Low recovery of Br− (31%), and therefore of injection solution, was due to a fairly high average pore water velocity (∼0.4 m day−1) at the site. However, the incubation time was chosen to allow detectable acetate and SO2−4 consumption, as determined in a previous study.
Acetate and SO2−4 were degraded at rates similar to what was found before in the same well with computed k of 0.31 ± 0.04 day−1 for acetate and 0.34 ± 0.05 day−1 for SO2−4 (previously reported k values were 0.60 ± 0.06 day−1 for acetate and 0.24 ± 0.01 day−1 for SO2−4). The computed k for SO2−4 consumption were higher than previously reported by Schroth et al. for the same well without any carbon substrate addition (k= 0.04 to 0.13 ± 0.01 day−1). Similarly, in other aquifers k for SO2−4 consumption without C addition was lower: 0.02–0.08 day−1 and 0.04 day−1. The comparisons detailed above are an indication that sulfate reduction was enhanced by the addition of acetate; therefore acetate oxidation was at least in part coupled to sulfate reduction. In addition, previous studies in this aquifer have shown isotopic evidence that sulfate reduction measured during PPTs was attributable to microbial activity [32,36].
Stoichiometric calculations indicated that 0.73–0.78 mol SO2−4 were consumed per mol of acetate. If we assume that all the SO2−4 consumed was used to mineralize acetate at the theoretical 1:1 stoichiometric ratio, about one quarter of the acetate was degraded by other processes. However, we cannot disregard the possibility that SO2−4 was also used as electron acceptor for oxidation of other organic substrates present in the aquifer (e.g., petroleum hydrocarbons). In that case, the proportion of acetate degraded by other processes would be even higher; the latter appears corroborated by our isotopic analyses. The increase of δ13C in DIC and CH4 indicates that acetate was degraded concomitantly by at least two processes: (a) it was completely oxidized to CO2 (presumably by SRB) and (b) it was reduced to CH4. A mass balance of the 13C content of the catabolic products of acetate revealed that approximately 43% of the absolute amount of 13C derived from acetate degradation went to DIC and 57% to CH4. The rapid and high 13C-enrichment in CH4 suggests that CH4 is directly derived from the methyl group of acetate (which bears the label) and not from reduction of labeled CO2. Therefore, we may speculate that a substantial part of the acetate was degraded by acetoclastic methanogenesis.
The two above-mentioned processes of acetate mineralization (oxidation and acetoclastic methanogenesis) were detected previously in this well, where stoichiometric calculations suggested that SO2−4 reduction accounted for only 29% of acetate degradation. Although the relative contribution of these processes is somewhat different in our study, the simultaneous consumption of SO2−4 and 13C-enrichment of DIC and CH4 provides unequivocal evidence that acetate oxidation coupled with sulfate reduction and acetoclastic methanogenesis occurred simultaneously at high concentrations of both electron donor and SO2−4, at least during the short period of time that this experiment lasted. Our results thus add to a growing body of geochemical and microbiological evidence that several terminal electron-acceptor processes may occur in spatial proximity in a bulk volume of aquifer material [37–40].
4.2Microbial community structure and its temporal stability
The addition of acetate in concentrations higher than normally measured for this environment could lead to an enrichment of microorganisms adapted to fast growth on acetate. Such enrichment would lead to a shift in the indigenous microbial community. However, PLFA patterns and FISH proportions in both attached and suspended bacteria remained almost identical from the beginning to the end of our experiment. This again shows, as in our previous studies [7,32], that there is no detectable change in the microbial community composition during the length of the experiment, and that the activities measured thus correspond to the potential activity of the microbial community present at that moment in the vicinity of the monitoring well.
We detected high percentages of gram-negative SRB and archaeal communities. Detection of Desulfotomaculum species was not attempted with probe S-G-Dtm-02929-a-A-18, because it cannot detect Desulfotomaculum acetoxidans, the only Desulfotomaculum species that could be suspected of acetate consumption in this well. Furthermore, using that probe we were not able to detect any Desulfotomaculum in groundwater from this aquifer in a previous study. The decrease in detectability of DAPI-stained cells in some samples (Table 2) could be attributed to the difficulty of visualizing the cells in the samples. This was due to the high clay content in both groundwater and sediment samples, which interfered with the detection of microorganisms at the wavelength used to observe DAPI-stained cells.
4.3Incorporation of 13C into PLFA
In this study, the enrichment of PLFA extracted from groundwater samples varied from 0‰ to 100‰. The enrichment obtained is in the expected range, taking into account the results we obtained in our previous study (where we observed enrichments of up to 5000‰ in some PLFA), the amount of label used, and the characteristics of the processes targeted in the present study. In our previous experiment ∼25% of the acetate-C was 13C in comparison to the present experiment with only ∼10% as 13C. Furthermore, nitrate reduction is energetically more favorable than sulfate reduction and therefore SRB incorporate lower relative amounts of carbon into their biomass. Growth yield on toluene, for example, is three times higher in Pseudomonas sp. strain K172 (a nitrate-reducing toluene-degrader) with a yield of 57 g cell mol−1 toluene than in Desulfobacula toluolica PRTOL1 (a sulfate-reducing toluene-degrader) with a yield of 19 g cells mol−1 toluene.
The pattern of PLFAs that were labeled is different to any known acetate-oxidizing SRB PLFA profile. However, the specific labeling of just some fatty acids suggests that only a few genera are involved in the degradation of acetate. One of the PLFA that incorporated a high proportion of the label, 16:1ω7c, is an important PLFA in most acetate-oxidizing SRB and is considered biomarker for Desulfotomaculum acetoxidans. In addition, D. acetoxidans grown on acetate produces almost exclusively even-carbon straight-chain PLFA with 16:1ω7c and 18:1ω7c as the major monoenoic acids. In our experiment, enrichment of monoenoic even-chain PLFA accounts for approximately 50% of total enrichment. A similar labeling pattern was found by Boschker et al. after incubating estuarine sediments with 13C-labeled acetate, and they concluded that the main acetate degrader in that sediment was D. acetoxidans. Hence, in our experiment, D. acetoxidans may have been an important acetate degrader.
On the other hand, 10Me16:0 that was also enriched and cy17:0 that was enriched to a lesser degree have not been detected in any Desulfotomaculum species, but have been found instead in high amounts in Desulfobacter species. Furthermore, other branched uneven-carbon PLFA, such as i15:0 and a15:0, that were also labeled in our experiment can account for 2 to 11% of PLFA in Desulfobacter grown in acetate but are not important in D. acetoxidans[11,44]. Hence, a 13C-enrichment in those PLFA together with a strong presence of Desulfobacter detected by FISH suggests that among the suspended populations in this well also Desulfobacter utilizes the labeled acetate. In summary, our observations suggest that among the suspended communities of this well both Desulfobacter and D. acetoxidans play an important role in acetate degradation.
In sediment samples, the 13C-incorporation was an order of magnitude lower than in groundwater samples (compare Fig. 5a and b). This could be explained if we consider that the total C present in attached bacteria is higher than the total C present in suspended bacteria in the same aquifer volume. Therefore, the label incorporated in the PLFA of attached biomass is diluted with more unlabeled C. Unfortunately, a precise estimate of relative contributions of suspended and attached communities to acetate degradation is not possible with the obtained data due to the experimental design. Nevertheless, both suspended and attached communities were active and degraded acetate. This contradicts previous observations by Alfreider et al., who did not detect significant activity in suspended communities. However, their study was conducted in pristine aquifers with extremely low nutrient content. Other authors instead have found that suspended communities are denser and more active in contaminated aquifers than in pristine groundwater, with suspended communities ranging from 1 to 50% of the total aquifer community [46,47].
The enrichment pattern in sediment-derived PLFA indicates activity of only D. acetoxidans, because of its similarity with the results obtained by Boschker et al.. In spite of the high abundance of Desulfobacter-like bacteria detected with FISH (∼10% of total DAPI-stained bacteria), the latter group was probably not degrading a substantial amount of acetate, because PLFA biomarkers for gram-negative SRB were not significantly enriched. Another reason for the failure in detecting enrichment in Desulfobacter fatty acids could be that the recovery efficiency of PLFA from sediments is not high enough (expected efficiency is 93 ± 5%), and with a low biomass sample some minor PLFA are likely not recovered. This could be avoided in future studies by extracting PLFA from bigger sediment samples.
In a previous microcosm study performed with sediment from the same aquifer, a Desulfobacter-like population increased in microcosms amended with acetate. This was not the case in our field experiment, neither in the suspended nor the attached community. In the field both populations Desulfobacter and D. acetoxidans, competed effectively for acetate in the suspended microbial fraction at least during the length of the experiment. In contradiction, in the attached community D. acetoxidans appeared to dominate acetate degradation.
As observed from the isotopic data, an important part of the acetate was rapidly transformed to CH4. This shows that the high percentage of Archaea detected with FISH (roughly 30% of total DAPI counts) were actively degrading acetate. One of the main populations of methanogens degrading acetate in this well could be Methanosaeta concilii, which is a known acetoclastic methanogen and was previously detected as an important component of the archaeal community in this aquifer [19,49].