VZJ sign up for etocs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 8 October 2007
Published in Vadose Zone J 6:705-712 (2007)
DOI: 10.2136/vzj2006.0185
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
Related Collections
Right arrow Isotopes
Right arrow Field-Scale Studies
Right arrow Microbial Processes

ORIGINAL RESEARCH

Quantification of Microbial Methane Oxidation in an Alpine Peat Bog

Karina Urmann*, Graciela Gonzalez-Gil, Martin H. Schroth and Josef Zeyer

Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, Universitätsstrasse 16, 8092 Zurich, Switzerland
* Corresponding author (karina.urmann{at}env.ethz.ch).

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


Received 21 December 2006.



   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
Methane (CH4) is an important greenhouse gas that is produced in different subsurface environments. Its main biological sink, microbial CH4 oxidation, can be quantified in situ in the vadose zone using gas push-pull tests (GPPTs). This field method is based on the comparison of breakthrough curves of the reactant CH4 and a nonreactive tracer. Under diffusion-dominated transport conditions, previously employed noble gases are unsuitable as tracers to calculate rate constants for CH4 oxidation due to differing diffusion coefficients. Here, we show that by performing two consecutive GPPTs and coinjecting acetylene (C2H2) as an inhibitor of CH4 oxidation in the second test, the reactant CH4 can be used as a substitute tracer. Applying this procedure, apparent first-order rate constants for CH4 oxidation ranging from 0.38 to 0.82 h–1 were obtained in the vadose zone of three hummocks in an alpine peat bog near Lucerne, Switzerland. Corresponding estimates of in situ rates ranged from 4 to 299 ng CH4 g dry weight–1 h–1. In all but one GPPT, strong stable carbon isotope fractionation due to diffusion masked isotope fractionation due to microbial oxidation. Therefore, stable carbon isotope fractionation is suitable only to a limited extent as an indicator of microbial CH4 oxidation during a GPPT with diffusion-dominated gas transport. In contrast, the presented procedure for the quantification of microbial CH4 oxidation using GPPTs can be applied without restrictions even in systems with high porosity. Furthermore, the presented method may be useful for quantifying other processes for which suitable inhibitors but no suitable tracers are available.

Abbreviations: GFC, gas flow controller • GPPT, gas push-pull test • TDR, time-domain reflectometry


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
Methane (CH4) is the second most important greenhouse gas after carbon dioxide (Ehhalt et al., 2001). To understand the factors that influence CH4 emissions from different subsurface environments such as wetlands or peat bogs (Segers, 1998; Whalen and Reeburgh, 2000), it is essential to quantify microbial CH4 oxidation, a process that was estimated to reduce overall CH4 emissions by approximately 58% (Reeburgh et al., 1993). In addition to controlled laboratory experiments, it is important to quantify microbial processes in situ at the field scale to obtain rates or rate constants that may be more representative for the studied environment (Madsen, 1998; Reeburgh, 1996).

A new method, the gas push-pull test (GPPT), was developed and applied to quantify methanotrophic activity in situ in the vadose zone above a petroleum-contaminated aquifer (Urmann et al., 2005). The GPPT consists of the injection of a gas mixture of the reactants CH4 and O2 and nonreactive tracers, such as neon (Ne) and argon (Ar), into the vadose zone at a specific location. While the injected gas mixture is transported away from the injection point, the reactants are consumed by indigenous microorganisms. Subsequently, the gas mixture is extracted together with soil air from the same location. First-order rate constants for CH4 oxidation can be calculated from breakthrough curves of extracted CH4 and a tracer using simplified methods that do not require transport modeling or knowledge of physical parameters of the investigated environment (Haggerty et al., 1998; Schroth and Istok, 2006). An important prerequisite for applying these methods is that CH4 and tracer exhibit similar transport behavior. During a GPPT, transport occurs largely via advection, induced by pumping, and gas-phase diffusion. The relative importance of both processes depends on the applied test conditions and the properties of the investigated subsurface environment.

In a first application, GPPTs were performed above a contaminated aquifer within a 70-cm-diameter concrete well casing (Urmann et al., 2005). Under these apparently restricted transport conditions, Ne was a suitable nonreactive tracer for CH4. However, in an open system with diffusion-dominated transport, noble gases or other tracer gases cannot be used directly as tracers for CH4 due to different transport behavior (Gonzalez-Gil et al., 2007) resulting from differing diffusion coefficients (Fuller et al., 1966). When simplified methods cannot be applied because of the lack of a suitable tracer, GPPTs can be evaluated by inverse modeling, that is, by inferring physical parameters from tracer breakthrough curves (Haggerty et al., 1998). However, this will be challenging in heterogeneous environments as the required physical parameters are spatially variable (Schroth et al., 2001). Determination of relevant parameters in a highly heterogeneous system, if at all possible, would likely require destructive sampling, which is counter to the idea of a nondestructive field method.

Inhibitors for CH4 oxidation can be used to directly compare the transport of CH4 and a tracer during a GPPT without the influence of microbial conversion and to verify that CH4 oxidation is microbially mediated. This is particularly important when applying the GPPT in a subsurface environment that has previously not been studied with this method. Acetylene (C2H2), a frequently applied gaseous inhibitor for methantrophic activity (Frenzel and Karofeld, 2000; McDonald et al., 1996) has been shown to effectively inhibit CH4 oxidation during GPPTs (Urmann et al., 2005).

A different approach to verify microbial CH4 oxidation is the use of stable carbon isotope analysis (Conrad et al., 1999; Liptay et al., 1998), which is based on the faster oxidation of 12CH4 over 13CH4 by methanotrophic bacteria causing isotope fractionation (Whiticar, 1999). However, the use of this approach is complicated by fractionation due to gas-phase diffusion of CH4 with a fractionation factor of similar magnitude as CH4 oxidation (De Visscher et al., 2004). During previous GPPTs in which the importance of diffusion was reduced because of restricted transport conditions, stable carbon isotope fractionation served as an additional indicator for microbial CH4 oxidation (Urmann et al., 2005). However, the usefulness of this approach during GPPTs with diffusion-dominated transport conditions remains to be investigated in the field.

The vadose zone of a peat bog represents an extreme case for an ecosystem with nearly unrestricted gas transport as peat bogs have a high porosity of up to 0.98 (Kellner and Lundin, 2001). Furthermore, the vadose zone is generally limited in vertical extend due to a high water table, so that losses of injected gases to the atmosphere may occur. Peat bogs are an important source of CH4 emissions (Bartlett and Harriss, 1993). In a number of studies CH4 oxidation has been quantified in peat material in laboratory incubations (e.g., Sundh et al., 1994; Whalen and Reeburgh, 2000). Furthermore, CH4 oxidation has been quantified in peat cores under close to in situ conditions (Pearce and Clymo, 2001), and rates have been inferred from CH4 profiles (Fechner and Hemond, 1992). However, data on in situ CH4 oxidation in peat bogs remain scarce, and to our knowledge, no attempt has been made to use tracer tests such as GPPTs to determine in situ rate constants in this type of ecosystem.

Here, we present a method to quantify CH4 oxidation in a highly porous system, such as a peat bog, by using the reactant CH4 as a substitute tracer during GPPTs. We achieved this by performing two consecutive GPPTs at the same location with the coinjection of C2H2 in the second test. Specifically, the aims of this study were (i) to show the feasibility to quantify CH4 oxidation using CH4 as a substitute tracer in the vadose zone of a peat bog as an example application and (ii) to assess if stable isotope fractionation can be used as an indicator of microbial CH4 oxidation during a GPPT under diffusion-dominated transport conditions.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion and Conclusions
 REFERENCES
 
Field Site Description
Gas push-pull tests were performed at a drained but partially regenerated raised peat bog located at 960 m above sea level in Eigenthal, above the city of Lucerne, Switzerland (47°01' N, 47°00' S, 8°14' E, 8°12' W). Tests were performed in three different hummocks during 2005 (H1, H2, GPPT 1 and 2) and 2004 (H3, GPPT 3). The top 20 to 30 cm of hummocks consisted of living and sparsely degraded Sphagnum mosses with increasing humification below 30 cm depth.

To determine CH4 gas concentration profiles before GPPTs, samples were collected through permanently installed stainless steel needles (1 mm i.d.) with three-way valves at 10-cm-depth intervals. Samples were taken slowly with a syringe that was flushed twice with 1-mL sample gas before injection of 4-mL sample into 8 or 13 mL N2–flushed vials. Dilution was determined by measuring pressure in the vials in the laboratory before and after sampling using a manometer (Keller AG, Winterthur, Switzerland) with a needle attached to it to account for pressure differences between the site and the laboratory.

Water table levels in the hummocks were similar in H1 and H2 but lower in H3 (Fig. 1 ). Gas sampling was not possible all the way down to the water table, likely because gas permeability in layers with a higher degree of humification and water saturation was too low. Only in H1 could a sample be taken very close to the water table where a high CH4 concentration was observed (Fig. 1). This was likely due to a less humified and therefore less saturated layer. The peat above this layer, however, was less permeable, and gas sampling between 45 and 55 cm depth was difficult. In contrast to H1, CH4 concentrations at the lowest measured depth at H2 were only slightly elevated and around ambient at H3 (Fig. 1). Methane concentrations in the water were 0.47 µmol mL–1 at 82.5 cm depth at H1 and 0.2 µmol mL–1 at 50 cm at H2, in a saturated layer above the water table. Volumetric water contents, estimated from time-domain reflectometry (TDR) measurements in the vicinity, were 0.48 and 0.75 cm3 cm–3 at 40 and 50 cm depth (H1), 0.65 cm3 cm–3 at 40 cm depth (H2) and 0.57 cm3 cm–3 at 50 cm depth (H3). Temperatures between 20 and 60 cm depth averaged 12°C at H1 and 9°C at H2. Water-phase CH4 concentrations and temperature were not determined in H3 because installations were not yet completed.


Figure 1
View larger version (17K):
[in this window]
[in a new window]

  		FIG. 1. Gas-phase CH4 concentration profiles at three hummocks (H1, H2, and H3) before gas push-pull tests. Note the logarithmic scale of the x axis. The depth of the water table measured from the peat surface in each hummock is indicated by a dashed line.

 
Gas Push-Pull Tests
At each of the three hummocks, an "active" GPPT (designated "a") to determine CH4 oxidation activity was followed by an "inactive" GPPT (designated "i") as a reference (Table 1). All injection gas mixtures contained the substrates of CH4 oxidation, CH4 and O2, and the nonreactive tracer gases He and Ne. The injection gas mixtures of the inactive tests additionally contained acetylene (C2H2) as an inhibitor of CH4 oxidation. Methane concentrations were adapted according to observed background CH4 concentrations and expected activity. Gas push-pull tests were performed as close to the water table as injection and extraction of gas was possible (Table 1) and performed as described previously (Urmann et al., 2005) with slight modifications. Briefly, background soil air was sampled through the respective needle at the injection–extraction depth by pumping 1 L of gas. Subsequently, 13 to 16 L of injection gas mixture (Table 1) was injected with an average flow rate of 0.20 L min–1. After a transition phase of 0.7 to 2 min, flow was reversed and 20 to 29 L was extracted with an average flow rate of 0.18 L min–1 from the same location. For GPPT 1, extraction of gas was not possible at the injection depth, likely because of high water saturation, and the extraction point was moved up by 5 cm. After each active GPPT, the procedure was repeated using an injection gas mixture of the same composition but also containing C2H2 (GPPT "i") (Table 1). Inactive tests were started 9 and 15 min after active tests for GPPT 1 and 3, respectively. After GPPT 2a, extraction was continued for 15 min at a pump rate of 0.6 L min–1, and a sample was taken before starting the injection phase of GPPT 2i to evaluate the stability of background CH4 concentrations. Similarly, after the extraction phase of GPPT 2i, extraction was continued for 30 min at a pump rate of 0.94 L min–1 before a final sample was taken.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Operational parameters for three pairs of gas push-pull tests (GPPTs) at three hummocks (H1, H2, and H3).

 
For injection and extraction, a gas flow controller (GFC) was used (Urmann et al., 2005). The core equipment of the GFC was a diaphragm pump and a mass flow meter. Injection samples were taken in the same way as extraction samples in clamped, pre-evacuated (–6 x 104 Pa) serum bottles closed with butyl rubber stoppers.

Analytical Methods
Samples were analyzed for CH4 and C2H2 by gas chromatography–flame ionization detector and stable carbon isotope ratios in CH4 by gas chromatography–isotope ratio mass spectrometry as described previously (Urmann et al., 2005). The isotope ratios were determined by reference to the standard Peedee belemnite (PDB) (Hoefs, 1997) in the {delta} notation (Mariotti et al., 1981) in per mil:

Formula 1[1]
In GPPT 3, He, Ne, and O2 were analyzed using the method described by Gonzalez-Gil et al. (2007). In GPPT 1 and 2, the latter gases were measured using a Trace GC Ultra gas chromatograph (Thermo Electron, Rodano, Italy) with a thermal conductivity detector and a capillary Molsieve 5A column (Varian, Palo Alto, CA; 50 m x 0.53 mm i.d., 50 µm) at 30°C with a micropacked ShinCarbon ST precolumn (Restek, Bellefonte, PA) at 50°C using H2 as carrier gas. To prevent adsorption of CO2 and H2O to the molecular sieve column, the precolumn was back flushed before the elution of CO2 and H2O. During back flush, carrier-gas flow on the main column was maintained.

Data Analysis
To obtain breakthrough curves of the different gases, relative concentrations (C*) were calculated by dividing concentrations in extraction samples by the concentration in the respective injection gas mixture (Table 1) and plotted versus time since end of injection. A simplified method was used to evaluate GPPTs (Schroth and Istok, 2006). With this method, one can account for reaction during injection even when only a segment of a GPPT is evaluated. It is based on the assumption that no mixing occurs between "parcels" of gas injected at different times of the injection phase. To apply this method, a residence time tR was calculated for each parcel j, which is the time from its injection until its extraction:

Formula 2[(2)]
where t* is time since end of injection, Qext is the extraction pump rate, text is time since extraction began, MCH4i is the total mass of CH4 in the inactive test, CCH4i is the CH4 concentration from the inactive test at time text, and Tinj is the injection time. Subsequently, the natural logarithm of the ratio of relative CH4 concentration C* in the active test and relative CH4 concentration in the inactive test was plotted versus residence time tR (Eq. [3]). The substitute tracer CH4, i.e., CH4 from the inactive test, thereby accounts for dilution of the injected gas with soil air:

Formula 3[(3)]
Apparent first-order rate constants k were calculated by linear regression from the segment of the data that showed a ln-linear relationship according to Eq. [3] with c as an arbitrary constant.

In situ CH4 oxidation rates per g dry peat (in ng CH4 g dry weight–1 [gdw–1] h–1) before GPPTs were estimated by multiplying apparent first-order rate constants k with in situ CH4 gas-phase concentrations measured before GPPTs using the estimated volumetric water content and assuming a porosity of 0.96 and a peat bulk density of 0.05 g cm–3 as measured by Kellner and Lundin (2001) at 30 to 40 cm depth in hummocks of a Sphagnum peat bog. To estimate the potential error for calculated rates, pairs of peat porosity and bulk density in the lower and upper range of values reported for peat (0.85 and 0.1 g cm–3 and 0.98 and 0.03 g cm–3) (Kellner and Lundin, 2001; Schachtschabel et al., 1998) were used in additional calculations.

Gas Push-Pull Test Simulations
We used the Subsurface Transport Over Multiple Phases (STOMP) simulator (White and Oostrom, 2000) to numerically simulate a GPPT without CH4 conversion. A two-dimensional, cylindrical domain was used (for a more detailed description, see Gonzalez-Gil et al. [2007], supporting information) consisting of 875 nodes, all spaced 2 cm apart, representing a 1-m-diameter hummock with a water table at 70 cm depth. Boundary and initial conditions were defined similar to Gonzalez-Gil et al. (2007), with the exception that CH4 and O2 concentrations at top and side boundaries were set constant to atmospheric concentrations, 0.0018 mL L–1 CH4 and 210 mL L–1 O2, and that the injection–extraction point was only one node high. Water boundary conditions were constant water pressure at the bottom and no water flow at the top and both sides; initial conditions were a static water pressure distribution. Different water saturation distributions were simulated by changing van Genuchten water-retention parameters (van Genuchten, 1980). Peat porosity was used as given above.

Parameter values from GPPT 3i (Table 1) were used as an example. Additional input parameters were diffusion coefficients (cm2 s–1) in air at 10°C: 0.208 for CH4 (Massman, 1998), 0.608 for He, 0.280 for Ne, 0.144 for C2H2 (all Fuller et al., 1966), and 0.190 for O2 (Massman, 1998; Perry and Green, 1997); and dimensionless Henry constants at 10°C: 22.7 for CH4, 111.5 for He, 88.3 for Ne, 0.8 for C2H2 (all Wilhelm et al., 1977), and 25.37 for O2 (Perry and Green, 1997). Equilibrium partitioning into the water phase was assumed. The Millington and Quirk (1961) model was used to calculate effective diffusion coefficients.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion and Conclusions
 REFERENCES
 
Gas Push-Pull Test Performance
During the extraction phase of all GPPTs, relative concentrations C* of CH4, Ne, and He decreased rapidly due to diffusion and dilution with soil air (Fig. 2 ). Background CH4 concentrations before active GPPTs were low, with 8.9 (H1), 13 (H2), and 1.9 µL L–1 (H3) (see also Fig. 1). At the end of active tests, CH4 concentrations again approached background values: the last extraction sample of active tests contained a CH4 concentration of 56 (H1), 30 (H2), and 9.8 µL L–1 (H3), respectively. As GPPT 1i and 3i were started 9 to 15 min after end of extraction of GPPT 1a and GPPT 3a, actual background concentrations before these tests were likely even lower, although they were not determined. At the end of an additional 15-min extraction phase after GPPT 2a, CH4 concentrations had decreased to 13 µL L–1, that is, to the same background concentration as before the active test. Furthermore, the CH4 concentration reached the same value at the end of an additional extraction phase after GPPT 2i. This confirmed stable CH4 background concentrations throughout the whole test procedure.


Figure 2
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 2. Breakthrough curves of He, Ne, and CH4 at the injection–extraction point during three pairs of gas push-pull tests (GPPTs); "a" designates curves from active GPPTs, and "i" designates curves from inactive tests. The part of each GPPT test pair where CH4 consumption was discernible is shown in more detail. Note the different scales of both axes. Noble gases could not be measured for the entire extraction period in GPPT 3 as the analytical method available at the time was less sensitive. In GPPT 1 He dropped below the quantification limit after 1.4 h, and Ne after 2.1 h.

 
In all GPPTs, breakthrough curves were clearly different for the different gases with highest relative concentrations for CH4 and lowest relative concentrations for He (Fig. 2). Neon and He breakthrough curves of active and inactive tests were in good agreement for all GPPT test pairs (Fig. 2, Details). Conversely, CH4 breakthrough curves from active GPPTs were slightly lower than those from inactive GPPTs with the largest difference in GPPT 3 (Fig. 2, Details). Lower CH4 curves in active tests indicated consumption of CH4. Neon, He, and CH4 breakthrough curves (Fig. 2) were integrated to obtain percentages of the injected mass of gases recovered during extraction. In all tests, between 2.8 and 10.4% of Ne, 0.8 and 5.4% of He, and 5.8 and 15.5% of CH4 were recovered. Relative C2H2 concentrations decreased during inactive GPPTs similar to the other gases (see section on simulations for data from GPPT 3i) with absolute concentrations dropping to 0.2 (GPPT 1i), 3.4 (GPPT 2i), and 4.6 mL L–1 (GPPT 3i) at the end of extraction. Oxygen concentrations fluctuated between 105 and 210 mL L–1 during the extraction of all tests.

First-Order Rate Constants and Methane Oxidation Activity
First-order rate constants were determined using CH4 from inactive GPPTs as a substitute tracer to account for a decrease in concentration due to diffusion and dilution with soil air. In all GPPTs, a ln-linear relationship between corrected relative CH4 concentrations and residence time, that is, the time the gas stayed in the subsurface (see "Materials and Methods"), indicated CH4 oxidation that followed approximately first-order kinetics (Fig. 3 ). This was apparent later during extraction in GPPT 1 and 2 than in GPPT 3. In GPPT 1 and 2, apparent first-order rate constants k were similar, whereas the k value determined in GPPT 3 was lower by a factor of two (Fig. 3). The 95% confidence interval was small for GPPT 1 and 3, but larger for GPPT 2 due to fewer, more scattered data points.


Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		FIG. 3. Plot for the determination of apparent first-order rate constants k for CH4 oxidation from three pairs of gas push-pull tests (GPPTs). Residence time is the total time the gas stayed in the subsurface (see "Materials and Methods"). Slopes from linear regressions (solid lines) are apparent first-order rate constants k. C*(CH4)a and C*(CH4)i are relative CH4 concentrations from an active and the corresponding inactive GPPT, respectively. For better readability, early data from GPPT 1 and 2 are not shown. First-order rate constants k are given with 95% confidence intervals.

 
Using apparent first-order rate constants obtained from GPPTs and in situ CH4 concentrations measured before GPPTs, we estimated in situ CH4 oxidation rates. Estimated rates at H1 ranged from 48 to 299 ng CH4 gdw–1 h–1 between 40 and 50 cm depth. At H2, 42 cm depth, the estimated CH4 oxidation rate was with 45 ng CH4 gdw–1 h–1, similar to the rate at H1, 40 cm depth. At H3 the estimated in situ CH4 oxidation rate was one to two orders of magnitude smaller (4.0 ng gdw–1 h–1).

Isotope Fractionation during Gas Push-Pull Tests
The {delta}13C value of CH4 in the injection mixture of all GPPTs was –45.9{per thousand}. During the first 0.2 h of extraction, CH4 became strongly enriched in {delta}13C with an average isotope shift of 19.6{per thousand} (Fig. 4 ). Subsequently, {delta}13C values increased further in all tests, with a maximum value of +20.5{per thousand} in GPPT 1a. Toward the end of extraction in GPPT 1a and 1i, {delta}13C values decreased again by 8.7 to 9.8 {per thousand}. In GPPT 2, the isotope shift in the active and the inactive test were similar. However, because of low CH4 concentrations, isotope data could only be obtained until just before the part of extraction where CH4 oxidation followed first-order kinetics. Conversely, in GPPT 1 and 3, there was a difference in isotope shifts with time between the active and inactive test, with a maximum difference of 3.5{per thousand} in GPPT 3 and a maximum difference of 15.5{per thousand} in GPPT 1. A larger isotope shift in GPPT 1a than in 1i was observed already before CH4 oxidation followed first-order kinetics, that is, after 1.04 h since end of injection (Fig. 4), which corresponds to 1.14 h residence time (Fig. 3). The {delta}13C value of CH4 in background soil air was –45.5{per thousand}, measured 2 d before GPPTs in H1.


Figure 4
View larger version (18K):
[in this window]
[in a new window]

  		FIG. 4. Stable carbon isotope ratios in CH4 over time during three pairs of gas push-pull tests (GPPTs) at the injection–extraction point; "a" designates active GPPTs, "i" designates inactive GPPTs.

 
Simulation of GPPTs
To evaluate the spatial distribution of gases during GPPTs and the possibility of using simulations as an additional tool for data analysis, we performed simulations using test parameters from GPPT 3i as an example. The van Genuchten water-retention parameters were varied from 2 to 4 for n, 0.05 to 0.1 cm–1 for {alpha} and 0.05 to 0.16 for residual saturation to obtain a similar water saturation distribution as observed during TDR measurements. Obtained water saturation curves bracketed the values observed during measurements (data not shown). The best fit to measured tracer breakthrough curves was achieved with van Genuchten parameters n = 4, {alpha} = 0.06 cm–1, and a residual saturation of 0.10 (Fig. 5 ) and could not be improved even by simulating two peat layers with different water saturation characteristics. The general trend of breakthrough curves and the relative position of breakthrough curves of different gases were represented well by simulations (Fig. 5), and mass recovery of gases was similar to the field experiment. However, simulated and measured data did not agree well enough to use simulations for breakthrough curve data analysis. Nevertheless, we could use simulations to analyze the transport of C2H2 relative to CH4, which is important for effective inhibition of CH4 oxidation. At a radius of 30 cm away from the injection point, C2H2 was delayed only by 10 min with respect to CH4 and reached a concentration of 0.01 mL L–1 18 min after injection started (Fig. 6 ). When only 40 mL L–1 instead of 80 mL L–1 C2H2 were injected, as was the case in GPPT 1i and 2i, the latter time increased only slightly, from 18 to 20 min (data not shown). At the end of injection, C2H2 reached the same concentration of 0.01 mL L–1 at the edge of the simulated domain, that is, at 50 cm radius (data not shown).


Figure 5
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 5. Measured and simulated breakthrough curves at the injection–extraction point during gas push-pull test 3i. Symbols represent measured data, lines represent simulations (van Genuchten water-retention parameters n = 4, {alpha} = 0.06 cm–1, and residual saturation 0.10).

 

Figure 6
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 6. Simulated CH4 and C2H2 concentrations at a 30-cm radius from the injection point over the entire duration of gas push-pull test 3i. Arrows indicate the time when each gas reached a concentration of 0.01 mL L–1.

 

   Discussion and Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
REFERENCES
 
Methane as a Substitute Tracer to Quantify Methane Oxidation
Different diffusion coefficients resulted in clearly different transport behavior of noble gases and CH4 during GPPTs with diffusion-dominated gas transport, similar to previous laboratory experiments and simulations (Gonzalez-Gil et al., 2007). Consequently, noble gases were unsuitable as tracers to quantify CH4 oxidation from GPPTs using simplified methods that do not require knowledge of physical parameters of the system. In contrast to GPPTs performed in a homogeneous sand (Gonzalez-Gil et al., 2007), it was also not possible to simulate breakthrough curve data accurately enough to use a modeling approach for data analysis, most likely because of the peat's heterogeneity. Instead, we presented an approach to use the reactant CH4 as a substitute tracer. We showed that during GPPTs in a peat bog, the three prerequisites to apply this approach were met. First, reproducibility of two consecutive GPPTs with respect to gas transport was confirmed by coinciding Ne and He breakthrough curves during all three test pairs (Fig. 2). Second, low and similar CH4 background concentrations before active and inactive tests indicated that background CH4 did not interfere with breakthrough curve analysis. In addition, an extended extraction phase after one active and inactive test confirmed the stability of CH4 background concentrations throughout one test pair. And third, the presence of sufficient inhibitor was confirmed by C2H2 concentrations at least 20 times higher than C2H2 concentrations required for 94% (Chan and Parkin, 2000) or 100% (Bodelier and Frenzel, 1999) inhibition in laboratory experiments. In the field, effectiveness of C2H2 as an inhibitor of CH4 oxidation has been shown during previous GPPTs (Urmann et al., 2005). In addition, simulations of GPPT data presented here showed that breakthrough of C2H2 was only slightly delayed compared with CH4 due to higher partitioning and a lower diffusion coefficient and that, at a 30-cm radius, a C2H2 concentration sufficient for inhibition (0.01 mL L–1 [Bodelier and Frenzel, 1999; Chan and Parkin, 2000]) was reached within a quarter of the injection time. At the end of injection, this concentration was even reached at the edge of the simulated hummock. Only up to 15% of injected CH4 were recovered, meaning that the CH4 injected first was most likely not retrieved during extraction.

The described procedure should be applicable for the quantification of CH4 oxidation in a range of ecosystems. However, care has to be taken that prerequisites are met. Rate constants for CH4 oxidation could, for example, be underestimated if inhibition efficiency was too low. Furthermore, results may become difficult to interpret if CH4 background concentrations strongly varied during one test pair. This could, for example, happen if significant methanogenesis occurred within the test zone and was inhibited by C2H2, which is also a known inhibitor for methanogenesis even if not as effective as for CH4 oxidation (Chan and Parkin, 2000). A further potential disadvantage of the use of C2H2 is that inhibition, that is, binding of C2H2 to the enzyme, is nonreversible (Prior and Dalton, 1985). Even though methanotrophs have been shown to recover fully from C2H2 inhibition in laboratory studies (Bodelier and Frenzel, 1999), this may complicate time-dependent studies of CH4 oxidation at the same location using GPPTs. Recovery of methanotrophic activity after C2H2 application in the field is currently under investigation. In this study, we used C2H2 as an inhibitor because of its efficiency at very low concentrations (Bodelier and Frenzel, 1999; Chan and Parkin, 2000). As an alternative to C2H2, a reversible gaseous inhibitor such as difluoromethane (Miller et al., 1998) could be used.

Quantification of Rate Constants
In all GPPTs, dilution and diffusive transport led to a fast decrease in relative gas concentrations and low recoveries of injected gases. This in itself does not impede accurate quantification of rate constants (Haggerty et al., 1998). However, a fast decrease in concentrations limits the available test time, which may complicate quantification of low activity. Despite fast decrease in CH4 concentrations, we were able to quantify apparent first-order rate constants for CH4 oxidation in the unsaturated zone of three different hummocks using CH4 as a substitute tracer. Moreover, the sensitivity of the GPPT was sufficient to detect activity, even though (i) apparent first-order rate constants were at the lower end of those previously determined above a contaminated aquifer and (ii) CH4 oxidation followed apparent first-order kinetics at lower CH4 concentrations than in previous GPPTs (Urmann et al., 2005). During GPPTs at all sites, CH4 oxidation following first-order kinetics was only observed in the later part of extraction. Most likely, in the earlier part, oxidation followed Michaelis–Menten kinetics as indicated by a curvature in the rate plots (Fig. 3, not well visible for GPPT 2 due to scale). Michaelis–Menten parameters were not quantified as this is difficult to achieve from a single active GPPT (unpublished data, 2007), and quantification of these parameters was not the purpose of this study.

Quantification of Methane Oxidation Rates
At H3 the estimated in situ CH4 oxidation rate before GPPTs—that is, the in situ rate uninfluenced by the addition of substrate—was one to two orders of magnitude lower than at H1 and H2. As the k value at H3 was only a factor of two lower, the difference was due to the low in situ concentration at H3, which was similar to ambient CH4 concentrations. At ambient concentrations, different kinetic parameters, such as higher affinity for CH4, are usually observed (Bender and Conrad, 1992; Gulledge et al., 2004). Therefore, the obtained in situ rate at H3 has to be interpreted with care since the first-order rate constant may not apply for measured in situ CH4 concentrations. However, the observation of apparent first-order kinetics at high CH4 concentrations during GPPT 3 suggests that higher in situ concentrations have been occurring at this site, inducing high activity.

Because high variability in CH4 gas-phase concentrations was observed at all sites in the peat bog (data not shown), CH4 oxidation rates could not be inferred from CH4 concentration gradients for comparison with those calculated from GPPT results. Therefore, we estimated rates per gram dry peat to compare our results with literature data. Although rates were potentially associated with a large error of up to ~70% due to the use of literature values for porosity and bulk density, the actual error was likely smaller because the site where the employed values were determined had characteristics similar to our site (Kellner and Lundin, 2001). For a field site, it would be more meaningful to report CH4 oxidation rates per volume of peat, which would also reduce the potential error because bulk density is not needed in these calculations. However, this would allow little comparison with data currently available in the literature.

Comparison with values reported in the literature showed that at H1 and H2, estimated in situ CH4 oxidation rates before GPPTs were two orders of magnitude lower than in situ rates calculated from profiles in a northern peat bog in summer but one order of magnitude higher than rates determined at the same site in autumn (Fechner and Hemond, 1992). Because reports of in situ oxidation rates are limited, we further compared our results with potential oxidation rates measured in laboratory incubations. In situ rates in our study were three orders of magnitude lower than potential oxidation rates at a temperate bog ranging from 1.3 to 37.9 µg CH4 gdw–1 h–1 (McDonald et al., 1996) and similar to a range of potential oxidation rates of 32 to 650 ng CH4 gdw–1 h–1 observed in a boreal bog (Whalen and Reeburgh, 2000). Considering that oxidation followed first-order kinetics in our study, potential activity may have been in a similar range as in the temperate but higher than in the boreal bog.

Isotope Fractionation during Gas Push-Pull Tests
Methane oxidation during GPPTs with restricted gas transport has previously been confirmed by a shift in CH4 stable carbon isotope ratios toward heavier values (Urmann et al., 2005). This shift was induced by faster oxidation of 12CH4 over 13CH4 (Coleman et al., 1981; Whiticar, 1999). At the same site, isotope shifts during a GPPT with the coinjection of C2H2 as inhibitor of CH4 oxidation were small. In contrast, in the highly porous peat bog in this study, large isotopic shifts toward heavier {delta}13C values were observed during GPPTs with the coinjection of C2H2, indicating strong isotope fractionation due to gas-phase diffusion (Fig. 4). This is in accordance with theory and experimental observations of gas-phase diffusion of CH4 (De Visscher et al., 2004). The reversal of the observed isotope trend toward more negative ("lighter") values at the end of GPPT 1 may be explained by mixing with CH4 in background soil air, which was comparatively light.

Isotope ratios in active GPPTs generally showed a similar trend with time as those in inactive tests, underlining the strong influence of diffusion on isotope fractionation under the test conditions in this study. Despite this strong influence, a clearly discernible difference in isotope ratios between active and inactive test was observed throughout a large part of extraction in GPPT 1. Consequently, isotopic shifts provided a valuable independent verification of microbially mediated CH4 oxidation that followed Michaelis–Menten and first-order kinetics. In GPPT 3, on the other hand, differences between active and inactive tests were small. In this case, the observed differences in {delta}13C with time could only be attributed to microbial oxidation because similar transport behavior of gases in both tests was verified by noble gas breakthrough curves. In GPPT 2 no difference in {delta}13C time trends was apparent between active and inactive test, although CH4 oxidation was observed in the active test. Consequently, similar shifts in isotope ratios do not necessarily indicate absence of CH4 oxidation.

The degree of fractionation from CH4 oxidation, described by the fractionation factor, is likely an important parameter determining whether isotope fractionation is a suitable indicator for CH4 oxidation during GPPTs. The fractionation factor is highly variable, with one possible determinant being the degree of mass transfer limitation of CH4 oxidation (Templeton et al., 2006). At high CH4 concentrations combined with the consumption of a small fraction of CH4 and therefore no mass transfer limitation, high fractionation was observed as enzymatic fractionation was probably fully expressed (Templeton et al., 2006; unpublished data, 2007). This could explain why in GPPT 1, with a very high CH4 injection concentration, isotope fractionation due to microbial oxidation could be clearly distinguished from isotope fractionation due to diffusion.

In this study, we presented an approach to quantify CH4 oxidation in situ based on GPPTs combined with inhibitors that can be applied in a range of ecosystems. Stable carbon isotope data obtained during GPPTs may provide independent verification of CH4 oxidation under certain conditions but must be interpreted with care. The presented procedure may be useful for the quantification of other microbial processes for which suitable inhibitors but no suitable tracers are available.


   ACKNOWLEDGMENTS
 
We thank Matthias Saurer and Rolf Siegwolf (PSI, Villigen, Switzerland) for help with stable carbon isotope measurements and Elena Norina (ETH Zurich) for help with field work. We also thank Niklaus Troxler from Pro Natura Lucerne for his cooperation. Helpful suggestions by three anonymous reviewers were highly appreciated.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Urmann, K.
Right arrow Articles by Zeyer, J.
Related Collections
Right arrow Isotopes
Right arrow Field-Scale Studies
Right arrow Microbial Processes

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome