Severe wildfire legacies suppress methane emissions after peatland rewetting Authors: Shihao Cui, Haonan Guo, Lorenzo Pugliese, Shubiao Wu Published: 2026 ABSTRACT Drained peatlands are major sources of greenhouse gases, but rewetting them to restore carbon storage often triggers substantial methane emissions and faces resistance from agricultural land users. Wildfires are becoming more frequent on these drained landscapes and can sharply reduce agricultural productivity. It remains unclear whether severely wildfire-affected peatlands can serve as effective targets for rewetting. Here we show that rewetting severely wildfire-affected peat soils suppresses methane emissions by more than 90% compared with rewetting non-fire-affected soils. Severe wildfire legacies increase soil carbon stability, raise pH and electrical conductivity, reduce methanogenic functional gene abundance, and can override typical hydrological controls, turning what is usually a methane surge into a strong net climate benefit. These changes position recent severe wildfire footprints as overlooked low-emission, high-priority targets for peatland rewetting that also reduce land-use conflict. Global scaling indicates that prioritizing rewetting on severely wildfire-affected degraded peatlands could avoid up to 0.8 million tonnes of CO2-equivalent methane emissions per year, with more than 70% of the potential in Asia. The findings reveal a practical pathway to accelerate restoration by harnessing wildfire legacies rather than fighting them. Introduction Peatlands cover only about 3% of Earth's land, yet store about 25% of global soil carbon, making them critical ecosystems for climate regulation [1]. However, over the past century, large areas of peatlands have been drained, mainly for agriculture, transforming them from net carbon sinks into major sources of CO2, which now account for approximately 5% of global anthropogenic greenhouse gas (GHG) emissions [2]. Peatland rewetting effectively restores hydrological functions, suppresses aerobic peat decomposition, and recovers long-term carbon storage capacity [3]. Consequently, this strategy has rapidly expanded in climate mitigation and restoration policies, with ambitious restoration targets now embedded in national and international policy frameworks [4,5]. Rewetting also involves a well-known climate trade-off: although raising the water table suppresses CO2 release, it often stimulates CH4, a more powerful GHG. This net climate benefit of rewetting is therefore not always immediate, slowing progress toward large-scale restoration. Rewetting also competes with farmers’ land-use interests. Most drained peatlands have been used for cropping or grazing, and rewetting typically implies reduced productivity or outright land-use change [6]. The resulting economic loss leads farmers to resist rewetting unless it is offset by adequate financial compensation or subsidies [7]. This conflict between climate goals and agricultural land-use interests is a bottleneck in peatland rewetting. Consequently, there is increasing interest in identifying areas where rewetting can be both climatically effective and socially feasible. One underexplored factor in this regard is wildfire. Drained peatlands are becoming increasingly susceptible to wildfires due to prolonged drainage and climate change [8,9]. Centuries of drainage have considerably lowered water tables, dried the peat, and shifted vegetation toward more flammable species, reducing peatland resilience to wildfire events [10]. Climate change intensifies these vulnerabilities, particularly in temperate and Arctic regions, where warming, recurrent droughts, and reduced groundwater recharge have led to widespread fires in recent years, documented across Europe (e.g., Russia, the United Kingdom, and Scandinavian countries) [11], North America (e.g., western Canada) [6], and Northeast China [12]. In tropical regions, fires set to clear vegetation during land-use conversion often ignite the peat layer itself [13], leading to near-annual dry-season fires in some areas [14]. Under extreme drought, these soils may smolder for weeks or months, with fire depths reaching several tens of centimeters [15,16]. Such fires can destroy soil structure, damage crops, and cause long-term declines in land productivity [17,18], resulting in financial losses that farmers may take years to recover from Ref. [19]. The resulting soil degradation may sometimes make the land less suitable for farming, likely reducing farmer opposition or compensation costs for rewetting. This raises a key question: Could fire-affected peatlands, having already lost much of their agricultural value, present a unique opportunity for prioritized rewetting? If these areas are more acceptable to farmers or landowners, they might offer a more pragmatic entry point for scaling up restoration. However, the viability of this strategy also effectively depends on understanding how wildfire history influences GHG emission dynamics in fire-affected peatlands after rewetting. Answering this question is essential for evaluating both the climate benefits and the feasibility of prioritizing burned peatlands for accelerated restoration. In this study, we assessed whether fire-affected, drained peatlands can have meaningful climate benefits. We first conducted laboratory incubation experiments comparing non-fire-affected and fire-affected peat soils across wildfire severities under controlled rewetting conditions. To elucidate the biogeochemical mechanisms underlying the observed GHG responses, we analyzed soil physicochemical properties, characterized microbial community composition and functional potential, and quantified the critical genes involved in CH4 cycling, given that CH4 often dominates the GHG balance after rewetting. Correlation analyses were conducted to link environmental changes with gas flux variations and to provide a mechanistic framework for the observed emission trends. Finally, we integrated our experimental results with global geospatial data on peatland distribution and burned area to estimate the potential CH4 mitigation of prioritizing rewetting in fire-affected peatlands. These findings are expected to elucidate how wildfire history influences GHG responses to rewetting and help guide restoration priorities in increasingly fire-prone drained peatlands. Methods Site description and soil sampling The peatland site is located within a Natura 2000 reserve in Viborg, Denmark (56°32′46″ N, 9°34′24″ E). This site, transitioning from a riparian fen, is dominated by vegetation such as Holcus lanatus and Festuca species. The site represents a typical converted peatland that remains under shallow, but unmaintained, drainage conditions, with very extensive surrounding land use since the 1950s. The basic physicochemical properties of the original soil are presented in Supplementary Table S1. Soil sampling was conducted using cylindrical metal rings (20 cm inner diameter, 30 cm height) with sharpened lower edges, which were mechanically pressed into the soil to extract intact soil cores. These cores were carefully placed on plastic trays and transported to the designated burning experiment facility at Aarhus University (Viborg campus) for further treatment. Experimental design To examine the effects of wildfire severity on post-rewetting GHG emissions, we established four treatments: drained soils, rewetted non-fire-affected soils, rewetted mildly non-fire-affected soils, and rewetted severely fire-affected soils. We simulated wildfire severity by applying a propane–butane torch directly to the soil surface from a distance of approximately 5 cm for 5 min or 10 min, representing mild and severe wildfire treatment, respectively. We did not monitor temperature profiles at different depths within the peat columns. However, previous studies on fire duration and soil heating show that short heating events, such as 5 min of heating, generally raise soil temperatures to less than 150 °C at a depth of 1 cm [20]. Heating of this magnitude mainly affects surface moisture and has little effect on organic matter properties [21]. In contrast, longer heating, such as 10 min of exposure, can raise surface temperatures to more than 400 °C and sustain temperatures above 250 °C several centimeters below the surface [22]. These conditions can induce deeper structural and biogeochemical changes in soil organic matter [23,24]. These treatments represent, to some extent, ecologically relevant analogs of field-observed mild versus severe wildfire severities in peatlands while acknowledging that they cannot capture the full complexity of natural fire behavior. After the simulated fires, the soil cores were allowed to cool. Smaller intact soil subsamples, 5 cm in diameter and 8 cm in height, were then collected from within the larger cores for incubation experiments. We focused on the surface peat layer because near-surface peat may exhibit relatively high CH4 production potential due to greater availability of labile carbon substrates that stimulate methanogenesis [25]. The smaller soil cores from each treatment were placed in airtight incubation jars, with an inner diameter of 10 cm and a height of 25 cm, for GHG measurements under controlled laboratory conditions. In the drained treatment, the water table was set at 8 cm below the soil surface to replicate drained conditions [26]. In the three rewetted treatments, the water table was maintained 8 cm above the soil surface to ensure sustained saturation and anoxic conditions, allowing for a clear assessment of CH4 production potential under postfire rewetting scenarios [27]. Under these conditions, the rewetted peat columns remained largely water-saturated, and substantial vertical differences in water content within the peat cores were therefore expected to be limited. All incubation jars were then placed in a climate-controlled room maintained at 20 °C for 3 months, with three replicates per treatment. The 3-month incubation period was chosen to capture the early post-rewetting phase, when shifts in microbial activity and gas dynamics are most pronounced in response to changes in substrate quality and soil physicochemical conditions. Similar incubation lengths (60−100 days) have been reported to capture peatland gas emission patterns at the correct order of magnitude relative to field observations [25]. Measurement of gas and soil physicochemical properties To evaluate treatment effects on CH4 and CO2 emission dynamics, we collected headspace gas samples from the glass jars twice weekly throughout the experimental period. At each sampling event, 10 mL of headspace gas was collected using a syringe and transferred into pre-vacuumed glass exetainers (Labco Limited, UK). The cumulative concentrations of CO2 and CH4 were then analyzed using a gas chromatograph (8890 GC System, Agilent, Santa Clara, USA). Instrument configuration and calibration followed established protocols [28]. Understanding the molecular characteristics of soil organic carbon is essential to identifying the stability changes induced by fire. Before incubation, surface soil samples (0−1 cm depth) from non-fire-affected, mildly fire-affected, and severely fire-affected treatments were collected, homogenized, and used for the characterization of major functional groups by Fourier-transform infrared spectroscopy (FTIR; Bruker INVENIO, USA). Differences in soil carbon molecular characteristics were assessed by comparing the normalized peak height variations of key functional groups across treatments. After incubation, soil pH was measured in a 1:2.5 soil-to-water suspension using a pH meter (PHM 210, Radiometer Analytical, France). Electrical conductivity was measured at a 1:5 soil-to-water ratio with an electrical conductivity meter (CDM 210, Radiometer Analytical, France). Total carbon and total nitrogen were analyzed with an elemental analyzer (VarioMax, Elementar GmbH, Hanau, Germany). For total organic carbon analysis, inorganic carbon was first removed using hydrochloric acid, and total organic carbon was then measured using an elemental analyzer. Inorganic nitrogen species (NH4+-N and NO3−-N) were extracted with potassium chloride solution and measured using a segmented flow analyzer (AA500 Autoanalyzer, SEAL Analytical, Southampton, UK). Elemental concentrations of iron (Fe), manganese (Mn), phosphorus (P), and sulfur (S) were analyzed using inductively coupled plasma optical emission spectrometry (ICP–OES; Thermo Scientific iCAP 6000 series, Thermo Fisher Scientific, Waltham, MA, USA). Microbiology To investigate the microbial mechanisms underlying GHG emissions across different treatments, we analyzed microbial community composition and the abundance of key functional genes associated with CH4 cycling. After incubation, DNA from fresh soil samples from each treatment was extracted using the DNeasy PowerSoil Pro Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. DNA concentrations were determined using a Qubit 3.0 fluorometer (Invitrogen, Carlsbad, CA, USA). We then performed real-time quantitative polymerase chain reaction (qPCR) on a QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) for different target genes. For methanogens, the methyl coenzyme M reductase (mcrA) gene was amplified using the primer pairs ML-F (5′-GGTGGTGTMGGATTCACACARTAYGCWACAGC-3′) and ML-R (5′-TTCATTGCRTAGTTWGGRTAGTT-3′) [29]. For aerobic methanotrophs, we used the primer pairs A189F (5′-GGNGACTGGGACTTCTGG-3′) and mb661R (5′-CCGGMGCAACGTCYTTACC-3′) to quantify the particulate methane monooxygenase subunit α (pmoA) gene [30]. The primer pairs mmoX-206F (5′-ATCGCBAARGAATAYGCSCG-3′) and mmoX-886R (5′-ACCCANGGCTCGACYTTGAA-3′) were employed to target the soluble methane monooxygenase subunit α (mmoX) gene [30]. Additionally, the primer pairs Type e F (5′-CHCTGGAAGATCACTTCGGTGGTTC-3′) and Type e R (5′-RTATCCGAAGAARCCSAGTCKRCC-3′) were used to quantify the anaerobic methanotrophic archaea (ANME-3) gene [31]. The thermal profiles for different genes are detailed in Supplementary Table S2. Detailed qPCR methods are provided in Supplementary Text S1. High-throughput sequencing of bacterial and archaeal 16S rRNA genes and subsequent bioinformatics analyses were performed using BMK (Biomarker Technologies GmbH, Münster, Germany). Detailed methodologies are described in Supplementary Text S2. Calculation and statistical methods We treated wildfire as a pre-existing disturbance in the background and therefore did not include direct carbon losses during wildfire. Instead, we focused on how wildfire history influences GHG emissions following rewetting. Cumulative CO2 and CH4 emissions during incubation were calculated using the following equation. To compare their climate impacts, CH4 emissions were converted to CO2 equivalents using the 100-year global warming potential (GWP100) from the Intergovernmental Panel on Climate Change Sixth Assessment Report for biogenic CH4 (27.0 kg CO2-eq kg−1 CH4) [32]. We also conducted a sensitivity analysis using the 20-year global warming potential (GWP20 = 79.7 kg CO2-eq kg−1 CH4) to evaluate how the choice of time horizon influences the relative importance of CH4.Fcum=CtVhVmS×10where Fcum is the cumulative emission flux (mmol m−2); Ct represents the headspace concentration (μL L−1) of CO2 or CH4 on sampling day t; Vh is the headspace volume (L) at different water levels; Vm indicates the gas molar volume (L mol−1) at the standard temperature and atmospheric pressure; S is the surface area of the soil core (19.6 cm2). Because all peat columns were collected using the same coring dimensions and incubation setup, cumulative emissions were expressed on an area basis (mmol m−2) to facilitate treatment comparisons [33]. However, simulated fires could influence peat mass and bulk density among treatments; because peat dry mass and bulk density were not measured in this study, these differences may not have been fully reflected. All statistical analyses were conducted using SPSS (Version 22.0). One-way analysis of variance (ANOVA) was used to assess differences in GHG emissions and soil properties among treatments, with statistical significance defined at p < 0.05. Significant ANOVA results were followed up by Tukey's honestly significant difference post hoc tests. Venn diagrams were generated using the EVenn platform [34]. Functional predictions from 16S rRNA gene sequences were performed with PICRUSt via the Wekemo Bioincloud platform (https://www.bioincloud.tech) [35]. Microbial co-occurrence networks were constructed based on Spearman's correlation matrix from the operational taxonomic unit tables. These networks were visualized and analyzed for topological parameters using Gephi 0.9.7. Pearson correlation analysis was employed to explore the relationships between GHG emissions from postfire rewetting treatments and soil properties. Nonnormally distributed variables were log-transformed to approximate normality. Spatial analysis of fire-affected peatlands To examine the global distribution of fire-affected peatlands and potential post-rewetting CH4 emissions, we conducted annual overlay analyses between the global extent of peatlands and burned areas in 2019, 2020, and 2021. This multiyear approach accounted for interannual variability in wildfire occurrence. The global peatland layer was obtained from Global Peatland Map 2.0 [36], a high-resolution dataset used here as a proxy for peatland extent because no harmonized global dataset is available. Burned peatlands were identified using annual global burned-area shapefiles for 2019–2021 (the most recent available) from the MODIS MCD64A1 Collection 6.1 burned area product [37]. For each year, burned-area polygons were spatially overlaid with the peatland layer by using the Intersect tool in ArcGIS Pro (v3.1.0), generating year-specific maps of fire-affected peatlands. To estimate fire impact at the country and continent levels, we calculated the area of each intersected polygon and assigned geopolitical attributes through a spatial join with a global country boundary layer. For transcontinental countries (e.g., Russia), continent attribution was refined based on the centroid coordinates of each polygon. To restrict the analysis to peatlands where rewetting is likely to be relevant, fire-affected peatland areas were further adjusted using continent-specific peatland degradation ratios [36], thereby excluding intact peatlands unlikely to require restoration. In addition, because wildfire severity cannot be reliably determined from currently available global remote sensing products in peatlands, we adopted a scenario-based approach. We assumed that 25%, 50%, 75%, or 100% of the fire-affected peatland area experienced wildfire severities comparable to those of the severe-fire treatment applied in the incubation experiments (i.e., a 10-min simulated fire). For each scenario, the per-area mitigation potential derived from the laboratory incubation results was extrapolated linearly to the corresponding fire-affected peatland area (Supplementary Text S3a). The resulting estimates should be interpreted as order-of-magnitude approximations intended to illustrate the potential contribution of prioritizing severely fire-affected peatlands to climate-effective rewetting rather than predictive assessments. Detailed continental distributions of fire-affected degraded peatlands and the corresponding potential CH4 mitigation are provided in Supplementary Tables S3 and S4. Results GHG emissions Cumulative soil-derived CO2 and CH4 emissions differed significantly among treatments during the incubation period (Fig. 1a and b). At the end of incubation, drained soils exhibited the highest cumulative CO2 emissions (3022 ± 185 mmol CO2 m−2, Fig. 1a). Compared with drained soils (control), emissions decreased to 445 ± 251 mmol CO2 m−2 in rewetted non-fire-affected soils (−85% [95% confidence interval (CI): −95% to −76%], p < 0.05), 962 ± 306 mmol CO2 m−2 in rewetted mildly fire-affected soils (−68% [95% CI: −80% to −57%], p < 0.05), and 304 ± 72 mmol CO2 m−2 in rewetted severely fire-affected soils (−90% [95% CI: −93% to −87%], p < 0.05). Compared with rewetted non-fire-affected soils, cumulative CO2 emissions were 116% higher in rewetted mildly fire-affected soils (95% CI: −42% to 275%; p < 0.05) and 32% lower in rewetted severely fire-affected soils (95% CI: −79% to 16%; Fig. 1a).Fig. 1Greenhouse-gas emissions and global warming potential across drainage and rewetting treatments. a–b, Cumulative CO2 (a) and CH4 (b) emissions of different treatments during the 90-day incubation period. Symbols represent the measured cumulative emissions at each sampling time, solid lines connect the mean values, and shaded bands indicate the standard error around the mean (n = 3). c, Contributions of cumulative CO2 and CH4 emissions to the 100-year global warming potential (GWP100 = 27.0 for biogenic CH4), expressed as CO2 equivalents. d, Total GWP100 of cumulative emissions across treatments. Bars represent the mean values, and error bars indicate the standard error (n = 3). Treatments include drained soils (DS), rewetted non-fire-affected soils (RNFS), rewetted mildly fire-affected soils (RMFS), and rewetted severely fire-affected soils (RSFS). Corresponding estimates based on the 20-year GWP (GWP20 = 79.7) are provided in Supplementary Fig. S1. Greenhouse-gas emissions and global warming potential across drainage and rewetting treatments. a–b, Cumulative CO2 (a) and CH4 (b) emissions of different treatments during the 90-day incubation period. Symbols represent the measured cumulative emissions at each sampling time, solid lines connect the mean values, and shaded bands indicate the standard error around the mean (n = 3). c, Contributions of cumulative CO2 and CH4 emissions to the 100-year global warming potential (GWP100 = 27.0 for biogenic CH4), expressed as CO2 equivalents. d, Total GWP100 of cumulative emissions across treatments. Bars represent the mean values, and error bars indicate the standard error (n = 3). Treatments include drained soils (DS), rewetted non-fire-affected soils (RNFS), rewetted mildly fire-affected soils (RMFS), and rewetted severely fire-affected soils (RSFS). Corresponding estimates based on the 20-year GWP (GWP20 = 79.7) are provided in Supplementary Fig. S1. CH4 emissions differed considerably among treatments (Fig. 1b). They were significantly higher in rewetted non-fire-affected soils (46.5 ± 15.7 mmol CH4 m−2) than in drained soils (1.2 ± 0.7 mmol CH4 m−2, p < 0.05). Rewetted mildly fire-affected soils exhibited an even larger increase, reaching 281 ± 84 mmol CH4 m−2. In contrast, rewetted severely fire-affected soils maintained relatively low CH4 emissions of 4.4 ± 1.2 mmol CH4 m−2, showing no significant difference from the drained control (p > 0.05). Compared with rewetted non-fire-affected soils, rewetted mildly fire-affected soils had 504% higher cumulative CH4 emissions (95% CI: 196% to 813%; p < 0.05), whereas rewetted severely fire-affected soils had 91% lower cumulative CH4 emissions by the end of the incubation (95% CI: 86% to 95%; p < 0.05, Fig. 1b). Regarding the soil-derived GHG emissions expressed in CO2 equivalents, the treatments exhibited clear differences (Fig. 1c and d). Compared with the drained soils, rewetted non-fire-affected exhibited a significantly lower total GWP100 (p < 0.05, Fig. 1d), as the increase in CH4 emissions was more than offset by a strong suppression of CO2 release (Fig. 1c). However, in rewetted mildly fire-affected soils, CH4-derived CO2-equivalent emissions increased sharply, resulting in a high total GWP100 that was not significantly different from that of the drained soils (p > 0.05, Fig. 1d). In contrast, rewetted severely fire-affected soils achieved the lowest total GWP100 across treatments (Fig. 1d), which was attributed to a strong suppression of CH4 release in addition to reduced CO2 emissions (Fig. 1c). Consequently, compared with rewetted non-fire-affected, rewetted severely fire-affected soils significantly reduced total GWP100 by approximately 65% (95% CI: −72% to −57%; p < 0.05), whereas rewetted mildly fire-affected soils showed a significant 287% increase (95% CI: 182% to 392%; p < 0.05, Fig. 1d). A complementary analysis using GWP20 produced the same ranking of treatments but with amplified differences, reflecting the greater weight of CH4 in near-term climate forcing (Supplementary Fig. S1). Soil physicochemical properties After incubation, soil pH decreased significantly in rewetted non-fire-affected and mildly fire-affected soils (p < 0.05, Supplementary Table S5) compared with drained soils, whereas rewetted severely fire-affected soils maintained comparable pH levels (p > 0.05). A similar trend was observed for electrical conductivity. Total carbon and total organic carbon were slightly but nonsignificantly lower in rewetted severely fire-affected soils than in other soils. Compared with treatments that did not undergo fire (i.e., drained soils and rewetted non-fire-affected soils), ammonium nitrogen levels increased significantly in rewetted mildly (p < 0.05) and severely fire-affected soils (p < 0.05), with the latter exhibiting the highest values across treatments. Relative to non-fire-affected soils, severely fire-affected soils showed higher FTIR peak heights of functional groups associated with recalcitrant components, such as phenols at approximately 1440 cm−1 and aromatics at approximately 1510 cm−1 and approximately 1670 cm−1. Mildly fire-affected soils exhibited an increase only at approximately 1670 cm−1 (Supplementary Fig. S2). Soil microorganisms Microbial community composition varied distinctly across treatments for both bacteria and archaea (Supplementary Fig. S3). Rewetted severely fire-affected soils contained the highest number of unique bacterial and archaeal operational taxonomic units (Supplementary Fig. S3a and b), yet most dominant bacterial families and methanogenic archaea declined compared with other rewetted soils (Supplementary Fig. S3c and d). In contrast, rewetted mildly fire-affected soils exhibited enriched methanogenic families (Methanobacteriaceae and Methanosarcinaceae) (Supplementary Fig. S3d) and more connected microbial co-occurrence networks (Supplementary Fig. S4 and Table S6), consistent with their higher GHG emissions compared with other rewetted treatments. Predicted functional profiles indicated that among the rewetted treatments, the highest overall metabolic potential was observed in mildly fire-affected soils and the lowest in severely affected soils (Supplementary Fig. S5). At the enzyme level, these differences translated into clear shifts in CH4-related pathways, including methanogenesis (Fig. 2a) and CH4 oxidation (Fig. 2b). Compared with drained soils, rewetted non-fire-affected and mildly fire-affected soils exhibited elevated abundances of core methanogenesis enzymes (e.g., Mtr, Mcr, and Hdr), with rewetted mildly fire-affected soils generally highest (Fig. 2a). In contrast, rewetted severely fire-affected soilssoils exhibited the lowest levels of these core enzymes across treatments, with the acetoclastic methanogenesis enzymes Ack and Pta being lower even those of drained soils (Fig. 2a). Regarding CH4 oxidation pathways (Fig. 2b), the predicted abundances of key enzymes, particularly those in the reverse methanogenesis pathway, were lower in rewetted severely fire-affected soils than in other rewetted soils.Fig. 2Methane-cycling pathways and functional-gene abundances involved in CH4 cycling across treatments. a–b, Relative abundance of key enzymes in CH4 metabolism pathways annotated using the Kyoto Encyclopedia of Genes and Genomes, including methanogenesis (hydrogenotrophic and acetoclastic; a) and CH4 oxidation (aerobic and anaerobic; b). Blue boxes indicate the key enzymes (or enzyme complexes) involved in each metabolic step. Values show abundance ratios for RNFS, RMFS, and RSFS relative to DS; ratios >1 (red) indicate higher abundance, where ratios <1 (green) indicate lower abundance. The absolute abundance of these enzymes is shown in Supplementary Fig. S6, and abbreviations are listed in Supplementary Table S7c–f, Abundance of functional genes, including the methyl coenzyme M reductase (mcrA) gene for methanogenesis (c), the particulate methane monooxygenase subunit α (pmoA) gene (aerobic CH4 oxidation; d), the soluble methane monooxygenase subunit α (mmoX) gene (aerobic CH4 oxidation; e), and the anaerobic methanotrophic archaea (ANME-3) gene (f), at the end of incubation. Data are presented as means ± standard error (n = 3). Different lowercase letters indicate statistically significant differences among treatments (p < 0.05). DS, drained soils; RNFS, rewetted non-fire-affected soils; RMFS, rewetted mildly fire-affected soils; RSFS, rewetted severely fire-affected soils. Methane-cycling pathways and functional-gene abundances involved in CH4 cycling across treatments. a–b, Relative abundance of key enzymes in CH4 metabolism pathways annotated using the Kyoto Encyclopedia of Genes and Genomes, including methanogenesis (hydrogenotrophic and acetoclastic; a) and CH4 oxidation (aerobic and anaerobic; b). Blue boxes indicate the key enzymes (or enzyme complexes) involved in each metabolic step. Values show abundance ratios for RNFS, RMFS, and RSFS relative to DS; ratios >1 (red) indicate higher abundance, where ratios <1 (green) indicate lower abundance. The absolute abundance of these enzymes is shown in Supplementary Fig. S6, and abbreviations are listed in Supplementary Table S7c–f, Abundance of functional genes, including the methyl coenzyme M reductase (mcrA) gene for methanogenesis (c), the particulate methane monooxygenase subunit α (pmoA) gene (aerobic CH4 oxidation; d), the soluble methane monooxygenase subunit α (mmoX) gene (aerobic CH4 oxidation; e), and the anaerobic methanotrophic archaea (ANME-3) gene (f), at the end of incubation. Data are presented as means ± standard error (n = 3). Different lowercase letters indicate statistically significant differences among treatments (p < 0.05). DS, drained soils; RNFS, rewetted non-fire-affected soils; RMFS, rewetted mildly fire-affected soils; RSFS, rewetted severely fire-affected soils. qPCR-based measurements revealed treatment effects on the functional groups that mediate CH4 production and oxidation. Regarding mcrA, which is responsible for CH4 production, both rewetted non-fire-affected and mildly fire-affected soils exhibited higher abundances than drained soils, with a significantly higher abundance in mildly fire-affected soils (p < 0.05, Fig. 2c). In contrast, rewetted severely fire-affected soils exhibited the lowest abundance (p < 0.05, Fig. 2c). For CH4 oxidation, pmoA was most abundant in drained soils but declined after rewetting, reaching the lowest levels in rewetted severely fire-affected soils (Fig. 2d). Similarly, mmoX was lowest in rewetted severely burned soils (Fig. 2e). The anaerobic CH4-oxidizing group ANME-3 also declined after simulated fires, with a slight decrease under mild fire but a significant decrease under severe fire (p < 0.05, Fig. 2f). Discussions Wildfires cause widespread damage to biodiversity and ecosystem services, including climate, hydrology, and nutrient regulation, and result in substantial loss of life and property [38]. In agricultural landscapes, in particular, wildfires destroy livestock and crops, degrade soil structure, reduce fertility, and diminish yield potential, all of which are not easily reversible and are costly [17,18]. These productivity-declining conditions and tangible fire risks can cause landowners to accept rewetting as a management option. Moderate water table restoration reduces flammability and fire recurrence, particularly in regions where drained peatlands have experienced repeated burning [39]. However, whether wildfire-affected peatlands also deliver distinct climate benefits after rewetting remains unclear, a critical gap for both stakeholder acceptance and the effective scaling of restoration strategies. Fire effects on post-rewetting CO2 emissions Across all treatments, rewetting reduced CO2 emissions relative to drained soils, consistent with the development of anoxic conditions that inhibit aerobic carbon decomposition [40]. However, the magnitude of CO2 emission suppression varied with fire severity. Rewetted mildly fire-affected soils exhibited weaker CO2 emission reductions (−68%) than rewetted non-fire-affected (−85%) and severely fire-affected soils (−90%) (Fig. 1a). This diminished mitigation effect was statistically significant when comparing rewetted mildly and severely fire-affected soils (Fig. 1c). This CO2 emission pattern suggests that mild fires may partially offset the CO2 mitigation capacity of rewetting. Although not measured directly, this may be related to greater preservation or release of labile carbon substrates in mildly fire-affected soils, as observed in permafrost peatlands [41]. Our FTIR analysis indirectly indicated that, unlike severe fires, mild wildfires did not clearly increase the recalcitrant component of soil carbon (Supplementary Fig. S2), but instead potentially caused superficial drying without substantial carbonization [42]. During drying, the vaporization of soil pore water exerted pressure that exceeded the yield strength of soil aggregates, leading to their structural breakdown, which can release previously protected labile carbon [43]. This is further supported by the enrichment of metabolic pathways related to organic matter degradation observed in rewetted mildly fire-affected soils (Supplementary Fig. S5). Low-severity burning can stimulate soil respiration by increasing the availability of readily decomposable carbon substrates after fire [41,44,45]. Fire-induced shifts in the microbial community structure may further accelerate soil carbon decomposition and CO2 release [46], likely also through increased nutrient availability and stimulated microbial activity [41,47]. In contrast, more severe fires may suppress microbial decomposition through greater thermal disturbance, particularly by reducing fungal decomposers that break down more recalcitrant organic matter [45]. However, in our study, both mildly and severely fire-affected soils exhibited elevated NH4+-N contents, whereas NO3−-N did not change significantly (Supplementary Table S5). This result suggested that differences in nitrogen availability alone are unlikely to account for the contrasting CO2 responses between mild and severe fire treatments. Thus, mild fire treatments may maintain more favorable conditions for post-rewetting microbial decomposition than severe fire treatments. Fire effects on post-rewetting CH4 emissions Although fire-induced soil changes did not override CO2 suppression following rewetting, CH4 emissions diverged far more sharply across the rewetted soils. Rewetting non-fire-affected soils resulted in an approximately 40-fold increase in CH4 fluxes compared with drained soils (Fig. 1b), consistent with the well-documented stimulation of methanogenesis under raised water tables during peatland restoration [48]. This post-rewetting CH4 surge indicates that rewetting fire-affected peatlands warrants careful evaluation. High CH4 fluxes can dominate the GHG balance in the initial months after rewetting, as observed here, and may persist for years [40] or even decades thereafter [49,50], thereby delaying or even negating the cooling benefits of peatland restoration for decades [51]. Fire severity profoundly reshaped this response. Our results revealed that rewetting mildly fire-affected soils resulted in a six-fold surge in CH4 emissions relative to rewetted non-fire-affect soils, whereas rewetting severely fire-affected soils exhibited a 91% decrease (Fig. 1b). Similar CH4 suppression following severe fires has also been observed in the field: in boreal bogs, hollow sites with deep burns (>9 cm) maintained CH4 fluxes that remained far lower than those of unburned hollows even 1−2 years after fire, despite water tables (−9.0 ± 0.9 cm in burned sites versus −6.9 ± 1.8 cm in unburned sites) being similarly close to the surface and within the range typically considered favorable for post-rewetting CH4 generation [52]. This indicates that severe fire-induced soil transformations may override the expected increase in CH4 emissions typically associated with elevated water tables after rewetting. These findings raise two questions: why CH4 responses diverge across fire severities, and how the biogeochemical legacies of fire should inform rewetting strategies. A key factor contributing to these divergent CH4 responses is the altered stability and bioavailability of soil organic carbon, as methanogenesis becomes carbon-substrate-limited once the water table reaches the level necessary to favor anoxic conditions [53]. In our study, severe fires clearly enhanced the proportion of chemical recalcitrant carbon components, such as phenols and aromatics (Supplementary Fig. S2). FTIR spectroscopy can capture only broad functional-group changes and lacks the molecular-level resolution achievable with more advanced techniques, such as Fourier-transform ion cyclotron resonance mass spectrometry [54]. Nevertheless, thermal alteration from severe fires can promote the condensation and cyclization of aliphatic compounds into more aromatic and polymerized structures, thereby increasing chemical recalcitrance [55]. These thermally modified compounds are less bioavailable and thus less likely to fuel methanogenesis in rewetted severely fire-affected soils and may even suppress methanogenesis by diverting electron flow away from CH4 production toward anaerobic respiration [56]. Shifts in soil pH and electrical conductivity may also have contributed to the observed CH4 patterns, as both were negatively correlated with CH4 emissions (Fig. 3). Severe fires increased both variables (Supplementary Table S5 and Fig. S7), probably reflecting ash deposition and the release of alkaline and soluble ions during organic matter combustion [57,58]. Mild fires did not induce comparable physicochemical changes. Previous studies have reported enhanced soil CH4 uptake following post-fire increases in pH, although the underlying mechanisms remain unclear [59]. In our study, these physicochemical shifts may have further influenced the microbial processes associated with CH4 cycling, potentially contributing to the lower CH4 emissions observed in severely fire-affected soils.Fig. 3Associations among greenhouse-gas emissions, microbial markers and soil properties. Pearson correlation matrix of greenhouse gas emissions, functional gene abundances, fire severity, and soil physicochemical properties across the rewetting treatments. Colors denote Pearson's correlation coefficients, and the size of the white squares is proportional to the absolute value of the correlation coefficient. Asterisks denote statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. TC, total carbon; TN, total nitrogen; TOC, total organic carbon; EC, electrical conductivity. Remaining abbreviations are explained in Supplementary Table S7. Associations among greenhouse-gas emissions, microbial markers and soil properties. Pearson correlation matrix of greenhouse gas emissions, functional gene abundances, fire severity, and soil physicochemical properties across the rewetting treatments. Colors denote Pearson's correlation coefficients, and the size of the white squares is proportional to the absolute value of the correlation coefficient. Asterisks denote statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. TC, total carbon; TN, total nitrogen; TOC, total organic carbon; EC, electrical conductivity. Remaining abbreviations are explained in Supplementary Table S7. CH4 emissions were also positively correlated with the methanogenic gene mcrA (Fig. 3). Given that mild fires caused neither substantial shifts in pH and electrical conductivity nor carbon stabilization, yet potentially increased labile carbon substrate, this likely maintained or even enhanced ecological niches favorable for methanogenic communities in the rewetted mildly fire-affected soils [60,61]. This is supported by both a higher mcrA abundance (Fig. 2c) and a more interconnected microbial network (Supplementary Fig. S4b and Table S6). In contrast, severe fires altered soil physicochemical properties and likely imposed direct thermal stress on microbial populations [62], thereby constraining the abundance and activity of methanogenic communities. This is further evidenced by a significant reduction in mcrA abundance (Fig. 2c) and a fragmented microbial network topology in rewetted severely burned soils (Supplementary Fig. S4c and Table S6). Our analyses focused on bacterial and archaeal communities. Because wildfire also affects fungal populations that degrade condensed aromatic compounds and mediate postfire carbon turnover [63], incorporating fungal sequencing or metagenomic profiling would further refine these microbial insights. Taken together, these results suggest that the interaction between wildfire severity and rewetting fundamentally alters the biogeochemical pathways that drive GHG emissions. Mild fires, due to their limited alteration of soil conditions, may preserve favorable environments for methanogenesis after rewetting. In contrast, severe fires likely suppress methanogenesis through stronger thermal disturbance, increased carbon recalcitrance, and clear shifts in soil biogeochemical conditions. These factors underlie the post-fire suppression of CH4 emissions across ecosystems [52,59,64]. Consequently, severely fire-affected soils may somewhat unexpectedly exhibit greater climate mitigation potential following rewetting. Scaling potential and associated uncertainties Our findings suggest that rewetting severely fire-affected peatlands can lead to an additional 91% reduction in post-rewetting CH4 emissions compared with rewetted non-fire-affected soils. We explicitly evaluated wildfire-affected soils within a peatland rewetting framework and found that the identified underlying mechanisms are consistent with established responses of peatlands to fire. Consistent signals were observed in our previous burning experiment conducted under extensive management in Denmark [65] and in a boreal bog in Canada [52], where post-fire water table rise coincided with suppressed CH4 emissions. Our findings highlight the central mechanism of reduced bioavailability of soil carbon substrates after wildfires, which has been widely documented [23,55,66], and its constraining effect on GHG emissions has been observed across a broad climatic gradient [22]. These findings suggest that fire-induced suppression of methanogenesis may represent a general response to postfire rewetting. We therefore evaluated the broader climate benefit that could arise if wildfire-affected soils are explicitly prioritized in rewetting strategies. Global mapping indicated that more than 6 million hectares of peatlands were affected annually by wildfires, of which 0.5–0.9 million hectares corresponded to fire-affected degraded peatlands, representing realistic candidates for rewetting (Supplementary Table S3). More than 70% of these candidate areas were concentrated in Asia, particularly Asian Russia, Indonesia, Kazakhstan, and China (Fig. 4). Other potential regions included parts of Africa (e.g., Zambia and Angola), Europe (e.g., Romania and the United Kingdom), South America (e.g., Brazil and Venezuela), North America (e.g., Canada and the United States), and Oceania (e.g., Australia).Fig. 4Global distribution of fire-affected degraded peatlands and potential CH4 mitigation from prioritized rewetting during 2019–2021. a, Year 2019; b, Year 2020; c, Year 2021. Presenting multiple years accounts for interannual variability in wildfire occurrence. For each continent, the four countries with the largest areas of fire-affected degraded peatland are labeled. Adjacent vertical bars represent the estimated areas of fire-affected degraded peatlands in each country, with bar height scaled by area. The bar for Asian Russia provides the full-height reference: a full-height dark blue bar represents 150,000 ha, whereas a full-height dark red bar represents 30,000 ha. Shorter bars represent proportionally smaller areas. Donut charts for each continent show each continent's share of the global fire-affected degraded peatland area, with the orange segment indicating the continental contribution. Insets show the estimated CH4 mitigation for rewetting shares of 25%, 50%, 75%, and 100%, expressed using the 100-year and 20-year global warming potentials (GWP100 and GWP20), respectively. Detailed numerical estimates are provided in Supplementary Tables S3 and S4. DRC, Democratic Republic of the Congo; RoC, Republic of the Congo; UK, United Kingdom; USA, United States of America. Global distribution of fire-affected degraded peatlands and potential CH4 mitigation from prioritized rewetting during 2019–2021. a, Year 2019; b, Year 2020; c, Year 2021. Presenting multiple years accounts for interannual variability in wildfire occurrence. For each continent, the four countries with the largest areas of fire-affected degraded peatland are labeled. Adjacent vertical bars represent the estimated areas of fire-affected degraded peatlands in each country, with bar height scaled by area. The bar for Asian Russia provides the full-height reference: a full-height dark blue bar represents 150,000 ha, whereas a full-height dark red bar represents 30,000 ha. Shorter bars represent proportionally smaller areas. Donut charts for each continent show each continent's share of the global fire-affected degraded peatland area, with the orange segment indicating the continental contribution. Insets show the estimated CH4 mitigation for rewetting shares of 25%, 50%, 75%, and 100%, expressed using the 100-year and 20-year global warming potentials (GWP100 and GWP20), respectively. Detailed numerical estimates are provided in Supplementary Tables S3 and S4. DRC, Democratic Republic of the Congo; RoC, Republic of the Congo; UK, United Kingdom; USA, United States of America. Upscaling based on our CH4 reduction indicated that prioritizing rewetting in these areas could achieve CH4 mitigation of 0.1–0.8 Mt CO2-eq yr−1 under GWP100 and 0.4–2.2 Mt CO2-eq yr−1 under GWP20, assuming rewetting shares of 25–100% (Supplementary Table S4). Because wildfires in degraded peatlands can persist for days to months, our conservative assumption that 25% of fire-affected areas experienced fire severity comparable to the 10-min experimental burn was unlikely to overestimate real-world conditions. Even so, neither global fire-severity distributions nor land-use–specific distributions of degraded peatlands are resolved at the global scale; therefore, these mitigation estimates should be interpreted as a conceptual approximation of what climate benefits from prioritizing fire-affected peatlands could achieve, rather than as a direct extrapolation to all degraded peatlands. The absolute CH4 mitigation potential identified here appears limited relative to global net-zero targets, but its value lies in the feasibility and speed of implementation. Unlike rewetting on actively productive land, interventions in fire-affected peatlands can be framed as recovery and risk-reduction measures rather than as competing land-use claims. Because burned land has already lost much of its agricultural value, this approach avoids much of the land-use conflict that has historically slowed restoration on productive peatland. Indeed, wildfire has already prompted national rewetting and governance initiatives in several regions, including Indonesia [67], Malaysia [68], Russia [69], and the United States [70]. Large-scale practical application of these insights nevertheless remains challenging. Natural wildfires vary widely in frequency, intensity, and duration, and not all fire-affected peatlands experience biogeochemical responses comparable to those observed under simulated burning. In addition, deeper peat layers and lateral hydrological connectivity may continue to support CH4 production after rewetting under field conditions, processes that may not be fully captured by our incubation experiment using short soil cores. However, post-rewetting vegetation succession may alter overall GHG dynamics beyond the soil-derived fluxes captured in our study. Long-term field studies are essential to elucidate the relationship between actual fire conditions and post-rewetting GHG emissions. Rewetting severely fire-affected peatlands may mobilize nutrients [71] and contaminants [72], which can be transported into nearby water bodies, affecting aquatic ecosystems through nutrient enrichment, shifts in microbial communities, or the introduction of toxic compounds [73]. To understand these environmental implications, future studies should adopt a catchment-scale perspective that connects terrestrial changes to downstream aquatic responses. In summary, severely fire-affected peatlands may serve as low-hanging fruit, both biogeochemically and socioeconomically, for accelerating peatland rewetting. As wildfire frequency increases and restoration progress remains slow, identifying such leverage points is essential. However, long-term field-based studies are required to determine whether the observed GHG responses reflect transient fluctuations or enduring shifts in ecosystem function, while also accounting for broader ecological impacts. Prioritization frameworks should integrate biogeochemical evidence, recent wildfire footprints, and land tenure considerations to ensure that future rewetting efforts are both climatically effective and practicable. CRediT authorship contribution statement Shihao Cui: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Conceptualization. Haonan Guo: Writing – review & editing. Lorenzo Pugliese: Writing – review & editing, Supervision. Shubiao Wu: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.