Skip to main content

Biomethanation processes: new insights on the effect of a high H2 partial pressure on microbial communities



Biomethanation is a promising solution to upgrade the CH4 content in biogas. This process consists in the injection of H2 into an anaerobic digester, using the capacity of indigenous hydrogenotrophic methanogens for converting the injected H2 and the CO2 generated from the anaerobic digestion process into CH4. However, the injection of H2 could cause process disturbances by impacting the microbial communities of the anaerobic digester. Better understanding on how the indigenous microbial community can adapt to high H2 partial pressures is therefore required.


Seven microbial inocula issued from industrial bioprocesses treating different types of waste were exposed to a high H2 partial pressure in semi-continuous reactors. After 12 days of operation, even though both CH4 and volatile fatty acids (VFA) were produced as end products, one of them was the main product. Acetate was the most abundant VFA, representing up to 94% of the total VFA production. VFA accumulation strongly anti-correlated with CH4 production according to the source of inoculum. Three clusters of inocula were distinguished: (1) inocula leading to CH4 production, (2) inocula leading to the production of methane and VFA in a low proportion, and (3) inocula leading to the accumulation of mostly VFA, mainly acetate. Interestingly, VFA accumulation was highly correlated to a low proportion of archaea in the inocula, a higher amount of homoacetogens than hydrogenotrophic methanogens and, the absence or the very low abundance in members from the Methanosarcinales order. The best methanogenic performances were obtained when hydrogenotrophic methanogens and Methanosarcina sp. co-dominated all along the operation.


New insights on the microbial community response to high H2 partial pressure are provided in this work. H2 injection in semi-continuous reactors showed a significant impact on microbial communities and their associated metabolic patterns. Hydrogenotrophic methanogens, Methanobacterium sp. or Methanoculleus sp. were highly selected in the reactors, but the presence of co-dominant Methanosarcinales related species were required to produce higher amounts of CH4 than VFA.


Anaerobic digestion (AD) is one of the core technologies contributing to the transition from a fossil fuel-based economy to a more renewable energy-based circular economy [1, 2]. This technology aims to valorize organic residues into an energetic biogas and a digestate that could be used as fertilizer or soil amendment [1]. The energy content of the biogas is proportional to the CH4 content. Usually, the AD biogas contains between 40–75% CH4 and 25–60% CO2, besides other components in minor quantities such as H2, N2, NH3, H2S, H2O and others trace organic and inorganic components [3]. Because of the presence of CO2, the biogas has a lower calorific value than natural gas, i.e. 21.5 MJ/Nm3 to 35.8 MJ/Nm3, respectively, and therefore, biogas cannot be directly injected into the natural gas grid [4]. The minimal purity of CH4 in the natural gas grid must be up to 95%, depending on the countries legislations [5]. Thus, the CH4 content in biogas needs to be upgraded prior to be injected into the natural gas grid, used as vehicle fuel or for energy storage.

In order to improve the CH4 content in biogas, several purification technologies can be used, and the valorization of the CO2 through biological technologies has been recently proposed. In a recent review [5], several biogas upgrading methods were described and discussed. The biological upgrading methods are distinguished as chemoautotrophic or photosynthetic processes. The ex-situ and in-situ biomethanation processes correspond to the chemoautotrophic methods relying on the AD process. In both biomethanation processes, H2 is injected into an anaerobic digester in order to upgrade the CH4 content by reducing the CO2. For the ex-situ biomethanation process, an additional bioreactor, physically separated from the anaerobic digester, is fed with H2 and biogas, while for the in-situ biomethanation process, the H2 is directly injected into the anaerobic digester [6]. Both processes are based on the capacity of hydrogenotrophic methanogens to use external H2 as electron donor for the reduction of CO2 into CH4 [7], following the reaction:

$$4 {\text{ H}}_{ 2} + {\text{ CO}}_{ 2} \to {\text{ CH}}_{ 4} + {\text{ 2 H}}_{ 2} {\text{O }}\Delta {\text{G}}^\circ = \, - 1 30. 7 {\text{ KJ}}/{\text{mol}}$$

H2 can have several origins, but to keep the process environment friendly, a renewable energy source should be used [8]. In this way, H2 can be generated by water hydrolysis using the energy surplus of wind and solar power plants. However, these renewable power plants produce energy in a fluctuant way, which could lead to imbalance between energy production and the energy demand. Even though H2 can also be stored or used as vehicle fuel, the current technologies are today too expensive and technically challenging [9]. Therefore, an alternative is to transform H2 into CH4, for which storage and transportation are cheaper and, CH4 can also be used as vehicle fuel or injected into the gas grid [10, 11]. This concept is named Power-to-Gas (PtG) and fulfils the requirement of linking electrical power and gas-grid networks in an environmentally friendly way.

In this context, both, ex-situ and in-situ biomethanation are suitable processes to be applied to purify biogas produced in AD. However, some challenges need to be addressed prior to develop this technology at industrial scale. One of the limitations concerns the increase of the H2 partial pressure within the digester that can alter the AD metabolic equilibrium [7]. The AD process is composed of 4 steps (hydrolysis, acidogenesis, acetogenesis and methanogenesis) which are carried on by different microbial groups that perform these reactions in a very coordinated way. H2 is an intermediate during the AD, whose partial pressure, mainly affects the acetogenesis step [12]. During acetogenesis, the volatile fatty acids (VFA) that were formed during the hydrolysis step and the acidogenesis step, are transformed into acetate, H2 and CO2 [13]. This step is carried out by syntrophic microorganisms, which are thermodynamically constrained by the H2 partial pressure, which it must be kept under 10−4 atm to allow VFA degradation and methanogenesis [14]. A higher H2 partial pressure leads to VFA accumulation in the media, which further inhibits the methanogenic archaeal populations. Nonetheless, the effect of the H2 injection over the microbial community is still not very well understood [15]. Mulat et al. [16] and Wahid et al. [17] have reported the inhibition of methane production due to VFA accumulation, caused by H2 injection. Moreover, Cazier et al. [18] reported an inhibitory effect due to a high H2 partial pressure in the methanogens community, even though the VFA accumulation and pH were not at inhibitory levels. In contrast, in the work of Bassani et al. [19], no hydrolysis inhibition or VFA accumulation was reported after an H2 addition and pH increase. Agnesseens et al. [20] detected a transient accumulation of acetate, which was overcome after several H2 injections by a microbial adaptation to high H2 partial pressure. Therefore, each microbial community configuration can adapt differently to the H2 partial pressure. A better understanding of the microbial community characteristics that can improve CH4 production under ex-situ and in-situ biomethanation conditions is a key step prior to process optimization [21,22,23]. This work aims to provide new insights on the influence of the inoculum origin on biomethanation performances, and more particularly on the response of indigenous microbial communities facing high H2 partial pressures.

Results and discussion

Metabolic patterns production during ex-situ and in-situ biomethanation

Seven microbial inocula coming from different configurations of industrial digestion facilities, and treating diverse types of substrates were inoculated into semi-continuously fed reactors (Table 1). These reactors were fed with (i) H2 (ex-situ biomethanation reactors) or (ii) glucose and H2 (in-situ biomethanation reactors) in order to compare the effect of H2 on the indigenous microbial community during ex situ and in situ biomethanation. Incubations without external CO2 addition were chosen in order to simulate a CO2-limiting environment.

Table 1 Origin of the tested inocula and operational characteristics of the digester of origin

The consumption of H2 started within 24 h after the first H2 injection, as consistently reported by Kern et al. [21], Agnesseens et al. [24] and Wahid et al. [17]. Concomitantly, CH4 production started 24 h after the first H2 injection, and declined along with CO2 detection in the head-space (Additional file 1). The sequential injection of H2 led to CO2 depletion in the head-space of most the reactors between days 4 and 6. At the end of the experiment, a pH increase was observed due to CO2 depletion as reported by Luo et al. [9]. Although the ideal pH for methanogens is close to the neutrality, the optimal pH remains variable according to the individual species [7].

The COD mass balance analysis shows that no major metabolites were omitted in this study (Additional file 2), if considering a reasonable variability error of 10% and that at least 10–12% of the electrons contributed to the production of biomass [25,26,27,28]. Nonetheless, the carbon conversion was overestimated in some reactors (10–34%), particularly in the ex-situ biomethanation reactors, being attributed to an endogenous methane production [29]. Although, the inocula were stored at ambient temperature (\(\cong\) 20 °C) 1 week before use, it was previously reported that an endogenous methanogenic activity was still detected even 10 days after incubation in biochemical methane potential (BMP) tests conditions [30]. In addition, Luo and Angelidaki [8] have attributed the excess of CH4 detected in their ex-situ biomethanation reactors to the degradation of the organic matter in the inocula even 10 days after the operation of their reactors have started. Besides, methanogens are considered to have a lower biomass yield than acidogens [27, 31].

The average metabolite production (i.e. VFA and methane) of each operating condition is shown in Fig. 1. In all the reactors, after the first H2 injection, the H2 addition was only made once the total pressure in the reactors was less than 1.2 bar, to maintain an H2 partial pressure approximately at 1.0 bar in the head-space (Additional file 1). As some reactors consumed a higher quantity of H2, they produced a higher amount of metabolites than other reactors. When the total pressure of the reactors was over 1.2 bar, H2 addition was stopped. Liu et al. [32] reported that in batch reactors fed with only H2/CO2, in a 4:1 proportion, at a H2 partial pressure of 0.96 bar, 60% of the H2 was used by the hydrogenotrophic methanogens and the other 40% was converted to methane via the association between homoacetogens and acetotrophic methanogens.

Fig. 1

Metabolite production of the ex-situ biomethanation reactors (fed with H2 only) and in-situ biomethanation reactors (fed with glucose and H2) inoculated with the different inocula and cluster analysis gathering the reactors according to their metabolite production

A partitioning clustering analysis using the k-means algorithm was performed based on the metabolite production patterns of the reactors (Fig. 1). Three clusters were distinguished according to the source of inoculum and their metabolite production patterns in biomethanation. Cluster 1 corresponds to the reactors (in-situ and ex-situ processes) inoculated with cattle manure leachate from a cattle manure dry anaerobic digestion facility (inoculum BM). These reactors produced almost only methane (94% in average of the total produced metabolites) all along the operation. Cluster 2 is composed of the reactors inoculated with a granular sludge coming from an up-flow anaerobic sludge blanket (UASB) reactor treating paper mill waste (inoculum GS) and the reactors inoculated with a sludge from an anaerobic digester treating aerobic sludge (inoculum AnS). Even though these reactors produced mostly methane (\(\cong\) 81% and \(\cong\) 78%, respectively, in average from the total produced metabolites), they have also accumulated VFA in their bulk phase, mostly acetate and in lower proportions propionate, butyrate, isobutyrate, and isovalerate. Cluster 3 represents the reactors that accumulated mostly VFA. These reactors were inoculated with the other four inocula used for this experiment (Table 1). One set of reactors was inoculated with the liquid fraction of the first-step reactor from a two-step anaerobic digestion process treating farm waste (principally poultry slurry, food waste and some green waste) (inoculum MFW1), while another set of reactors was inoculated with the liquid fraction of the second-step reactor of the same process (inoculum MFW2). The third set of reactors belonging to cluster 3 was inoculated with the liquid fraction of an anaerobic digester treating food waste (inoculum FW), while the fourth set of reactors belonging to this group was inoculated with an aerobic sludge from a sewage waste water treatment plant (WWTP) (inoculum AeS). A statistically significant difference was confirmed between the clusters regarding their methane production after performing a Kruskal–Wallis test, followed by a Wilcoxon’s test, in which the obtained p-values were adjusted with the Bonferroni’s correction method. However, regarding the VFA production only cluster 3 was statistically different from the others (Additional file 3). For the reactors inoculated with MFW1, MFW2 and FW, acetate represented the main accumulated VFA (93%, 94% and 95% in average from the total produced VFA for the reactors inoculated with MFW1, MFW2 and FW, respectively), while the rest of VFA was propionate and traces of butyrate, isobutyrate and isovalerate. Such acetate accumulation is consistent with previous biomethanation performances reported in the literature [9, 24, 33]. On the contrary, the reactors inoculated with AeS produced almost the same proportion of acetate (46%) and butyrate (39%) in average from the total produced VFA. A higher proportion of propionate (10%) was also observed, in comparison with the reactors inoculated with MFW1 (3%), MFW2 (4%) and FW (2%). Some traces of isobutyrate and isovalerate were detected in the reactors inoculated with AeS and fed with glucose and H2. In the reactors inoculated with AeS, H2 was hardly consumed and the accumulated VFA were likely produced from the organic substrate consumption, leading to the production of other VFA different from acetate. Besides, as these reactors had the lowest H2 consumption, H2 partial pressure was kept constantly high, likely inhibiting the VFA consumption by the syntrophic microorganisms. Regarding the inoculum AeS, another cluster inside cluster 3 could be detected regarding the accumulation of butyrate and a lower quantity of CH4 produced. However, when performing a cluster analysis using the k-means algorithm with k = 4, cluster 3 was divided into two groups regarding the amount of total VFA produced: one group containing the ex-situ biomethanation reactors inoculated with MFW1, MFW2, FW and the ex situ and in situ biomethanation reactors inoculated with AeS. While, the other group was formed by the in-situ biomethanation reactors inoculated with MFW1, MFW2 and FW. Nevertheless, the distribution of the groups is more likely related to the amount of COD received by the reactors than with the actual distribution of the metabolite patterns.

For the reactors belonging to cluster 3, a lower quantity of H2 was consumed, in comparison with the reactors that produced mostly CH4 (clusters 1 and 2). It is likely than in these reactors (cluster 3), when CO2 was available, some of it and some of the added H2 were transformed into CH4 or acetate by the hydrogenotrophic methanogens or homoacetogens, respectively. As a consequence of the CO2 depletion, H2 started to accumulate in the head-space, likely leading to the accumulation of the VFA coming from the degradation of the remaining organic matter in the inocula, for the ex-situ biomethanation reactors or from glucose degradation in the case of the in-situ biomethanation reactors. According to the observations made by Cazier et al. [18], CO2 depletion led to H2 and VFA accumulation inducing subsequently the inhibition of syntrophic interactions, and impeding more CO2 formation and higher inhibition. In some of the reactors from cluster 3 (inoculated with MFW1, MFW2 and FW) due to H2 addition and concomitant CO2 depletion, the final pH was over 8.5, which may have contributed to inhibit the methanogens, as already reported in other ex-situ and in-situ biomethanation works [8, 17]. The high accumulation of acetate in these reactors was likely due to the prevalence of homoacetogens over hydrogenotrophic methanogens as more acetate than methane was produced. Agneessens et al. [20] reported that homoacetogens could be more important than hydrogenotrophic methanogens in a low-CO2 environment due to their higher resistance to high pH.

The in-situ biomethanation reactors have received a higher amount of substrates (as H2 and glucose were provided) than the ex-situ biomethanation reactors (that received only H2), likely contributing to a higher production of metabolites as observed in Fig. 1, although the metabolite pattern distribution between in-situ and ex-situ biomethanation reactors was similar.

In the reactors belonging to cluster 1 and 2 (reactors inoculated with BM, GS and AnS) that mostly produced CH4, lower accumulation of VFA and better H2 assimilation were observed. Despite H2 consumption have decreased after CO2 depletion in the head-space, CH4 production positively correlated with H2 consumption (r2 = 0.8147) (Fig. 2), showing that CH4 was mostly produced by hydrogenotrophic methanogenesis. In contrast, VFA production did not correlate with H2 consumption (data not shown), probably because some of the VFA were consumed to form CH4 and, in some reactors, acetate was not exclusively issued from homoacetogenesis. However, acetate production from the reactors in cluster 3, and more particularly the ones inoculated with MFW1, MFW2, FW did positively correlate with H2 consumption (r2 = 0.7615), indicating that at least some of the acetate production came from homoacetogenesis (Additional file 4). The co-occurrence of methanogenesis and homoacetogenesis was already reported by Lay et al [34].

Fig. 2

CH4 production vs H2 consumption

Regarding these results, it is likely that the inocula microbial composition influences the capability of the inocula to use H2 in order to produce CH4. To our knowledge only the works of Luo and Angelidaki [8] and Bassani et al. [19] have evaluated the influence of the inocula origin during ex-situ and in-situ biomethanation, respectively. Both works have compared the biomethanation potential of a mesophilic inocula under an incubation temperature of 35 °C, to that of a thermophilic inocula under an incubation temperature of 55 °C. These works have reached to the conclusion that a thermophilic operation was more efficient than a mesophilic operation due to their higher methane productivity. However, the growth rate of microorganisms in anaerobic digestion increases linearly between 20 and 60 °C, which improves the efficiency of the process [35, 36]. Therefore, it cannot be determined if the higher methane productivities of the thermophilic inocula were due to the microbial composition or the incubation temperature of the reactors.

The composition of the indigenous microbial community determines the CH4 production and the VFA accumulation after H2 addition

The microbial communities of all the reactors were analysed to elucidate the ecological response during ex-situ and in-situ biomethanation. Two samples of each reactor were analysed: the initial inoculum and the final sample for each set of reactors. The effects of H2 injection on bacterial and archaeal diversities and dynamics were assessed by 16S rRNA gene sequencing and qPCR analysis.

Bacterial community

3872 OTUs were identified and grouped in 98 different classes. Among them, only 17 classes showed a relative abundance in the community higher than 1%. In Fig. 3, the Shannon diversity index of each bacterial community is shown. Most of the inocula have shown a statistically significant decrease of the bacterial community diversity after the H2 addition, due to the selective conditions given by the added H2. Treu et al. [22] reported a similar decrease in the Shannon entropy index after H2 addition in a thermophilic continuous stirred tank reactor (CSTR). Such decrease in diversity of the microbial community was attributed to the high selectivity of H2, towards a more specialized community able to use or resist to the added H2. Bassani et al. [19] also reported a decrease in alpha diversity of the microbial community when working with thermophilic and mesophilic two-stage biomethanation CSTR treating cattle manure. In the reactors inoculated with farm waste and manure (MFW1 and BM) no difference in the Shannon entropy index was observed between the community of the initial inocula and the communities at the end of the operation, whatever the condition.

Fig. 3

Shannon diversity index calculated for the bacterial community of each inocula. t0 indicates the index of the initial inocula. A t-test was performed between the inocula sample and the final sample of the ex-situ or the in-situ biomethanation reactors: statistically significant differences (p > 0.05) are shown with an asterisk. NS stands for not statistically significant

In Fig. 4, the relative abundance of the bacterial community of all reactors is shown. At the beginning of the experiment, the Clostridia class was dominant in almost all of the inocula, except in inocula AeS, where the most abundant bacterial classes were related to Betaproteobacteria and Sphingobacteria. At the end of operation, Clostridia class was the most abundant class in all the ex-situ biomethanation reactors and in some in-situ ones, except in the reactors belonging to cluster 2, where Bacilli class outcompeted the Clostridia class. Bacilli was the second most abundant class in all in-situ biomethanation reactors, with exception of the reactors belonging to cluster 1 where Bacteroidia class was the second most abundant. Clostridia, Bacilli and Bacteroidia classes are composed of bacteria which are able to hydrolyse polysaccharides, oligosaccharides, rest of plants and manure in complex environments [37]. Some members of Clostridia can also perform homoacetogenesis and syntrophic interactions with hydrogenotrophic methanogens [38]. Although Clostridia was positively selected in all conditions, in the reactors belonging to cluster 3 where VFA (mostly acetate) accumulated, no well-known homoacetogens were specifically identified among the OTU. This is in contradiction with the qPCR analysis results, which showed an increase in copies of the genes coding for the formyl-tetrahydrofolate synthase (FTHFS) a key enzyme of the Wood–Ljungdahl pathway involved in homoacetogenesis [39]. In Fig. 5 is shown the augmentation or the decrease of the number of copies of the genes coding for FTHFS. A decrease in the FTHFS copy number in all reactors inoculated with AnS (cluster 1) and MFW1 (cluster 3) was detected, although, these sets of reactors had a very different metabolite production profile. Likely the faster depletion of CO2 in the reactors inoculated with MFW1 favoured the acetate accumulation [20, 24]. Meanwhile, in the AnS reactors the methane production pathway was favoured because the hydrogenotrophic methanogens have outcompeted the homoacetogens in these reactors. However, the increase in the number of copies of this gene was detected in most of the inocula suggesting a significant presence of homoacetogens (Fig. 5). Homacetogenesis phenotype is widely distributed in the phylogeny, with at least 23 different genera within the Bacteria domain and more particularly within the Firmicutes phylum that were identified as containing homoacetogenic microorganisms [39]. Since the capacity for carrying a homoacetogenic activity is not always tested as a phylogenetic trait, it is highly probable that already known bacteria may also carry this attribute, but not yet identified [40].

Fig. 4

Relative abundance of Bacteria classes. The sum of all the OTU with less than one percent abundance is represented under “Others

Fig. 5

Difference between the initial number of copies of the gene FTHFS per mL of sample (copies FTHFS_t0) and the final number of copies of the FTHFS gene per mL of sample (copies FTHFS_tf) for each reactor

At the beginning of the operation, the most abundant OTU in the reactors inoculated with BM (cluster 1), MFW1 and MFW2 (cluster 3), was affiliated to a cluster named MBA03 at the order level. Although, a BLASTn search against the NCBI database using the 16S rRNA sequence database was performed, no further identification was possible. MBA03 belongs to the Clostridia class and was previously found in AD reactors treating cattle manure, silage waste and food waste [41,42,43]. At the end of the operation, this OTU relative abundance increased in all reactors inoculated with BM. In the ex-situ biomethanation reactors inoculated with MFW1, MFW2 and FW, MBA03 relative abundance was quite stable, it has varied from 38 to 44% in the reactors inoculated with MFW1, from 34% to 33% in the reactors inoculated with MFW2 and from 10 to 13% in the reactors inoculated with FW. In contrast, in the in-situ biomethanation reactors (inoculated with MFW1 and MFW2), the abundance of this OTU decreased from 38 to 14% and from 34 to 16%, respectively. Therefore, this microorganism was likely outcompeted by other members of the Bacilli or Clostridia classes for glucose.

Some of the OTU belonging to the most abundant groups in the final microbial community were affiliated at a species level, by performing a BLASTn search against the NCBI database using the 16S rRNA sequence database. The identified OTUs related to Bacilli were Amphibacillus xylanus [44], Paenibacillus ihumii [45] and Vagococcus acidifermentas [46]. Three of the major OTUs belonging to Clostridia were: (i) Natronincola peptidivorans [47], (ii) Proteiniborus ethanoligenes [48], (iii) Clostridium isatidis [49]. All these bacterial species are heterotrophs, mostly peptide- or carbohydrate-consumers, which are able to grow at pH ranging between approximately 6 and 10. Even though the pH of the reactors was measured at the beginning and at the end of the experiment, the pH increase likely started with the CO2 depletion as reported elsewhere [17]. Such pH increase probably favoured these species among others, due to their high tolerance to alkaline pH as they could grow even at pH equal to 10.

Archaeal community

461 OTUs belonging to the Archaea domain were detected in the sequencing analysis. They were grouped in 20 genera, from which 13 had an abundance higher than 1% in the archaeal community. The Shannon entropy index of the archaeal community in all reactors were smaller than the ones calculated for the bacterial communities, meaning a less diverse community dominated by few distinct OTUs (Fig. 6). A decrease in the diversity was also observed in the archaeal communities in almost all reactors, except in the reactors inoculated with GS and FW where a slightly increase in the diversity of their final microbial communities was noticed. Although, these increase was no statistically significant with regard to the diversity of the inocula (t test, p > 0.05).

Fig. 6

Shannon diversity index calculated for the archaeal microbial community of each inocula. t0 indicates the index of the initial inocula. The asterisks indicate that the samples are statistically different (t test, p < 0.05), while “NS” indicates that not a statistically significant difference exist between samples (t test, p > 0.05)

The relative abundance of the archaeal community in all reactors is shown in Fig. 7. In both types of reactors, ex-situ or in-situ ones, the proportion of hydrogenotrophic methanogens increased by the end of the operation, suggesting that a shift in the microbial communities towards a more specialized H2-utilizing ones occurred as shown by Agneessens et al. [24]. Meanwhile, the presence of Methanosarcinales was detected all along the operation in most of the reactors with a relative abundance greater than 1%, except in the reactors inoculated with FW. The presence and the increase in abundance of Methanosarcinales during biomethanation has already been reported [17, 20, 21].

Fig. 7

Relative abundance of Archaea genera. Only the genus representing more the one percent in at least one of the samples is shown. The term “Others” represent the sum of all OTU, which abundances were under 1%

Methanobacterium sp. was found to dominate in all the reactors belonging to cluster 2 and in most reactors belonging to clusters 3. Dominance of this genus in biomethanation processes has already been reported [16, 17, 33]. Other genera related to hydrogenotrophic methanogens, in particularly Methanobrevibacter sp. and Methanosphaera sp. were also found in greater proportion in the ex-situ biomethanation reactors belonging to clusters 3, while in the in-situ biomethanation reactors their proportion diminished with regard to the inocula. Meanwhile, in the reactors belonging to cluster 2, Methanobacterium sp. outcompeted the other genera of hydrogenotrophic methanogens (inocula AnS) or were not even present in the initial inocula (inocula GS).

The reactors inoculated with FW were dominated by the genus Methanoculleus sp. all along the operation. This genus is composed of hydrogenotrophic methanogens and was also found to dominate the archaeal community during biomethanation [19, 21, 23]. In the reactors belonging to cluster 1, both Methanobacterium sp. and Methanoculleus sp. increased in proportion in the microbial community by the end of the operation. However, these reactors were dominated by Methanosarcina sp. By the end of the operation, the genera Methanosaeta sp. and Methanosarcina sp. were more abundant or were stable in the reactors belonging to cluster 1 and 2, which produced mostly CH4. The presence of these genera probably contributed to avoid acetate accumulation and favoured CH4 production.

Methanosarcina sp. dominated all along the operation in the all the reactors belonging to cluster 1, although its abundance was reported to decrease because of H2 partial pressure inhibition [24]. Interestingly, Methanosarcina sp. are able to switch their metabolism from acetate-consuming to H2-consuming pathways as a response to the H2 partial pressure [50]. Therefore, a high H2 availability could be favourable to select hydrogenotrophic methanogens with a lower H2 affinity such as Methanosarcina sp. [51]. The reactors inoculated with MFW1 and MFW2 (belonging to cluster 3) were dominated by Methanosarcina sp. at t0, although acetate accumulated and the Methanosarcina sp. population decreased by the end of the experiment. It is not completely clear why Methanosarcina sp. was negatively affected in these reactors. It is possible that the faster CO2 depletion due to the H2 addition in the reactors inoculated with MFW1 and MFW2 (clusters 3) than in the reactors inoculated with BM (cluster 1) have led to the inhibition of Methanosarcina sp. Agneessens et al. [20] and Mulat et al. [16], have attributed such inhibition to low CO2 concentration and high H2 partial pressure. The loss of Methanosarcina sp. was probably the cause of acetate accumulation and low CH4 production.

The Archaea/Bacteria ratio was calculated for all reactors. The reactors belonging to the clusters 1 and 2, that grouped the reactors that have produced mostly methane (inoculated with AnS, GS, BM) had a significantly higher proportion in Archaea at t0 (1:21, 1:2.5 and 1:29, respectively), while the other reactors had ratio approximately lower than 1:200 (Table 2). Moreover, in inocula BM (cluster 1) and GS (cluster 2) a higher amount of hydrogenotrophic methanogens (one order of magnitude higher) than homoacetogens was detected (Table 2). Hence, the hydrogenotrophic methanogens were the major hydrogen-utilizing microorganisms in these inocula. Lay et al. [34] reported that the amount of homoacetogens and hydrogenotrophic methanogens in batch co-cultures had an impact in the H2 consumption and the favoured pathway to form CH4. Such high amount of hydrogenotrophic methanogens might have contributed, at a local scale, to the rapid consumption of H2. Moreover, if considering the fact that H2 has a low solubility in the liquid media, this could have helped to the decreasing distribution of the solubilized H2 from the gas–liquid interphase to the liquid media and therefore, leaving zones in the liquid media where the H2 partial pressure was low enough to allow syntrophic interactions [20]. Through these syntrophic microorganisms removing efficiently the VFA, VFA accumulation was avoided and CH4 production was enhanced. Nonetheless, the inoculum AnS had a higher amount of homoacetogens than hydrogenotrophic methanogens, although in the same magnitude order. Liu et al. [32] reported that homoacetogens had a lower H2 conversion rate than hydrogenotrophic methanogens, meaning that, under same proportions, hydrogenotrophic methanogens will contribute in higher amount to H2 consumption than homoacetogens. Since inocula AnS showed the highest amount of Methanosaeta sp. among all the inocula, some of the produced methane was likely produced though the homoacetogenic pathway followed by acetotrophic methanogenesis.

Table 2 The initial concentration of Bacteria and Archaea (number of copies of the 16S RNA from Bacteria or Archaea gene, respectively, per mL of sample), the initial concentration of homoacetogens (number of copies of the FTHFS gene/mL sample), the initial number of hydrogenotrophic methanogens (number of copies of the 16S archaea gene/mL sample*relative abundance of hydrogenotrophic methanogens in the inocula) and the initial number of acetotrophic methanogens (number of copies of the 16S archaea gene/mL sample*relative abundance of acetotrophic methanogens in the inocula) are shown. As well as, the calculated ratio between the amount of Archaea, respectively, to that of the Bacteria (ratio A:B), the amount of hydrogenotrophic methanogens in relation to the amount of homoacetogens (ratio HM:HA) and to the quantity of acetotrophic methanogens (ratio HM:AM). The sum of the relative abundance of Methanobacterium sp., Methanosarcina sp., Methanoculleus sp., Methanobrevibacter sp., Methanosphaera sp., Methanothermobacter sp. and Methanospirillum sp. was used as total relative abundance of hydrogenotrophic methanogens in the inocula, while the sum of the relative abundance of Methanosaeta sp. and Methanosarcina sp. was used as total relative abundance of acetotrophic methanogens

Overall, the final biomethanation performances of the reactors are the outcome of the high selectivity of the H2 addition in the initial inocula [22]. Not only the presence of these initial characteristics but also the persistence of them can lead to the production of CH4 or the accumulation of VFA, as the possibility of the microbial communities to cope with perturbations is related to the presence of specific tolerant species [52]. A Pearson correlation analysis (Fig. 8) was performed in order to elucidate which features of the inocula could have impacted the final biomethanation performances. To perform the correlation analysis, several features of the inocula were selected. The initial and final amounts of archaea in the inocula were taken into account for this analysis because the reactors grouped in clusters 1 and 2 were inoculated with the inocula that presented higher amounts of archaea in their community. The amount of hydrogenotrophic methanogens and homoacetogens in the inocula likely influenced the predominant metabolite pathway in the microbial community as they are H2-consumer microorganisms and their amount in the microbial communities was reported to determine which microorganisms is the principal H2 consumer [34]. Hence, their amount and ratio in the inocula and in the final microbial communities of the reactors were taken into account for the analysis. Initial quantity and persistence of the acetotrophic methanogens and the ratio of acetotrophic methanogens to hydrogenotrophic methanogens were also considered. The initial and final amounts of Methanobacterium sp., Methanosarcina sp., Methanoculleus sp. and Clostridia were also considered as they were the predominant microorganisms in the microbial communities of the reactors which are able to consume H2 to produce CH4 or VFA.

Fig. 8

Pearson correlation matrix between the reactors final performances and the microbial communities of the inocula and the final community of the reactors. The asterisk shows that the correlation has a statistical significance (p < 0.05). References: H2, consumed H2 (in gCOD); CH4, produced CH4 (in gCOD); VFA, produced VFA (in gCOD)

From the correlation analysis represented in Fig. 8, the amount of archaea in the inocula did not correlated (negatively or positively) to CH4 production or VFA accumulation, although, the persistence of the archaea in the microbial community led to a higher methane production. The initial or final amount of hydrogenotrophic methanogens and homoacetogens did not correlate with H2 consumption, nor with CH4 production or VFA accumulation, respectively. Interestingly, the ratio homoacetogens to hydrogenotrophic methanogens (HA/HM) in the initial inocula anti-correlated with the CH4 production and H2 consumption. Meanwhile, the final HA/HM ratio positively correlated to VFA accumulation. Therefore, a higher amount of hydrogenotrophic methanogens than homoacetogens probably favoured H2 consumption and methane production. While a higher amount of homoacetogens vs hydrogenotrophic methanogens was correlated with the accumulation of VFA and with the amount of Clostridia at the beginning and at the end of the operation. Likely, the homoacetogens present in the reactors belong to the class Clostridia. In addition, the amount of acetotrophic methanogens in the initial inocula as well as at the end of operation correlated with methane production and H2 consumption and anti-correlated with VFA accumulation, confirming that the persistence of this group of microorganisms is crucial to avoid VFA accumulation.

Even though Methanobacterium sp. was one of the most abundant methanogen in all reactors, the increase in its abundance did not correlate to CH4 production or H2 consumption. In fact, the abundance of Methanobacterium sp. increased in all reactors whatever the CH4 production. In the reactors from clusters 1 and 2 that have produced mainly CH4, Methanobacterium sp. was the most (cluster 2) or 2nd most abundant methanogen (in the reactors from cluster 1), although, Methanosaeta sp. or Methanosarcina sp. were also present, probably contributing to the CH4 production, as well. Meanwhile, in the rest of the reactors dominated by Methanobacterium sp., an accumulation of VFA was detected. Since CO2 was poorly available, this was probably limiting the CH4 production by Methanobacterium sp. It was previously reported that the CH4 production rate of Methanobacterium sp. was severely affected by the concentration of CO2 in the head-space [24, 53]. The final amount of Methanoculleus sp. also correlated with the CH4 production likely because, its abundance was highly increased in the inocula of cluster 1 (especially in the ex-situ biomethanation reactors) that mostly produced methane. The reactors grouped in cluster 1 were dominated all along the operation by Methanosarcina sp., which was positively correlated, to the CH4 production and the H2 consumption. Methanosarcina sp. is able to stabilize the AD process in adverse conditions such as high H2 partial pressure and shift its metabolisms to H2 consumption according to the H2 partial pressure in the media [50, 54].


This work has provided new insights on the microbial community response during biomethanation, which are relevant for a practical ex-situ or in-situ biomethanation operation, since the microbial composition of the 7 initial inocula and its maintenance determined the metabolic pathways observed during these processes. According to the results, three main characteristics in the composition of the inocula resulted in better CH4 production in biomethanation: (i) an increase or a stability of the Archaea proportion from the initial inocula, (ii) a high hydrogenotrophic methanogens/homoacetogens ratio in the inoculum and the persistence or increase of it, and (iii) the initial presence of Methanosarcinales, especially Methanosarcina sp. and their persistence all along the operation. These new insights would contribute to a more efficient operation of biomethanation reactors.


All experiments and physical–chemical analysis were performed at the Bio2E platform [55]

Inoculum sources

In order to assess how different microbial community structures would respond to H2 injections, seven inocula from different origins (Table 1) were exposed to H2 injections during 12 days.

Before measuring the concentration of VSS in the UASB inoculum, the granules were broken with immersion blender version WSB33E/K from Waring Commercial® at high speed (20.500 ± 500 rpm) for 10 min. The measurements of TSS and VSS of the inocula were made using the APHA (American Public Health Association) standard methods [56].

Operational conditions

Schott flasks of 500 mL with a working volume of 200 mL were used and sealed with rubber stopper. The inoculum concentration was 5 gVSS/L. All the inocula were incubated at the same temperature (35°) and agitation speed (370 rpm). The mineral medium was composed of: NH4Cl 859 mg/L, KH2PO4 323 mg/L, hexa-hydrated MgCl2 194 g/L, di-hydrated CaCl2 97 mg/L, and was supplemented with an oligo-element solution as described in Cazier et al. [18]. Buffer phosphate was also added at a final concentration of 50 mM, at pH 7.5. Two sets of reactors were prepared: (i) feed with H2 only (ex-situ biomethanation reactors), and (ii) feed with H2 and glucose (in-situ biomethanation reactors). Glucose control was not performed, as in preliminary test carried on in the same conditions, we have detected a pH difference in the media of the reactors fed only with glucose and the ones fed with glucose and H2 or H2 only (Additional file 5). pH highly influences the amount and the type of produced VFA [57,58,59,60], so we could not determine if the difference in the produced VFA would be due to pH effect or to H2 injection.

H2 injections were carried out with no previous enrichment in hydrogenotrophic methanogens to evaluate the response of the indigenous microbial community to high H2 partial pressure. H2 injection was made manually by applying a pressure of 1.2 bars. More precisely, H2 was injected once a day, only if the total pressure of the vial was lower than 1.2 bars, in order to keep the H2 partial pressure around 1 bar, similar to the experiment performed by Liu et al., [32]. The separate injection once a day was done in order to simulate an intermittent H2 addition, as if the provided H2 was issued from the energy surplus of a wind or solar power plant. Glucose was added every 3 days at final concentration of 0.75 g/L. Thus, the operation was carried on in a semi-continuous regime for 12 days. The experiment duration was based in a previous experiment (Additional file 5), in which after 12 days of operation, the pH has reached almost the value of 9 in some of the reactors and the methane production was nearly stopped. The experiments were performed in duplicates.

Physical–chemical analysis

Gas pressure and composition were measured twice a day, before and after H2 feeding. Liquid samples were taken every day and centrifuged (12,100g, 15 min). The supernatant was used to analyse the Volatile Fatty Acid (VFA) concentration while the pellet was kept at − 20 °C for further molecular biology analysis. Gas pressure was manually measured with a manometer Keller LEO 2 (KELLER AG, Winterthur, Switzerland), and gas composition was analysed by gas chromatography using GC Perkin Elmer model Clarus 580, with thermal conductivity detector as described elsewhere by Moscoviz et al. [61]. VFA were analysed by gas chromatography (Perkin Elmer, Clarus 580) coupled with a flame ionization detector as described in Cazier et al. [18]. Glucose concentration of the sample was analysed by YSI 2900D biochemistry analyser, with the corresponding membrane and buffer, according to manufacturer instructions (YSI Inc. Yellow Springs, USA).

Microbial community analyses

To analyse microbial community composition, Illumina Miseq sequencing and qPCR methods were used. From each reactor, two samples were analysed: the initial and the last-day-of-operation samples. DNA extraction was made with a FastDNA™ SPIN kit in accordance with manufacturer’s instructions (MP biomedicals, LCC, California, USA).

Bacterial and archaeal community sequencing

The bacteria members were identified by the amplification of the V3–V4 region of the 16S rRNA gene as reported by Carmona et al. [62]. For the identification of archaea members, degenerated primers designed by our laboratory amplifying the V4–V5 region of the 16S rRNA gene were used: 5′-CAGMGCCGCGGKAA-3′ (F504-519) and 5′-CCCGCCWATTCCTTTAAGT-3′ (R910-928). Adapters and bar codes for Miseq sequencing were already included in the primer sets. The PCR mix contained MTP™ Taq DNA Polymerase (Sigma-Aldrich, Inc., Merck, Germany) (0.05 u/µL) with its enzyme buffer, forward and reverse primers (0.5 mM), dNTP (0.2 mM), sample DNA (0.04 to 0.2 ng/µL) and water with a 60µL final volume. The PCR amplification program was the following: 35 cycles of denaturation (95 °C, 1 min), annealing (set at 59 °C, 1 min) and elongation (72 °C, 1 min). At the end of 35 amplification cycles, a final extension step was carried out for 10 min at 72 °C. PCR reactions were carried on in a Mastercycler® thermal cycler (Eppendorf, Hamburg, Germany). All PCR amplifications were verified by 2100 Bioanalyzer (Agilent, Santa Clara, California, USA). The sequencing reaction was carried on in Illumina Miseq sequencer using a 2 × 300 pb paired-end run at the GenoToul platform, Toulouse, France ( Reads cleaning, assembly and quality checking was performed in Mothur version 1.39.5. SILVA release 128 was used for alignment and as taxonomic outline [63]. The generated sequencing datasets are registered in the Sequence Read Archive ( under the BioProject accession number PRJNA624130, with SRA accessions SRR11528034 to SRR11528089 for the bacteria-targeted-sequencing dataset and SRR1159475 to 11529530 for the archaea-targeted-sequencing dataset.

qPCR analysis

Total bacteria, total archaea and the formyltetrahydrofolate synthetase (formate:tetrahydrofolate ligase (ADP-forming), EC; FTHFS) gene targeting homoacetogens were analysed by qPCR. All the amplification qPCR programs were performed in a BioRad CFX96 Real-Time Systems C1000 Touch Thermal Cycler (Bio-Rad Laboratories, USA). For total bacteria and total archaea qPCR analysis, primers 338F and 805R and primers 787F and 1059R, respectively, were used [64]. For the bacteria qPCR mix: SsoAdvanced™ Universal Probes Supermix (Bio-rad Laboratories, USA), 338F primer (100 nM), 805R primer (250 nM), TaqMan probe (50 nM), 2 μL of DNA and water was used until a volume of 12.5 μL. The qPCR cycle was the following: 40 cycles of dissociation (95 °C, 7 s) and elongation steps (60 °C, 25 s). The following mix was used for the archaea qPCR reactions: SsoAdvanced™ Universal Probes Supermix (Bio-rad Laboratories, USA), 787F primer and 1059R primer (200 nM), TaqMan probe (50 nM), 5 μL of DNA and water (final volume 25 μL). The qPCR cycle consisted of 40 cycles of denaturation (95 °C, 15 s) and elongation (60 °C, 1 min) (adapted from Braun et al., [64]). For the FTHFS gene, a PCR mix containing: SsoAdvanced™ Universal SYBR® Green Supermix, 500 nM of forward primer and reverse primer (described by Xu et al. [65]), 5 μL of DNA and water for a final volume of 25 μL. The qPCR program consisted in: 2 min at 98 °C, follow by 9 cycles of 45 s at 98 °C and 45 s at 63 °C, each cycle the second set temperature was decreased 1 °C. Finally, 30 cycles 98 °C, 45 s, hybridization (55 °C, 45 s) and elongation (72 °C, 1 min) (adapted from Xu et al. [65]). qPCR results are available in Additional file 1.

Statistics analysis

All statistical analyses were performed with R software v 3.6.2 using Rstudio v 1.2.1335. The Gap statistic to predict the optimal number of clusters from the used data set (Additional file 1) was calculated with the function “clusGap” of the “cluster” package v 2.0.8. The clusters were calculated using the k-means algorithm from the package “stats” v 3.4.4. The t-test analyses were made using the package “stats” v 3.4.4. The t-test analyses were made using the package “stats” v 3.4.4. The Kruskal–Wallis tests, the Wilcoxon test and the Bonferroni correction method to adjust the p-values for pairwise comparisons, were performed with the “rstarix” v 0.6.0. The Pearson correlation index was calculated with the function “rcorr” from the “Hmsic” v 4.2-0. The representation of the Pearson correlation matrix was made with the “corrplot” package v 0.89 was used. The diversity indexes were calculated with the PhyloSeq package v 1.28.0 [66].

Availability of data and materials

The generated sequencing datasets are registered in the Sequence Read Archive ( under the BioProject accession number PRJNA624130, with SRA accessions SRR11528034 to SRR11528089 for the bacteria-targeted-sequencing dataset and SRR1159475 to 11529530 for the archaea-targeted-sequencing dataset. The reactors’ performances data used for the analysis of the presented results are available in Additional file 1. The qPCR results are also presented in Additional file 1.



Anaerobic digestion


Acetotrophic methanogens


Basic local alignment search tool nucleotide


Biochemical methane potential


Continuous stirred tank reactor


Formyl-tetrahydrofolate synthase




Hydrogenotrophic methanogens


National Center for Biotechnology Information


Operational Taxonomic Unit


Polymerase chain reaction


Quantitative polymerase chain reaction


Sequence read archive


Up-flow anaerobic sludge blanket


Volatile fatty acids


Waste water treatment plant


  1. 1.

    Khan MA, Ngo HH, Guo WS, et al. Optimization of process parameters for production of volatile fatty acid, biohydrogen and methane from anaerobic digestion. Bioresour Technol. 2016;219:738–48.

    CAS  PubMed  Google Scholar 

  2. 2.

    Götz M, Lefebvre J, Mörs F, et al. Renewable power-to-gas: a technological and economic review. Renew Energy. 2016;85:1371–90.

    Google Scholar 

  3. 3.

    Kadam R, Panwar NL. Recent advancement in biogas enrichment and its applications. Renew Sustain Energy Rev. 2017;73:892–903.

    CAS  Google Scholar 

  4. 4.

    Sarker S, Lamb JJ, Hjelme DR, et al. Overview of recent progress towards in situ biogas upgradation techniques. Fuel. 2018;226:686–97.

    CAS  Google Scholar 

  5. 5.

    Angelidaki I, Treu L, Tsapekos P, et al. Biogas upgrading and utilization: current status and perspectives. Biotechnol Adv. 2018;36:452–66.

    CAS  PubMed  Google Scholar 

  6. 6.

    Lecker B, Illi L, Lemmer A, et al. Biological hydrogen methanation—a review. Bioresour Technol. 2017;245:1220–8.

    CAS  PubMed  Google Scholar 

  7. 7.

    Zabranska J, Pokorna D. Bioconversion of carbon dioxide to methane using hydrogen and hydrogenotrophic methanogens. Biotechnol Adv. 2018;36:707–20.

    CAS  PubMed  Google Scholar 

  8. 8.

    Luo G, Angelidaki I. Integrated biogas upgrading and hydrogen utilization in an anaerobic reactor containing enriched hydrogenotrophic methanogenic culture. Biotechnol Bioeng. 2012;109:2729–36.

    CAS  PubMed  Google Scholar 

  9. 9.

    Luo G, Johansson S, Boe K, et al. Simultaneous hydrogen utilization and in situ biogas upgrading in an anaerobic reactor. Biotechnol Bioeng. 2012;109:1088–94.

    CAS  PubMed  Google Scholar 

  10. 10.

    Balat M. Potential importance of hydrogen as a future solution to environmental and transportation problems. Int J Hydrogen Energy. 2008;33:4013–29.

    CAS  Google Scholar 

  11. 11.

    Deng L, Hägg MB. Techno-economic evaluation of biogas upgrading process using CO2 facilitated transport membrane. Int J Greenh Gas Control. 2010;4:638–46.

    CAS  Google Scholar 

  12. 12.

    Schink B. Energetics of syntrophic cooperation in methanogenic degradation. Microbiol Mol Biol Rev. 1997;61:262–80.

    CAS  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Amani T, Nosrati M, Mousavi SM. Using enriched cultures for elevation of anaerobic syntrophic interactions between acetogens and methanogens in a high-load continuous digester. Bioresour Technol. 2011;102:3716–23.

    CAS  PubMed  Google Scholar 

  14. 14.

    Angenent LT, Karim K, Al-Dahhan MH, et al. Production of bioenergy and biochemicals from industrial and agricultural wastewater. Trends Biotechnol. 2004;22:477–85.

    CAS  PubMed  Google Scholar 

  15. 15.

    Aryal N, Kvist T, Ammam F, et al. An overview of microbial biogas enrichment. Bioresour Technol. 2018;264:359–69.

    CAS  PubMed  Google Scholar 

  16. 16.

    Mulat DG, Mosbæk F, Ward AJ, et al. Exogenous addition of H 2 for an in situ biogas upgrading through biological reduction of carbon dioxide into methane. Waste Manag. 2017;68:146–56.

    CAS  PubMed  Google Scholar 

  17. 17.

    Wahid R, Mulat DG, Gaby JC, et al. Effects of H 2:CO 2 ratio and H 2 supply fluctuation on methane content and microbial community composition during in situ biological biogas upgrading. Biotechnol Biofuels. 2019;12:1–15.

    Google Scholar 

  18. 18.

    Cazier EA, Trably E, Steyer JP, et al. Biomass hydrolysis inhibition at high hydrogen partial pressure in solid-state anaerobic digestion. Bioresour Technol. 2015;190:106–13.

    CAS  PubMed  Google Scholar 

  19. 19.

    Bassani I, Kougias PG, Treu L, et al. Biogas upgrading via hydrogenotrophic methanogenesis in two-stage continuous stirred tank reactors at mesophilic and thermophilic conditions. Environ Sci Technol. 2015;49:12585–93.

    CAS  PubMed  Google Scholar 

  20. 20.

    Agneessens LM, Ottosen LDM, Andersen M, et al. Parameters affecting acetate concentrations during in situ biological hydrogen methanation. Bioresour Technol. 2018;258:33–40.

    CAS  PubMed  Google Scholar 

  21. 21.

    Kern T, Theiss J, Röske K, et al. Assessment of hydrogen metabolism in commercial anaerobic digesters. Appl Microbiol Biotechnol. 2016;100:4699–710.

    CAS  PubMed  Google Scholar 

  22. 22.

    Treu L, Kougias PG, de Diego-Díaz B, et al. Two-year microbial adaptation during hydrogen-mediated biogas upgrading process in a serial reactor configuration. Bioresour Technol. 2018;264:140–7.

    CAS  PubMed  Google Scholar 

  23. 23.

    Treu L, Campanaro S, Kougias PG, et al. Hydrogen-fueled microbial pathways in biogas upgrading systems revealed by genome-centric metagenomics. Front Microbiol. 2018;9:1079.

    PubMed  PubMed Central  Google Scholar 

  24. 24.

    Agneessens LM, Ottosen LDM, Voigt NV, et al. In-situ biogas upgrading with pulse H2 additions: the relevance of methanogen adaption and inorganic carbon level. Bioresour Technol. 2017;233:256–63.

    CAS  PubMed  Google Scholar 

  25. 25.

    Paillet F, Marone A, Moscoviz R, et al. Improvement of biohydrogen production from glycerol in micro-oxidative environment. Int J Hydrogen Energy. 2019;44:17802–12.

    CAS  Google Scholar 

  26. 26.

    Angelidaki I, Sanders W. Assessment of the anaerobic biodegradability of macropollutants. Rev Environ Sci Biotechnol. 2004;3:117–29.

    CAS  Google Scholar 

  27. 27.

    Cohen A, Zoetemeyer RJ, Deursen A, et al. Anaerobic digestion of glucose with separated acid production and methane formation. Water Res. 1979;13:571–80.

    CAS  Google Scholar 

  28. 28.

    Carrillo-Reyes J, Celis LB, Alatriste-Mondragón F, et al. Different start-up strategies to enhance biohydrogen production from cheese whey in UASB reactors. Int J Hydrogen Energy. 2012;37:5591–601.

    CAS  Google Scholar 

  29. 29.

    Angelidaki I, Alves M, Bolzonella D, et al. Defining the biomethane potential (BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays. Water Sci Technol. 2009;59:927–34.

    CAS  PubMed  Google Scholar 

  30. 30.

    Elbeshbishy E, Nakhla G, Hafez H. Biochemical methane potential (BMP) of food waste and primary sludge: influence of inoculum pre-incubation and inoculum source. Bioresour Technol. 2012;110:18–25.

    CAS  PubMed  Google Scholar 

  31. 31.

    Jia XS, Furumai H, Fang HHP. Yields of biomass and extracellular polymers in four anaerobic sludges. Environ Technol (United Kingdom). 1996;17:283–91.

    CAS  Google Scholar 

  32. 32.

    Liu R, Hao X, Wei J. Function of homoacetogenesis on the heterotrophic methane production with exogenous H2/CO2 involved. Chem Eng J. 2016;284:1196–203.

    CAS  Google Scholar 

  33. 33.

    Rachbauer L, Beyer R, Bochmann G, et al. Characteristics of adapted hydrogenotrophic community during biomethanation. Sci Total Environ. 2017;595:912–9.

    CAS  PubMed  Google Scholar 

  34. 34.

    Lay JJ, Li YY, Noike T. Interaction between homoacetogens and methanogens in lake sediments. J Ferment Bioeng. 1998;86:467–71.

    CAS  Google Scholar 

  35. 35.

    Hashimoto AG. Methane production from beef cattle manure: effects of temperature, hydraulic retention time and influent substrate concentration. Biotechnol Bioeng. 1982;24:2039–52.

    CAS  PubMed  Google Scholar 

  36. 36.

    Weiland P. Biogas production: current state and perspectives. Appl Microbiol Biotechnol. 2010;85:849–60.

    CAS  PubMed  Google Scholar 

  37. 37.

    Treu L, Kougias PG, Campanaro S, et al. Deeper insight into the structure of the anaerobic digestion microbial community; The biogas microbiome database is expanded with 157 new genomes. Bioresour Technol. 2016;216:260–6.

    CAS  PubMed  Google Scholar 

  38. 38.

    Hassa J, Maus I, Off S, et al. Metagenome, metatranscriptome, and metaproteome approaches unraveled compositions and functional relationships of microbial communities residing in biogas plants. Appl Microbiol Biotechnol. 2018;102:5045–63.

    CAS  PubMed  PubMed Central  Google Scholar 

  39. 39.

    Schuchmann K, Müller V. Autotrophy at the thermodynamic limit of life: a model for energy conservation in acetogenic bacteria. Nat Rev Microbiol. 2014;12:809–21.

    CAS  PubMed  Google Scholar 

  40. 40.

    Willems A, Collins MD. Phylogenetic relationships of the genera Acetobacterium and Eubacterium sensu stricto and reclassification of Eubacterium alactolyticum as Pseudoramibacter alactolyticus gen. nov., comb. nov. Int J Syst Bacteriol. 1996;46:1083–7.

    CAS  PubMed  Google Scholar 

  41. 41.

    Rago L, Zecchin S, Marzorati S, et al. A study of microbial communities on terracotta separator and on biocathode of air breathing microbial fuel cells. Bioelectrochemistry. 2018;120:18–26.

    CAS  PubMed  Google Scholar 

  42. 42.

    Calusinska M, Goux X, Fossépré M, et al. A year of monitoring 20 mesophilic full-scale bioreactors reveals the existence of stable but different core microbiomes in bio-waste and wastewater anaerobic digestion systems. Biotechnol Biofuels. 2018;11:1–19.

    Google Scholar 

  43. 43.

    FitzGerald JA, Wall DM, Jackson SA, et al. Trace element supplementation is associated with increases in fermenting bacteria in biogas mono-digestion of grass silage. Renew Energy. 2019;138:980–6.

    CAS  Google Scholar 

  44. 44.

    Niimura Y, Komagata K, Kozaki M. Anaerobic sporeforming xylan-digesting bacterium which lacks cytochrome, quinone, and catalase. Int J Syst Bacteriol. 1990;5:297–301.

    Google Scholar 

  45. 45.

    Togo AH, Khelaifia S, Lagier JC, et al. Noncontiguous finished genome sequence and description of Paenibacillus ihumii sp. nov. strain AT5. New Microbes New Infect 2016; 10: 142–150.

  46. 46.

    Wang L, Cui YS, Kwon CS, et al. Vagococcus acidifermentans sp. nov., isolated from an acidogenic fermentation bioreactor. Int J Syst Evol Microbiol. 2011;61:1123–6.

    CAS  PubMed  Google Scholar 

  47. 47.

    Zhilina TN, Zavarzina DG, Osipov GA, et al. Natronincola ferrireducens sp. nov., and Natronincola peptidovorans sp. nov., new anaerobic alkaliphilic peptolytic iron-reducing bacteria isolated from soda lakes. Microbiology. 2009;78:455–67.

    CAS  Google Scholar 

  48. 48.

    Niu L, Song L, Dong X. Proteiniborus ethanoligenes gen. nov., sp. nov., an anaerobic protein-utilizing bacterium. Int J Syst Evol Microbiol. 2008;58:12–6.

    CAS  PubMed  Google Scholar 

  49. 49.

    Padden AN, Dillon VM, Edmonds J, et al. An indigo-reducing moderate thermophile from a woad vat, Clostridium isatidis sp. nov. Int J Syst Bacteriol. 1999;49:1025–31.

    CAS  PubMed  Google Scholar 

  50. 50.

    Qu X, Mazéas L, Vavilin VA, et al. Combined monitoring of changes in δ13CH4 and archaeal community structure during mesophilic methanization of municipal solid waste. FEMS Microbiol Ecol. 2009;68:236–45.

    CAS  PubMed  Google Scholar 

  51. 51.

    Hao L, Lü F, Li L, et al. Response of anaerobes to methyl fluoride, 2-bromoethanesulfonate and hydrogen during acetate degradation. J Environ Sci (China). 2013;25:857–64.

    CAS  Google Scholar 

  52. 52.

    De Vrieze J, Gildemyn S, Vilchez-Vargas R, et al. Inoculum selection is crucial to ensure operational stability in anaerobic digestion. Appl Microbiol Biotechnol. 2014;99:189–99.

    PubMed  Google Scholar 

  53. 53.

    Chen X, Ottosen LDM, Kofoed MVW. How low can you go: methane production of methanobacterium congolense at low CO2 concentrations. Front Bioeng Biotechnol. 2019;7:1–10.

    Google Scholar 

  54. 54.

    De Vrieze J, Hennebel T, Boon N, et al. Methanosarcina: the rediscovered methanogen for heavy duty biomethanation. Bioresour Technol. 2012;112:1–9.

    PubMed  Google Scholar 

  55. 55.

    Bio2E, INRAE. Environmental Biotechnology and Biorefinery Platform. Epub ahead of print 2020.

  56. 56.

    APHA. Standard methods for the examination of water and wastewater analysis. Am Public Heal Assoc Washingt DC.

  57. 57.

    Zoetemeyer RJ, van den Heuvel JC, Cohen A. pH influence on acidogenic dissimilation of glucose in an anaerobic digestor. Water Res. 1982;16:303–11.

    CAS  Google Scholar 

  58. 58.

    Mosey FE. Mathematical modelling of the anaerobic digestion process: regulatory mechanisms for the formation of short-chain volatile fatty acids from glucose. Water Sci Technol. 1983;15:209–32.

    CAS  Google Scholar 

  59. 59.

    Costello DJ, Greenfield PF, Lee PL. Dynamic modelling of a single-stage high-rate anaerobic reactor-I. Model derivation. Water Res. 1991;25:847–58.

    CAS  Google Scholar 

  60. 60.

    Horiuchi JI, Shimizu T, Tada K, et al. Selective production of organic acids in anaerobic acid reactor by pH control. Bioresour Technol. 2002;82:209–13.

    CAS  PubMed  Google Scholar 

  61. 61.

    Moscoviz R, Trably E, Bernet N. Consistent 1,3-propanediol production from glycerol in mixed culture fermentation over a wide range of pH. Biotechnol Biofuels. 2016;9:1.

    CAS  Article  Google Scholar 

  62. 62.

    Carmona-Martínez AA, Trably E, Milferstedt K, et al. Long-term continuous production of H <inf>2</inf> in a microbial electrolysis cell (MEC) treating saline wastewater. Water Res. 2015;81:149–56.

    PubMed  Google Scholar 

  63. 63.

    Venkiteshwaran K, Milferstedt K, Hamelin J, et al. Anaerobic digester bioaugmentation influences quasi steady state performance and microbial community. Water Res. 2016;104:128–36.

    CAS  PubMed  Google Scholar 

  64. 64.

    Braun F, Hamelin J, Gèvaudan G, et al. Development and application of an enzymatic and cell flotation treatment for the recovery of viable microbial cells from environmental matrices such as anaerobic sludge. Appl Environ Microbiol. 2011;77:8487–93.

    CAS  PubMed  PubMed Central  Google Scholar 

  65. 65.

    Xu K, Liu H, Du G, et al. Real-time PCR assays targeting formyltetrahydrofolate synthetase gene to enumerate acetogens in natural and engineered environments. Anaerobe. 2009;15:204–13.

    CAS  PubMed  Google Scholar 

  66. 66.

    McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE. 2013.

    Article  PubMed  PubMed Central  Google Scholar 

Download references


This study was co-founded by the European Union under the Operational Programme FEDER-FSE Languedoc—Roussillon 2014–2020 (axes of interest: AP01-OT01-PI01a-OS01-3; Synergy: LR0012032) and The Direction of Research, Technological Transfer and Higher Education of the Région Occitanie (ALDOCT-000253). The authors would like to thank the Bio2E platform ( for experimental support.


This study was co-founded by the European Union under the Operational Programme FEDER-FSE Languedoc—Roussillon 2014–2020 (Synergy: LR0012032) and The Direction of Research, Technological Transfer and Higher Education of the Région Occitanie (ALDOCT-000253).

Author information




LBN, ET, JPD and RE designed the study. LBN performed lab work, analysed and interpreted the results of the reactors’ performance, the sequencing analysis and the qPCR analysis. LBN drafted the initial manuscript of this work. GSC designed the method for the amplification of the region V4–V5 of the microbial 16 S rRNA for the detection of archaea, performed all the qPCR analysis and contributed in the interpretation of the sequencing and qPCR data. ET, JPD and RE contributed to data interpretation and writing of the manuscript. LBN, ET, JPD, RE, GSC and NB reviewed and edited subsequent version of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Renaud Escudié.

Ethics declarations

Ethics and consent to participate

Not applicable.

Consent of publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Additional file 1:

The reactors’ performances results and qPCR results.

Additional file 2: Table S1.

The COD mass balance analysis of the reactors in MS Word document format.

Additional file 3: Figure S1.

The results of the statistical analysis of the clusters regarding: S1a) the methane production and S1b) the VFA production in PDF format.

Additional file 4: Figure S2.

The amount of total produced VFA in gCOD vs percentage of consumed H2 in MS Word document format.

Additional file 5:

The results of the preliminary experiment.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Braga Nan, L., Trably, E., Santa-Catalina, G. et al. Biomethanation processes: new insights on the effect of a high H2 partial pressure on microbial communities. Biotechnol Biofuels 13, 141 (2020).

Download citation


  • Hydrogen
  • Anaerobic digestion
  • Biogas upgrading
  • Power-to-gas
  • Hydrogenotrophic methanogens
  • Homoacetogens
  • Methanosarcinales
  • Methanosarcina sp.