Role of Scirpus mariqueter on Methane Emission from an Intertidal Saltmarsh of Yangtze Estuary

The role of wetland plant (Scirpus mariqueter) on methane (CH4) emissions from a subtropical tidal saltmarsh of Yangtze estuary was investigated over a year. Monthly CH4 flux and pore-water CH4 concentration were characterized using static closed chamber technique and pore-water extraction. Measured chamber CH4 fluxes indicated that saltmarsh of the Yangtze estuary acted as a net source of atmospheric CH4 with annual average flux of 24.0 mgCH4·m·day. The maximum chamber CH4 flux was in August (91.2 mgCH4·m·day), whereas the minimum was observed in March (2.30 mgCH4·m·day). Calculated diffusion CH4 fluxes were generally less than 6% of the chamber fluxes. Significant correlations were observed between the chamber CH4 flux and rhizospheric pore-water CH4 concentration (11–15 cm: p < 0.05, R = 0.732; 16–20 cm: p < 0.05, R = 0.777). In addition, chamber CH4 fluxes from July to September constituted more than 80% of the total annual emission and were closely correlated with aboveground biomass yield of S. mariqueter. The results indicated that S. mariqueter transportation was the dominant CH4 emission pathway and it provided an efficient route for the belowground CH4 to escape into the atmosphere while avoiding oxidation, leading to CH4 emissions.


Introduction
Methane (CH 4 ) is the second most important greenhouse gas in the atmosphere [1].One of important pathways of CH 4 formation is through fermentation mediated by microbiological process and controlled by oxygen availability and the amount of labile organic matter [2].Another CH 4 formation pathway is carbonate reduction, which is believed to be a main process of CH 4 production in marine systems [3].As CH 4 exchange at air-soil boundary layer is determined by its net result of production and consumption, CH 4 emissions at ecosystem level cannot be accurately quantified without accounting for the potential of CH 4 oxidation [4][5][6].It has been reported that most of the CH 4 formed in the anoxic environment is biologically oxidized to CO 2 before escaping out of the sediments [7,8].In wetland sediments, CH 4 oxidation mediated by methanotrophs could occur in both aerobic and anaerobic conditions [9][10][11].
There are three pathways for dissolved CH 4 in sediment pore-water to reach the atmosphere: molecular diffusion, bubble ebullition, and vascular plant transport [12,13].High belowground CH 4 stock resulted from rapid microbial CH 4 production creates concentration gradients at air-soil boundary layers [14], which drives CH 4 diffusion following the Fick's first law [15].In some cases, the diffusion fluxes were found to be smaller than bubble ebullition and vascular plants transport [16].To adapt to long-term water flooding, wetland plants develop aerenchyma as a pathway for internal gas transport.This pathway is bidirectional such that oxygen can be transported belowground tissues and CH 4 be vented to the atmosphere [17].Wetland plants can also greatly influence CH 4 production and consumption by secreting O 2 and exudates in rhizosphere [18][19][20].In addition, the plant transportation of CH 4 could play an important role in CH 4 emissions [14].This is especially true in vegetated wetlands where plant transportation usually acts as the main emission pattern of CH 4 [20,21], even though the magnitude of CH 4 emissions may vary significantly among different species [20,[22][23][24].
The Yangtze estuary, located in the subtropical area with a clear four seasons, is one of the biggest estuaries in the world.Annually, the Yangtze River transports 4.80 × 10 8 tons of sediment to the East China Sea, and about half of that settles in the estuary, thereby forming an extensive intertidal zone [25].Thus, intertidal marshes in Yangtze estuary may play an important role in carbon cycles including both carbon sequestration and release in the regional scale.Previously, Wang et al. [26] reported the CH 4 emission fluxes and seasonal influence in this area.Bu et al. [27] studied the semi-lunar tidal cycling on sediment CH 4 emissions.Yin et al. [28] revealed the influences of Spartina alterniflora invasions on both CH 4 and N 2 O fluxes from the salt marsh, just north of Yangtze estuary.However, few studies focused on the influence of the endemic plant (S.mariqueter) on CH 4 emissions in Yangtze estuary.Clarification of the relative contribution of two emission pathways, molecular diffusion and plant transport, would be very useful to reveal the plants' contribution to CH 4 emissions.Besides, Yangtze estuary is also faced with the invasion of the exotic Spartina alterniflora, which could greatly decrease the plant density of S. mariqueter, thus changing the CH 4 emissions.Therefore, the objectives of this case study were: (1) to elucidate the role of S. mariqueter on methane emission; (2) to identify the seasonal CH 4 emission patterns; and (3) to examine the relationship between CH 4 emissions and pore-water CH 4 concentrations, at representative point of an intertidal saltmarsh at Yangtze estuary.

Study Site
Yangtze estuarine wetlands continue extending to East China Sea rapidly because of the abundant sediment import from the upper reaches [25].Dongtan in the east Chongming Island is the largest and most conserved intertidal wetland in Yangtze estuary, which is about 100 km 2 and composed of saltmarsh and bare flat.The shape of saltmarsh is semilunar, and the widest middle part is about 4 km (Figure 1).Research and sampling point is selected at about the center of the saltmarsh (31 • 30.111N, 121 • 59.024 E).The point is not submerged during the neap tide and submerged for several hours during the spring tide.Sampling was carried out at low tide.Phragmites, S. mariqueter and the invasive Spartina alterniflora compose the vegetation community of saltmarshes.S. mariqueter is the native species in Yangtze estuary.In general, one axillary bud per corm sprouts to form a new shoot at the depth of 20 cm in spring, which determines that most of the roots distribute within the depth of 20 cm [29].At the base of the new shoot, constituted by a corm, usually one to three rhizomes develop and form new shoots [30].The growing season for S. mariqueter generally occurs from late April to early November with the most active growth occurring during the three summer months (July, August and September), and it dies off at the end of the growing season and they gradually fall over.Along with the continuous sediment deposition, the dead plant will be buried by the sediment and they cannot be flow out with the tide due to the deeper root fixation [31].

Sampling
A sediment core (60-cm long) was collected monthly from April to December 2011 using a PVC pipe (inner diameter: 12 cm, length: 100 cm).Core sampling was not carried out from January to March because the preparatory work was not being done during that period.The core was sealed with tapes and carried back to the laboratory for the measurement of pore-water CH 4 concentration within one day.
Static closed chamber technique was adopted to measure CH 4 fluxes 70 times from February to December 2011.Three stainless-steel collars were inserted into the soil to a depth of 5 cm including one bunch of moderate growth size of Scirpus mariqueter, half hour before gas sampling, and, then, triplicate transparent chambers made of 3-mm thickness Perspex cylinder were placed on collars, and an airtight closure was ensured by water sealing during the measurements.Considering that the vegetation height, even during the peak of the growing season, was no more than 50 cm, the dimensions of the chambers are 50-cm height and 30-cm inner diameter.Small electric fans and kerosene thermometers were fixed inside the chambers for the air blending in the chambers and the temperature measurement inside the chambers.For each chamber, a venting tube (5 mm inner diameter and 18.8 m long) was used to balance the air pressure between inside and outside chambers [32], and a polyethylene pipe connected with a three-way stopcock was fixed on the chamber top for sampling.All the connections and gaps were sealed by silicone gel to ensure chambers were air tight.Immediately after installing each chamber and again 30 min later, 180 mL gas sample was drawn out by using a syringe with a three-way airtight stopcock and injected into a pre-vacuumed air bag (plastic bag plated with Aluminum inside), which is inert to the greenhouse gas.Gas samples were carried back to the laboratory and analyzed for CH 4 .
Environmental factors including light intensity, air temperature and sediment temperature at different depths (5 cm, 10 cm and 15 cm) were recorded synchronously using a TES-1332 photometer, spirit thermometer and geothermometer at each time of gas sampling.The aboveground biomass of S. mariqueter was measured monthly from April to December.Vegetation samples were collected from seven 50 cm × 50 cm randomly selected quadrats, maintaining approximately 3-5 m spacing between quadrats.In each quadrat, the aboveground vegetation was cut carefully, then the number of living shoots were counted, and the shoot weight and cross-section area were measured and recorded.

Pore-Water CH 4 Extraction
The pore-water CH 4 concentration was measured in the laboratory based on static headspace method within one day after the core was sampled.Fifty milliliters of distilled water were transported into a headspace tube equipped with a screws cap which has a hole and silicone septum.The 60-cm long sediment cores were sliced at 1-cm interval, and then 10 mL of the subsamples in each layer were put into the prepared headspace tube quickly.The tube was capped with the silicone septum immediately and set on a shaker for 20 min at 150 rpm to make the sediment fully blend with the aerated distilled water forming slurry and gas equilibrium between slurry and headspace air.After 20 min of shaking, 10 mL gas sample was extracted with a 50 mL syringe from the headspace of each tube for CH 4 concentration measurement.

Sediment Properties
The CH 4 in samples was measured using an Agilent 7890A Gas Chromatography (GC) equipped with a FID detector.Sediment water content was calculated based on the mass method.The wet and dry weight of the sediment samples were separately measured before and after they were freeze-dried with a freeze dryer (CHRIST ALPHA 4-1/LD plus).Dried sediment samples were ground and sieved to analyze sediment organic carbon (SOC) based on K 2 Cr 2 O 7 -H 2 SO 4 oxidation method [33].The sediment water content (SWC) was obtained by using the difference between the wet and dry weights.Average particle size (APS) and medium diameter (MD) are measured by laser granularity meter (Coulter LS13320).Extractable nitrogen (NO 3 − -N and NH 4 + -N) was determined by extracting 10 g of dry sediment sample with 2 mol•L −1 KCl followed by detection of NO 3 − N using a continuous flow analyzer (FUTURA, Alliance Co.), and NH 4 + -N based on standard colorimetric method [34].
For acid volatile sulfide (AVS) measurement, approximately 2 g of sediment were added to the reaction flask and sparged for 2 min with pure N 2 (120 cm 3 •min −1 ).The sulfide in the sediment was liberated by extraction with 1 mol•L −1 HCl for 40 min at room temperature and then trapped in a solution of 0.2 mol•L −1 (CH 3 COO) 2 Zn and 0.1 mol•L −1 CH 3 COONa with a continuous N 2 flow to form ZnS. Each trap was quantified using the methylene blue method [35,36].

Data Calculating
Chamber flux was calculated based on the following equation: FistheCH 4 fluxes at the sediment-air interface (mgCH 4 •m −2 •h −1 ), V is the chamber volume(m 3 ), A is the sediment area in the base of sampling chamber(m 2 ), ∆c ∆t is the change of CH 4 concentration with time ( ), P is the atmospheric pressure in sampling field, R is the gas constant (8.314J•mol −1 •K −1 ), T is the temperature (Kelvin Temperature, K), and m is the molar mass of the gas molecule.
Pore-water CH 4 concentration was calculated by the following equation: C is the CH 4 concentration in pore-water (mgCH 4 •L −1 ), C AIR is the measured CH 4 concentration in the headspace of tubes (mgCH 4 •L −1 ), V AIR is the volume of the headspace of tubes (L), and α is the Bunsen coefficient.V WATER is the total volume of aerated distilled water in tubes and the pore water in sediment samples (L).V POREWATER is the volume of the pore water of the sediment sample (L).C BLANK is the measured CH 4 concentration of the pure water (mgCH 4 •L −1 ).
Diffusive fluxes of CH 4 were calculated from linear pore-water concentration gradients according to Fick's first law assuming steady state conditions.The diffusion equation is described as Equation (3): where F cal is the calculated diffusion CH 4 flux (µmolCH 4 •cm −2 •s −1 ), ∆C ∆Z represents the CH 4 concentration change with sediment depth (µmolCH 4 •cm −3 •cm −1 ), ∆Z is the diffusion distance (cm), C S is measured CH 4 concentration in the surface sediment pore water (µmolCH 4 •cm −3 ), C A is the saturation concentration of dissolved CH 4 in ambient air (µmolCH 4 •cm −3 ), and Ds is the effective diffusion coefficient for CH 4 in sediment pore-water.Ds is calculated by the polynomial regression equation obtained by the measured diffusion coefficients for CH 4 in water (Dw) in the range 0 • C to 35 • C (cm 2 •s −1 ) (83rd Edition of the Handbook of Physics and Chemistry) which yielded the relationship [37]: and then porosity corrected by Equation ( 5) from Lerman [38]: where φ is sediment porosity.To make the calculated diffusion flux be consistent with the total emission flux, F cal was finally expressed in mgCH 4 •m −2 •day −1 .

Statistical Analysis
The total CH 4 fluxes were presented as means of the replications.SPSS (Version 17.0) was used to perform all of the statistical data tests.Pearson correlation analyses were used to examine the relationship between fluxes and the measured environmental variables.The statistical results were regarded as significant if the P values were lower than 0.05 or 0.01.

Environmental Factors
Table 1 shows seasonal climatic, vegetation and soil physiochemical parameters of experimental site.Air temperature exhibited significant seasonal variation with the highest and lowest temperatures in July and February, respectively.Highest biomass of S. mariqueter occurred in August (692.6g•m −2 ).The number of living shoots increased with the biomass and also reached the highest value in August.Since then, the plant began to wither and almost all of them fell over on the sediment surface in December.SOC did not show substantial changes throughout the year.SWC stayed at a relatively low level from February to April but increased in the following months.APS and MD indicated that sediment was dominated by silt fractions.In terms of the extractable nitrogen, NH 4 + -N content was generally higher than that of NO 3 − -N, which probably indicated that the tidal marsh was generally under denitrification condition rather than nitrification.On the other hand, the irregular variation of AVS content throughout the year suggested that the redox potential of the microenvironment may differ a lot in this estuarine system (Table 1).AVS is mainly produced by the sulfate reduction process which occurred after the O 2 and NO 3 − were utilized.Thus, AVS can be seen as an index of the sediment redox condition.Monthly AVS distribution in the sediment profile is shown in Figure 2, which indicated that the AVS in the surface sediment is usually lower than the deeper sediment.Besides, the AVS showed complex variation along with depth and month.

CH 4 Fluxes
Measured chamber CH 4 fluxes showed a significant seasonal variation throughout the year with the highest (91.2 mgCH 4 •m −2 •day −1 ) and lowest (2.30 mgCH 4 •m −2 •day −1 ) in August and March, respectively (Figure 3).The annual average chamber CH 4 flux in the salt marsh was 24.0 mgCH 4 •m −2 •day −1 .The chamber CH 4 flux did not increase significantly until June when the biomass of S. mariqueter rather than temperature increased significantly, and the chamber CH 4 flux was significantly positively correlated with the aboveground biomass (p = 0.001, R = 0.928) (Figure 4).Chamber CH 4 flux during summer (July, August and September) constituted more than 80% of the total annual emission.Calculated diffusion CH 4 fluxes between the top 1 cm sediment and air were very low (<0.55 mgCH 4 •m −2 •day −1 ) due to the low pore-water CH 4 concentration in surface sediment (Figure 3).There was evident seasonal variation of diffusion CH 4 fluxes with the highest (0.55 mgCH 4 •m −2 •day −1 ) and lowest (0.0035 mgCH 4 •m −2 •day −1 ) diffusion flux in July and September, respectively (Figure 3).Overall, gas molecular diffusion method contributed a little to the total CH 4 emission with the highest proportion of 5.53% in May.
In addition, regression analysis of different temperatures (air temperature and the sediment temperature at different depths 5 cm, 10 cm and 15 cm) and chamber CH 4 fluxes revealed that CH 4 fluxes were best fitted with the temperature at the sediment depth of 15 cm where the roots of S. mariqueter were concentrated (Figure 5).

Pore-Water CH 4 Concentration
Top 1 cm pore-water CH 4 concentration did not show any apparent seasonal variation.It was higher in the first half year and decreased to a lower value in the second half year (see Figure 3).Pore-water CH 4 concentration in the top 1 cm was much lower than in the deeper layer sediment, resulting in a low diffusion flux.In the vertical profile, CH 4 concentration increased significantly with depth.Although the pore-water CH 4 concentration in deep sediment layers did not exhibit an apparent synchronously monthly variation with temperature, it did show an evident seasonal change (Figure 6).Pore-water CH 4 concentration of deep sediment layer increased from May to July, decreased through October, and then reached the highest concentration (8.30 mgCH 4 •L −1 ) in November.After that, deep layer sediment pore-water CH 4 concentration decreased again until the next April (Figure 6).Along with the deep pore-water CH 4 concentration variation, the depth of the steepest increase in CH 4 concentration fluctuated up and down regularly during the year.The transition point of CH 4 concentration increment usually occurred after the concentration exceeding 0.1 mgCH 4 •L −1 .The location of this interface moved up from April to July and consistently stayed at about 10 cm depth through September, and then it went down from October to November (Figure 6).
Relationship between the pore-water CH 4 concentration and total CH 4 fluxes were also analyzed through Pearson correlation analysis.Significant correlations were observed between CH 4 flux and pore-water CH 4 concentration only at the depths of 11-15 cm (p < 0.05, R = 0.732) and 16-20 cm (p < 0.05, R = 0.777).(see Table 2).

Dominant Mechanism of CH 4 Emission
Different types of wetlands may be dominated by different CH 4 emission patterns (plant mediated emission, bubble emission and the molecule diffusion) [39][40][41].Anaerobic conditions in wetlands are favorable for CH 4 production and accumulation, which finally create a CH 4 concentration gradient between the underground and atmosphere [42].Molecular diffusion is driven by the concentration gradient following Fick's first law of diffusion [37].Based on Fick's first law, we calculated the molecular diffusive CH 4 flux from the saltmarsh into the atmosphere.In this study, the calculated diffusion fluxes stayed at very low level with the highest diffusion flux of 0.55 mgCH 4 •m −2 •day −1 in July.The proportion of diffusion fluxes accounted for 0.0048-5.53% of total monthly CH 4 emission during the year.The specific reason for the weak diffusion flux is unclear due to the lack of ground water level, wind speed and the tide movement measurements.It is speculated that the wet condition in the saltmarsh is one of the most probable reasons, since it is known that water condition of the media can greatly influence the gas diffusion efficiency [43].The S. mariqueter marsh in Yangtze estuary is located at the lower position of the tidal flat and was regularly submerged by the tide water.Surface sediments are always kept under a wet condition (this can be seen from the surface sediment water content in Table 1), which greatly inhibits the gas diffusion from surface sediment into air.In addition, the surface sediment usually stayed at a relative oxic condition, which can be seen from the low AVS concentrations in the surface sediment.Pore-water CH 4 concentrations was significantly correlated with AVS concentrations (p < 0.01, R 2 = 0.265).It was also proven in the previous study that oxygen could easily penetrate sediment creating an oxic condition when the sediment surface exposes to the air [44].Previous studies reported that the existence of oxidizing zones can greatly reduce the diffusion CH 4 flux.According to Yun et al. [45], the aerobic methanotrophs could oxidize more than 90% of the CH 4 produced in the anoxic conditions.Similar results were also obtained by Liebner et al. [46] who found that the diffusive CH 4 flux from the sediment surface only accounted for less than 2% of the total flux into the atmosphere in an alpine fen due to the linearly decreased pore water CH 4 concentration at the depth of 0-15 cm.Our research also indicated that the intense oxidation of surface pore-water CH 4 directly led to the weak diffusive CH 4 flux.
In this study, S. mariqueter was found to facilitate CH 4 emissions and the plant emitted more than 94% of the total CH 4 , which was consistent with many previous studies (Table 3).The highest chamber CH 4 flux appeared in August when S. mariqueter was in the utmost abundant with the biggest aboveground biomass throughout the year of 692.6 g•m −2 .The chamber CH 4 fluxes of the three thriving months of S. mariqueter (July, August and September) constituted more than 80% of the total annual emission.The chamber CH 4 flux decreased greatly at the end of growing season, which was consistent with the research of Chmura et al. [47].The aerenchyma of wetland plants had two opposite effects on CH 4 emission.Negatively, it could transport oxygen downward, creating an aerobic zone in rhizosphere [19].The enhanced oxygen availability in sediment could inhibit CH 4 production and accelerate CH 4 consumption [46,48].On the other hand, presence of aerenchyma in wetland plants also provides gas channels for CH 4 escaping from underground to the atmosphere [16,49].In a previous research of similar salt marsh, the net effect of these two processes was found to change along with the S. mariqueter growth stage [26].It is indicated that the quantity of upward transported CH 4 to the air through S. mariqueter exceeded the oxidized CH 4 by downward transported O 2 through S. mariqueter.
The fact that chamber CH 4 fluxes were significantly correlated with aboveground biomass of S. mariqueter suggested significant effect of marsh plant growth on CH 4 emission.Besides, regression analysis of different temperatures (air temperature and the sediment temperature at different depths 5 cm, 10 cm and 15 cm) and chamber CH 4 fluxes revealed that CH 4 fluxes were best fitted with the temperature at the sediment depth of 15 cm where the roots of S. mariqueter were concentrated [29] (Figure 5).This indicates that the rhizospheric temperature would be a good predictor of CH 4 emission.While higher temperature usually leads to stronger methanogenesis and CH 4 accumulation [50], our result emphasizes the significance of the temperature increase in rhizosphere that enhanced the CH 4 emission.These observations further pointed to the great contribution of plant transportation to overall CH 4 emissions in this salt marsh ecosystem.Previous research showed that the wetland plant, no matter living or dead, could provide a continuous escaping route for CH 4 even in winter [51].Nonetheless, our results indicated that the strongest transportation capacity of CH 4 by individual plant appeared in August with the highest CH 4 flux at 2.34 × 10 −2 mgCH 4 •stem −1 •day −1 in these Yangtze estuary wetlands.

The Implication of Pore-Water CH 4 Concentration for CH 4 Emission
Static pore-water CH 4 concentration from the surface to deep sediment was a net result of CH 4 production and consumption, which could be affected by many environmental factors such as the redox potential, sediment organic carbon content, the distribution of methanogens and methanotroph, etc. [59][60][61].In this study, the deep pore-water CH 4 concentration did not exhibit synchronously seasonal variation with temperature (Figure 2), which was consistent with previous study in an alpine fen [46].Sediment pore-water CH 4 concentration profiles showed that there was an interface where the CH 4 concentration steeply increased after it exceeded 0.1 mgCH 4 •L −1 (Figure 6).Although the CH 4 concentration in the surface 10 cm sediment did not show apparent seasonal variations, the depth of this interface did vary and was shallower (~10 cm) in summer than in other seasons (Figure 6).Previous studies showed that the surface sediment pore-water CH 4 concentration was influenced by the oxygen concentration easily and more than 90% of the CH 4 produced in anoxic conditions could be oxidized in the upper aerobic conditions [62].Conversely, the pore-water CH 4 concentration could be seen as the indicator of environmental oxidation-deoxidation status.In our study, we attributed the upper layer of the sediment where CH 4 concentration was lower than 0.1 mg CH 4 •L −1 to the active CH 4 oxidization (aerobic and anaerobic).Above this interface, CH 4 concentration was controlled by the aerobic and anaerobic oxidation process and maintained at a low level (Figure 5).Below this interface, CH 4 concentration increased sharply, which was controlled by the anaerobic oxidation and production and diffusion from deeper layer.The effects of temperature on the depth of this "interface" could be explained from two aspects.On the one hand, under high temperature, oxygen penetration into sediment was generally reduced because the solubility of atmospheric oxygen decreased with temperature increase [44].On the other hand, intensive respiration took place at high temperature, which accelerated the oxygen consumption within the surface sediment zones [63,64].
The analysis of the correlation between chamber CH 4 fluxes and the pore-water CH 4 concentration in different depths showed that only the pore-water CH 4 concentration at the depths of 11-15 cm and 16-20 cm, where the dense roots of S. mariqueter are distributed, best predicted the chamber CH 4 fluxes (see Table 2).The increased temperature in summer raised the depth of reactive layer for CH 4 production to the depth of plant root distribution, which directly promoted the rhizospheric CH 4 uptake from sediment pore water and prevented the CH 4 oxidation in diffusion.The positively correlated chamber CH 4 fluxes and rhizospheric CH 4 concentration revealed that overall S. mariqueter acted as a promoter of CH 4 emission when taking both rhizospheric CH 4 oxidation and transportation into consideration.
The lack of significant relationship between the surface (0-10 cm) pore-water CH 4 concentration and chamber CH 4 fluxes further indicated the small contribution of gas molecular diffusion to CH 4 emission.Although the pore-water CH 4 concentration was apparently higher in the deeper layers, it had little effect on CH 4 emission, probably because most of the upward diffusive CH 4 was anaerobically oxidized due to the long-distance migration [64,65].In addition, the CH 4 in the deep sediment scantly contributed little to CH 4 fluxes through bubble emission because the CH 4 concentration was not high enough to form gas bubbles.Previous studies have indicated that bubbles could not form until pore-water CH 4 concentrations exceeded 7.1-8.0mgCH 4 •L −1 [66].In this research, pore-water CH 4 concentration in surface was only about 0.001-0.01mgCH 4 •L −1 , and even the highest pore-water CH 4 concentration in deep sediment layer (below 45 cm) was under 8.0 mgCH 4 •L −1 .

Conclusions
This study demonstrated that S. mariqueter marsh in Yangtze estuary acted as a net source of atmospheric CH 4 .The annual average chamber CH 4 flux was 24.0 mgCH 4 •m −2 •day −1 with the peak flux in August when the oxidization layer became shallow (about 10 cm) and the S. mariqueter thrived.The calculated diffusion CH 4 fluxes were no more than 6% of the total fluxes, indicating that molecular diffusion was not a major pathway of CH 4 emission in this salt marsh.Analysis of seasonal variation of pore-water CH 4 concentration identified the dominant emission pattern as plant transport, which was evident by significantly positive correlation between CH 4 concentrations in rhizosphere and chamber CH 4 fluxes.In addition, generally low pore-water CH 4 concentration (much less than 8.0 mg CH 4 L −1 ) throughout the year prevented the formation of CH 4 gas bubbles from ebullition.While wetland plant could exert an influence on rhizosphere CH 4 oxidation, the transport function of the plant played a more important role in CH 4 emission in the Yangtze estuary S. mariqueter marsh.

Figure 1 .
Figure 1.The sampling site in Yangtze estuary.The studied salt marsh was mainly dominated by the endemic species of S. mariqueter.
AAT: average air temperature; ATR: Air temperature range during the sampling day; GT5: Sediment temperature at the depth of 5 cm; GT10: Sediment temperature at the depth of 10 cm; GT15: Sediment temperature at the depth of 15 cm) are the mean value of the daily variation.PAR (Photosynthetically Active Radiation) is shown as the variation range of the daily continuous observations.SOC is the average soil organic carbon values of the surface 25 cm sediments.The rest of the sediment parameters (SWC, APS, MD, NH 4 + -N, NO 3 − -N and AVS) are the values of the surface 1 cm sediment.

Figure 2 .
Figure 2. AVS distribution in the sediment profile.

Figure 3 .
Figure 3.The total chamber CH 4 fluxes, calculated diffusion CH 4 fluxes, the top 1 cm pore-water CH 4 concentration and the biomass of S. mariqueter in each month during the year.CH 4 fluxes are presented as column graphs, while the pore-water CH 4 concentration and biomass are presented as line and symbol graphs.The black graphs in the figure are integrated with the left black Y axis and the blue graphs are integrated with the right blue Y axis.

Figure 4 .
Figure 4. Correlation relationship between the CH 4 fluxes and the aboveground biomass of S. mariqueter.

Figure 5 .
Figure 5. Relationship between chamber CH 4 fluxes and temperature.Abbreviations: 5 cm GT, 10 cm GT and 15 cm GT represent the sediment temperature at depths of 5 cm, 10 cm and 15 cm, respectively (n = 70).

Figure 6 .
Figure 6.Monthly variation of pore-water CH 4 concentration.A steep increase in pore-water CH 4 concentration occurred when it exceeded 0.1 mgCH 4 •L −1 for each month.The light color located above the depth of this steep increase indicates the low CH 4 concentrations, while the dark color indicates the high CH 4 concentrations.

Table 1 .
Climatic, plant and soil physiochemical parameters of experimental site.

Table 2 .
Correlation relationships between monthly methane (CH 4 ) fluxes and the pore-water CH 4 concentration at the different depths a , * p < 0.05.The pore-water CH 4 concentration was measured at 1-cm interval, but the mean of the pore-water CH 4 concentrations at 5-cm interval was used to analyze the correlation relationship between the CH 4 fluxes and the underground CH 4 concentrations at different depths of each month.b The correlation coefficient "R" for CH 4 fluxes and the averaged pore-water CH 4 concentrations in different depths. a

Table 3 .
Plant contributions of different species to CH 4 emissions.The proportion of plant transported CH 4 fluxes in the total CH 4 fluxes.b Different approaches to quantify the plants' effects on CH 4 emissions.c,d The most commonly adopted approaches in studying the plant transportation capacity for CH 4 .c The plant stems covered by the static closed chamber were clipped leaving the stem section on the sediment surface; d Plants covered by the static closed chamber were uprooted. a