Bioaugmentation and Adsorption Treatment of Coking Wastewater

Feb 3, 2011 - Peking University, Beijing 100871, People's Republic of China. ‡. Institute of Environmental Health and Related Product Safety, Chines...
0 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/est

Bioaugmentation and Adsorption Treatment of Coking Wastewater Containing Pyridine and Quinoline Using Zeolite-Biological Aerated Filters Yaohui Bai,† Qinghua Sun,‡,† Renhua Sun,† Donghui Wen,*,† and Xiaoyan Tang† †

College of Environmental Sciences and Engineering, The Key Laboratory of Water and Sediment Sciences (Ministry of Education), Peking University, Beijing 100871, People’s Republic of China ‡ Institute of Environmental Health and Related Product Safety, Chinese Center for Disease Control and Prevention, Beijing 100050, People’s Republic of China

bS Supporting Information ABSTRACT: Bioaugmented zeolite-biological aerated filters (Z-BAFs), i.e. adding isolated degrading bacteria into the BAFs with zeolite as fillings, were designed to treat coking wastewater containing high concentrations of pyridine and quinoline and to explore the bacterial community of biofilm on the zeolite surface. The investigation was carried out for 91 days of column operation and the treatment of pyridine, quinoline, total organic carbon (TOC), and ammonium was shown to be highly efficient by bioaugmentation and adsorption. Biomass determination and bacterial diversity detection based on 16S rDNA and rRNA techniques supported the treatment data and indicated that bioaugmentation could recover the bacterial richness and diversity from pyridine and quinoline loading shocks. However, bioaugmentation accelerated the shift of the bacterial community structure resulting in a more distinct difference from the starting community. Clone library analysis revealed that pyridine and quinoline were more harmful to Bacterodietes among all ingenious bacteria, and bioaugmentation promoted the growth of Planctomycetes in the biofilm. Moreover, the introduced bacteria did not remain dominant in the bioaugmented biofilm, indicating the indigenous degrading bacteria played the most significant role in the treatment. This bioaugmented Z-BAF method was shown to be an alternative technology for the treatment of wastewater containing pyridine and quinoline or other N-heterocyclic aromatic compounds.

1. INTRODUCTION Pyridine and quinoline are recognized as two hazardous N-heterocyclic aromatic compounds and are often present in coking, refinery, and pharmaceutical wastewater.1-3 Bioaugmentation with biodegradative strains has been shown to be a costeffective technology for pyridine- and/or quinoline-containing wastewater treatment.4,5 However, the ammonium biotransformed from pyridine and quinoline is still a serious problem which may lead to the disruption of biological treatment and eutrophication if such wastewater is discharged into the water environment. Our short-term experiments carried out in flasks showed that using highly efficient degrading bacteria and zeolite had abilities to remove pyridine, quinoline (biodegradation), and their byproduct - ammonium (adsorption).6 So one purpose of this study was to assess the long-term effectiveness of highly efficient degrading bacteria and zeolite-biological aerated filters (Z-BAFs) for the bioaugmented treatment of coking wastewater containing pyridine and quinoline and discuss further applications of this method. The bioaugmentation process has always been regarded as a “black box” model, until the introduction of culture-independent techniques in recent years. It is essential to obtain ecological data regarding the fate and activity of inoculated microorganisms and their interactions with indigenous microbial communities during the process.7 Therefore, the bacterial diversity and composition in the biofilm attached to the zeolite was investigated in this study using molecular ecological techniques. r 2011 American Chemical Society

Few studies investigated the bioaugmentation process and associated ecology in the biofilm. Therefore, another purpose of this study was to establish a relationship between the treatment efficiency and the bacterial ecological data to expand our knowledge of the bioaugmentation process and provide useful clues for the bioremediation of other N-heterocyclic aromatic compounds.

2. MATERIALS AND METHODS 2.1. Zeolite. Natural zeolite, composed mainly of clinoptilolite and quartz, was obtained from Jinyun, Zhejiang province, China. In order to allow more microorganisms to attach to the zeolite, the natural zeolite was modified to have more mesopores and macropores on its surface. The modified zeolite had little change in the crystal structure and chemical composition comparing with the natural zeolite. More details regarding the natural zeolite and modified zeolite were described previously.6 Both natural and modified zeolites were sterilized before use in the experiments. 2.2. Wastewater and Activated Sludge. Raw wastewater was obtained from the adjusting tank of the coking wastewater Received: September 16, 2010 Accepted: January 11, 2011 Revised: December 19, 2010 Published: February 03, 2011 1940

dx.doi.org/10.1021/es103150v | Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology

ARTICLE

Figure 1. Schematic diagram of the Z-BAFs.

treatment plant in Capital Iron and Steel Corporation, Beijing, China. The wastewater quality was as follows: 41.2 mg pyridine/ L, 28.0 mg quinoline/L, 533.9 mg total organic carbon (TOC)/ L, 1400 mg COD/L, 127.7 mg NH3-N/L, 149.8 mg NO3--N/ L, and 5.2 mg NO2--N/L. The raw wastewater was diluted 1:2(v/v) in deionized water, and pyridine and quinoline were added to give prominence to bioaugmentation and explore the hazardous impact on the indigenous bacterial community. This high pyridine- and quinoline-concentrated wastewater was the influent of each column. The dissolved oxygen (DO) in each column was 5-8 mg/L and the room temperature was 1620 °C during the operation. Activated sludge was gathered from the secondary sedimentation tank of the same coking wastewater treatment plant. 2.3. Bacteria and Their Enrichment. A pyridine-degrading bacterium (Paracoccus sp. BW001)8 and a quinoline-degrading bacterium (Pseudomonas sp. BW003)9 were used in the experiment. Both bacteria were isolated from the activated sludge in the coking wastewater treatment plant of Wuhan Iron and Steel Corporation, Hubei Province, China. The strain BW001 was cultivated in Luria-Bertani (LB) medium10 with 500 mg of pyridine/L, and the strain BW003 was cultivated in the same medium with 500 mg of quinoline/L. Both were incubated at 30 °C and shaken at 180 rpm in a rotary shaker until the bacteria reached the logarithmic phase of growth. The bacterial cells were harvested by centrifugation at 3000  g for 5 min. The cells were washed three times with 20 mL of sterile deionized water. The bacterial deposit was resuspended by vortex and diluted with 50 mL of water. The bacterial suspension was employed as the inoculum in the Z-BAFs experiment immediately after its preparation. 2.4. Z-BAFs Configuration and Operating Conditions. The experiment was conducted using a Z-BAFs system (Figure 1). The system consists of three identical and independent columns that are 40 mm in diameter and 400 mm in height. The heights of the graded gravel layer and the filling layer were 50 mm and 210 mm, respectively. Two columns were filled with natural zeolite and one with modified zeolite. At the beginning, three columns were seeded with coking activated sludge (0.58 g, dry weight at 105 °C) and domesticated with raw coking wastewater. After 15 d, two columns were inoculated with the two highly efficient bacteria (0.012 g, 1:1 in proportion). Therefore, the three filters were operated with different protocols: (1) Column 1, bioaug-

mented Z-BAF (natural zeolite); (2) Column 2, bioaugmented Z-BAF (modified zeolite); and (3) Column 3, nonbioaugmented Z-BAF (natural zeolite). The operation was divided into three phases, first phase (0-31 d): the concentrations of pyridine and quinoline were about 60 mg/L, and the hydraulic retention time (HRT) was 37.5 h; second phase (32-67 d): pyridine concentration was 101-113 mg/L, quinoline concentration was 95-134 mg/L, and the HRT was 37.5 h; and third phase (68-91 d): pyridine concentration was 119-130 mg/L, quinoline concentration was 99-128 mg/L, and the HRT was 20 h. The pH and DO values of the influent and effluent were monitored at a regular time interval during the operation. For analyzing the concentrations of pyridine, quinoline, TOC, NH3-N, NO2--N, and NO3--N, a portion of the effluent and influent samples was filtered through a 0.45 μm membrane. 2.5. Analytical Methods. Pyridine and quinoline concentrations were analyzed using a high performance liquid chromatography (HPLC) system (Shimadzu LC10ADVP, SPD10AVP UVvis Detector; Rheodyne 7725i manual injector; Diamonsil C18 reverse-phase column, 250  4.6 mm, 5 μm). Methanol and water solutions (1:1 for pyridine detection and 4:1 for quinoline detection) were used as the mobile phase in isocratic mode at a flow rate of 1.0 mL/min. Pyridine was detected at 254 nm, and quinoline was detected at 275 nm.5 The TOC was analyzed using a TOC analyzer (Shimadzu TOC-vCPH, Japan). Concentrations of NH3-N, NO2--N, and NO3--N were analyzed using standard methods, i.e. salicylatehypochlorous acid, N-1-naphthyl-ethylenediamine, and UVspectrophotometric determination, respectively.11 In addition, pH and DO were measured using a pH meter (Thermo Orion 868, USA) and a DO meter (Thermo 3-star benchtop), respectively. 2.6. Plate Count and Biofilm Observation. Zeolite samples (1.0 g, wet weight) from the top and bottom layers of the three columns were collected after the wastewater treatment. To extract biomass from the zeolite, the biofilm-attached zeolites were dispersed in 10 mL of NaCl solution (8.5 g/L) and vortexed for 60 s.12 Each suspension was serially diluted with sterile ddH2O and spread onto beef-extract peptone agar plates in duplicate. The number of bacterial colonies was counted after the plates were incubated at 30 °C for 3 days. Zeolite samples from the bottom layers were sent to the Institute of Microbiology, Chinese Academy of Sciences, for 1941

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology biofilm observation by a scanning electron microscope (SEM, FEI QUANTA 200, Holland). 2.7. DNA and RNA Extraction from Biofilm. Besides the activated sludge inoculated, triplicate zeolite samples (5.0 g, wet weight) from the top and bottom layers of the three columns were taken out after the treatment. Each sample was transferred to a 15 mL tube. Total RNA and total DNA were extracted simultaneously with an RNA PowerSoil total RNA isolation kit and an RNA PowerSoil DNA elution accessory kit (Mobio, USA). The extracted RNA was digested with RNase-Free DNase I (Tiangen, China) for DNA elimination and extracted with phenol:chloroform:isoamyl alcohol according to the recommendations of the Mobio kit protocol. The treatment efficiency was verified by PCR, which showed no amplification with the universal primers 27F and 1492R.8 The DNaseI-treated RNA was immediately reverse transcribed with random 6 mers by using a Primescript RT-PCR kit (Takara, China). The total DNA and synthesized cDNA were stored at -20 °C in a freezer for further application. 2.8. Length Heterogeneity PCR (LH-PCR) and Reverse Transcription LH-PCR (RT-LH-PCR). PCR amplification of the 16S rRNA gene from the total DNA and cDNA was performed with the universal primers 27F and 338R as described by Ritchie et al.13 The 27F primer was labeled with 6-carboxyfluorescein at the 50 end. Takara premix Taq Hot Start polymerase was used to increase the specificity of the PCR reaction. The PCR products were purified by a Qiaquick PCR purification kit (QIAGEN, Germany). The purified PCR products were analyzed using an ABI PRISM 3700 DNA analyzer (Applied Biosystems, Foster City, CA) in Genescan mode at SinoGenoMax Co., Ltd. (Beijing, China) with the GS500 Liz internal size standards (Applied Biosystems). LH-PCR and RT-LH-PCR electropherograms were inspected with the Genemapper 3.7 software. The minimum noise threshold was set at 50 fluorescent units of peak height after normalization of the sum of total fluorescence in each profile. Three different parameters of bacterial diversity including richness (total number of LH-PCR or RT-LH-PCR peaks), diversity P (Shannon-Weaver index, equal to - pi ln pi, where Pi is the relative peak height ratio detected in a sample), and evenness (equal to diversity/ln(richness)) were calculated.14 Also, the fingerprint data were processed to create binary data (presence, 1; absence, 0) matrices and analyzed with the additive main effects and multiplicative interaction (AMMI) model using T-REX online software.15 In addition, hierarchical cluster analysis of LH-PCR and RT-LH-PCR profiles was performed with Ward’s method and binary squared Euclidean distance measurements using SPSS 15.0 software. 2.9. Cloning, Sequencing, and Phylogenetic Analysis. PCR amplification of the 16S rRNA gene from the total DNA and cDNA of the biofilm bacteria was performed with 27F and 1492R primer pairs. The PCR products were separated by agarose (0.8%) gel electrophoresis, and the target DNA fragments were cut with a blade and purified using a Qiaquick gel extraction kit. The purified DNA fragments were cloned into pGEM-T Easy vectors (Promega, USA) according to the manufacturer’s protocol. The recombinant plasmids were transformed into competent E. coli JM109. Positive clones were randomly selected and sequenced with the vector-specific primers M13F and M13R using an ABI 3730xl DNA analyzer (Applied Biosystems, USA). Chimeric sequences were detected using the Bellerophon16 and Mallard17 programs and were excluded from subsequent

ARTICLE

analysis. Rarefaction, community similarity, and diversity statistics including library coverage18 and Shannon-Weaver index were calculated using Mothur software.19 LIBSHUFF was used to compare two libraries of 16S rDNA and 16S rcDNA gene sequences from column 1 and to determine if they were significantly different.20 Operational taxonomic units (OTU) were defined as groups in which the sequence similarities were greater than 97%. In order to acquire the taxonomic information and the closest matches of bacterial sequences in the present databases, nearly complete 16S rDNA and 16S rcDNA gene sequences were obtained for the representative OTUs and were subjected to the Ribosomal Database Project (RDP)21 or the National Center for Biotechnology Information (NCBI) databases when there was no close relative in the RDP database. Selected matched sequences and the detected OTU sequences were assembled using Bioedit software. Phylogenetic trees were constructed using the neighbor joining method with MEGA 4.0 software.22 All representative 16S rDNA and 16S rcDNA gene sequences (OTUs) originated from this study were deposited in the NCBI database under accession numbers HQ681945 to HQ682064 and HQ011386 to HQ01400, respectively.

3. RESULTS 3.1. Coking Wastewater Treatment by Z-BAFs. Three columns of Z-BAFs were used to investigate the efficiency and stability of bioaugmentation and adsorption during long-term and continuous operation. Figure 2 shows the long-term performance of the Z-BAFs. After adjustment in the first phase for 10 days, columns 1 and 2 reached a stable and high removal efficiency for pyridine. However, column 3 kept fluctuating at a lower efficiency for pyridine removal. The removal efficiency decreased rapidly when the influent concentration was increased in the second phase, and it became even worse when the HRT was further decreased in the third phase. The final percentage of pyridine removed in column 1 (90%) and column 2 (99%) was much higher than in column 3 (23%). For the removal of quinoline, columns 1 and 2 did not need time for adjustment; they reached a stable and high removal efficiency from the first day and kept removing the quinoline throughout the three operational phases. Column 3 also reached a stable and high removal efficiency as column 1 after one day of adjustment and kept the same efficiency as in the first phase. When the influent concentration was increased in the second phase, the removal efficiency of column 3 dropped at the first day (81%) but soon recovered; when the HRT was further decreased in the third phase, it dropped remarkably (45%) and did not recover to the previous efficiency level after several days of adjustment. Quinoline removal in columns 1 and 2 was 99%100%, except for one day in column 1 (the first day of the third phase) during the long-term operation. The removal in column 3 decreased significantly to an average of 50% in the third phase. Despite the negative impact of pyridine and quinoline, the removal of TOC in all columns was effective throughout the long-term operation except for a few days in the second phase. The final efficiencies were 97% for column 1, 98% for column 2, and 90% for column 3. Because of the ion-exchange adsorption of zeolite, the NH3-N removal was improved. The removal efficiencies in columns 1 and 3 were higher in the beginning but decreased during the operation because the adsorption capacity of the natural zeolites 1942

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology

ARTICLE

Figure 2. Comparison of treatment efficiencies for (A) pyridine, (B) quinoline, (C) TOC, and (D) NH3-N using three columns. Column 1, bioaugmented Z-BAF (natural zeolite); Column 2, bioaugmented Z-BAF (modified zeolite); Column 3, nonbioaugmented Z-BAF (natural zeolite).

reduced gradually. At the end of the operation, the NH3-N removal in columns 1 and 3 were 60% and 70%, respectively. However, the removal in column 2 was sharply lower, and even the effluent NH3-N concentration became higher than the influent concentration after 60 days. The influent pH values ranged from 8.75 to 8.95 during the operation, and the effluent pH from columns 1 (8.18 ( 0.13, the

mean value ( stand deviation) and 3 (8.20 ( 0.11) was always lower than that from column 2 (8.28 ( 0.05). In addition, the NO2--N and NO3--N removal in columns 1 and 2 were both considerably higher than in column 3 (Supporting Information, Figure S1). 3.2. Biomass. Culture-dependent (plate count; Supporting Information, Table S1) and culture-independent techniques 1943

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology

ARTICLE

Table 1. Diversity Indices Calculated from LH-PCR and RT-LH-PCR Profiles in the Three Z-BAFsb LH-PCR a

sample

richness

diversity

RT-LH-PCR richness

diversity

1T

21 ( 2

2.49 ( 0.01

19 ( 0

2.60 ( 0.02

1B

21 ( 1

2.45 ( 0.06

20 ( 1

2.38 ( 0.09

2T 2B

21 ( 1 20 ( 1

2.57 ( 0.02 2.55 ( 0.03

24 ( 2 23 ( 1

2.59 ( 0.09 2.43 ( 0.10

3T

12 ( 3

2.05 ( 0.07

11 ( 3

1.93 ( 0.07

3B

14 ( 2

2.16 ( 0.08

13 ( 2

1.95 ( 0.10

AS

20 ( 1

2.32 ( 0.02

17 ( 1

2.39 ( 0.02

a

The numbers represent the column number and the capital letters represent the sampling site in each column. T: top, B: bottom. AS: the activated sludge inoculated. b The numbers represent the means of 3 traces, plus/minus standard deviation.

(amount of extracted DNA and RNA) were used to compare the biomass among the three columns. The results displayed that the biomass was in the following order: column 2 > column 1 > column 3, and the biomass at the bottom was always higher than that at the top of each column. SEM observation also supported this argument (Supporting Information, Figure S2). Moreover, the biofilm in column 2 was much thicker than those in columns 1 and 3, and the configuration of the biofilm in column 3 was very different from that in columns 1 and 2. 3.3. LH-PCR and RT-LH-PCR. The effects of pyridine, quinoline, highly efficient bacteria (bioaugmentation), and zeolite (ion-exchange) on the bacterial community and active bacterial community during the coking wastewater treatment were monitored by LH-PCR and RT-LH-PCR techniques, respectively. Statistical analysis of rDNA- and rRNA-based fingerprint data for diversity indices (Table 1) demonstrated that the richness and diversity (Shannon-Weaver index) of the bacterial community and active bacterial community decreased significantly in column 3 in comparison with the activated sludge inoculated because of the pyridine and quinoline loading shocks. Bioaugmentation recovered the richness and diversity of community. In addition, the evenness of the bacterial community (0.78-0.85) and active bacterial community (0.75-0.88) had no obvious difference among all the samples. AMMI analysis of the LH-PCR and RTLH-PCR profiles (Figure 3) revealed that in the bioaugmented Z-BAFs, pyridine and quinoline were two primary drivers of the variations of the bacterial community and active bacterial community structures. These two factors were captured by the first and second interaction principal components (IPCA1 and IPCA2). Different materials (natural or modified zeolites) had a lower impact on the bacterial community. However, they had a comparatively higher impact on the active bacterial community as reflected by the IPCA2. Some differences between the top (with lower concentrations of pyridine and quinoline) and bottom (with higher concentrations of pyridine and quinoline) layers of each column could be observed, but they were not as significant as the factors mentioned above on the variation of bacterial community structure. In addition, some small differences among triplicate independent samples might have been caused by the heterogeneity of the biofilm and activated sludge and the bias of DNA and RNA extraction and PCR amplification. The clustering analysis results (Supporting Information, Figure S3) also supported the AMMI findings.

Figure 3. AMMI analysis of bacterial (A) LH-PCR and (B) RT-LHPCR data sets from the three Z-BAFs. In one set of LH-PCR and RTLH-PCR profiles, the data of independent triplicate samples from the same column-site were treated individually. For symbols, the number represents the column number, and the capital letter represents the sampling site in the column. T: top, B: bottom. AS: the activated sludge inoculated.

The peaks corresponding to the inoculated degrading strains were not present in the LH-PCR and RT-LH-PCR profiles of both bioaugmented columns, which indicated that the inoculated strains were not dominant in their own bacterial communities. 3.4. Phylogenetic Analysis. The four 16S rDNA clone libraries from the activated sludge inoculated and the biofilm of three Z-BAFs were constructed to evaluate the composition of bacterial community and investigate the community shift with the pyridine and quinoline loading shocks and bioaugmentation. In addition, a 16S rcDNA clone library was examined to investigate the composition of active bacterial community in the most effective column (column 1), according to the treatment data. A total of 569 cloned 16S rRNA gene sequences were obtained from 16S rDNA and 16S rcDNA clone libraries. All sequences were checked for chimeras, leaving 377 sequences (328 16S rDNA including 102 for column 1, 70 for column 2, 62 for column 3, 94 for the activated sludge inoculated; and 49 16S rcDNA for column 1) for further analysis. Estimation of species coverage and diversity were calculated for four 16S rDNA clone libraries (Supporting Information, Table S2). Good’s coverage ranged from 58.6% to 76.5%. 1944

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology A significant number of the OTUs comprised of a single clone, which indicated high levels of bacterial diversity. This was also reflected by the unsaturated rarefaction curves (Supporting Information, Figure S4). Shannon-Weaver indices revealed that the 16S rDNA sequences diversity in column 3 (2.38) was lower than in columns 1 and 2 (3.04 and 3.15, respectively) and in the activated sludge inoculated (3.08). In order to analyze the relationship of bacterial communities among three columns and the activated sludge inoculated, we combined the 16S rDNA sequences of four clone libraries. This resulted in 120 unique 16S rDNA phylotypes (OTUs) that were identified from the sequences. These results are presented in a figure of taxonomic classification (Supporting Information, Figure S5) and a phylogenetic tree (Figure 4A). The bacterial community in the activated sludge inoculated was dominated by Bacterodietes (51.1%) and Proteobacteria (44.7%). The abundance of phylum Bacterodietes was decreased significantly in column 3 (1.6%), even in the bioaugmented columns 1 (7.8%) and 2 (10.0%) after the pyridine and quinoline loading shocks. In contrast, the abundance of phylum Proteobacteria was increased in columns 1 (77.5%), 2 (80.0%), and 3 (96.8%). Among the different classes of Proteobacteria, the Betaproteobacteria was always dominant in the four clone libraries (37.2-67.7%). It was remarkable that the phylum Planctomycetes was only present in columns 1 (13.8%) and 2 (7.1%) because of the bioaugmentation (Supporting Information, Figure S5). Among 120 OTUs, the OTU AS-95 representing 21.3% 16S rDNA clones in the activated sludge inoculated had 95% sequence identity to uncultured Flexibacter sp. (AB076886, belonging to Bacterodietes). The OTU ZBAF239, which was dominant in columns 1 (27.5%), 2 (18.6%), and 3 (43.5%), had 100% similarity with Thiobacillus thioparus (HM535225, belonging to Betaproteobacteria) (Figure 4A). Furthermore, clustering analysis for the sequences in the four 16S rDNA clone libraries was shown in Supporting Information, Figure S6. It is clearly evident that the bacterial community of the activated sludge inoculated was distinctively different from those in the three columns, especially in columns 1 and 2, which formed a cluster. Good’s coverage for 16S rcDNA library was 83.7%, which indicated that this library represented the majority of bacterial cDNA sequences present in column 1 (Supporting Information, Table S2). The Shannon-Weaver index (2.11) showed that the diversity of 16S rcDNA sequences was comparatively lower than 16S rDNA sequences. In addition, Pairwise comparison of 16S rDNA and 16S rcDNA libraries in column 1 using LIBSHUFF revealed that the sequences of the two libraries were distinctively different (p = 0.005). Figure 4B shows that the active bacterial community in column 1 was comprised of three phyla: Proteobacteria (89.8%), Bacterodietes (6.1%), and Planctomycetes (4.1%). Among 15 OTUs, the dominant OTU representing 38.8% 16S rcDNA clones had 98% sequence identity to Acidovorax sp. (CP000539). After screening all 16S rDNA and 16S rcDNA sequences, we did not find any sequence that was identical (>97% similarity) to the inoculated strains.

4. DISCUSSION Bioaugmentation provides certain advantages for the removal of persistent organic pollutants from the environment in cases where pollutant toxicity or a lack of specific degrading microorganisms (both quantity and quality) are important.23 The survival and degradation efficiency of the introduced inocula as

ARTICLE

Figure 4. Neighbor-joining trees showing the phylogenetic affiliation of (A) selected 16S rDNA OTU sequences (only OTUs comprising g2 clones are shown) from three columns and the activated sludge inoculated and (B) all 16S rcDNA OTU sequences from column 1. The numbers on the branch nodes represent percentage of bootstrap resamplings based on 1000 replicates (only g50% are shown). The scale bar indicates the number of nucleotide substitutions per site. Aphanocapsa feldmani was used as the outgroup in the trees. The relative abundance (percentage) of each OTU (comprising g2 clones) or group (only for 16S rcDNA library, see Supporting Information Figure S5 for 16S rDNA) is shown in parentheses. Additional symbols for abundance in part (A) are (9) column 1, (2) column 2, (b) column 3, and (0) activated sludge inoculated. Two reference 16S rDNA sequences with red color are obtained from the inoculated strains. 1945

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology well as their associated community effects upon the indigenous microorganisms still remain to be controversial.24 Combining environmental engineering technology and modern molecular techniques, we attempted to achieve higher treatment efficiencies for hazardous pollutants and reveal the mechanism of bioaugmentation, since most previous studies have vague explanations regarding the relationship between treatment efficiency and microbial ecology. In this study, natural and modified zeolites were used as the ammonium exchangers as well as the BAF fillings, and highly efficient degrading bacteria were delivered as the inocula to develop biofilm on the surface of the zeolites. Then the biozeolite, i.e. the zeolite attached by the biofilm, was used to treat coking wastewater containing high concentrations of pyridine and quinoline. The bioaugmented degradation for organic pollutants (pyridine, quinoline, and other organic compounds) and ion-exchange adsorption for ammonium (those in the raw wastewater and transformed from N-heterocyclic compounds) were achieved simultaneously in the Z-BAFs. Comparison of the treatment efficiencies among the three columns showed that column 1 (bioaugmented BAF containing natural zeolite) had the best treatment effect in the long term operation when taking into account of the removal efficiencies of all pollutants. This indicates that the application of nature zeolite as a biofilm carrier could provide a sufficient amount of biomass for organic pollutants removal. Also, the attached biofilm did not significantly affect the adsorption capacity of the zeolite for NH3-N removal. The modified zeolite has more mesopores and macropores on its surface; therefore, the inocula and indigenous bacteria could easily attach and form a thicker biofilm on the surface of the modified zeolite. The column 2 (bioaugmented BAF containing modified zeolite) exhibited similar removal efficiencies of pyridine, quinoline, and TOC. However, the NH3-N removal efficiency decreased sharply because of the lower ammonium adsorption capacity of the modified zeolite. The decreased adsorption of the modified zeolite was caused by the other materials that were added during the processing of this zeolite.6 Nonbioaugmented column 3 had lower removal efficiencies of pyridine and quinoline. The remaining pyridine and quinoline in the column might eliminate some strains in the biofilm. As a result, the biomass in column 3 decreased, which further caused an adverse effect on TOC removal. In recent years, rRNA based methods are more prevalent than rDNA based methods in the detection of active bacterial communities, since DNA may persist in dead cells and spores, or extra-cellularly.25-27 Both rRNA and rDNA methods were used in this study to assess both the metabolically active bacterial community and the bacterial community, respectively. High loadings of pyridine and quinoline led to a decrease of bacterial richness and diversity because of selective pressure that placed on the bacterial community. Such a conclusion was also found in one of our previous studies that as well focused on the bioaugmented treatment for coking wastewater containing pyridine and quinoline. However, that study was carried out in a sequencing batch reactor without any zeolite.28 Likewise, similar results were obtained in previous studies on other pollutants.29,30 The LHPCR, RT-LH-PCR, and clone library analysis also showed that bioaugmentation recovered bacterial richness and diversity from the toxicity of pyridine and quinoline, indicating that bioaugmentation alleviated the impact of pyridine and quinoline and prevented the extermination of some bacteria. This was also reflected from the treatment efficiency of TOC as other pollutants

ARTICLE

in the coking wastewater were removed more efficiently in the two bioaugmented columns. A similar finding was reported in a previous study, which showed that bioaugmentation with a 3-chloroaniline-degrading inoculants protected the bacterial community and recovered the reactor function from 3-chloroaniline shocks.31 Furthermore, the results of this study showed that bioaugmentation did not recover the bacterial community structures from the pyridine and quinoline loading shocks but increased the variation of community with the activated sludge inoculated (Figure 3, Supporting Information Figures S3 and S6). It was also determined by previous studies that bioaugmentation could promote bacterial community shift and make the community distinct from the original system,31,32 because of changes in interior (such as invasion of inoculated strains) and exterior environment (such as change of pollutant concentrations). Our phylogenetic analysis of four 16S rDNA clone libraries revealed that pyridine and quinoline were more harmful to Bacterodietes among the indigenous bacteria, and bioaugmentation promoted the growth of Planctomycetes in the biofilm. Both phyla mentioned above and Proteobacteria that was dominant in all samples are frequently retrieved from the activated sludge or biofilm in wastewater treatment units.33 A number of OTUs in our four clone libraries were never obtained in culture nor described previously (>97% identity). Therefore, coking wastewater could be regarded as an extreme environment in which the microbial community remains largely unexplored. The direct comparison between 16S rcDNA and 16S rDNA sequences in column 1 revealed that there might be many species dominant in the bioaugmented columns, but only a few of them were active and contributed to pollutants degradation. Another interesting finding from the rcDNA detection was that the predominant phylotype had great similarity to Acidovorax, which was recognized as an important phenol-degrading genus.34 Phenols were always the major organic compounds in the coking wastewater, and phenol and methyl-phenol were detected by GC-MS as the top two organic compounds in this study (Supporting Information, Figure S7). Successful application of bioaugmentation technology for wastewater treatment depends on the appropriate microbial strains, and their subsequent survival and activity once released into the treatment units.7,35 On the other hand, the introduced microbial strains should not outcompete other indigenous degrading microorganisms due to the competition for survival space and nutrients. Our study showed that TOC removal remained at a high level (>98%); therefore, the introduced highly efficient bacteria could collaborate with other indigenous microorganisms to achieve good treatment results. LH-PCR and RTLH-PCR data and the clone library results demonstrated that the introduced pyridine- and quinoline-degrading bacteria did not become dominant in the biofilm even after the pyridine and quinoline loading shocks. This indicated that the indigenous bacteria were mainly responsible for the treatment of coking wastewater containing high concentrations of pyridine and quinoline. The similar results were present in the bioaugmentation studies on perchlorate36 and TCE,37 which revealed the inoculated strains were undetectable in the end and the removal of target pollutants was still in effect. Above all, the bioaugmented Z-BAFs facilitated the stable microbial community with elevated capacities for pyridine and quinoline degradation and ammonium adsorption. On the basis of our study, it can be concluded that bioaugmented Z-BAF can be considered as an alternative technology for the treatment of 1946

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology wastewater containing high concentrations of N-heterocyclic aromatic compounds and ammonium.

’ ASSOCIATED CONTENT

bS

Supporting Information. Tables S1 and S2 and Figures S1-S7. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: þ86 10 62751923. Fax: þ86 10 62751923. E-mail: [email protected].

’ ACKNOWLEDGMENT This study was supported by an “863” Follow-up Exploration Project (2009AA06Z309) granted by the Chinese Ministry of Science and Technology and a research project granted by the Chinese Postdoctoral Science Foundation (20100470109). We sincerely thank Mr. Anping L€u who kindly provided the natural zeolite and helped to prepare the modified zeolite at Jinyun, Zhejiang Province, China. We also thank Dr. Zbigniew Cichacz in the Biodesign Institute at Arizona State University for carefully checking the paper. ’ REFERENCES (1) Fetzner, S. Bacterial degradation of pyridine, indole, quinoline, and their derivatives under different redox conditions. Appl. Microbiol. Biotechnol. 1998, 49, 237–250. (2) Padoley, K. V.; Mudliar, S. N.; Pandey, R. A. Heterocyclic nitrogenous pollutants in the environment and their treatment options An overview. Bioresour. Technol. 2008, 99, 4029–4043. (3) Kaiser, J. P.; Feng, Y. C.; Bollag, J. M. Microbial metabolism of pyridine, quinoline, acridine, and their derivatives under aerobic and anaerobic conditions. Microbiol. Rev. 1996, 60, 483–498. (4) Padoley, K. V.; Rajvaidya, A. S.; Subbarao, T. V.; Pandey, R. A. Biodegradation of pyridine in a completely mixed activated sludge process. Bioresour. Technol. 2006, 97, 1225–1236. (5) Bai, Y. H.; Sun, Q. H.; Zhao, C.; Wen, D. H.; Tang, X. Y. Simultaneous biodegradation of pyridine and quinoline by two mixed bacterial strains. Appl. Microbiol. Biotechnol. 2009, 82, 963–973. (6) Bai, Y. H.; Sun, Q. H.; Xing, R.; Wen, D. H.; Tang, X. Y. Removal of pyridine and quinoline by bio-zeolite composed of mixed degrading bacteria and modified zeolite. J. Hazard. Mater. 2010, 181, 916–922. (7) Thompson, I. P.; van der Gast, C. J.; Ciric, L.; Singer, A. C. Bioaugmentation for bioremediation: the challenge of strain selection. Environ. Microbiol. 2005, 7, 909–915. (8) Bai, Y. H.; Sun, Q. H.; Zhao, C.; Wen, D. H.; Tang, X. Y. Microbial degradation and metabolic pathway of pyridine by a Paracoccus sp. strain BW001. Biodegradation 2008, 19, 915–926. (9) Sun, Q. H.; Bai, Y. H.; Zhao, C.; Xiao, Y. N.; Wen, D. H.; Tang, X. Y. Aerobic biodegradation characteristics and metabolic products of quinoline by a Pseudomonas strain. Bioresour. Technol. 2009, 100, 5030– 5036. (10) Sambrook, J.; Russell, D. Molecular cloning: a laboratory manual, 3rd ed.; Cold Spring Harbor Laboratory Press: New York, 2001. (11) State Environmental Protection Administration of China. Monitoring and analysis method of water and wastewater, 3rd ed.; China Environmental Science Press: Beijing, 1989 (in Chinese). (12) Emanuelsson, M. A. E.; Henriques, I. S.; Jorge, R. M. F.; Castro, P. M. L. Biodegradation of 2-fluorobenzoate in upflow fixed bed bioreactors operated with different growth support materials. J. Chem. Technol. Biotechnol. 2006, 81, 1577–1585.

ARTICLE

(13) Ritchie, N. J.; Schutter, M. E.; Dick, R. P.; Myrold, D. D. Use of length heterogeneity PCR and fatty acid methyl ester profiles to characterize microbial communities in soil. Appl. Environ. Microbiol. 2000, 66, 1668–1675. (14) Shannon, C. E.; Weaver, W. The mathematical theory of communication; University of Illinois Press: Urbana, 1963. (15) Culman, S. W.; Bukowski, R.; Gauch, H. G.; Cadillo-Quiroz, H.; Buckley, D. H. T-REX: software for the processing and analysis of T-RFLP data. BMC Bioinf. 2009, 10, 171. (16) Huber, T.; Faulkner, G.; Hugenholtz, P. Bellerophon: a program to detect chimeric sequences in multiple sequence alignments. Bioinformatics 2004, 20, 2317–2319. (17) Ashelford, K. E.; Chuzhanova, N. A.; Fry, J. C.; Jones, A. J.; Weightman, A. J. New screening software shows that most recent large 16S rRNA gene clone libraries contain chimeras. Appl. Environ. Microbiol. 2006, 72, 5734–5741. (18) Good, I. J. The population frequencies of species and the estimation of population parameters. Biometrika 1953, 40, 237–264. (19) Schloss, P. D.; Westcott, S. L.; Ryabin, T.; Hall, J. R.; Hartmann, M.; Hollister, E. B.; Lesniewski, R. A.; Oakley, B. B.; Parks, D. H.; Robinson, C. J.; et al. Introducing mothur: open-source, platformindependent, community-supported software for describing and comparing microbial communities. Appl. Environ. Microbiol. 2009, 75, 7537– 7541. (20) Singleton, D. R.; Furlong, M. A.; Rathbun, S. L.; Whitman, W. B. Quantitative comparisons of 16S rRNA gene sequence libraries from environmental samples. Appl. Environ. Microbiol. 2001, 67, 4374– 4376. (21) Cole, J. R.; Wang, Q.; Cardenas, E.; Fish, J.; Chai, B.; Farris, R. J.; Kulam-Syed-Mohideen, A. S.; McGarrell, D. M.; Marsh, T.; Garrity, G. M.; et al. The Ribosomal Database Project: improved alignments and new tools for rRNA analysis. Nucleic Acids Res. 2009, 37, D141–D145. (22) Tamura, K.; Dudley, J.; Nei, M.; Kumar, S. MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 2007, 24, 1596–1599. (23) Vogel, T. M. Bioaugmentation as a soil bioremediation approach. Curr. Opin. Biotech. 1996, 7, 311–316. (24) Ciric, L.; Griffiths, R. I.; Philp, J. C.; Whiteley, A. S. Field scale molecular analysis for the monitoring of bacterial community structures during on-site diesel bioremediation. Bioresour. Technol. 2010, 101, 5235–5241. (25) Kowalchuk, G. A.; de Bruijn, F. J.; Head, I. M.; Akkermans, A. D.; van Elasa, J. D. Molecular microbial ecology manual. 2nd ed.; Springer: Dordrecht, The Netherlands, 2008. (26) Roy, C. S.; Talbot, G.; Topp, E.; Beaulieu, C.; Palin, M. F.; Masse, D. I. Bacterial community dynamics in an anaerobic plug-flow type bioreactor treating swine manure. Water Res. 2009, 43, 21–32. (27) Eichler, S.; Christen, R.; H€ oltje, C.; Westphal, P.; B€ otel, J.; Brettar, I.; Mehling, A.; H€ofle, M. G. Composition and dynamics of bacterial communities of a drinking water supply system as assessed by RNA- and DNA-based 16S rRNA gene fingerprinting. Appl. Environ. Microbiol. 2006, 72, 1858–1872. (28) Bai, Y. H.; Sun, Q. H.; Zhao, C.; Wen, D. H.; Tang, X. Y. Bioaugmentation treatment for coking wastewater containing pyridine and quinoline in a sequencing batch reactor. Appl. Microbiol. Biotechnol. 2010, 87, 1943–1951. (29) Beaulieu, M.; Becaert, V.; Deschenes, L.; Villemur, R. Evolution of bacterial diversity during enrichment of PCP-degrading activated soils. Microb. Ecol. 2000, 40, 345–355. (30) Watanabe, K.; Teramoto, M.; Harayama, S. An outbreak of nonflocculating catabolic populations caused the breakdown of a phenol-digesting activated-sludge process. Appl. Environ. Microbiol. 1999, 65, 2813–2819. (31) Boon, N.; Top, E. M.; Verstraete, W.; Siciliano, S. D. Bioaugmentation as a tool to protect the structure and function of an activatedsludge microbial community against a 3-chloroaniline shock load. Appl. Environ. Microbiol. 2003, 69, 1511–1520. 1947

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948

Environmental Science & Technology

ARTICLE

(32) Morgante, V.; Lopez-Lopez, A.; Flores, C.; Gonzalez, M.; Gonzalez, B.; Vasquez, M.; Rossello-Mora, R.; Seeger, M. Bioaugmentation with Pseudomonas sp. strain MHP41 promotes simazine attenuation and bacterial community changes in agricultural soils. FEMS Microbiol. Ecol. 2010, 71, 114–126. (33) Wagner, M.; Loy, A. Bacterial community composition and function in sewage treatment systems. Curr. Opin. Biotechnol. 2002, 13, 218–227. (34) Manefield, M.; Griffiths, R. I.; Leigh, M. B.; Fisher, R.; Whiteley, A. S. Functional and compositional comparison of two activated sludge communities remediating coking effluent. Environ. Microbiol. 2005, 7, 715–722. (35) Boon, N.; Goris, J.; De Vos, P.; Verstraete, W.; Top, E. M. Bioaugmentation of activated sludge by an indigenous 3-chloroanilinedegrading Comamonas testosteroni strain, I2gfp. Appl. Environ. Microbiol. 2000, 66, 2906–2913. (36) Zhang, H.; Logan, B. E.; Regan, J. M.; Achenbach, L. A.; Bruns, M. A. Molecular assessment of inoculated and indigenous bacteria in biofilms from a pilot-scale perchlorate-reducing bioreactor. Microb. Ecol. 2005, 49, 388–398. (37) Munakata-Marr, J.; Matheson, V. G.; Forney, L. J.; Tiedje, J. M.; McCarty, P. L. Long-term biodegradation of trichloroethylene influenced by bioaugmentation and dissolved oxygen in aquifer microcosms. Environ. Sci. Technol. 1997, 31, 786–791.

1948

dx.doi.org/10.1021/es103150v |Environ. Sci. Technol. 2011, 45, 1940–1948