Relative Decay of Fecal Indicator Bacteria and ... - ACS Publications

Article pubs.acs.org/est

Relative Decay of Fecal Indicator Bacteria and Human-Associated Markers: A Microcosm Study Simulating Wastewater Input into Seawater and Freshwater L. Jeanneau,†,* O. Solecki,‡ N. Wéry,§ E. Jardé,† M. Gourmelon,∥ P. -Y. Communal,⊥ A. Jadas-Hécart,⊥ M. -P. Caprais,∥ G. Gruau,† and A. -M. Pourcher‡ †

CNRS, UMR 6118 Geosciences Rennes, Rennes, France CEMAGREF, Rennes, France § INRA, UR0050, Laboratoire de Biotechnologie de l’Environnement, Narbonne, France ∥ IFREMER, EMP, Laboratoire de Microbiologie, Plouzané, France ⊥ Université d’Angers, Angers, France ‡

S Supporting Information *

ABSTRACT: Fecal contaminations of inland and coastal waters induce risks to human health and economic losses. To improve water management, specific markers have been developed to differentiate between sources of contamination. This study investigates the relative decay of fecal indicator bacteria (FIB, Escherichia coli and enterococci) and six human-associated markers (two bacterial markers: Bacteroidales HF183 (HF183) and Bif idobacterium adolescentis (BifAd); one viral marker: genogroup II F-specific RNA bacteriophages (FRNAPH II); three chemical markers: caffeine and two fecal stanol ratios) in freshwater and seawater microcosms seeded with human wastewater. These experiments were performed in darkness, at 20 °C and under aerobic conditions. The modeling of the decay curves allows us (i) to compare FIB and markers and (ii) to classify markers according to their persistence in seawater (FRNAPH II < HF183, stanol ratios < BifAd, caffeine) and in freshwater (HF183, stanol ratios < FRNAPH II < BifAd < caffeine). Although those results depend on the experimental conditions, this study represents a necessary step to develop and validate an interdisciplinary toolbox for the investigation of the sources of fecal contaminations.

1. INTRODUCTION Human wastewaters are significant contributors to the degradation of the quality of inland and coastal waters. Such contaminations often threaten human health due to the presence of pathogens and may induce economic losses due to the closure of bathing areas and/or of shellfish harvesting areas.1,2 In the European Water Framework Directive (2000/ 60/EC), the directive on bathing waters (2006/7/EC) sets more stringent regulatory values for fecal indicator bacteria (FIB) and requires the acquisition of bathing water profiles as a tool to assess pollution risks. Moreover, the directive on shellfish harvesting (2006/113/EC) requires the assessment of potential pollution sources upstream of shellfish harvesting areas. These requirements imply the identification and the prioritization of the potential sources of fecal contamination. However FIB (namely Escherichia coli and enterococci) cannot distinguish between the potential sources of contamination.3 As a consequence, research efforts are needed to develop and validate source-specific markers that will allow discriminating sources of fecal contaminations indicated by FIB.4,5 To ideal host-associated marker should fulfill several requirements: (i) to be source specific, (ii) to occur in high © 2012 American Chemical Society

concentration in polluting matrices, (iii) to exhibit extraintestinal persistence similar to FIB, (iv) not to grow outside the host and (v) to exhibit similar behavior as FIB during the transfer of fecal pollution through the drainage network.4 For the third and the fourth requirements, potential markers are compared with FIB and not with pathogens since regulatory values are set for FIB. However, no single marker has so far fulfilled all these requirements. As a consequence, the use of a combination of several markers has been proposed to generate more reliable data.4,6 This has led to the development of a microbial source tracking (MST) toolbox including FIB and microbial and chemical markers to differentiate between human, bovine and porcine fecal contaminations.7 Among 17 specific markers investigated in this previous study, seven hostassociated markers (caffeine, tri(2-chloroethyl)phosphate (TCEP), benzophenone and the bacterial markers HF183, Pig-2-Bac, Rum-2-Bac and Lactobacillus amylovorus) and two Received: Revised: Accepted: Published: 2375

August 29, 2011 January 9, 2012 January 10, 2012 January 10, 2012 dx.doi.org/10.1021/es203019y | Environ. Sci. Technol. 2012, 46, 2375−2382

Environmental Science & Technology

Article

Table 1. Parameters Used for Modelling the Decay of FIB and Specific Markers in Water Microcosms Inoculated with WWTP influent (1:18 dilution). T90 is calculated for FIB and human specific bacterial and viral markers, and T1/2 is calculated for specific chemical markers. The modelling parameters, T90 and T1/2, are reported as mean values for three microcosms, and the uncertainties as standard errors. Tf is the final sampling date for FIB and specific markers k or k1b (day−1)

fa Seawater cEC cENT HF183 BifAd FRNAPH II caffeine Cop Etcop Sito Freshwater cEC cENT HF183 BifAd FRNAPH II caffeine Cop Etcop Sito

0.99 ± 0.01

0.82 ± 0.11 0.67 ± 0.04 0.55 ± 0.05 0.99 ± 0.00 0.99 ± 0.00 0.99 ± 0.00 0.66 0.96 0.90 0.60

± ± ± ±

0.16 0.01 0.01 0.02

1.32 ± 0.08 0.72 ± 0.16 1.03 ± 0.08 0.63 ± 0.01 2.28 ± 0.30 0.24d 0.08 ± 0.01 0.06 ± 0.02 0.07 ± 0.03 0.41 ± 0.02 0.79 ± 0.12 1.39 ± 0.02 0.66 ± 0.04 0.52 ± 0.16 0.17 ± 0.01 0.19 ± 0.00 0.18 ± 0.00 0.26 ± 0.04

k2a (day−1)

0.17 ± 0.04

0.06 ± 0.05 0.04 ± 0.01 0.02 ± 0.00 0.11 ± 0.00 0.13 ± 0.01 0.06 ± 0.02 0.07 0.03 0.03 0.02

± ± ± ±

0.04 0.00 0.00 0.00

R2

T90 or T1/2e (days)

Tf (days) 13 48 6 13 2 13 55 55 55

± ± ± ± ± ± ± ± ±

55 55 6 20 13 27 55 55 55

0.90 ± 0.97 ± 0.78 ± 0.75 ± −c 0.85d 0.86 ± 0.76 ± 0.79 ±

0.02 0.01 0.03 0.13

0.01 0.03 0.05

1.7 ± 0.1 3.6 ± 0.8 2.3 ± 0.2 3.7 ± 0.1 1.1 ± 0.1 2.9d 8.3 ± 0.9 15.5 ± 3.2 20.9 ± 4.6

± ± ± ± ± ± ± ± ±

0.00 0.00 0.01 0.01 0.02 0.01 0.00 0.00 0.03

5.8 3.1 1.7 3.6 5.2 4.9 3.8 4.4 5.4

0.96 0.99 0.90 0.92 0.96 0.94 0.96 0.94 0.87

0.2 0.5 0.0 0.2 1.3 0.5 0.1 0.1 0.9

a

The proportion of C0 consumed during the first phase of the decay ( f) and the decay constants k1 and k2 are calculated using the biphasic first-order model. f and k2 are not available when the data are modeled by the monophasic first-order decay model. bDecay constants in bold are calculated using the monophasic first-order decay model (k) and decay constants in normal type are calculated using the biphasic first-order decay model (k1). c 2 R for FRNAPH II in SW is not calculated since data are only available for day 0 and day 2. dStandard error is not calculated since the modeling is performed using the mean concentration. eThe values in italics are T1/2 calculated for chemical markers and the values in normal type are T90 calculated for FIB and microbial markers.

the evolution of a fecal contamination by a WWTP influent in a finite volume of water transferred from upstream to downstream by advection. Due to these simplifying assumptions, the provided data must be considered in those experimental conditions.

fecal stanol ratios (R1: coprostanol/(coprostanol+24-ethylcoprostanol) and R2: sitostanol/coprostanol) were found to be the most efficient for discriminating the origin of fecal pollutions.7 Moreover, Bif idobacterium adolescentis (BifAd) and F-specific RNA bacteriophages from the subgroup II (FRNAPH II) were slightly less discriminating, but were considered as useful complementary markers.7 One of the main conclusions from the development of the above toolbox was the need to study the persistence of these specific markers in comparison with the FIB. This information is an absolute prerequisite before applying the MST toolbox to access the quality of superficial waters. The present study was specifically undertaken to address this problem. One bacterial marker was selected among the Bacteroidales, namely HF183,8,9 and another among Bifidobacterium, namely Bif idobacterium adolescentis,10 two bacterial groups characterized by their host-specificity.11,12 Two genogroups (II and III) of FRNAPH were investigated as viral specific markers.13 TCEP, benzophenone, and caffeine were also investigated as indirect chemical markers.14 Finally, the stanol ratios were used as direct chemical markers to differentiate between effluents of wastewater treatment plant (WWTP) (R1 > 0.73; R2 = 0.10 ± 0.01) and animal feces.7,15,16 In this study, the persistence of FIB and human-associated markers was investigated in seawater (SW) and freshwater (FW) microcosms seeded with a WWTP influent for 55 days. These conditions were chosen (i) to determine their decay in both of the environmental matrices investigated by the MST toolbox and (ii) to compare the decay of FIB and humanassociated markers. Such an experimentation allows to model

2. MATERIALS AND METHODS 2.1. Microcosms Design. The experimental design is described in a previous paper.17 Untreated WWTP influent (5 L) was diluted in 90 L of unfiltered water (1:18 dilution), which corresponds to a low hydraulic impact of WWTP effluent on a receiving body.18 The microcosms were maintained under aerobic conditions (8.9 ± 0.1 mg/L dissolved O2) with constant mixing and at constant temperature (18.5 ± 0.2 °C). These experiments were conducted in darkness to avoid heterogeneous lighting due to the turbidity of the system. The decay of FIB and of specific markers was investigated in two types of recreational waters (marine and fresh water). Three microcosms were filled with seawater (SW) and three with freshwater (FW). Culturable Escherichia coli (cEC), culturable enterococci (cENT), a phylotype related to Bifidobacterium adolescentis (BifAd), the human-associated (HF183) Bacteroidales 16S rDNA marker, the chemical markers caffeine, coprostanol (cop), 24-ethylcoprostanol (etcop) and sitostanol (sito), as well as F-specific RNA bacteriophages (FRNAPH), and especially the human-associated genotype (FRNAPHII) were quantified in SW, FW and WWTP influent and in microcosms on the starting day and on days 2, 6, 13, 20, 27, 34, 41, 48, and 55. Sampling dates are denoted as Dn in the following, where n is the sample day. 2376

dx.doi.org/10.1021/es203019y | Environ. Sci. Technol. 2012, 46, 2375−2382


Article

Figure 1. Decay curves of fecal indicator bacteria in (A) SW and in (B) FW microcosms and of bacterial and viral human-associated markers in (C) SW and (D) FW. The values on the y axis are expressed as Log10 mean ± standard error (three experimental replicates) CFU/100 mL (cEC and cENT), PFU/100 mL (FRNAPH), and gene copies/100 mL (HF183 and BifAd). The dashed lines on figures A and B indicate the regulatory values (RV) for cEC and cENT in SW (A) and FW (B). The dotted lines in panels C and D indicate the limits of quantification of microbial markers.

2.2. Water and WWTP Influent Samples. The coastal SW sample was collected on 29 March 2010 (the day before the beginning of the experiment) in the Atlantic Ocean off Landunvez (48.54° N, 4.75° W; France). The measured salinity was 33‰, while dissolved organic carbon (DOC) and dissolved nitrogen (DN) concentrations were 0.4 and 4.2 mg/L, respectively. The lake FW was sampled from Commana (48.41° N, 4.01° W; France) on the same day, yielding DOC and DN concentrations of 2.2 and 3.8 mg/L, respectively. The WWTP influent was collected on 29 March 2010 at La Meignanne (47.52° N, 0.67° W; France) and was stored at 4 °C. This WWTP has a capacity of 1800 population-equivalent and returns 160 m3/day (annual mean value) of treated water to the Brionneau River. DOC and DN concentrations in the WWTP influent were 45.9 and 120.0 mg/L, respectively. 2.3. Analytical Methods. The methods used for the enumeration of FIB and for the quantification of humanassociated markers are described in the Supporting Information. Briefly, FIB were enumerated by cultural methods, BifAd and HF183 were quantified by real-time PCR using the methods described by Gourmelon et al.,7 FRNAPH were enumerated following the ISO 10705−1 method and were genotyped, as described in Ogorzaly et al.,19 by real-time PCR. Fecal stanols, namely coprostanol (5β-cholestan-3β-ol), 24ethylcoprostanol (5β-24-ethylcholestan-3β-ol) and sitostanol (5α-24-ethylcholestan-3β-ol) were investigated, according to Jeanneau et al.,20 by gas chromatography−mass spectrometry. Finally, indirect chemical markers, namely tri(2-chloroethyl)phosphate (TCEP), benzophenone and caffeine were investigated, were investigated by gas chromatography ion-trap tandem mass spectrometry. Among these compounds, only caffeine was quantified in the WWTP influent. As a consequence TCEP and benzophenone are not discussed further in this paper. 2.4. Modeling. Two models are used here to (i) approximate decay rates and (ii) calculate the length of time required to obtain a reduction of 90% and 50% of the initial inoculum (T90 and T1/2, respectively). T90 is calculated for FIB

and microbial markers, and T1/2 for chemical markers. The choice between the two models is based on the final sampling date (Tf), which corresponds to the last sample taken for quantification of the indicator or marker present in at least one microcosm (Table 1). For Tf ≤ 13, a first-order decay model (Chick model) is applied and the decay rates, T90 and T1/2 are calculated according to the following equation (eq 1), where C0 is the initial concentration, Ct the concentration at time t and k the decay constant.

⎛C ⎞ Ct = C0 × e−kt ⇔ ln⎜ t ⎟ = − k × t ⎝ C0 ⎠ ⇔ T90 =

ln(0.1) −k

or

T1/2 =

ln(0.5) −k

(1)

This model is applied for cEC, BifAd and caffeine in SW and for HF183 and FRNAPH II in both SW and FW. For Tf > 13 (Table 1), decay rates are calculated using a biphasic first-order decay model.21,22 This value of Tf (13 days) is chosen because a minimum of 5 points is needed to apply the biphasic decay model. Three parameters are calculated using the following equation (eq 2), where f is the proportion of C0 consumed during a first phase of decay, k1 is the decay constant for the first phase and k2 the decay constant for the second phase. The three parameters of this biphasic decay model are calculated using the nonlinear regression modeling of XLSTAT 2010.2. T90 and T1/2 are approximated by an iterative method.

Ct = C0 × (fe−k1t + (1 − f )e−k 2t ) ⎛C ⎞ ⇔ ln⎜ t ⎟ = ln(fe−k1t + (1 − f )e−k 2t ) ⎝ C0 ⎠

(2)

For both models, Ct and C0 are expressed in CFU/100 mL for FIB, in DNA copies/100 mL for HF183 and BifAd, in PFU/ 100 mL for FRNAPH II and in μg/L for chemical tracers. The decay of FIB and markers is thus modeled in each microcosm. 2377



Article

Figure 2. Decay curves of chemical markers in SW (black circles) and in FW (gray squares) microcosms: (A) coprostanol; (B) 24-ethylcoprostanol; (C) sitostanol; (D) caffeine. The values on the y axis are expressed as mean ± standard error μg/L (three experimental replicates).

WWTP influent, the concentrations of cEC and cENT are 8 × 105 and 5 × 106 CFU/100 mL, respectively, and the concentrations of HF183 and BifAd are 4.3 × 107 and 1.5 × 108 copies/100 mL, respectively. The concentration of FRNAPH is 3.6 × 104 PFU/100 mL with 96% of FRNAPH belonging to genogroup II, which is the predominant FRNAPH genogroup in WWTP effluents.23 The concentrations of caffeine, cop, etcop, and sito are 66.3, 98.7, 23.9, and 6.4 μg/ L, respectively.The decay curves are illustrated in Figure 1 (FIB and microbial markers) and Figure 2 (chemical markers). Table 1 reports the modeling parameters, as well as the T90 and T1/2 calculated by the modeling of these curves. 3.1. cEC. The initial concentration in the SW microcosms was 8.0 ± 1.2 × 104 CFU/100 mL (mean value ± standard error). cEC was detected up to D6 in two microcosms and up to D13 in the third. The amount of cEC decreased at the same rate in the three microcosms (0.7 Log10/day) between D0 and D6, and then remained stable until D13 in the microcosm with detectible levels of cEC up to D13. In FW, the initial concentration of cEC was 7.3 ± 1.3 × 104 CFU/100 mL (mean value ± standard error) and cEC was detected up to the end of the experiment (D55). This initial concentration was doubled between D0 and D2, and then decreased by 0.3 Log10/ day between D2 and D13. Then, it remained stable up to D27, decreasing by 0.2 Log10/day between D27 and D34, and finally remained stable until D55 at a mean concentration of 2.0 ± 0.6 CFU/100 mL. T90 was significantly (p < 0.05) higher in FW (5.8 days) than in SW (1.7 days). The modeling of these decay curves allows to determine the time required to reach the regulatory values set by the European legislation (2006/7/EC), in order to classify inland and coastal waters as being of sufficient quality (SQ) (500 and 900 CFU/100 mL for SW and FW, respectively). The regulatory value was attained after 3.6 and 11.3 days in SW and FW microcosms, respectively. 3.2. cENT. The initial concentrations were 4.1 ± 0.9 × 105 and 4.3 ± 0.7 × 105 CFU/100 mL in SW and FW, respectively. In SW, cENT was detected up to D41 in two microcosms and up to D48 in the third. In FW, cENT was detected up to D48 in

2.5. Statistical Analyses. Statistical analyses were performed to test the differences between (i) the T90 of a given target in SW and in FW and (ii) the decay of FIB and specific markers. In a previous study, the T90 of cEC in FW seeded with pig slurry was shown to have a normal distribution.17 From this, we infer that the decay rates of cENT and microbial markers also follow a normal law. According to these results, and assuming that the same is true for FIB and markers from a WWTP influent, we use parametric tests (Student’s t-test; Microsoft Office Excel 2007) to investigate the differences between (i) the T90 of FIB and microbial markers in SW and FW and (ii) the decay of FIB and microbial markers. On the contrary, there is no available evidence showing that the decay rates of chemical markers are normally distributed. Hence the differences between the T1/2 of chemical markers in SW and FW are studied using nonparametric tests (Mann and Whitney test; XLSTAT 2010.2). For statistical tests, we set an α value of 0.05 as the probability of having no significant difference even if there is a numerical difference. The differences between the decay of FIB and chemical markers are analyzed by linear regressions of FIB and marker concentration data. The correlation between the timeevolution of FIB and chemical markers is quantified by the regression coefficient R2 (Microsoft Office Excel 2007).

3. RESULTS The persistence of human fecal markers was investigated in six independent controlled water microcosms seeded with a WWTP influent; three were filled with SW and three with FW. In the WWTP-free SW, levels of cEC, FRNAPH II, HF183, BifAd, and caffeine are below the detection limits, while the concentration of cENT is 1 CFU/100 mL. The concentrations of cop, etcop and sito are 0.03, 0.03, and 0.04 μg/L, respectively. In the WWTP-free FW, levels of FRNAPH, BifAd, HF183 and caffeine remain below the detection limits, whereas the concentrations of cEC and cENT are 2 and 4 CFU/100 mL, respectively. The concentrations of cop, etcop, and sito are 0.04, 0.04, and 0.07 μg/L, respectively. In the 2378



Article

evolution. In SW, the initial concentrations of cop and etcop were 1.6 ± 0.5 and 0.4 ± 0.1 μg/L, respectively. Between D0 and D2, the concentrations of cop and etcop increased by 190 and 224%/day, respectively, then decreased by 17%/day between D2 and D6, fell to 12%/day between D6 and D13 and finally to 2%/day until D55. In SW, the initial concentration of sito was 0.23 ± 0.04 μg/L. The concentration of this stanol increased by 28%/day between D0 and D2, then decreased by 15%/day between D2 and D6 and remained stable up to D27. Between D27 and D34, the concentration of sito decreased by 8%/day and then remained stable until D55. In FW, the initial concentrations of cop and etcop were 5.7 ± 0.4 and 1.8 ± 0.2 μg/L, respectively. Between D0 and D2, the concentrations of cop and etcop increased, respectively, by 53 and 34%/day, then fell to 18 and 16%/day between D2 and D6, 12%/day between D6 and D13 and 2%/day until D55. In FW, the initial concentration of sito was 0.50 ± 0.04 μg/L. It increased by 5%/day between D0 and D2, then decreased by 17%/day between D2 and D6 to remain stable up to day D27. Between D27 and D34, the concentration of sito decreased by 9%/day, and then remained stable until D55. The initial increase observed for each stanol might be due to the breakdown of ester and ether bonds involving stanols in the initial WWTP influent, releasing free stanols that were analyzed by GC/MS.24 For cop, etcop,, and sito, T1/2 values in SW were 8.3, 15.5, and 20.9 days, respectively, while T1/2 in FW were 3.8, 4.4, and 5.4 days, respectively. For each fecal stanol, the difference between T1/2 in SW and FW was significant (p < 0.05). 3.7.b. Evolution in Fecal Stanol Ratios. Since the decay curves of cop, etcop, and sito evolved differently, the values obtained for fecal ratios R1 and R2 changed during the course of the experiments (Supporting Information Figure S1). In SW, R1 decreased regularly from 0.81 to 0.55, whereas R2 rose from 0.15 to 1.42. In FW, R1 decreased regularly from 0.76 to 0.55, whereas R2 rose from 0.09 to 1.52. In both SW and FW, R1 and R2 showed values characteristic of human fecal matter up to D13 and D6, respectively. As a consequence, the information provided by these ratios indicated that the distribution of fecal stanols remained in the range of WWTP effluent values up to D6.

two microcosms and up to D55 in the third. In SW and FW, the density of cENT decreased by 0.3 Log10/day between D0 and D13 and 0.1 Log10/day between D13 and D41 (SW) or D48 (FW). The modeling of these decay curves yielded a T90 value of 3.6 and 3.1 days for SW and FW, respectively. This difference was not significant. The regulatory values for cENT required to classify waters as SQ have been set at 185 and 330 CFU/100 mL in SW and FW, respectively. These regulatory values were attained after 15.2 and 11.7 days in the SW and FW microcosms, respectively. 3.3. HF183. The initial concentrations were 6.5 ± 2.5 × 106 and 2.8 ± 0.4 × 107 copies/100 mL in SW and FW, respectively. HF183 was quantified up to D6 in SW and FW microcosms. In both types of water, the number of copies remained stable between D0 and D2, and then decreased by 0.9 Log10/day between D2 and D6. T90 was significantly (p < 0.05) higher in SW (2.3 days) than in FW (1.7 days). 3.4. BifAd. In SW, the initial concentration of BifAd was 1.9 ± 0.9 × 106 copies/100 mL. This marker was quantified up to D6 (2 microcosms) and D13 (one microcosm). Its concentration remained stable between D0 and D2, and then decreased by 0.6 Log10/day until it fell below LQ. In FW, the initial concentration was 3.1 ± 0.4 × 106 copies/100 mL and BifAd was quantified in the three microcosms up to D20. The concentration remained stable between D0 and D2, decreased sharply by 0.6 Log10/day up to D6 and, finally, decreased by 0.1 Log10/day until D20. T90 was 3.7 and 3.6 days in SW and FW, respectively. This difference is not significant. 3.5. FRNAPH. Initial concentrations were 4.0 ± 3.0 × 104 (SW) and 1.9 ± 0.1 × 104 (FW) PFU/100 mL. FRNAPH was quantified up to D2 and D13 in SW and FW microcosms, respectively. The percentage of FRNAPH from genogroup II ranged between 96 and 100%. In SW, its concentration decreased by 1 Log10/day between D0 and D2. In FW, the concentration decreased by 0.2 Log10/day between D0 and D6, and then by 0.04 Log10/day up to D13. T90 was significantly (p < 0.05) higher in FW (5.2 days) than in SW (1.1 days). 3.6. Caffeine. In SW, the initial concentration was 3.8 ± 0.7 μg/L. Caffeine was quantified up to D6 (2 microcosms) and D13 (1 microcosm). Its concentration decreased by 20%/day between D0 and D2, remained stable until D6 and then decreased by 14%/day between D6 and D13. In FW, the initial concentration of caffeine was 3.2 ± 0.4 μg/L. This molecule was quantified up to D27 in all three microcosms. Its concentration decreased by 21%/day between D0 and D2, remained stable until D6 and then regularly decreased by 7%/ day between D6 and D20 and by 2.7%/day until D27. The modeling yielded T1/2 values of 2.9 and 4.9 days in SW and FW, respectively. In SW, due to scattering of the data, the modeling was performed on the mean decay curve of the three microcosms. On the contrary, in FW, the modeling was performed on each decay curve. As a result, one set of modeling parameters was calculated for SW and three sets for FW. Since only one value was calculated for T1/2 in SW, it was not possible to apply statistical tests to assess the difference of T1/2 between SW and FW. However, the confidence interval of T1/2 in FW ranged from 3.9 to 5.8. Since T1/2 in SW lay outside this interval, the difference between T1/2 in FW and SW was considered as significant. 3.7. Fecal Stanols. 3.7.a. Decay Curves. The three investigated fecal stanols were quantified up to the end of the experiment in the six microcosms (SW and FW). Among these compounds, cop and etcop followed the same pattern of

4. DISCUSSION 4.1. Intra-Marker Comparisons between Seawater and Freshwater. The T90 of cEC was 3.4 times lower in SW than in FW, whereas the T90 of cENT was not significantly different between these matrices, which confirms that cENT is a more suitable FIB than cEC for investigating fecal contamination in seawater.25,26 In our experimental conditions, HF183 exhibited slightly higher T90 in SW (2.3 days) than in FW (1.7 days). This result is in agreement with the positive effect of salinity on the persistence of Bacteroidales 16S rRNA genetic marker previously described.27,28 Concerning HF183 in particular, T90 in SW was 3.8 times lower than the T90 obtained by Walters et al.29 under similar experimental conditions, whereas T90 in FW was of the same order of magnitude as previous results.30,31 The T90 of BifAd in SW and in FW were not significantly different, which is concordant with the results of Rhodes and Kator.32 However, the decay rate between D6 and Tf was higher in SW (0.6 Log10/day) than in FW (0.1 Log10/day). Although this did not impact the T90, since its value was lower than 6 days in both SW and FW, this difference implied that BifAd 2379



Article

differences between (i) HF183T90 and cENTT90 and (ii) HF183T90 and cECT90 were significant, thus indicating the following order of persistence: HF183 < cENT < cEC, which corresponds to the data of Walters and Field31 from FW experiments. The differences between BifAdT90 and cECT90 were significant in both SW and FW, contrary to the differences between BifAd T90 and cENTT90. In SW, cECT90 was 45% lower than BifAdT90, whereas, in FW, cECT90 was 60% higher than BifAdT90. This latter result is in agreement with previous studies comparing the decay of fecal coliforms and sorbitol-fermenting bacteria in freshwater.37 Thus, the persistence of BifAd appears similar to that observed for cENT in SW and FW, while it is longer than cEC in SW and shorter than cEC in FW. In SW, FRNAPH IIT90 was 3.6 and 1.6 times lower than cENTT90 and cECT90, respectively. This is in agreement with previous results on the decay of FRNAPH II and FIB from sewage water in natural seawater conditions,38 which are corroborated by the comparison between cENT and FRNAPH II but not between cEC and FRNAPH II. As a consequence, according to the experimental conditions, FRNAPH IIT90 in SW can be higher38 or lower (this study) than cECT90, while cENT would appear to persist longer than FRNAPH. In FW, the differences between FRNAPH II T90, cENTT90 and cECT90 were not significant. Hence, the decay of FRNAPH II in FW, under the present experimental conditions, was closely related to the decay of both FIB, which is compatible with previous results on the decay of FIB and bacteriophages in FW under natural conditions.39 This result validates the third requirement for the application of FRNAPH II as a human-associated marker in FW. The regression tests performed on the average concentrations of caffeine and FIB indicated that, although not due to the same processes, the decay of caffeine was closely correlated to the decay of cENT (R2 = 0.95 and 0.94 in SW and FW, respectively) and cEC (R2 = 0.76 and 0.89 in SW and FW, respectively). The coefficients of correlation between the decay of fecal stanols in SW and FW and the decay of cENT (SW and FW) and cEC (FW) were lower than those of caffeine. However, these coefficients ranged from 0.69 to 0.79, with a mean value of 0.75 ± 0.01, which represents a good correlation. In SW, the Tf (the latest time at which a marker could be quantified in at least one microcosm) for HF183, BifAd, FRNAPH II, and caffeine (Table 1) were lower than Treg (15.2 days). In FW, the Tf for HF183 and FRNAPH II were lower than Treg (11.7 days), whereas the Tf values for BifAd and caffeine were higher. As already mentioned, the stanol ratios R1 and R2 remained in the field of WWTP effluents values up to 6 days, which was lower than Treg in both SW and FW. 4.3. Inter-Marker Comparisons. In this study, we test the differences between the decays of the specific markers in SW and in FW. In the experimental conditions of this study, the differences were significant (p < 0.05) in both types of recreational water, except in the case of BifAd and FRNAPH II in FW. Moreover, the different analytical methods do not have the same sensitivity, which leads to different Tf values (Table 1). Thus, the markers can be classified in order of increasing persistence in SW (FRNAPH II < HF183, stanol ratios < BifAd, caffeine) and in FW (HF183, stanol ratios < FRNAPH II < BifAd < caffeine). 4.4. Implication for the Application of the MST Toolbox. The above results show that, under the studied experimental conditions, in SW, the Tf of all the specific markers were lower than Treg. As a consequence, the application

remained quantifiable longer in FW (Tf = 20 days) than in SW (Tf = 13 days). FRNAPH II exhibited lower T90 in SW (1.1 days) than in FW (5.2 days). This value of T90 in SW is similar to the value previously reported for the persistence of FRNAPH incubated in SW in the dark at 20 °C (T90 = 1.5 days).33 The T90 of FRNAPH in FW is similar to the results obtained for the decay of group II FRNAPH in lake water incubated at 20 °C in the dark (calculated T90 = 6.0 days).34 However, our results are in contradiction with previous results highlighting that salinity does not significantly impact the persistence of FRNAPH from group II.23 The T1/2 of caffeine was 1.7 times lower in SW (2.9 days) than in FW (4.9 days). Those values in SW are in the order of the values obtained for a degradation study of caffeine performed in the dark at 15 °C that ranged from 3.5 to 13 days.35 However, the present T1/2 in FW is much lower than the value from Buerge et al.36 who reported a T1/2 of 115 days for an incubation study using lake water spiked with a solution of caffeine in the dark at 20 °C. The T1/2 of cop, etcop and sito in SW was 2.2, 3.5, and 3.9 higher than in FW, respectively. However, contrary to microbial markers and caffeine, the specificity of fecal stanols in tracking human fecal contamination is not based on their absolute concentrations but on their relative distribution expressed, in this study, by the ratios R1 and R2. Although these three molecules exhibited different decays in SW and in FW, the ratios R1 and R2 followed the same evolution in both of these matrices(Supporting Information Figure S1), which is in agreement with the results of Solecki et al.17 on the degradation of pig-specific markers. Our experiments were performed using unfiltered SW and FW and, although the microcosms were subject to constant mixing, some partial sedimentation occurred. As a result, the observed decay of markers was due to a combination of processes including death, inactivation, biodegradation and/or sedimentation. Due to the sensitivity of the FIB and the specific markers to the physicochemical conditions (light, temperature, dissolved organic matter, etc.), it is difficult to determine the major processes explaining their decay by comparing the present results with previous data. 4.2. Comparisons between Specific Markers and FIB. The decay of specific markers was compared to the decay of cEC and cENT in SW and FW to test the third required condition for the development of a host-specific marker, i.e., it should exhibit an extra-intestinal persistence similar to FIB. Moreover, the Tf of markers (Table 1) were compared with the time to reach the SQ regulatory values (Treg) for FIB. In SW, Treg was clearly longer for cENT (15.2 days) than for cEC (3.6 days). In FW, there was no significant difference between Treg for cENT (11.7 days) and Treg for cEC (11.3 days). Since Treg was longer for cENT than for cEC in both SW and FW, cENT was the most restrictive parameter. As a consequence, the Tf values of human-associated markers were compared with the values of Treg for cENT in SW and FW. In SW, the differences (i) between HF183T90 and cENTT90 and (ii) between HF183T90 and cECT90 were not significant, in contrast to the difference between cECT90 and cENTT90. As a result, under these experimental conditions, the decay of HF183 was closely related to the decay of each FIB. Although this result was in contradiction with the data of Walters et al.,29 which show HF183T90 3.5 times higher than cENTT90, it validates HF183 as a human-associated marker in SW. In FW, the 2380



Article

water use and other aquatic facility-associated health eventsUnited States, 2005−2006. In Morbidity and Mortality Weekly Report, Surveillance Summaries, 2008; Vol. 57, pp 1−29 (3) Scott, T. M.; Rose, J. B.; Jenkins, T. M.; Farrah, S. R.; Lukasik, J. Microbial source tracking: Current methodology and future directions. Appl. Environ. Microbiol. 2002, 68, 5796−5803. (4) Blanch, A. R.; Belanche-Munoz, L.; Bonjoch, X.; Ebdon, J.; Gantzer, C.; Lucena, F.; Ottoson, J.; Kourtis, C.; Iversen, A.; Kuhn, I.; Moce, L.; Muniesa, M.; Schwartzbrod, J.; Skraber, S.; Papageorgiou, G. T.; Taylor, H.; Wallis, J.; Jofre, J. Integrated analysis of established and novel microbial and chemical methods for microbial source tracking. Appl. Environ. Microbiol. 2006, 72, 5915−5926. (5) Hagedorn, C.; Weisberg, S. Chemical-based fecal source tracking methods: Current status and guidelines for evaluation. Rev. Environ. Sci. Biotechnol. 2009, 8, 275−287. (6) Sinton, L. W.; Finlay, R. K.; Hannah, D. J. Distinguishing human from animal fecal contamination in water: A review. N. Z. J. Mar. Freshwater Res. 1998, 32, 323−348. (7) Gourmelon, M.; Caprais, M. P.; Mieszkin, S.; Marti, R.; Wéry, N.; Jardé, E.; Derrien, M.; Jadas-Hécart, A.; Communal, P. Y.; Jaffrezic, A.; Pourcher, A. M. Development of microbial and chemical MST tools to identify the origin of the fecal pollution in bathing and shellfish harvesting waters in France. Water Res. 2010, 44, 4812−4824. (8) Bernhard, A. E.; Field, K. G. A PCR assay to discriminate human and ruminant feces on the basis of host differences in bacteroidesprevotella genes encoding 16S rRNA. Appl. Environ. Microbiol. 2000, 66, 4571−4574. (9) Seurinck, S.; Defoirdt, T.; Verstraete, W.; Siciliano, S. D. Detection and quantification of the human-specific HF183 Bacteroides 16S rRNA genetic marker with real-time PCR for assessment of human fecal pollution in freshwater. Environ. Microbiol. 2005, 7, 249− 259. (10) Wéry, N.; Monteil, C.; Pourcher, A.-M.; Godon, J.-J. Humanspecific fecal bacteria in wastewater treatment plant effluents. Water Res. 2010, 44, 1873−1883. (11) Dick, L. K.; Bernhard, A. E.; Brodeur, T. J.; Santo Domingo, J. W.; Simpson, J. M.; Walters, S. P.; Field, K. G. Host distributions of uncultivated fecal bacteroidales bacteria reveal genetic markers for fecal source identification. Appl. Environ. Microbiol. 2005, 71, 3184−3191. (12) Lamendella, R.; Domingo, J. W. S.; Kelty, C.; Oerther, D. B. Bifidobacteria in feces and environmental waters. Appl. Environ. Microbiol. 2008, 74, 575−584. (13) Ogorzaly, L.; Gantzer, C. Development of real-time RT-PCR methods for specific detection of F-specific RNA bacteriophage genogroups: Application to urban raw wastewater. J. Virol. Methods 2006, 138, 131−139. (14) Glassmeyer, S. T.; Furlong, E. T.; Kolpin, D. W.; Cahill, J. D.; Zaugg, S. D.; Werner, S. L.; Meyer, M. T.; Kryak, D. D. Transport of chemical and microbial compounds from known wastewater discharges: Potential for use as indicators of human fecal contamination. Environ. Sci. Technol. 2005, 39, 5157−5169. (15) Leeming, R.; Ball, A.; Ashbolt, N.; Nichols, P. Using fecal sterols from humans and animals to distinguish fecal pollution in receiving waters. Water Res. 1996, 30, 2893−2900. (16) Leeming, R. et al. Detecting and distinguishing sources of sewage pollution in australian inland and coastal waters and sediments. In Molecular Markers in Environmental Geochemistry; Eganhouse, R. P., Ed.; ACS Symposium series: Washington, DC, 1997; Vol. 671. (17) Solecki, O.; Jeanneau, L.; Jardé, E.; Gourmelon, M.; Marin, C.; Pourcher, A. M. Persistence of microbial and chemical pig manure markers as compared to faecal indicator bacteria survival in freshwater and seawater microcosms. Water Res. 2011, 45, 4623−4633. (18) Figuet, C.; Frangi, J.-P. Receiving fresh water from domestic wastewater treatment plant outlet: A case study for Mauldre (Ile-deFrance), a medium under very high pressure. Rev. Sci. Eau 2000, 13, 119−138. (19) Ogorzaly, L.; Tissier, A.; Bertrand, I.; Maul, A.; Gantzer, C. Relationship between F-specific RNA phage genogroups, fecal

of these specific markers should be restricted to the investigation of the origin of recent contaminations, as already pointed out for HF18331 and BifAd.40 Moreover, the increasing order of persistence (FRNAPH II < HF183, stanol ratios < BifAd, caffeine) implies that the occurrence of FRNAPH II could reflect a recent human contamination dating back as far as 2 days (FRNAPH IITf = 2 days in SW), whereas a BifAd positive/HF183 negative sample could be characteristic of an earlier human contamination going back to between 2 and 6 days (HF183Tf = 6 days in SW). In FW, all the specific markers except HF183 and fecal stanol ratios showed Tf higher than Treg. It means that BifAd, FRNAPH II and caffeine are not just limited to investigating the origin of recent contaminations but can also be used to monitor the evolution of contamination in terms of insufficient quality or sufficient water quality. The present study involves the simultaneous investigation of the decay of FIB and human-associated markers inherited from a WWTP influent in seawater and freshwater microcosms. Human-associated markers are microbiological and chemical markers of an interdisciplinary MST toolbox. Such a study is critical to the implementation of such tools in environmental regulatory activities. This experimentation was designed to model the evolution of a fecal contamination by a WWTP influent in a finite volume of water transferred from upstream to downstream by advection. Due to these simplifying assumptions, the provided data must be considered in those experimental conditions. Upscaling those results to natural environments would need to study the influence of natural processes such as (i) the diffusion and the dispersion due to turbulent flow, (ii) the photodegradation due to sun-light and (iii) the effect of temperature.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional figure and information on (i) the analysis of FIB and human-associated markers and (ii) the results of statistical tests and regressions. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +33 (0)2 23 23 39 69; e-mail: [email protected].

■

ACKNOWLEDGMENTS This research forms part of the “Marquopoleau” research programme supported by the French General Directorate for Competitiveness, Industry and Services (DGCIS), the Brittany Regional Council and the County Councils of Morbihan, Finistère and Ille-et-Vilaine. We thank E. Guillerm, O. Henin, C. Le Mennec, C. Marin, and P. Petitjean for their technical support and Dr. M. Carpenter for revising the English style of the manuscript.

■

REFERENCES

(1) Rabinovici, S. J. M.; Bernknopf, R. L.; Wein, A. M.; Coursey, D. L.; Whitman, R. L. Economic and health risk trade-offs of swim closures at a Lake Michigan Beach. Environ. Sci. Technol. 2004, 38, 2737−2745. (2) Yoder, J.; Hlavsa, M.; Craun, G.; Hill, V.; Roberts, V.; Yu, P.; Hicks, L.; Alexander, N.; Calderon, R. ; Roy, S.; Beach, M.Surveillance for waterborne disease and outbreaks associated with recreational 2381



Article

pollution indicators and human adenoviruses in river water. Water Res. 2009, 43, 1257−1264. (20) Jeanneau, L.; Jardé, E.; Gruau, G. Influence of salinity and natural organic matter on the solid phase extraction of sterols and stanols: Application to the determination of the human sterol fingerprint in aqueous matrices. J. Chromatogr., A 2011, 1218, 2513−2520. (21) Lee, D.-U.; Heinz, V.; Knorr, D. Biphasic inactivation kinetics of escherichiacoli in liquid whole egg by high hydrostatic pressure treatments. Biotechnol. Prog. 2001, 17, 1020−1025. (22) Cerf, O. A REVIEW tailing of survival curves of bacterial spores. J. Appl. Microbiol. 1977, 42, 1−19. (23) Schaper, M.; Jofre, J.; Uys, M.; Grabow, W. O. K. Distribution of genotypes of F-specific RNA bacteriophages in human and non-human sources of fecal pollution in South Africa and Spain. J. Appl. Microbiol. 2002, 92, 657−667. (24) Thoumelin, G.; Marty, Y.; Le Corre, P.; Aminot, A. Laboratory investigation of the degradation of organic matter in estuarine and coastal waters: Sterols variations. Oceanol. Acta 1990, 13, 53−60. (25) Guidelines for Safe Recreational Waters−Vol. 1−Coastal and Fresh Waters; World Health Organization: Geneva, Switzerland, 2003; www. who.int/water_sanitation_health/bathing/srwe1/en/. (26) Wade, T. J.; Pai, N.; Eisenberg, J. N. S.; Colford, J. M. Do U.S. Environmental Protection Agency water quality guidelines for recreational waters prevent gastrointestinal illness? A systematic review and meta-analysis. Environ. Health Perspect. 2003, 111. (27) Okabe, S.; Shimazu, Y. Persistence of host-specific BacteroidesPrevotella 16S rRNA genetic markers in environmental waters: Effects of temperature and salinity. Appl. Microbiol. Biot. 2007, 76, 935−944. (28) Schulz, C. J.; Childers, G. W. Fecal bacteroidales diversity and decay in response to variations in temperature and salinity. Appl. Environ. Microbiol. 2011, 77, 2563−2572. (29) Walters, S. P.; Yamahara, K. M.; Boehm, A. B. Persistence of nucleic acid markers of health-relevant organisms in seawater microcosms: Implications for their use in assessing risk in recreational waters. Water Res. 2009, 43, 4929−4939. (30) Dick, L. K.; Stelzer, E. A.; Bertke, E. E.; Fong, D. L.; Stoeckel, D. M. Relative decay of bacteroidales microbial source tracking markers and cultivated Escherichia coli in freshwater microcosms. Appl. Environ. Microbiol. 2010, 76, 3255−3262. (31) Walters, S. P.; Field, K. G. Survival and persistence of human and ruminant-specific fecal Bacteroidales in freshwater microcosms. Environ. Microbiol. 2009, 11, 1410−1421. (32) Rhodes, M. W.; Kator, H. Sorbitol-fermenting bifidobacteria as indicators of diffuse human fecal pollution in estuarine watersheds. J. Appl. Microbiol. 1999, 87, 528−535. (33) Pommepuy, M.; Caprais, M.-P.; Le Saux, J.-C.; Le Mennec, C.; Parnaudeau, S.; Madec, A.; Monier, M.; Brest, G.; Le Guyader, F. Evaluation of viral shellfish depuration in a semi-professional tank. In Proceedings of the 4th International Conference on Molluscan Shellfish Safety; Santiago de Compostella, Spain, 2002. (34) Long, S. C.; Sobsey, M. D. A comparison of the survival of F+RNA and F+DNA coliphages in lake water microcosms. J. Water Health 2004, 2, 15−22. (35) Benotti, M. J.; Brownawell, B. J. Microbial degradation of pharmaceuticals in estuarine and coastal seawater. Environ. Pollut. 2009, 157, 994−1002. (36) Buerge, I. J.; Poiger, T.; Müller, M. D.; Buser, H.-R. Caffeine, an anthropogenic marker for wastewater contamination of surface waters. Environ. Sci. Technol. 2003, 37, 691−700. (37) Ottoson, J. R. Bifidobacterial survival in surface water and implications for microbial source tracking. Can. J. Microbiol. 2009, 55, 642−647. (38) Moce-Llivina, L.; Lucena, F.; Jofre, J. Enteroviruses and bacteriophages in bathing waters. Appl. Environ. Microbiol. 2005, 71, 6838−6844. (39) Durán, A. E.; Muniesa, M.; Méndez, X.; Valero, F.; Lucena, F.; Jofre, J. Removal and inactivation of indicator bacteriophages in fresh waters. J. Appl. Microbiol. 2002, 92, 338−347.

(40) Morrison, C. R.; Bachoon, D. S.; Gates, K. W. Quantification of enterococci and bifidobacteria in Georgia estuaries using conventional and molecular methods. Water Res. 2008, 42, 4001−4009.

2382


Relative Decay of Fecal Indicator Bacteria and ... - ACS Publications

Recommend Documents