Journal of Bioscience and Bioengineering VOL. xx No. xx, 1e8, 2013 www.elsevier.com/locate/jbiosc

Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process Kurumi Hashimoto,1 Masami Matsuda,1 Daisuke Inoue,1, 2 and Michihiko Ike1, * Division of Sustainable Energy and Environmental Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan1 and Department of Health Science, Kitasato University, 1-15-1 Kitasato, Sagamihara-Minami, Kanagawa 252-0373, Japan2 Received 11 October 2013; accepted 10 December 2013 Available online xxx

To elucidate the bacterial community dynamics in a full-scale wastewater treatment plant (WWTP) and the relatedness among bacterial communities in the influent, effluent and sludge, the structure and metabolic ability of the bacterial community throughout a full-scale WWTP employing a conventional activated sludge process was investigated during a period of 10 months. The bacterial community structure was analyzed by terminal-restriction fragment length polymorphism targeting eubacterial 16S rRNA genes, while a Biolog assay was applied to assess the metabolic ability of the activated sludge. Influent bacterial community structure was generally stable. In contrast, the bacterial community structure in the effluent was similar to that in the influent in some cases, while in other cases it was unique and differed greatly from that in the influent and sludge. These results suggest that temporal variations of the effluent bacterial community may be useful to predict the wastewater treatment performance and settleability of activated sludge. The bacterial community structure in the sludge was relatively stable and was rarely impacted by the influent populations. Biolog assay also revealed that activated sludge maintained a remarkably similar metabolic potential of organic compounds over time due to functional redundancy, in which the minor populations played a significant role. Ó 2013, The Society for Biotechnology, Japan. All rights reserved. [Key words: Activated sludge; Bacterial community; Biolog assay; Full-scale wastewater treatment plant; Terminal-restriction fragment length polymorphism]

Activated sludge treatment has been used worldwide in secondary treatment systems in municipal and industrial wastewater treatment plants (WWTPs). The wastewater purification performance of activated sludge treatment depends heavily on the metabolism and interactions of the microbial community, with bacteria playing a key role in the purification process. Previous efforts have enabled efficient removal not only of suspended solids (SS) and biochemical oxygen demand (BOD), but also of nitrogen, phosphorus and xenobiotic compounds, via the appropriate design and operation of treatment systems. Nevertheless, activated sludge treatment is still subject to crucial problems, such as unsuccessful sludge settlement associated with sludge bulking and foaming caused by non-floc forming or filamentous microorganisms (1e3), as well as failure of wastewater treatment due to the loss or inactivation of key populations (4,5). Accordingly, it is necessary to understand the dynamics of the bacterial community to solve these operational problems and develop promising strategies for improved process performance and high-efficiency operation. There have been many attempts to evaluate the diversity and variations in the bacterial community during activated sludge treatment; however, previous studies have focused on lab-scale systems. Because the microbiology in full-scale WWTPs is more complex than that in lab-scale systems operated under artificial

* Corresponding author. Tel.: þ81 6 6879 7672; fax: þ81 6 6879 7675. E-mail address: [email protected] (M. Ike).

conditions (6), it is impossible to directly relate the microbial community dynamics observed in lab-scale studies to the operation of full-scale WWTPs. Consequently, more studies in full-scale systems are needed to obtain practically useful knowledge. To date, relatively few studies have explored the temporal dynamics of the entire activated sludge bacterial community (in aeration tanks) for relatively long periods (more than half a year) in full-scale WWTPs treating municipal wastewater (7e9) or pulp mill effluent (10) to clarify the influence of operational and environmental parameters on the diversity and composition of activated sludge bacterial communities. Moreover, to our knowledge, there have been no in-depth investigations of the long-term dynamic behavior of the bacterial community in activated sludge and the influent and effluent of a full-scale WWTP. The dynamics of the bacterial community in an activated sludge reactor are determined by the sum of the growth and decay of bacterial populations in the bioreactor, the inflow of wastewater-derived populations, and the discharge of unsettled populations in treated water. Therefore, whole-plant monitoring including the inflow and outflow populations is required to fully grasp the dynamics of the activated sludge bacterial community. Furthermore, there have been few attempts to understand structural variations in activated sludge bacterial communities by linking these variations with their metabolic ability, which can have considerable effects on wastewater treatment performance. To our knowledge, Yang et al. (11) conducted the only study that compared the bacterial community in four activated sludge samples collected from geographically

1389-1723/$ e see front matter Ó 2013, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

2

HASHIMOTO ET AL.

J. BIOSCI. BIOENG.,

distinct WWTPs by both phylogenetic and metabolic fingerprinting methods. The results of their study suggested that the activated sludge communities had broad carbon source utilization profiles because of the enrichment of various degraders by long-term acclimation to pollutant-rich wastewater, and that the phylogenetic diversity appeared to be associated with metabolic diversity in the activated sludge microbial community. However, a periodical monitoring study of a WWTP to clarify the dynamic changes of the structure and function of the activated sludge community and their possible links is still needed. In this study, the bacterial community throughout a full-scale WWTP was monitored over 10 months. The primary objectives of this study were to gain a thorough understanding of the dynamic behavior of bacterial community in a full-scale WWTP in light of the phylogenetic structure and metabolic ability of organic compounds and elucidate possible correlations among the bacterial communities in influent wastewater, effluent water and activated sludge. Bacterial community structure was analyzed by terminalrestriction fragment length polymorphism (T-RFLP) analysis (12) targeting eubacterial 16S rRNA genes. T-RFLP analysis has frequently been used in the long-term monitoring of the bacterial community in full-scale WWTPs (7,8,10). In addition, the carbon source utilization potential of the activated sludge community was evaluated by a Biolog assay (13), which is capable of examining the ability of the community to utilize 95 different carbon sources, to understand the activated sludge bacterial community based on its metabolic properties.

22A portable water quality checker (DKK-TOA, Tokyo, Japan). Concentrations of mixed liquor suspended solid (MLSS), SS, dissolved organic carbon (DOC), total nitrogen (T-N), NH4eN, NO2eN, NO3eN, PO4eP and heterotrophic bacteria were determined in the laboratory. Samples were filtered through a qualitative filter paper (pore size, 5 mm; Advantec, Tokyo, Japan) prior to measurement of the DOC, TN, NH4eN, NO2eN, NO3eN and PO4eP. DOC was analyzed using a total organic carbon analyzer (TOC-5000A: Shimadzu, Kyoto, Japan). MLSS/SS, T-N, NH4eN, NO2eN, NO3eN and PO4eP were analyzed according to the test methods of the Japan Industrial Standards K0102, with minor modifications. Heterotrophic bacteria were enumerated using a CGY agar plate (14). DNA extraction Approximately 2 ml of sludge samples and 10 ml of wastewater samples were used for DNA extraction. DNA in the samples was extracted using an ISOIL for Beads Beating kit (Nippon Gene, Tokyo, Japan) and subsequently purified with a MagExtractor-PCR&Gel Clean up-kit (Toyobo, Tokyo, Japan) according to the manufacturer’s instructions. T-RFLP analysis The bacterial community structure was analyzed by T-RFLP targeting eubacterial 16S rRNA genes as described in our previous study (15). Briefly, 16S rRNA genes were PCR amplified using the 27F primer with a 50 -end labeled with 6-FAM and the 1392R primer (16). PCR was terminated during the exponential amplification phase to maintain the composition ratios of the 16S rRNA genes from different bacteria in the DNA template. The PCR products were purified with a Montage PCR kit (Millipore, Billerica, MA, USA) and digested with HhaI, which is a restriction enzyme that can yield a great number of terminal-restriction fragments (T-RFs) (17). Fluorescently labeled T-RFs were separated and detected using an ABI Prism 310 genetic analyzer (Applied Biosystems, Foster City, CA, USA), and their size and abundance were determined with GeneScan ver. 3.7 (Applied Biosystems). Because the PCR products were nearly 1400 bp, GeneScan 2500 ROX Size Standard (Applied Biosystems) was used as the internal standard to measure the size of labeled T-RFs. T-RFs with a peak height of 100 fluorescence units were considered to be positive peaks. The phylogenetic positions representing predominant T-RFs were presumably identified using a database [Microbial Community Analysis (MiCA) III, http://mica.ibest.uidaho.edu] (18).

MATERIALS AND METHODS Wastewater treatment plant and samples A municipal WWTP in Osaka, Japan that employs a conventional activated sludge (CAS) process for secondary treatment was investigated in this study. This WWTP treats 82,400 m3/d of wastewater (over 90% domestic wastewater and small amounts of industrial discharges) for a population of nearly 110,000 people. The hydraulic retention time (HRT) in the biological treatment tank during the sampling period was 8.610.3 h, and the sludge retention time (SRT) was 5.2e7.2 d (Table 1). Samples for T-RFLP analysis and the Biolog assay were collected in July, August and November of 2008 and January and April of 2009. The following four grab samples were collected during each sampling event: influent (Inf), effluent (Eff), and activated sludge (AS) samples from the outlet of the primary settling tank, the final settling tank and the aeration tank, respectively, as well as a return sludge (RS) sample from the return sludge pipe. Hereafter, these samples are referred as the sample type (Inf, Eff, AS or RS)-sampling month (7, 8, 11, 1 or 4). For example, Inf-8 and AS-1 indicate the Inf sample taken in August 2008 and AS sample collected in January 2009, respectively. An AS sample was also collected for Biolog assay in April 2010. All samples were collected in 1-L sterile polyethylene bottles, transported to the laboratory on ice, and subjected to DNA extraction within 3 h and to Biolog assay within 12 h. Water temperature, pH and dissolved oxygen (DO) concentration at the points at which the Inf, Eff and AS samples were collected were recorded on site using a WQC-

Biolog assay The potential of AS samples to utilize 95 different carbon sources was evaluated with a Biolog GN2 plate (Biolog, Hayward, CA, USA) (Supplementary data, Table S1). Briefly, AS samples were washed twice with phosphate buffer (pH 7.5), homogenized with an ultrasonicator (50 W, 2.5 min), and diluted with saline solution (0.85% (w/v)) to obtain an MLSS concentration of 1 mg/L. Aliquots (150 ml) of the pretreated samples were then added to wells of a Biolog GN2 plate. Next, the plates were statically incubated at 28 C in the dark for 72 h, during which time the absorbance at 595 nm (A595) was measured periodically. All assays were performed in duplicate. The average well color development (AWCD) for all 95 carbon sources (13) was calculated as an index to evaluate the carbon source utilization potential. Wells with an A595 of over 0.25 were considered to be positive (utilizable), and the number of utilizable carbon sources was used as the other index. Numerical analysis Microbial community diversity was evaluated using the Shannon-Weaver index (H0 ). H0 (T-RFLP) (19) and H0 (Biolog) (20) were calculated from the T-RFLP and carbon source utilization profiles, respectively, by the following equations: H^0 (T-RFLP) ¼ S (Pi  ln Pi)

(1)

H^0 (Biolog) ¼ S (Ri  ln Ri)

(2)

TABLE 1. Operational parameters and wastewater treatment performance in WWTP investigated in this study.a Sampling month

Sample

July 2008

Inf-7 AS-7 Eff-7 Inf-8 AS-8 Eff-8 Inf-11 AS-11 Eff-11 Inf-1 AS-1 Eff-1 Inf-4 AS-4 Eff-4

August 2008

November 2008

January 2009

April 2009

a

SRT (day)

HRT (h)

Temp ( C)

pH

6.7

8.8

7.2

8.8

5.2

9.2

5.5

8.6

6.8

10.3

26.8 26.9 26.9 29.9 30.4 30.4 23.7 25.1 24.7 19.5 20.7 20.3 23.1 23.6 23.2

7.6 6.8 6.6 7.4 6.7 6.6 7.6 6.9 6.7 7.6 7.0 6.5 7.1 7.1 7.0

Heterotrophic bacteria (CFU/ml) 2.4 9.7 1.7 1.3 9.8 2.7 7.1 3.5 4.1 5.7 6.5 1.0 1.6 1.4 2.1

              

106 106 105 107 106 105 107 107 104 106 107 105 106 107 104

Concentration (mg/L) of

Removal (%) of

DO

MLSS/SS

DOC

T-N

NH4eN

NO2eN

NO3eN

PO4eP

SS

DOC

T-N

NH4eN

0.06 2.6 0.05 0.5 3.0 0.5 0.4 2.5 0.8 0 2.4 0.2 0 3.0 0

55.2 1212 2.7 35.0 1013 1.8 74.5 1332 13.0 74.5 1617 7.9 83.0 1239 ND

54.3 15.7 12.1 59.2 22.1 20.1 84.2 16.8 25.7 NA 13.7 5.8 56.6 13.1 9.1

14.7 10.3 11.2 12.9 10.4 10.2 29.0 12.2 15.9 NA NA NA 26.3 13.1 17.8

16.5 3.6 1.8 12.6 3.3 2.1 23.0 4.9 2.2 31.8 14.4 4.1 23.5 8.9 8.4

0.2 0.3 1.1 0.03 0.2 0.3 0.1 0.7 0.1 0.2 0.6 4.9 0.03 0.3 1.6

0.2 6.6 4.7 0.2 8.4 8.3 0.3 6.0 10.5 0.5 0.2 3.4 0.1 0.6 3.9

1.5 0.02 0.04 3.0 0.1 1.2 3.0 0.05 0.9 1.3 0.01 0.03 3.1 0.4 0.1

95.1

77.7

23.8

89.1

94.9

66.0

20.9

83.3

82.6

69.5

45.2

90.4

89.4

NA

NA

87.1

100

83.9

32.3

64.3

NA, not analyzed; ND, not detected.

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

VOL. xx, 2013

BACTERIAL COMMUNITY DYNAMICS IN A FULL-SCALE WWTP

numbers detected in Inf, AS, Eff and RS samples were 18e30, 28e37, 23e30 and 17e26, respectively. Higher bacterial diversity was found in AS and RS samples than in Inf and Eff samples, and in AS samples than RS samples. All of the Inf samples had similar T-RFLP profiles to each other, with common predominant T-RFs of 98 bp, 186 bp, 209 bp and 571e572 bp (Fig. 1A). The relative abundances (Pi values) of those TRFs were 0.15e0.33, 0.040e0.073, 0.047e0.078 and 0.069e0.095, respectively. Based on the MiCA database search, T-RFs of 98 bp and 571e572 bp were assumed to be derived from the CytophagaFlexibacter-Bacteroides (CFB) group and coliforms, respectively, while no specific groups were assigned for the other predominant T-RFs. The T-RFLP profiles of AS and RS samples were similar to each other (Fig. 1B, D), and largely different from those of Inf samples (Fig. 1A). T-RFs of 59e60 bp, 78 bp, 198 bp and 203 bp were commonly present in all of the AS and RS samples, with a T-RF of 203 bp being predominant at a Pi value of 0.054e0.18. In addition, a T-RF of 201 bp was predominantly present with a Pi value of 0.078e0.21 in all sludge samples but RS-4. T-RFs of 201e203 bp were also detected in AS samples of various treatment processes in our previous study (15). In contrast, predominant T-RFs in Inf samples (98 bp, 186 bp, 209 bp and 571e572 bp) were undetectable or minor in AS and RS samples. A database search suggested that TRFs of 201 bp and 203 bp may represent Actinobacteria and bProteobacteria or Firmicutes, respectively. Five Eff samples were divided into two groups, Eff-7 and Eff-4 and Eff-8, Eff-11 and Eff-1 (Fig. 1C). Eff-7 and Eff-4 exhibited similar T-RFLP profiles to Inf samples, with common predominant T-RFs of 98 bp (Pi: 0.16 and 0.11, respectively), 186 bp (Pi: 0.025 and 0.027, respectively), 209 bp (Pi: 0.15 and 0.13, respectively) and 571e572 bp (Pi: 0.11 and 0.076, respectively). In contrast, Eff-8, Eff11 and Eff-1 had specific T-RFLP profiles and most of their detected T-RFs were undetected in other samples. Indeed, although T-RFs of 117 bp, 118 bp, 122 bp and 123 bp were dominant (Pi > 0.05) in Eff8, Eff-11 and Eff-1, none of these were detected in the other samples. Among the specific T-RFs, those of 117 bp, 118 bp and 122 bp were assumed to be derived from g-Proteobacteria, Firmicutes and a-Proteobacteria, respectively. Cluster analysis of the T-RFLP profiles revealed that 20 samples analyzed in this study were separated into three clusters: (i) Eff-8,

where Pi is the relative abundance of the targeted T-RF in the T-RFLP analysis and Ri is the relative color development of the targeted well to total color development for all 95 substrates in a Biolog GN2 plate. To evaluate the similarities of the bacterial community compositions among samples, cluster analysis and principal component analysis (PCA) were conducted using the Pi matrix obtained from T-RFLP analysis using PAST ver. 1.3.4 (http://folk. uio.no/ohammer/past/). In the cluster analysis, a dendrogram was constructed from Dice’s coefficients (SD) using the unweighted pair group method with arithmetic averages (UPGMA).

RESULTS Wastewater treatment performance Operational parameters and wastewater treatment performance in the investigated WWTP are presented in Table 1. The MLSS concentration in the aeration tank fluctuated between 1013 mg/L and 1617 mg/L, and was slightly lower during the warm season (July and August) than the cold season (November and January). The concentrations of SS, DOC, T-N and NH4eN in influents discharging from the primary settling tank were 35.0e83.0 mg/L, 54.3e84.2 mg/L, 12.9e29.0 mg-N/L and 12.6e31.8 mg-N/L, respectively. More than 83% of the SS were removed after CAS treatment. The SS concentrations in the effluents were less than 13 mg/L, and were higher in November and January than during other months. The DOC removal by the CAS treatment varied between 66% and 84%; however, a report provided from the investigated WWTP showed that 98.9e99.4% of the BOD removal was accomplished during our monitoring period (data not shown). Additionally, 64e90% of the NH4eN was removed (oxidized) by the CAS treatment, and NO3eN occurred at levels of 3.4e10.5 mg/L in the effluents. Because of the process being fully aerated, the T-N removal was low (21e45%). Taken together, these results indicated that efficient removal of SS and readily-biodegradable organic compounds and nitrification, which is prerequisite for CAS treatment, was stably achieved in the investigated WWTP. In addition, low concentrations of SS and heterotrophic bacteria in the effluents compared with those in influents and activated sludge indicated relatively successful sludge settlement. Bacterial community profiles in influent, effluent and activated and return sludge Fig. 1 shows the T-RFLP profiles of bacterial communities in all of the Inf, AS, Eff and RS samples during the 10-month study period. As shown in Table 2, T-RF

A

T-RF length (bp) 100

200

300

400

500

600

700

800

900

1000

B 100

T-RF length (bp) 200

300

400

500

600

700

800

900

1000 AS-7

Inf-11

Inf-1

Fluorescence intensity

Fluorescence intensity

Inf-7

Inf-8

AS-8

AS-11

AS-1

AS-4

Inf-4

C

T-RF length (bp) 100

200

300

400

500

600

700

800

900

1000

D 100

200

300

T-RF length (bp)

400

500

600

700

800

Eff-11

Eff-1

900

1000 RS-7

Fluorescence intensity

Fluorescence intensity

Eff-7

Eff-8

3

Eff-4

RS-8

RS-11

RS-1

RS-4

FIG. 1. T-RFLP profiles of bacterial community in Inf (A), AS (B), Eff (C) and RS samples (D).

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

4

HASHIMOTO ET AL.

J. BIOSCI. BIOENG.,

TABLE 2. T-RF number and Shannon-Weaver index [H0 (T-RFLP)] obtained from TRFLP analysis. Sampling month July 2008

August 2008

November 2008

January 2009

April 2009

Sample

T-RF number

H0 (T-RFLP)

Inf-7 AS-7 Eff-7 RS-7 Inf-8 AS-8 Eff-8 RS-8 Inf-11 AS-11 Eff-11 RS-11 Inf-1 AS-1 Eff-1 RS-1 Inf-4 AS-4 Eff-4 RS-4

22 37 23 22 27 33 30 20 30 28 30 17 18 30 27 17 20 35 28 26

2.6 3.4 2.7 2.8 2.8 3.2 2.7 3.1 2.9 3.0 2.5 3.0 2.4 3.1 2.6 2.9 2.5 3.4 2.9 2.9

Eff-11 and Eff-1; (ii) the Inf samples of all five months and two Eff samples (Eff-7 and Eff-4); (iii) the AS and RS samples of all five monitoring months (Fig. 2). Samples in the first, second and third groups had similarity values (SD value) of 0.62e0.85, 0.41e0.73 and 0.52e0.74, respectively. PCA was also performed with the same dataset as cluster analysis. The first and second PCs (PC1 and PC2, respectively) in PCA accounted for 41.5% and 25.6% of the total

0

0.20

Similarity (SD) 0.60 0.40

0.80

1.0 Eff-8 Eff-11 Eff-1 Eff-7 Inf-1

variation in the dataset, and a scatter plot with PC1 and PC2 values clearly differentiated 20 analyzed samples into three core groups as also observed upon cluster analysis (Fig. 3). The results of both analyses confirmed the temporal stability of bacterial community structure in each of the Inf, AS and RS samples and variability of that in Eff samples during our monitoring period. Carbon source utilization potential of activated sludge community As shown in Fig. 4, the AWCD started increasing after an approximately 15 h lag period, regardless of the sampling month, after which it increased with incubation time, reaching 0.68e0.80 at 72 h. The number of utilizable carbon sources and H0 (Biolog) value also increased with time, reaching 70e78 and 4.3 at 72 h, respectively (Table 3). Cluster analysis using Biolog profiles revealed that all AS samples had similar carbon source utilization potential with SD values of 0.93, 0.95 and 0.96 after 24, 48 and 72 h of incubation, respectively (data not shown). Polysaccharides could be rapidly utilized by all AS samples, and rapid utilization of surfactants was also observed in AS-7 and AS-11 (Supplementary data, Fig. S1). In addition, monosaccharides were commonly utilized well by the tested AS samples, while the utilization of di- and trisaccharides varied depending on sampling month. Carboxylic acids, amino acids, alcohols and nucleotides did not appear to be utilized well by AS samples under the incubation conditions applied in this study. Additional experiments to try to identify the bacterial populations responsible for the utilization of carbon sources were performed using an AS sample collected in April 2010. DNA was extracted from the original AS sample and five wells of the Biolog plate that contained carbon sources that were utilized well after 72-h of incubation, and then subjected to T-RFLP analysis. A T-RF of 203 bp, a common predominant T-RF in AS samples (Fig. 1B), was also dominant after incubation on most carbon sources investigated (Fig. 5). In addition, minor or undetectable T-RFs in the original AS sample became dominant for most of the tested carbon sources, excepting D-gluconic acid (Fig. 5E), suggesting that minor populations in activated sludge may make considerable contributions to the degradation of various organic compounds and their metabolites in wastewater. However, possible specific groups for those minor populations could not be phylogenetically inferred by the MiCA database search conducted in this study.

Inf-4 Inf-8

DISCUSSION

Inf-7

This study applied T-RFLP analysis to investigate the bacterial community in different parts of a full-scale WWTP. To our knowledge, this is the first study in which samples were collected repeatedly for a prolonged period of time and used to compare the bacterial community in activated and return sludge, influent wastewater and effluent water in a WWTP. T-RFLP analysis is a simple, highly reproducible and robust molecular method for rapid profiling of the microbial community (21,22). However, the method has several limitations. Because several phylogenetically distinct bacteria can generate a T-RF of the same length (10,23), detailed analysis of microbial community composition and diversity requires the use of two or more restriction enzymes. In addition, identification of the species representing a single T-RF is often difficult and uncertain, and complete phylogenetic characterization of a given community needs the employment of sequence analysis of clone libraries or pyrosequencing. However, previous study reported that differentiation of soil communities by T-RFLP analysis was consistent with that obtained by analysis of clone libraries (24). Another study also demonstrated that both T-RFLP analysis and pyrosequencing are capable of yielding comparable microbial community structure (25). Those evidences support that the use of T-RFLP analysis is

Inf-11 Eff-4 AS-4 RS-4 AS-7 AS-11 RS-8 AS-1 RS-11 RS-7 RS-1 AS-8 FIG. 2. Dendrogram showing clustering of T-RFLP profiles.

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

VOL. xx, 2013

BACTERIAL COMMUNITY DYNAMICS IN A FULL-SCALE WWTP

5

FIG. 3. Ordination produced from principal component analysis of T-RFLP profiles. (A) Results of all Inf, AS, Eff and RS samples; (B) highlighted expression of a part of plot area showing AS and RS samples. Symbols: closed triangle, Inf sample; closed circle, AS sample; open square, Eff sample; open diamond, RS sample. The numbers near the plots represent the sampling month.

sufficient for the purpose of this study that compares bacterial communities in the samples collected from different parts of a fullscale WWTP and on distinct time points. The T-RFLP profiles of the Inf samples were similar to each other throughout the monitoring period (Figs. 1A, 2 and 3 and Table 2). These findings suggested that the bacterial community in the influent wastewater was generally stable against temporal changes. Given that the investigated WWTP primarily receives domestic wastewater, human feces was likely the main bacterial source. Previous pyrosequencing studies have revealed that members of Bacteroidetes, Firmicutes (mainly class Clostridia) and Actinobacteria accounted for the majority of bacteria in human feces, while members of Proteobacteria were rarely detected (26e28). In contrast, McLellan et al. (27) analyzed the bacterial community compositions in untreated sewage samples collected from two WWTPs in metropolitan Milwaukee, USA and compared them with those in human feces. They found that Proteobacteria (mainly band g-Proteobacteria) dominated the untreated sewage, but were minor in feces. In this study, although putative identification of all of the detected T-RFs was not possible because of the application of single endonuclease digestion in T-RFLP analysis, some common predominant bacteria were deduced to be members of the CFB group and coliforms. Thus, our results appeared to be partly in

accordance with those of McLellan et al. (27). The dissimilar bacterial community compositions in feces and in WWTP influent would be caused by the selective proliferation of specific populations, particularly some members of Proteobacteria, through the sewer system (27), as well as by the poor survival of other populations (i.e., predominant members in feces) not suited for persistence in the sewer system. Because our samples were collected from the outlet of the primary settling tank, the selective elimination of some populations attached to SS materials during the sedimentation also likely had a large effect on changes in the bacterial community composition. Consequently, it was suggested that the temporal stability of the influent bacterial community observed in this study could be attributed to the temporal (seasonal) invariance of human-derived feces and generally stable environment in the sewer system. Unlike the influent wastewater, the T-RFLP profiles of effluent water discharged from the final settling tank were divided into two groups (Figs. 1C, 2 and 3). One group had profiles similar to those of influent wastewater, while the other group had specific profiles that were very different from those of influent wastewater and sludge. The groupings were neither season-dependent nor gradually shifting with time. Because the SS concentration in effluent was low in the former group and relatively high in the latter group (Table 1), the dissimilarity between bacterial communities in the two groups could be attributed to settlement characteristics of the activated sludge community. Specifically, when the overall activated sludge maintains stable flocs and can settle readily, the bacterial community in the effluent (treated) water would resemble that in influent wastewater, as observed in July 2008 and April 2009. Conversely, when activated sludge no longer settles easily, the bacterial community in effluent water may vary from that in influent wastewater, as detected in August and November 2008 and January 2009 in this study. Since better DOC removal was observed in the sampling months when the effluent bacterial community was similar to the influent bacterial community (Table 1), comparison of the influent and effluent bacterial

TABLE 3. Utilized carbon source number, average well color development (AWCD) and Shannon-Weaver index [H0 (Biolog)] obtained from Biolog assay of AS samples. Sample

FIG. 4. Time courses of AWCD in Biolog assay of AS-7 (closed circles), AS-8 (open squares), AS-11 (open triangles), AS-1 (closed diamonds) and AS-4 (open diamond).

AS-7 AS-8 AS-11 AS-1 AS-4

Utilized carbon source number at 72 h 77 74 70 78 73

H0 (Biolog)

AWCD 24 h

48 h

72 h

24 h

48 h

72 h

0.17 0.15 0.21 0.14 0.21

0.52 0.48 0.43 0.49 0.57

0.80 0.71 0.68 0.68 0.75

3.8 3.9 3.9 3.9 4.1

4.2 4.2 4.2 4.3 4.2

4.3 4.3 4.3 4.3 4.3

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

6

HASHIMOTO ET AL.

J. BIOSCI. BIOENG.,

FIG. 5. T-RFLP profiles of AS sample, which was collected in April 2010, without (A) and with cultivation on glycogen (B), N-acetyl-D-galactosamine (C), D-cellobiose (D), D-gluconic acid (E) and D-glucose-6-phosphate (F) in Biolog assay.

community or detection of some bacterial species (T-RFs) specific to effluent may be used to predict the wastewater treatment performance and health status of activated sludge. Although effluent water has primarily been investigated for pathogenic microbes (29,30), our results suggest the need to identify the specific populations and understand the bacterial community in effluent for the improvement of effluent water quality, as shown in previous studies (31). The T-RFLP profiles of AS and RS samples were relatively similar to one another, regardless of the monitoring month (Figs. 1B, D, 2 and 3). These findings indicated that the bacterial community in sludge was generally stable, although it fluctuated slightly. Previous studies also reported the temporal stability of the activated sludge bacterial community in WWTPs used to treat pulp mill effluent (10,32) and municipal wastewater (7,9,33), although continuous marginal fluctuations occurred in all cases, depending on the variations in influent composition and operational parameters. Exceptionally, Yi et al. (9) found that bacterial community structure of activated sludge changed remarkably in response to the occurrence of Flexibacter sp. associated with a bulking problem. Therefore, it can be assumed that although the activated sludge bacterial community in a full-scale WWTP fluctuates continuously with influent composition and operational variables, its temporal variations are insubstantial and general structural similarity is maintained under given selective pressures in the WWTP, except under some specific situations such as the occurrence of sludge bulking. Comparison of the bacterial community in influent wastewater, activated/return sludge and effluent water revealed that the sludge bacterial community was completely different from that of the influent, suggesting that the influent populations have little impact

on the bacterial community of activated sludge in the bioreactor of a WWTP. Liu et al. (34) also found that the bacterial community in influent wastewater and activated sludge differed from each other in two WWTPs in Beijing, China, although no repeated investigation was conducted. The number of influent bacteria is generally smaller than that of activated sludge bacteria. In addition, the influent bacteria would not be suitable to grow and persist under the operational and environmental conditions of activated sludge reactor in the WWTP, and thus decrease their number in the activated sludge reactor. Consequently, influent bacterial community was likely to pose negligible impact on the activated sludge bacterial community. In contrast, the bacterial community in the effluent was considerably different from that in the activated sludge, indicating that bacterial selection by sedimentation can make a large contribution to changes in the bacterial community structure. Liu et al. (34) found that the bacterial community structure of activated sludge with bad settling performance (slight foam formation) was similar to that of effluent. Taking this and our suggestion described above into consideration, the temporal variations of the effluent bacterial community (especially the predominance of effluentspecific populations) and the similarity between the bacterial community structure in activated sludge and effluent may be used as a measure to determine the settleability of activated sludge and predict the occurrence of settling problems. Al-Mutairi (35) showed that the carbon source utilization potential of activated sludge from three WWTPs in Kuwait varied seasonally, with better utilization occurring in summer than in winter. However, our Biolog assays revealed that all of the AS samples had remarkably similar carbon source utilization potential

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

VOL. xx, 2013

BACTERIAL COMMUNITY DYNAMICS IN A FULL-SCALE WWTP

with regard to the utilizable carbon source number, AWCD, H0 (Biolog) and types of readily-utilizable and difficult/non-utilizable carbon source (Figs. 4 and S1 and Table 3). These results coincided with the removal performance of organic compounds (BOD and DOC) (Table 1). The stability of carbon source utilization potential may have resulted from the functional redundancy of the activated sludge community enabling it to maintain its function over time. Our trials to identify the populations relevant to the utilization of some specific carbon sources revealed that minor populations with appearance/disappearance that was not reflected in the overall bacterial community structure played significant roles in the utilization of carbon sources (Fig. 5) although they could not be phylogenetically identified. Thus, the activated sludge bacterial community could maintain equivalent organics degradation potential (carbon source utilization potential) over time through temporal fluctuations of minor populations along with the changes in operational and environmental conditions. Furthermore, small temperature variations in the WWTP (Table 1) may help maintain a high metabolic activity of the activated sludge community, even in cooler seasons. In conclusion, results of this study suggested that the influent populations impacted negligibly on the sludge bacterial community in the bioreactor, and that temporal changes in the effluent bacterial community, more specifically the predominance of some specific populations, may be useful for prediction of the wastewater treatment performance and settleability of activated sludge. This study also revealed that minor populations made large contributions to the utilization of carbon sources by the activated sludge. Thus, the functional redundancy that is attributable, at least in part, to the minor populations would enable the activated sludge to maintain equivalent potential to degrade organic compounds in wastewater independent from temporal structural fluctuations of the activated sludge community. Phylogenetic identification of the predominant populations in influent, effluent and sludge and the minor populations that pose a great impact on the degradation ability of organic compounds by clone library or pyrosequencing techniques will further deepen the understanding of the bacterial community dynamics in full-scale WWTPs. Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jbiosc.2013.12.008. ACKNOWLEDGMENTS We appreciate the anonymous WWTP officials for allowing us to collect the samples used in this study and for providing helpful information. We also thank Dr. Hirofumi Tsutsui at the Division of Sustainable Energy and Environmental Engineering, Osaka University (present affiliation: Agriculture Unit, Natural Science Cluster, Research and Education Faculty, Kochi University) for his assistance with sample collection. This study was partly supported by Grants-in-Aid for Young Scientists (A) (no. 24681010) from the Japan Society for the Promotion of Science (JSPS). References 1. de los Reyes, F. L., III and Raskin, L.: Role of filamentous microorganisms in activated sludge foaming: relationship of mycolata levels to foaming initiation and stability, Water Res., 36, 445e459 (2002). 2. Wagner, M., Loy, A., Nogueira, R., Purkhold, U., Lee, N., and Daims, H.: Microbial community composition and function in wastewater treatment plants, Antonie van Leeuwenhoek, 81, 665e680 (2002). 3. Martins, A. M. P., Pagilla, K., Heijnen, J. J., and van Loosdrecht, M. C. M.: Filamentous bulking sludge e a critical review, Water Res., 38, 793e817 (2004). 4. Yu, Z. and Mohn, W.: Bioaugmentation with resin-acid-degrading bacteria enhances resin acid removal in sequencing batch reactors treating pulp mill effluents, Water Res., 35, 883e890 (2001).

7

5. Graham, D. W. and Smith, V. H.: Designed ecosystem services: application of ecological principles in wastewater treatment engineering, Front. Ecol. Environ., 2, 199e206 (2004). 6. Wong, M.-T., Mino, T., Seviour, R. J., Onuki, M., and Liu, W.-T.: In situ identification and characterization of the microbial community structure of fullscale enhanced biological phosphorus removal plants in Japan, Water Res., 39, 2901e2914 (2005). 7. Wells, G. F., Park, H.-D., Eggleston, B., Francis, C. A., and Criddle, C. S.: Finescale bacterial community dynamics and the taxa-time relationship within a full-scale activated sludge bioreactor, Water Res., 45, 5476e5488 (2011). 8. Valentín-Vargas, A., Toro-Labrador, G., and Massol-Deyá, A. A.: Bacterial community dynamics in full-scale activated sludge bioreactors: operational and ecological factors driving community assembly and performance, PLoS One, 7, e42524 (2012). 9. Yi, T., Lee, E. H., Kang, S., Shin, J., and Cho, K. S.: Structure and dynamics of microbial community in full-scale activated sludge reactors, J. Ind. Microbiol. Biotechnol., 39, 19e25 (2012). 10. Gilbride, K. A., Frigon, D., Cesnik, A., Gawat, J., and Fulthorpe, R. R.: Effect of chemical and physical parameters on a pulp mill biotreatment bacterial community, Water Res., 40, 775e787 (2006). 11. Yang, C., Zhang, W., Liu, R., Li, Q., Li, B., Wang, S., Song, C., Qiao, C., and Mulchandani, A.: Phylogenetic diversity and metabolic potential of activated sludge microbial communities in full-scale wastewater treatment plants, Environ. Sci. Technol., 45, 7408e7415 (2011). 12. Liu, W. T., Marsh, T. L., Cheng, H., and Forney, L. J.: Characterization of microbial diversity by determining terminal restriction fragment length polymorphisms of genes encoding 16S rRNA, Appl. Environ. Microbiol., 63, 4516e4522 (1997). 13. Garland, J. L. and Mills, A. L.: Classification and characterization of heterotrophic microbial communities on the basis of patterns of community-level sole-carbonsource utilization, Appl. Environ. Microbiol., 57, 2351e2359 (1991). 14. Pike, E. B., Carrington, E. G., and Ashburner, P. A.: An evaluation of procedures for enumerating bacteria in activated sludge, J. Appl. Bacteriol., 35, 309e321 (1972). 15. Matsuda, M., Inoue, D., Anami, Y., Tsutsui, H., Sei, K., Soda, S., and Ike, M.: Comparative analysis of DNA-based microbial community composition and substrate utilisation patterns of activated sludge microorganisms from wastewater treatment plants operated under different conditions, Water Sci. Technol., 61, 2843e2851 (2010). 16. Amann, R. I., Ludwig, W., and Schleifer, K. H.: Phylogenetic identification and in situ detection of individual microbial cells without cultivation, Microbiol. Rev., 59, 143e169 (1995). 17. Marsh, T. L., Saxman, P., Cole, J., and Tiedje, J.: Terminal restriction fragment length polymorphism analysis program, a web-based research tool for microbial community analysis, Appl. Environ. Microbiol., 66, 3616e3620 (2000). 18. Shyu, C., Soule, T., Bent, S. J., Foster, J. A., and Forney, L. J.: MiCA: a web-based tool for the analysis of microbial communities based on terminal-restriction fragment length polymorphisms of 16S and 18S rRNA genes, Microb. Ecol., 53, 562e570 (2007). 19. Blackwood, C. B., Hudleston, D., Zak, D. R., and Buyer, J. S.: Interpreting ecological diversity indices applied to terminal restriction fragment length polymorphism data: insights from simulated microbial communities, Appl. Environ. Microbiol., 73, 5276e5283 (2007). 20. Al-Mutairi, N. Z.: Functional biodiversity of microbial communities in aerobic selector slaughterhouse wastewater, Water Environ. Res., 79, 660e666 (2007). 21. Osborn, A. M., Moore, E. R. B., and Timmis, K. N.: An evaluation of terminalrestriction fragment length polymorphism (T-RFLP) analysis for the study of microbial community structure and dynamics, Environ. Microbiol., 2, 39e50 (2000). 22. Kitts, C. L.: Terminal restriction fragment patterns: a tool for comparing microbial communities and assessing community dynamics, Curr. Issues Intest. Microbiol., 2, 17e25 (2001). 23. Dunbar, J., Ticknor, L. O., and Kuske, C. R.: Phylogenetic specificity and reproducibility and new method for analysis of terminal restriction fragment profiles of 16S rRNA genes from bacterial communities, Appl. Environ. Microbiol., 67, 190e197 (2001). 24. Dunbar, J., Ticknor, L. O., and Kuske, C. R.: Assessment of microbial diversity in four southwestern United States soils by 16S rRNA gene terminal restriction fragment analysis, Appl. Environ. Microbiol., 66, 2943e2950 (2000). 25. Pilloni, G., Granitsiotis, M. S., Engel, M., and Lueders, T.: Testing the limits of 454 pyrotag sequencing: reproducibility, quantitative assessment and comparison to T-RFLP fingerprinting of aquifer microbes, PLoS One, 7, e40467 (2012). 26. Lauber, C. L., Zhou, N., Gordon, J. I., Knight, R., and Fierer, N.: Effect of storage conditions on the assessment of bacterial community structure in soil and human-associated samples, FEMS Microbiol. Lett., 307, 80e86 (2010). 27. McLellan, S. L., Huse, S. M., Mueller-Spitz, S. R., Andreishcheva, E. N., and Sogin, M. L.: Diversity and population structure of sewage-derived microorganisms in wastewater treatment plant influent, Environ. Microbiol., 12, 378e392 (2010).

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008

8

HASHIMOTO ET AL.

28. Lei, F., Yin, Y., Wang, Y., Deng, B., Yu, H. D., Li, L., Xiang, C., Wang, S., Zhu, B., and Wang, X.: Higher-level production of volatile fatty acids in vitro by chicken gut microbiotas than by human gut microbiotas as determined by functional analyses, Appl. Environ. Microbiol., 78, 5763e5772 (2012). 29. Sahlström, L., Aspan, A., Bagge, E., Danielsson-Tham, M. L., and Albihn, A.: Bacterial pathogen incidences in sludge from Swedish sewage treatment plants, Water Res., 38, 1989e1994 (2004). 30. Wen, Q., Tutuka, C., Keegan, A., and Jin, B.: Fate of pathogenic microorganisms and indicators in secondary activated sludge wastewater treatment plants, J. Environ. Manag., 90, 1442e1447 (2009). 31. Langenbach, K., Kuschk, P., Horn, H., and Kästner, M.: Slow sand filtration of secondary clarifier effluent for wastewater reuse, Environ. Sci. Technol., 43, 5896e5901 (2009).

J. BIOSCI. BIOENG., 32. Smith, N. R., Yu, Z., and Mohn, W. W.: Stability of the bacterial community in a pulp mill effluent treatment system during normal operation and a system shutdown, Water Res., 37, 4873e4884 (2003). 33. Cui, D., Li, A., Zhang, S., Pang, C., Yang, J., Guo, J., Ma, F., Wang, J., and Ren, N.: Microbial community analysis of three municipal wastewater treatment plants in winter and spring using culture-dependent and culture-independent methods, World J. Microbiol. Biotechnol., 28, 2341e2353 (2012). 34. Liu, X., Zhang, Y., Yang, M., Wang, Z., and LV, W.: Analysis of bacterial community structures in two sewage treatment plants with different sludge properties and treatment performance by nested PCR-DGGE method, J. Environ. Sci, 19, 60e66 (2007). 35. Al-Mutairi, N. Z.: Variable distributional characteristics of substrate utilization patterns in activated sludge plants in Kuwait, Bioresour. Technol., 100, 1524e1532 (2009).

Please cite this article in press as: Hashimoto, K., et al., Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process, J. Biosci. Bioeng., (2013), http://dx.doi.org/10.1016/j.jbiosc.2013.12.008