Regulation of Methane Emissions in a Constructed Wetland by Water Table Changes

Abstract

Riparian wetlands release greenhouse gases and sequestration carbon as well, so their carbon source and carbon sink functions have become some of the key research issues of global climate change. In this present paper, the main controllable factors of the self-designed and constructed riparian wetland, namely hydrological conditions and additional carbon sources, were artificially regulated, and then methane fluxes were measured. The results proved that the methane emissions were significantly positively correlated with the water level heights, and the methane emissions increased exponentially with the rise of water level when the water level was between −20 cm and +20 cm. According to the −20~0 cm water level, a small number of methane emissions was significantly different from the 10 cm and 20 cm water levels, which indicated that higher water level could significantly promote methane emission. When the water level reached above 0 cm, the methane emission gradually increased as the flooding time became longer; it reached the peak value after more than 20 days of flooding after which it decreased, which provided a scientific basis for optimal design and effective management of restored and constructed riparian wetlands, minimizing the methane emissions of riparian wetlands.

1. Introduction

The total organic carbon storage of global wetlands accounts for 20–30% of the surface carbon storage of the terrestrial ecosphere, and wetlands have become the main source of methane emissions, occupying 70% of all natural methane emission sources [1]. Natural wetlands emit about 1.45 × 10¹¹ kg of methane to the atmosphere every year, which is equivalent to 25% of the total emissions from all natural and anthropogenic sources [2]. The riparian wetland ecosystem has high primary productivity, which fixes and stores a large amount of carbon, and plays a special role in global climate change [3]. Human activity not only changes wetland vegetation composition and hydrological conditions but also varies the soil permeability, oxidation reduction potential, and soil respiration rate, thus causing a certain amount of carbon loss [4,5,6,7]. Therefore, some countries and regions around the world are actively taking measures to construct and restore wetland ecological projects to recover the lost carbon.

Compared with natural riparian wetlands, constructed and restored riparian wetlands might reduce methane emissions to some extent. Nahlik and Mitsch [5] illustrated that by establishing fluctuating hydrological conditions and restoring vegetation, the annual methane emission (CH₄-C) of the wetland could be 17.00 g m⁻² y⁻¹. A natural riverside wetland located at the same latitude has 80.00 g m⁻² y⁻¹ of CH₄-C that is approximately five times higher [6]. In a comparison between CH₄-C and different types of wetlands at the Olentangy River Wetland Research Center, Sha et al. [7] also concluded that the annual CH₄-C from a constructed riverside wetland with planted vegetation, a lagoon with semi-artificial restoration and a natural river-side wetland were 68.00, 0.30 and 379.00 g m⁻² y⁻¹, respectively. The CH₄-C of the natural riverside wetland was higher than that of the constructed wetland. In a restoration experiment of a riverside wetland in Denmark, the CH₄-C range was −0.21~32.10 g m⁻² y⁻¹, which was slightly lower than the CH₄-C of 38.00 g m⁻² y⁻¹ in the natural riverside wetland in the same area [8].

In constructed and restored riverside wetlands, high water level is a hindering factor in the process of CH₄-C from soil to the atmosphere, and excessive water inundation will inhibit gas diffusion and reduce CH₄-C [9]. Many natural riverside wetlands have dynamic inter-annual flood intensity, duration, and frequency variability, which leads to the ground and underground water depth change or dry–wet circulation. Such wetlands tend to emit much less methane compared with continuously submerged wetlands [10] because the higher average water level and longer inundation time in the continuous inundation area result in the hypoxia of the bottom sediment, thereby increasing the activity of methanogens and strongly reducing soil conditions [11]. Accordingly, for the construction and restoration of riverside wetlands, the conditions of fluctuating water levels and intermittent inundation tend to cause fewer CH₄-C [12]. In the laboratory simulation experiment conducted by McLain and Martens [13], methane oxidation activity was remarkably weakened when the riverside wetland changed from submerged dry to surface dry.

Methane production requires anaerobic conditions that are provided by flooding and the water level range has a threshold value, while it has been difficult to reveal the water level effectiveness, especially when it is inundation, with results indicating the specific positive impacts of water level fluctuations on carbon emissions from wetlands [14]. To achieve this, our study sought: (1) To Assess the wetland methane emissions under different factors dominated by hydrological conditions and additional carbon sources that drive carbon emissions; (2) To understand CH₄-C regulation and explore the main factors affecting CH₄-C and the possible regulation ways through the exploration of hydrological conditions and vegetation changes of the restored riverside wetland and provide a theoretical basis for the optimal design and effective management of the constructed and restored riverside wetland.

2. Study Methods

2.1. Study Site Description

The study was carried out in the west of Shanghai and Suzhou River upstream region (longitude 121°236′, latitude 31°257′), Shanghai, China, and the average temperature of the area was 17.60 °C. The lowest monthly average temperature was 5.60 °C in January, the highest was up to 29.80 °C in August, and the average annual precipitation was 1104.40 mm. Rainy days were 130 days and mainly concentrated in the flood season from May to October. Soil parent material experienced natural soil-forming processes, such as salinization, marsh gley, meadow soil, and long-term cultivation, forming the existing soil type. The main types are blue mud and ditch dry mud, and the texture is slightly sticky. Local soil types were bruising mud and ditch dried mud, clay loam soil, and a slightly sticky texture.

2.2. Restoration Study Design

Restored wetlands consisted of inundated and constructed wetland that was 15 m long and 4.50 m wide, with horizontal bottom. Water was pumped from the adjacent rivers to the restored wetlands, which were in continuously inundated treatment with the same water table. Studies demonstrated that CH₄-C could be observed within the water table range of −20~+20 cm to simulate the dynamic hydrological conditions of natural riparian wetlands [15,16]. Three parallel sample microcosms (84 cm × 65 cm × 50 cm) were placed on water tables of −20 cm, −10 cm, 0 cm, 10 cm, and 20 cm, respectively (Figure 1). The microcosms, containing gravel to a depth of 10 cm on the bottom, were filled with 30 cm of soil on the gravel. Two equidistant holes with an inside diameter of 1 cm on both sides of 5 cm height from the bottom were set. Moreover, the soils from the riverbank were used to simulate the soil conditions of natural riparian wetlands. The initial physical and chemical properties of soils in every microcosm were the same.

In order to simulate the construction or restoration of riverside wetlands, river water will bring more organic matter and nutrients with the increase in time and the litters of artificial shelterbelts along the shore will also increase the organic matter in wetland soil. Therefore, the experiment sets the addition of organic carbon of different concentrations to simulate the change of soil organic carbon content in riverside wetlands. An additive carbon source (dried reeds that equal 130 g/kg carbon content) was added to the soil on each water table gradient, which contrasted with additive-free carbon source ones. During the growing season from May to September 2020, CH₄-C in the soil of additive carbon groups and additive-free groups were determined on various water tables at the beginning of every month, respectively. In early June 2020, research on the influence of inundation duration time on methane emissions by extraneous carbon groups was initiated. The CH₄-Crates of various water tables were measured on the 1st, 7th, 15th, 32nd, and 43rd day of the inundated treatment, respectively.

2.3. Gas Sampling and Analysis Methods

Gas sampling was conducted using static chambers. The static chamber included a pedestal, extending box, and top box that were made of an 8 mm PVC plate due to higher plants. The exterior of the extending box and the top box were sprayed with a layer of insulation material. The length, width, and height of the base were 50, 50, and 30 cm, respectively. The capacity of the extending and the top box was the same, whose length, width, and height were 50 cm. The top of the chamber was equipped with a 23 cm long needle, one end of which reached into the center of the chamber and the other end was connected to the three-way valve outside to collect gas. The top was installed with a stainless steel tube (23 cm long, with an internal diameter of 2 mm) to ensure the gas-pressure balance inside and outside the chamber when gathering gas. The temperature inside the chamber was measured using a PT1000 stainless steel temperature probe (23 cm long, with an internal diameter of 3 mm) when collecting gas. The inner sidewall was equipped with two axial flow fans to mix gas. The upper end of the pedestal and the extending box had sealing sinks that were 3 cm deep and 2 cm wide.

Sampling was conducted between 9:00 a.m. and 11:00 a.m. at the beginning of each month. The height of the water to the top (H) was measured and recorded, and the temperature inside (t), after placing static chambers and being stable for 5 min, was recorded. Gas samples were collected after rotating the fans for 1 min to mix air in static chambers. Sampling lasted 30 min and was conducted every 10 min at 0 min (T₀), 10 min (T₁), 20 min (T₂), and 30 min (T₃). The air samples were stored in a cooler at 4 °C and the concentration was analyzed by gas chromatography within 28 days.

The concentrations of CH₄ gas samples were determined on an Agilent 6890N gas chromatograph (Agilent Technologies, Inc., Palo Alto, CA, USA). CH₄ emission flux was calculated as follows:

$$F=6\times {10}^{-4}H\frac{(16\times P)}{R\times (273+T)}\frac{dC}{dt}$$

(1)

In the formula, F is the CH₄ emission flux (mg m⁻² h⁻¹), dC/dt refers to the CH₄-C rate (pp mv min⁻¹), and 60 is the time constant conversion from a minute to an hour. H is the effective height of the sampling container (cm), and 16 is the molar mass of CH₄ (g mol⁻¹). P represents the atmospheric pressure of the sampling site and generally standard atmospheric pressure (1.013 × 10⁵ Pa). R indicates the universal gas constant (8.31 J mol⁻¹ K⁻¹). T marks the average temperature inside when sampling (°C).

2.4. Environmental Factor Sampling and Analysis Methods

Collecting gas the ambient temperature, surface temperature, water temperature, soil temperature (5 cm, 10 cm), and groundwater levels were recorded by collecting gas. Soil samples were collected in August and October. The redox potential (ORP), electrical conductivity, and water content of soil were measured simultaneously by the FJA-6 ORP depolarization automatic tester and HH2 moisture meter, respectively. According to the soil chemical analysis method compiled by the Soil Science Society of China [LURKed analytical methods of soil agrochemistry Beijing Chinese agriculture science and technology press 1999], soil organic carbon was analyzed.

2.5. Data Analysis

SPSS20.0 pair wise t-test was used to analyze the differences in methane emissions in seasons, vegetation types, and water levels. The significance level of the test was 0.05. Pearson correlation analysis was used to test the influence of environmental factors, including temperature, water level, and soil organic carbon, on CH₄-C fluxes. Curve estimation was used to analyze the relationship between environmental factors and methane emissions.

3. Results

3.1. Seasonal Dynamics of CH₄-C

CH₄-C fluxes of additive-free carbon treatment and adding carbon treatment presented similar seasonal dynamics during the growing season. Methane fluxes were lower in spring, increased significantly in summer, and reduced to a lower emission level in September. Soil temperature continued to rise in May, and reached the highest value in August, which showed significant seasonal variation. The results were similar to previous studies [17,18,19], which indicated a time synchronization between CH₄-C and temperatures. Higher soil temperatures promoted methane emissions, while lower soil temperature inhibited the activity of methanogenic bacteria and caused lower CH₄-C.

In additive-free carbon groups, little CH₄-C was observed in sample boxes of water levels within −20~0 cm and the average CH₄-C rates were 0.01 ± 0.03 mgm⁻²h⁻¹, 0.02 ± 0.02 mgm⁻²h^−1, and 0.07 ± 0.06 mgm⁻²h⁻¹, respectively. Only 0.10 ± 0.06 mgm⁻²h⁻¹ of CH₄-C was measured in sample boxes of 10 cm water level in May. CH₄-C began to increase noticeably in June and the upward trend continued until August and reached a peak at 2.77 ± 0.78 mgm⁻²h⁻¹. CH₄-C was reduced to only 0.34 ± 0.09 mgm⁻²h⁻¹ in September, and was lower at 2.56 ± 0.31 mgm⁻²h⁻¹ in sample boxes of 20 cm water level in May and went up to triple in June. In summer, CH₄-C was maintained between 6.62 and 7.22 mgm⁻²h⁻¹ and had a narrower fluctuation. CH₄-C significantly reduced in September, with only 1.18 ± 0.49 mgm⁻²h⁻¹ (Figure 2). The soil temperature of different water levels showed a similar varying pattern in that temperature was lower in May, warm until August, reaching the maximum of 34.40~37.60 °C, and significantly reduced to draw near the temperature of May in September.

CH₄-C variation of additive carbon groups was similar to that of additive-free carbon groups. Slight CH₄-C was observed in sample boxes of water levels within −20~0 cm and the average emission rates were 0.07 ± 0.03 mgm⁻²h⁻¹, 0.09 ± 0.03 mgm⁻²h⁻¹, and 0.58 ± 0.21 mgm⁻²h⁻¹, respectively. A small amount of CH₄-C was measured in sample boxes of 10 cm water level in May. CH₄-C sharply increased from 0.77 ± 0.72 mgm⁻²h⁻¹ to 11.74 ± 4.08 mgm⁻²h⁻¹ in June and the upward trend continued until August and reached a peak of 16.82 ± 5.11 mgm⁻²h⁻¹. CH₄-C greatly reduced to only 1.95 ± 0.78 mgm⁻²h⁻¹ in September. 11.22 ± 4.69 mgm⁻²h⁻¹of CH₄-C were relatively lower in sample boxes of 20 cm water level in May and rose significantly to triple in June. Afterward, the CH₄-C rate slowly increased and came to a maximum value of 41.65 ± 10.02 mgm⁻²h⁻¹. CH₄-C decreased rapidly in September (7.01 ± 2.26 mgm⁻²h⁻¹) (Figure 3). Soil temperature of different water levels showed a similar variation because temperature was lower in May and warm until August, with the maximum value of 33.9~36.6 °C, and rapidly reduced to about 26 °C in September.

3.2. Relationship between Water Height and CH₄-C Fluxes

Organic carbon content, moisture content, pH, ORP, and soil temperature were measured in sample boxes of different water tables during the growing season. Based on the results of soil physiochemical properties, soil water content gradually increased with the rising water tables and had a significant relationship with water tables in additive carbon and additive-free carbon experiments (R² = 0.983; R² = 0.9583). Eh values had a gradual decrease trend with the rising water tables and there was a linear relationship between the two (R² = 0.8744; R² = 0.8296) (Figure 4).

Although the rates of CH₄-C were quite different between the two and mean values were 1.15 mgm⁻²h⁻¹ and 8.43 mgm⁻²h⁻¹, CH₄-C changed with a water level: methane fluxes increased significantly as the water level rose. Inundated conditions within −20~0 cm led to a significant reduction in CH₄-C, which indicated that higher water levels will promote the production and release of methane.

In additive-free carbon experiments, CH₄-C was shown in Figure 3 at various water levels during the growing season and increased as the water levels rose. There was only a small sample of CH₄-C in sample boxes of water levels within −20~0 cm and the average emission rates were 0.03 ± 0.02 mgm⁻²h⁻¹, 0.05 ± 0.03 mgm⁻²h⁻¹, and 0.07 ± 0.05 mgm⁻²h⁻¹, respectively. CH₄-C at a 10 cm water level increased significantly and the average emission rate was 1.36 ± 0.34 mgm⁻²h⁻¹. When the water level rose to 20 cm, CH₄-C increased rapidly, reaching 4.25 ± 0.77 mgm⁻²h⁻¹.

In the additive carbon experiment, CH₄-C was shown in Figure 5 at various water levels during the growing season and water level rise increased CH₄-C. Little CH₄-C was observed in sample boxes of water levels within −20~0 cm and the average emission rates were 0.31 ± 0.15 mgm⁻²h⁻¹, 0.32 ± 0.16 mgm⁻²h⁻¹ and 1.01 ± 0.22 mgm⁻²h⁻¹, respectively. When the water level rose to flooded soil, CH₄-C increased rapidly, and the average emission rate of 10 cm water level was 11.49 ± 3.11 mgm⁻²h⁻¹. A total of 10 cm water level of CH₄-C performed a significant difference, compared with water levels within −20~0 cm (p = 0.025; p = 0.026; p = 0.035). When the water level rose to 20 cm, CH₄-C increased to 4.25 ± 0.77 mgm⁻²h⁻¹ sharply and was nearly three times as much as that of the 10 cm water level.

3.3. Relationship between Inundated Time and Methane Flux

Variations of CH₄-C at different water tables were not the same with inundated time, but minimum fluxes all occurred on the first day of inundation (Figure 6). Methane fluxes were 0.02 ± 0.01 mgm⁻²h⁻¹, 0.04 ± 0.00 mgm⁻²h⁻¹, 0.02 ± 0.01 mgm⁻²h⁻¹, 0.77 ± 0.72 mgm⁻²h⁻¹, and 11.22 ± 4.69mgm⁻²h⁻¹ within −20~20 cm, respectively.

Variations of CH₄-C at the −20 cm water table were not obvious over inundated time. The maximum value of 0.18 ± 0.04 mgm⁻²h⁻¹ occurred on the 7th day of inundation. Methane flux decreased slightly on the 15th day of inundation and continued to reduce to 0.02 ± 0.01 mgm⁻²h⁻¹ on the 32nd day, but CH₄-C that was about two folds more than that of the 32nd day increased slightly on the 43rd day.

The methane flux of the −10 cm water table that was similar to that of the −20 cm water table was less obviously related to inundate time. Likewise, the maximum value of 0.18 ± 0.04 mgm⁻²h⁻¹ occurred on the 7th day of inundation, and methane flux decreased slightly on the 15th day of inundation. CH₄-C rebounded to 0.14 ± 0.02 mgm⁻²h⁻¹ on the 32nd day, which differed from that of the −20 cm water table. However, on the 43rd day, CH₄-C was reduced to the lowest emission level (0.05 ± 0.02 mgm⁻²h⁻¹).

Variations of CH₄-C within 0~20 cm water tables were similar with inundated time. On the 32nd day after inundation, methane flux gradually increased over time but began to decrease on the 43rd day. There was more obvious CH₄-C as 0.52 ± 0.27 mgm⁻²h⁻¹ at the 0 cm water table on the 7th day, and methane fluxes on the 15th and 32nd day were two- and seven-fold more than that of the 7th day, respectively. However, on the 43rd day, CH₄-C was reduced to the lowest emission level of 0.38 ± 0.05 mgm⁻²h⁻¹. There was higher methane flux for 10.23 ± 1.04 mgm⁻²h⁻¹ at the 10 cm water table on the 7th day after which the methane flux rapidly increased over time. A maximum value of 37.01 ± 11.84 mgm⁻²h⁻¹ occurred on the 32nd day of inundation. Methane flux significantly reduced on the 43rd day (5.78 ± 1.95 mgm⁻²h⁻¹), which had a decline of about 84%. CH₄-C of 20 cm water level was always higher than 10 mgm⁻²h⁻¹. Methane flux maintained a steady growth trend and reached a maximum value of 45.87 ± 6.10 mgm⁻²h⁻¹ on the 32nd day, and decreased slightly on the 43rd day, for 38.94 ± 12.40 mgm⁻²h⁻¹.

4. Discussion

4.1. The Impacts of Water Level Change on CH₄-C

During the construction and restoration of riverside wetlands, CH₄-C could be directly affected by adjusting the water level and changing the intensity and duration of flooding. The water flow process in the riverfront region changes and determines the physical and chemical properties of wetland soil, such as the availability of nutrients, soil and water salinity, soil anaerobic, pH and sediment properties, etc., and the two properties act together on specific biomes to influence CH₄-C [20]. New and repaired water level and water area in riverside wetland not only directly influence the methane emission intensity change, but also determine the wetland types, the dynamic change of plant community, and regulate the spatial structure of wetland soil organic carbon and the spatial pattern of water distribution, thus, making the wetland in the methane production and methane oxidation environment mutually transform between two kinds of states [21].

Therefore, water level is one of the factors that have the strongest impacts on CH₄-C from newly built and restored riverside wetlands [22]. High water levels, long hydraulic retention times, and slow water flow are generally favorable for methanogenesis. Regression analysis between different water tables and methane flux indicated that methane flux increased exponentially as water levels rose, which suggested that water level had significant effects on CH₄-C, and water level rise promoted CH₄-C. Little CH₄-C was observed underwater levels within −20 ~0 cm, and there was no significant difference among the three (p < 0.05). When the water level reached the groundwater level, the effect sizes gradually decreased with the increasing groundwater level, which was mainly due to the relatively higher soil oxygen diffusion rates as the groundwater level increased [23]. The water column not only hinders the diffusion of CO₂ but also dissolves part of the CO₂ in the water [24]. At the same time, flooding stress decreases the photosynthetic capacity of plants [25].

After water level rose to 10 cm, CH₄-C rapidly increased and the maximum value occurred at 20 cm water level. The results demonstrated that with the increase in water level, its contribution to the promotion of CH₄-C would increase. Furthermore, when the soil was not in an inundated condition, methane production and emission were less affected by water level change. In additive-free carbon experiments, CH₄-C had a significant relationship with the water tables (R² = 0.8791, p < 0.01), and there was only a small amount of emissions within −10~0 cm. After water level rose to 10 cm, CH₄-C increased noticeably, and CH₄-C rates showed greater growth when water level continued to rise to 20 cm. In additive carbon experiments, CH₄-C had a significant relationship, and it increased exponentially as water levels rose, which was similar to additive-free carbon experiments (R² = 0.915). CH₄-C had significant differences within −20~10 cm (p = 0.000; p = 0.000; p = 0.000; p = 0.001) (Figure 7). The rise of water level caused the reduction in that soil dissolved oxygen, and then suppressed the oxidation of methane and increased CH₄-C because methane was produced under anaerobic conditions.

Supported by field observations in a variety of water levels, our study revealed, maximum CH₄-C occurring for a critical level of inundation. This water level originated from two effects. One might be related to the availability of microbial activities and oxygen utilizations as influenced by the saturated soil. The other linked to the gas transport in and out of the soil and gas diffusion rates as well [26]. The correlation between elevated CH₄-C at higher water table levels was proved by previous studies in freshwater ecosystems [27,28] and saline soils [29,30]. A microcosm simulation study in a freshwater marsh showed that a higher water table condition (+2 to +14 cm) emits 75% more CH₄ than at low water table (−11 to 0 cm) [31]. A study on CH₄-C in lowland tropical peatlands confirmed that CH₄ emissions under 0 and −10 cm water table levels were significantly greater than those at −20, −30, and −40 cm over 150 days [32]. Liu et al. [33] reported that the CH₄-C with the highest water level exhibited a higher average CH₄ flux (100.3 ± 21.8, 264.9 ± 32.8 mmol m⁻² h⁻¹ for P. australis; 186.7 ± 44.9, 705.7 ± 76.0 mmol m⁻² h⁻¹ for S. alter-niflora;) in Chongming Dongtan. Altor and Mitsch [34] discovered that CH₄-C was maximum when the water level was near the soil surface and significantly reduced when the water level was higher than 25–30 cm. Meanwhile, almost no CH₄-C occurred when the water level was 20 cm below the surface in intermittently submerged areas. Other studies have also confirmed that lower CH₄-C levels can be ensured when the water table is 20–30 cm [35,36]. According to a study on CH₄-C from a typical hydro-fluctuation belt by Chen [37], the average CH₄-C flux was 0.98 mgm⁻²h⁻¹ within −5~7 m, and the average flux in -1m water level was 1.29 mgm⁻²h⁻¹. Hou et al. [38] reported that CH₄-C significantly increased as the water level rose, which indicated that flooding conditions promoted the production and emission of methane and its contribution enhanced as the water level rose. Li et al. [39] found that CH₄-C had a significantly positive correlation with water levels and CH₄-C flux exponentially increased as the water level rose, which was consistent with the results of this experiment.

4.2. Effects of Soil Moisture on CH₄-C

By measuring the moisture content of the soil, there was a good binomial relationship between CH₄-C and soil moisture content, which indicated that soil moisture was the main factor affecting CH₄-C (Figure 8). In additive-free carbon experiments, the average moisture contents of soil within −20~20 cm were 42.66 ± 1.54%, 49.01 ± 1.32%, 49.49 ± 1.74%, 55.71 ± 1.74%, and 60.38 ± 1.79%, respectively. There was a significantly positive relationship between CH₄-C and the soil moisture contents (R² = 0.7624). In additive carbon soil, the average moisture contents within −20~20 cm were 41.89 ± 1.75%, 45.36 ± 2.68%, 50.33 ± 2.51%, 58.27 ± 1.91%, and 65.26 ± 1.58%, respectively. There was a significant relationship between CH₄-C and the soil moisture contents (R² = 0.8319).

Related binomial analysis showed that the higher the soil moisture contents were, the more the CH₄-C was. Soil had the least amount of CH₄-Cat, about 45% of moisture contents. According to a study in a coastal wetland by Zhao et al. [40], CH₄-Cwas positively related to soil moisture. Many researchers also proved that soil moisture was the main factor affecting CH₄ production and consumption [41,42].

Furthermore, Cardoso et al. [43] also revealed that, compared with the moist soil, CH₄ oxidation in the dry soil increased, which was in line with the previous knowledge that the rate of CH₄ oxidation was higher in dry conditions than that in wet conditions. That high soil moisture helped methane production because high soil moisture contents inhibited the activity of methane-oxidizing bacteria and affected methane production and oxidation by impacting the amount of oxygen in soil [44]. From the above studies, it could be anticipated that CH₄ fluxes in wetlands are sensitive to the hydrologic condition. Lower water levels would not only reduce anaerobic zones and suppress methanogenesis, but also increase CH₄ oxidation with abundant oxygen in the overlaying zones. When the water table drops considerably below the surface, the wetlands may change from being a source to being a sink for methane due to increased methane oxidation.

4.3. Effects of Soil Redox Potential on CH₄-C

According to the results of the soil redox potential (Eh), there was a good binomial relationship between CH₄-C and soil Eh values (Figure 9). In additive-free carbon experiments, the average Eh values of soil within −20~20 cm were 360.87 ± 29.67 mV, 348.71 ± 16.62 mV, 286.21 ± 44.34 mV, −31.34 ± 52.48 mV, and −140.46 ± 54.83 mV, respectively. There was a significantly positive relationship between CH₄-C and soil Eh values (R² = 0.8693). In additive carbon soil, the average Eh values within −20~20 cm were 291.85 ± 33.60 mV, 276.74 ± 28.31 mV, 265.79 ± 29.91 mV, 129.91 ± 25.83 mV, and −18.47 ± 32.21 mV, respectively. CH₄-C had a significant relationship with soil Eh value (R² = 0.6979).

Related binomial analysis showed that the smaller the soil Eh value was, the more the CH₄-Cis. When the soil Eh value was below 0 mV, CH₄-C was generally greater. There was only a small amount of CH₄-C when Eh was 200~400 mV. The results might be due to the high water tables keeping anaerobic conditions, while lower water tables allow oxygen to more readily diffuse into the substrate, creating aerobic conditions in the substrate. Methanogenesis in wetlands is limited to zones under anaerobic conditions, thereby controlled by the water table. Different studies have different definitions for the methane-Eh-required threshold. Peters and Conrad found that Eh value was about 0~70 mV when producing methane in the beginning [45]; Huang et al. reported that Eh value was −110 mV [46]; Cicerone and Oremland considered that Eh value was −300 mV [47]. The results of this study were in these ranges. Lower redox potential would promote methane production, Eh values drop faster, and methane produces more because Eh determined soil microbial and all kinds of soil enzyme activity and low Eh values could provide the necessary reduction conditions for methanogens.

4.4. Impacts of Inundated Time on CH₄-C

The results illustrated that flooding time was correlated with methane flux under different water conditions. CH₄-C at −20 cm water level was less relevant to flooding time. Probably, soil moisture was too small and did not provide an adequate anaerobic environment for methanogens. CH₄-C was little and there were larger fluctuations. CH₄-C at −10 cm water level had a binomial correlation with flooding time (R² = 0.549), but CH₄-C was still low, and the significance level was not high. When the water level reached above 0 cm, there was a significant relationship between soil inundation time and CH₄-C (R² = 0.7724; R²= 0.7507; R²= 0.989). CH₄-C gradually increased with the increase in flooded time. The maximum value occurred on the 20th~30th day, and then there was a certain degree of reduction. Especially at a 20 cm water level, there was an extremely significant correlation between CH₄-C and inundated time, probably because adequate soil moisture caused extreme reduction and hypoxic conditions while inhibiting the activity of methane-oxidizing bacteria, reducing methane oxidation and increasing CH₄-C.

According to a study on CH₄-C from a typical hydro-fluctuation belt by Li [31], observation sites were out of water surface after a period of inundation, thus leading to the peak of CH₄-C. A long-term anaerobic environment caused sediment to produce a large number of methanogens and large amounts of decaying matter, which provide an ideal environment for methane production [48]. Conrad [49] reported that nutrient releasing, strong bacterial activity and unstable carbon decomposition may result in CH₄-C from the water–gas boundaries that increase in a short time after flooding. It can be observed that a long flooding time will result in increased CH₄-C, but flooding time has a threshold. That flooded time is too long, which may also hinder the release of methane into the atmosphere and reduce methane emissions. When soil column is inundated, the water column hinders the diffusion of O₂, and the effect sizes further demonstrated that microbial activity that causes soil mineralization and decomposition is significantly reduced [50]. It is also likely due to the diffusion barrier of the water column and methanotrophy, which limits the emission of CH₄ from the soil to the atmosphere [51]. Higher soil organic matter decomposition rates would lead to higher concentrations of labile fractions [52]. As part of organic carbon, soil labile organic carbon refers to unstable, easily oxidized and mineralizable. It occupies a relatively small proportion of the total soil carbon, but is of great significance to CH₄ emissions [53].

5. Conclusions

The results of the research on CH₄-C from riparian restored wetlands at different water levels during the growing season demonstrated that CH₄-C had a significant relationship with water tables, and increased exponentially as water levels rose, which indicated that water level rise significantly promoted CH₄-C. The inundation environment is a necessary condition for promoting methane production, and its contribution to the promotion of CH₄-C will increase with the rise of the water level. CH₄-C gradually increased with the increase in flooded time, and the peak occurred after more than 20 days of inundation. Afterward, CH₄-C showed a decreasing trend. Thus, the water level of wetlands can periodically drop below the soil surface and minimize CH₄-C. The suggestions for the design and construction of created and restored riparian wetlands are to simulate the dynamic hydrological conditions of natural riparian wetlands and create periodic water level fluctuations rather than continue flooded wetlands. Furthermore, the influence of water conditions on the decomposition process of litter by microorganisms is still insufficient. It is suggested that in future studies, strengthening the research on the litter decomposition mechanism under flooded environments, as well as the influence of community succession on microbial species and community structure is needed.