Microbial regrowth, microbial growth after disinfection, is an important problem that deteriorates water quality during the storage and distribution of reclaimed water. Biodegradable organic matter (BOM) that remains after water reclamation processes directly promotes microbial regrowth. In this study we propose a novel assay called the ‘bacterial growth fingerprint (BGF)’ to characterise BOM based on the maximum growth of bacterial strains, which is the extension of the conventional assimilable organic carbon assay for drinking water. Nine bacterial strains were selected from nearly 200 isolates from various reclaimed water systems. These selected bacterial strains exhibited unique substrate utilisation patterns. The BGF assay clearly reflected the difference in the quantity and quality of BOM between six different reclamation plants and the changes in BOM during a full-scale reclamation process. The information on BOM revealed by the BGF assay is useful to optimise the treatment processes or operational conditions for biologically stable reclaimed water.

INTRODUCTION

The reclamation and reuse of wastewater has become increasingly important to compensate for a lack of fresh water resources. An important hindrance in water reuse is bacterial regrowth during storage and distribution. Bacterial regrowth can potentially cause problems such as deterioration of aesthetic quality (Narasimhan et al. 2005), health concerns related to opportunistic pathogens (Jjemba et al. 2010) and reduction of heat transfer efficiency in cooling systems (Zan et al. 2010). Bacterial regrowth is usually managed by maintaining residual chlorine. However, residual chlorine can be depleted during long retention times. In a full-scale reclaimed water distribution system in Japan, an increase in number of microorganisms (heterotrophic plate counts) was observed whenever the residual chlorine concentration was low (Thayanukul et al. 2013a). It is necessary to control the substrate for heterotrophic bacteria, i.e., biodegradable organic matter (BOM), to control bacterial regrowth in reclaimed water at reduced disinfectant concentrations. However, the quality and quantity of BOM in reclaimed water, as well as the fate of BOM during the water reclamation process, remain largely unknown due to the limitation of chemical analysis.

The amount of BOM is often evaluated by biodegradable organic carbon (BDOC), which is total organic carbon (TOC) consumed by indigenous microorganisms during the sample incubation (Servais et al. 1987). However, it is difficult to evaluate BDOC at the concentration level close to the quantification limit of TOC measurement. In drinking water research, the assimilable organic carbon (AOC) assay has been widely used to quantify a portion of BOM. In the AOC assay, the growth of Pseudomonas fluorescens P-17 and Aquaspirillum sp. NOX in a pasteurised water sample is converted to the amount of substrate consumed for growth (van der Kooij et al. 1982b; APHA et al. 2005). P-17 and NOX, which were originally isolated from drinking water system, have different substrate utilisation spectra. P-17 can assimilate a wide range of organic matter, whereas NOX specifically utilises carboxylic acids (van der Kooij et al. 1982a; van der Kooij & Hijnen 1984).

Several studies have applied the conventional AOC assay using P17 and NOX to reclaimed water. The AOC concentration in eight reclaimed water samples produced from full-scale water reclamation plants in Japan ranged from 36 to 446 μg-C l−1 (median, 316 μg-C l−1) (Thayanukul et al. 2013a), depending on different treatment configurations. Another study in the United States reported that membrane bioreactors, activated sludge, sequencing batch reactors and rotating biological contactor treatments produced reclaimed water with different AOC concentrations (Weinrich et al. 2010). However, it is likely that the BOM composition of reclaimed water is different from that of drinking water. Therefore, it is possible that the bacterial strains originated from reclaimed water can utilise more kinds of organic compounds in wastewater than those used for AOC assay. The conventional AOC analysis might underestimate the actual content of BOM in reclaimed water. In fact, Zhao et al. (2013) demonstrated that some isolates obtained from reclaimed water (Stenotrophomonas sp. ZJ2, P. saponiphila G3 and Enterobacter sp. G6) produced higher acetate equivalent AOC concentrations than P-17 and NOX, suggesting that some BOM in reclaimed water were not available for P-17 and NOX.

Very few studies have been reported on the qualitative aspects of BOM. One possible approach is the chemical analysis before and after biological degradation. For instance, excitation-emission fluorescence spectroscopy was used to identify the degradation of protein-like and humic-like organic substances in wetland soil (Fellman et al. 2008). Another approach is to evaluate organic matter consumed by bacteria with different growth substrates. As the bacterial strains used for AOC analysis have different substrate utilisation spectra, it is possible to discuss the composition of the assimilable part of organic matter by their growth. One example is referred to by Lai et al. (2006) that investigated the effect of ozone treatment and filtration on AOC composition with strain P-17 and NOX. It would be possible to extend the resolution of the analysis by employing a higher number of bacterial strains with different substrate utilisation spectra.

In order to obtain the bacterial library used to characterise BOM quality for reclaimed water, we isolated nearly 200 bacterial strains from samples taken at seven water reclamation plants. The substrate utilisation patterns of those isolates were evaluated using Biolog, a commercial 96-well plate containing different single carbon sources in different wells (Thayanukul et al. 2013b). These indigenous strains can potentially cause regrowth by assimilating BOM in reclaimed water.

In the present study, we aimed to establish a novel methodology to characterise BOM composition in reclaimed water. First, substrate utilisation of the selected isolates was re-evaluated at a substrate concentration of 0.1 mg-C l−1, which is close to the organic matter concentration in reclaimed water. Second, we established a new method, called the bacterial growth fingerprint (BGF) assay to characterise BOM based on the growth patterns of the selected isolates. Finally, the proposed BGF assay, was applied to evaluate reclaimed water samples from six different water reclamation plants. In addition, changes in BOM during the water reclamation process were investigated using the BGF assay.

MATERIALS AND METHODS

Carbon-free material preparation

All glassware used to evaluate bacterial growth in the present study was prepared according to the Standard Method 9217B (APHA 2005) with some modifications. Carbon-free glassware was processed by washing with detergent, rinsing with 0.1 N HCl (w/v) and ultra-pure water and baking at 550 °C for 6 h. Plastic and rubber materials were soaked in 0.1 N HCl (w/v) overnight and rinsed several times with ultra-pure water.

Test microorganisms

The test bacteria were isolated from July to November 2010, from seven water reclamation plants and a reclaimed water distribution system in Japan (Thayanukul et al. 2013b), using R2A agar (BD, USA). We selected the test strains based on variations in the 16S rRNA sequences, the growth patterns in reclaimed water samples and the substrate utilisation in Biolog GN2 microplate tests (Biolog, USA; Supplementary material Figure S1, available with the online version of this paper). Finally, nine strains with diverse substrate utilisation characteristics were selected, as listed in Table 1. These strains have been deposited at the Biological Resource Center, National Institute of Technology and Evaluation, Japan, with the following deposition numbers: NBRC110197, NBRC110198, NBRC110199, NBRC110209, NBRC110210, NBRC110147, NBRC110148, NBRC110211 and NBRC 110780, for strains BGF-1 to BGF-9, respectively. In addition, P. fluorescens strain P-17 (ATCC 49642) and Aquaspirillum sp. strain NOX (ATCC 49643), the reference bacteria in the conventional AOC assay, were included for comparison. The stock inocula were prepared following the AOC protocol by APHA (2005) with some modification, by suspending each colony in sterile buffered water (final concentration: KH2PO4, 85 mg l−1; MgCl2·6H2O, 405 mg l−1; pH 7.2, tap water) with the addition of a mineral solution (KNO3, 0.36 mg l−1; NH4Cl, 0.19 mg l−1; K2HPO4, 0.04 mg l−1; for promoting consumption of biodegradable organic compounds) and sodium thiosulfate (25 mg l−1; for neutralising residual chlorine) around 10 days prior to the inoculation to eliminate any remaining organic matter. Cell enumeration was checked with a flow cytometer (Accuri C6; BD, USA) by staining cells with SYBR green I (Invitrogen, USA) at 37 °C for 10 min (Prest et al. 2013). The efficiency of the flow cytometer was validated regularly using the standard six and eight beads; following the manufacturer's instructions. The 0.2 μm filtered sample was always included for discriminating background particles.

Table 1

Description of bacterial isolates from several water reclamation plants

Isolate nameaClosest relative strains [Accession number]IdentitybIsolation source
 Alphaproteobacteria 
BGF-1 Sphingomonas yabuuchiae strain A1-18 [NR028634] 1434/1436 (99%) Sand filtration effluent 
 Betaproteobacteria 
BGF-2 Acidovorax sp. clone KWE55-19 [JQ670734] 1488/1491 (99%) Reclaimed water (toilet flushing) 
BGF-3 Herminiimonas saxobsidens strain: NS11 [NR042610] 1474/1476 (99%) Ozonation effluent 
 Gammaproteobacteria 
BGF-4 Pseudomonas sp. F(2012b) [JQ766121] 1498/1500 (99%) Secondary settling tank 
BGF-5 Nevskia ramosa [AJ001011] 1457/1479 (99%) Reclaimed water (no chlorination) 
 Actinobacteria 
BGF-6 Mycobacterium sp. JLS [CP000580] 1451/1486 (98%) Secondary settling tank 
BGF-7 Microbacterium sp. U13 [JN000344] 1473/1478 (99%) Ozonation effluent 
 Bacteroidetes 
BGF-8 Pedobacter insulae strain DS-39 [NR044083] 1425/1458 (98%) Secondary settling tank 
BGF-9 Chryseobacterium hominis NF802 [NR042517] 1402/1460 (96%) Secondary settling tank 
Isolate nameaClosest relative strains [Accession number]IdentitybIsolation source
 Alphaproteobacteria 
BGF-1 Sphingomonas yabuuchiae strain A1-18 [NR028634] 1434/1436 (99%) Sand filtration effluent 
 Betaproteobacteria 
BGF-2 Acidovorax sp. clone KWE55-19 [JQ670734] 1488/1491 (99%) Reclaimed water (toilet flushing) 
BGF-3 Herminiimonas saxobsidens strain: NS11 [NR042610] 1474/1476 (99%) Ozonation effluent 
 Gammaproteobacteria 
BGF-4 Pseudomonas sp. F(2012b) [JQ766121] 1498/1500 (99%) Secondary settling tank 
BGF-5 Nevskia ramosa [AJ001011] 1457/1479 (99%) Reclaimed water (no chlorination) 
 Actinobacteria 
BGF-6 Mycobacterium sp. JLS [CP000580] 1451/1486 (98%) Secondary settling tank 
BGF-7 Microbacterium sp. U13 [JN000344] 1473/1478 (99%) Ozonation effluent 
 Bacteroidetes 
BGF-8 Pedobacter insulae strain DS-39 [NR044083] 1425/1458 (98%) Secondary settling tank 
BGF-9 Chryseobacterium hominis NF802 [NR042517] 1402/1460 (96%) Secondary settling tank 

aIsolates BGF-1 to BGF-9 correspond to Iso-85, Iso-33, Iso-43, Iso-13, Iso-59, Iso-37, Iso-44, Iso-18 and Iso-52, respectively, in Thayanukul et al. (2013b).

bDNA sequence identity of the 16S rRNA gene.

Bacterial growth test with individual carbon substrates

Basal salt medium was prepared following the method described by van der Kooij & Veenendal (1995), with a slight modification. The medium contained KH2PO4, 2.7 mg; K2HPO4, 4.1 mg; Na2HPO4, 6.5 mg; CaCl2·2H2O, 50 mg; CoCl2·6H2O, 50 μg; H3BO3, 4 μg; MgSO4·7H2O, 50 mg; CaSO4, 65 μg; ZnSO4·7H2O, 0.1 mg; FeSO4·7H2O, 10 mg and NH4Cl, 0.764 mg, in 1 l of ultra-pure water. The medium was filtered through a baked 0.3 μm GF-75 glass filter (Advantec, Japan) after rinsing with 500 ml of ultra-pure water. The filtered medium was dispensed (15 ml) into 30 ml borosilicate tubes. A total of 25 carbon sources, including seven amino acids, three aromatics, seven carboxylic acids and eight carbohydrates (Table 2), were supplied to tubes at a concentration of 100 μg-C l−1. Pasteurisation was performed using a water bath at 70 °C for 30 min. Bacterial isolates were inoculated separately in duplicate at an initial concentration of 1,000 cells ml−1. A negative control (without inoculum) and a minimal growth control (without carbon source) were included. The cultures were incubated at 20 °C in darkness without shaking. The total cell concentrations were monitored with a flow cytometer.

Table 2

Utilisation of individual carbon substrates by BGF strains at a concentration of 100 μg-C l−1, determined using synthetic media

 Isolate
SubstrateBGF-1BGF-2BGF-3BGF-4BGF-5BGF-6BGF-7BGF-8BGF-9
Amino acid 
dl-Phenylalanine ++ − ++ − − ++ 
l-Arginine − − ++ − − − − 
l-Aspartate − − − − − − NA 
l-α-Alanine − − − − − ++ 
 Methionine − ++ − − ++ − − 
l-Proline − − − − − NA ++ 
l-Asparagine − − − − − 
Aromatic          
 Benzoate − − − − − − − 
 Salicylate ++ − − − − − − − 
 Phthalate − ++ − − − − − 
Carboxylic acid 
 Acetate ++ − − − ++ 
 Oxalate ++ − − − − − − 
 Formate ++ − − ++ − − − 
 Citrate − − − − − − − 
 Succinate ++ − ++ ++ − − 
 Propionate ++ − ++ ++ − − − ++ 
l(+)-Tartarate ++ − − − − − 
Carbohydrate 
 α-l(+)-Rhamnose ++ − − − − − 
d-Glucose − ++ − − − 
d-Xylose − − ++ − ++ − 
 Cellobiose − ++ ++ − − ++ 
 Arabinose − ++ − − − NA 
 Sucrose ++ − ++ − − − − − 
 Glycerol − − − − − 
d-Mannitol − − ++ − − − − 
 Isolate
SubstrateBGF-1BGF-2BGF-3BGF-4BGF-5BGF-6BGF-7BGF-8BGF-9
Amino acid 
dl-Phenylalanine ++ − ++ − − ++ 
l-Arginine − − ++ − − − − 
l-Aspartate − − − − − − NA 
l-α-Alanine − − − − − ++ 
 Methionine − ++ − − ++ − − 
l-Proline − − − − − NA ++ 
l-Asparagine − − − − − 
Aromatic          
 Benzoate − − − − − − − 
 Salicylate ++ − − − − − − − 
 Phthalate − ++ − − − − − 
Carboxylic acid 
 Acetate ++ − − − ++ 
 Oxalate ++ − − − − − − 
 Formate ++ − − ++ − − − 
 Citrate − − − − − − − 
 Succinate ++ − ++ ++ − − 
 Propionate ++ − ++ ++ − − − ++ 
l(+)-Tartarate ++ − − − − − 
Carbohydrate 
 α-l(+)-Rhamnose ++ − − − − − 
d-Glucose − ++ − − − 
d-Xylose − − ++ − ++ − 
 Cellobiose − ++ ++ − − ++ 
 Arabinose − ++ − − − NA 
 Sucrose ++ − ++ − − − − − 
 Glycerol − − − − − 
d-Mannitol − − ++ − − − − 

Growth level was categorised by the ratio of cell numbers between test samples (Ntest) and control samples without carbon source (Ncont) after the incubation: r = Ntest/Ncont. −: r < 3, +: 3≤ r < 10, ++ : r ≥ 10. NA: Data not available.

Reclaimed water samples

Batch water samples (2l) were collected from water reclamation plants A–F, as illustrated in Figure 1. Water samples were collected from February to October 2012. All plants received secondary effluents from municipal wastewater treatment facilities using activated sludge with Pseudo AO (anaerobic/oxic) and conventional configuration. The general water quality of the samples is shown in Supplementary material Table S1 (available with the online version of this paper). Dissolved organic carbon (DOC) and total dissolved nitrogen (TDN) were measured with a TOC-VCSH analyser (Shimadzu, Japan) and a TNM-1 analyser (Shimadzu, Japan), respectively. All water samples were filtered through 0.2 μm PTFE filters (JGWP; Millipore, USA). The limit of quantification of DOC was approximately 500 μg-C l−1. Ultraviolet absorbance was measured with a UV/VIS U-2000 Spectrophotometer (Hitachi, Japan) at a wavelength of 254nm. For specific UV absorbance (SUVA, lm−1 mg-C−1), the adsorption value was divided by the DOC concentration.
Figure 1

Schematic representation of the treatment processes of water reclamation plants A–F. ○: Treatment unit installed; no mark: no treatment unit installed.

Figure 1

Schematic representation of the treatment processes of water reclamation plants A–F. ○: Treatment unit installed; no mark: no treatment unit installed.

The BGF assay

Water samples were placed on ice and transferred to a laboratory within 4 h. The samples were then filtered through 0.2 μm PTFE filters (JGWP; Millipore, USA), which were soaked in ultra-pure water overnight and rinsed again with approximately 500 ml of ultra-pure water. The filtered reclaimed water was dispensed (15 ml) into 30 ml borosilicate tubes and supplied with mineral solution and sodium thiosulfate in the same manner as for the preparation of stock inoculum. Pasteurisation was performed at 70 °C for 30 min, and bacterial isolates were separately inoculated at an initial concentration of 1,000 cells ml−1. Samples from Plants C and D were tested in triplicate as indicated in data with standard error values. A negative control (without inoculum) was included. Samples were incubated at 20 °C in darkness for 10 days until most of the strains reached a stationary phase. Cell concentrations were evaluated using a flow cytometer.

RESULTS AND DISCUSSION

Substrate utilisation of the selected isolates

We selected nine strains with different substrate utilisation patterns illustrated by Biolog assay and different growth responses in reclaimed water samples (Thayanukul et al. (2013b), and Supplementary material Figure S1). However, Biolog employs a very high organic matter concentration (2 g l−1), much higher than the organic matter concentration in reclaimed water. The selected bacterial strains may have different responses to substrates at lower concentrations of organic matter. Therefore, we re-evaluated substrate utilisation at a concentration of 100 μg-C l−1, which is closer to the concentration of reclaimed water. Organic substrates such as amino acids, aromatics, carboxylic acid and carbohydrate groups were used as representative compounds for BOM based on the difference in chemical structures. The results of substrate consumption measured by cell growth are presented in Table 2 and Supplementary material Table S2 (available with the online version of this paper).

BGF-1 and BGF-5 used almost all the organic compounds tested, but neither grew on d-xylose or d-mannitol. BGF-1 could not use asparagine, benzoate or citrate, whereas BGF-5 could not use methionine or phthalate. BGF-3, BGF-4 and BGF-6 used approximately half of the organic matter tested. BGF-2, BGF-7 and BGF-9 used only a few of the compounds, and BGF-8 did not use any. The most used substrates were acetate, succinate, phenylalanine and cellobiose as they were consumed by six strains, whereas the least used substrates were aspartate, benzoate, salicylate and citrate as they were consumed by only two strains.

Interestingly, we found inconsistent results between the bacterial growth test in this study and the previous Biolog test. For instance, strain BGF-3 could use several carbohydrates in the substrate utilisation test but it did not use any carbohydrates in the Biolog test. The substantial differences in substrate concentrations may affect the metabolic activity of the test strains (Thayanukul et al. 2010). This might also happen when introducing a bacterial isolate into a new water sample with different BOM concentration.

The BGF of reclaimed water from different plants

The nine BGF strains were independently inoculated into sterilised reclaimed water samples collected from six water reclamation plants. The maximum cell numbers were recorded (Figure 2). The cell growth level, the pattern of which was defined as the BGF, was very different among samples. BGF-3 gave the highest cell numbers in samples from Plants A, B and E. The cell concentrations of BGF-3 were 4.2 × 106, 8.7 × 106, and 1.2 × 106 cells ml−1, respectively, and these concentrations indicated differences in the concentration of BOM that the strain can use in the water tested. The reclaimed water from Plant C was distinctly characterised by the growth of strains BGF-3 and BGF-5 at similar levels. BGF-1 and BGF-7 demonstrated the highest growth for Plants D and F, respectively. The cell concentration of BGF-1 and BGF-7 was much smaller in the other samples, and thus the BOM that these strains can use was prominent only in those samples.
Figure 2

The BGF of reclaimed water from six treatment plants. Error bars represent the standard error (S.E.) for triplicate samples (Plant C).

Figure 2

The BGF of reclaimed water from six treatment plants. Error bars represent the standard error (S.E.) for triplicate samples (Plant C).

The BGF was very different among samples. This indicates large differences in BOM composition in the reclaimed water samples. Such differences could be due to different water reclamation treatment processes (Figure 1), or to original differences in secondary effluent. It is noteworthy that the BOM compositions in reclaimed water were different among plants A–D even though they used similar treatment processes.

Quantitative evaluation of BOM composition

Growth yield for each BGF strain must be obtained to quantitatively compare BOM used by different strains. We selected acetate to calculate the growth yield of six strains that can use acetate; because acetate is one of the most commonly used substrates tested in this study and is also used for the AOC assay (Table 3). The growth yields for BGF-1, 3, 4, 5, 6 and 9 were in the range of 3.4 × 106 to 2.2 × 107 cells (μg acetate-C)−1, which were equivalent to the growth yields of P-17 (4.1 × 106 CFU (μg acetate-C)−1) and NOX (1.2 × 107 CFU (μg acetate-C)−1). With the yield values obtained, the BOM concentration for each strain was calculated for the reclaimed water samples (Figure 3). In addition, DOC, TDN concentration and SUVA for all reclaimed water samples were measured to compare with the BOM concentration (Supplementary material Table S1).
Table 3

Growth yield of BGF strains on acetate

BGF StrainBGF-1BGF-3BGF-4BGF-5BGF-6BGF-9
Yield (cells (μg acetate-C)−1)) 3.4 × 106 1.0 × 107 1.6 × 107 6.1 × 106 1.3 × 107 2.2 × 107 
BGF StrainBGF-1BGF-3BGF-4BGF-5BGF-6BGF-9
Yield (cells (μg acetate-C)−1)) 3.4 × 106 1.0 × 107 1.6 × 107 6.1 × 106 1.3 × 107 2.2 × 107 
Figure 3

Quantitative profiles of BOM assessed by BGF strains and reference bacteria of the AOC assay in reclaimed water from six treatment plants. Error bars represent the standard error (S.E.) for triplicate samples (Plant C).

Figure 3

Quantitative profiles of BOM assessed by BGF strains and reference bacteria of the AOC assay in reclaimed water from six treatment plants. Error bars represent the standard error (S.E.) for triplicate samples (Plant C).

The maximum BOM concentrations in Plants A–F were 439 μg acetate-C l−1 by strain P-17, 933 μg acetate-C l−1 by strain NOX, 1,301 μg acetate-C l−1 by strain P-17, 2,777 μg acetate-C l−1 by strain BGF-1, 122 μg acetate-C l−1 by strain NOX, and 1,849 μg acetate-C l−1 by strain BGF-1, respectively. Plant E contained the lowest maximum BOM concentration because DOC concentration in Plant E was as low as 0.8 mg-C l−1 (Table S1). However, no strong correlation was observed between DOC and maximum BOM concentration for the other samples.

At least one BGF strain determined BOM in reclaimed water comparable to or better than P-17 and NOX, except in Plant C. In the samples from Plants D and F, the BOM concentration determined by P-17 was higher than that determined by NOX, and BGF-1 demonstrated even higher BOM consumption. In reclaimed water from plants A, B and E, BOM concentrations determined by strain P-17 and NOX were at the same level, which was similar to the concentration determined by BGF-3. These results suggested that BGF can be the better method to cover BOM in reclaimed water than AOC method.

Changes in the BGF in a reclaimed water treatment plant

The fate of BOM during the water treatment process in Plant D was evaluated using six BGF strains that use acetate and two reference AOC strains (Figure 4 and Supplementary material Figure S2, available with the online version of this paper). Samples were collected between the secondary effluent and the point after microfiltration. The finished water after chlorination was not sampled because it was mixed with water from another treatment line (Figure 1).
Figure 4

Changes in BOM concentration in water samples collected during the treatment process D. 2ef: Secondary effluent. Bf: Sand filtration. Oz: Ozonation. Co: Coagulation. Mf: Microfiltration. All the tests were done in triplicate: error bars represent the standard error (S.E.).

Figure 4

Changes in BOM concentration in water samples collected during the treatment process D. 2ef: Secondary effluent. Bf: Sand filtration. Oz: Ozonation. Co: Coagulation. Mf: Microfiltration. All the tests were done in triplicate: error bars represent the standard error (S.E.).

The DOC concentration gradually decreased during the treatment process with a slight increase after biofiltration (Table S1). The increase in DOC may result from the release of residual organic compounds from the biofilter. However, the BGF results demonstrated that BOM concentrations measured by BGF strains decreased or remained unchanged after biofiltration. Biofiltration could reduce the portion of BOM available for strains BGF-5 and BGF-6 by 537 μg acetate-C l−1 and 152 μg acetate-C l−1, respectively, possibly due to biodegradation by the microorganisms on the biofilter. In contrast, BOM that could be used by P-17 increased by 268 μg acetate-C l−1. The metabolites of the microbial community from the filter could be evaluated by P-17. For the other indigenous strains, BOM concentrations were almost stable before and after biofiltration, suggesting that their substrates were neither removed nor produced by the process.

The BOM concentrations evaluated by BGF strains, except for BGF-6 and BGF-9, increased after ozonation, by 345 μg acetate-C l−1 on average of all BGF strains or by 1 to 42 folds. The SUVA, indicating the presence of unsaturated moieties, declined after ozonation. Therefore, the partial degradation or transformation of complex organic molecules into a bioavailable fraction accounted for the BOM increase. The increase in BOM and the decrease of SUVA are consistent with previous studies (Hammes et al. 2006; Thayanukul et al. 2013a). In contrast, the BOM identified by BGF-6 in the sample after biofiltration (60 μg acetate-C l−1) was almost completely removed by ozonation. The strain could utilise the portion of BOM that was removed by ozonation. Further investigation, including comprehensive chemical analysis, is necessary to understand the impacts of ozonation on DOM.

Coagulation could remove BOM by 209 μg acetate-C l−1 and 175 μg acetate-C l−1 for BGF-3 and NOX, while 102 μg acetate-C l−1 was produced for BGF-6. No large difference was observed for the other strains. The difference in BOM profiles between ozonated water and coagulation was basically small. This result was in contrast to the findings by Zhao et al. (2014), who observed a 55–667% increase in AOC concentration after the coagulation treatment of secondary effluent. They attributed this increase to the removal of growth inhibitors. Existence of pre-treatment steps and different DOC levels in the secondary effluent (3 mg l−1 in the present study vs. 8–13 mg l−1 in Zhao et al. (2014)) should be considered in order to understand the difference in findings.

After microfiltration, the BOM was comparable to that for coagulated water. BOM evaluated by BGF-6 and P-17 decreased by 103 and 365 μg-acetate C l−1, whereas BOM evaluated by BGF-5 and NOX increased by 280 and 155 μg-acetate C l−1, respectively. It is unlikely that the membrane physically removed DOM since the 0.1 μm pore size is relatively large. Further analysis is required to interpret the behaviours of BOM in microfiltration. Membrane-attached biofilm could remove some part of the BOM and may also release metabolites which can be assimilated by the microorganisms there.

This study clearly indicated that each treatment step has different influences on the BOM determined by different strains. One treatment condition may be effective to reduce the substrate for certain bacteria, but the condition may promote the regrowth of other bacteria. In order to control bacterial regrowth, it is essential to understand how treatment processes change those BOM. BGF assays can be a useful tool to evaluate the change.

CONCLUSIONS

We established a novel method to characterise BOM composition in reclaimed water using nine bacterial strains (BGF strains) with different substrate utilising spectra. BGF strains can be used to determine BOM composition in reclaimed water from different sources, and to monitor changes in BOM during the water reclamation process.

The more specific conclusions are as follows:

  • All selected BGF strains have unique substrate utilisation spectra under organic matter concentrations similar to those in reclaimed water.

  • BGF strains demonstrate that BOM composition of reclaimed water treated by different treatment processes varies considerably.

  • BOM composition can be quantitatively characterised using six BGF strains that are capable of utilising acetate. The six selected BGF strains can describe BOM equal to or better than P17 and NOX.

ACKNOWLEDGEMENTS

This study was conducted as part of a Core Research for Evolutional Science and Technology (CREST) project supported by the Japan Science and Technology Agency, entitled ‘Development of well-balanced Urban Water Use System adapted to climate change’. The authors would like to thank the water reclamation plants and personnel for their participation and co-operation.

REFERENCES

REFERENCES
APHA/AWWA/WEF
2005
Standard Methods for the Examination of Water and Wastewater
21st edn.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
Fellman
J. B.
D'Amore
D. V.
Hood
E.
Boone
R. D.
2008
Fluorescence characteristics and biodegradability of dissolved organic matter in forest and wetland soils from coastal temperate watersheds in southeast Alaska
.
Biogeochemistry
88
,
169
184
.
Jjemba
P. K.
Weinrich
L. A.
Cheng
W.
Giraldo
E.
LeChevallier
M. W.
2010
Regrowth of potential opportunistic pathogens and algae in reclaimed-water distribution systems
.
Applied and Environmental Microbiology
76
(
13
),
4169
4178
.
Narasimhan
R.
Brereton
J.
Abbaszadegan
M.
Ryu
H.
Butterfield
P.
Thompson
K.
Werth
H.
2005
Characterizing Microbial Water Quality in Reclaimed Water Distribution Systems
.
AWWA Research Foundation
,
Denver
.
Prest
E. I.
Hammes
F.
Kötzsch
S.
van Loosdrecht
M. C. M.
Vrouwenvelder
J. S.
2013
Monitoring microbiological changes in drinking water systems using a fast and reproducible flow cytometric method
.
Water Research
47
,
7131
7142
.
van der Kooij
D.
Hijnen
W. A. M.
1984
Substrate utilization by an oxalate-consuming spirillum species in relation to its growth in ozonated water
.
Applied and Environmental Microbiology
47
(
3
),
551
559
.
van der Kooij
D.
Veenendal
H. R.
1995
Determination of the Concentration of Easily Assimilable Organic Carbon (AOC) in Drinking Water with Growth Measurements using Pure Bacterial Cultures
.
The AOC manual SWE 95.022
,
KIWA
,
Nieuwegein
,
The Netherlands
.
van der Kooij
D.
Oranje
J. P.
Hijnen
W. A.
1982a
Growth of Pseudomonas aeruginosa in tap water in relation to utilization of substrates at concentrations of a few micrograms per liter
.
Applied Environmental Microbiology
44
(
5
),
1086
1095
.
van der Kooij
D.
Visser
A.
Hijnen
W. A. M.
1982b
Determining the concentration of easily assimilable organic carbon in drinking water
.
Journal of the American Water Works Association
74
(
10
),
540
545
.
Weinrich
L. A.
Jjemba
P. K.
Giraldo
E.
LeChevallier
M. W.
2010
Implications of organic carbon in the deterioration of water quality in reclaimed water distribution systems
.
Water Research
44
(
18
),
5367
5375
.
Zan
C.
Shi
L.
Yang
W.
Ma
X.
2010
Evolution of composite fouling on a vertical stainless steel surface caused by treated sewage
.
Frontiers of Energy and Power Engineering in China
4
(
2
),
171
180
.
Zhao
X.
Hu
H.
Liu
S.
Jiang
F.
Shi
X.
Li
M.
Xu
X.
2013
Improvement of the assimilable organic carbon (AOC) analytical method for reclaimed water
.
Frontiers of Environmental Science and Engineering
7
(
4
),
483
491
.

Supplementary data