Metabolic Model and Molar Balances
For all experiments, gas composition of the headspace was monitored prior to its replacement and H2 pressure was constantly kept at 2.15 ± 0.10 atm. H2 was the sole gas detected in the headspace in all cases, except for the control experiment, which showed a methane production of 0.24 mmol (5.37 mL at standard temperature and pressure (STP)). Initial and final concentrations of the main metabolites (added or produced) and of dissolved H2 and CO2 are presented in Table 3. Dissolved H2 was not measured but estimated through Henry’s law [42, 43], and its concentration was considered constant, due to its continuous replacement. Dissolved CO2 was calculated from the alkalinity value measured at the beginning of each assay. CO2, O2 and N2 were never detected at the headspace. Acetone, lactate, and methanol were only sporadically detected in the liquid phase, in traces concentrations (below 1 mg L−1). Neither propionate nor alcohols were detected at the beginning of each experiment.
Table 3
– Mean concentrations of all metabolites detected and of dissolved H2 and CO2.
Metabolitea | Concentration (mg L−1) |
Control | Acidic | Thermal | Acidic-thermal | Thermal-acidic |
Initial | Final | Initial | Final | Initial | Final | Initial | Final | Initial | Final |
Acetate | 1225 ± 1 | 1575 ± 40 | 1222 ± 3 | 1492 ± 24 | 1206 ± 2 | 1722 ± 120 | 1235 ± 3 | 1645 ± 44 | 1224 ± 2 | 1447 ± 26 |
Butyrate | 1831 ± 3 | 1773 ± 24 | 1828 ± 1 | 1935 ± 14 | 1791 ± 4 | 1796 ± 17 | 1840 ± 5 | 1664 ± 25 | 1831 ± 4 | 1738 ± 58 |
Propionate | - | 42 ± 12 | - | 35 ± 4 | - | 107 ± 25 | - | 52.8 ± 9.2 | - | 40.1 ± 5.4 |
Ethanol | - | 40 ± 6 | - | 14 ± 2 | - | 104 ± 7 | - | 94 ± 9 | - | 33 ± 6 |
Propanol | - | - | - | 0.3 ± 0.2 | - | - | - | 1.8 ± 0.1 | - | 1.0 ± 0.3 |
Butanol | - | 49 ± 8 | - | 11 ± 2 | - | 83 ± 3 | - | 154 ± 14 | - | 61 ± 25 |
Dissolved H2 | 1.6 ± 0.1 | 1.7 ± 0.0 | 1.6 ± 0.1 | 1.7 ± 0.1 | 1.7 ± 0.1 |
Dissolved CO2 | 3.07 | - | 2.40 | - | 3.30 | - | 1.56 | - | 2.19 | - |
Duration (d) | 29.2 | 29.1 | 29.1 | 29.1 | 29.0 |
aAll values are expressed in mg L−1 except for dissolved H2 and CO2 expressed in mmol L−1. Mean concentrations of all volatile acids and alcohols was calculated considering 5 replicates. The amount of H2 was kept constant in all experiments, there were no difference in its initial and final concentrations. Dissolved H2 mean concentration considered 50 replicates. |
Acetate and butyrate concentrations along the process demonstrated acetate production in all cases. Only minor variations of butyrate concentrations (slight decrease for the control and the acidic-thermal and thermal-acidic pretreatments, low production for the acidic and the thermal pretreatments) were observed. Such results suggested an alternative organic matter input in the system. This hypothesis is consistent with the decrease of the concentration of total volatile solids (TVS) that was observed for all conditions tested, as a direct effect of the different pretreatments on the biomass, which might have resulted in partial microbial cell death. Such non-living cells constituted non-soluble organic matter which was probably hydrolyzed and then consumed as a supplementary carbon source. Based on well-known anaerobic acidogenic and solventogenic metabolisms [1, 44], and considering the inoculum as the only possible alternative source of organic matter, an alternative metabolic model including this alternative contribution and considering the metabolites (added and produced) presented in Table 3, was developed, as shown in Figure 1.
Figure 1 depicts butanol as the only product from lysed inoculum (pathway 1). Pathway 1 seems the only way to explain butanol production maintaining the butyric acid concentration constant since it is thermodynamically unfeasible to form butanol from acetate. Acetate could only be produced from butyrate, considering the high concentration of hydrogen and lack of CO2 in all assays, indicating that acetate was produced through acetogenesis from butyrate (pathway 2) or homoacetogenesis (pathway 7) after initial dissolved CO2 was consumed.
The metabolic model depicted in Figure 1 was used to perform a molar balance of all carbon inputs and outputs during the process (Table 4) and estimates the butyrate input from inoculum. All balances were performed on a 1 L basis and derived from initial and final experimental values obtained in all conditions (Table 3), except for the organic matter from the inoculum (pathway 1) which was expressed in butyrate equivalent.
Table 4
– Molar balance of all metabolites detected in all metabolic pathways of each experiment.
Metabolite | Pathwaya | Molar Balance (mmol)b |
Control | Acidic | Thermal | Acidic-thermal | Thermal-acidic |
Butyrate | Initial* | 21.0 | 21.0 | 20.6 | 21.1 | 21.0 |
Final* | 20.4 | 22.2 | 20.6 | 19.1 | 20.0 |
1 | + 3.16 | + 3.56 | + 6.92 | + 4.73 | + 1.96 |
2 | - 2.53 | - 1.73 | - 4.34 | - 3.92 | - 1.57 |
3 | - 0.66 | - 0.15 | - 1.12 | - 2.08 | - 0.82 |
4 | - 0.57 | - 0.48 | - 1.46 | - 0.75 | - 0.57 |
Butanol | 3* | + 0.66 | + 0.15 | + 1.12 | + 2.08 | + 0.82 |
Propionate | 4* | + 0.57 | + 0.48 | + 1.46 | + 0.75 | + 0.57 |
5 | - | - | - | - 0.03 | - 0.02 |
Propanol | 5* | - | - | - | + 0.03 | + 0.02 |
Acetate | Initial* | 20.7 | 20.7 | 20.4 | 20.9 | 20.7 |
Final* | 26.7 | 25.3 | 29.2 | 27.9 | 24.5 |
2 | + 2.53 | + 1.73 | + 4.34 | + 3.92 | + 1.57 |
6 | - 0.87 | - 0.30 | - 2.26 | - 2.04 | - 0.72 |
7 | + 4.35 | + 3.17 | + 6.72 | + 5.08 | + 2.95 |
Ethanol | 6* | + 0.87 | + 0.30 | + 2.26 | + 2.04 | + 0.72 |
Dissolved CO2 | Initial* | 3.07 | 2.40 | 3.30 | 1.56 | 2.19 |
2 | + 5.05 | + 3.46 | + 8.68 | + 7.85 | + 3.14 |
4 | + 0.57 | + 0.48 | + 1.46 | + 0.75 | + 0.57 |
7 | - 8.69 | - 6.34 | - 13.4 | - 10.2 | - 5.90 |
*Values obtained experimentally (derived from initial and final concentrations as indicated in Table 3). |
aMetabolic pathways described in Figure 1. |
bBalance on a 1 L basis; “+” indicates a production; “-” indicates a consumption. |
According to the mass balance shown in Table 4, CO2 was produced by converting butyrate into acetate (pathway 2) and by acidogenesis from butyrate to propionate (pathway 4). This CO2 was then consumed to form acetate through a homoacetogenic pathway (pathway 7). This hypothesis was supported by the absence of gaseous CO2 in all experiments, followed by an increase in acetate concentration. This CO2 absence may be linked to microbial communities more adapted to convert butyrate and CO2 into acetate (pathways 2 and 7, respectively) rather than to the conversion of acetate into ethanol (pathway 6). This metabolic model also indicates that butyrate was mainly consumed to produce acetate (pathway 2) in all studied conditions and the organic matter from the inoculum (pathway 1) represented an external input since butyrate concentrations were almost constant in all assays, although consumed to form butanol (pathway 3). Ethanol and butanol were produced (pathways 3 and 6) in all experiments. These pathways were more active in the experiments with thermal and acidic-thermal pretreatments, as these conditions showed the highest ethanol and butanol production among all other experiments. Although acidogenesis from butyrate to propionate (pathway 4) was active in all conditions, conversion of propionate into propanol (pathway 5) was not an important metabolic pathway. Propanol production was not expressive in neither experiment since it showed the highest concentration of 1.82 mg L−1 in acidic-thermal pretreatment essay.
The successful closure of molar balances shown in Table 4, with a stoichiometrically balanced sum of all inputs and outputs (considering the contribution of the inoculum), indicates that the metabolic model proposed in Figure 1 accurately represents the solventogenic processes occurring in all studied conditions. An energy balance based on the mass balance depicted in Table 4 was calculated to estimate Gibb’s free energy values for each pathway for both initial and final conditions (Figure 2). Those values provide an overview of the thermodynamic feasibility of each pathway involved in the solventogenic process and tend to validate the metabolism proposed in Figure 1.
Energetic profiles could help to anticipate changes within metabolic pathways. As shown in Figure 2, the estimated ΔGor values indicate that acetogenesis from butyrate and acidogenesis from butyrate to propionate (pathways 2 and 4, respectively) are theoretically thermodynamically unfeasible metabolisms. Nevertheless, these pathways became thermodynamically feasible due to the absence (or low concentration) of dissolved CO2 at the course of each essay. Acetogenesis from butyrate (pathway 2) was likely inhibited at the beginning of each experiment due to the concentration of H2 and CO2. During the experiment, CO2 was progressively consumed through homoacetogenesis (pathway 7), which rendered pathway 2 thermodynamically feasible. At low concentrations of CO2 as found after the beginning of each essay, the homoacetogenic pathway might have become thermodynamically unfeasible. Gibb’s free energy values found for these pathways could explain the production of acetate observed in all studied conditions, all along the process. Metabolic pathways 3 (solventogenesis from butyrate to butanol) and 6 (solventogenesis from acetate to ethanol) were thermodynamically feasible in all studied conditions all along the process, explaining the observed production of such alcohols. Metabolic pathway 4 (acidogenesis from butyrate to propionate) was unfeasible at the ending of each condition, showing that propionate production was stationed in a low concentration at some point of the experiments.
As described previously, H2 has a significant role as a co-substrate in the anaerobic process energetics [45, 46]. As shown in Figure 3, solventogenesis of butanol from butyrate (pathway 3) is less sensitive to low ppH2 than solventogenesis of ethanol and propanol, respectively from acetate (pathway 6) and propionate (pathway 5). By extrapolation, equilibrium (ΔGor = 0) in pathway 3 would be achieved at a ppH2 of 4·10−4 atm (43.7 Pa), implying that solventogenesis of butanol could theoretically be carried out at very low ppH2. It is also possible to infer from Figure 3 that at ppH2 higher than 0.62 atm (62.5 MPa), acidogenesis of propionate (pathway 4) and acetogenesis of acetate (pathway 2) from butyrate will stop, shutting down the production of acetate, propionate and H2. In addition to low CO2 concentrations, high ppH2 thus represents an important factor that renders such process thermodynamically feasible and favours solventogenesis from VFAs.
Alcohols and VFAs metabolisms
As shown in Figure 4, the highest concentrations of alcohols were obtained for both thermal and acidic-thermal pretreatments, with the best rate observed for the acidic-thermal pretreatment, especially for butanol production. On the opposite, the lowest alcohol production was observed for the acidic pretreatment. Acetate and butyrate concentrations along the essays confirmed the contribution of the inoculum as an important source of organic matter as the substrate for alcohol production, since in all conditions tested, acetate concentration increased, and butyrate concentration increased or only slightly decreased. These results reinforced the hypothesis described in Figure 1 and Table 4 of an acetate production through acetogenesis of butyrate (pathway 2) and homoacetogenesis (pathway 7), and preferentially through pathway 2 due to low concentrations of CO2 at the beginning of all assays.
As shown in Figure 5, propionate was produced in all conditions tested. No significant difference was observed in the concentrations produced (42.4 ± 10.0 mg L−1), except for the thermal pretreatment essay, for which the level of production was 2.5 times higher than in all other conditions. As presented in Table 3, propionate was then consumed and converted into propanol (pathway 5, Figure 1) for the acidic, acidic-thermal and thermal-acidic pretreatments. Despite the favourable thermodynamics of this reaction (Figure 2), propanol was only produced in trace amounts. Oxidation of propionate into acetate was considered unlikely since such reaction is energetically unfeasible in standard conditions (ΔGor = 53.3 kJ mol−1), and that, as previously described [39], at high H2 concentrations ΔGr values are increased.
Table 5 presents the parameters of the modified Boltzmann model fitting the experimental data from Figure 4. Those parameters are related to the kinetics of alcohol production and compare efficiencies between all pretreatments. A high correspondence was obtained between replicates for all conditions tested, apart from the thermal-acidic pretreatment. The correlation coefficient was very low in that condition, only reaching 0.5 for ethanol and 0.4 for butanol, indicating that the process was unpredictable and could not be reproduced. Due to its instability and unpredictability, the thermal-acidic pretreatment was thus no longer considered for analyses.
Table 5
– Parameters of the modified Boltzmann model fitting for ethanol and butanol production.
Parameters | Pretreatment of the inoculum |
Control | Acidic | Thermal | Acidic-thermal | Thermal-acidic |
Ethanol | \({C}_{max}\) (mg L−1) | 38.5 ± 1.5 | 13.6 ± 0.7 | 122 ± 10 | 87.3 ± 1.9 | 27.6 ± 2.8 |
\({t}_{m}\) (d) | 12.7 ± 0.6 | 10.4 ± 0.8 | 21.0 ± 1.1 | 6.9 ± 0.3 | 6.1 ± 1.0 |
\({r}_{max}\) (mg L−1 d−1) | 1.9 ± 0.0 | 0.7 ± 0.0 | 7.0 ± 0.48 | 9.4 ± 0.7 | 3.1 ± 1.2 |
R2 | 0.96 | 0.89 | 0.94 | 0.95 | 0.50 |
\({t}_{i}\) (d) | 2.7 | 0.9 | 12 | 2.3 | 1.6 |
\({t}_{e}\) (d) | 20 | 19 | 18 | 9.3 | 9.0 |
Butanol | \({C}_{max}\) (mg L−1) | 50.7 ± 2.3 | 10.9 ± 0.5 | 96.7 ± 6.8 | 143 ± 2 | 53.8 ± 5.4 |
\({t}_{m}\) (d) | 17.6 ± 0.7 | 14.2 ± 0.8 | 21.1 ± 1.0 | 5.3 ± 0.2 | 5.1 ± 0.8 |
\({r}_{max}\) (mg L−1 d−1) | 2.4 ± 0.0 | 0.5 ± 0.0 | 5.5 ± 0.5 | 25 ± 2 | 26 ± 23 |
R2 | 0.97 | 0.94 | 0.96 | 0.96 | 0.40 |
\({t}_{i}\) (d) | 7.0 | 2.9 | 12 | 2.4 | 4.1 |
\({t}_{e}\) (d) | 21 | 23 | 18 | 6 | 2 |
\({C}_{max}\), maximum concentration; \({t}_{m}\), time when maximum production rate is achieved; \({r}_{max}\), maximum production rate; R2, correlation coefficient; \({t}_{i}\) and \({t}_{e}\), initial and ending time of exponential growth phase. |
As shown in Figure 4 and Table 5, both alcohol production and maximum production rates (\({r}_{max}\)) were improved for thermal and acidic-thermal pretreatments. Although the highest ethanol production was observed for the thermal pretreatment, the best results were obtained for the acidic-thermal pretreatment, which allowed the best butanol production and an increase of 4.5 times of the ethanol and of 10.2 times of the butanol maximum production rate. In opposition, a decrease in alcohol production was observed for the acidic pretreatment. Such results clearly indicate that an acidic-thermal pretreatment of the inoculum is a promising approach for designing more efficient and smaller-sized bioreactors.
Length of the lag phase (\({t}_{i}\)) and duration of the bacterial exponential growth phase (\({t}_{e}\)) were evaluated for each pretreatment and compared to the control by considering alcohol production curves as growth-associated curves (Table 3). The length of the lag phase is related to the time required for a bioreactor to initiate its process (start-up) and achieve higher rates of alcohol production. For both ethanol and butanol production, the shortest lag phases were observed for the acidic and acidic-thermal pretreatments. In both pretreatments, the lag phase was shorter than found in the control essay. The most extended lag phase was observed for the thermal pretreatment, which was considerably greater for both ethanol and butanol production than observed in control. Such results indicate that the thermal pretreatment had the highest impact on the inoculum, while acidic and acidic-thermal pretreatments selected microbial communities which were the best adapted to solventogenic processes. Variations in the duration of the bacterial exponential growth phases (\({t}_{e}\)) were also observed (Table 5). A shorter exponential growth phase is representative of a faster process. However, this parameter must be evaluated concomitantly with the maximum rates of production (\({r}_{max}\)). For example, a short \({t}_{e}\) occurring at low \({r}_{max}\) indicates a process occurring with low efficiency. Among all the conditions tested, low \({t}_{e}\) values and high \({r}_{max}\) values were obtained for the acidic-thermal pretreatment for ethanol and butanol production. Taken all together, results obtained for all conditions tested indicate that the acidic-thermal pretreatment was the best approach to generate a rapid and efficient process.
pH was initially set to a value of 5.92 ± 0.09 for all essays, and their value increased during the experiments (Figure 6). This rise in pH values likely reflects the consumption of H2 as an electron donor to form alcohols since solventogenic pathways (3, 5 and 6) require the consumption of H+ and alcohols show a low ionization on water. Such a pH increase occurred to a lesser extent in all conditions, including an inoculum acidic pretreatment (acidic, acidic-thermal and thermal-acidic pretreatments). This difference might be explained by a more drastic initial pH drop resulting from the acidic addition on these pretreatments, probably lowering the buffering capacity of the inoculum. This hypothesis is reinforced by the concentrations of dissolved CO2, which were higher for both control and thermal pretreatment than acidic and acidic-thermal pretreatments (Table 3).
A decrease in total volatile solids (TVS) was also observed for all conditions tested (Figure 7). This decrease occurred probably due to a direct effect of the different pretreatments on the biomass, resulting in partial microbial cell death. Such non-living cells constituted non-soluble organic matter which was probably hydrolyzed and then consumed as a carbon source, contributing to the organic matter input proposed in pathway 1 of Figure 1. The TVS concentration stopped decreasing and started to increase after 6.3 days in the acidic-thermal pretreatment slightly. Such evolution is probably linked to the higher values of \({r}_{max}\) observed for this pretreatment (Table 5), which reflect an increase in the biomass growth’s rate, and consequently, the TVS concentration.
Bacterial communities
Based on the results presented above and focusing on alcohol production, both thermal and acidic-thermal pretreatments experiments were submitted to microbial community analyses. Bacterial populations were thus characterized and monitored all along with those two processes. Thirteen phyla were detected, namely Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, Synergistetes, Cyanobacteria, Tenericutes, Spirochaetes, Deinococcus-Thermus, Fibrobacteres, Verrucomicrobia, Chloroflexi and Nitrospirae. Among those, three (Firmicutes, Proteobacteria and Bacteroidetes) represented up to 94.6% of the total OTUs detected in each sample for both pretreatments (Figure 8). Firmicutes was the most abundant phylum detected in the initial non-treated inoculum (corresponding to 66.6% of the detected OTUs). Both pretreatments generated a shift in the bacterial population by stimulating the development of Proteobacteria and strongly reducing the number of Firmicutes. Bacterial populations evolved differently along the process for the two pretreatments. In the thermal pretreatment experiment (Figure 8A), a decrease of Proteobacteria was observed concomitantly with an increase in Bacteroidetes and Firmicutes, the latest becoming the dominant phylum at the end of the process. This shift started after 200 hours of experiment, which correspond to the early beginning of the bacterial exponential growth phase (Figure 4C and Table 5). In opposition, in the acidic-thermal pretreatment experiment (Figure 8B), relative stability was observed for the bacterial population, with Proteobacteria remaining the main phylum all along the process (72.0 ± 4.99% of the detected OTUs) and Firmicutes constantly being the second phylum of importance (24.0 ± 5.10% of the detected OTUs). Bacteroidetes increased during the process, before starting to decrease after 500h of experiment, to reach their initial level at the end of the experiment.
Deeper phylogenetic analyses were performed down to the genus level to infer the potential metabolic pathway(s) that could be associated with the enhancement of alcohol production observed for both thermal and acidic-thermal pretreatments (Figure 9). Substantial differences were noticed between the bacterial population from the initial inoculum (before any pretreatment) and those from the thermal and acidic-thermal pretreated inocula.
A total of 317 genera were detected in all samples tested, but only 31 of them, representing up to 85% of the bacterial population, were considered for further analysis (Figure 9). One of the most noticeable changes consisted of an increase of Pseudomonas in pretreatment experiments compared to the initial population. Such increase was observed as soon as the first hours of experiments. Pseudomonas then remained a dominant genus of the bacterial population, despite a slight decrease observed for the thermal pretreatment in the second half of the process. Several Pseudomonas species have been genetically well-characterized and have been shown to possess the genetic components for both ethanol (from pyruvate) and butanol (from glycerol and pyruvate) production [47]. The prevalence of Pseudomonas is likely linked with the higher levels of alcohol production observed in thermal and acidic-thermal pretreatments. Some Pseudomonas species also possess genes responsible for the degradation of ethanol into acetyl-CoA [48]. A pathway related to ethanol degradation might explain better butanol production compared to ethanol, observed for the acidic-thermal pretreatment. Among the other noticeable results, two bacterial genera, namely Acinetobacter and Paenibacilus, which were not detected in the initial population, appeared to be positively affected by both pretreatments. It is being reported [49, 50] that Acinetobacter strains are related to alcohol consumption as they can express alcohol dehydrogenase (ADH) to convert ethanol into acetate, reverting the solventogenic pathway 6. As shown in Figure 2, this reversed pathway probably not occurred during the essays since in all experiments, in their beginning and ending, ΔGr for pathway 6 were thermodynamically feasible.
Conversely, ADH is also a molecule that is strongly related to bacterial quorum sensing with a key role in biofilm formation [51, 52]. The growth of Acinetobacter was more stimulated in the thermal pretreatment, so this quorum sense mechanism could be linked with the best performance of this pretreatment to produce alcohols, considering that the presence of ethanol could have stimulated the expression of ADH, although the alcohol degradation metabolism was probably shut down. Acidic-thermal pretreatment preferentially stimulated the growth of Paenibacilus, which is directly related to alcohol production [53].
In addition to those changes, several other bacterial genera evolved differently depending on the pretreatment applied (Figure 9). The thermal pretreatment appeared to stimulate the growth of Brevundimonas, Bacteroides, Butyrivibrio, Bacillus, Alcaligenes, Eubacterium, Clavibacter, Psychrobacter, Serratia and Microbulbifer. The latter three being even exclusively detected in this pretreatment. On the opposite, the acidic-thermal pretreatment appeared to improve the growth of Tissierella, Novosphingobium, Sedimentibacter and Yersinia. The latter one being exclusively detected in this pretreatment.
The genus Brevudimonas started to increase in the thermal pretreatment after 8.3 days, coinciding with the very beginning of the exponential growth phase (Figure 4 and Table 5). Members belonging to this genus have already been shown to possess metabolic pathways related to the acidogenesis of alcohols [54]. High H2 partial pressures applied during the process might have rendered solventogenesis thermodynamically favourable through the reverse same metabolic pathway (Figure 3). The genus Yersinia started to increase in the acidic-thermal pretreatment after approximately 4.2 days, coinciding with the early exponential growth phase (Figure 4 and Table 5). Members belonging to this genus have already been shown to possess metabolic pathways to convert pyruvate into butanol and ethanol [55]. The genus Sedimentibacter is one of the predominant genus found in the fermentation of Baijiu, and it could be directly related to alcohol production [56]. It is noticeable that the genus Clostridium, which is composed of a high number of known alcohol producers [57–60], was detected for both pretreatments all along the process, without any significant variations. The genus Clostridium usually produces alcohol by converting glycerol and pyruvate into ethanol and butanol. Thus, this genus probably has an essential role in producing ethanol and butanol in both thermal and acidic-thermal pretreatments. There is no reporting on alcohol-producing metabolism for all other genera found within the experiments.
Other results were observed at higher taxonomic levels, notably decreasing microorganisms from the Peptostreptococcaceae family and the Epsilonproteobacteria class, in both pretreatments. Peptostreptococcaceae family represented almost half (45% of total bacterial OTUs) found in the initial inoculum (before any pretreatment), strongly decreased its presence to 5.4 ± 2.2% in thermal and 9.2 ± 3.1% in acidic-thermal pretreatments. Epsilonproteobacteria class represented up to 6% of total bacterial OTUs in the initial inoculum but significantly decreased to less than 0.3% of total bacterial OTUs in both thermal and acidic-thermal pretreatments. Such results indicate that neither bacteria from the Peptostreptococcaceae family nor Epsilonproteobacteria class played a significant role in the processes studied.