Introduction

Methane (CH4) in the atmosphere contributes ∼16–25% of global warming to date1,2. Importantly, over the past decade, there has been a continuous increase in atmospheric CH4 concentration, coupled with notable variations in growth rates3,4,5,6. Isotope fractionation-based analysis suggests that biogenic CH4 emissions may strongly contribute to this dynamic7,8,9,10. Wetlands, as the largest natural biogenic source of CH411, are considered as the main contributor to the recent atmospheric CH4 growth rate anomaly12,13. However, their emissions exhibit high uncertainty and vary across spatial and temporal scales14,15.

Variations in wetland CH4 emissions are also implicated as the primary driver of observed interannual variability in global CH4 fluxes16,17. Critical gaps arise from the fact that existing observational networks and their methodologies, whether bottom-up approaches based on process modeling or top-down approaches based on atmospheric inversions, often exhibit substantial differences. This variance makes it challenging to pinpoint specific changes in CH4 sources at large scales, particularly when dealing with conditions of hydrological fluctuations4,18,19,20. Indeed, diurnal, seasonal, and annual fluctuations in meteorological factors such as air temperature, precipitation, and evapotranspiration21,22, along with extreme events under climate change conditions like flooding and droughts23, result in dynamic hydrological processes at the regional or catchment scale. These variations introduce considerable uncertainties in large-scale CH4 emission assessments, especially for wetlands in different climatic zones, thus complicating the understanding and prediction of atmospheric CH4 dynamics5. Nowadays, advancements, such as the use of ecosystem-scale eddy covariance measurement systems, enable continuous monitoring and capture of changes in CH4 fluxes across various time scales, particularly under disturbances caused by hydrological, meteorological, and management activities24,25,26,27. This may help us understand the response of CH4 fluxes to hydrological fluctuations.

Despite longstanding knowledge about the limitations of microbial thermodynamics and energetics, traditional wetland CH4 flux predictions rigidly assign CH4 production to the anoxic zone below the water table and CH4 oxidation to the oxic zone above it28,29. Recent discoveries in microbial CH4 processes over the past decade are challenging these boundaries. For instance, it is now understood that methanogenesis can take place in the microaerobic zone, often referred to as the methane paradox, although its presence in freshwater wetlands remains somewhat speculative30,31,32. Additionally, CH4 oxidation can largely occur in reductive zones using alternative electron acceptors occurring in wetlands, such as nitrate33, nitrite34, sulfate35, iron36, and more, from both autochthonous and allochthonous sources (e.g., anthropogenic eutrophication37,38, saltwater intrusion39, and atmospheric deposition40). Of greater significance, variations in wetland hydrological conditions at different time scales (spanning from hourly to annual)41, as well as changes in intensity of period, amplitude, and the ratio of oxic to anoxic periods42, can lead to even more intricate and nonlinear responses in CH4 fluxes. Moreover, climate warming influences wetland CH4 cycle not only through precipitation-driven hydrological fluctuations, but also through directly influencing the kinetics of CH4 biogeochemical reactions due to rising temperatures43 and indirectly affecting plant community composition and plant traits such as biomass44,45. This complexity adds an additional layer of uncertainty, which in turn diminishes our ability to accurately assess the impact of wetlands on climate change. Thus, a complete understanding of the connection between hydrological fluctuations and the CH4 cycle in wetlands, influenced by both biotic and abiotic environmental factors, is lacking. This knowledge gap poses a substantial barrier to accurately reporting CH4 emissions and implementing effective mitigation measures, especially in the face of climate change, marked by increasingly frequent extreme events.

In this review, we aim to bridge this gap by: (i) presenting a comprehensive exploration of the intricate relationship between CH4 flux and hydrological dynamics, and (ii) providing a detailed analysis of the impacts of hydrological fluctuations on the CH4 cycle, along with the underlying mechanisms. By shedding light on these aspects, we hope to contribute to a more comprehensive understanding of the complex interplay between wetland hydrology and CH4 emissions, facilitating improved management and mitigation in the context of our changing climate.

Hydrological fluctuations in wetland ecosystems

Prevalence of hydrological fluctuations

Wetlands, serving as the typical transitional zones between the permanent wet environment of aquatic ecosystems and the dry environment of terrestrial ecosystems, are susceptible to longitudinal, lateral, and vertical hydrological fluctuations46,47,48,49. These dynamics are shaped by the interplay of hydrometeorological factors (e.g., precipitation and evapotranspiration) and key topographic and hydrogeologic features (e.g., elevation, microtopography, and connections to groundwater and adjacent water systems)50.

Climate change adds a layer of complexity to wetland hydrology. Worldwide, rising temperatures are amplifying both the overall precipitation and its variability, fundamentally altering the hydrological dynamics within wetlands51. Notably, in tropical regions, meteorological changes at a large scale are accentuating the hydrological cycle of tropical wetlands52,53,54,55,56. This phenomenon is triggering hydroclimatic events, such as seasonal rainfall anomalies, fostering the expansion in certain tropical wetlands57 while others experience loss58. Overall, in a warmer future, the increased evaporative demand may outweigh changes in precipitation, potentially leading to a substantial reduction (∼34,000 km2) in tropical wetland area by the end of the century59. North American inland wetlands are also expected to experience a 5–10% reduction in average area by the end of the century compared to historical contemporaries60. Moreover, the accelerated glacial melting (including ice caps and mountain glaciers) due to global warming greatly alters water inputs to wetlands61, while thawing permafrost can increase the hydrogeological connectivity of wetlands by opening up previously blocked vertical and lateral flow pathways, in addition to recharging groundwater and reducing surface runoff62,63,64. It is also important to note that glacier melting and thermal expansion of oceans are driving sea level rise65, which is projected to increase 0.26–0.82 m by 210066,67. Although wetlands can naturally keep pace with changes in sea level through accretion processes, climate change may upset this balance, accelerating water table rise in coastal and estuarine wetlands, and eventually resulting in wetlands being submerged, pushed inland or covered with salt.

Beyond these natural factors, anthropogenic activities such as land-use change68, groundwater extraction69, dam construction70, and petroleum exploration71,72,73,74 can also exert strong impacts on the water balance in wetlands. In particular, land-use change has led to the long-term degradation of 21% of wetlands globally from 1700 to 2020 due to drainage primarily for intensive agriculture development75. However, large-scale wetland restoration and conservation initiatives are working to re-raise water table76. As a result, hydrological fluctuations have become a prevalent occurrence in wetlands under the influence of natural processes, climate change, and anthropogenic activities.

Patterns of hydrological fluctuations

Hydrological fluctuation patterns in wetlands vary widely across wetlands due to the spatiotemporal variations of different hydrometeorological and hydrogeologic factors77. Two key variables, amplitude, and periodicity/frequency are commonly used to characterize hydrological fluctuations in wetlands23. Seasonal fluctuations in wetlands can vary widely, ranging from several tens of centimeters78,79 to over a meter80,81,82, with the extent largely depending on climate conditions and wetland types. Tropical wetlands, for instance, often experience more dramatic fluctuations compared to boreal wetlands83, primarily due to pronounce seasonal rainfall patterns within tropical regions. Additionally, rainfall patterns in wetlands are subject to the impact of interannual and interdecadal climate fluctuations, further amplifying the variability in fluctuation amplitudes84. By synthesizing water table data from 31 wetland sites in the FLUXNET database (including six bog sites, seven fen sites, three swamp sites, seven marsh sites, three salt marsh sites, and five rice sites)26,27, Fig. 1 illustrates differences in the water table fluctuation amplitudes and density distributions of water table data across various wetland types85,86, typically attributed to variations in water supply pathways87, proximity to water sources88, vegetation communities89, and soil properties90.

Fig. 1: Water table fluctuations in six wetland types.
figure 1

These wetland sites include six bog sites with 6717 water table observations (green), seven fen sites with 10,679 water table observations (orange), three swamp sites with 1543 water table observations (blue), seven marsh sites with 9108 water table observations (pink), three salt marsh sites with 947 water table observations (light green), and five rice field sites with 3495 water table observations (yellow). Data are daily mean values and were collected from FLUXNET-CH4 database26,27. Site information is in Supplementary Table 1. The raincloud plot for each wetland type includes a jitter plot showing individual data points (colored dots), a half-violin plot presenting density distributions of data (shaded areas), and a box plot illustrating summary statistics (boxes with whiskers). The box plots show the median (line within the box), interquartile range (box edges), and whiskers (1.5 times the interquartile range). Note: positive values of the water table represent distance below the ground surface and negative values represent surface flooding.

Full size image

Moreover, water regime can determine wetland vegetation structure and composition. In turn, vegetation can influence wetland water balance through rainfall interception, stemflow, and transpiration91. Besides, differences in vegetation ecological characteristics and adaptation strategies can further impact hydrological fluctuations. For instance, the encroachment of woody shrubs (e.g., Carolina willow) into graminoid-dominated subtropical wetlands can increase transpiration water loss, which may drive a sustained water table decline92. In tropical papyrus wetlands, the papyrus mat system can enhance water storage capacity during the rainy season, allowing it to withstand high water volume change and water table fluctuations93. Additionally, Sphagnum can form a dense cover on some boreal wetlands, with its unique cellular structure enabling it to absorb and retain water, acting like a sponge94. This can impede the infiltration of low-intensity rainfall and thus influence the fluctuation amplitude95.

Water table in wetlands generally exhibits periodic fluctuations of different intensities over various time scales, attributable to diurnal, seasonal, and annual water inputs and outputs. Diurnal water table fluctuations, usually a few centimeters, are generally described as a function of evapotranspiration and periodic changes in atmospheric pressure96,97. During the summer, wetlands often experience active evapotranspiration throughout the day, leading to a typical pattern of gradual decline in water table starting at dawn (Supplementary Fig. 1)98. Seasonal water table fluctuations are primarily governed by the temporal dynamics of precipitation and temperature. Water table rises substantially during wet seasons such as the rainy season or snowmelt, and declines accordingly during dry seasons when precipitation decreases and evapotranspiration increases (Fig. 2a, b). This implies that the seasonal fluctuation pattern of water table is closely related to the local climatic conditions at the time.

Fig. 2: Seasonal water table and soil temperature fluctuations with corresponding CH4 emission patterns.
figure 2

a Seasonal water table fluctuations (blue circles) and corresponding CH4 emission patterns (bluish green circles) at the SE-Deg site (Västerbotten, Sweden) in 2016. b Seasonal water table fluctuations (blue circles) and corresponding CH4 emission patterns (bluish green circles) at the FI-Sii site (Pirkanmaa, Finland) in 2013. c Seasonal soil temperature fluctuations (vermillion circles) and corresponding CH4 emission patterns (bluish green circles) at the SE-Deg site (Västerbotten, Sweden) in 2016. d Seasonal soil temperature fluctuations (vermillion circles) and corresponding CH4 emission patterns (bluish green circles) at the FI-Sii site (Pirkanmaa, Finland) in 2013. Site information is in Supplementary Table 1. Note: positive values of the water table represent distance below the ground surface and negative values represent surface flooding. Soil temperature probe depth at SE-Deg and FI-Sii were 0.05 cm and 0.02 cm below the surface, respectively.

Full size image

Moreover, climate change can reshape rainfall patterns, such as increasing rainfall in northern Europe and reducing rainfall in southern Europe1. Increased rainfall generally leads to increased flooding frequency in inland wetlands99, especially in low-elevation wetlands rather than high-elevation wetlands100. Over long timescales, this interannual variability and gradual alterations in climate conditions contribute to long-term changes in water table. On the other hand, changes in rainfall patterns may increase the risk of extreme weather events, causing dramatic fluctuations in water table in the short term101,102.

Uncertainty of methane fluxes under hydrological fluctuations

In wetlands, water table has long been considered a primary driver of CH4 flux variations103,104,105. This recognition stems from the straightforward association observed between CH4 flux and water table in specific scenarios. Particularly, it has been repeatedly observed that CH4 fluxes from wetland soils increase as the water table approaches the soil surface106,107,108,109,110,111. For instance, a significant positive linear correlation between water table and CH4 flux (p < 0.01) emerged in a blanket bog with water table ranging from 50 cm below the soil surface to 0 cm106. Additionally, CH4 emissions in a coastal wetland exhibited an exponential increase with rising water table in the 40 cm below the soil surface to 0 cm range (p < 0.001)107. This correlation under sub-surface water table conditions has also been verified in cross-wetland analyses, including different wetland types108,112,113,114, diverse climate zones115, and different soil management practices113,114,116.

However, nonmonotonic relationships between water table and CH4 flux have also been widely reported in wetlands. For example, CH4 fluxes from permanently inundated wetlands appear to lack a consistent or strong direct relationship with water level108,117. In other words, CH4 fluxes do not consistently increase with water table rise, but rather there exists a critical water table where CH4 emissions increase as the water table rises up to this threshold, followed by high variability19,118,119. Based on data analysis of 42 wetland sites (including bogs, fens, marshes, swamps, and ponds/lake), the critical water table could be 50 cm above the soil surface118. Notably, such an extensive analysis may not fully capture the specific influence of wetland type on the critical water table. The critical water table, as it turns out, is highly dependent on the specific wetland ecosystem, as evident in separate studies exploring the relationship between water table and CH4 fluxes in different wetland types, involving 51 and 83 wetland sites113,120, respectively. The critical water table tended to be below the soil surface in bogs, close to the soil surface in fens and marshes, while in swamps, it was usually above the soil surface.

On the other hand, hydrological fluctuations in wetlands can amplify CH4 flux uncertainty through lag effects on different time scales. During the transition of hydrological state in wetlands, the temporal lag in CH4 emissions could be a few days24,119,121,122 or a few months123,124,125, which is closely related to factors such as the intensity and duration of fluctuation events and soil properties like availability of alternative electron acceptors, soil texture, and initial moisture122. For instance, seasonal CH4 flux changes could lag water table fluctuations by about 17 ± 11 days across 23 wetland sites24, while the temporal lag between CH4 fluxes and precipitation in tropical wetlands was up to eight months125. This short-term lag effect can even lead to a 5-fold lower CH4 emission during the transition from drought to inundation compared to the steady wet conditions126. In addition, extreme drought events can exert long-term effects on CH4 emissions through potential impacts on vegetation composition127 and microbial community structure and activity128, leading to temporal lags of up to several years129.

Moreover, episodic CH4 pulses, such as ebullition events, associated with hydrological fluctuations like dramatic drops in water table130, transient flooding131, spring thaw in most northern peatlands132, and rainfall anomalies in tropical regions133, can lead to short-term outbursts of CH4 fluxes, thereby complicating the relationship with water table. For example, the release of trapped CH4 in the ice layer of Finnish mires during the spring thaw contributed 11% of the annual CH4 emissions134. Further, the seasonal CH4 pulses in East Africa due to rainfall anomalies accounted for over 25% of the global CH4 emission growth in 2019133. This implies that, when exclusively relying on the average water table over an extended period (e.g., annual mean water table), the assessment of CH4 emissions may ignore the considerable impact of high emissions from episodic events135.

Thus, subjected to typical periodical and atypical dramatic hydrological fluctuations, CH4 fluxes readily exhibit nonmonotonicity, lagged response, and pulsation. Additionally, under climate change, rising temperatures may override the influence of water table on CH4 emissions, especially at high latitudes136, by affecting the reaction rates of CH4-related biogeochemical processes. Even in some cases where two wetland sites have completely opposite seasonal water table trends, their CH4 flux patterns do not necessarily mirror the fluctuations in water table85 (Fig. 2a, b), but rather align with soil temperature variations85 (Fig. 2c, d). In tropical wetlands, increased temperatures and reduced precipitation together explain 49% of interannual CH4 flux variations125. This further complicates the relationship between CH4 flux and water table, highlighting its strong dependence on the specific wetland types and local environment, as demonstrated across the 6 wetland types in this study85 (Fig. 3). Nevertheless, water table remains a key explanatory variable for variations in CH4 fluxes within current multivariate models137,138. This is partly due to the simplicity and cost-effectiveness of water table measurements, as well as the convenience of applying simple binary relationships at large scales97. More importantly, in some practical engineering scenarios, such as peatland rewetting/restoration, water table remains the primary consideration for realizing the potential climate benefits that rewetting could offer114. Based on this logic, it is imperative to comprehend and modify the relationship between water table and CH4 cycle before identifying a more suitable explanatory variable. In the following sections, we aim to gain insights into how hydrological fluctuations specifically decouple the subsurface CH4-related biogeochemical processes from surface CH4 fluxes.

Fig. 3: CH4 fluxes at different water tables in six wetland types.
figure 3

a Bog: six sites with 4593 paired observations of CH4 flux (green circles) and water table. b Fen: seven sites with 8130 paired observations of CH4 flux (orange circles) and water table. c Swamp: three sites with 1437 paired observations of CH4 flux (blue circles) and water table. d Marsh: seven sites with 7875 paired observations of CH4 flux (pink circles) and water table. e Salt marsh: three sites with 835 paired observations of CH4 flux (light green circles) and water table. f Rice: five sites with 3070 paired observations of CH4 flux (yellow circles) and water table. Site information is in Supplementary Table 1. Note: positive values of the water table represent distance below the ground surface and negative values represent surface flooding.

Full size image

Response of soil redox conditions to hydrological fluctuations

The occurrence of CH4 production and oxidation in wetlands is determined by the soil oxygenation status that can be indicated by the redox potential (Eh). Basically, Eh levels in wetland soils is largely determined by oxygen availability, which is controlled by diffusion from the atmosphere into the soil profile and by consumption through soil respiration139. In seasonally flooded or chronically submerged wetland soils, soil respiration and organic matter degradation rapidly deplete available oxygen, thereby forming the anoxic environment. Subsequent anaerobic respiration processes continue the organic matter decomposition using alternative electron acceptors (e.g., nitrate, iron, manganese, and sulphate), resulting in a gradual decline in soil Eh until a highly reduced environment (typically below –200 mV) for methanogenesis140. As water table declines, atmospheric oxygen penetrates into the reducing zone through diffusion pathways in water-unsaturated soils, thereby elevating soil Eh at depth and suppressing CH4 production127,141. In addition, oxygen availability plays a crucial role in determining the pathway of CH4 oxidation, influencing whether it occurs aerobically in the presence of oxygen or anaerobically with alternative electron acceptors142.

However, oxygen diffusion in wetland soils is a function of both soil properties (e.g., soil water content, soil structure, and texture) and environmental conditions (e.g., hydrological status and vegetation physiological characteristics). For example, oxygen diffuses 10,000 times faster in air-filled pores than in water-filled pores143, indicating that the original moisture content of wetland soils largely determines the oxygen diffusion rate. In addition, compared to macropores with low tortuosity and high pore connectivity in soil structural units, microporous with high tortuosity and low pore connectivity are more likely to retain water and elongate the diffusion pathways for oxygen, thereby hampering oxygen supply144,145. In these conditions, anoxic microsites can be established even within well-drained soils, thus forming steep Eh gradients between macropores and micropores41. Furthermore, the abundance of anoxic microsites is positively correlated with clay content. This is because soils with higher clay content exhibit elevated levels of microaggregates, whose micropore networks can strongly reduce the oxygen diffusion rate144. Thus, anoxic microsites may be prevalent in fine-textured wetland soils with high organic carbon content during periodic declines in water table, allowing for the maintenance of methanogenic activity146.

Conversely, the influx of oxygen-rich water, such as precipitation and lateral subsurface flow147,148, can establish oxic microsites within otherwise anoxic zones, even in extremely wet or submerged soil conditions. In addition, the formation of oxic microsites is easier in coarser-textured wetlands, attributed to more macropores and faster oxygen transport rates compared to finer-textured wetland soils146,149. In the hyporheic zone of wetlands, the mixing of oxygen-rich surface water and shallow groundwater can also facilitate the deep penetration of dissolved oxygen into the reduced zone150,151. The input of oxygen-rich water implies that transient water table rises might not necessarily result in increased CH4 production, but in some cases can even reduce CH4 emission147. Besides, the aerenchyma in wetland aquatic plants can transport oxygen from the air or from plant photosynthesis to the root tip and the rhizosphere, thus developing oxic microsites in the rhizosphere152.

Thus, redox conditions in wetland soils under hydrological fluctuations are highly dynamic, characterized by pronounced shifts, periodic oscillations between oxic and anoxic states, or high heterogeneity due to the prevalence of aggregate-scale oxic and anoxic microsites. This implies that CH4 production and oxidation are not strictly depth-stratified (Supplementary Table 2). In addition, the presence of anoxic microsites is even considered to be a potential explanation for CH4 production in wetland oxic zones/water columns (i.e., methane paradox)30,31. Similarly, anaerobic CH4 oxidation is not restricted to deep soils, but can be prevalent in surface soils due to the presence of anoxic microsites153.

Effects of hydrological fluctuations on methane production

Substrate availability

Early investigations have repeatedly shown that the availability of substrate, primarily derived from plant residues and root exudates, is one of the critical environmental variables controlling CH4 fluxes in wetlands118,154,155,156. Notably, common substrates for methanogens are typically found in the dissolved organic matter (DOM) fraction, rather than in the organic matter in the bulk soil142. Besides, characteristics of DOM are closely linked to the composition of wetland vegetation communities154,157. For example, DOM derived from different plant species contains different levels of organic acids, which are the more bioavailable fraction of DOM, thus leading to differences in DOM reactivity. Both long-term (often years, decades, or even centuries)158,159 and short-term (often days to months)160 hydrological fluctuations can drive vegetation community succession161. For example, prolonged water table drawdown gradually leads to the transformation of vascular plant communities into woody plants162,163 and the long-term loss of aquatic Sphagnum species, such as S. cuspidatum163. Consequently, water table fluctuations inevitably lead to differences in DOM composition and decomposition by indirectly controlling vegetation communities164.

Hydrological fluctuations can also provide additional substrates for CH4 production. Examples include the input of exogenous organic carbon (e.g., plant debris and terrestrial DOM)165 and the release of previously sequestered labile organic carbon within permafrost166 during thawing contributed to increased methanogenic activity in wetlands. This activity can be further exacerbated by rising soil temperatures167. Lateral water exchange in floodplains also led to higher CH4 fluxes compared to adjacent tundra wetlands, attributed to the additional supply of particulate organic matter168. Similar phenomena can manifest during rainstorms and flooding events169.

On the other hand, the penetration of oxygen under hydrological fluctuations can alter substrate availability through biotic and abiotic pathways and therefore affect methanogenic activity. For instance, redox fluctuations could trigger functional changes in the peat microbiome that contribute to the biotic degradation of organic carbon rich in aromatic and polyphenolic compounds170. Additionally, the co-existence of Fe(II)-containing minerals and oxygen can trigger Fenton-like reactions that generate hydroxyl radicals (•OH) to intensify abiotic organic carbon mineralization in wetland soils170,171,172. Similarly, tidal hydrology-driven redox cycling in intertidal soils can activate thermodynamically stable iron minerals (e.g., pyrite and goethite) to generate redox-active metastable iron phases with efficient electron transfer capacity, which can effectively activate oxygen to produce •OH173. Moreover, during hydrological fluctuation periods, reactive iron minerals can precipitate or dissolve in response to changes in redox conditions, leading to the capture or mobilization of organic carbon62,174,175.

Alternative electron acceptors

Even lacking environmental studies, prolonged periods of inundation (e.g., 50 days) are believed to largely deplete alternative electron acceptors through anaerobic respiration176. However, hydrological fluctuations can induce the repeated renewal of alternative electron acceptor pools, enabling alternative electron acceptors to re-engage in anaerobic respiration processes during post-drought inundation177. This dynamic process inhibits and delays CH4 production, potentially explaining the lagged response of CH4 emissions to water table fluctuations176,178. The inhibitory effect of anaerobic respiration is closely linked to the size of alternative electron acceptor pools, with a focus on iron, manganese, sulfate, and organic alternative electron acceptors due to their significance in natural wetlands.

Regarding iron content in wetlands, groundwater-recharged fens exhibit higher concentrations than precipitation-recharged bogs179,180,181. In such cases, iron can serve as a crucial alternative electron acceptor under anoxic conditions, but readily undergoes redox cycling through various biotic and abiotic pathways under hydrological fluctuations (Supplementary Fig. 2)182,183,184,185,186. The iron turnover due to water table fluctuations has been shown to inhibit CH4 production by stimulating microbial anaerobic respiration176,178. Shorter fluctuation periods generally lead to higher iron reduction rates than long fluctuation periods187,188. In addition, compared to the low fluctuation amplitude (1% oxygen), the high fluctuation amplitude (21% oxygen) during the oxic phase induces faster Fe(II) oxidation and produces more short-range ordered iron minerals, while possessing a higher Fe(III) reduction rate during the subsequent anoxic phase42. However, it should also be noted that the continuously fluctuating redox environment can induce short-range ordered iron minerals to gradually transform into thermodynamically more stable iron-containing minerals, thereby reducing reactivity189. Similarly, manganese turnover, prevalent in mineral-rich wetlands, can competitively inhibit CH4 production190.

Bacterial sulfate reduction is the most important mechanism of anaerobic decomposition and methanogenesis suppression in brackish wetlands due to the continuous supply of sulfate by tidal exchange142,191. Saltwater inundation is therefore considered a useful strategy to mitigate CH4 emissions from rewetted wetlands192,193,194. For instance, compared to freshwater-flooded soils emitting around 303 Mg CH4 yr−1, the emissions of CH4 from saltwater-flooded soils can be negligible193. Besides differences in water levels and substrate availability, substantial differences in sulfate concentrations are the key factor in this phenomenon. Despite small sulfate pools in freshwater wetlands, periodic hydrological fluctuations, transport of oxygen-rich water in hyporheic zones, and alternating oxic-anoxic conditions in plant root zones can fuel rapid sulfur cycling by active iron cycling or direct oxidation via aerobic sulfur-oxidizing organisms151,195,196,197,198,199, potentially reducing CH4 production198.

DOM is considered an important class of organic alternative electron acceptors in organic-rich wetlands, containing electron-donating phenolic and electron-accepting quinone moieties200. They can inhibit CH4 production either directly via the competition between quinone-reducing bacteria and methanogens for electron donors201, or indirectly by chemically oxidizing H2S to oxidized sulfur species to fuel the dissimilatory sulfate reduction process202, or by acting as an electron shuttle to stimulate microbial Fe(III) reduction203. Reduced DOM during the inundation period can be oxidized after water table drops204 or by oxygen transported through the vascular aerenchyma205, thereby competitively inhibiting CH4 production. Importantly, DOM enable reversible and sustainable electron transfer even under continuous redox cycling206,207. This ability allows DOM to reduce 10–166% of the average CH4 flux in the northern peatlands by inhibiting hydrogenotrophic methanogenesis207. Similarly to natural DOM, pyrogenic carbon, produced by wildfire or prescribed burning and constituting for 5–13% or even up to 50% of the soil organic carbon pool in peatlands208,209, is likewise recognized as a redox-active substance that can be recharged repeatedly, which allows it to competitively inhibit CH4 production by 13–24%210.

In the context of changing climate conditions, redox conditions characterized by spatiotemporal dynamics may strongly drive the regeneration of alternative electron acceptors. This hints that current models that do not include redox cycling of iron, manganese, sulfur, and organic alternative electron acceptors may misestimate CH4 emissions211.

Effects of hydrological fluctuations on methane oxidation

Aerobic methane oxidation (MOx) is a biogeochemical process driven by aerobic methanotrophs using oxygen as the terminal electron acceptor. During hydrological fluctuations, MOx inevitably exhibit spatial variations as the availability of oxygen and CH4 varies with water table. Generally, the overlapping region between the oxygen and CH4 gradients, where both gases are present in sufficient concentrations, show increased MOx rates212,213. This region is often located close to the water table fluctuation zone (e.g., the mesotelm in peatlands), implying that pulsating soils with CH4 oxidation hotspots possess higher CH4 consumption capacity than permanently flooded soils140,214. For example, the highest MOx rate (∼160 nmol CH4 g−1 dry soil h−1) in a drained fen with seasonal flooding was observed at the water table fluctuation layer (50–60 cm below the soil surface)215. However, this is not always the case, and to some extent CH4 availability appears to play a more important role in regulating MOx than oxygen availability, which results in MOx sometimes being higher in wet soils than in dry soils216,217,218,219.

CH4 oxidation can also occur under anoxic conditions (i.e., anaerobic methane oxidation, AOM) using alternative electron acceptors220,221,222,223,224,225,226,227. Compared to dry conditions, the submerged conditions during hydrological fluctuations can create a conducive anoxic environment for AOM as well as abundant electron donors, i.e., CH4. The availability of alternative electron acceptors is another key determinant of AOM efficiency. Hydrological fluctuations can introduce exogenous alternative electron acceptors, such as increased sulfate availability due to sea level rise39 and saltwater rewetting193, and nitrate/nitrite inputs from agricultural runoff and leaching228, thereby intensifying corresponding AOM processes. For instance, high sulfate concentrations like 20 mmol L−1 could consume up to 96% of CH4 in the sulfate-CH4 transition zone of coastal wetlands through sulfate-dependent AOM processes220. Nitrogen inputs have been shown to increase the contribution of denitrifying AOM processes to CH4 consumption in coastal wetlands and rice paddies to ∼34% and ∼72%, respectively229. In addition, throughout hydrological fluctuations, the regeneration of alternative electron acceptors not only enhances anaerobic respiration but also offers a continuous supply of electron acceptors for AOM processes. This is why, despite the fact that sulfate-driven AOM is not thermodynamically more favorable than AOM driven by other alternative electron acceptors such as nitrate, nitrite, iron, and manganese, sulfates that rely on rapid sulfur cycling can still be the most important electron acceptor for AOM in wetland hyporheic zones230. This dynamic interaction between CH4, alternative electron acceptors, and AOM under hydrologic fluctuations is also well documented in estuarine/coastal wetlands. Material fluxes between oxygenated tidal water and intertidal surface sediments induced by periodic tidal pulses can provide sufficient carbon and nitrogen sources thereby facilitating denitrification-driven AOM processes231.

Effects of hydrological fluctuations on methane cycling microbes

Methanogens

While methanogens have been identified in various anoxic environments worldwide, only Methanobacteriales, Methanomassiliicoccales, Methanocellales, Methanomicrobiales, and Methanosarcinales are commonly found in wetlands232,233. Generally, methanogens are divided into three groups according to their catabolic pathways, including hydrogenotrophic methanogens, methylotrophic methanogens, and acetotrophic methanogens233,234,235. Traditionally, it is assumed that the acetotrophic and hydrogenotrophic methanogenesis are dominant pathways for controlling CH4 production in most freshwater wetlands, followed by methylotrophic methanogenesis with a smaller contribution233,236. Hydrogenotrophic methanogenesis may be particularly important in acidic peatlands with incomplete organic matter decomposition and high H2 partial pressures237.

The relative abundance of methanogens generally increases with increasing water table and saturation duration238,239, and decreases after water table drops240,241,242. For instance, the abundance of methanogens in submerged soils was 4.2 and 43.2 times higher than that in the soil-water interface and emergent soils, respectively243. Conversely, the diversity of methanogens is lower in hydrologically stable zones that are chronically saturated or submerged compared to water table fluctuation zones239,244,245, which is associated with heterogeneous environments with both anoxic and oxic ecological niches caused by water table fluctuations246. Moreover, water table fluctuations can alter the community structure of methanogenic populations. These differences occur due to changes in soil environmental conditions (e.g., pH, nutrients, and organic carbon content)244,247, variations in vegetation distribution248, and particular physiological characteristics exhibited by certain methanogens (e.g., relative higher resistance to oxygen toxicity/exposure/stress)249,250,251.

Although the abundance and community structure of methanogens are closely affected by water table, they can exhibit certain resistance to drought, and re-establish and further reactivate methanogenic processes after water table has re-emerged, but with varying recovery times ranging from days to years. For example, methanogenesis was rapidly recovered within 3–6 days after wetland inundation in summer121, or more than 3 months after the water table exceeded a threshold of 5 cm below soil surface in shrub bogs252. Even 10–12 years after the rewetting of forestry drainage peatlands, methanogens were still not fully re-established213. Differences in the recovery time of methanogens’ viability are related to a range of environmental variables, such as the duration and intensity of water table drop, the content of organic substrate, and the regeneration of alternative electron acceptor pools253,254.

Methanotrophs

Generally, aerobic methanotrophs are categorized into Gammaproteobacteria, Alphaproteobacteria, and phylum Verrucomicrobia255. According to the phylogenetic variability, genera of aerobic methanotrophs can be further grouped into type Ia (Methylobacter, Methylomonas, Methylomicrobium, and Methylosarcina species within Gammaproteobacteria class), type Ib (Methylocaldum and Methylococcus species within Gammaproteobacteria class), and type II (Methylocystis and Methylosinus species within Alphaproteobacteria class)216,256.

Aerobic methanotrophs are somewhat resilient to stress caused by water table fluctuations (including prolonged drought or dry-wet alternation)257, while their activity, community composition, and abundance can be greatly affected and related to the frequency and intensity of water table fluctuations. For example, under the context of increased frequency of wet-dry alternation (i.e., fortnightly drying to weekly drying), type Ia methanotrophs exhibited the greatest resilience, being the least affected compared to other methanotrophs. In contrast, the abundance of type II and type Ib methanotrophs decreased by about two orders of magnitude, but the former gradually recovered over time while the latter kept declining258. In extreme cases, such as the severe drought experienced in Europe in 2018, the abundance of aerobic methanotrophs and the transcript abundance of the pmoA gene encoding the β-subunit of the particulate methane monooxygenase enzyme showed an overall downward trend in two rewetted fens, and did not recover after the end of the drought259. In addition, the interaction network of the recovered aerobic methanotrophs was profoundly altered compared to the pre-disturbance period, with a reduction in complexity and modularity in a rewetted bog260, or an increase in complexity but a reduction in modularity in dry-wet rice soils261. This implies that water table fluctuations also affect the interactions between methanotrophs and accompanying microorganisms, which in turn affects the recovery of methanotrophs activity and resilience to future environmental disturbances. On the other hand, water table fluctuations can affect aerobic methanotrophs hosted on Sphagnum mosses. Although the symbiotic relationship between methanotrophs and Sphagnum mosses actively contributes to in situ CH4 oxidation in Sphagnum-dominated wetlands, this effect is more pronounced when the water table is close to Sphagnum moss layer rather than during dry conditions262,263.

The AOM process is first associated with sulfate-reducing bacteria in the sulfate-CH4 transition zone of marine ecosystems. Here, the microbial consortium, consisting of sulfate-reducing bacteria and ANME belonging to ANME-1, ANME-2a/b/c, and ANME-3 clades264,265,266,267,268,269, utilizes downward-diffusing sulfate as electron acceptors to oxidize CH4220. The sulfate-dependent AOM process is also found in brackish wetlands220,270. In the sulfate-poor freshwater wetlands, the AOM process can be coupled with nitrate221 and nitrite34,222, which are catalyzed by Candidatus Methanoperedens nitroreducens and Candidatus Methylomirabilis oxyfera (phylum NC10), respectively. Candidatus M. nitroreducens are also capable of coupling the reduction of iron and manganese to the AOM process223. Moreover, Candidatus Methanoperedens ferrireducens271 and Candidatus Methanoperedens manganireducens/manganicus272 have been identified to independently mediate iron- and manganese-dependent AOM processes, respectively. The reductions of chromium225, DOM224, and arsenic226,227 are also successively found to involve ANME-2d clade-driven AOM processes.

Considering that anaerobic methanotrophs prefer to inhabit oxygen-limited environments, they are expected to show a positive correlation between their relative abundance with water table and saturation duration230,239. For instance, the abundance of NC10 bacteria in the water table fluctuation zones can increase from 3.2 × 103 to 5.3 × 104 copies g−1 dry soil after about six months of inundation time273. However, continuous inundation could reduce their abundances due to substrate depletion231. On the other hand, prolonged oxygen exposure can result in anaerobic methanotrophs being subjected to oxidative stress, leading to a decrease in AOM rates274,275. However, intermittent shifts between oxic and anoxic conditions, which occur in oxygen minimum zones, may create suitable living conditions for anaerobic methanotrophs. In periodically flooded mangrove wetlands, for example, NC10 was observed to dominate in the upper soil layers (i.e., 0–20 cm) due to the presence of trace oxygen276. In addition to oxygen, the distribution of electron acceptors (e.g., nitrite and nitrate) also affects the abundance and activity of anaerobic methanotrophs. In Zoige peatlands, for instance, lower nitrite concentrations in deeper soils can result in the relative abundance of NC10 first increasing and then decreasing with soil depth, peaking around 50–60 cm277. Similarly, the effect of water table on the distribution of electron donors (i.e., CH4) inevitably shapes the spatial distribution of anaerobic methanotrophs278. Thus, hydrological fluctuations can affect anaerobic methanotrophs by influencing the formation of anoxic environments and the distribution of both electron acceptors and donors.

Effects of hydrological fluctuations on methane transport

CH4 fluxes from wetlands are the result of the balance between methanogenic and methanotrophic processes in soils, sediments, and water columns5. The CH4 transport pathways, which include plant-mediated transport, ebullition, and diffusion through the water column, determine whether and what proportion of CH4 can be oxidized, as well as the CH4 emission rate130,279. Both plant-mediated transport and ebullition allow CH4 to be rapidly released into the atmosphere almost without undergoing oxidation280, in contrast to the slower diffusion process.

Molecular diffusion, pressurized flow, and transpiration-driven flow constitute the three main pathways for plant-mediated transport. The vegetation composition largely depends on wetland hydrology and varies along a gradient from dry to wet soil conditions281, leading to corresponding changes in the primary CH4 transport mechanisms282. The efficiency of molecular diffusion depends on CH4 concentration gradients, but is usually lower than pressurized flow. As water levels rise, pressurized flow becomes evident in some emergent plants inhabiting shallow water areas, albeit with a relatively lower flow rate. With further increases in water levels, some emergent plants, as well as floating or even submerged plants in deep water areas, can exhibit higher pressurized flow rates282. Consequently, plant-mediated transport dominates CH4 transport from some emergent plant-dominated wetlands compared to ebullition and diffusion137,283,284. Trees also play an important role in CH4 transport by molecular diffusion and transpiration-driven flow that is associated with sap flow in the xylem tissue285. Notably, water table fluctuations are a key driver of changes in tree-mediated CH4 fluxes286,287,288. In forested wetlands, tree-mediated CH4 fluxes can contribute up to ∼50–70% of total CH4 fluxes during wet periods, much higher than the ∼3–30% during dry periods286. Thus, not only differences in wetland vegetation composition induced by long-term water table changes can affect CH4 fluxes via plant-mediated transport, but also short-term seasonal water table changes impact the contributions of plant-mediated transport289.

Ebullition is considered a critical release mechanism especially when the water table is above the soil surface290,291,292,293. Ebullition events are closely linked to the potential for CH4 production in sediments, and can be further exacerbated (6–20% per 1 °C increase) by climate warming due to enhanced microbial metabolism and substrate availability, as well as reduced CH4 solubility294. Another key contributor to ebullition is hydrostatic changes caused by hydrological fluctuations293,295,296. For example, ebullition caused by falling water table was the main driver of the episodic CH4 fluxes in some fens297,298, and even contributed up to 65% of the seasonal CH4 fluxes299. However, when the water table drops below the soil surface, the ebullition pathway would stop as the water-filled pores above the water table drain out and the pore spaces become filled with air instead. Differently, diffusion processes, especially Fickian transport, plays an important role in CH4 transport when the water table is below the surface. This is mainly because the diffusion coefficient in unsaturated soils is much higher than that in water300. Additionally, the diffusion coefficient is negatively related to the diffusion path length, which increases as the water table rises118.

Conclusions and perspectives

This work summarized the possible pathways through which hydrological fluctuations prevalent in wetlands can interfere with the CH4 cycle (Fig. 4). Compared to prolonged drought with CO2 as the dominant greenhouse gas (Fig. 4a) and prolonged flooding where CH4 takes that role (Fig. 4c), the assessment of CH4 fluxes under hydrological fluctuations appears more challenging. Several factors caused by those fluctuations contribute to this complexity (Fig. 4b): (i) the spatial heterogeneity of soil redox conditions, closely linked to wetland soil properties; (ii) alterations in substrate availability for methanogenesis due to changes in vegetation composition, exogenous substrate supply, and substrate degradation efficiency through chemical reactions and mineral transformations; (iii) changes in the size of alternative electron acceptor pools for anaerobic CO2 production and the AOM process through the regeneration of alternative electron acceptors or the introduction of exogenous ones; (iv) variations in community structure and activity of CH₄-cycling microorganisms; (v) changes in CH4 transport pathways and their respective contributions. This reminds us of the need for advanced scientific understanding of hydrological fluctuations, particularly in the context of climate change. To this end, we believe that the following key knowledge gaps still need to be addressed:

  1. (i)

    To better comprehend the hydrology of wetlands, it is essential to increase the monitoring of water table changes, particularly in tropical wetlands with distinct seasonality15. This requires transitioning from manual measurements to remote sensing technologies like satellite imagery and Interferometric Synthetic Aperture Radar301,302, which offers a comprehensive view of water table changes at large scales and high temporal resolutions, along with real-time data from automated sensors and telemetry systems.

  2. (ii)

    To deeply understand the biogeochemical reactions of CH4 production and consumption under different patterns of hydrological fluctuations, we need to explore the impact of electron acceptor pool renewal on the electron flow pathways of organic matter mineralization and AOM processes. This requires more extensive measurements of electron acceptor pools, revealing which biogeochemical processes are occurring in wetlands. Such an electron perspective differs from simple redox potential measurements, which might fail to capture local heterogeneity such as oxic and anoxic microsites123.

  3. (iii)

    To better understand seasonal variation patterns and microbial driving mechanisms of CH4 fluxes, more research is needed to explore the seasonal activity of methanogens and methanotrophs in response to seasonal dynamics of environmental conditions43,303,304,305,306.

  4. (iv)

    To clearly understand the complex interactions between hydrological fluctuations and CH4 transport mechanisms, more research is needed to observe the impact of water table fluctuations on three CH4 transport pathways, i.e., plant-mediated transport, ebullition, and diffusion. Particularly, the current understanding of wetland tree-mediated CH4 emissions, including transport mechanisms, control factors, and quantitative methods, is still in its infancy. In addition, our knowledge regarding the impact of emergent macrophytes, such as papyrus, on CH4 production and transport in tropical wetlands307 remains limited.

  5. (v)

    To accurately assess global CH4 budgets, existing process-based models need to be updated by incorporating the understanding of CH4-related biogeochemical processes under hydrological fluctuations. Furthermore, it is evident from our review that the current approach of relying solely on annual mean water table depth for calculating CH4 budgets is insufficient. Instead, we propose implementing alternative approaches that account for more detailed temporal variability (such as daily and seasonal) in water table depth and consider other relevant hydrological variables, such as hydroperiod and inundation events.

  6. (vi)

    The impact of hydrological fluctuations on CH4 cycling may vary under different climatic conditions with different temperatures308. However, the current information remains challenging in deciphering the interactive effects of these two factors. Thus, further research is essential to explore the temperature-affected hydrological effect on CH4 emissions.

Fig. 4: Conceptualization of the biogeochemical CH4 cycle in wetlands under different hydrological scenarios.
figure 4

a Prolonged drought. b Hydrological fluctuations. c Prolonged flooding. The diagrams illustrate the CH4-related biogeochemical processes (red arrows) and the impact pathways of hydrological fluctuations on the CH4 cycle (blue dashed arrows). The different zones are marked as oxic zone (light brown), fluctuation zone (blue), anoxic zone (dark brown), and root zone (patterned area). Plant-mediated transport is indicated by plant diagrams, ebullition by bubbles, and diffusion by curved arrows. Aerobic respiration tends to be more active in oxic zones during prolonged drought, while methanogenesis is more prevalent in anoxic zones during prolonged flooding. Under hydrological fluctuations, CH4-related biogeochemical processes become more complex due to changes in soil redox state, substrate availability, the pool of alternative electron acceptors, contribution of different CH4 transport routes, and the community structure and activity of methanogens and methanotrophs. Note: rhizosphere oxygen is particularly critical for aerobic processes under prolonged flooding conditions and is therefore highlighted in the panel c. AOM anaerobic methane oxidation, MOx aerobic methane oxidation, DOM dissolved organic matter, Corg organic carbon.

Full size image

In conclusion, understanding the impact of hydrological fluctuations on CH4 cycle are essential to accurately estimate CH4 emissions from wetlands and reduce their contribution to global greenhouse gas emissions. Ignoring this issue would compromise our ability to achieve climate change mitigation targets and sustainable management of wetland ecosystems.