OUP user menu

Contribution of hydrogen to methane production and control of hydrogen concentrations in methanogenic soils and sediments

R. Conrad
DOI: http://dx.doi.org/10.1111/j.1574-6941.1999.tb00575.x 193-202 First published online: 1 March 1999


Hydrogen is, with acetate, one of the most important intermediates in the methanogenic degradation of organic matter and serves as substrate for methanogenic archaea. Hydrogen should theoretically account for 33% of total methanogenesis when carbohydrates or similar forms of organic matter are degraded. Many methanogenic environments show both much lower and much higher contributions of H2 to CH4 production than is considered normal. While the lower contributions are relatively easily explained (e.g. by the contribution of homoacetogenesis), the mechanisms behind higher contributions are mostly unclear. In methanogenic environments H2 is rapidly turned over, its concentration being the result of simultaneous production by fermenting plus syntrophic bacteria and consumption by methanogenic archaea. The steady-state concentration observed in most methanogenic environments is close to the thermodynamic equilibrium of H2-dependent methanogenesis. The threshold is usually equivalent to a Gibbs free energy of −23 kJ mol−1 CH4 that is necessary to couple CH4 production to the generation of 1/3 ATP. Methanogenesis from H2 is inhibited if the H2 concentration decreases below this threshold. Concentrations of H2 can only be decreased below this threshold if a H2-consuming reaction with a lower H2 threshold (e.g. sulfate reduction) takes over at a rate that is equal to or higher than that of methanogenesis. The instantaneous and complete inhibition of H2-dependent CH4 production that is often observed upon addition of sulfate can only be explained if a comparably high sulfate reduction potential is cryptically present in the methanogenic environment.

  • H2
  • CH4
  • Acetate
  • Fermentation
  • Syntrophy
  • Methanogenesis
  • Homoacetogenesis
  • Threshold
  • Gibbs free energy
  • Km

1 Introduction

Methanogenic archaea utilize only a limited number of substrates, the most important ones being acetate and H2/CO2 (or formate) [1]. Most methanogenic archaea are able to utilize H2/CO2 and such methanogens can be found in every methanogenic environment. Indeed, H2 is a ubiquitous compound in anaerobic environments where it exhibits a fast turnover but usually occurs at only very low concentration [24]. Low H2 concentrations are a thermodynamic prerequisite for the degradation of alcohols and fatty acids by H2-producing syntrophic bacteria [5]. In methanogenic environments where inorganic electron acceptors other than CO2 are not available, consumption of H2 is only possible by methanogenic archaea and homoacetogenic bacteria. There, degradation of alcohols and fatty acids is usually accomplished by syntrophy between H2-producing syntrophic bacteria and H2-consuming methanogenic archaea [5].

In this MiniReview I will address the following two questions. (1) What is the percentage contribution of H2 to the production of CH4? (2) How is the H2 concentration and methanogenesis controlled by competition? I do not address the possibility that formate may replace H2 in many of the processes [6] which, however, should have no consequences for the principal conclusions.

2 Contribution of H2 to methanogenesis

Hydrogen is a product of the anaerobic degradation of organic matter by fermenting and syntrophic bacteria. The most abundant source of dead organic matter in natural environments is usually plant material consisting of lignin and polysaccharides. Some aquatic sediments receive a large input of dead crustaceans consisting of chitin. Lignin is largely recalcitrant under anaerobic conditions [7], but methanol may be released from the methoxy groups and thus may support methanogenesis to a limited extent. In general, however, we may assume that the anaerobic degradation process is largely driven by carbohydrates as the dominant substrate. This assumption is valid for aquatic sediments, peat, other wetlands, ruminants, arthropods feeding on plant material, and for many types of sewage sludge.

The anaerobic degradation pathway of dead organic matter is in principle well known [1]. Different groups of microorganisms participate in the degradation which basically proceeds in three steps. (1) Fermenting bacteria excrete enzymes that hydrolyze organic polymers (e.g. polysaccharides) and catabolize the resulting monomers to alcohols, fatty acids and H2. (2) Syntrophic bacteria further degrade the alcohols and fatty acids to acetate, H2 (alternatively formate) and CO2. (3) Acetate and H2 (alternatively formate) plus CO2 finally serve as substrates for methanogens. Alternatively, many of the monomers (e.g. sugars) can be catabolized by homoacetogenic bacteria to acetate which then serves as substrate for acetotrophic methanogens converting it to CH4 and CO2 (Fig. 1).


Pathway of anaerobic degradation of organic matter to methane.

Using the degradation of glucose as an example, most of the standard Gibbs free energy content is utilized during the first stage, i.e. the fermentation to alcohols and fatty acids (Figs. 1 and 2; Table 1). The next stage, i.e. the syntrophic degradation of alcohols and fatty acids to acetate and H2, is usually endergonic under standard conditions (Table 1) and is only possible when combined with H2-consuming methanogenesis. Less than half of the Gibbs free energy content of glucose is available for the syntrophic degradation of the alcohols and fatty acids to CH4 and CO2 (Fig. 2; Table 2) and this energy has to be shared among the syntrophs and the methanogens. Only if the fermentation step is homoacetogenesis (reaction 1.5), the residual free energy (about a quarter of the total) is exclusively available for acetotrophic methanogenesis (Fig. 2). In fact, there is no thermodynamic reason why homoacetogenic degradation of carbohydrates coupled to acetotrophic methanogenesis should not be a major pathway in anoxic environments. At the moment, however, the role of homoacetogenesis in methanogenic environments is unclear.


Residual standard Gibbs free energy (ΔG°′) after each reaction stage in the methanogenic degradation of glucose utilizing different glucose fermentation reactions (equation numbers from Table 1 in parentheses).

View this table:

Standard Gibbs free energies (ΔG°′) of defined stages in the degradation of glucose to CH4 (calculated after [38] using CO2 in gaseous state)

#ReactionΔG°′ (kJ mol−1 substrate)
1.1C6H12O6→2 CH3CHOHCOOH−198.1
1.2C6H12O6→2 CH3CH2OH+2 CO2−235.0
1.3C6H12O6→2/3 CH3CH2CH2COOH+2/3 CH3COOH+2 CO2+8/3 H2−248.0
1.4C6H12O6→4/3 CH3CH2COOH+2/3 CH3COOH+2/3 CO2+2/3 H2O−311.4
1.5C6H12O6→3 CH3COOH−311.2
2.2CH3CH2OH→CH3COOH+2 H2+9.6
2.3CH3CH2CH2COOH+2 H2O→2 CH3COOH+2 H2+48.3
2.4CH3CH2COOH+2 H2O→CH3COOH+CO2+3 H2+31.8
1-2C6H12O6+2 H2O→2 CH3COOH+2 CO2+4 H2−216.1
Hydrogenotrophic methanogenesis
34 H2+CO2→2 H2O+CH4−32.7
1-3C6H12O6→2 CH3COOH+CO2+CH4−346.8
Acetotrophic methanogenesis
1-4C6H12O6→3 CO2+3 CH4−418.1
View this table:

Examples of the contribution of H2 to CH4 production in different methanogenic sediments

EnvironmentContribution (%)Ref.
Contribution normal
Kichier Lake32–46[39]
Lake Mendota36–46[40]
Lake Washington15–39[41]
Anoxic paddy soil17–31[42]
Contribution low
Colne Pt. Salt marsh8[9]
Knaack Lake4[10]
Lake Constance0[12]
Contribution high
Kuznechika lake97[39]
Octopus Spring mat74–86[15]
Blelham Tarn76–82[43]
Cape Lookout Bight71–80[44]
Kings Lake Bog100[45]
Bunger Hills, Antarctica95–97[46]
Lake Baikal, deep sediment99–100[47]

Hydrogen can be produced in the first fermentative degradation stage (e.g. reaction 1.3), and it is obligatorily formed in the second syntrophic stage of organic matter degradation. The syntrophic stage is sensitive to inhibition by H2 for thermodynamic reasons. The maximum amount of H2 relative to acetate that can be produced from the degradation of carbohydrates is 4 mol H2 plus 2 mol acetate per mol glucose (reaction sum 1-2), i.e. a ratio of H2/acetate of 2:1. Any contribution of homoacetogenesis (reaction 1.5) decreases this ratio. Degradation of chitin (monomer=N-acetylglucosamine) results in one more acetate (ratio of H2/acetate of 4:3) than in the case of the degradation of glucose and thus decreases the possible contribution of H2.

Since 4 H2, but only 1 acetate, are required to produce 1 CH4, the contribution of H2 to methanogenesis during anaerobic degradation of carbohydrates can maximally be 33% of the total CH4 formed. Indeed, this percentage is consistent with data obtained from many studies of methanogenic environments (Table 2). However, lower contributions are also found in some methanogenic environments. Usually, they are easily explained. In marine sediments the low contribution of H2 can be due to the dominance of sulfate reduction for degradation of organic matter, while methanogenesis depends on non-competitive precursors such as trimethylamine [8,9]. In acidic lake sediments the low contribution of H2 may be explained by a larger contribution of homoacetogenesis [10]. In Lake Constance sediment, CH4 production occurs exclusively from acetate. This observation is explained by sulfate reducers which consume H2 in the upper sediment layers. Because of the lack of acetotrophic sulfate reducers [11], acetate is not consumed, and thus it diffuses into deeper layers where it is consumed by methanogens [12]. In methanogenic rice field soil, the contribution of H2 decreases when the temperature is shifted to lower values (30 to 15°C), so that CH4 is then mainly produced from acetate [3,13]. Most probably, homoacetogenesis becomes the main fermentation reaction under this condition.

There are many studies in the literature which report much higher contributions of H2 than the expected 33%. Conceivable explanations for these exceptions include (i) additional sinks of acetate, (ii) additional sources of H2, or (iii) measurements under non-steady-state conditions. Additional sinks of acetate are not uncommon, e.g. in the rumen, acetate is largely absorbed into the blood stream of the host, leaving H2 as the predominant source for methanogenesis [14]. Similar observations were made in microbial mats where acetate is assimilated by the phototrophs [15]. Transient phenomena must occur when H2 and acetate are sequentially produced or utilized. For example, the low amounts of CH4 produced immediately after flooding of paddy soil are mainly due to H2-dependent methanogenesis, since the H2-dependent methanogens apparently become active before the acetotrophic ones [16]. Eventually, however, steady state is reached and H2 then contributes about 30% to CH4 production as theoretically expected (Table 2).

In most cases, however, where H2/CO2-dependent methanogenesis dominates (up to 100%) CH4 production in sediments of lakes, marine bights and peat bogs (Table 2), an explanation for elevated contributions of H2 to methanogenesis is more difficult to find. Additional sources of H2 are as yet undescribed except where there is a geological input of H2, such as in Lake Kivu [17]. In most of the deep sediments and peat bogs, detailed carbon and electron balances for CH4 and its precursors acetate and H2 do not exist and thus it is unclear, for instance, whether the preferential production of CH4 from H2/CO2 is balanced by an equivalent accumulation of acetate or by non-methanogenic consumption of acetate. Clearly, more research is required to explain the high contribution of H2 to CH4 production in these anoxic environments.

3 Control of the environmental H2 concentration

Hydrogen is an intermediate in the methanogenic degradation of organic matter and is rapidly turned over (turnover times of minutes [3,4]). Any change of H2 concentration (C) is caused by a change of either its rate of production (p) or utilization (u): Embedded Image 1 Steady state is reached if p=u. If p>u, H2 concentration will increase, thus also resulting in increased H2 utilization. Assuming Michaelis-Menten kinetics (with umax and Km as parameters) the new and higher steady-state H2 concentration will then be given by: Embedded Image 2

However, this higher H2 steady state will not persist for long, since the H2 utilizers will eventually start to increase their biomass X, e.g. according to the Monod equation: Embedded Image 3 (with μmax=maximum growth rate; Ks=H2 concentration at μmax/2). Since Embedded Image 4 (with vmax=specific maximum H2 utilization rate), this adaptation will return the H2 concentration to the original value that existed before the increase of H2 production. In other words, the H2 steady-state concentration is basically under the control of the H2 utilizers and their kinetic characteristics [18].

The parameters μmax, vmax, Ks and Km are specific for a given microorganism. Thus, it has been proposed that the parameters of competing H2 utilizers should determine which organism finally wins the competition. Indeed, it was shown that sulfate reducers utilize H2 faster than methanogens because of their lower Km[19]. Similarly, it was shown that sulfate reducers have a lower Ks (H2 concentration at half-maximum growth rate μ) than methanogens and thus are able to outgrow the latter [19].

Indeed, it has repeatedly been demonstrated that H2-dependent CH4 production is inhibited in the presence of sulfate [19,20]. This inhibition has usually been explained by the more efficient H2 utilization kinetics in sulfate reducers than in methanogens. However, this model provides no explanation of why the resident methanogens should not continue H2 utilization, albeit at a reduced rate. Complete inhibition can only be achieved after the methanogenic population has been outgrown by the sulfate reducers [19,20]. Thus, methanogenic populations may be replaced by sulfate reducers, iron reducers or nitrate reducers in systems that have been exposed to sulfate, Fe(III) or nitrate for a long time, e.g. aquatic sediments or aquifers. These environments are largely in steady state with respect to concentrations of sulfate, Fe(III) and nitrate and consequently exhibit H2 concentrations that are characteristic for methanogenesis, sulfate reduction, iron reduction, etc. [21]. However, the kinetic model does not provide an explanation for the instantaneous and complete inhibition of H2-dependent CH4 production that has been observed in some methanogenic environments upon addition of sulfate [2224]. An alternative model, one which incorporates a threshold concept, on the other hand, does provide such an explanation [21,2527].

The threshold concept of anaerobic H2 utilization assumes that there is a certain H2 concentration below which utilization is no longer possible because of thermodynamic constraints. Theoretically, the H2 threshold should be given by the conditions at which reactants and products are in thermodynamic equilibrium (ΔG= 0). Thus, the H2 threshold should be defined by the equilibrium constant (K): Embedded Image 5 for example, the H2 threshold partial pressure (pH2) of H2-dependent methanogenesis is given by the equilibrium constant and the partial pressures of CO2 and CH4: Embedded Image 6

Indeed, it has been found that H2 thresholds for various anaerobic H2-utilizing reactions and bacteria decrease with decreasing ΔG° (increasing K) of the H2-utilizing reaction [21,26,28]. In reality, however, the H2 thresholds were found to be slightly higher than those indicated by the equilibrium constant [27,28]. Obviously, H2 utilization stops at a value which still allows for a small negative Gibbs free energy, the critical Gibbs free energy (ΔGc). This critical value is probably explained by the coupling to the energy-generating system of the cell which has a threshold of about 1/3 ATP or approximately −23 kJ mol−1 of the energy-generating reaction [5]. Interestingly, the values of ΔGc increase (less negative) in the order sulfate reducers>methanogens>homoacetogens, indicating that sulfate reducers need more free energy than homoacetogens to allow H2 utilization [28].

Reaction kinetics close to the thermodynamic equilibrium become increasingly reversible. Therefore, they are not well described by Michaelis-Menten kinetics which are based on irreversible reactions. Hoh and Cord-Ruwisch [29] recently modified the Michaelis-Menten model. Their equilibrium model takes into account the relative difference of the actual H2 concentration to that at the thermodynamic equilibrium by amending the Michaelis-Menten equation with the term Γ/K: Embedded Image 7 with Γ=Π (actual concentration of products)/Π (actual concentration of reactants), and K=Π (concentration of products at equilibrium)/Π (concentration of reactants at equilibrium).

Thus, Γ is equivalent to the equilibrium constant, but uses the actual concentrations instead of the concentrations at thermodynamic equilibrium. The authors were able to show that their model fitted experimental data well for both H2-producing reactions (e.g. propionate degradation by syntrophs) and H2-utilizing reactions (e.g. homoacetogenesis and methanogenesis) [29]. An important result of this modeling approach is that the H2 conversion rates at environmentally relevant H2 concentrations are much more sensitive to the thermodynamic conditions in the environment (i.e. Γ/K) than to the kinetic parameters of the microorganisms (i.e., vmax and Km). This response is because the H2 concentrations are much closer to the thermodynamic equilibrium than to the microbial Km values. The model of Hoh and Cord-Ruwisch [29] may be further improved by using Γ/Kc instead of Γ/K, where Kc is the equilibrium constant based on ΔGc rather than ΔG° to account for the fact that H2 utilization (also H2 production) stops short of the thermodynamic equilibrium.

In contrast to the Michaelis-Menten model, the threshold concept easily explains why a H2-utilizing process is rapidly and completely outcompeted when another process with a lower threshold becomes possible. As soon as the H2 concentration decreases below the threshold for a process, activity stops. Measurements in methanogenic environments indicate that in situ H2 concentrations correspond to ΔG values of approximately −23 kJ mol−1 CH4, i.e. equivalent to the energetic threshold of 1/3 ATP, or less (Table 3). Only one study found ΔG values that were much higher than −20 kJ mol−1 CH4[30]. In many cases, H2-dependent methanogenesis obviously operates at its thermodynamic threshold. If we assume that the steady-state concentration of H2 in methanogenic environments is identical to the H2 threshold of the resident methanogenic flora, then we can consider what would happen if a second H2 utilization process becomes active, e.g. H2-dependent sulfate reduction after addition of sulfate. Let the rates of methanogenic and sulfate-reducing H2 utilization be um and us. Then, the steady-state conditions (dC/dt=0) would change from the methanogenic H2 utilization: Embedded Image 8 to the simultaneous utilization by methanogenesis and sulfate reduction: Embedded Image 9 and the steady-state H2 concentration would consequently decrease below the threshold of the methanogens, so that CH4 production would stop. Now, the H2 production would have to be balanced by the sulfate reducers (us alone). Such a balance is only possible if the instantaneous potential of H2-dependent sulfate reduction is equal to or higher than that of H2-dependent methanogenesis (usum). If this is not the case, then p>us, and consequently, H2 concentrations will increase again until H2-dependent methanogenesis resumes and balances H2 production. Then the same cycle would repeat itself. Macroscopically, this chain of events should result in a partial but instantaneous inhibition of methanogenesis without any concomitant decrease of the H2 concentration. Only much later, the population of the sulfate reducers would have eventually grown up. Increasing X of sulfate reducers would result in increasing us (Eqs. (3) and (4)) until us=p, then also resulting in decreasing H2 concentrations until a new steady state characteristic of sulfate reducers would be attained.

View this table:

Gibbs free energies of H2-dependent methanogenesis under steady-state conditions in various environments and at the threshold of H2 consumption in methanogens

Methanogenic system−ΔG (kJ mol−1 CH4)Reference
Sewage sludge28–32[2,48]
Lake Mendota; Knaack Lake27–35[2]
Canal with detritus and leaves8–18[30]
Alder swamp12–19[30]
Littoral sediment, Lake Constance33–39[49]
Profundal sediment, Lake Constance23–34[12]
Upland soils turned methanogenic25–50[50]
Italian rice field soil24–38[13]
Methanobacterium bryantii29–37[27,28]
Other methanogenic archaea29–50[27,28]

One example which may fit this pattern is that of sediment of Lake Mendota where 2 days of incubation were required for a decrease of the H2 concentration although the partial inhibition of H2-dependent methanogenesis was immediate [31]. Methanogenic rice field soil, on the other hand, on sulfate addition shows an instantaneous and complete inhibition of H2-dependent methanogenesis with concomitant decrease of the H2 concentration to values that are thermodynamically no longer permissive for methanogens (Fig. 3). Similar results have also been obtained with Lake Wintergreen sediment [22]. The instantaneous and rapid decrease of H2 concentration indicates that the potential for H2-dependent sulfate reduction must be as high as that of CH4 production.


Effect of sulfate addition on the H2 partial pressure, the Gibbs free energy (ΔG) of H2-dependent methanogenesis and the accumulation of CH4 in slurries of anoxic Italian rice field soil (adapted from [24]).

Plentiful evidence indicates that most H2-dependent methanogenesis operates in microbial aggregates in which H2 producers are juxtaposed to H2 consumers [3,4]. It has been proposed that sulfate reducers may act as syntrophic H2 producers in the absence of sulfate, e.g. during syntrophic degradation of lactate, ethanol or propionate [31]. The syntrophic propionate oxidizers that have so far been isolated are all able to reduce sulfate (e.g. [32]). Addition of sulfate would switch these bacteria from acting as syntrophs to acting as sulfate reducers, stop the production of H2, and starve the juxtaposed methanogens. Also, H2 concentrations would decrease where H2 production by sulfate reducers was one of the main H2 sources. Interestingly, circumstantial evidence indicates that sulfate reducers may indeed be involved in the syntrophic propionate degradation in methanogenic rice field soils [33], where a rapid decrease of H2 concentrations has been observed upon addition of sulfate.

Analogously to addition of sulfate, addition of ferrihydrite or nitrate should also inhibit methanogenesis by competition for H2. Indeed, H2 concentrations decrease and CH4 production is inhibited when ferrihydrite or nitrate are added to methanogenic rice field soil [23,34]. However, the microbes utilizing Fe(III) or nitrate as electron acceptors probably compete not only for H2 and acetate, but also for fermentation products that are precursors for H2 and acetate production and probably also for carbohydrates directly. Therefore, the effects of these electron acceptors on H2 turnover and methanogenesis are not comparable to those of sulfate. In addition, the effects of nitrate on methanogenesis were shown to be due to toxicity of denitrification products (nitrite, NO and/or N2O) to the methanogens in rice field soils [35].

4 Control of the environmental acetate concentration

Another question which is currently unresolved is to what extent acetate turnover follows similar principles as H2 turnover. Most experiments show that addition of sulfate, ferrihydrite or nitrate also inhibited acetate-dependent methanogenesis. As in the case of H2, this inhibition is thought to be due to sulfate, iron and nitrate reducers competing successfully for acetate [20,36]. The threshold concept has occasionally been applied to acetate utilization but less rigorously than in the case of H2. Methanogens have dramatically different thresholds for acetate due to different activation mechanisms. Thus, Methanosarcina species, which activate acetate (input of 1 ATP) with an acetate kinase, have a much higher threshold (0.2–1.2 mM) for acetate than Methanosaeta species (7–70 μM), which activate acetate (input of 2 ATP) with an acetyl-CoA synthetase [37]. If the acetate steady-state concentration observed in methanogenic environments is equivalent to the threshold of the resident methanogenic population, then inhibition of methanogenesis upon addition of sulfate, iron or nitrate does not necessarily require an instantaneous decrease of the acetate concentration (see conjecture above). Indeed, in experiments with anoxic rice field soil, such a decrease has not been observed, although acetate-dependent CH4 production was inhibited [24,34]. The observed inhibition would be consistent with an acetate-utilizing potential of the sulfate, iron and nitrate utilizers that is lower than that of the acetate-utilizing methanogens. More research is needed to confirm this possible conclusion.


I thank H. Scholten for critically reading the manuscript.


View Abstract