Climate and water-table levels regulate peat accumulation rates across Europe

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 1 / 16 OPEN ACCESS Citation: Swindles GT, Mullan DJ, Brannigan NT, Fewster RE, Sim TG, Gallego-Sala A, et al. (2025) Climate and water-table levels regulate peat accumulation rates across Europe. PLoS One 20(7): e0327422. https://doi.org/10.1371/ journal.pone.0327422 Editor: John Toland Van Stan II, Georgia Southern University, UNITED STATES OF AMERICA Received: October 25, 2024 Accepted: June 13, 2025 Published: July 23, 2025 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal. pone.0327422 Copyright: © 2025 Swindles et al . This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, RESEARCH ARTICLE Climate and water-table levels regulate peat accumulation rates across Europe Graeme T. Swindles 1,2*, Donal J. Mullan3, Neil T. Brannigan4, Richard E. Fewster 1, Thomas G. Sim5, Angela Gallego-Sala6, Maarten Blaauw 7, Mariusz Lamentowicz8, Vincent E.J. Jassey9, Katarzyna Marcisz 8, Sophie M. Green3, Thomas P. Roland6, Julie Loisel10, Matthew J. Amesbury11, Antony Blundell12, Frank M. Chambers13, Dan J. Charman14, Callum R.C. Evans3, Angelica Feurdean15,16, Jennifer M. Galloway17, Mariusz Gałka18, Edgar Karofeld19, Evelyn M. Keaveney1, Atte Korhola20,

Łukasz Lamentowicz8, Peter Langdon21, Dmitri Mauquoy22, Michelle M. McKeown23, Edward A. D. Mitchell24,25, Gill Plunkett26, Helen M. Roe3, T. Edward Turner27, Ülle Sillasoo28, Minna Väliranta29, Marjolein van der Linden30, Barry Warner31 1 Geography and 14Chrono Centre, School of Natural and Built Environment, Queen’s University Belfast, Belfast, United Kingdom, 2 Ottawa-Carleton Geoscience Centre and Department of Earth Sciences, Carleton University, Ottawa, Ontario, Canada, 3 Geography, School of Natural and Built Environment, Queen’s University Belfast, Belfast, United Kingdom, 4 School of Geography and Environmental Sciences, Ulster University, Coleraine, Northern Ireland, United Kingdom, 5 Forest Research, Northern Research Station, Roslin, Midlothian, United Kingdom, 6 Geography, Faculty of Environment, Science and Economy, University of Exeter, Exeter, United Kingdom, 7 Archaeology and Palaeoecology and 14Chrono Centre, School of Natural and Built Environment, Queen’s University Belfast, Belfast, United Kingdom, 8 Climate Change Ecology Research Unit, Faculty of Geographical and Geological Sciences, Adam Mickiewicz University, Poznań, Poland, 9 Université de Toulouse, Toulouse INP, CNRS, IRD, CRBE, Toulouse, France, 10 Department of Geography, University of Nevada in Reno, Reno, Nevada, United States of America, 11 Research Services, University of Exeter, Innovation Centre Phase 2, Exeter, United Kingdom, 12 School of Geography, University of Leeds, Leeds, United Kingdom, 13 Centre for Environmental

Change and Quaternary Research, School of Natural and Social Sciences, University of Gloucestershire, Cheltenham, United Kingdom, 14 Geography, College of Life and Environmental Sciences, University of Exeter, Exeter, United Kingdom, 15 Goethe University, Institute of Physical Geography, Frankfurt am Main, Germany, 16 Department of Geology, Babeş-Bolyai University, Cluj-Napoca, Romania, 17 Geological Survey of Canada/Commission géologique du Canada, Calgary, Alberta, Canada, 18 Department of Biogeography, Paleoecology and Nature Conservation, Faculty of Biology and Environmental Protection, University of Lodz, Lodz, Poland, 19 Institute of Ecology and Earth Sciences, University of Tartu, Tartu, Estonia, 20 Environmental Change Research Unit, Faculty of Biological and Environmental Sciences, University of Helsinki, Helsinki, Finland, 21 School of Geography and Environmental Science, University of Southampton, Southampton, United Kingdom, 22 School of Geosciences, The University of Aberdeen, Aberdeen, United Kingdom, 23 Department of Geography and Environmental Research Institute, University College Cork, Cork, Ireland, 24 Laboratory of Soil Biodiversity, Institute of Biology, University of Neuchâtel, Neuchâtel, Switzerland, 25 Jardin Botanique de Neuchâtel, Neuchâtel, Switzerland, 26 Archaeology and Palaeoecology, School of Natural and Built Environment, Queen’s University Belfast, Belfast, United Kingdom, 27 Forestry and Land Scotland, South Region, Dumfries, United Kingdom, 28 Gaia School in Tallinn, Tallinn, Estonia, 29 Environmental Change Research Unit, Ecosystems and

environment research programme, University of Helsinki, Helsinki, Finland, 30 BIAX Consult, Zaandam, the Netherlands, 31 Department of Earth and Environmental Sciences, University of Waterloo, Waterloo, Ontario, Canada * g.swindles@qub.ac.uk

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 2 / 16 Abstract Background Peatlands are globally-important carbon sinks at risk of degradation from climate change and direct human impacts, including drainage and burning. Peat accumulates when there is a positive mass balance between plant productivity inputs and litter/ peat decomposition losses. However, the factors influencing the rate of peat accumu- lation over time are still poorly understood. Methodology/Principal Findings We examine apparent peat accumulation rates (aPAR) during the last two millennia from 28 well-dated, intact European peatlands and find a range of between 0.005 and 0.448 cm yr-1 (mean = 0.118 cm yr-1). Our work provides important context for the com- monplace assertion that European peatlands accumulate at ~0.1 cm per year. The highest aPAR values are found in the Scandinavian and Baltic regions, in contrast to Britain, Ireland, and Continental Europe. We find that summer temperature is a signif- icant climatic control on aPAR across our European sites. Furthermore, a significant relationship is observed between aPAR and water-table depth (reconstructed from testate-amoeba subfossils), suggesting that higher aPAR levels are often associated with wetter conditions. We also note that the highest values of aPAR are found when the water table is within 5–10 cm of the peatland surface. aPAR is generally low when water table depths are < 0 cm (standing water) or > 25 cm, which may relate to a decrease in plant productivity and increased decomposition losses, respectively.

Model fitting indicates that the optimal water table depth (WTD) for maximum aPAR is ~10 cm. Conclusions/Significance Our study suggests that, in some European peatlands, higher summer temperatures may enhance growth rates, but only if a sufficiently high water table is maintained. In addition, our findings corroborate contemporary observational and experimental studies that have suggested an average water-table depth of ~10 cm is optimal to enable rapid peat growth and therefore carbon sequestration in the long term. This has important implications for peatland restoration and rewetting strategies, in global efforts to mitigate climate change. Introduction Peatlands comprise only ~3% of the global landmass but are globally important habitats, carbon (C) stores and valuable archives of past environmental and climatic change [1–3]. European peatlands contain approximately half of the continent’s total soil organic carbon, and five times more carbon than its forests [4,5]. Peatlands and reproduction in any medium, provided the original author and source are credited. Data availability statement: All relevant data are within the manuscript and its Supporting Information files. Funding: We acknowledge all the organiza- tions that have funded the data used in this analysis: Academy of Finland; Department for Employment and Learning (Northern Ireland); European Commission (Fifth Framework); INTERACT (European Community’s Seventh Framework Programme); Irish Discovery Programme; Leverhulme Trust; National Science Centre (Poland); Natural Environment

Research Council (UK); Natural Sciences and Engineering Research Council of Canada; Netherlands Organization for Scientific Research; Polish National Science Centre (2021/41/B/ST10/00060 and 2021/03/Y/ ST10/00093 [BIODIVRESTORE]); Quaternary Research Association; Swiss Contribution to the enlarged European Union; Swiss Federal Office for Education and Science; Swiss National Science Foundation; World University Network; Wüthrich Fund (University of Neuchâtel); and Yorkshire Water. T.G.S. was funded by the Leeds–York Natural Environment Research Council (NERC) Doctoral Training Partnership (grant no. NE/L002574/1). T.E.T. acknowledges NERC Doctoral Training Grant no. NE/G52398X/1. JMGs contribution rep- resents NRCan contribution number/Numéro de contribution de RNCan: 20230392. This paper is a contribution to the PAGES C-PEAT group. PAGES is supported by the Swiss Academy of Sciences and Future Earth. This research was supported by a grant to G.T.S from the UK Leverhulme Trust (Grant No. RPG-2021-354) and a grant to E.M.K from UKRI NERC (Grant No. UKRI182). Competing interests: The authors have declared that no competing interests exist.

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 3 / 16 accumulate peat when plant inputs exceed peat losses through decomposition. Peat accumulation can therefore be con- sidered a mass balance where net growth or loss is dependent on the difference between inputs and outputs of organic carbon [6,7]. Natural peatlands exhibit a near-surface water table, which slows down C losses from microbial catabolic activities and enables peat accumulation [8]. Vegetation composition is a strong control on peat accumulation rates, with Sphagnum-dominated bogs often the fastest peat accumulators [e.g., 9–11]. Many European peatlands have been damaged in recent centuries through human activities including burning, drain- age, peat extraction, forestry, nutrient pollution and land-use changes [12]. In addition, climate change may have nega- tive consequences for some peatland ecosystems as warmer temperatures and shifts in precipitation patterns may lead to deepening water tables, which could in turn lead to greater peat decomposition and loss of soil organic carbon [13], although climate change may also increase peat accumulation in certain regions [1,14]. Major efforts are underway to conserve and restore degraded peatlands across Europe (e.g., LIFE Peat Restore). The primary method for restoring peatlands involves “re-wetting” or re-establishing the natural flow of water and soil saturation to the peatland ecosystem, which may include blocking artificial drains or canals [15]. The anticipation is that drain block-

ing will lead to restoration of a healthy peatland ecosystem, and in time, the ‘stabilised’ peatland will begin to accumulate peat and/or act again as a long-term carbon store [16]. There is also evidence to suggest that damaged peatlands can self-repair, including the spontaneous recovery of degraded peatlands where an ecological regime shift from erosion to renewed carbon accumulation has been observed [17,18]. Peatlands can thus be considered complex adaptive systems, demonstrating the capacity to self-regulate in terms of vegetation and hydrological functioning [19,20]. However, there are many gaps in scientific understanding of the factors behind peat formation and what causes the peatland carbon sink capacity to destabilise and re-stabilise. If the key processes involved can be elucidated, researchers will be better placed to model how peatlands respond to ongoing climate changes and to advise how to manage peatland carbon stocks more effectively. Current efforts to restore peatlands are only partly based on scientific evidence and, with- out process-based understanding, it is difficult to predict the long-term outcome of management approaches – including the possibility that restoration may be ineffective in transforming peatlands into carbon sinks [see 21]. This study uses palaeoenvironmental data from 28 well-dated, intact European peat bogs (Fig 1) to determine the climatic and hydrological controls on the apparent rates of peat accumulation (aPAR) (Table 1, Fig 1). An understanding of how fast peatlands grow

vertically (i.e., aPAR) may also have important implications for: i) robust determination of carbon (C) accumulation rates; ii) determining the temporal resolution of palaeoenvironmental profiles; and iii) validation of peatland development models. Results Peat-core chronologies were determined using Bayesian age models (S1) and used to calculate temporal changes in aPAR at a 1-cm resolution for each core. Trends in aPAR data were then analysed against hydrological (water-table depth) and climatic (gridded reanalysis) datasets. We find that aPAR in our European sites range from 0.005 to 0.448 cm yr-1 (Fig 2) and the average of the inter-site means (n = 28) is 0.118 cm yr-1 (average of all sample depths (n = 1732) is 0.140 cm yr-1). To reveal long-term trends that reflect the typical functioning of each site, we consider the site-based aPAR aver- ages, maxima and minima for further statistical analysis against climatic and WTD data (Table S7 in S1 File). We find that higher summer temperatures (June, July, August; JJA) are generally associated with higher aPAR values, a pattern evident in both contemporary and palaeo-climatic datasets. For the contemporary data, the Theil-Sen slope estimates indicate positive associations between summer temperature and aPAR (average aPAR: β = 0.30; maximum aPAR: β = 0.33; Table S8 in S1 File Fig 3). Spearman’s rank correlation supports this trend, showing a weakly significant rela- tionship between average aPAR and JJA temperature (p = 0.0875), and a more robust association for minimum aPAR (p = 0.0123; S2).

Using the palaeo-climate dataset, the relationship is even stronger, with Theil-Sen β of 0.52 for average aPAR and 0.55 for maximum aPAR. Spearman’s rank correlation confirms the significance of these associations (minimum

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 4 / 16 aPAR: p = 0.0101; maximum aPAR: p = 0.0025; average aPAR: p = 0.0004). In contrast, relationships between aPAR and other climatic variables are ambiguous (Table S8 in S1 File, Fig 3, S2,5 & 6 in S1 File). We find that the highest aPAR values are found in the Scandinavia and Baltics region (S3) and that there is no discernible difference in aPAR between notable, continent-wide climate phases of the late Holocene, including the Little Ice Age (~1500– ~1850 cal. yr. CE), Medi- eval Warm Period (~950– ~1250 cal. yr. CE) and Roman Warm Period (~1– ~400 cal. yr. CE) (S4). We observe a weak negative relationship between average aPAR and average WTD (Theil-Sen beta = −0.31) (Fig 3). Spearman’s rank correlation confirms a weakly significant relationship (p = 0.0854) (S2). aPAR is generally lower under deeper water tables (higher WTD), although a degree of non-linearity is suggested through the presence of the highest aPAR in the 5–10 cm WTD range (Fig 3). It is notable that none of our sites has high average aPAR when WTD is deeper than ~11 cm, suggesting a hydrological threshold. To explore this relationship in greater detail we re-examined the data from the 1-cm thick layers. Fig 4 shows there is a clear distribution of accumulation rates in relation to WTD – the highest box median values of AR occur when the reconstructed WTD is in the 5–10 cm range, which is in alignment with the site average-based analysis (Fig 3). Application of a Loess Smoothing Model indicates an optimal WTD range between 7 and

Fig 1. Location of study sites. Site descriptions are provided in Table 1. https://doi.org/10.1371/journal.pone.0327422.g001

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 5 / 16 Table 1. European study sites. Site number Site name Region Country Latitude Longitude Reference 1 Cloonoolish Britain & Ireland Ireland 53.1865 −8.2569 Blundell et al. (2008) Journal of Quaternary Science 23, 59–71. 2 Ballyduff Britain & Ireland Ireland 53.0807 −7.9925 Swindles et al. (2013) Earth Science Reviews 126, 300–320. 3 Derragh Britain & Ireland Ireland 53.7667 −7.4083 Langdon et al. (2012) Quaternary International 268, 145–155. 4 Ardkill Britain & Ireland Ireland 53.3653 −6.9532 Blundell et al. (2008) Journal of Quaternary Science 23, 59–71. 5 Dead Island Britain & Ireland Ireland 54.8862 −6.5487 Swindles et al. (2010) Quaternary Science Reviews 29, 1577–1589. 6 Slieveanorra Britain & Ireland Ireland 55.0848 −6.1921 Swindles et al. (2010) Quaternary Science Reviews 29, 1577–1589. 7 Butterburn Britain & Ireland England 55.0875 −2.5036 Mauquoy et al. (2008) Journal of Quaternary Science 23, 745–763. 8 Malham Britain & Ireland England 54.0964 −2.1750 Turner et al. (2014) Quaternary Science Reviews 84, 65–85. 9 Keighley Britain & Ireland England 54.4253 −2.0369 Blundell et al. (2016) Palaeogeography, Palaeo- climatology, Palaeoecology 443, 216–229. 10 Praz-Rodet Continental Europe Switzerland 46.5667 6.1736 Mitchell et al. (2001) The Holocene 11, 65–80. 11 Combe des Amburnex Continental Europe Switzerland 46.5397 6.2317 Sjögren & Lamentowicz (2008) Vegetation His- tory and Archaeobotany 17, 185–197.

12 Mauntschas Continental Europe Switzerland 46.4900 9.8544 van der Knaap et al. (2011) Quatermary Science Reviews 30, 3467–3480. 13 Lille Vildmose Scandinavia Denmark 56.8391 10.1896 Mauquoy et al. (2008) Journal of Quaternary Science 23, 745–763. 14 Izery Continental Europe Poland 50.8519 15.3602 Kajukało et al. (2016) European Journal of Proti- stology 55, 165–180. 15 Słowińskie Continental Europe Poland 54.3619 16.4785 Lamentowicz et al. (2009) Boreas 38, 214–229. 16 Bagno Kusowo Continental Europe Poland 53.8078 16.5872 Lamentowicz et al. (2015) Palaeogeography, Palaeoclimatology, Palaeoecology 418, 261–277. 17 Åkerlänna Römosse Scandinavia Sweden 60.0167 17.3667 van der Linden et al. (2008) Palaeogeography, Palaeoclimatology, Palaeoecology 262, 1–31. 18 Jelenia Wyspa Continental Europe Poland 53.5918 17.9821 Lamentowicz et al. (2007) The Holocene 17, 1185–1196. 19 Stążki Continental Europe Poland 54.4244 18.0833 Lamentowicz et al. (2011) Studia Quaternaria 28, 3–16. 20 Linje Continental Europe Poland 53.1880 18.3098 Marcisz et al. (2015) Quaternary Science Reviews 112, 138–152. 21 Stordalen I Scandinavia Sweden 68.3568 19.0484 Gałka et al. (2017) Permafrost and Periglacial Processes 28, 589–604. 22 Stordalen II Scandinavia Sweden 68.3568 19.0484 Gałka et al. (2017) Permafrost and Periglacial Processes 28, 589–604. 23 Lappmyran Scandinavia Sweden 64.1647 19.5828 van der Linden et al. (2008) Palaeogeography, Palaeoclimatology, Palaeoecology 258, 1–27. 24 Gązwa Continental Europe

Poland 53.8726 21.2201 Gałka et al. (2015) The Holocene 25, 421–434. 25 Mechacz Continental Europe Poland 54.3314 22.4419 Gałka et al. (2017) Quaternary Science Reviews 156, 90–106. 26 Kontolanrahka NE Europe Finland 60.7833 22.7833 Väliranta et al. (2007) The Holocene 17, 1093–1107. (Continued)

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 6 / 16 9 cm. Fitting a Gaussian Response Curve and Generalized Additive Model to the data suggests that the optimal WTD for the highest aPAR is 10.0 cm for both models (Fig 5). However, the Generalized Additive Model was not statistically signif- icant, and it is worth noting that the estimated optimal water table depth (WTD) is close to the overall mean of the WTD data (10.4 cm). aPAR is generally low when water table depths are < 0 cm (standing water) or > 25 cm. Site number Site name Region Country Latitude Longitude Reference 27 Tăul Muced Continental Europe Romania 47.5739 24.5450 Feurdean et al. (2015) The Holocene 25, 1179–1192. 28 Männikjärve NE Europe Estonia 58.8667 26.2500 Väliranta et al. (2012) Quaternary International 268, 34–43. https://doi.org/10.1371/journal.pone.0327422.t001 Fig 2. Boxplot showing aPAR for each site. Blue triangles indicate the mean values for each site. The blue line indicates the average of site means (0.118 cm yr-1). https://doi.org/10.1371/journal.pone.0327422.g002 Table 1. (Continued)

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 7 / 16 Discussion Peatlands exist because of an imbalance between plant productivity and decomposition. However, the mechanisms behind peatland growth, stability and degradation have been debated for many years [19,22]. The role of restoration and management in maximising peatland carbon sequestration and storage potential – to maintain and enable their contribu- tions as nature-based climate solutions to the climate crisis – is increasingly recognised in public and policy narratives. However, to optimise the benefits and minimise undesirable or ineffective outcomes, the driving mechanisms of peatland C accumulation need to be better understood. It is known that peatlands can self-regulate to a steady state [e.g., 23], but there is a concern that the magnitude of recent climate change and land-use degradation can lead to peatlands being pushed beyond a threshold, leading to loss of their carbon sink function and ultimately their carbon stock [13]. Although our findings support the common assertion that peat accumulates at ~0.1 cm yr-1, they also suggest that aPARs vary widely from ~0.005–0.448 cm yr-1 within Europe. We use contemporary and palaeo- climatic data to quantify the relative difference in climatic space between our sites. Significant relationships between peat accumulation rates and climate variables have been found in previous studies [24–26]. In our study, the sites experiencing warmer temperatures – in particular, warmer summers – are asso-

ciated generally with greater peat accumulation rates (Fig 3 & Table S8 in S1 File). This finding aligns with previous research that also indicated a positive association between elevated summer temperatures and increased peat accu- mulation rates [27], albeit with variations in temporal resolution. In high- to mid-latitude peatlands with adequate mois- ture, productivity is sometimes regarded as a more dominant driver of carbon accumulation than decomposition [28,29]. Warmer temperatures are well known to increase plant productivity, resulting in a greater input of organic matter into peat formations [1,4,30]. While microbial activity may also rise with increasing temperatures, increased productivity commonly outpaces decomposition, resulting in a net increase in peat accumulation [31]. However, for net peat accu- mulation there must still be adequate moisture in the summer months or elevated temperatures can lead to negative effects on Sphagnum structure and moisture holding capacity, desiccation, and even loss of peat [11,32]. In terms of the Köppen climate classification, the three sites with the highest aPAR values (Lille Vildmose, Mechacz and Åkerlänna Römosse) are in the warm-summer humid continental climate (Dfb), whereas the two sites with the lowest aPAR values (Stordalen I and II) are in the subarctic classification with cool summers and year-round rainfall (Dfc) and dark winter months with minimal insolation. Fig 3. Theil–Sen robust regression scatterplot of aPAR versus climatic data and reconstructed water-table depth (WTD). Palaeo-climatic data

are derived from CHELSA-TraCE21k [63], while contemporary data are sourced from the NOAA–CIRES–DOE 20th Century Reanalysis Version 3 [61,62]. https://doi.org/10.1371/journal.pone.0327422.g003

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 8 / 16 However, our findings suggest a lack of consistent or statistically significant correlations between peat accumulation and precipitation or GDD, as shown in Fig 3, S2 and S8 Tables in S1 File. While deglaciated regions have previously not shown a clear correlation with peat initiation and effective precipitation [e.g., 27], there may also be a degree of non-linearity, with an apparent optimum around 950 mm annual precipitation and 90 mm JJA precipitation (S5, S6). In such cases, overly wet conditions may reduce plant productivity by causing the expansion of open water areas within peatlands. This expansion can limit plant growth owing to reduced oxygen availability [11,33]. In waterlogged conditions, anaerobic decomposition processes dominate, which are much slower and less efficient, further favouring peat accumu- lation via decreased decomposition [33]. It is particularly intriguing in the case of GDD where previous studies [1,27,29] have intuitively indicated that a longer growing season leads to increased long-term carbon accumulation in northern peat- lands. The apparent absence of a strong correlation between aPAR and GDD in this study may be explained by the need for more substantial variations in GDD values to establish a statistically significant relationship with peat accumulation Fig 4. Boxplot of aPAR separated into 5-cm bins across the water-table depth gradient. https://doi.org/10.1371/journal.pone.0327422.g004

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 9 / 16 rates. This limitation could be attributed to the relatively limited seasonal distinctions within our study areas, which might not encompass the full range of GDD values necessary to establish a connection with peat accumulation definitively. Our research represents the first study using long-term peat accumulation data to examine the correlation between water-table depth and peat growth rates. We observe that the highest aPAR values occur when the water table is between 5 and 10 cm below the peatland surface. In contrast, aPAR tends to be low when the water table is above the surface (< 0 cm, ~standing water) or deeper than 25 cm, likely reflecting reduced plant productivity and increased decomposition losses, respectively. It is interesting that previous studies in peatland research have consistently identified an optimal WTD of approximately 10 cm, albeit with vastly contrasting approaches to this study, and over shorter timescales. Vari- ous studies focusing on the rewetting of peatlands suggest that maintaining a WTD of ~10 cm optimally mitigates green- house gas emissions. Studies include [34,35], drawing from earlier research [36–38], and [39], who refer to an ‘optimal Fig 5. Hexagonally-binned point density plots of aPAR data against mean water-table depth (width = 5 cm on the x-axis). The Loess Smoothing Model (green), Gaussian Response Model (red) and Generalized Additive Model (blue) are illustrated. The dashed line represents a water-table depth of

10 cm. ‘Rugs’ are shown along the axes to illustrate data density. https://doi.org/10.1371/journal.pone.0327422.g005

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 10 / 16 rewetting scenario’. Furthermore, palaeoecological records from Polish peatlands suggest a plant community tipping point at approximately 11.7 cm WTD [40]. This WTD value represents a hydrological threshold where marked shifts in the dominant peat-forming species (vascular plants and mosses) occur, leading to changes in peat accumulation patterns and potentially increased carbon loss [40,41]. Indeed, increasing vascular plant cover usually promotes C loss by increasing heterotrophic respiration and decomposition of peat carbon owing to rhizosphere priming effects [41–43]. Independently of vascular plants, a drop in WTD can accelerate the growth and activity of microbial communities, causing a rise in hetero- trophic respiration and decomposition rates [44]. [45] also demonstrated that peat formation rates peak when the acro- telm thickness is approximately 10 cm, while [46] found that photosynthesis reaches its maximum efficiency at a WTD of around 11 cm. A water-table depth value of 10 cm appears to also be an important hydrological threshold for the functional traits of testate amoebae [47]. Although [34,35,39] specifically mention an optimal peatland WTD of approximately 10 cm, their analyses are based on short-term observations, likely indicative of immediate peatland responses. Our approach provides information regarding the optimum WTD for long-term mean peat accumulation over millennia across multiple European sites. This approach

eliminated the potential problem of incomplete decomposition of the uppermost peats (the ‘acrotelm effect’) high- lighted by [51] through removal of data from the period 1850 cal. yr. CE to present from our analysis, following [29]. However, a recent article has suggested instances where apparent carbon accumulation rates (aCAR) may contradict the true behaviour of the peatland [52]. The authors highlight the “ageing problem” – which stipulates that the use of aCAR from peatlands is erroneous because of slow, long-term decomposition of peat in the catotelm over millennia. This issue is also highly relevant for peat accumulation rates (aPAR) as discussed here, in that the thickness of the peat layer measured at the time of coring reflects its current state rather than what necessarily accumulated initially [52]. We note, however, that our peat core profiles for the past 2,000 years do not exhibit a long-term decay trend nor do they show statistically significant increases in bulk density downcore, suggesting that the ageing effect is unlikely to be a major problem in this study. Furthermore, [52] discuss how accumulation rates lower in the peat profile may have been affected by what happened later in time as water tables fluctuate (the ‘secondary decomposition’ problem). However, our use of a sufficiently large number of sites, and a single time period (the last two millennia), should allow the detection of a signal from the noise inherent to these systems (including a degree of secondary decomposition).

Another potential criticism is the use of absolute reconstructed water-table measurements (cm) rather than directional shifts [53]. However, the transfer function used here has been shown to provide accurate mean annual water table predictions [54].

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 11 / 16 Despite these potential limitations, the consistent patterns observed in the accumulation rate data suggest that useful infor- mation may remain in peat archives. Our results may prove useful for parameterisation and testing of peatland development models such as Millennia [55], Holocene Peat Model [56] and Digibog [57]. Such models are becoming increasingly important for understanding the impacts of future climate- and land-use change on peatland ecosystems. Our results may support a grow- ing body of research based on contemporary monitoring and palaeoecological investigation that the optimal WTD for maximum peat accumulation is ~10 cm in European peatlands, and to steer away from simply ‘rewetting’ without prior knowledge of opti- mal WTD for each specific system. This provides a better-informed target condition that can be used in the restoration process if the main priority is to encourage maximum peat accumulation and therefore carbon sequestration in the long term. Materials and methods Data processing We analysed the European network of sites presented in [3] with each dataset spanning the last 2000 years (Table 1, Fig 1). A flexible Bayesian age–depth modelling approach [58] was used to generate an age model for each site (S1), using radiocarbon, 210Pb, tephra and spheroidal carbonaceous particle-based dating techniques. aPAR was calculated by dividing depth of peat accumulated (in cm) by time (years). [51] showed that recently formed peat cannot be compared to

older, deeper peats as decomposition means that most of the newly added material will not become part of the long-term carbon store (the ‘acrotelm effect’). We therefore removed aPAR data from the period AD1850-present from our analysis (following [29]) as this will contain uppermost peats where rapid aerobic decay is still taking place (the ‘acrotelm’ – [59]). Water-table reconstruction WTDs were reconstructed from subfossil testate amoebae with the pan-European transfer function of [54] using a weighted averaging tolerance-downweighted model with inverse deshrinking (S9). Sample-specific errors of prediction (maximum and minimum reconstruction ranges) were generated through 1,000 bootstrap cycles. The pan-European transfer function model has been shown to generate accurate mean annual water table predictions for surface samples with associated automated instrumental mean annual and summer WTD measurements [54]. Therefore, we used the absolute reconstructed water-table values rather than standardised values in this study. Reconstructions were carried out on the subfossil testate amoeba datasets following the removal of weak silicic idiosomic tests (Corythion-Trinema type, Euglypha ciliata type and Euglypha rotunda type) [60]. Climate analysis Daily mean temperature and precipitation data were taken from the NOAA-CIRES-DOE 20th Century Reanalysis V3 data- set covering the period 1836–2015 [61,62]. The dataset is the first ensemble of sub-daily global atmospheric conditions

spanning over a century, making it ideal for climate analyses extending as far back as the 19th century. We downloaded these data from KNMI Climate Explorer (https://climexp.knmi.nl/), focusing on the 1° longitude x 1° latitude grid box within which each of the 31 study sites is located. Growing degree days (GDD) were calculated for base temperatures above 0°C (GDD0) and 5°C (GDD5). Mean temperature, GDDs and precipitation totals were then aggregated to annual and northern hemisphere summer (June, July, August) temporal resolution for the entire record. Mean monthly and annual palaeo-climate simulation data were taken from the CHELSA-TraCE21k long-term climatology covering the period 100 B.C.E. to 1990 C.E. [63]. The data are available at 100-year timesteps at high spatial resolution. We downloaded these data from: https://chelsa-climate.org/chelsa-trace21k/. Mean annual temperature (°C) and total annual precipitation (mm yr-1) data were taken directly from the variables bio1 and bio12. Mean monthly temperature was calculated from the mean of tasmin and tasmax. Mean monthly temperature and precipitation totals were then aggregated for northern hemisphere summer (June, July, August) for each 100-yr timestep. For GDD0 and GDD5, we multiplied mean monthly tem- peratures by the days of each month and summed these values relative to thresholds of 0°C and 5°C respectively.

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 12 / 16 Statistical analysis Relationships between site-based aPAR (average, highest, lowest), climate variables and reconstructed WTD were explored for each site using Theil-Sen robust regression in the R package deming v.1.4 [64]. Theil-Sen regression coefficients were then standardised to produce beta coefficients. Spearman’s rank correlation was used to further clarify relationships among variables because it is a non-parametric method that does not assume a linear relationship or normally distributed data. This makes it particularly suitable for exploring monotonic associations between ecological and climatic variables, especially when dealing with reconstructed or heterogeneous datasets that may include outliers or non-linear trends. The climatic variables analysed for each site included both contemporary (NOAA-CIRES-DOE 20th Century Reanalysis V3) and palaeo-climate (CHELSA-TraCE21k) datasets. These comprised annual and summer (JJA) temperature, annual and summer (JJA) precipi- tation, and growing degree days above 0°C (GDD0) and 5°C (GDD5). Water table depth (WTD) variables were represented by average, maximum, and minimum values from each reconstruction, based on bootstrapped error estimations. To determine relationships between aPAR and WTD, a Loess Smoothing Model [65,66], a Gaussian Response Model [67], and a Generalized Additive Model (GAM) [68] were utilised. The GRM calculates an initial estimation of optimum and

tolerance based on the weighted average, followed by a nonlinear optimization by the Levenberg-Marquardt method. All statistical analyses were carried out in either PAST v.4 [69] or R v.4.2.1 [70]. Graphics were produced using the ggplot2 R package v.3.4.4 [71]. The GAM was calculated using the mgcv R package v.1.9.0 [72]. Supporting information S1 File. Supporting Information 1 (Figure): Bayesian age-depth models. Supporting Information 2 (Table): Spearman’s rank correlation information (Rs (bottom left) and p-values (top right) are shown). Supporting Information 3 (Figure): Boxplot showing aPAR for each region (Britain and Ireland, Continental Europe; Scandinavia and Baltics).Supporting Information 4 (Figure): Boxplot showing aPAR across different climate phases including the Little Ice Age (LIA: 1500–1850 CE), Medieval Warm Period (MWP: 950–1250 CE) and Roman Warm Period (RWP: 1–400 CE). All data points are shows as well as the non-Little Ice Age data points (nLIA). Supporting Information 5 (Figure): Theil–Sen robust regression scatterplot of aPAR versus modern climatic data (NOAA–CIRES–DOE 20th Century Reanalysis Version 3). Supporting Information 6 (Figure): Theil–Sen robust regression scatterplot of aPAR versus palaeo-climatic data (CHELSA-TraCE21k). Supporting Information 7 (Table): Site-based data for aPAR, WTD, contemporary climate and palaeo-climate. Supporting Information 8 (Table): Theil–Sen robust regression information for aPAR regressions with WTD, contemporary climate, and palaeo-climate. Supporting Information 9 (Dataset). All data. (ZIP) Acknowledgments

We thank the two reviewers for their constructive comments on an earlier version of this manuscript. JMGs contribution represents NRCan contribution number/Numéro de contribution de RNCan: 20230392. This paper is a contribution to the PAGES C-PEAT group. Author contributions Conceptualization: Graeme Swindles, Richard E. Fewster, Mariusz Lamentowicz, Evelyn M. Keaveney, Michelle M. McKeown. Data curation: Graeme Swindles, Angela Gallego-Sala, Mariusz Lamentowicz, Matthew J. Amesbury, Antony Blundell, Frank M. Chambers, Dan J. Charman, Angelica Feurdean, Mariusz Galka, Edgar Karofeld, Atte Korhola, Lukasz Lamentowicz, Peter Langdon, Dmitri Mauquoy, Edward A. D. Mitchell, Gill Plunkett, Helen M. Roe, T. Edward Turner, Ulle Sillasoo, Minna Valiranta, Marjolein van der Linden, Barry Warner.

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 13 / 16 Formal analysis: Graeme Swindles, Donal J. Mullan, Neil T. Brannigan, Thomas G. Sim, Angela Gallego-Sala, Vincent E.J. Jassey. Investigation: Graeme Swindles, Donal J. Mullan, Neil T. Brannigan, Richard E. Fewster, Thomas G. Sim, Maarten Blaauw, Mariusz Lamentowicz, Vincent E.J. Jassey, Katarzyna Marcisz, Sophie M. Green, Thomas P. Roland, Julie Loisel, Jennifer M. Galloway. Methodology: Graeme Swindles, Donal J. Mullan, Neil T. Brannigan, Richard E. Fewster, Thomas G. Sim, Angela Gallego-Sala, Callum R.C. Evans. Project administration: Graeme Swindles. Resources: Graeme Swindles. Supervision: Graeme Swindles. Validation: Graeme Swindles. Visualization: Graeme Swindles, Donal J. Mullan, Neil T. Brannigan. Writing – original draft: Graeme Swindles, Donal J. Mullan, Neil T. Brannigan, Richard E. Fewster, Thomas G. Sim, Angela Gallego-Sala, Maarten Blaauw, Mariusz Lamentowicz, Vincent E.J. Jassey, Katarzyna Marcisz, Sophie M. Green, Thomas P. Roland, Julie Loisel, Callum R.C. Evans, Jennifer M. Galloway, Michelle M. McKeown. Writing – review & editing: Graeme Swindles, Neil T. Brannigan, Richard E. Fewster, Thomas G. Sim, Maarten Blaauw, Mariusz Lamentowicz, Katarzyna Marcisz, Julie Loisel, Matthew J. Amesbury, Antony Blundell, Frank M. Chambers, Dan J. Charman, Angelica Feurdean, Jennifer M. Galloway, Mariusz Galka, Edgar Karofeld, Evelyn M. Keaveney, Atte Korhola, Lukasz Lamentowicz, Peter Langdon, Dmitri Mauquoy, Michelle M. McKeown, Edward A. D. Mitchell, Gill

Plunkett, Helen M. Roe, T. Edward Turner, Ulle Sillasoo, Minna Valiranta, Marjolein van der Linden, Barry Warner. References 1. Gallego-Sala AV, Charman DJ, Brewer S, Page SE, Prentice IC, Friedlingstein P, et al. Latitudinal limits to the predicted increase of the peatland carbon sink with warming. Nature Clim Change. 2018;8(10):907–13. https://doi.org/10.1038/s41558-018-0271-1 2. Xu J, Morris PJ, Liu J, Holden J. PEATMAP: Refining estimates of global peatland distribution based on a meta-analysis. CATENA. 2018;160:134– 40. https://doi.org/10.1016/j.catena.2017.09.010 3. Swindles GT, Morris PJ, Mullan DJ, Payne RJ, Roland TP, Amesbury MJ, et al. Widespread drying of European peatlands in recent centuries. Nat Geosci. 2019;12(11):922–8. https://doi.org/10.1038/s41561-019-0462-z 4. Limpens J, Berendse F, Blodau C, Canadell JG, Freeman C, Holden J, et al. Peatlands and the carbon cycle: from local processes to global impli- cations – a synthesis. Biogeosciences. 2008;5(5):1475–91. https://doi.org/10.5194/bg-5-1475-2008 5. Panagos P, Van Liedekerke M, Yigini Y, Montanarella L. Contaminated sites in Europe: review of the current situation based on data collected through a European network. J Environ Public Health. 2013;2013:158764. https://doi.org/10.1155/2013/158764 PMID: 23843802 6. Malmer N, Wallén B. Peat formation and mass balance in subarctic ombrotrophic peatland around Abisko, northern Scandinavia. Ecol Bull. 1996;1996:79–92.

7. Malmer N, Svensson G, Wallén B. Carbon and mass balance in a south Swedish ombrotrophic bog: processes and variation during recent centu- ries. Mires Peat. 2011;8(1). 8. Holden J. Peatland hydrology and carbon release: why small-scale process matters. Philos Trans A Math Phys Eng Sci. 2005;363(1837):2891– 913. https://doi.org/10.1098/rsta.2005.1671 PMID: 16286296 9. Thormann MN, Szumigalski AR, Bayley SE. Aboveground peat and carbon accumulation potentials along a bog-fen-marsh wetland gradient in southern boreal Alberta, Canada. Wetlands. 1999;19(2):305–17. https://doi.org/10.1007/bf03161761 10. Belyea LR, Malmer N. Carbon sequestration in peatland: patterns and mechanisms of response to climate change. Global Change Biology. 2004;10(7):1043–52. https://doi.org/10.1111/j.1529-8817.2003.00783.x 11. Rydin H, Jeglum JK. Peatlands around the world. The Biology of Peatlands. 2006;216–38. 12. Page SE, Baird AJ. Peatlands and Global Change: Response and Resilience. Annu Rev Environ Resour. 2016;41(1):35–57. https://doi. org/10.1146/annurev-environ-110615-085520

PLOS One | https://doi.org/10.1371/journal.pone.0327422 July 23, 2025 14 / 16 13. Ise T, Dunn AL, Wofsy SC, Moorcroft PR. High sensitivity of peat decomposition to climate change through water-table feedback. Nature Geosci- ence. 2008;1:763–6. https://doi.org/10.1038/ngeo331 14. Loisel J, Yu Z. Recent acceleration of carbon accumulation in a boreal peatland, south central Alaska. JGR Biogeosciences. 2013;118(1):41–53. https://doi.org/10.1029/2012jg001978 15. Parry LE, West LJ, Holden J, Chapman PJ. Evaluating approaches for estimating peat depth. J Geophys Res Biogeosci. 2014;119(4):567–76. https://doi.org/10.1002/2013jg002411 16. Green SM, Baird AJ, Holden J, Reed D, Birch K, Jones P. An experimental study on the response of blanket bog vegetation and water tables to ditch blocking. Wetlands Ecol Manage. 2017;25(6):703–16. https://doi.org/10.1007/s11273-017-9545-z 17. Swindles GT, Morris PJ, Wheeler J, Smith MW, Bacon KL, Edward Turner T, et al. Resilience of peatland ecosystem services over millennial times- cales: evidence from a degraded British bog. Journal of Ecology. 2016;104(3):621–36. https://doi.org/10.1111/1365-2745.12565 18. Milner AM, Baird AJ, Green SM, Swindles GT, Young DM, Sanderson NK, et al. A regime shift from erosion to carbon accumulation in a temperate northern peatland. Journal of Ecology. 2020;109(1):125–38. https://doi.org/10.1111/1365-2745.13453 19. Belyea LR, Baird AJ. Beyond “the limits to peat bog growth”: cross-scale feedback in peatland development. Ecological Monographs.

2006;76(3):299–322. https://doi.org/10.1890/0012-9615(2006)076[0299:btltpb]2.0.co;2 20. Dommain R, Dittrich I, Giesen W, Joosten H, Rais DS, Silvius M, et al. Ecosystem services, degradation and restoration of peat swamps in the Southeast Asian tropics. In: Peatland restoration and ecosystem services: Science, policy and practice. Cambridge: Cambridge University Press; 2016;253–88. https://www.cambridge.org/core/books/abs/peatland-restoration-and-ecosystem-services/ ecosystem-services-degradation-and-restoration-of-peat-swamps-in-the-south-east-asian-tropics/24B1396EE935F2842372C47A53DB45E0 21. Loisel J, Gallego-Sala A. Ecological resilience of restored peatlands to climate change. Commun Earth Environ. 2022;3(1). https://doi.org/10.1038/ s43247-022-00547-x 22. Clymo RS. The limits to peat bog growth. Phil Trans R Soc Lond B. 1984;303(1117):605–54. https://doi.org/10.1098/rstb.1984.0002 23. Waddington JM, Morris PJ, Kettridge N, Granath G, Thompson DK, Moore PA. Hydrological feedbacks in northern peatlands. Ecohydrology. 2014;8(1):113–27. https://doi.org/10.1002/eco.1493 24. Yu Z, Loisel J, Brosseau DP, Beilman DW, Hunt SJ. Global peatland dynamics since the Last Glacial Maximum. Geophysical Research Letters. 2010;37(13). https://doi.org/10.1029/2010gl043584 25. Turetsky MR, Benscoter B, Page S, Rein G, van der Werf GR, Watts A. Global vulnerability of peatlands to fire and carbon loss. Nature Geosci. 2014;8(1):11–4. https://doi.org/10.1038/ngeo2325

26. Wilson RM, Hopple AM, Tfaily MM, Sebestyen SD, Schadt CW, Pfeifer-Meister L, et al. Stability of peatland carbon to rising temperatures. Nat Commun. 2016;7:13723. https://doi.org/10.1038/ncomms13723 PMID: 27958276 27. Morris PJ, Swindles GT, Valdes PJ, Ivanovic RF, Gregoire LJ, Smith MW, et al. Global peatland initiation driven by regionally asynchronous warm- ing. Proc Natl Acad Sci U S A. 2018;115(19):4851–6. https://doi.org/10.1073/pnas.1717838115 PMID: 29666256 28. Turetsky MR, Mack MC, Hollingsworth TN, Harden JW. The role of mosses in ecosystem succession and function in Alaska’s boreal forestThis arti- cle is one of a selection of papers from The Dynamics of Change in Alaska’s Boreal Forests: Resilience and Vulnerability in Response to Climate Warming. Can J For Res. 2010;40(7):1237–64. https://doi.org/10.1139/x10-072 29. Charman DJ, Beilman DW, Blaauw M, Booth RK, Brewer S, Chambers FM, et al. Climate-related changes in peatland carbon accumulation during the last millennium. Biogeosciences. 2013;10(2):929–44. https://doi.org/10.5194/bg-10-929-2013 30. Malmer N, Johansson T, Olsrud M, Christensen TR. Vegetation, climatic changes and net carbon sequestration in a North‐Scandinavian subarctic mire over 30 years. Global Change Biology. 2005;11(11):1895–909. https://doi.org/10.1111/j.1365-2486.2005.01042.x 31. Robroek BJM, Jassey VEJ, Payne RJ, Martí M, Bragazza L, Bleeker A, et al. Taxonomic and functional turnover are decoupled in European peat bogs. Nat Commun. 2017;8(1):1161. https://doi.org/10.1038/s41467-017-01350-5 PMID: 29079831

32. Dorrepaal E, Aerts R, Cornelissen JHC, Callaghan TV, Van Logtestijn RSP. Summer warming and increased winter snow cover affect Sphagnum fuscum growth, structure and production in a sub‐arctic bog. Global Change Biology. 2003;10(1):93–104. https://doi. org/10.1111/j.1365-2486.2003.00718.x 33. Moore TR, Roulet NT, Waddington JM. Uncertainty in predicting the effect of climatic change on the carbon cycling of Canadian peatlands. Climatic Change. 1998;40(2):229–45. https://doi.org/10.1023/a:1005408719297 34. Tiemeyer B, Freibauer A, Borraz EA, Augustin J, Bechtold M, Beetz S, et al. A new methodology for organic soils in national greenhouse gas inven- tories: Data synthesis, derivation and application. Ecological Indicators. 2020;109:105838. https://doi.org/10.1016/j.ecolind.2019.105838 35. Tanneberger F, Appulo L, Ewert S, Lakner S, Ó Brolcháin N, Peters J, et al. The Power of Nature‐Based Solutions: How Peatlands Can Help Us to Achieve Key EU Sustainability Objectives. Advanced Sustainable Systems. 2020;5(1). https://doi.org/10.1002/adsu.202000146 36. Couwenberg J, Thiele A, Tanneberger F, Augustin J, Bärisch S, Dubovik D, et al. Assessing greenhouse gas emissions from peatlands using vege- tation as a proxy. Hydrobiologia. 2011;674(1):67–89. https://doi.org/10.1007/s10750-011-0729-x 37. Wichtmann W, Schröder C, Joosten H. Paludiculture-productive use of wet peatlands. Stuttgart: Schweizerbart Sche Vlgsb. 2016. 38. Gaudig G, Tanneberger F. Peatland Science and Conservation: Contributions of the Greifswald Mire Centre, Germany. Innovations in Landscape