Biogas Emission from an Anaerobic Reactor

Anaerobic denitrification is accompanied by biogas emissions, which may orient greenhouse gases or air pollutants. Ferrous iron play an important role in wastewater treatment. However, the effect of ferrous ion on biogas emission and the relationship between microbial community and nitrogen removal performance are not fully understood, especially in upflow anaerobic sludge blanket (UASB) reactors. Here, the results revealed that nitrogen gas increased rapidly with Fe(II) addition. And ferrous ions enhanced denitrification and COD removal efficiency with more than 60% and 10%, respectively. Moreover, Ferrous ions addition evidently increased Hydrogenophaga, Methylotenera, Zoogloea and Fluviicola during the whole UASB reactor operation progress. The correlation coefficient analysis further confirmed that Fe(II) was positively correlated with the abundances of Hydrogenophaga, Methylotenera and Fluviicola. Based on these results of chemicals transformation kinetics, microbial community and the correlations coefficient analysis, a hypothetical mechanism is proposed: In the UASB system with Fe(II) and NH4, firstly Paludibacter made a contribution to NH4 oxidation and generated NO3, and then Hydrogenophaga, played an important role in NO3 reduction, Zoogloea coupled NO3 reduction to Fe(II) oxidation, Fluviicola combined with Methylotenera conducted NO2 reduction together with Fe(II) oxidation. This study will improve our understanding of ferrous ion’s influence on biogas emission, denitrification process and corresponding microbial community.


INTRODUCTION
Up-flow anaerobic sludge blanket (UASB) reactors are commonly used in municipal wastewater treatment, because the reactor with small occupied area, low sludge generation, low cost of investment and maintenance (Powar et al., 2013).The effluent of UASB reactor is required to meet the effluent standards, especially in terms of nutrients.In general, nitrogen removal via aerobic nitrification and anaerobic denitrification by bacteria (Kartal et al., 2010).Generally, nitrogen removal efficiency in UASB reactors does not reach the requirement.In order to develop fast and effective approaches for nitrogen removal, many investigations are performed to unveil the influence of operational parameters on nitrogen removal efficiency and microbial communities (Kim et al., 2011;Reddy et al., 2017).Recently, more and more research points to greenhouse gas emission and benefits of biogas electricity generation during anaerobic sludge digestion (Yang et al., 2017;Gingerich and Mauter, 2018).
Iron is the most abundant transition metal element on the earth (Kappler et al., 2005;Li et al., 2016), and it is also the necessary trace element for microorganisms.In recent years, it was found that iron strongly influence many nutrients and contaminants removal performance and their degradation products (Lalonde et al., 2012;Melton et al., 2014).Ferrous iron could improve chemical oxygen demand (COD) and nonbiodegradable contaminant removal efficiency a stable level by adjusting redox potential.In addition, the settle ability of microorganisms was also enhanced (Chen et al., 2007;Vlyssides et al., 2009).However, previous research has been concentrated on the physicochemical changes of reactor for wastewater treatment after catalysts addition (Li et al., 2017;Liu et al., 2017), but the influence on biogas emission and its mechanism for microorganism changes are not clear.As we known, many gases, such as NO x , CO 2 and CH 4 , are released to the atmosphere by biological processes occurring in anaerobic environments (Rajab et al., 2012).Considering their environmental effects, it is important to investigate underlying mechanisms on these biogas emissions.
In this study, after the influences of ferrous iron on the stability of COD and nitrogen removal efficiency were evaluated, the roles of ferrous iron on biogas emission and corresponding microbial community were investigated in UASB reactor across different conditions.Then, potential mechanisms were further proposed based on the observed relationships among microorganisms, biogas emission and ferrous iron.

The UASB Reactor Set up and Operating Condition
The inoculum of UASB reactor was anaerobic sludge obtained from the Wangtang municipal wastewater treatment plant in Anhui province, China.The configuration of UASB reactor was shown in Fig. 1, working volume was 4.5 L, the operating temperature maintained at 35 ± 2°C by a constant temperature chamber.The medium composition was , 2500 mg L -1 glucose was fed as carbon source, the medium pH was adjusted to 6.8-7.2 with NaHCO 3 or hydrochloric acid.FeSO 4 •7H 2 O with 5 mM concentration was added to the influent for one UASB reactor on Day19 until stop on Day 67 and then recovered Fe(II) addition on Day 105.The other UASB reactor as the control system, the influent composition was not changed during the whole operation period without Fe(II) supplement.The whole operating process of UASB reactor was divided four stages, before Fe(II) addition period was named as Con; after Fe(II) addition was named as AdF; Fe(II) addition stopped and pH decreased to acid condition, this period was named as Aci; Fe(II) addition was recovered period was named as Rev.

Chemical Analytical Methods
During the reactor operation, the concentration of COD, Fe 2+ , Fe 3+ , NO 3 -, NO 2 -, NH 4 + in influent and effluent were measured daily.All chemicals were purchased from Sigma-Aldrich (USA), and their analysis carried out in accordance with standard methods (Clesceri et al., 2012).Samples Liquid sample were sampled from different ports and mixed with equal volume.And after centrifuged at 8500 rpm for 20 min, the samples were filtered with mixed cellulose ester membrane (0.22 µm pore size) to remove suspended solids (Liu et al., 2014).NO 3 -and NO 2 -were analyzed with ion chromatography (Dionex ICS-90), ion column was IonPac AS14A 4 × 250 mm, mobile phase was the mixed liquid 8 mM Na 2 CO 3 and 1 mM NaHCO 3 , the flow rate was 1 mL min -1 NH 4 + was measured with spectrophotometry at 420 nm after a colorimetric reaction with Nessler's reagent (Paul et al., 2007).The concentration of Fe 2+ was extracted with 0.5 M HCl and analyzed with 1,10-phenanthroline colorimetric assay as previously described (Li et al., 2010).

DNA Extraction and High Throughput Sequencing
The DNA was extracted from the sludge in UASB using a PowerWater™ DNA Isolation Kit (MO BIO Laboratories, USA) according to the manufacturer's instructions.Amplicon libraries were prepared as previously reported (Fadrosh et al., 2014;Lax et al., 2014).Briefly, the V3-V4 hypervariable region of bacterial and archaeal 16S rRNA genes was amplified using this primer pair 338F and 806R.The primer sequences are 5'-ACTCCTACGGGAGGCAGCA-3' and 5'-GGACTACVSGGGTATCTAAT-3', respectively.In addition, a sample specific 12-bp barcode was added to the reverse primer.According to a primer coverage test using TestPrime 1.0 (Lee et al., 2011;Fadrosh et al., 2014), the primers have high target coverage (> 90%).Polymerase chain reaction (PCR) was conducted as follows: initial denaturation at 95°C for 3 min; 30 cycles of denaturation at 94°C for 30 s, primer annealing at 61°C for 1 min, extension at 72°C for 1 min; and a final extension at 72°C for 10 min.Replicate amplicons were pooled for purification with an AMPure®XP PCR purification kit (Agencourt Bioscience Corp., Beverly, MA).The purified PCR products from each sample and then sequencing on an Illumina HiSeq 2500 platform in the Beijing Genomics Institute in China.

Analysis of Sequencing Data
Raw data was firstly removed ambiguous bases, then combined using the Flash software with default parameters to obtain sequences (Magoc et al., 2013).The obtained sequences were processed using the Quantitative Insights Into Microbial Ecology (QIIME) software pipeline, to remove chimeric and low-quality sequences and assign sequences to individual samples (Caporaso et al., 2010).Operational taxonomic units (OTUs) were identified with 97% similarity (Edgar, 2010), and the representative sequence of each OTU was classified with Ribosomal Database Project (RDP) database (Wang et al., 2007).The alpha diversity of samples was also compared with Shannon index, Chao1 index, and Simpson index.The raw sequences of this study have been deposited in Sequence Read Archive database, the under deposited number was PRJNA347292.

Statistical Analysis
All the experiments in this study were conducted in triplicate, and the results were showed as average value added standard deviation.The correlation coefficients were analyzed by SPSS 18.0.Values of P were determined using Student's t-test.

Ferrous iron Enhanced COD Removal and Affected pH of UASB Reactor
Two UASB reactors were set up to investigate the effect of Fe(II) on COD removal performance.In control reactor, the COD of influent was about 500 mg L -1 , the COD of effluent gradually increased from 70 mg L -1 to 250 mg L -1 , the COD removal efficiency decreased from 90% to 45% and stabilized in this level (Fig. 2(A)), accompanied with COD removal, the pH of effluent decreased from 7.5 to 4.5 (Fig. 2(B)).After 5 mM Fe(II) addition on Day 105, COD removal efficiency gradually recovered to 90%.In experiment reactor, the operating condition was same to control reactor (Con period), 5 mM Fe(II) was added on from Day 19 to Day 67 (AdF period), COD removal efficiency slightly increased from 80% to 88% and the pH was stable before and after Fe(II) addition (Figs.2(C) and 2(D)).After 67 days, Fe(II) no longer added and COD removal efficiency sharply decreased from 88% to 45% companied with pH decreased from 7.2 to 4.9 (Aci stage).Comparable to the results in Rev stage, both the COD removal efficiency and pH gradually increased when Fe(II) addition was recovered.
The operational and environmental variations influence the typical responses include a decrease in performance, the accumulation of volatile fatty acids, and a drop in the pH and alkalinity in anaerobic wastewater treatment systems (Leitão et al., 2006).In this study, the COD removal efficiency declined to below 50% following the pH decrease, especially since pH lower than 6.The lower pH caused microbial activity inhibition and led to bad reactor performance, which was similar to previous studies (Haandel, 1994).The pH could be an indicator of stability or the COD removal efficiency of the UASB.
Ferrous ion could slightly improve COD removal efficiency and maintain the stable reactor performance.Importantly, UASB reactor with low COD removal efficiency and acid pH condition could be recovered to high COD removal performance and normal pH conditions.It has been reported that Fe(II) addition significantly affect reactor performance.Fe(II) addition was an effective method to obtain a high removal rate of nitrobenzene in UASB reactor (Chen et al., 2007).Increasing Fe(II) concentration could decrease the specific activity of sludge granules in UASB reactor (Yu et al., 2000).

Ferrous iron Affected Nitrogen Removal Performance
NO 3 -, NO 2 -, NH 4 + , Fe 3+ and Fe 2+ concentrations were analyzed to explore the relationship between iron addition and nitrogen removal performance in experiment reactor (Fig. 3).The 5 mM Fe(II) in influent was partly oxidized to Fe(III), the concentration of Fe(III) and Fe(II) in effluent was about 3.5 mM and 1.5 mM, respectively.In this study, the influent contained NH 4 + and NO 3 -and NO 2 -, and NH 4 + was the dominant component with concentration about 14 mM, NO 3 -concentration was less 2 mM and NO 2 concentration was less 1 mM.In effluent NH 4 + concentration was below 1 mM, when Fe 2+ was added to influent, NO 3 and NO 2 -concentration was close to influent, almost all the NH 4 + was oxidized to NO 2 -and NO 3 -by nitrification process, and then they were removed by denitrification.When Fe 2+ addition was stopped, NO 3 -and NO 2 -concentration was higher than influent, almost all the NH 4 + was removed, but a part of NO 2 -(22%) and NO 3 -(61%) was failed to remove by denitrification.Whether Fe(II) added or not the NH 4 + in effluent was no observed difference, this indicated that Fe(II) addition does not affect nitrification, but seriously affected denitrification.
The concentration of NO 3 -in effluent was about 6 mM, after Fe(II) addition it decreased to 2 mM similar to the influent level, when stopped Fe(II) addition, NO 3 -in effluent gradually increased to 8 mM, Fe(II) was added again on Day 105, and NO 3 -in effluent gradually decreased and recovered to the influent level.The effect of Fe(II) on NO 2 -removal trend was similar to Fe(III).The bad NO 3 -and NO 2 -removal performance was shown after Fe(II) addition stopped, it was not only affected by Fe(II) but also  related to the low pH, because no Fe(II) addition led to pH decrease to below 6 from Day 89 to Day 105, the acid condition was negative for microbial activity.These results indicated that Fe(II) addition contributed to denitrification.Ammonium in domestic wastewater was oxidized to nitrite or nitrate via nitrification in aerobic condition, nitrite or nitrate was reduced to nitrogen gas via denitrification in anaerobic condition, and denitrification is usually regarded as one of the determining steps of the nitrificationdenitrification process (Mac Conell et al., 2013;Wu et al., 2016).The results revealed that NO 3 -and NO 2 -could be rapidly reduced to nitrogen gas with Fe(II) addition, which was consistent with previous study (Li et al., 2016).Ferrous ions (Fe2+) are required for metalloenzymes such as hydrogenases and ferredoxins.And Fe2+ ions are transported by the FeoAB transporter encoded by genes implicated in ferrous iron uptake (Kammler et al., 1993).The expression feo operon was down-regulated under aerobic conditions and up-regulated under anaerobic conditions through transcription factors ArcA and FNR.Feo-mediated import of Fe2+ promotes Fur-Fe2+ occupancy and contributes to Fur regulation under anaerobic conditions (Beauchene et al., 2017).Although many other hidden mechanisms warrant further investigations, our results suggested that Fe(II) addition is an effective method to improve nitrogen removal efficiency and maintain the high reactor performance.

Ferrous iron Affected Microbial Community
To study the effect of Fe(II) on microbial community, the samples of the experiment reactor were analyzed with high throughput sequencing during four different periods, samples were named as Con, AdF, Aci and Rev.The sequence number of these samples were 22,593, 23,908, 26,222 and 23,656, respectively (Table 1).In total, 1412 OTUs were identified in the complete data set, and the average OTUs number of each sample was 718.Over 98.7% of the OTUs was assigned to a taxonomic group (phylum), and over 89.5% was identified at the order level.The Shannon index, ACE index and Chao index for Con sample tended to be higher than other samples, which indicated that bacteria diversity was most abundant in Con sample.
The dominant phyla in Con sample were Proteobacteria, Bacteroidetes, Chloroflexi, Firmicutes and Verrucomicrobia (Fig. 4).After the addition of Fe(II), Proteobacteria  increased from 32.77% to 47.33% and Firmicutes also increased from 5.89% to 15.36%.Fe(II) could stimulate Proteobacteria and Firmicutes increase, which was consistent with previous studies investigating iron oxidation and NO 3 -reduction (Coby et al., 2011;Melton et al., 2014;Li et al., 2016).However, Chloroflexi sharply decreased from 18.86% to 1.36% after Fe(II) addition, which indicated that Fe(II) may down-regulate Chloroflexi growth.These results suggested that Fe(II) remarkably affected the microbial community.
In the genus level, the dominant genera in Con sample were Clostridium, Hydrogenophaga, Flavobacterium, Prevotella, Achromobacter.After Fe(II) addition, the relative abundances of Hydrogenophaga, Methylotenera, Zoogloea and Fluviicola evidently increased, while the relative abundances of Clostridium, Flavobacterium, Prevotella, Achromobacter and Paludibacter, dramatically decreased in AdF sample.When stopped Fe(II) addition, the pH decreased below 6, the relative abundances of Paludibacter, Zoogloea and Methylomonas notably increased in Aci sample.When the pH of the reactor returned to the range from 7 to 8, the relative abundances of Hydrogenophaga, Methylotenera, Zoogloea and Fluviicola increased again in Rev sample, which was similar to the observations in AdF (Fig. 5).Interestingly, previous related studies have documented that some denitrifying bacterial strains with high nitrogen removal efficiency were isolated from biofilters and identified as Pseudomonas aeruginosa and Chelatococcus daeguensis according to 16S rRNA gene homology (Wu et al., 2013;Yang et al., 2013).The results are not agreement with ours.This implies that different strains may acquire similar functions through convergence evolution.

Correlations between Chemicals and Abundances of Genera
The composition of influent and pH significantly affected the UASB reactor performance and stability by influencing the bacterial community and diversity.Therefore, the correlation coefficients between abundant genera and chemicals reaction rate were analyzed (Table 2).The results reconfirmed that the abundances of Hydrogenophaga, Zoogloea, Fluviicola and Syntrophus positively correlated with COD removal efficiency; the abundances of Hydrogenophaga, Methylotenera and Fluviicola also positively correlated with Fe(II); the abundances of Paludibacter and Syntrophus were related to low pH condition.
Paludibacter, Hydrogenophaga, Zoogloea, Fluviicola, and Methylotenera were the dominant bacteria in the UASB reactor with Fe(II) addition.When the pH was below 6 and without addition of Fe(II), the relative abundances of Paludibacter increased notably in Aci, which was in accordance with a previous study and may result from that Paludibacter produces many organic acids (Ueki et al., 2006).Hydrogenophaga, an important bacterium for denitrification, is an autotrophic facultative aerobe genus that uses hydrogen as an electron donor, with the metabolites of N 2 and H 2 O, which was consistent with our results that the NO 3 -concentration was decreased after the addition of Fe(II) (Chen et al., 2013).Zoogloea has been previously reported to have NO 3 -reduction and nitrogen-fixing capabilities with Fe(II) oxidation (Shao et al., 2009;Oosterkamp et al., 2011).Both Fluviicola and Methylotenera have been proven to play roles in denitrification with Fe(II) oxidation (Kalyuhznaya et al., 2009;Gonzalez-Martinez et al., 2016).
Based on the correlation coefficient analysis and known function of dominant bacteria, the hypothetical mechanism was proposed in Fig. 6.Firstly, Paludibacter made a great contribution to nitrification utilizing COD to oxidize NH 4 + and generated NO 3 -; secondly, Hydrogenophaga and Zoogloea reduced NO 3 -to NO 2 -companied with Fe 2+ oxidized to Fe 3+ ; Fluviicola and Methylotenera reduced NO 2 -Fig.6.A proposed mechanism of the accelerated biogas production with Fe(II) addition by the dominant bacteria in UASB reactor.Table 2. Correlation coefficients between chemicals and abundance of genus.

Methylotenera
Zoogloea to N 2 with Fe 2+ addition.Recently, it was documented that maritime emissions have a notable influence on air pollution over coastal areas, especially in summer (Ding et al., 2018).Apart from VOC control, NO x control is also critical to reduce peak ozone concentrations (Li et al., 2013).As reported by Liu et al. (2017), our study may provide a hint for the design and preparation of Fe catalysts for biologicallyderived NO x control.

CONCLUSION
This study investigated the roles of ferrous ion in UASB reactor.Ferrous ion not only play an important role in maintaining appropriate pH and stable reactor performance, but also significantly influence biogas emission and corresponding microbial community.The findings will improve understanding the mechanisms of ferrous ions on denitrification process and provide hints for biogas emission management.

Fig. 2 .
Fig. 2. Profiles of pH and COD removal efficiency with time in different UASB reactors.(A) and (B): control UASB reactor without the addition of Fe(II); (C) and (D): UASB reactor with Fe(II) addition; the yellow vertical solid lines represented the beginning of Fe(II) addition; the orange vertical solid lines represented the stop of Fe(II) addition; the box represented the sample for high-through sequencing.

Fig. 3 .
Fig. 3. Concentration of (A) Fe 3+ and Fe 2+ , (B) NO 3 -, (C) NO 2 -and (D) NH 4 + , in the UASB reactor.The yellow vertical solid lines represented the beginning of Fe(II) addition; the orange vertical solid lines represented the stop of Fe(II) addition.

*
Correlation is significant at P < 0.05.

Table 1 .
Summary of the 16S rRNA sequences, operational taxonomic units (OTUs), and microbial diversity indices for all examined samples.