Membrane Biofouling in Pilot-Scale Membrane Bioreactors (MBRs

Dec 7, 2006 - Membrane Biofouling in Pilot-Scale Membrane Bioreactors (MBRs) Treating Municipal Wastewater: Impact of Biofilm Formation. Yuki Miura ...
2 downloads 8 Views 540KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 632-638

Membrane Biofouling in Pilot-Scale Membrane Bioreactors (MBRs) Treating Municipal Wastewater: Impact of Biofilm Formation YUKI MIURA, YOSHIMASA WATANABE, AND SATOSHI OKABE* Department of Urban and Environmental Engineering, Graduate School of Engineering, Hokkaido University, North-13, West-8, Kita-ku, Sapporo 060-8628, Japan

For more efficient control and prediction of membrane biofouling in membrane bioreactors (MBRs), a fundamental understanding of mechanisms of membrane biofouling is essential. In this study, we operated full-scale submerged MBRs using real municipal wastewater delivered from the primary sedimentation basin of a municipal wastewater treatment facility over 3 months, and the adhesion and formation of biofilms on 0.4-µm pore size polyethylene hollowfiber microfiltration (MF) membrane surfaces, separated from simple deposition of sludge cake, were monitored using scanning electron microscopy (SEM). In addition, the compositions of planktonic and biofilm microbial communities in the MBR were analyzed using culture independent molecular-based methods (i.e., fluorescent in situ hybridization (FISH) and 16S rRNA gene sequence analysis). The SEM and LIVE/DEAD staining analyses clearly showed that the biofilms gradually developed on the membrane surfaces with time, which had a strong positive correlation with the increase in trans-membrane pressure (TMP). This indicated that the biofilm formation induced the membrane fouling. The FISH results revealed that the microbial communities on membrane surfaces were quite different from those in the planktonic biomass in the mixed liquor. Moreover, FISH and 16S rRNA gene sequence analyses revealed that a specific phylogenetic group of bacteria, the Betaproteobacteria, probably played a major role in development of the mature biofilms, which led to the severe irreversible membrane biofouling.

Introduction Membrane separation technology is increasingly becoming an important innovation in biological wastewater treatment. Despite its many advantages over conventional treatment (i.e., activated sludge systems) such as smaller footprint, reduction of sludge production, and better-treated water quality, membrane fouling, particularly biofouling, in submerged membrane bioreactors (MBRs) restricts their widespread application. Membrane biofouling results in reduced performance, severe flux decline, high energy consumption, and frequent membrane cleaning or replacement, which directly lead to increases in maintenance and operating costs. Biofouling in MBRs treating real wastewater is the result of * Corresponding author phone: +81-11-706-6266; fax: +81-11707-6266; e-mail: [email protected]. 632

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

interactions of biomass (including bacterial cells, cell debris, extracellular polymeric substances (EPS), and so on) with the membranes; that is, formation of biofilms or accumulation of microbial products on the membrane surfaces and/ or in the membrane pores. This is separate from simple deposition of sludge cake on the membrane, which can be readily removed by physical washing (defined as reversible fouling). Because of the nature of wastewater and mixed liquor suspended solids (MLSS), internal fouling caused by the adsorption of dissolved organic and inorganic matter into the membrane pores (defined as irreversible fouling) probably occurs simultaneously with biofouling in MBRs (13). Therefore, the control of complex membrane biofouling is one of particular concern in modern MBR plant designs and operations. The attachment and development of biofilms on the membrane surfaces is usually controlled by physical washing, such as air scrubbing, crossflow filtration, and backwashing. However, the precise relationship between the strength of adhesion and the vigor of physical washing (i.e., hydrodynamic conditions) is not completely understood. Consequently, these control strategies are not always successful. The development of more efficient strategies to control the biofouling requires a more thorough understanding of adhesion and formation of biofilms on membrane surfaces. The membrane biofouling in the MBRs treating real wastewater is caused by complex physicochemical and biological processes, which are highly dependent on feedwater compositions, biomass characteristics, membrane characteristics, and operation conditions. Previous studies have focused on various factors affecting the membrane fouling in MBRs, including MLSS concentrations (4, 5), food/ microorganisms (F/M) ratio (2), sludge characteristics (1, 6), and the amount and composition of microbial products (1, 3, 7). However, unified and well-described mechanisms of membrane biofouling that can lead to the development of appropriate strategies to control membrane biofouling are not currently available. This is partly because of the diverse range of operating conditions, membranes, and feedwater matrices (mostly synthetic media were used) that has been employed. In addition, only a few studies have focused on microbial aspects, such as microbial colonization, biofilm formation, and microbial community structures on the membrane surfaces in MBRs (8, 9). The primary objectives of this study were, therefore, to obtain experimental evidence of membrane biofouling in pilot-scale MBRs treating real municipal wastewater and to identify key bacteria responsible for formation of biofilms on membrane surfaces during a long-term operation. To achieve these objectives, culture independent molecularbased methods (i.e., fluorescence in situ hybridization (FISH) and 16S rRNA gene sequence analysis) were used to analyze the microbial community structures of mixed liquor and biofilms firmly attached on membrane surfaces in the MBRs.

Materials and Methods Operation of MBR. Continuous operation of submerged MBRs was carried out at the Soseigawa municipal wastewater treatment facility, Sapporo, Japan. The MBR was operated with the feed wastewater delivered from the primary sedimentation basin of the facility. The MBR was equipped with hollow-fiber microfiltration (MF) membrane modules made of polyethylene that had a total surface area of 3 m2 and a nominal pore size of 0.4 µm (Mitsubishi Rayon, Tokyo, Japan). In the MBR, aeration was continuously carried out at the flow rate of 3500 L/h until day 38, and thereafter was 10.1021/es0615371 CCC: $37.00

 2007 American Chemical Society Published on Web 12/07/2006

TABLE 1. 16S and 23S rRNA-targeted Oligonucleotide Probes Used in this Study probe

target group

probe sequence (5′ f 3′)

target sitea

FAb(%)

ref

EUB338c EUB338-IIc EUB338-IIIc ALF1b BET42a GAM42a SRB385d SRB385Dbd GNSB941e CFX1223e

Eubacteria Planctomycetes Verrucomicrobia Alphaproteobacteria Betaproteobacteria Gammaproteobacteria Deltaproteobacteria Deltaproteobacteria Chloroflexi Chloroflexi

GCTGCCTCCCGTAGGAGT GCAGCCACCCGTAGGTGT GCTGCCACCCGTAGGTGT CGTTCGYTCTGAGCCAG GCCTTCCCACTTCGTTT GCCTTCCCACATCGTTT CGGCGTCGCTGCGTCAGG CGGCGTTGCTGCGTCAGG AAACCACACGCTCCGCT CCATTGTAGCGTGTGTGTMG

16S, 338-355 16S, 338-355 16S, 338-355 16S, 19-35 23S, 1027-1043 23S, 1027-1043 16S, 385-402 16S, 385-402 16S, 941-957 16S, 1223-1242

0-50 0-50 0-50 20 35 35 35 35 35 35

22 23 23 24 24 24 25 26 27 28

a Position in the rRNA of Escherichia coli. b Formamide concentration in the hybridization buffer. c ,d,eApplied in combination. d These probes are used as Deltaproteobacteria because sulfate-reducing bacteria are a major group in the Deltaproteobacteria subdivision in the MBR sludge.

changed to the rate of 5000 L/h, which was the maximum rate of the MBR used in this study. The reactor volume was 180 L. Filtration was carried out in the constant flow rate (0.4 m3/m2/day) mode of operation using suction pumps. The filtration was intermittently carried out (12 min filtration and 3 min pause). The mixed liquor suspended solid (MLSS) concentration in the reactor was maintained at 20 g/L by extraction of excess sludge. The average total organic carbon (TOC), dissolved organic carbon (DOC), total nitrogen (T-N), NH4+-N, NO2--N, NO3--N, total phosphorus (T-P), turbidity, pH, and temperature of the feed wastewater were 54.4 ( 24.1 mg/L, 35.5 ( 14.5 mg/L, 35.9 ( 12.0 mg/L, 21.0 ( 3.5 mg/L, 0.0 ( 0.1 mg/L, 0.3 ( 0.5 mg/L, 4.1 ( 1.0 mg/L, 137.7 ( 45.7 NTU, 8.0 ( 0.3, and 17.8 ( 2.9 °C, respectively. On the other hand, the average TOC, DOC, T-N, NH4+-N, NO2--N, NO3--N, T-P, turbidity, pH, and temperature of the treated water during the operation were 4.3 ( 1.8 mg/L, 4.3 ( 1.8 mg/L, 17.2 ( 7.6 mg/L, 2.5 ( 0.3 mg/L, 0.0 ( 0.0 mg/L, 13.7 ( 9.9 mg/L, 0.5 ( 0.4 mg/L, 0.0 ( 0.0 NTU, 6.6 ( 0.4, and 16.8 ( 3.6 °C, respectively. When membrane fouling became significant, membrane modules were taken out of the reactor and were cleaned physically and/or chemically. Physical membrane cleaning was carried out by spraying pressurized water on the membrane surface. Chemical membrane cleaning was carried out by submerging the membrane module in solutions of sodium hypochloride (500 ppm) overnight and hydrochloric acid (pH 2) for 1 h. Scanning Electron Microscopy. For scanning electron microscopy (SEM) analysis, hollow-fiber membranes with biofilms were taken from the MBR and immersed in 2% (v/v) glutaraldehyde in 0.1 M phosphate buffer for 2 h. The samples were then washed twice for 10 min and then for 1 h in 0.1 M phosphate buffer. The samples were fixed in 1% OsO4 in 0.1 M phosphate buffer for 2 h and then washed again in the same way. The fixed samples were dehydrated in an ethanol series (sequentially in 50, 70, 80, 90, 95, 100, 100, and 100% ethanol for 15 min each) and substituted with isoamyl acetate. The fixed samples were dried with a critical-point drier using liquid CO2 and coated with platinum and palladium for 2 min. The coated samples were examined under a SEM (model S-4000; Hitachi, Japan) at 3.5 kV. LIVE/DEAD Staining. Bacterial viability was determined by using the LIVE/DEAD bacterial viability staining kit, (BacLight, Molecular Probes Inc., Eugene, OR). Two stock solutions of stains (SYTO 9 and propidium iodide) were diluted to a concentration of 3 µL/mL. The hollow-fiber membranes with biofilms were cut into ca. 30-mm-long pieces using a sterile cutter knife. The membrane biofilms were stained with 1 mL of the diluted stain solution in 1.5mL micro tubes (MCT150C, Axygen Inc., CA) at room temperature in the dark for 15 min. The samples were rinsed twice for 1 min with PBS and were mounted on glass slides.

Live SYTO9-stained cells and dead propidium iodide-stained cells on the hollow-fiber membranes were visualized with an LSM 510 confocal scanning laser microscope (CLSM, Carl Zeiss) equipped with an argon laser (488 nm) and two HeNe lasers (543 and 633 nm), respectively. To quantitatively determine biofilm development on the membrane surfaces, the membrane surface area covered with biofilms was measured from CLSM digital projection images using the image analysis software provided by Zeiss. The average biofilm surface coverage was determined from at least 20 representative microscopic images for each sample. FISH. Biofilms attached on hollow-fiber membranes and biomass in mixed liquor were taken from the MBR and fixed in a 4% paraformaldehyde solution. Washing, dehydration, and in situ hybridization were performed according to the procedure described by Amann (10) with the following modifications. For biomass in mixed liquor, the samples spotted on glass slides were air-dried and dehydrated by successive ethanol washes. The 16S and 23S rRNA-targeted oligonucleotide probes used in this study are shown in Table 1. The probes were labeled with fluorescein isothiocyanate (FITC), tetramethylrhodamine 5-isothiocyanate (TRITC), or the sulfoindocyanine dye Cy5 at the 5′ end. The divisionspecific probes were applied simultaneously with probe EUB338 mixed probes, specific for almost all Eubacteria. The hollow-fiber membranes with biofilms were cut into about 10-mm pieces using a sterile cutter knife. In situ hybridizations of the membrane samples were performed in 180 µL of hybridization buffer (0.9 M NaCl, 20 mM Tris-HCl [pH 7.6], 0.01% sodium dodecyl sulfate and formamide at the concentrations shown in Table 1) at 46 °C for 2-3 h in 200µL micro tubes (Continental Lab Products Inc., CA). In the same way, in situ hybridization of the glass slides was performed in 8 µL of hybridization buffer in an equilibrated sealed moisture tube. The final probe concentration was approximately 5 ng/µL. Subsequently, a stringent washing step of the hollow-fiber membranes was performed at 48 °C for 20 min in 1 mL (50 mL for glass slides) of prewarmed washing solution. After hybridization and washing steps, the membrane fibers and the glass slides were air-dried and mounted in antifading solution (Slow Fade Light; Molecular Probe, Eugene, OR). Fluorescent and phase-contrast images were recorded with an LSM 510 confocal scanning laser microscope (Carl Zeiss) equipped with an argon laser (488 nm) and two HeNe lasers (543 and 633 nm). The total biomass area and probe-stained area were measured from CLMS projection images by using image analysis software provided by Zeiss (11). At least 20 representative CLMS projection images of each sample were analyzed. Since fluorescence intensity derived from probestained cells varied slightly for each image, the highest fluorescence intensity of background was first determined (11). This value was used as a threshold low value. All pixels VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

633

FIGURE 1. Changes in TMP during the operation. The aeration rate was changed from 3800 to 5000 L/h on day 38. Physical cleaning (white arrow) was performed on days 20, 23, 30, 38, 44, 55, 82, 93, and 97, respectively. Chemical cleaning (black arrow) was performed on days 55 and 97, respectively. with fluorescence intensity above the threshold value were counted as probe-stained area. 16S rRNA Gene-Cloning and Phylogenetic Analysis. The biofilm samples taken on day 55 before chemical cleaning were homogenized and subjected to DNA extraction. Total DNA was extracted using a Fast DNA Spin kit (BIO 101, USA) as described in the manufacturer’s instructions. The nearly full-length 16S rRNA genes from mixed bacterial DNA were amplified by PCR using the universal primer set for Bacteria, 11f (12) and 1492r (13). One microliter of the PCR-amplified bacterial 16S rRNA fragments was directly ligated into the pCR vector using TOPO XL PCR Cloning Kit (Invitrogen, USA) and transformed into competent cells (high-efficiency Escherichia coli TOP10; Invitrogen, USA) as described in the manufacturer’s instructions. Plasmids were extracted and purified from clones with the Wizard Plus Minipreps DNA purification system (Promega, USA) in accordance with the manufacturer’s instructions. The 16S rRNA gene inserts were sequenced by using an ABI model 310 genetic analyzer with a BigDye terminator v3.1 Cycle Sequencing kit (Applied Biosystems, USA). All sequences were checked for chimeric artifacts by the CHECK_CHIMERA program in Ribosomal Database Project II (14) and were compared with the sequences available in public databases (GenBank and DDBJ) by the BLAST system (15). Sequence data were aligned with the CLUSTAL W package (16). Phylogenetic trees were constructed by the neighbor-joining method (17). Bootstrap resampling analysis for 100 replicates was performed to estimate the confidence of tree topologies.

Results and Discussion Reactor Performance. The submerged MBR was stably operated for more than 100 days. During the operation, the MBR removed nearly 90% of the organic impurities and NH4+-N. In addition, the MBR removed nearly 50% of the total nitrogen. Although the influent had rather low pollutant strength, an adequate hydraulic retention time (average 7 h) and a long solids retention time (approximately 80 days) effectively maintained a sufficient biomass in the MBR (ca. 20 g/L of MLSS concentration). The membrane fouling of the MBR was demonstrated by the increase in transmembrane pressure (TMP) (Figure 1). When the TMP reached a level of around 25 kPa, the membrane modules were physically cleaned to remove the sludge cake deposited on the membrane surfaces. The TMP right after physical cleaning, TMP0, gradually increased with time, indicating occurrence of irreversible membrane fouling (Figure 1) even after several physical cleanings were applied. On days 55 and 97, the membranes were therefore chemically washed with HCl and NaClO solutions to remove the irreversible foulants from the membrane. After the chemical cleaning, 634

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

FIGURE 2. SEM photographs showing the surfaces of clean and fouled membranes after physical washing: (A and B) a new hollowfiber MF membrane; (C) a partially fouled hollow-fiber membrane with some deposits at day 16; (D-H) a completely fouled membrane with mature biofilms just before chemical washing at day 97. Panels G and H show that microorganisms firmly attached on the hollowfiber MF membrane surfaces with anchor-like structures (maybe pili or curli net) even after physical washing. These images indicate that microorganisms grew and formed mature biofilms. the TMP dramatically dropped and the filtration capacity of the membrane was almost fully recovered. During the initial 38 days, the aeration rate was set at 3500 L/h. However, the viscosity of mixed liquor in the MBR increased up to 360 mPa‚s (the average was approximately 210 mPa‚s). As a consequence, membrane fouling indicated by rapid increases in the TMP occurred due to the accumulation of sticky cake layers (a thin layer of brown deposit) on the membrane surfaces, which could not be physically removed by the aeration rate of 3500 L/h. We therefore increased the aeration rate from 3500 L/h to 5000 L/h (the maximum aeration rate) on day 38. SEM Analysis. New hollow-fiber membrane surfaces were porous and free of particles (Figure 2A and B). SEM images of fouled membrane surfaces were taken after physical cleaning on days 16 and 97 (Figure 2C-H). On day 16 when the TMP was relatively low (below 10 kPa), the membrane pores were still visible even though some microorganisms were present on the membrane surface (Figure 2C). These microorganisms were most likely deposition of sludge from mixed liquor and were not actively growing on the membrane surfaces. The SEM analysis demonstrated that the further operation gradually caused pore blocking and attachment

FIGURE 3. CLSM images showing the development of biofilms on PE hollow-fiber MF membrane surfaces with time. The biofilms on (A) day 9, (B) day 38, (C) day 55, (D) day 64, (E) day 85, and (F) day 95 were stained by LIVE/DEAD stain (Molecular Probes, Inc.). The biofilm on day 55 was hybridized with FITC-labeled EUB338 mixed probes (G). All bars in panels (A-F); 20 µm. of more microcolonies on the membrane surface (Figure 2D-H). On day 97 when the TMP exceeded 20 kPa again, the membrane pores were completely fouled and many coccusor bacillus-shaped microorganisms firmly attached on the membrane surfaces with anchor-like structures even after physical washing (Figure 2D-H). The closer observation of these attached microcolonies indicated that some bacteria seem to have pili or curli net to promote cell-to-cell interaction or adhesion to the membrane surfaces (Figure 2F-H). These microorganisms were also embedded in a large amount of extracellular polymeric substance (EPS), suggesting the formation of biofilms. It should be noted that EPSs are commonly produced when biofilms are actively growing or maturing, which significantly increases the biofilm integrity (rigidness) and adhesion strength to the membrane surface (18). Furthermore, EPS that might be produced by the attached bacterial cells might also directly clog the membrane pores. Once the mature biofilms were developed, it would become difficult to remove the biofilms from the membrane surface by routine aeration and physical cleaning. Consequently, the formation of mature biofilms caused severe irreversible membrane fouling in this study. LIVE/DEAD Visualization Biofilms on the Membrane Surfaces. Figure 3 shows representative CLSM images to determine time-dependent development of microcolonies on the membrane surfaces. Viable bacteria with intact cell membranes were stained with SYTO 9 (green), whereas dead bacteria with damaged cell membranes were stained with propidium iodide (red). Figure 3 clearly demonstrated that the number of microcolonies increased with time. At a high magnification we observed single cells and big clumps of cells attached to the membrane surfaces (Figure 3G). The development of the biofilms on the hollow-fiber membrane surface was quantified by measuring membrane surface area covered with biofilms (Figure 4). At the beginning of MBR operation, the biofilm surface coverage was low (less than 1% on day 9) and the viability (live cells/total cells) was also low (less than 15% on day 9). However, the biofilm surface coverage and the viability steadily increased to 18 ( 8% and 30% on day 55, respectively. This is probably because a small fraction of MBR sludge in the mixed liquor, whose viability was low (data not shown), first attached or stuck on the membrane surface, and then specific groups of microorganisms gradually grew and formed irreversible mature biofilms during the MBR operation as described below. These actively growing microorganisms attached on the membrane surfaces

FIGURE 4. Changes in biofilm surface coverage (white circle) and viability (live cells/all cells) (black square) during the operation. The points are average values of 20 time measurements for each sample. The aeration rate was changed from 3800 to 5000 L/h on day 38. Physical cleaning (white arrow) was performed on days 20, 23, 30, 38, 44, 55, 82, 93, and 97, respectively. Chemical cleaning (black arrow) was performed on days 55 and 97, respectively. could not be easily removed by physical washing. Moreover, the increase in biofilm surface coverage synchronized with the increase in TMP (Figure 1). After chemical washing performed on day 55, the similar synchronism in both parameters was repeatedly observed. The other MBR that was operated in parallel with lower MLSS concentration (10 g/L) and the same municipal wastewater also showed exactly the same trend (data not shown), indicating a good reproducibility of the synchronized relationship between biofilm formation and the increase in TMP. This result clearly VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

635

FIGURE 5. Relative abundance of major phylogenetic groups in biofilms formed on the membrane surfaces during MBR operation as determined by whole-cell hybridization with rRNA-targeted fluorescent probes.

FIGURE 6. Relative abundance of major phylogenetic groups in suspended biomass in mixed liquors during MBR operation as determined by whole-cell hybridization with rRNA-targeted fluorescent probes. indicated that the biofilm formation was associated with irreversible membrane fouling. Previous reports suggested that membrane fouling in the MBR treating the municipal wastewater was mainly caused by irreversible adsorption of some fractions of organic matter (i.e., carbohydrate) on the membrane and into membrane pores (1, 2). However, they could not elucidate that the biofilm was developed on the membrane. In this study, the results of SEM and LIVE/DEAD staining analyses clearly demonstrated that the biofilm formation on the PE hollow-fiber membrane surfaces occurred simultaneously with internal fouling caused by the adsorption of dissolved matter into the membrane pores. Microbial Community Structure of the Biofilm. FISH with probes for the major phylogenetic groups revealed that bacterial populations of biofilms were initially composed of mainly Alpha-, Beta-, Gamma-, and Deltaproteobacteria (Figure 5). It should be noted that probes GNSB941- and CFX1223-hybridized filamentous Chloroflexi accounted for approximately 20% of total EUB338-mixed probe hybridized cells in the mixed liquor biomass (Figure 6). They were not detected at all in the biofilms on membrane surfaces. During the initial 38 days when aeration rate was 3500 L/h, the biofilm community structure remained relatively unchanged. After the aeration rate was increased to 5000 L/h on day 38, the relative abundance of Betaproteobacteria significantly increased and became the most dominant phylogenetic group in the mature biofilms (accounting for more than 70% of total EUB338 mixed-stained cells) on days 51 and 55 (Figure 5). After chemical washing on day 55, the similar domination of Betaproteobacteria in the biofilms was repeatedly observed (days 93 and 97). On the other hand, bacterial community 636

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

structure in the mixed liquors remained relatively unchanged (Figure 6). Consequently, our result suggested that a specific phylogenetic group of bacteria (i.e., Betaproteobacteria) was the key microorganism and played an important role in development of mature biofilms on PE hollow-fiber MF membrane surfaces in the MBR treating municipal wastewater. Bacterial population shifts after the change in aeration rate indicate that shear force over membrane surfaces directly influences the composition of biofilm communities, because weakly adhering bacteria could not stay on the membrane surfaces under high shear forces provided by aeration. The high shear stress selected mainly Betaproteobacteria in this study (Figure 5). Rickard et al. have reported that fluid flow velocity moderated the diversity of bacterial community of biofilms; the bacterial diversity decreased with the increase in fluid shear (19). In their study, Actinobacteria and Betaproteobacteria were selected by high shear rates. However, different experimental systems (e.g., different configurations of membrane modules, different substrate, and so on) influence hydrodynamic conditions on the membrane surfaces and microbial populations in mixed liquor, which result in different dominant bacterial populations in the biofilms. For example, Zhang et al. reported that Alphaproteobacteria and Acinetobacter were major bacterial populations in the biofilms formed on MF membrane surfaces in a MBR treating synthetic paper mill wastewater that was operated for only 4 h (8). Moreover, Alphaproteobacteria were predominant bacterial populations in the biofilms formed on flat sheet type of polyvinylidenefluoride (PVDF) MF membrane surfaces in a MBR treating secondary effluent of a local municipal wastewater treatment plant (9). Thus, although the Betaproteobacteria were perhaps associated with biofilm formation and irreversible fouling in this study, other conditions (i.e., shear force, liquid viscosity, and aeration rate) might support different populations that are equally detrimental to filter performance. Analysis of the 16S rRNA Clone Library. The phylogenetic affiliation of clones retrieved from the mature biofilm developed on the membrane surface on day 55 is shown in Figure 7. Forty-seven clones were randomly selected, and partial sequences (approximately 500 bp) were analyzed. All clones were grouped into 36 operational taxonomic units (OTUs) on the basis of more than 97% sequence similarity within an OTU. Nearly complete 16S rRNA gene sequences of the representative 36 OTUs were analyzed. The majority of OTUs (22/36) were affiliated with the Betaproteobacteria subdivision. Other OTUs ( 8, 1, and 1) belonged to the divisions of the Gamma-, Delta-, and Alphaproteobacteria, respectively. This result was co-incident with the FISH results. This confirmed that the Betaproteobacteria were the dominant bacterial population in the biofilm. As the biofilm developed, certain bacterial populations became dominant and persisted within the biofilm, stabilizing the overall community (20). Among the Betaproteobacteria, the clones related to the genera Dechloromonas were most frequently detected (detection frequency of 50%). Similarly, a previous research has showed that only Dechloromonas was detected in late successional stages of natural river biofilms (21). Therefore, we suggest that Dechloromonas-related bacteria are probably one of the particularly important bacterial species for formation of the mature biofilms. In summary, we successfully demonstrated that the biofilm development on hollow-fiber MF membrane surfaces caused severe irreversible fouling during a long-term operation of pilot-scale MBRs treating municipal wastewater. Further identification of the bacterial populations firmly attached on the membrane surfaces suggested that a specific phylogenetic group (i.e., Betaproteobacteria) could be responsible for development of the mature biofilms, which

FIGURE 7. Phylogenetic distance tree representing the affiliation of the 16S rRNA clone sequences retrieved from biofilm samples taken on day 55 (OTU-#). The tree was generated by using nearly full length of 16S rRNA gene sequences and the neighbor-joining method. The scale bar represents 5% estimated divergence. The numbers at the nodes are bootstrap values (100 replicates) with more than 50% bootstrap support. caused irreversible membrane fouling. This result implies that to efficiently control biofilm formation on membrane surfaces, one should focus upon these specific bacterial groups rather than the total microbial community present in the mixed liquor of MBR systems. Since the Betaproteobacteria are most likely key players in this study, further research on their identification and ecophysiology including biofilm formation ability must be performed to provide fundamental information that environmental engineers can use to develop effective biofouling control strategies in MBR systems.

Acknowledgments We gratefully appreciate Nobuhiro Yamato, Taro Miyoshi, and Katsuki Kimura, Department of Urban and Environmental Engineering, Hokkaido University, Sapporo Japan, for MBR operation, and Tsukasa Ito, Department of Civil

Engineering, Faculty of Engineering, Gunma University, Japan, for technical support for molecular-based analyses. In addition, we thank Yoshinobu Nodasaka, Laboratory of Electron Microscopy, Graduate School of Dental Medicine, Hokkaido University, Sapporo, Japan, for kind advice for SEM analysis.

Literature Cited (1) Yamato, N.; Kimura, K.; Miyoshi, T.; Watanabe, Y. Difference in membrane fouling in membrane bioreactors (MBRs) caused by membrane polymer materials. J. Membr.Sci. 2006, 280, 911919. (2) Kimura, K.; Yamato, N.; Yamamura, H.; Watanabe, Y. Membrane fouling in pilot-scale membrane bioreactors (MBRs) treating municipal wastewater. Environ. Sci. Technol. 2005, 39, 62936299. (3) Rosenberger, S.; Laabs, C.; Lesjean, B.; Gnirss, R.; Amy, G.; Jekel, M.; Schrotter, J.-C. Impact of colloidal and soluble organic VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

637

(4) (5)

(6)

(7)

(8) (9)

(10)

(11) (12)

(13) (14)

(15)

(16)

638

material on membrane performance in membrane bioreactors for municipal wastewater treatment. Water Res. 2006, 40, 710720. Itonaga, T.; Watanabe, Y. Performance of membrane bioreactor combined with pre-coagulation/sedimentation. Water Supply 2004, 4, 143-149. Le-Clech, P.; Jefferson, B.; Judd, S. J. Impact of aeration, solids concentration and membrane characteristics on the hydraulic performance of a membrane bioreactor. J. Membr. Sci. 2003, 218, 117-129. Itonaga, T.; Kimura, K.; Watanabe, Y. Influence of suspension viscosity and colloidal particles on permeability of membrane used in membrane bioreactor (MBR). Water Sci. Technol. 2004, 50, 301-309. Ng, H. Y.; Tan, T. W.; Ong, S. L. Membrane fouling of submerged membrane bioreactors: impact of mean cell residence time and the contributing factors. Environ. Sci. Technol. 2006, 40, 2706-2713. Zhang, K.; Choi, H.; Dionysiou, D. D.; Sorial, G. A.; Oerther, D. B. Identifying pioneer bacterial species responsible for biofouling membrane bioreactors. Environ. Microbiol. 2006, 8, 433-440. Chen, C.-L.; Liu, W.-T.; Chong, M.-L.; Wong, M.-T.; Ong, S. L.; Seah, H.; Ng, W. J. Community structure of microbial biofilms associated with membrane-based water purification processes as revealed using a polyphasic approach. Appl. Microbiol. Biotechnol. 2004, 63, 466-473. Amann, R. I. In situ identification of micro-organisms by wholecell hybridization with rRNA-targeted nucleic acid probes. In Molecular Microbial Ecology Manual; Akkerman, A. D. L., van Elsas, J. D., de Bruijn, F. J., Eds.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1995; pp 1-15. Okabe, S.; Satoh, H.; Watanabe, Y. In situ analysis of nitrifying biofilms as determined by in situ hybridization and the use of microelectrodes. Appl. Environ. Microbiol. 1999, 65, 3182-3191. Kane, M. D.; Poulsen, L. K.; Stahl, D. A. Monitoring the enrichment and isolation of sulfate-reducing bacteria by using oligonucleotide hybridization probes designed from environmentally derived 16S rRNA sequences. Appl. Environ. Microbiol. 1993, 59, 682-686. Weisburg, W. G.; Barns, S. M.; Pelletier, D. A; Lane, D. J. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 1991, 173, 697-703. Cole, J. R.; Chai, B.; March, T. L.; Farris, R. J.; Wang, Q.; Kulam, S. A.; Chandra, S.; McGarrell, D. M.; Schmidt, T. M.; Garrity, G. M.; Tiedje, J. M. The Ribosomal Database Profect (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy. Nucleic Acids Res. 2003, 31, 442443. Altschul, S. F.; Madden, T. L.; SAchaffer, A. A.; Zhang, J.; Zhang, Z.; Miller, W.; Lipman, D. J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997, 25, 3389-3402. Thompson, J. D.; Higgins, D. G.; Gibson, T. F. CLUSTAL W: improving the sensitivity of progressive multiple sequence

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

(17) (18) (19) (20) (21) (22)

(23)

(24)

(25)

(26)

(27)

(28)

alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994, 22, 4673-4680. Saito, N.; Nei, M. The neighbor-joining method: a new method for constructing phylogenetic trees. Mol. Biol. Evol. 1987, 4, 406-425. Splendiani, A.; Livingston, A. G.; Nicolella, C. Control of membrane-attached biofilms using surfactants. Biotechnol. Bioeng. 2006, 94, 15-23. Rickard, A. H.; McBain, A. J.; Stead, A. T.; Gilbert, P. Shear rate moderates community diversity in freshwater biofilms. Appl. Environ. Microbiol. 2004, 70, 7426-7435. Jackson, C. R.; Churchill, P. F.; Roden, E. E. Successional changes in bacterial assemblage structure during epilithic biofilm development. Ecology 2001, 82, 555-556. Lyautey, E.; Jackson, C. R.; Cayrou, J.; Rols, J. L.; Garabe´tian, F. Bacterial community succession in natural river biofilm assemblages. Microb. Ecol. 2005, 50, 589-601. Amann, R. I.; Krumholz, L.; Stahl, D. A. Fluorescent-oligonucleotide probing of while cells for determinative, phylogenetic, and environmental studies in microbiology. J. Bacteriol. 1990, 172, 762-770. Daims, H.; Bru ¨ hl, A.; Amann, R.; Schleifer, K.-H.; Wagner, M. The domain-specific probe EUB338 is insufficient for the detection of all bacteria: Development and evaluation of a more comprehensive probe set. Syst. Appl. Microbiol. 1999, 22, 434444. Manz, W.; Amann, R.; Ludwig, W.; Wagner, M.; Schleifer, K.-H. Phylogenetic oligodeoxynucleotide probes for the major subclasses of Proteobacteria: problems and solutions. Syst. Appl. Microbiol. 1992, 15, 593-600. Amann, R.; Binder, B. J.; Olson, R. J.; Chisholm, S. W.; Devereux, R.; Stahl, D. A. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial population. Appl. Environ. Microbiol. 1990, 56, 1919-1925. Rabus, R.; Fukui, M.; Wilkes, H.; Widdle, F. Degradative capacities and 16S rRNA-targeted whole-cell hybridization of sulfatereducing bacteria in an anaerobic enrichment culture utilizing alkylbenzenes from crude oil. Appl. Environ. Microbiol. 1996, 62, 3805-3613. Gich, F.; Garcia-Gil, J.; Overmann, J. Previously unknown and phylogenetically diverse members of the green nonsulfur bacteria are indigenous to freshwater lakes. Arch. Microbiol. 2001, 177, 1-10. Bjo¨rnsson, L.; Hugenholtz, P.; Tyson, G. W.; Blackall, L. L. Filamentous Chloroflexi (green non-sulfur bacteria) are abundant in wastewater treatment processes with biological nutrient removal. Microbiology 2002, 148, 2309-2318.

Received for review June 29, 2006. Revised manuscript received October 16, 2006. Accepted November 1, 2006. ES0615371