Response of CH 4 emissions to moss removal and N addition in boreal peatland of northeast China

Boreal peatlands are an important natural source of atmospheric methane (CH 4). Recently, boreal peatlands have been experiencing increased nitrogen (N) availability and decreased moss production. However, little is known about the interactive effect of moss and N availability on CH 4 emissions in boreal peatlands. In this study, the effects of moss removal and N addition (6 g N m −2 yr−1) on CH4 emissions were examined during the growing seasons of 2011, 2012 and 2013 in a boreal peatland in the Great Hinggan Mountain of northeast China. Notably, the response of CH 4 emissions to moss removal and N addition varied with experimental duration. Moss removal and N addition did not affect CH4 emissions in 2011 and 2012, but respectively reduced CH4 emissions by 50 % and 66 % in 2013. However, moss removal and N addition did not produce an interactive effect on CH4 emissions. Consequently, moss removal plus N addition had no effect on CH 4 emissions in 2011 and 2012, but decreased CH 4 emissions by 68 % in 2013. These results suggest that the effects of moss removal and N enrichment on CH4 emissions are time-dependent in boreal peatlands, and also imply that increased N availability and decreased moss growth would independently inhibit CH 4 emissions in the boreal peatlands of northeast China.


Introduction
Methane (CH 4 ), as the second most important greenhouse gas after carbon dioxide, contributes 18 % to the overall global radiative force and is predicated to play a key role in determining future climate change (IPCC, 2007).Boreal peatlands are recognized as a primary natural source of at-mospheric CH 4 and contribute 1/10 of total CH 4 emissions to the atmosphere, despite covering a small area of the earth's surface (Wahlen, 1993;Moore et al., 1998;Baird et al., 2009).In boreal peatlands, CH 4 is produced by methanogens in the anaerobic layer and is then consumed by methanotrophs in the aerobic layer (Whalen, 2005).The amount of CH 4 released from peat to the atmosphere depends on the difference of CH 4 -producing and CH 4 -oxidizing processes in peat (Sundh et al., 1994).In boreal peatlands, CH 4 flux dynamics are influenced by soil temperature, water table position, substrate quality, microtopography and vegetation distribution (Bubier et al., 1995;Bellisario et al., 1999).
The moss layer is usually dominant in peatland ecosystems and is probably the only aerobic layer for CH 4 consumption before it enters the atmosphere (Basiliko et al., 2004).Moss provides a good thermal layer for the underlying soils and may play a role in controlling CH 4 oxidation (Basiliko et al., 2004;Turetsky, 2004).About 90 % of the CH 4 produced in peat could be consumed in the moss layer and the soil (Bubier and Moore, 1994;Whalen, 2005).It has been reported that the rate of CH 4 oxidation was > 0.2 µl mol CH 4 g dry weight −1 h −1 by submerged brown moss (Liebner et al., 2011).However, climate change inhibits moss growth and decreases moss production in boreal peatlands (Rustad et al., 2001;Limpens et al., 2011).This could influence the CH 4 emissions from the boreal peatlands, given the important role of moss in CH 4 oxidation.Human activities have already increased nitrogen (N) input to boreal ecosystems (Vitousek et al., 1997;Kaiser, 2001) and climate warming would further stimulate the soil N mineralization rate and increase N availability in soils (Rustad et al., 2001).Previous studies regarding the effects of increased

H. N. Meng et al.: CH 4 emissions in boreal peatland
N availability on CH 4 emissions have yielded inconsistent results; some studies showed that increased N input increased CH 4 emissions (Saarnio et al., 2000;Granberg et al., 2001), and other studies found that N enrichment either decreased CH 4 emissions (Granberg et al., 2001) or had no effect on CH 4 production and oxidation (Saarnio and Silvola, 1999).As with climate warming, the effects of agricultural activities will expand with the increase of latitude, and the effect of high ammonium loading to boreal peatland will increase gradually, which we simulated in our study.To accurately develop the CH 4 budget in boreal peatlands, further studies are needed to examine the effect of N enrichment on CH 4 emissions.
Although previous studies have independently examined the effects of moss and N availability on CH 4 emissions (Ferenci et al., 1975;Conrad, 1999;Riutta et al., 2007;Larmola et al., 2010), there is little information about the interactive effect of moss and N addition on CH 4 emissions in boreal peatlands.Given the wide co-occurrence of declined moss growth and increased N availability in boreal peatlands, determining the effects of moss and N availability on CH 4 emissions would help to better understand CH 4 dynamics, especially in light of future climate change.In this study, a field experiment was established in a boreal peatland in the Great Hinggan Mountain in northeast China, and a 3year (2011 to 2013) continuous observation was conducted to assess the effects of moss removal and N addition on CH 4 emissions during the growing season of a boreal peatland, to simulate the effect of environmental changes of moss degradation and N deposition on CH 4 emissions in the context of global change.

Study site
The research was conducted in a boreal peatland ecosystem located in the north of the Great Hinggan Mountain (52 • 56 N, 122 • 52 E; 457 m a.s.l.) in northeast China.The study site is located in the continuous permafrost zone, and belongs to the cool continental climate (Miao et al., 2012).The mean annual precipitation (1991-2010) is ∼450 mm with 45 % falling from July to August, and the mean annual air temperature is ∼ −3.9 • C with monthly mean ranging from −31.9 • C in January to 19.8 • C in July.The soil of the study site is a typical peat soil and the depth of the peat layer ranges from 40 to 100 cm, with a mean soil bulk density of 0.16 g cm −3 , pH of 5.0, soil organic carbon of 371.68 g kg −1 , and total N content of 17.2 g kg −1 at 0-20 cm depth (Sun, 2012).The amount of N deposition in the study area is approximately 0.45 g N m −2 yr −1 (Zhan et al., 2014).The dominant plant species are Betula fruticosa, Ledum palustre, Chamaedaphne calyculata, Vaccinium uliginosum, Rhododendron parvifolium, Eriophorum vaginatum, Sphag-num moss and Aulacomnium androgynum.Hummocks are covered by continuous moss with some shrubs and occupy ∼50 % of the ground surface.The height of shrubs, sedges and moss are 45-50, 30-33 and 10-12 cm, respectively.The coverage of moss is nearly 90 %, and moss biomass ranges from 190 to 400 g m −2 .

Experiment design
A complete randomized block design with control (CK), moss removal (MR), N addition at 6 g N m −2 yr −1 (N) and moss removal plus N addition (MR × N) treatments was used.Each treatment was replicated three times, resulting in 12 50 cm × 50 cm plots.Plots were separated from adjacent plots by ∼1 m buffer zones, to avoid horizontal movement and lateral loss of the added N. In 2011, plots were placed on flat hummocks with a Sphagnum moss-dominated community.Moss was removed by cutting the green part of the moss layer (∼10 cm) in May from 2011 to 2013.The N was added as urea and applied twice a year (mid-May and mid-July).The urea was dissolved in 1 L purified water and sprayed.The control treatments were sprayed with 1 L purified water without N fertilizer.

CH 4 flux measurement
CH 4 emissions were measured by static chamber and gas chromatography at 7-day intervals between 09:00 a.m. and 11:00 a.m. during the growing periods of 2011 to 2013.The removable open-bottom chambers (stainless steel, two small fans fixed symmetrically inside, 50 cm × 50 cm × 50 cm) were put on the base flumes (stainless steel, 50 cm × 50 cm × 30 cm) during sample collecting, and immediately removed after collection.The grooves (2 cm wide) of the base collar were filled with water to ensure gas tightness.Gas samples were taken at 0, 10, 20 and 30 min from the chamber headspace following closure by 60 mL plastic syringes attached to three-way stopcocks.Immediately, the samples were stored in 100ml vacuum Tedlar ® air sample bags, and analyzed within a week in the laboratory by modified gas chromatography (Agilent HP-7820A, USA), which was modified by adding an independent sample injector by the Institute of Atmospheric Physics, Chinese Academy of Sciences, and equipped with a flame ionization detector.Details and configurations of the measuring system for analyzing concentrations of CH 4 and the associated method for calculating the flux have been described by Wang and Wang (2003) and Song et al. (2009).Where the linear regression with coefficients of determination (R 2 ) were < 0.8, the samples were rejected for CH 4 .

Precipitation, soil moisture and soil temperature
Precipitation was measured by a rain gauge located near the experimental area.Soil moisture at 5 and 10 cm depth were recorded using a portable time-domain reflectometry May to 22 September in 2013.Soil temperatures at 5 cm below the peat surface were collected in the center of each plot using the portable digital thermometer (JM 624, Jinming Instrument CO., Ltd, Tianjin, China).

Statistical analysis
The seasonal mean values were calculated by averaging the monthly mean values from May to October, and then it was multiplied by the number of experimental days and the CH 4 -C transformed was the seasonal carbon (C) budget.A p value of < 0.05 was considered significant unless otherwise stated.Dependent variables were tested for normality by the Kolmogorov-Smirnov test, and were log-transformed when data were not following the normal distribution.One-way ANOVA was used to examine the differences in seasonal C budget among treatments, followed by Tukey's or Tamhane's multiple comparison test.Repeated measures of ANOVAs were used to examine the effects of sampling dates, moss removal and N addition on CH 4 flux and soil moisture.In each year, two-way ANOVAs were used to assess the effects of moss removal, N addition and their interactions on CH 4 budgets.Linear regression analysis was conducted to examine the relationship between CH 4 flux and soil moisture or soil temperatures.All the statistical analyses were tested using SPSS package 16.0 (SPSS Inc., Chicago, IL, USA), and figures were conducted by Origin 8.0 (Origin Lab Corporation, USA) and SigmaPlot 12.0 (Systat Software, Inc. USA) for Windows.

Precipitation, soil moisture, soil temperature and biomass
Precipitation showed great annual variations during the sampling periods.The precipitation during the growing seasons of 2011 (496.2 mm) and 2012 (347.1 mm) was 28.8 % higher and 9.9 % lower than the 20-year (1991-2010) mean annual value (385.4 mm), respectively, whereas total precipitation during the growing season in 2013 (621.4 mm) was much higher than the 20-year mean annual value.Annual fluctuations in precipitation resulted in the highest soil moisture in 2013 (p < 0.001).Moss removal significantly increased soil moisture at 5 cm (p < 0.001) in all years, whereas N addition significantly decreased soil moisture (p < 0.001) in 2012 and 2013.Both moss removal and N addition significantly increased soil moisture (p < 0.01) in   1, Fig. 1a, d, g).The aboveground biomass of the re-growth of moss was 1.48 g m −2 yr −1 without N addition, and 0.69 g m −2 yr −1 with N addition.The total biomass of sedges decreased 11.8 and 16.3 % by moss removal and N addition, respectively.The total biomass of shrubs increased 18.6 % by moss removal, and 94.4 % by N addition.Moss removal decreased the total biomass of moss by 33.5 %, and N addition decreased 13.9 % (Table 3).However, moss removal and N addition did not produce significant effects on total biomass of sedges, shrubs and moss, (p > 0.05).

Effects of moss removal and N addition on CH 4 flux
Moss removal significantly reduced CH 4 emissions by 50.4 % in 2013 (p < 0.05; Table 1, Fig. 2c), but had no significant effects in 2011 and 2012 (p > 0.05; Table 1, Fig. 2a, b).The N addition showed a significant negative effect on CH 4 emissions in 2013 (65.8 %, p < 0.05), but did not significantly affect it in 2011 and 2012 (p > 0.05).However, moss removal and N addition did not produce an interactive effect on CH 4 emissions during the whole sampling period (Table 2).Moss removal and N addition decreased CH 4 emissions by 68.5 % in 2013, but had no effect in 2011 and 2012.
The N addition interacted with the sampling dates to significantly affect CH 4 flux in 2013 (p< 0.05).Similarly, N addition significantly interacted with moss removal or sampling dates to affect CH 4 flux (p < 0.05).However, moss removal and sampling dates did not produce an interaction on CH 4 flux (p > 0.05, Table 1).

Discussion
Moss removal and N addition were found to have no effects on CH 4 emissions in 2011 and 2012, but to significantly decrease CH 4 emissions in 2013.These results imply that the effects of moss removal and N addition on CH 4 emissions vary with experiment duration and suggest that long-term studies are needed to accurately develop the CH 4 budget in boreal peatlands.
In this study, moss removal had no effect on CH 4 emissions in 2011 and 2012, and produced a negative effect in 2013.Methanogens produced CH 4 in strictly anaerobic conditions and were limited by substrate availability (Yavitt et al., 2012).Moss removal decreased belowground biomass (Table 3) that in turn may reduce C substrates for methanogens.Moss would absorb rainfall, saturating the moss layer and forming anaerobic conditions.Therefore, moss removal would decrease substrate for methanogens in soil and moss mats (Riutta et al., 2007) and decrease CH 4 production.Meanwhile, moss removal would decrease soil moisture through evaporation and form aerobic conditions (Amaral and Knowels, 1995).Larmola et al. (2010) observed that methanotrophy was less frequent on high hummocks, and the oxidation rates were not detectable.It could have already been oxidized belowground near the water table.Therefore, moss removal may lead to more aerobic conditions for methanotrophy and stimulate CH 4 oxidation in soil.Therefore, moss removal would decrease CH 4 emissions in boreal peatlands.The results from this study showed that the effect of moss removal on CH 4 emissions were time-lagged in boreal peatlands.
Similar to moss removal, N addition had no effect on CH 4 emissions in 2011 and 2012, but inhibited CH 4 emissions in 2013 in the boreal peatland in this study.Previous studies also found that N addition effects on CH 4 emissions were inconsistent.Saarnio et al. (2000) and Granberg et al. (2001) reported that N input increased CH 4 emissions in the peatland.In contrast, Granberg et al. (2001) showed that N enrichment decreased CH 4 emissions in a fen.In addition, Saarnio and Silvola (1999) found no response of CH 4 production and oxidation to N addition.The major environmental factors that affect CH 4 emissions in peatland include temperature, water table and substrate properties, for instance, mineral nitrogen content (Melling et al., 2006).Climate warming is predicted to increase microbial activity, and  further urge urea to be mineralized to ammonium in soils.Meanwhile, the change of water environment in soil may also affect the dissolving process of urea and microbial activity.These would lead to high ammonium loading in boreal peatlands in northeast China.Moss and vascular plants inter-cepted the added N in the initial 2 years, which made N unavailable to soil microbes (Nordin et al., 1998;Saarnio and Silvola, 1999;Bobbink et al., 2010).Growth reduction would occur as moss saturated with N (Baxter et al., 1992;Gunnarsson and Rydin, 2000;Limpens et al., 2011) and the loss of biomass production would be further stimulated (Lamers et al., 2000;Limpens et al., 2011).However, N deposition would stimulate the growth of vascular plants (Tomassen et al. 2004).Therefore, substrate for microbes may be balanceable in the moss-dominated ecosystem.Hence, N addition did not affect CH 4 emissions in 2011 and 2012 in the boreal peatland.The subsequent negative effect of N addition on CH 4 emissions in 2013 was explained by the following mechanisms.Firstly, N addition inhibited methanogenesis, due to competition for hydrogen with some microbes and toxicity of denitrification products to the methanogens, including nitrite, NO and/or N 2 O (Conrad, 1999).Secondly, N addition promoted CH 4 oxidation by methanotrophs (Bodelier et al., 2000;Bodelier and Laanbroek, 2004).Previous studies found that ammonium enhanced the development and activity of microorganisms, especially methanotrophs (Bodelier and Laanbroek, 2004) and hence stimulated CH 4 oxidation in the rhizosphere of the plant (Bodelier et al., 2000).
Notably, moss removal and N additions did not produce an interactive effect on CH 4 emissions in the boreal peatland of this study.CH 4 emissions depended on the balance among methanogenesis, CH 4 oxidation and CH 4 transport.In peatlands, methanogenesis and CH 4 oxidation were controlled mainly by soil temperature, soil moisture (or water table below the surface) and substrates (Yavitt et al., 2012).In this study, the mechanisms that controlled the combined effects of moss removal and increased N input on CH 4 emissions in a boreal peatland ecosystem have not been fully elucidated.Nevertheless, the results suggest that moss removal and N addition independently affect CH 4 emissions in boreal peatlands.
The mean CH 4 budget in the control plots ranged from 0.39 g C m −2 in 2011 to 4.49 g C m −2 in 2013 (Fig. 3), and varied substantially with annual precipitation over the study period in boreal peatland.These results imply that altered precipitation regimes and increased extreme weather would exert profound influences on CH 4 emissions in boreal peatland in the context of global climate change.

Conclusions
In this study, we simultaneously assessed the impact of moss removal and N enrichment on CH 4 emissions during the growing seasons of 2011, 2012 and 2013 in a boreal peatland of northeast China.Neither moss removal nor N addition affected CH 4 emissions in 2011 and 2012, but suppressed them in 2013.Moreover, moss removal and N addition did not produce an interactive effect on CH 4 emissions.These results suggest that the effects of moss removal and N addition on CH 4 emissions are time-dependent, and long-term studies are needed to accurately develop knowledge of CH 4 emissions in boreal peatlands in the context of global climate change.Meanwhile, these results also imply that moss removal and N addition independently suppressed CH 4 emissions in boreal peatlands in northeast China.
The Supplement related to this article is available online at doi:10.5194/bg-11-4809-2014-supplement.

Figure 1 .
Figure 1.Temporal dynamics of soil temperature at 5 cm depth (a, d, g), soil moisture at 5 cm depth (b, e, h) and soil moisture at 10 cm depth (c, f, i) during the growing seasons in 2011, 2012 and 2013.Data are daily averages for each treatment (± SE, n = 3).CK, control; MR, moss removal; N, N addition; MR × N, moss removal plus N addition

Figure 2 .
Figure 2. Temporal dynamics of CH 4 flux during the growing seasons in 2011, 2012 and 2013.Data are daily means (±SE, n = 3).Insets represent the seasonal means.CK, control; MR, moss removal; N, N addition; MR × N, moss removal plus N addition.Different lowercase letters indicate significant differences among treatments (p < 0.05).

Figure 4 .
Figure 4. Temporal dependence of CH 4 flux on soil temperature (a) and soil moisture (b, c) across the three growing seasons.

Figure 5 .
Figure 5.The relationships between mean CH 4 flux and soil temperature at 5 cm depth (a, d, g), soil moisture at 5 cm depth (b, e, h) and soil moisture at 10 cm depth (c, f, i), respectively.

Table 2 .
Results (F Values) of two-way ANOVAs on the effects of N addition (N), moss removal (MR) and their interactions on CH 4 flux (g CH 4 m −2 h −1 ).