Bioaugmentation of Syntrophic Acetate-Oxidizing Culture in Biogas Reactors Exposed to Increasing Levels of Ammonia

Abstract

The importance of syntrophic acetate oxidation for process stability in methanogenic systems operating at high ammonia concentrations has previously been emphasized. In this study we investigated bioaugmentation of syntrophic acetate-oxidizing (SAO) cultures as a possible method for decreasing the adaptation period of biogas reactors operating at gradually increased ammonia concentrations (1.5 to 11 g NH₄⁺-N/liter). Whole stillage and cattle manure were codigested semicontinuously for about 460 days in four mesophilic anaerobic laboratory-scale reactors, and a fixed volume of SAO culture was added daily to two of the reactors. Reactor performance was evaluated in terms of biogas productivity, methane content, pH, alkalinity, and volatile fatty acid (VFA) content. The decomposition pathway of acetate was analyzed by isotopic tracer experiments, and population dynamics were monitored by quantitative PCR analyses. A shift in dominance from aceticlastic methanogenesis to SAO occurred simultaneously in all reactors, indicating no influence by bioaugmentation on the prevailing pathway. Higher abundances of Clostridium ultunense and Tepidanaerobacter acetatoxydans were associated with bioaugmentation, but no influence on Syntrophaceticus schinkii or the methanogenic population was distinguished. Overloading or accumulation of VFA did not cause notable dynamic effects on the population. Instead, the ammonia concentration had a substantial impact on the abundance level of the microorganisms surveyed. The addition of SAO culture did not affect process performance or stability against ammonia inhibition, and all four reactors deteriorated at high ammonia concentrations. Consequently, these findings further demonstrate the strong influence of ammonia on the methane-producing consortia and on the representative methanization pathway in mesophilic biogas reactors.

INTRODUCTION

Biogas has promising potential as a substitute for fossil fuel and could contribute to the reduction of greenhouse gas emissions and global warming. Methane is the energy-rich component of biogas and is formed as the end product during anaerobic degradation of organic material. In addition to valuable production of renewable energy, anaerobic degradation also represents a suitable method for waste and wastewater treatment and may bring indirect environmental benefits such as reduced spontaneous emissions of ammonia and methane otherwise occurring during composting or storage of untreated animal manure (7, 8). Furthermore, utilization of the residue as a crop fertilizer contributes to recirculation of nutrients, lowering the use of additional mineral nitrogen fertilizers (5). Among many different materials that can be used for biogas production, protein-rich materials such as slaughterhouse waste, animal manure, and distiller's waste are highly interesting due to their high methane potential. Furthermore, the digestion residues contain large amounts of plant-available ammonia, thus representing a valuable fertilizer. However, the high content of ammonia is a concern due to its association with unstable process performance and increased risk of process failure (10). The potential inhibitory effects on the biogas-producing consortia and in particular the acetate-utilizing methanogens are considered to be the main causes of decline in process performance (21, 35).

In several studies, investigators have considered strategies for improving process operations with protein materials, e.g., dilution of substrate, use of additives, change of operational temperature, etc. (2, 3, 14, 20). In addition, acclimatization has been reported to increase tolerance and retain microbial viability at ammonia concentrations far exceeding the initial inhibitory concentrations (4, 21, 22, 43). One possible explanation for this acclimatization effect is enhanced growth of ammonia-tolerant microorganisms able to produce methane through syntrophic acetate oxidation (SAO) (39). Methanogenesis through SAO involves a two-step reaction consisting of acetate oxidation to hydrogen and carbon dioxide by syntrophic acetate-oxidizing bacteria (SAOB), followed by conversion of these products to methane by hydrogen-utilizing methanogens (44). In addition to elevated levels of ammonia, high operation temperatures, high acetate concentrations, long retention times, and absence of the aceticlastic Methanosaetaceae are factors suggested to have an impact on the dynamic transition from aceticlastic methanogenesis to SAO (1, 15, 18, 26, 34).

The conversion from aceticlastic methanogenesis to SAO may comprise a decline in gas production and methane yield (31). However, as the shift may involve development of an ammonia-tolerant biogas-producing community, the process can continue even at ammonia levels reported as being inhibitory for the aceticlastic methanogens (42).

Bioaugmentation, in terms of adding specific microorganisms or enriched consortia to anaerobic processes to enhance a desired activity, has been reported to lead to improvements in degradation of specific organic compounds (9, 13, 16), such as cellulose-containing biomass (6) and manure (25), start-up of new reactors (28), odor reduction (12, 38), and recovery after organic overload (37) and toxicant exposure (30) at a laboratory scale. Furthermore, improved reactor performance in terms of increased methane production and decreased accumulation of volatile fatty acids has been observed after the addition of hydrogen-utilizing methanogens to mesophilic (26 to 35°C) reactors degrading distillery wastewater (29). To our knowledge, bioaugmentation with the intention to improve reactor operation at high ammonia concentrations has not been assessed previously.

Thus, the aim of this study was to examine bioaugmentation of the natural biogas-producing consortia with syntrophic acetate oxidizers as a method for decreasing the adaptation period of biogas reactors operating at gradually increased ammonia concentrations. The influence of bioaugmentation and potential associations between the acetate conversion pathway, microbial abundance dynamics, and operational parameters and process performances were analyzed.

MATERIALS AND METHODS

Culture used for bioaugmentation.

The syntrophic acetate-oxidizing (SAO) cultures used for bioaugmentation included SAOB isolated at the Department of Microbiology, Swedish University of Agricultural Sciences. These were Clostridium ultunense sp. Esp JCM16670 (41), Syntrophaceticus schinkii JCM16669 (41), Tepidanaerobacter acetatoxydans DSM 21804 (42), and the hydrogen-utilizing methanogen Methanoculleus sp. strain MAB1 (33). Cultivation was conducted in modified bicarbonate-buffered basal medium (41) containing 0.2 M NH₄Cl and 100 mM acetate, added on two occasions. Complete degradation of acetate was confirmed before addition to the reactors.

The average gene abundances of the SAOB in the SAO culture were determined to be about 2.5 ± 0.3 × 10⁷, 3 ± 0.5 × 10⁹, and 2.7 ± 1.1 × 10¹⁰ per ml for C. ultunense, S. schinkii, and T. acetatoxydans, respectively.

Reactor operation and substrate.

Four identical laboratory-scale continuously stirred tank reactors (Belach Bioteknik, Stockholm, Sweden), designated E1, E2, R1, and R2 (E, experimental; R, reference) and with a working volume of 5 liters, were operated for 458 days under mesophilic conditions (37°C). The reactors operated semicontinuously and were fed 6 days a week with a mixture of cattle manure and whole stillage produced from fermentation of cereals at an ethanol production plant. The inoculum used for setting up the processes, the origin and collection of substrates, and the total solids (TS), volatile solids (VS), carbon/nitrogen (C/N) ratios, chemical characteristics, and methane potentials of the whole stillage and the manure are specified by Westerholm et al. (40). The substrate mixtures varied during the experimental period and consisted of 16% manure and 84% whole stillage (based on total VS) during the first 375 days. Between day 375 and 409, some of the whole stillage was replaced with egg albumin powder (Källbergs Industries, Sweden). The resulting substrate mixture consisted of 16% manure, 59% whole stillage, and 25% egg albumin (based on total VS). In the final operating period, i.e., from day 410, whole stillage was completely omitted from the substrate and exchanged for egg albumin. At this stage the mixture consisted of 16% manure and 84% egg albumin based on total VS. Reactor performance was monitored regularly. Measurements of volumes and compositions of the gases produced were determined daily, whereas pH was monitored once a week. Volatile fatty acid (VFA) concentrations, bicarbonate alkalinity, and ammonium-nitrogen were measured continuously. Initially, the reactors operated at a hydraulic retention time (HRT) of 57 days and an organic loading rate (OLR) of 0.8 g VS/day for about 150 days, but later the operation was changed in order to increase the ammonium level in the reactors (Fig. 1).

Fig 1.

OLR (A) and HRT (B) of the reference (R) and experimental (E) reactors throughout the operation period of about 460 days.

Bioaugmentation of experimental reactors E1 and E2 with the SAO cultures was initiated on day 150, concurrently with the gradual increase in loading rates. A 10-ml portion of the culture was added daily immediately before addition of the substrate. In order to compensate for any improvement in process performance due to the addition of medium components to the experimental reactors, 10 ml of sterile medium was added to the reference reactors R1 and R2.

Analytical and tracer analyses.

Analyses of the VFA composition, TS, VS, C/N ratio, bicarbonate alkalinity, trace element concentrations, total gas production, and the methane and carbon dioxide content of the gases were performed as described previously (40). Ammonium-nitrogen was analyzed as described by Westerholm et al. (40) or by Tekniska Verken i Linköping AB according to FOSS Tecator application sub note 3502 with a Kjeltec 8200 auto distillation unit (FOSS, Scandinavia, Sweden). The pathway of acetate degradation to methane was determined by tracer analysis involving incubation of reactor sludge (20 ml) with [2-¹⁴C]acetate (final concentration, 0.11 μCi/ml) and monitoring of labeled gases by scintillation counting as described by Schnürer and Nordberg (31). A ¹⁴CO₂/¹⁴CH₄ ratio above 1 indicates dominance of syntrophic acetate oxidation, while aceticlastic methanogenesis is the main pathway at ratios below 1 (31).

Molecular analysis.

Samples for molecular analyses were withdrawn from the reactors at different time points throughout the operating period and stored at −20°C until analysis. Extraction of total genomic DNA, construction of DNA standards, and performance of the quantitative PCR (qPCR) analysis were conducted as reported elsewhere (39). For the quantification of Thermacetogenium phaeum, the primer pair TH795f/985r was used (L. Sun, B. Müller, M. Westerholm, and A. Schnürer, unpublished data). To target the methanogen MAB species, the primers MAB62f (5′-GGAATGCCCTGTAATCCAAA-3′) and MAB301r (5′-CACCTGAACAGCCTGCATT-3′) were developed as described by Westerholm et al. (39). This primer pair targeted 16S rRNA genes of Methanoculleus bourgensis and Methanoculleus sp. strains MAB1, MAB2, MAB3, and BA1 (33). To control the specificity, the new primers were used in PCR with genomic DNA from the closely related methanogens Methanoculleus submarinus DSM 15122, Methanoculleus palmolei DSM 4273, and Methanoculleus chikugoensis DSM 13459. The quantifications of Methanomicrobiales, Methanobacteriales, Methanosaetaceae, and Methanosarcinaceae were conducted with the primer sets and qPCR protocols assigned by Westerholm et al. (40). The mean similarities of gene abundances were compared using Student's t test, and a P value of <0.05 was regarded as indicating a statistically significant difference. Triplicate samples from each reactor and sampling point were included in the analyses.

RESULTS AND DISCUSSION

Reactor performance.

According to the chemical analyses, all the reactors operated with similar performances throughout the operating period (Fig. 1, 2, and 3 and Table 1). The performances were initially stable (days 0 to 150) but later involved two instability periods (around days 230 to 300 and days 400 to 458). In the initial operational period (<150 days), the four reactors operated at an OLR of 0.8 g VS/day and an HRT of 57 days (Fig. 1). The gas produced was composed of about 61 to 66% CH₄, the specific methane production varied between 285 and 343 ml CH₄/g VS, and the alkalinity and the pH in the digester sludge varied between 7.5 and 10 g CaCO₃/liter and 7.5 to 7.8, respectively. Ammonium-nitrogen was found to be 1.5 g NH₄⁺-N/liter (107 mM) in all the reactors. The VFA levels were low, illustrating that the performance was stable (Fig. 3). A gradual increase in OLR and simultaneous decrease in HRT, starting at day 150, to 3.6 g VS/day and 26 days, initially resulted in an increased gas production but later in reactor instability, as indicated by a rapid decrease in methane yield (Fig. 2) and pH (to about 6.5 to 7) and increased VFA (Fig. 3). The operational instability was most likely caused by the increase in OLR and decrease in HRT, combined with the relatively high concentration of ammonia-nitrogen (about 3.0 g NH₄⁺-N/liter; 214 mM). By decreasing the OLR during the following period (days 244 to 305), it was possible to stabilize the reactor operation, as indicated by a reduction in the acetic acid concentration and a recovery of the methane yield. However, the levels of propionic acid did not decrease and at day 301 concentrations of about 50 to 60 mM and 4 to 26 mM were established in the experimental and reference reactors, respectively. After 305 days of operation OLR was once again increased, but at this point with extended sludge retention time (HRT 41 to 45 days). The longer HRT increased the degree of mineralization, which was reflected in increased ammonium-nitrogen concentrations (Table 1). After the increase in OLR, the specific methane yield stabilized at lower levels than those during the initial period and varied between 164 and 204 ml CH₄/g VS between days 367 and 400. A change in feed composition, with retained OLR, performed by exchanging part of the whole stillage for egg albumin powder (day 375) and by the use of manure and egg albumin powder as the sole substrates (from day 410) further increased the ammonium-nitrogen content in the reactors, which reached levels around 11 g NH₄⁺-N/liter (790 mM) at day 449. This operation resulted in a complete failure of the process, with decreasing methane yield and increasing VFA levels. At this point, the high ammonia level was the likely factor behind the instability, especially since the reactors operated with comparatively lower OLR and longer HRT. This result is in line with several previous studies illustrating process instability as a consequence of increasing ammonia levels (10). The acetic acid concentrations in the reactors were relatively stable between days 310 and 392, 11 to 61 mM, but rapidly increased to 247 to 274 mM at the final sampling point. In contrast to acetic acid concentrations, the propionic acid concentrations showed no tendency toward decrement between these two instability periods. Instead, the concentrations continued to increase and finally reached relatively high levels at day 442 (120 to 136 mM). Worth considering in this regard is that during this period, higher levels of propionic acid were reached earlier in the experimental reactors than in the reference reactors.

Fig 2.

Methane yield of the reference and experimental reactors.

Fig 3.

Concentrations of acetic acid (left) and propionic acid (right) in the reference and experimental reactors.

Table 1.

Average ammonium-nitrogen concentrations in the reference and experimental reactors

Day of operating period	NH4+ (mM)	NH3 (mM)
76	107 ± 3	4.1 ± 0.4
170	76.5 ± 0.4	2.9 ± 0.3
185	170 ± 4	8.5 ± 0.2
245	214 ± 9	11 ± 2
295	236 ± 12	9.7 ± 3
350	278 ± 6	19 ± 9
385	336 ± 5	20 ± 2
449	790 ± 48	30 ± 10

Methanogenic pathway.

The tracer analyses showed relatively similar results for all four reactors, and the labeling recovery on the different sampling occasions was between 93 and 110%. At day 138, the average ¹⁴CO₂/¹⁴CH₄ ratio in the reactors was 1.2 ± 0.3, indicating activities of both SAO and aceticlastic methanogenesis in the operational phase preceding the first increase in OLR. The presence of SAO at this stage was somewhat surprising due to the low levels of ammonia and acetate but could possibly be explained by the prevailing long HRT (57 days). This result confirms claims by Shigematsu et al. (34) that SAO is the primary pathway at low dilution rates, whereas the aceticlastic pathway dominates at higher dilution rates. In agreement with these results, aceticlastic methanogenesis in the present study became dominant when HRT decreased, which was seen at day 241 when a ¹⁴CO₂/¹⁴CH₄ ratio of 0.3 ± 0.1 was determined. One possible explanation for these findings is that a relatively short HRT constrains the growth and activity of mesophilic SAOB. At day 374, the monitored ratios were distinctly higher, 1.8 ± 0.4, reflecting a change in the primary pathway for acetate degradation to SAO in both the experimental and reference reactors. This shift in pathways confirms findings in a previous study in which SAO was activated at ammonia levels exceeding 3 g NH₄⁺-N/liter (31). The inhibition of aceticlastic methanogens by ammonia probably gives ammonia-tolerant syntrophic acetate oxidizers a competitive advantage (32, 41, 42). Otherwise, these organisms are considered to be less competitive for acetate than the aceticlastic methanogens (27).

As the results were equal in all digesters, the bioaugmentation of the SAO cultures in the biogas reactors seemed to have no influence on the dominant pathway for acetate degradation.

Molecular analysis.

All standard dilution series provided high correlation coefficients (>0.99), similar calibration slopes (between −3.3 and −4.0), and qPCR efficiencies above 86%. The quantitative values of 16S rRNA genes affiliated with the SAOB and the methanogens are displayed in Fig. 4.

Fig 4.

Average log gene abundance of SAOB (A to C) and methanogens (D to H) in reference (dashed line) and experimental (continuous line) reactors.

Abundances of SAOB.

The gene abundances of the different SAOB were below the detection limits in the substrates used for feeding the reactors, and T. phaeum was not detected in any of the reactor samples. The reason for the absence or low abundance of T. phaeum is probably that the mesophilic conditions in the reactor did not support this thermophilic bacterium, which requires a temperature range of 40 to 65°C for growth.

The average gene abundance of S. schinkii was initially (days 0 to 266) between 0.6 × 10⁷ and 1.9 × 10⁸/ml but increased to 6.1 × 10⁸ to 1.4 × 10¹⁰/ml after about 300 days of operation. The relatively similar abundances in both the reference (R) and experimental (E) reactors throughout the operational period illustrated that at this level this species was evidently unaffected by bioaugmentation. In contrast, the quantitative assessment of SAOB in the reactors established distinctly higher levels of C. ultunense and T. acetatoxydans in the experimental reactors than in the reference reactors in the operational phase following bioaugmentation (day 150 to 400) (Fig. 4B and C). The C. ultunense-related species were not abundant above the detection limits in the reference reactors before day 250 (except in R1 at day 153), but were prominent in reactors E1 and E2 just after the initiation of bioaugmentation (>150 days of operation), with average gene abundances around 0.7 to 1.4 × 10⁵/ml. In the reference reactors, C. ultunense was detected first at day 301 (average gene abundance, 4.7 × 10²/ml) in R1 and day 392 (average gene abundance, 1.3 × 10⁶/ml) in R2. In the experimental reactors, T. acetatoxydans had already been distinguished before bioaugmentation was initiated (average gene abundance, 1.5 × 10⁴ to 2.0 × 10⁶/ml) but was not detected in R1 and R2. However, the average gene abundances increased significantly (P < 0.05) in the E1 and E2 reactors after day 212 and then varied between 1.1 × 10⁵ and 1.2 × 10⁸/ml. In R1 and R2, the species was first detected only after 300 days of operation. These findings indicate that bioaugmentation might be considered successful for C. ultunense and T. acetatoxydans. However, considering the expected theoretical levels of SAOB, based on the amount of added microorganisms and the dilution rates, it appears that the organisms did not become established and grow in the process during the first 250 days of operation. The levels from the addition were lower than expected, possibly suggesting that the bacteria were broken down to some degree.

Remarkably, a significant increase (P < 0.05) in genes affiliated with all mesophilic and thermotolerant SAOB occurred in all the reactors, both the experimental and the reference reactors, after approximately 250 to 300 days of operation, i.e., during the first instability period. At this stage of the operation, the ammonium-nitrogen concentration had reached 3 g NH₄⁺-N/liter (215 mM), a level previously shown to induce increased numbers of SAOB (39). The selective pressure of ammonia in combination with the lower HRT at this time probably allowed the establishment of the relatively slow-growing SAOB. The introduction of SAO was also further supported by the tracer analysis, illustrating dominance of SAO after 374 days of operation. The levels of SAOB in the experimental reactors increased by about two log units, concurrently with the increase in abundance in the reference reactors. Thus, at the end of the operating period, the abundances of C. ultunense, T. acetatoxydans, and S. schinkii were at equal levels in the reference reactors and in the experimental reactors. Apparently, these species were present in the nonbioaugmented reactors and were able to grow and become established as important components in the microbial community under appropriate operational conditions. Possibly, in the early phase of operation, when aceticlastic methanogenesis was still predominant, the SAOB grew with organic compounds other than acetate, like amino acids, alcohols, and sugars, and produced acetate as the main degradation product (41, 42). However, the low abundances of these organisms (below the detection limits) indicate inferior competitiveness for these substrates compared to that of other, likely more efficient, fermentative bacteria. Alternatively, the SAOB performed acetate oxidation but at very low activity levels. Interestingly, the increase of C. ultunense and T. acetatoxydans to the final levels occurred earlier in the experimental reactors than in the reference reactors, probably due to the relatively higher abundances at the starting points for the increase (day 250; Fig. 4B and C). This suggests that the added organisms to some extent actually had become established in the process and, when conditions allowed, quickly increased in number. However, this rapid increase in abundances did not seem to influence the performance of the experimental reactors. Possibly, the slightly higher levels of propionic acid in E1 and E2 may have been associated with the higher amounts of SAOB observed in these bioaugmented reactors.

Despite this initial divergence in abundances between the reactors, the levels of SAOB were similar in all of the reactors at the final sampling points. This indicates a strong dependence on certain operating conditions, most likely high acetate concentration, high ammonia concentration, and/or the methanogenic population structure for the abundances of the different SAOB. The findings in the present study together with results from recent publications indicate that the levels of specific abundant SAOB are indicative of the prevalent pathway for the methanization of acetate. Corresponding levels of SAOB-related species have been revealed in studies of several laboratory- and full-scale biogas reactors operating through SAO or the aceticlastic pathway (19, 39; L. Sun, B. Müller, M. Westerholm, and A. Schnürer, unpublished data). However, to fully establish specific levels of SAOB gene abundance as likely indicators for SAO, further studies are required.

Abundance of methanogens.

Examination of the methanogenic populations (Methanosaetaceae, Methanosarcinaceae, and Methanobacteriales) revealed similar abundances in both the experimental and reference reactors. Despite changes in operational parameters and reactor performances in the present study, species belonging to the families Methanosaetaceae and Methanosarcinaceae remained at relatively constant levels during the initial 200 days of operation. This is in accordance with an earlier study, where apparently stable methanogenic abundance was demonstrated in biogas reactors despite altered OLR, HRT, and substrate composition (40). In addition, the results of the present study showed that the addition of the acetate-oxidizing culture did not cause observable influences on the abundances of Methanosaetaceae and Methanosarcinaceae. However, the abundance of Methanosaetaceae decreased significantly (P < 0.05) concurrently with the increase in SAOB, with average gene abundances of 1.4 to 3.7 × 10¹⁰/ml (day 212) to 0.7 to 2.3 × 10⁸/ml during the late stage of operation (days 345 to 445). The levels of Methanosarcinaceae were more stable, but a small decrease was observed from 4.6 to 6.5 × 10⁵ at day 63 to 0.8 to 2.3 × 10⁴ at day 392. However, the levels again increased to 0.7 to 3.3 × 10⁵/ml at the final sampling point at day 442. This pattern in abundance is likely to be attributable to the elevated levels of ammonia, causing specific inhibition of the aceticlastic methanogens (39). Furthermore, members of the Methanosaetaceae have been proven to be more sensitive to ammonia than members of the Methanosarcinaceae (17, 24). Ammonia concentrations exceeding 3 g NH₄⁺-N/liter (215 mM) have been reported to inhibit Methanosaeta species, whereas Methanosarcina species have been shown to tolerate concentrations up to 7 g NH₄⁺-N/liter (500 mM) (11).

Interestingly, the relatively high abundances of Methanosarcinaceae coincided with high abundances of the SAOB, in particular S. schinkii and T. acetatoxydans, even after inception of SAO as the dominant pathway for methane production. Since both these groups use acetate for growth, they are probably strongly competitive with one another, which was not reflected in the present results. However, there have been recent indications that members belonging to the order Methanosarcinales may act as hydrogen-utilizing methanogens during SAO (19). Molecular studies of the abundant methanogenic population in biogas reactors, with SAO as the main mechanism for methane formation, revealed that more than 90% of the methanogenic population consisted of Methanosarcinales. In addition, it has been suggested that Methanosarcinaceae may be able to perform acetate oxidation (23), indicating that these methanogens could mediate the entire process of acetate oxidation and subsequent methanogenesis (18).

The abundance of the hydrogenotrophic methanogens belonging to Methanomicrobiales, Methanobacteriales, and Methanoculleus species fluctuated somewhat during the operating period, but overall no significant changes were observed between the reactors (Fig. 4D to F). An initial increase in the abundances of Methanoculleus species at day 183 indicated that the establishment of these methanogens may have occurred by the time of the initiation of bioaugmentation, but this was contradicted by the rapid decrease at day 212. After day 212 a more persistent increase in abundances occurred, and at day 253 the levels had increased significantly by about one to two log units (Fig. 4F). From day 392, the abundances seemed to decrease rapidly, but the changes were calculated to be insignificant (P > 0.05). These changes were also reflected in the number of abundant Methanomicrobiales, comprising hydrogenotrophic methanogens, among other species belonging to the genus Methanoculleus. Comparing the gene abundances of Methanomicrobiales and Methanoculleus suggested a proportional increase in the latter in line with the increase in ammonia levels. This result is in line with results of a previous study, which suggested that Methanoculleus is an important partner organism during SAO under mesophilic conditions (33).

The similar performances of the reactors showed that the bioaugmentation did not improve their operations during periods of increasing ammonia levels. Apparently the positive effects of bioaugmentation illustrated in previous studies were not applicable in this experiment. The fact that the bioaugmentation here was performed in a continuous stirred tank reactor (CSTR) while the majority of previous studies were performed with batch reactor systems might be one explanation for the difference in results. However, in accordance with these results, a review regarding bioaugmentation reported low or no benefit of bacterial augmentation, biomass enhancement, or inoculum addition (36). A variety of reasons for unsuccessful application were suggested, such as limitations of available substrate, overrestriction of the amounts of added microorganisms, and growth inhibition due to competition with other microorganisms. These factors may also be applicable here as potential reasons for the absence of apparent effects of bioaugmentation with SAO culture. However, an investigation of two-phase anaerobic reactors suggested that a short retention time and acidic conditions could favor SAO in the first-stage reactor. The constant input of SAOB (affiliated with C. ultunense, T. phaeum, and S. schinkii) from the first-stage reactor was then proposed to promote the prevalence of SAO in the second-stage reactor and to allow the bacteria to compete successfully with the aceticlastic methanogens (32).

Conclusions.

This study further emphasizes the strong impact of ammonia on the occurrence of SAO and the increased appearance of SAOB and the partner hydrogenotrophic methanogens. Previous studies have revealed a significant increase in the abundance of SAOB concurrently with an ammonia-induced shift from aceticlastic methanogenesis to SAO in a laboratory-scale reactor (39). Furthermore, the lack of influence of the syntrophs that were added to the pathway for acetate degradation in the reactors and the identification of ammonia as the deciding factor for the appearance of SAOB indicate that only microbes with the ability to thrive and proliferate in the reactor will have any impact on the process in a longer-term perspective. This also highlights the need to establish optimized growth conditions for SAOB and the hydrogenotrophic partner in order to optimize methane production at high ammonia concentrations. However, information about syntrophic acetate oxidation, the organisms responsible, and their role in the methanogenic environment is currently limited. A greater understanding of their response to different environmental conditions would assist further development and optimization of the anaerobic treatment processes proceeding through the SAO pathway.

The results of the present study also indicate that the SAOB known today are important components of the biogas-producing microbiota and are abundant, although at relatively low levels, even in processes dominated by aceticlastic methanogenesis. High abundances and supplies of SAO microorganisms are not crucial factors for the onset of the SAO pathway. Instead, operation under certain environmental conditions such as high ammonia levels promotes the dynamic transition to SAO.