Journal of Bioscience and Bioengineering VOL. 119 No. 1, 95e100, 2015 www.elsevier.com/locate/jbiosc

Nitrate reduction pathway in an anaerobic acidification reactor and its effect on acid fermentation Li Xie,* Chi Ji, Rui Wang, and Qi Zhou State Key Laboratory of Pollution Control and Resources Reuse, Tongji University, Siping Road No. 1239, Shanghai 200092, PR China Received 19 November 2013; accepted 31 May 2014 Available online 28 June 2014

This study investigated the performance of a reactor in which denitrification was integrated into the anaerobic acidogenic process. Industrial wastewater cassava stillage was used as the carbon source, and the nitrate reduction pathway and its effects on acid fermentation were examined. Results from batch and semi-continuous tests showed that the presence of nitrate did not inhibit anaerobic acidification but altered the distribution of volatile fatty acid (VFA) species. Nitrate reduction was attributable to denitrification and to dissimilatory nitrate reduction to ammonia (DNRA). The ratio of DNRA to denitrification was proportional to the ratio of COD=NO3 L eN. After 130 days of semi-continuous operation, denitrification removal efficiency accounted for about 60% at a COD=NO3 L eN of 50. The proportional distribution of VFAs was acetate, followed by propionate and then butyrate. The polymerase chain reactionedenaturing gradient gel electrophoresis results confirmed the contributions of denitrification and DNRA in the nitrate-amended reactor and showed that the addition of nitrate enriched the structure of the bacterial community, but did not suppress the activity of acid-producing bacteria. Ó 2014, The Society for Biotechnology, Japan. All rights reserved. [Key words: Acidogenesis; Denitrification; Dissimilatory nitrate reduction to ammonia; Simultaneous; Microbial community]

Anaerobic digestion processes are widely applied in ethanol wastewater treatment and have become well established. During anaerobic digestion, most of the organic pollutants in the highstrength wastewater can be biodegraded to biogas (CH4, CO2). However, the degradation of organic nitrogen results in an effluent containing high levels of ammonia, which is hardly removed under anaerobic conditions and requires further treatment in the followup biological nitrification and denitrification processes. In 1994, Akunna and co-workers reported that carbon and nitrate could be simultaneously removed in the anaerobic filter, with direct recirculation of the aerobic effluent (1). Since then, the effect of nitrate and its reduction products on simultaneous denitrification and methanogenesis has been extensively studied (2e4). For example, it was shown that nitrate reduction is superior to methanogenesis in the anaerobic reactor but that nitrogen-containing metabolic compounds could temporarily inhibit methanogenesis. Moreover, the calorific value of the generated biogas is lowered by the presence of nitrogen gas. In attempts to resolve this problem, the simultaneous removal of carbon and nitrogen was investigated, by coupling two-phase acidogenesis and methanogenesis. In those studies, anaerobic activity was not inhibited when the effluxed nitrates were recycled to the acidogenic reactor and complete denitrification occurred (5,6). The nitrate reduction pathway is known to be influenced by the carbon source and by the C/N ratio, either through denitrification or dissimilatory nitrate reduction to ammonia (DNRA). The latter occurs prior to denitrification under

* Corresponding author. Tel.: þ86 21 6598292; fax: þ86 21 659886313. E-mail addresses: [email protected], [email protected] (L. Xie).

conditions of high C/N or easily biodegradable COD, e.g., in wastewater with a high carbohydrate content (7,8). However, since these findings were obtained in studies of simultaneous denitrification and methanogenesis, the roles of the C/N ratio and carbon source in simultaneous denitrification and acidogenesis require further investigation. In addition, synthetic organic wastewater and pure substrates were often used with the aim of investigating the C/N ratio and the nature of the carbon substrates. Consequently there is a lack of experience with practical wastewater. Therefore in this work, an industrial wastewater, cassava stillage, was used to determine the feasibility of simultaneous denitrification and acidogenesis. Specifically, the effect of the C/N ratio on the nitrate reduction pathway and acid fermentation was examined in batch tests, followed by the realization of simultaneous denitrification and acidification in a single acidogenic bioreactor. The performance of this system and the compositions of the microbial communities were then evaluated. MATERIALS AND METHODS Inoculum and substrates Anaerobic granular sludge acquired from the fullscale, mesophilic, anaerobic internal circulation reactor of a paper mill (Jiangsu, China) was used directly as inoculum without acclimation to nitrate or nitrite. Raw cassava stillage (CS) wastewater, obtained as the effluent from a cassava ethanol plant (Jiangsu, China), was chosen because of its high solid organic wastewater content and good biochemical degradability. Table 1 summarizes the characteristics of the CS used in this study. Nitrate was added in the form of sodium nitrate. Both CS and sodium nitrate were stored at 4
C until needed. Batch experiments The effects of nitrate and COD=NO3  eN on acidification and the nitrate reduction pathway were investigated in batch assays. A series of identical serum bottles were used as reactors with a working volume of 500 mL. 50 mL of anaerobic granular sludge and 250 mL of CS were added to each reactor and

1389-1723/$ e see front matter Ó 2014, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2014.05.027

96

XIE ET AL.

J. BIOSCI. BIOENG.,

TABLE 1. Characteristics of cassava ethanol wastewater. Parameter Total solids Volatile solids Total COD Soluble COD Total nitrogen NH4 þ eN

Average value (mg/L) 71500 58600 73200 40800 1700 3.1

then diluted to the desired working volume. The initial pH of each reactor was controlled at 8.00  0.1, using 5 N NaOH and then pH was not adjusted after sampling during batch assays. The bottles were purged with argon for 1 min to avoid residual oxygen interference. The initial total and soluble chemical oxygen demand (TCOD and SCOD) were maintained at around 45,000 mg/L and 21,000 mg/L, respectively. The concentrations of total and volatile solids (TS and VS) in the reactor were, respectively, around 49,100 mg/L and 35,900 mg/L. NaNO3 from stock solutions was added to each reactor to obtain C/N ratios of 30, 41, 77 and 152 in the CSfed cultures. A control culture without nitrate amendment was also prepared. All cultures were incubated at 35
C in a reciprocating water bath shaker set at 120 rpm. Gas bags were used to collect the evolved biogas and its amount and composition were determined at regular intervals. In addition, 20-mL samples were taken for analysis. Semi-continuous experiments A semi-continuous assay was performed in serum bottles with a working volume of 210 mL. The hydraulic retention time (HRT) was 3 days. The reactors were operated under a complete mixing condition, and sludge retention time (SRT) was the same as HRT. The initial TCOD was 30,000 mg/L; SCOD ranged from 14,000 to 20,000 mg/L. Aliquots of NaNO3 stock solutions were added to the influent, such that the initial NO3  eN concentration was 600 mg/L and the COD=NO3  eN ratio was around 50 in the anaerobic CS-fed cultures. A blank reactor of CS alone was prepared for comparison. Every 24 h, 70 mL of the mixture was removed for analysis, substituting the same amount of fresh substrate in the bottles. The initial pH was controlled after sampling at 8.00  0.1, using 5 N NaOH. The final pH of reactors amended with nitrate or not was observed to be 6.6  0.2 and 6.1  0.2 respectively at every 24 h during steady state. The bottles were purged with argon gas for 1 min to maintain anaerobic conditions. The cultures were incubated at 35
C in a reciprocating water bath shaker set at 120 rpm. In this study, steady state was defined as a sustained concentration of nitrogen compounds within 10% deviation. During this period, COD, volatile fatty acids (VFAs), and other parameters were determined for ten consecutive days. Analytical methods Standard methods (9) were used to measure TS, VS, COD, total nitrogen (TN), and NH4 þ eN. The collected liquid samples were centrifuged at 11,000 rpm for 10 min and then filtered through a 0.45-mm filter membrane prior to analysis. Nitrate and nitrite were analyzed by ion chromatography in a system equipped with two conductivity detectors and two sets of columns (Dionex ICS-3000). Separation and elution of the anions were carried out on an IonPac AG11-HC (4  50 mm) guard column and an IonPac AS11-HC (4  250 mm) analytical column. The eluent was 18 mM KOH and the isocratic flow rate was 1.2 mL/min. Cations were analyzed on an IonPac CG12A (4  50 mm) guard column and an IonPac CS12A (4  250 mm) analytical column. The eluent was 20 mM methanesulfonic acid and the isocratic flow rate was 1.0 mL/min. Auto suppression mode was used during the detection. Samples for VFA determinations were diluted with 3% (v/v) H3PO4 (sample/acid, 1:1 v/v) and then analyzed on a gas chromatograph (Agilent, 6890 N) equipped with a flame ionization detector and a CPWAX52CB analytical column (30 m  0.25 mm  0.25 mm). The temperature of the injector and detector were 200
C and 220
C, respectively. Nitrogen served as the carrier gas; the flow rate was 50 mL/min. The GC oven was programmed to begin at 110
C and to remain at that temperature for 2 min, followed by increases at a rate of 10
C/min to 220
C, with a final hold at 220
C for an additional 2 min. The sample injection volume was 1.0 mL. When the TN removal efficiency in the continuous reactor reached steady state, 1.5 ml of mixed culture was withdrawn from the reactors for microbial community analysis. The culture was centrifuged (12,000 g, 4
C, 5 min) and the solids were separated. DNA was extracted using the QIAamp DNA stool mini kit (Qiagen, Germany) according to the manufacturer’s instructions. The w170-bp fragments of the V3 region in the 16S rRNA gene of the bacteria were amplified from the extracted DNA by polymerase chain reaction (PCR) with the forward primer F357-GC (50 -CGC CCG CCG CGC GCG GGC GGG GCG GGG GCA CGG GGG GCC TAC GGG AGG CAG CAG30 ) and the reverse primer R518 (50 -ATT ACC GCG GCT GCT GG-30 ) (10). The PCR conditions, targeted for bacteria, were as follows: 4 min initial denaturation at 94
C, 30 s at 94
C, 1 min at 50
C, and 1 min at 72
C for 30 cycles, with a final extension for 7 min at 72
C. Denaturing gradient gel electrophoresis (DGGE) analysis of the PCR amplicons was performed with the Dcode universal mutation detection system (Bio-Rad, USA) on 8% (v/v) polyacrylamide gels in 1 TAE and a denaturant gradient of 30e55%. A 100% denaturing solution consisted of 7 M urea and 40% (v/v) formamide. Same volume of PCR products (nearly 400 ng) was loaded into each gel lane; however, DNA concentration was not controlled in two samples. Electrophoresis was run at

60
C at a constant voltage of 60 V for 16 h. After electrophoresis, the gel was stained with SYBR Green I, photographed under UV transillumination, and documented using the GelDoc system (Bio-Rad). The target bands were removed from the DGGE gels with sterile pipette tips and placed in sterile vials. DNA was extracted from the DGGE gels with the abovedescribed DNA extraction kit. The extracted DNA was re-amplified using the primers and conditions described above. The PCR products were then sent for sequencing. Sequencing similarity searches were performed using the Basic Local Alignment Search Tool (BLAST) and used to search the National Centre for Biotechnology Information Sequence Database (http://www.ncbi.nlm.nih.gov/BLAST/).The diversity of the microbial communities was expressed by the ShannoneWiener index of diversity H0 . H0 was calculated on the basis of the band intensity on the gel tracks, which was reflected as peak heights in the densitometric curve. The equation for the Shannon index is:

H0 ¼ 

X .   .  ni N log ni N

(1)

where ni was the height of the peak and N was the sum of all peak eights of the densitometric curve.

RESULTS AND DISCUSSION Nitrate-amended acid fermentation and the nitrate reduction pathway in the batch experiments Fig. 1 shows the changes in the VFA profiles with reaction time in CS-fed cultures amended or not with 600 mg NO3  eN=L. Initially, the amount of total VFAs in the nitrated-amended culture was similar to that in the nitrate-free culture. After 48 h of fermentation, VFA concentrations were about 50% lower but they progressively increased such that by the end of fermentation they had reached about 13,000 mg/L in both cultures. The main VFA species in the nitrate-free culture after 108 h of fermentation were propionate, acetate, and butyrate, in that order. However, with the addition of nitrate into the system, the butyrate concentration decreased significantly such that the VFAs consisted almost exclusively of acetate and propionate. Compared to the nitrate-free culture, the acetate content increased from 33% to 48%, propionate was present in similar amounts, but butyrate decreased from 17% to 0.04%. This change in the VFA distribution may have reflected carbon source utilization caused by nitrate reduction. Fig. 2 shows the change in nitrate, nitrite, ammonium, and TN as a function of the reaction time in nitrate-amended cultures. The initial nitrate concentration of around 600 mg-N/L was reduced completely at a reaction time of 60 h. Nitrite, as the metabolic product, initially accumulated but by 96 h was completely degraded. The ammonium concentration increased continuously, reaching 265 mg-N/L after 96 h; by contrast, in the blank control culture the measured ammonium concentration was only 40 mg-N/ L after 96 h (Fig. 2A), indicating that nitrate was partially reduced to ammonium and that DNRA had occurred. The change in the TN concentration as a function of the reaction time is further described in Fig. 2B. The significant decrease in TN in the nitrate-amended culture corresponded to a removal efficiency of 38.9% at the end of the reaction, compared to the less than 5% TN removal in the nitrate-free culture. This decrease in TN was due to denitrification. Thus, in this study, the nitrate reduction pathway was completed with denitrification, as the main pathway, but also by DNRA. The nitrate reduction pathway and VFA production at different COD=NO3 L eN ratios To further evaluate the effect of the COD=NO3  eN ratio on nitrate reduction, the changes in the concentrations of nitrogen compounds at different ratios were determined (Table 2). In the blank reactor, the ammonium concentration did not change significantly whereas in the test reactors it increased, indicative of DNRA. Regardless of the small amount of microbial ammonia assimilation, the decrease in the TN concentration was attributable to denitrification. As shown in Table 2, the nitrate reduction pathway was sensitive to the COD=NO3  eN ratio. As the ratio increased from 30 to 152,

VOL. 119, 2015

NITROGEN AND CARBON REMOVAL IN ACIDOGENIC SYSTEM

16000

Blank

3

Nitrogen compound conc (mg/L)

HPr HAc

12000

-

600mg/L NO3 -N

600

HBu

14000

VFAs Conc (mgCOD/L)

A

-

600mg/L NO -N

10000 8000 6000 4000 2000

97

Blank

-

NO -N 2

500

-

NO -N 3 +

400

NH -N 4

300 200 100 0

0 0

48

24

72

96

108

0

20

40

60

80

100

120

Time (h)

Time (h)

B

FIG. 1. Comparison of VFA profiles in reactors amended or not with nitrate.

1000 900

nitrogen removal through denitrification decreased from 85.5% to 43.7% whereas DNRA became much more important, increasing from 14.5% to 56.3%. Denitrification was the main nitrate reduction pathway at a COD=NO3  eN ratio below 77. At a ratio of 117, DNRA accounted for 56.3% of nitrogen removal. This result was in agreement with that of a previous study showing a decrease in denitrification activity as the COD=NO3  eN ratio increased, such that DNRA became the main reduction process when the ratio climbed above 130 (11). The difference in the critical COD=NO3  eN ratio to retard DNRA might be due to the wastewater species and compositions, considering that pure and synthetic glucose wastewater were used as carbon source in the work by Rustrian et al. (5). Thus, to achieve optimal TN removal from CS, denitrification is the desired process, in which case the COD=NO3  eN ratio should be maintained below 77. The analyses of VFA production and carbon source utilization during denitrification at different COD=NO3  eN ratios are summarized in Table 3. Assuming that denitrification was the only nitrate reduction pathway, the minimal amount of VFAs utilized as carbon source could be estimated by VFAs ¼ 2:86CNO3  eN, where CNO3  eN is the amount of reduced NO3  eN. Based on the further assumption that the amount of total VFAs is the sum of the amounts of residual VFAs and the VFAs utilized as carbon source during denitrification, then the amounts of total VFAs at COD=NO3  eN ratios of 152 and 77 were higher in the nitrate-amended than in the blank reactor. Accordingly, the introduction of the appropriate amount of nitrate into the acidogenic phase did not suppress acidification. Total VFA concentrations at COD=NO3  eN ratios of 41 and 30 were about 20% lower than the concentration in the blank, probably because acidogenesis was suppressed by newly formed nitrite during nitrate reduction. The accumulated nitrite concentration was 403 mg-N/L and 586 mg-N/L respectively. Rustrian and co-workers also found that complete denitrification in the

(mg/L)

800 -

700

600mg/L NO3 -N

600

TN

Blank

TN

500 400 300 200 0

20

40

60

80

100

120

Time (h) FIG. 2. Changes in NH4 þ eN (A) and TN (B) concentrations in the reactors.

acidogenic reactor did not suppress anaerobic activity and that the generated VFAs could be used as the carbon source for denitrification (5). In this study, the presence of nitrate in the acidogenic reactor did not permanently suppress acidogenesis; instead, suppression was temporary, occurring under conditions of relatively high nitrite concentrations of about 400e600 mg-N/L. With a decrease in nitrite, acidogenic activity recovered. Fig. 3 shows the variations in the individual VFA species with reaction time. The acetate profiles at the different COD=NO3  eN ratios were initially similar to the profile in the blank; however, as the reaction time preceded, acetate levels at COD=NO3  eN ratios of 152 and 77 were higher, corresponding to the completion of denitrification. At COD=NO3  eN ratios of 41 and 30, the acetate concentrations were lower than in the blank control, suggesting the use of acetate as carbon source during nitrate reduction. As shown in Fig. 3B and C, the propionate and butyrate contents at the tested ratios were much lower than the value in the blank. pH is also an

TABLE 2. Changes in the concentrations of nitrogen compounds at different COD=NO3  eN ratios. 

COD=NO3 eN

30 41 77 152

Initial concentration (mg/L)

Final concentration (mg/L)

NO3  eN

NH4 þ eN

TN

NO3  eN

NH4 þ eN

TN

1512 1008 612 301

27.6 7.4 15.0 11.2

1865 1377 910 653

3.3 3.0 4.1 3.0

234 272 265 172

592 574 587 475

DNRA%a

Denitrification%b

14.5 26.1 48.1 56.3

85.5 73.9 51.9 43.7

Dr*TN is the net TN variation, which is equal to the variations in TN at different COD=NO3  eN ratios minus the TN variation in the blank. a DNRA% ¼ DrNH4 þ eN =ðDr*TN þ DrNH4 þ eN Þ: b Denitrification% ¼ Dr*TN =ðDr*TN þ DrNH4 þ eN Þ:

98

XIE ET AL.

J. BIOSCI. BIOENG., TABLE 3. VFA production and nitrate utilization rate at different COD=NO3  eN ratios. 

COD=NO3 eN

NO3 eN removal amount (mg/L)

VFA utilized for NO3  eN removal (gCOD/gN)

Residual amount of VFAs in effluent (gCOD/gVFA)

Total VFA production (gCOD/gVFA)

VFA production vs. the blank

300 609 1005 1508 0

857 1741 2875 4314 0

13742 13361 8356 6606 14162

14599 15102 11230 10919 14162

þ þ  

152 77 41 30 Blank

A

important factor to determine VFAs composition according to the previous paper (12). However, in this study, pH variations with reaction time at COD=NO3  eN ratio of 152 did not present much difference with that in control group, and even at COD=NO3  eN ratios of 41 and 30 the pH was in the range of 7e7.5. Our previous study showed that distribution of VFAs compositions under controlled pH of 6e8 was similar. Thus, an explanation for the above result is that nitrate addition altered the fermentation type from butyrate to acetate dominated. Noike et al. also reported an increase in the acetate concentration following the introduction of nitrate into acidified municipal solid waste (13). In a further study we will examine the acidification pathway under nitrate- and/or nitrite-amended conditions.

Blank

6000

-

COD/NO3 -N=152 -

5000

COD/NO3 -N=77

Acetate(mg/L)

-

COD/NO3 -N=41

4000

-

COD/NO3 -N=30

3000

2000

1000

0 0

20

40

60

80

100

120

80

100

120

Time (h)

B

5000

Blank -

4000

COD/NO3 -N=152 -

COD/NO3 -N=77

Propionic (mg/L)

-

3000

COD/NO3 -N=41 -

COD/NO3 -N=30

2000

1000

0

0

20

40

60

Time (h)

C

1400

Nitrate reduction and its effect on acidogenesis in semicontinuous experiments To further evaluate the influence of nitrate introduction on both the acidification fermentation pathway and nitrate reduction, semi-continuous experiments were conducted in which the amended nitrate concentration was 600 mg-N/L. Fig. 4 shows the effluent ammonia content and the calculated denitrification efficiency during the operation time of 130 days. Nitrate in the effluent was not detected at a hydraulic retention time of 3 days whereas the ammonia concentration was higher than the nearly 1 mg-N/L in the blank, indicating that nitrate was partially reduced to ammonia through DNRA. TN removal in the amended system was attributable to denitrification regardless of microbial ammonia assimilation. In this study, the denitrification removal efficiency could be calculated as DTN=DNO3  eN, where DTN is the variation in the TN concentration minus the DTN in the blank and DNO3  eN is the amount of reduced NO3  eN. As depicted in Fig. 4, the initial ammonia concentration was within 30 mg-N/L and then increased continuously until, after 130 days operation, it reached about 180 mg-N/L. The denitrification removal efficiency was about 60%, which was in agreement with the results from batch

Blank

Butyrate (mg/L)

1000

90

-

COD/NO3 -N=77

80

-

150

COD/NO3 -N=41

NH4+-N Conc (mg/L)

1200

-

800

100

200

-

COD/NO3 -N=152

COD/NO3 -N=30

600 400

70 60 50

100

40 COD/NO -N=50

50

Denitrification efficiency

NH -N

CS alone

30

NH -N

20

200

Denitrification efficiency



10 0

0 0

20

40

60

80

100

120

Time (h) FIG. 3. Changes in acetate (A), propionic (B) and butyrate (C) concentrations at different COD=NO3  eN ratios.

0 20

40

60

80

100

120

Time (d) FIG. 4. Ammonia concentrations in reactors with or without nitrate and the denitrification efficiency of the nitrate-supplemented reactor.

VOL. 119, 2015

NITROGEN AND CARBON REMOVAL IN ACIDOGENIC SYSTEM

99

7000 6500

CODmg/L

5500

VFAs Conc

6000

3500

5000

Acetate Propiomic Butyrate Total VFAs

4500 4000 3000 2500 2000 1500 1000 500 0

Nitrate addition

CS wastewater alone

FIG. 5. VFA concentration in reactors amended with nitrate and with CS wastewater alone.

tests and implied that in our system denitrification was the main nitrate reduction pathway. The average VFA concentrations in the nitrate-amended acidification reactor and the blank reactor between operation days 22 and 120 are shown in Fig. 5. The change in the TN removal efficiency did not affect either the production amount or the composition of VFAs, suggesting that denitrification and DNRA proceeded via similar VFA utilization patterns. As also seen in Fig. 5, residual VFAs in the blank were 738 mg COD/L higher than in the nitrateamended reactor. As noted in Nitrate reduction pathway and VFA production, the amount of total VFAs is the sum of the utilized and residual VFAs; thus, the amount of VFAs consumed during denitrification was about 1608 mg COD/L at a nitrate concentration of 600 mg NO3  eN=L. The amount of total VFAs actually produced in the nitrate-amended acidification reactor was higher than in the blank, further indicating that anaerobic fermentation was not suppressed following the introduction of nitrate in the continuous experiment. The VFA compositions in the nitrate-amended and blank reactors were similar, but the percentages of the individual species were different, in agreement with the results from the batch tests. The acetate concentration was higher than in the blank, but the propionic and butyrate concentrations were much lower. Microbial communities When the conditions in each reactor reached steady state, the microbial community structure was sampled and analyzed by PCReDGGE (Fig. 6). Each band on the DGGE profile corresponds to a fragment of unique 16S rRNA gene sequences and thus represents a distinct species in the microbial community. As expected, the microbial communities differed in terms of diversity and relative abundance in the reactor containing CS alone vs. the reactor in which CS was amended with nitrate. The ShannoneWiener indexes for reactors with and without nitrate amendment were 2.773 and 2.079 respectively, revealing that there was a significantly greater biological diversity in both reactors. Band 4 and band 13 were relatively abundant in both reactors with or without nitrate amendment. Sequencing results (Table 4) showed that these bands were representative of Parabacteroides distasonis (91% similarity), an anaerobic acidogenic species whose primary metabolic product is acetate (14), and Chloroflexi (88% similarity), respectively. Both are common in anaerobic digestions. The relative abundance of acidproducing bacteria in the nitrate-amended reactor further suggested that the addition of nitrate did not suppress the activity of acid-producing bacteria. Other bands were not detectable in the reactor containing CS wastewater alone. For

FIG. 6. DGGE bands of microbial communities in reactors with CS wastewater alone (lane A) or amended with nitrate (lane B).

example, bands 12 and 14 were closely related to Bacillus, one of the main genera involved in denitrification and DNRA (15), and bands 1 and 3 were, respectively, related to Corynebacterium (88% similarity) and Bacteroides (93% similarity), two genera known to

TABLE 4. Sequencing results of 16S rRNA fragments obtained from DGGE gels. Band 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Organism affiliation Corynebacterium Prevotella buccalis Bacteroides helcogenes P 36-108 Parabacteroides distasonis ATCC 8503 Clostridium sporosphaeroides Prevotella oris Clavibacter michiganensis subsp. sepedonicus Clostridium nexile DSM 1787 Bacillus lentus Bacillus pumilus SAFR-032 Prevotella bivia Bacillus megaterium QM B1551 Dehalococcoides sp. GT Bacillus herbersteinensis Proteiniphilum acetatigenes Bifidobacterium thermacidophilum subsp. porcinum [Brevibacterium] frigoritolerans

Identity

Phylum

Accession no.

88% 89% 93% 91%

Actinobacteria Bacteroidetes Bacteroidetes Bacteroidetes

NR_074826 NR_044630 NR_074546 NR_074376

94% 96% 89%

Firmicutes Bacteroidetes Actinobacteria

NR_044835 NR_044628 NR_074600

98% 96% 94% 92% 98% 88% 94% 92% 97%

Firmicutes Firmicutes Firmicutes Bacteroidetes Firmicutes Chloroflexi Firmicutes Bacteroidetes Actinobacteria

NR_029248 NR_040792 NR_074977 NR_044629 NR_074290 NR_074288 NR_042286 NR_043154 NR_025672

95%

Firmicutes

NR_042639

100

XIE ET AL.

J. BIOSCI. BIOENG.,

participate in denitrification (15,16). Bands 5 and 8 were closely related to Clostridium, previously suggested to play a role in DNRA (17). The numerous bands in lanes from the reactor containing nitrate-amended CS were indicative of the greater diversity of the microbial communities in this system than in the reactor containing CS alone. Moreover, our results show that in the former system both denitrification and DNRA contributed to the nitrate reduction pathway. Conclusions In this study, industrial wastewater CS was used as the carbon source for the integration of denitrification and acidogenesis. Batch results showed the complete reduction of nitrate generated in the acidification phase, either through denitrification or DNRA. The proportion of these two processes in nitrate reduction was related to the COD=NO3  eN ratio, with denitrification as the main pathway at a ratio below 77. The VFAs produced in the acidification process were utilized as the carbon source in the nitrate reduction process. The presence of nitrate in the anaerobic fermentation system did not inhibit acidogenesis but changed fermentation metabolites, such that acetate, rather than propionate, was the dominant species. The semi-continuous operation of the nitrate-amended reactor for nearly 130 days indicated that denitrification was the main pathway for nitrate reduction at a COD=NO3  eN of 50 and that the presence of nitrate did not suppress anaerobic microbial activity but instead changed the proportion of the main VFA components, with acetate as the main VFA species. PCReDGGE analysis further showed that nitrate addition enriched the structure of the bacterial community, and did not suppress the activity of acid-producing bacteria. Bacteria able to carry out denitrification and DNRA were also detected. Thus, our study showed the integration of heterotrophic denitrifiers with anaerobic acidogenesis at the proper COD=NO3  eN ratio provides a strategy for the simultaneous removal of organic carbon and nitrogen, in which heterotrophic denitrification occurs without inhibiting acidogenesis.

ACKNOWLEDGMENTS This research was supported by the National Natural Science Foundation of China (51178326), and the National Hi-Tech Research and Development Program of China (2011AA060903).

References 1. Akunna, J., Bizeau, C., Moletta, R., Bernet, N., and Heduit, A.: Combined organic-carbon and complete nitrogen removal using anaerobic and aerobic upflow filters, Water Sci. Technol., 30, 297e306 (1994). 2. Mosquera-Corral, A., Sanchez, M., Campos, J. L., Mendez, R., and Lema, J. M.: Simultaneous methanogenesis and denitrification of pretreated effluents from a fish canning industry, Water Res., 35, 411e418 (2001). 3. Baloch, M. I., Akunna, J. C., and Collier, P. J.: Carbon and nitrogen removal in a granular bed baffled reactor, Environ. Technol., 27, 201e208 (2006). 4. Xie, L., Chen, J. R., Wang, R., and Zhou, Q.: Effect of carbon source and COD/ NO 3 eN ratio on anaerobic simultaneous denitrification and methanogenesis for high-strength wastewater treatment, J. Biosci. Bioeng., 113, 759e764 (2012). 5. Rustrian, E., Delgenes, J. P., Bernet, N., and Moletta, R.: Simultaneous removal of carbon, nitrogen and phosphorus from wastewater by coupling two-step anaerobic digestion with a sequencing batch reactor, J. Chem. Technol. Biotechnol., 73, 421e431 (1998). 6. Park, S. M., Park, N. B., Seo, T. K., and Jun, H. B.: Nitrogen removal in a phaseseparated up-flow anaerobic sludge blanket reactor combined with biological nutrient removal, Environ. Eng. Sci., 26, 397e405 (2009). 7. Akunna, J. C., Bizeau, C., and Moletta, R.: Denitrification in anaerobic digesters: possibilities and influence of wastewater COD/N-NOX ratio, Environ. Technol., 13, 825e836 (1992). 8. Ruiz, G., Jeison, D., and Chamy, R.: Development of denitrifying and methanogenic activities in USB reactors for the treatment of wastewater: effect of COD/N ratio, Process Biochem., 41, 1338e1342 (2006). 9. APHA, AWWA, and WPCF: Standards methods for the examination of water and wastewater, 19th ed. APHA, Washington, D.C. (1995). 10. Muyzer, G., Dewaal, E. C., and Uitterlinden, A. G.: Profiling of complex microbial-populations by denaturing gradient gel-electrophoresis analysis of polymerase chain reaction-amplified genes-coding for 16s ribosomal-RNA, Appl. Environ. Microbiol., 59, 695e700 (1993). 11. Rustrian, E., Delgenes, J. P., Bernet, N., and Moletta, R.: Nitrate reduction in acidogenic reactor: influence of wastewater COD/N-NO3 ratio on denitrification and acidogenic activity, Environ. Technol., 18, 309e315 (1997). 12. Yu, H. Q. and Fang, H. H. P.: Acidogenesis of dairy wastewater at various pH levels, Water Sci. Technol., 45, 201e206 (2002). 13. Noike, T., Goo, I. S., Matsumoto, H., and Miyahara, T.: Development of a new type of anaerobic digestion process equipped with the function of nitrogen removal, Water Sci. Technol., 49, 173e179 (2004). 14. Nelson, M. C., Morrison, M., and Yu, Z.: A meta-analysis of the microbial diversity observed in anaerobic digesters, Bioresour. Technol., 102, 3730e3739 (2011). 15. Philippot, L. and Germon, J. C.: Contribution of bacteria to initial input and cycling of nitrogen in soils, pp. 159e176, in: Varma, A. and Buscot, F. (Eds.), Microorganisms in soils: Roles in genesis and functions, vol. 3. Springer, Berlin, Heidelberg (2005). 16. Wallenstein, M. D., Myrold, D. D., Firestone, M., and Voytek, M.: Environmental controls on denitrifying communities and denitrification rates: insights from molecular methods, Ecol. Appl., 16, 2143e2152 (2006). 17. Caskey, W. H. and Tiedje, J. M.: The reduction of nitrate to ammonium by a Clostridium sp. isolated from soil, J. Gen. Microbiol., 119, 217e223 (1980).