Biogeosciences Relation between methanogenic archaea and methane production potential in selected natural wetland ecosystems across China

Methane (CH4) emissions from natural wetland ecosystems exhibit large spatial variability at regional, national, and global levels related to temperature, water table, plant type and methanogenic archaea etc. To understand the underlying factors that induce spatial differences in CH 4 emissions, and the relationship between the population of methanogenic archaea and CH 4 production potential in natural wetlands around China, we measured the CH 4 production potential and the abundance of methanogenic archaea in vertical soil profiles sampled from the Poyang wetland in the subtropical zone, the Hongze wetland in the warm temperate zone, the Sanjiang marsh in the cold temperate zone, and the Ruoergai peatland in the Qinghai-Tibetan Plateau in the alpine climate zone. The top soil layer had the highest population of methanogens (1.07–8.29 ×109 cells g−1 soil) in all wetlands except the Ruoergai peatland and exhibited the maximum CH4 production potential measured at the mean in situ summer temperature. There is a significant logarithmic correlation between the abundance of methanogenic archaea and the soil organic carbon ( R2 = 0.72, P < 0.001, n = 13) and between the abundance of methanogenic archaea and the total nitrogen concentrations ( R2 = 0.76,P < 0.001, n = 13) in wetland soils. This indicates that the amount of soil organic carbon may affect the population of methanogens in wetland ecosystems. While the CH 4 production potential is not significantly related to methanogen population ( R2 = 0.01, P > 0.05, n = 13), it is related to the dissolved organic carbon concentration ( R2 = 0.31,P = 0.05, n = 13). This suggests that the methanogen population might be not an effective index for predicting the CH 4 production in wetland ecosystems. The CH 4 production rate of the top soil layer increases with increasing latitude, from 273.64 μg CH4 kg−1 soil d−1 in the Poyang wetland to Correspondence to: W. X. Ding (wxding@mail.issas.ac.cn) 664.59 μg CH4 kg−1 soil d−1 in the Carex lasiocarpamarsh of the Sanjiang Plain. We conclude that CH 4 production potential in the freshwater wetlands of Eastern China is mainly affected by the supply of methanogenic substrates rather than temperature; in contrast, low summer temperatures at high elevations in the Ruoergai peatland of the Qinghai– Tibetan Plateau result in the presence of dominant species of methanogens with low CH 4 production potential, which in turn suppresses CH 4 production.


Introduction
The methane (CH 4 ) concentration in the atmosphere has increased by 153% since 1750.Natural wetlands have emitted 100-231 Tg CH 4 year −1 , which accounts for 20-39% of the annual global CH 4 emission (IPCC, 2007).The CH 4 emission from wetlands increased by 7% from 2003 to 2007 (Bloom et al., 2010).Up to date, many studies have focused on the CH 4 emission from wetland ecosystems, in an attempt to quantify the CH 4 source strengths from wetlands (Whalen and Reeburgh, 1992;Christensen, 1993;Christensen et al., 2003).Meanwhile, many process-based models have been developed to predict CH 4 emissions from natural wetlands (Walter and Heimann, 2000;Walter et al., 2001a, b;Petrescu et al., 2008Petrescu et al., , 2010)).However, there is great uncertainty in the magnitude and distribution of CH 4 sources from regional to global scales because of the large spatial and temporal variations in emissions across individual wetland types (Melling et al., 2005;Wang and Han, 2005;Chen et al., 2008).Several factors that affect CH 4 emissions have been identified, such as temperature (Westermann, 1993), plant type (Bartlett et al., 1992), primary production (Whiting and Chanton, 1993), the water table (Moore and Dalva, 1993;Frenzel and Karofeld, 2000;Ding et al., 2010), and the thaw depth at permafrost sites (Walter and Heimann, 2000).
Published by Copernicus Publications on behalf of the European Geosciences Union.D. Y. Liu et al.: Relation between methanogenic archaea and methane production potential Saarnio et al. (1998) concluded that the decomposition of root exudates is suppressed at low temperatures, and thus, CH 4 production and emission are reduced.Similarly, Ding et al. (2004) suggested that the high elevation (height above sea level: > 3400 m) of the Ruoergai peatland in the Qinghai-Tibetan Plateau leads to low temperatures in summer, which in turn lowers the supply of methanogenic substrates and CH 4 production.Shannon and White (1996) and Chasar et al. (2000) concluded that the CH 4 production in wetlands is affected by the acetate supply through acetate fermentation and/or the CO 2 reduction potential.On the basis of a study conducted in an ombrotrophic bog in Michigan, Avery et al. (2003) pointed out that the exponential increase in the rate of CH 4 production with temperature is due to an increase in the number of available substrates and is not associated with changes in the composition and populations of methanogens.On the other hand, Duddleston and Kinney (2002) argued that temporal accumulation of acetate in the Turnagain bog of Alaska is attributable to the absence of acetoclastic methanogenesis.Horn et al. (2003) found that no active acetoclastic methanogenesis occurred in an acidic peatland in Germany and that the addition of acetates or other volatile fatty acids reduced rather than increased CH 4 production, possibly because of the inhibition of hydrogenotrophic methanogenesis (Brauer et al., 2004).Cadillo-Quiroz et al. (2006) found that in the case of vertical profiles of the McLean and Chicago bogs in central New York State, USA, the difference between the methanogenesis in different soil layers is strongly associated with the population and activity of methanogens using the quantitative polymerase chain reaction (qPCR).To date, little is known about the size of the methanogenic archaea population and the CH 4 production potential in freshwater natural wetlands in China.
We have examined the spatial variation in the CH 4 production potential and methanogenic population in four typical freshwater wetlands across China.The objectives of this study were (1) to understand the relationship between the population of methanogenic archaea and CH 4 production potential, and (2) to evaluate the underlying factors that determine the spatial variation of the CH 4 production potential in wetland ecosystems.For this purpose, a method based on the qPCR of methanogens was used and soil slurries were incubated at mean summer temperatures in order to simulate in situ conditions.

Site description and soil sampling
Four typical natural wetlands in China were selected for this study (Fig. 1).The eutrophic freshwater marsh is located at the Chinese Academy of Sciences Sanjiang Mire Wetland Experimental Station in Tongjiang City of the Heilongjiang province (Table 1).The elevation is 56 m and the mean an- nual precipitation is approximately 600 mm.The mean annual temperature is 1.9 • C, ranging from −18.8 in January to 20.8 • C in July.Marsh initiation began during the late-Pleistocene epoch due to convergence of the Heilongjiang River, Songhuajiang River and Wusulijiang River, and blockage of water seepage by clayey soil.Wetland vegetation varies from Calamagrostis angustifolia to Carex lasiocarpa as standing water depth increases (Ding et al., 2002).Vegetation in the C. lasiocarpa marsh is composed of 90% C. lasiocarpa and 10% Glyceria spiculosa.The soil profile is composed of standing water, root layer, peat layer and grey soil layer.Vegetation in the C. angustifolia marsh is purely C. angustifolia, but its profiles are absent from the peat layer.The Ruoergai peatland is situated in Ruoergai county, Sichuan province, a typical part of the Qinghai-Tibetan Plateau in the alpine climate zone, and is the largest peatland in China.The dominant vegetation is Carex muliensis and Carex meyeriana.Similar to the Sanjiang Plain, the mean annual temperature is 1 • C and the mean annual precipitation is 650 mm.Ruoergai highland reaches 3400 m and shows little annual variation of temperature.The lowest and highest average monthly temperatures were −10.5 • C in January and 10.7 • C in July (Zhao, 1999).Low temperatures in summer greatly reduce the decomposition rate of plant litters and accelerate the deposit of peat, thus the depth of peat at the sampling site is more than 1 m.Hongze Lake, with an area of 2069 km 2 and an elevation of 12 m, is located in the northern part of the Jiangsu province and is the fourth largest freshwater lake in China.The mean annual precipitation is approximately 926 mm and the mean low and high monthly temperatures are 0 • C and 27.5 • C, respectively.This lake was formed on an alluvial plain in the middle reaches of the Huaihe River due to river course blockage during the Tang dynasty (Zhu, 1991).At present, the average water depth is approximately 4 m.There is a variety of wetland plants in the land-water ecotone, such as Phragmites australis, Zizania caduciflora, Soil samples were collected from 25 September to 10 October, 2009.Three 5 m×5 m sampling plots were randomly established at each site.Five soil samples were taken for each layer at different positions in each plot using a 2.5 cm diameter stainless steel soil sampler.All samples of the same layer from each plot were then carefully mixed to form a composite.All samples were immediately stored in sterile bags, kept on ice in coolers, and directly transported to the laboratory.One subsample was stored at −20 • C for DNA extraction, another subsample was used for the measurement of CH 4 production potential, and the other was air-dried for the measurement of soil properties.

Methane production potential measurement
Methane production potential of wetland soils was determined using a slightly modified method from Galand et al. (2003).Ten grams of fresh soil sample (on an oven dried basis) were weighed and placed in a 100 ml glass jar.The ratio of soil to water was adjusted to be 1:5 with distilled water.The contents of the jar were mixed thoroughly using a glass rod.The jars were vacuumed and then injected with pure nitrogen gas (N 2 ) using an atmospheric pressure balance.The above procedure was replicated three times to obtain completely anoxic conditions.The jars were incu-bated in the dark at different temperatures.Incubation temperatures were chosen based on mean summer temperatures of wetlands sampled and was 10 • C for Ruoergai, 20 • C for Sanjiang, 28 • C for Hongze, and 30 • C for Poyang.We also measured the CH 4 production potential of the Ruoergai peatland at temperature of 20 • C to evaluate the effect of temperature increases on CH 4 production.All treatments were performed with three replicates.During the incubation, the CH 4 concentration in the jar headspace was measured daily by sampling 1.0 ml headspace gas with a precision sampling syringe (Valco Instruments, Baton Rouge, LA, USA) and applied to a Shimadzu GC12A with FID and a 2-m Porapak Q (80/100 mesh) column.The oven, injector, and detector temperatures were 80 • C, 200 • C, and 200 • C, respectively.The carrier gas (N 2 ) flow rate was 30 ml min −1 and flame gases (H 2 and O 2 ) were set at 20 and 30 ml min −1 , respectively.The standard gas was cross-checked by the National Institute for Agro-Environmental Sciences, Japan.The rate of CH 4 production was calculated from the slope of the linear regression given by the graph of CH 4 concentration increase over time.The amount of CH 4 produced in the jars was sum of CH 4 in the jar headspace and CH 4 dissolved in water; the latter was calculated based on Bunsen solubility coefficients of CH 4 (Ding et al., 2010).

Soil analysis
Soil pH was determined using a glass electrode and a soil to water ratio of 1:5.Soil organic carbon (SOC) was measured by wet oxidation using dichromate in acid medium followed by the FeSO 4 titration method, and total nitrogen (TN) was determined by the Kjeldahl method (Lu, 2000).For dissolved organic carbon (DOC), 10 g fresh soil (on an www.biogeosciences.net/8/329/2011/Biogeosciences, 8, 329-338, 2011 D. Y. Liu et al.: Relation between methanogenic archaea and methane production potential oven dried basis) was incubated with 50 ml distilled water for 30 in on an end-over-end shaker at 25 • C, then centrifuged for 20 min at 8000 rpm.The extracted solutions were passed through a 0.45-µm filter paper and analyzed on a Shimadzu C analyzer (TOC Vcph, Shimadzu, Kyoto, Japan).

DNA extraction and real-time PCR
Total DNA from three replicates of soil samples was extracted using the FastDNA SPIN Kit for soils (BIO 101, Qbiogene, Carlsbad, CA, USA) according to the manufacturer's instructions (Cahyani et al., 2008).Real-time PCR was carried out to quantify the methanogenic archaea 16S rRNA genes in soil samples using LightCycler ST300, Light-Cycler Software Version 3.5 (Roche Diagnostics, Germany) and SYBR Premix Ex Taq (TaKaRa, Japan).The primer pair 1106F (forward) and 1378R (reverse) was used for PCR amplification targeting the 16S rRNA gene of methanogenic archaea (Watanabe et al., 2006(Watanabe et al., , 2009)).Each reaction mixture (25 µl) consisted of 12.5 µl 1×SYBR Premix Ex Taq, 0.25 µl of each primer, 1 µl of DNA template diluted 20 times, and sterilize distilled water.The real-time PCR program was initiated by a denaturation step at 95 • C for 10 min, followed by 35 cycles of denaturation at 95 • C for 10 s, annealing at 57 • C for 10 s, and extension at 72 • C for 6 s.A standard curve based on known methanogenic archaea copy numbers (1.97 to 19.7×10 8 copies µl −1 ) was generated using purified PCR product (Jia and Conrad, 2009).

Statistical analyses
All data were expressed on the basis of oven-dried soil.The means and standard errors were calculated with three to six replicates.All statistical analyses were performed with SPSS 11.0 software.Statistically significant differences of means in all soils were judged by one-way analysis of variance (ANOVA) and least significant difference (LSD) calculations at a 5% significance level.The P-values for the effects between different wetlands and between different depths of the same wetland were adjusted using a Bonferroni correction.Regression analyses were used to test relationships between population of methanogenic archaea and SOC, TN, and the average CH 4 production potential.

Biogeochemical characteristics of soils
In Table 1, we present the site characteristics and soil properties of the selected five wetlands.Soil pH in the top layer of the Sanjiang marsh and Poyang wetland is significantly lower than the top-layer pH of the Hongze wetland and the Ruoergai peatland.SOC in the top layer of the Ruoergai peatland is 134.83 g C kg −1 soil, significantly higher than that of other wetlands in this study, and the lowest SOC is measured in the  Poyang wetland in the subtropical zone.SOC concentration in the top layer of the Sanjiang marsh in the cold temperate zone varies greatly with plant types and is 38.73 g C kg −1 soil in the C. angustifolia marsh, which is only one third of the value in the C. lasiocarpa marsh, and also dramatically lower than in the Hongze wetland in the warm temperate zone.SOC concentration in the vertical profile sharply increases with depth in the Ruoergai peatland, but to some extent decreases in other wetlands.Soil TN shows a similar pattern to SOC (Table 1).The concentrations of DOC increase with soil depth in all wetlands except the Sanjiang Plain C. lasiocarpa marsh where most roots are distributed in the top layer.The highest top layer soil DOC concentration was measured in the C. lasiocarpa marsh, whereas the lowest in the Poyang wetland.The sulfate (SO 2− 4 ) concentration ranges from 0.04 to 1.37 g kg −1 soil, and is much higher in the Hongze wetland than in other wetlands, which were below 0.12 g kg −1 .

Methane production potential
In the vertical profile, the highest CH 4 production potential occurs in the top soil layer, and is significantly higher than the values in the lower soil layers in all wetlands except the Ruoergai peatland (Fig. 2).The CH 4 production potential of the 0-20 cm soil in the C. angustifolia marsh amounts ) to 191 µg kg −1 soil d −1 , and is 8.86 and 7.23 times greater than those of the 20-40 and 40-60 cm soils, respectively.Methane production potential of 0-30 cm soil in the C. lasiocarpa marsh is 184-fold higher than that of the lower layer soils, and is 3.48 times as much as the value in the corresponding layer of soil in the C. angustifolia marsh.In the top soil layer of the Hongze wetland, CH 4 production potential is 6.40 times that of the lower soil layer, and 5.12-15.76times greater than that of the Poyang wetland.In the Ruoergai peatland, CH 4 production potential decreases as depth increases at both 10 • C and 20 • C, however difference between soil layers is not significant at the 10 • C incubation temperature, but is significant at 20 • C. When comparing the same soil layer at different temperatures, no significant difference is observed between 10 • C and 20 • C (Fig. 2).The highest CH 4 production potential in the top soil layer is observed in the Sanjiang Plain C. lasiocarpa marsh and was 665 µg CH 4 kg −1 soil d −1 , whereas the lowest values occur in the Ruoergai peatland (10.73 µg CH 4 kg −1 soil d −1 ).The CH 4 production potential in the top soil layer does not increase with the increase of wetland in situ temperature and water table depth in China (Fig. 3).However, it increases with the water table depth in wetlands if Hongze Lake (HZ1) is excluded for analysis.

Methanogen population
The population of methanogenic archaea in the top soil layer of the Hongze wetland is highest relative to the other wetlands counting 8.29×10 9 cell g −1 soil, whereas the lowest was measured in the Poyang wetland (1.07×10 9 cell g −1 soil) (Fig. 4).In the Sanjiang Plain, the population of methanogens in the top soil layer of the C. lasiocarpa marsh is statistically significantly lower than that in the corresponding soil layer of C. angustifolia marsh.Both, however, are dramatically lower than those in the Ruoergai peatland.In the vertical profile, the population of methanogens in the top soil layer is significantly higher than that in the Sanjiang marsh lower soil layer, as well as of the Hongze and Poyang wetlands.In contrast, the diversity of the methanogenic community increases with depth in the Ruoergai peatland.
Regression analysis shows that DOC significantly increases exponentially with SOC concentration in the soils of all wetlands (R 2 = 0.32, P = 0.04, n = 13), while CH 4 production potential is significantly correlated with DOC (R 2 = 0.31, P = 0.05, n = 13) rather than SOC (R 2 = 0.001, P = 0.92, n = 13).When integrating all factors including pH, SOC, DOC and temperature for the multiple-regression www.biogeosciences.net/8/329/2011/Biogeosciences, 8, 329-338, 2011 D. Y. Liu et al.: Relation between methanogenic archaea and methane production potential analysis, we identify DOC is the primary factor and could explain 26.74% of the variation in the CH 4 production potential between different wetlands (P < 0.05).There is a significant logarithmic correlation between the population of methanogens and the SOC and TN concentrations (Fig. 5), however the population of methanogens is not significantly associated with CH 4 production potential (Fig. 6).The specific CH 4 production potential in the top soil layer of the Sanjiang Plain C. lasiocarpa marsh and the Poyang wetland is significantly higher than those in corresponding soil layers of other wetlands (Fig. 7).It is considerably reduced in the lower soil layer relative to the top layer for all wetlands.

Discussion
Methane emissions from natural wetland ecosystems show large spatial variations (Yavitt et al., 1988;Crill et al., 1992;Saarnio et al., 1998).Ding et al. (2004) found that CH 4 emissions from the Ruoergai peatland vegetated with Carex muliensis and Carex meyeriana in the Qinghai-Tibetan Plateau average 2.87 mg CH 4 m −2 h −1 over the growing season, which correlates to only one sixth of the values measured at the Carex freshwater marsh in China's Sanjiang Plain.In this study we find that the highest CH 4 pro- duction potential occurred in the C. lasiocarpa marsh of the Sanjiang Plain and the lowest occurred in the Ruoergai peatland (Fig. 2).This is completely consistent with results of previous field measurements (Ding et al., 2004).
The high CH 4 emission from freshwater marshes and the low emission from peatlands have also been measured in Canada and USA (Harriss et al., 1985;Bubier et al., 1993).
However, CH 4 production potential measured at the mean in situ summer temperature roughly decreased as the air temperature increased from the Sanjiang Plain freshwater marsh in the cold temperate zone, to the Hongze wetland in the warm temperate zone, and then to the Poyang wetland in the north subtropical zone (Fig. 3).Previous studies have shown that the mean CH 4 flux from wetlands over the growing season was 19.65 mg CH 4 m −2 h −1 in the Sanjiang Plain of Northeast China (Ding et al., 2004), 9.56 mg CH 4 m −2 h −1 in Inner Mongolia (Duan et al., 2007), and 4.74-5.13mg CH 4 m −2 h −1 in Southeast China (Tong et al., 2009), which roughly decreases from high-latitude to low-latitude.This indicates that CH 4 production ability of wetlands may increase as the latitude increases in the eastern part of China.Freitag and Prosser (2009) observed that the CH 4 production rate was significantly correlated with the mcrA transcript:gene ratio in an ombrotrophic peatland in the North Wales, UK.In the present study, a simultaneous decrease in the CH 4 production rate and methanogen population with decreasing soil depth in all wetlands except the Ruoergai peatland suggests that CH 4 production potential is likely determined by the abundance of methanogenic archaea (Figs. 2 and 4).On the national scale, however, no significant relationship is apparent between CH 4 production rate and the population of methanogens in wetlands (Fig. 6).This difference may be due to discrepancy in primers used in two studies.In the present study, primers (1106F-1378R) target the methanogenic archaeal 16S rRNA genes rather than the mcrA genes.However, this finding is in agreement with the results of Galand et al. (2003) and Cadillo-Quiroz et al. (2006), who reported that the variation in community and population of methanogens did not change potential CH 4 production.Colwell et al. (2008) also failed to accurately estimate the emission rate using a newly-launched model based on the relationship between in situ CH 4 production rates and the abundance of mcrA genes in marine sediments.Our finding, together with previous results (Galand et al, 2003;Freitag and Prosser, 2009), suggest that the discrepancies in CH 4 production potential among wetlands around China do not completely result from the differences in the abundance of methanogenic archaea.However, it should be mentioned that the high proportion of dead or inactive methanogens in wetland soils may mask the relationship between the population of functional methanogenic archaea and the CH 4 production (Thauer, 1998).A further study is required to evaluate the population of functional methanogenic archaea and their relationship to CH 4 production potential in wetlands.Bergman et al. (1998Bergman et al. ( , 2000) ) found that the absence of labile organic carbon in peatland heavily suppressed CH 4 production at low temperature.In a Michigan ombrotrophic bog, Avery et al. (1999Avery et al. ( , 2003) ) observed that acetate accumulation stimulated CH 4 production, which contributed >80% of total CH 4 production.In the present study, CH 4 production rate was found to be significantly correlated with the presence of DOC (R 2 = 0.31, P = 0.05, n = 13).This indicates that the supply of methanogenic substrates may be responsible for the level of CH 4 production potential in wetlands.We found that both CH 4 production rate and population of methanogenic archaea in top soil layer of the Poyang wetland are significantly lower than in the other Eastern China wetlands (Figs. 2  and 4).Apparently, low SOC in the Poyang wetland not only suppressed the growth of methanogens (Fig. 5), but also reduced the supply of substrates for methanogens (Table 1), resulting in a low CH 4 production rate.The water table in the Poyang wetland fluctuates dramatically with increases or decreases in the Yangtze River water level and is generally near or below the soil surface in winter, resulting in strong decomposition of SOC in the subtropical zone.Altor and Mitsch (2006) verified that CH 4 emissions in the intermittently-flooded wetlands were remarkably lower than those in the permanently-inundated natural and anthropogenic wetlands.They suggested that the higher emissions in continuously anoxic wetlands were due to greater availability of methanogenic substrates and consistently low redox potential.And some opposite special cases were observed in wetlands (Hargreaves et al., 2001;Cheng et al., 2007).We argue that intermittently inundated wetlands such as the Poyang wetland may not be a hot source of atmospheric CH 4 compared to permanently inundated Sanjiang Plain marsh in China (Ding et al., 2004(Ding et al., , 2010)).
In this study, the population of methanogens in the top soil layer of wetlands was measured to be between 1.07-8.29×109 cells g −1 soil (Fig. 4), which is higher than the  ∼1×10 8 cells g −1 soil in an acidic bog and a calcareous fen in the UK (Kim et al., 2008) and the 0.5-0.9×10 7cells g −1 fresh peat in an acidic peat bog in West Siberia, Russia (Kotsyurbenko et al., 2004), but similar to the value (∼1.3×10 9 cells g −1 soil) recorded in Japanese paddy field soil (Watanabe et al., 2007).The highest population of methanogen archaea was measured in the Hongze wetland (Fig. 4), indicating that this wetland is particularly favorable to the growth of methanogens.However, the CH 4 production rate and specific CH 4 production potential of per methanogen cell in the top soil layer of the Hongze wetland was significantly lower than in the Sanjiang Plain C. lasiocarpa marsh (Figs. 2 and 7).This implies that the methanogenic community in wetlands located in different latitudes shows a distinct capacity for CH 4 production.However, a relatively high SO 2− 4 concentration was measured in the Hongze wetland but not in other wetlands (Table 1).Sulfate-reducing bacteria could outcompete methanogens for substrates such as acetate and H 2 (Chin and Conrad, 1995), thereby lowering CH 4 production or possibly masking potential CH 4 production of individual metahnogen cells in the Hongze wetland.This inhibition on CH 4 production also masked the relationship between CH 4 production potential and the water table depth in wetlands across China (Fig. 3).Alternatively, Thauer (1998) andCadillo-Quiroz et al. (2006) observed a higher proportion of dead or inactive methanogens in wetland soils with low levels of functional mcrA mRNA (methyl coenzyme M reductase).Thus, it is possible that a difference exists between the composition and the physiology of dominant methanogens in different wetlands (Galand et al., 2003).Further experiments are necessary to evaluate the methanogen community in Chinese wetlands.
Although there are more methanogenic archaea in the C. angustifolia marsh than in the C. lasiocarpa marsh of the Sanjiang Plain (Fig. 4), a significantly lower amount of DOC in the top soil layer results in a low CH 4 production rate (Fig. 2).Previous studies have shown that Eriophorum plants provide more root exudates for methanogens than Carex and Juncus (Ström et al., 2005;Koelbener et al., 2010).Ding et al. (2005) verified that C. angustifolia may make a larger contribution to CH 4 production than to CH 4 oxidation, while C. lasiocarpa has the opposite effect.This indicates that higher CH 4 production in the C. lasiocarpa marsh is not attributable to organic materials such as root exudates released by living C. lasiocarpa.Ding et al. (2002) suggested that the deep standing water in the C. lasiocarpa marsh inundates more plant litter, which in turn provides more substrates for methanogens.Therefore, we propose that SOC in Eastern China wetlands affects the population of methanogens and the supply of DOC, rather than temperature controlled CH 4 production potentials, while SOC in wetland soils is mainly affected by the position and stability of the water table through varying vegetation cover and SOC decomposition rate (Galand et al., 2003).
On the contrary, DOC concentration in the Ruoergai peatland is much higher than that in the Hongze and Poyang wetlands (Table 1), indicating that low CH 4 production potential in the Ruoergai peatland is not be attributable to the absence of substrates for methanogenesis.Saarnio et al. (1998) and Frenzel and Karofeld (2000) recognized that methanogens are quite sensitive to soil temperature, especially temperatures lower than 20 • C. Chin et al. (1999) observed a significant reduction in CH 4 production and a subsequent increase in the accumulation of acetates as incubation temperature was lowered from 30 • C to 15 • C in an Italian paddy soil.Further, they found that the reduction in CH 4 production potential resulted from changes in dominant species of methanogens.Zhang et al. (2008) identified a dominant uncultured methanogen in the Ruoergai peatland and named it a Zoige cluster I (ZC-I).They found that this type of methanogen is suitable only for living at low temperature.It follows that low CH 4 production potential in the Ruoergai peatland is mainly due to a low cell-specific rate of CH 4 production by distinct methanogens (Fig. 7).Results of the present study indicate that CH 4 production potential in the Ruoergai peatland did not change significantly as the incubation temperature increased from 10 • C to 20 • C (Fig. 2).This is in agreement with the result of Zhang et al. (2008), who found that there is no significant difference in the CH 4 production rate and the increased rate of population of ZC-I between incubation temperatures of 15 • C and 30 • C.These results suggest that the predicted future increase (0.6-4.0 • C) of global temperatures (IPCC, 2007) may not strongly stimulate CH 4 emissions from peatlands in the Qinghai-Tibetan Plateau as in natural wetlands with low elevation (Kettunen et al., 1996) unless the shift in dominant species of methanogens occurred in a peatland.A detailed study is required to evaluate the influence of rising temperatures on CH 4 emission from peatlands in the Qinghai-Tibetan Plateau.

Conclusions
The top soil layer has the maximum population of methanogenic archaea in all wetlands except the Ruoergai peatland and exhibits the highest CH 4 production potential, as measured at the mean in situ summer temperature for all wetlands in China.There is a significant relationship between the population of methanogens and SOC or TN concentrations in wetland soils.This indicates that soil organic carbon and/or nitrogen may control the abundance of methanogens in wetland ecosystems.However, the CH 4 production potential is not significantly related to the methanogen population; it is related to the DOC concentration, and this indicates that DNA-based methanogen abundance might not be an effective indicator for predicting the CH 4 production potential in wetlands.In China, the CH 4 production potential in the top soil layer of wetlands increases with increasing latitude.This suggests that the CH 4 production potential in the wetlands of Eastern China is not affected by temperature and depends on the supply of substrates for methanogens, which may depend on the wetland niche such as the position and stability of the water table.In contrast, low temperatures at high elevations may result in methanogens with low CH 4 production potential to become the dominant species, which in turn results in the suppression of the CH 4 production in the Ruoergai peatland, rather than the deficiency of substrates for methanogensis.

Fig. 3 .
Fig. 3. Relationship between CH 4 production potential of the top soil layer in different wetlands across China and incubation temperature (a), or the water table depth (b).Vertical bars denote standard errors of means (n = 3).SJA1 = 0-20 cm soil in the Sanjiang Plain C. angustifolia marsh, SJL1 = 0-30 cm soil in the Sanjiang Plain C. lasiocarpa marsh, HZ1 = 0-20 cm soil in the Hongze wetland, PY1 = 0-10 cm soil in the Poyang wetland, and REG1 = 0-10 cm soil in the Ruoergai peatland.

Table 1 .
Sampling site characteristics and soil properties.