Bioresource Technology 102 (2011) 7265–7271

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Microbial community structure and dynamics in a mixotrophic nitrogenremoval process using recycled spent caustic under different loading conditions

Sora Park a, Jaecheul Yu a, Imgyu Byun b, Sunja Cho a, Taejoo Park a,c, Taeho Lee a,⇑a Department of Civil and Environmental Engineering, Pusan National University, Jangjeon-Dong, Geumjung-Gu, Busan 609-735, Republic of Koreab Institute for Environmental Technology and Industry, Pusan National University, Jangjeon-Dong, Geumjung-Gu, Busan 609-735, Republic of Koreac Korea Environment Institute, 290 Jinheungno, Eunpyeong-Gu, Seoul 122-706, Republic of Korea

a r t i c l e i n f o

Article history:Received 7 February 2011Received in revised form 20 April 2011Accepted 27 April 2011Available online 30 April 2011

Keywords:Autotrophic denitrificationMicrobial communityNitrogen loading ratePyrosequencingSpent caustic

0960-8524/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.biortech.2011.04.091

⇑ Corresponding author. Tel.: +82 51 510 2465; faxE-mail address: leeth55@pusan.ac.kr (T. Lee).

a b s t r a c t

A laboratory-scale Bardenpho process was established to investigate the proper nitrogen loading rate(NLR) when modified spent caustic (MSC) is applied as electron donor and alkalinity source for denitri-fication. MSC injection induced autotrophic nitrogen removal with sulfur as electron donor and hetero-trophic denitrification. The nitrogen removal rate (NRR) did not increase proportionally to NLR. Basedon the total nitrogen concentration in the effluent observed in the trials with MSC, the NLR in the influentshould not exceed 0.15 kg N/m3 d in order to satisfy water quality regulations. Microbial communities inthe anoxic reactors were characterized by pyrosequencing of 16S rRNA gene sequences amplified by thepolymerase chain reaction of DNA extracted from sludge samples. Microbial diversity was lower as MSCdosage was increased, and the injection of MSC caused an increase in SOB belonging to the genus Thio-bacillus which is responsible for denitrification using sulfur.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Nitrogen removal is essential in wastewater treatment becausenitrogen is a primary pollutant which causes eutrophication (U.S.EPA, 2001) and is harmful to human health (Metcalf and Eddy,1991). Heterotrophic denitrification, which has been widely usedfor eliminating nitrogen, causes high operating costs since it re-quires external carbon sources in treating wastewater with a lowCOD/N ratio (Rittmann and McCarty, 2001). As an alternative toheterotrophic denitrification, autotrophic denitrification whichgenerally uses inorganics as an electron donor has been studied(Oh et al., 2001; Moon et al., 2008). However, autotrophic denitri-fication that uses sulfur sources consumes alkalinity as shown inthe Eqs. (1)–(3) (McCarty, 1975);

1:1Sþ NO�3 þ 0:4CO2 þ 0:76H2Oþ 0:08HCO�3 þ 0:08NHþ4! 0:08C5H7O2Nþ 0:5N2 þ 1:1SO2�

4 þ 1:28Hþ ð1Þ

0:844S2O2�3 þ NO�3 þ 0:347CO2 þ 0:434H2Oþ 0:086HCO�3

þ 0:086NHþ4! 0:086C5H7O2Nþ 0:5N2 þ 1:689SO2�

4 þ 0:697Hþ ð2Þ

ll rights reserved.

: +82 51 514 9574.

0:421H2Sþ 0:421HS� þ NO�3 þ 0:346CO2 þ 0:086HCO�3þ 0:086NHþ4 ! 0:086C5H7O2Nþ 0:5N2 þ 0:842SO2�

4

þ 0:434H2Oþ 0:262Hþ ð3Þ

Therefore, the alkalinity lost during sulfur-based autotrophicdenitrification has to be replenished. Spent caustic (SC) producedby the petrochemical industry can be used for this purpose sinceit contains a sulfur source as well as an alkalinity source (Rajga-nesh et al., 1995). SC application proved to be efficient for nitrogenremoval from the sewage with a low COD/N ratio (Park et al.,2009), SC subjected to ultrasound treatment fortified with thiosul-fate (Na2S2O3�5H2O) was an electron donor for nitrogen removal(Park et al., 2010a).

However, chemical oxygen demand (COD) and total nitrogen(TN) concentration of wastewater in Korea range from 16.6 to274.6 mg/L and 6.5 to 326.2 mg/L, respectively. Therefore, an eval-uation of the permissible nitrogen loading rate is required toachieve a discharge water quality of 20 mg/L as TN (Water Qualityand Ecosystem Conservation Act, Ministry of Environment, Repub-lic of Korea) and to determine the proper MSC dosage for practicalapplication in the wastewater treatment process.

The microbial communities carrying out the various biochemicalreactions in SC-fed processes has already been studied using molec-ular techniques such as polymerase chain reaction–denaturing gelgradient electrophoresis (PCR–DGGE) and fluorescent in situ hybrid-ization (FISH) (Park et al., 2010a,b), but the information obtained

Table 1Operating conditions of laboratory scale Bardenpho process.

A B C

Influent sCODCr concentration (mg/L) 60.2–67.8 (63.9)a 93.5–137.7 (108.1) 131.5–139.3 (136.0)Influent ammonia concentration (mg/L) 32.2–36.2 (34.2) 48.5–55.5 (50.8) 63.4–68.7 (66.8)Nitrogen loading rate (kg/m3 d) 0.10 0.15 0.20MSC dosage (mL/ L influent) 2.0 3.0 4.0HRT (h) 8 8 8

a The numbers in the brackets are the mean values.

Fig. 1. Influent (d) and effluent (s) concentration of sCODCr (a), NH4+-N (b) and

Total nitrogen (c) under different loading rate conditions. Nitrate (5) and nitrite (.)concentration in the effluent are also represented in (c).

7266 S. Park et al. / Bioresource Technology 102 (2011) 7265–7271

was limited. Sequencing based on the pyrosequencing methodologyhas emerged as a tool for analyzing complex microbial communitiessince it can generate massive-scale 16S rRNA gene sequence libraries(Edwards et al., 2006). This technique was used to obtain more in-sight into the microbial community in the laboratory-scale reactorsoperated in the current study.

Therefore, the main objectives of this study are to determine thepermissible nitrogen loading rate (NLR) of the nitrogen removalprocess fed with MSC to satisfy discharge water quality and attaina more detailed insight into the microbial community by using thepyrosequencing technology.

2. Methods

2.1. Reactor configuration

A laboratory scale reactor for the Bardenpho process describedpreviously (Park et al., 2010a) was operated at a room temperatureunder different influent loading rates with a fixed COD/N ratio. Thereactor consists of anoxic tank 1, aerobic tank 1, anoxic tank 2, aer-obic tank 2 and a settling tank. The effective volume of each tank is5.2 L. The reactor was inoculated with activated sludge (AS) from afull-scale municipal sewage treatment plant in Busan, Korea, andthe internal and external recycle ratios were fixed at 1.0 and 1.5,respectively.

2.2. Operational condition

A synthetic wastewater containing (mg/L) KH2PO4, 17.6; NaH-CO3, 450; KCl, 17.5; CaCl2�2H2O, 17.5; NaCl, 37.5; MgSO4�7H2O,12.5 was introduced into the reactor for an influent. The syntheticwastewater also contained 1 mL/L of a trace element solution (g/L)containing EDTA, 0.5; FeSO4�7H2O, 0.2; ZnSO4�7H2O, 0.01;MnCl2�4H2O, 0.003; H3BO3, 0.03; CoCl2�6H2O, 0.02; CuCl2�2H2O,0.01; NiCl2�6H2O, 0.002; Na2MoO4�2H2O, 0.003. Glucose and NH4Clwere added depending on the desired influent loading rate. To eval-uate the NLR that would satisfy the discharge water quality, the NLRwas increased from 0.10 kg N/m3 d to 0.20 kg N/m3 d. The organicloading rate (OLR) also increased as shown in Table 1, because theCOD/N ratio was fixed at 2.0. The MSC dosage was also proportionalto NLR, and MSC dosage was determined theoretically on the basis ofEqs. (1) and (2). The mean concentrations of S2�, S2O3

2� and alkalin-ity of MSC are 13,653 mg/L, 25,400 mg/L and 65,200 mg/L,

Table 2Characteristics of community diversity based upon diversity indices.

Conditions Numbers of sequencereads

Numbers ofOTUsa

Chao1

Shannon(H’)

AS 20,132 3132 4397 6.75A 6680 2187 4266 6.36B 13,298 1155 1647 5.51C 6611 605 910 4.81

AS: activated sludge, A, B, C: MSC-fed conditions.a OTUs were defined at a 97% similarity level.

S. Park et al. / Bioresource Technology 102 (2011) 7265–7271 7267

respectively, and MSC contains suspended solids of 30 mg/L, NH4+-N

of 195 mg/L and TOC of 1937 mg/L (Park et al., 2010a). The MSC wasinjected into the anoxic tank 1 and anoxic tank 2 for denitrification.To acclimate microorganisms in the reactor to the changes in theinfluent organics and the nitrogen loading rate, MSC was not injectedfor 10–15 days between conditions A and B, and conditions B and C.

2.3. DNA extraction, PCR amplification and pyrosequencing

Genomic DNAs for pyrosequencing were obtained from an AS ofa wastewater treatment plant and the sludge from anoxic tank 1operating under the different influent conditions. The sludge sam-ples of 1 mL were centrifuged at 10,000g for 1 min and resus-pended in 1 mL of distilled water. A PowerSoil DNA kit (Mo BioLabs, Carlsbad, CA) was used to extract DNA. PCR amplificationwas performed as follows: an initial denaturation at 95 �C for10 min; 35 cycles of denaturation (45 s at 94 �C), annealing (60 sat 55 �C), extension (60 s at 72 �C); and a final extension at 72 �Cfor 10 min. Amplification of 16S rDNA gene fragments of eubacte-ria from the extracted DNA was performed with the combinationsof 27F (50-GAGTTTGATC(A/C)TGGCTCAG-30) (Edwards et al., 1989)and 518R (50-(A/T)TTACCGCGGCTGCTGG-30) (Muyzer et al., 1993).Additional barcode sequences were added to the front of the for-ward primer for sorting each sample from the pyrosequencing re-sults. The reactions were carried out in the following PCR solution;3 lL of 100 ng template DNA, 1 lL of 10 pmol each primer, 5 lL of10� Taq buffer, 5 lL of 10 mM dNTP, and 0.3 lL of Taq polymerase(Solgent, Daejeon, Korea). Distilled water was added to a final vol-ume of 50 lL. PCR products of 491 bp were isolated by gel electro-phoresis and purified twice with QIAquick Gel extraction kit(Qiangen, CA, USA) and QIAquick PCR purification kit (Qiangen,CA, USA). Amplicon pyrosequencing was performed by MacrogenInc. (Seoul, Korea) using 454/Roche GS-FLX Titanium (Roche, NJ,USA). The sequences were passed through quality filters to reducethe overall error rate; sequences shorter than 100 nucleotides andless than 20 of Average Quality score were removed. Multiple se-quence alignment and complete linkage clustering were done tocluster the sequences from 0% to 3% of dissimilarity using the Ribo-somal Database Project’s (RDP) pyrosequencing pipeline (http://rdp.cme.msu.edu). The Shannon and Chao 1 diversity indexes werealso calculated using the RDP’s pyrosequencing pipeline.

Fig. 2. The relative phylogenetic distribution of OTUs in samples from activated sludge aMSC-fed conditions.

2.4. Analytical methods

The concentrations of soluble chemical oxygen demand (sCODCr)and NH4

+-N were determined using an auto analyzer (AA3, Bran+-Luebbe, Germany) after filtration of the sample through a 0.45 lmmembrane filter. NO2

�-N, NO3�-N concentrations were analyzed

using an ion chromatography (DX-300, DIONEX, Sunnyvale, USA).The pH and dissolved oxygen (DO) concentrations were measuredusing an Orion Research pH meter (230A, Thermo Fisher ScientificInc., Waltham, USA) and a DO meter (YSI 58, Fondriest Environmen-tal, Inc., Dayton, USA), respectively.

3. Results and discussion

3.1. Organic and ammonia removal

The removal efficiencies of sCODCr increased as the sCODCr con-centration in the influent increased in the range from 63.9 mg/L to136.0 mg/L, and they were 59.5%, 73.3% and 80.2% in conditions A,B and C, respectively as shown in Fig. 1(a). The effluent sCODCr con-centration in each condition was in the range of 25.9–28.9 mg/L.The organic removal rates (ORR) closely followed the inflow load-ing patterns, and the ORRs increased from 0.11 kg/m3 d to 0.33 kg/m3 d as the OLRs increased from 0.20 kg/m3 d to 0.40 kg/m3 d. Theorganic removal is attributed to the activity of heterotrophic since96.0% of eliminated organic matter was removed in anoxic tank 1(data were not shown). The increase in the MSC dosage that fol-lowed the NLR increase did not affect the nitrification efficiency,as shown in Fig. 1(b) since it was over 97% throughout the operat-ing time, as it was in our previous studies which applied SC to thebiological nitrogen removal (BNR) process (Park et al., 2009,2010b).

3.2. Denitrification with and without MSC

Comparison of TN removal efficiencies under conditions withand without MSC demonstrated that MSC injection induced addi-tional TN removal (Fig. 1(c)). This observation suggests that someportion of nitrate was removed heterotrophically. The remainingnitrate was denitrified by using MSC without inhibition by organicsunder mixotrophic conditions where the sulfur source and organ-ics were present simultaneously (Oh et al., 2001).

nd MSC-fed conditions as the result of pyrosequencing. AS: activated sludge, A, B, C:

7268 S. Park et al. / Bioresource Technology 102 (2011) 7265–7271

Most nitrogen existed in the form of nitrate in the effluent, andthe nitrite concentration in the effluent was below 0.83 mg/L dur-ing the entire operating period (nitrite accumulation was not ob-served). The NLRs were 0.10 kg/m3 d, 0.15 kg/m3 d and 0.20 kg/m3 d, and the nitrogen removal rates (NRR) were 0.08 kg/m3 d,0.12 kg/m3 d and 0.12 kg/m3 d, respectively. The TN removal effi-ciencies were 73.8%, 78.2% and 62.0% in conditions A, B and C,respectively. The NRR did not increase in condition C, which meantthat not all nitrogen fed to the reactor was converted into nitrogengas. These findings suggest that nitrogen levels above 0.15 kg/m3 d

Fig. 3. Neighbor-joining phylogenetic tree. The 20 OTUs which are five most abundantwith MEGA 4.0. The accession numbers from Genbank are shown by the cluster names an5% difference in nucleotide sequences.

overload the system, and this value can be used as design criteriaof wastewater treatment plant that applied MSC.

NRR values depend on the type of reactors, electron donors andNLRs. When the denitrification was carried out in the presence oforganics such as methanol, ethanol and acetate, the NRR valuesare usually higher than those in the presence of sulfur sources (Bo-ley et al., 2000; Soares, 2002; Zhang and Lampe, 1999). For exam-ple, methanol and ethanol resulted in NRR values as high as 1–27 kg/m3 d and 0.4–1.2 kg/m3 d, respectively (Boley et al., 2000).The NRRs in the studies that applied sulfur ranged from 0.004

OTUs in AS and MSC-fed conditions and their closest microorganisms were alignedd the abundance of each OTU was presented in the brackets. The scale bar represents

S. Park et al. / Bioresource Technology 102 (2011) 7265–7271 7269

kg/m3 d to 0.560 kg/m3 d depending on reactor type and inflowNLR (Hashimoto et al., 1987; Kimura et al., 2002; Soares, 2002;Zhang and Lampe, 1999). Interestingly, higher NRR values were ob-served under mixotrophic conditions with a combination of sulfurand organics source; sulfide with acetate (Gommers et al., 1988)and sulfur with methanol (Lee et al., 2001) resulted in NRR valuesof 5 kg/m3 d and 5.05 kg/m3 d, respectively. Therefore, a higherNRR can be expected when MSC is applied in treating wastewaterwith a low COD/N ratio because both autotrophic denitrificationusing sulfur from MSC and heterotrophic denitrification usingorganics in wastewater can take place.

3.3. Bacterial diversity analyzed by pyrosequencing

3.3.1. Composition of bacterial communityPyrosequencing analysis yielded a total 46,721 sequences from

four genomic DNA samples after removal of short and low-qualityreads. The sequence library size ranged from 6611 sequences forcondition C to 20,132 sequences for AS. The numbers of operationaltaxonomic units (OTUs) (3% dissimilarity) for each library areshown in Table 2. The highest number of OTUs was found in AS,and the lowest number in the library from condition C whichhad the highest loading rate and MSC dosage. The Shannon (H0)and Chao 1 diversity indices also decreased with increasing MSCdosage (Table 2). The difference in microbial diversity betweenthe AS and the samples derived from the laboratory-scale reactorsmight have been due to the relative simplicity of the reactor feeds.The bacterial diversity would likely be higher in the presence ofmore complex influents (Borole et al., 2011; Cho and Cho, 2008).

At the phylum level, sequences belonging to the Proteobacteriawere the most abundant with over 50% of the total bacterial se-quences in all samples (Fig. 2) and among them, betaproteobacteriasequences were the most common. The second most abundantgroup of sequences originated from Bacteroidetes. Other sequences

Table 3Bacterial identification of the major OTUs and their potential functions in activated sludg

Condition OTUnumber

Relativeabundance (%)

Closest strain Accenum

AS OTU AS1 10.2 Dokdonella koreensis EF58OTU AS2 6.2 Haliscomenobacter hydrossis

DSM 1100AJ78

OTU AS3 3.2 Thioflavicoccus mobilis AJ01OTU AS4 2.9 Caenimonas koreensis DQ3

OTU AS5 2.9 Simplicispira sp. R-23033 AM2A OTU A1 19.3 Dechloromonas sp. JJ AY03

OTU A2 7.4 Zoogloea caeni DQ4

OTU A3 6.7 Uncultured Azospira sp. GQ1OTU A4 4.4 Janthinobacterium sp. Lc51-6 GU7OTU A5 3.2 Uncultured bacterium AY30

B OTU B1 9.7 Rhodobacter sp. HCD-7052 FN81OTU B2 9.3 Rhodobacter sp. TUT3732 AB25OTU B3 7.7 Kouleothrix aurantiaca AB07OTU B4 6.8 Thiobacillus thioparus AF00OTU B5 6.2 Tetrasphaera elongata NR_0

C OTU C1 22.8 Thiobacillus denitrificans EU54

OTU C2 17.5 Uncultured Zoogloea sp. GQ2

OTU C3 7.9 Uncultured Thiobacillus sp. FJ480OTU C4 5.5 Uncultured bacterium FM2OTU C5 5.0 Aeromonas hydrophilia GU1

AS: activated sludge, A, B, C: MSC-fed conditions.

present stemmed from the phyla Chloroflexi, Actinobacteria, Plan-ctomycetes, Firmicutes, Acidobacteria and Nitrospira.

This bacterial composition was consistent with the previousstudies which investigate microbial diversity of AS from wastewa-ter treatment plants (Amann et al., 1996; Snaidr et al., 1997; Jure-tschko et al., 2002).

The five most abundant OTUs in each condition and their rela-tionship with sequences from related microorganisms are pre-sented in Fig. 3. Some of the sequences obtained from reactorsoperating under conditions B and C were closely related to thosefrom Thiobacillus denitrificans, an autotrophic denitrifier. Table 3summarizes the potential function of the bacteria represented bythe different OTUs. Overall, the OTUs represent bacteria that arepotentially capable of organics and phosphorus removal, nitratereduction and sulfur oxidization.

3.3.2. Nitrifying bacterial communityThe relative abundance of sequences representing nitrifying

bacteria as a percentage of the total eubacterial sequences arelisted in Table 4. The AS sample had the most diverse nitrifyingpopulations. The abundance of sequences from nitrifying bacteriadid not increase with the increase in NLRs. The differences betweenthe numbers and types of sequences from nitrifying bacteria re-trieved from the AS and the laboratory-scale reactor samples arelikely due to differences in the influents including the presenceof MSC since the DO concentration (0.18 ± 0.01 mg/L) and the tem-perature were not varied in MSC-fed conditions.

The presence of Nitrosomonas sequences in samples from condi-tions B and C were most dominant among AOB species, which par-allels results obtained in studies using FISH (Schramm et al., 1996;Wagner et al., 1995). The disappearance of sequences related tothose of the genera Nitrobacter and Nitrococcus upon introductionof MSC could have been due to the inability of these types ofbacteria to tolerate MSC and low DO concentrations. Sequences

e and MSC-fed conditions.

ssionber

Similarity(%)

Potential function Reference

9679 884892 97 – Degradation of soluble

organics– Floc formation

Gerardi, 2006

0125 9449098 98 – Nitrate reduction

– Phosphorus removalRyu et al., 2008

36310 99 Nitrification2611 98 Nitrate reduction Coates et al., 2001

13150 99 Degradation of solubleorganic

Gerardi, 2006

83381 96 Ammonia oxidation Allen et al., 201033479 99 Nitrate reduction Hashidoko et al., 20082119 99 Heterotrophic

denitrificationGinige et al., 2005

3451 971408 979640 100 Nitrate reduction Kohno et al., 20025628 97 Sulfur oxidization Beller et al., 200624735 98 Phosphorous removal Hanada et al., 20026130 97 Autotrophic denitrification Baalsrud and Baalsrud,

195449359 94 Heterotrophic

denitrification831 95 Autotrophic denitrification

13064 99 Sulfur utilization69711 99 Nitrate reduction Firth and Edwards,

2000

Table 4The relative abundance of sequences classified as genera of nitrifying bacteria to eubacteria detected by pyrosequencing.

Condition Ammonia oxidizing bacteria Nitrite oxidizing bacteria

Nitrosomonas Nitrosospira Nitrosococcus Nitrobacter Nitrococcus Nitrospira

AS 0.06 0.18 0.03 0.18 0.03 0.09A 1.0 0.42 0.12 – – 0.28B 0.60 0.10 –a – – 3.40C 0.27 0.03 0.03 – – 1.24

a No sequence was detected.

7270 S. Park et al. / Bioresource Technology 102 (2011) 7265–7271

representing the genus Nitrospira persisted in the presence of MSCperhaps because Nitrospira is able to CO2 and grow mixotrophically(Daims et al., 2001), which would have been advantageous in thepresence of organic matter concentrations ranging between 63.9and 136.0 mg/L. Nitrospira-like nitrite oxidizing bacteria are alsowell-adapted to low nitrite concentrations (Schramm et al., 1999).

3.3.3. Heterotrophic denitrifying bacterial communitySome studies have shown that the members of genera Achromo-

bacter, Agrobacterium, Alcaligenes, Bacillus, Chromobacterium, Flavo-

Table 5Relative abundance of genus-level bacteria capable of heterotrophic denitrificationdetected by pyrosequencing in activated sludge and each condition.

Condition Relative abundanceof HD bacteria (%)

Major genera Relative abundancein HD bacteria (%)

AS 14.1 Aquaspirillum 16.5Thauera 14.4Rhodobacter 13.1Dechloromonas 11.8Zoogloea 9.1

A 13.4 Zoogloea 64.7Flavobacterium 12.2Dechloromonas 9.1Thauera 6.8Rhodobacter 4.5

B 29.4 Rhodobacter 32.9Zoogloea 31.5Chromobacterium 17.1Flavobacterium 7.6Hyphomicrobium 4.0

C 38.9 Zoogloea 45.0Chromobacterium 20.4Aeromionas 12.9Rhodobacter 9.5Thauera 5.8

HD: heterotrophic denitrification.The NLRs in each condition are as follows; A: 0.10 kg/m3 d, B: 0.15 kg/m3 d, C:0.20 kg/m3 d.

Table 6Sulfur oxidizing bacteria and their number of sequences detected in activated sludge and

Class Order Family

Alphaproteobacteria Rhodobacterales RhodobacteraceaeBetaproteobacteria Hydrogenophilales Hydrogenophilaceae

Burkholderiales Burkholderiales_ incertae_

Gammaproteobacteria Chromatiales Chromatiaceae

EctothiorhodospiraceaeThiotrichales Thiotrichaceae

Epsilonproteobacteria Campylobacterales Helicobacteraceae

AS: activated sludge, A, B, C: MSC-fed conditions.The MSC dosages in each condition are as follows; A: 2.0 mL/L influent, B: 3.0 mL/L influ

bacterium, Hyphomicrobium, Paracoccus and Pseudomonas areresponsible for heterotrophic denitrification (Knowles, 1982). Therelative abundance of sequences originating from members ofthese genera in AS was 14.1% which is comparable to condition A(13.4%).

The relative abundance of sequences from potential heterotro-phic denitrifiers increased to 29.4% and 38.9% in condition B andcondition C (Table 5) likely due to the increased NLR and OLR.The genera Zoogloea and Rhodobacter were detected in all condi-tion, and Zoogloea sequences increased in numbers after MSC injec-tion (Table 5). Under these conditions, sequences representingAeromonas, Chromobacterium and Thauera were present. Most ofthese genera include bacteria able to use nitrate as an electronacceptor and to reduce nitrate to nitrogen gas (Brenner et al.,2009).

3.3.4. Sulfur-utilizing autotrophic denitrifying bacterial communityIn contrast to previous FISH and PCR–DGGE studies (Park et al.,

2010a,b), the high throughput pyrosequencing approach readilyrevealed the presence of sequences originating from SOB (Table 6).Interestingly, the highest diversity of sequences potentially origi-nating from SOB was found in the AS sample. In the presence ofMSC, Thiobacillus sequences were highly represented and OTU C1contained sequences 97% similar to those of T. denitrificans, a bac-terium previously observed in a FISH-based study with SC (Parket al., 2010b). This bacterium is widely distributed in the environ-ment, especially under anoxic conditions. It oxidizes thiosulfate,tetrathionate or sulfide under anaerobic condition using nitrate, ni-trite or nitrous oxide as the terminal electron acceptor (Brenneret al., 2009). While PCR–DGGE analysis showed that epsilonproteo-bacteria were dominant (Park et al., 2010a), pyrosequencing resultsshowed betaproteobacteria were dominant when MSC was injected.The pyrosequencing results also provided overall sequence li-braries from bacteria inhabiting a MSC-applied wastewater treat-ment plant.

anoxic tank 1 of Bardenpho process in each condition.

Genus AS A B C

Thioclava 13 3 2 2Thiobacillus 7 6 898 1506

sedis Thiobacter 41 21 37 76Thiomonas 3 – – 1Thiobaca 1 – – –Thiococcus 5 1 – –Thiodictyon 2 – – –Thioflavicoccus 642 – – –Thiohalocapsa 85 4 59 6Thiorhodovibrio 4 – 1 –Thiorhodospira 8 1 – –Leucothrix 4 – – –Thiothrix 110 2 67 3Sulfurovum – – 1 –

ent, C: 4.0 mL/L influent.

S. Park et al. / Bioresource Technology 102 (2011) 7265–7271 7271

4. Conclusions

The laboratory-scale studies with reactors carrying out the Bar-denpho process suggest that a maximum NLR of 0.15 kg/m3 d canbe introduced into the reactors in the presence of MSC as electrondonor without affecting discharge water quality. The determinedNLR value would be used as the design criteria for setting upfull-scale wastewater treatment plants that applies MSC. Thepyrosequencing of 16S rRNA sequence libraries obtained fromthe reactors suggest that MSC resulted in changes to the SOB pop-ulations reflected in the presence of large numbers of sequencesrepresenting the genus Thiobacillus.

Acknowledgements

This work was financially supported by Pusan National Univer-sity Research Grant for two years and second stage of Brain Korea21 Project in 2010.

References

Allen, J.G., Beutel, M.W., Call, D.R., Fischer, A.M., 2010. Effects of oxygenation onammonia oxidation potential and microbial diversity in sediment from surface-flow wetland mesocosms. Bioresour. Technol. 101, 1389–1392.

Amann, R., Snaidr, J., Wagner, M., Ludwig, W., Schleifer, K.H., 1996. In situvisualization of high genetic diversity in a natural microbial community. J.Bacteriol. 178, 3496–3500.

Baalsrud, K., Baalsrud, K.S., 1954. Studies on Thiobacillus denitrificans. Arch.Microbiol. 20, 34–62.

Beller, H.R., Chain, P.S.G., Letain, T.E., Chakicherla, A., Larimer, F.W., Richardson,P.M., Coleman, M.A., Wood, A.P., Kelly, D.P., 2006. The genome sequence of theobligately chemolithoautotrophic, facultatively anaerobic bacteriumThiobacillus denitrificans. J. Bacteriol. 188, 1473–1488.

Boley, A., Müller, W.R., Haider, G., 2000. Biodegradable polymers as solid substrateand biofilm carrier for denitrification in recirculated aquaculture systems.Aquacult. Eng. 22, 75–85.

Borole, A.P., Hamilton, C.Y., Vishnivetskaya, T.A., 2011. Enhancement in currentdensity and energy conversion efficiency of 3-dimensional MFC anodes usingpre-enriched consortium and continuous supply of electron donors. Bioresour.Technol. 102, 5098–5104.

Brenner, D.J., Krieg, N.R., Staley, J.T., Garrity, G.M., 2009. Bergey’s manual ofsystematic bacteriology: Volume 2 The Proteobacteria, second ed. Springer.

Cho, W.S., Cho, K.S., 2008. Effect of methyl tert-mutyl ether and its metabolites onmicrobial activity and diversity in tidal mud flat. Korean J. Microbiol.Biotechnol. 36, 336–342.

Coates, J.D., Chakraborty, R., Lack, J.G., O’Connor, S.M., Cole, K.A., Bender, K.S.,Achenbach, L.A., 2001. Anaerobic benzene oxidation coupled to nitratereduction in pure culture by two strains of Dechloromonas. Nature 411, 1039–1043.

Daims, H., Nielsen, J.L., Nielsen, P.H., Schleifer, K.H., Wagner, M., 2001. In situcharacterization of Nitrospira-like nitrite-oxidizing bacteria active inwastewater treatment plants. Appl. Environ. Microbiol. 67, 5273–5284.

Edwards, R.A., Rodriguez-Brito, B., Wegley, L., Haynes, M., Breitbart, M., Peterson,D.M., Saar, M.O., Alexander, S., Alexander Jr, E.C., Rohwer, F., 2006. Usingpyrosequencing to shed light on deep mine microbial ecology. BMC Genomics 7,57–69.

Edwards, U., Rogall, T., Blocker, H., Emde, M., Bottger, E.C., 1989. Isolation and directcomplete nucleotide determination of entire genes – characterization of a genecoding for 16s-ribosomal RNA. Nucleic Acids Res. 17, 7843–7853.

Firth, J.R., Edwards, C., 2000. Denitrification by indigenous microbial populations ofriver water measured using membrane inlet mass spectrometry. J. Appl.Microbiol. 89, 123–129.

Gerardi, M.H., 2006. Waste water bacteria. Microbiology Series. Wiley, Hoboken, NJ.Ginige, M.P., Keller, J., Blackall, L.L., 2005. Investigation of an acetate-fed denitrifying

microbial community by stable isotope probing, full-cycle rRNA analysis, andfluorescent in situ hybridization–microautoradiography. Appl. Environ.Microbiol. 71, 8683–8691.

Gommers, P.J.F., Buleveld, W., Gus Kuenen, J., 1988. Simultaneous sulfide andacetate oxidation in a denitrifying fluidized bed reactor – I. Start-up and reactorperformance. Water Res. 22, 1075–1083.

Hanada, S., Liu, W.T., Shintani, T., Kamagata, Y., Nakamura, K., 2002. Tetrasphaeraelongata sp. nov., a polyphosphate-accumulating bacterium isolated fromactivated sludge. Int. J. Syst. Evol. Microbiol. 52, 883.

Hashidoko, Y., Takakai, F., Toma, Y., Darung, U., Melling, L., Tahara, S., Hatano, R.,2008. Emergence and behaviors of acid-tolerant Janthinobacterium sp. thatevolves N2O from deforested tropical pealand. Soil. Biol. Biochem. 40, 116–125.

Hashimoto, S., Furukawa, K., Shioyama, M., 1987. Autotrophic denitrification usingelemental sulfur. J. Ferment. Technol. 65, 683–692.

Juretschko, S., Loy, A., Lehner, A., Wagner, M., 2002. The microbial communitycomposition of a nitrifying–denitrifying activated sludge from an industrialsewage treatment plant analyzed by the full-cycle rRNA approach. Syst. Appl.Microbiol. 25, 84–99.

Kimura, K., Nakamura, M., Watanabe, Y., 2002. Nitrate removal by a combination ofelemental sulfur-based denitrification and membrane filtration. Water Res. 36,1758–1766.

Knowles, R., 1982. Denitrification. Microbiol. Mol. Biol. Rev. 46, 43.Kohno, T., Sei, K., Mori, K., 2002. Characterization of type 1851 organism isolated

from activated sludge samples. Water Sci. Technol.: J. Int. Assoc. Water Poll. Res.46, 111–114.

Lee, D.U., Lee, I.S., Choi, Y.D., Bae, J.H., 2001. Effects of external carbon source andempty bed contact time on simultaneous heterotrophic and sulfur-utilizingautotrophic denitrification. Process Biochem. 36, 1215–1224.

McCarty, P.L., 1975. Stoichiometry of biological reactions. Prog. Water Technol. 7,157–172.

Metcalf, Eddy, 1991. Wastewater Engineering. Treatment, Disposal, Reuse, third ed.McGraw-Hill Int. Ed, Singapore.

Moon, H.S., Shin, D.Y., Nam, K., Kim, J.Y., 2008. A long-term performance test on anautotrophic denitrification column for application as a permeable reactivebarrier. Chemosphere 73, 723–728.

Muyzer, G., Dewaal, E.C., Uitterlinden, A.G., 1993. Profiling of complex microbial-populations by denaturing gradient gelelectrophoresis analysis of polymerasechain reaction-amplified genes-coding for 16s ribosomal-RNA. Appl. Environ.Microbiol. 59, 695–700.

Oh, S.E., Yoo, Y.B., Young, J.C., Kim, I.S., 2001. Effect of organics on sulfur-utilizingautotrophic denitrification under mixotrophic conditions. J. Biotechnol. 92, 1–8.

Park, J.J., Byun, I.G., Park, S.R., Lee, J.H., Park, S.H., Park, T.J., Lee, T.H., 2009. Use ofspent sulfidic caustic for autotrophic denitrification in the biological nitrogenremoval processes: lab-scale and pilot-scale experiments. J. Ind. Eng. Chem. 15,316–322.

Park, S., Seon, J., Byun, I., Cho, S., Park, T., Lee, T., 2010a. Comparison of nitrogenremoval and microbial distribution in wastewater treatment process underdifferent electron donor conditions. Bioresour. Technol. 101, 2988–2995.

Park, S., Lee, J., Park, J., Byun, I., Park, T., Lee, T., 2010b. Characteristics of nitrogenremoval and microbial distribution by application of spent sulfidic caustic inpilot scale wastewater treatment plant. Water Sci. Technol.: J. Int. Assoc. WaterPoll. Res. 62, 1440–1447.

Rajganesh, B., Sublette, K.L., Camp, C., Richardson, M.R., 1995. Biotreatment ofrefinery spent sulfidic caustics. Biotechnol. Prog. 11, 228–230.

Rittmann, B.E., McCarty, P.L., 2001. Environmental Biotechnology: Principles andApplications. McGraw-Hill, New York.

Ryu, S.H., Lee, D.S., Park, M., Wang, Q., Jang, H.H., Park, W., Jeon, C.O., 2008.Caenimonas koreensis gen. nov., sp. nov., isolated from activated sludge. J. Syst.Evol. Microbiol. 58, 1064–1068.

Schramm, A., De Beer, D., Van Den Heuvel, J.C., Ottengraf, S., Amann, R., 1999.Microscale distribution of populations and activities of Nitrosospira andNitrospira spp. along a macroscale gradient in a nitrifying bioreactor:quantification by in situ hybridization and the use of microsensors. Appl.Environ. Microbiol. 65, 3690–3696.

Schramm, A., Larsen, L.H., Revsbech, N.P., Ramsing, N.B., Amann, R., Schleifer, K.H.,1996. Structure and function of a nitrifying biofilm as determined by in situhybridization and the use of microelectrodes. Appl. Environ. Microbiol. 62,4641–4647.

Snaidr, J., Amann, R., Huber, I., Ludwig, W., Tandoi, V., Schleifer, K.H., 1997.Phylogenetic analysis and in situ identification of bacteria in activated sludge.Appl. Environ. Microbiol. 63, 2884–2896.

Soares, M., 2002. Denitrification of groundwater with elemental sulfur. Water Res.36, 1392–1395.

U.S. EPA, 2001. Memorandom: development and adoption of nutrient criteria intowater quality standards. Available from:<http://oaspub.epa.gov/waters/national_rept.control#TOP_IMP>.

Wagner, M., Rath, G., Amann, R., Koops, H.P., Schleifer, K.H., 1995. In situidentification of ammonia-oxidizing bacteria. Syst. Appl. Microbiol. 18, 251–264.

Zhang, T.C., Lampe, D.G., 1999. Sulfur: limestone autotrophic denitrificationprocesses for treatment of nitrate-contaminated water: batch experiments.Water Res. 33, 599–608.