Acetoclastic methanogens have been described to be inhibited at much lower concentrations of methyl fluoride, CH3F, than H2/CO2-utilizing methanogens. Therefore, we tested whether CH3F inhibition may be used to determine, in anoxic rice field soil, the contribution of H2/CO2-dependent methanogenesis to the total CH4 production by comparing this technique with the incorporation of 14CO2 into CH4. In general, addition of 0.01–1% CH3F to the gas phase resulted in an immediate partial inhibition of the total CH4 production which lasted for at least 200 h. Inhibition increased with the logarithm of the initial CH3F concentration up to about 0.2–0.6%. The initial CH3F concentration slowly decreased with time, probably due to decomposition. CH4 production sometimes completely recovered during the course of the experiment. The presence of CH3F resulted in the accumulation of acetate, the final concentration of which was usually stoichiometrically related to the deficit in CH4 production and increased with the initial CH3F concentration. In some experiments, acetate accumulation was larger than expected from the CH4 deficit and a substantial incorporation of 14CO2 into acetate was observed. Hydrogen, on the other hand, was only slightly elevated in the presence of CH3F. Addition of increasing CH3F resulted in an increase of the percentage of H2-dependent methanogenesis (measured by conversion of 14CO2 to 14CH4) demonstrating that acetoclastic methanogenesis was preferentially inhibited by CH3F. However, the conversion of 14CO2 to 14CH4 was also slightly inhibited by CH3F. Apparently, CH3F inhibited the H2-dependent methanogenesis to some extent, depending on the concentration of CH3F applied. Indeed, the ratio between the residual CH4 production rate and the fraction of CH4 produced from 14CO2 decreased with an increasing CH3F concentration. A ratio of unity was obtained at initial CH3F concentrations of 0.2–0.6% (58–174 μM). Both methods, i.e. inhibition using 0.5% CH3F and conversion of 14CO2 to 14CH4, were applied to determine the temporal change of the contribution of H2/CO2-dependent methanogenesis to the total CH4 production in two different batches of Italian rice field soil during a 120-days anoxic incubation period. The results of the two methods agreed well within the error of the methods and showed a relatively constant contribution of H2/CO2-dependent methanogenesis of about 25–30% as soon as CH4 was produced at a steady rate and H2 partial pressures had stabilized at about 1.5–2.5 Pa.
The breakdown of organic matter in anaerobic environments results in the formation of CO2 and CH4 as final gaseous products. Polysaccharides, for example, are degraded to equal amounts of CH4 and CO2, when oxidants such as nitrate, Mn(IV), Fe(III) and sulfate have been exhausted [1,2]. The most important immediate precursors of methanogenesis are acetate and H2/CO2. The degradation pathway of polysaccharides is such that about 2/3 of the produced CH4 should be derived from acetate and 1/3 from H2/CO2 if steady state conditions exist [2,3]. Nevertheless, the contribution of these two precursors can range between 0 and 100% depending on which methanogenic system is studied . The reason for such a large variability is not always obvious . For more research, it is necessary to have a reliable technique which allows to measure the contributions of H2/CO2 and acetate to CH4 production. The most widely applied method is the measurement of the conversion of radioactive precursors to CH4. However, the application of radioactive materials is not always possible. Furthermore, it is necessary to have access to radio-gas chromatography or to use cumbersome chemical purification steps [4–6].
Methyl fluoride has originally been introduced as a ‘specific’ inhibitor for aerobic CH4-oxidizing bacteria [7,8]. Subsequently, CH3F was also found to inhibit methanogenesis in aquatic environments and anoxic root preparations provided that CH4 was mainly produced from acetate . Studies with pure cultures of different methanogenic archaea confirmed that acetate-dependent methanogenesis was inhibited at much lower CH3F concentrations than H2/CO2-dependent methanogenesis . Hence, ‘specific’ inhibition of acetate-dependent methanogenesis by CH3F may be a useful tool for determining its contribution to the total CH4 production in methanogenic environments.
Therefore, we have systematically studied the inhibition of CH4 production by CH3F in two different rice field soils and compared the thus determined contribution of acetate-dependent methanogenesis to that determined with the radiotracer technique.
2Materials and methods
Two different soil samples were used, both originating from a paddy field at the Italian Rice Research Institute near Vercelli, Italy. For characterization of the location and the soil, see . The ‘fresh soil’ was taken from the Italian field in March 1993, air-dried and stored as dry lumps until the beginning of the experiment . The ‘recycled soil’ was used for growing rice in large microcosms (40×40×20 cm) at our greenhouse in Marburg. After harvest, the soil was drained, air-dried and stored as dry lumps.
For preparation of soil slurries, the dry soil lumps were broken in a mechanical grinder and sieved through a 0.2-mm mesh size. The sieved soil was used for several experiments until the stock was depleted, then, a new stock of sieved soil was prepared. For experiments with the ‘fresh soil’, 10 g dry weight (g-dw) was placed into a 26-ml glass pressure tube, mixed with 10 ml sterile anoxic distilled water, closed with a butyl rubber stopper and incubated under a N2 atmosphere without shaking. The same procedure was for the ‘recycled soil’, but in addition, 20 mg dry rice straw was added to each tube. The straw was previously cut into small pieces (<2×1×0.2 mm) using a blender.
All experiments were done in triplicate at 30°C. Data are given as arithmetic mean±S.D.
Inhibition by CH3F (>98%: Fluorochem, Old Glossop, Derbyshire, UK) was done with soil slurries that were anoxically incubated for about 8–25 days to ensure that CH4 was produced at a constant rate. CH3F was injected into the headspace of the tubes and mixed by vigorous shaking (by hand) to give different initial CH3F concentrations. Both the CH4 and the CH3F concentrations were measured by gas chromatography with a flame ionization detector  and followed with the incubation time. Using a Bunsen solubility coefficient of 0.72 at 30°C , a concentration of 1% CH3F in the gaseous headspace is in equilibrium with 290 μM dissolved CH3F. Before taking gas samples (0.2–1.0 ml) from the headspace with gas-tight syringes, the tubes were vigorously shaken by hand for equilibration of the gas and the liquid phases. The CH4 production rate was determined from the linear increase of CH4 with time using linear regression analysis. Hydrogen concentrations were measured using gas chromatography with a reductive gas detector (RGD2) .
For measurement of the conversion of 14CO2 to 14CH4, the pre-incubated soil slurries were flushed with N2 to remove the produced CH4, then, 1–2 μCi (37–74 kBq) of NaH14CO3 (Amersham, Braunschweig, Germany) was injected into each tube and the incubation continued. Radioactive and unlabelled CH4 and CO2 were repeatedly measured in the gas phase of the tubes by using radio-gas chromatography . The fraction (f) of CH4 that was produced from the reduction of H14CO3− was calculated from the specific radioactivities (SR) of 14CH4 (SRCH4) and 14CO2 (SRCO2) measured in the gas phase using f=SRCH4/SRCO2. At the end of incubation, the soil slurry was retrieved and the concentrations of radioactive and unlabelled acetate (and other fatty acids if present) were analyzed using a high pressure liquid chromatograph equipped with a radiodetector . The recovery of radioactivity in the soil aqueous phase plus gas phase was in the order of 70%. The fraction of acetate that was produced from reduction of H14CO3− was calculated analogously to that of CH4 (see above) using the SRs of the acetate-carbon and of the 14CO2 and assuming that each of the carbon atoms of acetate was equally labelled.
A total of nine different experiments were conducted, six with recycled (Rad3, Rad4, Rad10, Rad11, Bod6, Bod7) and three with fresh soil (Rad6, Rad7, Bod4). In six of the experiments (labelled ‘Rad’), the conversion of 14CO2 to 14CH4 was measured in parallel to the inhibition of CH4 production by different CH3F concentrations.
Long term incubations (up to 140 days) were initiated with many parallel soil slurry tubes. One set of triplicate tubes was used to follow the gaseous CH4 and H2 concentrations throughout the incubation. The inhibition by CH3F at a particular time of the incubation was assayed by taking triplicate tubes from the initial set of incubations flushing their headspace with N2 prior to addition of CH3F and then incubating them for another 9–14 days to follow the production of CH4. Then, the tubes were discarded. Uninhibited control tubes were also assayed in triplicate and treated in the same way, except that 1 μCi (37 kBq) of NaH14CO3 was added in addition and the temporal change of radioactive CH4 and CO2 was followed as well.
Addition of CH3F to methanogenic rice soil (batch Rad10) resulted in the immediate inhibition of part of the CH4 production (Fig. 1A). The extent of inhibition increased with the initial CH3F concentration being maximal at about 0.2%. After addition, the concentration of CH3F slowly decreased, indicating that CH3F was decomposed (Fig. 1B). We assume that the CH3F decrease was due to microbial decomposition, since CH3F did not decrease in autoclaved controls. The results of similar experiments are summarized in Fig. 2, showing that inhibition increased linearly with the logarithm of the initial CH3F concentration up to about 0.2–0.6% and then became increasingly weaker. Although the slope of residual activities versus logarithm of initial CH3F concentrations was similar for all soil batches, the absolute CH3F concentration at which a particular inhibition was reached was different.
In contrast to the total CH4 production, addition of CH3F inhibited the production of 14CH4 from 14CO2 by <42% (Fig. 1C), while the total CH4 production was inhibited by <71% (Fig. 1A). Consequently, the percentage contribution of H2/CO2-dependent methanogenesis to total CH4 production increased in the presence of CH3F (Fig. 1D). Similar results were obtained with two other batches of rice field soil (Rad7, Rad11). One batch of soil (Rad4), however, showed a pronounced inhibition of H2/CO2-dependent methanogenesis by addition of CH3F (data not shown). At an initial CH3F concentration of 0.15%, the residual rate of H2/CO2-dependent methanogenesis was only 35% of the rate of the control. Nevertheless, this soil batch also exhibited a stimulation of the percentage contribution of H2/CO2-dependent methanogenesis to the total CH4 production (from 24% in the absence to 55% in the presence of CH3F), indicating that acetate-dependent CH4 production was even more inhibited by CH3F than H2/CO2-dependent CH4 production.
Inhibition of CH4 production by CH3F resulted in accumulation of acetate, but the steady state partial pressure of H2 increased only slightly (Fig. 3). In this particular batch of soil (Bod7), the accumulation of acetate was equal to the deficit in CH4 production (on a molar basis), indicating that CH3F specifically inhibited the acetate-depending methanogenesis. The same result was obtained with another soil batch (Bod6) and also in the earlier experiment by Frenzel and Bosse , but in most of the soil batches tested (Bod4, Rad6, Rad7, Rad10), the accumulation of acetate was higher than expected from the deficit in CH4 production. An example is shown in Fig. 4. In this soil batch (Rad6), at initial CH3F concentrations ≥0.46%, acetate concentrations were about 30–40% higher than expected from the CH4 deficit (Fig. 4C), whereas H2 partial pressures were again only slightly elevated in the presence of CH3F (Fig. 4D). In addition, propionate concentrations were almost 10-fold higher in the presence than in the absence of CH3F, but were not proportionally related to the initial CH3F concentrations (Fig. 4D). We also observed incorporation of 14CO2 into acetate, but only at initial CH3F concentrations ≥0.46% (Table 1). The incorporation of 14CO2 into acetate was so high that both the carboxyl and the methyl group of acetate must have been labelled. About 40% of the accumulated acetate was completely synthesized from CO2. Similar results were obtained with soil batches Rad7 and Rad10.
Table 1. Incorporation of 14CO2 into the acetate pool of anoxic rice field soil (Rad6) in the presence of CH3F
Initial CH3F (%)
SR-CO2 (Bq μmol−1)
SR-ac (Bq μmol−1)
Mean±S.D. (n=3) of the SR of CO2 and acetate (ac) and the fraction of acetate produced from CO2 (=0.5 SR-ac/SR-CO2).
Prolonged incubation (>10 days) in the presence of CH3F usually resulted in a recovery of the inhibited methanogenic activity (e.g. Fig. 4A). Possibly, the CH3F concentration had decreased to a threshold at which it was no longer inhibitory. For example, CH4 production in batch Rad6 only recovered at the two lowest initial CH3F concentrations (Fig. 4A, B). In batch Rad10, on the other hand, CH4 production did not recover although a relatively large percentage of the initial CH3F concentration was depleted (Fig. 1A, B). Possibly, recovery of activity also requires acclimatization of the methanogenic populations.
The fractions of H2/CO2-dependent methanogenesis that were determined from the conversion of 14CO2 to 14CH4 ranged among the different soil batches between 10–24% (legend of Fig. 5). These fractions were compared to the residual activities of CH4 production in the presence of CH3F. If CH3F would specifically inhibit only the CH4 production from acetate but not from H2/CO2, the ratio between the residual activity and the fraction of CH4 produced from CO2 should be unity. Our results obtained with different soil batches show that this ratio decreased with increasing an CH3F concentration (Fig. 5). Obviously, acetate-dependent methanogenesis was not completely inhibited when the CH3F concentration was too low and H2/CO2-dependent methanogenesis was in addition inhibited, at least partially, when the CH3F concentration was too high. A ratio of one was only achieved within a relatively small range of initial CH3F concentrations, i.e. between 0.2 and 0.5%. One soil batch (Rad4) out of six different soil batches was an outlier, since a ratio of one was already achieved at an initial CH3F concentration of 0.07% (Fig. 5). In this soil batch, we also observed a relatively high sensitivity of the H2/CO2-dependent methanogenesis against CH3F (see above). Obviously, the methanogenic population in this particular batch of soil behaved differently for unknown reasons.
The temporal change of the contribution of H2/CO2-dependent methanogenesis to the total CH4 production was followed during anaerobic incubation of recycled and fresh Italian rice field soil using both inhibition by 0.5% CH3F and conversion of 14CO2 to 14CH4. The results obtained with the two soils were similar, those with the fresh soil are shown in Fig. 6. Production of CH4 started after a lag phase of about 5–10 days. Methane accumulated at a constant rate for about 40–50 days and then, the CH4 production rate slowed down. Hydrogen partial pressures showed a transient maximum after about 20 days of incubation and then stabilized at about 1.5–2.5 Pa. The percentage contribution of H2/CO2-dependent methanogenesis to total CH4 increased from a relatively low 10% at the beginning (i.e. day 10) to about 25–30% at day 30 and then stayed more or less constant at this value until the end of incubation. The values determined with the CH3F inibition and the 14CO2 conversion technique agreed fairly well.
Our study has shown that methyl fluoride can be used to determine the percentage contribution of H2/CO2-dependent methanogenesis to the total CH4 production. At an initial concentration of 0.2–0.6% CH3F in the gas phase, which at 30°C is equivalent to 58–174 μM dissolved CH3F, acetate-dependent methanogenesis was rather specifically inhibited, indicating that the residual rate of CH4 production was only due to H2/CO2-dependent methanogenesis. This residual rates agreed fairly well with the rates of 14CO2 conversion to 14CH4 and both methods allowed equally well the determination of the temporal pattern of H2/CO2-dependent methanogenesis in anoxic rice soils incubated for up to 120 days.
However, the CH3F inhibition technique should be used with caution. CH3F was not an absolutely specific inhibitor, as H2/CO2-dependent methanogenesis was also increasingly inhibited when CH3F concentrations were increased. It was necessary to adjust the CH3F concentration within a relatively narrow range to obtain results comparable to those of the 14CO2 conversion technique. The effective range of CH3F may be different in other methanogenic systems and thus would have to be carefully adjusted. A similar problem has been noted in pure cultures of methanogenic archaea in which the degree of inhibition of aceticlastic methanogenesis is dependent on the strain tested and in which at higher CH3F partial pressures, hydrogenotrophic methanogenesis is also inhibited .
In addition, it should be checked whether CH3F is decomposed during the incubation. We found a gradual decrease in the CH3F concentration with time and ultimately a recovery of the CH4 production to its previous activity. For systems with a very rapid decomposition of CH3F, the adjustment of the effective CH3F concentration may be difficult. Microbial decomposition of CH3F has been demonstrated for aerobically incubated sediment and for methanotrophic and nitrifying bacterial cultures [8,16]. In anoxic systems, on the other hand, decomposition of CH3F has so far not been reported. Krone et al.  demonstrated the methanogenic degradation of CH2F− corrinoids suggesting that CH3F may be decomposed by microorganisms containing corrinoids. However, more research is necessary to learn about anaerobic decomposition of CH3F.
We would also like to point out that the residual methanogenic activity in the presence of CH3F is not necessarily due to only H2/CO2-dependent methanogenesis. In many methanogenic environments, CH4 is produced from trimethyl amine and then seems to be insensitive to CH3F inhibition [7,9]. Analogously, methanogenesis from methanol or dimethyl sulfide may not be inhibited by CH3F, but explicit studies are lacking. Hence, the residual methanogenic activity in the presence of CH3F can only be taken for the H2/CO2-dependent methanogenesis in those anaerobic systems for which acetate and H2/CO2 are known to be the exclusive or at least the dominant precursors of CH4 production.
Our observation that H2/CO2-dependent methanogenesis was partially inhibited by CH3F is consistent with the result that rates of acetate accumulation in the presence of CH3F were often larger than expected from the CH4 deficit due to the inhibition of acetate-dependent methanogenesis. Since we observed a substantial conversion of 14CO2 to acetate under these conditions, H2 was possibly channelled into homoacetogenesis as H2-dependent methanogenesis was increasingly inhibited by CH3F. On the other hand, H2 concentrations did not increase very much and it remains unclear how homoacetogens would manage to produce acetate from these relatively low H2 partial pressures [18,19]. The only explanation which comes to our mind is that H2 was possibly available at a higher concentration within soil microsites than measured. We measured H2 in the gas headspace which was kept in equilibrium with the dissolved H2 by frequent shaking, so that the measured H2 partial pressures were characteristic for the bulk soil solution. Nevertheless, H2 concentrations may be higher within soil microsites and microbial aggregates due to ongoing H2 production. There is numerous evidence for disequilibrium of H2 between soil microsites and the gas phase (e.g. [20,21]). Unfortunately, it is not possible to permanently agitate soil suspensions because of adverse affects on microorganisms and microbial aggregates  and thus, it is not possible to keep the H2 within microsites in equilibrium with the gas phase. Anyway, our observations indicate a substantial potential of chemolithotrophic homoacetogenesis in anoxic rice field soil. This conclusion is in agreement with earlier studies [23,24].
Propionate levels were also increased in the presence of CH3F suggesting that propionate degradation was inhibited. Indeed, the thermodynamic conditions of propionate degradation to acetate, CO2 plus 3H2 were only slightly exergonic, especially in the inhibited soil samples (with ΔG=−17.7, −16.9, −19.7, −8.7 and −5.5 kJ mol−1 propionate for the 0, 0.08, 0.18, 0.46 and 0.93% CH3F, respectively). However, relatively small negative ΔG values are not unusual for propionate degradation in methanogenic paddy soil [24,25]. Again, however, one may argue that thermodynamic conditions in microsites may be even less favorable and that there, acetate and H2 concentrations are too high to allow the degradation of propionate which then accumulates [21,26].
Using the CH3F inhibition technique, we determined that the contribution of H2/CO2-dependent methanogenesis increased with the time of incubation of anoxic rice field soil and soon reached a constant value of about 25–30% of the total CH4 production. This percentage is similar to that determined at the Italian field site  and is only slightly less than the value of 33% that is theoretically expected when polysaccharides are methanogenically degraded in the absence of homoacetogenic fermentation . However, as already discussed in earlier studies [15,24,28], we assume that homoactogenic bacteria do contribute to the breakdown of organic matter under anaerobic conditions and thus increase the relative contribution of acetate-dependent methanogenesis to the total CH4 production.
We thank Dr Peter Frenzel for methodological advice and helpful discussion and the European Union (ENV4-CT97-04098) for financial support.