatures in the northern part of the research area should be more favourable to the formation of peatlands. Nonetheless, the spatial distribution of peatlands shows no such relationship (Fig. 2). It should also be noted that in RDA, the climatic factor (precipitation) was detected as an insignificant variable for the changes in use of peatlands in the period 1879-1986 ( Fig. 9 and Table 4).
Therefore, it should be assumed that the features recorded on the 1879 map as active peatlands indicate that the studied features meet the following conditions: 1 -the presence of a depression required for the development of a peatland and 2 -favourable water supply conditions. The spatial distribution of the identified peatlands indicates that such conditions are met not only along large depressions (river valleys, glacial channels) Słowiński et al. 2015), but also in the border zone between moraine areas and outwash plain. In these zones, sand and gravel formations are shallowly underlain by boulder clay (Szumińska 2014), and the circulation of groundwater favours flow from moraine areas towards the central part of outwashes drained by major rivers (Suppl. Mat., Fig. S3) (Kachnic and Kachnic 2006).

Human impacts
Losses of wetlands and peatlands increased during the Anthropocene (inter alia : Crutzen 2006;Gimmi et al. 2011;Dóka et al. 2019;Wittnebel et al. 2021;Diensen et al. 2022), and climate-induced losses coincided with those caused by humans (inter alia Ojala and Louekari 2002;Lamentowicz et al. 2007Lamentowicz et al. , 2009Booth et al. 2010). The main anthropogenic determinants of processes causing peatland transformation include drainage, large-scale lowering of ground water level (Glina et al. 2019b), peat subsidence (e.g. due to infrastructure established on it, agricultural use or structural amelioration of peat soils) (Tobolski 2003;Lamentowicz et al. 2009;Klimkowska et al. 2010;Grzybowski and Glińska-Lewczuk 2020), changes to natural vegetation, peat extraction (Glina et al. 2019a;Grzybowski and Glińska-Lewczuk 2020) and beaver activity (Rurek et al. 2016;Rurek 2021;Śnieszko et al. 2021). Research by Lamentowicz et al. (2011) indicates that at the end of the period 1800-1950, on the southern Baltic coast, as a result of, inter alia, drainage works, there were significant changes in the water management of peatlands. This resulted in a drop of water table, which in turn inhibited the development of peat-forming vegetation. Only at the beginning of the twenty-first century were measures taken to delay water runoff from peatlands (by blocking outflow in ditches). Changes in land use usually entail transformation of the entire local natural environment, especially features as sensitive to all external influences as peatlands (Koff et al. 1998;Lamentowicz et al. 2011;Glina et al. 2019aGlina et al. , 2022. As the presented results of research in the Tuchola Pinewoods in 1879-1986 show, as much as 53.5% of the total number of 744 peatlands studied (and 73.4% of the area) were transformed into agricultural areas, and this direction of change was particularly apparent in peatlands of more than 15 ha. It should also be noted that this change was facilitated/accompanied by the drainage of peatlands to reclaim agricultural land, which took place both before 1879 and in the twentieth century. In Poland and Europe, peatlands were drained on a large scale in the nineteenth and twentieth centuries (the most intensive drainage in Poland was in the 1960s and 70 s), and this was accompanied or preceded be other forms of anthropogenic transformation of the peatlands (Ilnicki 2002;Gimmi et al. 2011;Swindles et al. 2019;Kiryluk 2020;Glina et al. 2022).
Intensive drainage may lead to peat compaction, which in the case of fen peat under Polish climatic conditions has been estimated at up to 3-10 mm year −1 (Oleszczuk et al. 2008), 20 mm year −1 (Klimkowska et al. 2010) to even 30 mm year −1 (Glina et al. 2019a).
The specificity of the studied Tuchola Pinewoods, the remoteness from any major towns, the prevalence of nutrient-poor sandy soils and the history of the region resulted in the occurrence of the largest complex of pine forests in Poland. The prevalence of woodland use in the entire region (70.3% of the area is covered by) is probably one of the reasons why as much as 38.8% of the total number of peatlands (or 23.5% of the area) was partially or completely forested. It should be noted that not only the peatlands themselves were repurposed for forestry, but so were their immediate surroundings, which may explain the observed directions of change. Klimkowska et al. (2010) noted that the introduction of a pine monoculture may cause changes in evapotranspiration, due to the variation of this parameter from stand to stand. The twentieth century marked a general increase in the percentage of woodland in the Tuchola Pinewoods (Dysarz et al. 2005;Giętkowski 2012), which have caused an increase in total evapotranspiration across the region (Klimkowska et al. 2010). However, in the earlier period of 1750-1910, the share of pine monoculture (Słowiński et al. 2019), which may have causes the lowering of evapotranspiration, compared to natural, more species-diverse stands. The most recent study related to forest management in the region indicated several factors impacting changes in land use between the forest and non-forest type, as well as transformation in the species composition of the forest (Łuców et al. 2021): introduction of Pinus sylvestris monoculture, active removal of Betula, intended clear-cutting, logging after insect outbreaks and extensive forest damage connected with extreme tornado events. The above mentioned indicators should be considered 'local' factors affecting peatland changes in the Tuchola Pinewoods. Economic purposes (pine plantation) paired with natural phenomena (tornado, insect outbreaks) led to both direct and indirect changes of hydrological conditions in peatlands. As shown by the RDA analysis carried out in this study, land use in the vicinity of peatlands is a key determinant of the direction of changes in peatland use in the Tuchola Pinewoods. Another important factor is the drainage of the peatland, as is its location in a catchment area with a well-developed surface runoff network. The prevalent uses of peatlands in the Tuchola Pinewoods reflect worldwide trends. Globally, peatlands are nowadays commonly repurposed for forestry and agriculture, covering a total area of more than 1 million km 2 (∼25% of global peatland area) (Joosten and Clarke 2002;Page and Baird 2016), and this area doubles if peatlands used as pasture are taken into account (Montanarella 2014). Wetlands show the highest share of transformation to woodland areas, in particular at high and tropical latitudes, while in temperate areas, the changes result mainly from agricultural use (Ghajarnia et al. 2020;Wittnebel et al. 2021). In Europe, around 125,000 km 2 of peat soils are used for agricultural production across large areas of Russia, Germany, Belarus, Poland and Ukraine (Joosten and Clarke 2002). In the case of Poland, peatland management is focused predominately on agricultural areas (mainly meadows and pastures, which account for 69.6% of all area), followed by wasteland (13.8%) and peatlands for forestry use (11.7%). 4.4% of the peatland areas constitute peat extraction sites, and only 0.27% arable land (Lipka 1984). According to Kotowski and Piórkowski (2003), the loss of undisturbed mires in the twentieth century in Poland was estimated to be more than 80%. Wittnebel et al. (2021) claim that in Germany, 70% of peat and other organic soils are used for agriculture. Inappropriate agricultural management of drained peatlands led to a gradual loss of soil fertility (Wittnebel et al. 2021;Glina et al. 2022). Thus, degraded peatland soils are included, along with sandy soils, in the group of marginal soils with a low organic level (Ilnicki 2002). Since the 1980s, thanks to surplus food production, there has been a tendency in developed countries to discontinue the agricultural use of peatlands. This tendency is also evident in the Tuchola Pinewoods, where agricultural use of meadows away from towns, whether it in river valleys or within depressions found in woodlands, has been gradually abandoned (Szumińska 2014).
As noted in the introduction, there are few published studies that would address changes in peatland use directly, especially in the context of comparably long study period and scope (region). Analyses conducted in the scale of mesoregion that encompass similarly long periods (over 100 years) concentrate mostly on wetlands (Gimmi et al. 2011;Godet and Thomas 2013;Skaloš et al. 2017;Wittnebel et al. 2021). Gimmi et al. (2011) estimated that the share of wetlands in the surface area of Canton Zurich (Switzerland) decreased from 13,759 ha in 1850 (over 8% of total surface area) to 1233 ha in the year 2000 (less than 1%). Similar results were obtained in the Czech Republic, where the total surface area of wetlands drastically declined from 5762 ha (over 9.5% of the study area) in the years 1825-1843 to 54 ha (0.9%) in 2014 (Skaloš et al. 2017). Wetlands disappeared almost completely in west France, diminishing their share in the study surface area from 7.4% in 1820 to 0.2% in 1950 (Godet and Thomas 2013). Regional study on peatlands specifically was conducted in Germany, and the results suggest that approximately 70% of peat and other organic soils are transformed (drained and used for agriculture) (Wittnebel et al. 2021). The main factors underlying the disappearance of peatlands and loss of ecosystem services include drainage (Wittnebel et al. 2021) and transformation into agricultural areas, mostly grasslands (Godet and Thomas 2013;Wittnebel et al. 2021). In this study, we found that relatively early drainage (nineteenth century) as well as changes in the use of areas surrounding peatlands exerts impact on changes in peatland management. The importance of the adjacent areas may be connected with high share of small peatlands in the total number of such objects in the Tuchola Pinewoods. The presented study indicates that small peatlands may be strongly dependent on the large-scale changes of the landscape.

Climate drivers
Taking into consideration climate changes during the twentieth century, several study results show climate warming in the area of central Poland (Marosz et al. 2011;Wójcik and Miętus 2014;Pospieszyńska and Przybylak 2019;Okoniewska and Szumińska 2020), including the Tuchola Pinewoods. Pospieszyńska and Przybylak (2019), based on the data regarding a 140-year period for the meteorological station in Toruń, indicated an increase in MAAT amounting to 0.1 °C per decade. The authors show a higher frequency of very cool and extremely cool years before 1900, and in the 1940s, 1950s and 1960 of the twentieth century. It should be emphasised that the bulk of drainage works carried out in the Tuchola Pinewoods coincides with said cool periods. Following intensive drainage, since the late 1980s of the twentieth century, very warm years have been increasingly more frequent (Wójcik and Miętus 2014;Pośpieszyńska and Przybylak, 2019). The observed upward trends in air temperature (Michalska 2011) and terrestrial evaporation (Somorowska 2022) are the highest in western Poland and decrease eastwards. Potential evaporation (E 0 ) increased also in the region encompassing the study area from 440 mm in the years 1952-1980 to 530-560 mm in the years 2001-2018 (Okoniewska and Szumińska 2020). The increase in MAAT and E 0 resulted in more frequent occurrence of droughts (Marosz et al. 2011;Michalska 2011;Wójcik and Miętus 2014), which in turn exerts strong influence on peatlands (Słowiński et al. 2016).
Climatic changes that have been recorded since the late 1980s caused an overall change in the water balance (Wrzesiński et al. 2022), and hydrological responses to this changes are observable in central Poland as a decrease in water level of lakes (inter alia: Ptak 2013; Choiński et al. 2016;Skowron andJaworski 2017, Sojka et al. 2022) and drought condition on rivers (Bartczak et al. 2022;Wrzesiński et al. 2022). Small streams show either a dramatic decrease in flows or a periodic occurrence thereof (Banasik and Kazanowska 2016). In the peatlands located in watershed zones, as have been commonly observed in the Tuchola Pinewoods, drought conditions may occur earlier and more frequently than hydrological drought on the main rivers. In the shallow peatlands in these areas, water resources and water levels may be shaped mainly by the climatic conditions due to limited ground water supply. In view of a lack of apparent trend regarding precipitation in the twentieth century (Żmudzka 2009;Przybylak 2011), we can assume that the changes in peatland conditions (moisture, vegetation) in the study period were likely caused not only by the drainage works, but also by the increase in air temperature and potential evaporation. These climate drivers influenced changes in the use of peatlands, including woodland succession. Furthermore, drainage systems constructed in the cool climate periods may reinforce degradation of peatlands in warm and extremally warm conditions observed in the last decades.
It is worth noting that there is no regular and wide-range, regional or national monitoring of water levels in Polish peatlands. Such a programme would be invaluable in light of climate forecasts for the twenty-first century, which predict an increase in air temperature, particularly in northern latitudes (Solomon 2007;Kundzewicz and Matczak 2012;Meresa et al. 2016). Modelling of future climatic conditions has indicated an air temperature increase and changes in the seasonal pattern of atmospheric precipitation in the twenty-first century. Christensen and Hewitson (2007) forecast an increase in sums of precipitation in winter, but a decrease in summer, and the largest increase in evaporation and air temperature in the winter months. In light of projected climate changes, the subsurface water content in peat may drop drastically (Bertrand et al. 2021), which may result in reduced growth of peat-forming vegetation and, consequently, reduced carbon sequestration in peatlands (Słowiński et al. 2016;Wittnebel et al. 2021). According to Lunt et al. (2019), the sequestration of CO 2 in mires ranged between 5.7 and 21.6 t CO 2 ha −1 year −1 at the end of the nineteenth century and the early twentieth century. Therefore, all peatlands in the studied region sequestered approx. 59.8 and 226.5 Gt CO 2 year ˗1 at the turn of the twentieth century. Currently, the least transformed areas (showing no distinctive change) store from about 1.5 to 5.9 Gt CO 2 year ˗1 .
Due to a significant amount of organic carbon being sequestered in peatlands (Harenda et al. 2018;Li et al. 2019;Lunt et al. 2019;Krüger et al. 2021), peatland degradation may lead to increased CO 2 emission. The scale of the aforementioned emissions from European drained peatlands ranges on average from 8.6 to 15.1 t CO 2 ha −1 year −1 (Hooijer et al. 2010;Juszczak and Augustin 2013). Taking into account these values, it can be estimated that all peatlands drained by 1986 in the Tuchola Pinewoods collectively emitted an average of 79.3 to 139.2 Gt CO 2 year −1 at the end of the twentieth century.
According to Hooijer et al. (2010), each decrease of 10 cm in the groundwater level results in an increase in CO 2 emissions of 9.1 t CO 2 ha −1 year −1 . Therefore, in view of the projected climate warming, there is a significant risk of a surge in CO 2 emissions in the Tuchola Pinewoods, especially for the a significant number of peatlands where drainage networks have been found (the numbers being 182 as of 1879, 381 as of 1986, representing a total area of 6362 ha as of 1879, and 9219 ha as of 1986). The volume of CO 2 emissions in a degraded peatland can drop significantly provided that measures are implemented to raise the water table, slow down the rate of peat mineralisation and promote peat-forming vegetation (Stocker et al. 2013;Campbell et al. 2014; Ramsar Convention on Wetlands 2018; Junttila et al. 2021). According to Barthelmes et al. (2015), these values for mid-latitudes can be reduced by about 6 t CO 2 ha −1 year −1 for peatlands converted to forest areas, from 28 to 20 t CO 2 ha −1 year −1 for degraded peatlands used for agriculture and by 9 t CO 2 ha −1 year −1 for inactive peatlands (wasteland).

Conclusions
Based on our study covering a period of more than 100 years and 744 peatlands (with a total area of 10,762 ha) located in the young glacial area of the Tuchola Pinewoods (area 3550 km 2 ), the following conclusions can be drawn: -peatlands occur in clusters, the location of which is related to geomorphological (the presence of depressions and favourable geological structure) and hydrological conditions; numerous clusters of peatlands occur in watershed zones; -small peatlands (< 15 ha) are the most numerous, but very large ones (with an area of > 100 ha) account for the largest share of all identified peatlands, which together account for 32% of the area of all peatlands; -the analysis showed a quite clear tendency of land management transformation in the study period of 1879-1986, towards agricultural use for large peatlands, and to use as woodland in the case smaller areas; -the RDA analysis showed that the most important environmental factors influencing the directions of changes in peatland transformation are as follows: (i) the landuse type of the surrounding area, (ii) the implementation of anthropogenic drainage systems and (iii) location in a lake catchment or catchments drained by rivers or streams. Interestingly, however, no relationship was found between the type of peatland transformations and the spatial variability of precipitation sums; -considering that 85.7% of peatland in the TP region had been drained by 1986, the risk of releasing greenhouse gases from peatlands in the study area should be assessed as high.
The lack of meteorological stations, characterised by long-term air temperature observation series in the Tuchola Pinewoods (as opposed to stations with rainfall data), made it impossible to include air temperature in the RDA analysis conducted in this study. Another limitation of the obtained results is the uncertainty of the peatland delimitation based on historical maps. However, heterogeneity of data sources is common in studies at the meso-or macro-regional scale. Given the ecological importance of peatlands, including their ability to sequester and release carbon dioxide and methane, further research is advised, including monitoring to determine whether the peatlands in TP are being dried out and degraded at a regional scale on account of human impact and climate warming. Hence, the authors intend to continue their study of the selected areas indicated within TP based on larger number of cartographic sources, remote sensing data (Czapiewski and Szumińska 2022;Czapiewski 2022) and field research.
At the same time, it is necessary to consider the possibility of using the natural diversity land relief and hydrological conditions in the Tuchola Pinewoods to counteract process of peatlands degradation by increasing retention. Peatlands, due to their ability to store water and organic matter, may be used in sustainable water management and greenhouses emission control, both at regional and global scale. In this view, peatlands should be regarded not only as objects in need of protection, but also as objects that may help protect other elements of the environment: climate, hydrology, ecosystems. It should be noted that in young glacial areas, where peat-filled depressions are in many cases connected to a natural or artificial drainage network, it is quite easy to form small retention systems that support the preservation and/or restoration of peat-forming communities. The supporting features consist in land management in the areas surrounding peatlands, including the transformation of the species composition of forest stands towards those that support water retention in the environment. The results and conclusions presented in relation to the study area may be implemented for other young glacial areas of temperate climatic zone, which are under the influence of adverse effects of climate change, as well as anthropogenic changes in water runoff conditions.