United States
                          Environmental Protection
                          Agency
                          Wastewater Technology Fact Sheet
                          Bacterial Source Tracking
INTRODUCTION
Pathogens are a major pollutant of water bodies
nationwide according to many states' Clean Water Act
303(d) reports. Various sources contribute pathogens
to contaminated waters, including fecal pollution from
humans, wildlife, and livestock. Besides being potential
pathogens, fecal bacteria (such as Escherichia coif)
can  indicate the  presence  of  other  waterborne
pathogens. Bacteria from human sources may indicate
the presence of human viruses, while bacteria from wild
and domestic animals may indicate the presence of the
parasites Giardia or Cryptosporidia. The presence
of any fecal bacteria in drinking water is considered a
health hazard. Knowing the source(s) of bacteria in a
water body or water supply is of great value in the
remediation  and  prevention  of  further  bacterial
contamination. However, it can be difficult to address
water quality impairment effectively without a reliable
method to determine  the  source of contamination.
Bacterial Source Tracking (BST) is new methodology
used  to  determine  the source of fecal pathogen
contamination in environmental samples.

There are many BST methods available and more are
under  development.  Interest in  applying   these
techniques stems from EPA's recent implementation of
the  Total Maximum Daily Load (TMDL) study,  as
BST techniques appear to provide the best method to
determine the origins of fecal contamination in water
bodies. Projects to develop TMDLs for fecal coliforms
and  to design and  implement  best  management
practices (BMPs) to reduce fecal loading in water may
benefit from BST technology (Hager, 2001). This fact
sheet discusses BST methods and presents examples
of BST  application  to TMDL  development  and
implementation.
DESCRIPTION

Potential sources of  fecal bacteria  are  generally
grouped into three major categories: human, livestock,
or wildlife.   In more urban watersheds,  a fourth
category of pets or dogs may be added. Each source
produces unique, identifiable strains of fecal bacteria
because the  intestinal environments  and  selective
pressures to which the bacteria are subjected differ
from source to source.

BST may use one of several methods to differentiate
between potential sources of fecal contamination, all of
which follow a common sequence of analysis. First, a
differentiable characteristic, or fingerprint  (such  as
antibiotic resistance  or DNA), must be selected  to
identify various strains of bacteria.  A representative
library of bacterial strains and their fingerprints must
then be generated from the human and animal sources
that may impact the water body. Indicator bacteria
fingerprints from the polluted water body are compared
to those in the library and assigned to the appropriate
source category based on fingerprint similarity. BST
methods  can  be grouped  as molecular  or non-
molecular  methods,  according to the characteristic
used to identify or fingerprint the bacteria.   Table 1
summarizes the classification of various BST  methods.

Molecular methods are also referred to as "DNA
fingerprinting" and are based on the unique genetic
makeup of different strains of fecal bacteria. Molecular
methods rely on genetic variation as the fingerprint to
identify the source of fecal  contamination.  Three
molecular BST methods are commonly used,  including
ribotyping (RT), polymerase chain reaction (PCR), and
pulsed-field gel electrophoresis (PFGE). Procedures
for the RT and PFGE methods are relatively  similar
among multiple  studies,  but substantially  different
variations  are  reported when using PCR  methods
(Hagedorn, 2001).

-------
    TABLE 1 CLASSIFICATION OF BST
                  METHODS
 Molecular methods (DMA fingerprinting)
        Ribotyping (RT)
        Polymerase chain reaction (PCR)
        Pulsed-field gel electrophoresis (PFGE)
 Non-molecular methods
        Biochemical methods
              Antibiotic resistance analysis (ARA)
              Cell wall fatty acid methyl ester
 (FAME)
              F-specific coliphage typing
              Carbon utilization (BIOLOG)
        Chemical methods
              Caffeine detection
	Optical brightener detection	
 Source:  Parsons, 2001.

Non-molecular   methods   use   non-genetic
characteristics as the fingerprint or basis to differentiate
the source  of fecal bacteria, and may  be further
subdivided  between  biochemical  and  chemical
methods.  Biochemical methods are based on the
ability of an organism's genes to actively produce a
biochemical substance. The type and quantity of the
substance(s) produced form the bacterial fingerprint.
Antibiotic  resistance  analysis (ARA)  is the  most
commonly used biochemical BST method.   Other
biochemical methods, such as cell wall analysis of fatty
acid methyl ester (FAME), F-specific coliphage typing,
and carbon source utilization (BIOLOG system), are
under development.  Chemical methods do not detect
the presence  of  fecal  bacteria, but rely  on the
identification of compounds that co-occur with fecal
bacteria in human  wastewater  to differentiate the
source of fecal pollution. Thus, chemical methods can
only determine whether or not the source of fecal
pathogens is human (Hagedorn, 2001).  Examples of
compounds used in chemical BST include caffeine and
optical  brighteners  commonly used   in   laundry
detergents.
APPLICABILITY

BST is  intended to aid in identifying sources (e.g.,
human, livestock, or wildlife) of fecal contamination in
water bodies. Several states have started to use DNA
fingerprinting to target water quality problems  and
formulate  a  mitigation  strategy  (Pelley,  1998;
Blankenship, 1996).  These techniques  can also be
used to  direct implementation of effective BMPs to
remove  or reduce fecal contamination. For example,
two New Hampshire communities are performing BST
surveys  (using  the  RT  method) to determine  the
contribution of bacterial  contamination from several
specific  sources so that BMPs may be put in place to
help rehabilitate water quality.  The following is  a
summary of one representative survey.

Hampton Harbor, New Hampshire

Hampton  Harbor is a tidally dominated, shallow
estuary located at the extreme southeast corner of New
Hampshire. The Hampton Harbor clamflats are closed
for clam harvesting during September and October due
to elevated fecal coliform levels.  The flats are open
from November through May but close temporarily if
the rainfall exceeds 0.25 inches.  These clamflats are
popular, productive, and accessible to the public.
Despite  the  construction  of  a  new  wastewater
treatment facility in the Town of Seabrook, the bacteria
levels often exceed the limits  set  by the  New
Hampshire Shellfish Program, resulting in flat closures
and frustrated clam diggers.  The potential sources of
bacterial contamination  include birds  (cormorants,
starlings, gulls), domestic animals (cats,  dogs, goats,
horses),   sanitary   wastewater  from  wastewater
treatment plant failures, and wildlife.  The intent of the
survey  was  to provide information  to  support
implementation  of  specific  source  controls and to
reduce the bacterial contamination to a level  that
increases the number of days  that the clamflats  are
open for recreational harvesting.

Source classification provided by BST is  often used in
the development and  implementation  of  TMDL
projects. The information can be used to assign load
reduction allocations to sources in a watershed.  For
example, BST techniques have been very useful to
regulatory officials in Virginia, where the ARA method

-------
has been used in seven TMDL watershed projects to
date.

Virginia Department of Conservation and
Recreation TMDLs

Over  300 stream  segments were listed  on  the
Commonwealth  of  Virginia's 303(d)  list  for fecal
coliform bacteria violations.  The uncertainty inherent in
identifying specific sources of fecal coliform bacteria in
the streams has hindered development of the TMDLs.
BST studies were applied  to three stream segments
(Accotink Creek, Blacks Run, and Christians Creek)
to provide more  accurate waste-load allocations and
enhance the  development  and  defensibility of  the
TMDLs.   In each, the RT  method  was  used to
determine  the dominant sources of fecal coliform in the
impaired  stream  segments.   The  source-tracking
distribution determined in each segment were used to
modify and strengthen the waste-load allocations in the
TMDL watershed model. In addition, DNA testing is
underway in the Muddy Creek, Lower Dry River, Mill
Creek, and Pleasant Valley watersheds as part of their
TMDL implementation plans.

Cedar Creek, Hall Creek, Byers  Creek, Hutton
Creek, and Lower Blackwater River were also placed
on the Commonwealth of Virginia's 303(d) list because
of violations of the fecal coliform bacteria water quality
standard. In fulfilling the state requirement to develop
a  TMDL  Implementation Plan, a  framework was
established to reduce fecal coliform levels and achieve
the water  quality goals for  which TMDL allocations
were developed.  BST analysis using the ARA method
was performed as part of the TMDL implementation.
Results indicated contributions of fecal coliform from
livestock,  human, and  wildlife sources. The wildlife
contribution alone was enough to push fecal coliform
levels  beyond the standard at five  sampling sites, while
human sources alone were high enough to violate the
standard at five sampling sites. Livestock sources were
sufficient to violate the standard at  eight of  eleven
sampling sites.   In the Cedar  and Hutton Creek
watersheds, livestock appeared to be  an  issue
throughout the watershed,  while in the  Hall/Byers
Creek watershed, livestock  problems appeared limited
to  smaller tributaries  (e.g. Indian  Run  and  Tattle
Branch).  Human sources seemed most significant in
the Hall/Byers and Hutton Creek watersheds.  The
quantity  of  control  measures  required  during
implementation was determined and progress toward
end  goals will  be assessed during implementation
through tracking  control measure installations and
continued water  quality monitoring.  Water quality
monitoring will include fecal coliform enumeration and
BST analysis. BST will provide an indication of the
effectiveness of specific groups of control measures,
specifically agricultural and urban. Implementation was
scheduled to begin in July 2001, with the final goal
being the delisting of the impaired segments from the
Commonwealth of Virginia's 303(d) List of Impaired
Waters by 2011.

ADVANTAGES AND DISADVANTAGES

In general, molecular BST methods may offer the most
precise identification  of specific types of sources, but
are limited by high per-isolate costs and detailed, time-
consuming procedures. They are also not yet suitable
for assaying large numbers of samples in a reasonable
time frame. Biochemical BST methods are simpler,
faster,  less  expensive, and  allow large numbers of
samples to be assayed in a short period of time.

BST development is  so  new  that little  research
comparing individual  methods is complete.  Results of
initial studies should become available over the next
few years.

The United States Department of Agriculture recently
funded  a two-year  study to compare three BST
methods using E. coll and Enterococcus: ARA,
PFGE, and RT.  The merits of these  methods will be
compared by a) accuracy, cost, and processing time;
b) determining the geographic range of the libraries;
and c) assessing  the  utility  of each  method in field
experiments.  This comparison and  development of
BST  methodology   is  intended  to  refine  BMP
implementation and  focus  resources  on  pollution
sources for water quality impairments.

The United States Geologic Survey is developing a
program to identify sources  of fecal bacteria in the
waters of Berkeley County, West Virginia.  At least
five methods will be tested for their ability to determine
animal sources of fecal bacteria in water samples (RT,

-------
PFGE, ARA, PCR, and BIOLOG carbon-utilization).
The three objectives of this project include building
source  libraries for the five methods,  comparing
methods to see which is best to determine sources, and
using the best method to identify sources of bacteria in
water resources of Berkeley County.  This study will
provide source libraries for five promising methods to
identify bacteria sources, quantitative information on
which method(s) works best, determination of bacteria
sources for ten domestic wells that contain bacteria,
and determination of bacteria sources for five large
public-supply springs. The libraries and methods will
be applicable to both surface and ground  water in
Berkeley County and surrounding areas.

PERFORMANCE

Many  BST  techniques  are  undergoing intensive
research that leads to rapid change in existing methods
and the creation of new methods. BST technologies
are quickly becoming proven and should be used by
federal  and  state regulatory  agencies to address
sources of fecal bacterial pollution in water. Although
they are still experimental, BST methods represent the
best tools available to determine pathogen TMDL load
allocations   and   TMDL   implementation   plan
development.  The  following are examples of BST
technique performance in specific watershed studies.

Antibiotic Resistance Analysis (ARA) Method

Holmans Creek, Virginia

Holmans  Creek  watershed   was  listed  on  the
Commonwealth of  Virginia's  1998 303(d) TMDL
Priority List of Impaired Waters based on violations of
the fecal  coliform bacteria water quality standard.
There are several potential fecal coliform sources in this
watershed, including the non-point sources of wildlife,
livestock, individual residential sewerage systems, and
land  application of manure and litter.   Beef cattle,
poultry and dairy are the major livestock operations in
the Holmans Creek watershed. Residential sewerage
in the watershed  consists of direct discharges from
straight pipes (homes without facilities to treat their
waste discharge), privies, and failing septic systems.
BST analysis using the ARA method was used to
classify sources of  the fecal  bacteria found in the
polluted water.  Results of the BST analysis suggest
that the primary source of fecal pollution is human,
constituting just under half of the total fecal coliform
deposited into the waters of Holmans Creek.  Wildlife
and cattle sources each contribute approximately one-
fourth of the total fecal coliform loads in the watershed.
Poultry were determined to be a minor contributor to
fecal coliform pollution in Holmans Creek, contributing
one-tenth of the total fecal load.

Stevenson Creek, Florida

The   Stevenson  Creek   basin  encompasses
approximately 6,000 acres in central Pinellas County,
Florida.   In keeping  with  the objectives of the
Stevenson Creek Watershed Management Plan, a
BST  study was  initiated to  identify  the  dominant
source(s) of fecal contamination to Stevenson Creek in
Clearwater, Florida.  The ARA method was chosen
because it can assess the source of indicator organisms
based  on a  much  larger subset  of the  bacterial
population than molecular methods can. The dominant
sources of fecal coliform over the course of the study
were wild animal, dog, and human, with the overall
trend indicating that wild animal isolates comprised the
majority  of  fecal coliforms obtained when colony
forming units (CFU)  counts exceeded the acceptable
limit of 200 CFU per 100 mL. While human input was
not the major cause of elevated fecal coliform levels for
most  of the  samples  analyzed for this study, the
domination of  some  small  populations by human
isolates suggests that human  sources contribute to
low-level background  contamination.   This occurs
when fecal coliform populations are  low,  near the
transition to dry season, and perhaps few isolates are
washed into surface waters from draining storm water.
Lowering water tables may also draw wastewater from
small, otherwise innocuous leaks. Overall, there was
little evidence of acute human fecal contamination on a
large scale; however, human sources may influence two
sampling sites, detectable despite the presence of fecal
coliforms from other sources.  The human input  alone
for these two sites in one month was high enough to
violate water quality standards.

-------
Pulsed-Field Gel Electrophoresis (PFGE)
Method

Eastern Shore, Virginia

DNA fingerprinting using PFGE proved helpful when
an oyster farmer on Virginia's Eastern Shore was faced
with the closure of his shellfish beds due to elevated
levels of E. coli. Failing septic tanks were assumed to
be the primary source of the fecal pollution, but a
survey of septic systems in the  sparsely populated
watershed indicated that they were not the cause, and
it became necessary to identify other potential sources.
The highest levels of coliform bacteria were measured
in the small tidal inlets and rivulets of the wetlands
located  upstream of local houses, shifting suspected
sources from human to other sources. Researchers
collected fecal samples from raccoon, waterfowl, otter,
muskrat, deer, and humans in the area and used DNA
fingerprinting to confirm the suspicion that the source
was  not anthropogenic in nature.  Comparing E. coli
from the shellfish beds against the fingerprints of known
strains in the DNA library, the researchers linked the
in-stream  E. coli to deer and raccoon (mostly
raccoon).  Several  hundred animals, including  180
raccoon, were removed from areas adjacent to the
wetlands.  E. coli levels subsequently declined by 1 to
2 orders of magnitude  throughout the watershed,
allowing threatened areas of the tidal creeks  to be
reopened to shellfishing.

Four Mile Run, Virginia

Four Mile Run is listed  on the  Commonwealth of
Virginia's  303(d) listing  for elevated levels of fecal
coliform bacteria.  The Northern Virginia Regional
Commission is currently developing a TMDL for the
Four Mile Run watershed, with the final  draft to be
submitted  to Virginia Department of Environmental
Quality by March 1, 2002. Four Mile Run is an urban
stream with no agricultural runoff.  The watershed is
home to 183,000 people, just over 9,000 per square
mile.  The  dominant land use in the watershed is
medium to high density residential housing.  Seven
central business districts exist within this 20 square mile
watershed, and two high-capacity  interstates  pass
through the watershed along with numerous primary
and  secondary  roadways.    The  watershed  is
approximately 40 percent impervious.  A large pet
population accompanies the dense human population in
the watershed. As to potential fecal sources, there is
little manufacturing industry to generate point source
discharges and there are no combined sewers in the
majority of the watershed. Sanitary sewers serve more
than 99.9 percent of the watershed population.  The
number of septic  systems in the watershed is believed
to be less than 50. The PFGE method of BST analysis
was conducted on E. coli DNA from seasonally varied
stream and sediment samples in the watershed. Results
of the  analysis show that waterfowl contribute over
one-third (38 percent) of the bacteria, humans and pets
together account for over one-fourth (26 percent), and
raccoons account for 15 percent of the contamination,
with deer (9  percent)  and  rats (11  percent)  also
contributing.  The predominant non-human sources
include wildlife species with intimate association with
the waterways.

Ribotyping (RT) method

Little  Soos Creek, Washington

A BST survey was designed to help characterize
sources of fecal  coliform bacterial contamination in
Little  Soos  Creek in  southeast  King  County,
Washington, in response to the impact of existing and
anticipated urban development in the area. Little Soos
Creek  has historically been categorized as a Class A
stream, but violates fecal coliform standards for this
classification.  The goal of the BST survey was to help
determine  the contribution to contamination of the
stream from two potential sources: livestock on hobby
farms and ranches adjacent to the stream and on-site
septic systems close to the stream in highly permeable
soils.   Other  animal sources were also  considered.
Genetic fingerprinting (using ribosomal RNA typing or
RT) was performed on E. coli isolates to effectively
match  specific strains of E. coli from a contaminated
site in  the stream to its source.  The intent was to
provide information to  support implementation of
specific source controls.   The  study  identified the
sources of approximately three-fourths of the  fecal
coliform contamination, with the  primary sources
determined to be cows, dogs, and horses. Although
septage was  identified  as  a  contributor to  the
contamination problem, it was not indicated as a major

-------
source. However, even low levels of contribution from
septage suggest the potential for Little Soos Creek to
harbor a number of human viral, bacterial, and parasitic
pathogens associated with human sources.  For this
reason, further investigation of the contribution from
septic systems and of human exposure (particularly
children) to the stream may be warranted.

Lower Boise River, Idaho

The Lower Boise River watershed from Lucky Peak
Reservoir to the Snake River near Parma contains
almost one-third of Idaho's population and four major
municipalities, including the city of Boise.  An arid
climate (approximately 10 inches of annual rainfall)
makes irrigation a requirement on most farmland. This
irrigation coupled with reuse of pasture water on
irrigated fields results in the contribution of non-point
discharge of fecal coliform bacteria to the Lower Boise
River.     In   1994,  the  Idaho  Department   of
Environmental Quality (IDEQ) placed the Lower Boise
River on the 303(d) list for impairment of primary and
secondary  contact designated  uses  because fecal
coliform  levels exceeded state  standards.  A draft
TMDL was completed and submitted to the USEPA
on December 1998 and approved on January 2001
with implementation plan due July 2001. The TMDL
indicates that bacteria discharge loads will require more
than 95  percent reductions from non-point  source
bacteria loadings to meet the primary contact bacteria
standard.  A DNA fingerprinting of coliform bacteria
was   conducted  to  focus   bacteria  reduction
improvement. E. coll cultures were grown from fecal
samples of cows, sheep, humans, ducks, and geese,
and DNA from these samples was identified.  The
maj or bacteria sources in the watershed identified using
the RT method were waterfowl, humans, pets, and
cattle/horses.   Waterfowl  were clearly the  largest
source.   The major  advantage  of using  the DNA
fingerprinting tool is the ability  to develop accurate
control measures (BMPs) in terms of bacterial sources.
Prior to this study, IDEQ knew there were bacteria
problems, but did not know where to focus control
measures. The results of the BST analysis identify the
major sources, allowing IDEQ to strategically place
BMPs.
University of Georgia/USDA RT comparison

BST methods, including RT, rely on a database of
known source fingerprints to identify environmental
isolates of fecal bacteria. It is not well understood to
what  degree these  known  source fingerprints are
biogeographically variable.  This is important because
a fingerprint database developed for one state or region
may or may not be  applicable to another.   The
objective of a University of Georgia/USDA study
(Hartel et al, 2002) was to use the RT method of BST
analysis to determine the geographic variability of the
fecal bacterium, E. coli, from one location in Idaho and
three  locations in Georgia for four animals: cattle,
horse, swine, and poultry. The study identified distance
from the source sample to the watershed as a key
variable for cattle and horses, but not for swine and
poultry. When theE1. coll ribotypes among the animals
were compared at one location, the relative percent
difference between them was 86, 89, 81, and 79 for
each of the four locations, suggesting good ribotype
separation among host animal species at one location.
Achieving  a high degree  of accuracy in matching
environmental isolates of fecal bacteria to a host origin
database depends on having a large number of isolates
for comparison and using a distance of 175 km or less
(at least for certain host animal species).

COSTS

Given the fact that many BST methods are still in the
research  and  development  phase, there  is  great
variation for cost per sample (or  per isolate) among
different laboratories. Factors that affect cost include
the following:

Analytical method - Molecular  BST methods  (e.g.
RT) are generally more expensive  than non-molecular
methods. In addition, automated techniques are more
expensive   but less  labor-intensive  than  manual
techniques for the same method (such as RT).

Size of the database - It is not known what size
database or library of bacterial isolates from known
fecal sources is required for accurate source prediction
in  a  given  watershed.     Considerations  in  the
development of the BST library include the size of the
watershed, the diversity of animal species and human

-------
sources that may significantly impact water quality, and
the heterogeneity of the population within  a given
source species. In many studies, the number of isolates
required to develop the known source database may
make up  the  majority of total isolates  analyzed,
constituting a large  fraction of the total cost for the
study.

Number of environmental isolates - The number of
isolates that must be analyzed from the water body of
interest varies  among  study  sites.  There may be
multiple isolates from each water sample taken, with
costs generally calculated per isolate.

Level of accuracy - Cost increases in proportion with
accuracy  or the percentage  of isolates  classified
correctly.  In some cases 80 percent is considered the
lowest acceptable level of accuracy. More studies are
needed to determine the level  of accuracy achievable
by each BST method.

The cost for BST analysis ranges from $25 to $100
per isolate using molecular methods and from $10 to
$30 per isolate for non-molecular methods.  These
costs  are based on classifying a sample  within  an
accuracy  range of 70 to 90 percent or higher.
However,  there is little firm guidance on the required
number of reference  fecal  samples and  isolates
extracted from each sample, causes wide variance in
the total cost for a fecal source tracking project. For
example,   the   cost for TMDL  developments for
Accotink Creek, Blacks Run and Christians Creek in
Virginia by  the   USGS  Richmond  office  was
approximately  $617,000 (total for the three TMDLs),
while  the  New  Hampshire  Department   of
Environmental Services spent approximately $225,000
to establish the ribotyping laboratory and partially
support the two source tracking surveys.  In two
ongoing comparison studies, the cost of the San Juan
Creek Watershed  Bacteria  Study  (California)  is
$274,000  (excluding  the expenses  for laboratory
analysis), while the USD A grant to compare three BST
methods (RT, PFGE and ARA) is $310,000.
REFERENCES

Other Related Fact Sheets

Other EPA Fact Sheets can be found at the following
web address:
http://www.epa.gov/owm/mtbfact.htm

Overview

1.     Blankenship, K.  1996. DNA library would
       give investigators inside poop on pollution
       sources.    Bay   Journal    6(6),
       http://www.bayjournal.com/96-09/
       DNA.HTM.

2.     Blankenship,  K.   1996.  Masked   bandit
       uncovered in water quality theft / Team tails
       pollution to unlikely culprit. Bay Journal 6(6),
       http://www.bayjournal.com/
       96-09/NUTRIENT.HTM.

3.     Hager, M. C.,  2001. Stormwater Magazine,
       Vol. 2 No. 3,  p!6-25, http://www.forester.
       net/sw_0105_detecting.html, Vol. 2 No. 4
       May  / June,  p22-27,  http://www.forester.
       net/sw_0106_detecting.html.

4.     Northern  Virginia  Regional   Commission
       (NVRC),  2000.  Bacteria Research Menu,
       http://www.novaregion.org/4milerun/
       bacteria.html.

5.     Parveen,   S. and Tamplin,  M.  L.,  2002.
       Sources   of  fecal   contamination,   in
       Encyclopedia of Environmental Microbiology
       (Editor: Bitton,  G.), John Wiley and Sons, Inc.,
       New York, NY.

6.     Pelley, J.  1998.  DNA fingerprinting holds
       promise for identifying nonpoint  sources of
       pollution.   Environmental   Science  and
       Technology. 32(21): 486A

7.     Sargeant,  D.,  1999.   Fecal  contamination
       source identification methods in surface water,
       Washington Department of Ecology, Report #
       99-345.

-------
8.     Tynkkynen,  S., Satokari, R., Saarela, M.,
       Mattila-Sandholm, T., and Saxelin, M., 1999.
       Comparison  of  Ribotyping,  Randomly
       Amplified Polymorphic DNA Analysis, and
       Pulsed-Field Gel Electrophoresis in Typing of
       Lactobacillus  rhamnosus  and  L.  caser
       Strains,   Appl.   Environ.   Microbiol.
       65:3908-3914.

9.     U.S. EPA,  1997.  DNA fingerprinting aids
       investigation-fecal coliform sources traced to
       unlikely suspects.  Nonpoint Source  News
       Notes    April/May   48:19-20.
       http://www.epa.gov/owow/info/NewsNotes/
       issue48/nnh48 .htm#c.

10.    U.S. EPA,  2001.   Protocol for developing
       pathogen  TMDLs,  EPA  841-R-00-002,
       Washington, D.C.

11.    U.S. EPA,  2002. Workshop on Microbial
       Source Tracking (February  5, 2002; Irvine,
       CA).   http://www.sccwrp.org/tools/
       workshop/source_tracking_workshop.html.

12.    USGS, 2001. Identifying Sources of Fecal
       Coliform Bacteria in Selected Streams on
       Virginia's   TMDL   Priority   List,
       http://va.water.usgs.gov/projects/val29. html

13.    Virginia  Polytechnical  University,   2001.
       Bacterial Source Tracking (BST), Identifying
       Sources   of   Fecal   Pollution,
       http://www.bsi.vt.edu/
       biol_4684/BST/BST.html.

Antibiotic Resistance Analysis (ARA)

Peer-reviewed Journal Publications

14.    Hagedorn, C., S. L. Robinson, J. R. Filtz, S.
       M. Grubbs, T. A. Angier, and R. B. Reneau,
       Jr. 1999. Using antibiotic resistance patterns in
       the fecal streptococci to determine sources of
       fecal pollution in a rural Virginia watershed.
       Appl. Environ. Microbiol. 65:5522-5531.
15.    Harwood, V.  J.,  J. Whitlock,  and V. H.
       Wilhington.  2000.  Classification  of  the
       antibiotic  resistance  patterns of indicator
       bacteria  by discriminant  analysis:  use  in
       predicting the source of fecal contamination in
       subtropical Florida waters.  Appl. Environ.
       Microbiol. 66:3698-3704.

16.    Kaspar, C.W., Burgess, J.L., Knight, IT., and
       Colwell, R.R., 1990.  Antibiotic resistance
       indexing  of Escherichia  coli  to  identify
       sources of fecal contamination in water, Can.
       J. Microbiol. 36:891-894.

17.    Parveen, S., R., L. Murphree, L. Edmiston, C.
       W.  Kaspar, K.  M. Portier,  and  M. L.
       Tamplin.   1997.   Association  of
       multiple-antibiotic-resistance profiles with point
       and nonpoint sources of Escherichia coli in
       Apalachicola Bay. Appl.  Environ. Microbiol.
       63:2607-2612.

18.    Wiggins, B. A.,  1996. Discriminant analysis of
       antibiotic  resistance  patterns   in   fecal
       streptococci, a  method to differentiate human
       and animal sources of fecal pollution in natural
       waters.   Appl.   Environ.   Microbiol.
       62:3997-4002.

19.    Wiggins, B. A.,  R. W. Andrews, R. A.
       Conway, C. L. Corr, E. J. Dobratz, D. P.
       Dougherty, J. R. Eppard, S. R. Knupp, M. C.
       Limjoco, J. M.  Mettenburg, J. M. Rinehardt,
       J.  Sonsino,  R.  L. Torrijos,  and  M. E.
       Zimmerman. 1999. Identification of sources of
       fecal  pollution using discriminant  analysis:
       supporting evidence from large datasets. Appl.
       Environ. Microbiol. 65:3483-3486.

Non-Journal Publications

20.    Bower, R. J.  2000. M.S.  Thesis.  Source
       identification of fecal pollution in the Tillamook
       watershed:  antibiotic discriminant  analysis.
       Oregon State University,  Corvallis, OR.

-------
21.    Bowman, A. M., C. Hagedorn, and K. Hix.
       2000. Determining sources of fecal pollution in
       the Blackwater River watershed, p. 44-54. In
       T. Younos  and J. Poff (ed.), Abstracts,
       Virginia Water Research Symposium 2000,
       VWRRC   Special  Report  SR-19-2000,
       Blacksburg, VA.

22.    Graves, A. K. 2000. M.S. Thesis. Determining
       sources of fecal pollution in water for a rural
       Virginia  community.  Virginia  Polytechnic
       Institute and State University, Blacksburg, VA.

23.    Parveen, S., Tamplin, M.  L., Portier, K. M.,
       Lukasik, G., Scott, T., Sheperd, S., Tobia, S.,
       Braun, K.R., Koo,  P., and  Farrah, S. R.,
       2001.   Geographic variation in Antibiotic
       Resistance  Patterns of  Escherichia  coll
       isolated from swine, poultry, beef and dairy
       cattle farms in Florida, American Society for
       Microbiology 101th General Meeting, May
       20-May 24, Orlando, FL.

24.    Wagner, V. and V. J. Harwood. 1999. Use of
       antibiotic resistance  patterns  to  discriminate
       between sources of fecal coliform bacteria in
       surface waters of northeast Florida. American
       Society for  Microbiology  99th  General
       Meeting, May 30-June 3, Chicago, IL.

Ribotyping (RT)

Peer-reviewed Journal Publications

25.    Carson, A. C., B. L. Shear, M. R. Ellersieck,
       and A. Asfaw.  2001. Identification of fecal
       Escherichia coli from humans and animals by
       ribotyping.  Appl.   Environ.  Microbiol.
       67:1503-1507.

26.    Hartel, P. G,  Summer, J. D.,  Hill, J. L.,
       Collins, J. V, Entry, J. A., and Segars, W. I,
       2002.     Biogeographic  variability   of
       Escherichia coli ribotypes from Idaho  and
       Georgia, Journal of Environmental Quality (in
       press).
27.    Popovic, T., Bopp,  C.,  Olsvik,  O,  and
       Wachsmuth,  K.,  1993.    Epidemiologic
       application of a standardized ribotype scheme
       for Vibrio cholerae Ol, J. Clin. Microbiol,
       31(9): 2474-2082.

28.    Parveen, S., K. M. Portier, K. Robinson, L.
       Edmiston,  and  M.  L.  Tamplin.  1999.
       Discriminant analysis of ribotype profiles  of
       Escherichia coli for differentiating human and
       nonhuman sources of fecal pollution. Appl.
       Environ. Microbiol. 65:3142-3147.

29.    Tee,  W.,  Lambert,  J.,  Smallwood,  R.,
       Schembri, M., Ross B. C., and Dwyer, B.,
       1992.   Ribotyping of Helicobacter pylori
       from clinical specimens, J. Clin. Microbiol,
       30(6): 1562-1567.

30.    Wheeler, A. L., Hartel, P. G, Godfrey, D. G,
       Hll, J. L. and Segars, W. L, 2002. Combining
       Ribotyping   and  Limited  Host  Range  of
       Enterococcus faecalis for Microbial Source
       Tracking,   Journal   of  American   Water
       Resources Association (in press).

Non-Journal Publications

31.    Hartel, P. G, W. I. Segars, N. J. Stern,  J.
       Steiner,  and A. Buchan.  1999. Ribotyping  to
       determine host  origin of Escherichia  coli
       isolates in different water samples, p.377-382.
       In D. S. Olsen and J. P. Potyondy (Eds.),
       Wildland   Hydrology.  American   Water
       Resources Association Technical Publication
       Series TPS-99-3, Herndon, VA.

32.    Hll, J. L., P. G. Hartel, W. I. Segars, and P.
       Bush.   2001. Ribotyping  to determine the
       source of fecal coliform contamination in three
       household wells near Cochran, Georgia,  p.
       743-746. In: K. J. Hatcher (ed.) Proceedings
       of  the  2001  Georgia Water  Resources
       Conference, March  26-27, University  of
       Georgia, Athens, GA.

-------
33.    Samadpour, M., and N. Chechowitz. 1995.
       Little Soos Creek microbial source tracking: a
       survey, University of Washington Department
       of Environmental Health, Seattle, WA, 49p.

Pulsed-Field Gel Electrophoresis (PFGE)

34.    Barrett, T. J., Lior, H., Green, J. H., Khakhria,
       R., Wells, J. G., Bell, B. P., Greene, K. D.,
       Lewis J., and Griffin, P.M., 1994. Laboratory
       investigation  of  a   multistate  food-borne
       outbreak of Escherichia  coli O157:H7  by
       using  pulsed-field gel  electrophoresis and
       phage  typing, J.  Clin.  Microbiol,  32(12):
       3013-3017.

3 5.    Buchrieser, C., Gangar, W., Murphree, R. L.,
       Tamplin, M. L., and Kaspar, C. W., 1995.
       Multiple Vibrio vulnificus strains in oysters as
       demonstrated   by  clamped  homogeneous
       electric  field  gel  electrophoresis,  Appl.
       Environ. Microbiol., 61(3): 1163-1168.

36.    Johnson, J. M., Weagant, S. D., Jinneman, K.
       C.,  and  Bryant,  J. L.,  1995.    Use  of
       pulsed-field   gel   electrophoresis  for
       epidemiological  study of Escherichia coli
       015 7 :H7 during a food-borne outbreak, Appl.
       Environ. Microbiol., 61(7): 2806-2808.

37.    Kariuki,S., Gilks, C., Kimari, J., Obanda, A.,
       Muyodi, J., Waiyaki, P.,  and Haiti, C. A.,
       1999. Genotype Analysis of Escherichia coli
       Strains Isolated from Children and Chickens
       Living  in Close  Contact, Appl.  Environ.
       Microbiol. 65(2): 472-476.

38.    Parveen,  S.,  Hodge, N.  C.,  Stall, R. E.,
       Farrah, S. R., and Tamplin, M. L., 2001.
       Phenotypic and genotypic characterization of
       human and nonhuman Escherichia coli, Water
       Research, 35(2): 379-386.
39.    Simmons, G. M., Jr., S. A. Herbein, and C.
       M.  James. 1995. Managing nonpoint fecal
       coliform  sources to tidal inlets.  Universities
       Council  on   Water   Resources.  Water
       Resources Update, Issue 100:64-74.

40.    Simmons, G. M.,  Jr.,  and  S. A. Herbein.
       1998. Potential sources of Escherichia coli to
       children's pool in La Jolla, CA. Final Report
       for the City of San Diego and the County of
       San  Diego  Department  of Environmental
       Health.

41.    Simmons, G. M., D. F. Waye, S. Herbein, S.
       Myers,  and E.  Walker.  2000.   Estimating
       nonpoint  fecal coliform sources  in Northern
       Virginia's  Four  Mile  Run  watershed,  p.
       248-267.  In  T.  Younos and J. Poff (ed.),
       Abstracts,  Virginia   Water   Research
       Symposium 2000, VWRRC  Special Report
       SR-19-2000, Blacksburg, VA.

Polymerase Chain Reaction (PCR)

42.    Bernhard, A. E., and K. G.  Field. 2000. A
       PCR assay to discriminate human and ruminant
       feces  based   on  host  differences  in
       Bacteroides-Prevotella 16S ribosomal DNA.
       Appl. Environ. Microbiol. 66: 4571-4574.

43.    Dombek,  P. E.,  L.  K.  Johnson, S.  T.
       Zimmerley, and M. J. Sadowsky. 2000. Use
       of repetitive DNA sequences  and the PCR to
       differentiate Escherichia coli isolates  from
       human and animal sources.  Appl. Environ.
       Microbiol. 66(6): 2572-2577.

44.    Farnleitner, A. H., Kreuzinger, N., Kavka, G.
       G, Grillenberger, S., Rath, J., and Machl, R.
       L.,   2000.   Simultaneous   Detection   and
       Differentiation of Escherichia coli Populations
       from Environmental Freshwaters by Means of
       Sequence Variations in a  Fragment of the
       -D-Glucuronidase  Gene,  Appl.  Environ.
       Microbiol. 66(4):  1340-1346.

-------
45.    Koenraad, P.M.F.J., F. M. Rombouts, and
       S.H.W.  Notermans. 1997. Epidemiological
       aspects of thermophilic  Campylobacter in
       water-related environments: A review. Water
       Environ. Res. 69(1): 52-63.

46.    Kostman, J.R, Edlind, T. D., LiPuma, J. J.,
       and  Stull,  T.  L.,  1992.     Molecular
       epidemiology   of  Pseudomonas  cepacia
       determined by  polymerase chain reaction
       ribotyping,  J.   Clin.  Microbiol,   30(8):
       2084-2087.

Other methods

47.    Bernhard,  A.  E.,  and K. G. Field.  2000.
       Identification of nonpoint sources of fecal
       pollution   in   coastal  waters   by  using
       host-specific 16S  ribosomal  DNA genetic
       markers from fecal anaerobes. Appl. Environ.
       Microbiology 66(4): 1587-1594.

48.    Rhodes,  M.  W.,  and  H. Kator.  1999.
       Sorbitol-fermenting  bifidiobacteria   as
       indicators of diffuse human faecal pollution in
       estuarine  watersheds. J.  Appl. Microbiol.
       87:528-535.

49.    Souza,  V., Rocha, M.,  Valera,  A, and
       Eguiarte, L. E., 1999. Genetic Structure of
       Natural Populations of Escherichia coli in
       Wild Hosts on Different Continents,  Appl.
       Environ. Microbiology 65(8): 3373-3385.

ADDITIONAL INFORMATION

Dr. C. Andrew Carson
Department of Veterinary Pathobiology
College of Veterinary Medicine
University of Missouri
201 Connaway Hall
Columbia, MO 65211-5120

Dr. Charles Hagedorn
Department of Crop and Soil Environmental Sciences
Virginia Tech University
Blacksburg, VA 24061-0404
Dr. Valerie J. Harwood
Department of Biology
University of South Florida
4202 East Fowler Ave.
Tampa, FL 33620

Dr. Peter G. Hartel
Department of Crop and Soil Sciences
University of Georgia
3111 Plant Sciences Building
Athens, GA 30602-7274

Dr. Salina Parveen
USDA - Agricultural Research Service
Delaware State University
1200N. DuPontHwy
W.W. Baker Center
Dover, DE 19901

Dr. Michael J. Sadowsky
Department of Soil, Water, and Climate
University of Minnesota
258 Borlaug Hall
1991 Upper Buford Circle
St. Paul, MN 55108

Dr. Mansour Samadpour
Department of Environmental Health
School of Public Health & Community Medicine
University of Washington
Health Sciences Building
Box 357234
Seattle, WA 98195-7234

Dr. Bruce A. Wiggins
Department of Biology
James Madison University
Harrisonburg, VA 22807

The mention of trade names or commercial products
does not constitute endorsement or recommendation
for use by the U. S. Environmental Protection Agency
(EPA).

                 Office of Water
              EPA 832-F-02-010
                  May 2002

-------
For more information contact:

Municipal Technology Branch
U.S. EPA
1200 Pennsylvania Avenue, NW
Mail Code 4204M
Washington, D.C. 20460
          * 2002 *
          THEYEAR'OF
          CLEAN WATER
 yvrre
 Excellence in compliance through optimal tedmicat solutions
 MUNICIPAL TECHNOLOGY    *

-------