- Email: [email protected]
Fecal coliform accumulation within a river subject to seasonally-disinfected wastewater discharges
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Fecal coliform accumulation within a river subject to seasonally-disinfected wastewater discharges Azalea A. Mitch, Katherine C. Gasner, William A. Mitch* Department of Chemical Engineering, Yale University, Mason Lab 313b, 9 Hillhouse Avenue, New Haven, CT 06520, United States
article info
abstract
Article history:
As pathogen contamination is a leading cause of surface water impairment, there has been
Received 17 December 2009
increasing interest in the implications of seasonal disinfection practices of wastewater
Received in revised form
effluents for meeting water quality goals. For receiving waters designated for recreational
17 May 2010
use, disinfection during the winter months is often considered unnecessary due to reduced
Accepted 18 May 2010
recreational usage, and assumptions that lower temperatures may reduce pathogen
Available online 12 June 2010
accumulation. For a river subject to seasonal disinfection, we sought to evaluate whether fecal coliforms accumulate during the winter to concentrations that would impair river
Keywords:
water quality. Samples were collected from municipal wastewater outfalls along the river,
Fecal coliform
as well as upstream and downstream of each outfall during the winter, when disinfection
Seasonal disinfection
is not practiced, and during the summer, when disinfection is practiced. During both seasons, fecal coliform concentrations reached 2000e5000 CFU/100 mL, nearly an order of magnitude higher than levels targeted for the river to achieve primary contact recreational uses. During the summer, wastewater effluents were not significant contributors to fecal coliform loadings to the river. During the winter, fecal coliform accumulated along the river predominantly due to loadings from successive wastewater outfalls. In addition to the exceedance of fecal coliform criteria within the river, the accumulation of wastewaterderived fecal coliform along the river during the winter season suggests that wastewater outfalls may contribute elevated loads of pathogens to the commercial shellfish operations occurring at the mouth of the river. Reductions in fecal coliform concentrations between wastewater outfalls were attributed to dilution or overall removal. Combining discharge measurements from gauging stations, tributaries and wastewater outfalls to estimate seepage, dilution between wastewater outfalls was estimated, along with the percentage of the river deriving from wastewater outfalls. After accounting for dilution, the residual reductions in fecal coliform concentrations observed between outfalls were attributed to actual fecal coliform removal. The estimated rate of removal of 1.52 d1 was significantly higher than die-off rates determined by previous researchers at similarly low temperatures in laboratory batch experiments, indicating the potential importance of other removal mechanisms, such as predation or sedimentation. ª 2010 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ1 203 432 4386; fax: þ1 203 432 4387. E-mail address: [email protected] (W.A. Mitch). 0043-1354/$ e see front matter ª 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2010.05.060
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
1.
Introduction
Pathogens are amongst the most important causes of water body impairments in the United States. High levels of pathogenic bacteria found in impaired waterways present a health risk to bathers (Boehm et al., 2002; Liu et al., 2006), non-contact recreational users such as anglers and boaters (Donovan et al., 2008; Hellweger and Masopust, 2008), and consumers of raw or undercooked shellfish such as oysters (Ueki et al., 2005; LeGuyader et al., 1996; Formiga-Cruz et al., 2002). In addition to the health risks to humans, pathogen contamination can incur serious economic hardships on communities due to the closure of beaches or shellfish beds. Pathogen loadings derive from non-point source inputs from urban or agricultural activities (Petersen et al., 2005; Dorner et al., 2004), dry weather flows from storm sewers (Boehm et al., 2002), wet weather loadings from combined or sanitary sewer overflows (Donovan et al., 2008), or discharges from wastewater treatment plants. Attempts to develop water quality models to predict violations of pathogen standards have combined pathogen inactivation and hydrodynamic models. Inactivation models incorporate results obtained from laboratory (Medema et al., 1997) or mesocosm (Sinton et al., 2002; Dutka and Kwan, 1980; Easton et al., 2005) studies to predict pathogen die-off as a function of nutrient and toxin concentrations, pH, temperature, salinity, sunlight and predation. Hydrodynamic models are employed to track pathogen plumes and account for their dilution. These models assess advection and dispersion as well sedimentation and resuspension of pathogens from sediments (Hellweger and Masopust, 2008). In the cases of lakes and oceans, these models have also incorporated the effects of wind-driven currents and tides (Boehm et al., 2002; Liu et al., 2006). At their most complex, these models have solved two or three dimensional advection and dispersion equations within a finite element framework (Hellweger and Masopust, 2008; Liu et al., 2006). Evaluating the fate of pathogens discharged to rivers can be even more complex. In oceans, lakes (Liu et al., 2006) and dammed rivers that behave like lakes (Hellweger and Masopust, 2008), hydrodynamic models have tracked plumes associated with individual wastewater outfalls or other point discharges. Rivers, however, feature multiple pathogen loading inputs from both point and non-point sources along their course such that pathogens may accumulate within the same plume. Furthermore, water withdrawals along the river may remove pathogens while water inputs from uncontaminated tributaries and groundwater seepage may result in dilution. A previous study addressed these complications by tracking the transport downstream of a non-native bacterial culture spiked into a river; however, only presence/absence data were obtained (Dutka and Kwan, 1980). Seasonal disinfection of municipal wastewater effluents is permitted in many locations. Where the designated beneficial use of the receiving water is for recreational uses, regulatory agencies may deem disinfection of wastewater effluents during the winter months unnecessary, due to the reduction in recreational activities. However, the implications of these policies for water quality in both the primary receiving waters
4777
and in other waters to which these receiving waters discharge have not been a focus of previous research. In this paper, we evaluated the fate and transport of wastewater-derived fecal coliform within a river system. Designated for recreational use, the river system is subject to seasonal disinfection of municipal wastewater effluents. We measured fecal coliform accumulation along the river during the winter months, and evaluated overall fecal coliform removal rates within the river. Our results raised questions about the implications of seasonal disinfection for public health risks associated with recreational use along the river and shellfish harvesting activities conducted at the mouth of the river.
2.
Experimental section
2.1.
Study site
The Quinnipiac River is located in central Connecticut, originating in Deadwood Swamp north of the Town of Southington and flowing southerly for 61 km to New Haven, where it empties into Long Island Sound (Fig. 1). In a previous study, we had employed boron as a conservative tracer to evaluate the fate of wastewater-derived precursors of nitrosamine disinfection by-products within this river (Schreiber and Mitch, 2006). Registered and permitted activities include 103 water withdrawals and 19 discharges. Five of the dischargers are municipal wastewater treatment plants; from north to south: Southington, Cheshire, Meriden, Wallingford and North Haven (Schreiber and Mitch, 2006). Only the four most northerly wastewater treatment plants were evaluated for this study because the section of the river near the North Haven facility is influenced by tidal action. Eight tributaries empty into the Quinnipiac River between Southington and Wallingford (Table 1): Misery and Honeypot Brooks between the Southington and Cheshire outfalls, Broad, Sodom and Harbor Brooks between the Cheshire and Meriden outfalls, and Meetinghouse Brook and two small unnamed brooks between the Meriden and Wallingford outfalls. One active USGS gauging station is located upstream of the Southington outfall, prior to the conjunction of the Quinnipiac River with the Eightmile and Tenmile Rivers. An additional USGS gauging station is located just upstream of the Wallingford outfall. Shellfish breeding beds are located at the mouth of the Quinnipiac River in New Haven harbor.
2.2.
Regulatory framework
The Quinnipiac River falls under two types of regulatory frameworks. Under the authority of the Connecticut Department of Environmental Protection (CT DEP), the Quinnipiac River is designated for recreational uses including swimming, fishing and boating. Used as indicator bacteria, Escherichia coli (E. coli) levels in the river must remain below a geometric mean of 126 CFU/100 mL and a single sample maximum of 576 CFU/100 mL. To achieve these criteria, the CT DEP issued wastewater treatment plants National Pollutant Discharge Elimination System (NPDES) permits whereby fecal coliform levels, used as indicator bacteria, must be below a geometric
4778
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
Table 1 e Streamflow Measurements. Site
4/21/2010 m3/s
2/22/09-2/25/09 m3/s
0.96 0.87 1.49
0.96 0.87 1.49 0.29
Southington USGS Eightmile River Tenmile River Seepage Southington WWTP Discharge Miser Brook Honeypot Brook Seepage
0.38 0.21
0.24 0.38 0.21 0.34
Cheshire WWTP Discharge Broad Brook Sodom Brook Harbor Brook Seepage
0.16 0.05 0.38
0.13 0.16 0.05 0.38 0.38
Meriden WWTP Discharge Meetinghouse Brook Unnamed Brook #1 Unnamed Brook #2 Seepage
0.06 0.04 0.04
0.54 0.06 0.04 0.04 0.56
Wallingford USGS Wallingford WWTP Discharge Wallingford Downstream Total Flow Without Seepagea
7.47
7.12 0.29 7.41 5.56
a To the Wallingford USGS Station.
National Shellfish Sanitation Program (NSSP), which requires that the geometric mean of fecal coliform concentrations of waters in approved shellfish growing regions exhibit levels below 14 CFU/100 mL (NSSP, 2007). Shellfish harvesting in the Quinnipiac River and most of New Haven Harbor is prohibited, but the seeding of oysters and clams is allowed in these areas. After seeding, the clams and oysters from these prohibited areas must be transferred to approved areas that are located 4 miles offshore, which is believed to permit depuration of pathogens prior to harvest.
Fig. 1 e Sampling locations along the Quinnipiac River. , [ municipal wastewater treatment plant effluent. C [ sampling point on the river upstream or downstream of a wastewater outfall. SU [ Southington upstream, SE [ Southington effluent, SD [ Southington downstream, CU [ Cheshire upstream, CE [ Cheshire effluent, CD [ Cheshire downstream, MU [ Meriden upstream, ME [ Meriden effluent, MD [ Meriden downstream, WU [ Wallingford upstream, WE [ Wallingford effluent, WD [ Wallingford downstream. : [ USGS gauging station 01194500.
2.3.
mean of 200 CFU/100 mL and a single sample maximum of 400 CFU/100 mL. However, disinfection of wastewater effluent is only required between May 1 and September 30 (CT DEP, 2008). The CT DEP considered the summer months to be the most challenging under the assumption that indicator bacteria survive longer in waters with temperatures closer to human body temperature. However, the Quinnipiac River is currently considered impaired for recreational use due to elevated E. coli levels. The CT DEP considers the likely cause of these elevated levels to be storm sewers and non-point source runoff and has developed a Total Maximum Discharge Level (TMDL) analysis to reduce E. coli concentrations in the Quinnipiac River. Commercial shellfish breeding in the State of Connecticut occurs year round. The State Department of Agriculture participates in the U.S Food and Drug Administration’s
Samples were collected from the 4 most northerly municipal wastewater treatment plant effluents, as well as locations located immediately upstream and at least 100 m downstream of the discharges (Fig. 1). These sampling sites were a subset of the ones employed in our previous study (Schreiber and Mitch, 2006); the locations had been identified using the method of Fischer et al. (1979) to achieve mixing of discharges throughout the river’s cross-section. Samples were collected in fluorinated high-density polyethylene containers, and were stored at 4 C pending analysis. Samples were collected once per day from each site on 4 consecutive days during the winter (February 22e25, 2009), when disinfection was not being conducted at the wastewater treatment plants. Sampling over 4 consecutive days, approximately the time frame of travel down the river, enabled an evaluation of sample variability. During this period, non-point source inputs to the river were minimized as no precipitation occurred, although there was snow on the ground. The river sample temperatures ranged
Sample collection and analysis
4779
Results and discussion
3.1.
Fecal coliform
Fig. 2 presents the geometric mean and standard deviation of fecal coliform concentrations measured daily at each of the sampling locations over four consecutive days from February 22e25, 2009. No precipitation occurred immediately before or during the sampling period and no disinfection of wastewater effluents was practiced. The influence of non-point sources was minimal. The wastewater discharges were clearly the most significant loading to the river during this time, featuring concentrations ranging from 1800 to 17 000 CFU/100 mL. The relatively small error bars indicate that the concentrations were repeatable during the sampling period, resulting in a stable concentration profile along the river. Fecal coliform concentrations increased significantly along the river to reach approximately 5000 CFU/100 mL. Fecal coliform sampling conducted in the summer on August 24, 2009, when disinfection of wastewater effluents
60000 40000
5000 4000 3000 2000 1000 0
SU SE SD CU CE CD MU ME MD WU WE WD
20000
Fig. 2 e Geometric mean fecal coliform concentrations measured in wastewater effluents and the river (A) during the spring along the Quinnipiac River (CT). Sampling sites: Southington (S), Cheshire (C), Meriden (M), Wallingford (W) wastewater treatment plant effluents (E) or either upstream (U) or downstream (D) of the discharges. Samples were collected at each sampling location for four consecutive days: February 22e25, 2009. Error bars represent 1 geometric standard deviation.
was being practiced, indicated concentrations in the river ranging from approximately 500e2200 CFU/100 mL (Fig. 3). During both seasons, the concentrations were comparable, and similar to fecal coliform concentrations measured in other studies of impaired surface waters (Dutka and Kwan, 1980; Hellweger and Masopust, 2008; Liu et al., 2006; Donovan et al., 2008). However, during the summer lower concentrations were observed in the wastewater effluents than in the river, and upstream locations exhibited higher concentrations than downstream. As it had rained the day prior to sampling in the northern section of the sampling area, it is likely that non-point sources were the most significant contributors to fecal coliform loadings in the river. When pathogen loadings from municipal wastewater effluents are reduced by disinfection, previous studies have indicated the importance of pathogen loadings to surface waters from non-
2500 2000 1500 1000 500 0 SU SE SD CU CE CD MU ME MD WU WE WD
3.
80000
Sampling Site
Fecal Coliform (CFU/100 mL)
from 2 to 5 C. For fecal coliform samples, 3 dilutions were prepared and analyzed by the membrane filtration method in triplicate (APHA, 1998). The dilutions exhibiting colony counts closest to 20 CFU per plate were selected for counting. Deionized water controls exhibited no colonies. Geometric mean and geometric standard deviation calculations were employed. An additional day of sampling was conducted during the summer (August 24, 2009), when disinfection of wastewater effluent was being conducted. Although no precipitation occurred on that day, light rain had fallen on the previous day in Southington. The river sample temperatures ranged from 21 to 23 C. Grab samples were collected from the river at locations upstream and downstream of each of the 4 wastewater treatment plants. Whether grab samples capture diurnal variations in fecal coliform concentrations is a concern. Diurnal variations in riverine concentrations of wastewaterderived constituents are anticipated to be less severe than in wastewater effluents due to dispersion within the river and non-uniform spacing of discharges. For example, a morning pulse of fecal coliform discharged from an upstream wastewater treatment plant would be unlikely to meet a morning pulse from a downstream plant on following days due to non-uniform spacing of the discharges along the river. To evaluate the importance of diurnal variations within wastewater discharges, we collected a mixture of grab and 24-h composite final effluent samples at the treatment plants. No significant differences in fecal coliform concentrations were observed between the two types of samples. For example, grab samples collected from the Meriden treatment plant on Days 1 and 2 were 11 900 CFU/100 mL and 24 700 CFU/ 100 mL, respectively, while 24-h composite samples collected on Days 3 and 4 were 20 800 CFU/100 mL and 24 800 CFU/ 100 mL, respectively. Although the recommended holding time for fecal coliform analyses is 6 h, previous studies have indicated no significant differences in E. coli (Pope et al., 2003) or fecal coliform (Standridge and Lesar, 1977) concentrations for samples stored in the refrigerator for up to 24e48 h.
Fecal Coliform (CFU/100 mL)
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
Sampling Site Fig. 3 e Geometric mean fecal coliform concentrations measured on August 24, 2009 along the Quinnipiac River (CT). Sampling sites: Southington (S), Cheshire (C), Meriden (M), Wallingford (W) wastewater treatment plant effluents (E) or either upstream (U) or downstream (D) of the discharges. Error bars represent 1 geometric standard deviation of triplicate analyses.
4780
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
point sources and dry weather flows from storm drains (Hellweger and Masopust, 2008; Petersen et al., 2005; Dorner et al., 2004; Donovan et al., 2008).
3.2.
Dilution analysis
During the February sampling period, although there was an overall increase in fecal coliform concentrations proceeding downriver, there was a slight decline in fecal coliform concentrations between each of the wastewater treatment plant outfalls (Fig. 2). These declines could result either from dilution by tributaries or groundwater seepage, or from overall removal of fecal coliform within the river by processes including death, predation, and sedimentation. To distinguish these possibilities, dilution was quantified by measuring the discharges of all of the tributaries to the Quinnipiac River upstream of the Wallingford outfall on April 21, 2010. At each tributary, the depth (di) and flow velocity were measured at one-foot intervals across the tributary, according to the method of Harrelson et al. (1994). The average velocity at each one-foot cross-section (vi) was taken as the flow velocity measured at 60% of the total depth from the surface using a Marsh-McBirney Flowmate 2000 model electromagnetic velocity meter. The discharge of the tributary (Q) was calculated as (Table 1): Q¼
n X
di vi
1
The 0.96 m3/s discharge recorded by the Southington USGS gauging station (site number 01195490) was the same on April 21, 2010 as the average measured during the February 22e25, 2009 period over which fecal coliform samples were collected. Precipitation had not occurred within 2 d prior to either sampling period. Accordingly, we assumed that the discharges from the tributaries were comparable over the two periods. For the period February 22e25, 2009, the sum of the discharges at the Southington USGS gauging station, and all tributaries and wastewater discharges between this gauging station and the Wallingford USGS gauging station was 5.56 m3/s, a value within 22% of the 7.12 m3/s average discharge measured at the Wallingford USGS gauging station (site number 01196500) over this period. The difference between these values could arise from either the combined
error of the discharge measurements, or additional contributions by groundwater seepage. Assuming that the difference arose from seepage, seepage contributions within river sections were estimated by assuming that seepage occurred evenly along the river. Accordingly, the 1.56 m3/s difference between the discharge recorded and calculated at the Wallingford USGS gauging station was divided among the following river sections according to their length (Tables 1 and 2): Southington USGS gauging station to the Southington outfall, the Southington to Cheshire outfalls, the Cheshire to Meriden outfalls, the Meriden outfall to the Southington USGS gauging station. Using these discharges, the percentage dilution of the river occurring between outfalls ranged from 10.8% between the Meriden and Wallingford outfalls to 24.2% between the Southington and Cheshire outfalls (Table 2). The percentage of the river that derived from wastewater, calculated for locations upstream and downstream of each outfall, increased to 16.2% downstream of the Wallingford outfall. This value was comparable to the 16% measured on March 16e19, 2004, but lower than the 41% measured on August 10e13, 2004, the low flow season (Schreiber and Mitch, 2006). To evaluate total fecal coliform removal rates within the river between wastewater outfalls, the fecal coliform concentration measured just upstream of the downstream outfall was compared to the value anticipated at this location for the case where only dilution, not removal, occurred along this river section. The anticipated concentrations (Canticipated) were calculated as the flow-weighted averages of the concentrations measured just downstream of the upstream outfalls, and the dilution waters entering between the outfalls (Table 2); since no municipal wastewater plants discharge to the tributaries, the fecal coliform concentrations in the dilution waters were assumed similar to the 206 CFU/100 mL measured in the Quinnipiac upstream of the Southington outfall: Canticipated ¼
Cupstream Q upstream þ CSouthington Q dilution Q upstream þ Q dilution
Assuming that total fecal coliform removal rates between outfalls are first order, we used the following equation to estimate these removal rates using the measured fecal coliform concentrations just upstream of the downstream
Table 2 e Fecal Coliform Calculations. Municipal Wastewater Treatment Plant
Dilution between this plant and upstream plant Wastewater fraction upstream Wastewater fraction downstream Measured upstream fecal coliform (CFU/100 mL) Measured downstream fecal coliform (CFU/100 mL) Fecal coliform upstream modeled for dilution without decay (CFU/100 mL) Distance between this plant and upstream plant (km) Fecal coliform decay rate (km1) Fecal coliform decay rate (d1)
Southington
Cheshire
Meriden
Wallingford
0.0% 6.2% 206 1213
24.2% 5.0% 7.5% 765 1387 1017
19.6% 6.3% 14.2% 679 2425 1193
10.8% 12.8% 16.2% 1454 5073 2209
6.8 0.042
7.6 0.074
11.3 0.037 1.52
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
outfalls (Cmeasured) and those anticipated if only dilution, but not removal, occurred between the outfalls, where x ¼ distance downstream (km), and k ¼ first order degradation rate (km1):
4781
model, alternative fecal coliform removal mechanisms must be important. Because these calculations already account for dilution, and as the winter sampling season likely reduced the importance of sunlight inactivation, predation or sedimentation of fecal coliform may have dominated this system.
Cmeasured ¼ Canticipated ekx
3.3. With a distance between the Southington and Cheshire wastewater treatment plant outfalls of 6.8 km, the estimated removal rate was 0.042 km1 (0.018 km1 standard deviation; Table 2). To estimate the error of this calculation, we used the 22% deficit between the measured total discharges of tributaries and wastewater effluents into the Quinnipiac, and the discharge recorded at the Wallingford USGS gauging station as an estimate of the relative error on discharge measurements. This value was combined with the geometric standard deviation calculations for fecal coliform concentrations to propagate the error through the calculations described above. The estimated removal rates between Cheshire and Meriden, over a distance of 7.6 km, and between Meriden and Wallingford, with a distance of 11.3 km, were 0.074 km1 (0.033 km1 standard deviation) and 0.037 km1 (0.012 km1 standard deviation), respectively. These rates were fairly similar. Although the reason for the somewhat higher estimated removal rate between the Cheshire and Meriden outfalls is unclear, any enhanced removal may be related to increased removal as the Quinnipiac River passed through Hanover Pond. Unfortunately, to compare these distance-based removal rates to the time-based rates obtained in previous laboratory batch experiments, a hydraulic model is needed to compute water travel times. Using an average flowrate of 6.5 m3/s within an existing HEC-RAS hydraulic model for the section of the Quinnipiac River between the Meriden and Wallingford wastewater treatment plant outfalls (FEMA, 2001), the predicted average travel time for water between the two treatment plant outfalls was 6.6 h; the selected flowrate lies within the range of discharges occurring within this section of the river during the sampling period (Table 1). The associated estimate for the time-based total fecal coliform removal rate between these outfalls would be 1.52 d1 (0.50 d1 standard deviation). Removal mechanisms may include sedimentation and die-off, including by predation or sunlight inactivation, although our data do not allow us to distinguish these mechanisms. Previous laboratory experiments involving batch cultures maintained at 15e25 C indicated initial die-off rates of E. coli of 0.48e1.44 d1 (Medema et al., 1997; Menon et al., 2003), although die-off rates declined over time (Petersen et al., 2005; Dutka and Kwan, 1980). While these rates are comparable to the estimated total removal rate in the Quinnipiac River, removal rates often decline with temperature (Easton et al., 2005; Medema et al., 1997; Dorner et al., 2004). Laboratory batch experiments conducted at 5 C, closer to the 2e5 C occurring in the river during sampling, observed E. coli removal rates of 0.24 d1 in a river water, but only 0.023 d1 in a river water previously autoclaved to prevent predation (Medema et al., 1997). These die-off rates are at least 6 times lower than the estimated total fecal coliform removal rate observed in the river. As this difference is greater than could be attributed to uncertainty in the HEC-RAS
Environmental implications
At up to 5000 CFU/100 mL, the levels of fecal coliform measured in the river during the winter were comparable to the 500e2200 CFU/100 mL measured during the summer. During neither the summer nor the winter is it likely that concentrations of E. coli in the river would meet the 126 CFU/ 100 mL geometric mean or even the 576 CFU/100 mL single sample maximum levels designated by the CT DEP as recreational criteria for the Quinnipiac River (CT DEP, 2008). As part of the TMDL analysis, the CT DEP has focused on the reduction of non-point source loadings to meet the recreational criteria. Moreover, they assumed that meeting the recreational criteria during the summer season would also be protective during the winter season, believing that pathogens die-off more quickly at colder temperatures. However, fecal coliforms accumulated to significant concentrations within the river during the winter due to wastewater discharges. The reduction in non-point source loadings observed during the winter was counterbalanced by the fecal coliform loadings from non-disinfected municipal wastewater effluents. Accordingly, continuous wastewater disinfection may be necessary to meet recreational criteria throughout the year. Although not explicitly considered by the CT DEP as part of the TMDL analysis, commercial shellfish seeding operations are conducted year round at the mouth of the Quinnipiac River in New Haven harbor. The Connecticut Department of Agriculture monitors shellfish for shellfish diseases (e.g., MSX and Dermo), and requires that they be moved to approved offshore areas prior to harvest for times anticipated to allow for depuration of pathogens. However, viral concentrations in shellfish currently are not monitored, and the lack of disinfection of municipal wastewater effluents along the Quinnipiac River during the winter may burden commercial shellfish operations by contributing to shellfish contamination by pathogens. Because shellfish live for several years, they may accumulate pathogens to which they are exposed during the winter. Monitoring of shellfish exposure to pathogens that employs indicator bacterial levels may not be conservative. Shellfish may be particularly prone to infection by wastewater-derived viruses, such as Norovirus (Ueki et al., 2005). Previous research has indicated that depuration rates of viruses may be lower than that of E. coli, and that Norovirus contamination may be more prevalent in the winter (FormigaCruz et al., 2002).
4.
Conclusions
We observed an accumulation of fecal coliform along the Quinnipiac River to approximately 5000 CFU/100 mL during the winter, when disinfection of wastewater effluents is not practiced. Wastewater effluents were found to serve as the dominant source of fecal coliform loading. While
4782
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
concentrations were similar during the summer, when disinfection of wastewater effluents was practiced, municipal wastewater effluents were not significant contributors to fecal coliform loadings. During both seasons, fecal coliform concentrations were nearly an order of magnitude higher than those targeted for the river for recreational uses. Even though recreational usage of the river may decline significantly during the winter, the river discharges to New Haven harbor, which is employed for shellfish culture throughout the year. Achieving protection of recreational uses along the river as well as commercial shellfish operations at the mouth of the river may necessitate reduction of pathogen loadings both by control of non-point sources and by year-round disinfection of municipal wastewater effluents. Estimates of overall removal rates of fecal coliform within rivers are relatively rare. Correcting for dilution by tributaries and groundwater seepage, the total fecal coliform removal rate within a river section was estimated to be 1.52 d1. As this value was nearly 6-fold higher than those observed for bacterial cultures in batch laboratory experiments at the 5 C temperatures prevailing in the river during the winter, fecal coliform removal by sedimentation or predation may have been important.
Acknowledgements We would like to thank Ning Dai and Katherine McKinstry for help with measuring tributary discharges, the staff members at the Towns of Southington, Cheshire, Meriden and Wallingford municipal wastewater treatment plants for help with sample collection, and Mr. Christopher Ziemba and Drs. Jordan Peccia, and Emily Viau for help with the fecal coliform analyses, and Dr. Shimon Anisfeld for use of the flowmeter. We would like to thank Kristin Frank of the Connecticut Department of Agriculture for helpful discussions.
references
APHA, 1998. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association, American Water Works Association, Environmental Federation Publishers, Washington D.C. Boehm, A.B., Grant, S.B., Kim, J.H., Mowbray, S.L., McGee, C.D., Clark, C.D., Foley, D.M., Wellman, D.E., 2002. Decadal and shorter period variability of surf zone water quality at Huntington Beach, California. Environ. Sci. Technol. 36, 3885e3892. CT DEP, 2008. A total maximum daily load analysis for the Quinnipiac River regional basin. Conn. Department Environ. Prot. June 4, 2008; retrieved April 20, 2010. http://www.ct.gov/ dep/lib/dep/water/tmdl/tmdl_final/quinnipiac_tmdl_final.pdf. Donovan, E., Unice, K., Roberts, J.D., Harris, M., Finley, B., 2008. Risk of gastrointestinal disease associated with exposure to pathogens in the water of the lower Passaic River. Appl. Environ. Microbiol. 74, 994e1003. Dorner, S.M., Huck, P.M., Slawson, R.M., 2004. Estimating potential environmental loadings of Cryptosporidium spp. and Campylobacter spp. from livestock in the Grand River watershed, Ontario, Canada. Environ. Sci. Technol. 38, 3370e3380. Dutka, B.J., Kwan, K.K., 1980. Bacterial die-off and stream transport studies. Water Res. 14, 909e915.
Easton, J.H., Gauthier, J.J., Lalor, M.M., Pitt, R.E., 2005. Die-off of pathogenic E. Coli O157:H7 in sewage contaminated waters. J. Am. Wat. Res. Assoc., 1187e1193. FEMA, 2001. Hydraulic Model of the Quinnipiac River in Wallingford, Connecticut. Federal Emergency Management Agency. Fischer, H.B., List, J.E., Koh, C.R., Imberger, J., Brooks, N.H., 1979. Mixing in Inland and Coastal Waters. Academic Press, NY. Formiga-Cruz, M., Tofin˜o-Quesada, G., Bofill-Mas, S., Lees, D.N., Henshilwood, K., Allard, A.K., Conden-Hansson, A.C., Hernroth, B.E., Vantarakis, A., Tsibouxi, A., Papapetropoulou, M., Furones, M.D., Girones, R., 2002. Distribution of human virus contamination in shellfish from different growing areas in Greece, Spain, Sweden, and the United Kingdom. Appl. Environ. Microbiol. 68 (12), 5990e5998. Harrelson, C.C., Rawlins, C.L., Potyondy, J.P., 1994. Stream Channel Reference Sites: an Illustrated Guide to Field Technique. Gen. Tech. Rep. RM-245. U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. Hellweger, F.L., Masopust, P., 2008. Investigating the fate and transport of Escherichia coli in the Charles River, Boston, using high-resolution observation and modeling. J. Am. Wat. Res. Assoc. 44 (2), 509e522. LeGuyader, F., Neill, F.H., Estes, M.K., Monroe, S.S., Ando, T., Atmar, R.L., 1996. Detection and analysis of a small roundstructured virus strain in oysters implicated in an outbreak of acute gastroenteritis. Appl. Environ. Microbiol. 62, 4268e4272. Liu, L., Phanikumar, M.S., Molloy, S.L., Whitman, R.L., Shively, D.A., Nevers, M.B., Schwab, D.J., Rose, J.B., 2006. Modeling the transport and inactivation of E. coli and Enterococci in the near-shore region of Lake Michigan. Environ. Sci. Technol. 40, 5022e5028. Medema, G.J., Bahar, M., Schets, F.M., 1997. Survival of Cryptosporidium parvum, Escherichia coli, faecal enterococci and Clostridium perfrigens in river water: influence of temperature and autochthonous microorganisms. Water Sci. Technol. 35 (11e12), 249e252. Menon, P., Billen, G., Servaisa, P., 2003. Mortality rates of autochthonous and fecal bacteria in natural and aquatic ecosystems. Water Res. 37 (17), 4151e4158. NSSP, 2007. National Shellfish Sanitation Program. Guide for the Control of Molluscan Shellfish. retrieved April 20, 2010. http:// www.fda.gov/Food/FoodSafety/Product-SpecificInformation/ Seafood/FederalStatePrograms/ NationalShellfishSanitationProgram/ucm046353.htm. Petersen, T.M., Rifai, H.S., Suarez, M.P., Stein, A.R., 2005. Bacteria loads from point and nonpoint sources in an urban watershed. J. Environ. Eng. 131 (10), 1414e1425. Pope, M.L., Bussen, M., Feige, M.A., Shadix, L., Gonder, S., Rodgers, C., Chambers, Y., Pulz, J., Miller, K., Connell, K., Standridge, J., 2003. Assessment of the effects of holding time and temperature on Escherichia coli densities in surface water samples. Appl. Environ. Microbiol. 69, 6201e6207. Schreiber, I.M., Mitch, W.A., 2006. Occurrence and fate of nitrosamines and nitrosamine precursors in wastewaterimpacted surface waters using boron as a conservative tracer. Environ. Sci. Technol. 40, 3203e3210. Sinton, L.W., Hall, C.H., Lynch, P.A., Davies-Colley, R.J., 2002. Sunlight inactivation of fecal indicator bacteria and bacteriophages from waste stabilization pond effluent in fresh and saline waters. Appl. Environ. Microbiol. 68, 1122e1131. Standridge, J., Lesar, D.J., 1977. Comparison of four-hour and twenty-four-hour refrigerated storage of nonpotable water for fecal coliform analysis. Appl. Environ. Microbiol. 34, 398e402. Ueki, Y., Sano, D., Watanabe, T., Akiyama, K., Omura, T., 2005. Norovirus pathway in water environment estimated by genetic analysis of strains from patients of gastroenteritis, sewage, treated wastewater, river water and oysters. Water Res. 39, 4271e4280.
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Fecal coliform accumulation within a river subject to seasonally-disinfected wastewater discharges Azalea A. Mitch, Katherine C. Gasner, William A. Mitch* Department of Chemical Engineering, Yale University, Mason Lab 313b, 9 Hillhouse Avenue, New Haven, CT 06520, United States
article info
abstract
Article history:
As pathogen contamination is a leading cause of surface water impairment, there has been
Received 17 December 2009
increasing interest in the implications of seasonal disinfection practices of wastewater
Received in revised form
effluents for meeting water quality goals. For receiving waters designated for recreational
17 May 2010
use, disinfection during the winter months is often considered unnecessary due to reduced
Accepted 18 May 2010
recreational usage, and assumptions that lower temperatures may reduce pathogen
Available online 12 June 2010
accumulation. For a river subject to seasonal disinfection, we sought to evaluate whether fecal coliforms accumulate during the winter to concentrations that would impair river
Keywords:
water quality. Samples were collected from municipal wastewater outfalls along the river,
Fecal coliform
as well as upstream and downstream of each outfall during the winter, when disinfection
Seasonal disinfection
is not practiced, and during the summer, when disinfection is practiced. During both seasons, fecal coliform concentrations reached 2000e5000 CFU/100 mL, nearly an order of magnitude higher than levels targeted for the river to achieve primary contact recreational uses. During the summer, wastewater effluents were not significant contributors to fecal coliform loadings to the river. During the winter, fecal coliform accumulated along the river predominantly due to loadings from successive wastewater outfalls. In addition to the exceedance of fecal coliform criteria within the river, the accumulation of wastewaterderived fecal coliform along the river during the winter season suggests that wastewater outfalls may contribute elevated loads of pathogens to the commercial shellfish operations occurring at the mouth of the river. Reductions in fecal coliform concentrations between wastewater outfalls were attributed to dilution or overall removal. Combining discharge measurements from gauging stations, tributaries and wastewater outfalls to estimate seepage, dilution between wastewater outfalls was estimated, along with the percentage of the river deriving from wastewater outfalls. After accounting for dilution, the residual reductions in fecal coliform concentrations observed between outfalls were attributed to actual fecal coliform removal. The estimated rate of removal of 1.52 d1 was significantly higher than die-off rates determined by previous researchers at similarly low temperatures in laboratory batch experiments, indicating the potential importance of other removal mechanisms, such as predation or sedimentation. ª 2010 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: þ1 203 432 4386; fax: þ1 203 432 4387. E-mail address: [email protected] (W.A. Mitch). 0043-1354/$ e see front matter ª 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2010.05.060
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
1.
Introduction
Pathogens are amongst the most important causes of water body impairments in the United States. High levels of pathogenic bacteria found in impaired waterways present a health risk to bathers (Boehm et al., 2002; Liu et al., 2006), non-contact recreational users such as anglers and boaters (Donovan et al., 2008; Hellweger and Masopust, 2008), and consumers of raw or undercooked shellfish such as oysters (Ueki et al., 2005; LeGuyader et al., 1996; Formiga-Cruz et al., 2002). In addition to the health risks to humans, pathogen contamination can incur serious economic hardships on communities due to the closure of beaches or shellfish beds. Pathogen loadings derive from non-point source inputs from urban or agricultural activities (Petersen et al., 2005; Dorner et al., 2004), dry weather flows from storm sewers (Boehm et al., 2002), wet weather loadings from combined or sanitary sewer overflows (Donovan et al., 2008), or discharges from wastewater treatment plants. Attempts to develop water quality models to predict violations of pathogen standards have combined pathogen inactivation and hydrodynamic models. Inactivation models incorporate results obtained from laboratory (Medema et al., 1997) or mesocosm (Sinton et al., 2002; Dutka and Kwan, 1980; Easton et al., 2005) studies to predict pathogen die-off as a function of nutrient and toxin concentrations, pH, temperature, salinity, sunlight and predation. Hydrodynamic models are employed to track pathogen plumes and account for their dilution. These models assess advection and dispersion as well sedimentation and resuspension of pathogens from sediments (Hellweger and Masopust, 2008). In the cases of lakes and oceans, these models have also incorporated the effects of wind-driven currents and tides (Boehm et al., 2002; Liu et al., 2006). At their most complex, these models have solved two or three dimensional advection and dispersion equations within a finite element framework (Hellweger and Masopust, 2008; Liu et al., 2006). Evaluating the fate of pathogens discharged to rivers can be even more complex. In oceans, lakes (Liu et al., 2006) and dammed rivers that behave like lakes (Hellweger and Masopust, 2008), hydrodynamic models have tracked plumes associated with individual wastewater outfalls or other point discharges. Rivers, however, feature multiple pathogen loading inputs from both point and non-point sources along their course such that pathogens may accumulate within the same plume. Furthermore, water withdrawals along the river may remove pathogens while water inputs from uncontaminated tributaries and groundwater seepage may result in dilution. A previous study addressed these complications by tracking the transport downstream of a non-native bacterial culture spiked into a river; however, only presence/absence data were obtained (Dutka and Kwan, 1980). Seasonal disinfection of municipal wastewater effluents is permitted in many locations. Where the designated beneficial use of the receiving water is for recreational uses, regulatory agencies may deem disinfection of wastewater effluents during the winter months unnecessary, due to the reduction in recreational activities. However, the implications of these policies for water quality in both the primary receiving waters
4777
and in other waters to which these receiving waters discharge have not been a focus of previous research. In this paper, we evaluated the fate and transport of wastewater-derived fecal coliform within a river system. Designated for recreational use, the river system is subject to seasonal disinfection of municipal wastewater effluents. We measured fecal coliform accumulation along the river during the winter months, and evaluated overall fecal coliform removal rates within the river. Our results raised questions about the implications of seasonal disinfection for public health risks associated with recreational use along the river and shellfish harvesting activities conducted at the mouth of the river.
2.
Experimental section
2.1.
Study site
The Quinnipiac River is located in central Connecticut, originating in Deadwood Swamp north of the Town of Southington and flowing southerly for 61 km to New Haven, where it empties into Long Island Sound (Fig. 1). In a previous study, we had employed boron as a conservative tracer to evaluate the fate of wastewater-derived precursors of nitrosamine disinfection by-products within this river (Schreiber and Mitch, 2006). Registered and permitted activities include 103 water withdrawals and 19 discharges. Five of the dischargers are municipal wastewater treatment plants; from north to south: Southington, Cheshire, Meriden, Wallingford and North Haven (Schreiber and Mitch, 2006). Only the four most northerly wastewater treatment plants were evaluated for this study because the section of the river near the North Haven facility is influenced by tidal action. Eight tributaries empty into the Quinnipiac River between Southington and Wallingford (Table 1): Misery and Honeypot Brooks between the Southington and Cheshire outfalls, Broad, Sodom and Harbor Brooks between the Cheshire and Meriden outfalls, and Meetinghouse Brook and two small unnamed brooks between the Meriden and Wallingford outfalls. One active USGS gauging station is located upstream of the Southington outfall, prior to the conjunction of the Quinnipiac River with the Eightmile and Tenmile Rivers. An additional USGS gauging station is located just upstream of the Wallingford outfall. Shellfish breeding beds are located at the mouth of the Quinnipiac River in New Haven harbor.
2.2.
Regulatory framework
The Quinnipiac River falls under two types of regulatory frameworks. Under the authority of the Connecticut Department of Environmental Protection (CT DEP), the Quinnipiac River is designated for recreational uses including swimming, fishing and boating. Used as indicator bacteria, Escherichia coli (E. coli) levels in the river must remain below a geometric mean of 126 CFU/100 mL and a single sample maximum of 576 CFU/100 mL. To achieve these criteria, the CT DEP issued wastewater treatment plants National Pollutant Discharge Elimination System (NPDES) permits whereby fecal coliform levels, used as indicator bacteria, must be below a geometric
4778
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
Table 1 e Streamflow Measurements. Site
4/21/2010 m3/s
2/22/09-2/25/09 m3/s
0.96 0.87 1.49
0.96 0.87 1.49 0.29
Southington USGS Eightmile River Tenmile River Seepage Southington WWTP Discharge Miser Brook Honeypot Brook Seepage
0.38 0.21
0.24 0.38 0.21 0.34
Cheshire WWTP Discharge Broad Brook Sodom Brook Harbor Brook Seepage
0.16 0.05 0.38
0.13 0.16 0.05 0.38 0.38
Meriden WWTP Discharge Meetinghouse Brook Unnamed Brook #1 Unnamed Brook #2 Seepage
0.06 0.04 0.04
0.54 0.06 0.04 0.04 0.56
Wallingford USGS Wallingford WWTP Discharge Wallingford Downstream Total Flow Without Seepagea
7.47
7.12 0.29 7.41 5.56
a To the Wallingford USGS Station.
National Shellfish Sanitation Program (NSSP), which requires that the geometric mean of fecal coliform concentrations of waters in approved shellfish growing regions exhibit levels below 14 CFU/100 mL (NSSP, 2007). Shellfish harvesting in the Quinnipiac River and most of New Haven Harbor is prohibited, but the seeding of oysters and clams is allowed in these areas. After seeding, the clams and oysters from these prohibited areas must be transferred to approved areas that are located 4 miles offshore, which is believed to permit depuration of pathogens prior to harvest.
Fig. 1 e Sampling locations along the Quinnipiac River. , [ municipal wastewater treatment plant effluent. C [ sampling point on the river upstream or downstream of a wastewater outfall. SU [ Southington upstream, SE [ Southington effluent, SD [ Southington downstream, CU [ Cheshire upstream, CE [ Cheshire effluent, CD [ Cheshire downstream, MU [ Meriden upstream, ME [ Meriden effluent, MD [ Meriden downstream, WU [ Wallingford upstream, WE [ Wallingford effluent, WD [ Wallingford downstream. : [ USGS gauging station 01194500.
2.3.
mean of 200 CFU/100 mL and a single sample maximum of 400 CFU/100 mL. However, disinfection of wastewater effluent is only required between May 1 and September 30 (CT DEP, 2008). The CT DEP considered the summer months to be the most challenging under the assumption that indicator bacteria survive longer in waters with temperatures closer to human body temperature. However, the Quinnipiac River is currently considered impaired for recreational use due to elevated E. coli levels. The CT DEP considers the likely cause of these elevated levels to be storm sewers and non-point source runoff and has developed a Total Maximum Discharge Level (TMDL) analysis to reduce E. coli concentrations in the Quinnipiac River. Commercial shellfish breeding in the State of Connecticut occurs year round. The State Department of Agriculture participates in the U.S Food and Drug Administration’s
Samples were collected from the 4 most northerly municipal wastewater treatment plant effluents, as well as locations located immediately upstream and at least 100 m downstream of the discharges (Fig. 1). These sampling sites were a subset of the ones employed in our previous study (Schreiber and Mitch, 2006); the locations had been identified using the method of Fischer et al. (1979) to achieve mixing of discharges throughout the river’s cross-section. Samples were collected in fluorinated high-density polyethylene containers, and were stored at 4 C pending analysis. Samples were collected once per day from each site on 4 consecutive days during the winter (February 22e25, 2009), when disinfection was not being conducted at the wastewater treatment plants. Sampling over 4 consecutive days, approximately the time frame of travel down the river, enabled an evaluation of sample variability. During this period, non-point source inputs to the river were minimized as no precipitation occurred, although there was snow on the ground. The river sample temperatures ranged
Sample collection and analysis
4779
Results and discussion
3.1.
Fecal coliform
Fig. 2 presents the geometric mean and standard deviation of fecal coliform concentrations measured daily at each of the sampling locations over four consecutive days from February 22e25, 2009. No precipitation occurred immediately before or during the sampling period and no disinfection of wastewater effluents was practiced. The influence of non-point sources was minimal. The wastewater discharges were clearly the most significant loading to the river during this time, featuring concentrations ranging from 1800 to 17 000 CFU/100 mL. The relatively small error bars indicate that the concentrations were repeatable during the sampling period, resulting in a stable concentration profile along the river. Fecal coliform concentrations increased significantly along the river to reach approximately 5000 CFU/100 mL. Fecal coliform sampling conducted in the summer on August 24, 2009, when disinfection of wastewater effluents
60000 40000
5000 4000 3000 2000 1000 0
SU SE SD CU CE CD MU ME MD WU WE WD
20000
Fig. 2 e Geometric mean fecal coliform concentrations measured in wastewater effluents and the river (A) during the spring along the Quinnipiac River (CT). Sampling sites: Southington (S), Cheshire (C), Meriden (M), Wallingford (W) wastewater treatment plant effluents (E) or either upstream (U) or downstream (D) of the discharges. Samples were collected at each sampling location for four consecutive days: February 22e25, 2009. Error bars represent 1 geometric standard deviation.
was being practiced, indicated concentrations in the river ranging from approximately 500e2200 CFU/100 mL (Fig. 3). During both seasons, the concentrations were comparable, and similar to fecal coliform concentrations measured in other studies of impaired surface waters (Dutka and Kwan, 1980; Hellweger and Masopust, 2008; Liu et al., 2006; Donovan et al., 2008). However, during the summer lower concentrations were observed in the wastewater effluents than in the river, and upstream locations exhibited higher concentrations than downstream. As it had rained the day prior to sampling in the northern section of the sampling area, it is likely that non-point sources were the most significant contributors to fecal coliform loadings in the river. When pathogen loadings from municipal wastewater effluents are reduced by disinfection, previous studies have indicated the importance of pathogen loadings to surface waters from non-
2500 2000 1500 1000 500 0 SU SE SD CU CE CD MU ME MD WU WE WD
3.
80000
Sampling Site
Fecal Coliform (CFU/100 mL)
from 2 to 5 C. For fecal coliform samples, 3 dilutions were prepared and analyzed by the membrane filtration method in triplicate (APHA, 1998). The dilutions exhibiting colony counts closest to 20 CFU per plate were selected for counting. Deionized water controls exhibited no colonies. Geometric mean and geometric standard deviation calculations were employed. An additional day of sampling was conducted during the summer (August 24, 2009), when disinfection of wastewater effluent was being conducted. Although no precipitation occurred on that day, light rain had fallen on the previous day in Southington. The river sample temperatures ranged from 21 to 23 C. Grab samples were collected from the river at locations upstream and downstream of each of the 4 wastewater treatment plants. Whether grab samples capture diurnal variations in fecal coliform concentrations is a concern. Diurnal variations in riverine concentrations of wastewaterderived constituents are anticipated to be less severe than in wastewater effluents due to dispersion within the river and non-uniform spacing of discharges. For example, a morning pulse of fecal coliform discharged from an upstream wastewater treatment plant would be unlikely to meet a morning pulse from a downstream plant on following days due to non-uniform spacing of the discharges along the river. To evaluate the importance of diurnal variations within wastewater discharges, we collected a mixture of grab and 24-h composite final effluent samples at the treatment plants. No significant differences in fecal coliform concentrations were observed between the two types of samples. For example, grab samples collected from the Meriden treatment plant on Days 1 and 2 were 11 900 CFU/100 mL and 24 700 CFU/ 100 mL, respectively, while 24-h composite samples collected on Days 3 and 4 were 20 800 CFU/100 mL and 24 800 CFU/ 100 mL, respectively. Although the recommended holding time for fecal coliform analyses is 6 h, previous studies have indicated no significant differences in E. coli (Pope et al., 2003) or fecal coliform (Standridge and Lesar, 1977) concentrations for samples stored in the refrigerator for up to 24e48 h.
Fecal Coliform (CFU/100 mL)
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
Sampling Site Fig. 3 e Geometric mean fecal coliform concentrations measured on August 24, 2009 along the Quinnipiac River (CT). Sampling sites: Southington (S), Cheshire (C), Meriden (M), Wallingford (W) wastewater treatment plant effluents (E) or either upstream (U) or downstream (D) of the discharges. Error bars represent 1 geometric standard deviation of triplicate analyses.
4780
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
point sources and dry weather flows from storm drains (Hellweger and Masopust, 2008; Petersen et al., 2005; Dorner et al., 2004; Donovan et al., 2008).
3.2.
Dilution analysis
During the February sampling period, although there was an overall increase in fecal coliform concentrations proceeding downriver, there was a slight decline in fecal coliform concentrations between each of the wastewater treatment plant outfalls (Fig. 2). These declines could result either from dilution by tributaries or groundwater seepage, or from overall removal of fecal coliform within the river by processes including death, predation, and sedimentation. To distinguish these possibilities, dilution was quantified by measuring the discharges of all of the tributaries to the Quinnipiac River upstream of the Wallingford outfall on April 21, 2010. At each tributary, the depth (di) and flow velocity were measured at one-foot intervals across the tributary, according to the method of Harrelson et al. (1994). The average velocity at each one-foot cross-section (vi) was taken as the flow velocity measured at 60% of the total depth from the surface using a Marsh-McBirney Flowmate 2000 model electromagnetic velocity meter. The discharge of the tributary (Q) was calculated as (Table 1): Q¼
n X
di vi
1
The 0.96 m3/s discharge recorded by the Southington USGS gauging station (site number 01195490) was the same on April 21, 2010 as the average measured during the February 22e25, 2009 period over which fecal coliform samples were collected. Precipitation had not occurred within 2 d prior to either sampling period. Accordingly, we assumed that the discharges from the tributaries were comparable over the two periods. For the period February 22e25, 2009, the sum of the discharges at the Southington USGS gauging station, and all tributaries and wastewater discharges between this gauging station and the Wallingford USGS gauging station was 5.56 m3/s, a value within 22% of the 7.12 m3/s average discharge measured at the Wallingford USGS gauging station (site number 01196500) over this period. The difference between these values could arise from either the combined
error of the discharge measurements, or additional contributions by groundwater seepage. Assuming that the difference arose from seepage, seepage contributions within river sections were estimated by assuming that seepage occurred evenly along the river. Accordingly, the 1.56 m3/s difference between the discharge recorded and calculated at the Wallingford USGS gauging station was divided among the following river sections according to their length (Tables 1 and 2): Southington USGS gauging station to the Southington outfall, the Southington to Cheshire outfalls, the Cheshire to Meriden outfalls, the Meriden outfall to the Southington USGS gauging station. Using these discharges, the percentage dilution of the river occurring between outfalls ranged from 10.8% between the Meriden and Wallingford outfalls to 24.2% between the Southington and Cheshire outfalls (Table 2). The percentage of the river that derived from wastewater, calculated for locations upstream and downstream of each outfall, increased to 16.2% downstream of the Wallingford outfall. This value was comparable to the 16% measured on March 16e19, 2004, but lower than the 41% measured on August 10e13, 2004, the low flow season (Schreiber and Mitch, 2006). To evaluate total fecal coliform removal rates within the river between wastewater outfalls, the fecal coliform concentration measured just upstream of the downstream outfall was compared to the value anticipated at this location for the case where only dilution, not removal, occurred along this river section. The anticipated concentrations (Canticipated) were calculated as the flow-weighted averages of the concentrations measured just downstream of the upstream outfalls, and the dilution waters entering between the outfalls (Table 2); since no municipal wastewater plants discharge to the tributaries, the fecal coliform concentrations in the dilution waters were assumed similar to the 206 CFU/100 mL measured in the Quinnipiac upstream of the Southington outfall: Canticipated ¼
Cupstream Q upstream þ CSouthington Q dilution Q upstream þ Q dilution
Assuming that total fecal coliform removal rates between outfalls are first order, we used the following equation to estimate these removal rates using the measured fecal coliform concentrations just upstream of the downstream
Table 2 e Fecal Coliform Calculations. Municipal Wastewater Treatment Plant
Dilution between this plant and upstream plant Wastewater fraction upstream Wastewater fraction downstream Measured upstream fecal coliform (CFU/100 mL) Measured downstream fecal coliform (CFU/100 mL) Fecal coliform upstream modeled for dilution without decay (CFU/100 mL) Distance between this plant and upstream plant (km) Fecal coliform decay rate (km1) Fecal coliform decay rate (d1)
Southington
Cheshire
Meriden
Wallingford
0.0% 6.2% 206 1213
24.2% 5.0% 7.5% 765 1387 1017
19.6% 6.3% 14.2% 679 2425 1193
10.8% 12.8% 16.2% 1454 5073 2209
6.8 0.042
7.6 0.074
11.3 0.037 1.52
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
outfalls (Cmeasured) and those anticipated if only dilution, but not removal, occurred between the outfalls, where x ¼ distance downstream (km), and k ¼ first order degradation rate (km1):
4781
model, alternative fecal coliform removal mechanisms must be important. Because these calculations already account for dilution, and as the winter sampling season likely reduced the importance of sunlight inactivation, predation or sedimentation of fecal coliform may have dominated this system.
Cmeasured ¼ Canticipated ekx
3.3. With a distance between the Southington and Cheshire wastewater treatment plant outfalls of 6.8 km, the estimated removal rate was 0.042 km1 (0.018 km1 standard deviation; Table 2). To estimate the error of this calculation, we used the 22% deficit between the measured total discharges of tributaries and wastewater effluents into the Quinnipiac, and the discharge recorded at the Wallingford USGS gauging station as an estimate of the relative error on discharge measurements. This value was combined with the geometric standard deviation calculations for fecal coliform concentrations to propagate the error through the calculations described above. The estimated removal rates between Cheshire and Meriden, over a distance of 7.6 km, and between Meriden and Wallingford, with a distance of 11.3 km, were 0.074 km1 (0.033 km1 standard deviation) and 0.037 km1 (0.012 km1 standard deviation), respectively. These rates were fairly similar. Although the reason for the somewhat higher estimated removal rate between the Cheshire and Meriden outfalls is unclear, any enhanced removal may be related to increased removal as the Quinnipiac River passed through Hanover Pond. Unfortunately, to compare these distance-based removal rates to the time-based rates obtained in previous laboratory batch experiments, a hydraulic model is needed to compute water travel times. Using an average flowrate of 6.5 m3/s within an existing HEC-RAS hydraulic model for the section of the Quinnipiac River between the Meriden and Wallingford wastewater treatment plant outfalls (FEMA, 2001), the predicted average travel time for water between the two treatment plant outfalls was 6.6 h; the selected flowrate lies within the range of discharges occurring within this section of the river during the sampling period (Table 1). The associated estimate for the time-based total fecal coliform removal rate between these outfalls would be 1.52 d1 (0.50 d1 standard deviation). Removal mechanisms may include sedimentation and die-off, including by predation or sunlight inactivation, although our data do not allow us to distinguish these mechanisms. Previous laboratory experiments involving batch cultures maintained at 15e25 C indicated initial die-off rates of E. coli of 0.48e1.44 d1 (Medema et al., 1997; Menon et al., 2003), although die-off rates declined over time (Petersen et al., 2005; Dutka and Kwan, 1980). While these rates are comparable to the estimated total removal rate in the Quinnipiac River, removal rates often decline with temperature (Easton et al., 2005; Medema et al., 1997; Dorner et al., 2004). Laboratory batch experiments conducted at 5 C, closer to the 2e5 C occurring in the river during sampling, observed E. coli removal rates of 0.24 d1 in a river water, but only 0.023 d1 in a river water previously autoclaved to prevent predation (Medema et al., 1997). These die-off rates are at least 6 times lower than the estimated total fecal coliform removal rate observed in the river. As this difference is greater than could be attributed to uncertainty in the HEC-RAS
Environmental implications
At up to 5000 CFU/100 mL, the levels of fecal coliform measured in the river during the winter were comparable to the 500e2200 CFU/100 mL measured during the summer. During neither the summer nor the winter is it likely that concentrations of E. coli in the river would meet the 126 CFU/ 100 mL geometric mean or even the 576 CFU/100 mL single sample maximum levels designated by the CT DEP as recreational criteria for the Quinnipiac River (CT DEP, 2008). As part of the TMDL analysis, the CT DEP has focused on the reduction of non-point source loadings to meet the recreational criteria. Moreover, they assumed that meeting the recreational criteria during the summer season would also be protective during the winter season, believing that pathogens die-off more quickly at colder temperatures. However, fecal coliforms accumulated to significant concentrations within the river during the winter due to wastewater discharges. The reduction in non-point source loadings observed during the winter was counterbalanced by the fecal coliform loadings from non-disinfected municipal wastewater effluents. Accordingly, continuous wastewater disinfection may be necessary to meet recreational criteria throughout the year. Although not explicitly considered by the CT DEP as part of the TMDL analysis, commercial shellfish seeding operations are conducted year round at the mouth of the Quinnipiac River in New Haven harbor. The Connecticut Department of Agriculture monitors shellfish for shellfish diseases (e.g., MSX and Dermo), and requires that they be moved to approved offshore areas prior to harvest for times anticipated to allow for depuration of pathogens. However, viral concentrations in shellfish currently are not monitored, and the lack of disinfection of municipal wastewater effluents along the Quinnipiac River during the winter may burden commercial shellfish operations by contributing to shellfish contamination by pathogens. Because shellfish live for several years, they may accumulate pathogens to which they are exposed during the winter. Monitoring of shellfish exposure to pathogens that employs indicator bacterial levels may not be conservative. Shellfish may be particularly prone to infection by wastewater-derived viruses, such as Norovirus (Ueki et al., 2005). Previous research has indicated that depuration rates of viruses may be lower than that of E. coli, and that Norovirus contamination may be more prevalent in the winter (FormigaCruz et al., 2002).
4.
Conclusions
We observed an accumulation of fecal coliform along the Quinnipiac River to approximately 5000 CFU/100 mL during the winter, when disinfection of wastewater effluents is not practiced. Wastewater effluents were found to serve as the dominant source of fecal coliform loading. While
4782
w a t e r r e s e a r c h 4 4 ( 2 0 1 0 ) 4 7 7 6 e4 7 8 2
concentrations were similar during the summer, when disinfection of wastewater effluents was practiced, municipal wastewater effluents were not significant contributors to fecal coliform loadings. During both seasons, fecal coliform concentrations were nearly an order of magnitude higher than those targeted for the river for recreational uses. Even though recreational usage of the river may decline significantly during the winter, the river discharges to New Haven harbor, which is employed for shellfish culture throughout the year. Achieving protection of recreational uses along the river as well as commercial shellfish operations at the mouth of the river may necessitate reduction of pathogen loadings both by control of non-point sources and by year-round disinfection of municipal wastewater effluents. Estimates of overall removal rates of fecal coliform within rivers are relatively rare. Correcting for dilution by tributaries and groundwater seepage, the total fecal coliform removal rate within a river section was estimated to be 1.52 d1. As this value was nearly 6-fold higher than those observed for bacterial cultures in batch laboratory experiments at the 5 C temperatures prevailing in the river during the winter, fecal coliform removal by sedimentation or predation may have been important.
Acknowledgements We would like to thank Ning Dai and Katherine McKinstry for help with measuring tributary discharges, the staff members at the Towns of Southington, Cheshire, Meriden and Wallingford municipal wastewater treatment plants for help with sample collection, and Mr. Christopher Ziemba and Drs. Jordan Peccia, and Emily Viau for help with the fecal coliform analyses, and Dr. Shimon Anisfeld for use of the flowmeter. We would like to thank Kristin Frank of the Connecticut Department of Agriculture for helpful discussions.
references
APHA, 1998. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association, American Water Works Association, Environmental Federation Publishers, Washington D.C. Boehm, A.B., Grant, S.B., Kim, J.H., Mowbray, S.L., McGee, C.D., Clark, C.D., Foley, D.M., Wellman, D.E., 2002. Decadal and shorter period variability of surf zone water quality at Huntington Beach, California. Environ. Sci. Technol. 36, 3885e3892. CT DEP, 2008. A total maximum daily load analysis for the Quinnipiac River regional basin. Conn. Department Environ. Prot. June 4, 2008; retrieved April 20, 2010. http://www.ct.gov/ dep/lib/dep/water/tmdl/tmdl_final/quinnipiac_tmdl_final.pdf. Donovan, E., Unice, K., Roberts, J.D., Harris, M., Finley, B., 2008. Risk of gastrointestinal disease associated with exposure to pathogens in the water of the lower Passaic River. Appl. Environ. Microbiol. 74, 994e1003. Dorner, S.M., Huck, P.M., Slawson, R.M., 2004. Estimating potential environmental loadings of Cryptosporidium spp. and Campylobacter spp. from livestock in the Grand River watershed, Ontario, Canada. Environ. Sci. Technol. 38, 3370e3380. Dutka, B.J., Kwan, K.K., 1980. Bacterial die-off and stream transport studies. Water Res. 14, 909e915.
Easton, J.H., Gauthier, J.J., Lalor, M.M., Pitt, R.E., 2005. Die-off of pathogenic E. Coli O157:H7 in sewage contaminated waters. J. Am. Wat. Res. Assoc., 1187e1193. FEMA, 2001. Hydraulic Model of the Quinnipiac River in Wallingford, Connecticut. Federal Emergency Management Agency. Fischer, H.B., List, J.E., Koh, C.R., Imberger, J., Brooks, N.H., 1979. Mixing in Inland and Coastal Waters. Academic Press, NY. Formiga-Cruz, M., Tofin˜o-Quesada, G., Bofill-Mas, S., Lees, D.N., Henshilwood, K., Allard, A.K., Conden-Hansson, A.C., Hernroth, B.E., Vantarakis, A., Tsibouxi, A., Papapetropoulou, M., Furones, M.D., Girones, R., 2002. Distribution of human virus contamination in shellfish from different growing areas in Greece, Spain, Sweden, and the United Kingdom. Appl. Environ. Microbiol. 68 (12), 5990e5998. Harrelson, C.C., Rawlins, C.L., Potyondy, J.P., 1994. Stream Channel Reference Sites: an Illustrated Guide to Field Technique. Gen. Tech. Rep. RM-245. U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. Hellweger, F.L., Masopust, P., 2008. Investigating the fate and transport of Escherichia coli in the Charles River, Boston, using high-resolution observation and modeling. J. Am. Wat. Res. Assoc. 44 (2), 509e522. LeGuyader, F., Neill, F.H., Estes, M.K., Monroe, S.S., Ando, T., Atmar, R.L., 1996. Detection and analysis of a small roundstructured virus strain in oysters implicated in an outbreak of acute gastroenteritis. Appl. Environ. Microbiol. 62, 4268e4272. Liu, L., Phanikumar, M.S., Molloy, S.L., Whitman, R.L., Shively, D.A., Nevers, M.B., Schwab, D.J., Rose, J.B., 2006. Modeling the transport and inactivation of E. coli and Enterococci in the near-shore region of Lake Michigan. Environ. Sci. Technol. 40, 5022e5028. Medema, G.J., Bahar, M., Schets, F.M., 1997. Survival of Cryptosporidium parvum, Escherichia coli, faecal enterococci and Clostridium perfrigens in river water: influence of temperature and autochthonous microorganisms. Water Sci. Technol. 35 (11e12), 249e252. Menon, P., Billen, G., Servaisa, P., 2003. Mortality rates of autochthonous and fecal bacteria in natural and aquatic ecosystems. Water Res. 37 (17), 4151e4158. NSSP, 2007. National Shellfish Sanitation Program. Guide for the Control of Molluscan Shellfish. retrieved April 20, 2010. http:// www.fda.gov/Food/FoodSafety/Product-SpecificInformation/ Seafood/FederalStatePrograms/ NationalShellfishSanitationProgram/ucm046353.htm. Petersen, T.M., Rifai, H.S., Suarez, M.P., Stein, A.R., 2005. Bacteria loads from point and nonpoint sources in an urban watershed. J. Environ. Eng. 131 (10), 1414e1425. Pope, M.L., Bussen, M., Feige, M.A., Shadix, L., Gonder, S., Rodgers, C., Chambers, Y., Pulz, J., Miller, K., Connell, K., Standridge, J., 2003. Assessment of the effects of holding time and temperature on Escherichia coli densities in surface water samples. Appl. Environ. Microbiol. 69, 6201e6207. Schreiber, I.M., Mitch, W.A., 2006. Occurrence and fate of nitrosamines and nitrosamine precursors in wastewaterimpacted surface waters using boron as a conservative tracer. Environ. Sci. Technol. 40, 3203e3210. Sinton, L.W., Hall, C.H., Lynch, P.A., Davies-Colley, R.J., 2002. Sunlight inactivation of fecal indicator bacteria and bacteriophages from waste stabilization pond effluent in fresh and saline waters. Appl. Environ. Microbiol. 68, 1122e1131. Standridge, J., Lesar, D.J., 1977. Comparison of four-hour and twenty-four-hour refrigerated storage of nonpotable water for fecal coliform analysis. Appl. Environ. Microbiol. 34, 398e402. Ueki, Y., Sano, D., Watanabe, T., Akiyama, K., Omura, T., 2005. Norovirus pathway in water environment estimated by genetic analysis of strains from patients of gastroenteritis, sewage, treated wastewater, river water and oysters. Water Res. 39, 4271e4280.