Comparison of Enterococcm Measurements in Marine Beach and Bay
Samples by Quantitative (real-time) Polymerase Chain Reaction,
Membrane Filtration and Enterolert®
James A. Ferretti, USEPA Region 2
USEPA Region 2, Edison, NJ 08837
Hiep V. Iran, USEPA Region 2
USEPA Region 2, Edison, NJ 08837
Virginia K. Loftin, NJDEP
NJDEP, Trenton, NJ 08625
Becky Cosgrove, MCHD
Monmouth County Health Department, Freehold, NJ 07728
John Protonentis, OCHD
Ocean County Health Department, Toms River, NJ 08755
Revision 1.0
October 2008
-------
TABLE OF CONTENTS
LIST OF TABLES AND FIGURES 3
1.0 PROJECT SUMMARY 5
2.0 INTRODUCTION 6
3.0 METHODS AND MATERIALS 8
3.1 STUDY SITES 8
3.2 WATER SAMPLING 10
3.3 MICROBIOLOGICAL PROCEDURES 11
3.3.1 Method 1600, Membrane Filtration (MF) 11
3.3.2 Enterolert ® Method, IDEXX 12
3.4 qPCR PROCEDURES 13
3.4.1 Test Sample Filtration Procedure 13
3.4.2 Test Filter Sample Extraction Procedure 13
3.4.3 Enterococcus/aecalis culture Procedure 14
3.4.4 Calibrator and DNA extraction procedure 14
3.4.5 qPCR assay preparation and detection procedure 15
3.5 qPCR QUALITY CONTROL 16
3.6 qPCR DATA ANALYSIS 17
3.7 STATISTICAL ANALYSIS 18
4.0 RESULTS AND DISCUSSION 18
4.1 TEMPORAL FACTORS AFFECTING ENTEROCOCCUS 31
4.2 SPATIAL FACTORS AFFECTING ENTEROCOCCUS 34
4.3 PHYSIO-CHEMICAL PARAMETERS AFFECTING ENTEROCOCCUS....34
4.4 qPCR QUALITY CONTROL 35
5.0 CONCLUSIONS 37
6.0 ACKNOWLEDGEMENTS 38
7.0 REFERENCES 46
-------
LIST OF TABLES AND FIGURES
List of Tables
1. Summary of Arithmetic Mean Enterococcus Concentrations of Ocean and
Monmouth County Bathing Beach and Bay Samples using qPCR and Membrane
Filtration or Enterolert ®, June 18 - August 20, 2007
2a. Summary of Enterococcus beach and bay water analysis results from qPCR and
MF analysis, Monmouth County, NJ
2b. Summary of Enterococcus beach and bay water analysis results from qPCR and
MF analysis, Ocean County, NJ
3a-3c Summary of Chemical and Physical Parameters For Each Sampling Visit at
Beaches and Bays in Monmouth County, New Jersey, June - August 2007.
4a-4c Summary of Chemical and Physical Parameters For Each Sampling Visit at
Beaches and Bays in Ocean County, New Jersey, June - August 2007.
List of Figures
1. Map of Beach and Bay Sampling Stations Sampled Between
June 18 - August 20, 2007, Monmouth and Ocean Counties, New Jersey
2a-2d. Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at a) Army Recreation Beach; b) Surf Beach; c) Newark Ave.;
and d) The Terrace.
2e-2h. Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at e) Village Beach Club; f) Cedar Ave.; g) Broad Street; and
h) Brown Avenue.
2i-2j. Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at i) Recreation Center and j) Myron/Wilson Bay.
3a-3d. Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at a) Ocean Area 1; b) Sheridan Avenue; c) Broadway Ave.;
and d) Bay Beach.
3e-3h. Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at e) Windward Beach; f) Anglesea Avenue; g) Avon Road
West; and h) Money Island.
-------
3i -3j Geometric means of Enterococcus densities per lOOmL of water from all
sampling locations at i) Beachwood Beach and j) Central Avenue.
4. Scatter plot and regression analysis results of geometric mean Enterococcus
densities, Determined by MF (Method 1600) or Enterolert Methods vs.
Enterococcus determined by qPCR Method for all sampling visits to
beaches, bays, and environmental sampling areas in Ocean and Monmouth
Counties, NJ, June 18 - August 20, 2007.
5a-5b. Densities as geometric means (n=4) of Enterococcus using qPCR Method (CCE)
and Enterolert (MPN/100 mL) on samples collected at Central Avenue and
Beachwood Beach on 7/30/07.
5c. Densities as geometric means (n=4) of Enterococcus using qPCR Method (CCE)
and Enterolert ® (MPN/100 mL) on samples collected at Windward Beach on
7/30/07.
6a-6b. Comparison of qPCR results and MF results along a three station transect
sampled at 8:00 am and 2:00 pm on August 6, 2007 at Myron/Wilson sampling
area, Monmouth County.
-------
1.0 PROJECT SUMMARY
Cell densities of the fecal pollution indicator genus, Enterococcus, were determined by a
rapid (4 h or less) quantitative polymerase chain reaction (qPCR) analysis method in 50
mL water samples collected from recreational and environmental beaches and bays in
New Jersey during the summer of 2007. Measurements by qPCR were compared with
counts of Enterococcus colony-forming units (CPU) determined by Metho.d 1600,
membrane filter (MF) analysis using mEI agar and Most Probable Number per 100 mLs
(MPN/100 mL) using the defined substrate technology test Enterolert ® (IDEXX,
Westbrook, ME). At all sampling sites, the geometric means of ambient Enterococcus
concentrations in water samples (collected every two weeks at each sampling station for a
total of five events for each County), exhibited lognormal distributions over the study
period among all methods evaluated. The geometric means ranged from 1.3 to 37.7
calibrator cell equivalents (CCE) by qPCR analysis and 5.2 to 64.9 CPU by MF analysis
in Monmouth County Beach/Bay Samples (Af=200). The geometric means from the
samples collected in Ocean County were 1.3 to 357 cells by qPCR and 6.2 to 150
MPN/100 mL by Enterolert (#=196).
The sampling sites consisted of bathing beaches, bays, and environmental (non-
swimming) areas to target samples which would potentially exhibit a wide range of
Enterococcus concentrations. In general, when Enterococcus concentrations were low
using MF and Enterolert ®, qPCR results followed the same trend. Less than a 1 fold
difference was documented between qPCR and the corresponding conventional methods
in 19 of the 20 sites sampled. The endpoints for qPCR and MF/Enterolert are not directly
comparable. Estimates by qPCR are expressed as Calibrator Cell Equivalents, which is a
mathematical computation based on comparison of gene sequences recovered from the
test samples with those from spikes of known numbers of Enterococcus cells in similarly-
processed and analyzed calibrator samples. However, the results presented in this paper
do allow a comparison of relative trends in indicator densities determined by the different
methods. MPN and CPU per mL are the benchmark endpoints and comparison to CCE is
provided as a way to express the relative changes between qPCR and the conventional
methods. Regression analysis of these results showed a significant positive correlation
-------
between qPCR and MF/Enterolert ® methods with an overall correlation coefficient (r)
ofO.71.
Sampling was conducted between 6:00 am and 9:00 am during the study. There was one
sampling event in each County in the afternoon to characterize temporal differences in
Enterococcus concentrations among the methods. Also, sampling across a transect was
performed at one of the sampling stations in Ocean County. There were differences in
Enterococcus concentrations along the transect, but the differences were not statistically
significant. There were no defined trends in Enterococcus concentrations in the morning
versus the afternoon sampling events.
Before qPCR can be a stand alone technology for beach management decisions,
additional data regarding intra- and interlaboratory variability, especially use of different
qPCR platforms and reagents, must be evaluated. The qPCR protocols are more
complicated as compared to the traditional Membrane Filtration techniques, and a higher
level of expertise is needed to collect and evaluate the data. Other site-specific variables
such as precipitation, tide height, cloud cover, salinity, temporal and spatial variability
need to be further evaluated using qPCR technology at marine recreational areas.
Additional side by side comparisons of qPCR data and the conventional microbiological
procedures are recommended to further define and explain variability among the
methods. Also, there is a need to evaluate epidemiological data in conjunction with
qPCR data to help formulate appropriate risk values. Epidemiological studies are being
performed by USEPA as part of the NEEAR program (National Epidemiological and
Environmental Assessment of Recreation Water) study using qPCR data and Method
1600 MF procedures. The objective of the NEEAR program is to evaluate water quality
based on indicator densities determined by the qPCR method at a number of beaches and
use these data to establish risks for gastrointestinal illness and potentially other adverse
health effects associated with the recreational use of these waters.
2.0 INTRODUCTION
There is a need for more rapid methods for the determination of microbial water quality
at bathing beaches. It has been demonstrated that densities of the bacteria of the genus
-------
Enterococcus in both marine and freshwater samples are directly correlated with
gastroenteritis illness rates in exposed swimmers (Cabelli 1982, Dufour 1984). USEPA
requires that recreational waters across the United States be monitored routinely for
Enterococcus spp. and /or Escherichia coll. While neither of these organisms is
pathogenic, both are considered to be surrogates for the presence of bacterial and viral
pathogens found in fecal material. Currently, approved methods for measuring
concentrations of Enterococcus and E. coll in recreational waters include Membrane
Filtration (MF), Most Probable Number (Multiple Tube Fermentation (MTF) techniques
and Defined Substrate Technology (DST®) tests. Although these methods have been
refined over the years, results are not available for at least 18 hours. Due to the
fluctuating nature of microbial contamination, this delay makes it difficult for beach
managers to make decisions regarding beach closures and/or swimming restrictions. At
best, decisions are made using one day old information; or a decision regarding safe
beach usage is not made until results of a confirmation test are available, which may be
up to 72 hours after the initial "failed" test was sampled. Because microbial water quality
can change rapidly (Boehm et. al, 2002), guidelines based on indicator organisms that
require 18-24 hours to develop, may result in both unnecessary beach closings or
exposure of swimmers to poor microbial water quality. A recent study estimates that up
to 40% of beach closures are in error (Kim and Grant, 2004).
The use of qPCR assays has shown promise as an alternative technology for monitoring
microbial water quality at recreational beaches (Haugland, et. al. 2005; Wade, et. al,
2006). Primer sets and probes are available for the specific detection of Enterococcus as
well as other fecal indicator and pathogenic microorganisms using real time or
quantitative PCR (qPCR) (Ludwig and Schleifer (2000), Lyon (2001), Brinkman et al.
(2003), Foulds et al. (2002), Blackstone et al. (2003), Frahm and Obst (2003), Guy et al.
(2003), Noble et al. (2003). Protocols for qPCR are now available for quantifying
indicator bacteria in recreational waters in approximately 3-4 hours. Because these
methods provide a more rapid assessment of water quality, they have the potential to
improve the timeliness of decision making for those responsible for beach management
decisions. A positive correlation was observed between Enterococcus qPCR and the MF
-------
results at two freshwater beaches in a 10 week study conducted by Haugland et al., 2005.
The purpose of this study was to compare Enterococcus measurements using qPCR
technology to the MF and DST® (Enterolert) testing methods using marine beach and bay
waters in an attempt to evaluate a more rapid assessment protocol for recreational water
quality. The sensitivity, accuracy and precision of qPCR relative to established methods
was examined, as well as the ability of qPCR to measure ambient concentrations of
Enterococcus at varying levels in the marine environment. qPCR results were compared
with corresponding Enterococcus CFU and/or MPN counts obtained by USEPA Method
1600 and Enterolert® tests. Correlations between qPCR and MF/Enterolert were
performed as an initial assessment of the utility of qPCR for use in marine beach and bay
Enterococcus analysis. Within and between sample and within station variability were
assessed and compared for qPCR, MF (USEPA Method 1600) and the Enterolert®
techniques. Temporal and spatial variability were evaluated from results over the course
of the nine week study. Also, additional studies comparing qPCR measurements to
currently accepted MF and DST® values, as described in this report, are important for
further evaluating the applicability of qPCR as a tool for monitoring bathing beaches as
well as building public awareness and confidence in this new technology.
3.0 METHODS AND MATERIALS
3.1 Study Sites
Twenty beach or bay locations in Ocean and Monmouth Counties, New Jersey were
sampled once every two weeks between 6/18/07 and 8/20/07 for this study (Figure 1).
Sampling sites were selected based on historical microbiological monitoring data. The
primary objective of the experimental design was to establish sampling areas with a
-------
South'!
m
Sampling Locations
Monmouth County
1 Broad St., Keyport
2 Army Recreation Beach, Sandy Hook
3 Surf Beach, Sandy Hook
4 Recreation Center, Highlands
5 Cedar Ave., Allenhurst
6 Village Beach Club, Loch Arbour
7 MyronAWIsonBay, Neptune City
8 Nevsark Ave., Spring Lake
9 Browi Ave., Spring Lake
10 The Terrace, Sea Girt
Ocean County
11 Broadvwy, Pt. Pleasant Beach
12 Windward Beach, Brick
13 Central Ave., Island Heights
14 Sheridan Ave., Seaside Heights
15 Money Island, Toms River
16 AvonRd., Pine Beach
17 Beacrwood Beach West, Beach wood Borough j
18 Anglesea Ave., Ocean Gate
19 Ocean Area 1, Island Beach
20 Bay Beach, Bamegat
Nave
„-—
vV^
IrjR
!<*l
Tiraon I
"4;
Rums!
.1—'
-V*£-
jintPle
Beat
'Pod
xg;. >, ; ;-Y-Y>- v '^:>t: -V
**x/ ' ,v ; F'*)*''•,<;'- "f~:;-. 'is";;*.-1'.
•^ i>'.V1r4fc'*';"<'-'J".;-'"::V'
,- ^V\:>|!t:'^V;;pvfv:
Mttflfff;, ;J», ''I'tV"'''"'"'-/' * -%;-.,""'".': ;j''
JnAcres
Figure 1. Map of Beach and Bay Sampling Stations Sampled Between
June 18 -August 20,2007, Monmouth and Ocean Counties, New Jersey
-------
gradient of microbial densities to fully compare the sensitivities of the methods.
Historical data from the Ocean and Monmouth County beach monitoring program were
used to select the sampling stations (NJDEC CCMP Summary Report, 2005, 2006). The
study sites were comprised of open ocean bathing beaches, bay areas, or environmental
areas with documented levels of known microbial contamination. Environmental areas
were included because these areas were characterized as having higher Enterococcus
concentrations than the public bathing beach areas.
3.2 Water Sampling
Samples were collected following procedures outlined in NJDEP, Cooperative Coastal
Monitoring Program (CCMP), Quality Assurance Project Plan, FY07/FY08, and Section
12.1, Sample Collection; in Chapter IX (Public Recreational bathing) of the State
Sanitary Code, N.J.A.C. 8:26-1 et seq. (amended April 2004) and described briefly here.
Samples were collected in sterile HDPE containers in an area with a stabilized water
depth between the sampler's lower thighs and chest. The sample container (250 or 500
mL sterilized HDPE wide mouth jars, Nalgene® or equivalent) was placed approximately
8-12 inches below the water surface with the lid and stopper still attached. With the
collector's arms extended to the front, the container was held near its base and downward
at a 45-degree angle. The cap was removed and the container filled in one slow sweeping
motion. The mouth of the container was kept ahead of the collector's hand and the
container recapped while it is was still submerged. The cap remained submerged during
sample collection. Sample remaining from microbiological analysis was used for
turbidity and salinity analyses. A total of four independent (true) replicate samples were
collected at each station for MF, Enterolert ®, and qPCR water samples. Conventional
parameters (MF or Enterolert®) and qPCR analyses were analyzed from each replicate
bottle.
Time and date of sample collection, tidal conditions, air and water temperature, rainfall,
wind direction, and other general conditions were documented and recorded. Following
collection, all samples were placed in coolers with ice during transport to the laboratory
and stored at 1-5 °C prior to filtration in the laboratory. Sample filtration was performed,
-------
and MF and Enterolert® tests were initiated within 6 h of collection. The filters for the
qPCR analysis were frozen immediately at -20 to -70°C until analysis. The turbidity and
pH of each water sample were determined by standard methods, American Public Health
Association (20th Edition). Salinity was measured via conductance bridge (YSI, Model
85) or refractometer.
All samples were collected in the morning (between 6:00 and 9:00 am). There were
additional samples collected in the afternoon at three stations in Ocean County on
7/30/08 and one stations in Monmouth County on 8/6/08. All samples were collected
from a single point from each location with the exception of one sampling event across a
transect in Monmouth County established at the Myron Wilson Bay site. At this transect,
additional sampling stations approximately 40 meters away on each side of the existing
sampling station were established. Samples were collected at this transect on 8/6/07 at
approximately 8:00 am, and repeated at 2:00 pm.
3.3 Microbiological Procedures
3.3.1 Method 1600, Membrane Filtration (MF)
Enterococcus was enumerated by EPA Method 1600 on mEI agar plates, US EPA
(2002). Volumes of 10 mL from each water sample were filtered on 47-mm
diameter, 0.45 (am pore size, membrane filters (Millipore Corp., Billerica, MA).
The filters were incubated on plates of mEI agar for 24 hours at 41±0.5 °C before
determining colony numbers. Enterococcus by MF was expressed as CPUs per
100 mL of water. Monthly verification tests of 10 typical and 10 atypical were
performed for each batch of water samples collected over the nine week study
period. Each preparation of mEI agar was tested for performance (i.e., correct
enzyme reaction) using pure cultures of target and non-target organisms. Sterility
of the filters and phosphate-buffered water used for rinsing the filtration apparatus
was also tested with each batch of samples received by the laboratory.
BioBalls™, TCS Biosciences, LTD, which contain a certified number of
bacterium, were used routinely for determination of Ongoing Precision and
Recovery.
-------
3.3.2 Enterolert ® Method, IDEXX
Enterolert ® provides an MPN result based on the presence or absence of
fluorescence in 101 individual wells. Each well contains a sample-nutrient
indicator mixture. Enterolert analysis on all samples collected in Ocean County
was performed by the Ocean County Utilities Authority (OCUA), Bayville, New
Jersey. A 1:10 dilution of the test water sample was prepared (90 mLs of sterile
Buffer plus 10 mL of sample) in a sterile container. A package of powdered
Enterolert regent was then added to the container and the sample solution was
mixed and poured into a Quantitray, a sterile plastic disposable panel containing
101 wells. The tray was then mechanically sealed after distribution of the mixture
into the wells and incubated for 24 h at 41.0 ± 5°C. Enterolert ® uses a nutrient-
indicator to detect Enterococcus. This nutrient-indicator fluoresces when
metabolized by Enterococcus. Test results were read in a darkened room. The
tray was placed under a 365-nm-wavelenth UV light with a 6 W bulb as supplied
by IDEXX, Westbrook, Maine, and the number of positive wells was enumerated.
Any fluorescence in a well was considered a positive reaction for that well. MPN
tables were used, and the number of positive wells with a 10X dilution factor,
were used to determine the density of Enterococcus per 100 mL of sample.
Ongoing Precision and Recovery samples were performed on a monthly basis in
addition to the required media viability checks. The sterility of the sample
containers used was checked by OCUA prior to first use. Sterile water used for
the Enterolert® testing was evaluated once per lot prior to first use.
Enterolert ® is an EPA approved method for wastewaters and ambient waters. A
strong positive correlation with Membrane Filtration has been established for both
fresh and marine water samples using this method (Budnick, 1996). MF and
Enterolert ® results from Monmouth and Ocean County samples were treated
similarly and compared directly to the qPCR results.
-------
3.4 gPCR Procedures
DNA extraction, amplification and detection of Enterococcus were based on previously
reported protocols, Brinkman (2002) and Haugland (2005).
3.4.1 Test Sample Filtration Procedure
Fifty mLs of each test sample were filtered through a 0.4 micron, 47 mm diameter
polycarbonate filter fitted in a pre-sterilized disposable 250 mL filter funnel
within 6 hours of collection. The filter paper was folded in half and folded
longitudinally 2-3 more times before being placed into a 2.0 mL polycarbonate
preloaded bead tube (Gene-Rite S0201-50) using sterile forceps. The tubes with
the polycarbonate filter paper were frozen at -20 to -70 until ready for use in the
qPCR analysis.
3.4.2 Test Filter Sample Extraction Procedure
Salmon testes DNA extraction buffer was prepared in advance of the DNA
extraction procedure. Salmon testes DNA extraction buffer acts as an exogenous,
positive control and reference. Initially, the concentrated salmon testes DNA
(Sigma, D1626, and St. Louis, MO) was resuspended in water and was diluted
with AE buffer (Qiagen, Cat No. 19077, Valencia, CA) to obtain the target
concentration required for the procedure. 590 uL of a 0.2 ug/mL of salmon testes
DNA extraction buffer mix was added to 2.0 mL tubes containing silica beads
(GeneRite, #80205-50, North Brunswick, NJ) and the negative control filter blank
or test sample filter. The extraction tubes were subjected to bead beating in an
eight position mini bead beater (Biospec Corp., Bartlesville, OK) for 1 minute at a
rate of 5,000 rpm and were then centrifuged at 12,000 x g for 1 minute to pellet
the glass beads and debris. The DNA in the supernatants from the extraction
tubes was transferred to sterile 1.7 mL microcentrifuge tubes and then centrifuged
for additional 5 minutes at 12,000 x g to further remove any sediments. The final
genomic supernatant was either analyzed immediately or stored at -20 °C until
analysis by qPCR.
-------
3.4.3 Enterococcus faecalis culture procedure
A pure culture of E. faecalis, ATCC 29212, was inoculated in a 20 mL test tube
with 10 mL brain heart infusion broth (BHI, Difco, #Ref 237500, Sparks, MD)
and was incubated on a shaker for 24 ± 2 hours at 35 °C ± 0.5 °C. Also, an
uninoculated tube was placed in the incubator to test the sterility of BHI broth.
The cell culture was transferred to 15 mL conical tubes and centrifuged at 6000 x
g for 5 minutes to pellet the cells. The supernatant was discarded, and the cell
pellet was washed twice with 10 mL of a Ix phosphate-buffered saline (PBS)
(Invitrogen, Cat. No. 14190, Carlsbad, CA) and resuspended in 5 mL of Ix PBS
solution. The optical density of the E. faecalis cell suspension was quantified on
aNanoDrop spectrophotometer ND-1000 v3.3.1 (Wilmington, DE). The E.
faecalis cell suspension was divided into 6 microcentrifuge tubes, each one
containing 500 uL for preparation of purified genomic DNA standards. The
remaining cell suspension was dispensed by 10 uL aliquots into 100-200
microcentrifuge tubes, which were used to prepare calibrator samples for
subsequent qPCR analyses.
3.4.4 Calibrator and DNA extraction and preparation procedure
A 10 uL aliquot of E. faecalis cell suspension was spotted onto a blank
polycarbonate filter which was then transferred to an extraction tube containing
pre-loaded glass beads and 590 uL of 0.2 ug/mL Salmon testes DNA extraction
buffer as described in section 3.4.2. The tube was shaken by a mini-bead beater
for 1 minute at 5,000 rpm and then centrifuged at 12,000 x g for 1 minute to pellet
the glass beads and debris. The genomic DNA in the supernatants from the
extraction tubes was transferred to sterile 1.7 mL microcentrifuge tubes and then
centrifuged for additional 5 minutes at 12,000 x g. The final DNA-containing
supernatant was either analyzed immediately or stored at -20 °C until analysis.
The E. faecalis cell suspensions used to prepare the calibrator samples were also
used to create purified genomic DNA standards for E. faecalis. Two 500 uL
undiluted E. faecalis cell suspensions were placed into a 2.0 mL preloaded tube
-------
containing glass beads, extracted in the mini-bead beater, and centrifuged. The
supernatant was transferred to another tube. The genomic DNA supernatant was
then digested with 1 uL of 5 ug/uL RNase A (Sigma, R-4642, St. Louis, MO) for
1 hour at 35 °C. The RNase A was used to digest the RNA in the sample to
facilitate purification of the genomic DNA of E. faecalis. After RNase digestion,
the DNA was purified by DNA-EZ purification kit (GeneRite, K102-02C-50,
North Brunswick, NJ). The concentration of DNA was then measured on the
NanoDrop spectrophotometer. The DNA was considered to be acceptable if the
o reading was > 1.75.
3.4.5 qPCR assay preparation and detection procedure
Each reaction tube contained assay mix with a total volume of 25 uL. The qPCR
assay mix had the following components: 12.5 uL of TaqMan Universal Master
Mix (Applied Biosystems, Part Number 4304437); 2x concentrated; 1 uL of
forward primer (1 uM); 1 uL reverse primer (1 uM); 1.5 uL of a fluorogenic
probe (0.08 uM); 2.5 uL of 2 mg/mL ultra pure bovine serum albumin (Ambion,
Cat # AM2616); 1.5 uL of sterile water and 5 uL of diluted DNA template (5 fold
dilution). TaqMan Universal Master Mix consisted of AmpliTaqGold DNA
polymerase, AmpErase UNG, dNTPs with UTP, passive reference 1 and
optimized buffer components. The published primer sequences were ECST748F:
5'-AGAAATTCCAAACGAACTTG, ENC854R: 5'-
CAGTGCTCTACCTCCATCATT and GPL813TQ: 5'-6FAM-
TGGTTCTCTCCGAAATAGCTTTAGGGCTA-TAMRA was used for the probe
(Ludwig and Schleifer, 2000). These sequences are homologous to the large
subunit ribosomal RNA genes of all reported species within the Enterococcus
genera. Published primers and hybridization probe sequences for salmon DNA
assay were SketaF2: 50-GGTTTCCGCAGCTGGG for the forward primer;
SketaR3: 50-CCGAGCCGTCCTGGTCTA for the reverse primer; and SketaP2:
50-6FAM-AGTCGCAGGCGGCCACCGT- TAMRA for the probe. These
sequences are homologous to internal transcribed spacer region 2 of the ribosomal
RNA gene operon of chum salmon, Oncorhynchus keta, Domanico et al. (1997).
-------
Primers and fluorescently labeled probes were purchased from Applied
Biosystems Inc. (Foster City, CA).
Each reaction tube containing 25 uL of the assay mix was then placed in a Smart
Cycler II (Cephied, Sunnyvale CA) for 45 cycles under the following conditions:
2 min at 50 °C, 10 min at 95 °C, 15 s at 95 °C and 2 min at 60 °C. Cycle
threshold (Or) values were calculated by the instrument. Cycle threshold values
occurred when the amplification fluorescence growth curves crossed a threshold
of 8 units that was established for this qPCR method. CT values for each sample
were collected at the end of each run and saved in Excel format. A No Template
Control (NTC), which tests the assay mix for contamination, was included with
each batch of samples analyzed along with one field blank. Any positive
amplification for no-template control and filter blank samples were reanalyzed for
verification. A sample was considered below the limit of detection when the
fluorescence threshold was not reached within 45 cycles.
3.5 qPCR Quality Control
Maintaining a contamination free process and environment is an important component of
qPCR analysis. Decontamination of workstations, pipettes and equipment after each use
was performed using a 70% ethanol solution. Presterilized pipets manufactured with
aerosol resistant filters were used in the steps of the DNA testing process. An ultraviolet
workstation was used to maintain sterility of tips. Pre-sterilized disposable filters and
housings were used during the test sample filtration process to prevent cross-over
contamination. All reagents and supplies were tested and certified by the manufacture for
specificity, sensitivity and to be free of contamination. The primers and probes were
dispensed in small aliquots to avoid contamination and degradation.
Contamination or misleading qPCR results can be detected by using positive and negative
quality control samples, which were implemented throughout this study. A negative
control (method blank), was performed in the lab after sample collection to test for proper
filtering technique and reagent sterility. A method blank was performed by using
-------
DNase/RNase free sterile water as a sample, which was processed in parallel with the
water samples. Another negative control was a No Template Control (NTC) which was
used in each batch of samples tested by qPCR to verify the purity of the master mix,
reagents, and ensure no contamination occurred during the processing of the test samples.
A NTC consisted of DNase/RNase free sterile water or buffer.
The calibrator samples served as the positive control to validate that the master mix and
reagents were prepared properly by producing amplification of the target nucleic acid.
Salmon DNA, added to each sample, served as the sample processing control (SPC) to
indicate that there were no significant losses of DNA during sample extraction or PCR
interferences from the test samples.
3.6 qPCR Data Analysis
The amplification efficiency of the Enterococcus qPCR assay was determined as the first
step in the qPCR data analysis process. Amplification efficiency is defined as the rate at
which a PCR amplicon is generated, normally doubling during each cycle (Applied
Biosystem, 2004). The amplification efficiency is normally equal to 2, however, the
reagents, assay preparation, purity of the samples and the inherent features of the primers,
probe and target sequence can alter the efficiency to less than 2. Initially, purified and
quantified E. faecalis genomic DNA was serial diluted to estimated concentrations 4 x
104, 4 x 103, 4 x 102, 2 x 102, and 1 x 102 IsrRNA gene sequences per 5 uL. These
standards were analyzed by qPCR in triplicate. CT values were obtained, averaged and
subjected to regression analysis against the log 10-transformed target sequence per
reaction in order to obtain the equation of the line for the standard curve. The DNA
standard curve for this study was y = -3.44x + 35.792, where -3.44 is the slope. The
slope value from the standard curve was used to calculate the amplification efficiency
using the following formula AF = 10 A (1/ Hslope value). The calculated AF was 1.954. The
r2 value from the DNA standard curve used during this study was 0.999 . This
amplification efficiency value was used in conjunction with the comparative cycle
threshold method to estimate the target cell densities in the water filtrate extracts as
calibrator cell equivalents (CCE) as previously described (Applied Biosystem (2004),
-------
Haugland et al (1999) Haugland et al (2005). Target cell estimates in each test sample
were then multiplied by 2 to express results as CCE per 100 mL sample volume.
Five-fold dilutions of the test sample filters and calibration extracts were analyzed in this
study to minimize potential interferences from undiluted extracts of the saltwater
samples. The CT values for salmon DNA assay in water filter samples with higher than 3
CT units above the mean values from the calibration extracts were reanalyzed. If
reanalysis did not fall within expected results, the data may be qualified.
3.7 Statistical Analysis
Arithmetic and geometric means were calculated on all microbiological results collected
during the study. A Logio transformation was performed on all raw data. Standard
deviation between and within sampling visits was determined on the Log 10 transformed
data. Coefficient of Variation (C.V.) calculations of within sampling visit raw geometric
means were performed on data from each sampling station. A linear regression was
calculated using the geometric means of MF/Enterolert versus qPCR results. Two-way
unpaired t-tests were performed on samples collected in the morning and afternoon
(p=0.05) to determine significant differences. Differences in Enterococcus
concentrations across a sampling transect were tested for assumptions of Normality and
Variance using Log 10 transformed data. Significant differences of Enterococcus
concentrations across transects were determined using either Tukey's Method of Multiple
Comparisons (p=0.05, critical value = 4.49, parametric) or Kruskal Walis/Dunns Multiple
Comparison Test, (p=0.05; crit value = 2.936, non-parametric).
4.0 RESULTS AND DISCUSSION
Sampling stations were selected to provide a range of Enterococcus concentrations.
Environmental stations, i.e., non-bathing beach stations, were included in the study
design because the established bathing beach sampling areas typically do not exhibit
Enterococcus concentrations near or above the single sample maximum water quality
criterion of 104 CFU/mL.
-------
Arithmetic mean Enterococcus concentrations using MF/Enterolert® ranged from 5.3 to
261 CPU or MPN/lOOmL in Ocean and Monmouth Counties over the entire sampling
period (Table 1). Arithmetic mean CCE values for qPCR analysis ranged from 3.0 to 549
CCE/100 mL (Table 1).
Enterococcus were detected by either MF or Enterolert® methods in 60.4% of the
samples (Total N=404). Enterococcus DNA was detected in 77% of the samples via
qPCR. Samples with no detectable bacterium were assigned values of 5 CPU or MPN
per 100 mLs (equivalent to one-half the detection limit of the MF and Enterolert
Methods), and those with no detectable CCE's were assigned 0.6 CCE for qPCR analysis,
for calculating the geometric means and standard deviations of individual sample results
for the study. Approximately 10% of the samples collected from Monmouth county
(n=204) resulted in a CPU concentration of 104 or greater. This percentage was doubled
in Ocean County where 41 out of the 200 samples analyzed for Enterococcus by MF
(20.5%) resulted in values greater than 104 MPN/lOOmL.
Figures 2a-j and 3a-j list the geometric means of Enterococcus densities at all sampling
locations measured over the course of the study. The charts are arranged from lowest to
highest geometric means for MF or Enterolert ®. In general, sampling areas with low
concentrations of Enterococcus as measured by MF or Enterolert ®, also had low levels
of Enterococcus via qPCR. At most stations with low levels of Enterococcus by the
culture methods, PCR results were lower due to the lower reporting limits for PCR. The
reporting limit for qPCR was one order of magnitude lower than MF and Enterolert®.
At stations with detectable levels of Enterococcus, qPCR concentrations were less than 1
order of magnitude from MF/Enterolert ® results (Figures 2 and 3). An important
observation is that the relative changes in MF or Enterolert® from week to week were
reflected in the qPCR values, which generally changed in the same direction and
magnitude.
-------
Table 1. Summary of Arithmetic Mean Enterococcus Concentrations of Ocean and Monmouth
County Bathing Beach and Bay Samples suing qPCR and Membrane Filtration or Enterolert®
June 18-August 20, 2007.
Sampling Area
Ocean Area 1
Sheridan Ave.
Bay Beach
Broadway
Windward Beach
Anglesea Ave.
Avon Road
Money Island
Beachwood Beach
Central Ave.
Army Rec Center
Surf Beach
Newark Ave.
The Terrace
Village Beach
Club
Cedar Ave.
Brown Ave.
Broad Street
Rec Center
Myron/
Wilson Bay
MF(CFU/100mL)or
Enterolert®
(MPN/100 mL)
n = 20
6.5
8.5
13.5
13.3
86.5
101
124
135
145
261
5.3
7.3
7.8
13.5
15.5
28.0
42.3
40.0
77
123
qPCR
(CCE/lOOmL)
n=20
3.3
7.9
9.0
9.1
47.8
160
156
157
184
549
3.0
8.8
9.2
4.6
23.2
30.2
15.6
45.8
73.0
79.6
County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Ocean County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Monmouth
County
Recreational
Or
Environmental
Station
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Environmental
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Recreational
Environmental
Recreational
Environmental
-------
1000 -i
a
o>
o
'k_
a!
E
o
a>
100
NJ
CO
o
^1
00
2
to
o
o
10
o
o
o
Sampling Date
a) Army Recreational Beach
Sampling Date
b) Surf Beach
E 1000 i
c
re
v
o
o
0)
O
100
10
o>
to
C£l
to
o
o
to
o
o
10
CO
to
o
o
22
to
o
i\5
o
o
Sampling Date
Sampling Date
c)Newark Avenue d) The Terrace
Figure 2a. - 2d. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at a) Army Recreation
Beach; b) Surf Beach; c) Newark Ave.; and d) The Terrace. The geometric means are calculated using four replicates for each
method. qPCR Results are designated with a (•) on a solid line; and MF results are designated with a (>) on a dotted line.
-------
• 1000 -
e
o
^ 100
ft
«
1 10
'C I
w
V
S 1-
^^—
^ * ^+^^~^
, ,
1000
•; 100
o
o
--J
10
O
O
fo
s
s
N)
O
3
Sampling Date
e) Village Beach Club
Sampling Date
J) Cedar Avenue
1000
1000
100
Sampling Date
Sampling Date
g) Broad Street
h) Brown Avenue
Figure 2e. - 2h. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at e) Village Beach
Club; f) Cedar Ave.; g) Broad Street; and h) Brown Avenue. The geometric means are calculated using four replicates for each
method. qPCR Results are designated with a (•) on a solid line; and MF results are designated with a (4) on a dotted line
-------
1000 n
100
1000 -,
100
Sampling Date
K>
to
NJ
O
o
Sampling Date
00
3>
ro
o
o
M
O
o
o
i) Recreation Center
j) Myron/Wilson Bay
Figure 2i. - 2j. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at i) Recreation Center
and j) Myron/Wilson Bay. The geometric means are calculated using four replicates for each method. No results reported on 7/23 for
the Recreation Center site due to non-quantified MF results. qPCR Results are designated with a (•) on a solid line; and MF results
are designated with a (+) on a dotted line.
-------
E 1000 -p
I
0)
I
O
Sampling Date
a) Ocean Area 1
10
1—
en
"*^
g
o
-4
~~— ^-"""
-4
IO
ro
o
o
^•^
^ * ^-,
v
g
o
3
-<^*
^
CO
0
IO
%
-4
#
m
oc
_]
fc
N
C
c
^
Sam pling Date
b) Sheridan Avenue
1000 -
100 •
10 •
1
k
k "-^
o> -J 2!
i s e
3^3
j*
CO
o
o
-4
...^
^
CO
c.
5
-g
1000
> 100
o
ID
O
Sampling Date
c) Broadway Avenue
Sampling Date
d) Bay Beach
Figure 3a. - 3d. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at a) Ocean Area 1; b)
Sheridan Avenue; c) Broadway Ave.; and d) Bay Beach. The geometric means are calculated using four replicates for each method.
qPCR Results are designated with a (•) on a solid line; and Enterolert® results are designated with a (4) on a dotted line.
-------
1000
1 100
I
I 10
1000
_l
I 100
c
10 :r
o>
oo
fo
o
o
o
o
o
-J
0>
fo
o
o
-J
Sampling Date
W
O
fo
o
o
-g
oo
w
fo
o
o
-J
ro
o
o
-g
M
O
o
-g
ro
o
o
-g
Sampling Date
ro
o
o
-g
S
o
o
-g
e) Windward Beach
J) Anglesea Avenue
1000
1000
t
I 100
(B
Ol
1 10
ro
o
o
Sampling Date
^J
o>
O
O
Sampling Date
CO
o
ro
o
o
10
o
o
g) Avon Road West
h) Money Island
Figure 3e. - 3h. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at e) Windward Beach;
f) Anglesea Avenue; g) Avon Road West; and h) Money Island. The geometric means are calculated using four replicates for each
method. qPCR Results are designated with a (•) on a solid line; and Enterolert® results are designated with a (4) on a dotted line.
-------
1000
100
1000
I) Beachwood Beach
j) Central Avenue
Figure 3i. - 3j. Geometric means of Enterococcus densities per lOOmL of water from all sampling locations at i) Beachwood Beach
and j) Central Avenue. The geometric means are calculated using four replicates for each method. qPCR Results are designated with
a (•) on a solid line; and Enterolert® results are designated with a (+) on a dotted line.
-------
The average geometric means at each sampling area are presented by county in Tables 2a
(Monmouth County) and 2b (Ocean County). The sampling areas are arranged in order
of low to high based on the MF/Enterolert ® geometric means from all analysis at each
area. A major objective of the sampling plan was to select stations which would be
represented by Enterococcm over a wide range of concentrations. qPCR geometric mean
concentrations were relatively consistent as compared to MF/Enterolert® at all sampling
areas. Each sampling area was ranked from 1 to 10 based on the enterococci geometric
means using MF or Enterolert (with 1 being the lowest geometric mean and 10 the
highest) (Tables2a. and 2b). qPCR relative ranks were within 2 ranks of Enterolert
concentrations at all Ocean County stations (Table 2b) and 3 ranks within MF results in
Monmouth County (Table 2a). In general, sampling areas with low concentrations of
enterococcus as measured by MF or Enterolert, also had low levels of enterococcus via
qPCR. In general, qPCR concentrations varied less than 50 percent compared with
corresponding MF/Enterolert® results (Figures 2 and 3). The only exception was Central
Avenue in Ocean County. The qPCR average geometric mean at this sampling area was
twice as high as Enterolert®.
A scatter plot and regression analysis of qPCR versus MF/Enterolert ® geometric mean
densities si Enterococcus from all sampling visits is presented in Figure 4. The overall
correlation coefficient (f) between the results of qPCR and MF/Enterolert ® was 0.71.
This strong correlation is similar to the value reported by Haugland et. al. 2005 in a qPCR
method comparison study of two freshwater bathing beaches.
-------
1000
0
o
c
cs
100
E
o
o>
O
LU
O
O
10
100
1000
CPU Geometric Mean / 100 ml_
Figure 4. Scatter plot and regression analysis results of geometric mean Enterococcus CPU densities (n=100),
Determined by MF (Method 1600) or Enterolert Methods vs. Enterococcus CCE, determined by qPCR for all sampling visits to
beaches, bays, and environmental sampling areas in Ocean and Monmouth Counties, NJ, June 18 - August 20, 2007.
-------
Table 2a. Summary of Enterococcus beach and bay water analysis results from qPCR and MF analysis,
Monmouth County, NJ, in increasing Membrane Filtration order.
Army Rec
Surf Beach
Newark Ave.
The Terrace
Village Beach
Club
Cedar Ave.
Brown Ave.
Broad Street
Rec Center
Myron/
Wilson Bay
Column ID
Average Geometric Mean
n=20
Membrane Filt.
(CFU/100 mL)
5.2
6.8
7.1
8.2
10.0
15.7
15.9
16.0
27.7
64.9
A
qPCR
(CCE/lOOmL)
1.3
2.7
3.5
1.5
12.1
9.4
2.9
15.4
21.2
37.7
B
Rank
(ILow- 10 High)
Enterolert
1
2
3
4
5
6
7
8
9
10
C
qPCR
1
3
5
2
7
6
4
8
9
10
D
LoglO SD
AMONG
Sampling Visits
Enterolert
.07
.03
.06
.40
.33
.46
.59
.47
.70
.57
E
qPCR
.49
.32
.42
.56
.38
.62
.51
.50
.65
.51
F
LoglO SD
WITHIN
Sampling Visits
Enterolert
.07
.15
.17
.12
.24
.27
.26
.38
.13
.23
G
qPCR
• .51
.69
.62
.36
.53
.63
.75
.57
.58
.43
H
C.V. WITHIN
Sampling Visits
Raw Data
Enterolert
.21
.34
.46
.59
.79
.62
.59
1.00
.30
.48
/
qPCR
1.43
1.20
.83
.87
.83
1.07
1.30
1.42
.41
.84
J
-------
Table 2b. Summary of Enterococcus beach and bay water analysis results from qPCR and MF analysis,
Ocean County, NJ, in increasing Membrane Filtration order.
Ocean Area 1
Sheridan
Ave.
Bay Beach
Broadway
Windward
Beach
Anglesea
Ave.
Avon Road
Money Island
Beachwood
Beach
Central Ave.
Column ID
Average Geometric Mean
n=20
Enterolert
(MPN/100 mL)
6.2
7.6
9.8
10.4
29.9
31.1
46.7
54.9
94.9
150.2
A
qPCR
(CCE/lOOmL)
1.3
2.3
3.1
2.2
27.7
27.5
72.0
65.4
132.7
357.1
B
Rank
(1 Low- 10 High)
MF
1
2
3
4
5
6
7
8
9
10
C
qPCR
1
3
4
2
6
5
8
7
9
10
D
LoglO SD
BETWEEN
Sampling Visits
MF
.11
.13
.23
.09
.68
.67
.62
.72
.40
.49
E
qPCR
.10
.53
.40
.58
.51
.77
.49
.69
.26
.49
F
LoglO SD
WITHIN
Sampling Visits
MF
.10
.16
.25
.31
.25
.23
.23
.19
.16
.21
G
qPCR
.51
.51
.58
.50
.46
.78
.40
.37
.27
.23
H
C.V. WITHIN
Sampling Visits
Raw Data
MF
.24
.44
.85
.83
.46
.59
.38
.21
.43
.40
/
qPCR
1.79
1.49
1.36
1.19
.43
1.53
.57
.35
.59
.31
J
-------
The standard deviation values shown in Tables 2a and 2b (Columns G & H) indicate that
qPCR within sampling visit variability was generally higher than MF/Enterolert. This
difference between the methods was consistent with the results of Haugland et. al. 2005
and was most pronounced at the locations with relatively low enterococci levels. The
differences in standard deviation between sampling visits was analogous between qPCR
and MF/Enterolert results (Tables 2a and 2b).
4.1 Temporal Factors Affecting Enterococcus Concentrations
Samples collected in the morning at three locations in Ocean County (July 30, 2007;
7:00 am) and one location in Monmouth County (August 6, 2007; 07:45 am) were
resampled in the afternoon (between 12:00 pm and 2:00 pm) to evaluate temporal factors
related to MF, Enterolert ®, and qPCR results. At Beachwood Beach in Ocean County,
both qPCR and Enterolert ® values were higher in the afternoon (Figure 5b.). A
contrasting result was observed at Central Avenue in Ocean County (Figure 5a.). At
Windward Beach in Ocean County, qPCR was higher in the afternoon sample, but
Enterolert ® results were lower in the afternoon (Figure 5c). Overall, there was not a
discernable trend with regard to Enterococcus densities between morning and afternoon.
Data from the BMP ACT Beaches Project (USEPA 2005), using E. coli and Enterococcus
by MF, found that indicator levels generally decreased by the afternoon at four of the
beaches studied, and there was no discernable trends at one freshwater beach which
generally had low levels of E. coli and Enterococcus at all times. Wade, et. al. (2006)
found Enterococcus concentrations were higher in the afternoon in a recent study of
bathing beach areas at the Great Lakes. There are many variables which could affect
temporal bacterial densities including cloud cover, tidal stage (and water level as
measured in feet above the mean low tide mark), wind direction and speed, water
temperature, bather density, sampling location (bay versus ocean area) and sources of
point and non-point source contributors of bacteria. Relative levels of bacteria in
morning versus afternoon need to be evaluated on a site by site basis. Tidal stage, as
measured by water level above mean low water mark, was seen as a significant
determinant of Enterococcus density within the swimming areas (Wymer, et. al. 2005).
-------
a) Central Avenue, Ocean County NJ
1000
_1
E
o
o
Is 100
Q.
z>
u.
O
o
_l
E
es
o
7 10 -
o
o
545
qPCR-AM qPCR-PM Enterolert®-AM Enterolert® - PM
b) Beachwood Beach, Ocean County NJ
10000
1615
o
o
o
Q.
ID
U.
O
o
o
-------
C) Windward Beach, Ocean County
1000
E
o
IK 100 -
LL.
O
_
E
§
a>
Q.
LLI
8
286
307
109
10
qPCR - AM qPCR - PM
Enterolert®- AM Enterolert® - PM
Figure 5 c. Densities as geometric means (n=4) of Enterococcus using qPCR Method
1606 (CCE) and Enterolert® (MPN/100 mL) on samples collected at Windward Beach
on 7/30/07. Morning samples collected between 6:25 and 6:29 am; Afternoon samples
collected at 1:00 pm.
-------
4.2 Spatial Factors Affecting Enterococcus Concentrations
Data from the EMPACT Study (Wymer, et. al. 2005) found that bacterial density was
similar along a beach front transect as long as the distance from shore was similar. In this
study, samples were collected at three stations along an approximate 80m transect in the
morning and afternoon. While there were some differences in CCE and CFU/MPN
values, the results were not significantly different along the transects (Figure 6a. and 6b.).
There are many factors which may affect patterns of bacterial contamination along a
transect. Tidal cycle, in-shore currents, point source discharges, and physical barriers in
or near the bathing area (i.e., sandbar, pier, and reef) will be a factor at some beaches and
not others. Determination of spatial variability was not a large part of the experimental
design of this project. USEPA, state environmental agencies, and local Health
Departments have guidelines established for determination of the most appropriate areas
to sample to be representative and protective at a certain bathing beach. However, spatial
variability is an important component to a beach monitoring program when evaluating a
new monitoring technology or protocol and will be included in the experimental design in
any follow-up studies.
4.3 Physio-Chemical Parameters Affecting Enterococcus Concentrations
Turbidity was low for all samples collected throughout the study (Tables 3 and 4). Only
one station had a result which exceeded 10.1 NTUs (Recreation Center, Monmouth
County, 30.2 NTUs) (Table 3a). This station also was characterized by the highest levels
of Enterococcus by both qPCR and Enterolert ® (Table 3a).
Tables 3 and 4 summarize the physical/chemical data collected for this study. The data
are arranged in a row based on the date sampled. Differences or trends of these
measurements / observations can be evaluated versus changes in qPCR or
MF/Enterolert® results over the course of the study.
There was no rainfall encountered during any of the sampling days of the 10 week study
period, so differences in Enterococcus during wet weather conditions could not be
evaluated. Salinity, water temperature, air temperature, cloud cover, wind direction and
-------
wind speed, and tidal cycle were recorded every time test samples were collected (data
not presented). All of these parameters potentially can affect the concentrations of
bacteria present. There was insufficient data or differences in results for these parameters
to evaluate their effects on the Enterococcus measurements from this study and remain an
area of further investigation in any follow-up studies.
4.4 qPCR Quality Control
There were no qPCR sample extracts which failed to meet the salmon DNA assay based
acceptance criterion of being within 3 ct units of the mean of the control samples as
outlined in previous studies (Haugland et al. 2005).
There were 23 of 104 qPCR assays which exhibited a measurable amount of fluorescence
during the qPCR analysis of the no template control (NTC) samples. The NTC
contamination was not considered high enough to affect the quality of the associated test
results. Two of the contaminated no template controls had CT values of 35.59 and 35.64.
The remaining had CT values of 38.22 to 43.35. In this study the quality control criterion
that was adopted was that for any NTC CT value lower than 35, the analysis should be
repeated and, if necessary, a new master mix should be prepared. However, all of the CT
values for the NTC in our study were greater than 35.
AE buffer was used as a filter blank and was processed with the test samples. One
filtration blank was performed for every 6 test samples collected in this study. One of 44
filtration blanks collected for the Monmouth County samples and 3 of the 36 filtration
blanks collected for the Ocean County samples exhibited a positive result for
Enterococcus during the qPCR analysis. The fluorescent signal was relatively low except
one of the qPCR positive filtration blanks.
35
-------
6a. qPCR Transect Results
100
in
E
0)
Q.
Ill
o
o
1
Station Left
Mid Station
Station Rieht
6b. Membrane Filtration Transect Results
100.T
Station Left
Mid Station
Station Riaht
Figures 6a. and 6b. Comparison of qPCR results and MF results along a three station
transect sampled at 8:00 am and 2:00 pm on August 6, 2007 at Myron/Wilson sampling
area, Monmouth County.
-------
5.0 CONCLUSIONS
qPCR was found to provide accurate and sensitive measurements of Enterococcus sp.
concentrations and was performed in less than 4 hours per sample. This study supports
the continued evaluation of qPCR as a potentially effective monitoring tool for bathing
beach management. Before a reliable and defensible qPCR risk value for Enterococcus is
developed, additional data are needed.
One perceived drawback to qPCR is the inability to distinguish between viable and non-
viable genetic target (PCR may amplify DNA from dead or non culturable target
organisms). However, a recent report by Wade et al. (2008) indicated that Enterococcus
measurements by qPCR were more strongly predictive of gastrointestinal illnesses than
CPU measures across four different potentially POTW-impacted freshwater beaches.
This finding could have been related to different relative impacts of the treated POTW
wastes on these beaches as well as the relative persistence of pathogens versus culturable
and non-culturable enterococci in these wastes and in other potential fecal sources. While
it is reassuring to demonstrate strong correlations between qPCR and the currently
accepted culture based methods, such relationships may not exist at all beaches.
There is a need to evaluate epidemiological data in conjunction with qPCR data to help
formulate appropriate risk values. Epidemiological studies are being performed by
USEPA as part of the NEEAR Program Study (National Epidemiological and
Environmental Assessment of Recreation Water) study using qPCR data and the Method
1600 MF procedure. The objectives of NEEAR program are to evaluate the water quality
at one or two beaches per year and ultimately obtain a new set of relationships between
health and water quality as determined by the qPCR results for Enterococcus and
potentially other fecal indicator organisms. Inter- and Inner laboratory variability, as well
as effects of physical and chemical parameters as they relate to qPCR performance is
needed. Protocol refinement and comparisons of the performance of the many different
qPCR platforms available are additional identified needs prior to establishment of water
quality criteria based on qPCR results.
-------
. In summary:
• Estimates of Enterococcus densities by the qPCR method and measurements by the
approved MF and Enterolert ® methods showed similar levels of between visit variability
and within visit variability was generally higher for qPCR results from 20 bay and ocean
recreational beaches and environmental sampling areas over a six week study period.
• A significant positive correlation was observed between the qPCR and both MF and
Enterolert ® results at all sampling areas, suggesting that the qPCR method has the
potential to be used as a tool for beach management.
• The qPCR protocols are more complicated and the procedures are more sophisticated as
compared to the traditional Membrane Filtration and defined substrate technology
procedures (i.e., Enterolert ®). A higher level of expertise and initial outlay of money for
equipment is needed to perform the qPCR analysis.
• Epidemiological data collected in conjunction with qPCR data is needed to formulate
appropriate risk values.
6.0 ACKNOWLEDGEMENTS
We wish to convey our gratitude to all of the individuals whose contributions were
invaluable for this project.
Debra Waller of NJDEP, and Donna Ringel and Sumy Cherukara of USEPA Region 2,
through their extensive experience in water QA and/or beach monitoring, provided
comments and guidance during development of the QAPP which contributed to the
success of this project; Rich Haugland, USEPA, ORD, Cincinnati, OH for his review of
our protocol and DNA calculations; Larry Wymer, USEPA, ORD, Cincinnati, OH and
Bob Stewart, SAIC (under EPA contract, Office of Environmental Information) for their
assistance and recommendations on portions of the statistical tests for this project; and
Erwin Smieszek, USEPA Region 2 for his preparation of the study area map.
38
-------
We wish to thank the scientists and field personnel for their contributions of collecting
the samples and performing the filtration, and analyses of the samples. A special thank
you to Robert N. Grant and Carol S. Conklin of the Ocean County Utilities Authority, as
the contract laboratory for the Ocean County Health Department and Ann Marie Fournier
of the Monmouth County Health Department for their organization of the fieldwork and
laboratory analysis for the sample filtration and/or performance of the Membrane
Filtration and Enterolert tests. The quality of the field work and sample filtration was
reflected in the excellent test and QC results documented during this study. The review
comments provided by Deborah Szaro, Marcus Kantz, Randy Braun, and Helen Grebe of
USEPA Region 2; and Robert N. Grant and Carol Conklin of OCUA were very much
appreciated and improved the quality of the manuscript. Finally the authors wish to
thank the management team of all the stakeholders for supporting this effort and
providing the resources to complete this study.
Mention of trade names or commercial products does not constitute endorsement or
recommendation by the EPA for use.
39
-------
Table 3a. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Monmouth County,
New Jersey, June - August 2007.
Monmouth County
Sampling Week #
qPCR
Geometric Mean
MF
Geometric Mean
Salinity
Turbidity
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
ARMY RECREATION BEACH
1
1
5
28.1
1.8
75°
65°
ND
high
SW
ND
8:20
2
2
5
27.3
2.4
87°
76°
ND
low
W
ND
8:30
3
3
6
28.0
5.72
63°
73°
ND
low
N
ND
8:30
4
1
5
27.0
3.21
75°
75°
ND
low
SW
ND
8:20
5
1
5
27.2
6.33
65°
70°
ND
low
E
ND
8:20
SURF BEACH
1
1
8
29.2
2
75°
65°
ND
high
SW
ND
7:40
2
2
8
27.4
3.3
87°
76°
ND
low
W
ND
7:30
3
18
6
27.8
7.35
63°
73°
ND
low
N
ND
7:45
4
2
6
25.7
5.44
75°
75°
ND
low
SW
ND
7:30
5
1
6
26.8
6.91
65°
70°
ND
low
E
ND
7:30
BROAD STREET
1
5
5
21.8
5.7
73°
70°
ND
high
NW
ND
7:30
2
41
10
23.4
5.4
78°
75°
ND
low
SSW
ND
7:30
3
58
72
22.3
12
64°
75°
ND
low
NNW
ND
7:30
4
3
7
23.5
4.62
75°
75°
ND
low
S
ND
8:40
5
26
41
20.9
5.32
65°
72°
ND
low
E
ND
7:30
RECREATION CENTER
1
6
7
23.6
4.8
73°
70°
ND
high
NW
ND
9:15
2
11
10
24.6
3.3
78°
75°
ND
low
SSW
ND
9:15
3
3479
>600
24.5
30.2
64°
75°
ND
low
NNW
ND
9:10
4
236
242
24
5.84
75°
75°
ND
low
S
ND
10:10
5
13
33
24.2
6.03
65°
72°
ND
low
E
ND
9:08
Table Notes:
Air and Water Temperature = "Fahrenheit; Turbidity = NTUs; ND = No Data; S = South, W= West, N= North, E= East; qPCR and MF/Enterolert were rounded to the nearest
whole number; Time Collected = All AM
Sampling Week 1 = 6/25/07; Week 2 = 7/9/07; Week 3=7/23/07; Week 4=8/6/07; Week 5=8/20/07
-------
Table 3b. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Monmouth County, New Jersey, June - August
2007.
Monmouth County
Sampling Week#
qPCR
Geomean
MF Geometric
Mean
Salinity
Turbidity
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
Cedar Ave
1
4
6
29.9
8.2
70°
68°
ND-
high
SW
ND
8:55
2
5
14
29.5
7.3
80°
68°
ND
low
WSW
ND
8:45
3
39
59
28.9
18
64°
71°
ND
low
N
ND
8:55
4
1
5
29.7
4.89
75°
75°
ND
low
S
ND
9:00
5
63
38
28.9
7.25
62°
70°
ND
low
ENE
ND
8:45
Village Beach Club
1
15
5
30
3.6
70°
68°
ND
high
SW
ND
8:40
2
37
6
28.3
3.9
80°
68°
ND
low
WSW
ND
9:00
3
20
29
28.7
11.2
64°
71°
ND
low
N
ND
8:40
4
2
6
29.1
4.37
75°
75°
ND
low
S
ND
8:45
5
11
20
28.9
10.8
62°
70°
ND
low
ENE
ND
9:00
Myron/Wilson Bay
1
9
29
28.3
3.4
70°
70°
ND
high
NW
ND
7:37
2
78
97
28.6
2.7
80°
70°
ND
low
W
ND
7:42
3
191
307
28.4
5.68
60°
70°
ND
low
WNW
ND
7:37
4
12
10
28
10.1
75°
75°
ND
low
SW
ND
7:38
5
48
132
26.8
6.91
65°
70°
ND
low
NE
ND
7:37
Newark Ave.
1
i
5
30.2
1.4
70°
71°
ND
high
WS
W
ND
9:05
2
3
7
30
1.3
80°
68°
ND
low
W
ND
7:20
3
7
7
28.6
4.38
62°
68°
ND
Low
NN
W
ND
9:20
4
1
7
29.2
1.67
72°
75°
ND
low
S
ND
9:05
5
18
10
28.4
5.08
62°
70°
ND
low
ENE
ND
7:26
Table Notes:
Air and Water Temperature = "Fahrenheit; Turbidity = NTUs
ND = No Data
S = South, W= West, N= North, E= East
qPCR and MF results were rounded to the nearest whole number
Time Collected = All AM
Sampling Week 1 = 6/25/07; Week 2 = 7/9/07; Week 3=7/23/07; Week 4=8/6/07; Week 5=8/20/07
41
-------
Table 3c. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Monmouth County,
New Jersey, June - August 2007.
Monmouth County
Sampling Week #
qPCR
Geomean
MF Geometric
Mean
Salinity
Turbiditv
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
Brown Ave
1
i
6
30.2
1.1
70°
71°
ND
high
WSW
ND
8:35
2
2
5
28
1.5
80°
68°
ND
low
W
ND
7:55
3
15
153
27.7
7.12
62°
68°
ND
low
NNW
ND
8:40
4
1
11
29.4
1.4
72°
75°
ND
low
S
ND
8:30
5
12
20
29.4
6.55
62°
70°
ND
low
ENE
ND
8:05
The Terrace
1
i
6
30.1
1.1
70°
71°
ND
high
WSW
ND
8:00
2
1
6
29.6
1.1
80°
68°
ND
low
W
ND
8:15
3
1
5
29.3
3.95
62°
68°
ND
low
NNW
ND
8:30
4
1
5
29.8
2.04
72°
75°
ND
low
S
ND
8:15
5
16
42
28.8
4.98
62°
70°
ND
low
ENE
ND
8:30
Table Notes:
Air and Water Temperature = "Fahrenheit; Turbidity = NTUs
ND = No Data
S = South, W= West, N= North, E= East
qPCR and MF results were rounded to the nearest whole number
Time Collected = All AM
Sampling Week 1 = 6/25/07; Week 2 = 7/9/07: Week 3=7/23/07; Week 4=8/6/07; Week 5=8/20/07
-------
Table 4a. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Ocean County, New
Jersey, June - August 2007.
Ocean County
Sampling Week #
qPCR
Geometric Mean
Enterolert
Geometric Mean
Salinity
Turbidity
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
Broadway
1
i
14
ND
4.2
73°
68°
No
ND
W
PC
7:01
2
1
7
34
2.4
62°
65°
No
ND
calm
S
7:02
3
2
9
34
10
78°
65°
No
ND
calm
C
6:58
4
1
9
34
3.5
73°
65°
No
ND
calm
C
6:57
5
27
14
32
1.1
70°
70°
No
ND
calm
C
6:42
Windward Beach
1
3
12
ND
3.3
71°
74°
No
ND
calm
PC
6:19
2
31
12
22
5.3
60°
67°
No
ND
calm
S
6:27
3
31
7
25
4.9
76°
76°
No
ND
calm
C
6:22
4
105
300
17
1.1
71°
75°
No
ND
calm
C
6:28
5
49
79
23
5.8
70°
75°
No
ND
calm
C
6:20
Sheridan Avenue
1
4
12
ND
3.4
73°
70°
No
ND
calm
PC
6:28
2
1
7
35
4.7
56°
62°
No
ND
calm
S
6:30
3
18
6
35
5
72°
70°
No
ND
calm
PC
6:25
4
1
5
34
7.2
70°
70°
No
ND
calm
C
6:35
5
2
10
32
2.4
75°
70°
No
ND
calm
C
6:31 •
Central Avenue
1
58
73
ND
3.4
70°
68°
No
ND
calm
S
6:28
2
257
51
16
6.7
59°
70°
No
ND
S
S
6:35
3
505
527
18
5.7
750
74°
No
ND
N
S
6:31
4
682
521
12
5.4
71°
73°
No
ND
Calm
C
6:33
5
1136
75
15
4.5
74°
73°
No
ND
Calm
C
6:35
Table Notes:
Air and Water Temperature = °Fahrenheit; Turbidity = NTUs; Salinity=ppth; ND = No Data; S = South, W= West, N= North, E= East;
S=Sunny, C=Cloudy, PC= Partly Cloudy; qPCR and Enterolert® were rounded to the nearest whole number; Time Collected = All AM; Need to Verify Rec. Center Salinity and
Turbidity
Sampling Week 1 = 6/18/07; Week 2 = 7/2/07; Week 3=7/16/07; Week 4=7/30/07; Week 5=8/13/07
-------
Table 4b. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Ocean County. New Jersey, June - August
2007.
Ocean County
Sampling Week #
qPCR
Geomean
Enterolert
Geomean
Salinity
Turbidity
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
Anglesea Avenue
1
3
7
ND
4
80°
77°
No
ND
calm
Clear
noon
2
14
12
18
4.8
59°
70°
No
ND
N
S
7:25
3
67
19
20
6.6
75°
74°
No
ND
N
S
7:27
4
516
385
10
4.2
71°
70°
No
ND
calm
C
7:26
5
11.6
48
18
3.5
74°
73°
No
ND
calm
C
7:24
Avon Road West
1
19
13
ND
4.5
80°
76°
No
ND
calm
Clear
11:45
2
50
37
13
4.9
59°
70°
No
ND
N
S
7:08
3
57
69
14
5
75°
74°
No
ND
N
S
7:07
4
512
462
5
1.4
71°
68°
No
ND
calm
C
7:08
5
71
14
13
4.7
74°
73°
No
ND
calm
C
7:07
Money Island
1
7
8
ND
3.6
70°
68°
No
ND
calm
S
6:15
2
35
12
16
5.3
59°
70°
No
ND
NE
S
6:09
3
47
76
17
6
75°
74°
No
ND
N
S
6:27
4
251
404
8
4.4
70°
73°
No
ND
calm
C
6:25
5
415
164
15
6
74°
73°
No
ND
calm
C
6:26
Beachwood Beach
l
99
75
ND
5.2
80°
73°
No
ND
calm
Clear
11:40
2
99
68
12
6.5
59°
70°
No
ND
N
S
7:03
3
141
117
12
5.1
75°
74°
No
ND
N
S
7:01
4
414
405
5
4
71°
68°
No
ND
calm
C
6:58
5
72
32
10
3.8
74°
73°
No
ND
calm
C
6:54
Table Notes:
Air and Water Temperature = "Fahrenheit; Turbidity = NTUs; Salinity=ppth
ND = No Data
S = South, W= West, N= North, E= East, S=Sunny, C=Cloudy, PC= Partly Cloudy
qPCR and Enterolert® were rounded to the nearest whole number
Time Collected = All AM
Need to Verify Rec. Center Salinity and Turbidity
Sampling Week 1 = 6/18/07; Week 2 = 7/2/07; Week 3=7/16/07; Week 4=7/30/07; Week 5=8/13/07
44
-------
Table 4c. Summary of Chemical and Physical Parameters For Each Sampling Visit at Beaches and Bays in Ocean County, New
Jersey, June - August 2007.
Ocean County
Sampling Week #
qPCR
Geomean
Enterolert
Geomean
Salinity
Turbiditv
Air
Temperature
Water
Temperature
Precipitation
Tidal Cycle
Wind Direction
Cloud Cover
Time Collected
Ocean Area 1
1
i
5
ND
4.4
77°
70°
No
ND
calm
C
8:01
2
1
6
34
1.4
75°
70°
No
ND
calm
Clear
8:21
3
3
6
34
4.6
74°
70°
No
ND
calm
C
8:30
4
2
5
34
1.1
70°
70°
No
ND
calm
C
7:57
5
1
10
32
1.8
75°
70°
No
ND
calm
C
9:03
Bay Beach
1
i
5
ND
3
70°
72°
No
ND
calm
clear
7:42
2
2
8
32
6
77°
71°
No
ND
calm
clear
7:28
3
8
7
33
5.7
80°
70°
No
ND
calm
clear
8:15
4
4
25
31
7.1
77°
70°
No
ND
calm
clear
7:37
5
10
12
31
6
82°
71°
No
ND
calm
C
730
Table Notes:
Air and Water Temperature = "Fahrenheit; Turbidity = NTUs; Salinity=ppth
ND = No Data
S = South, W= West, N= North, E= East, S=Sunny, C=Cloudy, PC= Partly Cloudy
qPCR and Enterolert® were rounded to the nearest whole number
Time Collected = All AM
Sampling Week 1 = 6/18/07; Week 2 = 7/2/07; Week 3=7/16/07; Week 4=7/30/07; Week 5=8/13/07
-------
7.0 REFERENCES
American Public Health Association, American Water Works Association and Water
Environment Federation, 1999. Standard Methods for the Examination of Water
and Wastewater, 20th Edition. American Public Health Association, Washington,
DC.
Applied Biosystems, 2004. Guide to Performing Relative Quantification of Gene
Expression Using Real-Time Quantitative PCR. Applied Biosystems Corporation,
Foster City, CA.
Blackstone, G.M., Nordstrom, J.L., Vickery, M.C., Bowen, M.D., Meyer, R.F. and
DePaola, A., 2003. Detection of pathogenic Vibrio parahaemolyticus in oyster
enrichments by real time PCR. J. Microbiol. Methods 53, 149-155.
Boehm, A.B., Grant, S.B., Kim, J.H., Mowbray, S.L., McGee, C.D., Clark, C.D., Foley,
D.M., and Wellman, D.E., 2002. Decadal and shorter period variability and surf
zone water quality at Huntington Beach, California. Environ. Sci. Technol. 36,
3885-3892.
Brinkman, N.E., Haugland, R.A., Wymer, L.J., Byappanahalli, M., Whitman R.L., and
Vesper, S.J., 2003. Evaluation of a rapid, quantitative real-time PCR method for
cellular enumeration of pathogenic Candida species in water. Appl. Environ.
Microbiol. 69, 1775-1782.
Budnick, G.E., Howard, R.T., and Mayo, D.R., 1996. Evaluation of Enterolert for
enumeration of enterococci in recreational waters. Appl. Environ. Microbiol., 62:
3881-3884.
Cabelli, V., Dufour, A.P., McCabe, L.J., and Levin, M.A., 1982. Swimming-associated
gastroenteritis and water quality, Am. J. Epidemiol. 115, 606-616.
-------
Domanico, M.J., Phillips, R.B. and Oakley, T.H., 1997. Phylogenetic analysis of Pacific
salmon (genus Oncorhynchus) using nuclear and mitochondrial DNA sequences.
Canad. J. Fisheries Aquat. Sci. 54,1865-1872.
Dufour, A.P., 1984. Health effects criteria for fresh recreational waters. EPA-600/1-84-
004, Office of Research and Development, US Environmental Protection Agency,
Cincinnati, OH.
Foulds, I.V., Granacki, A., Xiao, C., Krull, U.J., Castle, A. and Horgen, P.A., 2002.
Quantification of microcystin-producing cyanobacteria and E. coli in water by 5'-
nuclease PCR. J. Appl. Microbiol. 93, 825-834.
Frahm, E. and Obst, U., 2003. Application of the fluorogenic probe technique (TaqMan
PCR) to the detection of Enterococcus spp, and Escherichia coli in water
samples. J. Microbiol. Methods 52, 123-131 .
Guy, R.A., Payment, P., Krull, U.J., and Horgen, P.A., 2003. Real-time PCR for
quantification of Giardia and Crytosporidium in environmental water samples
and sewage, Appl. Environ. Microbiol. 69, 5178-5185.
Haugland, R.A., Vesper, S.J. and Wymer, L.J., 1999. Quantitative measurement of
Stachybotrys chartarum conidia using real time detection of PCR products with
the TaqMan™ fluorogenic probe system, Mol. Cell. Probes 13, 329-340.
Haugland, R.A., Varma, M., Wymer, L.J. and Vesper, S.J., 2004. Quantitative PCR
analysis of selective Aspergillus, Penicillium, and Paecilomyces Species.
Systematic and Applied Microbiology, 27. 198-210.
Haugland, R.A., Siefring, S.C., Wymer, L.J., Brenner, K.P., and Dufour, A.P., 2005.
Comparison of Enterococcus measurements in freshwater at two recreational
beaches by quantitative polymerase chain reaction and membrane filter culture
analysis. Water Research. 39, 559-568.
47
-------
Kim, J.H., Grant SB. 2004. Public mis-notification of coastal water quality: a
probabilistic evaluation of posting errors at Huntington Beach, California.
Environ Sci Technol 38(9):2497-2504.
Ludwig, W. and K.-H. Schleifer K.H., 2000. How quantitative is quantitative PCR with
respect to cell counts?, Syst. Appl. Microbiol. 23, 556-562.
Lyon, W.J. 2001. TaqMan PCR for detection of Vibrio cholerae Ol, O139, non-Ol, and
non-0139 in pure cultures, raw oysters, and synthetic seawater, Appl. Environ.
Microbiol. 67,4685-4693.
New Jersey Department of Environmental Protection. Division of Watershed
Management. Office of Watershed Education, Estuaries and Monitoring.
Cooperative Coastal Monitoring Program. Summary Report for 2005 and 2006
Noble, R.T., Allen, S.M., Blackwood, A.D., Chu, W., Jiang, S.C., Lovelace, G.L.,
Sobsey, M.D., Stewart, J.R. and Wait, D.A., 2003. Use of viral pathogens and
indicators to differentiate between human and non-human fecal contamination in
a microbial source tracking comparison study, J. Water Health 1, 195-207.
US Environmental Protection Agency, 2002. Method 1600: Enterococci in water by
membrane filtration using membrane-Enterococcus indoxyl-D-glucoside agar
(mEI), EPA 821/R-02/022, 2002. US Environmental Protection Agency, Office of
Water (4303T), Washington DC.
Wade, T.J., Calderon, R.L., Brenner, K.P., Sams, E., Beach, M., Haugland, R., Wymer,
L., and Dufour, A.P., 2008. High Sensitivity of Children to Swimming-
Associated Gastrointestinal Illness: Results Using a Rapid Assay of Recreational
Water Quality. Epidemiology. 19:375-383.
Wade, T.J., Calderon, R.L., Sams, E., Beach, M., Brenner, K.P., Williams, A.H., Dufour,
A.P., 2006. Rapidly measured indicators of recreational water quality are
48
-------
predictive of swimming-associated gastrointestinal illness. Environmental Health
Perspectives. 114. 24-28
Wymer, L.J., Dufour, A.P., Brenner, K.P., Martinson, J.W., Stutts, W.R., Schaub, S.A.,
2005. The BMP ACT beaches project. Results and recommendations from a study
on the microbiological monitoring of recreational waters. EPA 600/R-04/023,
2004. US Environmental Protection Agency, Office of Research and
Development, Washington DC.
49
------- |