EPA/600/J-03/262
PARTICLE-ASSOCIATED MICROORGANISMS IN
STORMWATER RUNOFF
Michael Borst1 and Ariamalar Selvakumar1*
JUrban Watershed Management Branch, United States Environmental Protection Agency (MS-
104), Edison, NJ 08837, USA
AbstractThis research investigated the effects of blending and chemical addition before
analysis of the concentration of microorganisms in stormwater runoff from a single summer
storm to determine whether clumped or particle-associated organisms play a significant role.
The standard membrane filtration method was used to enumerate the microorganisms. All
organisms, except for Escherichia coli, showed an increase in the measured concentration after
blending samples at 22,000 rpm with or without the chemical mixture. Other than fecal
streptococci, the organism concentrations decreased with the addition of the Camper's solution
in both blended and unblended samples before analyses. There was a statistically significant
interaction between the effects of Camper's solution and the effects of blending for all the
organisms tested, except for total coliform. Blending did not alter the mean particle size
significantly. The results show no correlation between increased total coliform, fecal coliform,
and fecal streptococcus concentrations and the mean particle size.
Keywordsstormwater runoff, particle-associated microorganisms, blending, Camper's
solution, particle size, chemical addition
*Author to whom all correspondence should be addressed. Tel.: 732-906-6990; fax: 732-321-
6640; e-mail: selvakumar.ariamalar@epa.gov
ID
-------
INTRODUCTION
According to the EPA's 1998 National Water Quality Inventory Report to Congress, about 40%
of assessed U.S. streams, lakes, and estuaries did not support the criteria for locally-designated
uses such as fishing and swimming. High bacteria concentrations in stormwater runoff from
agricultural and urban areas are a leading cause in the failures to meet designated use criteria
(USEPA, 2000). Investigators have documented large concentrations of fecal coliform and
pathogens such as Pseudomonas aeruginosa and Staphylococcus aureus in urban stormwater
(Oliveri et a/., 1977). Schillinger and Gannon (1982) reported that about 15 to 20 percent of
fecal coliform cells present in untreated stormwater are adsorbed to larger suspended particles,
most of which were greater than 30|im in diameter. They further noted that more than half the
organisms were not attached and remained suspended in water. Traditionally, monitoring and
research programs quantify the microorganism concentrations in samples using standard
methods (e.g., membrane filtration or multiple tube fermentation). By design, these methods
target public health and do not completely measure either clumped organisms or organisms
associated with particles; therefore, they may not fully enumerate the organism concentrations.
This research gauges the degree that clumped and particle-associated organisms exist in
stormwater runoff by using sample pretreatments such as blending, adding a mixture of
chemicals believed to help separate organisms from particulates, or both to estimate the degree
that these phenomena affect measurements of bacteria in stormwater.
The literature supports the phenomena of clumped/aggregated and particle-associated
organisms in drinking and municipal wastewaters and outlines general procedures to document
the severity by treating samples before traditional analysis. Conceptually, the pretreatment
processes use mechanical or chemical techniques to free bacteria from either paniculate matter
or other bacteria in the sample. Each separated organism forms a separate colony during
incubation, allowing more complete enumeration.
Researchers and practitioners have selected various blending conditions and studied
different organisms. Early investigations on primary clarified combined sewer overflow (CSO)
by Glover and Herbert (1973) and Moffa etal. (1975) established mechanical separation as a
technique to more completely enumerate the bacteria concentrations. Camper et al. (1985)
2D
-------
evaluated microorganism desorption from granular activated carbon using a mixture of
chemicals resulting in concentrations of 1(H> M Zwittergent 3-12, 10~3 M EGTA, and 0.01 M
Tris buffer with 0.01 wt% peptone and about pH 7. In these desorption experiments, blending
samples at 16,000 rpm for 3 minutes with the chemical mixture gave the highest recovery of
heterotrophic plate count (HPC) organisms from the spent carbon. This process increased the
measured HPC concentration as much as 50-fold and coliform concentrations as much as 1200-
fold in drinking water treated with carbon, compared with hand shaken samples. Parker and
Darby (1995), using multiple tube fermentation (MTF) methods to study secondary effluent
disinfection, found that blending samples with the same chemical mixture used by Camper et al.
at 19,000 rpm for 1.5 minutes produced the greatest recovery of particle-associated coliform
organisms. They concluded that particle association and organism shielding significantly affect
MTF coliform density measurements, and effluents may contain many more coliform bacteria
than measured using the standard enumeration procedure.
Perdek and Borst (2000) evaluated blending CSO samples and diluted sanitary sewage to
release particle-associated microorganisms before measuring microbial indicator concentrations
by membrane filtration (MF). After screening the samples to remove solids greater than 2 mm,
the samples were pretreated to reduce the concentrations of free-swimming microorganisms by
ultraviolet irradiation. The analysis of blended samples showed as much as a 10-fold increase in
measured fecal coliform and enterococcus concentrations, compared with concentrations
measured in the unblended samples without Camper's solution. These experiments showed that
a blending speed from 14,500 to 22,000 rpm and a blending time from 0.5 to 3 minutes affected
the measured concentrations. Samples blended for about 2 minutes at the highest speed
evaluated, 22,000 rpm, showed the greatest concentration increases. These results also showed
that the blending decreased mean particle size, but showed no correlation between increased
indicator microorganism concentration and decreased particle size. The New York City
Department of Environmental Protection (NYCDEP) blended raw treatment influent, primary
effluent, and chlorinated primary effluent and reported an increase in concentration with an
increase of blending time in the range of 0 to 60 seconds, with peak total coliform counts within
the first 10 seconds of blending (COM, 1997). Ridgway and Olson (1982) found that particle-
associated bacteria in drinking water were found mostly on particles greater than 10 : m in
3D
-------
diameter. Scheible et al. (1986), as part of a disinfection study, selected much shorter blending
times.
Although several researchers documented the release of particle-associated or clumped
organisms in drinking and wastewaters using blending and chemical addition before analysis,
none have focused on stormwater runoff, and few have studied a varied collection of organisms.
This research uses both techniques to decide whether clumped or particle-associated organisms
play a significant role in stormwater samples and if sample pretreatment using these techniques
will give a more thorough picture of the organism concentration. The research uses both
techniques (blending with and without the Camper's solution) because of the inconsistent
treatments reported in the literature and the lack of a clear mechanism for the release process.
MATERIALS AND METHODS
Sample collection
An automatic sampler (Model #900 max, American Sigma, Loveland, CO) collected a
flow-weighted stormwater sample from a 15-inch diameter, concrete storm sewer outfall. The
storm sewer drains a small, slightly sloping, high-density residential area in Monmouth County,
New Jersey. Earlier evaluations following the procedures developed by Pitt et al. (1993) showed
the storm sewer was unlikely to have sanitary cross connections. The automatic sampling began
when the flowing water depth in the storm sewer reached 2.54 cm. The sampler collected one
1-L sample after each 1,350 L of stormwater flow was measured by the attached flow meter
(Model #960, American Sigma, Loveland, CO). A calibrated peristaltic pump transferred the
samples to a precleaned (Standard Methods 9040), 5-gallon HDPE container. The sample was
collected during a rain event on July 10, 2000. The event produced 1.8 mm total rainfall over 74
minutes. Rainfall was recorded using a tipping bucket rain gage (Model #RGD-04,
Environmental Sensors, Inc., Escondido, CA), positioned near the sampler within the drainage
area. The runoff was slightly acidic (pH from 6.03 to 6.86), with a conductivity of 0.1 to 0.2 mS,
and a temperature from 20.5 to 23.8°C. The gage recorded no rain at the site during the
preceding 140 h. The nearly 6-day dry period should be sufficient for normal terrestrial build-up
processes, the net effect of time-dependant deposition and loss processes, to reach equilibrium
4D
-------
(Sartor and Boyd, 1972). The sample was recovered, placed in a cooler with ice and transported
to the laboratory for processing.
Experimental methods
In the laboratory, the sample was thoroughly mixed by shaking the container and was
divided into six subsamples. The subsamples were prepared using a three-by-two experimental
design (Refer to Figure 1). The design evaluates adding the Camper's solution (yes or no) and
three blending times (0, 1, or 2 minutes) with a fixed blending speed (22,000 rpm). Camper's
solution is a mixture of chemicals (Zwittergent 3-12, ethyleneglycol-bis-($ amino-ethyl ether)-
N,N'-tetra acetic acid (EGTA), and Tris buffer). Work by Camper et a/., (1985) showed that the
mixture enhances bacteria dissociation from solids. The chemicals were added to final
concentrations of 10'6 M Zwittergent 3-12, 10'3 M EGTA, 0.01 M Tris buffer, and 0.01 wt%
peptone, buffered to a pH of about 7 before blending.
When called-for in the experimental plan, the samples were blended at 22,000 rpm
(Manufacturer's reported speed) in a 7-speed commercial laboratory blender (Blend Master #
57199, Hamilton Beach, New Hartford, CT) for the designated time as measured with a
stopwatch. The 1.2-L blender has 4 mixing blades: 2 rounded blades pointing upward and 2
pointed blades tilting downward. Each blade is approximately 2.5-cm long and 1-cm wide at the
base. The blender jar was washed with soap between uses and autoclaved at 15 psi pressure for
15 minutes to assure sterility. The plastic lid was rinsed with isopropyl alcohol. The sample
temperature increase during blending was monitored during separate studies and found
negligible for the 2-minute period.
The resulting samples were analyzed for four indicator organisms (total coliform, fecal
coliform, fecal streptococcus, and Escherichia coll (E. coli)). These organisms are commonly
used or proposed bacterial indicators in water quality monitoring.
Particle size distributions of the samples before and after blending the sample for one or
two minutes were measured using a Coulter Particle Characterization Unit (Model # Delsa 440
SX, Beckman Coulter, Miami, FL).
Analysis of microorganisms
-------
All samples were analyzed using membrane filtration methods following Standard
Methods for the Examination of Water and Wastewater (APHA et a/., 1998). Three serial
sample dilutions were prepared in sterile phosphate buffered solution before filtration. A series
of three dilutions were selected for each organism using analytical results of samples previously
collected at the same outfall, in order to obtain a colony count in the preferred 20 to 60 colony
range. Quadruplicate analyses for each organism at each dilution were completed to monitor the
analytical variability. Reference cultures were used in the laboratory to evaluate the test
procedures, including media and reagents. Blanks were run before and after each analytical set.
Total coliforms were determined by incubation on M-Endo agar for 24 h at 35°C and
confirmed by gas formation in lauryl tryptose broth and brilliant green lactose broth. Fecal
coliform was incubated on M-FC agar for 24 h at 44.5°C and confirmed by gas formation in
lauryl tryptose broth and EC broth. E. coli levels were measured by transferring the membrane
from the Endo-type medium to a nutrient agar containing 4-methylumbelliferyl-$-D-glucuronide
(NA-MUG), incubating for 4 h at 35°C, and checking for blue fluorescence on the colony
periphery under long-wavelength UV. Similarly, E. coli levels can also be measured by
transferring the membrane from the fecal coliform positive sample to a nutrient agar containing
NA-MUG. Fecal streptococci concentrations were determined by incubation on m-Enterococcus
agar for 48 h at 35°C. Colonies were transferred to brain heart infusion (BHI) agar incubated for
24-48 h at 35°C. Transfers were made to BHI broth and incubated at 35°C for 24 h, with
confirmations made by retransfer to bile esculin agar incubated at 35°C for 48 h, BHI broth
incubated at 45°C for 48 h, and BHI with 6.5% NaCl incubated at 35°C for 48 h.
Chemicals
Zwittergent 3-12 was purchased from Calbiochem-Novabiochem Corp. (LaJolla, CA).
M-Endo agar, M-PA agar, Baird-Parker agar, NA-MUG, and M-Endo media were obtained from
Difco Labs (Detroit, MI). M-FC media and lauryl tryptose broth were purchased from Beckton
Dickinson (Sperks, MD). All other chemicals were obtained from Sigma Chemical Corp. (St
Louis, MO). All the chemicals were stored according to manufacturers' recommendations.
Data analysis and statistical methods
-------
Organism concentrations were calculated and expressed as colony forming units per 100
mL (CFU/100 mL). Previous evaluation at this outfall (data not shown) supported a log-normal
distribution (logio-transformed) of organism concentrations as suggested by other researchers
(APHA et a/., 1998; USEPA, 1983). Organism concentrations were logio-transformed before
data analysis. Raw colony counts obtained from the set of plates with enumerations in the
preferred 20- to 60-count range were used for data analysis. When no set of plates contained
countable colonies in the target range, countable plates on both sides were used for data analysis.
Multiple analysis of variance was used to compare data groups. Statistical significance
was set at the 95% level of confidence. Analysis was done using Statistica software (StatSoft,
Inc., 1998).
RESULTS
Analytical variability
Analytical variability of microorganisms was calculated using the standard deviation of
the logio-transformed data from the four replicate analyses in the dilution sets used. The
standard deviation of the log^o-transformed data is relatively constant and generally less than
0.25 units.
Effects of blending on microorganism concentrations
Total Coliform: Table 1 summarizes the results for the total coliform analyses. Adding
Camper's solution decreased the measured total coliform concentration in both blended and
unblended samples (p<0.01). Blending the sample before analysis increased the measured
concentrations in both the samples with Camper's solution and without Camper's solution. The
0.26-log increase in mean concentration of samples analyzed without Camper's is not significant
(p=0.56). The 0.67-log difference in samples analyzed with Camper's solution is significant
(p=0.02). The interaction between the effects of Camper's solution and the effects of blending is
not significant (p=0.15). The difference between 1- and 2-minute blending time is not
significant (p=0.25). Figure 2 presents the concentrations of total coliform in both blended and
unblended samples with and without Camper's solution.
7D
-------
Fecal Coliform: Figure 3 and Tables 2 and 3 give the results of the analyses. Adding
Camper's solution decreased the measured concentration of fecal coliform in both blended and
unblended samples (p<0.01). Blending the sample before analysis increased the measured
concentrations in both the samples with Camper's solution and without Camper's solution. The
0.87-log difference without Camper's is significant (p<0.01) and the 1.42-log difference in
samples analyzed with Camper's solution is also significant (p<0.01). An interaction exists
between Camper's solution and blending (p<0.01). Increasing the blending time from 1 to 2
minutes increased the measured concentration. The 0.27-log increase in mean concentration
without Camper's solution is not significant (p=0.11), but the 0.33-log increase with Camper's is
significant (p<0.01).
Fecal Streptococcus: Adding Camper's solution decreased the measured concentration
of fecal streptococcus in blended samples, but increased the concentration in unblended samples
(p<0.01) (Table 4). Blending the sample before analysis increased the measured concentrations
in both the samples with and without Camper's solution. The 0.82-log difference in mean
concentrations without Camper's is significant (p<0.01). The 0.06-log difference in the means
of samples analyzed with Camper's solution is not significant (p=0.67). The interaction between
the effects of Camper's solution and the effects of blending is significant (p<0.01). Increasing
the blending time from 1 to 2 minutes does not increase the measured concentration in samples
without Camper's solution. Figure 4 presents the concentrations of fecal streptococcus in both
blended and unblended samples with and without Camper's solution.
E. coli: Adding Camper's solution decreased the measured E. coli concentration in
unblended samples (p<0.01), but increased the concentration in blended samples (p<0.01) (Table
5). Blending the sample before analysis decreased the measured concentrations in samples
without Camper's solution (p<0.01), but increased the measured concentration in samples with
Camper's solution (p=0.12). The 2.29-log difference without Camper's is significant (p<0.01).
The 0.94-log difference in the means of samples analyzed with Camper's solution is not
significant (p=0.12). The interaction between the effects of Camper's solution and the effects of
blending is significant (p<0.01), and increasing the blending time does not affect the measured
sample concentration. Figure 5 presents the concentrations of E. coli in both blended and
unblended samples with and without Camper's solution.
8D
-------
Effects of blending on particle size
Table 6 lists the mean particle size of the samples before and after blending for 1 and 2
minutes. The mean particle size remains essentially constant with 1 and 2 minutes blending.
The results show no correlation between increased total coliform, fecal coliform, and fecal
streptococcus concentrations and the mean particle size. Earlier studies by Perdek and Borst
(2000) with CSO samples and diluted sanitary sewage showed decrease in mean particle size
with blending. However, no correlations between increased fecal coliform and enterococcus
indicator concentrations and decreased mean particle size were observed.
Summary
Blending samples before analysis increased the measured concentration of all bacteria
except for E. coli. Adding Camper's solution to the sample but not blending decreased the
concentration of all measured bacteria concentrations other than fecal streptococcus. Adding
Camper's solution before blending decreased all measured concentrations other than E. coli,
however, these decreases were smaller than the decreases observed with Camper's alone. An
interaction exists between blending and adding Camper's solution. Blending the sample does
not affect the measured mean particle size.
DISCUSSION
In many, perhaps most measurements, including the particle-associated organisms in the
measured aqueous bacterial load will have little direct effect on the intended use of the analytical
result. The comparative difference in concentration measured in raw combined or sanitary
sewage accounted for by attached organisms, for example, will generally be negligible if
considered with the magnitude of the measurement and the application. Increasing the reported
concentration of a few orders of magnitude in these applications say from 106 to 109CFU/100 mL
will not effect decisions based on the data even if the differences are statistically significant. In
9D
-------
these conditions, understanding of the propensity of organisms attached to particles may have
little more than academic interest. In selected applications, however, the relative differences can
be not only statistically significant, but also physically significant. Evaluating of the full or
partial disinfection processes of these same two streams after pretreatment may require
consideration of the associated organisms to fully understand the process effectiveness.
Stormwater runoff in municipal separate storm sewer systems (MS4s) can become a
major part of the total stream flow in some low-order receiving waters. In these applications, the
fully-diluted Stormwater flow can readily raise the bacteria concentration in the monitored
receiving water to approach water quality standards. Including the particle-associated organisms
can shift the analysis from "pass" to "fail." Similarly, when Stormwater flows through retention
controls, the particles and associated organisms can, depending on effective holding time and
settling velocity, settle and accumulate. Knowing that the organism decay processes have long
time constants in sediment (Schillinger and Gannon, 1982), may suggest that sediment removal
frequencies should increase to prevent microorganism-rich washouts to receiving waters.
Similarly a Stormwater management strategy could be developed promoting aqueous conditions
that induce passive particulate attachment making retention an effective substitute for
Stormwater disinfection to protect receiving waters. These results, although based on a single
event from a single outfall, support the supposition that particle-associated bacteria exist in
Stormwater. Watershed managers and supporting practitioners must consider the potentially
advantageous and disadvantageous effects that result.
CONCLUSIONS
Stormwater runoff contains organisms not readily identified using standard MF analysis.
Each tested organism, except for E. coli, showed an increase in measured concentration after
blending. The relative increases varied with the specific organism. The chemical mixture
developed by Camper et al. (1985) for releasing organisms from activated carbon did not
consistently promote, and appears to usually suppress, the release for later enumeration. In some
cases, the apparent suppression was two orders of magnitude. There is a statistically significant
interaction between the effects of Camper's solution and the effects of blending for all the
ion
-------
organisms tested, except for total coliform. Blending had negligible effects on mean particle
size.
Although based on a single storm event from a single outfall, these results suggest that
particle-associated microorganisms play an important, if often unmeasured, portion of the total
organism count in storm water.
ACKNOWLEDGMENT
The sample collection and analyses were performed by US Infrastructure, Inc. of Edison,
New Jersey under EPA Contract 68-C98-157. Mano Sivaganesan of EPA's National Risk
Management Research Laboratory (NRMRL) in Cincinnati provided statistical support, and Sam
Hayes of NRMRL provided quality assurance advice and review.
DISCLAIMER
Use of trade, brand, or firm names in this report is for identification purposes only and
does not constitute endorsement by the U.S. Environmental Protection Agency.
REFERENCES
American Public Health Association, American Water Works Association, and Water
Environment Federation. (1998) "Standard Methods for the Examination of Water and
Wastewater," 20th Edition.
Armon, R. Recharge of Groundwater by Infiltrated Urban Runoff: Quality Aspects (Microbial
and Chemical). Internal Report. Technion-Israel Institute of Technology, Israel.
Berman, D., E.W. Rice, and J.C. Hoff. (1988) Inactivation of Particle-Associated Coliforms by
Chlorine and Monochloramine. Appl. Environ. Microbiol. 54, 507.
11D
-------
Camper, A.K., M.W. LeChevallier, S.C. Broadaway, and G.A. McFeters. (1985) Evaluation of
Procedures to Desorb Bacteria from Granular Activated Carbon. Journal of
Microbiological Methods 3, 187 - 198.
COM. 151977) Spring Creek AWPCP Upgrade Capital Project No. WP-225: CSO Disinfection
Pilot Study. Final Report. November.
Glover, G.E. and G.R. Herbert. (1973) Microstraining and Disinfection of Combined Sewer
Overflows - Phase II. EPA-R2-73-124, Washington, D.C.
Moffa, P.E., E.C. Tifft, Jr., S. L. Richardson, and J.E. Smith. (1975) Bench-Scale High-Rate
Disinfection of Combined Sewer Overflows with Chlorine and Chlorine Dioxide. EPA-
670/2-75-021, Cincinnati, OH.
Oliveri, V.P., C.W. Kruse, K. Kawata, and J.E. Smith. (1977) Microorganisms in Urban
Stormwater. USEPA Report No. EPA-600/2-77-087. July.
Parker, J.A. and J.L. Darby. (1995) Particle-Associated Coliform in Secondary Effluents:
Shielding from Ultraviolet Disinfection. Water Environment Research 67, 1065 - 1075.
Perdek, J.M. and M. Borst. (2000) "Particle Association Effects on Microbial Indicator
Concentrations and CSO Disinfection," ASCE's 2000 Joint Conference on Water
Resources Engineering and Water Resources Planning & Management, July - August,
Minneapolis, MN.
Pitt, R., M. Lalor, D.D. Adrian, R. Field, and D. Barbel. (1993) Investigation of Inappropriate
Pollutant Entries into Storm Drainage System: A Users Guide. Office of Research and
Development, U.S. Environmental Protection Agency, Cincinnati, Ohio. EPA/600/R-
92/238.
12D
-------
Ridgway, H.F. and Olson, B.H. (1982) Chlorine Resistance Patterns of Bacteria from Two
Drinking Water Distribution Systems. Appl. Environ. Microbiol. 44, 972.
Sartor, J. D. and G. B. Boyd. (1972). Water Pollution Aspects of Street Surface Contaminants.
U. S. Environmental Protection Agency, Washington D.C. EPA-R2-72-081.
Scheible, O.K., M.C. Casey, and A. Forndran. (1986) Ultraviolet Disinfection of Wastewaters
from Secondary Effluent and Combined Sewer Overflows. U. S. Environmental
Protection Agency, Cincinnati, Ohio. EPA-600/2-86-005.
Schillinger, I.E. and JJ. Gannon. (1985) Bacterial Adsorption and Suspended Particles in
Urban Stormwater. J. WPCF 57, 384-389.
StatSoft, Inc. (1998) STATISTIC A for Windows [Computer Program Manual]. StatSoft, Inc.,
2300 East 14th Street, Tulsa, OK 74104, phone: (918) 749-1119, fax: (918) 749-2217,
email: info@statsoft.com, WEB: http://www.statsoft.com.
USEPA. (1983) Results of the Nationwide Urban Runoff Program, Volume 1: Final Report.
Water Planning Division, Washington, DC. December. NTIS Publication No. 83-
185552.
USEPA. (2000) Water Quality Conditions in the United States: A Profile from the 1998
National Water Quality Inventory Report to Congress. Office of Water, Washington,
D.C. http://www.epa.gov/305b/98report/98summary.html.
13
-------
Table 1. Summary of Results for Total Coliform Analysis
Unblended Samples
Camper's
Added
No
Yes
Significance
Concentration,
C
(CFU/100 mL)
3.7 xlO4
S.lxlO2
Log C Number of
Samples
N
4.56 5
2.50 4
p<0.01
Blended Samples
Concentration,
C
(CFU/lOOmL)
6.7 xlO4
l.SxlO3
Log C Number of
Samples
N
4.82 5
3.17 6
p<0.01
Significance
P
p=0.56
p=0.02
14D
-------
Table 2. Effects of Blending Time on Fecal Coliform Concentrations
1 -Minute
Camper's
Added
No
Yes
Significance
Concentration,
C
(CFU/lOOmL)
7.3x 104
2.0 xlO2
LogC
4.86
2.31
pO.Ol
2 -Minute
Number of Concentration,
Samples C
N (CFU/lOOmL)
4 1.3xl05
4 4.3 xlO2
Log C Number of
Samples
N
5.13 3
2.64 4
pO.Ol
Significance
P
p=0.11
p=0.01
15D
-------
Table 3. Summary of Results for Fecal Coliform Analysis
Unblended Samples
Camper's
Added
No
Yes
Significance
Concentration,
C
(CFU/100 mL)
1.3xl04
l.lxlO1
Log C Number of
Samples
N
4.11 10
1.05 2
pO.Ol
Blended Samples
Concentration,
C
(CFU/lOOmL)
9.5 xlO4
3.0 xlO2
Log C Number of
Samples
N
4.98 7
2.47 8
pO.Ol
Significance
P
pO.Ol
pO.Ol
16D
-------
Table 4. Summary of Results for Fecal Streptococcus Analysis
Unblended Samples
Camper's
Added
No
Yes
Significance
Concentration,
C
(CFU/lOOmL)
1.0 xlO3
2.7 xlO3
Log C Number of
Samples
N
3.00 5
3.43 5
pO.Ol
Blended Samples
Concentration,
C
(CFU/lOOmL)
6.6 xlO3
S.lxlO3
Log C Number of
Samples
N
3.82 8
3.49 9
pO.Ol
Significance
P
pO.Ol
p=0.67
17D
-------
Table 5. Summary of Results for E. coli Analysis
Unblended Samples
Camper's
Added
No
Yes
Significance
Concentration,
C
(CFU/lOOmL)
1.7 xlO4
2.0x10'
Log C Number of
Samples
N
4.23 5
1.29 4
pO.Ol
Blended Samples
Concentration,
C
(CFU/100 mL)
8.8x10'
1.7xl02
LogC
1.94
2.23
pO.Ol
Significance
Number of P
Samples
N
4 pO.Ol
17 p=0.12
18
-------
Table 6. Mean Particle Size
Condition Mean Particle Size (: m)
Unblended 0.768
Blended for 1 minute 0.758
Blended for 2 minutes 0.775
19
-------
Figure 1. Test Design for Blending Study
20
-------
Total Coliform
5.2
4.8
4.0
2.4
Without Camper's Solution
Not Blended
With Camper's Solution
5.2
4.4
3 3.6
u_
O
2.8
2.0
Not Blended
Figure 2. Total Coliform Concentrations
21
-------
Fecal coliform
Without Camper's solution
With Camper's solution
J.-J
5.0
4.5
4.0
3.5
E
8
§ 3.0
o.
o
g 2.5
2.0
1.5
1.0
n s
-T-
' 1 '
o
I
^izr^
o
J.-J
5.0
4.5
4.0
3.5
E
8
§ 3.0
0
O
g 2.5
2.0
1.5
1.0
n s
"T" MeanħS[
I I MeanħSE
n Mean
o Outliers
* Extremes
|
I ° I
o
-^
Not Blended
Not Blended
Figure 3. Fecal Coliform Concentrations
22
-------
Fecal Streptococcus
Without Camper's Solution
With Camper's Solution
4.0
3.8
3.6
3.4
I
8 3.2
u_
" 3.0
-------
Without Camper's Solution
EColi
With Camper's Solution
MeanħS[
I I MeanħSE
D Mean
Extremes
Figure 5. E. coli Concentrations
24
------- |