Linking rhizospheric CH4 oxidation and net CH4 emissions in an arctic wetland based on 13CH4 labeling of mesocosms
Poorly drained arctic ecosystems are potential large emitters of methane (CH4) due to their high soil organic carbon content and low oxygen availability. In wetlands, aerenchymatous plants transport CH4 from the soil to the atmosphere, but concurrently transport O2 to the rhizosphere, which may lead to oxidation of CH4. The importance of the latter process is largely unknown for arctic plant species and ecosystems. Here, we aim to quantify the subsurface oxidation of CH4 in a waterlogged arctic ecosystem dominated by Carex aquatilis ssp. stans and Eriophorum angustifolium, and evaluate the overall effect of these plants on the CH4 budget.
A mesocosms study was established based on the upper 20 cm of an organic soil profile with intact plants retrieved from a peatland in West Greenland (69°N). We measured dissolved concentrations and emissions of 13CO2 and 13CH4 from mesocosms during three weeks after addition of 13C-enriched CH4 below the mesocosm.
Most of the recovered 13C label (>98 %) escaped the ecosystem as CH4, while less than 2 % was oxidized to 13CO2.
It is concluded that aerenchymatous plants control the overall CH4 emissions but, as a transport system for oxygen, are too inefficient to markedly reduce CH4 emissions.
KeywordsCarex Greenhouse gases Methane Oxidation Stable isotopes Tundra
Methane (CH4) is a greenhouse gas produced in arctic wetlands during anaerobic decomposition of organic matter. Production and emission of CH4 from wetlands is controlled by temperature, water level, plant cover, and plant species composition (Elberling et al. 2008; Kutzbach et al. 2004; Olefeldt et al. 2013; Schimel 1995; Ström et al. 2005, 2012). The climate change-related increasing temperatures, changing precipitation patterns, and permafrost thaw in the Arctic (Collins et al. 2013; Hartmann et al. 2013; Schuur et al. 2015) could result in increased emissions of methane to the atmosphere (Johnston et al. 2014; Klapstein et al. 2014; Turetsky et al. 2008).
The potential amount of CH4 that can be emitted to the atmosphere depends on the balance between subsurface CH4 production, and consumption, and methanotrophic oxidation kinetics are typically an order of magnitude faster than methanogenic production kinetics (Segers 1998). Therefore, the overall CH4 transport efficiency out of the system is important for the net emission. Plants with aerenchyma play an important role in regulating wetland CH4 emissions by serving as conduits for CH4 from deeper soil layers, bypassing upper potentially oxidizing soil layers (King et al. 1998). The plant functional contribution to CH4 emissions is species specific and arctic species of Carex have been reported to contribute with up to 99 % of ecosystem CH4 emissions (Schimel 1995). Another influence of aerenchyma-containing plants is the diffusion of oxygen from the atmosphere, through rhizomes and roots, into the rhizosphere, a process known as radial oxygen loss (ROL) (Armstrong and Armstrong 1988) which has been documented under laboratory (Askaer et al. 2010) and field conditions (Elberling et al. 2011). ROL promote the oxidation of CH4 to carbon dioxide (CO2) in an otherwise anoxic zone, and may consequently reduce CH4 emissions (Askaer et al. 2010; Ding et al. 2004; Frenzel and Rudolph 1998). This aspect has been less studied in the Arctic, and to our knowledge no direct studies of rhizospheric methane oxidation in arctic wetlands exist, though rates can be inferred from measurements on saturated areas by Moosavi and Crill (1998) and Preuss et al. (2013).
Studies in temperate ecosystems have shown that the potential for plant-mediated rhizospheric CH4 oxidation is species-specific (Calhoun and King 1997; van der Nat and Middelburg 1998). Species-specific differences in rhizospheric CH4 oxidation may be related to the oxygen transport capacity of the plant species, but also other factors may affect the oxidation; several studies report temporal variations in rhizospheric CH4 oxidation with higher oxidation rates early in the growing season (Ding et al. 2004, 2005; Popp et al. 2000; van der Nat and Middelburg 1998). This correlates well with the theory of largest oxygen output from young, growing root tips (Armstrong and Armstrong 1988; Laanbroek 2010), shown in mesocosm experiments (Askaer et al. 2010). Furthermore, temperature could be an important factor for variations in rhizospheric CH4 oxidation (Saarnio et al. 1997).
Several approaches have been used to study rhizospheric methane oxidation. Indirect approaches include incubations of soil slurries and roots (Gerard and Chanton 1993), and mass balances calculated from differences in production and emission. Direct methods include mass balances of fluxes from well-lit and oxic, and dark and anoxic conditions (Gerard and Chanton 1993; Kankaala and Bergström 2004), and comparisons of in situ fluxes with and without inhibitors (acetylene, methyl fluoride) (Ding et al. 2005; Lombardi et al. 1997; van der Nat and Middelburg 1998). Stable isotope signatures of methane have also been used to quantify methane oxidation (Gerard and Chanton 1993; Popp et al. 1999), as well as addition of isotope tracers (Groot et al. 2003; Ström et al. 2005). The methods used have different strengths and weaknesses. Mass balances from incubations likely overestimate the rhizospheric methane oxidation and are referred to as a potential methane oxidation rate (Gerard and Chanton 1993). On the other hand, the inhibitors methyl fluoride and acetylene partly inhibits methane production and the use of this inhibitor could therefore lead to underestimation of rhizospheric methane oxidation (King 1996; Popp et al. 2000). Hence, there is a need for assessing the impact of rhizospheric CH4 oxidation on arctic ecosystem fluxes of CH4 using precise, non-invasive methods.
The purpose of this study was i) to quantify the effect of plant-mediated oxidation of methane to CO2 on net CH4 emissions in an arctic ecosystem and ii) to estimate the overall effect of presence of aerenchyma-containing plants on methane emissions. We conducted a mesocosm study on intact wetland peat blocks with plants from Disko Island, Greenland, to which 13C-labeled CH4 was added as a tracer. Subsequently, fluxes of 13CH4 and 13CO2 from the whole ecosystem and from individual plants, as well as depth-specific soil water concentrations of dissolved 13CH4 and 13CO2, were measured repeatedly over 18 days. We hypothesized that most of the labeled 13CH4 would be transported through plants, but significant amounts would be oxidized in the rhizosphere and released as 13CO2.
Materials and methods
Mesocosm excavation and incubation
A block of intact, water-saturated peat (25 × 25 × 30 cm3) including vegetation was excavated from a wet fen in Blæsedalen, Disko Island, Greenland (69°18′40.9″N; 53°30′40.9″W), and transferred to the laboratory at University of Copenhagen on September 14 to 16, 2013. The site is characterized by glacial reworked basalt parent material and covered by an approximately 40 cm peat layer. The water table fluctuates between 15 above and 20 cm below the soil surface.
The annual mean soil temperature at 5 cm depth is −0.9 °C (1991–2004) and the warmest monthly mean air temperature is 7.9 ± 1.6 °C during July (D’Imperio et al. 2016), whereas the coldest monthly mean is −14.0 ± 5.0 °C during February–March (Hollesen et al. 2015). Frozen soil conditions prevail from October to late May. The area is characterized by the presence of discontinuous permafrost and at the site the active layer is about 40 cm deep.
The vegetation was dominated by Carex aquatilis ssp. stans, and further included Eriophorum angustifolium, Salix arctophila, and mosses as Tomentypnum nitens and Paludella squarrosa. The peat block was stored in dark in a climate chamber at −6 °C for four months and 22 days mimicking winter conditions. The temperature was then raised to 5 °C combined with 12 h of light a day (251 ± 1 μmol photons m−2 sec−1) for a week. The temperature was subsequently raised to 10 °C for a week, and finally the climate chamber was adjusted to 15 °C and constant light. The peat was pre-incubated at these conditions 48–75 days until start of the experiment. One replicate was incubated first and monitored intensively with measurements three times a day for the first seven days and two times a day for the following twelve days. The two other replicates were measured once a day for twelve days. Thus, data are presented with three replicate measurements on every second to third measurement point. We did not observe any visual differences in plant development by the onset of incubation. The experimental incubation took place in 15 °C and constant light (194 ± 11 μmol photons m−2 sec−1). The mesocosms were watered regularly with deionized water to maintain the water table at 1 cm above soil surface.
Prior to labeling (18 days for the intensively monitored replicate and 45 days for the extensively monitored replicates) the peat block was divided into three minor blocks of 20 × 7.4 cm2 surface area and 20 cm depth, and each block was placed in a two-compartment polycarbonate enclosure. The lower compartment consisted of a 663 mL water-filled reservoir achieved by lowering a firmly fitted box (19.8 × 7.7 × 6.8 cm3 outer dimensions) inside the main enclosure. The reservoir was sealed against the upper compartment holding the mesocosm (inner dimensions of 20 × 7.4 × 30 cm3) by a 3 mm silicone membrane. The outer enclosure wall and interior reservoir wall was pierced by aligned butyl septa for gas injections into the reservoir. The mesocosm enclosure was equipped with Rhizon water samplers (Rhizosphere Research Products, The Netherlands) at depths 2, 4, 6, 8, 13 and 18 cm through a plastic fitting with a butyl septum. The mesocosm sides were covered with metal foil to keep out light.
At the start of the experiment (i.e. time = 0), 20 mL (833.4 μmol) of pure 99 % 13C enriched CH4 (13CH4) was added to each mesocosm by injecting the gas into the lower reservoir. This established a diffusive supply of isotopically enriched CH4 across the silicone membrane to the peat mesocosm. The gas fluxes from the mesocosms were measured with a chamber (see method below) during injection of the enriched CH4 to ensure that there was no leak of 13CH4. The 13CH4 addition to the reservoir below the membrane was made once, as a pulse, reflecting a short period with methane production. Therefore, the experiment represents a transient system wherein diffusion of 13CH4 into the soil decreased over time.
Gas exchange from the whole mesocosm was measured by fitting a transparent polycarbonate chamber (7 × 20 × 25 cm3) gas tight atop the mesocosm enclosure by the aid of a water-filled collar. The chamber was equipped with a small fan in the top, and connected in a closed loop to a Picarro G2201-i Analyzer to determine synchronous emissions of the 12C and 13C isotopologues of CO2 and CH4 (Picarro Inc., Santa Clara, USA). Gas fluxes of individual plants were measured by attaching a transparent, top-sealed pipe (17.7 cm length × 3.5 cm diam.) to a rubber stopper sitting around the base of the plant shoot and connecting the pipe in a loop to the gas analyzer. The transparent chamber and pipe was used for flux measurements to simulate the constant light during arctic summer and to avoid affecting the radial oxygen loss (ROL) and stomata closure. For the intensively monitored replicate, each measurement cycle consisted of a whole mesocosm flux measurement followed by flux measurements from two C. aquatilis ssp. stans shoots and one E. angustifolium shoot (hereafter referred to as Carex and Eriophorum) all of which were fully developed but had not flowered by the start of the incubation. In the period 214 to 323 h after labeling a fourth plant, an emerging shoot of Carex, was measured too. For the two less intensively monitored replicates each measurement cycle consisted of one mesocosm flux measurement and one measurement on a fully developed, but not yet flowering Carex shoot.
Soil water sampling
Immediately after completing each flux measurement cycle, soil water was sampled in the six depths by syringe and needle in a two-step procedure. First c. 0.5 mL was extracted and discarded to clean the Rhizon tubing, and then 1 mL soil water was extracted and transferred to a 12 mL Exetainer vial (Labco Scientific, High Wycombe, UK) containing N2. The overpressure was equalized using a needle and the water samples were frozen down immediately.
Dissolved gas measurements
Prior to analysis for headspace gas concentrations, the soil water samples were weighed, thawed, and adjusted to room temperature (20 °C), and shaken manually to reach steady state. The headspace concentration of 12CO2 and 13CO2 was measured on an Isoprime isotope ratio mass spectrometer coupled to a Multiflow gas inlet system (Isoprime Ltd., Cheadle Hulme, UK). Similarly, the headspace concentration of 12CH4 and 13CH4 was measured by transferring 5 mL of the headspace gas to a multilayer foil gas bag (Jensen Inert, Florida, USA) filled with 35 mL air with ambient (and known) content of 12CH4 and 13CH4, and measuring the concentration of 12CH4 and 13CH4 on the Picarro G2201-i Analyzer.
Redox and pH
The redox potential of the soil was measured before incubation. Using a microelectrode profiling system (Unisense, Aarhus, Denmark), redox measurements were made in 40 profiles of 1000 μm intervals starting from the water surface and down to 40 mm.
Measurements of soil pH were made at the end of the experiment by inserting a MI-410 Micro-Combination pH Probe (Microelectrodes Inc, New Hampshire, USA) connected to a PHM 80 Portable pH Meter (Radiometer, Copenhagen, Denmark) directly into the soil at soil surface and depths 1 and 2 cm.
The outer 3 cm soil, in each side of the mesocosm, was cut into 6 depth sections (0–3, 3–5, 5–7, 7–9, 9–15, and 15–20 cm) and hand-sorted to separate roots. The inner part of the soil was sorted to obtain intact roots from the plants used for flux measurements. The soil from the outer sides was pooled for the specific depths and extracted with water (1:5) and analyzed for ammonium (NH4+), nitrate (NO3−) and total dissolved N on a FIA STAR 5000 flow injection analyzer (FOSS Tecator, Höganäs, Sweden), and dissolved organic C (DOC) on a TOC-5000A total organic analyzer (Shimadzu, Kyoto, Japan). A subset of samples were chloroform fumigated for 24 h, and microbial C and N was calculated as the difference between fumigated and non-fumigated concentrations of DOC and total dissolved N, respectively, using 0.45 and 0.4 respectively as extractability factors for microbial C and microbial N (Jonasson et al. 1996; Michelsen et al. 1999).
Depth-specific soil samples from 0–5, 5–10, 10–15 and 15–20 cm were collected at four places at the site in July 2015 and the roots were sorted. In the beginning of October one Carex and one Eriophorum specimen were collected and separated in root and shoot.
The mesocosm plants used for shoot flux measurements were sorted individually and divided in above- and belowground biomass. The aboveground plant material was sorted in species, while the biomass of live roots + rhizomes was sorted in the six depths.
All plant material was rinsed in demineralized water, oven-dried (60 °C), and weighed. The depth-specific belowground plant biomasses from the outer sides of each replicate mesocosm were pooled together (giving six root + rhizome samples per mesocosm), and the aboveground plant material was pooled in species. Samples were finely ground on a ball mill and weighed (2 mg) into tin combustion cups for total C and 13C/12C isotope ratio analysis by Dumas combustion (1020 °C) on an elemental analyzer (EA Flash 2000, Thermo Scientific, Bremen, Germany) coupled in continuous flow mode to a Thermo Scientific Delta V Advantage Isotope-Ratio Mass spectrometer (Thermo Scientific, Bremen, Germany).
Fluxes of CH4 and CO2 from the mesocosms and individual plants were calculated by fitting a second order polynomial regression to data and using the slope at time zero. Only significant slopes (p < 0.05) were used for flux calculation. Keeling plots, (ie the δ 13C value on the reciprocal concentration) were established in order to obtain the δ13C value of the emitted CO2 and CH4 derived from the Y-axis intercept. Only Keeling correlations with significant slopes (p < 0.05) were used. Moreover, as fluxes were measured under light conditions a number of CO2 fluxes were very low (i.e. balanced photosynthesis and respiration) and fluxes within the range from −0.27 to +0.27 μmol CO2 m−2 sec−1 for ecosystem, and from -0.00055 to +0.00055 μmol CO2 plant sec−1 for plant fluxes were discarded, as such low fluxes resulted in large uncertainties in the Keeling δ13C estimates. The total number of discarded CO2-flux observations was 8.9 %; since 13CO2-enrichments of the headspace during these events were small/negligible (not shown), the 13CO2 flux estimates are considered reliable. Time lags (time since addition of label) among replicates were accounted for by using the average time.
Calculation of dissolved gasses
The concentration of gasses dissolved in the soil water was calculated by the application of the Bunsen absorption coefficient assuming equilibrium between water and headspace gas (0.343 and 0.871 for CH4 and CO2 at 20 °C, respectively). In order to estimate total amounts of dissolved gases in the mesocosm profile linear interpolations between sampling points were applied, and it was assumed that the concentration measured at −2 cm depth was valid for the 1 cm water above the soil surface. The amounts of water in the different layers (+1 to −3 cm, −3 to −5 cm, −5 to -7 cm, −7 to −10.5 cm, −10.5 to −15.5 cm and −15.5 to −20 cm) were calculated based on soil water content, and corrected for offsets between the increments into which the soil was divided for post-experimental processing and the depths applied for water sampling. The difference in time since labeling for the six depths and three replicates at each time point was accounted for by taking the average.
where δ is the δ13C value of the dissolved gas (measured in headspace of water samples) or the flux (estimated from Keeling plots), and RST is the 13C/12C ratio of the international reference material (0.01118, PeeDee Belemnite). For comparison, the 13C-content in dissolved gases (CH4 and CO2) are reported by the same notation (δ13C), and above calculations applied to establish the 13C-isotope budget.
The cumulated emissions of excess 13C as CH4 and CO2 were calculated by linear interpolations between individual time points and adding fluxes for each time step. The excess 13C dissolved in the soil water as CO2 was calculated by adding the amount of excess 13C from the six depths at each time point.
The partitioning of the excess 13C label was applied to four pools: dissolved 13CH4, dissolved 13CO2, emitted 13CH4 and emitted 13CO2. This partitioning was calculated for five specific time points, and the sum of the four pools was considered 100 %. The five time points, ie 88, 164, 260, 347 and 429 h after labeling, were established as the average time of adjacent water sampling and gas flux measurement events at roughly 2-hour intervals. Measurements at the first three time points were carried out in triplicate, whereas for the last two time points, data are from one mesocosm, as described above.
Rhizospheric CH4 oxidation
where CH4 dissolved is the concentration of dissolved CH4 and CH4 cumulated is the cumulated fluxes of CH4, and R is the ratio of the excess 13C recovered as CO2 (dissolved pool + cumulated fluxes) to the total excess 13C pool (dissolved pool + cumulated fluxes).
Data are reported as average ± standard error (SE) or as single observations of the intensively monitored replicate. In the case of measurements on the two Carex plants in the intensively monitored replicate, data are reported as the average of the two without standard error. Temporal trends in gas fluxes and dissolved gasses over the measurement period were tested by fitting a linear regression in Microsoft Excel and calculating the corresponding p-value. P-values < 0.05 were considered significant. Graphs were made in Sigma Plot 13.0.
Soil extractable pools of dissolved organic carbon (DOC) total dissolved nitrogen (TDN), ammonium (NH4+), nitrate (NO3−), and microbial C and N in six depths of the mesocosms. Data are averages ± standard error
DOC (mg C g−1 soil)
TDN (μg N g−1 soil)
NH4+ (μg N g−1 soil)
NO3− (μg N g−1 soil)
Microbial C (mg C g−1 soil)
Microbial N (μg N g−1 soil)
1.8 ± 0.3
446 ± 210
11.04 ± 4.04
6.0 ± 3.0
19.0 ± 1.0
1062 ± 269
1.9 ± 0.1
234 ± 52
11.83 ± 3.63
15.1 ± 11.4
13.6 ± 4.0
1151 ± 494
1.7 ± 0.4
214 ± 14
6.78 ± 0.57
8.1 ± 4.4
10.5 ± 2.7
757 ± 368
1.7 ± 0.2
149 ± 30
9.36 ± 3.66
8.8 ± 7.3
12.3 ± 0.1
369 ± 89
1.7 ± 0.2
97 ± 5
18.71 ± 7.12
16.4 ± 15.3
10.6 ± 0.7
351 ± 82
1.2 ± 0.2
87 ± 6
15.47 ± 5.78
4.7 ± 3.9
6.2 ± 0.6
237 ± 92
Dissolved CH4 and CO2
Concentrations of dissolved CH4 in soil water in the six depths averaged 87.5 ± 10.8 μg L−1 over the season and did not show a significant temporal change (p = 0.203). Concentrations of dissolved CO2 in soil water in the six depths averaged 83.5 ± 4.9 mg L−1 and increased significantly over time (p < 0.01) from c. 70 mg L−1 to 130 mg L−1.
Total CH4 and CO2 fluxes
Ecosystem CH4 emissions averaged 512 ± 10 μmol m−2 h−1 (Fig. S2a) and increased from c. 460 μmol m−2 h−1 to c. 560 μmol m−2 h−1 during the measurement period (p < 0.01), while ecosystem CO2 fluxes averaged 2.74 ± 0.21 mmol m−2 h−1 (Fig. S2b) and did not show a significant temporal change (p = 0.231). The CH4 emissions from individual shoots of Carex and Eriophorum averaged 1.73 ± 0.06 μmol shoot−1 h−1 and 3.21 ± 0.09 μmol shoot−1 h−1 respectively (Fig. S3a). While the emissions from Carex did not change over time, the emissions from Eriophorum increased significantly (p < 0.001) from c. 2 μmol shoot−1 h−1 to c. 3.6 μmol shoot−1 h−1 over the measurement period. CO2 fluxes from shoots of Carex and Eriophorum averaged 6.79 ± 0.79 μmol shoot−1 h−1 and 13.45 ± 0.65 μmol shoot−1 h−1, respectively (Fig. S3b), and decreased significantly over time (p < 0.01) from c. 10 μmol shoot−1 h−1 to c. 2 μmol shoot−1 h−1, and from c. 20 μmol shoot−1 h−1 to c. 10 μmol shoot−1 h−1, respectively. CH4 and CO2 fluxes from the small Carex shoot averaged 0.430 ± 0.118 μmol shoot−1 h−1 and 7.33 ± 2.03 μmol shoot−1 h−1 respectively (Fig. S4).
Carbon isotope composition of dissolved gasses
Carbon isotope composition of emitted gases
The δ13C values of the CH4 fluxes from ecosystems and individual plant shoots increased immediately after labeling and peaked after about 100 h (Fig. S6a). The δ13C values of CH4 fluxes from Eriophorum shoots (n = 1) were numerically higher than the δ13C values of fluxes from Carex shoots and whole ecosystems with a maximum of 33077 ‰ (27.6 atom% 13C), but this could not be tested statistically. The maximum δ13C values of (repeated) CH4 fluxes from Carex shoots and whole ecosystems were 13391 ‰ (13.9 atom% 13C) and 17307 ‰ (17.0 atom% 13C). The δ13C values of the CO2 fluxes from ecosystems and individual Eriophorum shoots increased linearly and highly significantly over time from −20 to 18 ‰ (p < 0.0001) and from −25 to 0 ‰ (p < 0.0001), respectively. The δ13C values of the CO2 fluxes from individual Carex shoots did not change over time (Fig. S6b). The δ13C values of the CH4 and CO2 fluxes from the small Carex shoot averaged 6131 ± 448 ‰ and 21.5 ± 0.7 ‰ respectively (Fig. S7).
Excess 13C of dissolved gasses
Excess 13C of ecosystem gas fluxes
Carbon isotope composition of plant material
The δ13C of root samples from the field averaged −26.4 ± 0.3 ‰ (n = 16) and for aboveground Carex and Eriophorum material the value was −26.1 ± 0.5 ‰ (n = 4). Roots from the mesocosm had a δ13C of −30.0 ± 0.3 ‰ (n = 18) and for the Carex and Eriophorum biomass the δ13C was −30.8 ± 0.5 ‰ (n = 22).
The recovery of the 13C label and the distribution of recovered 13C label among dissolved 13CH4 and 13CO2 and cumulated fluxes of 13CH4 and 13CO2
Time (hours since labeling)
Recovery of label (%)
13C dissolved as 13CH4 (%)
13C dissolved as 13CO2 (%)
13C emitted as 13CH4 (%)
13C emitted as 13CO2 (%)
% CH4 oxidized
When plant-mediated fluxes were up-scaled from shoot to mesocosm level, based on abundance of Carex and Eriophorum, plant transport accounted for 245 % and 26 % of ecosystem 13CH4 and 13CO2 emissions respectively, and 243 % of total ecosystem excess 13C emissions.
The fluxes of CH4 from the mesocosms were 36 times higher compared to in situ fluxes from wetlands at Disko Island (Christiansen et al. 2014). This was expected as the water table was kept constantly above the soil surface as opposed to an in-situ fluctuating water table. The mesocosm fluxes were comparable with the growing season fluxes reported by Tagesson et al. (2012) from NE Greenland, consistent with the fact that the incubation temperature was within the natural range of arctic canopy summer temperatures.
The δ13C values of the emitted CH4 from plants and ecosystem show that the labeled CH4 quickly diffused up into the soil and out of the system. The apparent trend of higher δ13C values of CH4 emitted from Eriophorum shoots than from Carex shoots and the whole ecosystem suggests that the CH4 enters the Eriophorum at a deeper level closer to the silicone membrane. However, as the flux measurements from Eriophorum were not replicated this remains speculative. Yet, it correlates well with results by Shaver and Billings (1977) who observed higher root elongation rates in Eriophorum than in C. aquatilis in deep soil layers (15 – 30 cm) and higher root elongation rates in Eriophorum in deep soil layers as compared to shallow.
By extrapolating plant shoot specific excess 13CH4 emissions to ecosystem level (based on no. of shoots per mesocosm), it appears that the plants contribute 245 % of the ecosystem excess 13CH4 emissions (at 430 h). This emphasizes, as hypothesized, that the plants, in this representative wet ecosystem of West Greenland, are the main pathways of transporting CH4 from the rhizosphere to the atmosphere. It also highlights that up-scaling of plant CH4 emissions based on plant abundance is not straightforward. Indeed, the CH4 fluxes from the small Carex shoot were on average four times lower than those from the fully developed Carex shoots, and the δ13C values of the fluxes were 2–3 times lower for the same time interval. Thus it seems that plant size and well as rooting depth would be important for up-scaling 13CH4 fluxes in this case.
Rhizospheric CH4 oxidation
The increases in δ13C values of dissolved CO2 demonstrate the oxidation of 13CH4 to 13CO2 (Fig. 2b). This is supported by the concentrations of dissolved excess 13CO2 (Fig. 3) and the excess 13CO2 fluxes (Fig. 4). We presume that this oxidation is primarily a result of rhizospheric oxygenation by the plants, as the alternative, anaerobic CH4 oxidation, seems less likely due to the generally low NO3− and sulfate (SO42−) (Table 1; and Nielsen et al. submitted) concentrations in the soil (Beal et al. 2011; Norði and Thamdrup 2014; Segers 1998). Yet, the presence of NO3− in the rhizosphere column (Table 1) indicates a supply of oxygen to the soil. As dissolved 13CO2 increased linearly over the entire measurement period, the supply of electron acceptors for 13CH4 oxidation was not exhausted but rather renewed continuously. This supports the assumption of oxidation by a plant facilitated rhizospheric O2 supply. The importance of pure oxygen diffusion in the water phase cannot be ruled out, but as highest concentrations of dissolved excess 13CO2 were found in 4, 8 and 18 cm, it implies that the oxygen used for the oxidation was not supplied primarily by diffusion from the water surface.
Over the incubation period, between 1.0–1.3 % of the 13CH4 present was oxidized to 13CO2 (as calculated from the five time points; Table 2), which is lower than the 13–38 % observed by Moosavi and Crill (1998) and 27 – 52 % measured by Kankaala and Bergström (2004). It is, however, in the range reported by Popp et al. (1999) who found that rhizospheric CH4 oxidation reduced CH4 emissions with 0 - 35 % in a Carex dominated fen in Alberta, Canada. Also, Ding et al. (2004) found that Carex spp. reduced potential CH4 fluxes from a freshwater marsh in NE China with 3.2 - 38.5 %. Frenzel and Rudolph (1998) found no rhizospheric CH4 oxidation in an Estonian bog ecosystem dominated by Eriophorum ssp., and this explains partly the low CH4 oxidation rates reported here based on Carex and Eriophorum. Also the timing and growth phase of the plants in the current study can affect the actual oxidation rates, as higher oxidation rates early in the growing season, compared to later, have been observed (Ding et al. 2004, 2005; van der Nat and Middelburg 1998). Popp et al. (2000) even found rhizospheric CH4 oxidation to be undetectable in the end of the growing season. Finally, when rhizospheric CH4 oxidation is expressed as a percentage of the potential CH4 flux, then higher CH4 fluxes will make the same rate of oxidation appear less important (Lombardi et al. 1997), and our CH4 fluxes were in the high range. Indeed, our observed absolute rates of CH4 oxidation (5.6 ± 035 μmol m−2 h−1) were 41 % of the late season oxidation rates of bulrush reported by van der Nat and Middelburg (1998), but when calculated in % oxidation they constitute only 20 % of reported fluxes.
At the end of the incubation period 98 % of the recovered label had been emitted as CH4 (Table 2), showing that methane emissions by far exceeds the CH4 oxidation potential in this ecosystem. Thus, it is concluded that although the plant oxygen transport to the rhizosphere very likely occurred, it plays only a minor role in the CH4 budget. This may however change with time of growing season and fluctuating water table, which would affect the roots oxygen demand and hence the oxygen transport to the rhizosphere (Visser et al. 2000).
The peat block used for the mesocosms was collected in the fall and therefore had to be frozen down and gradually thawed and warmed up to mimic a winter and the following spring and summer. Although there may have been effects on plant physiology and hence on rhizospheric methane oxidation, this procedure is considered the most realistic approach given the sampling time.
The flux measurements on both ecosystems and individual plant tillers were carried our using transparent chambers, due to the simulation of arctic summer with continuous light, and the connection between photosynthesis and ROL. Thereby photosynthesis was possible, and plants could have taken up some of the oxidized 13CO2. No 13C enrichment was however detected in the plant material, and the incorporation of 13C label into this pool has likely been negligible. Indeed, plant from the mesocosms showed a more depleted δ13C signal than plant samples from the field, which was probably due to mesocosm plants growing in an urban environment (Lichtfouse et al. 2003). We do however encourage future studies to measure gas fluxes under dark conditions as well, as it would help assess the potential 13CO2 emissions from the system, and the possible 13CO2 transport through plants from root zone to atmosphere.
Part of the 13C label could also have been incorporated into the DOC pool. We did not measure the 13C enrichment of the DOC pool, but consider it of minor importance since it was not possible to detect enrichment in the plant material. Future studies are however encouraged to include this component.
The recovery of 55.8 % of the added label by the end of the experiment indicates that some label could have been entering other pools than the measured dissolved and emitted gasses. It is however assumed, that the majority of the unrecovered labeled CH4 remained dissolved in the lower reservoir. Future studies could benefit from sampling the 13C concentration in the lower reservoir some times during the incubation.
We looked into at which depth the highest CH4 oxidation took place by examining the soil profile data at 260 h after labeling. At this time, the 13CO2/13CH4 ratio was highest in 2 and 4 cm depth, and it was 9–12 times higher than the ratio in 18 cm depth. At the same time, the δ13C values of the CH4 fluxes from Carex shoots, Eriophorum shoots, and the whole ecosystem correspond to the δ13C values of the CH4 dissolved in 15, 17, and 14 cm (calculated by fitting a logarithmic regression to the depth vs. δ13C values) (Fig. S9). Assuming steady state conditions, this indicates that the largest CH4 oxidation takes place in the top soil, while the transport of CH4 from the rhizosphere seems to mainly take place at deeper soil layers. This might explain why oxidation only mediates a minor fraction of the CH4 emissions in this ecosystem, but more research is needed to tell the importance of spatial distribution of plant rhizosphere oxygenation and CH4 transport out of the system.
Future measurements should thus include measurements of depth specific pH, O2 Fe3+, SO42− and bicarbonate in order to account for depth specific oxygen transport, possible anaerobic CH4 oxidation and proportion of oxidized label (as 13CO2) in the bicarbonate pool.
It appears that in this wet tundra ecosystem, under water-saturated conditions, rhizospheric CH4 oxidation exerts only a minor control on ecosystem CH4 emissions. Plant-mediated CH4 transport seems to be the major factor controlling CH4 emissions, possibly because the CH4 emission is faster than the oxidation to CO2, and possibly because the CH4 enters the root system at lower depths than where the major part of the oxidation takes place. However, care should be taken when upscaling these results, obtained under controlled laboratory conditions, to ecosystem level, as changes in growing conditions such as temperature, precipitation and wind likely affect the processes. Furthermore, more research is needed to reveal the importance of plant growth phases on plant O2 transport capacities and CH4 oxidation, and the spatial distribution of CH4 oxidation and transport out of the system.
The Danish National Research Foundation is gratefully acknowledged for the funding of this study through the funding of CENPERM (DNRF100). We thank Frida Lindwall and Nynne Rand Ravn for peat block collection and transport to Copenhagen, Katrine Wulff for help during the incubation experiments, and technicians at CENPERM for help with analyses.
- Askaer L, Elberling B, Glud RN, Kuhl M, Lauritsen FR, Joensen HR (2010) Soil heterogeneity effects on O2 distribution and CH4 emissions from wetlands: In situ and mesocosm studies with planar O2 optodes and membrane inlet mass spectrometry. Soil Biol Biochem 42:2254–2265. doi:10.1016/j.soilbio.2010.08.026 CrossRefGoogle Scholar
- Christiansen J, Romero A, Jørgensen NG, Glaring M, Jørgensen C, Berg L, Elberling B (2014) Methane fluxes and the functional groups of methanotrophs and methanogens in a young Arctic landscape on Disko Island, West Greenland. Biogeochemistry: 1–19. doi: 10.1007/s10533-014-0026-7
- Collins M, Knutti R, Arblaster J, Dufresne J-L, Fichefet T, Friedlingstein P, Gao X, Gutowski WJ, Johns T, Krinner G, Shongwe M, Tebaldi C, Weaver AJ, Wehner M (2013) Long-term climate change: projections, commitments and irreversibility. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: The physical science basis contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USAGoogle Scholar
- D’Imperio L, Nielsen CS, Westergaard-Nielsen A, Michelsen A, Elberling B (2016) Methane oxidation in contrasting soil types: responses to experimental warming with implication for landscape-integrated CH4 budget. Glob Chang Biol. doi:10.1111/gcb.13400
- Elberling B, Nordstrøm C, Grøndahl L, Søgaard H, Friborg T, Christensen TR, Ström L, Marchand F, Nijs I (2008) High-arctic soil CO2 and CH4 production controlled by temperature, water, freezing and snow. In: Meltofte H, Christensen TR, Elberling B, Forchhammer MC, Rasch M (eds) Advances in ecological research, Vol 40: High-Arctic ecosystem dynamics in a changing climate. Elsevier Academic Press Inc, San DiegoGoogle Scholar
- Hartmann DL, Klein Tank AMG, Rusticucci M, Alexander LV, Brönnimann S, Charabi Y, Dentener FJ, Dlugokencky EJ, Easterling DR, Kaplan A, Soden BJ, Thorne PW, Wild M, Zhai PM (2013) Observations: atmosphere and surface. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: The physical science basis contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USAGoogle Scholar
- Johnston CE, Ewing SA, Harden JW, Varner RK, Wickland KP, Koch JC, Fuller CC, Manies K, Jorgenson MT (2014) Effect of permafrost thaw on CO2 and CH4 exchange in a western Alaska peatland chronosequence. Environ Res Lett 9. doi:10.1088/1748-9326/9/8/085004
- King JY, King JY, Reeburgh WS, Reeburgh WS, Regli SK, Regli SK (1998) Methane emission and transport by arctic sedges in Alaska: Results of a vegetation removal experimentGoogle Scholar
- Michelsen A, Graglia E, Schmidt IK, Jonasson S, Sleep D, Quarmby C (1999) Differential responses of grass and a dwarf shrub to long-term changes in soil microbial biomass C, N and P following factorial addition of NPK fertilizer, fungicide and labile carbon to a heath. New Phytol 143:523–538. doi:10.1046/j.1469-8137.1999.00479.x CrossRefGoogle Scholar
- Schuur EAG, McGuire AD, Schädel C, Grosse G, Harden JW, Hayes DJ, Hugelius G, Koven CD, Kuhry P, Lawrence DM, Natali SM, Olefeldt D, Romanovsky VE, Schaefer K, Turetsky MR, Treat CC, Vonk JE (2015) Climate change and the permafrost carbon feedback. Nature 520(7546):171–179CrossRefPubMedGoogle Scholar
- Tagesson T, Molder M, Mastepanov M, Sigsgaard C, Tamstorf MP, Lund M, Falk JM, Lindroth A, Christensen TR, Strom L (2012) Land-atmosphere exchange of methane from soil thawing to soil freezing in a high-Arctic wet tundra ecosystem. Glob Chang Biol 18:1928–1940. doi:10.1111/j.1365-2486.2012.02647.x CrossRefGoogle Scholar