v>EPA
NATO CDSM
OTAN CCMS
NATO-CCMS
United States
Environmental Protection
Agency
Office of
Drinking Water
Washington DC 20460
EPA 570/9-84-006
CCMS-128
Water
Committee on the
Challenges of Modern Society
(NATO/CCMS)
Drinking Water Microbiology
NATO/CCMS Drinking Water
Pilot Project Series
CCMS 128
-------
-------
COMMITTEE ON THE CHALLENGES OF MODERN SOCIETY
(NATO/CCMS)
DRINKING WATER MICROBIOLOGY
Edited by:
Dean O. Cliver and Ruth A. Newman
With assistance from
Ross D. Pickford and Paul S. Berger
Office of Drinking Water
U.S. Environmental Protection Agency
NATO/CCMS Drinking Water Pilot Project Series
Joseph A. Cotruvo, Chairman
CCMS #128
1984
-------
DISCLAIMER
This report has been reviewed by the Office of Drinking
Water, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
-------
FOREWORD
Many of the industrial Nations today face problems related
to population, energy, and protection of the environment. in
order to optimize use of the scientific and technical expertise
from different countries, the Committee on the Challenges of
Modern Society (CCMS) was created between the Allied Nations of
the North Atlantic Treaty Organization (NATO). This international
society of scientists strengthens ties among members of the
North Atlantic Alliance and permits NATO to fill a broader social
role with non-member countries. CCMS has been responding to the
increasingly complex, technological problems facing modern society.
The Drinking Water Pilot Study was initiated by the U.S.
Environmental Protection Agency (EPA) in order to address a broad
spectrum of drinking water quality and health related issues.
Six subject areas have been studied by a number of groups
representing individuals from eleven NATO countries and three non-
alliance countries with technical participation from many others.
The conclusions and recommendations reached by the participants
are hoped to allow national programs to focus on specific areas
of water supply research and to bring out the most up-to-date
technology and practices. However, the recommendations and
conclusions do not necessarily reflect the policy of the U.S. or
any other participating countries. The work of the Pilot Project
has been covered in a summary report.
In addition three other major publications round out the
work of the pilot study - two being reports of international
symposia "Oxidation Techniques in Drinking Water Treatment".
(CCMS-111; EPA-570/9-79-020) and "Adsorption Techniques in Drinking
Water Treatment" (CCMS-112; EPA-570/9-84-005); and the third
being this document. EPA has been proud and pleased to associate
with such a fine group of scientists in this import work and
hopes to see these relationships continued into the future.
Joseph A. Cotruvo, USEPA
Chairman, Drinking Water Pilot Project
NATO/CCMS
-------
-------
FRAU PROFESSOR DR. GERTRUD MULLER
1919 - 1980
Frau Professor Dr. Gertrud Muller, Leitender Direktor
and Professor, Abt. "Spezielle Umwelthygiene, Humanfikologie
und Gesundheitstechnik," Institute fur Wasser-, Boden- und
Lufthygiene des Bundesgesundheitsamtes, Berlin, Federal
Republic of Germany, departed this world and the community
of scientists on 30 April 1980, ending a long and outstanding
career as a water and wastewater microbiologist and environ-
mental hygienist. She was a recognized leader in her fields
of endeavor in her own country and a frequent participant in
international professional activities. Among her more than
200 scientific publications were several books that are
regarded as classics in their fields. Those who knew and
worked with her were invariably impressed by the enthusiasm
and the positive approach with which she addressed whatever
she did. We, her colleagues in the writing, compilation,
and editing of Project Area III - Microbiology of the CCMS
Pilot Study on Drinking Water Supply Problems, fondly dedicate
this report of our work to her. She has been an inspiration
to us in this undertaking and will surely continue to be one
for our work in the years to come.
-------
-------
TABLE OF CONTENTS
Pages
List of participants .................. x
Preface .............. ......... xviii
Introduction ....... . .......
List of tables
List of figures . .
A. RAW WATER ..................... !
1. Microbiology of groundwater .......... 7
a. Definition of groundwater ... ...... 9
b. Occurrence of indigenous microorganisms
in groundwater .... ........... 10
xxvi
xxxi
c. Fate of organic substances in groundwater
13
New trends in the disposal of sludge and
sewage effluent ...... ........ 14
Factors affecting survival in soil
and groundwater .... ..... ..... 18
f . Occurrence of endotoxins in groundwater
25
g. Criteria for evaluating groundwater
quality .................. 27
h. Protection of groundwater ......... 30
i. Conclusion and recommendations . . ..... 31
-------
11
TABLE OF CONTENTS Continued
Pages
2. Microbiology of surface water 31
a. Occurrence of microorganisms in
surface water 31
b. Lakes and reservoirs 40
c. Toxic cyanobacteria in raw water supplies . 43
d. Sampling, transport, and microbiological
requirements for raw surface water 46
e. Protection of surface water 48
3. Survey of the bacteriological quality of
raw water supplies from nine countries ..... 50
a. Size of waterworks, source of raw
water, and sampling frequencies . 51
b. Bacteriological quality of raw
water supplies 55
c. Conclusion 55
4. Summary 61
v
B. MICROBIAL PATHOGENS TRANSMITTED BY WATER 63
1. Categories and properties of waterborne
microbial pathogens of humans 65
a. Bacteria 65
(i) Salmonella 65
(ii) Shigella 71
(iii) Yersinia enterocolitica 73
(iv) Enteropathogenic Escherichia
' coli 76
(v)
Francisella tularensis 78
-------
Ill
TABLE OF CONTENTS Continued
Pages
(vi) Leptospira 80
(vii) Vibrio cholerae .......... 82
(viii) Campylobacter ........... 85
(ix) Opportunistic pathogens 87
b. Viruses 90
c. Protozoa and metazoa 105
(i) Entamoeba histolytica 105
(ii) Giardia lamblia . . . . . . . . . .107
(iii) Naegleria fowleri Ill
(iv() Acanthamoeba species 113
(v) Helminths JL14
d. Gastroenteritis of undeter-
mined etiology ............... 116
2. Sources of waterborne pathogens ........ 120
3. Persistence and death of pathogens
in water . . . . ... ... . 126
4. Infectivity of waterborne pathogens ...... 135
5. Epidemiology of waterborne
infectious diseases 139
6. Summary 157
7. Recommendations . 158
-------
IV
TABLE OF CONTENTS Continued
Pages
C. INDICATOR SYSTEMS FOR MICROBIOLOGICAL
QUALITY AND SAFETY OF WATER 168
1. Established viable indicator systems . ...... 169
a. Colony count ... 169
b. Total coliforms 174
c. Thermo-tolerant coliforms and
Escherichia coli 180
d. Fecal streptococci 184
e. Sporeformers 193
2. Proposed viable indicators 197
a. Coliphages 198
b. Vaccine polioviruses 202
c. Pseudomonas aerugiriosa 206
d. Klebsiella . 209
e. Bif idobacterium 213
f. Candida albicans 219
g. Vibrio 222
h. Aeromonas 225
3. Other indicators of microbial quality 232
a. Limulus amebocyte lysate . . . 232
b. Adenosine triphosphate 235
c. Fecal sterols 240
-------
V
TABLE OF CONTENTS Continued
5,
6,
Pages
Potentials for mechanization, automation,
and shorter read-out times . 246
a. Radioactive isotope methods 247
b. Impedance methods . . 252
c. Automated sampling, plating
and incubation 257
d. Automated plate counting 259
e. Automated enzymatic methods ........ 263
f. Rapid thermo-tolerant coliform
MF method . 269
g. Presence absence test 271
h. Membrane filter procedure for US
Standard plate count ... 281
i. Epifluorescence total cell count .... 283
Summary ...... 287
Recommendations 291
D. TESTING AND STANDARDS 293
1. Sampling and sample storage 294
2. Total coliform testing, membrane
filter technique ... 294
3. Total coliform testing, multiple tube
technique ...... 305
4. Thermo-tolerant (fecal) coliform testing,
membrane filter technique 313
5. Thermo-tolerant (fecal) coliform testing,
multiple tube technique 313
6. Colony count 322
-------
vi
TABLE OF'CONTENTS Continued
Pages
7. Miscellaneous ................. 326
8. Summary and conclusions ............ 326
9. Recommendations ................ 327
E. MICROBIOLOGY OF WATER TREATMENT
AND DISINFECTION 329
1. Storage reservoirs 329
2. Coagulation ............ 339
3. Sand filtration . 348
4. Activated carbon filtration . . 358
5. Disinfection . . . . . . . . .... . . . . . . 361
a. Chlorination 361
b. Ozone treatment 375
c. Chlorine dioxide .............. 386
d. Iodine 389
6. Summary .................... 393
7. Recommendations 396
F. DISTRIBUTION SYSTEMS 397
1. Service reservoirs . . . . . . . ... . ; . . . 398
a. Direct contamination from sewage 398
b. Contamination of open
service reservoirs . . 398
c. Contamination from leakage ... 399
d. Growth in bottom deposits 400
-------
TABLE OF CONTENTS Continued
Vll
Pages
2. Aftergrowth 401
a. Influence of nature of raw water and
methods of treatment on aftergrowth . . . .401
b. Accumulation of organic debris in
dead ends and other protected areas .... 404
c. Interrelationships between organic
matter, bacterial growth, and
animals in the distribution systems .... 405
d. Chemical changes in distri-
bution systems ....... 406
3. Growth on materials
a,
409
Ability of various microorganisms to
grow on some materials used in dis-
tribution and plumbing systems ....... 409
b. Problems that can be caused by
microbial growth 411
c. Methods of testing materials . 413
. . . 416
d. Standards and regulations relating
to materials in contact with water .
e. Growth on atmospheric volatile
organic chemicals ..... . 417
4. Back-siphonage and cross-connections ...... 419
a. Types of back-siphonage and cross-
connections and risks involved 419
b* Regulations concerning back-siphonage
421
c. Mechanical devices and
hydraulic safeguards . 421
-------
viii
TABLE OF CONTENTS Continued
Pages
5. Main laying and repair 422
a. Codes of practice to prevent
contamination . . 422
b. Problems with jointing materials 423
c. Cleaning, disinfection, and
sampling procedures 424
d. Acceptable standards 426
6. Summary 426
7. Recommendations 429
G. TECHNOLOGICAL ASPECTS 431
1. Biodeterioration of materials 432
a. Biodeterioration of rubber
sealing rings 432
b. Biodeterioration of sealants
and mastics 437
2. Microbial growth on resins and
filter media 438
a. Filters 439
b. Ion exchangers ....439
c. Portable filters 439
d. Water softeners that add phosphorus .... 440
' e. Conclusion 440
3. Hydraulic effects of microbial growth 441
a. Types and diameters of pipes ........ 441
b. Microbial oxidation of manganese
in pipes 441
-------
IX
TABLE OF CONTENTS Continued
c. Microbial growth from improper
installation of household equipment
Pages
442
d. Microbial growth from
cross-connections 442
4. Microbially mediated chemical cycles ...... 443
a. Microbial manganese oxidation
affecting wells and water
distribution systems 443
b. Iron and sulfur bacteria corroding
well casings and other structures ..... 444
5. Drinking water supply for ships 449
6. Water in containers 453
7. Summary ... ......455
8. Recommendations . . 456
GENERAL SUMMARY . 457
GENERAL RECOMMENDATIONS 471
REFERENCES 474
-------
LIST OF PARTICIPANTS
Project Area III: Microbiology
Canada
Mr. J. A. Clark, Supervisor
Microbiology Section, Laboratory Services Branch
Ministry of the Environment
P.O. Box 213
Rexdale, Ontario
Dr. B. J. Dutka, Head
Microbiology Laboratories
Applied Research Division
Canada Centre for Inland Waters
867 Lakeshore Rd., P.O. 5050
Burlington, Ontario L7R 4A6
Dr. Paul R. Gorham
Department of Botany
University of Alberta
Edmonton, Alberta T6G 2E9
Dr. Richard S. Tobin - Leader, Topic C & National Contact
Criteria Section
Environmental Standards Division
Environmental Health Centre
Tunney's Pasture
Ottawa, Ontario K1A OL2
Denmark (not a participating country)
Prof. Dr. Gunnar G. Bonde
Institute of Hygiene
University of Aarhus
Aarhus
-------
XI
Denmark (cont.)
Dr. Kaj Krongaard Kristensen
Vandkvalitetsinstitutet
Agern Alle 11
2970 H^rsholm
Federal Republic of Germany
Professor Dr. Gertrud Mulier - Leader, Topic G
Institut fur Wasser-, Boden- und Lufthygiene
des Bundesgesundheitsamtes
Postfach
D-1000 Berlin 33
Dr. Peter Scheiber - National Contact
Medizinal-Untersuchungsamt
6100 Darmstadt
Wilhelminstr. 2
Postfach 11 07 61
Professor^Dr. Reinhard Schweisfurth
Universitat des Saarlandes
Institut fur Hygiene und Mikrobiologie
Universitatskliniken, Medizinische Fakultat
D-665 Homburg
Professor Dr. R. Schubert
Department of General & Environmental Hygiene
Centre of Hygiene
Paul-Erlich-Strasse 40
6 Frankfurt a/M
France "
Dr. Jean-Claude Block
Laboratoire de Microbiologie
U. E. R. d'Ecologie
Universite'de Metz
1, rue des Recollets
57000 Metz
Professor Dr. J. M. Foliguet - Leader, Topic B,
Laboratoire d1Hygiene et de Recherche en Sante Publique
Universite de Nancy 1
Facultlf de Medecine "B"
Avenue de.la Foret de Haye
54500 Vandoeuvre-les-Nancy
-------
xii
France (cont.)
Dr. P. Hartemann - Leader, Topic B ^
Laboratoire d1Hygiene et de Recherche en Sante Publique
Universitef de Nancy 1
Faculte' de Medecine "B"
Avenue de la Fore*t de Haye
54500 Vandoeuvre-les-Nancy
Professor H. Leclerc ^
Institut National da la Sante et de la Recherche Medicale
Unite No. 146
Ecotoxicologie Microbienne
Domaine du C.E.R.T.I.A.
369, rue Jules Guesde
59650 Villeneuve-D'Ascq
P. A. Trinel ^ ^
Institut National de la Sante et de la Recherche Medicale
Unite No. 146
Ecotoxicologie Microbienne
Domaine du C.E.R.T.I.A.
369, rue Jules Guesde
59650 Villenueve-D'Ascq
Dr. J. Vial - National Contact
Laboratoire Regional d1Hygiene
de 1'Homme et de son Environnement
Institut Pasteur de Lyon
77, rue Pasteur
69300 - Lyon Cedex 2
Greece
Professor J. A. Papadakis
Athens School of Hygiene
L. Alexandras 196
Athens
- National Contact
Israel
Dr . U . Bachr ach
Department of Molecular Biology
The Hebrew University - Hadassah Medical School
Jerusalem, Israel
-------
X1Z1
Israel (cont.)
Professor Dr. Yehuda Kott - National Contact
Environmental & Water Resources Engineering
Technion-Israel Institute of Technology
Technion City, Haifa 32 000
The Netherlands
Dr. A. H. Havelaar
National Institute of Public Health
P.O. Box 1
Bilthoven
Dr. H. J. Kool - Leader, Topic E and National Contact
Rijksinstituut voor Drinkwatervoorziening
Postbus 150
2260AD Leidschendam
Dr. M. van Schothorst
Laboratory for Zoonoses & Food Microbiology
Rijks Instituut voor de Volksgezondheid
Antonie van Leeuwenhoeklaan 9
Postbus 1, Bilthoven
Norway
Dr. J. Kvittingen
Moholtlia 30
7000 Trondheim
Dr. J^rgen Lassen
National Institute of Public Health
Postuttak
Oslo 1
LTC, D.V.M. Jens J. Nygard - Leader, Topic A
Assistant Chief Army Veterinarian
Norwegian Institute for Water Research
Gaustadalleen 25
P.O. Box 333 Blindern
Oslo 3
-------
XIV
Norway (cont.)
Dr. Tov Omland - National Contact
Norwegian Defence
Microbiological Laboratory
National Institute of Public Health
Postuttak
Oslo 1
Mrs. Kari Ormerod
Norwegian Institute for Water Research
P.O. Box 333, Blindern
Oslo 3
Sweden (not a participating country)
Dr. Thor Axel Stenstrom - Observer
The National Bacteriological Laboratory
S-105 21 Stockholm
United Kingdom
Dr. Norman P. Burman - Leader, Topic F
Formerly, Manager of Thames Water Authority
Metropolitan Water Services
9, Derby Road
Surbiton
Surrey, KT5 9AY
Mrs. L. M. Evison
Department of Civil Engineering
University of Newcastle Upon Tyne
Clairmont Road
Newcastle Upon Tyne
NE1 7RU
Dr. J. W. Ridgway
District Division
Water Research Centre
P.O. Box 16, Henley Road
Medmeriham, Marlow, Buckinghamshire
SL 7 2HD
-------
XV
United Kingdom (cont.)
Dr. J. S. Slade, Virologist
Thames Water Authority
New River Head Laboratories
177 Rosebery Avenue
London, EC1R 4TP
Miss J. Stevens - Leader, Topic F
Chief Microbiologist
Thames Water Authority
New River Head Laboratories
177 Rosebery Avenue
London, EC1R 4TP
& National Contact
United States
Dr. Elmer W. Akin
Research Virologist
Environmental Protection Agency
HERL (Health Effects Research Lab.)
26 W. St. Clair St.
Cincinnati, Ohio 45268
Dr. J. D. Buck
Marine Sciences Institute
Marine Research Lab.
P.O. Box 278
Noank, Connecticut 06340
Dr. S. L. Chang
1035 Juanita Drive
Walnut Creek, California 94595
Pfofessor Dean O. Cliver - Project Leader
Food Research Institute
University of Wisconsin
1925 Willow Drive
Madison, Wisconsin 53706
Professor R. R. Colwell
Department of Microbiology
University of Maryland
College Park, Maryland 20742
-------
XV X
United States (cont.)
Professor Koby T. Crabtree
Professor of Microbiology
Chairman, Center System - Department of Biological Sciences
University of Wisconsin
518 So. 7th Avenue
Wausau/ Wisconsin 54401
Mr. Edwin E. Geldreich, Chief
Microbiological Treatment Branch
Water Supply Research Division
Municipal Environmental Research Lab.
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Dr. B. L. Green
Department of Environmental Sciences
Marshall Hall
University of Massachusetts
Amherst, Massachusetts 01003
Mr. Arnold E. Greenberg - Leader, Topic D
Chief, Bioenvironmental Laboratories
California Department of Health
2151 Berkeley Way
Berkeley, California 94704
J. E. Hobbie
The Ecosystems Center
Marine Biological Laboratory
Woods Hole, Massachusetts 02543
Dr. John C. Hoff
Microbiological Treatment Branch
Water Supply Research Division
Municipal Environmental Research Lab.
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Professor Warren Litsky
Department of Environmental Sciences
Marshall Hall
University of Massachusetts
Amherst, Massachusetts 01003
-------
XV11
United States (cont.)
Ruth Newman - Project Editor
Food Research Institute
University of Wisconsin
1925 Willow Drive
Madison, Wisconsin 53706
Melvin P. Silverman
National Aeronautics & Space Administration
Ames Research Center
Moffett Field, California 94035
Professor Otis J. Sproul, Chairman
470 Hitchcock Hall
Ohio State University
Department of Civil Engineering
2070 Neil Avenue
Columbus, Ohio 43210
Susan Stramer
Food Research Institute - Department of Bacteriology
University of Wisconsin
1925 Willow Drive
Madison, Wisconsin 53706
Dr. S. W. Watson
Woods Hole Oceanographic Institute
Woods Hole, Massachusetts 02543
Dr. C. K. Wun
Department of Environmental Sciences
Marshall Hall
University of Massachusetts
Amherst, Massachusetts 01003
-------
XV111
PREFACE
The Pilot Study on Drinking Water Supply Problems was
performed under the auspices of the Committee on the
Challenges of Modern Society, to examine problems or special
concern to industrialized nations. -Increasing technology
has "helped developed countries deal with some of the drink-
ing water supply problems that still trouble less affluent
nations, but other problems arising from industrialization
and urbanization are just beginning to surface.
This report on the microbiology of drinking water was
compiled by approximately 50 experts from 11 countries (cf.
preceding list). Each of the seven topics comprising the
report was led by a person from a different country, and
most of the working groups were international in compo-
sition.
We intended to incorporate into this project all
aspects of drinking water microbiology that have practical
significance. We considered the entire continuum from the
source water, through treatment and delivery to the con-
sumer. We included organisms which may affect human health
directly, those which may have an indirect influence (such
as indicator organisms), and those which cause technologic
problems in the provision of safe drinking water. Thus, the
scope of our undertaking may well be broader than that of
any previous project in the area of drinking water micro-
biology.
On the other hand, extreme breadth necessarily limits
depth. We chose to try to define the problems, identify
available solutions, and propose research that is required
to produce solutions that do not yet exist. Both published
and unpublished research results have been included; but we
have often directed the reader to a review on some broad
subject, rather than a multitude of older source articles.
We have tried to produce a text which is useful and meaning-
ful to a career microbiologist who may not be an expert on
the aspects of drinking water microbiology. The summary of
the report is directed to those with a less specialized
technical background.
-------
XIX
The seven topics are intended to cohere as much as
possible, but the topic leaders have chosen a variety of
formats on the basis of suitability to the kinds of material
they had to present. The references cited are principally
those which are likely to be available in technical li-
braries of industrialized nations.
Topic A, led by Dr. J. J. Nygard of Norway, deals with
the microbiology of raw water. Groundwater and surface
water are considered separately, and the report includes a
survey of microbiologic experience with raw water in several
countries.
Topic B, led by Prof. J. M. Foliguet with Dr. P.
Hartemann and assisted by Dr. J. Vial, all of France, con-
cerns the pathogens which are transmissible through drinking
water. Following a survey of the properties of individual
pathogenic species, the report addresses the general ques-
tions of sources, persistence, and infectivity of waterborne
pathogens and the compilation of statistics on waterborne
disease.
Topic C, led by Dr. R. S. Tobin of Canada, discusses
indicator systems. This includes both indicators which are
already established and in general use as well as others
which are still at the research stage. Some are primarily
indicators of fecal contamination, whereas others relate to
different aspects of microbiologic quality. The discussion
also includes techniques presently under study which are
designed to be potentially more rapid or convenient and do
not rely upon the growth of viable organisms.
Topic D, compiled by Mr. A. E. Greenberg of the United
States from materials supplied by respondents in thirteen
countries, concerns testing and standards for drinking
water. Tables are used to make comparisons among nations as
to their requirements or recommendations for sampling and
testing procedures. Methods for determining coliforms and
thermo-tolerant coliforms (Escherichia coli) are emphasized.
The existence and effects of standards are compared.
Topic E, led by Dr. H. J. Kool of the Netherlands,
addresses processes used in drinking water treatment. The
effects of various unit processes are considered from the
microbiologic standpoint. Where possible, effects on patho-
gens are contrasted with effects on indicator organisms.
Topic F, which concerns the distribution of drinking
water, was written by Dr. N. P. Burman and Miss J. K. Stevens,
-------
XX
both of the United Kingdom. The discussion emphasizes
protecting the microbiologic quality of finished water
during storage and on the way to the consumers. Means of
preventing growth of organisms within the system as well as
contamination by organisms from without are considered.
Topic G, led by Prof. G. Muller of the Federal Republic
of Germany, deals with technological aspects of potable
water microbiology. One broad area of discussion concerns
the influence of microorganisms upon the structures or
functions of devices used in treating and distributing
drinking water. Another area is concerned with water held
in containers. These include drinking water reservoirs
aboard ships, or cans and bottles used either for commercial
or emergency use.
The summary which concludes the text of the report is
intended to be of value to both policymakers and the public.
It attempts to provide an overview, from the microbiologic
standpoint, of the technical and institutional problems
involved in providing a safe and quantitatively adequate
supply of drinking water from whatever raw water is avail-
able. It also emphasizes the need for continuing research
to fill the gaps in present knowledge.
-------
XXI
INTRODUCTION
Throughout the course of history, the development of a
safe drinking water supply has been crucial to the public
health of nations. Drinking water is a necessity and approxi-
mately 2 liters/day are required to sustain life. Despite
the body's many special mechanisms for conserving water,
substantial quantities are lost each day and must be re-
placed promptly. Much of the replacement is derived from
water that is present in food, some is a product of metabo-
lism; the rest must be obtained by drinking water or other
beverages, all of which contain water. Water must be drunk,
even if what is available is unsafe, in the sense that
drinking it may produce adverse effects upon health. Post-
ponement of meeting the body's needs for beverage water can
lead to extreme illness and death within a very few days.
Because people have virtually no option but to drink tne
water that is available to them, it is a fundamental respon-
sibility of governments to ensure a continuous supply of
safe drinking water, to provide alternative sources, or
provide for conservation.
Water does not occur in an absolutely pure state any-
where in the world. By and large, the chemical impurities
in both ground and surface waters govern, through their
chemistry, the kinds of native bacterial populations that
ordinarily grow in the water. The organisms are generally
found in low numbers and are of little public health con-
sequence as long as the water contains minimal concen-
trations of organic, nitrogen, and phosphorus compounds as basal
nutrients. Pollution of both ground and surface water
supplies, much of which is domestic sewage, is becoming a
serious problem for industrialized countries. These wastes
not only contain pathogenic microorganisms, but also varying
amounts of basal nutrients which can support the growth of
various microorganisms and as a result, cause taste, odor,
or fouling problems for some water supplies.
Historically, the agents causing typhoid fever, bacil-
lary dysentery, and cholera have been of primary concern to
the sanitary microbiologist and public health officials.
-------
XX11
These diseases are only rarely observed now in industrial
countries which efficiently practice modern means of water
treatment and waste disposal. Other disease producing
bacteria, viruses, and protozoa, including Giardia lamblia,
are receiving attention probably because of our general
awareness of the organisms in some source waters and our
ability to recover them from the water. Attention has
focused, too, on the opportunistic pathogens which are
primarily saprophytic and can cause disease in debilitated
individuals. These bacteria have the ability to multiply
in situ and in distribution systems while the more fas-
tTdious bacteria, viruses, and protozoan cysts cannot com-
pete in the aquatic environments. ,
Pathogens cannot usually be sought as a means of
determining the microbiologic quality or safety of drinking
water. Levels of pathogens in finished water and in highly
protected source waters are essentially nil, most of the
time, because pathogens are likely to be so fastidious that
only highly specialized laboratory techniques may allow them
to be detected. Therefore, bacterial species which are
generally nonpathogenic, but are more readily cultivated in
the laboratory, have come to be used as indicators of water
quality and safety. Some of these indicators may be taken
as evidence that water is contaminated with feces. Some of
these indicators can show that water treatment has been
inadequate or that an abnormal or undesirable situation
exists, but none can be directly correlated with the pres-
ence of specific pathogens. Other indicator systems, some
of which do not entail the growth of viable organisms from a
sample, have been proposed to detect endotoxin or ATP or to
provide results more rapidly than is possible by techniques
presently in routine use. No microbiologic indicator system
will serve all of these purposes, nor does any system serve
even a single purpose perfectly. Nevertheless, the coliform
indicator system has been used for many years in monitoring
water quality and, with minor refinements, has stood the
test of time reasonably well.
Routine testing for microbiologic indicators has been
the traditional "means of evaluating the safety and reli-
ability of public water supplies for more than 80 years.
Sampling and microbiologic testing procedures to determine
drinking water quality have been established in every indus-
trialized nation, yet the procedures used are not always the
same. Potentially significant differences exist among
nations as to where and with what frequency water samples
are to be taken, how samples are to be tested when they
-------
xxiii
reach the laboratory, how the laboratory test results are to
be interpreted, and what remedial action is appropriate when
indicator test results exceed established norms. In com-
paring activities it is important to note that commendable
and much-needed efforts are already under way toward stan-
dardizing microbiologic testing of water through a number of
existing organizations.
Water for potable use must be taken from the best
source available to the community. If the quality of the
source does not meet appropriate criteria and standards for
safety, treatment must be developed in order to protect the
public health. Water treatment techniques are well estab-
lished and sophisticated in industrialized nations. This
supports the observation that waterborne disease is rela-
tively rare in these countries, even though many of the
suppliers are obliged to prepare drinking water from sources
that are not ideal in purity. However, even procedures that
have long been in use at a given location may have to be
modified if raw water quality suddenly deteriorates, if the
public that is served demands a higher degree of water
purification than was previously practiced, or if reductions
in the treatment costs, in terms of money or energy, are
sought. As a result of long experience, we know the
effects each of the major water treatment procedures are
expected to produce upon the pathogens that may be present
in the water and what effect changes in treatment may have
on the indicator systems by which the quality of the water
is monitored. Few unit processes in water treatment, other
than disinfection, have been designed specifically to act
upon the microorganisms present in the water. Therefore,
reductions of pathogens as a result of these processes are
largely fortuitous. Furthermore, a given unit process acts
differently upon different classes of pathogens, so it is
not at all surprising that bacterial indicators of sanitary
significance may not be affected in the same way or to the
same degree as particular pathogens. Information concerning
treatment efficacy can aid in assessing the microbiologic
effects of these water treatment processes and in predicting
the results of changes that may be contemplated.
No matter how excellent the quality of drinking water
may be when the treatment plant is finished, the respon-
sibility of the water utility must extend to safe storage
and distribution of the water to the service connection at
the point where the water is used. Where proper dis-
tribution and storage facilities do not exist, they must be
-------
XXIV
built. Indeed, the epidemiologic record of disease out-
breaks involving public water supplies suggests that recon-
tamination of finished water during distribution is one of
the major problems confronting water supplies in all coun-
tries. Community distribution systems are inevitably com-
plex and have frequently been constructed in stages over
long periods of time, so that construction practices pres-
ently regarded as standard may be represented in only a
portion of an existing network. Use of substantial volumes
of water for firefighting, industrial applications, or the
sudden incidence of natural or manmade catastrophe, can
disrupt well-considered and established water distribution
procedures. Therefore, water distribution systems must be
adaptable to a variety of uses, and decisions for new
construction must be integrated into the overall land use plan-
ning exercise for the community.
Though the principal emphasis in this report has,
necessarily, been on the effects of water and of drinking
water technology upon microorganisms, it is also obvious
that some microorganisms are capable of significant effects
upon their immediate environment and that these effects must
be considered in discussing water treatment and dis-
tribution. Problem areas include the degradation of sur-
faces in contact with drinking water, as a result of organic
or inorganic reactions induced by microbes, and the influ-
ence of accumulated microbial cells upon the functions of
water treatment resins and on the mains through which water
is expected to flow. The technological problems which may
result from microbial growth during treatment and dis-
tribution of public water supplies are part of the task of
providing water to the consumer; however, drinking water, in
some situations, must be held in a static condition for
substantial periods of time. Two such applications are that
of supplying drinking water aboard ships and that of pack-
aging drinking water in containers for emergency use or for
commercial distribution. Each of these technological prob-
lems may appear to lie outside the traditional scope of
drinking water microbiology, but ,are included because they
are significant to the task of providing safe, palatable
drinking water to the consuming public.
This report on drinking water microbiology is, as was
previously stated, intended to treat the subject as broadly
as possible. Although it is clear that there are many
problems in water supply which are not of a microbiologic
nature,
-------
XXV
it must be noted that microbes are probably the oldest form
of life and that they almost certainly originated in water.
Therefore, the human species, as a late and opportunistic
arrival on the face of the earth, must to some degree base
its strategies for survival upon understanding and learning
to deal with the interactions between microorganisms and
water. In this context, we trust that the report which
follows will contribute significantly to our collective
potential to survive.
-------
XXVI
LIST OF TABLES
A. 1-1
A.1-2
A. 1-3
A.1-4
A. 1-5
A.1-6
A. 1-7
A.1-8
A.1-9
A. 2-1
A. 2-2
A. 2-3
Composition of the Hydrosphere
and Rate of Turnover
Pages
8
Content of Bacteria and Organic Matter
at Different Soil Depths 12
Survival of Bacteria on or in Soil 19
Survival of Bacteria in Groundwater .... 21
Movement of Bacteria through Soil 22
Median Bacterial Counts (per gram) in
Four Replicate Samplings of a Single
Site at Various Times after Application
of Digested Sludges
24
Bacteria in Groundwater beneath Infil-
tration Ponds Sampled in January and
February of 1977 26
Microbiological Summary of Completed
Groundwater Surveys - 28
Microbiological Requirements for Drink-
ing Water in the European Economic
Community 29
Colony Counts in Four Rivers
of Czechoslovakia 34
River Water Quality as Determined by
Numbers of Chemoorganotrophs at 22°C
35
Classification of Chemolithotrophic
Bacteria 36
-------
XXV11
LIST OF TABLES Continued
A. 2-4
A.3-1
A.3-2
A.3-3
A.3-4
A.3-5
A.3-6
A. 3-7
A.3-8
B.l.b-1
B.2-1
B.3-2
B.4-1
Pages
Proposed Water Quality Criteria to be
Used for Raw Surface Water Supplies ..... 47
Recorded Waterworks by Population Served ... 52
Recorded Waterworks Serving More
than 1,000 People 53
, Frequency of Bacteriological Analyses
of Raw Water at the Recorded Waterworks
54
Bacteriological Quality of Surface
Water: Total Coli forms 56
Bacteriological Quality of Surface
Water: Thermo-tolerant Coliforms 57
Bacteriological Quality of Ground-
water: Total Coliforms
58
Bacteriological Quality of Ground-
water: Thermo-Tolerant Coliforms ...... 59
Bacteriological Quality of Raw
Water Receiving Minimum or no Treatment
60
Human Enteric Viruses that may be
Present in Contaminated Water . 92
Sources of Waterborne Pathogens 122
Persistence of Some Enteroviruses
in Water 134
Median Peroral Infective Dose of Water-
borne Pathogens cited in this Section .... 137
-------
LIST OF TABLES Continued
xxviil
Pacres
B.5.a-l
B.5.a-2
B.5.a-3»
C.l.a-1
C.l.d-1
C.2.e-l
C.2.e-2
C.2.e-3
C.2.h-l
C.4.e-l
C.4.f-l
C.4.g-l
C.4.h-l
D-l
Waterborne Disease Outbreaks in the U.S.,
1971-1975
Waterborne Outbreaks in the U.S.,
1971-1975, by Type of System . .
Deficiencies resulting in Outbreaks of
Waterborne Disease in U.S. ......
Colony Count Criteria for Various
Types of Water in Several European
Comtunities
Thermo-Tolerant Coliform (TTC)/Fecal
Streptococcus (FS) Ratios and FS
Distributions in Warm-Blooded Animals
Modified Gyllenberg and Niemela
Medium Formulation
YN-6 Medium Formulation and Preparation
Groupings of Bifidobacterium Species
Distribution of Aerogenic Versus
Anaerogenic Aeromonas in Water .
BLB Formulation
M 7-h FC Medium Formulation
Identification and Relative Frequency of
Cultures from Raw and Drinking Water Samples
M-SPC Medium Formulation
Sampling and Sample Storage, Drinking
Water Analysis
142
143
145
175
186
215
216
218
2.29
268
270
276
282
295A
-------
XXIX
LIST OF TABLES Continued
D-2
Total Coliform Testing, Membrane
Filter Technique, Drinking Water Analysis
Pages
. 297A
D-3 Total Coliform Testing, Multiple Tube
Technique, Drinking Water Analysis . . . . . . 306A
D,-4 Thermo-tolerant (Fecal) Coliform
Testing, Membrane Filter Technique,
Drinking Water Analysis . 314A
D-5 Thermo-tolerant Coliform Testing,
Multiple Tube Technique, Drinking
Water Analysis ............... . 319A
D-6 Colony Count, Drinking Water Analysis .... 323A
Scheme of Treatment for the Preparation
E.Intro-1 of drinking Water 33°
E.2-1 Removal of Bacteria by Coagulation
With Aluminum Sulfate and Sedimen-
tation 341
E.2-2 Removal of Enteric Viruses by
Coagulation and Sedimentation 344
E.2-3 Removal of Parasitic Cysts by Coagulation
with Aluminum Sulfate and Sedimentation . . . 349
E.5.a-l Lethality Coefficients for Different
Microorganisms Based on Treatment with
Free and Combined Chlorine at 5°C 366
E.5.a-2 Dosages of Various Chlorine Species
Required to Inactivate 99% of Escherichia
coli and Poliovirus 1 367
E.5.a-3 The Oxidation-Reduction Potential of
Chlorine in Water . 369
-------
LIST OF TABLES Continued
XXX
Pages
E.5.b-l
E.5.b-2
E.6-1
Summary of Data From Studies of
Bacterial Inactivation by Ozone in Water . . . 380
Summary of Data from Studies of Viral
Inactivation by Ozone in Water ' . 382
Estimation of the Percentage of Micro-
organisms Removed by Various Water
Treatment Processes .
395
-------
XXXI
LIST OF FIGURES
Pages
A.1-1 Degradation of Organic
Substances in the Presence of Suf-
ficient Amounts of Oxygen 15
A.1-2 Degradation of Organic Substances
with Insufficient or no Oxygen . 16
B.l.c-1 Relationship between Cysticidal
Residual , Titratable Iodine (I2)
and Contact Time at 3°, 10°, 23^
and 35°C 108
B.l.c-2 Relationship of Contact Time to
Concentration of I2 and HOI for
Destruction of 99.9 percent of
Cysts, Viruses, and Bacteria at 18°C 109
C.4.e-l Automatic Apparatus for Con-
tinuous Testing for 13. coli
in Water: Schematic Diagram 265
C.4.e-2 Sequential Diagram of Intro-
duction of Samples, Culture
Medium and Decontaminating
Solution A & B: Pneumatic
Lifters Alternately Closing
Off the Tubes Supplying
Culture Medium, Water Sample
and Decontaminatina Solution 266
Scheme for Identification of Cultures
C.4.g-l from P-A Test 277
E.5.a-l Diagrammatic Representation
of Completed Breakpoint Reaction for Chlorine. 364
E.5.a-2 Time-Concentration Relation-
ship in Destruction of Cysts
of Entamoeba Histolytica by
HOC1 at 3° and 23°C 372
-------
XXX11
LIST OF FIGURES Continued
E.5.a-3
E.S.b-1-
G.4.b-l
G.4.b-2
G.4.b-3
Pages
Effect of Temperature on
Cysticidal Efficiency of
NH2C1 and NHC12
Diagrammatic Representation of Completed
Breakpoint Reaction Ozone
373
378
Cycle of Biological Corrosion . .
. 446
Formation of Different Forms of
Iron Deposits in Water, Pipes 447
Schematic Representation of a
Section of a So-called Rust Knob
448
-------
A. RAW WATER
The ultimate source of water in various freshwater
environments is precipitation, be it rain, snow, or hail.
From there, it evaporates, is transpired, runs off directly
into streams, rivers, and lakes or it percolates through the
soil to travel eventually into groundwater aquifers. As
water makes its descent to the ocean, it may pass through both
the groundwater state and the surface water state. The increas-
ing demand for potable water to supply domestic and commercial
needs has prompted many communities of the world to use new
source waters, including those receiving wastewater from
homes, communities, and industries (Dykes, e_t a_l., 1967).
As a rule any body of-fresh water, whether or not it
receives wastewater, contains a great number of microbes,
among which may be bacteria, algae, protozoa, fungi, and
viruses (including phages). Some of these microbes are
indigenous to natural bodies of water whereas others are
transient,, entering the water from air or soil and, more
significantly, from domestic or industrial wastewater. The
presence, activities, and interactions of both indigenous
and transient microbes .are extremely important because they
may affect the health of humans and other animals. The
persistence of these organisms in any given water varies
from hours to months and is influenced by a variety of
environmental conditions (e.g., available nutrients, pH,
temperature, presence of toxic substances, competition,
predation, etc.). The complexity with which these factors
interact is even greater if the body of receiving water is
flowing, as in the case of a river. ,
Chemolithotrophs, as the name implies, obtain their
energy from the oxidation of.simple inorganic elements or
compounds (i.e., ammonia, sulfur, reduced iron and/or manganese)
Thus, all of these organisms play an important role in the
cycling of the elements. Chemolithotrophic bacteria are
divided into: (1) obligate chemolithotrophs (able to oxidize
only certain specific inorganic substances as sources of
energy but can derive their carbon from any of several
-------
sources); (2) heterotrophic chemolithotrophs (can utilize
several different sources for energy but are wholly re-
stricted to carbon dioxide as their carbon source); and (3)
mixotrophic chemolithotrophs (can obtain both energy and
carbon from different sources).
The prevalence of chemolithotrophs is influenced by
various factors. For example, nitrifiers (members of the
family Nitrobacteraceae) are strict aerobes, preferring a pH
slightly above 7.01. On the other hand, some sulfur-oxidizers
require a pH of 1.0 to 3.5. Whereas the growth of some
microorganisms is inhibited by the presence of organic
compounds, others are mixotrophic and can function both as
chemolithotrophs and as chemoorganotrophs (Alexander, 1971).
Chemolithotrophs play various important ecological
roles within source waters. For example, Nitrobacteraceae
convert ammonia to nitrite and nitrite to nitrate. Nitrate
is a universal nitrogen source for green plants and, when
present, stimulates the growth of aquatic weeds which, in
turn, may temporarily immobilize the nitrate by removing it
from circulation. However, the subsequent death and decom-
position of such weeds within source waters result in taste
and odor problems along with oxygen depletion.
Sulfur bacteria, such as members of the genus Thio-
bacillus, are facultative and as such, can thrive in
aerobic as well as anaerobic environments [See Section
G.4.c] They oxidize both reduced sulfur (such as hydrogen
sulfide) and reduced iron (ferrous compounds). Their
activities, therefore, are largely beneficial since reduced
sulfur compounds (such as hydrogen sulfide) not only impart
a distinctive odor to potable water, but may be toxic to
animals, given sufficient quantities.
Members of the family Siderocapsaceae can deposit iron
and/or manganese oxides on their capsules. Their presence
in source water may be beneficial where large amounts of
reduced iron and/or manganese (agents responsible for
fouling of pipes) are also present and will, as a result,
undergo biological oxidation [See Sections F.2 to 3 and
G.4.b].
Chemoorganotrophs, more commonly referred to as hetero-
trophs, obtain energy from the oxidation of organic com-
pounds. They include not only saprophytic, parasitic, and
pathogenic bacteria, but also fungi and protozoa.
-------
Whereas human or animal pathogens usually require rich
nutrient substrates, aquatic saprophytic bacteria are less
fastidious and can thrive in an environment with extremely
low nutrient concentrations. They include both gram-nega-
tive nonsporeformers (members of the Spirillaceae and Pseudo-
monadaceae), and gram-positive cocci and endospore-forming
bacteria (members of the Micrococcaceae and Bacillaceae).
Some are sheathed (members of the genera Sphaerotilus,
Leptothrix, Crenothrix, etc.); some possess appendages
(Hyphomicrobium, Caulobacter, Gallionella, etc.). Several
members of the sheathed and appendaged bacteria are well
known among sewage and water treatment operators. For
example, species of Gallionella have been responsible for
plugging sand filters and other water treatment media [See
Sections E.4 and G.2] (Warren, 1971).
Among the myriads of fungi, only a few species thrive
to any extent in an aquatic environment. Those classes of
fungi most prevalent in the aquatic environment (chytridio-
mycetes, hypochytridiomycetes, and oomycetes) do not appear
to cause disease in man or other mammals (some chytrids may
parasitize certain species of algae). However, saprolegnias,
notably S. parasitica, have been known to cause a disease of
fish and~fish eggs and may cause significant economic damage
(Alexopoulos, 1962).
Aquatic protozoa ingest minute algae and bacteria in
addition to large quantities of organic debris found in
their habitat; they are regarded as consumers or grazers.
Of the five phyla (Mastigophora, Sarcodina, Ciliata, Cnido-
spora, and Sporozoa), the first three are thought to be
indigenous to unpolluted waters. They play an important
role in the aquatic environment as scavengers, but the
majority are considered harmless to humans.
The occupancy of different niches, the cycling of
elements, and the scavenger activities, all of which support
a diversity of life forms and maintain the stability of the
source water ecosystem, can be thrown off balance or dras-
tically altered when increased volumes of untreated or
inadequately treated wastewater are discharged into source
waters (Cairns, 1977).
Most source waters, whether lake, reservoir, river, or
groundwater will, at some point, receive wastewater from
human activity (Fair and Morrison, 1967). Nearly 200 to 450
1 (50 to 100 gal) of wastewater are produced daily per
individual in a developed country. These wastewaters are
discharged to receiving bodies of water as untreated or
-------
treated effluent. Wastes may be derived from domestic
sewage, from agricultural or industrial discharge, from
cesspool seepage or leachate, from a refuse disposal site,
or from wild animals and birds. The effect of effluent
discharge upon the receiving water, expressed in micro-
biologic terms or as BOD and suspended solids, varies with
the kind of wastewater, the treatment it has received, the
quantity discharged, and the dynamics of subsequent dilu-
tion. Understandably, the BOD or microbial content of
wastewaters derived from slaughterhouses and meat packing
plants is considerably different from wastewaters of paper
manufacture, and these are different from domestic waste-
waters (Tabor, 1976).
When large numbers of thermo-tolerant coliforms, fecal
streptococci, Clostridium perfringens, Bifidobacterium
bifidum, Bacteroides, Veillonella, and Peptostreptococcus
are encountered in a source water, they signify fecal pol-
lution [See also Sections C.l.c to e and D.4 to 5]. In
addition to human intestinal microbes, wastewater derived
from municipal sources may contain all of the indigenous
aquatic and soil microflora and -fauna. As a rule, the
great majority of microbes derived from wastewaters are
harmless chemoorganotrophic saprophytes. However, wastes
entering a waterway, whether treated or not, increase the
opportunities for contamination with disease agents.
The agents of typhoid fever, bacillary dysentery, and
cholera are members of the family Enterobacteriaceae, are
associated with wastewater, and are potentially transmissible
by contaminated drinking water [See Sections B.l.a(i),
(11), and (vii)]. Outbreaks of gastroenteritis caused by
Yersinia enterocolitica [See Section B.l.a(iii)] have been
reported, and the organism is apparently more widespread in
source waters receiving wastewater than once thought (Highsmith,
et al., 1977a). Enteropathogenic or toxigenic strains of
Escherichia coli [See Section B.I.a(iv)] also have occurred
in source waters. Francisella tularensis (agent of tularemia)
[See Section B.l.a.(v)] is another pathogen associated with
surface water contamination on rare occasions (Safe Drinkinq
Water Committee, 1977).
Recently, gram-negative, antibiotic-resistant strains
of Pseudomonas aeruginosa (cause of melioidosis) have been
transmitted to humans via source water receiving wastewater
[See Section B.l.a(ix)]. The organisms are well known as
pathogens of both animals and plants. Antibiotic-resistant
species of Aeromonas (e.g., A. hydrophila) not commonly
-------
identified as agents of human infection (but rather of
certain species of fish and frogs) have been transmitted to
humans via source waters contaminated by wastewater and have
caused cases of acute cellulitis (Hanson, e_t al. . 1977).
Populations of this organism in numbers as high as 105 to
10 cells per ml have been found in aquatic environments
which receive wastewater [See Section C.2.h].
Other bacteria (e.g., Klebsiella pneumoniae, Coryne-
bacterium diphtheriae, Streptococcus pyogenes, S_. pneu-
moniae, and others), whose epidemiologic significance has
been traditionally associated with food or air, have been
isolated from improperly treated wastewater effluent.
Pathogenic members of the orders Actinomycetales, Myco-
plasmatales, Rickettsiales, and Chlamydiales might also be
transmissible from wastewater to sources of drinking water.
Similarly, raw water derived from surface water which re-
ceives human or animal waste (either raw sewage or effluent
from improperly treated sewage) may be a potential source of
the Legionnaires' disease agent, Legionella pneumophila.
The agent was thought to be closely associated with water in
an air-conditioning system or wind-blown dust from an exca-
vation site close to the outbreak of the disease. However,
further investigation revealed that L. pneumophila has been
isolated from nonepidemic-related aquatic habitats including
streams, lakes, and soil (FliermansP e_t al. , 1979).
More than 100 different enteric viruses have been
isolated fromfithe feces of infected humans in concentrations .
as high as 10 plaque-forming units £pfu) per gram. Raw
sewage has yielded as many as 5 x 10 pfu per liter of
enteric viruses (Buras, 1976). Since viruses are not alto-
gether destroyed during wastewater treatment (though the
majority will have been removed by coagulation, sedimen-
tation, and filtration procedures) [See Section E], they
may enter surface or groundwater via treated waste effluent
(Wellings, e_t al., 1975) and find their way to another
community's source water [See Sections B.l.b and B.l.d].
Viruses have been recovered recently from both treated and
raw wastewater effluents; from soil receiving digested
sludge; from lakes and rivers which receive treated sewage
effluent (Gerba, et al., 1975a); and from estuaries in which
water salinity may vary from less than 1 percent to 12
percent. Some of the more important viruses which occur in
wastewater and other aquatic environments and might be
transmitted by contaminated drinking water are enteroviruses,
reoviruses, parvoviruses, hepatitis virus type A, adeno-
viruses, and gastroenteritis viruses [See Sections B.l.b and
B.l.d].
-------
Only free-living protozoa may flourish in unpolluted
waters, so facultative or obligately parasitic protozoa may
be considered as transient populations. Yet, both parasitic
and facultatively parasitic protozoa produce cysts and
trophozoites in their life cycles. Protozoan cysts are
frequently found in human feces and therefore, also in
wastewater [See Section B.l.c]. Hence, source waters re-
ceiving wastewater discharges may harbor forms of these
protozoa. Most cysts resist disinfection by chlorine,
bromine, or ozone; but the great majority are apparently
removed by coagulation, sedimentation, or subsequent fil-
tration. The ma-jority of waterborne protozoan disease
outbreaks in the U.S. have been associated with drinking water
derived from surface waters receiving wastewater which was
treated only by natural sedimentation and disinfection.
More efficient treatment, including coagulation and sedi-
mentation to remove cysts, followed by disinfection, must be
used routinely when processing source water receiving waste-
water, to avoid epidemics of amoebic dysentery, giardiasis,
and menlngoencephalitis [See Section E].
There is little information on the habitats of some
potentially pathogenic microbes such as Mycoplasma, patho-
genic yeasts, and other fungi (Safe Drinking Water Com-
mittee, 1977). However, fungi do not appear to be important
causes of waterborne disease in man, though they are fre-
quently incriminated as the cause of undesirable tastes and
odors.
Certain pathogens are more virulent than others.
Though virulence is a genetic trait, it is also related to
infectious dosage (the number of pathogenic organisms re-
quired to cause infection) [See Section B.4]. For example,
ingestion of a relatively small number of Shigella dysen-
teriae may cause acute disease, whereas exposure to a some-
what larger population of Vibrio cholerae is required to
infect a susceptible individual. Pathogens also vary in
that: some are able to donate or receive antibiotic resis-
tance; some have traditionally been associated with food-
borne or airborne disease; some contain in their cell walls
1ipid-polysaccharide complexes (endotoxins) that can produce
illness; and some propagate or cause illness only under very
special conditions [See Sections B and G.2].
Furthermore, there is nothing constant or necessarily
predictable about the numbers and kinds of pathogenic micro-
flora and -fauna associated with wastewater effluent. The
numbers and kinds vary with the degree of treatment and
disinfection, and these factors are subject to daily or even
-------
hourly change. In addition, some pathogens are capable of
resisting chlorination, or displaying a phenomenon known as
aftergrowth [See Section F.2]; that is, multiplication
within the receiving body of water. At one mechanical-
biological treatment plant in Denmark, reduction of fecal
and pathogenic organisms was found to be 90 to 99 percent,
with the exception of Clostridium perfringens [See Section
C.l.e], which was reduced by only 63 to 98 percent (Kristen-
sen, 1974). C^. perfringens, considered primarily a food-
borne pathogen is virtually always present in raw sewage as
well as in mechanically and biologically treated sewage
effluent. For all of the reasons just mentioned, kinds as
well as numbers of pathogens per volume of raw water must be
determined before the potential health hazard of any source
water can be assessed and the degree of treatment prescribed.
To summarize, microbes within source water that receives
wastewater effluents are either indigenous or transient.
These microbes include bacteria, algae, fungi, protoz.oa, and
viruses. The bacteria may be further characterized on the
basis of modes of obtaining energy, such as phototrophic,
chemolithotrophic, and chemporganotrophic; each set occupies
a specialized niche in the aquatic ecosystem.
As is true of organisms indigenous to source water, the
majority of microbes derived from wastewater are harmless
saprophytes. However, as greater quantities and more diver-
sified qualities of wastes are discharged to source waters,
pathogens, whether obligate or opportunistic, are more
likely to be present. Hence, it is the degree to which raw
waters are contaminated with waste that determines, in large
measure, the risk involved in producing finished water that
is free of disease agents.
1. Microbiology of Groundwater
Historically, groundwater has been a source of relatively
clean, high quality drinking water, needing little, if any,
treatment. In contrast, surface water almost always has
required some level of treatment as a result of ever increasing
human interactions, in the form of industry, agriculture,
aquaculture, recreation, sewage disposal, and other activities,
with the hydrologic cycle. In many countries, pollution of
surface waters has led to a decrease in their suitability
for potable use or rendered them more costly and complicated
to treat. The relative purity of groundwater has led to
greater world reliance on this source for supplying potable
water. On a global scale, groundwater sources appear to be
much more plentiful than surface water sources [See Table
A.1-1].
-------
TABLE A.1-1
COMPOSITION OF THE HYDROSPHERE
AND RATE OF TURNOVER
Parts of the
Hydrosphere
World oceans
Groundwater
(groundwater zones
of active turnover)
Polar ice caps
Surface water
Rivers
Soil moisture
Atmospheric vapor
Total hydrosphere
Volume in
Thousands of
Cubic
Kilometers
1,370,000
60,000
4,000
24,000
280
1.2
80
14
1,454,000
Rate of Water
Turnover
(Years)
3,000
5,000b
330C
8,000
7
0.031
1
0.027
2,800
Prom Ilvovitch, 1977.
aln round numbers.
V"\
Includes groundwater runoff to oceans bypassing rivers
4,200 years.
Q
Includes groundwater runoff to oceans bypassing rivers
280 years.
-------
About 98 percent of Denmark's population obtains its
drinking water from minimally treated or untreated ground-
water. In the U.S., over 50 percent of the population
receives its drinking water from groundwater sources, roughly
half of which have been treated minimally or not at all.
However, between 1946 and 1974, over 50 percent of all
waterborne disease outbreaks in the U.S. were attributed to
contaminated groundwater (Allen and Geldreich, 1975). One
of the most important diseases spread via groundwater is
hepatitis A [See Section B.l.b]. In a review of 48 hepa-
titis outbreaks occurring worldwide and involving 31,357
people, groundwater was implicated in 21 of these and in-
volved 1,378 people (Taylor, e_t al^. , 1966).
Recent trends in land disposal of sewage effluent and
sludge have prompted scientists and hygienists to examine
possible new and adverse chemical and microbiological effects
upon groundwater. Manure'has been used extensively in
agriculture for hundreds of years without any reported
disease outbreaks resulting (except where excessive fertiliza-
tion or cattle feedlot impoundments led to nitrate contamina-
tion). However, lately, many new methods -have been used to
dispose of wastes on land, which has consequently increased
the volumes disposed there. Moreover, as sewage treatment
methods improve, production of sludges will increase, resulting
in greater impacts to the land and to aquifers below. The
growing practice of groundwater recharge with wastewater is
also of concern.
Volumes of wastewater produced in the year 2000 are
expected to be ten to 15 times higher than those produced in
1970 (Ilvovitch, 1977). It therefore would appear that we
are dealing with problems whose solutions will become increas-
ingly urgent in the next few decades. The protection of our
groundwater resources will become more important as our
reliance on land disposal of wastes increases as an alter-
native to discharging them into surface water.
a. Definition of Groundwater. Water resources differ
from other natural resources in that the hydrologic cycle
and hydrosphere are two inextricably bound systems with one
constantly exerting an influence over the other. Taking
part in this cyclic continuum are oceans, ice caps and
glaciers, lakes, streams, groundwater, and the atmosphere.
Surface water and groundwater are in contact through river
and lake beds. Most precipitation evaporates and returns to
the atmosphere; a lesser portion returns to the sea through
-------
10
streams and rivers; and only a minor part sinks into the
ground and percolates through the soil system to become
groundwater.
That water which percolates through the soil and stops
its downward movement upon reaching an impermeable layer
comprises the groundwater. Water in this zone fills the
soil pores and flows horizontally. Because it has passed
through the soil system, where processes such as ion ex-
change, adsorption, precipitation, and chemical alterations
including^biodegradation take place, groundwater will differ
both chemically and microbiologically from surface water.
Above the groundwater is an unsaturated zone where
water percolates both horizontally and vertically while some
of it remains there held by adhesion and capillarity. It is
in this unsaturated stratum that various physical, chemical,
and biological activities take place which normally will
determine groundwater quality. Climate, the depth and
geochemistry of the unsaturated zone, and activities at the
surface (such as waste disposal) are all major factors
affecting groundwater quality.
Normally, groundwater is virtually free of organisms
because they are filtered out or adsorbed to particles in
the soil system. If there is sufficient soil depth, tex-
ture, structure, and biological activity, the groundwater *
should be of acceptable microbiological quality. On the
other hand, more than 40 groundwater samples from deep
aquifers in Jordan were found to contain coliforms; this was
attributed to poor filtration of the overlying soil (Shehabi,
1976). The rate of groundwater turnover is much slower than
that of surface water [See Table A.1-1] and therefore, once
groundwater has become contaminated, it remains contaminated
much longer. For this reason, the issue of waste disposal
on land becomes that much more serious.
This groundwater discussion pertained to that which
occurs in deep aquifers and did not take into account well
water which may become contaminated with surface water when
leaky covers are situated below surface level, or as a
result of improper construction or placement.
b. Occurrence of Indigenous Microorganisms in Groundwater.
Groundwater usually is extremely low in organic matter (or
that which is present is in an unavailable form), and this
combined with filtration in the unsaturated zone normally
-------
11
precludes the presence of any organisms except possibly a
few chemolithotrophic bacteria. These bacteria usually
range in numbers between 20 and 50 per ml (Hvid, 1955). The
ratio of chemoorganotrophs (also called chemoheterotrophs)
to chemolithotrophs is a function of the organic content of
the soil, which decrease with depth [See Table A.1-2]. The
total numbers of microorganisms decrease with the depth of
the soil, but below one meter, both the amount of organic
matter and numbers of chemoorganotrophs are low while the
proportion made up of chemolithotrophs rises. Samples
taken from deep borings may be sterile or contain only
chemolithotrophic bacteria, such as methane, sulfur, and/or
iron bacteria.
The presence of methane in groundwater indicates activ-
ity of methane-generating bacteria such as Methanobacterium,
which reduce carbon dioxide and organic acids to methane.
Such bacteria may cause problems because methane is a nutrient
source for certain slime bacteria which may grow in layers
on filters, along pipes, and in containers (Volker, et al.,
1977) [See Sections G.2 and F.3]. Sulfates, if present in
groundwater, may be reduced to hydrogen sulfide by bacteria
such as Desulfovibrio and thereby render the water unacceptable
for potable use[See Sections F.2.d and G.4],
Iron and manganese bacteria, including the Sphaero-
tilus-Leptothrix group and Gallionella, may be present in
groundwater.These bacteria obtain energy by oxidation of
ferrous and bivalent manganese compounds [See Section G.4].
Hydrogenomonas, which can oxidize molecular hydrogen to
obtain energy, has also been recovered from aquifers (Hvid,
1955).
Samples taken from deep borings normally contain few,
if any, chemoorganotrophic bacteria unless there has been
some mixing with surface water resulting in availability of
organic material. Danish investigators found anaerobic
actinomycetes, clostridia, and anaerobic gram-positive non-
motile cocci (unclear taxonomy) in samples from deep borings
(Hvid, 1955). These cocci and actinomycetes are important
in denitrification of nitrates in the unsaturated zone and
as such, are instrumental in preventing occurrence of ni-
trates in groundwater. Also found within the groundwater in
Denmark were nonsporeforming, motile, gram-negative, glu-
cose-negative, and gelatin-positive bacteria (possibly
belonging to the genus Vibr_io) (Bonde, personal communication).
Such bacteria may not always toe detected by conventional
methods and may propagate in slow moving or stagnant parts
of the distribution system [See Section F.2].
-------
12
TABLE A.1-2
CONTENT OF BACTERIA AND ORGANIC MATTER
AT DIFFERENT SOIL DEPTHS
Depth
(cm)
0-6
6-12
12-28
28-48
48-80
Organic Matter
(%)
8.04
3.18
2.41
1.76
0.80
Million
Ae robes
49.2
131.8
158.3
45.3
6.0
Bacteria per Gram
Anaerobes
1.0
1.0
10.0
1.0
0.001
From Jepsen, 1972.
-------
13
Bacteria also can propagate in water of initial potable
quality that has been stored. Investigators in Norway
stored groundwater with a potassium permanganate value of
1.5 at 22°C for 3 to 13 days, after which time original
bacterial numerals of 2 per ml had swelled to over 100,000
per ml. However, such marginal levels of organic matter
content would be highly selective for the less fastidious
bacteria [See Section G.6].
Although fungi (of which about 200 genera have been
isolated from soil) are more prevalent than bacteria in
soil, very little is known about their presence in ground-
water. Fungi, pathogenic to humans, also may occur in soil.
All fungi are aerophilic and do not grow under strict an-
aerobic conditions such as would occur in water-logged soil.
Protozoa, including flagellates, amoebae, and ciliates,
may appear in the upper^part of the unsaturated zone in
numbers varying from 10 to 10 per gram of soil. However,
owing to their size, they are readily entrapped within the
soil matrix and prevented from entering groundwater.
c. Fate of Organic Substances in Groundwater. As a
rule, undisturbed groundwater bears organisms and chemicals
indigenous to that environment. But, accompanying the rapid
increases in human populations have been more intense and
varied uses of land and water. This has meant, in some
instances, that groundwater may be derived in part from
water percolated from surfaces to which treated or untreated
sewage effluents have been applied. Or, it. has led to the
practice of disposing sewage effluent or sludge on land near
sources of groundwater used in producing potable water. The
extent to which these practices pose a threat to the safety
of groundwater depends on the types and quantities of waste
applied and on the climate and hydrogeology of the site.
Pollutants may enter the aquifer directly through
crevices and fractures in the bedrock or through porous
subsoil. They may percolate through shallow soils, or enter
the aquifer during bank infiltration practices used in some
areas for treating surface waters (Federal Ministry of the
Interior, 1975) [See Section E.4]. Pollutants also may
enter groundwater via deep injection wells used for under-
ground storage or disposal of industrial or municipal waste-
water or of septage. Toxic or obnoxious pollutants may
enter the aquifer as a result of accidental spills during
transport (Federal Ministry of Health, 1969); such events
-------
14
can no longer be considered extremely rare. Once these
substances enter groundwater, they may travel great dis-
tances and pollute uncontaminated water in nearby aquifers.
The fate of organic substances in groundwater depends
largely on the presence of saprophytic microbes and on the
availability of dissolved oxygen. Organic substances, when
in^the presence of sufficient available oxygen, may be
oxidized or mineralized to such products as carbon dioxide,
nitrate, and sulfate [See Figure A.1-1]. However, any
aerobic biooxidation of organic substances in an environment
such as an aquifer almost always depletes the supply of
dissolved oxygen, thereby decreasing the oxidation-reduction
potential of the water. Such conditions favor the microbial
reduction of certain substances (e.g., ferric and tetra-
valent manganese ions, nitrate, sulfate, etc.) to a number
of unacceptable products including ferrous and manganous
ions, ammonium, nitrite, methane, hydrogen sulfide, and
other reduced sulfur compounds (Schweisfurth and Ruf, 1976)
[See Figure A.1-2]. These biogeochemical conversions are
Known to occur even in aquifers deep underground.
Reduced inorganic or organic substances in groundwater
subsequently may be spontaneously (chemically) or micro-
bially oxidized when the water is brought to the surface to
be processed through a water treatment facility, for example.
Here, various microorganisms such as species of Sphaero-
tJ-lus, Leptothrix, Phragmidiothrix, Gallionella, and espe-
cially Crenothrix polyspora (Schweisturth, 1974; Volker, et
al., 1977) proliferate, producing abundant sheaths or slimes
which may be encrusted with oxidized minerals. Sheaths and
slimes not only discolor and impart tastes and odors to the
water, but also cause considerable problems in treatment
since they (especially slimes encrusted with iron or manganese)
may plug sand filters and otherwise interfere with the
treatment and distribution of drinking water [See Section
E.4 and G.2 to 4].
__ d< New Trends in the Disposal of Sludge and Sewage
Effluent. Opportunities for contamination ot groundwater,
including water in wells, seem to be tied closely to new
trends in land disposal techniques. Increasing populations
produce more wastes and at the same time withdraw more water
for domestic, agricultural, and industrial purposes. As a
result, more water is being converted to wastewater, only to
be re-introduced into the hydrologic cycle where some will
invariably end up as groundwater.
-------
FIGURE A.l-1
DEGRADATION OP ORGANIC SUBSTANCESIN THE PRESENCE
OF SUFFICIENT AMOUNTS OF OXYGEN
Org.C \ ,
Energy n .. \
Substances Uiy. IN }
Org.SH/
Energy
lHan<^350 to»A5QmV \
MO
( Microoraani
sms ^
~Water humus
A .(difficult to degrade)
Water after mineralization low In microorganisms.
-------
FIGURE A.1-2
DEGRADATION OP ORGANIC SUBSTANCES WITH INSUFFICIENT OR NO OXYGEN
Org.C
Org.N
Alcohols
Fatty acids
,co?
nm -150 nqV/ other products
N0'
N02
/Other organic products
i^T^r(Mass of bacteria
from -200mV
Microblal and chemical
org.SH
MO
NHt
NH^ - Eb +350 to +100 mV
Mn : Eb: from approx.
2+ +300 mV
: Eb: from approx.
Q +200 mV
^ : Eb: from approx.
H,S ±0 "v
Note.: MOfimicroorganisms Q; determined in the laboratory culture
LJ»chemically determined
-------
17
Hazards associated with septic tank systems, cesspools,
and leaky sewage pipes have been known for a long time.
Sewerage systems may lose influent to the surrounding soil;
and, in fact, it is estimated that the wastewater arriving
at the treatment plant is generally 20 percent to 60 percent
lower than the amount originally supplied to the municipal
system (Shuval, 1977). The reverse (infiltration) also has
been known to occur with some regularity. Home septic
systems that incorporate a sand-filled seepage pit or leach
field for adsorption can be very effective in removing
pathogens from the septic tank effluent; yet, many of these
individual systems are poorly constructed, receive little
maintenance, and become overloaded by the householder so
that the domestic sewage essentially bypasses treatment.
The potential hazard therefore is that poor quality sewage
effluent may seep into groundwater supplies of the home or
community.
Sewage effluent disposal to the land has, in recent
years, been largely in the form of irrigation, wastewater
treatment lagoons, infiltration ponds, and wastewater dis-
posal in underground pits (Cleary and Warner, .1970). Volumes
of irrigation water used in some countries are expected to
double in the next 20 to 25 years. Such activities may
modify water circulation patterns, soil ecology, and ulti-
mately, groundwater quality,. Heavy irrigation may cause
stronger vertical seepage through soil increasing the poten-
tial for groundwater contamination (Kovacs, 1977).
Infiltration ponds are designed for disposal of sewage,
but often are used for groundwater recharge besides. How-
ever, they should be viewed as waste treatment processes
only, since seepage of pollutants to groundwater is a defi-
nite possibility, depending on geochemical conditions,
climate, level of the water table, and infiltration rate of
the wastewater (which can be rather high in some cases).
Danish investigators found that the infiltration rate varied
between 0.18 m per day during the first year to 0.05 m per
day after a few years (Bonde, et al., in press).
The use of treatment lagoons is common, especially in
semi-arid regions. Lagoons may be used in conjunction with
conventional primary or secondary treatment, but may also be
used exclusively to dispose of waste by evaporation, seepage,
and direct re-use through irrigation (Klock, 1971).
In areas with a high water table and a low adsorption
capacity of the unsaturated zone, the use of wastewater
treatment lagoons or infiltration ponds may result in con-
tamination of groundwater with fecal (including pathogenic)
-------
18
and soil organisms. Moreover, where a sewage stabilization
pond has been placed improperly over river alluvium or
limestone sink holes, leakages into groundwater or an entire
collapse of the pond is possible.
Alterations in the ecology of the soil and groundwater
are possible with increased reliance on wastewater treatment
lagoons and infiltration ponds. Such changes as the appear-
ance of large numbers of iron and sulfur bacteria in aquifers
have already been noted. More important are the potentials
for changes in survival, propagation, and movement of patho-
gens. If pathogens survive longer in soil than the time it
takes to pass from a treatment lagoon through the unsat-
urated zone, these organisms may gain entry into the ground-
water. Creation of a new set of variables within soil and
groundwater from new waste disposal technologies may change
conditions enough that the previously applied principles
will no longer hold true. Therefore, it is essential that
more be known about factors affecting survival of intestinal
microorganisms in soil and groundwater as compared to their
survival in treatment lagoons and infiltration ponds.
.e- Factors Affecting Survival in Soil and Groundwater.
Survival ot pathogens in soil and water is determined largely
by: (1) species of microorganisms; (2) temperature; (3)
sunlight; (4) rainfall; (5) pH; (6) soil type; (7) moisture
and organic content of the soil; (8) microbial antagonisms
(including production of antibiotics such as actinomycin and
streptomycin), predation, and parasitism (bacteriophages);
and (9) infiltration rates. Different organisms exhibit
different survival rates in soil and water [See Tables A.1-3
to 5]. For example, mycobacteria and sporeformers survive
longer than thermo-tolerant coliforms in soil. Climate also
is a major factor in that survival is generally greater in
winter than in summer [See Table A.1-3], and contamination
of groundwater is much more likely during periods of heavy
rainfall. Microbial antagonisms (especially from actino-
mycetes) are very important in the reduction of pathogenic
bacteria, but appear not to be with viruses (Gerba, 1974).
Organisms at the soil surface are exposed to UV and
desiccation and therefore are less likely to survive than
those within moist soil. Less than 15 percent humidity
greatly reduces virus numbers (Moore, et al., 1976). In
general, oacteria survive longer in alkTlTne (limestone)
than in acid (peat) soils. Irrigation with sewage effluent
-------
TABLE A.1-3
SURVIVAL OF BACTERIA ON OR IN SOIL
Organism
Soil Type
Time of
Year
Survival Time
(Days)
Total coliforms
Escherichia coli
Escherichia coli
Escherichia coli
Escherichia coli
Fecal streptococci
Fecal streptococci
Streptococcus sp.
Streptococcus sp.
Streptococcus sp.
Surface soil
Coarse loam (rich in
Coarse loam (rich in
Coarse loam (rich in
Coarse loam (rich in
Coarse loam (rich in
Coarse loam (rich in
Clay loam
Clay loam
Sandy
organic material)
organic material)
organic material)
organic material)
organic material)
organic material)
Winter
Summer
Autumn
Summer
Winter
Summer
38
40-65
15 - 30
13.4 (90% reduction)
3.3 (90% reduction)
20.1 (90% reduction)
2.7 (90% reduction)
63
49
35
-------
TABLE A.1-3 Continued
Organism
Soil Type
Time of
Year
Survival Time
(Days)
Salmonella sp.
Salmonella sp.
Salmonella sp.
Salmonella typhi
Salmonella typhi
Salmonella typhi
Salmonella typh i
Salmonella typhi
Salmonella typhi
Salmonella typhi
Salmonella typhi
Brucella abortus
Mycobacterium bovis
Mycobacterium bovis
Non-sterile soil
Surface soil
Sterile soil
Garden soil
Garden soil
Sandy soil
Sandy soil
Loam (wet weather)
Loam (dry weather)
Soil
Soil
Surface garden soil
November
October
October
October
Winter
280
100
40
216
60 - 70
58
50
36
29
25 (90% reduction)
5 (90% reduction)
125
178
8-11 (90% reduction)
-------
21
TABLE A.1-4
SURVIVAL OF BACTERIA IN GROUNDWATER
Organism
Medium
Survival Time
Escherichia coli
Escherichia coli
Escherichia coli
Total coliforms
Salmonella
Shigella
Shigella flexneri
Vibrio cholerae
Groundwater held in the
laboratory
Groundwater in the field
Recharge well
Well water
Water infiltrating sand
columns
Water infiltrating sand
columns
Well water
Well water
4-4.5 months
3 -3.5 months
63 days
17 h (50% reduction)
44 days
24 days
26.8 h (50% reduction)
7.2 h (50% reduction)
From Gerba, et. al., 1975b.
-------
TABLE A.1-5
MOVEMENT OF BACTERIA THROUGH SOIL
Organism
Medium
Nature of Pollution
Maximum
Observed
Distance
Travelled
(in Meters)
Time of
Travel
(Hours)
Total coliforms
Total coliforms
Total coliforms
Total coliforms
Total coliforms
Total coliforms
Thermo-tolerant
coliforms and
Streptococcus sp,
Thermo-tolerant
coliforms
Escherichia coli
Sandy gravel
Sand and pea gravel
Sand and gravel mix
Fine to medium sand
Fine-grained sand
Fine sandy loam
Coarse gravel
Fine loamy sand to
gravel
Sand dunes
Secondary sewage
Diluted settled sewage (into
injection well)
Oxidation pond effluent
Tertiary treated wastewater
Sewage introduced through a
perforated pipe
Primary and treated sewage
effluent
Tertiary treated wastewater
Secondary sewage effluent
on percolation beds
Canal water on percolation
beds
0.9
31
830
6.1
1.8
0.6-4.0
460
9.2
3.1
35
48
NJ
to
From Gerba, et al., 1975b.
-------
23
on soil and plant surfaces allows for direct exposure of
accompanying flora to UV light and for desiccation between
irrigation periods. Death rates under such conditions are
relatively high. Waste stabilization ponds, on the other
hand, protect microorganisms against exposure to UV and so
increase the likelihood of survival. The presence of organic
compounds also seems to increase survival of microorganisms.
For example, the data in Table A.1-6 suggest that indicator
bacteria survive for extended periods in soil to which
digested sludge has been applied.
Adsorption of viruses is influenced mainly by pH, the
type and concentration of cations, organic matter, and the
surface area of the soil particles. Clay soils or, to a
somewhat lesser extent, consolidated sands facilitate virus
adsorption. The efficiency of adsorption increases the more
acid the soil is or the greater the ionic strength is in the
suspending medium (obtained, for example, by adding calcium
chloride to the soil). Adsorption may be hindered or the
process reversed with high concentrations of organic matter.
Sites consisting largely of highly fractured rock or mix-
tures of gravel and sand are unacceptable for waste disposal
because viruses and other microorganisms may be transported
over long distances in such soils [See Table A.1-5].
In experiments using 250-cm columns of calcareous sand,
Lance and coworkers (1976) showed that reductions in the
ionic strength of the soil solution, from flooding with
deionized water, caused desorption and movement of viruses
through the soil (Lance, e_t auk. , 1976). However, most
viruses were re-adsorbed, and showed a peak concentration at
a depth of 10 cm, after flooding with sewage. It is evident,
from these and other data, that the best results are achieved
by alternating applications of sludge or sewage effluent
with dry periods and by avoiding disposal of wastes during
periods of heavy rain. If done correctly and at the appro-
priate site, land application of municipal wastewater or
sludge can be a highly satisfactory method of waste treat-
ment.
Once they have entered the groundwater, bacteria no
longer have to contend with UV light or microbial antago-
nisms and this, along with low temperatures of groundwater,
prolongs survival (E. coli has been found to survive up to
100 days in subsoiT~water: Gerba, e_t al. , 1975) [See Table
A. 1-4] When sewage effluent was pumped directly into
aquifers, fecal streptococci and thermo-tolerant coliforms
were recovered at a distance of 33 m (Reviewed by Jepsen,
1972). Viruses, as well, persist longer in groundwater
-------
TABLE A.1-6
MEDIAN BACTERIAL COUNTS (PER GRAM) IN FOUR
REPLICATE SAMPLINGS OF A SINGLE SITE AT VARIOUS TIMES
AFTER APPLICATION OF DIGESTED SLUDGES
Days After
Application
of Sludge E.
Colia
15 250
22 4,602
29 8,550
36 1/200
Soil
Total
Coliforms
770
1,774
950
1,225
Fecal
Streptococci
17
20
23
250
Grass
E. Total
Coli Coliforms
689 4,720
464 > 1,415
12 144
32 175
Fecal
Streptococci
0.8
1.3
1
< 0.2
a
Escherichia coli.
From Carrington, in press.
to
*»
-------
25
which is clean compared to river or seawater where viral
reductions are greater [See Section B.3], Investigators
recovered hepatitis A virus from a well situated 23 m (75
ft) away from a cesspool (Neefe and Stokes, 1945).
From tests made of treatment lagoons (Klock, 1971), it
appeared that after about one week (a shorter time during
the summer), only 1 percent of coliforms had survived; and
only about .001 percent were viable after two weeks (Klock,
1971). Survival rates in infiltration ponds probably are
the same. From this, one could surmise that a large major-
ity of pathogens would be eliminated before reaching the
soil. Nevertheless, when soil beneath infiltration ponds
became saturated with sewage effluent, thermo-tolerant
coliforms were recovered at a depth of four, but not eight,
meters (Baars, 1957). Since infiltration ponds often are
placed in areas of sandy soil, movement of organisms tends
to be a problem.
By studying a sedimentation-infiltration pond system in
operation for about ten years, researchers in Denmark were
able to evaluate long-term consequences to underlying soil
and groundwater (Bonde, e_t ajL^., in press) [See Table A.l-7] .
The system consists of two sedimentation ponds and four
infiltration ponds receiving raw sewage on an infiltration
surface equalling 10 m per capita waste output. The depth
to groundwater in the area is about 2 m and the soil is
characterized as sandy. Average retention times were eight
days in the sedimentation ponds and five days in the infil-
tration ponds. As can be seen in Table A.l-7, thermo-tolerant
coliforms were isolated from all samples collected at a
depth of 16.5 m beneath infiltration ponds. Colony counts
at 20 and 37°C were quite high at all depths and the presence
of P. aeruginosa in nearly all groundwater samples indicates
proEable occurrence of pathogenic bacteria in groundwater
below these infiltration ponds.
f. Occurrence of Endotoxins in Groundwater. Gram-
positive bacteria are the predominant bacterial microflora
of soil, whereas gram-negative bacteria normally predominate
in water. Irrigation and disposal of sewage water on soil
consistently contribute enormous numbers of gram-negative
bacteria, many of which are subsequently eliminated by
antagonisms in the soil.
The confrontation in soil between the indigenous gram-
positive soil bacteria and the gram-negative water bacteria
may result in the release of endotoxins, which are produced
-------
TABLE A.1-7
BACTERIA IN GROUNDWATER BENEATH INFILTRATION PONDS
SAMPLED IN JANUARY AND FEBRUARY OF 1977
BACTERIA
Total coliforms
Thermo-tolerant
coliforms
Colony counts
at 20°C
Colony counts
at 37°C
Fecal streptococci
Clostridium
perfnngens
Pseudomonas
aeruginosa
Coliphages
aBacteria per 100 ml.
DEPTH (in meters)
1.9 2.5 4.5
5.4xl03-2.2xl04 >1.6xl03-1.6xl05 >1.6xl03-1.7xl04
3.5xl03-2.2xl04 >1.6xl03-5.4xl04 >1.6xl03-l.lxl04
2.2xl08-12.9xl08 1.8xl08-13.2xl08 5.0xl07-4.1xl08
6.0xl06-4.4xl08 1.0xl07-4.5xl08 9.0xl06-1.3xl08
1.8xl03-4.6xl03 5.4xl02-3.9xl04 3.6xl02-4.8xl03
1.0xl02-3.0xl02 1.4xl03-1.7xl03 4.0xl01-9.1xl02
+b ' + .
2.0xl01-3.5xl02 2.0xl01-5.4xl02 1.0xl01-9.2xl02
9.5 16.5
>1.6xl03-1.6xl04 2.3xl01-2.5xl02
>1.6xl03-1.6xl04 8xlO°-4.3xl01
1.3xl07-4.7xl08 4.0xl06-4.0xl07
4.0xl06-9.6xl07 1.0xl06-4.6xl07
1.0xl02-5.9xl02 1.0xlO°-1.0xl01
- 200 ^ 100
+
S.OxlO^S.OxlO1 ^ 5.0
+ = Pseudomonas aeruginosa present.
to
a\
(BQNDE, §T AL.f IN PRESS)
-------
27
only by gram-negative bacteria. Endotoxins may then travel
from the unsaturated zone ,in the soil into the groundwater.
Groundwater situated beneath infiltration ponds in Denmark
and sampled at depths of from 1.9 to 16.5 m wa| found_|o
contain endotoxins in amounts varying from 10~ to 10 g
per ml (Kristensen, 1978). Maximum endotoxin concentrations
found during this study were over 10,000 times the minimum
dose (2 ug endotoxin per kg body weight) needed to produce
clinically measurable effects by parenteral injection in
humans (Mikkelsen, 1977). Since they are toxic only if they
enter the bloodstream, endotoxins in potable water pose
special problems for production of solutions for infusion
[See also Sections B.l.a(ix) and C.S.a].
g. Criteria for Evaluating Groundwater Quality.
Groundwater supplies often are tested by the same microbio-
logical criteria used for potable water (Allen and Geldreich,
1975). Though there are differences from country to country,
most include tests for total and thermo-tolerant coliforms
and colony counts at 21 and 37°C [See Sections C.l.a-c and
D]. Tests made in the U.S. have demonstrated the presence
of total and even thermo-tolerant coliforms in some com-
munity groundwater supplies [See Table A.1-8]. Yet, the
apparent absence of coliforms does not necessarily ensure
that pathogens, such as I?, aeruginosa [See Sections B.l.a(ix)
and F.3.b] and Y. enterocolitica [See Section B.l.a(iii)]
are not present in the groundwater (Nemedi and Lanyi, 1971;
Lassen, 1972).
There is very little information available on the
spread of viruses in groundwater primarily because the most ,
important groundwater-borne virus, the hepatitis A virus,
cannot as yet be cultured in the laboratory and remains
extremely difficult to study [See Section III.B.l.b]. Virus
criteria for drinking water normally are not specified for
routine drinking water analyses. Although, in a preliminary
proposal submitted by the Commission of the European Com-
munities in 1974, recommendations included a statement that
enteroviruses should be absent in 10 1 of drinking water,
this was removed from the final Council Directive (Commis-
sion of the European Communities, 1975) [See Table A.1-9 and
Section C.2.b]. It has been suggested that bacteriophages
be used as an indicator of enteroviruses [See Section C.2.a];
but Gerba and coworkers (1974) state that removal of bacterio-
phages during percolation through soil is much greater than
removal of enteroviruses.
Before improved indicator systems for assessing ground-
water quality can be designed however, more needs to be
-------
28
TABLE A.1-8
MICROBIOLOGICAL SUMMARY OF COMPLETED GROUNDWATER SURVEYS
Survey
Community water supply study
Tennessee-Georgia rural
water supplies
Interstate highway drinking
water systems
Umatilla Indian Reservation
Number of
Samples
621
1,257
241
498
Samples (+)
for Total
Coliforms
(%)
9.0
51.4
15.4
35.9
Samples ( + )
for
Thermo-Tolerant
Coliforms
(%)
2.0
27.0
2.9
9.0
>^ 1 organism per 100 ml.
From Allen and Geldreich, 1975.
-------
TABLE A.1-9
MICROBIOLOGICAL REQUIREMENTS FOR DRINKING WATER
IN THE EUROPEAN ECONOMIC COMMUNITY
Parameter
Volume of Guide Level (GL)
Maximum Admissible
Concentration (MAC)
Multiple Tube
uumt-xi. v«u.j ^"^ - -yu. plate Count Agar Membrane Method
Total coliforms
Thermo- tolerant
coliforms
Fecal streptococci
Clostridium sp.
Colony count
(for water supplied
for direct
consumption)
100
100
100
20
1
- 0 MPN < 1
- 0 MPN < 1
0 MPN < 1
MPN < 1
10 (37°C)a -
100 (22°C) -
Values for disinfected water should be considerably lower at the point where it leaves the
processing plant.
Footnote: In addition to the above analyses, microbiologic examinations should, in cases where
warranted, include tests for Salmonella, enteroviruses, and coliphages. All
finished water should be free of algae and parasitic protozoa and helminths.
vo
-------
30
known about soil-groundwater ecology and survival and move-
ment of enteric organisms released to the soil via current
waste disposal practices. Whether or not to establish
quality criteria for chemolithotrophic bacteria is another
matter to, be considered. Chemolithotrophs, though not
indicators of fecal contamination, are an important cause of
operational difficulties at the treatment plant and in the
distribution network.
h. Protection of Groundwater. In an effort to divert
some of the tremendous volumes of wastewater from dis-
charging into rivers and lakes, many countries are turning
to land disposal. At the same time, it is estimated that
groundwater will, in the future, satisfy over 50 percent of
drinking water needs (Ilvovitch, 1977). Although the soil
acts as a buffer against groundwater contamination, just how
far this protection can be counted on is not known given
these new and intensified circumstances.
Recommendations for protecting groundwater, submitted
by the Commission of the European Communities (1979), make a
distinction between pollution that occurs in the saturated
zone and pollution that occurs in the unsaturated zone.
Unfortunately, the proposed measures deal only with chemical
and toxicological problems even though over 99 percent of
outbreaks of waterborne disease are caused by microorgan-
isms. Whereas it has been shown that groundwater beneath
infiltration ponds need only be diluted less than 100-fold
to fulfill chemical requirements (e.g., NH. -N and anionic
detergents) for drinking water, it must be diluted several
thousand fold to satisfy microbiologoical requirements.
Legislation enacted in Denmark for the protection of
groundwater requires that a license be obtained prior to
land disposal and specifies conditions under which land
disposal of wastes may be considered. Containers whose
contents could be construed as potentially contaminating
cannot be placed in the soil. Seepage drains and cesspobls
must be at least 2.5 m above the water table; and wastewater
percolation facilities must be no less than 300 m from the
nearest water catchment. Pollutants in all such wastewaters
must be easily biodegradable. It is recommended that, for
municipal sewage effluent, the equivalent of one person's
daily waste output be applied to 100 to 400 m of farmland.
One hectare is required for one day's disposal of 10 m of
dairy wastewater.
-------
31
High groundwater quality is best maintained by intel-
ligent preventative measures aimed at protecting the integ-
rity of the soil-aquifer ecosystem. Primary treatment with
disinfectants such as chlorine does not serve as an adequate
substitute but rather, renders more complex the economic and
technical props needed to produce a potable product.
i. Conclusion and Recommendations. Pollution of our
groundwater supplies is one of the most serious problems
confronting the 20th Century. In order to devise and imple-
ment means for safeguarding this resource, we need to know a
great deal more about long-term effects to soir and aquifers
from inundations with wastes. Besides studying altered
conditions for survival, propagation, and transport of
pathogens, water microbiologists should pay more attention
to oxidation and mineralization processes and their effects
on the growth of chemolithotrophic and chemoorganotrophic
organisms. Also needed is more research into potential
groundwater contamination with endotoxins and other bio-
toxins from land disposal activities. Finally, better
indicator systems are sorely needed to take into account
properties unique to soil-aquifer ecosystems.
2. Microbiology of Surface Water
In Western Europe, agricultural, industrial, and urban
expansion have led to a four to five-fold increase in water
consumption over the past 100 years. For Denmark, as for
many countries, increased demands for potable water have
necessitated turning to nontraditional raw water sources.
Presently, Denmark derives over 90 percent of its water from
groundwater, which accounts for 20 percent of the total
groundwater formation and 3 percent of the total precipi-
tation over the country's surface area. However, in order
to avoid markedly lowering the water table, Denmark is now
attempting to meet demands by exploiting its surface water
resources to a greater extent than previously. At the same
time, surface water is increasingly being used as a reposi-
tory for chemical and biological wastes; this trend could
put many municipalities, and even nations, on' a disaster
course if corrective steps are not taken soon.
a. Occurrence of Microorganisms in Surface Water.
Bacteria in lakes and streams are either part of the aquatic
community or they are of exogenous origin (runoff, waste
effluent, animals, or aerosols). Whereas types and numbers
-------
32
of the indigenous flora are influenced mainly by the avail-
ability of nutrients, light, pH, temperature, and dissolved
gases, most transient bacteria (e.g., indicators, pathogens,
and soil bacteria) live a more or less passive existence in
surface water, their survival rates being determined more by
negative factors including UV light, predation, microbial
antagonisms, and sedimentation [See Section B.3 and E.I].
Therefore, one may expect a rise in the indigenous popu-
lations and a reduction in transitory microbes during the
summer months. Conversely, winter temperatures tend to
prolong survival of indicators and pathogens while curbing
activity of the endogenous aquatic community. Heavy winter
rains can cause soil runoff in an unprotected watershed and
convey nutrients (that stimulate the growth of indigenous
flora) and soil bacteria and fungi (that may temporarily
mask actual bacterial numbers) to surface waters. However,
the effects are generally short-lived.
(i) Chemoorganotrophic Bacteria. Chemoorganotrophic
bacteria depend on organic matter for energy as a principal
carbon source [See Section A.Intro.]. Aquatic chemoorgano-
trophs differ from soil chemoorganotrophs in being mostly
gram-negative rods. Dominant surface water bacteria belong
to the Flavobacterium-Cytophaga, Achromobacter-Acineto-
bacter-Alcaligenes-Moraxella and Vibrio-Aeromonas groups and
to the genus Pseudomonas; arid their occurrence in surface
waters is related to depth. Together these bacteria can
comprise up to 90 percent of colony counts in surface waters
(Druce and Thomas, 1970; Yoshimizu, e_t a.1. , 1976).
Chemoorganotrophs in lakes usually are encountered in
numbers of around 10 .to 10 per ml (at 20°C). In Scotland,
three la^es were found to contain respectively 5.6 x 10 ,
3.0 x 10 , and 4.1 x 10 chemooorganotrophs per ml at 20°C
incubation. However, colony counts at 20°C can exceed 10
per ml in highly eutrophic lake water. Even highly oligo-
trophic lakes (with 1 mg per liter or less organic matter)
will have a colony count of around 10 per ml (Godlewska-
Lipowa and Jablonska, 1972); and finished drinking water has
been shown to contain nutrients sufficient for growth of
bacteria when stored in containers at a favorable tempera-
ture and in the absence of microbial antagonisms [See Section
G.6].
Colony counts on plate count agar at 20°C are often
subject to considerable variation depending largely on the
organic matter content. The more fastidious of the chemo-
organotrophs require a minimum of 10 to 100 mg of organic
-------
33
matter per liter (Zo Bell, 1943) and the less fastidious
will grow in water containing as low as (or lower than) 1 mg
of organic matter per 1. Benthic chemoorganotrophs, espe-
cially those in the upper 10 mm of sediment, occur in much
higher numbers and are composed largely of gram-positive
bacteria such as Micrococcus, Corynebacterium, Actinomyces,
Bacillus, and Clostndium.
Colony counts in streams and rivers vary according to
dilution from rainfall and pollution from wastewater. When
dilution is low and pollution high, as often occurs with
heavy irrigation, bacterial levels will be high and flow
rates low. The same holds true for urban communities that
remove more water than they replace or that replace it in -...
the form of wastewater. Colony counts at 20°C usually are
higher for rivers and streams than for lakes and are higher
at 20°C than at 37°C incubation [See Table A.2-1]. Quality
criteria for river water are given in Table A.2-2. [See
also Sections C.I.a, C.4.h, and D.6].
Certain pathogens and indicator bacteria can grow in
the presence of low nutrient concentrations. Both Entero-
.bacter aerogenes and Escherichia coli grew at nutrient
levels up to 50 times lower than those required by Strepto-
coccus faecalis as long as conditions were aerobic (Zo Bell,
1943).
(ii) Chemolithotrophic Bacteria. Chemolithotrophic
bacteria can grow in aquatic environments of purely inor-
ganic content [See Section A.Intro.]. Soluble salts and
carbon dioxide are often present from mineralization and
other processes when organic matter has settled in the
sediment, depriving chemoorganotrophs of a nutrient source.
However, chemolithotrophs can thrive under such relatively
clean conditions (Terney, 1973). Table A.2-3 lists some of
the important chemolithotrophs, classified on the basis of
their growth requirements. Chemolithotrophs are instru-
mental in the nitrogen and sulfur cycles and can oxidize
reduced iron and manganese compounds. Though not responsible
for disease, they often are responsible for operational
problems at treatment plants and taste, odor, and color
problems at the consumer's tap [See Sections F.3.b and G.4].
Nitrification is carried out in the benthos under
aerobic conditions, primarily by two chemolithotrophs,
Nitrosomonas and Nitrobacter. Denitrification is usually
-------
34
TABLE A.2-1
COLONY COUNTS IN FOUR RIVERS OF CZECHOSLOVAKIA
River
Colony Counts (per ml)
20°C
37°C
Danube
March
Vah
Hron
6,200
7,500
2,200
6,100
1,800
2,800
330
740
From Daubner, 1972,
-------
35
TABLE A.2-2
RIVER WATER QUALITY AS DETERMINED BY
NUMBERS OF CHEMOORGANOTROPHS AT 22°C
Quality of Water
Cheraoorganotrophs per ml
Pure river water
Relatively pure river water
Moderately polluted river water
Polluted river water
Highly polluted river water
< 500
5,000 - 10,000
25,000 - 50,000
- 100,000
- 1,000,000
From Daubi.ar, 1972.
-------
TABLE A.2-3
CLASSIFICATION OF CHEMOLITHOTROPHIC BACTERIA
Category
Obligate chemolithotrophs
Mixotrophs
Heterotrophic chemolithotrophs
From Rittenberg, 1972.
Genera and Species
Nitrosomonas europaea, Thiobacillus
thiooxidans, Thiobacillus thioparus.
Thiobacillus neopolitanus/~Thiobacillus
denitrificans, Thiobacillus ferrooxidans
Hydrogenomonas, Micrococcus denitrificans,
Thiobacillus intermedius, Thlgbacillus
novellus, Nitrobacter agilTs"
Desulfovibrio desulfuricans,
rniobacillus perometaboITs~
-------
37
the result of reductions by chemoorganotrophs such as pseudo-
monads, although Paracoccus denitrificans (a facultative
nhemolithotroph), Thiobaciflus denitrificans (a strict
chemolithotroph), and a tew other chemolitnotrophs are
denitrifiers. Levels of nitrogen and the form in which it
enters surface water will influence the numbers of nitri-
fying bacteria; in turn, relative numbers of nitrifiers and
denitrifiers will decide the ultimate form nitrogen is to
take: whether ammonia, nitrate, nitrite, or nitrogen gas
which escapes to the atmosphere.
Sulfur, upon entering surface waters by way of the
atmosphere or through wastewater discharge, sets in motion a
number of microbial oxidations and reductions carried out by
chemolithotrophic and chemoorganotrophic bacteria. These
reactions can lead to the formation of various undesirable
sulfurous compounds that not only produce bad tastes and
odors, but that also can degrade concrete structures used
for water storage and distribution [See Section G.4.b].
There may occur in surface waters both sulfur oxidizing
and sulfate reducing chemolithotrophs. Thiobacillus species
and Beggiatoa can oxidize sulfides, elemental sulfur, or
thiosulfate to acceptable or .unacceptable forms. Although
all species of Thiobacillus, except T. denitrificans, are
obligate aerobes, they inhabit primarilygthe benthic zone
(in numbers as high as or higher than 10 per g: Fjerding-
stad, 1969; Terney, 1973) and can oxidize hydrogen sulfide
(which is found only in the oxygen-poor hypolimnion) to
sulfuric acid. The presence of hydrogen sulfide may be due
to a microbial decomposition of proteins or to the anaerobic
reduction of sulfates (or other reducible sulfur compounds)
carried out by the gram-negative chemoorganotrophs Desulfo-
vibrio (nonsporeforming) and Desulfotomaculum (sporeforming).
Counts of IQ Desulfovibrio and 1(T Desulfotomaculum per
gram have been found in bottom sediments in Denmark (Fjerding-
stad, 1969). The thermophile Desulfotomaculum is respon-
sible for hydrogen sulfide odors emanating from hot water
tanks where conditions have become anaerobic.
In view of the problems caused worldwide by iron bac-
teria in surface water, well water, and groundwater sup-
plies, it is surprising that so little is known about actual
numbers occurring in raw source waters. Biological oxi-
dation of ferrous compounds to insoluble ferric compounds
with consequent encrustations along pipes and in reservoirs
and organoleptic changes in finished water, has been reported
in Canada, the U.S., the U.K., Scandinavia, Australia,
-------
38
India, and some African countries (Cullimore and McCann,
1977). In addition, iron oxide or iron hydroxide deposits
on bacterial cell surfaces form a protective layer that can
hamper the effectiveness of disinfection [See also Sections
F.3 and G.3].
The chemolithotroph Thiobacillus ferrooxidans has been
found in numbers as high as 10b per ml in iron-containing
waters of Denmark (Fjerdingstad, 1956). Gallionella,
Crenothrix, Clonothrix, Toxothrix, and Lieskeela are proba-
bly chemolithotrophic and aerobic or microaerophilic; and
all have been isolated from surface waters (Cullimore and
McCann, 1977). Iron bacteria known to be chemoorganotrophs
are Sphaerotilus and Leptothrix, both of which can oxidize
ferrous and manganous compounds. They have been isolated
from both uncontaminated and contaminated iron-containinq
waters throughout the world.
_ Under acidic conditions, only the biological oxidation
of iron is significant, whereas at a neutral pH, both bio-
logical and chemical iron oxidation can occur. Iron bac-
teria grow within a temperature range of 5 to 34°C if there
is more than 0.2 mg of ferrous ions per liter and if the pH
is within a range of 5.4 to 7.2. Some iron bacteria, notably
Sphaerotilus, can utilize inorganic compounds for energy and
carbon when the required vitamins also are present.
_ (111) Phototrophic Bacteria. Photosynthetic organisms,
primarily the aquatic bacteria, can inhabit surface waters.
included in this diverse physiological community are genera
from the families Rhodospirillaceae (nonsulfur purple and
brown bacteria), Chromatiaceae (sulfur purple bacteria), and
Chlorobiaceae (green sulfur bacteria). Bacteria from these
families oxidize reduced inorganic substances such as molecular
hydrogen, hydrogen sulfide, elemental sulfur, or low molecular
weight organics. The presence of iron seems important to
their metabolism. They require a highly reduced or anaerobic
environment in which to carry on anoxygenic photosynthesis
(where free oxygen is not a by-product) and therefore occupy
a stratum of minimal light penetration near the benthic zone
(Buchanan and Gibbons, 1974).
In contrast, the cyanobacteria (or blue-green bacteria,
formerly classified as blue-green algae) inhabit the aerobic
zone and carry on oxygenic photosynthesis. Unlike the
ma3ority of anoxygenic phototrophs, cyanobacteria can fix
-------
39
atmospheric nitrogen; hence, they can thrive in aquatic
environments that lack a combined nitrogen source such as
ammonia or nitrate. Their prevalence in oligotrophic,
mesotrophic, and especially in phosphate-rich eutrophic
bodies of water is attributed to these dual abilities to
photosynthesize and to fix nitrogen. It is generally agreed
that the growth of cyanobacteria is limited only by the
availability of phosphates. Hence, any waters rich in
phosphates, such as those receiving phosphate-laden waste-
water, tend to accelerate the growth of the organisms.
Massive blooms of species from some genera (e.g.,
Anabaena, Oscillatoria, Microcystis, Lyngbya, Gleotrichia,
Anacystis) have created considerable problems in water
treatment facilities by clogging filters and imparting
noxious tastes and odors (Benson-Evans and Williams, 1975)
[See Section F.S.b]. Some of the cyanobacteria have also
been reported to produce toxins that, if ingested, may cause
gastrointestinal disturbances (Mackenthun, e_t al. , 1967)
[See accompanying section]. Some of the genera mentioned
above have been used as indicator organisms for assessing
the degree of eutrophication. Great numbers of fish in-
habiting shallow waters have perished in winter as a result
of frequent severe summer blooms: when the photosynthetic
activity of aquatic plants is hindered due to reduced
surface aeration and solar insulation (especially in the
case of ice cover), dissolved oxygen is depleted as decom-
posing masses of cyanobacteria settle to the benthic zone.
(iv) Other Microorganisms. Numbers of algae, protozoa,
and fungi occurring in a river or lake reflect its nutrient
state. Predominant classes of algae occurring in lakes are
the Chlorophyceae (green algae) and Bacillariophyceae (diatoms)
Up to 100 species of algae have been found in lakes; yet,
high diversities and low numbers of algae signify relatively
clean conditions whereas low diversities and high numbers
signify eutrophic conditions. Algae are photoautotrophs in
the light, but can be chemoorganotrophs in the dark. Flag-
ellates of the classes Euglenophyceae and Chrysophyseae are
most often implicated as the cause of tastes described as
bitter or sweetish and odors described as "fishy", "grassy",
"musty", and "earthy" (Palmer, 1962). Freshwater protozoa
include genera of the classes Flagellata, Ciliata, and
Sarcodina, the latter containing in it the pathogens
Acanthamoeba, Entamoeba histolytica, and Naegleria fowleri
[See Section B.l.c].
-------
40
^ Fungi, including yeasts, are chiefly soil organisms and
their mostly accidental presence in water does not appear to
affect significantly the quality of drinking water. Yeast
cells normally comprise < 1 percent of the microbial flora,
but may exceed this level in streams and rivers that receive
effluent from certain industries such as breweries or metal
works. Yeasts, predominantly Rhodotorula, Candida [See
Section C.2.f], Cryptococcus, and ToruTopsis were isolated
from the St. Lawrence River (Canada) in numbers of from 0 to
9,500 cells per ml (Simard, et a^., 1970). Also common to
freshwaters are the fungi TrTchosporon, Cladosporium, Endo-
myces, and Deboromyces (Buck, 1975; Daubner, 1972).
b- Lakes and Reservoirs. Lakes and reservoirs (which
include any depressions of land created by natural causes,
such as glacial and volcanic action, or through human effort,
that retain water from rain, rivers, or springs) are the
bodies of water most frequently used as sources of drinkinq
water in the U.S. Yet, very little is known about the
ecology of microorganisms, and even less about viruses, in
such waters.
The waters in lakes and reservoirs, as a group, differ
markedly in chemical composition and are frequently clas-
sified on the basis of nutrient content or degree of bio-
logical productivity. Biological productivity here refers
to the amount of organic matter synthesized from inorganic
substances through photosynthetic activity; and produc-
tivity, as a rule, is a function of the levels of nutrients
(e.g., C, N, P, and S) present in the water. Oligotrophic
lakes are those lakes relatively low in nutrients or rela-
tively infertile in which nutrient recycling is of the
autochthonous (self-nourishing) type. Eutrophic lakes, on
the other hand, are productive lakes rich in nutrients of
which the major portions are derived from the external
environment (allochthonous). Lakes that fall between these
two extremes are considered mesotrophic.
Microbial populations within oligotrophic lakes are
rich in species diversity but low in numbers; in excessively
eutrophic lakes, as a rule, there is less species diversity
but a massive increase in the populations of a few species,
as in the case of algal blooms. These few typically domi-
nant species (of the genera Anabaena, Microcystis, Lyngbya,
and Anacystis) are frequently referred to as index species
of eutrophication or pollution.
-------
41
Some of the major causes of accelerated eutrophication
are the discharge of nutrient-rich effluents from sewage
plants and septic tanks and of farm runoff. Obvious undesir-
able consequences of accelerated cultural eutrophication are
an increase in turbidity, promotion of a toxic algal bloom
[See Section A.2.c], excessive stratification of dissolved
oxygen in the summer, or depletion of dissolved oxygen in
the winter. Depletion of dissolved oxygen during the winter
is attributable to aerobic decomposition of accumulated
organic matter (e.g., dead algae and weeds) which removes
dissolved oxygen from the water. Consequences of the deple-
tion of oxygen include a drastic alteration in distribution
of micro- and macrocommunities within the lake. Anaerobic
digestion of sediment organic material by indigenous bottom-
dwelling organisms (such as species of Desulfovibrio, Desul-
fotomaculum, Butyrivibrio , Selenomonas, Clostndium, etc.)
generally results in production of reduced inorganic ions
(e.g., Fe+* and Mn*+) as well as products that impart undesir-
able tastes and odors.
Most lakes and reservoirs serving as community water
sources have characteristic zonation and temperature strati-
fication. Most lakes are rimmed by a relatively shallow
littoral zone where nutrient accumulation is highest; further
offshore is the limnetic or photic zone where sufficient
light is available for photosynthetic activities by primary
producers (the phytqplankton), and a profundal zone where
photosynthetic activities cease as a result of insufficient
light penetration.
The bacteria within the littoral zone are largely
representative of soil microflora, but conditions favor the
survival of nutritionally nonfastidious gram-negative organisms.
Among the notable bacteria frequently encountered here are
oxygenic, nitrogen-fixing cyanobacteria (formerly known as
blue-green algae); anoxygenic, photosynthetic bacteria '
(members of Rhodospirillales); chemolithotrophic nitrogen-
fixing bacteria (members of Nitrobacteraceae); organotrophic
bacteria (such as species of Pseudomonas, Aeromonas, Alcaligenes,
and Flavobacterium); gram-positive bacteria (such as species
of Bacillus, Micrococcus, Mycobacterium, and Corynebacterium);
and various species of Actinomyces.The discharge of sewage
effluent may add members of the family Enterobacteriaceae,
such as species of Escherichia, Proteus, Enterobacter,
Yersinia, and others, some of which may be pathogenic (species
of Salmo'nella, Shigella, Vibrio, etc.).
-------
42
The viability of pathogens, or of bacteria considered
as potential or opportunistic pathogens, is dependent on
such factors as available nutrients, temperature, and pH, as
well as biotic factors such as predation and antagonism.
The grazing of bacteria by protozoa, bacteriovorous bacteria
(?dellovibrio), and bacteriophage in water may destroy
millions of bacteria. Further, antibiotic substances produced
by some microorganisms may destroy other species. The role
of cyanophages, which attack cyanobacteria, remains to be
evaluated. In addition, lakes rich in suspended solids may
harbor considerable numbers of epiphytes or peritrophic
bacteria (such as species of Caulobacter, Hyphomicrobium,
Seliberia, Thiodendron, Gallionella, Leptothrix, Crenothrix,
and Clonothrix). Some of these have been known to cause
considerable problems in the water treatment process, either
by fouling sand filters or imparting undesirable tastes and
odors.
Lakes that are sufficiently deep generally display
temperature stratification during summer. The region of
rapid temperature drop with depth is known as the thermo-
cline? water above the thermocline is called epilimnion, and
below, the hypolimnion. As a rule, the thermocline layer
prevents mixing of epi- and hypolimnion water during summer,
but in spring and fall when the temperature of the epilimnion
reaches approximately 4°C (the temperature at which water is
densest), the cold water begins to sink and causes extensive
mixing of both layers of water resulting in complete circula-
tion (holomixis). Lakes which turn twice a year are called
dimictic, but there are lakes which fail to turn (amictic),
ones which turn once a year (monomictic), or those which
turn constantly (polymictic). The last group of lakes are
rarely used as source water because they are either under
permanent ice cover or are located in colder regions.
As a rule, distribution of micro- and macroflora within
the lake ecosystem is neither constant nor necessarily
predictable except within wide limits. This is because the
conditions of the littoral zone change hourly, from week to
week, or season to season, as this region of the lake is
subject to the intrusion of sewage effluent, soil runoff,
and other external disturbances, such as wind and wave
action, utilization of the water by wild animals, and abstrac-
tion of water for various human uses.
The total number of microorganisms within the limnetic
zone may be a few hundred per milliliter during the quiescent
period, but the microbial population immediately after
-------
43
circulation may reach a few million per milliliter. This
change is primarily attributed to disturbance of the bottom
sediment and upwelling of nutrients as a result of the
circulation. The organic matter stirred up from the benthic
zone temporarily furnishes a rich and varied pabulum.
Numerous species of aerobic and facultative organisms
(Bacillus, Micromonospora, Micropolyspora, Chromobactenum,
and Flavobacterium) as well as some strict anaerobes
(Bacteroides, DesUlfovibrio, Desulfotomaculum, Sporosarcinaf
Clostridium, Methanobacterlum, etc.) previously present in
undetectable numbers may multiply enormously, frequently
reaching millions per milliliter, and some previously numerous
species may be suppressed by newly multiplying antagonistic
species. The circulation may be accompanied by increased
turbidity and the lakes with an excessively eutrophic condi-
tion may actually show an increase in BOD content as well as
production of a noticeable hydrogen sulfide odor. Circu-
lation may continue for a few days or weeks, depending on
such factors as rate of water temperature change, average
depth of the lake, etc.
Microbial communities in lakes and reservoirs are
extremely complex and fragile. They are readily altered by
internal as well as external influences. External influ-
ences not under local control, such as "acid rain", are of
considerable concern in the deterioration of water quality
especially in those lakes which lack a limestone base.
However, cultural eutrophication of a lake is the single
most important factor in the deterioration of water quality.
Eutrophic lakes tend to promote the bloom of various algae,
including toxigenic species; lengthen the viability period
or even promote "aftergrowth" of undesirable bacteria; cause
increased turbidity; and impart undesirable odors and tastes.
The water derived from such lakes requires extensive and
costly treatment. Thus, lakes and reservoirs serving as
source waters must be protected from external influences
that tend to deteriorate water quality.
c. Toxic Cyanobacteria in Raw Water Supplies. There
are numerous reports about poisonings of livestock, pets,
and wildlife by ingestion of waterblooms of cyanobacteria;
and mounting evidence indicates that man may also be affected
by toxic cyanobacteria through water supplies (Dillenberg
and Dehnel, 1960). Outbreaks of human gastroenteritis which
occurred in Charleston, West Virginia, and the area served by
the Anacostia Reservoir near Washington, D.C., in the drought
years of 1930 and 1931, were attributed to growths of cyano-
bacteria in the water supplies. The usual bacterial causes
-------
44
of gastroenteritis could be excluded, but direct toxic
effects of the cyanobacteria or associated bacteria were not
clearly established.
Recent cases of cyanobacterial toxicity to humans have
been reported by Dillenberg and Dehnel (1960) who described
cases of human gastroenteritis resulting from the ingestion
of heavy blooms of Anabaena, Microcystis, and/or Aphanizomen
from a number of lakes and reservoirs in the province of
Saskatchewan, Canada. The most convincing case cited in
this review is of a physician who accidentally fell into a
lake containing a heavy bloom of cyanobacteria and swallowed
an estimated half-pint (250 ml) of water. A few hours
later, he suffered stomach pains, nausea, vomiting, painful
diarrhea, fever, headache, and pains in limb muscles and
points which lasted for two days. Samples of slimy green
stool subjected to laboratory analysis for pathogens (includ-
ing virus) revealed no Salmonella or Entamoeba, but many
cells of Anabaena and Microcystis.
By isolating, growing, and testing (by intraperitoneal
or oral administration to laboratory mice) cultures of the
most common bloom-forming species (Anabaena flos-aquae [L.]
de Breb., Microcystis aeruginosa Kutz. emend Elenkin, and
Aphanizomenon flos-aquae [L.] Ralfs.) from a number of
different lakes, ponds, and reservoirs, investigators have
shown that there are toxic strains of these organisms. They
have, in addition, been able to determine fully or partially
the chemical structure of at least one toxin from one strain
of each of these three species. A number of toxins from
newly isolated strains of Anabaena flos-aquae have recently
been described (Carmichael and Gorham, 1978), and a toxin
from a new strain of Microcystis aeruginosa has recently
been found. _Slower-acting toxins produced by bacteria
associated with these cyanobacteria have also been noted
(Gorham, 1965).
Known strains of Anabaena flos-aquae produce four
(perhaps six) toxins (Carmichael and Gorham, 1978). The
first of these, called anatoxin-a, is an alkaloid with both
pre- and post-synaptic neuromuscular blocking activity.
Anatoxins-b and d are fast-acting and are suspected of being
alkaloids. Anatoxin-c is slower-acting and is suspected of
being a peptide.
The effects of a toxin produced by a new strain of
Microcystis aeruginosa, called type c, are indistinguishable
from those produced by anatoxin-c. Microcystin, the cyclic
polypeptide toxin produced by Microcystis aeruginosa NRC-1
-------
45
affects the cardiovascular system and produces character-
istic lesions of the liver when administered by intraperi-
toneal injection or by the oral route to mice, guinea pigs,
rabbits, chickens, ducks, calves, and a lamb (Konst, et
al., 1965). Toxic mixed blooms of Aphanizomenon flos-aguae
and Microcystis aeruginosa produced signs that were indis-
tinguishable trom microcystin poisoning (Gorham, 1965).
Since no cultures were successfully established, it could
not be decided whether the toxin came from the Aphanizomenon,
the Microcystis, or from both. Blooms of Microcystis toxica
Stephens have been reported to produce an alkaloid hepa-
totoxin of unknown structure.
An atypical strain of Aphanizomenon flos-aguae produces
an ichthyotoxin, called aphantoxin, wnich kills mammals
(Gentile and Maloney, 1969). It is a mixture of which one
component is, surprisingly, the known alkaloid saxitoxin.
Saxitoxin is the paralytic shellfish toxin which affects
man and is produced by the marine dinoflagellate Gonyaulax
catenella (Schantz, e_t ad. , 1975).
The question of whether or not toxins from cyanobacteria,
if present in raw water supplies in significant quantities,
would survive water purification practices has been considered
Cases in the province of Saskatchewan of poisoning associated
with the presence of cyanobacteria led to tests for the
occurrence of toxic cyanobacteria in blooms of two impound-
ments that served as the water supply for three of the
province cities. Raw untreated water was either mildly or
fatally toxic as determined by intraperitoneal injections
into mice (Dillenberg and Dehnel, 1960). These same authors
reviewed tests conducted by Wheeler and coworkers in which
they found that toxins produced by Microcystis aeruginosa
blooms maintained their toxicity after the laboratory
equivalent of water purification processes, including alum
coagulation, filtration, chlorination, and activated carbon
treatment. Massive amounts of activated carbon were needed
to render the effluent non-toxic.
Gorham (1962) considered that two of the known cyano-
bacterial toxins, microcystin and anatoxin-a, would normally
be removed or inactivated by the usual water treatment
procedures. Moreover, the large doses and high percentage
(75 percent or higher) of toxic strain necessary to produce
symptoms in animals suggest that the comparatively small
amounts of toxins that might remain in a drinking water
supply after treatment probably would not be sufficient to
cause human poisoning.
-------
46
One must also consider that strains of different species
of cyanobacteria are capable of producing a great variety of
toxins, individually and in mixtures. Since some are toxic
to man, it could be that others, which may not be eliminated
by standard water treatment practices, may also prove toxic.
This suggests a need for a comprehensive study of the effects
of water treatment practices upon toxins produced by different
strains and species of cyanobacteria.
If one could select a set of representative water treatment
practices to be evaluated as a model laboratory-scale process,
the suggested investigation could proceed in three stages.
For each of several known toxic strains of cyanobacteria,
one would produce a mass culture and determine the extent of
detoxification which resulted from the model process. Then
one would concentrate naturally-occurring blooms of cyano-
bacteria from raw water supplies and measure the amount of
detoxification, by the model process, of those blooms which
were found to be toxic. Finally, one would grow laboratory
cultures from the toxic natural blooms and attempt to confirm
the detoxification results more critically.
Such a study would go far towards evaluating the risk
to man of consuming treated water produced from sources
containing toxic cyanobacterial blooms. It would also help
to identify specific treatment practices which are capable
of detoxifying such source waters and to suggest possible
improvements.
d- Sampling, Transport, and Microbiological Requirements
tor Raw Surtace water.A council directive submitted to the
director of the Commission of the European Communities
(Commission of the European Communities, 1975) has proposed
that raw surface waters be monitored using microbiological
parameters; and a more recent directive (Commission of the
European Communities, 1978) has included suggestions for
sampling protocol, frequency, and transport, along with ways
to ensure analytic reproduceability [See Table A.2-4].
According to this proposal, samples are to be collected from
a zone 50 cm below the surface and 50 cm above the bottom;
or, if not possible, samples are to be taken at a depth
midway between the surface and the bottom. These speci-
fications are meant to ensure that samples will contain
bacterial numbers that accurately reflect levels in the
water phase and not those in resuspended sediment, from
which much higher bacterial counts would be obtained.
However, most saprophytic bacteria are concentrated in the
few millimeters below the surface.
-------
TABLE A.2-4
PROPOSED WATER QUALITY CRITERIA TO BE USED FOR
RAW SURFACE WATER SUPPLIES
Category
of Raw
Surface
Water
Aia
A2b
A3C
Samples/Yr, for
Population Served
x in5 > in5 Coliforms/
< 1U " 1U 100 ml
1 3 50
2 6 5,000
4 12 50,000
Maximum Concentration Allowed
Thermo- tolerant
Coliforms/100 ml
20
2,000
20,000
Fecal
Streptococci
100 ml
20
1,000
10,000
Salmonella
0/5 liter.s
0/1 liter
aSurface water requiring only simple physical treatment (filtration) and disinfection.
Surface water requiring physical and chemical treatment (coagulation, flocculation,
filtration) and disinfection.
GSurface water requiring more extensive physical and chemical treatment (coagulation,
flocculation, filtration, along with supplementary treatment such as active carbon)
and disinfection with chlorine and possibly ozone.
(COMMISSION OF THE EUROPEAN COMMUNITIES, 1975 AND 1978)
£»
-J
-------
48
On the other hand, strong correlations were found
between presumptive total coliform MPN's in sediments and
virus numbers in recreational coastal waters (Goyal, et
&1i 1978). Moreover, sediments can serve as a reservoir
for enteroviruses which may become resuspended in the water
phase during heavy rains (la Belle and Gerba, 1979). The
microbiological quality of sediments also needs periodic
evaluation, possibly by surveying sediment samples for
coliforms and thermo-tolerant coliforms as well as fecal
streptococci or Clostridium perfringens [See Section C.I].
As wastewater treatment practices become more complex,
there is a greater possibility of reducing indicators without
necessarily producing concomitant reductions in some of the^
more resistant organisms such as viruses and protozoa. Both
viruses and parasites may persist for extended periods in
water and are fairly resistant (especially parasites) to
chlorine [See Section B.3, E.I, and E.S.a]. By not includ-
ing tests for viruses, and perhaps certain parasites, the
Commission of the European Communities proposal does not
take into account consequences from contamination of surface
water with treated wastewater.
_Another pollution parameter not dealt with in the
Commission's proposal is that of colony counts at 20 to 22°C,
which would enable detection of psychrophilic organisms.
Tests for these organisms could serve to indicate any heavy
influx of organic and inorganic pollutants into surface
waters [See Sections C.I.a and D.6]. Such an occurrence
could give rise to algal and cyanobacterial blooms and could
increase opportunities for biotoxin production. In addition,
the presence of several potentially pathogenic bacteria,
and others of special concern to the food industries, has
been associated with these blooms (potentially pathogenic
species of Aeromonas have been observed in large masses of
dying algae: Simudi, e_t a_l., 1971). Species of Klebsiella,
an opportunistic pathogen, are known to propagate in water
containing wastes from textile, lumber, and sugar industries.
Strains of these bacteria frequently are adapted to colder
environments and are able to grow at temperatures ranging
between 0 and 15°C [See Section C.2.d].
e; Protection of Surface Water. It is important to
recognize the connection between surface water contamination
problems and waste disposal problems. One will not be
solved without the other, for even adequately treated surface
waters have caused disease when recontamination occurred
-------
49
during distribution. Out of 54,935 cases in the U.S. (1946
to 1970) attributed to surface water contamination, 8,992 of
these were associated with "treated" surface water (Craun
and McCabe, 1973).
Surface water, usually to a far greater extent than
groundwater, is exposed to various kinds of contamination.
In countries where the drinking water supply is derived from
surface waters, measures to protect against chemical, physical,
and biological deterioration need upgrading. Laws to protect
surface water should be more restrictive and more specific;
and they must be backed up with adequate provision for
enforcement. The discharge (including sub-surface seepage)
of raw or inadequately treated waste into raw source waters
should be prohibited; and minimum distances%from freshwater
sources for disposal of wastes need precise delineation.
Dogs and motor boats should not be allowed, nor is swimming
acceptable in waters designated for potable use. Tight
restrictions should be placed on camping, sailing, and
other recreational activities in or around surface water
sources [See also Section E.I].
However, no amount of regulation will eliminate all
microbiological or chemical problems because surface waters
will always be in unavoidable contact with wild animals,
especially migratory birds, and because some organisms
(e.g., cyanobacteria, Pseudomonas, Alcaligenes, Chromobacter,
Acinetobacter, Vibrio, Aeromonas, Yersinia, and Bacillus)
may continue to grow (albeit in lower numbers) after all
waste discharge has ceased. The most difficult, yet the
most critical, issue to deal with is that of initiating
steps to curb entry of pollutants into surface waters.
Beyond this, authorities can only ensure hygienic safety by
conducting routine and appropriate analyses of raw source
waters and'by making sure that all surface waters used for
drinking purposes receive complete and continuous treatment
before distribution.
Programs for routine analysis should be selected on the
basis of knowledge about: (1) what conditions favor propaga-
tion of pathogens, potential pathogens, and cyanobacteria;
(2) what conditions favor production of undesirable compounds
by chemolithotrophs (notably iron bacteria) and chemoorgano-
trophs; (3) types of-information that can be obtained from
microbiological examination of sediments; and (4) what
hazards are associated with the presence of viruses, proto-
zoan cysts, metazoan eggs, and endotoxins and other bio-
toxins in surface water. Mapping of waterborne disease
-------
50
outbreaks [See Section B.5] should be carried out in con-
junction with microbiological monitoring, with special
attention paid to possibilities for growth of hitherto
unrecognized pathogens such as Yersinia enterocolitica [See
Section B.I.a(iii)], Campylobacter fetus I See Ejection
B.l.a(viii)], and Legionella pneumophila (Center for Disease
Control, 1978e). As different pathogens exhibit different
virulence in the host and different stabilities in the
aquatic environment [See Sections B.3 and B.4], the kinds as
well as numbers of pathogens in a given source water should
be determined prior to decisions about type and rigor of
drinking water treatment [See Section E].
All of these issues are part of one overriding concern:
that of finding ways, some of which have been discussed in
this report, to ensure that surface water used for drinking
purposes shall be protected, to the utmost extent possible,
from the discharge, runoff, or seepage of wastes.
3.
Survey of the Bacteriological Quality of Raw Water
Supplies from Nine Countries~~
An important consideration to be dealt with in Topic
Area A, and one which impinges on every other section of
Project Area III, concerns the differences from one water-
works to the next and from one country to the next, in
bacteriological quality of the raw water supply. The tables
and text that follow are the result of surveys taken of
waterworks facilities from nine developed countries (Norwegian
Institute of Water Research, 1980). In order to draw a
basis for comparison, the parameters chosen were total and
thermo-tolerant coliforms, the widely used and accepted
indicators of water quality.
One problem encountered was the variation among coun-
tries in methods for sampling and analysis [See Section D.I
to 5J. Another weakness of the study is that, except for
the Netherlands, Norway, and Sweden, information from water-
works in the various countries surveyed represents only a
small part of the total raw water supply and population
served. On the other hand, each country's national contact
for the drinking water microbiology project, who was respon-
sible for collecting data, was instructed to choose repre-
sentative raw water supplies wherever feasible. Though the
flaws in this study cannot be ignored, neither must the
reader discount the knowledge to be gained from this type of
study (one that no international organization has as yet
undertaken). It is hoped that documentation and comparisons
-------
51
of raw water quality in the various countries participating
in this survey will lead to improvements where warranted.
Data were gathered principally from sampling and analyses of
raw water during the year 1977, except in cases where no
samples had been taken that year. Results were then re-
quested for the previous or the following year.
a. Size of Waterworks/ Source of Raw Water, and
Sampling Frequencies"The number of people served by a
waterworks will be reflected in the size of that facility
[See Table A.3-1]. Thus, small waterworks predominate in
Norway, most facilities in Sweden serve a medium-sized
population, and the other countries surveyed (except France)
presented data from waterworks, most of which serve > 20,000
people. Apparently all but one waterworks (UK) in this
study serve > 1,000 people. Data from Norway and Sweden are
the most complete, representing the majority of waterworks
in these countries serving > 1,000 people.
Norway derives most of its drinking water from surface
water sources. Although there are more suppliers of ground-
water than of surface water in Sweden, a larger proportion
of the population receives water from surface sources;
likewise, in the Netherlands, 50 percent or more of the
population uses surface-derived water [See Table A.3-2].
Whereas most waterworks in Sweden sample one to six
times per year, the sampling frequency in Norway falls more
evenly into all the categories listed in Table A.3-3. A
large number of suppliers take no samples, but there also is
a sizeable group of them that sample > 26 times per year.
In Sweden, only groundwater may bypass routine monitoring
and treatment, but this also can be true of some surface
water supplies in Norway. Although epidemiological and
environmental conditions in Norway appear stable, Norwegian
regulations specify that treatment and routine bacterio-
logical monitoring of all surface water supplies are required.
One explanation for the apparent discrepancy is that many
Norwegian waterworks have not yet received governmental
licensing. The other countries (except France) reported
high sampling frequencies, which probably is due either to
the size of the waterworks or the level of pollution.
-------
TABLE A.3-1
RECORDED WATERWORKS BY POPULATION SERVED
52
Country
Canada
Denmark
UK
Prance
FRG
Netherlands
Norway
Sweden
US
< 5,000
5
-
1
4
-
-
222
24
2
- 5,000 to < 20,000
8
-
5
_
-
-
91
126
1
- 20,000
14
"" »
17
2
9
3
31
60
7
Total Number
of Waterworks
27
13
23
6
9
3
344
216
10
-------
TABLE A.3-2
RECORDED WATERWORKS SERVING MORE THAN 1,000 PEOPLE
Groundwater
Country
Canada
Denmark
UK
France
FRG
Netherlands
Norway
Sweden
US
People
Served
168,000
_
357,000
30,000
3,000,000
0
61,000
1,967,000
236,000
Number
Water-
works
11
11
12
5
2
0
19
118
5
Surface Water
People
Served
3,204
-
7,528
240
5,600
7,000
2,750
2,977
360
,000
,000
,000
,000
,000
,000
,000
,000
Number
Water-
works
16
2
11
1
7
3
325
98
5
Total
People
Served
3,372,
860,
7,885,
270,
8,600,
7,000,
2,811,
4,944,
596,
000
000
000
000
000
000
000
000
000
Number
Water-
works
27
13
23
6
9
3
344
216
10
Total
National
Populations
23 x
5 x
46 x
53 x
61 x
14 x
4 x
8 x
217 x
IO6
IO6
IO6
IO6
IO6
i°6
IO6
IO6
IO6
% of
Population
Served
14.7
17.2
17.1
0.5
14.1
50.0
70.3
61.8
0.3
Ul
CJ
-------
TABLE A.3-3
FREQUENCY OF BACTERIOLOGICAL ANALYSES OF RAW WATER
AT THE RECORDED WATERWORKS
54
Number of Analyses per Year
Country
Canada
Denmark
UK
France
FRG
Netherlands
Norway
Sweden
US
0
0
-
0
0
0
0
107
37
0
1-6
2
-
0
4
0
0
82
120
0
7-12
1
-
3
1
0
0
51
40
0
13-26
11
2
0
0
0
0
28
5
2
> 26
13
-
20
1
9
3
76
14
8
Total
27
13
23
6
9
3
344
216
10
-------
55
b. Bacteriological Quality of Raw Water Supplies.
Many surface water sources in Norway, and some in Sweden,
show high standards of bacteriological quality (Sweden tests
for total, but not for thermo-tolerant, coliforms). Except
for Canada, which reported relatively low coliform levels at
some waterworks, raw surface water in the other countries
surveyed appears to be of low bacteriological quality [See
Tables A.3-4 and A.3-5]. In contrast, but to be expected,
groundwater sources in most countries were of high bacterio-
logical quality (results from the FRG are based only on
tests for thermo-tolerant coliforms in groundwater) [See
Tables A.3-6 and A.3-7].
Most of the waterworks surveyed, in most of the countries,
reported extensive treatment of water prior to distribution
to the consumer. In cases where the raw water received only
minimal (aeration or microstraining whereby water is filtered
through finely woven fabrics of stainless steel to remove
suspended solids) or no treatment, the bacteriological
standards usually indicated water of potable quality [See
Table A.3-8]. Exceptions were noted for France, Norway, and
Sweden. All minimally treated or untreated source waters
reported for Canada, France, and the U.S., and most in
Sweden are derived from aquifers; in Norway, however, they
are derived largely from surface water sources.
c. Conclusion. This tabulation serves to point up
the large variation in raw water quality in the different
countries surveyed. Where the raw water quality is good,
sampling frequencies often tend to be low and the water
receives little or no treatment. However, infrequent .sampling
and testing cannot be expected to alert' a supplier quickly
to changes in conditions in the community or to a sudden
influx of contaminants into the water supply.
Countries such as Norway and Sweden, where waterworks
are numerous and predominantly of small or medium size, have
greater flexibility in finding means to supply safe, potable
water. At the same time, however, there is less centralized
control and uniformity of policy for managing and treating
raw water supplies.
Those countries fortunate enough to possess water of
high bacteriological quality are encouraged to enact policy
that will assure its continued excellence.
-------
TABLE A.3-4
BACTERIOLOGICAL QUALITY. OF SURFACE WATER; TOTAL COLIFORHS
Number
< 2
Country
Canada
Denmark
UK
France
FRG
Netherlands
Norway
Sweden
US
People
Served
20,000
0
0
0
-
0
257,000
224,000
170,000
Number
Water-
works
1
0
0
0
-
0
113
16
1
of Total Coliforms
- 2 to
People
Served
128,000
-
41,000
0
-ft
0
938,000
970,000
44,000
< 30
Number
Water-
works
4
1
1
0
-
0
106
40
1
per 100 ml
^ 30
People
Served
3,056,000
-
7,487,000
240,000
5,600,000
7,000,000
249,000
1,389,000
146,000
Number
Water-
works
11
1
10
1
7
3
.32
32
3
No Data Submitted Total
Number
People Water- People
Served works Served
3,204,000
-
7,528,000
240,000
5,600,000
7,000,000
306,000 74 2,750,000
394,000 10 2,977,000
360,000
Number
Water-
works
16
2
11
1
7
3
325
98
5
Ul
o\
-------
TABLE A.3-5
BACTERIOLOGICAL QUALITY OP SURFACE WATER: THERMO-TOLERANT COLIFORMS
Number of Thermo-Tolerant Coliforms per 100
Country
Canada
Denmark
UK
France
FRG
Netherlands
Norway
Sweden
US
< 2
Number
People Water-
Served works
22,000 2
0 0
0 0
0 0
00
0 0
1,365,000 150
-
_ _
^ 2 to < 30
Number
People Water-
Served works
2,747,000 9
2
41,000 1
0 0
0 0
0 0
404,000 50
-
_ -
>
People
Served
422,000
0
7,487,000
240,000
5,600,000
7,000,000
59,000
-
_
ml
30
Number
Water-
works
4
0
10
1
7
3
7
-
-
No Data Submitted
People
Served
13,000
-
- .
-
-
-
923,000
2,977,000
360,000
Number
Water-
works
1
-
-
- "
-
-
118
98
5
Total
People
Served
3,204,000
-
7,528,000
240,000
5,600,000
7,000,000
2,750,000
2,977,000
360,000
Number
Water-
works
16
2
11
1
7
3
325
98
5
U»
J
-------
TABLE A.3-6
BACTERIOLOGICAI, QUALITX OP GROUNDWATER: TOTAL COLIFORHS
Dumber of Total Coliforms per 100 ml
Country
Canada
Denmark
OK
Prance
FRG
Netherlands
Norway
Sweden
OS
<
People
Served
151,000
-
337,000
2,000
-
0
50,000
731,000
223,000
1
Number
Water-
works
9
11
11
2
-
0
15
32
2
* 1 to
People
Served
16,000
0
0
0
-
0
6,000
511,000
13,000
< 2
Number
Water-
works
2
0
0
0
-
0
1
43
3
^ 2
People
Served
0
0
20,000
28,000
-
0
0
304,000
0
No Data Submitted
Number Number
Water- People Water-
works Served works
0 - -
0 - -
1 -
3 - -
3,000,000 2
0 0 -
0 5,000 3
12 421,000 31
0 -
Total
People
Served
167,000
-
357,000
30,000
3,000,000
0
61,000
1,967,000
236,000
Number
Water-
works
11
11
12
5
2
0
19
118
5
in
CD
-------
TABLE A.3-7
BACTERIOLOGICAL QUALITY OP GROUNDWATERs THERMO-TOLERANT COLIPORMS
Number of Thermo-Tolerant Col i forms per 100 ml
Country
Canada
Denmark
UK
France
PRG
Netherlands
Norway
Sweden
US
< 1
People
Served
142,000
-
337,000
29,000
3,000,000
0
54,000
-
M
Number _
Water- '
works
9
11
11
4
2
0
14
-
_
*.l to
People
Served
0
0
0
0
0
0
0
7000
_
< 2
Number
Water-
works
0
0
0
0
0
0
0
5
_
* 2
Number
People Water-
Served works
0 0
0 0
20,000 1
1,000 1
0 0
0 0
0 0
-
-
No Data Submitted
Number
People Water-
Served works
25,900 2
-
-
-
-
-
-
1,967,000 118
236,000 5
Total
People
Served
168,000
-
357,000
30,000
3,000,000
0
61,000
1,967,000
236,000
Number
Water-
works
11
11
12
5
2
0
19
118
5
U1
to
-------
TABLE A.3-8
BACTERIOLOGICAL QUALITY OF RAW WATER
RECEIVING MINIMUM3 OR NO TREATMENT
60
Number of Total Coliforms per 100 ml
Country
Canada
Denmark
UK
France
FRG
Netherlands
Norway
Sweden
US
< 2
5
-
0
2
0
0
72
35
3
- 2 to < 30
0
-
0
2
0
0
25
3
0
> 30
0
-
0
1
0
0
9
2
0
No Value
0
-
0
0
0
0
34
18
0
Total
5
-
0
5
0
0
140
58
3
Minimum treatment microstraining or aeration.
-------
61
4. Summary
As the use of water and the disposal of wastes become
more diversified and intensive, those in positions of authority
will be required to effect needed changes in resource manage-
ment policy. In order to make correct decisions, a water
supplier will have greater need of information about the
impact of water from aspects of human activity, from the
siting of waste discharges, meat processing plants, and
stockyards, to the local incidence of disease.
Currently most drinking water is derived from surface
water sources; but in many cases, deteriorating quality has
led to the search for water, such as ground water, that is
better protected from human activities. At the same time, the
increasing use of soil in treatment and disposal of wastewater
raises safety questions in that the dynamics of microorganisms
in soil and groundwater are, as yet, imperfectly understood '
and are inadequately controlled by bacteriological monitoring
of aquifers.
Each of these entities, surface water and groundwater,
brings with it a distinctive ecological system that responds
differently to inputs of waste at different seasons; hence,
byproducts of microbial interactions with organic and inorganic
matter can be vastly dissimilar, as can conditions for
survival and propagation of indicators and pathogens.
Surface water, although more vulnerable and more enriched
with nutrients and aquatic flora, is essentially a hostile
environment to human intestinal microorganisms. Groundwater,
which usually occurs as a result of some form of soil filtra-
tion, is relatively free of microorganisms. However, once
pollution of the aquifer has taken place, conditions favor
the persistence of whatever organism initially gained entry.
It is important, therefore, that water authorities regularly
monitor groundwater, as is generally recommended for surface
water, and that they be vested with some legal recourse to
impose sanctions in the event of any infringement of regula-
tions.
Where large populations receive their water from one
central supplier there is an inherent susceptibility to
total disruption of the water supply as a result of a war or
natural disaster, especially if alternative raw water sources
are limited. On the other hand, if circumstances such as
war necessitated that many people gather where normally only
a few people lived, it would be incumbent on the responsible
agency to have on hand sufficient volumes of potable water.
-------
62
A comparable problem, in terms of availability of potable
water, arises whenever large numbers of people are assembled
together at resort areas during holidays.
Protection of our source waters will require more than
just good intentions if future generations are to receive
from their governments a sufficient quantity of safe and
palatable drinking water. Careful and appropriate micro-
biologic monitoring along with prompt correction of problems
are essential.
-------
63
B. MICROBIAL PATHOGENS TRANSMISSIBLE
BY WATER
Drinking water is a potential vehicle for pathogenic
organisms if it becomes contaminated with human or animal
feces. Fecal contamination can occur at the raw water
source or anywhere along a distribution system. The pres-
ence of fecal contaminants in finished water can also result
from accidental or routine inefficiency during the treatment
and disinfection of a raw water. A result of such fecal
contamination may be that the consumer will ingest some
quantity of bacterial, viral, or parasitic organisms. The
effects of these will depend both on the demographic and
epidemiologic conditions, and on the overall numbers and
virulence of the organism ingested.
The problem of transmission of microbial pathogens
through drinking water has changed with time; our perception
of the problem has also been changed by the, development of
new research results. Certain bacteria of fecal origin
(e.g., Salmonella, Shigella, and Vibrio cholerae) are less
prevalent than they once were, but are now more likely to
have acquired some resistance to antibiotics. Other micro-
organisms, such as several of the enteric viruses, have been
held in check over the last 25 years. However, there are
those (such as the hepatitis A virus) which still pose
problems that are far from being solved. Likewise, there
appear to be more frequent occurrences of gastroenteritis
in which the causative agent is undetermined but strongly
suspected of being a virus. Parasitic protozoa, such as
Giardia lamblia, also seem to be an occasional cause of
waterborne gastroenteritis.
The most important epidemiolgic issue to be addressed
by the developed countries is the extent to which various
microorganisms can cause sporadic cases or inapparent infec-
tions in the consuming population via the water supply
route. To be able to predict accurately, one must not only
understand epidemiologic facts, but must also take into
account conditions under which pathogens survive in water
-------
64
and retain their capacity to infect. There are deficiencies
in present information gathering systems and in methods for
identifying the responsible waterborne agent. Therefore, if
decisions are to be made on the basis of current epidemio-
logical data, the limitations to this approach must be
recognized.
Research on and detection of pathogenic microorganisms
in polluted drinking water is hampered by various factors
associated with waterborne spread of disease:
1. A mass infection leading to epidemic manifesta-
tions in the consumers is generally sporadic and
isolated rather than regular and repetitive.
2. Several days or even weeks can intervene between
contamination of the drinking water and appearance
of the first symptoms of illness. By the time
water is suspected of being the vehicle of trans-
mission the pathogenic agents most often will have
disappeared from the contaminated water.
3. Even when the causative agent remains present in
the water, its numbers are usually quite low
compared to total bacterial densities in the
water.
4. It is often necessary to eliminate other interfer-
ing organisms present in the sample which could
otherwise mask the presence of pathogens.
5. Finally, the choice of recovery methods must take
into account probable stressed states of pathogens
derived from a more or less hostile aqueous environ-
ment .
For all of these reasons, the following protocol is
especially important when culturing for microorganisms in
water:
1. Concentration of an appropriate volume of water
sample.
2. Enrichment (in certain cases pre-enrichment) in
specific liquid media and transfer to solid selec-
tive media.
3. Use of an optimal incubation temperature, chosen
on the basis of the suspected causal agent.
-------
65
The incidence of pathogenic organisms in drinking
water, and the means by which information about them can be
obtained, are the concerns of this section. Most papers in
the first part, dealing -with individual pathogens, are
organized according to the format: (1) characterization and
pathogenesis, (2) stability and sources, (3) epidemiology,
and (4) detection methods.
1. Categories and Properties of Waterborne Microbial
Pathogens of Humans ~~ ~~
a. Bacteria
(i) Salmonella. The genus Salmonella is made up of
gram-negative, nonsporeforming, generally motile bacilli
having many biochemical reactions in common with other
members of the family Enterobacteriaceae (Buchanan and
Gibbons, 1974). It is a very large group consisting of 1,200
known serotypes, classified according to possession of 0 and
H antigens. Biotypes, determined by different sugar fermen-
tation patterns, or serotypes demonstrated to be resistant
to antibiotics can serve as markers in epideiriiological
investigations. Plasmid transfers with various strains of
Enterobacteriaceae over the last several years have con-
ferred antibiotic resistance to certain of the Salmonella
serotypes.
All serotypes of Salmonella are pathogenic to humans,
causing mild to acute gastroenteritis and very occasionally
death. Typhoid fever, caused by S. typhi and paratyphoid
fever, caused by S_. paratyphi A or B are both enteric fevers
that occur only in humans. The other Salmonella serotypes
are responsible for foodborne intoxication accompanied by
mild to acute gastroenteritis, but rarely death. Referred
to as salmonellosis, these milder forms occur frequently in
humans and wild or domestic animals.
The organism grows either continuously or intermit-
tently in the intestines of sick individuals or clinically
healthy carriers and is excreted through the feces. Sal-
monella is excreted by infected individuals in the human
population (exclusively so for S. typhi), by infected farm
animals, by domestic pets, and b"y other warm-blooded animals
in the wildlife population. The average number of indi-
vidual humans excreting Salmonella at any given time will
vary, from < 1 percent to 3.9 percent based on information
from studies in several countries throughout the world [See
Section B.2]. As a result, a large constant reservoir of
-------
66
Salmonella: exists in the environment, although accurate
carrier rates are difficult to obtain since infected yet
asymptomatic animals can be included among the healthy
carrier population.
Salmonella densities, as well as the number of Sal-
monella serotypes present in sewage discharging to receiving
waters, vary with the number of people served and the extent
of community infection. Salmonella strains were regularly
found in the sewage system of a residential area of 4,000
persons (Harvey, e_t al^, 1969). A sewage collection network
of 50 to 100 homes is considered by Callaghan and Brodie
(1968) to be a minimal size for detecting salmonellae.
Streams, lakes, and rivers receiving discharges of meat
processing wastes or effluents of untreated or ineffectively
treated community sewage may contain substantial numbers of
salmonellae (Geldreich, 1972). Researchers recently re-
covered 32 serotypes both from sewage effluent and from
downstream sections on the Oker River, Germany (Popp, 1974).
Kampelmacher and Jansen (1973) calculated that the Rhine and
Meuse Rivers carried approximately 50 and seven million
salmonellae per second, respectively [For persistence of
Salmonella in water, See Section B.3].
Fish living in polluted water may ingest Salmonella and
become vectors of pathogen transport. Salmonella presence
in animal feed, notably fishmeal, is a case in point.
Experiments with carp suggest that S_. typhi survived for
five to six days in the intestinal tract when the water
temperature was 24 to 27°C, and for six to seven days at
water temperatures of 1 to 5°C (Lee, 1972). When researchers
sought evidence for possible Salmonella multiplication
within the fish gut, they instead discovered a progressive
die-off of S_. typhi in intestinal material from carp and
bluegill at water temperatures of both 10 and 20°C. However,
substantial multiplication of £. typhimurium occurred in 72
h at 20°C in fecal material from both fish species (Geldreich
and Clarke, 1966).
Drinking untreated, unprotected surface water presents
the greatest risk to the consumer. S_. typhi or S. para-
typhi are the agents responsible for most waterborne out-
breaks due to Salmonella contamination of water supplies;
other salmonellae produce illness only after growth in food
at ambient temperatures. Water supplies implicated in
Salmonella outbreaks have included individual water systems,
semi-public water systems, systems on passenger cruise
liners, and small community municipal supplies. While large
-------
67
municipal water supplies are not a common source of Salmonella
outbreaks, the number of cases involved is often large and
represents a major epidemic in the community.
The 1965 salmonellosis outbreak, traced to an unchlori-
nated water supply in Riverside, California, is to date the
largest epidemic in modern times to be caused by salmonella
contamination of a municipal water system. More than 16,000
people became ill, and three of them died (Boring III, et
al., 1971). Negative coliform readings, reported prior to
the outbreak may have been the result of suppression by high
densities;iof non-coliform bacteria (Geldreich, e_t aJ. ,
1972). Muller (1964) found Salmonella in chlorinated drink-
ing water that showed an absence of coliforms, yet had been
contaminated with 100,000 1 of floodwaters during distribu-
tion.
Properly constructed wells, located in high quality
aquifers protected from surface contamination, generally
afford a good untreated potable water supply free of micro-
bial pollution. Unfortunately, not all wells meet such
standards. Five cases of typhoid fever occurred in an area
served by individual shallow wells drilled into a layer of
river bed gravel. Septic tanks or cesspools were located
nearby. In one instance, only 60 meters separated a well
site from the home sewage disposal system of a known S_.
typhi carrier (Center for Disease Control, 1972).
If groundwater is subject to surface contamination,
continuous disinfection of water abstracted for potable use
is essential. Neglect of this critical requirement led to
the largest reported outbreak of typhoid fever (210 cases)
to occur in the U.S. since 1939. This outbreak occurred at
a farm labor camp in Bade County, Florida (Pfeiffer, 1973).
Engineering evaluations revealed that chlorination of the
wells (known to have had a history of intermittent con-
tamination) was interrupted and consequently the water
supply became a vehicle of infection.
Distribution pipe slimes, rich in iron, manganese and
organic materials, support the growth of Salmonella and E.
coli (Woratz and Bosse, 1968). Researchers, investigating
an outbreak of 13. bareilly in a maternity hospital, isolated
the organism from one-third of the water taps examined; and
from all of the water storage tanks (Mendis, et a_l. , 1976).
Upon further examination, S_. bareilly was found growing on
the interior surfaces of water taps and on the walls of the
water storage tanks. After the tanks and lines were cleaned
and disinfected, no further evidence of Salmonella was found
in the water system.
-------
68
Salmonella occurs and can be anticipated in polluted
water, but other organisms present during sampling can
interfere with Salmonella detection. There is no single
method that will ensure recovery of all Salmonella serotypes
present in a given sample, so a combination of methods is
necessary. Negative results by any method chosen do not
necessarily imply the absence of all salmonellae nor the
absence of other pathogens. There are presently no stan-
dardized _ methods for isolating Salmonella from water, although
the Committee of the European Economic Community, the World
Health Organization, and the International Standards Organi-
zation are working to establish standardized procedures.
The following, therefore, presents summaries of procedures
that may need modification to fit a particular set of cir-
cumstances .
Since Salmonella ordinarily occur in lower numbers than
expected for sanitary indicator bacteria, it must initially
be concentrated from large volumes of water samples. A
qualitative analysis may be performed by submerging ,gauze swabs
in flowing water, thus exposing the swabs to a large volume
of water (American Public Health Assoc., 1976). After three
to five days of submergence at the sampling site, the swab
is retrieved, placed in a sterile plastic bag inside a
container with ice, and sent to the laboratory for processing
within 6 h after sample collection. The swab is not a
perfect entrapment device since some salmonellae may pass
through, others may desorb from the swab during exposure
period, and the volume of water in contact with the swab is
unknown. Furthermore, since the swab is submerged for several
days, what results is a composite which does not reflect
changes in or cycling of Salmonella densities at the sampling
site.
The diatomaceous earth procedure often produces better
results than the swab method when floating solids are present,
such as in sugar beet effluents, paper mill wastes, and
waters containing massive algal (and. cyanobacterial) blooms.
In this procedure, the water sample is filtered through a
plug of diatomaceous earth to concentrate the organisms
(Brezenski and Russomanno, 1969). since known volumes of
water or wastewater are filtered through the diatomaceous
earth, the procedure provides a quantitative approach to
salmonellae detection.
A_membrane filter may be used to concentrate organisms,
including salmonellae, from a water sample provided sample
turbidity does not clog membrane pores and prevent filtra-
tion. Samples taken from suspect potable water supplies
(including wells) can be anywhere from 100 ml to several
-------
69
liters, depending on the concentration and characteristics
of the suspended solids.
Using a cartridge filter permits 20 1 or more of sample
to be filtered and is particularly applicable to waters that
have low concentrations of organisms (Levin, e_t al_. , 1974).
Water is drawn (under negative pressure) through a filter of
borosilicate glass microfibers bonded with epoxy resin.
After filtration, the filter cartridge is separated asep-
tically from the holder and placed in the enrichment broth.
If high numbers of background organisms are present, this
technique will not easily recover Salmonella. As with other
filtration procedures, sample turbidity will slow the filtra-
tion rate. Pre-enrichment for Salmonella can be accom-
plished with buffered peptone water. Selective enrichment
for Salmonella after concentration requires selenite, Rappa-
port, or tetrathionate broth. Selenite enrichment broth may
be combined with dulcitol to improve selectivity for salmon-
ellae. However, dulcitol selenite may retard the recovery
of S. typhi, S. cholerae-suis, S_. enteritidis biotype para-
typhi A, and S. pullorum because these species ferment
dulcitol slowly (Raj, 1966). Mannitol selenite broth has
been recommended for isolation of S_. typhi.
Tetrathionate and selenite broths may be modified by
the addition of brilliant green dye to enhance selectivity
for salmonellae other than S_. typhi. However, tetrathionate
broth is reportedly toxic to salmonellae at a temperature of
43°C (McCoy, 1962).
Salmonella strains have been further selected and
separated from other bacteria in polluted water samples with
a variety of enrichment media at various incubation tempera-
tures. Because other bacteria in the sample may interfere
with Salmonella detection, temperatures above 37°C have been
used to inhibit the growth of background microorganisms. In
general, most researchers agree that 41 to 42°C'(preferably
41.5°C) is an optimum incubation temperature for recovery of
the largest number of Salmonella in the enrichment pro-
cedure. However, since some salmonellae grow more vig-
ourously at 35 to 37°C, parallel enrichment cultures at
these temperatures may be performed when feasible; also,
when isolating, purifying, or cultivating salmonellae for
biochemical testing, a 35 to 37°C incubation temperature is
recommended (American Public Health Assoc., 1976).
The fluorescent antibody (FA) technique can be used to
screen rapidly for cultures taken directly from prior enrich-
ment (American Public Health Assoc., 1976). This technique
-------
70
requires careful interpretation of fluorescence; any posi-
tive FA results must be confirmed by the conventional bio-
chemical and serotyping techniques.
Pure cultures of Salmonella may be isolated from the
enrichment broths by streaking every 24 h for three days
onto the surface of selected differential plating media.
These media are: xylose lysine desoxycholate (XLD) agar,
brilliant green (BG) agar, xylose lysine brilliant green
(XLBG) agar, and bismuth sulfite medium. Because bacteria
other than salmonellae may grow and possibly interfere with
isolation and differentiation of suspect Salmonella strains,
using brilliant green agar at an elevated temperature of
41.5°C will reduce the number of these interfering or-
ganisms. However, the elevated temperature and this medium
may inhibit development of some Salmonella serotypes. After
incubation, the cultures are examined for typical colonies
of salmonellae which are then characterized biochemically
and serologically.
Biochemical reactions are used to characterize the
suspect Salmonella isolates recovered and permit a separa-
tion from closely related bacteria. Four major Salmonella
sub-genera can be differentiated by selecting tests to obtain
reactions for dulcitol, lactose,^galactosidase, d-tar-
trate, mucate, malonate, gelatin, and KCN. Ordinarily,
salmonellae do not ferment lactose, sucrose, malonate, and
salicin, but do ferment glucose, inositol, and dulcitol.
These tests, using pure culture isolates, can be made
in single tube media, or in commercial multitest systems.
The multitest systems permit the examination of large num-
bers of isolates in a relatively short time. All ten-
tatively identified Salmonella isolates are then submitted to
serological confirmation.
^To prepare for serological verification of Salmonella
strains, the pure culture isolate is transferred to a brain
heart infusion agar slant and incubated for 18 to 24 h to
insure maximum strain vigor. A range of polyvalent and
individual somatic or flagellar antisera should be used in
agglutination tests for determining Salmonella serotypes.
Especially when testing for Salmonella, the dedication
and skill with which the microbiologist conducts his ex-
aminations, and his critical review of culture reactions are
what determine the validity of the results.
-------
71
(ii) Shigella. This genus is composed of a group of
closely related gram-negative bacteria that cause mild to
acute bacillary dysentery in man. These bacteria have many
biochemical reactions in common with other members of the
family Enterobacteriaceae to which they belong. In addi-
tion, they are non-motile, do not normally ferment lactose,
yield positive methyl-red and negative Voges-Proskauer
reactions, and do not grow on citrate medium (Wilson and
Miles, 1975).
The Shigella genus is divided into four main subgroups
based on a combination of biochemical and serological char-
acteristics: Subgroup A (Shigella dysenteriae several
serotypes), Subgroup B (53. flexneri several serotypes),
Subgroup C (S. boydi several serotypes), and Subgroup D
(S. sonnei one species or serotype). The shigellae are
not particularly heat resistant and are killed by temperatures
of 55°C in one hour. They are even more sensitive to higher
temperatures and they will die within several hours when in
moist stool specimens that have become acidic from bacterial
growth. However, Shigella can surviv.e for many days in
clean, cool water.
All Shigella species cause bacillary dysentery (also
called shigellosis) exclusively in humans and some primates.
Infection normally is restricted to the intestinal tract and
the incubation period is only 48 h. The organism enters the
small bowel, multiplies, then proceeds to the terminal xleum
and colon where it penetrates the epithelial cells and
multiplies again. The result is inflammation, sloughing of
cells, and ulceration. Dysentery is a disease of the large
bowel and the mesenteric lymph nodes. Symptoms include
fever and diarrhea characterized by watery feces tinged
with blood, mucus, and groups of polymorphonuclear leuko-
cytes. The case-fatality rate varies widely according to
the age and general health of infected individuals and also
the Shigella species involved. S_. dysenteriae is the most
dangerous, possibly because it produces a particularly
cytotoxic exoenterotoxin not known to be produced by any of
the other species (Hollister, et a_l., 1955).
Infection is transmitted by the fecal-oral route. The
organism may be present in enormous numbers in the feces of
clinically ill persons, but it can also be spread by asymp-
tomatic carriers as well as convalescents. Shigella h,as
been isolated from clothing, toilet seats, and contaminated
food; and flies are known to carry and Spread infective
material. Infection with Shigella occurs endemically in
most communities and may be maintained by a few symptomless
carriers in the absence of clinical cases.
-------
72
Bacillary dysentery occurs most commonly during summer
months in tropical countries where lower standards of hy-
giene increase opportunities for transmission by many routes.
Many different serotypes generally are present in these
areas with no single one playing a much greater role than
any of the others. Water supplies in tropical countries
tend to receive shigellae that are water-washed from many
sources, whereas in developed countries, transmission by
water^is most often the result of contamination from one
identifiable source. Stricter sanitation measures, proper
sewage disposal, and public health standards enforced in the
developed countries have led to a shift in peak incidence
from summer to winter, as low temperatures favor survival
of Shigella. One species, S_. sonnei, predominates, probably
because it can survive longer under adverse conditions
(Feachem, 1977).
In the U.S., waterborne outbreaks of shigellosis have
increased somewhat from 1961 to 1977, although' shigellae are
not often isolated from water and the organism poses no
unique problems to the protection of water supplies. The
majority of outbreaks were from inadequately maintained and
monitored semi-public systems, especially those serving
recreational areas that are in use for only part of the
year. Fecal contamination and improper disinfection of a
well, situated approximately 46 m (150 feet) from a septic
tank, were blamed for an outbreak of shigellosis in Florida
(U.S.) involving 1,200 people (reviewed by Craun, 1977).
Thermo-tolerant coliforms were isolated from drinking water
after 690 people became ill with shigellosis during passage
aboard a cruise ship sailing from the U.S. The water was
found to have been stored improperly and not given suffi-
cient contact time with the chlorine disinfectant.
(Reviewed by Hughes and Merson, 1975). An outbreak in-
volving hundreds of cases occurred in 1977, in Germany, when
a reservoir was contaminated with waste from a nearby water-
flush toilet (Bohm, et aJL. , 1978). These incidences serve
to underscore that the lesson still to be learned in de-
veloped countries is that of increased vigilance, especially
in connection with semi-public water supplies.
As with Salmonella, no standardized procedures have
been established for isolating Shigella from water. Ten-
tative methods, subject to modification, are offered in
Standard Methods (American Public Health Assoc., 1975) and
elsewhere. Concentration techniques described in Standard
Methods are the same as those for Salmonella [See Section
III.B.l.a(i)]. GN broth affords good enrichment for
-------
73
Shigella while suppressing coliforms and fecal streptococci.
The formulation includes dextrose and d-mannitol which are
balanced to limit the growth of Proteus and encourage that
of enteric pathogens. High concentrations of sodium citrate
and sodium desoxycholate inhibit gram-positive organisms,
and the latter ingredient is far less toxic to Shigella than
brilliant green dye which is unacceptable for use with
Shigella. Xylose lysine desoxycholate agar is a differential media
suggested by Standard Methods for isolation of Shigella species
when used in conjunction with GN enrichment broth. Finally,
suspect colonies can be purified and subjected to various
biochemical and serological procedures.
(iii) Yersinia Enterocolitica. Evidence of infections
due to YersinTa enterocolitica has been mounting since the
early 1960's, especially from the world's cold or temperate
regions. Illness caused by this organism has been reported
in recent years from North America, Japan, South Africa, and
almost every European country. Some cases also have been
noted in South America, North Africa, and the Middle East;
and the disease probably would be found common in many other
parts of the world if there were more laboratories in these
areas capable of detecting the causative organism (Alonso,
et al., 1976; Bottone, 1977; Toma and LaFleur, 1974).
(iii.l) Characterization of Y. Enterocolitica. The
organisms comprising the genus Yersinia (Y. pestis, Y.^
pseudotuberculosis, and Y. enterocolitica), of the family
Enterobacteriaceae, were formerly considered members of the
genus Pasteurella. Although no consensus on the precise
definition of Y. enterocolitica has been reached, it is
described in general terms as a gram-negative, facultatively
anaerobic, motile (below 30°C) coccobacillus. Mollaret and
Thai (Buchanan and Gibbons, 1974) provide some biochemical
differentiations among the three species. Y. entero-
colitica has been divided into many serobio-, and phage-
types; and these have all been classified into three or four
groups based on DNA homology (Brenner, et al., 1976). How-
ever, it is more likely that this wide spectrum of types
eventually will be separated into different, more precisely-
defined species. The most noteworthy characteristic of this
bacterium is its stability and even growth down to tempera-
tures as low as 4°C (optimum, 20°C) [See Section B.S.c].
Yersiniosis, caused by both Y. enterocolitica and Y. pseu-
dotuberculosis, is thought to be contracted perorally; its
appendicitis, mesenteric lymphadenitis, or acute terminal
-------
74
ileitis. Y. enterocolitica has been isolated from lymph
nodes and feces of both sick and healthy humans along with a
growing number of animal species. Though there are human
and animal carriers, their role in the disease's trans-
mission is as yet undetermined.
(iii.2) Sources and Occurrence of Y. Enterocolitica.
Broadly speaking, the organisms that are classified Y.
enterocolitica fall into two main groups (Mollaret, T976).
The first group causes typical illnesses and consists of
host-specific strains with constant and stable bio-, sero-,
and phage types. Predominant types within this group may
differ from place to place suggesting that incidences of the
disease in different areas of the world may have no epi-
demiological connection. For example, in Japan, Canada, and
most European and African countries, serotype 0:3 predomi-
nates. However, strains of this serotype belong to phage
type 8 in Europe and Japan; whereas in South Africa, these
strains are classified as phage type 8 or 9a; and phage type
9b is regarded as the "specific Canadian" type elaborated by
these strains (reviewed by Toma and LaFleur, 1974) . The
serotype 0:9 predominates in Finland and serotype 0:8, which
predominates in the U.S., does not occur at all in Europe.
Sources of the infection have remained largely unknown.
The human serotype 0:3 has been found to occur in animals
(including pigs and household pets) suggesting a zoonotic
epidemiology; however, no clear connection between human and
animal infections has ever been established. In fact, there
is reason to believe that Y. enterocolitica infections occur
independently in humans and animals as a result of exposure
to an environmental source common to both (Mollaret, 1976).
Little has been written about the second group whose
occurrence and significance in different ecosystems is not
well understood. Strains of this group have been isolated
from small mammals (Kapperud, 1977), poikilotherms (Kapperud
and Jonsson, 1976), water (Harvey, e_t al. , 1976), and food
(Morris and Feeley, 1976). They do not appear to require a
specific host and they vary widely in biochemical and anti-
genie characteristics. Humans and animals can carry them
with no clinical symptoms (Mollaret, 1976).
Different Y. enterocolitica types have been isolated so
frequently from untreated surface water in some areas that
they most likely represent part of the normal microbial
flora of the water and surrounding terrestrial environments
-------
75
(Kapperud, 1977). Alternately, the same Y. enterocolitica
biotype and/or serotype is readily recoverable in these_
areas from a wide range of animal species (including poi-
kilotherms) indicating a nonspecific role in these eco-^
systems. Most of these organisms belong to the rather ill-
defined second group and generally do not appear to act as
pathogens. Although members of this group also have been
isolated occasionally from people with diarrhea (Kapperud,
1977; Toma, 1973), their presence is probably only inci-
dental .
(iii.3) Epidemiology of Waterborne Y. Enterocolitica
Infections. Reports referring to waterborne Y. entero-
colitica infections are few in number. Lassen (1972) de-
scribed a case of violent gastroenteritis caused by Y. en^
terocolitica serotype 0:13-7 in an 18 year-old woman. This
same sero- and biotype had been isolated only a few days
earlier from the well next to her house. Keet (1974) re-
ported a septicemia in a 75 year old man from Y. entero-
colitica serotype 0:8, in which the source of infection was
later traced to a mountain stream. He postulated that wild
animals inhabiting a watershed may transmit the disease
indirectly through the water to farm or domestic animals.
In December 1975, waterborne gastroenteritis occurred in
of approximately 1,500 guests at a ski resort in Montana
(U.S.). Although Y. enterocolitica was isolated from the
unchlorinated well~water, what, ±£ any, role it played in
the outbreak remains unclear (Center for Disease Control,
1975a).
It has been proposed that, to infect humans, Y. entero-
colitica may require an intermediate period of multipli-
cation at reduced temperatures (after passage through a
vector) wherein the organism can reach infective levels.
Such a "hot-cold" mode of transmission could account for the
observed increase in cases of yersiniosis in colder climates
and seasons (Bottone, 1977). Moreover, Y. enterocolitica
grown at 37°C is less resistant to normal cellular bacteri-
cidal defenses than when grown at 20°C (Nilehn, 1973); this
could explain why infection through direct person-to-person
contact is relatively rare, and again, why an intermediate
cold phase could be critical to its spread.. If so, the life
cycle for Y. enterocolitica would stand in sharp contrast to
that of others in the family Enterobacteriaceae whose abil-
ity to be transmitted through water represents only an
exceptional phenomenon.
There is no standard method for the isolation and
enumeration of Y. enterocolitica in water. This micro-
organism is similar to other enteric bacteria except that it
grows better at 25°C than at 36°C, and it is able to grow at
-------
76
4°C. For these reasons, a procedure analogous to those for
food examination has been proposed. This method described
recently by Highsmith and coworkers (1977b) includes:
membrane filtration; liquid enrichment in M-Endo broth at
25°C and 36°C for 72 h and plating on Salmonella-Shigella
(SS) or MacConkey agar at 4°C for 14 to 21 days; isolation
on the same plating media or enrichment broth and incubation
at 25°C; biochemical identification of the lactose-negative
colonies on Triple Sugar Iron agar for urea and motility
determinations before further identification and typing.
(iv) Enteropathogenic Escherichia Coli. Transmission
of enteropathogenic E. coli through drinking water was
frequently reported during the 1950's (Monnet, et al., 1954;
Seigneurin, et a^L., 1955; and Lanyi, e_t al. , 19~59)~
Certain enteropathogenic strains of 13. coli are now
known to cause acute diarrhea, especially in infants, in
travelers to foreign countries, and in consumers of con-
taminated foods. These organisms include 14 distinct an-
tigenic types, in addition to other recently implicated
strains, some of which are not serologically typable.
_ Disease may occur by either of (at least) two mech-
anisms: tissue invasion or toxin production. When cells
penetrate the intestinal epithelium, the result is a syn-
drome like the bacillary dysentery caused by virulent Shigella
strains. The second mechanism entails the production oi
potent heat-labile or heat-stable enterotoxins which induce
cholera-like symptoms or a salmonellosis-like enteritis. If
the organism colonizes the upper bowel it will manifest
enterotoxigenicity; but, it may be carried in the colon
without producing any harmful effects. However, because
toxin production is a plasmid-borne trait, organisms of-a
given serotype may or may not be toxigenic.
Of the 99 waterborne outbreaks of gastroenteritis
reported in the U.S. from 1971 to 1974, no causal agent was
identified in almost half of these incidents. The extent to
which enteropathogenic E. coli was involved in these out-
breaks is largely unknown, primarily because serotyping of
isolates is not routinely performed during outbreaks.
Two of the best documented outbreaks took place in
Sweden in the autumn of 1965, one in a residential area of
Uppsala, and the other in the small community of Gimo situ-
ated 50 kilometers north of Uppsala (Danielsson, et al.,
1968).
-------
77
Within a period of 14 days, 442 residents of Uppsala
fell ill; 261 of them experienced the onset of symptoms
during a two day peak. High coliform densities were dis-
covered in the community's water supply. A rapid (2 to 8 h)
preliminary screening technique, based on membrane filtra-
tion combined with fluorescent antibody staining (Danielsson,
et al., 1968), was applied to water samples and identified
thYee serotypes of enteropathogenic _E. coli. Freshly col-
lected water samples and enriched water samples (incubated
at 37°C for 4, 6, and 14 h in nutrient broth) were filtered
or centrifuged, and filter sections or sediment smears were
stained with fluorescent antibody. In conjunction with
this, fecal specimens from sick patients were serotyped
according to conventional culture and serological procedures.
Both water samples and fecal specimens were found to contain
the same three serotypes 026:B6, 0128:B12, and 0111:B4
the latter serotype predominated in both fecal and water
sources.
Only a few cases of gastroenteritis were reported for
the community of Gimo, although the drinking water supply
was also found to be highly contaminated with coliforms
after a spell of heavy rains. Serotypes 026:B6, 0114:B?,
and 0125:B15 were identified from samples subjected to
Danielsson's membrane filtration and fluorescent staining
technique. However, since no clinical specimens were col-
lected for bacteriological examination, etiology could not
be confirmed. Danielsson and coworkers (1968) cautioned
that their rapid screening technique could yield false
positive results due to serological cross reactions with
other bacteria, and they advised that conventional methods be
used for confirmation.
A large waterborne outbreak of illness caused by entero-
toxigenic JE. coli occurred in June and July of 1975, at
Crater Lake National Park, U.S.A. (Rosenberg, et al., 1977).
More than 200 staff members and 2,000 visitors to the park
experienced prolonged diarrhea, cramps, nausea, and vomiting.
Telephone surveys and questionnaires strongly associated the
illness with ingestion of the park's drinking water, sup-
plied by a shallow spring. The water supply, although
chlorinated before distribution, was not routinely monitored
for chlorine residuals once it had entered the distribution
system. Later tests, using fluorescein dye, revealed that
raw sewage from an overflowing manhole was contaminating the
water.
Lactose-positive and -negative colonies, picked from
cultured stool specimens, were tested for toxin production
-------
78
using mouse adrenal tumor cell, Chinese hamster ovary tissue
culture, and infant mouse assays (Rosenberg, e_t al. , 1977).
Of 120 patients examined, enterotoxigenic E. co"lT~(exclu-
sively serotype 06:H16) was isolated from 20 of them. The
organism produced both heat-labile and heat-stable toxins.
No Salmonella, Shigella, pathogenic vibrios, or Yersinia
enterocolitica were recovered from either the spring water
or ±rom clinical specimens.
These investigators pointed out the relative insen-
sitivity of current culture methods. They explained that
the lack of recovery of enterotoxigenic E. coli from many
fecal specimens was, in all probability, due to the patho-
gen's short duration (averaging less than four days) after
diarrhea had ceased. Also criticized was the practice of
testing only five to ten E_. coli isolates randomly chosen
from nonselective media. The serum antitoxin immunity test
was held to work well only in areas where a particular
pathogen, such as cholera, is endemic and/or introduced at
high antigenic levels. This was not the case with the
Crater^Lake outbreak. The authors recommended intensive
screening of a substantial portion of diarrheal fecal spec-
imens for presence of the enteropathogenic agent. Once a
predominant serotype of E_. coli was shown to be entero-
toxigenic by criteria listed earlier, the remainder of fecal
specimens would then be tested serologically for the presence
of that strain.
The definition of the coliform group includes E. coli,
so enteropathogenic E. coli is unlikely to be present Tn
water in which coliforms are undetectable; this relationship
is far more direct than is usual between indicators and
pathogens. However, even if E_. coli is shown to be present,
its^nteropathogenicity can be determined only by highly
refined techniques. The presence of enteropathogenic E.
coli is not a great deal easier to monitor than that of"
several pathogens, including viruses, which may cause water-
borne outbreaks of acute diarrhea.
(v) Francisella Tularensis. Tularemia is a zoonotic
disease, transmitted to humans from blood-sucking arthropods,
domestic animals, and primarily by a number of wild animal
species, many of which lead semi-aquatic lives. It is this
semi-aquatic existence of such susceptible hosts as voles,
beavers, muskrats, and water voles that accounts for most
waterborne outbreaks of the disease in humans. Tularemia is
endemic on three, possibly four, continents and is ubiquitous
throughout the northern hemisphere, including North .America,
-------
79
Europe (except the British Isles, Spain, and Portugal), the
U.S.S.R., Turkey, and Japan (Hopla, 1974).
The causative agent, Francisella tularensis, is de-
scribed as a tiny gram-negative, aerobic, singly occurring,
non-motile, nonencapsulated, nonsporulating, bipolar rod
requiring special media to which glucose and cysteine or
cystine have been added. Historically, it has occupied
several generic designations these being Bacterium,
Bacillus, Brucella, and Pasteurella. But, the organism was
finally placed in a genus by itself, Francisella, based on
its small size and because it does not grow on simple nutrient
media. Nonetheless, investigators in the U.K. generally
refer to this pathogen as Brucella tularensis because of
cross-reactions with agglutinating antigens of the genus
Brucella. P. tularensis is killed by heating for 30 minutes
at 50°C, but" it can survive over three weeks at -14°C or
after desiccation. ,
The disease occurs as an anginose-bubonic infection
when contracted through ingestion (as opposed to inhalation
or skin contact with animals), producing buboes and areas of
necrosis in organs and tissues of man and animals.
It was first suggested that tularemia was transmissible
through water in 1935 (Pollitzer, 1967). The first reported
outbreak of waterborne tularemia, in which 45 people, all ot
whom drank from a small stream where F_. tularensis was later
isolated, came down with the disease was documented in 1936.
Lesions observed on the mucosa of the tonsils and the oral
cavity implicated water as the vehicle.
Most references to waterborne tularemia are integrated
into a very extensive body of tularemia literature contributed
by Russian workers in response to a series of explosive
tularemia outbreaks which ravaged the U.S.S.R. in 1928 and
over the next 30 years. Several investigators have directed
their studies to the westward penetration of tularemia in
Eastern Europe during and immediately after the Second World
War. The disease had calamitous effects on the Soviet army
beginning in 1941 and was traced to drinking water contaminated
with rodent carcasses and excrement as a vehicle of infection.
Outbreaks of waterborne tularemia have been reported
throughout the U.S.S.R. Leningrad was struck with about 200
cases in 1944 - 1945 of which 20 percent were shown to be
waterborne (Pollitzer, 1967). Drinking water contaminated
with infected rodent carcasses accounted for 22.8 percent of
tularemia cases in White Russia (Pollitzer, 1967). Three
river cities in Central and Eastern Siberia experienced
-------
80
tularemia epidemics in the 1940's in which contaminated
drinking water played an important role second only to that
of insect and tick bites. It appeared that a peak in the
epidemic was reached during the warmer months of June to
September. Cases of waterborne tularemia have been reported
for the southern part of the U.S.S.R., notably Armenia and
Kazakhstan, where the anginosebubonic form of the disease
was greatly in evidence. Investigation of outbreaks in
Kazakhstan placed the blame on direct or indirect contact
with water rats. The Altai Mountain region was also hit
with tularemia in 1955 through 1957, reportedly from drink-
ing water contamination (Pollitzer, 1967).
Recent cases of waterborne tularemia in northern Norway
have brought about renewed interest in water as a vehicle of
infection in the Scandinavian peninsula. Mair (1973) studied
a series of what he termed otolaryngological manifestations
of the disease involving hunters who drank from a mountain
brook during a period of peak rodent (notably lemmings)
populations. Short-term contamination of rural streams is
the more common, well documented source of waterborne tula-
remia outbreaks, and is attributed to epizootics among semi-
aquatic wildlife, mainly voles, beavers, muskrats, and water
voles.
There is no standardized method for the examination of
water for IT. tularensis. The genus is characterized by
growth requirements for cysteine or cystine; no growth is
obtained on ordinary culture media. Semi-solid media which
allow for slow growth, such as gelatinized egg yolk and
media containing cysteine, glucose, and defibrinated rabbit
blood (or serum) can be used. Identification is accom-
plished by means of biochemical tests.
(vi) Leptospira. The genus Leptospira is composed of
finely coiled, spiral organisms including a-number of
strains that cause leptospirosis in humans and many animals.
The organisms are of slender appearance, have numerous coils
with bent or hooked ends, and are very actively motile.
They do not stain well in conventional stains, but require
special preparations.
^At present, all leptospirae are grouped within one
species, L. interrogans. L. interrogans is divided into two
complexes: 1) interrogans, all strains that are pathogenic
or^parasitic; and 2) biflexa, free-living or water lepto-
spirae that commonly occur in fresh surface waters. The
-------
81
interrogates complex contains 18 serogroups and about 150
serotypes. The most important serogroups are: Ictero-
haemorrhagiae, Canicola, Hebdomadis, Pomona, Grippotyphosa,
Autumnalis, Australis, Mitis, and Bataviae. More detailed
studies are available (Fuchs and Burger, 1974). It was
proposed in 1973 by the Leptospira subcommittee that these
complexes be designated species: L. jnterrogans and L.
biflexa.
Leptospirae can be grown on various media containing
peptone and inactivated serum. They are aerobic, their
optimal temperature is 28 to 30°C, and they grow well at pH
7.2 and 7.4. Reliable diagnostic methods are available, and
most laboratories are equipped to distinguish leptospirosis
from other diseases as well as to identify the specific
serogroup responsible.
The leptospirae penetrate the surface epithelium of the
host and enter the bloodstream, where they multiply and are
carried to all organs of the body (hence, the diversity of
symptoms). They finally settle in the convoluted tubules of
the kidney and from there are excreted in the urine.
Incubation time of the disease is usually from seven to 13
days. Symptoms may range from fever, headache, chills,
malaise, vomiting, and muscular aches to very severe and
even fatal illness including meningitis, encephalitis,
haemorrhagic states, and infrequently, jaundice. Symptoms
formerly ascribed to Weil's disease may be caused by all the
pathogenic strains, depending on the virulence of the strain
and on the reaction of the host. Severe cases may have
fatality rates of as high as 50 percent (Turner, 1973).
The leptospirae are not particularly heat-resistant
(Chang , e_t al. , 1948) and are killed in 10 min at a tempera-
ture of 5U°"C. They are sensitive to desiccation, acid (such
as gastric juice), and to disinfectants; and they are de-
stroyed by bile. They do not tolerate high salt levels and
are soon eliminated in polluted water or sewage. They can,
however, withstand freezing for some time.
Leptospirosis is essentially a zoonosis that can occur
in great numbers of domestic and wild animals depending on
climate and availability of a food supply. The organism is
maintained in the environment by both carrier and diseased
hosts, and transmission is particularly favored by popu-
lation explosions of animal carriers, especially rats. In
temperate regions, animal hosts most likely to transmit
disease to man are rats, dogs, and swine. Human exposure to
-------
82
the disease has been from direct or indirect contact with
urine from infected animals (Kathe, 1945; Muller, 1969a).
Humans are considered to be a dead-end host, as person-to-
person spread is rare. Mild anicteric forms are now recog-
nized as more common than clinically severe icteric forms of
the disease; consequently, human infection with pathogenic
leptospirae may pass completely unnoticed.
In the U.S., the most likely sources of infection with
Leptospira are surface water, dogs, rodents, cattle, and
swine (Reasoner, 1977). However, infection by the water
route is mainly the result of swimming in polluted ponds or
creeks; or, in the case of miners, of walking through stag-
nant water infested with rodents. Warm (/^22°C), slow
moving streams of neutral or slightly alkaline pH provide an
ideal environment that can prolong Leptospira survival for
many weeks. Infection with Leptospira is primarily associated
with certain occupations such as mining, dairy farming, and
sanitary engineering. Only rarely is it associated with
drinking water (Heath, et al., 1965), as in times of disaster
such as during wars, when tEere has been a breakdown in
sanitation and public health systems. In post-war Berlin,
the incidence of the infection was very high (Ranker
1969). ttsecKer,
The Detection of Leptospira is made difficult by the
competitive growth of other organisms and the need to differ-
entiate between pathogenic and saprophytic strains. There
is no standard method for isolating Leptospira from environ-
mental sources, but a tentative procedure entails analysis
of bottom sediments by concentration, enrichment in Fletcher's
semi-solid medium containing 10 percent rabbit serum, and
incubation at 30°C for six weeks. During incubation, the
medium should be checked repeatedly for the appearance of
leptospirae and for any culture contamination. Upon detection,
the Leptospira isolates must be further characterized by
various biochemical and serological tests (American Public
Health Assoc., 1975).
(vii) Vibrio Cholerae. The genus Vibrio comprises a
large number of species, only a few of which are of medical
importance: V. cholerae and its biotype V. cholerae El Tor
cause cholera exclusively in man; V. paraEaemolyticus (a
marine organism which typically inEabits estuaries); and
nonagglutinable (NAG) vibrios can cause cholera-like disease
or mild diarrhea in humans (Heiberg, 1935), but are not
normally transmitted by the water route. Only under excep-
tional circumstances (e.g., contamination with surface water
-------
83
due to leakage or cross-connections) can NAG vibrios be
transmitted through drinking5water. However, numbers suf-
ficient to cause illness (10 ) are obtained only after
subsequent enrichment in food (Muller, 1977a).
The genus is composed of polymorphic, gram-negative,
nonsporeforming rods that are slightly curved, actively
motile, and that possess a single thick polar flagellum.
They are not fastidious and grow readily on a number of
simple media which need only contain a balanced mixture of
utilizable carbohydrate, inorganic nitrogen, sulfur and
phosphorous sources, simple minerals, and an adequate
buffer. Vibrios grow best between 30 and 40°C and at a pH
near neutrality. They are classified according to their
biochemical reactions in tests that include slide hemag-
glutination, Voges-Proskauer test, polymyxin-B sensitivity,
and resistance or susceptibility to Mukerjee's group IV
cholera phage (Finkelstein, 1973; Wilson and Miles, 1975).
Two major serotypes, Ogawa and Inaba, and one rarely en-
countered serotype, Hikojima, are recognized in both V.
cholerae and the El Tor biotype.
The vibrios are sensitive to heat, sunlight, and drying;
and they are especially intolerant of acid but are able to
withstand alkalinity up to pH 10.2. They survive well at
low temperatures and can persist for weeks in cool, pure
drinking water. The El Tor biotype seems hardier than the
classical Vibrio biotypes and can survive longer in nature
(Finkelstein, 1973).
Cholera, whether caused by classical V. cholerae bio-
types or the El Tor biotype, produces the same clinical
symptoms of profuse diarrhea, leading to rapid loss of fluid
together with loss of bicarbonate, sodium, potassium, and ^
chloride ions. In severe cases, this can be accompanied by
a drop in body temperature and blood pressure, followed by
prostration, kidney failure, and death. Untreated case
fatalities are over 60 percent, but death always can be
averted by early medical intervention. The time of incuba-
tion is anywhere from a few hours to six days. Infection is
confined to the gastrointestinal tract, and formation of
enterotoxin is what triggers host response. Diagnostic
procedures are standardized throughout the world, and labora-
tories can easily identify the disease if alerted to the
possibility of cholera and suitably equipped.
Fecally contaminated water is the primary vehicle of
cholera transmission, although vibrios are also spread by a
-------
multitude of other routes including food, soiled clothing,
flies, and direct person-to-person contact. Extremely high
numbers of cholera vibrios can be found in human feces and
vomitus. The disease has a high rate of secondary spread
when there is contact with sick individuals who shed the
organism in enormous numbers, particularly during the first
days of illness. Generally, the number of inapparent infec-
tions or mild cases greatly exceeds that of severe cases by
a factor of ten up to several hundred. Major determinants
of this ratio are levels of hygiene and nutrition, occurrence
of other enteric infections (which can permit vibrios to
pass unharmed through the gastric juices), and quantity of
vibrios ingested [See Section B.4]. The carrier status is
presumed to be transitory; but, especially in patients over
50 years of age, prolonged infection in the gall bladder
has caused chronic and intermittent shedding (Zwadyk, 1976).
Cholera is still one of the most severe and contagious
of infections, having spread to over 171,000 people in 42
countries since 1961. In 1971, it became clear that the El
Tor biotype was responsible for this seventh recorded cholera
pandemic. The agent has been widely disseminated due to a
high frequency of unrecognized El Tor infections. The
disease has characteristically been sporadic and endemic in
areas of poor hygiene and warm humid climates (Barua, 1970;
Feachem, 1977; Felsenfeld, 1967; and Mosley, 1970). However,
isolated cases have been reported in Czechoslovakia, Spain,
Prance, Portugal, Australia, New Zealand, Sweden, the U.K.,
and the U.S., mainly from importation through tourism
(Finkelstein, 1973).
Crabs caught in a Louisiana (U.S.) lake caused six
known cases of cholera in 1978 (Center for Disease Control,
1978c). Water sampled from the lake and from sewage of a
nearby town showed presence of the Inaba serotype as did
clinical specimens from patients. Except for these and a
single case in Texas, the U.S. has had no reported cases of
cholera since 1911.
During the last century, when standards of hygiene were
lower, Europe suffered from several devastating epidemics.
However, it is now well understood that the spread of cholera
can be arrested quickly and entirely by adequate sewage and
water treatment, together with prompt detection and treat-
ment of cases and carriers. Moreover, a healthy and well
nourished population is unlikely to succumb to infection
with V. cholerae, particularly that caused by the El Tor
biotype.
-------
85
Methods for isolating vibrios from water have not been
standardized. Sampling procedures have included the use of
membrane filters, gauze pads, and grab devices. Enrichment
in alkaline peptone water at pH 8.4 or in taurocholate
tellurite peptone water at pH 9.2 yielded good results after
incubation initially for 4 to 6 h at ambient temperatures
(A/25°C) and then for 18 h at 37°C (Colwell and Kaper,
1977). Enrichment cultures were streaked onto TCBS agar and
suspect colonies were screened by testing for presence of
oxidase. Further differentiation was achieved by biochemical
and serological testing.
(viii) Campylobacter. Campylobacter is a small, gram-
negative, curved or spiral-shaped rod possessing a long,
single polar flagellum at one or both ends of the cell. It
has a characteristic corkscrew type of motility, is micro-
aerophilic, and grows easily on a conventional culture
medium at 37°C in a special atmosphere of 5 percent oxygen,
10 percent carbon dioxide, and 85 percent nitrogen (Smibert,
1978). Included in this genus are pathogens for both humans
and animals as well as saprophytic forms normally o&curring
in the oral cavity, reproductive organs, and intestinal
tract.
The organism was initially observed as the agent of
infectious abortion in cattle and was named Vibrio fetus.
However, because its G + C content is much lower than that
of members of the genus Vibrio, and because it does not
ferment or oxidize carbohydrates, V. fetus has been reclas-
sified as Campylobacter fetus. Of the three species C.
fetus, C. sputorum, and C_. fecalis only C_. fetus is
pathogenic for man.
In addition to causing abortion and sterility in cattle
and sheep, C_. fetus ss. jejuni and intestinalis also may
cause gastroenteritis in humans, cattle, sheep, and swine.
Human infections with Campylobacter can lead to gastro-
intestinal disease characterized by abdominal pains, diarrhea,
malaise, headache, and fever lasting from one to four days.
All age groups and both sexes appear to be equally susceptible
to infection. The organism inhabits the genitourinary and
intestinal tracts and is excreted in the feces.
In several studies, Campylobacter was isolated from
stools of 4 to 8 percent of patients with diarrhea an
isolation rate comparable to those of the more commonly
recognized enteric pathogens such as Salmonella [See Section
-------
86
B.l.a(i)] and Shigella [See Section B.l.a(ii)] (Butzler, et
al., 1973; Skirrow, 1977; Steel and McDermotts, 1978; and
Pai, et al., 1979). Campylobacterial gastroenteritis with
diarrEeVEas been described in Rwanda (Africa) (De Mol and
Bosniansr 1978), Canada (Laboratory Centre for Disease Control,
1978), and the U.S. (Center for Disease Control, 1978).
Feces from infected humans, animals, and fowl may
contaminate surface waters destined to be used for drinking
water supplies. If such source waters were inadequately
treated, viable Campylobacter cells could theoretically gain
entry to finished water. The exact mode of transmission to
humans is unknown, but contaminated food and water probably
serve as major vehicles (Joklik and Willett, 1976).
The best evidence for waterborne transmission of this
organism to humans came from an outbreak in Bennington,
Vermont (U.S.) (Center for Disease Control, 1978b). During
two weeks in May 1978, 2,000 of the town's 10,000 residents
became ill with gastroenteritis. An epidemiologic investiga-
tion showed a strong association between gastroenteritis and
consumption of water from the town supply. C_. fetus ss.
jejuni was isolated from individuals with dia"rrhea by cultur-
ing fecal samples obtained from rectal swabs. In this
case, water samples taken from several areas of the town
showed no residual chlorine. The report went on to state
that investigators were seeking evidence for contamination
from wild and domestic animals within the watershed.
Prior to the Bennington outbreak, isolation of this
pathogen was occasionally successful only if blood from an
infected individual was cultured (King, 1962). Later, a
procedure for isolating Campylobacter from stools was de-
scribed by Butzler and coworkers (1973). Also, a selective
growth medium containing vancomycin, polymyxin B, and tri-
methoprim was described in England by Skirrow (1977). Both
media require incubation at 43°C and microaerophilic culture
conditions. Failure to isolate Campylobacter is more often
the result of improper atmospheric adjustment than an incor-
rect choice of growth media. The organism also has been
grown using Albimi Brucella broth (Pfizer Diagnostics,
Flushing, N.Y.), supplemented with 10 percent animal blood.
This formulation is used either as a semi-solid broth (0.16
percent agar) or a solid medium. Because it is microaero-
philic, Campylobacter will grow only within the upper first
few millimeters of a semi-solid growth medium.
At present, there are no standard methods for detecting
Campylobacter in water. When such methods become available
-------
87
for routine use, it will then be possible to determine the
significance of Campylobacter in water as it relates to
outbreaks of waterborne gastroenteritis.
(ix) Opportunistic Pathogens. Several other micro-
organisms present in water can cause disease infrequently,
and almost always under unusual circumstances, either in
abnormal hosts or in situations where the normal flora have
been supplanted. These organisms are called opportunistic
pathogens and are ubiquitous in nature, very resistant in
water, can grow with only a few nutritional requirements,
and are able to accept and transfer plasmids which carry the
determinants for resistance to antibiotics. Concentrations
of opportunistic pathogens found in drinking water are not
normally sufficient to lead to infection in a healthy consumer.
The endotoxins produced by these agents, however, pose
a significant health hazard to the hospitalized, immuno-
depressed patient who might suffer from septicemia, wound
infection, urinary infection, or be in an otherwise weakened
state. Also susceptible are post-operative patients,
individuals being treated with immunosuppressive or antineo-
plastic agents, cancer or leukemia patients, newborn babies
(especially when premature), and the very old and infirm.
The organisms are most frequently transmitted by contact
with material, water, disinfectants, cleansing fluids, or
via intravenous therapy; airborne spread occurs less frequently.
These modes, of transmission are to be distinguished from the
definite possibility of self-infection, whereby the host,
being in a weakened state, may become susceptible to the
growth of an organism that is already present in the body in
low numbers. Of all hospitalized patients, 5 to 18 percent
can be expected to have contracted an infection from opportun-
istic pathogens; the risk is higher in hospital wards using
more technical equipment, for example, in emergency care or
neurosurgery.
The dangers resulting from bacterial metabolites, such
as pyrogens, should not be overlooked; and problems caused
by these organisms may extend to households and to scientific
laboratories, where de-ionizer and soft water units have
been shown to contain gram-negative bacteria. Opportunistic
pathogens also have been found growing in humidifiers and
ventilation plants.
-------
88
Opportunistic pathogens whose presence in water can be
anticipated include the following bacteria:
- Pseudomonas species, especially P_. aeruginosa,
present special problems in stored water and
with medical procedures and equipment that use
water, such as dialysis units, water baths,
sprinkler heads, scrub sinks, and others (to be
discussed further at the end of this section).
- Aeromonas hydrophila has been found in feces
from healthy individuals and those suffering
from diarrhea, in domestic animal feces, and in
drinking water (Annapurna and Sanyal, 1977).
This bacterium has drawn increased attention
recently because of its ability to cause infec-
tion in man, animals, fish, and reptiles, and
because it produces enterotoxins. Antibiotic-
resistant species of Aeromonas, including A.
hydrophila, not commonly identified as agents
of human infection (but rather of certain species
of fish and frogs), have been transmitted to
humans via source water receiving wastewater and
have caused cases of acute cellulitis (Hanson,
ejb al^., 1977). Populations of this organism in
numbers as high as 10 to 10 cells per ml have
been found in aquatic environments which receive
wastewater.
- Edwardsiella tarda has been isolated from human
patients and is associated with gastroenteritis.
It is harbored in the intestines and can cause
septicemia (Koshi and Lalitha, 1976).
- Flavobacterium can cause meningitis, bactereniia,
or septicemia. It is composed of a variety of
species, some of which tolerate high levels of
chlorine, are resistant to antibiotics, and can
produce pyrogenic metabolites (Colwell, et al.,
1978).
- The genera Klebsiella [See Section III.C.2.d],
Enterobacter, and Seirratia of the tribe
Klebsiellae have been recovered from human feces
and water. They are able to grow in water, are
often resistant to antibiotics, and can cause
urinary infections or septicemia.
-------
89
- The genera Proteus and Providencia of the tribe
Proteae can infect the urinary tract and cause
bacteremia in debilitated patients. They are
found in the feces or urine of infected indi-
viduals.
- Citrobacter and Acinetobacter can cause Infections
of the digestive tract or produce bacteremia or
septicemia in debilitated persons.
Staphylococcus aureus, Bacillus cereus, and Clostridium
species are unable to multiply in water, but can cause
illness (notably diarrhea) after incubation and growth in
food to which contaminated water was added.
p. aeruginosa has received the greatest attention in its
role as all opportunistic pathogen and also was the first organism
to be described as such. Pseudomonas species (pseudomonads) are
gram-negative bacilli, straight or curved, and are motile with
polar flagella. They are strictly aerobic, but denitrifying
species do exist. They are chemoorganotrophs and some are
facultative chemolithotrophs [See also Section A].
P. aeruginosa is a bacillus of 1 to 3 urn in length and
0.4 uH in diameter, with a polar flagellum. It is ubiqui-
tous in nature and may actively reproduce impotable and
distilled water to populations as high as 10 per ml (Carson,
et al 1975). This bacterium has few requirements and can
EFr"ow~with only carbon as its energy source in a broad range
?f temperatures (4 to 42°C) and PH (5 to 8). P. aeruginosa
is ordinarily a strict aerobe but can grow under anaero-^
JioSis* Stilizing nitrates or arginine as electron acceptors.
The identification of P. aeruginosa includes determining its
production of pyocyanin on King's medium and some biochemical
characteristics; for example, as proposed by Brodsky and
Cieben (1978). Strains can be differentiated on the bases
of the antigens and pyocyanins that they produce, as well as
by phage typing.
P. aeruginosa can cause illness, especially in immuno-
depre₯sed patients. Ingestion or contact with large amounts
of P. aeruginosa in water can cause enteric, eye, ear, and
upper respiratory tract infections. The enteric infections
are most often manifested as stomach upsets and nausea, but
they are not usually reported to medical practitioners. To
-------
90
become infected, a consumer must be weak (very young or old
or immunodepressed) and ingest an effective dose (Hunter,
1947). The eye, ear, and upper respiratory tract infections
often cause relatively long-term symptoms, as reported by
Hoadley (1977).
A very important problem associated with P. aeruginosa
is its increasing resistance to antibiotics (Van Dijck and
Van der Vorde, 1976). This resistance is extrachromdsomic
and the plasmid can code for multiple resistances (Leclerc,
e_t a^., 1977a). These plasmids can be passed to the human
enteric bacilli that occur with P_. aeruginosa in water. P.
aeruginosa can also be very resistant to antiseptics. ~~
Even though the pathogenicity of P_. aeruginosa is
determined more by the patient's state of resistance than by
any inherent virulence, the possibility of creating con-
ditions for this microorganism to proliferate in distri-
bution and plumbing systems must be considered. About 10
percent of the healthy human population are carriers and can
eliminate P_. aeruginosa in feces (Hugh and Gilardi, 1974).
This organTsm can survive in raw water and can be present in
drinking water with a frequency varying from 1 or 2 percent
to more than 50 percent, depending on the quality of the
water (potable or not potable) and the type of water treat-
ment practiced. P_. aeruginosa may also persist and grow in
water for extended periods and so is frequently found in the
absence of coliforms; one study reported that half of the
finished water samples containing P_. aeruginosa lacked
coliform bacteria (Nemedi and Lany, 1971).
For all these reasons, consideration should be given to
possible incorporation into routine drinking water analyses
of tests for the presence of P_. aeruginosa [See Sections
F.2 to 3 and G.6], since suitable methods exist for its
enumeration [See Section C.2.c].
b. Viruses. Viruses are ultramicroscopic intra-
cellular parasites, incapable of replication outside of a
host organism. They consist of a nucleic acid genome en-
closed in a protective protein coat and some viruses have
lipid-containing outer envelopes. They have been classified
on the basis of their size and shape, the composition of
their nucleic acid (single or double stranded DNA or RNA),
and by their antigenic properties.
Large numbers and diversities of bacterial, plant, and
animal viruses may be present in both polluted and unpol-
luted waters. Most of these viruses are harmless to humans;
-------
' ..;' 91
they may, however, be economically damaging to agriculture
and certain industries such as breweries and dairies.
The viruses of greatest concern are those -of'human-
origin which are capable of infecting and causing disease in
humans. In general, they are shed in the feces and are
known as the human enteric viruses. The most important
groups of human enteric viruses will be discussed below.
The virus particles (virions) in all these groups are roughly
spherical, acid-stable, and lack envelopes [See Table
B.l.b-1]. These viruses may cause symptoms in the digestive
tract (vomiting and diarrhea), generalized symptoms such* as
fever, and occasionally respiratory illnesses; some of the
more virulent enteric viruses may affect the central nervous
system (meningitis or poliomyelitis) or the liver.
(i) Pathogenesis of Viruses. Infectivity resides in the.
nucleic acid portion of the virus particle. When a sus-
pension of infectious particles is inoculated into a culture
of susceptible cells, the particles are engulfed by, or
penetrate, host cells, and the cells produce progeny virus.
Death of the host cells often results as the viruses con- .
tinue to replicate.
Infection usually takes place after viruses are in-
gested, possibly in contaminated water or food. They can
pass unharmed through the stomach and infect cells lining
the lower alimentary canal. Infection may also start in the
throat or, in some cases, in the upper respiratory tract and
then spread downwards to the gastrointestinal tract.
The severity of the disease depends on the health, and
immune status of the patient and on the virulence of the
virus. Fortunately, symptoms are negligible or even non-
existent with the majority of infections, even for the most
virulent strains of virus. Diagnosis can often be difficult
as one virus may cause a variety of symptoms in different
patients and different viruses may produce 'the same symptoms.
(ii) Occurrence and Stability of Viruses. Because
they are excreted through the feces of an infected indi-
vidual, human enteric viruses are the ones most often encoun-
tered at sewage treatment plants, and therefore, they are
the ones most likely to be released to environmental waters.
To a lesser extent, viruses may enter wastewaters from other
sources such as blood and urine. Additionally, animals,
-------
TABLE B.l.b-1
HUMAN ENTERIC VIRUSES THAT MAY BE PRESENT IN CONTAMINATED WATER
Virus Group
(Size and Composition)
Number
of Types
Disease or Symptoms Caused
Host Range for Principal
Natural Infection Cultivation Methods
ENTEROVIRUS (About.20-30 run
diameter. Single-stranded
RNA in a protein shell)
Poliovirus 3
Coxsackievirus A 24
Coxsackievirus B 6
Echovirus 34
New enteroviruses 4
HEPATITIS VIRUS type A
(probably an enterovirus)
GASTROENTERITIS VIRUS 2 .
(possibly an enterovirus)
Poliomyelitis, meningitis, fever Man
Herpangina, respiratory disease, Man
meningitis, fever
Myocarditis, congenital heart Man
anomalies, meningitis, respiratory
disease, pleurodynia, rash, fever
Meningitis, respiratory disease, Man
rash, diarrhea, fever
Meningitis, encephalitis, respira- Man
tory disease, acute haentorrhagic
conjunctivitis, fever
Infectious hepatitis Man
Vomiting and diarrhea Man
Human and monkey cells
Newborn mice, human cells
Newborn mice, human and
monkey cells
Human and monkey cells
Human and monkey cells
No known method except
possibly in marmosets
and chimpanzees
No method available
to
-------
TABLE B.l.b-1 Continued
Virus Group
(Size and Composition)
Number
of Types
Disease or Symptoms Caused
Host Range for Principal
Natural Infection Cultivation Methods
ROTAVIRUS (About 70 nm diame-
ter. Double-stranded RNA in
a single or double protein
shell
REOVIRUS (About 75 nm in dia-
meter. Double-stranded RNA
in a double protein shell)
ADENOVIRUS (About 70-80 nm in
diameter. Double-stranded DNA
molecule in a protein shell)
PARVOVIRUS (Small diameter.
Single-stranded DNA molecule)
Adeno-associated
Vomiting and diarrhea, mainly
with children
Not clearly established
>30 Respiratory disease, eye
infections
Not clearly established but
associated with respiratory
disease in children
Human infants
and calves
Man
Man
Man and some
domestic
animals
Limited replication in
pig kidney, continuous
monkey kidney, and
human embryo kidney
Human and monkey cells
Human cells
Human embryo kidney cells
coinfected with
adenovirus
1C
to
-------
94
contribute viruses to water, but as yet viruses from this
source do not appear to present a public health hazard.
Over 100 different enteric viruses are known to be shed
in human feces [See Table B.l.b-l]f and new ones are still
being discovered. They are often excreted in very large
numbers, > 10 per gram of feces, and > 5 x 10 per liter
have been found in sewage. Shellfish that are grown in
waters contaminated with sewage effluent can bioaccumulate
enteric viruses, including the virus of hepatitis A. Viruses
may be harbored as long as two months or more in shellfish
digestive tracts after all indications of water contamination
have ceased.
The human enteric viruses, in general, are stable in an
acid environment and in cold temperatures. They are resistant
to bile, ether, and chloroform; but they are sensitive to
heat, desiccation, and to various oxidants. All are relatively
stable outside the host organism, and some can persist in
water over long periods of time. As a result, human enteric
viruses have been recovered from water long distances from
the initial point of contamination.
(iii) Human Enteric Viruses Likely to be Recovered
from Water. Viruses of concern to water authorities are
those capable of causing major epidemics, as happened in New
Delhi in 1955 - 1956 when the water supply became grossly
contaminated with sewage, spreading infectious hepatitis to
28,000 people (Viswanathan, 1957). Yet, except for the
hepatitis A virus, the public health significance of human
enteric viruses in water remains unclear due possibly to the
inapparent or latent nature of viral infections and the
difficulty of detecting waterborne viruses.
(iii.l) Enteroviruses. These viruses comprise an
acid-stable subgroup of the picornaviruses. Human entero-
viruses include the polioviruses, the coxsackieviruses
groups A and B, and the echoviruses [See Table B.l.b-1], In
addition, several new members (yet to be fully classified)
recently have been discovered. Members of the picornavirus
group contain single-stranded RNA and are further charac-
terized by the absence of a lipid-containing envelope,
average 28 nm size (most subgroups), and increased thermal
stability in the presence of divalent cations.
-------
95
The polioviruses are the most widely studied because
they are easy to culture and they include vaccine strains
that are safe to handle. Still, they are the most poten-
tially dangerous of the enteroviruses, causing the major
crippling disease of infectious origin.
Polioviruses are transmitted primarily through direct
contact with pharyngeal secretions or feces of an infected
individual. The seasonal peak incidence of poliomyelitis is
in the summer and fall (as is true for most all of the
enteroviruses) although sporadic cases do occur in any month
of the year. In recent years, however, occurrence of polio-
myelitis has been reduced significantly by the use of highly
effective vaccines. Nevertheless, the virus has not been
eliminated from the community, and vaccination programs must
be maintained to prevent the resurgence of a susceptible
group of children.
Infection with poliovirus and probably most of other
enteroviruses normally confers lifelong immunity against the
infecting virus strain (this is the basis of the live atten-
uated vaccine). The older a person is before becoming
infected, the more severe his symptoms are likely to be. In
many developing countries where sanitary conditions tend to
be poor, waterborne spread of polioviruses and other entero-
viruses almost certainly occurs. There, infection usually
takes place early in life with severe symptoms being few and
limited to young children. However, in developed countries
where more hygienic conditions prevail, infection may be
delayed leading to a greater proportion of serious disease
in older age groups. This pattern, characteristic in devel-
oped countries, has been observed with poliovirus and can be
expected with other viruses.
Eight outbreaks of poliomyelitis, which occurred during
the period of 1900 - 1953, in Europe and North America, were
attributed to transmission by water, however, Mosley (1967)
believed that only one of them was adequately documented.
In this instance, 45 cases of poliomyelitis occurred among
families living in a cluster of temporary houses on the
outskirts of Lincoln, Nebraska (U.S.) in June and July of
1952. The water supply was known to have wide fluctuations
in pressure, and vacuum breakers were missing from a number
of the toilets.
Where sanitation is good, the oral-oral spread of
poliovirus may be more important than the fecal-oral route.
For this reason and others mentioned earlier, water would
-------
96
not appear to be involved with the transmission of polio-
viruses in developed countries.
The other enteroviruses are less well known, particu-
larly the group A coxsackieviruses. They also can cause
serious illnesses [See Table B.l.b-1] but, like the polio-
viruses, usually only in a small proportion of those in-
fected (generally, poliovirus causes significant central
nervous system disease in only 1 percent or less of those
infected).
The most common of the diseases associated with entero-
virus infections may well be aseptic meningitis: over 3,000
cases of this disease were reported in the U.S. during 1974.
The disease can be paralytic, mimicking poliomyelitis;
however, the cases are generally not as severe as those
caused by the polioviruses (reviewed in Safe Drinking Water
Committee, 1977). Many of the infections are so mild as to
escape clinical recognition; others produce encephalitic
manifestations and transient paresis. Aseptic meningitis
also occurs most frequently in late summer and early autumn,
but its incidence tends to be sporadic rather than epidemic.
It is possible that the viral agents of meningitis might be
transmitted via water, but there have been no documented
instances of this as yet.
Although large numbers of enteroviruses have been
isolated consistently from fecally contaminated water, only
a few reports have implicated water as the vehicle of trans-
mission. This may be due to lack of suitable detection
methods and difficulty in demonstrating disease incidence
epidemiologically. Or, it could be that the viruses re-
covered from sewage were the consequence of viral infection
in a population rather than the source. The ambiguities
surrounding the question of cause or effect with respect to
virus occurrence in water will be elaborated on later in the
discussion.
(iii.2) Hepatitis A Virus. Hepatitis A is the only
form of viral hepatitis known to be transmitted through
water and it is also the most prevalent waterborne disease
attributable to a specific etiologic agent. Its spread, via
sewage-polluted water, is well documented due to its clear-
cut symptomatology and the explosive nature of its occur-
rence. It is therefore considered the most important viral
disease transmitted through water.
-------
97
The virus particle is apparently small enough to be
either an enterovirus or a parvovirus and buoyant density
studies suggest that it resembles an enterovirus. Its
nucleic acid composition has not yet been determined. The
agent seems to withstand acidity and heat, but perhaps not
more than what the enteroviruses will endure. Though
chimpanzees (and perhaps other primates) are infected acci-
dentally on occasion, the only natural host of the virus
seems to be man. Transmission is predominantly fecal-oral.
The incubation period of the disease in humans ranges
generally from 15 to 50 days, with a median of 28 days.
Some people shed the virus in their feces as early as seven
days before onset of symptoms; others may transmit the virus
without ever becoming perceptibly ill. Anorexia, nausea, and
vomiting are seen in most cases; and jaundice is common but
certainly not universal. Laboratory diagnosis is often
based on demonstration of abnormally high levels of glutamic
oxalacetic and glutamic pyruvic transaminases (SCOT, SGPT)
in the patient's blood serum. The carrier state does not
seem to be prolonged; the virus is usually absent from
feces after the symptoms have abated and serum transaminase
levels have become normal.
The hepatitis A virus has caused several serious epi-
demics and numerous small outbreaks in various parts of the
world, some of which have been attributed to the consumption
of polluted water or contaminated shellfish. Hepatitis A
has been epidemiologically implicated in 66 waterborne
outbreaks from 1946 to 1975, but its incidence in developed
countries does not appear to be increasing. Moreover, the
occurrence of hepatitis A as a whole (of which waterborne
spread makes up only a small fraction), has shown a downward
trend since 1971, in the U.S. Of the 22 waterborne out-
breaks associated with municipal systems, three resulted
from either inadequate or interrupted disinfection and five
were related to the use of contaminated, untreated surface
or groundwater. Half (eleven) of the outbreaks traced to
municipal systems occurred as the result of contamination of
the distribution system, primarily through cross-connections
and back-siphonage (Craun, 1978) [S'ee Section F.4].
Attempts to culture the hepatitis A virus in cell lines
have consistently failed so that all available evidence
about its transmission by polluted drinking water has been
obtained from epidemiological surveys. Human volunteers and
chimpanzees have, however, been infected with the virus a d
-------
98
recently developed methods employing immune electron micro-
scopy (IBM) have afforded some progress. These investi-
gations indicate that the hepatitis A virus may soon be
classified as an enterovirus. If so classified, the hepa-
titis A virus could be expected to react to water treatment
and purification processes in a manner similar to the other
enteroviruses [See Section E.].
(iii.3) Gastroenteritis Viruses. Another group of small
viruses potentially transmissible through water, some of which
may in fact be enteroviruses, are called gastroenteritis
viruses. Included in this heterogenous group are the Norwalk
agent and the related Montgomery County and Hawaii agents,
rotaviruses, astroviruses, coronaviruses, caliciviruses, the
W agent, the cockle virus, and the Ditchling agent [See
Section B.l.d]. These have all been detected in diarrheal
feces by direct electron microscopy. They do not replicate
in routine cell cultures, although some have been shown to
replicate to some extent in particular cells (reviewed by
Madeley, 1979).
Gastroenteritis is probably the most common waterborne
disease in developed countries and was reported to have
involved 45,255 cases in 178 outbreaks between 1946 and 1970
within the U.S. (Craun and McCabe, 1973). Gastroenteritis
epidemics of viral origin are known to be spread by personal
contact; and it seems reasonable to suspect that many water-
borne gastroenteritis outbreaks might stem from the same
etiologic agents. [See Section B.l.d.]
(iii.4) Rotaviruses. These viruses are related to the reo-
viruses and belong to the same family, Reoviridae. They have been
shown to be antigenically related to viruses of similar morphology
isolated from various animal species; but they are anti-
genically distinct from reoviruses and orbiviruses. Rota-
viruses are primarily associated with gastroenteritis in
children.
Characterizations of rotaviruses have been made pos-
sible by the use of electron microscopy, and recent investi-
gations suggest the possibility of more than one human
serotype (reviewed by Madeley, 1979). The viruses have been
concentrated and purified from fecal extracts for bio-
chemical and biophysical characterization. Rotaviruses are
described as being 70 nm in diameter and appear to contain
double-stranded RNA and a protein component with eight
-------
99
distinct polypeptides observed by electrophoresis. Par-
ticles with single- and double-shelled capsids have been
detected at bouyant densities of 1.36, and 1.37 g/ml in
cesium chloride gradients (Rodger, et aJ., 1975).
The route(s) of transmission of rotaviruses associated
with diarrheal disease in infants and children has not been
identified, although direct person-tq-person spread has been
indicated In some hospital nurseries. In temperate cli-
mates, infections with rotaviruses increase during the cold
winter months; and at peak times, rotaviruses have been
observed in fecal specimens of as many as 90 percent of
young hospital patients. A few studies have indicated that
adults may,also be infected with rotaviruses, usually resulting
in relatively mild gastroenteritis. The isolation of rota-
viruses from U.S. students who experienced "travelers'
diarrhea" while attending a Mexican university and from a
community-wide outbreak in Sweden suggests that water trans-
mission of the agent should be considered (Bolivar, et al.,
1978; Lycke, et al., 1978). -Rotaviruses may be excreted in
very large numbers (up to 10 per gram of feces Madeley,
1979) and they are almost certain to be present in polluted
water, but there is presently no suitable method for their
detection at low concentrations [See also Section B.l.dJ.
(iii.5) Reoviruses. Reoviruses are members of the
Reoviridae and share a number of properties with the entero-
viruses fe~.g., spherical shape, lack of an envelope) only
reoviruses are larger and possess double-stranded RNA cores
[See .Table B.l.b-1]. They are unusual in that they infect a
wide range of animals; furthermore, they have been recovered
from persons with a wide variety of illnesses including
upper and lower respiratory tract disease, gastrointestinal
disease, steatorrhea, exanthems, and some central nervous
system diseases. Although they have been found to be highly
infectious and have been isolated frequently from sewage and
contaminated water, their etiologic role in man has not been
established (Horsfall and Tamm, 1965). ,
(iii.6) Adenoviruses. Adenoviruses are icosahedral
and contain double-stranded DNA.[See Table B.l.b-1]. At
least 28 types are recognized and many serotypes have been
isolated from human sources. The adenoviruses resemble
enteroviruses only in their resistance to ether and absence
of a lipid-containing envelope.
-------
100
Members of this group generally are considered to be
respiratory viruses; they can infect the upper respiratory
tract as well as the intestines, typically without evidence
of disease. Adenoviruses are often shed for prolonged
periods in the feces of young children. They have been
isolated frequently from stool specimens, sewage, and con-
taminated water. Curiously, only a few of those seen in
feces by electron microscopy have yielded to culturing in
cell lines.
The adenoviruses produce sporadic infections in the
general population which tend to be highest in the colder
months; but, overt disease is rare. Although they occasion-
ally appear to have been involved with outbreaks of diarrhea,
reports of their transmission through water have been con-
nected only with outbreaks of conjunctivitis in persons
frequenting swimming pools. Adenoviruses are readily inacti-
vated by free chlorine at levels commonly used to protect
water supplies; hence, they should be entirely absent from
adequately disinfected drinking water.
(iii.7) Adeno-Associated Viruses. The adeno-asso-
ciated viruses are members of the parvovirus group and, as
such, are the smallest of animal viruses, some having diam-
eters of as little as 19 nm. They are polyhedral, lack a
lipid-containing envelope, and contain one molecule of
single-stranded DNA. Buoyant density studies have placed
them between 1.38 and 1.46 g/ml in cesium chloride. They
comprise four known serotypes which have been found to
chiefly infect the lower animals such as rats, mice, cats,
dogs, swine, and birds; however, infectivity appears depen-
dent on concurrent infection with adeno- or (less readily)
herpesviruses. By themselves, the adeno-associated viruses
are not known to cause any symptoms. They have, however,
been recovered in association with' childhood respiratory
diseases.
.These viruses are quite stable outside of the host
organism and can withstand heat at 60°C for at least 1 h.
They are highly resistant to many physical and chemical
agents and are unaffected by ether, chloroform, and anionic
detergents.
Since they have been isolated from feces together with
adenoviruses, they are almost certain to be present in
fecally contaminated water, although no method for their
detection in water is currently available for confirmation.
-------
101
There are, at present, insufficient data on these viruses
for determining their pathogenicity and hence, for judging
what, if any, threat they might pose if ingested with drinking
water.
(iv) Methods for Detecting Viruses in Water. As more
and more water is used to receive waste discharges and run-
off from domestic, agricultural, industrial, and recreational
uses, the opportunities for fecal, and therefore viral,
contamination will increase accordingly. Most enteric
viruses are shed in large numbers in the feces of infected
persons, although it should be noted that the majority of
people in industrialized nations are not infected with
enteric viruses on any given day. Because viruses multiply
only within susceptible host cells, they cannot increase in
sewage. Their numbers are further reduced as a result of
sewage treatment, dilution, natural inactivation, and
water treatment practices [See Section E.]. Therefore,
barring gross contamination of finished water, only low
numbers of virus, if any, are likely to occur in properly
treated supplies.
(iv.l) Concentration Techniques. The possible presence
of only very low numbers of virus necessitates applying
highly efficient methods, including a preliminary concentra-
tion step, to demonstrate and quantitate any viruses present.
Some source waters may contain viruses at such low levels
that sample volumes of 400 1 or more would be required to
enable them to be detected.
Adsorption methods employing glass microfibers (Jakubowski,
1974) and glass powder (Sarrette, et a_l. , 1977) appear
promising for handling large volumes of water containing few.
viruses. Concentration with glass microfibers permits sampling
of 2,000 1 or more of water and is recommended as a tentative
procedure in Standard Methods (American Public Health Assoc.,
1975). Many of the enteroviruses can be adsorbed at pH 3.5
to 4.5 in the presence of A1C1~ followed by elution with a
protein solution at pH 9.5 or 0.05 M glycine buffer at pH
11.5. Excellent reviews of these methods have been published
(Foliguet, e_t al. , 1973; and Sobsey, 1976).
Raw water containing suspended matter and organic
substances may require more conservative analytical pro-
cedures, such as the aqueous polymer two-phase separation
-------
102
technique described in Standard Methods (American Public
Health Assoc./ 1975). If enteric viruses are entrapped in
fecal clumps, special procedures must be employed to disrupt
the clumps, releasing the virus.
(iv.2) Cell Culture Techniques. Human enteric viruses,
if they will replicate at all in the laboratory, must do so
in primate cells, although no one cell culture will accom-
modate all of the various enteric viruses capable of propa-
gation in cell lines. Therefore, selection of host cell
systems will, of necessity, be a compromise between what one
expects to find, what one knows will yield to culturing, and
what costs one can afford.
Enteroviruses, reoviruses, and some adenoviruses can be
detected by conventional cell culture techniques using human
primary amnion or fetal cells (such as embryonic kidney) and
suckling mice. For the coxsackieviruses A, all but three of
the 24 types are isolated only in suckling mice. Human
diploid embryonic lung cells were used successfully to
recover 31 echoviruses (reviewed in Safe Drinking Water
Committee, 1977). Unfortunately, there are yet many known
enteroviruses that have no known tissue culture hosts.
Likewise, there is no laboratory host system for detection
of the virus of hepatitis A.
Neither are there cell culture systems for detecting
gastroenteritis viruses, although rotaviruses will repli-
cate, to a limited extent, in pig kidney, continuous monkey
kidney, and in human embryo kidney (reviewed by Madeley,
1979). The Norwalk, Montgomery, and Hawaii agents, along
with rotaviruses, adenoviruses, caliciviruses, the W agent,
cockle virus, and the Ditchling agent are detectable, when
present in patients' stools at high enough levels (at least
10~6 pfu per g feces), using electron microscopy.
Adenoviruses can be cultured only in human cells, and
adeno-associated viruses are cultured on human embryo kidney
cells concurrently infected with adenoviruses. Quantita-
tion of adeno-associated viruses is thereby handicapped by
the necessary inclusion of a helper virus to induce infec-
tivity.
(iv.3) Virus Identification Techniques. Serologic
methods for virus identification can be based on serum
-------
103
neutralization or antigen-antibody interactions. The antigen-
antibody tests (comprising immunofluorescence and immune
enzymatic procedures) yield results sooner than the visible
cytopathic effects required for interpretation of serum
neutralization tests. However, these immunochemical tech-
niques have not been adapted to use pooled antisera and so,
unlike serum neutralization tests, they require as many
individual reactions as there are virus types to be iden-
tified.
(v) Public Health Significance of Viruses in Drinking
Water. The presence of viruses, pathogenic to humans, in
groundwater, streams, rivers, and lakes, is certain. The
majority of these emanate from domestic sewage but are present in
widely fluctuating numbers, anywhere from no viruses per 100
ml to over 45,000 viruses per 100 ml of sewage (Gerba, et
al., 1975a). The season, hygienic level of the population,
and incidence of disease in a particular community are the
major determinants of viral presence, and these factors all
are subject to change.
Except for the occurrence of waterborne hepatitis A,
however, there is little evidence to connect the presence of
enteric viruses in water with incidence of viral disease.
Although isolated regularly from sewage and environmental
waters, the picornaviruses have not been demonstrated,
epidemiologically, to cause disease by this route. Con-
versely, those outbreaks of gastroenteritis that have been
traced to water have not been shown to be of viral origin.
If the epidemiologic data on transmission of hepatitis
A through water are representative, it may be useful to
consider the kinds of circumstances which have led to water-
associated outbreaks of this disease. The overwhelming
majority of reported waterborne outbreaks of hepatitis A
were attributable to untreated or inadequately treated water
supplies. In the case of semipublic and private water
supplies, the fault lay usually with sewage contamination of
the source coupled with little or no treatment. With large
municipal supplies, the problem usually was one of cross-
connections or back-siphonage during distribution. The
solutions to these obvious deficiencies are straight-forward
and attainable.
On the other hand, it is believed by some that viral
infections may be spread insidiously by way of continual low
level transmission through water. It is the nature of most
-------
104
viral diseases that they elude the epidemiologist in his
quest for definitive data. Many viruses can cause inap-
parent or latent infections which go unrecognized until
secondary person-to-person spread finally leads to overt
disease in scattered, virtually untraceable pockets of the
population. One case in point is the picornaviruses, the
majority of which cause subclinical infection much more
frequently than overt disease.
Inadequacies in current water treatment techniques have
been demonstrated on several occasions when viruses were
isolated from conventionally treated potable supplies which
had received treatment adequate to remove bacterial patho-
gens (Berg, ej: al., 1975; Jenkins and Mendia, 1967; Hoehn,
et al_., 1977). Whether the extremely small numbers of viruses
that manage to get through into drinking water are suffi-
cient to initiate infection in a community, or whether other
modes of transmission play a much more prominent role in
disseminating them , is not yet known.
Virus interactions with tissue cultures and with the
intact host do not appear to be entirely comparable, although
such comparisons have been made. There is evidence to'
suggest that anywhere from a few to a great number of viral
particles must be present to account for one tissue culture
dose (TCD) that quantity of virus required to produce in
a tissue culture an infection that is perceptible on the
basis of cytopathic effect, plaque formation, or some other
manifestation. The number of virus particles corresponding
to a TCD sometimes ranges in the thousands but is always
greater than one. Although tissue cultures vary in their
susceptibility to any given virus, it is clear that the
process by which animal viruses initiate infection is inher-
ently inefficient; there is no known system in which every
single animal virus particle succeeds in initiating infec-
tion. The number of viral particles that equals a TCD
varies with the cell line as well as with the virus; the
laboratory procedures used to make these determinations
impose further uncertainties. Finally, host cells in culture
may not be strictly representative of intestinal cells in
vivo? hence, experiments which try to relate TCD's mani-
fested in cell lines to human infectious doses have produced
quite disparate results.
Infection and disease are not synonomous. Viral infec-
tions may be highly prevalent when the incidence of viral
disease is low. The occurrence of disease among infected
persons is strongly dependent upon the individual hosts'
-------
105
susceptibilities. A detailed summary of this topic is
presented .by the Safe Drinking Water Committee (1977) .
For all of the reasons outlined above, reported levels
of viral contamination in water cannot be used directly to
assess the risk to human health. If surveillance programs
were equipped and staffed to detect virus -infections, there
would be an increased likelihood of tracing sources of viral
disease. However, such an approach would be almost impos-
sible to implement because of the considerable expense and
personnel time it would consume. Studies to determine virus
levels in drinking water, in conjunction with epidemiologic
surveys of viral disease in the community, are sorely needed,
Epidemiologists and public health laboratories must have
available to them the expertise for conducting extensive
virologic studies.whenever disease is suspected of being
waterborne and the etiologic agent is unknown. Until infor-
mation gathered from such investigations is forthcoming, the
public health significance of small numbers of viruses in
drinking water will remain undetermined.
(i) Entamoeba histolytica. This protozoon causes amebic
dysentery as well as such other clinical manifestations as .
diarrhea, abscesses of the liver (or less commonly of the lung or
brain), and skin ulceration. It can also cause chronic infections
with minimal symptoms, or the organism may act as a commensal
within its asymptomatic host carrier. Amebiasis (infection
with 13. histolytica) occurs world-wide; and dysentery epidemics
are, by and large, waterborne -- usually the result of poor
sanitation.
(i.l) Characterization and Pathogenesis. Entamoebidae
is the only exclusively parasitic family in the order
Amoebida; and 13. histolytica (referred to in some American
literature as Endarooebal is the only true pathogen among :f-_
five species of amebae which parasitize the human intestine.
Two other members of the genus Entamoeba, E. coli and .E.
hartmanni, are harmless intestinal parasTtes, the latter"
morphologically resembling 13. histolytica and easily con-
fused in clinical diagnoses.
_E. histolytica exists in both the actively growing
trophozoite (9 x 11 urn or 18 x 25 urn) and dormant, thick-
walled cyst (6 to 8 um or 12 to 15 urn in diameter) states,
as do the other intestinal amebae (except Dientamoeba
fragilis which exists only as a trophozoite). The organism
-------
106,
is not infectious when passed in the feces of patients with
acute dysentery, owing to its occurrence at this time as a
fragile trophozoite. However, once lodged in the host,
trophozoites can attack the walls of the intestine, feeding
characteristically (but not exclusively) upon red blood
cells. They may penetrate the intestinal lining, causing
intense inflammation and ulcers (amebic dysentery). They
may also gain entry to the lymph and blood vessels and then
be carried to the liver, lungs, brain, and other organs
where abscesses will result. Cysts are passed in the feces
of asymptomatic carriers or of patients with chronic or mild
cases of amebiasis; in this way, they may spread the disease,
As early as the 1940's, most researchers agreed that
other factors beyond the solitary presence of 13. histo-
lytica were needed to provoke an infection (Chang, 1948).
When axenic E_. histolytica cultures were administered to
gnotobiotic guinea pigs, no symptoms were apparent. How-
ever, animals whose intestines harbored either E. coli or
A. aerogenes developed ulcerations after receiving the
pathogen (Phillips, et aJ., 1955). The importance (but not
the exact role) of an accompanying bacterial flora was
realized: one proposed mode of mixed etiology was supposed
to involve a profound anaerobiosis occurring as a result of
disturbance to the normal bacterial flora (Chang, 1948).
Such anaerobiosis would ostensibly favor penetration of the
mucosa by IS. histolytica. In its vegetative stage (tropho-
zoites), E. histolytica is extremely fragile and is killed
in 35 min at 45°C, in 2 to 5 h at 32°C, in 6 to 12 h at
25°C, and in 18 to 96 h at 15°C. The cysts are more resis-
tant, being able to survive in the environment for three to
four days at 25°C and more than four weeks in slightly
alkaline water.
Although the U.S. Center for Disease Control has, in
the last decade, received from 2,000 to 3,000 annual reports
of symptomatic amebiasis (Center for.Disease Control,
1977c), none since 1953 has been waterborne. Those
which did occur prior to 1953 involved sewage contamination
of water supplies that primarily serviced private dis-
tribution systems. It is clear that cross-connections or
back-flow of sewage into water supply lines must be pre-
vented in order to preclude transmission of IE. histolytica
through drinking water [See Section F.4]. No" waterborne
outbreak of symptomatic amebiasis has been reported in
Western Europe during the 1970's..
-------
- . 107
(i.2) Effects of Treatment Practices. In the U.S.,
improved sanitary sewerage systems have been largely respon-
sible for the curtailment of waterborne amebic dysentery.
Also important has been the introduction 6f coagulation [See
Section E.2] and filtration during water treatment. Sand
filtration of water [See Section E.3] removes nearlyiall
cysts, and diatomaceous earth filters remove them all. Data
on survival of E.. histolytica exposed to iodine [See Section
E.S.d] are summarized in Figures B.l.c-1 and 2. The para-
meter of cysticidal residual used in these calculations is
defined as that concentration required to kill 99.999 percent
of cysts. High pH decreases the cysticidal activity of
chlorine [See Section E.S.a], bromine, and iodine. But at
pH 4, a bromine residual of 1.5 ppm produced 9,9.9 percent
cyst mortality, whereas 5 ppm iodine residual and 2 ppm
chlorine residual were required for the same results under
the same conditions. Ozone, applied at 0.5 ppm, can kill 97
to > 99 percent IB. histolytica cysts suspended in tapwater
(Newton and Jones', 1949) [See Section E.S.b]. Also, heating
to 52°C for 2 min destroys E. histolytica cysts (Chang,
1950); and drying causes instantaneous death (Chang, 1948).
Finally, water storage for four weeks at 10°C and for one
week at 20°C destroys cysts.
There is at present no standardized method for de-
tecting E. histolytica in water. Various methods for .concen-
trating T5y filtration have been recommended. Detection is
essentially accomplished by careful microscopic examination
of concentrated or unfiltered water samples that have or
have not been stained. However, laboratory diagnosis for
13. histolytica is often difficult and requires highly spe- w
clalized procedures. ;
(ii) Giardia Lamblia. This protozoon infects the
small bowel and causes prolonged diarrhea and several other
intestinal symptoms. It is the most frequently diagnosed
parasitic pathogen in U.S. public health laboratories;
moreover, giardiasis appears to be emerging as a major
waterborne disease, having been responsible for 14 water-
borne outbreaks in the U.S. between 1965. and 1975, 12 of
which occurred between 1971 and 1974 (Horwitz, et al.,
1976)., It was not until 1966 that water was recognized as
a vehicle of giardiasis transmission, and since that time,
increased awareness on the part of physicians may have
accounted for the more frequent reporting of outbreaks.
-------
108
FIGURE B.l.c-1
E
a.
ex
OJ
CO
Ul
ce
2
o
p
CO
>~
o
RELATIONSHIP BETWEEN CYSTICIDAL RESIDUAL3,
TITRATABLE IODINE, (I0), AND CONTACT TIME
AT 3°, 10°, 23°, AND 35*CB
I I I I 11II
2 3 4 5 6 7 8 9 10 2O !30
CONTACT TIME t IN MINUTES
Concentration needed to kill 99.999% of cysts.
40 50 60 70 80 100
pH < 6.5 and iodine ion concentration < 20 ppm.
-------
FIGURE B.l.c-2
RELATIONSHIP OF CONTACT TIME TO CONCENTRATION OF I- AND HOI
FOR DESTRUCTION OF 99.9% OP CYSTS, VIRUSES, AND BACTERIA AT 18eC
I I I I I
|2 ON POLIOVIRU3
HOI ON CYSTS OF
HISTOLYTICA
12 ON CYSTS OF
ISTOLYTICA
HOI ON POLIOVIRUS
TYPE I
O.I
1.0 2 3 4 5 678 10
CONTACT TIME, min
II INI
60 80 100
o
vo
-------
110
(ii.l) Characterization and Pathogenesis. The flagel-
late (3. lamblia exists as both trophozoite (9 to 18 um x 6
to 9 um) and cyst (9 to 12 um x 5 to 8 um). After being
ingested as cysts, the organisms quickly develop into infec-
tious trophozoites, coating the mucosa of the duodenum and
upper jejunem and often causing such changes as flattening
of villi and roughing of mucosal surfaces; the flagellate
bodies may imbed deeply in the villi (Burke, 1975; Nath, e_t
al., 1974). The large number of trophozoites lining the
mucosa interferes mechanically .with absorption (especially
of fats) and enzyme activity, and this can lead to weight
loss, malnutrition, and anemia. Symptoms, which include
chronic diarrhea, steatorrhea, and abdominal cramps, are
?enerally more severe in children (especially those with any
mmunodeficiency) than in adults.
(ii.2) Epidemiology. Beavers were implicated as the
source of a waterborne giardiasis outbreak which took place
in Camas, -Washington (U.S.) in 1976, which included 128
confirmed cases in a population of 6,000 people (Dykes, et.
al. / 1977). Beavers, but not coyote, opossums, nutrias,~~or
porcupines, were found to be carrying G. lamblia cysts which
were recovered from both raw and finished surface waters.
The beavers inhabited a pond bordering a heavily used state
park and were within foraging distance of water intakes for
Camas [See also Section B.2]. A similar epidemic occurred a
year later in Berlin, New Hampshire (U.S.) where at least
205 residents (an attack rate of/x/25 percent) were -con-
firmed to be ill with giardiasis. Sewage contamination of
one raw water source was demonstrated, and unrestricted
public access to the other source provided opportunity for
fecal contamination (Center for Disease Control, 1977c).
Beavers trapped nearby were found with intestinal lesions
produced by Giardia trophozoites.
The first laboratory confirmed (using pathogen-free
beagle dogs) outbreak of giardiasis in the U.S., also the
largest outbreak to date, involved an estimated 5,300 persons
(in a population of 46,000) in Rome, New York; 359 of the
cases were clinically confirmed (Shaw, e_t al. , 1977). Raw
mountain stream water from the heavily wooded, sparsely
populated watershed had routinely shown low coliform counts
over a two-year period, but very high total bacterial counts
were recovered from water in the distribution system. This
led investigators to conclude that raw water, though not
contaminated with human sewage, was insufficiently chlori-
nated during treatment. However, they went on to caution
against reliance on chlorine for destroying Giardia cysts.
-------
Ill
Every U.S. outbreak involving municipal water supplies/
except one, has been associated with surface water sources
where disinfection was the only treatment. The one excep-
tion occurred in Aspen, Colorado in 1965 and 1966 (Moore,
e_t al. , 1969). Wells provided the town's water, but old,
broken and leaking sewer lines passing close by served as
the source of contamination.
(ii.3) Effects of Treatment Practices. G. lamblia
cysts are not destroyed by chlorination at dosages and
contact times commonly used in water treatment [See Section
E.S.a], nor do closed pressurized filtration systems afford
adequate protection, as was evidenced in the Camas outbreak.
Beyond ensuring against opportunities .for fecal contamina-
tion, water treatment regimes should include coagulation [See
Section E.I], sedimentation [See Section E.2], and filtra-
tion [See Section E.3]. When aluminum sulfate was applied
to water maintained at usual pH levels and turbidities en-
countered at a water treatment facility, 99.999 percent of
suspended Giardia cysts were removed. The addition of
calcium and/or magnesium salts to aid floe formation is
recommended for very cold or very soft waters. Filters
should be of adequate depth and the rates of filtration and
back-flushing should be adjusted to ensure entrapment of
cysts and to prevent turnover or channeling of the filter.
Proper maintenance is essential. The greatest risks
will occur in water systems that use only disinfection and
in contamination of water in the distribution system by
sewage from broken lines or from cross-connections. More
study is needed to define the conditions required for destruc-
tion of G. lamblia cysts.
Concentration techniques, using microporous filters,
for recovery of G. lamblia cysts are based on established
methods for concentrating viruses. When culture methods
become available for inducing G. lamblia to excyst, propagate,
form trophozoites, and encyst Tn one complete life cycle,
the organism will then be accessible for closer scrutiny and
possible eradication as a waterborne agent of disease.
(iii) Naegleria Fowleri. The fa-cultatively parasitic
ameba, Naegleria fowleri, has, in the recent past, been
recognized as "one of two agents (also Acanthamoeba) responsible
for causing primary amebic meningoencephalitis(PAM), mostly
in children and young adults a disease which is usually
-------
112
fatal within three to seven days. The organism is a free-
living ameba, ordinarily found in water, soil, and decaying
vegetation. There is no evidence of any person-to-person
transmission; instead, the infection appears to take hold,
in most instances, when water containing N. fowleri cysts
enters through the nose, usually as a result of swimming in
fresh water lakes, swimming pools, or hot springs. Though
associated with recreational waters, the ameba has more
recently been isolated from tapwater in connection with
incidences of PAM (Safe Drinking Water Committee, 1977).
(iii.l) Characterization and Pathogenesis. The first
two known cases of PAM occurred in 1962, in Orlando, Florida
(U.S.), but the causative agent was not definitely identi-
fied until several years later when a Maegleria species was
isolated from cerebrospinal fluids grown on inorganic agar
seeded with coliform bacteria (Butt, 1966; Butt, et al.,
1968).
The genus Naegleria belongs to the family Schizopy-
rendae, of which N. fowleri (which has been successively
named N. fowleri, N. aerobia, and N. invades by different
investigators)is the only species known to be pathogenic.
The ameba measures approximately 8-15 urn and has a spher-
ical nucleus and one or more lobate pseudopodia. It exists
either as a cyst or a trophozoite; the latter becomes flag-
ellated after a few hours in water. Trophozoites are indis-
tinguishable from those of the nonpathogenic N. gruberi in
freshly prepared specimens. Both species have a conspicuous
centrally located nucleolus and are slightly smaller than
E. histolytica; they must be differentiated by serological
tests.
Pathological findings indicate that, once insinuated
into the nasopharynx, the ameba can "chew" its way from the
roof of the nose through the cribiform plate into the cranial
cavity. A phospholipase-like enzyme of the ameba can decom-
pose syringomyelin in a chemically defined medium; this
enzyme is probably the means by which the ameba can pene-
trate the olfactory miicosa and find its way into the brain
through the olfactory nerve plexus (Chang, 1978). Extensive
demyelination of the white matter in the central nervous
system has also been attributed to the action of this
enzyme. The typical syndrome comprises fulminating meningo-
encephalitis with severe frontal headache, nausea, vomiting,
high fever, nuchal rigidity, and somnolence, with death on the
fifth or sixth day.
-------
113
(iii.2) Sources and Occurrence. Cases of PAM have
been reported in Australia, Czechoslovakia, the U.K., Ireland,
Belgium, New Zealand, India, Africa, and the U.S. As of
1974, there were 64 cases reported worldwide, most have
occurred during the hottest time of the year. In South
Australia, infection is believed to have occurred as the
result of school children washing their noses with ameba-
laden tap water (Fowler and Carter, 1965). In Czech-
oslovakia, a swimming pool (and possibly a stream) was the '
source of infection (Cerva and Novick, 1968); a swimming
pool and thermally polluted stream served as vehicles of
infection in Belgium. All cases in the southeast U.S.,
except one, were traced to swimming in lakes; in California,
two victims had a history of swimming in a creek fed by a
hot spring. An extensive survey of lakes in the Orlando,
Florida (U.S.) area (Wellings, et _ajL. , 1977) revealed the
presence of N. fowleri in 12 out of 26 lakes examined (single
sample). Moreover, viable cysts, identified as Naegleria
and Acanthamoeba species, have been found in eight of 15
finished water samples in a survey of large city supplies
across the U.S. (reviewed in Safe Drinking Water. Committee,
1977) . .;';;;
Increasing eutrophication and thermal pollution of
water are, likely to enhance the growth of N. fowleri.
However, evidence of its rather widespread presence has not
been associated with widespread occurrence, of PAM. In fact,
there has been no case in Czechoslovakia since 1968, in
Australia since 1972, in Belgium since 1973, and in New
Zealand since 1975. Other unknown factors may be important
accompaniments to the occurrence of infection and should be
investigated in view of the disease's high fatality.
(iii.3) Persistence. Little is known about the sta-
bility of N. fowleri cysts and trophozoites. Cysts very
likely are resistant to chlorine [See Section E.S.a] as are
those of other parasitic amebae,' therefore, emphasis should
foe placed on coagulation [See Section E.2],and filtration
[See Section E.3].processes to remove these organisms from
waters in which they may be proliferating.
(iv) Acanthamoeba Species. Acanthamoeba, also associ-
ated with PAM, is the most widely distributed, free-living
ameba in all types of fresh water and in sewage effluents
(Chang, 1971a).
-------
114
This facultative parasite measures 10 - 25 um and has
multiple, spine-like filiform pseudopodia. Cysts have a
double layered wall and, upon aging, give a wrinkled appear-
ance from retraction of the inner wall. Acanthamoeba is
able to grow axenically on ordinary media such as tryptic
digest soy broth or even peptone solution; in contrast,
Naegleria generally requires the presence of living bac-
teria, tissue culture cells, fresh tissue, or fetal calf
serum.
Species of Acanthamoeba were formerly classified within
the genus Hartmanella (Singh, 1953). There are still problems
in the differentiation of individual species due to morpho-
logical and cultural similarities and because single strains
oftentimes exhibit morphological variability, even under the
same growth conditions (Chang, 1971a). Species must there-
fore be classified on the basis of serology. Acanthamoeba
cysts are extremely resistant to chlorine [See Section
E.5.a3, drying, and freezing. t
Experimental evidence indicates that all Acanthamoeba
species, except A. palestinensis, are at least mildly patho-
genic. Infection does not occur through any specific portal
of entry,' rather, the organism is believed to be a secondary
invader taking advantage of abrasions, ulcers, or other
infections initiated by other organisms (Chang, 1971a).
When an ulcer is well established, the amebae may consume
all the bacteria and feed on the exudate and cell debris.
Acanthamoeba has been implicated in PAM cases of patients
who were debilitated or undergoing immunosuppressive therapy.
Like Naegleri fowleri, Acanthamoeba species have been
isolated from tapwater in association with PAM cases
(reviewed in Safe Drinking Water Committee, 1977). However,
their presence in finished drinking water seems unlikely to
lead to PAM in the consumer. Ac anthamo eba is detected by
microscopic examination of the water sample.
(v) Helminths (Worms). In the U.S., metazoan parasites
that may be transmitted by ingestion of water are limited to
a few intestinal nematodes (roundworms) and even fewer
trematodes and cestodes (flatworms). Of the nematodes, the
following merit consideration: Ascaris lumbricoides (the
stomach worm), Trichuris trichura (the whipworm), Necator
americanus, Ancylostoma duodenale, A. brasiliense (hookworms),
and Strongyloides stercoralis (the threadworm).Cestodes of
-------
115
public health significance are Hymenolepis nana (the dwarf
tapeworm) and Echinococcus granulosus (the.dog tapeworm).
Eggs of the intestinal nematodes originate from'faces '
and may be shed directly into surface waters or indirectly
through surface water run-off or drainage from polluted soil
where conditions would favor development of infective larvae.
In order to infect, a hookworm larva normally must penetrate
the host's skin, most commonly as a result of walking bare-
foot on polluted soil. However, infection can also be
acquired through penetration of oral mucosa by waterborne
larvae which, in the case of Strongyloid larvae, have been
found in drinking water as well as on leafy vegetables
(Strong, 1944). The guinea worm Dracunculus medinensis is a
nematode that develops as an infective larva in copepods,
usually Cyclops. Humans can become infected by drinking
water containing the larvae, and eggs may subsequently exit
through the feet as a result of wading. D. medinensis is
not endemic in the U.S.
Due to the complex life cycle of .trematodes, infection
occurs and is endemic only in areas where the appropriate
intermediate snail host is present. Most cestodes that
parasitize humans require vertebrates (cattle,swine, fish) as
intermediate hosts. Cestode life cycles are relatively
simple and the organisms are more .widely disseminated in
nature, particularly Hymenolepis nana and H. diminuta. H_.
nana is a tapeworm of man, but can incidentally infect r₯ts
and mice; 1H. diminuta is a tapeworm of rats and mice, but
occasionally is found in humans. Water contaminated with
human feces or rat droppings may serve as the vehicle of
transmission, and athe fully embryonated eggs are infective
upon ingestion.
Eggs of most parasitic helminths are large, about 50 urn
in diameter or larger, and being heavier than water they are
easily removed by standard water treatment processes that
include storage, coagulation, sedimentation, and filtration
[See Sections E.I through 3]. However, the infective larvae
of hookworms and threadworms are highly motile and can
sometimes move through sand filters. Consequently, sand
filtration and flocculation have not proved to be adequate
safeguards against these larval forms. They are also much
more resistant to disinfectants than are enteric bacteria,
viruses, or even protozoan cysts. If such larvae are sus-
pected in drinking water, heating to boiling is the only
safe measure. On the other hand, the chances of obtaining a
-------
116
serious worm infection via drinking water are extremely slim
considering that: (1) eggs and most larvae settle out in
water; and (2) each egg or larva develops into only one worm
in the hostj and this inability to replicate, coupled with
the enormous dilution effect in water, makes transmission
through drinking water highly unlikely. Such conditions of
relative safety do not, however, apply to wells polluted
with surface water nor to systems that use chemical dis-
infection as the sole treatment. .
Since hookworm and threadworm larvae are about the same
size as the harmless free-living adult nematodes (0.75 to
1.0 mm long, 40 to 60 um wide) and since these latter are
frequently found in water supplies and breed well in slow
sand filters, there are opportunities for mistaken identity.
The harmless, free-living nematodes include members of the
genera Cheilobus, Diplogaster, Diplogasteroides, Trilobus,
Aphelenchus, and Rhabditis. They normally inhabit the soilj
and if present in wastewater influent, will flourish during
aerobic sewage treatment such as by trickling filters or acti-
vated sludge. Effluents from such treatment plants, upon
entering receiving waters, may contain large numbers of
these organisms. Sand filters used to process raw water for
drinking purposes may become infested with these organisms^
which ultimately will be carried through the distribution
system to the consumer. Nevertheless, it should be borne in
mind that these nematodes are a subject of concern only
insofar as they reportedly produce a gummy substance, small
quantities of which confer an unpleasant taste to finished
water.
The only effective way to remove both the parasitic
hookworm and threadworm larvae and the free-living nuisance
nematodes from water is to chlorinate heavily, allowing at
least 2 to 4 h contact time to immobilize them prior to
flocculation and sand filtration. Nematodes are detected by
filtering the water sample through a woven nylon strainer of
25 to 30 um pore diameter, then pipetting in phosphate
buffer to dislodge them and finally examining the sample
microscopically.
d. Gastroenteritis of Undetermined Etiology. Gastro-
enteritis, resulting in nausea, vomiting, and diarrhea,
accounts for over half the reported cases of waterborne
disease. The cause of illness cannot be determined in the
majority of these cases.
-------
117
Many causes of gastroenteritis are possible: among
them, only the infectious agents are transmitted, with any
frequency, by contaminated water or food. Fecally contami-
nated water is a likely source of these agents since diar-
rhea is one of the major symptoms of gastroenteritis.
Indeed, sewage contamination of drinking water supplies has
produced the largest number of documented cases of water-
borne disease. Such contamination often results in explo-
sive outbreaks of acute gastroenteritis affecting a large
percentage of the individuals who ingest the water.
When gastroenteritis of undetermined etiology occurs in
otherwise healthy individuals, it is generally self-limiting;
and in the majority of cases, there is no need for medical
attention. However, outbreaks that occur among malnourished
or otherwise weakened persons can lead, especially when they
occur repeatedly, to more serious consequences; as a result,
this disease is considered to be a major cause of infant
mortality, particularly in the tropics.
(i) Categories and Properties of Agents of Gastro-
enteritis. Many of the infectious agents which sometimes
cause waterborne gastroenteritis are discussed in other
sections of this report. Those most often incriminated are
enteric bacteria of the genera Salmonella and Shigella [See
Sections B.l.a(i) and (ii)], perhaps because these are often
the only organisms sought in laboratory investigations of
waterborne outbreaks. More extensive investigations may also
include testing for enteric viruses [See Sections B.l.b and
C.2.b], enteropathogenic E. coli [See Section B.l.a(iv)]
and the cysts of Entamoeba and Giardia [See Section B.l.c].
Nonetheless, even when laboratories do thorough tests of
diarrheal stool specimens and water samples they usually
fail to identify an etiological agent. The illness in such
outbreaks is often called "acute infectious nonbacterial
gastroenteritis" (AING) if bacteria-free fecal extracts from
affected individuals cause gastroenteritis when ingested by
volunteers. The agents causing AING therefore were often
presumed to be viruses, although laboratories were (and
still are) unable to cultivate these presumed viruses. It
is only recently that means have been developed for iden-
tifying some of the viruses responsible for AING. Most of
the viruses that can be isolated from feces by standard
cultural methods are not agents of gastroenteritis, but may,
however, cause a variety of other symptoms [See Section
B.l.b].
-------
118
(i.l) Parvovirus-Like Agents of Gastroenteritis. A
breakthrough in identifying AING agents came in the early
1970's when viral particles were detected in connection with
a gastroenteritis outbreak that had a 50 percent attack rate
at a school in Norwalk, Ohio, U.S. (Kapikian, et al,., 1972).
The outbreak was characterized by nausea, vomiting, and
stomach cramps, and symptoms typically persisted for 12 to
24 h. Secondary cases occurred in 32 percent of family
contacts.
The viral particles were seen by means of immune elec-
tron microscopy (IEM) in fecal extracts from a volunteer who
had ingested stool filtrates from an outbreak victim. The
well from which the school's water was obtained was sus-
pected, but could not be definitely incriminated because the
IEM method is not sensitive enough to detect viral particles
in environmental samples. The virus was called the Norwalk
agent and was characterized as a parvovirus-like 27 nm
particle similar in appearance, but unrelated serologically,
to the virus of hepatitis A.
Further investigation has revealed similar particles in
stool specimens from two separate family outbreaks of AING,
one in Montgomery County (U.S.) and the other in Hawaii
(U.S.) (Thornhill, £t aJL. , 1977). The Montgomery County
(MC) agent was found to be antigenically similar and the
Hawaii (H) agent dissimilar to.the Norwalk agent, suggesting
that more than one type of parvovirus-like agent may be
associated with AING. This conclusion is supported by work
conducted in England where parvovirus-like particles desig-
nated "W", Ditchling, and cockle agents were associated with
two distinct gastroenteritis outbreaks (Appleton and Buckley,
1977; Appleton and Pereira, 1977). These viruses are simi-
lar in morphology, size, and density, but are antigenically
dissimilar to the Norwalk agent. They show considerable
similarities to enteroviruses- o-nd may soon be classified
as such. The cockle agent is of special interest because
shellfish, the probable vehicle of transmission, were taken
from water where sewage pollution was known to occur
(Appleton and Pereira, 1977).
(i.2) Rotavirus. A second category of viral agents
associated with gastroenteritis also was identified in the
early 1970's. Reovirus-like particl.es were visualized in
fecal extracts from infants and young" children experiencing
diarrheal illness (Bishop, et al., 1973). They were also
reported by Riley to be the cause of several non-bacterial
diarrheal outbreaks (reviewed in Craun, 1979). These par-
ticles have been found worldwide in association with diar-
rheal disease in infants and young children. They have been
-------
119
isolated from as many as 75 percent of children with gastro-
enteritis (Madeley, 1979). Infections appear to be more
common during colder winter months. The incubation period
is about 48 h and the disease is short-lived; viruses are
seldom isolated after the eighth day. Infection normally
leads to long-lived immunity.
These viruses are now generally called rotaviruses,
although they have also been referred to in the literature
as orbiviruses, duoviruses, reovirus-like agents, and infan-
tile gastroenteritis viruses. The human rotavirus, like the
parvovirus-like agents, has not yet been isolated in labora-
tory cell cultures. However, it is often excreted by in-
fected individuals in concentrations sufficient to be de-
tected by direct electron microscopic examination of fecal
extracts [See Section B.l.b].
(i.3) Other Viruses and Bacteria Associated with
Gastroenteritis. In addition to parvovirus-like particles
and rotaviruses, other yet unidentified viruses have been
visualized by electron microscopy of diarrheal stools.
These include particles resembling coronaviruses, myxoviruses,
astroviruses, adenoviruses, and caliciviruses, ranging in
diameter from 28 to over 100 nm.
Electron microscopy studies have revealed the presence
of adenoviruses in feces, while normal cultural methods
yielded negative results. The presence of these adeno-
viruses has occasionally been associated with mild outbreaks
of gastroenteritis (Flewett, et al., 1975). However, it is
difficult to assess the extent to which the illness can be
attributed to the virus because these viruses have commonly
been isolated, using both cultural methods and electron
microscopy, from healthy persons. It is not know^why some
of these viruses apparently cannot be cultured, nor is it
clear whether or not the nonculturable strains are of dif-
ferent serotypes.
There is still considerable confusion in identifying
viruses in feces, especially when variations in methods,
possible artifacts, and the presence of bacteriophages are
considered. Furthermore, all of these viruses" have" been
isolated from healthy persons as well as from those witri
gastroenteritis and, on some occasions, more than one virus
has been present. Therefore, the role of these agents in
AING is unclear at present, but it would appear that AING
might have several viral causes.
-------
120
Investigations by the U.S. Public Health Service on
waterborne outbreaks of acute gastroenteritis have indicated '
that parvovirus-like agents are most likely responsible for
many of these episodes in the U.S. However, the nonbac-
terial nature of AING outbreaks is sure only in those in-
stances where the disease has been transmitted to volunteers
in bacteria-free filtrates of diarrheal stool. The recent
association of Yersinia enterocolytica [See Section B. l-.a(ni) ]
and of Campylobacter fetus [See Section B.l.a(viii)] with
waterborne gastroenteritis outbreaks in the U.S. (Craun,
1976? Center for Disease Control, 1978b) must also be kept
in mind. These bacteria are more difficult to cultivate in
the laboratory than are the more familiar enteric bacterial
pathogens and tests for their presence are not routinely
conducted. For this reason they may sometimes be respon-
sible for outbreaks that are mistakenly designated AING.
(ii) Conclusion and Recommendations. Now that elec-
tron microscopic and serologic procedures are available to
detect them, it is clear that there are indeed viruses
capable of causing AING. As more information is forth-
coming, many once unidentified infectious agents will proba-
bly be classified as identifiable viruses. For example,
infections now known to be caused by rotaviruses can no
longer be referred to as "gastroenteritis of undetermined
etiology." The incidence of gastroenteritis viruses in
humans and in water and wastewater might be determined
better if laboratory means of cultivating these agents could
be developed. Meanwhile, culture procedures capable of
isolating the less-frequently suspected bacterial causes of
gastroenteritis should be employed in investigating all
waterborne outbreaks.
2. Sources of Waterborne Pathogens
The reservoirs for microbiological diseases trans-
missible to man, consist of man himself as well as both
domestic and wild animals (including some poikilotherms).
The microorganisms responsible for causing disease generally
are excreted in the feces or urine, whereupon they may gain
access to water. If drinking water treatment is inadequate
or lacking altogether, these organisms may pass freely into
water en route to the consumer, thereby engendering a risk of
infection and possibly disease [See Section E].
-------
..'' 121
The human intestines are the .source of most of the
bacteria [See Section B.1.a], virtually all of the viruses
[See Sections B.l.b and B.l.d], and the great majority of
protozoan and metazoan parasites [See Secton B.l.c] which
can be transmitted through water to man. Nonhuman sources
are significant for a few waterborne pathogens, as is seen
in Table B.2-1. ,
a. The Role of Man. Pathogenic microorganisms are
excreted not only by individuals with clinical symptoms, but
also by asymptomatic carriers. The average number of
individual humans excreting Salmonella at any given time
will vary from < 1 percent to 3.9 percent according to
studies conducted in several countries throughout the world
[See Section B.l.a(i)]. In a great number of cases, these
people were outwardly healthy yet excreted about 10 organ-
isms per gram of feces.
Enteric viral infections are common in children, espe-
cially those under five years of age. The high frequency of
occurrence is obscured by the fact that only a small percent
of those infected manifest serious disease. Prior to the
introduction of the poliovirus vaccine, it was estimated
that only one of every 1,000 children infected with the wild
virus contracted paralytic disease (Melnick and Ledinko,
1951). A similar ratio probably exists for many of the
other enteric viruses, therefore, an indication of the
prevalence of enteric viruses in a community must be deter-
mined from viral excretion data. A recent study conducted
in Seattle, Washington (U.S.) indicated a rate of virus
isolation in feces (excluding poliovirus) of 2 to 4 percent
among family members selected for the study (Cooney, et
al. , 1972).
Poliovirus isolates were eliminated from the evaluation
because they were assumed to be primarily of vaccine origin.
Gelfand and colleagues (1957) in earlier work in Louisiana
found that as many as 16 percent of the healthy children
included in their study were excreting other than polio-
viruses during the summer months. They also found excretion
rates to be inversely related to socio-economic status.
Interestingly, Chin and coworkers (1967) were able to demon-.
strate the presence of vaccine strain poliovirus in sewage
when as few as 0.3 percent of the local population has received the
live-virus vaccine shortly prior to the examination of the
sewage. The Seattle study also found that children less
than one year of age had an average of 1.5 enteric viral
infections per year which dropped to 0.58 for those two to
-------
TABLE B.2-1
SOURCES OP WATERBORNE PATHOGENS
MAN
Agent
Patients
Typical Average - Multiplication
Healthy per g of Feces fr. Outside
Carriers an Infected Person Human Host
Major Animal
Reservoir
Other than Man
Domestic
Animals
WILD-ANIMALS
Reptiles,
Mammals Fishes,
& Birds Shellfish
BACTERIA
Salmonella typhi
Other salmonellae
Shigella
10
10
10
f(food)
f(food)
+(food)
Yersinia enterocolitica + +
Enteropathogenic E. coli + +
Francisella tularensli + ?
Leptospira + ?
Vibrio cholerae + +
Campylobacter fetus + ?
Pseudomonas aeruginosa + +
Other opportunistic
pathogens + +
VIRUSES
Enterovirus + +
Hepatitis A + +
Rotaviruses * + ?
PARASITES
Entamoeba histolytica + +
Giardia lamblia + +
Naegleria fowleri -(?) -
Acanthamoeba -(?) +
Nematodes/Helminths + +
10" ?(food)
108 +
?{urine)
10 unlikely
If I
106-108 +
6
109 "
10 ~
10*
105
? , ?
10-10^ +(*)
+ + + -
+(?) - +(?)
:{7) ; ; !?
-(water) - - -
H
to
M
± In some cases Yes, in some cases No.
* = Presence of an intermediate host.
-------
123
->
five years of age and fell considerably lower for those over
five [See also Sections B.l.b and B.l.d].
Protozoan cysts are passed in the feces of asymptomatic
carriers or patients with chronic or mild cases of amebiasis.
Patients with acute dysentery shed trophozoites that are
non-infectious. Asymptomatic carriers of the protozoon _E.
histolytica usually outnumber those with symptomatic ame-
biasis. Tn~ the U.S., where the 1976 carrier rate for E.
histolytica was 0.6 percent (Center for Disease Control,
1978a),there is evidence of one clinical case for every 500
carriers. Horster (1943) reported that many German soldiers
who developed amebic, dysentery in association with bacillary
dysentery [See Section Ill.b.l.a(ii)] in North Africa had
been 13. histolytica carriers for years before in Germany.
LikewTse, Wenyon (1947) hypothesized that the 1933, Chicago
epidemic of amebic dysentery was, in fact, an outbreak of
bacillary dysentery in a population with a high rate of E_.
histolytica carriage. The source, in this outbreak, was
traced to cross-connections between sewage and water lines
in a hotel (Bundesen, e_t. al_., 1936).
Among populations exposed in three separate waterborne
outbreaks of amebic dysentery, the percentage of asympto-
matic persons passing cysts was quite high: 57 percent of
those exposed were found to be carriers in the 1953, South
Bend, Indiana (U.S.) epidemic (Of futt, e_t al_. , 1955); 62.9
percent in the Chicago epidemic (Hardy and Spector, 1935);
and 50 percent in the 1947, Tokyp Mantetsu-apartment-building
outbreak (Ritchie and Davis, 1948). Of these subclinical
carriers, only a small fraction developed clinical symptoms
[See also Section B.l.c(i)].
G_. Iambiia occurs throughout the world, especially in
areas of poor sanitation. In 1948, G. lamblia was estimated
to be present in 7.2 percent of the world's population,
based on 84 stool surveys. Giardiasis is endemic and primar-
ily asymptomatic in the Canadian arctic and has been reported
in as high as 72 percent of Greenla'hd populations. Cur-
iously, Canadian travelers overseas have a relatively high
rate of frank giardial illness and experienced a 59 percent
morbidity in a recent outbreak among tourists to Leningrad
(Center for Disease Control, 1976a). Whether those tourists
who were stricken did not have prior exposure, or whether
other factors such as variation in virulence were involved,
remain matters for speculation and further study. Leningrad
tapwater was implicated in a series of outbreaks among U.S.
tour groups which visited Leningrad between 1969 and 1973,
and which experienced a 23 percent (324 ill) attack rate
[See also Section B.l.c(ii)].
-------
124
It is not known whether there are carriers for Naegleria
fowleri [See also Section B.l.c(iii)]. Acanthamoeba species
have been recovered from the respiratory tract of healthy
individuals and have been found in eye infections (Sotelo-
Avilo, 1974; Nagington, 1974). They have also been isolated
from throat cultures of asymptomatic children home from camp
where activities included swimming [See also Section
B.l.c(iv)]. Man can become an intermediate host in the life
cycles of nematodes and other helminths and be severely or
minimally infected, depending on geography and climate [See
also Section B.l.c(v)].
b. The Role of Domestic Animals. Many waterborne
human pathogens can be excreted by either sick or apparently
healthy animals, such as domestic pets or farm animals.
Excellent reviews are given by Reasoner (1977, 1978).
The incidence of Salmonella in dogs ranges from 6.9 to
18 percent; cats, 6.0 to 13.6 percent; pet turtles, 7.3 to
20 percent; and pet birds, 8.3 percent. Studies of large
groups of clinically healthy farm animals indicate that
Salmonella carrier rate may vary from 1 to 15 percent for
sheep, 0.1 to 17 percent for cattle, 2 to 35 percent for
pigs, 2 to 4 percent for goats, 22 to 29 percent for chickens,
and 0.7 to 37 percent for ducks and geese.
It is well known that enteropathogenic 13. coli can
cause intestinal disturbance in domestic animals, especially
in pigs; it also can be present without any symptomatology.
This organism's ability to exchange plasmidic resistance
factors with pathogens such as Salmonella and Sh-'gella
heightens the disease risk for individuals in close prox-
imity to domestic animal carriers, but there is at least
some evidence that enteropathogenic IS. coli strains do not
pass readily from one host species to" another [See also
Section B.l.a(iv)].
Yersinia can'be derived from pets, cattle, and more
frequently, from pigs [See also Section B.l.a(iii)].
Prancisella tularensis has been identified in dogs, cats,
sheep and horses[See also Section B.l.a(v)]. Studies
reviewed by Reasoner (1977) reported the isolation of Lepto-
spira from pet dogs (42 percent), cattle (1.2 and 59 percent),
pigs (43 and 70 percent), horses (5 - 74.6 percent), sheep (7
and 65 percent), and goats (25 percent). These animals can
become carriers after inapparent or acute infections and
shed the bacteria in their urine [See also Section B.l.a(vi)].
Human carriers of Vibrio cholerae serve as the source of new
-------
125
cases of cholera; and the role of animals, if any, seems to
be minimal [See also Section B.l.a(vii)]. It has been
suggested, however, that in the absence of human carriers,
intermittent excretion of vibrios by cows and chickens may
be a source of new infections. Campylobacter fetus can
originate from cattle, sheep, pigs, and poultry [See also
Section B.1.a(viii)]. Likewise, all opportunistic patho-
gens, especially the Enterobacteriaceae, can be found in
domestic animals [See also Section B.l.a(ix)].
. Man is thought to be the only important reservoir for
members of the human enteric viruses. However, viruses that
appear to be human viral serotypes have been isolated from
the feces of domestic animals (Grew, et al., 1970; Graves
and Oppenheimer, 1975). It is assumed that these animals
are ony passive shedders of viruses ingested via grossly
contaminated food or water and that the viruses do not
multiply in these hosts. Rotaviruses may be an exception to
this viewpoint. Serologically related rotaviruses have been
isolated from the feces of children, calves, piglets, mice
and foals having acute gastroenteritis (Woode, et al.,
1976). The role of these animals in the natural trans-
mission of rotavirus disease in man has not been explored.
In fact, the role of domestic and wild animals in the trans-
mission of enteric viral disease to man has not been exten-
sively studied, and further work in this area is needed.
Horses, cows, pikas, and sheep were found to be in-
fected with Giardia species, according to unpublished animal
surveys conducted in Colorado in 1975 and 1976. Since only
one of two Giardia species infects humans, and since sources
for these animal infections were undetermined, data from
this survey raise more questions than they answer, such as:
- Which Giardia species was involved?
- Was there one source (perhaps domestic sewage)
or multiple sources?
- Were there any cross-transmissions?
G. lamblia of human origin has been shown to be highly
infectious for dogs (Craig and Faust, 1964), and has been
successfully transmitted from man to rats (Hegner, 1930).
Conversely, the organism appears to be transmissible from
dog to man (Padchenko and Stolyarchuk, 1969). Giardiasis
has recently been found to be apparently prevalent in various
species of psittacine birds, but any cross-infectivity to
humans has yet to be determined (Panigrahy, et al., 1979).
-------
126
c. The Role of Wild Animals. Wild animals can serve
as important reservoirs of some waterborne pathogens. For
example, the carrier incidence of Salmonella in grain fed
pigeons ranges from 1.4 to 3.8 percent. The occurrence of
Salmonella in 6.9 to 26 percent of various species of sea-
gulls may be directly related to their scavenger activities
in garbage refuse and ocean dumping sites [See also Sections
B.l.a(i) and E.l]. Among cold-blooded animals notably
snakes, turtles, and lizards a high incidence of Sal-
monella (12.7 to 46 percent) may be due to feeding on dung-
associated insects or on infected rodents or small birds.
Y. enterocolitica has been identified in several wild
mammals including deer, hare, primates, rodents, foxes,
rabbits, and wild birds. Contact with feces appears to be
the major transmission route. _F. tularensis has been
identified in rodents, rabbits, voles, muskrats, beaver,
deer, foxes, moles, wild birds, and many species of arthro-
pods. There is no good evidence that animal hosts maintain
carrier states) however, Bell (1975) demonstrated that
rodents with a tularemia-induced nephritis chronically shed
bacilli a matter of great theoretical interest. Lepto-
spira interrogans has been identified in mice, opposums,
field voles, badgers, rats, and squirrels. They constitute
a major reservoir for this organism. It has been suggested
that contaminated shellfish have served as sources of human
infection with Vibrio cholerae (reviewed by Zwadyk, 1976).
Campylobacter fetus has been isolated from birds and rats.
As yet, there are insufficient data for ascertaining the
role of wild animals in disseminating opportunistic patho-
gens. It appears that wild animals are not involved in the
spread of human enteric viruses (including reoviruses and
rotaviruses). Giardia lamblia has been recovered from deer,
elk, marmots, raccoons, coyotes, beavers, and muskrats.
3. Persistence and Death of Pathogens in Water
Water is a hostile environment to most human pathogens
which, once introduced, will die at varying rates. This
discussion will deal mainly with factors that influence this
death rate. The available data are sparse and scattered in
the literature. Moreover, much of the data cannot be di-
rectly compared because the experimental protocols differ.
An excellent review on the spread of pathogens through water
was prepared by Geldreich (1972). As the present work is
mainly directed towards drinking water, attention will be
directed primarily to intestinal pathogens. In addition,
-------
127
some data on indicator -bacteria will be included to eluci-
date mechanisms of survival or inactivation.
a. Death of Bacteria. Before discussing the various
factors that influence the death of pathogens in water, it
will be necessary to consider the phenomenon of death itself.
This is best done by citing Postgate (1976) who, in a very
instructive paper, pointed out that death of a microbe is
not the result of aging as it is in macrobes. A microbial
cell divides and its contents live on indefinitely; death
will occur only as the result of an environmental stress.
Death is usually diagnosed retrospectively. That is, a
population is exposed to a recovery medium, incubated, and
those cells that do not divide to form progeny are con-
sidered to be dead. A marginal state exists between life
and death, wherein a cell will grow in one recovery medium,
but will fail to grow in another. As a consequence the
bacterium will be counted as dead, or alive, depending on
the experimental procedure.
In food microbiology, the term "sublethal injury" has
been introduced to describe this phenomenon. General prin-
ciples for the recovery (resuscitation) of sublethally
injured bacteria have been developed and are reviewed by van
Schothorst (1976). Several effects, not directly related to
the aqueous environment, will influence the observed death
rates in experimental systems. First, a dense population
will die more slowly than a sparse one. This is partly
because dying bacteria leak small molecules which not only
protect neighbors from stress, but can actually be metabo-
lized for further growth. Second, the use of inappropriate
diluents can sensitize microorganisms to experimentally
applied stresses. Even mild stresses which would normally
elicit no reaction could, after an organism was exposed to
an overt stress, have a marked effect. For example, E.
coli became sensitive to warming at 50°C for 5 min after
being starved for two weeks or more in stream water (Klein
and Wu, 1974). This suggests that the pour plate method,
which entails a certain amount of heat shock when the melted
agar contacts the organisms, may be unsuitable for the
enumeration of starved bacteria. On the other hand, no
increased sensitivity to heat was observed for Aeromonas
after starvation. Third, the use of different media, incu-
bation temperatures, etc., will result in the recovery of a
different proportion of the sublethally damaged bacteria.
Also, the presence in the medium of a limiting nutrient that
was lacking in the environmental system may lead to death in
part of the population.
-------
128
t>. Factors that Influence the Death of Bacterial
Pathogens"It has long been recognized that bacterial death
is influenced by a number of factors (Rudolfs, et al.,
1950a), some of which will be considered here.
In general, death rates increase with increasing
temperatures (McFeters, et al., 1972; Mitchell and Starzyk,
1975; and Verstraete and Voets, 1976). Death rates in
summer are higher than in winter, not only because of higher
temperatures but possibly from the action of sunlight. Even
though close correlations between the death rate and insola-
tion or ultraviolet solar radiation have been found, it is
impossible to distinguish between the direct effects of
sunlight and the many indirect effects such as heating,
promotion of algal growth, etc. (Verstraete and Voets, 1976;
Calkins, e_t al. , 1976) [See Section E.I].
It is well known that pathogens prefer conditions close
to neutrality and survive best between pH values of six and
eight. High acidity or alkalinity are deleterious, and
sudden changes in pH from neutrality to either acidity or
alkalinity accelerate the death rate considerably. Hanes
and coworkers (1964) performed detailed studies and con-
firmed the earlier evidence that indicator bacteria survive
better under aerobic than under anaerobic conditions.
The role of nutrients in bacterial survival in water is
complex. High nutrient levels (as would occur in sewage
effluent or fecal suspensions) may prolong survival (Tannock
and Smith, 1971) or even promote bacterial growth. On the
other hand, nutrients can also encourage the growth of
bacterial predators (Kittrell and Furfari, 1963). Low nu-
trient levels will generally result in bacterial reductions
as has been shown in numerous papers. Wuhrmann (1972)
stated that pathogen multiplication is negligible in environ-
mental waters where nutrients are sparse, although Cherry
and coworkers (1972) found Salmonella in 44 percent of
apparently unpolluted waters. This could have been a conse-
quence of the high Salmonella carrier rates in birds and
wild animals [See Sections B.l.a(i) and B.2].
The ecosystem of a receiving water has an inherent
capacity to restore conditions to their previous norm. One
important mechanism is predation by protozoa such as Vexil-
lifera (Mitchell, 1972) and Colpoda steinii (Drake and
Tsuchiya, 1976), and another is competition and antagonisms
from the indigenous aquatic flora. Bacteriophages are
frequently mentioned as important contributors to bacterial
-------
129
die-oftf but no data have been published to confirm this
hypothesis. Algae and cyanophages can elso enhance bac-
terial death by a number of poorly understood mechanisms,
including production of antibacterial substances (also true
with higher plants), changes in pH, and limitation of carbon
dioxide (reviewed by Davis and Gloyna, 1972). Filtration or
autoclaving (the more efficient of the two), of raw river
water removed its bactericidal effect on Shigella flexrieri
(Mohadjer and Mehrabian, 1975). Another indirect indication
for the role of predators is the influence of stream mor-
phology. In general, death rates in small streams (with a
great surface-to-volume ratio) will be greater than those in
large streams (Kittrell and Furfari, 1963; Wuhrmann, 1972).
c. Survival of Bacterial Pathogens in Water. The
following data were obtained by investigators using-dif-
ferent experimental procedures and are not meant to be
construed as a basis for quantitative comparisons between
pathogens.
The survival of Salmonella after having entered a
receiving water depends upon several conditions. Salmonel-
lae Were regularly detected in surface water up to .25 km
downstream from a wastewater treatment plant, but never at
sample sites 1.5 to 4 km downstream (Kampelmacher and van
Noorle Jansen, 1976). In this case, dilution was the main
factor in reducing bacterial numbers. Salmonellae, trans-
ported by stormwater through a wastewater drain at a Uni-
versity of Wisconsin experimental farm, were isolated
regularly .8 km downstream of a swimming beach (Claudon, et
al., 1971). .
Laboratory experiments suggest that cold water and
available nutrients in wastes are critical to Salmonella
persistence. When £^. typhimurium was held in samples of
river water, 90 percent reduction took place in 13 days at
5°C, five days at 20°C, and two days at 30°C (Ahmed, 1975).
Ruys (1940) isolated Salmonella typhi from water in winter
but not in summer. Studies of ice covered rivers showed
survival times of from four to seven days (Spino, 1966;
Davenport, et al., 1976).
Salmonellae were isolated during September from the Red
River of the North, 36 km downstream of sewage discharges
from Fargo, North Dakota, and Moorhead, Minnesota (U.S.
Department of Health, Education, and Welfare, 1965). By
November, salmonellae were found 100 km downstream of these
-------
130
two cities. High levels of bacterial nutrients entered the
river under ice cover during a January peak in sugar beet
processing and accompanying waste discharges. Salmonella
strains were then isolated 118 km downstream (or four days'
flow time) from the nearest sources of warmblooded animal
pollution. Similarly, water samples taken from the Vltava
and Danube rivers, polluted with sugar beet wastewaters,
supported Salmonella growth even when diluted 1,000-fold at
20°C and 30°C (Adamek, et a^., 1977).
On the other hand, Shrewsbury and Barson (1952) con-
ducted experiments which showed that Salmonella survives
longer in pure water than in water contaTnTngTdissolved
inorganic substances. They inoculated 5 x 10 and 109 S.
typhi Per ml into tapwater, distilled water, and normal"
saline. Samples were kept in daylight at temperatures
between 7.2 and 35°C (mean, 21.1°C) under aerobic conditions.
Viable organisms were recovered for up to 211 days in tap '
water, up to 443 days' in distilled water, but only up to 153
days in normal saline.
Shigella flexneri survived more than a month at -8°C in
tap water, but untreated river water was bactericidal for
S. flexneri at 4 to 6°C and 23 to 25°C (reviewed by Reasbner,
1977).
Whereas most bacterial pathogens appear to show some
similarity in survival rates in water, the stability of Y.
enterocolitica in water differs considerably from that of"
the other Enterobacteriaceae. Y. enterocolitica is better
able to survive in nature than are other human pathogens
because it prefers lower temperatures (/^O20°C optimum). It
has been hypothesized (Kristensen, 1977) that the ability of
Y. enterocolitica to multiply at 4°C may explain its appar-
ent increased occurrence in industrialized countries where
refrigeration is used extensively. Dominowska and Malottke
(1971) found that Y. enterocolitica survived longer in autumn
and winter than in spring and summer, and that it survived
for as long as 157 days in water free of other microor-
ganisms. Krogstad (1974) found that during the first few
hours and days in water, Y. enterocolitica and' E. coli were
reduced at nearly the same rate, but that Y. enterocolitica
survived longer than E_. coli at low temperatures^ In sea-
water held at 15°C, the survival time was nearly the same "
for Y. enterocolitica serotype 0:3, Y. enterocolitica of the
type frequently isolated from water (and showing an atypical
rhamnose-positive reaction), and E. coli (Kristensen, 1977)
-------
131
But, at 6°C, both Y. enterocolitica types persisted con-
siderably longer than E. coli. Schillinger and coworkers
(1978) found that Y. enterocolitica survived longer, com-
pared to E_. coli, Tn oligotrophic stream water at tempera-
tures of "5.0 to 8.5°C. On the other hand, Y. enteroco-
litica died at a faster rate than E. coli in a continuous
flow of chlorinated tap water.
Y. enterocolitica has also been shown to grow under low
nutrient conditions.Highsmith and coworkers (1977) re-
ported growth of Y. enterocolitica in sterile distilled
water at 4, 25, and 37°C, but not at 42°C. Optimum growth
was at 25°C. After 18 months of storage at 4°C, the flasks
still contained 10 cells per ml.
According to Hopla (1974), whether or not tularemia
organisms persist in water over long periods has not been
satisfactorily resolved. Hopla described tests performed by
Bell wherein he found that £. tularensis isolated from
streams grew poorly or not at all when incubated at stream
temperatures. Parker and coworkers (reviewed in Hopla,
1974) recovered £. tularensis over a 16-month period in
certain areas of"northwestern U.S. and postulated that
specific properties of mud and water enhanced the organism's
survival potential.
Enteropathogenic 13. coli, unlike the non-pathogenic 1C.
coli types, does not appear to survive long in water; whereas
E. coli can persist in water for over 60 days, enteropatho-
genic E. coli dies within ten days, and this may be why it
is very rarely isolated from environmental waters (Mailer,.
1967).
Survival of Leptospira is enhanced by temperatures of
22°C or above and a neutral to slightly alkaline environment
[See Section B.l.a(vi)]. Vibrio cholerae survives best at
low temperatures, but only at a neutral or alkaline pH [See
Section B.l.a(vii)]. There is no information as yet on the
survival of Campylobacter [See Section B.l.a(viiiH in
water, nor are "there any precise quantitative data- for sur-
vival of opportunistic pathogens in water [See Section
B.l.a(ix)] .
The most promising types of experiments comparing
pathogen survival in water appear to be based on the mem-
brane filter chambers described by McFeters and coworkers
(1974). Table B.3-1 summarizes results from these experi-
ments on survival of different bacteria in well water at 9.5
-------
132
TABLE B.3-1
COMPARATIVE DIE-OFF RATES OF
FECAL INDICATOR BACTERIA AND ENTERIC PATHOGENS
Bacterial Species
Half-Time (Hours)*
Aeromonas
Shigella
Enterococci
Salmonella enteritidis ser. paratyphi D
Coliform bacteria
Salmonella enteritidis ser. paratyphi A
Salmonella enteritidis ser. typhimurium
t
Streptococcus equinus
Vibrio cholerae
Salmonella typhi
Streptococcus bovis
Salmonella enteritidis .ser. paratyphi B
no death
22 - 27
22
19
17
16
16
10
7
6
4
2
*The time required for a 50% reduction in the initial
population.
Taken from: McFeters, et al., 1974.
-------
133
to 12.5°C. In particular, the longevity of Shigella species
is striking and should be of concern to water hygienists.
In recent years, several workers have shown that bac-
teria can sustain sublethal injury after starvation in the
aqueous environment. Hoadley and Chang (1974) found counts
of P. aeruginosa, S_. faecalis, and E. coli on selective
media to be consistently lower when these organisms were
suspended in tapwater than when they were suspended in
stream water or 0.1 percent peptone water (both containing
A-"550 mg per 1 of organic carbon). Bisonnette and coworkers
(1975; 1977) studied the effect of environmental stress
using'membrane filter chambers. Both survival and death of
E. coli and S_. faecalis were found to be dependent on water
quality as reflected by differences between eight sampling
points. In general, the percentage of sublethally injured
bacteria increased with time. Resuscitation for 2.5 to 3 h
in enriched Trypticase Soy Broth was sufficient to restore
the cell's ability to grow in selective media. Andre and
coworkers (1967) noted that both colony size and morphology
of Salmonella and Shigella grown on SS agar changed with
increased exposure to water. Sublethal injury does not only
manifest itself in the growth characteristics of a bacterial
cell, but in many biochemical characteristics as well,
including reduced oxygen consumption and dehydrogenase
activity (Daubner, 1975). Oger and coworkers (1976) starved
different Salmonella serotypes in river water and found that
antigenic properties of the bacteria disappeared after two
weeks at a survival of approximately 0.5 percent.
d. Persistence of Viruses in Water. Virus persistence
in water has been the subject of a number of reviews.
Clarke and colleagues (1964) reviewed the early literature
that was published after the introduction of tissue culture
techniques. A more recent summary of results appears in
Table B.3-2.
There is no difference in opinion among the various
authors on the role of temperature: lower temperatures
favor persistence, whereas virus inactivation is accelerated
at higher temperatures. The following features also are
generally agreed upon: (1) viruses persist longest in
distilled or deionized water; (2) viruses persist longer in
autoclaved drinking water than in autoclaved river water;
(3) viruses persist longer in water heavily polluted with
sewage, in autoclaved and filter-sterilized river water, in
tap water, and in distilled water than they do in untreated
-------
TABLE B.3-2
PERSISTENCE OP SOME ENTEROVIRUSES IN WATER
Type of Water
Sea or estuary
River
Impounded fresh
Tap
Deionized or distilled
Viruses
Coxsackie Bl, B3
Echo 6
Polio 1, 2, 3
Coxsackie A2, A9, B3, B5
Echo 6, 7, 12
Polio 1, 2, 3
Coxsackie A9, B3
Echo 6, 7, 12
Polio 1, 2, 3
Coxsackie A2, A9
Echo 7, 12
Polio 1, 2, 3
Coxsackie Bl
Echo 6
Estimated Days before 99.9% Titer
Reduction at a Temperature of
4 -
90
30 -
30 -
10 -
15 -
19 -
6 -
5
21 -
98
85 -
140 -
_
-
6°C
88
130
75
60
75
18
42
52
130
168
15 - 16"C 20 -
8-14 2 -
15-16 4 -
8-15 2 -
8 2 -
15 3 -
8-45 3 -
3 -
4 -
3 -
15 -
10 -
95
- 5
14
258C
28
15
8
8
16
20
6
24
22
100
11
U)
Modified from Akin et al., 1977.
-------
135
environmental water, regardless of temperature; and (4)
t enteroviruses persist longer than E. coli in water.
Apart from the seeming paradox of prolonged survival in
water heavily polluted with sewage, the evidence suggests
that the cleaner the water, the longer the survival time
[See also Sections B.l.b and C.2.a to b] .
e. Survival of Parasites in Water. Recently, a
cesspool in Amersfoort (Netherlands) was uncovered that had
been in use at a monastery around the year 1600. The mate-
rial was almost free of bacteria, except for some anaerobic
sporeformers, but it contained large numbers of Ascaris and
Trichuris eggs (van Knapen, 1978). Although none of these
eggs were viable, this observation indicates that parasitic
eggs are very stable. The same holds true for parasites in
water, as concluded by Rudolf and coworkers (1950b). Both
protozoan cysts, such as those of Entamoeba histolytica, and
parasitic eggs can still retain their infectivity for weeks
or even months in water (reviewed by Geldreich, 1972) [See
also Section B.l.c].
f. Conclusion. The fate of pathogenic organisms in
water is subject to a great many influences including
temperature, nutrients, pH, indigenous and accompanying
flora, and a number of others. Furthermore, some pathogens
under certain conditions persist longer in water than the
standard bacterial indicators. Therefore, once a source
water has been contaminated, specific treatment to remove
pathogens is warranted; it is not enough to rely solely on
natural die-off during storage.
4. infectivity of Waterborne Pathogens
The presence of a viable, pathogenic microbe in drink-
ing water is always undesirable, but it does not always
guarantee that infection, and especially disease, will
result if someone drinks the contaminated water. There is a
paucity of information concerning the infectivity of water^
borne pathogens. Information that is available is mainly
based on controlled feeding studies of healthy unstressed
volunteers and probability of illness studies by Dudley and
coworkers (1976). Dudley's work indicated that if a person
were to swallow 506 S_. typhosa, there would be a 0.0051
percent probability that illness would ensue, and that if
-------
136
219 J3. typhosa were swallowed, the probability of illness
occurring would be .0003 percent.
_For the present purpose, an ideal study would have been
one in which a fully virulent infectious agent was adminis-
tered to a large number of human volunteers representing a
broad range of ages, if not states of health. The agent
should have been administered with drinking water and the
dose determined in terms of some absolute unit such as
viable bacterial cells, viral particles, or protozoan cysts.
Each recipient should have been observed carefully to deter-
mine whether infection (as contrasted to illness) occurred
and, if illness resulted, the length of the incubation
period and the duration and severity of the symptoms. A few
studies with human subjects may have met one of these stipu-
lations almost none seems to have met more than one.
Obviously, there is a great need for research in this area.
Unfortunately, it is extremely difficult to determine abso-
lute numbers of agents such as virus particles in an adminis-
tered dose; and it is virtually inconceivable that infants
or the elderly would be used in this type of research,
especially as recipients of virulent infectious agents.
The probability that a given agent will cause infection
if ingested with drinking water is, almost certainly, a
function of the dose; however, the mathematical nature of
the function is not known and may well not be the same for
all kinds of infectious agents. Mathematical rigor, in the
analysis of results of studies to date, is discouraged by
uncertainties in the measurement of administered dose. It
seems safe to say that the probability of becoming infected
is a function of dose ingested, but it is by no means clear
that the probability of becoming overtly ill is enhanced by
ingesting some dose greater than that which would have
produced infection. If an agent is competely attenuated,
no multiple of the infectious dose should produce disease.
If an agent is virulent, the probability that it will produce
disease may be a function of ingested dose, of host factors,
or of a combination of the two. Inasmuch as most studies
with human subjects have looked only for illness or only for
infection, depending on the agent studied, the gaps in
present knowledge are not surprising.
The data given in Table B.4-1 have been compiled by
Bradley and Feachem (1979) and are presented on the premise
that some information is better than none. The studies from
which these numbers derive are inevitably subject to the
limitations discussed previously, so the data must not be
-------
137
TABLE B.4-1
MEDIAN PERORAL INFECTIVE DOSE OF
WATERBORNE PATHOGENS CITED IN THIS SECTION
Campylobacter
Opportunistic pathogens
(Pseudomonas)
Quantity Needed to Infect
50% of Human Volunteers
BACTERIA:
Salmonella
Shigella ,
Yersinia
Enteropathogenic E. coli
Francisella tularensis
Leptospira
Vibrio
6
io2
6
10
s-
>io6
<102
10
/I
io4
or more
7
io7
6
10
VIRUSES:
Enteroviruses
Hepatitis A virus
Rotavirus
2
2
2
PARASITES:
Entamoeba histolytica
Giardia lamblia
10'
Taken f rom: Feachem 'et al> ] 978.
-------
138
taken too seriously either as absolute or relative pre-
dictors. A look at research on peroral virus infection
illustrates these points.
There have been a few published studies regarding the
infectivity of ingested poliovirus (reviewed by Safe Drinking
Water Committee, 1977), but none on pathogenicity. Each was
done with oral poliomyelitis vaccine virus and was subject
to the criticisms that: (1) the attenuated virus may not be
representative of "wild" types occurring in water and waste-
water; and (2) the quantities of virus administered to the
human subjects were measured in tissue culture infectious
doses, rather than in numbers of particles, so that all of
the results are expressed in relative terms. The results of
the various studies appear to differ by as much as a factor
of 100,000 in the relative infectivity of virus in the body
and in tissue cultures. Nothing in these findings rules out
initiation of an infection, either in a tissue culture or in
the body, by a single virus particle: they do suggest that
the probability of this happening is quite small in a tissue
culture and smaller still in the human body. There are no
data available on the relationship of viral dose to patho-
genesis.
The incidence of waterborne disease in most indus-
trialized nations appears to be quite low. The potential
practical use of further studies with human subjects would
be to allow more accurate assessment of the risk to public
health associated with a given level of a pathogen in drink-
ing water. In theory, an accurate set of risk assessments
could be used by a. government to array its public health
resources in a way that would minimize transmission of
infectious diseases by all routes, including the water
vehicle. However, it must be recognized that there are at
least two limitations to applying data on the infectivity of
waterborne pathogens in this way. First, the data cannot be
used to justify intensification of water treatment so as to
reduce the concentration of a pathogen to some tolerable
level: the presence of pathogens is not tolerable at any
level, though there are, of course, practical limits to what
can be done to ensure their complete absence. Second, most
known outbreaks of waterborne disease have resulted from
episodes in which the water produced during some short
period had not received any treatment or in which properly
treated water became contaminated by raw sewage during
distribution: in such instances the outbreak results from
the incident, and the precise level of contamination with
the pathogen is not a preeminent concern. In summary, there
-------
139
is a need for more information concerning the peroral infec-
tivity of waterborne pathogens, but it is clear that such ^
information can produce only limited changes in the practice
of drinking water treatment and distribution and that it is
unlikely to lead to testing water for pathogens rather than
indicators in routine surveillance.
5. Epidemiology of Waterborne Infectious Diseases
An infectious disease should be designated "waterborne"
only if water has, in fact, served as the vehicle of- trans-
mission for the infectious agent. A distinction should be
made between waterborne diseases and water-associated diseases.
Describing a disease as waterborne in a general sense implies
that water is a principal .means of its transmission._ The
role played by water in the transmission of certain infec-
tious agents has been overstressed at times to the point of
being quite unrealistic (Henderson, 1968).
This discussion concerns the epidemiology of waterborne
diseases, that is to say, the rates of infection and character-
istics of the diseases principally associated with waterborne
transmission in developed countries. There is a great
difference between the situations in developed arid developing
countries. In many developing countries worldwide, more_
illnesses and deaths result from sheer lack of availability
of water in quantities sufficient for personal and household
hygienic uses than from impurities in drinking ;water. These_.
deficiencies in quantity or availability, along with malnutri-
tion and lack of medical care, are responsible for the millions
of deaths ascribed annually to diarrhea and enteritis, which
are water-associated, but much leas commonly waterborne.
In developed countries, the quantity of water available
for drinking and hygienic purposes is sufficient; the
nutritional status and medical care are generally good, and
the incidence of waterborne disease is low, due to more than
a century's progress in sanitation. The lepidemiological
data all agree that there has been'a dramatic regression of
classical waterborne disease. For example, in the U.S. in
1908, 40 percent of the typhoid was transmitted via drinking
water, compared to 1.4 percent between 1946 and 1964.
There is an important further distinction to be made
between the waterborne diseases transmitted by drinking_
water and contaminated drinking water. Contaminated drink-
ing water is drinking water, having been made potable accord-
ing to accepted hygienic standards, that is later contami-
nated accidentally with wastewater. Such contamination
-------
140
accidents have the potential of producing epidemics that
include thousands of cases; whereas the extremely low numbers
of infectious agents that might be present in uncontaminated
drinking water, after proper treatment and disinfection, are
unlikely to produce an appreciable incidence of disease. A
third possibility is that untreated or inadequately treated
~* ^K33 beenudistributed stiH carrying significant levels
of pathogens that were present in the source waters; as will
be shown, substantial outbreaks have occurred in this way.
a- Reported Outbreaks of Waterborne Infectious Diseases
The classic and traditional waterborne diseases typhoid*
fever, bacillary dysentery, and cholera have declined
considerably in the developed countries. Development of
methods for treatment and disinfection of drinking water, as
well as construction of wastewater treatment facilities, has
made it possible to safeguard against contamination of
drinking water with bacteria of fecal origin, except in a
few extraordinary cases. Hence, in the developed countries
the majority of bacterial disease outbreaks reportedly
associated with water consumption have not arisen from a
state of endemicity in the population,' rather, they have
occurred as small localized outbreaks resulting from a
temporary deficiency in water treatment or distribution.
Infectious diseases are generally declining in relative
importance in developed countries; most of the infectious
diseases which are still prominent are not transmitted by
the fecal-oral route and are thus unlikely to be waterborne.
Whereas bacterial diseases known to be waterborne are definitely
less prevalent, the same may not be true for viral diseases.
Hepatitis A, which is sometimes transmitted through water,
shows a fairly 'constant overall incidence in the U.S. for
the past several years (Center for Disease Control, 1977a)
Laboratory methods by which the viral etiology of some
gastroenteritides can be confirmed are so new that trends in
annual incidence cannot be examined. A fundamental difficulty
in comparing the incidence of diseases is that all of them
tend to be underreported, in varying degrees, to the agencies
that are charged with compiling statistics. This applies to
diseases in general and to waterborne diseases in particular.
(i) Incidence in the U.S. The most complete data on
waterborne disease are those reported for the U.S. by the
Center for Disease Control and the Environmental Protection
-------
141
Agency (Craun, 1978). During the period 1971 through 1975,
123 waterborne outbreaks, resulting in almost 28,000 cases
of illness and perhaps two deaths, were documented in the
U.S. [See Table B.5.a-l]. The three largest outbreaks
involved municipal water systems: Sewickley, Pennsylvania
(5,000 cases of gastroenteritis) in 1975; Rome, New York
(4,800 cases of giardiasis) in 1974; and Pico Rivera,
California (3,500 cases of gastroenteritis) in 1971.
The mean annual number of waterborne outbreaks reported
in the U.S. during 1971 through 1975, was 25. This is twice
as many outbreaks as the mean annual number reported during
the period 1951 through 1970, and equals the mean annual
number reported during the period 1920 through 1936. The
reason for this apparent increase in number of outbreaks is
difficult to ascertain but is thought to be primarily the
result of increased reporting and surveillance activities.
An etiologic agent was implicated in only 49 percent of
the outbreaks (including just 36 percent of the cases) from
1971 through 1975. Among these, 12 outbreaks (511 cases)
were attributed to chemical poisoning. Outbreaks of infec-
tious diseases included 21 (4,062 cases) caused by the
Enterobacteriaceae (shigellosis, typhoid, other salmonello-
sis, and enterotoxigenic 13. coli), 14 outbreaks (368 cases)
of hepatitis A, and 13 outbreaks (5,136 cases) of giardiasis
(ii) Water Systems Involved in Outbreaks. Water
systems may be classified as municipal, semipublic, or
individual. Municipal water systems are defined as public
or investor-owned water supplies that serve communities.
Individual water systems are those used exclusively by
single residences or by persons traveling outside of populated
areas. Semipublic water systems, located in areas not
served by municipal systems, are developed and maintained
for a group of residences (e.g., subdivisions and trailer
parks) or at locations where the general public has access
to drinking water (e.g., industries, camps, parks, resorts,
institutions, and hotels).
The 123 waterborne outbreaks were classified by type of
water system [See Table B.5.a-2]. More outbreaks occurred
in semipublic water systems (57 percent) than municipal
systems (30 percent) or individual systems (13 percent);
however, outbreaks in municipal systems affected an average
of 504 persons compared to 129 persons per outbreak in semi-
public and nine persons per outbreak in individual systems.
Therefore, most of the illness (67 percent) resulted from
outbreaks in municipal systems.
-------
142
TABLE B.5.a-l
WATERBORNE DISEASE OUTBREAKS IN THE U.S., 1971-1975
Outbreaks
Cases of illness
1971 1972 1973 1974 1975 Totals
19 29 26 25 24 123
5,182 1,638 1,774 8,356 10,879 27,829
Prom Craun, 1978.
-------
143
TABLE B.5.a-2
WATERBORNE OUTBREAKS IN THE U.S.,
1971-1975, BY TYPE OF SYSTEM
Outbreaks
Cases of Illness
Municipal systems
Semipublic systems
Individual systems
37
70
16
123
18,633
9,058
138
27,829
From Craun, 1978.
-------
144
(iii) Deficiencies Leading to Outbreaks. Outbreaks of
waterborne disease can also be classified on the basis of
the kinds of deficiencies in the water supply system that
led to transmission of the causal agent. This has been
done, for a longer period of U.S. experience and with water
supplies categorized simply as public or private, in Table
B.5.a-3. One finds again that the greatest number of
outbreaks and the greatest number of cases do not neces-
sarily coincide. Whereas contaminated groundwater accounts
for the majority of outbreaks (52 percent) and cases (47
percent) involving private systems, most outbreaks (40
percent) involving public supplies resulted from distri-
bution deficiencies, but the most cases (45 percent) asso-
ciated with public supplies resulted from treatment defi-
ciencies.
It is extremely difficult to attribute disease con-
tracted as a result of consuming contaminated drinking water
directly to the discharge of treated wastewater. Commu-
nities that knowingly take their drinking water supplies
from contaminated (or potentially contaminated) sources
normally expend greater efforts to remove microbial patho-
gens that are expected to be present. The study of out-
breaks is seldom sensitive enough to implicate specific
sources of pollution, though they are likely to be the same
as some of those that give rise to shellfish contamination.
That is to say, the discharge to surface waters of untreated
or inadequately treated wastewater always carries with it at
least some potential for transmission.of disease.
b- Risk Assessment. In addition to investigations of
outbreaks, the data gathered as a result of epidemiologic
surveys can be used as a basis for estimating risks. In
some countries, groups such as the Center for Disease Control,
Atlanta, Georgia (U.S.) collect data on many notifiable
diseases, including several listed previously. Although
this type of surveillance suffers from the problem of incom-
plete reporting, it can be useful in identifying trends and
may^also be of help in determining the order of magnitude of
various risks of disease. These risks can be expressed
alternatively as cases per 100,000 population, as percent of
population affected, or as a probability of illness in a
single individual at risk. The time dimension is normally
one year. *
-------
TABLE B.5.a-3
DEFICIENCIES RESULTING IN OUTBREAKS OF WATERBORNE DISEASE IN U.S.
Cause
Untreated surface
water
Untreated ground
water
Treatment
deficiencies
Distribution
deficiencies
Miscellaneous
Total
Public Supply Private Supply Total
Outbreaks Cases Outbreaks Cases Outbreaks Cases
19 7,363 38 1,349 57 8,712
1 19 24,402 188 10,816 207 35,218
35 36,972 64 7,522 - 99 44,494
60 12,246 22 994 82 13,240
17 1,155 52 2,436 69 3,591
150 82,138 364 23,117 514 105,255
From Center for Disease Control, 1976b, 1977b ; Craun and McCabe, 1973; Craun, et al.,
1976.
Ul
-------
146
During the period from 1967 through 1976, 36 to 51
cases of enteric disease per 100,000 population were re-
ported annually (Center for Disease Control, 1977a) (these
included amebiasis, aseptic meningitis, hepatitis A, lepto-
spirosis, poliomyelitis, salmonellosis, shigellosis, and
typhoid fever, the agents of which may infect perorally
after having been shed with feces or urine)". About half of
the cases were hepatitis A. Even if the reported cases only
represent one out of 20 actual cases (Marier, 1977), it is
likely that each year less than 1 percent of the U.S. popu-
lation contracts infectious disease that is characterized as
being spread by fecal-oral route. This compares to the
rates observed in eastern Europe where reporting is thought
to be much more complete (Szmuness and Dienstag, 1971). in
1974, Czechoslovakia reported a rate of about 290 cases per
100,000 population; Poland, 245; and Yugoslavia, 360 (World
Health Organization, 1977). Estimates of the proportion of
enteric disease that is waterborne range from about 1 to 33
percent (Mosley, 1967; Singley, et aJ., 1975), so the proba-
bility of contracting waterborne infectious disease in any
year might range from 0.0001 to 0.0033 per person.
Over the past 30 years there have been over 100,000
reported cases of disease associated with water supply
deficiencies [See Table B.5.a-3], or an average of approxi-
mately 3,500 cases per year. If one out of 20 are reported,
as assumed previously, then this represents 70,000 actual
cases or a probability of 0.0003 (1976 U.S. population = 214
million) which is within the range estimated above. For
illustrative purposes assume that there are 10 cases per
year in the U.S. This figure can then be compared to average
water consumption in the U.S. it has been estimated that
the average person drinks about 1.6 1 of water per day (Safe
Drinking Water Committee, 1977^,which yields a national
consumption figure of 1.2 x 10 1 per year. This leads to
the estimate that there is less than one case of infectious
disease for every 1,000,000 1 of water consumed in the U.S.
It would seem that U.S. water supplies are quite safe in
terms of microbial hazards, although this admittedly repre-
sents .. a social judgment.
The purpose of the preceding exercise Is to demonstrate
that absolute values for risk can be estimated from general
disease statistics, albeit not with a great deal of confi-
dence. However, one can gain an idea of the magnitude of
the problem in this way. The drawback to this approach is
that the value obtained is insensitive to change because of
-------
147
the high level of uncertainty. There Is also no way of
knowing what portion of the risk can be attributed to indi-
vidual sources of infectious agents, such as effluents,
although the outbreak record may eventually prove helpful in
that regard.
Another approach to determining the risk associated .,.
with a given source of infection is to study specific popu-
lations that are most likely to be exposed to the hazard.
Ideally, the level of the exposure is determined so as to be
able to relate it to the observed response. That is to say,
it should be possible to demonstrate a dose-response rela-
tionship. For sewage and effluents as a source, a declining
order of probable exposure to infectious agents might be by
way of: occupational exposure in sewers, occupational
exposure in sewage treatment plants, shellfish consumption,
beach pollution, land irrigation, and drinking water contami-
nation. Studies intended to relate the frequency of enteric
disease to some of these modes of exposure are now in pro-
gress. ; ..-.-'.
c. Special Problems in Epidemiology of Waterborne
Disease. At least two cases of infectious disease must have
occurred before an outbreak can be considered to have taken
place and a common source of the infectious agent sought.
Except in unique circumstances, such as a case of chemical
poisoning in which the chemical was identified in ,the water,
a single case cannot be recognized as having been caused by
drinking water. The waterborne outbreaks reported above are
those in which drinking water has been implicated epidemio-
logically as the vehicle of transmission of the illness. In
most of the outbreaks, the water was found to be bacterio^
logically or chemically .contaminated. Nevertheless, the
etiologic agent of the illness was identified in only about
half of the reported outbreaks.
(i) Proving Transmission via Water. It is rare that a
simple epidemiologic study can, of itself, develop proof of
a cause and effect relationship. The strongest proof, of
course, is that provided by the time sequence whereby removal.
of a suspected cause is followed by a reduction in the
illness. Even here, however, it may be difficult to insure
that an effect will be seen if the disease has a long latent
period. False positive results, in the sense that not every
event that follows an action has surely been caused by that
action, have also to be anticipated.
-------
148
In that epidemiologic studies are done with people
having multiple activities and living in a complex world,
many things affect the outcome of exposure to an infectious
agent. The essence of a valid epidemiologic study is to
compare an exposed group to an unexposed group in which all
important factors other than exposure are identical. To
find an appropriate control (unexposed) group which meets
these needs is sometimes very difficult when the factor to be
studied is drinking water. A truly comparable control group
may be very difficult to identify when almost everyone is
"exposed"1
Furthermore, the nature of a disease may of itself
impose difficulties in undertaking prospective studies. For
example, the movements of the study population become impor-
tant problems when investigating a disease with a relatively
long incubation period, such as hepatitis A.
In order to prove, irrefutably, that drinking water has
acted as a vehicle in transmitting a particular disease
agent, it is necessary to: (1) be relatively certain of the
presence and levels of potentially pathogenic bacteria or
viruses occurring in finished water (taking into account the
limits of current techniques); (2) recover these agents from
the water; (3) have accurate knowledge of the minimum infec-
tious dose (i.e., to know at least approximately, the numbers
of organisms one must ingest to provoke infection and disease);
and (4) increase the probability of detecting sporadic and
isolated illness in a population by conducting well-planned
prospective studies.
(ii) Virus Transmission. The majority of hepatitis A
and viral gastroenteritis outbreaks reported have been
associated with drinking water supplies which, for reasons
of poor management or inadquate or interrupted disinfection,
became contaminated with fecal waste. There have not been
any reported manifestations of enteroviral disease where the
water in public distribution met conventional bacterio-
logical standards of water quality. In view of this epidemic-
logical void, three hypotheses have been offered: (1) either
there is no waterborne transmission of virus at all (however,
this seems unlikely, given the presence of viruses in waste-
water, in surface water, and even in groundwater); (2) the
impact to public health of virus transmission through water
is negligible; or (3) the 'current inadequacy of methods of
epidemiologic investigation has prevented the gathering of
substantial evidence in support of viral transmission through
water.
-------
149
Assumptions have been made that lead to the estimate
that one tissue culture infectious unit (TCIU) per 400,000 1
would be the maximum level of viruses likely to be present
in drinking water under currently accepted treatment and
sanitary standards (Environmental Protection Agency, 1978).
At this concentration, the residents of a typical city of
100,000 population would be exposed to about 100 TCIU of
viruses each day in the water used for domestic purposes
based on U.S. average community use of 640 1 per person per
day and 69 percent residential use (Murray and Reeves,
1977). If we assume that an "average" person drinks about
one 1 of tap water per day, under the circumstances outlined
above, one TCIU of virus in infectious condition would be
ingested with water by a single individual in the typical
city about every four days. The number of cases of clinical
disease that would result from the infections produced by
this level of ingestion is unknown but would most likely be
several orders of magnitude lower than the frequency of
ingestion. Factors important in the manifestation of clini-
cal disease from viral ingestion include: (1) the immune
state of the consumer as a result of previous exposure to
the virus; (2) the number of viral particles required to'
produce infection; and (3) the virulence of the virus (i.e.,
the disease:infection ratio).
It is probable that current methods (or the current
absence of methods) for obtaining epidemiologic information
about a population would not allow detection of waterborne
disease occurring at this level. Only in exceptional cases
(e.g., when wastewater contamination of a supply is substan-
tial) will numbers of affected consumers be sufficiently
high to register epidemiologically as having resulted from a
common source.
It also has been suggested that sporadic cases of viral
disease may be transmitted indirectly by water supplies.
That is, even though the levels of pathogens found in water
may not cause disease, they could produce infections which,
in turn, are spread through close contact of the infected
individuals (Berg, 1973). However, these cases should not
be considered as waterborne if there is a probability that
the diseased individual could have contracted the infection
elsewhere, nor should it be assumed that intervention with
the water route of transmission will necessarily reduce the
incidence of disease.
-------
150
Berg (1967) has indicated that it may be impossible to
prove, by current epidemiologic methods, that focal viral
infections are contracted from water and then spread in a
community by person-to-person contact. However, he con-
cludes that, given the uncertainties inherent in the situa-
tion, future decision-making must err, if at all, on the
side of safety. This point is certainly well taken, but it
is also true that decision-making should not take place
outside the context of the total picture of community health,
including the kinds of illnesses that are really occurring
in the population and the kinds of preventive measures
(including those unrelated to water) that are most likely to
exert a significant beneficial effect. Inasmuch as it may
be impossible to demonstrate that focal waterborne infec-
tions do or do not occur, the controversy remains unresolved
and may be expected to continue so for some time to come.
(iii) Parasites. Amebiasis has long been a major
concern, both as regards public health in general and drink-
ing water safety in particular. Certainly there have been
enough incidents of Entamoeba histolytica transmission by
water to show that the problem is a real one. However,
another significant problem that is now being perceived in
the U.S. and may be present, though undetected, in other
developed countries is that of endemic and sometimes epi-
demic giardiasis [See Section B.l.c(ii)]. The total inci-
dence of giardiasis in the U.S. may exceed that of amebiasis,
but the figures are difficult to compare because amebiasis
is a notifiable disease, whereas giardiasis is not (Center
for Disease Control, 1977a). There is no doubt that trans-
mission of Giardia lamblia through water is significant in
the U.S. As was reported above for the period 1971 through
1975,' waterborne outbreaks recorded in the U.S. included 13
of giardiasis with 5,136 cases (Craun, 1978). This was more
cases than were attributed to any other waterborne agent, or
.even to the entire group of Enterobacteriaceae combined.
Perhaps significantly, this case total compares to no re-
ported outbreaks of waterborne amebiasis.
Reported outbreaks of waterborne giardiasis in the U.S.
appear to have resulted primarily from use of surface water
for^drinking, after no treatment at all or only chemical
disinfection by small municipal or semipublic water systems.
Giardia cysts are relatively resistant to disinfection.
However, Giardia-free water can probably be produced con-
sistently by complete water treatment, including at least
coagulation and sedimentation followed by some form of sand
Filtration before disinfection. Giardia contamination may
not be a problem in groundwater from well-protected sources,
-------
151
but any surface water should probably be suspected, inasmuch
as contamination may sometimes derive from feral, nonhuman
reservoirs as well as from wastewater containing human
feces. For various reasons, it seems likely that waterborne
giardiasis will eventually be shown to occur in other indus-
trialized nations besides just the U.S.
(iv) Antibiotic Resistance in Bacteria. A current
major concern with regard to bacteria in drinking water
relates to their increasing resistance to antibiotics.
Bacterial cells are endowed with a remarkable capacity for
accepting genes of other strains, for incorporating these
traits, and then transmitting them. Antibiotic-resistant
enteric bacteria comprise 0.01 percent to 1 percent of the
total flora present in feces (in the absence of antibiotic
therapy); they make up 10 percent of cells isolated from
urban wastewater, 50 percent of cells recovered from river
water, and 82 percent of cells isolated from drinking water
(Leclerc, et al., 1977a). It would seem, therefore, that
over the years, antibiotic-resistant bacteria have been
diffused to and concentrated in water, which might be ex-
plained in terms of a strong selective pressure in the
aquatic environment. The hypotheses offered to explain this
are that antibiotic-resistant bacteria enjoy a selective
advantage, or that resistance is transferred directly from
one strain to another in water; neither has yet received a
direct experimental test.
It appears that in order to determine the seriousness
of the risk to public health due to increased antibiotic
resistance in waterborne bacteria, it would be necessary to
perform a survey calling for collaboration between epidemiol-
ogists and public health specialists, water authorities,
doctors, pharmacists, and diagnostic clinicians who see
patients and isolate organisms responsible for producing
disease manifestations. Such a survey ought to take into
account the total population of one geographic area.
Physicians should report all cases of confirmed infectious
illness, especially when symptoms include diarrhea. Pharma-
cists should record all medicines prescribed to clients for
treating infectious illness or diarrhea. Clinical diagnos-
tic laboratories should list organisms isolated from patients
on the basis of their antibiotic resistances. For their
part, water authorities should analyze water for more than
just the standard indicators of pollution. Finally, epidemiol-
ogists should seek better ways of ascertaining the role of
water in the spread of disease within a population. However,
-------
152
without substantial financial support and good collaboration
among these professions, such a survey could not be conducted,
d. Modeling and Monitoring. An epidemiological model
is a conceptual system that operates by resembling the
natural course of a disease and its transmission. It
incorporates major epidemiological factors that determine
the dynamics of the spread of infection. A model necessar-
ily represents a simplification of natural processes; never-
theless, if properly constructed it can simulate the natural
evolution of an epidemic or an endemic situation, thus
permitting the study of the disease dynamics and the effect
of deliberate interventions on the natural course of trans-
mission of the infection and hence, on the incidence of the
disease (Cvjetanovic, e_t ai_., 1978).
Observations of epidemics of infectious diseases led
epidemiologists to the conclusion that they present some
regular features and that there must be some definite
principles that determine the evolution of infectious pro-
cesses. Efforts have therefore been made to express, in
precise quantitative terms, time-related changes in the
dynamics of infections and to formulate a mathematical
theory of epidemics. One of the general principles of the
mechanism of epidemics was established in 1927 by Kermack
and McKendrick in mathematical terms when they formulated
their theory of the initial "threshold of density" of sus-
ceptible populations as the determining factor in epidemics.
Later models have been proposed for individual diseases
on the basis of general theories (stochastic, catalytic, and
deterministic models) which are excellently reviewed by
Cvjetanovic and coworkers (1978). Some of these models can
be applied to diseases transmitted partly though water as
well as by food. These models are constructed by identi-
fying categories of individuals and factors that play a
well-defined and important role in the dynamics of the
disease. In order to make the model relatively simple and
manageable it is desirable to eliminate unimportant factors
and retain only those that significantly influence the
epidemiological processes. It is obvious that the models
must be based on the natural history of the disease and
should be an expression of that history. The aim of the
model construction is to arrive at a system that is able to
"mimic" the natural processes, such as the past outbreaks
and trends of disease, and thus to simulate various real
or hypothetical situations.
-------
153
(i) Potential for Modeling Waterborne Infectious
Disease. The word "model" can take on many meanings. Even
when it is clear that one is referring to a mathematical
model, the great variety and complexity of models may still
make the definition unclear. To 'many/; "modeling may seem to
be a mathematician's art, interesting, but not very practi-
cal. Yet all models have one thing in common: they are
attempts to represent reality and as such (if they are
logically constructed) may be extremely useful in inferring
information about real situations. ,
To be useful to decision-makers, a mathematical model
must incorporate the following features: (1) it must be ;
clearly defined; (2) the assumptions made in its construc-
tion must be clearly stated; (3) those variables which have
the greatest impact on its output must be identified (sensi-
tivity analysis); (4) data inputs must be carefully selected;
and (5) it must be validated (Breidenbach, 1976). As a
modeler tries to describe large areas, his model can become
very complex; but as long as he adheres to the above tenets,
the model should still be understandable to those policy
makers who must use it. One of the major advantages of
modeling is that it promotes the systematic collection and
organization of information. On the other hand, if modeling
is to have any practical value, representative field data
will have to be collected in quantities sufficient to yield
unbiased estimates for use in the model.
A major reason for the, apparent difference in the
"incidence of a disease between areas is often the variation
in the efficiencies of passive reporting systems. Several
studies have indicated that from 60 to 90 percent of the
clinical cases of hepatitis are not reported to health
authorities (Center for Disease Control, 1975b; Koff, et
al., 1973; Levy, et al., 1977; Liao, et al., 1954; Marier,
1977). During nonepidemic periods as few as 10 percent of
the actual cases may be reported (Tolsma and Bryan, 1976).
However, with the advent of screening for the hepatitis B
surface antigen and the availability of immunoglobulin
prophylaxis, it is likely that a higher percentage of cases
will be reported in the future (Bernier, 1975). Inasmuch as
reporting of even notifiable diseases is clearly inadequate,
there is an obvious need for alternate, active methods of
gathering data for use in epidemiologic modeling and risk
assessment.
-------
154
(ii) Dose-Response Relationships. One of the more
basic approaches to estimating risk is the dose-response
relationship which predicts the percent of a population
likely to exhibit a specific response (e.g., infection,
disease, or death) when exposed to a given dosage of a
disease agent. Once this type of information has been
obtained for representative populations, all that is neces-
sary to assign risk is to measure the level of exposure to a
given agent among the subject population. Further, if the
sources of the agent are known, then the risk from each of
these sources can be determined. However, there are several
limitations to this approach: (1) there are many potential
disease agents to be evaluated; (2) there can be a wide
range of undesirable responses (for example, variety and
intensity of symptoms); (3) it is difficult to measure the
concentration of some disease agents; and (4) acquisition of
the needed data normally requires experimentation on human
beings of varying backgrounds, which has become a visible
ethical problem (Katz, 1972).
The response to infection is normally acute and gener-
ally can be classified into the various states of: infec-
tion, disease, and death. Not all infections result in
disease which, in the idealized form, means that the dose-
infection curve represents lower doses than the dose-disease
curve for a given percent response. The vertical difference
between the two curves is a measure of the proportion of the
infections that result in disease (for hepatitis A, this
ratio is about 1:5 to 1:20). However, it cannot always be
shown that a dose larger than that required to cause infec-
tion is necessary or more likely to produce disease. At
present, very few cases of enteric disease result in death.
(iii) Monitoring Methods. There are, of course,
active means of gathering needed data without having to
challenge human subjects with infectious agents:
- Monitoring "is an integrated system of making
observations on.health and environmental factors,
and of scrutinizing, storing, and retrieving
these data for specified purposes of protecting
and improving human health.
- Health Surveillance is the closely associated
system of collating and interpreting data col-
lected from monitoring programs (and from any
other relevant sources) with a view to the detec-
tion and evaluation of health problems so as to
provide a basis for remedial action.
-------
155
In addition, -monitoring may often prove a valuable tool
for the epidemiologist endeavoring to relate exposure to
response in order to elucidate the causes of widespread
diseases or to explain changes in the total disease and
mortality patterns. In regard to vital statistics such as
births and mortality, monitoring is already established in
all developed and in many developing countries throughout
the world, but two-thirds of the world's population still
lack adequate birth and death registration. Communicable
diseases are subject to surveillance, with the object of
alerting public health authorities, even at the stage of
suspicion, to take preventive action. Although highly
developed computerized systems have been designed to provide
the quickest possible warning, simpler systems can be devised
to suit the needs and capabilities of the country.
It is possible, in looking at environmental hazards, to
study" a population of children. They tend to remain in the
same area at least for a certain known period of time and
since they usually go to school near their homes, it is
relatively easy to determine the effect of an environmental
factor.
The ultimate in this approach is to study a target
population with a history of weak immunologic defenses, one
that is easy to keep under surveillance, and one for which
the source of drinking water is strictly controlled that
is to say, a population of infants attending child care
centers. By studying a population whose drinking water
consumption is restricted to bottled water exclusively, and
barring any person-to-person spread, one can obtain a good
picture of the impact of a water sample on the health of the :
consumers.
In comparing different groups of the population exposed
to environmental hazards, it is extremely important to
choose appropriate control groups. In considering an environ-
mental hazard in an industry, for example, it is possible to
examine illness or morbidity among husbands and wives, using
the wives as controls. It is unlikely that they will be
exposed to the same occupational hazards as their husbands,
whereas both are likely to be exposed to the same environ-
mental' hazards in their homes.
It is also possible to compare data from .epidemiologic
surveys of two geographic areas with populations of roughly
equivalent size and activity; one in which the drinking
water is of poor quality and the other using water'of high
-------
156
quality. By noting whether there has been any increase in
the sale of antidiarrheal medications, one can obtain valu-
able data without requiring that there be notification of
disease through official channels. The special merit of
this approach is that it does not require direct, conscious
participation by the populations that are included in the
study. *
Otherwise, whatever population is chosen, it is impor-
tant to investigate that population as completely as pos-
sible. Information on those who do not participate is
essential; they may, for example, have refused because they
are ill or because they fear the consequences of having the
disease discovered. Significant bias may be introduced in
this way if adequate precautions are not taken.
Once the population has been defined and the variables
to measure have been decided, the next step is to collect
the necessary information. One method of doing this is by
using a questionnaire. This is usually better than a clini-
cal history, which may not identify and record in a repeat-
able way what the individual actually says. The importance
of using a standardized questionnaire has been amply demon-
strated; whatever questionnaire is used, it is vital that
the sensitivity, specificity, and precision of the indi-
vidual questions be validated. Interviewing techniques must
also be standardized.
Physiological tests are another possible method of
obtaining the necessary information. Such tests must be
simple and if possible cheap to administer, acceptable to
those who receive them, and must produce accurate and repeat-
able results. They must also be "sensitive" (give a posi-
tive finding in those who have the condition under investi-
gation) and "specific" (give a negative finding in those who
do not) . Physiologica.1 tests, capable of direct and rapid
assessment of the functional state or normality of any
member of a study population, would obviously be of great
"help in minimizing the subjective aspect of an epidemiologic
study. Unfortunately, such tests may not always be readily
applicable to studies of enteric diseases. If they were,
they would yield results well suited to the application of
the threshold, or no-permissible-adverse-health-effects,
approach that is the basis for all national regulations
establishing drinking water quality standards. That is,
exposure limits based on bacteriologic indicators in water
are set at levels where no adverse health effect to con-
sumers can be anticipated. This is feasible because the
-------
: is?
relationsliip between the pathogenic bacteria and classical
microbiological indicators is fairly well known, but may be
insufficient for waterborne diseases because of the low
correlation between the presence of bacterial indicators and
viruses in water, as will be discussed in Topic G.
e. Summary. An extensive record of waterborne out-
breaks of disease in the U.S. indicates that most such
incidents have resulted from the use of untreated or under-
treated water and from episodes of contamination of drinking
water. Inasmuch as comparable records do not seem to be
available from other countries, it is not entirely certain
that the U.S. experience is typical of other industrialized
nations. Better reporting, and recording, of disease out-
breaks in other countries would certainly be desirable.
There is continuing concern about the possible inci-
dence of sporadic illness in the consuming public, espe-
cially as a result of viruses that might be present at
extremely low levels in finished drinking water. Existing
systems of compiling disease statistics are passive in the
sense that the compilers must wait and hope that clinicians
who observe waterborne illnesses will report them. Alter-
nate, active approaches to gathering information on water-
borne disease are needed: several possible methods have
been discussed in this section.
Meanwhile, decisions regarding drinking water safety
and public health must be made with the information that is
now available. The epidemiologic record (at least of the
U.S.) appears to make a strong case for complete treatment
and disinfection of drinking water. "Worst-case" assump-
tions may lead to predictions that a certain amount of
disease will be transmitted even by completely treated
water. The public interest may instead lie in determining
the validity of these predictions through better epidemio-
logic work than in devising new water treatment techniques
in hopes of solving a problem that has not yet been proven
to exist.
6. Summary
Drinking water may transmit pathogenic organisms if it
becomes contaminated with human or animal feces. Fecal
contamination may occur at any point from the raw water
source to the ends of the distribution network. If the
-------
158
pathogens thus introduced are not subsequently removed or
destroyed, a result may be that the consumer will ingest
some quantity of bacterial, viral, or parasitic organisms.
Geographic and epidemiologic conditions, as well as the
overall numbers and virulence of the organism ingested,
determine the result of such an incident.
a. Microbial Pathogens. The transmission of microbial
pathogens through drinking water in industrialized nations
has changed with time. Some of the most historically
significant waterborne pathogens (notably those that cause
typhoid fever, bacillary dysentery, and cholera) have been
on the decline in recent years. Other bacteria, all of the
enteric viruses, and the protozoon, Giardia Iambiia, have
evoked more concern of late, probably due instead to greater
awareness and better reporting than because they present a
greater threat than previously. This trend is associated
with the development of better methods of detecting poten-
tial waterborne pathogens in clinical specimens and, in some
cases, in water samples.
The genus Salmonella is a very large group consisting
of 1,200 known serotypes that are pathogenic to humans,
causing mild to acute gastroenteritis and very occasionally
death. Typhoid fever, caused by J3. typhi, and paratyphoid
fever, caused by S. paratyphi A or B, are both enteric
diseases that occur only in humans. The other Salmonella
serotypes are responsible for foodborne illnesses accom-
panied by mild to acute gastroenteritis, but rarely death.
These milder forms are referred to as salmonellosis and
occur frequently in humans and wild or domestic animals.
Salmonellae are excreted by infected humans (exclu-
sively so for ST typhi), farm animals, domestic pets, and
warm-blooded wild animals. Salmonella strains were regu-
larly found in the sewage system of a residential area of
4,000 persons. Streams, lakes, and rivers receiving dis-
charges of meat processing wastes or effluents of untreated
or inadequately treated community sewage may contain sub-
stantial numbers of salmonellae. Fish living in polluted
water may ingest salmonellae and become vectors of pathogen
transport. Drinking untreated, unprotected surface water
presents the greatest risk to the consumer. There are
presently no standardized methods for isolating salmonellae
from water. Since these microorganisms ordinarily occur in
lower numbers than those of sanitary indicator bacteria,
they must initially be concentrated from large-volume samples
of water.
-------
159
The Shigella genus is divided into four main subgroups.
All species cause bacillary dysentery (also called shigel-
losis) exclusively in humans and some primates. Infection
is transmitted by the fecal-oral route: shigellae have been
isolated from clothing, toilet seats, and contaminated
foods. Infection with Shigella occurs endemically in most
communities and may be maintained by a few symptomless
carriers in the absence of clinical cases. Stricter sani-
tation measures, proper sewage disposal, and public health
standards enforced in the developed countries have led to a
.shift in peak incidence from summer to winter, as low tem-
peratures favor survival of Shigella. As with Salmonella,
no standardized procedures have been established for iso-
lating Shigella from water.
Evidence of infections due to Yersinia enterolitica has
been mounting since the early 1960's, especially from the
world's cold or temperate regions. Yersiniosis is thought
to be contracted perorally. Y. enterocolitica has been
isolated from lymph nodes and feces of both sick and healthy
humans and of a growing number of animal species. Different
Y. enterocolitica types have been isolated so frequently
IFrom untreated surface water in some areas that they most
likely represent part of the normal microbial flora of the
water and surrounding terrestrial environments. Reports
referring to waterborne Y. enterocolitica infections are few
in number. Moreover, Y. enterocolitica grown at 37°C is
less resistant to normal cellular bactericidal defenses than
when grown at 20°C; this could explain why transmission of
infections through direct person-to-person contact is
relatively rare, and again, why an intermediate cold phase
could be critical to its spread. If so, the life cycle of
Y. enterocolitica would stand in sharp contrast to that of
others in the family Enterobacteriaceae, which are trans-
mitted through water less frequently than by contact. There
is no standard method for the isolation and enumeration of
Y. enterocolitica in water.
Transmission of enteropathogenic Escherichia coli
through drinking water was frequently reported during the
1950's. Certain enteropathogenic strains of E. coli are now
known to cause acute diarrhea, especially in infants, in
travelers to foreign countries, and in consumers of contami-
nated foods. The definition of the coliform group includes
E. coli, so enteropathogenic E. coli is unlikely to be
present in water in which colTforms are undetectable? this
relationship is far more direct than is usual between indica-
tors and pathogens. However, even if E_. coli is shown to be
-------
160
present, its enteropathogenicity can be determined only by
highly refined techniques.
Tularemia is a zoonotic disease transmitted to humans
from blood-sucking arthropods, domestic animals, and pri-
marily from a number of wild animal species, many of which
lead semi-aquatic lives. The causative agent, Francisella
tularensis, can be contracted through ingestion and may
produce buboes and areas of necrosis in organs and tissues
of man and animals. Outbreaks of waterborne tularemia have
been reported throughout'the U.S.S.R. Recent cases of water-
borne tularemia in northern Norway have brought about renewed
interest in water as a vehicle of infection in the Scan-
dinavian peninsula. There is no standardized method for the
examination of water for F. tularensis.
The genus Leptospira is composed of finely coiled
spiral organisms, including a number of strains that cause
leptospirosis in humans and many animals. Leptospirosis is
essentially a zoonosis that may involve domestic and wild
animals in great numbers depending on the climate and avail-
able food supply. The organism is maintained in the en-
vironment by both carrier and diseased hosts, and trans-
mission is particularly favored by population explosions of
animal carriers, especially rats. Human exposure to the
disease has been from direct or indirect contact with urine
from infected animals. Infections with Leptospira are
primarily associated with certain occupations, such as
mining, dairy farming, and sanitary engineering. It is
associated with drinking water only rarely, when there has
been a breakdown in sanitation and public health systems.
There is no standard method for isolating Leptospira from
environmental sources.
The genus Vibrio comprises a large number of species,
only a few of which are of medical importance: V. cholerae
and its biotype V. cholerae El Tor cause cholera exclusively
in man; V. parahaemolyticus and nonagglutinable (NAG) vibrios
can cause cholera-like disease or mild diarrhea in humans,
but are not normally transmitted by the water route. Cholera,
whether caused by classical V. cholerae biotypes or the El
Tor biotype, produces the same clinical symptom, profuse
diarrhea. Fecally contaminated water is the primary vehicle
of cholera transmission, although vibrios are also spread by
a multitude of other routes including food, soiled clothing,
flies, and direct person-to-person contact. Cholera vibrios
originate from human feces or vomitus. The disease has
characteristically been sporadic and endemic in areas of
poor hygiene and warm humid climates. However, isolated
-------
7 .. 161
cases have been reported in Europe, and the U.S., mainly
from importation through tourism. Methods for isolating
vibrios from water have not been standardized.
Campylobacter was initially observed as the agent of
infectious abortion in cattle; it also may cause gastro-
enteritis in humans, cattle, sheep, and swine. The organism
inhabits the genitourinary and intestinal tracts and is
excreted in the feces. Feces from infected humans, animals,
and fowl may contaminate surface waters destined to be used
for drinking water supplies. If such source waters were
inadequately treated, viable Campylobacter cells could,
theoretically, gain entry to finished water. At present,
there are no standard methods for detecting Campylobacter in
water. When such methods become available for routine use,
it will then be possible to determine the significance of
Campylobacter in water as it relates to outbreaks of water-
borne gastroenteritis.
Several other microorganisms present in water can
infrequently cause disease, and almost always under unusual
circumstances, either in abnormal hosts or in situations
where the normal flora have been supplanted. These organisms
are called opportunistic pathogens; Pseudomonas aeruginosa
is a most notable case in point, along with some Entero-
bacteriaceae. Opportunistic pathogens are ubiquitous in
nature, very resistant in water, and can grow with only a
few nutritional requirements. Concentrations of opportu-
nistic pathogens found in drinking water are not normally
sufficient to lead to infection in a healthy consumer.
Another very important property of these microorganisms is
their great ability to accept and transfer plasmids which
carry determinants for resistance to antibiotics.
Opportunistic pathogens present in drinking water
include the following bacteria: Pseudomonas sjjecies, Aero-
monas hydrophila, Edwardsiella tarda, Flavobacterium, Kleb-
siella, Enterobacter, Serratia, Proteus, Providencia, Citro-
bacter, Acinetobacter and Staphylococcus aureus. Examina-
tion of drinking water for the presence of Pseudomonas
should be incorporated into routine analyses of the water
since suitable methods exist for the enumeration of P_.
aeruginosa.
Viruses are ultramieroscopic intracellular parasites,
incapable of replication outside of a host organism. Great
numbers and varieties of bacterial, plant, and animal viruses
may be present in both polluted and unpolluted waters. The
viruses of greatest concern are those of human .origin, which
-------
162
are capable of infecting and causing disease in humans. In
general, they are shed in the feces and are known as the
human enteric viruses. Infection usually takes place after
viruses are ingested, possibly in contaminated water or
food. Except for the hepatitis A virus, the public health
significance of human enteric viruses in water remains
unclear due possibly to the apparent or latent nature of
viral infections and the difficulty of detecting waterborne
viruses.
Human enteroviruses include the polioviruses, the cox-
sackieviruses groups A and B, and the echoviruses. Although
large numbers of enteroviruses have been consistently isolated
from fecally contaminated water, only a few reports have implicated
water as the vehicle of transmission. Hepatitis A is the only
form of viral hepatitis known to be transmitted through water,
and it is also the most prevalent waterborne disease attributable
to a specific etiologic agent. The incubation period of the
disease in humans generally ranges from 15 to 50 days, with a
median of 28 days. Some people shed the virus in their feces as
early as seven days before onset of symptoms; others may transmit
the virus without ever becoming perceptibly ill. Attempts to
culture the hepatitis A virus in cell lines have consistently
failed, so that all available evidence about its transmission by
polluted drinking water is obtained from epidemiological studies.
Another group of small viruses potentially transmissible through
water, some of which may in fact be enteroviruses, are called
gastroenteritis viruses. Rotaviruses are primarily associated
with gastroenteritis in children. . Because rotaviruses may be
excreted in very large numbers, they are almost certain to be
present in polluted water; there is presently no suitable method
for their detection at low concentrations. Reoviruses, adeno-viruses,
and adeno-associated viruses can also be found in water.
Because viruses multiply only within susceptible host cells,
they cannot increase in sewage. Their numbers are further reduced
as a result of sewage treatment, dilution, natural inactivation,
and water treatment practices. Therefore, barring gross
contamination of finished water, only extremely low numbers of
viruses, if any, are likely to occur in properly treated supplies.
Large samples of water must be tested to enable detection of
whatever low level of viruses might be occurring in any given
source water. Adsorption methods employing glass microfibers and
glass powder appear promising for handling large volumes of water
containing few viruses. Until further information is forthcoming,
the public health significance of small numbers of viruses detected
in drinking water will remain undetermined, and it should be
noted again that the hepatitis A virus cannot be detected in water.
-------
163
Certain species of protozoa are now an increasing cause
of waterborne disease. Entamoeba histolytica causes amebic
dysentery, as well as such other clinical manifestations as
diarrhea, abcesses of the liver, and chronic infections with
minimal symptoms. Improved sanitary sewerage systems have
been largely responsible for the curtailment of waterborne
amebic dysentery. Another protozoan, Giardia Iambiia, now
appears to be emerging as a major causative agent of water-
borne disease. It was not until 1966 that water was recog-
nized as a vehicle of Giardia transmission; since that time,
increased awareness on the part of physicians may have
accounted for the more frequent reporting of outbreaks. The
overriding point to be emphasized is that outbreaks in-
volving municipal water supplies are associated with surface
water sources where disinfection was the only treatment.
G. Iambiia cysts are not destroyed by chlorination at dosages
and contact times commonly used in water treatment. Nae-
gleria fowleri has been recognized as one of the agents
responsible for causing primary amebic meningoencephalitis
(PAM). Acanthamoeba is also associated with PAM; like
Naegleri fowleri, Acanthamoeba species have ben isolated
from tap water in association with PAM cases. Metazoan
parasites that may be transmitted in drinking water are
limited to a few nematodes and helminths whose presence in
water is only incidental to their life cycles. Standard
chlorination is not effective against parasitic or nuisance-
causing metazoa discussed in this/section. Well-regulated
flocculation, sedimentation, and filtration practices afford
reliable protection against their occurrence in finished
water. ; ,
Many causes of gastroenteritis are possible, including
bacteria, viruses, and other unknown agents. The illness in
such outbreaks is often called "acute infectious nonbac-
terial gastroenteritis" (AING). It is only recently that
means have been developed for identifying some of the viruses
responsible for AING, such as parvovirus-like agents, rota-
viruses, and particles resembling coronaviruses, myxoviruses,
astroviruses, adenoviruses, and caliciviruses.
b. Sources of Waterborne Pathogens. Reservoirs of.
diseases transmissible to man are man himself, as well as
domestic and wild animals. The microorganisms responsible
for causing disease generally are excreted in the feces or
urine, whereupon they may gain access to water. If drinking
water treatment is inadequate or lacking altogether, these
organisms may pass freely into water en route to the con-
sumer, thereby engendering a risk of infection and possibly
-------
164
disease. Pathogenic microorganisms are excreted not only by
individuals with clinical symptoms, but also by asymptomatic
carriers.
tC. Persistence and Death of Pathogens. Water is a
hostile environment to most human pathogens which, once
introduced, will die at varying rates depending on factors
that influence this death rate: temperature, pH, nutrients,
and predation. Under certain conditions, some pathogens
persist longer in water than the standard bacterial indi-
cators. Therefore, once a source water has been contami-
nated, specific treatment to remove pathogens is warranted;
it is not enough to rely solely on natural die-off during
storage.
d. Infectivity of Waterborne Pathogens. The presence
of a viable pathogenic microbe in drinking water is always
undesirable, but it does not always guarantee that infec-
tion, and especially disease, will result if someone drinks
the contaminated water. There is a paucity of information
concerning the infectivity of waterborne pathogens. The
probability that a given agent will cause infection if
ingested with drinking water is, almost certainly, a func-
tion of the dose; however, many other factors can affect the
answer of the ingestor (immunity, nutrition, intestinal
flora, intercurrent illnesses).
e. Epidemiology of Waterborne Infectious Diseases.
The epidemiology of waterborne diseases in developing coun-
tries deals simply with the infection rates and charac-
teristics of the diseases principally associated with water-
borne transmission. However, in developed countries, the
quantity of water .available for drinking and hygienic pur-
poses is sufficient; the nutritional status and medical care
are generally good; and the incidence of waterborne disease
is low,^as a result of more than a century's progress in
sanitation. Contaminated drinking water (water that is
ordinarily treated so as to be safe to drink, but is acci-
dentally contaminated by wastewater containing feces) is an
important vehicle for epidemic outbreaks because of its
potential for mass transmission. On the other hand, drink-
ing water that is treated and disinfected so as to contain
only a few residual microorganisms might produce only a
barely perceptible or imperceptible rate of disease in the
consuming population. More recorded outbreaks have been
-------
165
associated with semipublic.water systems than with muni-
cipal or individual systems; however, outbreaks from muni-
cipal systems have affected more people than those from
semipublic or individual systems.
In some countries, an agency collects data on many
notifiable diseases, including several previously listed.
Although this type of surveillance suffers from the problem
of incomplete reporting, it can be useful in identifying
trends and may also be useful in determining the order of
magnitude of various risks of disease. During the period
from 1967 through 1976, in the U.S.A., 36 to 51 cases of
enteric disease per 100,000 population were reported annually,
About half of the cases were hepatitis A. Even if the
reported cases only represent one out of 20 actual cases, it
is likely that each year less than 1 percent of the popu-
lation contracts infectious disease that is characterized as
being spread by the fecal-oral route. Estimates of the
proportion of enteric disease that is waterborne range from
about 1 to 33 percent, so the probability of contracting
waterborne infectious disease in any year might range from
0.0001 to 0.0033. Absolute values for risk can be estimated
from general disease statistics, but not with a great deal
of confidence. One can gain an idea of the magnitude of the
problem, however. The drawback to this approach is that the
value obtained is insensitive to change because of the high
level of uncertainty. Another approach to determining the
risk of a given activity is to study specific populations
that are most likely to be exposed to the hazard. Ideally,
the level of the exposure is determined in order to relate
it to the observed response (dose-response relationship).
The record of waterborne outbreaks, even though^they
are underreported, reveals the most common deficiencies in
water supply systems. Contaminated groundwater accounts for
the majority of outbreaks involving private systems, whereas
water distribution deficiencies are the most common causes
among municipal systems. It is extremely difficult to
associate the discharge of treated wastewater with disease
contracted as a result of consuming contaminated drinking
water. It has been suggested that sporadic cases of viral
disease may be indirectly transmitted by water supplies.
The importance of the so-called "focal infection" (caused by
ingestion of water containing extremely low level of virus)
in the spread of enteric viral disease in a community has
been neither proven nor disproven. This is mainly because
epidemiological research methods are not sufficiently sen-
sitive to determine whether or not "waterborne focal in-
fections" occur. In most of the reported outbreaks, the
-------
166
water was found to be bacteriologically or chemically con-
taminated. An etiologic agent is determined in about 50
percent of the incidents, but the rate of reporting is very
low.
In any individual, the result of ingesting an infec-
tious agent with water will depend on a great many factors.
It is rare that a simple epidemiologic study can, of itself,
develop proof of a cause-effect relationship. The purpose
of an epidemiologic study is to compare an exposed group to
an unexposed group in which all factors other than exposure
are identical. Finding an appropriate control (unexposed)
group that meets these needs is sometimes very difficult
when the factor to be studied is water. The nature of the
disease sometimes imposes further difficulties in prospec-
tive studies, as in the case of hepatitis A, where the long
incubation period affords much opportunity for subjects to
move about before becoming ill. Efforts have, therefore,
been made to derive precise quantitative expressions for
time-related changes in the dynamics of infections.
Models are attempts to represent reality, and as such,
may be_extremely useful in inferring information about real
situations if they are logically constructed. To be useful
to decision makers, a mathematical model must incorporate
the following: (1) it must be clearly defined; (2) the
assumptions made in its construction must be clearly stated;
(3) those variables which have the greatest impact on its
output must be identified (sensitivity analysis); (4) data
inputs must be carefully selected; and (5) it must be vali-
dated. One of the major advantages of modeling is that it
promotes the systematic collection and organization of
information.
It is possible, in looking at environmental hazards, to
use a population of children. They tend to remain in the
same area at least for a certain known period of time and,
since^they usually go to school near their homes, it is
relatively easy to determine the effect of an environmental
factor, such as pathogens present in drinking water.
7. Recommendations
It is not to be expected that community water
suppliers will invariably start with pure
source water, treat and disinfect it appro-
priately, and distribute it through a flawless
-------
167
network. Even where source water is protected,
as is true of many groundwaters, disinfection
is recommended. This is especially important
where finished water is to be stored rather
than being used immediately.
2. Aftergrowth of opportunistic pathogens during
distribution of finished water should be pre-
vented by maintaining a chlorine residual
throughout the distribution network, especially
in large systems.
3.^ More sensitive methods for the detection of
waterborne pathogens need to be developed,
along with epidemic-logic techniques to afford a
more precise .understanding of the effects of
these agents on public health. Especially
needed are well-planned prospective epidemiol-
ogic studies that are capable of establishing a
valid causal relationship between the public
water distribution system and the incidence of
infection (and not only overt disease) in the
consuming community.
4. Other situations in which it would be appro-
priate to test for pathogens in distribution
water are: (a) after contamination is found to
have occurred; (b) to trace the source of an
outbreak; and (c) in analyzing disinfectant
efficiency.
5. Better techniques must be developed for the
detection and enumeration of all viruses that
may be present in drinking water. Studies are
needed to determine the quantities of water-
borne viruses that must be ingested to produce
infection and disease. Given the lack of
correlation between viruses and the bacterial
indicator systems, more research on the anti-
viral effectiveness of various water treatment
processes is needed.
6. Epidemiologic surveillance networks should be
established to determine the incidence of
infections in populations of drinking water
consumers, with an eye to either strengthening
or relaxing specific standards.
-------
168
C. INDICATOR SYSTEMS FOR MICROBIOLOGICAL
QUALITY AND SAFETY OF WATER
Good quality drinking water should be free of poten-
tially harmful organisms and substances. Although this
continues to be the main concern of drinking water micro-
biology, there is much more to the quality of drinking
water, by present-day standards, than mere harmlessness.
Today's good quality drinking water should be of acceptable
color, odor, taste, and (if possible) temperature, and should
be prepared and distributed under continuously controlled
microbiological conditions.
In the present context, "indicator systems" are defined
as quality control methods for water procurement, treatment,
and distribution, even when these are not directly related
to the presence of potential pathogens. In the broad sense,
consideration has been given to methods which range from the
traditional reliance on fecal indicators, to methods for
recovering other more fastidious organisms, to those which
employ chemical and physical means of determining microbial
populations and activities.
There are few established routine methods for detecting
human pathogens in drinking water due to the large diver-
sities and low numbers encountered in potable waters. Since
most of the important waterborne human pathogens are of
fecal origin, tests for determining the presence of organisms
normally found in human feces provide a means for alerting
the microbiologist to any potential hazard from pathogens in
the water. These tests for recovering normal fecal organisms,
which occur in far greater quantities, need not be as sensi-
tive and specific as methods which measure pathogens directly.
There is always room for improvement in the methods for
ensuring the absence of pathogens from drinking water.
However, we also intend to consider here microbiologically-
based methods for the control of all phases of drinking
water quality. Some of these methods are well established
and routinely used in water testing while others are in
-------
169
various experimental stages
special cases. ;
Some may only be applied in
Probably no indicator system will be found ideal for
all of the possible applications. Some may seem most
appropriate for daily, quality control? others may be most
useful in measuring gradual deterioration in water quality
over the long term. Some of the rapid or automated methbds
may find widespread application in determining water quality
during emergencies such as cross-connections or line breaks.
Other automated methods may. be better applied to on-line
monitoring of water quality.
This section will discuss both the theoretical bases of
indicator systems and improvements in testing technology in
an effort to evaluate their present and future roles in
ensuring the quality and safety of drinking water.
1. Established Viable Indicator Systems
The indicator systems in this category are the classical
microbiological parameters of water quality which have
proved useful for indicating the presence or potential
presence of pathogens originating from fecal sources. It is
generally agreed that a good indicator of these types of
pathogens should fulfill;the following criteria:
- It should be present and occur in greater numbers
than pathogens.
- Any potential.it may have for growth in the aquatic
environment should not surpass that of pathogens.
- It should yield characteristic and simple reactions
enabling, as far as possible, the unambiguous
identification of the group.
The indicators in this group are dealt with in terms of
their ability to satisfy these criteria and the extent to
which they:will continue to be relied upon, in the coming
years, as the primary microbiological indicator systems of
water quality.
a. Colony count. The colony count (also referred to
as standard plate count, total microbial count, viable
count, water plate count, total bacterial count, aerobic
mesophilic viable bacteria, etc.) measures the number of
-------
170
aerobic and facultative anaerobic heterotrophic bacteria in
1 or 0.1 ml of water that form colonies on nutrient medium
at a specified incubation time and temperature; Nutrient
media used and time and temperature of incubation .vary from
country to country [See Section D.6]. Colonies formed with
this method do not represent all bacteria in the sample
since no single growth medium will accommodate the nutri-
tional requirements of all bacteria. Also, strict anaerobes
and slower growing bacteria are missed by this method.
Clusters or chains of bacteria or bacteria adsorbed to small
particles do not necessarily yield colonies that were devel-
oped from each bacterium contained in such groupings.
Plates are incubated at 35 to 37°C for 24 to 48 h,
favoring bacteria adapted to the body temperature of warm-
blooded animals, or at 20 to 22°C for 48 to 72 h, favoring
saprophytic bacteria which may be capable of causing opera-
tional disturbances in water treatment and distribution.
Organisms able to grow at 20°C are found in tap water in
much higher numbers than organisms that grow at 35°cl This
is partly because the adverse conditions inherent in any
water environment, such as lack of nutrients and sub-optimal
temperatures, would tend to select for bacteria of little or
no public health significance [See Section B.3].
The vast majority of bacteria are of this latter cate-
gory and take part in virtually all cycles of nature: in
the self-purification processes that occur in surface water,
in antagonistic reactions among different species, and in
the mineralization of organic matter (e.g., ammonia to
nitrite, to nitrate), to name a few [See Section A]. Al-
though the genera of bacteria detected by the colony count
method may not be pathogenic to healthy humans, many of
these organisms can produce acute or chronic infections in
special cases for example, during medical therapy [See
Section B.l.a(ix)]. Bacteria other than coliforms also are
important because they can hamper filtration efficiency at
the treatment plant [See Sections E.4 to 5] and their pres-
ence in large numbers often signifies a deterioration in
finished water quality brought on when, for example, changes
in pressure in a distribution system cause a release of
microorganisms from dead ends and other protected sections
CSee Section P.2]. .
(i) Media Used for the Colony Count. Various nutrient
media used in the colony count differ mainly with respect to
the solidifier gelatin, agar, or silica gel. All nu-
trient media have to be enriched with some form of peptone
-------
171
and meat or yeast extract, the type and concentration of
which influence the number of developing bacteria [See also
Section D.6H-
Agar media contain either yeast or meat extract and
differ mainly in the quality of peptone used and in the
presence, in some cases, of glucose. However, the incor-
poration of glucose entails an additional analysis,and calls
for a number of precautions during sterilization in order to
avoid hydrolysis of the sugar. For this reason , the European
standardization committees INSTA and CEC do not recommend
glucose media for the colony count technique (Commission of
the European Communities, 1977).
Gelatin, sometimes used in conjunction with gelose,
contains meat extract with or without sodium chloride. The
gelatin is a nutrient for gelatin-liquifying bacteria, and
as such, will influence the colony count. Thus, gelatin not
only provides a quantitative measure, but also distinguishes
certain "nuisance" bacteria, such as green-fluorescent
pseudomonads [See Section C.2.cH and other bacteria that
liquify gelatin rapidly. Gelatin nutrient media should be
incubated only at 20°C, since they will liquify again at_
higher temperatures. Gelatin combined with agar in nutrient
media can be incubated at 37°C and enables differentiation
between bacteria that liquify gelatin and those that do not.
Selective counting of gelatin liquifiers is accomplished
after determination of the colony count by placing 5 ml of
ammonium sulfate solution on the nutrient medium and count-
ing colonies that become surrounded by clear halos after a
few minutes. Silica gel retains its gelling properties
regardless of variations in quality, but is not commonly
'used, mostly from lack of familiarity and consequent dif-
ficulty in preparing.
Studies compared yeast extract with meat extract at 3.7
and 20°C and found a more rapid development at 37°C and
higher bacterial numbers at 20°C using the yeast extract.
Numbers were either the same or lower when glucose was added
(Commission of the European Communities, 1977). In addition,
results from pour plates using plate count agar (American
Public Health Association, 1976) were in/agreement with . . .:
peptone yeast extract medium, proposed by the ISO (Commis-
sion of the European Communities, 1977).
(ii) Method for Determining Colony Counts. Very
simply, the procedure entails mixing the sample, then trans-
ferring 1.0 or 0.1-ml volumes of diluted or undiluted sample
-------
172
into culture plates, and adding 10 ml of melted nutrient
medium. The covered plates are swirled in a motion describ-
ing a figure 8, are allowed to solidify (within 10 min) on a
level surface, and are then incubated. Colony counts are
recorded from plates containing 30 to 300 colonies, and
counts are made using a Quebec counter or a counter that
gxves equivalent magnification and illumination. Densely
grown colonies may be counted by taking an average of the
colony numbers within 1 cm at six different sites on the
dish, provided the colonies are evenly distributed.
The colony number is calculated by using the formula:
ab/c = G
where a = number of colonies counted per cm2; b = area of
the culture dish in cm ; c = volume of the water sample
transferred into the dish, in ml; and G = number of colonies
per ml of the water sample. The colony count is expressed
as a Dumber per 1 ml of sample water, to not more than two
significant digits. The colony number should be recorded
along with information on type of culture medium and incu-
bation time and temperature applied. Numbers of gelatin
liquifiers, if determined, also are reported [See Section
C.4.hJ.
(iii) Applications of the Colony Count. The colony
count is not, by itself, an indicator of fecal pollution
[See Sections C.l.b to d]. However, when used routinely to
monitor water supplies at least every three months it
can^provide baseline data on the general bacterial popu-
lation and aid in assessing the degree of bacterial pollu-
tion of that supply. it is an especially valuable tool,
when used in conjunction with other tests, for assessing the
purity of a possible new raw water source.
Used as part of routine analyses, the colony count
would reveal any changes in the bacteriological quality of
finished water in storage reservoirs and distribution systems.
Colony counts taken repeatedly at specific sites in a water
treatment plant, at different times of the year, and at
various points throughout the distribution system offer a
valid procedure for maintaining ongoing quality control [See
also Section C.4]. Application of this method would in-
directly limit the occurrence and level of Pseudomonas,
'Flavobacterium, and other secondary pathogenic invaders that
could be harmful in a hospital environment [See Section
B.l.a(ix)j. It would, in addition, serve to monitor the
-------
173
effectiveness of chlorine throughout the distribution network,
Also, it would alert water authorities to any deterioration
in filtration efficiency at the treatment plant or sediment
accumulation in the distribution network with consequent
build-up of bacteria.
Many filter media (e.g., charcoal, polysterol, and
asbestos) and ion exchange resins used in the home and
elsewhere have been known to support large increases in
bacterial populations (Muller and Herzel, 1973) [See Section
G.2]. The colony count method is well suited as an indi-
cator of such disturbances.
A sudden increase in the colony count from a given
source can serve as an early indication of contamination.
Indeed, sudden increases in colony counts which had been low
for several years have, in the past, indicated the sources
of waterborne outbreaks in the FRG. Such an increase in the
colony count occurred prior to the typhoid fever outbreak in
Hannover in 1926 (Mohrmann, 1927), the cause of which was
flooding of untreated well water with heavily contaminated
river water. High colony counts were observed in distri-
bution system water before it was found to be positive for
E. coli and total coliforms. Some 40,000 people became ill
with nonspecific gastroenteritis ("water disease") [See
Section B.l.d] two weeks before the first cases of typhoid
fever were reported. The same observations were made during
the typhoid fever outbreaks at Pforzheim in 1919 and at
Gelsenkirchen in 1889.
Large populations (^1,000 per ml) of noncoliform
genera, including Pseudomonas, Bacillus, Streptomyces,
Micrococcus, Flavobacterium, Proteus, and various yeasts can
suppress the growth of coliform bacteria below detectable
levels. This may well have been the reason that coliform
readings were negative just prior to a salmonellosis out-
break involving 16,000 cases in Riverside, California (U.S.)
[See Section B.l.a(i)]. The colony count, used in conjunc-
tion with coliform tests, would reveal such interferences by
noncoliform bacteria and thereby afford a more accurate
interpretation of results.
(iv) Recommended and Mandatory Maximum Limits. Opinion
varies in different countries as to the necessity or value
of incorporating total microbial counts into routine water
quality surveillance programs [See Section D.6]. Even in
places where these tests are routinely performed, water
-------
174
quality criteria will differ with respect to maximum allow-
able concentrations, evaluation of the counts, and the kind
of water that shall be characterized by the count (Miiller
1977b) [See Table C.l.a-1]. Except for bottled water (which
has legal limits imposed of 1,000 colonies per ml) [See
Section G.6], all other types of drinking water in the FRG
are subject only to recommended guidelines; however, German
drinking water law specifies the method by which the colony
count is to be conducted (Aurand, et al. , 1976).
The colony count is not a required test in the U.S.,
but many health departments, water supply agencies, and
local jurisdictions observe limits commonly applied within a
range of 100 to 500 colony-forming units per ml. Geldreich
(reviewed in Safe Drinking Water Committee, 1977) proposed
that a 500 per ml limit (at 35°C for 48 h) be placed on the
colony count and that immediate investigations of water
treatment and distribution systems be undertaken whenever
this limit was exceeded. The U.S. government has acknow -
ledged the _importance of the colony count in recommendations
of the National Academy of Sciences, entered into the Federal
Register (Environmental Protection Agency, 1977): "It [the
colony count] is, however, a valuable procedure for asses-
sing the bacterial quality of drinking water.... The SPC
[colony count] has major health significance for surface-
water systems that do not use sedimentation-flocculation-
filtration, and chlorination, and for those systems and
Csic] do not include chlorination."
(v) Summary. The colony count is not a substitute for
total coliform measurements of the sanitary quality of
potable water. However, if it is used in conjunction with
other tests, bacterial numbers and types determined by this
method can provide an integrated picture of conditions upon
which to base decisions for further microbiological testing,
epidemiological surveys, and repair or upgrading of treat-
ment plants and distribution systems.
b: Total Coliforms. The relationship between the
ingestion of polluted water and the occurrence of certain
diseases has been recognized since the beginning of recorded
history. Even before specific agents had been detected,
epidemiologists were able to show that some human activities
could give rise to disease. Von Fritsch's observation in
1882, that the presence of Klebsiella pneumoniae and K.
rhinoscleromatis signified fecal contamination, and Escherich's
-------
175
TABLE C.l.a-1
COLONY COUNT CRITERIA FOR VARIOUS TYPES OF WATER
IN SEVERAL EUROPEAN COUNTRIES
Country
Maximum
Allowable
Numbers/ml
Incubation
Temperature
Type of Water-
Poland
25
Czechoslovakia
Yugoslavia
20
100
100
500
10
100
300
37°C
20°C
37°C
20°C
37°C
37°C
37°C
Public supply
Public supply
Well water
Well water
Treated water
Raw groundwater
Raw surface water
Romania
20
100 to 300
Public water
supply for
70,000 customers
Other supplies
Switzerland
100
300
20
300
Raw water
entering works
Raw water during
distribution
Treated water
immediately
after treatment
Treated water in
the distribution
system
-------
176
TABLE C.l.a-1 Continued
Country
Spain
Netherlands ,
Sweden, GDR
Fed . Republic
of Germany
Maximum
Allowable
Numbers /ml
50 to 65
100
100
20
( recommended )
100
( recommended )
1,000
( recommended )
1,000
(mandatory)
Incubation
Temperature
37°C
37°C
20°C
20 + 2°C
20 + 2°C
20 + 2°C
20 + 2°C
Type of Water
Good quality
finished water
Tolerable quality
finished water
__
Disinfected water
All other kinds
of drinking
water
Water tanks
,
Bottled water
France,
Austria, UK,
Greece,
Israel,
Italy,
Netherlands,
US
No recommended
or mandatory
values
-------
177
isolation of Bacillus coli (Escherichia coli) from faces in
1885, mark the beginning of the science of sanitary water
bacteriology. . .
During the next twenty years, investigators were quick
to recognize the difficulties in isolating pathogens com-
pared to the relative ease in isolating B_. coli from pol-
luted waters (Heathman, et al., 1936). Before the turn of
the century, Theobald Smith had observed that the presence
of B. coli in water must be due to fecal discharges. Sur-
vival studies indicated that B. coli persisted longer in
water than did the typhoid bacillus, the most feared patho-
gen of the day (Heathman, et al., 1936) [See also Sections
B.l.a(i), -B.3, and E.I]. Thus, the rationale for the use of
B. coli as an indicator of the sanitary quality of water was
established, being based on: its regular association with
feces; its presence in water in numbers greater than those
of pathogens; and its superior survival capabilities. In
1905, when the first edition of Standard Methods of Water
Analysis was issued ,by the American Public Health Associa-
tion,B7 coli was recommended for use as an indicator of the
bacteriological condition of water supplies.
There are several operational definitions of the coli-
form group. These differ in a number of procedural details
[See Sections D.2 to 3]; but all tend to result in the
inclusion of organisms from the genera Escherichia, Kleb-
siella, Enterobacter, and Citrobacter, not all of which are
of fecal origin. For example, in North America, the 14th
edition of Standard Methods for the Examination of Water
and Wastewater (American Public Health Association, 1976)
defines the total coliform group as: all aerobic and facul-
tative anaerobic, gram-negative, nonsporeforming, rod-shaped
bacteria that ferment lactose with gas formation within 48 h
at 35°C; or, all organisms that produce a colony with a
golden-green metallic sheen within 24 h on an Endo-type
medium containing lactose. Members of the coliform group
also are motile by peritrichous flagellation or immotile,
they reduce nitrate ions to nitrite ions under anaerobic
conditions, and they are oxidase-negative,- a feature which
distinguishes them from Aeromonas. These definitions are
not to be regarded as identical but rather, as indicative of
organisms roughly equivalent in sanitary significance.
The total coliform group is used today in many coun-
tries to indicate the microbial quality of raw and finished
drinking water C'See Sections D.2 to 3], and apparently this
indicator system continues to ensure, with very few excep-
tions, that drinking water shall be safe for human consump-
tion.
-------
178
(i) Isolation Methods. Two standard methods are
available for coliform determinations: the multiple tube
fermentation procedure and the membrane filter technique.
In the multiple tube fermentation procedure, tenfold dilu-
tions of the water to be tested are inoculated into tubes
containing the appropriate medium (five tubes per dilution).
Dilution is unnecessary, however, for drinking water exami-
nation because of the expected low bacterial counts, so it
is possible to set up 1 x 50 ml and 5 x 10 ml samples. The
presence of gas (and acid) after incubation for 48 h at 35
or 37°C constitutes a positive presumptive test for coli-
forms and must be followed by confirmatory procedures.
Results are reported as a most probable number (MPN). The
MPN is a statistical estimate of the number of bacteria
that, more probably than any other number, would give the
observed result; it is not an actual count of the bacteria
[See Section D.3].
The membrane filter (MF) technique has, in recent
years, largely replaced the multiple tube fermentation
procedure for the routine examination of drinking water.
However, the multiple tube fermentation method is still of
value when conditions render the membrane filter technique
unusable (e.g., for water that is very turbid or heavily
populated with noncoliform organisms capable of growth on
the medium), and as a reference standard procedure.
The MF technique became accepted in bacteriological
water analysis when it was demonstrated to be capable of
producing results equivalent to those obtained by the
multiple tube fermentation procedure. With the MF tech-
nique, the water sample is passed through a filter of 0.45
urn pore diameter; the filter is placed on an appropriate
selective/ differential medium and incubated for 24 h, after
which time .the coliform colonies are counted. The technique
allows for large quantities of non-turbid water to be exam-
ined, thus increasing sensitivity and reliability while
markedly reducing time, labor, equipment, space, and mate-
rial needed [See Section D.2]. A current area of contro-
versy with the MF technique concerns the variable perfor-
mance of filters in the recovery of coliforms. Brown (1973)
reported that numbers of 13. coli were 15 to 18 percent less
on filters than on standard plate cultures. He attributed
the discrepancy to lack of nutrients. Variation also has
been ascribed to differences in filter sterilization proce-
dures, surface pore size, and the source of coliforms (Sladek,
et al_. , 1975; Tobin and Dutka, 1977).
-------
179
(ii) Evaluation as an Indicator. There are several
serious problems with the general use of total coliforms as
indicator organisms (except when testing finished drinking
water). Total coliforms can grow in water of rather low
organic matter content as well as in low temperature water
such as cold mountain streams. They have been recovered in
soil, on vegetation, in forest and farm products, and in
many other environments (Geldreich, e_t _al. , 1964) including
those almost untouched by humans. Although these organisms,
clearly are not specific for fecal pollution, they are
considered by some to be preferable to thermo-tolerant
coliforms [See Section C.l.c] as quality indicators for ,
finished water because of their greater resistance to dis-
infection. "Lin (1977) and others have shown that the total
coliform group is more abundant and relatively more resis-
tant to chlorine and other stresses than are thermo-tolerant
coliforms, and thus it serves as a more stringent measure of
water quality. If total coliforms are detected in finished
water their presence should not be ignored, but should
signal an investigation of the type and source of pollution.
There have been a few documented cases in which pathogenic
organisms were isolated from water that was negative for
coliforms (Dutka, 1973). This may have been due to in-
hibition of the coliforms by other microorganisms, a phe-
nomenon known to occur when colony counts at 35°C exceed 500
per ml (Geldreich, et aJL. , 1972) [See Section C.I.a].
(iii) Conclusion. The total coliform group was one of
the first indicator systems used and is still used in many
surveys. It has limited value as a direct indicator of
fecal pollution in raw source waters because of the ability
of some coliforms to multiply readily in the environment and
because of the methodological difficulties involved. Hence,
it is more appropriate to monitor raw water for the presence
of E. coli or thermo-tolerant coliforms [See Section C.l.c],
and~~to test finished water for the presence of total coli-
forms which serve as a sensitive indicator of disinfection
inadequacy or problems during distribution. It is used in
many countries as the primary microbiological test for
evaluating the quality of finished water and may continue to
be for some time to come, or at least until easier, faster,
and more sensitive indicator systems are well established
[See Sections C.2 to 4j.
-------
180
(iv) Summary. The total coliform measurement is one
of the oldest and most frequently used methods in sanitary
microbiology. It enumerates bacteria of the genera Escheri-
chia, Klebsiella, Enterobacter, and Citrobacter.
oJ
r Since some
these bacteria are not restricted to fecal origin, their
presence should not be construed as a specific indication of
contamination with fecal or pathogenic organisms, but rather,
as a cause for further investigation. On the other hand,
the total coliform group is more resistant to chlorination
and environmental stresses than many other indicators, and
so continues to be preferred by many water authorities as
the primary indicator system for evaluating the microbio-
logical quality of finished water.
c. Therjno-Tolerant Coliforms and Escherichia Coli.
For many years, the total coliform group [See Section C.l.b]
served as the main indicator of water pollution. However,
because many of the organisms in this group are not limited
to fecal sources, methods were developed to restrict the
enumeration to those coliforms which are more clearly of
fecal origin. A modified method devised by Eijkman (1904)
called for a higher incubation temperature and this was
refined further to distinguish what are known as thermo-
tolerant coliforms (also referred to as fecal coliforms; it
was the consensus of this working group that .the descriptor
"thermo-tolerant" is more accurate and more representative of
the thinking of scientists in the participating countries)
(Hajna and Perry, 1943? Geldreich, et _al. , 1958). This
indicator group has all the properties of the total coliform
group and, in addition, is able to ferment lactose with the
production of gas in 24 h at 44.5°C (American Public Health
Assoc., 1976). In general, this test enumerates organisms
of the genera Escherichia and Klebsiella. Klebsiella [See
Section C.2.d3 may derive from nutrient-rich, non-fecal
sources; a further procedural refinement permits testing
specifically for E_. coli.
E. coli is the only member of the coliform group that
unquestionably is an inhabitant of the intestinal tract;
hence, it has come to be the definitive organism for demon-
strating fecal pollution of water. E. coli has the defin-
itive properties of all members of the coliform group [See
S-ection C.l.b], but possesses the following additional
-------
181
characteristics: production of acid and gas from lactose at
44°C, glutamic acid-decarboxylase activity [See also Section
C.4.e], no pigmentation on nutrient agar during an incu-
bation of eight days (20°C), inability to utilize citrate as
a sole carbon source/ negative cyanide test, no urea or
gelatin decomposition, positive methyl red reaction, and nega-
tive Voges-Proskauer reaction. Most of the strains form
indole from tryptophan contained in tryptone broth and
produce acid and gas from glucose at 448C. It must be
understood that not all these tests are carried out for
routine water analyses, so the bacteria enumerated could
perhaps be best described as "presumptive E. coli".
E. coli meets the criteria of a valid fecal indicator
in that: it is present in the intestine in numbers larger
than those of enteric pathogens; it behaves similarly to
enteric pathogens within the aquatic environment; and it is
less susceptible than most enteric pathogens to treatment or
disinfection procedures. The presence of E_. coli in a water
supply indicates contamination with fecal material from
warm-blooded animals such as birds and humans. One must
assume that, if E. coli has gained access to a waterway,
enteric pathogens also may have entered this water. Indeed,
results from recent epidemiological studies (Cabelli, et
al., 1976) show a higher statistical relationship between
presence of E. coli and incidence of gastrointestinal illness
than other members of the Klebsiella-Enterobacter-Citrobacter
group.
(i) Methods for Enumerating Thermo-Tolerant Coliforms
and 12. Coli. Since it is impossible in water investigation
to test all characteristics essential for identifying
species, simplified procedures have been developed. These
must be followed precisely, for any deviation from the
prescribed procedure (e.g., use of a different nutrient
medium) may invalidate the results.
Methods used in various countries for enumerating
thermo-tolerant coliforms are essentially high temperature
modifications of the total coliform tests. As with the
total coliform test, there are two methods for enumerating
thermo-tolerant coliforms the membrane filter technique
[See Section D.4] and the multiple tube fermentation pro-
cedure [See Section D.5] upon which are based two separate
(but hygienically equivalent) definitions. The IMViC tests
-------
182
(indole, methyl red, Voges-Proskauer/ and sodium citrate)
are often used to differentiate between coliform isolates,
particularly to determine the presence of El. coli (American
Public Health Assoc., 1976). In the.FRG, as in several
other countries [See Sections D.4 to 5], there are very
specific protocols for identifying IS. coli (Muller, 1977b) .
Here, 12. coli is differentiated from coliform bacteria by
means of tryptone broth cultures according to the following
procedure.
A single colony of a pure culture is inoculated into
tryptone broth; after the culture becomes slightly turbid (4.
to 7 h), it is inoculated into the following nutrient media:
ammonium-citrate agar slant (the material should be distrib-
uted homogeneously on the agar slope), glucose broth, lac-
tose broth, nutrient agar plate, and Endo agar plate. All
nutrient media, including tryptone broth, are then incubated
for 20 +^ 4 h glucose broth at 44 + 0.5°C in a water bath,
the other nutrient media at 37 + l.CPc in ari air incubator.
After incubation, the Endo and nutrient agar plates are
examined to determine whether all colonies show the same
morphology. If not, fresh-Endo agar plates must.be inocu-
lated from well isolated single colonies picked from the
nutrient agar plate, and then incubated.
If a pure culture is present on the nutrient agar
plate, the cytochrome oxidase test is performed by putting
two to three drops of dimethyl-p-phenylenediamine on the
colonies. The reaction is considered positive if the colo-
nies turn blue within 1 to 2 min; if there is no color
change, the reaction is negative. A positive cytochrome
oxidase reaction means that colonies are not coliform bac-
teria (they may well be Aeromonas [See Section C.2.h] or
Pseudomonas CSee Section C.2.c], and no further testing is
required. A negative cytochrome oxidase reaction and a
positive reaction in lactose broth (i.e., gas and acid pro-
duction) confirms the presence of coliform bacteria in the
water sample. If acid is detected (brom cresol purple
indicator turns from purple to yellow) without gas being
present, the lactose broth is incubated another 24 h (48 h
total elapsed time) and checked for gas, the presence of
which is considered a positive result. If the result of the
lactose broth culture is negative, there are no coliform
bacteria present and no further tests are indicated. When
coliforms are found to be present, the following biochemical
-------
183
reactions are performed to determine the presence of E_.
coli; ~
1. Glucose Broth. If gas and acid are detected
after incubation at 44 +_ 0.5°C (indicator
changes from purple to yellow), the reaction is
considered positive; no gas production, whether
or not acid is present, indicates a negative
result.
2. Tryptone Broth. Several drops of indole reagent
are added to the incubated broth and the tube
is agitated gently. The reaction is considered
positive if a red surface layer forms in the
culture within 1 to 2 min. If the surface
layer remains yellow, the reaction is negative.
3. Ammonium Citrate Agar Slant. If visible colo-
nies appear on the culture medium or if the
indicator has changed from green to blue (even
without detectable growth), the reaction is
positive. Otherwise it is considered negative.
12. coli is confirmed to be present in the water under inves-
tigation by positive results in glucose broth (gas production)
and in tryptone broth (indole positive), and by negative
results on ammonium citrate agar.
(ii) Evaluation of Thermo-Tolerant Coliforms and E_.
Coli as Indicators. Recent research has confirmed the
validity of the thermo-tolerant coliform test as an indi-
cator of the potential presence of enteric pathogens in
water. It has been shown that the frequency of Salmonella
detection in water is related to the density of thermo-
tolerant coliforms (Geldreich, 1970; Van Donsel and Geldreich,
1971). At thermo-tolerant coliform densities of 1 to 200
per 100 ml, Salmonella was detected in 28 percent of water
samples examined; this frequency rose 85 to 98 percent in
waters with therjno-tolerant coliform counts above 2,000 per
100 ml. Studies on survival in river water (Mitchell and
Starzyk, 1975), well water (McFeters, e_t al., 1974), and
septic tank effluent (Calabra, 1972) have shown that thermo-
tolerant coliforms persist longer than salmonellae CSee
Sections B.3 and E.I]. Because they are more specific for
fecal contamination, thermo-tolerant coliforms and IS. coli
are preferred over total coliforms for monitoring raw water
for the potential presence of pathogens [See also Section
C.l.b]. IS. coli also is preferred in several European
-------
184
countries for use in monitoring finished drinking water [See
Section D].
(iii) Conclusion. Thermo-tolerant coliforms and E.
coli are valuable indicators of fecal pollution in many
situations where total coliforms are less applicable because
of their widespread occurrence in the environment. When the
means are available, E. coli is a preferred indicator for raw
water because it excludes most of the Klebsiella organisms
[See Section C.2.d] which may or may not originate from
fecal sources. Because of its unrivaled specificity for
fecal pollution, !E. coli also is used for routine monitoring
of finished drinking water, even though it is somewhat les-s
resistant than total coliforms to chlorine and other dis-
infectants. .
(iv) Summary. Coliform organisms can be further
subdivided into thermo-tolerant coliforms (those capable of
growth at 44°C) or IS. coli, one of the principal species
making up the thermo-tolerant coliform group. These latter
two indicators are considered more specific for fecal pollution
than total coliforms and are preferred for monitoring raw
water quality. They also are the primary indicators used
for finished water in several countries.
d. Fecal Streptococci. The three most commonly
employed indicator groups for establishing the presence of
sewage in a water supply, are total coliforms, thermo-tolerant
coliforms, and fecal streptococci. All three groups occur
in feces and feces-polluted waters in greater numbers than
pathogens and fit all the other criteria for a valid indicator
system [See Section C.I.Intro.]. However, drawbacks to
total coliforms as indicators in raw water have been noted,
due to their presence in the environment remote from any
fecal pollution, and growth in nutrient-enriched waters [See
Section C.I.b] (Clausen, 1977; Geldreich, 1970). Thermo-
tolerant coliforms, a more restricted subgroup of the total
coliform group, occur only in association with feces and
thus are a better indicator of raw water quality. However,
thermo-tolerant coliforms are more sensitive to conditions
outside the animal intestinal tract than total coliforms,
and possibly some pathogens including viruses, and thus may
be subject to excessive die-off [See Section C.l.c].
-------
185
Fecal streptococci serve as an index of fecal pollution
of raw water and may provide valuable supplementary
data when used in conjunction with thermo-tolerant coliforms.
A ratio of thermo-tolerant coliforms to fecal streptococci
can be used to identify the source of pollution as being
either human- or nonhuman-derived. This is possible because
humans have greater numbers of thermo-tolerant coliforms than
fecal streptococci in their feces, whereas the reverse is
true of other warm-blooded animals. In addition, the fecal
streptococci are more resistant than coliforms to disinfection,
are capable of longer survival in the environment, and do
not exhibit aftergrowth, as is true of the coliforms. The
fecal streptococci should therefore be considered a valuable
indicator group for raw water quality when used as an adjunct
to thermo-tolerant coliforms.
Fecal streptococci originate in the intestinal tracts
of warm-blooded animals and are discharged with feces. They
are easily detected in water because of characteristics
which readily separate the fecal streptococci from other
groups of bacteria including other gram-positive cocci. The
fecal streptococci belong to the family Streptococcaceae and
include catalase-negative, non-motile, gram-positive cocci
that are indifferent to oxygen (Buchanan and Gibbons, 1974).
The term fecal streptococcus in this discussion, as well as
in most of the literature, is used as a synonym for Lancefield's
group D streptococcus, but these terms are not precisely
equivalent. Streptococci (e.g., S_. mitis and S_. salivarius
and the group Q streptococci of fowl!other than those
possessing the group D polysaccharide antigen between their
cytoplasmic membrane and cell wall may occur in feces and
thus may also be termed fecal streptococci (Clausen, et
al., 1977); however these are not the streptococci that
originate in the intestinal tracts of most warm-blooded
animals. The group D streptococci (or fecal streptococci of
sanitary significance; American Public Health Association,
1976) include:
Group D Streptococci
Streptococcus faecalis
Streptococcus faecalis variety liquefaciens
Streptococcus faecalis variety zymogenes
Streptococcus faecium
Streptococcus faecium variety durans
Streptococcus faecium variety casseliflavus
Streptococcus bovis .
Streptococcus equinus.
-------
186
The first two species and their varieties are termed the
"enterococci" and are the fecal streptococci that may be of
human origin. In addition to these two species, feces'of
nonhuman warm-blooded animals also contain varying concentra-
tions of S_. bovis and £3. equinus (Geldreich and Kenner,
1969) [See Table C.l.d-1], The latter two species have been
associated with pollution involving meat processing plants,
dairy wastes, cattle feedlots, duck farms, and runoff from
livestock pasture land (American Public Health Association,
1976? Geldreich, 1970); the detection of either of them
indicates that the spurce of pollution is from animals other
than humans.
(i) Isolation Methods. As is true for coliform testing,
determinations of fecal streptococcus densities in water may
be done by the multiple tube fermentation procedure or the
membrane filter procedure. The mechanics and comparisons
of these two general procedures are described elsewhere [See
Section C.l.b]. In addition, fecal streptococci may be
detected by a pour plate procedure (American Public Health
Association, 1976).
The most probable number (MPN) method for determining
fecal streptococcus concentrations, like the procedures for
total or thermo-tolerant coliforms, includes both a presumptive
and confirmed step (American Public Health Association,
1976). In the U.S., the presumptive medium used is azide
dextrose (AD) broth. Replicate sets of three or five tubes
of the medium are inoculated with tenfold dilutions of the
water sample and incubated at 35°C. Growth constitutes a
positive test after either 24 or 48 h of incubation. All
tubes demonstrating turbidity are transferred to ethyl
violet azide (EVA) broth. Growth in EVA, after 24 to 48 h of
incubation at 35°C, is confirmed evidence of the presence of
fecal streptococci. The MPN is taken from the number of
positive EVA tubes.
Although the media in the U.S. for the multiple tube
procedure have been adopted as part of their standard method
(American Public Health Association, 1976), other media may
be substituted for AD and/or EVA. Kibbey and coworkers
(1978) have achieved higher fecal streptococcus recoveries
when substituting KF streptococcus (KFS) agar (Geldreich and
Kenner, 1969) or m-Enterococcus (ME) agar for EVA in the
confirmed procedure, following incubation in either AD, KFS
broth, or ME broth. The combination of KFS or ME presumptive
broths followed by KFS or ME agar as the confirmed media
-------
TABLE C.l.d-1
THERMO-TOLERANT COLIFORM (TTC)/FECAL STREPTOCOCCUS (FS) RATIOS AND
FS DISTRIBUTIONS IN WARM-BLOODED ANIMALS
Fecal Source
Hujnan
Animal pets
Cat
Dog
Rodents
Livestock
Cow
Pig
Sheep
Poultry
Duck
Chicken
Turkey
Ratio
TTC/FS
4.4
0.3
0.02
0.04
0.2
0.04
0.4
0.6
0.4
0.1
Occurrence (%)
Enterococci
73.8
89.9
44.1
47.3
29.7
78. "7
38.9
51.2
77.1
76.7
S. Bovis
S. Equinus
None
1.5
32.0
17.1
66.2
18.9
42.1
48.8
1.1
1.6
Atypical
S. Faecalis
None
2.2
14.4
0.4
Hone
None
None
None
None
None
S. Faecalis Variety
Liquefaciens
26.2
6.3
9.6
35.3
«- .
4.1
2.4
19.0
None
21.8
21.8
From Geldreich and Kenner, 1969.
00
-------
188
have yielded the greatest number of streptococcal isolates
(Kibbey, et al., 1978). The use of ME medium, however, has
been discouraged by the American Public Health Association
(1976) due to its poor recoveries of £>. bovis and £3. equinus;
whereas these two species may be recovered on KPS medium
(Clausen, et. al^., 1978).
The use of membrane filtration (MF) or agar pour plates
for fecal streptococcus detection is recommended over the
MPN procedure because recoveries on the MF and pour plate
media are higher and less affected by interfering organisms,
and greater numbers of false positive reactions occur in the
AD/EVA broth MPN system (American Public Health Association,
1976). Also the MPN method using AD/EVA broths detects only
the enterococci (Clausen, jet ail., 1978). For the MF procedure,
filters are incubated on KFS agar at 35°C for 48 h, at which
time all dark red to pink colonies (due to the reduction of
the tetrazolium salt indicator) are counted (American Public
Health Association, 1976). Turbid waters or chlorinated
effluents cannot be analyzed by the MF procedure and one of
the other methods must be employed.
An alternative method to either the MPN or MF method
for fecal streptococcus detection is a pour plate procedure
(American Public Health Association,>1976). The procedure
should be used preferentially over the MF technique for
those samples containing few fecal streptococci associated
with significant turbidity. The only disadvantage to this
procedure is that the sample volume that may be tested is
usually limited to 1 ml. Two media have been recommenced
for use in the U.S.: KFS agar and Pfizer selective entero-
coccus (PSE) agar. Surface and subsurface colonies produced
by fecal streptococci after 48 h of incubation at 35°C are
dark red to pink. Few nonfecal streptococcus colonies will
be observed on KFS agar because of the selectivity of the
medium (American Public Health Association, 1976). Stream
samples may sometimes contain gram-positive soil 9rganisms
that may grow on this medium, such as Corynebacterium or
Bacillus species (American Public Health Association, 1976).
Other false positive organisms reported to grow on this
medium include: Pediococcus, Lactobacillus, and Staphylococcus
(Clausen, et al., 1977); but none of these organisms reduce
tetrazolium and thus, will not form a red colony. Fecal
streptococci on PSE agar produce brownish-black colonies
with brown halos, due to the formation of iron salts of
hydrolyzed esculin, after 24 h of incubation at 35°C. The
only gram-positive cocci that will grow and exhibit esculin
hydrolysis are the group D streptococci. Listeria monocytogenes,
-------
189
the only other organism capable of both growth and esculin
hydrolysis on PSE agar, shows only pinpoint colonies at 24
h, and coloration of the colonies is less marked with Listerja
than with the fecal streptococci (American Public Health
Association, 1976). The advantage of PSE agar over KFS agar
is the shorter incubation time of 24 h for PSE agar as
compared to 48 h for KFS agar. Unfortunately PSE agar does
not lend itself to use with the MF technique unless a PSE
agar overlay is added after the MF has been applied to the
medium. The overlay allows visualization of esculin hydro-
lysis directly above the colonies without having to turn
over the plate to see the blackening of the medium under the
MF (Clausen, et al., 1977).
All the media used for the detection of fecal strep-
tococci are dependent on the inclusion of the metabolic
inhibitor, sodium azide, which poisons hemeporphyrin com-
ponents (e.g., the enzyme catalase and the cytochromes of
the respiratory chain) of cells. Because the fecal strep-
tococci lack any porphyrin-containing enzymes or coenzymes,
they survive in the presence of this poison, whereas any
aerobic or facultative anaerobic organisms should be com-
pletely inhibited. ,
Although the detection of fecal streptococci is rela-
tively straightforward, and the selective media which are
recommended yield few false positives, their speciation is
extremely tedious and several different taxonomic schemes
have been proposed (American Public Health Association,
1976; Buchanan and Gibbons, 1974; Geldreich and Kenner,
1969; Kibbey, et al., 1978) which have led to considerable
confusion. The results of any identification are determined
by the taxonomic scheme which is employed (Kibbey, et al.,
1978). Characteristics of the enterococci used in most
identification schemes include: growth at 10 and 45°C,
growth at pH 9.6, tolerance to 60°C for 30 min, tolerance to
6.5 percent NaCl, growth in 40 percent bile broth, and
reduction of'0.1 percent methylene blue. To identify the
varieties of enterococci and to separate out J3. bovis and
S. equinus, other biochemical tests must be employed.
(ii) Evaluation as an Indicator of Raw Water Quality.
The fecal streptococci are a diverse group of organisms that
have variable fecal origins and survival characteristics;
they include several biotypes of limited sanitary signif-
icance (Geldreich, 1970; Geldreich, 1979). Usually the
detection of fecal streptococci in water denotes fecal
pollution; but if they were to be used as the sole indicator
-------
190
system, the existence of strains that have an environmental
origin should be noted. £5. faecium variety casseliflavus is
described by Mundt and .Graham (1968) as being plant-derived
and possessing a yellow pigment (unlike other streptococci).
It occurs in high numbers on vegetation, but rarely in
association with, fecal material (Mundt and Graham, 1968),
although it also "has been reported to be of human origin
(Ator and Starzyk, 1976). It appears to be a transitional
organism that "has properties of both S_. faecalis and j3.
faecium (Ator and Starzyk, 1976). _S. faecalis variety
liquefaciens also occurs in the environment in association
with plants, insects, and certain types of soil, in addition
to occurring in animal feces in high proportions (American
Public Health Association, 1976 ; Clausen, et al., 1977;
Geldreich, 1970). Geldreich (1970) has found high numbers
of this organism in unpolluted wells, perhaps due to contami-
nation of the wells by soil or insects. An atypical biotype
of _S. faecalis capable of starch hydrolysis may be isolated
in association with vegetation, but rarely from feces.
Recovery of the starch-hydrolyzing strain is uncommon from
water, soil, insects, or warm-blooded animals. Group D
isolates of human origin, by contrast, were found not to
hydrolyze starch (Clausen, e_t aJL., 1977).
The plant-associated streptococci of nonfecal origin
closely resemble the streptococci of fecal origin (Kibbey,
et a^., 1978), and low numbers of certain biotypes (e.g.,
S_. faecalis variety liquefaciens) are present even in good
quality waters free of fecal pollution (Geldreich, 1970);
although in waters from vegetable-processing plants their
numbers may be increased. Therefore, if high numbers of
fecal streptococci are detected in the absence of thermo-
tolerant coliforms, misinterpretation may be avoided if the
species of the fecal streptococci are determined, so as to
learn whether environmental rather than enteric varieties
are involved.
Fecal streptococcus densities may be best used in
association with thermo-tolerant coliforms as a ratio used
to assign the probable source of waste discharge as being
either human or from farm animals and wildlife. Fecal
streptococcus densities are significantly higher than thermo-
tolerant coliform densities in all warm-blooded animal feces
except for those of humans as shown in Table C.l.d-1 (Geldreich
and Kenner, 1969). In human feces, the thermo-tolerant
coliform to fecal streptococcus ratio is always greater than
four, whereas ratios for all other warm-blooded animal feces
-------
191
are always less than 0.7. However, this ratio is influenced
by die-off and is only valid at the waste outfall or during
the first 24 h after discharge into the receiving waters
(Geldreich'and Kenner, 1969). Due to environmental factors
such as death, survival, or aftergrowth, the proportions of
fecal streptococci and thermo-tolerant coliforms in water-
may be altered (Geldreich, 1979). .
Table C.l.d-1 illustrates the distribution of the fecal
streptococci in warm-blooded animals including man. _S.
bovis and £3. equinus do not occur in the feces of humans and
therefore lire considered specific indicators of npnhuman
animal pollution; however, S_. bovis and S. equinus are the
indicator organisms most sensitive to rapid die-off outside
the animal intestinal tract (Geldreich, 1970; Geldreich,
1978; Geldreich and Kenner, 1969). After 24 h at either 10
or 20°C water temperatures, less than 10 percent of £3.
bovis can be recovered, and even shorter survival times are
seen with £». equinus (Geldreich and Kenner, 1969) .
Although limited survival of S. bovis and £3. equinus
occurs when outside the intestinal tract, survival of the
other fecal streptococci in the environment is greater.
During favorable temperatures and in waters with high elec-
trolyte contents, survival of fecal streptococci is enhanced
(Geldreich, 1970; Geldreich, 1978) . S. faecalis and J3.
faecalis variety liquefaciens persist longer in water than
either thermo-tolerant coliforms or Enterobacter aerogenes
(Clausen, e_t al. , 1977). Unlike the total or thermo-tol-
erant coliforms, the fecal streptococci rarely mutiply in
polluted water, with the exception of the naturally-occur-
ring _S. faecalis biotypes that may replicate in vegetable
processing wastewaters (Clausen, e_t .al_. , 1977; Geldreich,
1978). Fecal streptococci are also more resistant to chlo-
rine, and populations surviving disinfection are not as
subject to aftergrowth as are total and thermo-tolerant
coliforms. The apparent resistance of fecal streptococci to
disinfectants and adverse environmental conditions is probably
due to the clustering of the organism in chains: in a
reaction with a disinfectant, all cells of a chain or cluster
must be destroyed, or the surviving cells will again mani-
fest growth of the chain or cluster (Berg, 1978). With the
extended survival of the fecal streptococci beyond that of
total or thermo-tolerant coliforms, and because they are
heavily clumped, they more closely parallel survival of the
enteric viruses (Berg and Metcalf, 1978; Clausen, et al.,
1977; Kenner, 1978). Similar rates of destruction for both
,fecal streptococci and enteric viruses have been observed in
mesophilic digested sludge, raw fresh waters, and possibly
may occur in ocean waters (Berg and Metcalf, 1978; Kenner,
1978).
-------
192
(iii) Conclusions. Like the total and thermo-tolerant
coliforms, the fecal streptococci may be used as fecal
pollution indicators in water because all three of these
groups are present in the feces of humans and warm-blooded .
animals in higher concentrations than pathogens and also are
present in feces-contaminated Waters. In raw water, fecal
streptococci are less numerous than total coliforms, which
are ubiquitous in nature and do not necessarily indicate
fecal contamination. Although the fecal streptococci occur
in low numbers within environmental sources not necessarily
associated with feces, their presence in high numbers,
especially if thermo-tolerant coliforms also are present,
indicate fecal additions have occurred. Because of the
environmentally-associated biotypes, £^. faecalis variety
liquefaciens, the atypical j3. faecalis, and £^. faecium
variety casseliflavus, it is important to include thermo-
tolerant coliform densities with fecal streptococcus den-
sities to determine whether fecal pollution actually has
occurred, and to define the polluting source by the thermo-
tolerant coliform to fecal streptococcus ratio.
Because fecal streptococci are not often found asso-
ciated with plants in pristine environments, it appears that
plant populations of fecal streptococci originate from a
fecal source and establish a reservoir which then may be
spread to other vegetation (Clausen, et al., 1977). To
compensate for the low levels of the naturally occurring
biotypes, fecal streptococcus densities of less than 100 per
100 ml should be considered of little sanitary significance
(Geldreich, 1970; Geldreich, 1978).
Unlike the total coliforms, the fecal streptococci
rarely multiply in polluted waters, but survive in waters,
especially those containing levels of high electrolyte, at
favorable temperatures (Geldreich, 1970; Geldreich, 1978).
Although the two species occurring in the feces of nonhuman
warm-blooded animals (i.e., £3. bovis and £3. equinus) survive
for only short periods in the environment, the other fecal
streptococci (i.e., the enterococci) have much greater
survival capabilities outside the animal host than either
£>. bovis or j3. equinus and also as compared to either total
or thermo-tolerant coliforms.
Due to the species-specific survival characteristics
and ubiquitous distribution of certain biotypes in the
environment, it is not recommended that the fecal strepto-
cocci be used as the sole fecal pollution indicator. Other
fecal indicators should be used concurrently if using the
-------
. 193
focal streptococci to investigate water quality (American
Public Health Association, 1976). Improved assurance of the
microbiological safety of raw water can be gained by using
both the thermo-tolerant coliforms and fecal streptococci as
fecal pollution indicators.
e. Sporeformers
(i) Definitions and Source of Indicators. Anaerobic
(genus Clostridium) and aerobic (genus Bacillus) spore-
forming species can be applied as indicators of pollution in
the examination of drinking water. Bacillus species often
constitute a considerable part of the colonies counted on
plates, especially at 37°C; these organisms can accumulate
in sand filters, and a sudden increase in numbers can indi-
cate a breakthrough.of the filter. However, an increased
number of aerobic sporeformers may also indicate specific
pollution (Schubert, 1975a); and the species demonstrated
may give an indication as to 'the source, E. licheniformis
being associated mainly with sewage; 33. cereus, JB. mega-
terium, and B. sphaericus with soil; and IJ. brevis with
filter sand TBonde, 1977).
The enumeration of anaerobic sporeformers (clostridia)
is based either on counting all sulfite-reducing clostridia
that grow at 37°C or on a specific count of £. perfringens.
A short definition of C. perfringens applicable in
water quality examination is: gram-positive, anaerobic,
sporeforming (although spores are infrequently seen), non-
motile rods, which reduce sulfite to sulfide, give stormy
fermentation on milk, produce lecithinase, hydrolyze gela-
tin, ferment sucrose but not mannitol, and do not form
indole. Demonstration of sulfite reducers necessitates
pasteurization of the sample to avoid interference from non-
sporeforming organisms (Vibrio, Salmonella, Arizona,
Edwardsiella, Citrobacter, Proteus mirabilis and P_. _
rettgeri) that can form H^S from inorganic or organic sulfur
compounds. A wide variety of clostridia can reduce sulfite
at 37°C (Bonde, 1962), including C. perfringens, C. sporo-
genes, C. feseri, C. fallax, C. septicum, C. sphenoides, C.
bifermentans, C. parasporogenes, C. botulinum, £. oedematiens,
C. roseum, C. tertium, and C. cochlearium. Not^all of
these specie's are as suggestive of fecal pollution as £.
perfringens, which on the other hand is often capricious
with regard to sporulation and might in some instances be
-------
194
lost or suppressed by the other sulfite reducers. The
aerobic sporeformers 13. polymyxa and J3. macerans have some-'
times been mentioned as sulfite reducers, but this has not
been confirmed (Bonde, 1977).
Applying selective media and procedures, combined with
incubation at 48°C, reduces the number of species enumerated
to five, of which C_. perfringens is more important in pol-
luted waters than the other four species (C.-. bifermentans,
£. feseri, C. sporogenes, and C. tertium), all of which
generally grow more slowly.
(ii) Sources arid Significance of C2. Perfringens in
Water Examinations. Bonde (1962) gives a review of the
classification, production of toxins, and pathogenicity, as
well as of the occurrence in soil, water, and sewage of this
organism. The most controversial point is whether this
organism is really ubiquitous and mainly found as spores in
nature, or dependent upon pollution and found as vegetative
forms in the neighborhood of the source of pollution.
Before 1900, £. perfringens was considered to be pri-
marily a fecal and pathogenic organism and, as such, to have
greater indicator significance than the coli bacteria.
Pollution by sewage could be detected at a ratio of 1:500,000
by means of the "enteritidis test" (for C. perfringens)
which was 25 times more sensitive than chemical tests.
Wilson and Blair (1925) considered the demonstration of
cells in the negative form to be particularly valuable, and
stressed that CJ. perfringens is an organism of indisputably
fecal origin that ,is of great importance for the detection
of intermittent and occasional pollution. Many others (cf.
survey by Bonde, 1962) considered a test for £. perfringens
to be a valuable supplement to other examination methods.
Windle Taylor (1958) emphasized that when water is satis-
factory from a sanitary point of view according to all
other criteria, it has never been possible to demonstrate
the presence of CJ. perfringens, and that, this organism is a
particularly useful indicator in cases where the coli-aero-
genes group fails.
Willis (1956) expressed moire negative assessment of
£. perfringens as an indicator after finding this bacterium
in large numbers in soil samples from areas around bore-
holes. He also found poor correlation between counts of
coliforms and of C. perfringens and showed that filter sand
contained this organism in great numbers. He concluded from
-------
195
his counts of vegetative cells that £. perfringens is able
to multiply in tapwater. Windle Taylor (1958)also found
C. perfringens in samples of filter sand, vegetative cells
being predominant. However, considering that waterbearing
strata are subject to contamination arid that defects in sand
filters are common, it is surprising that such adverse
interpretations are made from these findings.
C. perfringens is also very important in water samples
becau₯e it may cause "false positives" in the lactose fermen-
tation test for coliforms more frequently than any other
single organism, sometimes in nearly half of the samples.
Windle Taylor (1958) considered this species a predominant
cause of false positive results in testing filtered and
chlorinated waters: C. perfringens accounted for 22.7
percent of the total of 32.6 percent false positive reactions
Bonde (1962) did not find a regular correlation between
counts of C. perfringens and E. coli, but stressed that such
a correlation should not be expected. The initial numbers
of both species are functions of fecal pollution, but the
ecology, persistence, and survival of the two differ mark-
edly. C. perfringens and E. coli must both be considered
fecal organisms and are excreted together with potentially
pathogenic organisms. On the whole, the counts of thermo-
tolerant coliforms are usually greater, by a factor of about
100, than counts of £. perfringens regardless of whether the
latter are done with nonpasteurized samples or as counts of
spores. In some cases, however, C. perfringens may give
higher counts tnan 12. coli and may even be present despite
negative tests for ~E_. coli in as much as 46 percent of
tapwater samples, whereas among water samples in general
that contain E. coli, 85 percent lacked C. perfringens
(Bonde, 1962).
Those opposed to the use of C. perfringens as an indi-
cator have suggested that -.its.: numbers, in samples, are too
small to be useful. However, improved methods now often
reveal higher numbers of C. perfringens than of E_. coli
(Bonde, 1962).
(iii) Enumeration Techniques. Several methods and
media are available for C:. perfringens counts (Bisson and
Cabelli, 1979; Bonde, 1962; British Department of Health and
Social Security, 1969; Him and Raevuori, 1978; Windle
Taylor, 1958). The methods differ mainly in using media
that are liquid or solid and rich or lean in nutrients, in
incubation at 37°G or at 45 to 48°C, and in pasteurization
-------
196
of the inoculum or not. Liquid media are used for most
probable number methods and solid media are used for colony
counts.
An improved version of the Bonde pour-tube method uses
a lean medium (meat extract 1 percent, peptone 1 percent,
agar 0.75 percent, pH 7.2, in 10-ml amounts in tubes to
which ^ is added 2 ml 1 percent anhydrous sodium sulfite
solution, two drops of 5 percent ferrous alum, and 5 ml
inoculum). Ten tubes of pasteurized (at 80°C for 5 min) as
well as of nonpasteurized samples are incubated at 48°C for
24 h. Based on a study of about 13,000 counts and confir-
mation by gram stain, motility, stormy fermentation-reaction,
and fermentation of mannitol, colonies that had attained a
size of more than two mm were confirmed as C. perfringens
in more than 90 percent of cases, except in~samples from
polluted fresh water courses (75 percent). Lack of confir-
mation was, in most cases, due to failure to grow and not to
the presence of other sulfite-reducers. The method can,
therefore, be applied with good results as a presumptive
test without confirmation. The same medium can also be used
with membrane filters laid onto it.
(iv) Conclusions. Enumeration of C_. perfringens
compares to the requirements for an ideaT indicator system
as follows:
1. An indicator bacterium must always be present
when pathogenic organisms are present. C.
perfringens is a fecal organism and will~be
excreted together with possible enteric patho-
genic organisms.
2. It must be present only when the presence of
pathogenic organisms is a possibility. This
rigorous requirement is not fulfilled by any of
the known indicator species; Bonde (1977) has
shown that C. perfringens is widespread but
not ubiquitous, so it meets this criterion as
well as other established fecal indicators.
3. It must be excreted in greater numbers than the
pathogenic organisms. This appears to be true
of C. perfringens now that the methods devel-
oped" are yielding numbers comparable to those
for E. coli in feces;
-------
197
4. It must be mpre resistant, both to disinfec-
tants and to hostile aqueous environments, than
the pathogens. This condition is fulfilled by
C. perfringens to a greater extent than by any
other indicator.
5. It must display characteristic and simple
reactions during growth, enabling rapid and,
preferably, unambiguous identification; this is
the case with C. perfringens.
6. The growth of the indicator bacterium should be
independent of other species in the presumptive
medium. This requirement is fulfilled by C.
perfringens in sulfite-alum agar at 48°C.
C. perfringens may be preferable to IS. coli as an
indicator wheri .examining samples that may contain substances
toxic to microorganisms, including samples of chlorinated
water, and in the examination of samples for which transpor-
tation problems are such that the samples cannot be analyzed
within 12 h of collection. In all single examinations and
all first examinations of untreated water, determinations of
C. perfringens should be added to the usual tests. This
₯pecies may be an especially valuable indicator in exception-
ally hot or cold climates.
(v) Summary. Anaerobic, sporeforming clostridia can
be applied as indicators of pollution in the examination of
drinking water: the species of choice for this purpose is
£. .perfringens. Techniques for the enumeration of this
fecal organism have been so refined as to allow them to be
used with as much sensitivity as the methods for E. coli, in
most cases. In general, good correlations are not found, .
nor are they expected, between C. perfringens and E. coli ,
in environmental samples. C. perfringens is evidently
widespread, but not ubiquitous in the environment; its
detection in water must be considered evidence of fecal
contamination. ,
2. Proposed Viable Indicators
For any of several reasons, the organisms in this group
are not presently included under the designation, "established
viable indicators." In some cases, the detection.method is
-------
198
too complex or time-consuming, or too recently developed.
Not enough is known about some of the organisms to enable
correct interpretation of results in all situations. What-
ever drawbacks currently preclude their use may or may not
be subject to revision in the future. The following criti-
cal evaluations will attempt to discern their utility and
scope of application as indicators in light of the latest
data.
a. Coliphages . Twort in 1915, and d'Herelle in 1917,
were the first to report bacterial lysis caused by a virus.
d|Herelle named these viruses "bacteriophage" (i.e., bacte-
ria-eating); the name is frequently shortened to "phage."
Phages, each of which are specific for a single species
of bacterium, have been isolated from fresh and seawater, as
well as from wastewater. Those which attack Salmonella,
Shigella, Pseudomonas, Staphylococcus , Vibrio , and other
bacteria are described in the literature, as are phages that
attack yeasts and cyanobacteria.
Best characterized are the "coliphages" (phages which
infect E. coli), especially the T-phages; others designated
MS,,/ f2, and {6X174 have also been studied extensively.
Coliphages are shed at high levels by humans and other warm-
blooded animals and are present in the feces whether or not
enteroviruses are.
^ coliphages are not usually specific as to the
species of animal in whose feces they occur, but they are
quite selective as to which strains of E. coli they will
infect. This specificity is determined~~by the coat protein
(e.g., the tail structures in the case of the T-phages) of
the virus particle, which must attach to a homologous
receptor on the surface of the bacterial cell before the
viral nucleic acid can be injected into the host. Only
those coliphages that express themselves overtly (i.e.,
carry out a full replicative cycle) in cell culture can be
used as indicators. Those that do not cause cell lysis
within the 6 h incubation period lysogenic phages are
not useful indicators .
The presence of coliphages (shed in all human feces)
indicates that fecal material, possibly containing surviving
pathogenic enteroviruses, is present. Since at least some
coliphages are more resistant to environmental conditions
and chlorination than most enteroviruses, the elimination of
-------
199
the latter can be assumed to have occurred if these coli-
phages cannot be found. Coliform (especially thermo-
tolerant coliform) bacteria are not considered good indi-
,cators of enteroviruses because they are much less resistant
'than enteroviruses to environmental conditions and chlori-
nation (Scarpino, et al., 1972). Detection of waterborne
coliphages is easier than enteroviruses because:
1. They are present in greater numbers than entero-
viruses in fecal material and therefore, in sewage;
and
2. They can be isolated and counted by relatively
simple means, and in less time than is required
for enteroviruses. Coliphage enumerations can
often be accomplished within 4 to 6 h, whereas
seven or eight days are required for enterovirus
enumerations (Kott, et al., 1978b).
(i) Coliphage Enumeration Techniques. The agar layer
method (Adams, 1959) provides an accurate means of enu-
merating coliphage, and does so within 24 h. The sample is
inoculated directly onto a bacterial lawn grown on soft agar
(or mixed together with the sample and molten agar) in petri
plates, and plaques are counted. However, a preliminary
concentration step must be applied before this technique can
be used with large volumes of water containing small numbers
of bacteriophage.
The most probable number (MPN) method has been used
successfully for animal virus tissue culture enumerations
and could detect coliphage in fresh water, seawater, and
wastewater in numbers as low as two plaque-forming units
(pfu) per 100 ml.when E. coli B was the host strain (Kott,
1966). This technique has, accordingly, been recommended
for use with larger volumes of up to one liter of water and
provides for enrichment of phages. It calls for inoculation
of sample into a 5-tube, 3-dilution series, using double and
single strength Phage Assay Base (PAB) broth (Fisher Scien-
tific Co., Pittsburgh, Pa.) as the enrichment medium. Cul-
ture tubes are incubated for 16 h at 35°C after which time a
loopful from each tube is transferred to freshly seeded host
E. coli plates and incubated for 6 h at 35°C. Results are
computed from coliform MPN tables (Kott, 1966). Although
the MPN method can be used with any E_. coli strain, the best
results in the cited study were obtained when 12. coli B
served as host.
-------
200
(ii) Preparatory Concentration Techniques. Several
techniques are available for concentrating viruses, includ-
ing phages, which occur in water at extremely low levels
[See also Section III.C.2.b]. *Gilcreas and Kelly (1955)
used alum flocculation and obtained almost 99.99 percent
isolation of coliforms and coliphages; they accordingly
recommended it for concentrating coliphages. Preformed
floes of aluminum hydroxide, aluminum phosphate, and calcium
phosphate are generally used for the concentration and
isolation of a great variety of viruses, and have been
routinely employed as a step in drinking water treatment.
Sorber and coworkers (1972a) found that the polelec-
trolyte (PE) 60 (a cross-linked copolymer of isobutylene
maleic anhydride) adsorbed 100 percent of T2 coliphage from
large volumes (up to 18 1) of water. These same authors
(l972b) employed asymmetric cellulose-acetate membranes
(commonly used in reverse osmosis) to collect coliphage T2
and poliovirus from large volumes of water.
Bachrach and Friedmann (1971) used a polyethylene
glycol and dextran sulfate 2-phase system as well as zone
centrifugation in sucrose gradients to isolate T-even and T-
odd coliphages. When acid precipitation was employed, good
recoveries were obtained with T2 and T4 phages, but not
with T3 and T7 phages due to their susceptibility to the
acid.
The method found most suitable for recovering low
coliphage numbers from large volumes of water entails adsorp-
tion of the phage to a cellulose nitrate membrane filter
after increasing the electrolyte concentration (for better
adsorptive capacity) by adding 0.1 M MgCl2 - 6Ho° to tlie
sample. The sample is filtered through a 50 mm membrane of
0.45 um pore size (or larger membranes if over 20 1 are
applied); phages are eluted with/10 to 15 ml of 3 percent
beef extract (pH 9.0). The eluate is examined either by the
direct plaque count or by the MPN method (Vajdic, 1970).
E. coli B host cultures were used with this method.
(iii) Occurrence and Persistence of Coliphage and
Enterovirus in Different Water Sources. Types and numbers
of coliphages present will vary depending on the source of
the water to be tested whether taken, for example, from
stream water, wastewater effluent, or a sewage lagoon.
Therefore, the most appropriate bacterial host strain will
depend on which coliphage types are likely to predominate,
-------
201
depending upon their relative resistance to environmental _
(and disinfection) stresses, among other factors. Although
most investigators use host bacterial strains susceptible
to the T-phages, only 11 percent, of hundreds of E. coli
(differentiated by IMViC) strains isolated demonstrated
susceptibility to the T-series. Moreover,. the RNA bacte-
riophages f2 and MS- are more resistant than the other
coliphages to chlorlnation. ,
Vaughn and Metcalf (1975) studied the presence of
coliphages in an estuary polluted with sewage effluent. By
testing samples with three host E. coli strains over a
three-year-, period, they found that the dominant coliphage
type shifted from one year to the next; this suggested that
a host system selected on the basis of past experience might
prove ill-suited to present needs. Their studies also
revealed that coliphages were capable of multiplying in
polluted estuarine waters containing susceptible host cells.
They suggested that, due to the potential for coliphage
replication in certain waters, coliphages might not always
be used reliably to indicate the presence of enteric viruses
in water. Cleaner quality waters, however, may show no such
multiplication because coliphages infect and replicate only
in growing cells and only under very restrictive conditions
of .pH, temperature, and more.
Ratio of coliphages to enteroviruses also vary among
water sources. A three-year study of wastewater effluents
found that while the comparatively small number of entero-
viruses underwent seasonal fluctuations4 the comparatively
large number of coliphages (/vio to 10 times as many)
remained steady (Kott, et al., 1978a). However, levels of
both poliovirus 1 and coliphage f2, isolated from an experi-
mental oxidation pond, remained stable over a period of
several months; following oxidation pond treatment the
coliphages exhibited higher resistance to chlorination.
Sewage lagoons appeared to reduce enteroviruses and coli-
phages to approximately the same degree (about two orders of
magnitude). But, when poliovirus 1 (strain LSc) and coli-
phage f2 were kept in tapwater and oxidation .pond samples in
the dark at room temperature, the poliovirus was totally
inactivated in < 150 days in both sample waters, whereas
the f2 remained viable for over 300 days in both samples
[See also Section E.l.d]. Creek and stream waters often-
contain coliforms, coliphages, and enteroviruses: there may
be from 10 to 10 times as many coliphages as enteroviruses
present. Poliovirus survived < 217 days in dry sand, whereas
coliphage persisted for several hundred days. In effect,
-------
202
the physical and chemical quality of the water will influ-
ence, to a much greater extent, the numbers of enteroviruses
recovered than it will the numbers of coliphages.
(iv) Summary. Coliphages (phages that infect E.
coli) are indicators of the possible presence of entero-
viruses in water because they are shed in the feces of
humans and warm-blooded animals and may be more persistent
than enteroviruses under adverse conditions (including
disinfection). Coliphage numbers may range from a few per
liter in streams and rivers to as many as 10 per ml in
wastewater (where there has not been substantial dilution).
Coliphages can be quantified directly by the plaque tech-
nique if numbers are high; low numbers necessitate use of
the MPN method; either technique may be preceded by some
method for concentrating the phages. The simplest method
is that of adsorption to a cellulose nitrate membrane filter
followed by elution in a small volume of fluid. Using
coliphages to indicate the absence of enterovirus from
drinking water is inexpensive, does not require a high level
of operator skill, and yields results within 24 h. There
are, however, circumstances under which coliphage detection
may not accurately indicate the presence of enteric viruses.
b. Vaccine Polioviruses. A criticism frequently
leveled at the microbiologic indicators of water quality
presently in use today, is that their occurrence is poorly
correlated with the incidence of human enteric viruses in
water (Berg and Metcalf, 1978). The assertion may be true,
but the criticism is not entirely valid, for indicators now
in general use are expected to signal a variety of undesir-
able conditions without necessarily being absolutely related
to the presence of any one specific pathogen. If one wants
an indicator specifically to predict the incidence of human
enteric viral pathogens, the indicator in question should
always be present when these pathogens are present (Berg,
1978). The nonpathogenic, vaccine strains of the polio-
viruses are human enteric viruses and might thus serve as
specific indicators of the virologic safety of water
(Katzenelson, 1976).
More than 100 different virus types may be encountered
in waters contaminated with human wastes, although only the
hepatitis A virus has been demonstrated to be transmitted by
water (Berg and Metcalf, 1978; Katzenelson, 1978; Melnick,
et al^., 1978). Most commonly encountered in water are the
-------
203
enteroviruses (polioviruses, echoviruses, and coxsackie-
viruses); but the reoviruses, adenoviruses, hepatitis A _
virus, and gastroenteritis viruses may also be present LSee
Sections B.l.b and B.l.d] (Center for Disease Control, _
1978d; Katzenelson, 1978; and Melnick, et al., 1978). The
concentration of such viruses in feces-contaminated water
depends upon the number of infected individuals in the
population, independent of the incidence of clinical symp-
toms .
The concentration of any virus, once introduced into
water, can only remain at the initial level or decline
(Berg, 1978; Katzenelson, 1978). Viruses do not replicate
outside a susceptible living host, whereas indicator bacte-
ria can multiply in nutrient enriched waters and thus, exag-
gerate the perceived risk of viral contamination (Berg, 1978;
Geldreich, 1978). Enteric viruses are also more resistant
to environmental extremes and to disinfection, so t^at _
enteric viruses may persist in water long after all indi-
cator bacteria have been eliminated (Berg, 1978; Berg and
Metcalf, 1978; Katzenelson, 1976; U.S. Environmental Protec-
tion Agency, 1978). These are general reasons that the
occurrence of bacterial indicators cannot always be used to
signal the presence of viruses in water, as has been demon-
strated by outbreaks of hepatitis A resulting from con-
sumption of shellfish grown in contaminated surface waters
that were believed to be free of total coliforms (Berg and
Metcalf, 1978; Portnoy, et al. , 1975). The presence of
viruses in water may best be monitored using a viral indi-
cator group.
In comparison to 'the traditional coliform indicators of
water quality, the concentration of viruses in wastewater is
quite low. Estimates of 100 to 400 virus units per liter in
the U.S. (U.S. Environmental Protection Agency, 1978) and
100 to 10,000 per 100 ml in other countries (Fattal and
Nishmi, 1977; Katzenelson, 1978) have been reported. After
wastewater treatment and dilution in the environment fol-
lowed by exposure to environmental extremes, the con-
centration of the remaining viruses in the water supply
will be still lower. Thus, it is difficult to estimate the
extremely low numbers of viruses that might occur in drinking
water. " . ; -/-.. > : - ., ; -^ .-"'.-.*.
Given the many types of viruses that may enter water,
the detection of any single viral pathogen is a formidable
task, and the detection of a great range of virus types
would be too complicated for routine laboratory analyses.
-------
204
What is needed is a simple, reliable virus indicator system
that consistently occurs in feces-contaminated waters and
would dependably serve notice of the presence of viral
pathogens.
^m^Tr?n c°untries Usin9 the live trivalent oral polio vaccine
(TOPV), the predominant enteric viruses isolated from
sewage have been the three serotypes of the vaccine polio-
virus (Fattal and Nishmi, 1977; Katzenelson, 1976; Katzenelson
fn?0? X/ '" U'S* Environmental Protection Agency,
19/8;. These virus isolates are shown to be vaccine-
associated on the basis of the presence of distinctive
genetic markers. Vaccine-derived poliovirus is assumed to
be nonpathogenic, although of 142 cases of poliomyelitis
reported in the U.S. from 1970 to 1978, 44 have been vaccine-
associated; thus, the vaccine virus may be pathogenic for a
small susceptible portion of the population (U.S. Environ-
mental Protection Agency, 1978). The vaccine virus should
not exhibit seasonal fluctuations in concentration, whereas
the occurrence of wild viruses peaks in the late summer and
early fall (Fattal and Nishmi, 1977; Katzenelson, 1976- U.S
Environmental Protection Agency, 1978).
Vaccine polioviruses are hardy, safe to handle, and
easy to enumerate in the laboratory, whereas simple cell
culture recovery systems for many other viruses are non-
existent. Because they are convenient laboratory tools,
much is known about their survival in the environment, their
responses to disinfectants, and,methods for their recovery
from_environmental sources. it is for these reasons that the
vaccine-derived polioviruses are attractive candidates to
serve as viral indicators. Other investigators (Melnick,
et alL., 1978) also have realized the value of using the
readily demonstrable enteroviruses as indicators.
The major disadvantage of using a virus system for
determining water safety is the time involved in performing
virological assays, which usually takes more than 24 h and
often requires three to seven days after a 24 to 48 h concen-
tration step (Katzenelson, 1976). if polioviruses are to
serve as useful indicators, a more rapid detection method is
T^?1 ' .A Pr°P°sed method uses the fluorescent antibody
(FA) technique which identifies the virus after 18 to 24 h
(Katzenelson, 1976). With the FA method, poliovirus anti-
gens can be demonstrated after 6 to 9 h, but quantitative
results require the complete 18 to 24 h incubation. An
improvement of the original FA method incorporates tragacanth
gum into the micro tissue culture medium, analogous to the
-------
205
incorporation of agar into the plaque assay medium (Kedmi
and Katzenelson, 1978). The resulting increase in viscosity
allows the resolution of closely spaced plaques (preventing
overlapping of neighboring plaques) and eliminates scat-
tering of infected cells. Thus, a more accurate enumeration
of plaque-forming units is achieved.
In countries routinely administering the TOPV to infants,
in three to four challenges during the first year of life, a
constant seeding of wastewater occurs (Berg and Metcalf,
1978; Katzenelson, 1976; Katzenelson and Kedmi, 1979; U.S.
Environmental Protection Agency, 1978). The vaccine virus,
like the wild virus, replicates in the human6intestinal
tract and is shed in high numbers (10 to 10 plaque-forming
units per g of feces) for several weeks after the initial
administration. Vaccine-derived poliovirus should thus, be
expected to occur consistently in the sewage of countries
using the live attenuated virus.
In one study (U.S. Environmental Protection Agency,
1978), vaccine-derived poliovirus could be isolated from
sewage when as few as 0.3 percent of the local population
were vaccinated. In Israel, viruses were isolated from
sewage at different locations; among the 489 isolates, 74 ^
percent of those identified were polioviruses. However, in
Hungary only 19 percent of the identified viral isolates
were vaccine-strain polioviruses (Palfi, 1971). In this
study, poliovirus was the prevalent type found in_sewage
only after the compulsory polio vaccination campaigns were
carried out during the winter months. Reoviruses were
isolated throughout the year, but were isolated in highest
concentration only after excretion of the vaccine virus
ceased. The same pattern was observed with other wild
viruses, suggesting that vaccination of populations with
live polioviruses alters the incidence of wild viruses
within a population. This observation may be due to the
very rapid destruction of cells, by polioviruses, during
enumeration assays (Palfi, 1971) or to the displacement_of
other virus types in the human population. Administration
of live vaccine virus also alters the ratio of indicator
bacteria to viruses entering sewage (Berg and Metcalff
1978). Intermittent vaccination programs would cause polio-
virus predominance only seasonally, whereas continuous
vaccination programs, as in the U.S. and Israel, should
cause poliovirus to predominate throughout the year (Berg
and Metcalf, 1978).
In Israel, where TOPV is administered year-round, raw
sewage samples from both large cities and small communities
were analyzed for poliovirus (Katzenelson and Kedmi, 1979).
-------
206
In this study, pbliovirus was isolated from only 50 percent
of the samples assayed. Negative results may have occurred
because the wastewater contained a low concentration of
polioviruses that was beneath the limit of sensitivity of
the investigators' detection system. Detectable viruses
were absent not only in small communities where the infant
population could be expected to be low, but also in the
large communities that had a significant number of infants
who had received the TOPV.
In order for vaccine polioviruses to serve as indi-
cators for enteric viruses, they must be present in all
waters_receiving fecal discharges. Although vaccine-derived
poliovirus is isolated from sewage more often than other
viruses, it is not always present (as is true of E. coli) in
detectable quantities; therefore, negative test results
would not guarantee virus-free water (Katzenelson, 1976).
Vaccine polioviruses could not be expected to serve as
suitable indicators in small communities lacking a sig-
nificant infant population, in countries administering TOPV
only seasonally, or in countries and communities using a
killed virus vaccine instead of TOPV. However, the vaccine-
derived polioviruses would have been expected to predominate
in raw sewage from large cities in Israel, where the vaccine
is in continuous use. The inability to detect vaccine
polioviruses in many of such samples suggests that this
potential indicator system is not predictably better cor-
related, compared to others already in use, with the inci-
dence of human enteric viruses in water (Katzenelson and
Kedmi, 1979).
Summary. Vaccine polioviruses are prevalent in receiv-
ing waters in areas where live trivalent oral polio vaccine
is used for immunization. These viruses have been con-
sidered for use as indicators of human pathogenic enteric
viruses. They are frequently detected in sewage, but recent
data suggest that there are significant numbers of cases
where no polioviruses were detectable in raw sewage in urban
areas where the oral polio vaccine was in constant usage.
As a result, the vaccine polioviruses do not appear to offer
increased reliability over conventional bacterial indicators
for enteric viruses in water.
c' Pseudomonas Aeruginosa. Within the genus Pseudo-
monas , the species of concern to public health is P. aeru-
ginosa CSee Section B.l.a(ix)]. The most common diseases"
-------
207
caused are eye, ear, nose, and throat infections, although
general infections and septicemia are also possible (Caselitz,
1966). Given the opportunity to multiply in foods, ]?. .
aeruginosa may cause diarrhea after ingestion of the contami-
nated food (Caselitz, 1966; Kubota and Liu, 1971). Since
this organism can grow under low nutrient concentrations
(i.e., in distilled water), it warrants consideration for
use in routine monitoring of finished water.
p. aeruginosa has been defined by a working group of
the International Standards Organization as:
"Gram-negative, non-sporing rods which are oxidase
and catalase positive; capable of growth at 42°C,
but not 4°C; usually produces a water-soluble,
fluorescing pigment; exhibits oxidative metabolism
as indicated by the Hugh and Liefson test; reduces
nitrate beyond the stage of nitrite, and produces^
ammonia from the breakdown of acetamide; gelatin is
liquified, casein is hydrolyzed, but starch is not
hydrolyzed. Pyrocyanine is produced by more than 90
percent of strains." .
The ecology of P. aeruginosa is not well known. Although
the organism is found, regularly, in surface waters that
receive wastewater effluents (mainly of urban origin) LSee
Sections A.Intro, and A.2.a],' it is not clear whether this
is necessarily the source of all £. aeruginosa. Recovery of
Pseudomonas biotypes in areas untouched by humans indicates
that environmental factors alone may support and even favor
mass development of p. aeruginosa. Results from groundwater
surveys show that P. aeruginosa is rarely found in undis-
turbed groundwater and that the presence of P. aeruginosa
nearly always indicates contamination by surface water
(Hoadley, 1977; Schubert and Blun, 1974a; Schubert and
Scheiber, 1975). In addition, the organisms are not fre-
quently found in groundwaters infiltrated with surface water
or river water, which suggests that they are effectively
eliminated during passage through the soil (Schubert and
Blun, 1974a; Schubert and Scheiber, 1975) [See also Section
A.I.6].
(i) Enumeration Methods. The most probable number
(MPN) method is routinely used to enumerate P. aeruginosa,
with enrichment either in nutrient broth containing malachite
qreen dye incubated at 37°C (Schubert and Blun, 1974b) or in
asparagine broth incubated at 35 to 37°C (American Public
Health Association, 1976). Enrichment in Drake's medium 10
-------
208
also has been shown to yield good results (reviewed by
Hoadley, 1977). From there, cells are streaked onto a
selective medium (MacConkey's agar, cetrimide agar, etc.),
and finally identified by tests for fluorescein formation
(on^King's medium A) and/or pyocyanin formation (on King's
medium B) as well as ammonia production from acetamide
(which is of particular importance with apyocyanogenic
strains (Schubert, et al., 1975).
Several membrane filtration methods, based on the mPA
medium of Levin and Cabelli (1972), are in current use.
The mPA medium contains antibiotics to reduce background
growth (sulfapyridine, kanamycin, nalidixic acid, and acidione)
along with indicator systems to distinguish I?, aeruginosa
from colonies that ferment lactose, sucrose,"and xylose, and
produce hydrogen sulfide. Membrane filters are placed on
the agar medium and plates are incubated at 41.5°C for 48 h.
A recent modification of this procedure, known as mPA-C
»(Brodsky and Ciebin, 1978), allows presumptive identifi-
cation of P. aeruginosa within 24 h at an incubation tem-
perature oT 41.5°C.
(ii) Evaluation as an Indicator. Pseudomonas is an
organism worthy of attention in the area of finished drink-
ing water, largely because it has been found capable of
growth along portions of the distribution system, particu-
larly under conditions of low water flow and where dead end
pipes are situated [See Sections F.2.a and F.S.a]. Treat-
ment processes such as ion-exchange systems [See Section
G.2] and many filtering devices [See Section E.4], in ad-
dition to situations where warm water flows through some
types of plastic tubing or containers, can be especially
favorable to the growth of Pseudomonas.. It has been sug-
gested (Labonde and Festy, 1979) that, in certain cases
(e.g., in bottled water, in epidemiological studies, in new
homes, in high risk areas, and in water destined tor thera-
peutic purposes), drinking water should be monitored for
Pseudomonas.
(iv) Conclusions. ]?. aeruginosa is a potential patho-
gen, especially if ingested in high numbers, as can occur
after_enrichment in foods. Therefore, it must be considered
undesirable in drinking water and should be absent from
sample volumes of at least 100 ml of finished water. The
presence of jp. aeruginosa in drinking water almost always
can be traced either to direct contamination by surface
-------
209
water or to dissemination from a primary locus of contami-
nation to other parts of the distribution system, reser-
voirs, or treatment facilities.
(v) Summary. Pseudomonas aeruginosa is an important
opportunistic pathogen of humans. It may or may not be of
fecal origin in water, but it can grow even at low nutrient
levels, sometimes reaching high numbers under favorable
conditions, some of which have been described here. Methods
for the detection of Pseudomonas are sufficiently developed
that the organism should be routinely monitored in high risk
areas, in the distribution system, and in water destined for
therapeutic use.
d. Klebsiella. The genus Klebsiella is a member of
the familv~Enterobacteriaceae and is composed of non-
sporeforming, non-motile, capsulated gram-negative rods
(0rskov, 1974).
Some species of the present genera Klebsiella and
Enterobacter, with IMViC type ++, were traditionally
classified as Aerobacter aerogenes. Later, the motile
strains were designated Enterobacter (including A. aero-
genes), while the non-motile organisms were placed in the
genus Klebsiella. Another major difference between Entero-
bacter and Klebsiella is that the latter lack ornithine
decarboxylase.
According to the Edwards and Ewing classification,
which is widely accepted in the U.S. (especially in the
clinical laboratory), the klebsiellae are divided into three
species: K. pneumoniae, K. ozaenae, and K. rhinosclero-
matis. This classification has been adopted by Bergey's
Manual of" Determinative Bacter_ioiogy^fi (Buchanan and Gibbons, 1974);
however~alternate classification systems exist. The type species
K. pneumoniae, is the predominant species isolated from environmental
Sources and is identified in 95 percent of clinical KjLebsiella
isolates in the U.S. .
Biochemical procedures for differentiating K. pneu-
moniae are discussed elsewhere (Gellman, 1975; Dufour and
Cabelli, 1976; Vlassoff, 1977). This organism is char- _
acterized as a non-motile rod that ferments lactose with the
production of gas and does not decarboxylate ornithine. Its
typical IMViC reaction is --++. Other tests useful for its
identification are urease (slow +),
H2S
(-), oxidase (-),
-------
210
gelatin liquefaction (-), KCN (+), lysine decarboxylase ( + ),
arginine dihydrolase (-), and esculin hydrolysis (+).
Klebsiella isolates may be classified, serologically,
on the basis of their heat-labile capsular (K) and heat-
stable somatic (O) antigens. Presently, there are 72 known
capsular serotypes and five 0 antigens. Rennie and Duncan
(1974) found that biotypes and serotypes vary independently so
that the two methods, used in conjuction, distinguish among
more types than either method used alone.
Klebsiellae are widespread in normal humans and animals,
but are also known to be opportunistic pathogens capable of
causing infections when normal host defenses are weakened
(Bagley and Seidler, 1978; Gellman, 1975). Furthermore,
since the advent of antibiotic therapy, these organisms have
shown increasing resistance to antibiotics because of their
ability to accept and transfer plasmids (R factors) which
carry determinants for multiple drug resistance.
(i) Sources of Klebsiella in the Environment. Kleb-
siella may originate from vegetable as well as animal sources
(Duncan and Razzell, 1972). Klebsiella of animal origin, in
the environment, most frequently derives from feces, but not
all fecal_samples yield Klebsiella. On the other hand,
colonization of the human intestines, so that Klebsiella
predominates, is a known, but abnormal phenomenon (Dufour
and Cabelli, 1976; Knittel, et al., 1977; Vlassoff, 1977).
Regardless of its ultimate source, Klebsiella is ubiqui-
tous in the environment (Dufour and Cabelli, 1976; Knittel,
et: al., 1977; Seidler, et al., 1977; and Vlassoff, 1977).
These organisms are commonly associated with living trees
and have been found in samples of water, soil, needles, and
bark from different forest environments, including virgin
forests of British Columbia. High densities of coliform
bacteria, predominantly Klebsiella, can be found in pulp and
paper mill effluents. Other carbohydrate-rich nutrient
sources such as waste effluents from sugarcane, sugar re-
fining, and kelp processing also yield high numbers of
Klebsiella, far outnumbering the !E. coli population. Signifi-
cant numbers of Klebsiella have been isolated from a variety
of market vegetables and seeds, dairy products (with posi-
tive coliform counts), and pet turtles.
The ability of Klebsiella to grow in nutrient-rich
waters may be the cause of its frequent predominance over
-------
211
E. coli in surface waters receiving wastewater discharges
TDufour and Cabelli, 1976; Seidler, et al^. , 1977). The
capsule on the Klebsiella cell seems to impart some relative
chlorine resistance, so that Klebsiella may predominate
over IS. coli in chlorinated, water derived from raw water in
which~E. coli was the more numerous (Ptak, ejt aJL. , 1973).
Total coliform counts exceeding U.S. federal regulations
were detected by Seidler and coworkers (1977) in public and
private drinking water systems that utilized redwood storage
tanks. The coliform isolates were most often identified as ,
K. pneumoniae and Enterobacter species. El. coli was isolated
only rarely and was most likely due to more immediate fecal
contamination. It appeared to the authors that specific
nutrients in wood, trees, vegetables, and other plant matter
selectively support coliforms of the tribe Klebsiellae.
(ii) Significance of Klebsiella in Water Bacteriology.
Given the definition of the genus, virtually all Klebsiella
organisms will yield a positive result in the total coliform
test [See Section C.l.b]. Not all klebsiellae give a
positive result in the thermo-tolerant coliform test [See
Section C.I.c], apparently because some cannot grow at
44.5°C, however, most thermo-tolerant coliforms other than
E. coli are Klebsiella (Dufour, 1977). Those klebsiellae
which can grow at 44.5°C and are thermo-tolerant coliform-
positive are not necessarily of fecal origin. Nevertheless,
those which have been selected or adapted to growth in low
temperature environments tend to be negative in the fecal
coliform test (Naemura and Seidler, 1978), so there is a
measure of correlation between the test result and fecal
origin.
Members of the genus Klebsiella are" not reliable indi-
cators of sanitary qualilty or fecal contamination of water,
for they do not meet several criteria of ideal indicator
organisms. Detection of Klebsiella in water is more probably
indicative of certain types of industrial pollution. Many
studies have demonstrated high densities of Klebsiella, and
few E. coli', in wood processing and other nutrient-rich
effluents that were not known to have received human or
animal wastes. Therefore, high coliform or fecal coliform ,
counts in waters receiving such industrial effluents (unless
sanitary wastes from the plant are included) should not. be
considered indicative of the presence of othermicrobial
pathogens. Alternately, it has been suggested that ,
-------
212
nutrients arising from vegetable material may selectively
favor multiplication of the klebsiellae and that this could
mask the presence of other fecal organisms (Dufour and
Cabelli, 1976; Seidler, et ai., 1977). Klebsiella may also,
inhibit the growth of other bacteria (Ptak, et al., 1973).
Klebsiella, used in conjunction with thermo-tolerant
coliforms or IS. coli, can be a useful indicator for moni-
toring certain industrial discharges and their receiving
waters. Primary isolation methods, by which members of the
genus Klebsiella can be enumerated and differentiated for
this purpose, are obviously needed. Proposed media and
methods have been reviewed by Vlassoff (1977). Sufficient
data, on which to base a selection, are not yet available.
(iii) Conclusion. The widespread presence of Kleb-
siella in the environment is well known. Although carried
by many healthy individuals, members of the genus Klebsiella
are opportunistic human and animal pathogens and have become
increasingly important as the cause of nosocomial infec-
tions. Some Klebsiella strains, readily recovered from
environmental sources, have been found to be indistin-
guishable from clinical isolates when tested biochemically,
serologically, and for their ability to grow at elevated
temperatures (44.5°C) (Bagley and Seidler, 1978).
As is the case with any opportunistic pathogen, both a
compromised host and a sufficient number of organisms are
necessary for infection to occur. There is, as yet, no
epidemiological evidence to connect the incidence of Kleb-
siella in drinking water or recreational waters with occur-
ence of human disease.
The klebsiellae cannot be considered reliable indi-
cators of fecal pollution because: (1) they are not always
present or found in high numbers in feces; (2) they are
found in large numbers in certain industrial wastes; (3)
they are ubiquitous in the environment; and (4) they are
able to multiply in nutrient-rich effluents.
The klebsiellae are not thermo-tolerant coliforms;
thus, since a large proportion are able to grow at 44.5°C,
there has been some controversy about the validity of total
and thermo-tolerant coliform tests which include these
organisms. Insofar as the organisms detected are recognized
as members of the genus Klebsiella, their presence is most
likely to signify high nutrient levels in receiving waters
or to give evidence of inadequate treatment of industrial
wastes.
-------
213
(iv) Summary. The genus Klebsiella is a member of the
coliform group, Klebsiella pneumoniae being the species most
predominant in both environmental and clinical isolates. In
spite of its significance as a secondary pathogen, it is not
entirely suitable as a fecal pollution indicator because:
it is not always found in large numbers in feces; it is
often found in large numbers in certain industrial wastes;
and it is able to multiply in environmental waters rich in
carbohydrates.
e. Bifidobacterium. Bifidobacterium, previously
known as Lactobacillus bifidus, may have the necessary
characteristics of an ideal bacterial indicator of fecal
pollution in raw water (Evison and James, 1975). Bifido-
bacteria derive exclusively from feces and so, would occur
only where there has been fecal contamination. Moreover the
organisms are present only in very high numbers of 10 to
10 per gram of feces in humans (Levin, 1977; Cabelli,
1979) and occur, though \j.n markedly lower numbers, in some
warm-blooded animals, especially pigs. Both coliforms and
13. coli are now recognized as being capable of regrowth in
warm, organically polluted waters; conversely, under these
same conditions, numbers of fecal streptococci tend to
decline rapidly. Bifidobacteria, on the other hand, are
anaerobic organisms with fairly specific growth requirements
and thus, are highly unlikely to find all the conditions
necessary for regrowth in raw waters.
Members of the genus Bifidobacterium have posed clas-
sification problems in the past and the nomenclature has
been very confused. However, in the most recent edition of
Sergey's Manual of Determinative Bacteriology (Buchanan and
Gibbbns,1974),the genus has been divided into eleven
clearly defined species on the basis of fermentation reac-
tions with eleven carbohydrates. 13. bifidum, J3. adoles-
centis, 13. infantjs, and IJ. breve are regularly associated
with human feces; 13. longum, and perhaps 13. pseudolongum,
may also be isolated from human feces, but are more often
found in feces of lower animals, as are 13. thermophilum and
B. suis; 13. asteroides, B. indicum, and 13. coryneforme are
associated only with insects.
Bifidobacteria are defined as anaerobic, nonspore-
forming, non-motile/gram-positive, thickjpleomorphic rods.
They may exhibit branching bulbs, clubs, coryneforms, buds,
spheroids, and bifurcated Y and V forms when freshly iso-
lated from fecal sources. However, their morphology is
-------
214
related to nutritional conditions and may be affected by
exposure in an aquatic environment and by composition of the
culture medium. A calcium deficiency has been said to
encourage pleomorphism, but culturing on normal laboratory
media/ such as MRS broth (De Man, et al., 1960) or differen-
tial reinforced clostridial media TDRCM), greatly reduces
pleomorphism. Although young cultures of bifidobacteria are
gram-positive, cultures older than four days gradually
become gram-variable or even apparently gram-negative
(Buchanan and Gibbons, 1974). The organisms also are cata-
lase negative, do not reduce nitrates, and produce mainly
lactic and acetic acid, but not gas, from glucose fermen-
tation.
(i) Methods for Enumerating Bifidobacterium. Several
media have been proposed for isolating Bifidobacterium from
water, but only two are in use at present. Both of these
media are used with membrane filters, which allows concen-
tration of the sample, as may sometimes be necessary in the
case of raw water. The more fully defined isolation medium
is based on that proposed by Gyllenberg and Niemela
(1959), but has been modified considerably and improved
(Bhoonchaisri, 1979)(MGN-Modifled Gyllenberg and Niemela medium)
The isolation medium [See Table C.2.e-l for formula-
tion] is prepared in two parts. After Part A is melted and
cooled to 45°C, Part B is added in proportions of 10 ml Part
B to 90 ml Part A. Raw water samples are filtered through
cellulose nitrate membrane filters (0.45 um pore diameter,
47 mm filter diameter), the membranes are placed on the
solidified medium in incubation tins, and incubated at 37°C
for three days. Bifidobacterium colonies appear pink to red
and are raised, with a diameter of 0.5'to 1.5 mm. They also
will grow equally well in conventional anaerobic jars or in
jars containing gas generator packets,
A second medium, YN-6, has been proposed by Resnick and
Levin (1977) CSee Table C.2.e-2 for medium formulation and
preparation]. Membrane filters are incubated, anaerobically,
for two days at 37°C. Bifidobacterium colonies will measure
1 to 2 mm in diameter and appear light to dark green,
convex, glistening, and smooth.
Colonies from either MGN or YN-6 medium can be iden-
tified to the generic level by a simple series of tests.
-------
215
TABLE C.2.e-l
MODIFIED GYLLENBERG AND NIEMELA MEDIUM FORMULATION
Part A
Lactose
Cysteine hydrochloride
MgS04 ' 7H20
Tween 80
FeS04 * 7H20
NaCl
MnSO4 ' 4H2O
Nalidixic acid
Biotin
Riboflavin
Pantothenic acid
Neutral red
Oxoid agar No. 3
Distilled water
10.0 g
5.0 g
4.0 g
0.4 g
0.2 g
1 ml Adjust pH to 6.5.
Portion out in 90 ml
,0.01 g aliquots. Autoclave at
121°C for 15 min.
0.01 g
0.007 g
0.007 g
5 ug
250 ug
500 ug
0.1 g
12 g
900 ml
Part B
Ascorbic acid
Distilled water
10 g Adjust pH to 6.5. Sterilize
by pressure filtration
100 ml through membrane filter of
0.45 u pore -diameter
(GYLLENBERG AND NIEMELA, 1959: BHOONCHAISRI, 1979)
-------
216
TABLE C.2.6-2
YN-6 MEDIUM FORMULATION AND PREPARATION
Yeast extract
Peptone
Lactose
Casamino acids
Sodium chloride
Brom cresol green
Distilled water
20 g
10 g
10 g
8 g
3.2 g
0.3 g
1 liter
Boil all ingredients for 10 min and cool. Add cysteine
hydrochloride (0.4 g), Nalidixic acid (80 mg), adjust to
pH 6.9, add agar (15 g) and autoclave at 121°C for 15 min,
Cool to 60°C, add 1 ml stock Neomycin (2.5 mg/ml) and
dispense into incubation tins.
(LEVIN, 1977)
-------
217
Colonies are grown on solidified DRCM medium and Gram stained
(Gram +), spore stained (negative), tested for motility
(negative), aerotolerance (negative), and catalase activity
(negative). They also are grown in MRS-nitrate broth to
check for nitrate reduction (negative). Species of bifido-
bacteria are identified by their carbohydrates fermen-
tations . Purified isolates are inoculated into MRS-sugar
broths that have been modified by replacing 2 percent glu-
cose with other carbohydrates. Positive fermentation is
indicated by acid production (chlorophenol red indicator),
but no gas production. The isolates are divided initially
into four groups, on the basis of arabinose and gluconate
fermentation, as in Table C.2.e-3. identification
within groups is achieved by additional tests for lactose,
ribose, xylose, mannitol, and starch fermentation (Levin,
1977; Bhoonchaisri, 1979).
(ii) Comparative Evaluations of Proposed Bifido-
bacterium Culture Media. Results obtained when raw water
was examined for Bifidobacterium differ according to the
isolation medium used. When the MGN and YN-6 media were "
compared, using the same sample, the MGN medium yielded /
higher counts of confirmed Bi fidobacterium than the YN-6
medium. Although the YN-6 medium showed higher initial
colony counts, a larger proportion of these were aerotolerant
(non-Bifidobacterium) organisms. Bhoonchaisri (1979) also
showed that the incorporation of a reducing aqent (0.05
percent cysteine hydrochloride) into the sample diluent in-
hibited the growth of Bifidobacterium.
(iii) Evaluation of Bifidobacter ium as an Indicator of
Fecal Contamination. Studies of Bifidobacterium in fecal
specimens from humans have shown that the population is
higher and more constant than IS. coli or fecal streptococci
across all age and ethnic group boundaries (Drasar, 1974;
Evison and Morgan, 1978). To quote Cabelli (1979), "Of all
the indicator systems, Bifidobacterium appears to be the one
most exclusively associated with human, as opposed to lower
animal, fecal wastes." It seems that this group may offer
potential as a means of distinguishing human from animal
pollution and may be very useful, for example, in deter-
mining whether domestic waste has entered a stormwater
system.
Studies with pure cultures of Ii. bifidum, E. adoles-
centis, and B. pseudolongum have demonstrated clearly that
-------
TABLE C.2.e-3
Groupings of Bifidobacterium species
Group Arabinose Gluconate
Species
I
II
.I'll
IV
E. adolescentis, J3. asteroides, and 13. coryneforme
E. longum, J3. pseudolongum, and B. suis
B. bifidum, B. breve, B. infantis, and B. thermophilum
B. indicum
to
M
00
-------
219
bifidobacteria are unlikely to multiply in raw waters (Evison
and Morgan, 1978) because: very little growth occurs above
or below the extremely narrow temperature range of 30 to
40°C; no growth can occur above 7 percent atmospheric oxygen;
and maximum growth rates can occur only in the presence of
high concentrations of proteins or carbohydrates.
One problem that has become apparent in recent studies
is that the Bi fidobacterium isolation media are not as
selective when used with surface waters as they are for
sewage samples. Gram-positive rods are commonly found in
surface waters and are able to grow on the selective media,
so further development of the media is required to overcome
this interference.
A further problem, yet to be resolved, is that of the
survival potential of bifidobacteria in the aquatic environ-
ment. Evison and James (1975), and Opara (1978), found that
survival of bifidobacteria was equal to, and sometimes
greater than, survival of j|. coli in laboratory tests.
However, Resnick and Levin (1977), and Cabelli (1979), have
suggested that survival of bifidobacteria is poor in the
aquatic environment. A possible explanation for this dis-
crepancy might be that the YN-6 medium used by these latter
authors contains high concentrations of antibiotics which
may interfere with recovery of attenuated bifidobacteria.
(iv) Summary. The genus Bifidobacterium offers poten-
tial as a fecal indicator of raw water quality because it
derives exclusively from feces (with high numbers in human
feces and low numbers in animal feces) and it does not multi-
ply in water (narrow temperature range, low tolerance of
dissolved oxygen, high nutrient requirements). Better
selective media are urgently needed in order to study the
distribution and survival characteristics of the Bifido-
bacterium group in raw water.
f. Candida Albicans. .--.Candida albicans is a dimorphic
yeast which can be differentiated from other fungi by its
formation of germ tubes and chlamydospores (Ahearn, 1974).
The former are elongated cell extensions with no con-
strictions at the point of origin and are induced when young
cells are incubated for 1 to 3 h, at 37°C, in calf serum.
Chlamydospores are large, thick-walled, refractile bodies
formed on pseudomycelia produced by growth of C_. albicans on
corn meal or oxgall agar or other appropriate media, at 22
to 26°C.
-------
220
Although, other species of Candida (viz., _C. tropi-
calis, CI. parapsilosis, C. guilliermondii, (2. krusei, (2.
pseudotropicalis ) are involved in human infection [Ahearn,
1978), C. albicans (syn. (2. stellatoidea ) is by far the most
frequently encountered. While classical pathology involves
oral, vaginal, and skin loci (Winner and Hurley, 1964), more
recent awareness of the role of C_. albicans as a serious
opportunistic human pathogen has led to the increasing
frequency of reports of systemic infections affecting a wide
variety of internal organs (Kashkin, 1974; Ahearn, 1978).
(i) Sources of _C. Albicans in the Environment. .Human
feces contain C_. albicans in numbers ranging up to 10 per g
while its occurrence in raw sewage may be as high as 25,000
per 1, although a few thousand per 1 is more common (Buck
and Bubucis, 1978). In addition, large numbers of C_. albi-
cans are shed by individuals with active cases of diaper
rash or vaginitis caused by the yeast. In the latter case,
it is estimated that over two million cases of vulvovagi-
nitis of Cl. albicans etiology occur annually in the U.S.
(Ahearn, 1978) . Increasing numbers of infections can be
correlated with the popularity of oral contraceptives .
The organism is known to occur in domestic and farm
animals, although little information is_available for these
hosts. Feces of beagle dogs contain 10 - 10 cells per g,
and over 10 per g have been recorded from pigeons. C.
albicans also occurs in sea gulls and marine mammals ~~
( dolphins , porpoises, etc.) although no quantitation has
been recorded.
Englebrecht and coworkers (1977) found yeast densities
of up to 400 per 1 in chlorinated sewage effluent. They
showed that C. parapsilosis, C. krusei, and other yeasts had
a high chlorTne resistance compared with coli forms. C_.
albicans has also been isolated frequently from chlorinated
( ^u/t 0 . 3 - 0.5 mg per 1) aquarium water containing marine
mammals (J.D. Buck, unpublished data), and chlorine toler-
ance has been noted (Jones and Schmitt, 1978) . The organism
is capable of surviving in natural seawater, in situ, for at
least six days (Buck, 1978).
(ii) Significance of C_. Albicans in Water Microbiology.
Prom the above, it is clear that C2. albicans occurs in
natural environments; but little is known concerning mainte-
nance of infectivity and infective dose. Since there is no
-------
221
direct evidence of candidiasis in humans as related, to the
consumption of water, C. albicans probably does not repre-
sent a public health threat in drinking water, although
McCabe (1977) suggests the possible use of yeasts as indi-
cators of drinking water quality, based on their chlorine
resistance. Bonde (1977) includes "Candida, and yeasts"
among agents of waterborne diseases. Brisou (1975) has
noted increased incidence of vaginal infections in women
using polluted recreational waters. With the necessity for
recycling water in some areas, potential microbiological
problems must be considered (Lund, 1978). Environmental
stress may have a profound effect on the ability to recover
indicator bacteria (Bissonnette, et al., 1975). Some reports
are available on the occurrence of pathogens in water in
which no thermo-tolerant coliforms were detected (Bonde,
1977; Berg, et al., 1978). In fact, the concept of bio-
indicators is being reexamined (Hoadley and Dutka, 1977;
Pipes, 1978).
(iii) Enumeration of C. Albicans. The definition of
C. albicans is clear-cut, unlike that of total and thermo-
tolerant coliforms, fecal streptococci or other indicator
groups. A standard description and characterization is
provided by van Uden and Buckley (1970). In addition, a new
method for detection and enumeration of C. albicans (Buck
and Bubucis, 1978) has been incorporated into the Environ-
ment Canada Microbiology Methods Manual for the Analysis of
Waters, Wastewaters, and Sediments and will soon undergo
testing in accordance with the American Society for Testing
and Materials (ASTM) prior to eventual establishment as a
standard procedure. The method has proved useful for assess-
ing bathing water quality (Sherry, et al., 1979).
(iv) Recommendations for Future Research. The sig-
nificance of C_. albicans in raw and finished water would be
better understood if some additional research were done.
More should be known of the rate at which C. albicans is
shed in feces of normal humans and animals and of the role
of sea gulls and other birds as sources of this organism in
water. Second, the effects on £. albicans from sewage
treatment (including chlorination), temperature, salinity,
pH, industrial wastes, etc., need further study. ^It would
also be well to determine whether sediments in drinking
water collection basins concentrate or protect C_. albicans,
as occurs with coliforms. Finally, ratios between densities
-------
222
of C. albicans and accepted bioindicators should be measured
for as many situations as possible.
(v) Summary. Candida albicans is a yeast derived from
the feces of humans, animals, and birds. Although this
organism has not yet been implicated in any drinking water
disease outbreaks, it has emerged as a significant opportun-
istic human pathogen with common occurrence in environmental
and wastewaters. Therefore, it would be prudent to occa-
sionally monitor for Candida albicans in raw and finished
water. A membrane filter method is now available that is
suitable for these analyses.
g. Vibrio. Vibrio species are widely distributed in
aquatic environments. Recent studies have shown that Vibrio
cholerae, Vibrio parahaemolyticus, and other pathogenic
vibrios are found in brackish water and estuarine habitats
(Colwell, et al., 1977). Cholera has long been recognized
as a waterborne disease; the infectious agent enters water,
directly from the infected host or indirectly via waste-
water, from areas in which are found clinical cases of
cholera, persons in the incubation stage, or healthy car-
riers. The first documented outbreak of cholera in the
U.S. since 1911 (Center for Disease Control, 1978c) has
brought into focus the potential for cholera outbreaks in
those geographical areas hitherto considered cholera-free
because of the assumption that good sanitation practices and
high standards for waste treatment provide protection for
the community. Two newly described groups, the lactose-
positive and Group F vibrios, have been found to be asso-
ciated with disease in man. Thus, the potential health
hazard of Vibrio species, transmitted via the water route,
has been well documented and there is now a recognized need
to consider vibrios as potentially useful indicator organisms
of water quality.
(i) Taxonomy of the Genus Vibrio. The description of the
type species, V. cholerae of the genus Vibrio, is as follows:
short, curved or straight rods, single or united into spirals,
that grow well and rapidly on the surfaces of standard culture
media, are asporogenous and gram-negative, and produce L-lysine
and L-ornithine decarboxylases, but not L-arginine dehydrolase
or hydrogen sulfide (Kligler iron agar). The overall deoxyri-
bonucleic acid (DNA) base composition is about 48 + 1 percent
-------
223
guanineplus cytoslne. V. cholerae includes strains that
may or may not be hemolytic, may or may not be agglutinated
by Gardner and Venkatraman O group I antiserum, and may or
may not be lysed by Mukerjee V. cholerae bacteriophages I,
II, III, IV, and V (Buchanan and Gibbons, 1974; Sebald and^
Veron, 1963).
V. cholerae strains are grouped into 53 serotypes,
based on possession of O antigens. However, the nori O-I
vibrios, termed NCV, NAG, or non O-I of Heiberg groups I and
II, also have been implicated in cholera-like diseases. In
addition to the difficulties in assessing potential patho-
genicity of the non O-I V. cholerae, Vibrio species often
are confused with species of other genera such as Aeromonas,
Spirillum, Plesiomonas, and other related groups. A table
of features useful in separating these genera has been
published elsewhere (Colwell, and Kaper, 1977).
(ii) Enumeration Methods. In order to enumerate
Vibrio species in water, enrichment must be done prior to
enumeration. For this MPN procedure, a three dilution,
five-tube replication series in which 1 liter and 100 ml
volumes of water are filtered, is recommended. The volumes
employed will vary according to the species of Vibrio (viz.
V. parahaemolyticus, V. anguillarum, etc.) and the numbers
in which they are present in the sample. The filters are
placed in 50 ml Of alkaline peptone broth after filtration
is completed (Colwell and Kaper, 1977). A third dilution is
obtained by inoculation of a 10-ml sample into 10 ml double-
strength alkaline peptone broth containing 10 g per 1 pep-
tone (Difco Laboratories, Detroit, Michigan), 10 g per 1
sodium chloride (NaCl), pH 8.4. After inoculation, the
enrichment flasks are incubated at ambient temperature (-""V
25°C) for 6 to 8 h, followed by 18 h incubation at 35 to
37°C, yielding a combined incubation period of approximately
24 h. The cultures should be streaked onto TCBS agar
(Baltimore Biological Laboratories/ Div. of Bioquest,
Cockeysville, MD) at 6 h and after overnight incubation.
Plates should be heavily inoculated from enrichment cultures
and incubated at 35°C for 18 to 24 h. V. cholerae typically
appear as large, smooth, yellow colonies, slightly flat-
tened, with opaque centers, and translucent peripheries.
For slide agglutination and other identification
procedures, colonies should be picked from TCBS and trans-
ferred to gelatin neopeptone agar containing no added NaCl.
The gelatin neopeptone agar plates should be incubated at
35°C for 18 to 24 h. Gelatinase positive strains that are
-------
224
also oxidase positive can be tested with polyvalent O group
I antiserum by slide agglutination. Non O group I strains
should be tested fpr lysine and ornithine decarboxylase,
arginine dehydrolase, and other selected characteristics'
(Colwell and Kaper, 1977).
Several enrichment broths for V. parahaemolyticus have
been suggested. A modification of the arabinose, ethyl
violet broth of Horie and coworkers, containing (in g per
1): peptone (Difco Laboratories Inc., Detroit, Michigan),
5.0; beef extract (Difco), 3.0; NaCl, 30.0; bromothymol
blue, 0.03; ethyl violet, 0.001; and galactose, 5.0, pH 9.0,
yields good results (Kaper, et _al., 1979). After incubation
for 24 h at 37°C, the enrichment broth cultures are streaked
onto TCBS and incubated at 37°C for 24 h. Blue-green colo-
nies typical of V. parahaemolyticus can be picked to Brain
Heart Infusion Agar for further testing. See Colwell and
Kaper (1977) for recommended differentiating features.
Because of their halophilic nature, the lactose posi-
tive and Group F vibrios must be cultured on heart infusion
agar, which usually contains 0.5 percent or more Nad, or on
a suitable marine agar. Enrichment may be achieved, in the
case of the lactose-positive vibrios, by using a marine
salts solution '(Kaneko and Colwell, 1978) with lactose added
as a fermentable carboyhydrate in a properly buffered medium.
Enrichment methods for both the lactose positive and Group F
vibrios, however, have yet to be fully documented.
(iii) Significance of Vibrio Species as Water Quality
Indicators. With the geographically widespread occurrence
of cholera and cholera-like disease, it is clear that Vibrio
cholerae and related vibrios are of immediate, practical~
interest to public health authorities responsible for water
quality. The diagnosis of vibrio-induced disease has in-
creased in recent years, due, in part, to greater awareness
of the potential pathogenicity of Vibrio species.
The primary indicator organism for water quality has
long been E. coli, the presence of which is considered to
signify fecal contamination and the presence of potentially
pathogenic bacteria. in contrast to IS. coli, however, V.
cholerae, V. parahaemolyticus, and the lactose-positive~and
group F vibrios have been demonstrated to be pathogens and
are associated with foodborne disease and wound, ear, and
eye infections in man. it would be very useful to develop a
"Vibrio Index" for raw quality assessment based on total
-------
225
vibrio counts. The evidence available to date would
suggest that when the total count of vibrios is elevated,
the presence of the pathogenic vibrio species is more prob-
able and the human risk factor much greater. The signifi-
cance of large numbers of vibrios in the aquatic environ-
ment, both freshwater and marine, has yet to be determined,
so no absolute limits can be set at this time. It is neces-
sary to point out that a correlation between standard
pollution indices and numbers of Vibrio species in water or
food has not been documented. Quite to the contrary, it has
been shown that no relationship exists between the counts of
V. parahaemolyticus and IS. coli in estuaries. Thus, periodic
analyses for Vibrio in raw waters could be a useful adjunct
to routine water analyses.
(iv) Conclusions. Vibrios are a potential health
hazard in the environment, although their fate and role as
pathogens in finished drinking water is largely unknown. In
order to determine levels in the environment and to deter-
mine whether or not a possibility of this pathogen in fin^
ished drinking water exists, the periodic analysis for total
vibrios is considered useful.
(v) Summary. Vibrios are geographically widespread-
pathogens not necessarily associated with fecal pollution.
Recent outbreaks of cholera and other Vibrio-associated
illness have indicated that these pathogens have significant
health implications. Methods for their enumeration are
available for freshwater and estuarine species. The presence
of large numbers of vibrios in a water sample suggests that
pathogenic species of Vibrio may also be present. Inasmuch
as established microbial indicator systems are poorly
correlated with the incidence of pathogenic vibrios, a
Vibrio-specific test affords the only potentially valid
basis for predicting the presence of these pathogens [See
also Section B.l.a(vii)].
h. Aeromonas. The genus Aeromonas includes straight,
gram-negative, facultatively anaerobic rods that closely
resemble members of the Enterobacteriaceae, but instead, are
included in the family Vibrionaceae (Buchanan and Gibbons,
1974). Unlike the enterics, Aeromonas is motile by means of
a polar flagellum; and, more important diagnostically, it is
cytochrome oxidase positive. For these reasons, it resembles
Pseudomonas; however, the metabolism of Aeromonas is
-------
226
fermentative as well as respiratory. Like Pseudomonas, it
is also capable of utilizing "high molecular weight sub-
strates such as proteins (e.g., casein and gelatin), deoxy-
ribonucleic acids, starch, dextrin, and glycerol (Buchanan
and Gibbons, 1974; Lennette, et al . , 1974). Aeromonas
occurs in uncontaminated waters as well as in sewage and
sewage-contaminated waters . The organism may be pathogenic
for humans, other warm-blooded animals, and cold-blooded
animals including fish (Buchanan and Gibbons, 1974; Lennette,
et al . , 1974) . Because of its consistent occurrence in
appreciable, but variable, numbers in raw and treated sewage
and sewage contaminated waters, the use of Aeromonas as a
potential indicator of raw water quality has been considered
(Cabelli, 1977).
Among the three species belonging to the genus Aero-
monas , only A. hydrophila and A. punctata, as well as their
subspecies, are of sanitary importance. Only these two
species occur as free-living organisms in water, whereas A.
. salmonicida an<^ its subspecies are fish pathogens that do
not occur xn surface waters, unless associated with infected
fish. Several subspecies of A. hydrophila and A. punctata
may be pathogenic not only for fish, amphibians (e.g.,
frogs), and reptiles (e.g., snakes, turtles, and alligators),
but also for warm-blooded animals including humans, mice,
and guinea pigs, when the organisms are ingested or enter
the body through a wound; however, only A. hydrophila,
subspecies hydrophila , is of concern to human health
(Buchanan and Gibbons, 1974; Lennette, et aJL. , 1974; Schubert,
1967a) . This organism has been isolated from a wide variety
of specimens of human origin, including blood, wounds,
ulcers, pus, throat swabs, urine, bile, and feces of persons
with diarrheal disease, as well as from normal stools
(Lennette, et al . , 1974) ; however, the organism is isolated
from feces of normal individuals only infrequently, and then
only in low numbers (Cabelli, 1977). Cases of dysentery
caused by A. hydrophila have been documented (Annapurna and
Sanyal, 1977) [See Section B.I. a]. Dysentery caused by
Aeromonas-like organisms may also be due to Plesiomonas
shigelloides (formerly Aeromonas shigelloides ) . The genus
Plesiomonas, and its single species, P. shigelloides was
created for those organisms resembling Aeromonas, but
lacking some essential features of the genus. This includes
the possession of a restricted carbohydrate metabolism and
the absence of the enzymes diastase, DNase, lipases, and
proteinases. Also ]?. shigelloides has multiple polar
flagella grouped into a tuft, rather than possessing a
single flagellum (Buchanan and Gibbons, 1974; Lennette
et al. , 1974^- 2* shigelloides has been isolated from the
-------
227
^
feces of humans and other mammals, and was responsible for
causing two known outbreaks of acute infectious gastrb-
enteritis. It has also been found in association with
Shigella in persons with dysentery; in fact> some strains of
P. shigelliodes possess an 0-antigen identical to Shigella
sbnnei (Buchanan and Gibbons, 1974; Lennette, et. al. , 1974).
(i) Enumeration Methods. Membrane filtration using
DSF medium (Schubert, 1976) is capable of detecting A.
hydrophila and A. punctata in drinking water. The multiple
tube method employing Rimler-Schotts medium has been exten-
sively used to enumerate A. hydrophila in natural waters
(Schqtts and Rimler, 1973).
Aeromonas is not fastidious in its growth requirements,
and abundant growth occurs on nutrient agarand blood agar.
Aeromonas grows on-the differential-selective media used for
coliform plating, such as eosin-methylene blue (EMB) agar,
MacConkey agar, and Salmonella-Shigella (SS) agar, mimicking
coliform colonies, and occasionally on brilliant green agar
producing colonies resembling Salmonella (Hendricks, 1978;
Lennette, et al., 1974; Mack, 1977).The oxidase test is a
rapid and easy means to distinguish the Aeromonas colonies
from the coliform or Salmonella colonies, and its use is
thus of cardinal importance in total coliform detection
procedures (Mack, 1977).
(ii) Evaluation as an Indicator. Because A. hydro-
phila and A. punctata are capable of degrading high molecular
weight organic wastes such as starches, lipids, and proteins,
their enrichment occurs in waters that have received waste-
water additions. Massive multiplication of Aeromonas has
been noted to occur even in pipes draining household wastes
(Schubert, 1967b).
In raw sewage, it is not uncommon for Aeromonas concen-
trations to reach 10 colony-forming units per ml. During
primary treatment their numbers are not reduced, but a
significant decline in their numbers occurs during secondary
treatment, the degree depending on the biological treatment
method employed (Schubert, 1967c). Reductions in the numbers
of Aeromonas could possibly be used to indicate the extent
of pathogen removal during treatment of wastewaters and in
the final effluent. It has been shown that the elimination
of Aeromonas correlates well with the elimination of Sal-
monella and other pathogens (Schubert .and Schafer, 1971).
-------
228
Prom these and other results, it appears that the enumera-
tion of Aeromonas from treated sewage effluents may be used
as an index to determine whether pathogens remain in the
wastewater. Its detection from*surface waters or ground-
waters indicates sewage contamination of that water.
Once introduced into water, Aeromonas may grow on the
organic pollutants contained in the wastewater. Growth of
Aeromonas has been reported to occur on filters at potable
water treatment works as well as in other locations con-
taining sufficient organic material (Hendricks, 1978) [See
Sections E.3 and 4]. The detection of high numbers of
Aeromonas has been reported to be a reliable indicator of
the degree of contamination of a water source with organic
wastes (Schubert and Schafer, 1971).
The proportions of aerogenic to anaerogenic subspecies
of A. hydrophila may also be used to evaluate the degree of
wastewater contamination. Table C.2.h-l describes the sub-
species of the two important Aeromonas species from the
standpoints of their aerogenicity and their distributions in
raw waters (Buchanan and Gibbons, 1974).
In wastewaters, or in waters containing wastewaters and
solid wastes, the Aeromonas population consists almost en-
tirely of the anaerogenic varieties, with very few aerogenic
organisms isolated. The reverse is true of good quality
surface waters in which Aeromonas densities are low, but of
those organisms isolated, 95 to 98 percent are aerogenic
(Schubert, 1975b). in groundwaters, aerogenic Aeromonas
varieties are found at the surface, whereas the anaerogenic
varieties occur in deep or undisturbed areas (Schubert,
1976).
Studies of the water from river bank infiltration wells
demonstrates that the proportion of aerogenic aeromonads
increases with increasing distance from the point of appli-
cation, so that as purification of the river water occurs by
passage through the soil, the population of Aeromonas shifts
from predominately anaerogenic to predominately aerogenic
(Schubert, 1976). Thus, the isolation of anaerogenic aero-
monads from raw water indicates contamination with inade-
quately treated wastes.
Aeromonas resembles coliforms in all respects except
that it possesses polar flagella and is cytochrome oxidase
positive (Buchanan and Gibbons, 1974; Hendricks, 1978;
Lennette, et al., 1974). During total coliform isolation,
-------
TABLE C.e.h-1
DISTRIBUTION OF AEROGENIC VERSUS ANAEROGENIC AEROMONAS IN WATER
A. Hydrophila Subspecies
A. Punctata
Subspecies
Hydrophila Anaerogenes Proteolytica Punctata Caviae
Gas produced from:
glucose
glycerol
Isolated from:
uncontaminated
water
contaminated
water
ND
ND
No data available.
to
to
-------
230
the lactose-positive, aerogenic strains of Aeromonas are
also recovered, since the multiple tube test detects gas
production. Therefore, the presence of Aeromonas elevates
the apparent total coliform concentration (Leclerc, et al.,
1977b; Mack, 1977). In the FRG, coliform analysis of~dlFTnk-
ing^waters revealed that 30 percent of the samples having
positive presumptive tests also contained Aeromonas (Leclerc,
et al., 1977b). Thus, it is essential to distinguish these
organisms from coliforms. Unfortunately, Aeromonas dif-
ferentiation from coliforms is neglected in the current
standard procedures for total coliform testing used in the
U.S. (Leclerc, et al., 1977b).
(iii) Conclusions. Aeromonas is a saprophytic organism
that occurs in both marine and fresh waters. Its presence
in raw waters does not signal any significant health hazard;
however, these organisms are opportunistic pathogens espe-
cially if ingested in large quantities (e.g., after enrich-
ment in foods, although as yet few food-related diseases due
to Aeromonas or the closely related Plesiomonas have been
reported).High numbers of Aeromonas also are undesirable
in bathing waters, for they may act as opportunistic patho-
gens if they gain entry into the body via a wound.
^Because Aeromonas is a potential pathogen, it must be
considered undesirable in drinking waters. Aeromonas is
generally^sensitive to disinfection treatments, so its
presence in finished waters indicates inadequate disinfec-
tion or subsequent contamination [See Section P.2.a]. As an
indicator of finished water quality, Aeromonas detection has
no advantages over the detection of total coliforms.
Aeromonas survives in environmental waters considerably
longer than E. coli or most coliforms. However, because it
is capable of aftergrowth, as is true of some members of the
total coliform group, but not of E. coli, Aeromonas detec-
tion may falsely signify contamination with sewage. Small
numbers of aerogenic Aeromonas are common in groundwater, at
the surface, and in contaminated surface waters, so they do
not connote any hazard to public health; whereas anaerogenic
Aeromonas varieties are predominant in wastewater and their
detection indicates contamination of the water with sewage.
Because the current Aeromonas enumeration methods do not
distinguish aerogenic from anaerogenic varieties, detection
Of Aeromonas could lead to false condemnation of a raw
water.
The aerogenic varieties, although present in smaller
quantities than the anaerogenic types, are already being
-------
231
detected along with total coliforms in the-.MPN total coli-
form procedure, and all types are probably isolated when
using the membrane filtration procedure. Thus, it_seems
that the easiest way of determining the presence of Aero-
monas would be to do an oxidase test following total coli-
form isolation. The most important feature of Aeromonas is
that its presence may yield inaccurately elevated estimates
of coliform densities, so any total coliform detection _
procedure should also include the oxidase test to determine
their presence. Because of its natural occurrence in the
environment, potential for aftergrowth in nutrient-enriched
water, and confusion with total coliforms, Aeromonas would
not be a very specific indicator of sewage discharge or
pathogen occurrence; where raw water quality is concerned,
Aeromonas should be considered as a general indicator in
relation to nutrient discharge.
(iv) Summary. Some members of the genus Aeromonas,
particularly A. hydrophila and A. punctata are potential
pathogens that are widely distributed in the environment.
They thrive in heavily contaminated waters, particularly _
those derived from kitchen wastes. The anaerogenic varieties
are common to contaminated waters, therefore, their presence
may be considered an indicator of the degree of organic _
contamination of raw waters. However, because Aeromonas is
capable of aftergrowth in environmental waters, its detec-
tion may exaggerate or falsely signal a public health hazard.
Because Aeromonas is very sensitive to disinfection, its
presence in drinking water is an indication of inadequate
treatment or a breakdown in the distribution system. Due to
its similarity with the total coliforms, Aeromonas, if
present, will be detected along with them and may even
constitute a significant proportion of the detected total
coliforms. The extent to which this occurs may be deter-
mined with the aid of the oxidase test.
If for some reason, drinking water is to be used without
disinfection, aerogenic Aeromonas may cause false positive
results in both the multiple tube and membrane filter total
coliform tests with water that is not unsafe. Anaerogenic
Aeromonas seen in the membrane filter coliform test with
sucSsSmples would, in fact, signal a hazard. No Aeromonas
should be present in finished drinking water that has been
disinfected, so both false (due to Aeromonas) and true
positive results in the total coliform test are significant
with such samples. In a sense, the presence of Aeromonas
Augments the sensitivity of the total coliform test for
finished drinking water.
-------
232
3. Other Indicators of Microbial Quality
Three relatively new biochemical methods have been
employed to determine water quality. They do not measure
microbial numbers directly, as do the indicator systems
presented in the previous two sections. instead, microbial
numbers are determined indirectly in the case of Limulus
lysate and adenosine tr iphosphate,- these tests for fecal
sterols serve as a chemical index of fecal pollution.
Although these methods certainly must be measured against
the more traditional microbiological indicator systems,
their validity as indicators ultimately must be decided on
the basis of their own merits.
a- Limulus Amebocyte Lysate. Limulus amebocyte lysate
(LAL) is an aqueous extract of blood cells (amebocytes)
from the horseshoe crab, Limulus polyphemus (Levin and
Bang, 1969). LAL forms a clot or becomes turbid when incu-
bated with lipopolysaccharides (LPS), also known as endo-
toxins, which are components of the cell wall of all gram-
negative bacteria.
An elevated LPS or bacterial biomass concentration in a
water sample does not, by itself, constitute a health hazard
to a normal, healthy human. However, an elevated LPS or
bacterial biomass level indicates contamination of a water
source by gram-negative bacteria and in certain areas, for
example, near sewage outfalls, may indicate the presence of
potentially pathogenic bacteria (Evans, £t al., 1978) [See
also Sections A.l.f and B.l.a(ix)]. LAL currently is used
to detect and quantitate endotoxins in a variety of solu-r-
tions by a technique commonly referred to as the LAL or LPS
test (Watson, et al., 1977).
The LPS test has been modified for use in determining
bacterial biomass in aquatic environments (Watson, et al.,
1977). The concentration of LPS in a sample is mealured
before and after centrifugation at 10,000 x g. The level of
LPS that is free (in the supernatant after centrifugation)
is subtracted from the total LPS (in the sample before
centrifugation) to determine how much LPS was bound to
cells. Finally, the bacterial biomass is obtained by multi-
plying the concentration of bound LPS by a factor of 6.35
(Watson, et al., 1977) to obtain the grams of bacterial
cellular carbon in a given volume of water.
(i) Methods of Reading the LPS Test. The LPS test
can be read either by a clot or a spectrophotometric method,
-------
233
With the clot method, 0.1 ml of the sample is added to 0.1
ml of the LAL and incubated for 1 h at 37°C. -If a firm
clot is formed after incubation, the test is considered
positive. This can be applied as a quantitative test by
reacting an LPS standard, prepared in twofold serial dilu-
tions, with LAL. The endpoint of the test is the lowest LPS
concentration that will trigger clot formation,of LAL,
Appropriate twofold dilutions of the water sample also are
made and reacted with LAL. The highest dilution in which
clotting occurs is multiplied by the sensitivity of the LAL
to obtain the LPS equivalents in the test water, expressed
as ng per ml.
The clot method is easy to perform and requires minimum
operator skill. However, it does not allow for exact quan-
'titation of LPS in a water sample'; instead, this can be done
using a spectrophotometer by the method of Watson and
coworkers (1977). In this method, 0.2 ml of LAL is added to
a 1.0-ml portion of the sample. For freshwater samples,
dilutions should be made with pyrogen-free distilled water,
whereas seawater dilutions are made with a 3 percent pyrogen-
free sodium chloride solution. After incubation for 1 h at
37°C, the samples are gently mixed by rotary agitation and
the absorbance of each read in a spectrophotometer at 360
nm.
A standard curve is generated by incubating LAL with
known concentrations of LPS in a dilution series that can
range from 0.1 to 100 pg per ml. The limits of the range
vary according to the sensitivity and range of linearity of
the particular LAL preparation used. The concentration of
LPS can then be determined by making serial dilutions of the
sample and reacting these with LAL. LPS concentrations in
samples are determined by comparing their absorbance read-
ings with those found in the linear region of the standard
curve.
(ii) Evaluation of the LPS Test as an Indicator of
Water Quality. Some studies have demonstrated the presence
of LPS in raw and finished water. DiLuzio and Friedman
(1973) detected LPS in 16 of 18 finished water samples using
an insensitive assay with a detection limit of 10 ug per_ml
(the two negative drinking water samples came from artesian
wells). When the positive samples were measured for endo-
toxin content by taking serial dilutions, they were found to
contain LPS levels of between 1 and 10 ug per ml. Raw
surface water, used as a drinking water source, was tested
-------
234
in the same study and found to have LPS levels of 20 to 400
ug per ml. In other studies (S. Watson, unpublished results),
LPS levels of between 1 and 10 ng per ml were found in
finished water of Woods Hole (Massachusetts, U.S.).
A number of samples of drinking water and renovated
wastewater from advanced wastewater treatment plants .was
analyzed by Jorgensen and coworkers (1976). Drinking water,
sampled from ten cities in the U.S., had LPS concentrations
ranging from < 0.625 to 500 ng per ml. When the processes
employed by the water treatment plants were examined, no
association was found between use of activated charcoal and
high endotoxin levels. But, high LPS levels were found in
the effluents of five out of eight advanced wastewater
treatment plants surveyed., and positive results did corre-
spond with those plants that treated with activated carbon
[See Section E.5]. One of the other three wastewater treat-
ment plants, in which only low endotoxin levels were re-
vealed, also used activated carbon, but followed it with
breakpoint chlorination [See also Sections E.2 and F.2.a].
One plant, which employed a reverse osmosis unit as the
final treatment step, produced water containing no detect-
able LPS. As more and more water is being reclaimed due to
limited water resources in the world, application of ad-
vanced wastewater treatment for producing potable water will
increase.
Evans and coworkers (1978) used a spectrophotometric
assay with tapwater and found the sample to contain 1.19 ng
per ml of LPS, of which 1.10 ng per ml was free LPS and 0.09
ng^per ml was bound or cell-associated LPS. They also used
this assay with several environmental samples to determine
the correlation between endotoxin levels and bacteriological
parameters. The correlation coefficients between bound LPS
and the other parameters were: coliform (0.907), enteric
bacteria (0.946), gram-negative bacteria (0.934) and hetero-
trophic bacteria (0.952) [See also Sections C.I.a to c and
D]. Whether all environmental water exhibits this high
correlation, and whether the same holds true for drinking
water, has not been established.
(iii) Potential Applications of the LPS Method. The
LPS method could be applied i'n long-term routine monitoring
and also as an emergency method to determine whether there
has been an increase or decrease in gram-negative bacteria
in_raw or finished water. Another application for which
this test may be particularly well suited pertains to the
-------
235
increased use of aranular and powdered activated carbon in
the treatment of drinking water to remove organic compounds.
Since bacteria are likely to grow on the adsorbed organic
material, the LPS assay could prove useful for monitoring
bacterial growth [See Project Area II.b] . Likewise, this -.
assay could serve as a fast and simple method for determining,
periodically, the LPS or bacterial levels in home point-of-
use water treatment devices, such as activated carbon and
reverse osmosis units [See Section G.2]. Even though these
devices are known to eventually foster bacterial growth if
not properly maintained (Wallis, et al., 1974), their_use in
the home has increased. Commercially available LAL kits
could easily be adapted to serve as convenient household
monitors. /
The endotoxin assay will not replace coliform and
colony counts or other methods routinely used to survey
bacterial populations' in municipal water supplies [See
Sections C.I.a to c]. However, it could serve as a fast and
convenient additional technique for measuring parameters
related to public health [See also Section C.4.gJ. The
values could be used to indicate long-term trends, reveal
differences from city to city, or rapidly determine sources
of contamination in emergency situations, even in the
field, if necessary.
(iv) Summary. The LAL or LPS assay for LPS quanti-
tation is fast, accurate, and inexpensive. LPS, a component
of the cell wall'of all gram-negative bacteria^is of poten-
tial harm to humans and other animals only.after entering
the bloodstream. However, the LPS concentration in a volume
of water is directly related to the level of gram-negative
bacteria contained in that water. Thus, the LPS test has
considerable potential for use in monitoring raw and finished
water for the presence of gram-negative bacteria or endo-
toxin, including assessing bacterial contamination due to
growth on activated carbon filters. It .also could be of
great assistance in evaluating the efficacy of advanced
wastewater treatment employed as a means of producing potable
water.
b. Adenosine Triphosphate. There have been many
attempts to quantify cellular components in an effort to
estimate cell numbers and, in this manner, to develop a
nonspecific, quantitative indicator of biota in drinking
water. Most of these attempts have been less than success-
ful because of the difficult problem of distinguishing among
-------
236
compounds present in viable, active cells and those present
in dead cells or detrital material. Adenosine-5'-triphos-
phate (ATP) is a compound which, to a large extent, circum-
vents this problem. ATP shows good potential as a non-
specific indicator of viable organisms in drinking water
because it is: (1) present in all living cells; (2) rapidly
hydrolized by cellular enzymes upon cell rupture or death;
and (3) relatively constant in proportion to the biomass of
many studied organisms.
^A figure of 0.4 percent of total carbon is usually
applied when 'determining the ATP content of a cell; hence,
the total carbon is estimated as 250 times the ATP mass.
Although this represents a moderate concentration within the
organism, direct chemical analysis of ATP is not usually
possible because the numbers of organisms most often encoun-
tered in drinking water are low.
The^most commonly used method to determine ATP levels
is the firefly luciferin-luciferase assay. ATP is extracted
from the cells and reacted with firefly lantern luciferin
and luciferase, in the presence of magnesium ions and at the
proper pH (which depends on the extractant used). This
reaction results in the production of light and may be shown
as follows:
M 2 +
E + LH2 + ATP -2 >
E
LH,
AMP + PP
E
LH
AMP +
----- > E + AMP + CO, + hv- + T
where E is the enzyme luciferase, LH is reduced luciferin,
AMP is adenosine-5'-monophosphate, PP is pyr ophosphate , ht/
is light, T is thiazolinone (which is decarboxylated reduced
luciferin) . The light emitted when ATP is injected into the
enzyme reaction mixture rises in intensity quickly and then
decays exponentially. Both the maximum intensity, measured
as peak intensity, and the total light output are propor-
tional to the amount of ATP added, as-long as it falls
within a range of 1 x 10 to 2 x 10 g per 1.
(i) Methods for extracting and quantifying ATP. When
testing environmental samples, extracting ATP from micro-
organisms and stabilizing it can become more difficult. Not
only can the growth rate and physiological state of cells
from various sources affect the content per cell of ATP, but
also the extraction process itself may alter final ATP
readings . ,
-------
237
The extractant must be capable of immediately inacti-
vating intracellular and extracellular ATP.ae.es and Hos-
ass-.- £?^^r.f^.ft
capable of efficiently extracting the ATP from the large
variety of cells which may be encountered. Both organic
materials (such as humic substances) and inorcranw- «tio.nB
often associated with drinking water (such as Ca++, Mg ,.
Al*?₯ -?e and Mn ) may interfere to,varying degrees,
depending on the method of extraction used. A judicious
choice of extractants will usually prevent such interference
(Tobin, et al., 1978).
Extractants include tris-(hydroxy methyl)-aminomethane
(Tris) buffer, sulfuric acid, nitric acid, dimethyl sul-
foxide, glycine buffer, phosphate buffer, and many others.
Of the most commonly used extractants, boiling alkaline (pH
10) glycine buffer, containing magnesium-ethylene-diamine-
tetraacetate, offered the most efficient_combination^of
extracting ATP (in the presence of organic and inorganic
constituents) and stabilizing it (Tobin,. ft al. , 1978).
When automated systems were employed to concentrate^the
sample and extract ATP, the use of nitric acid afforded
adequate stabilization (Picciolo, 1977; 1978). The most
Sommon extractant currently in use for environmental samples
is boiling Tris buffer (American Public Health Assoc.,
1976). It is the basis of one of the original extraction
techniques (Holm-Hansen and Booth, 1966), and one on which
much of the ATP data is based.
The enzyme reaction and apparatus for detecting and
measuring luminescence have been fairly well established.
Several tvpes of photometric equipment are commercially
avI!lailiywSLh quantify ATP by integrating the lightoutput
over a period of less than 1 min, or by determining the peak
light output during the first few seconds of the reaction,
or both. Other equipment, such as liquid scintillation
counters, fluorometers, and reaction-rate analyzers may also
be used to quantify the light output.
Perhaps the greatest problem with using the biolumi-
nescent method for determining ATP levelj^iJnSv^h^
is the need to concentrate the sample sufficiently. The
linear range of ATE detection for a typical photometer is_ ^
down to about lO'1 g ATP under ideal sample conditions. The
averaglbacterium his been variously estimated to contain
3 x 101 g ATP (Picciolo, et al., 1977a), or from 2 to 40 x
10"l6g ATP, depending upon growth phase ^Hamilton and Holm-
Hansen, 1967). If the value of 3 x 10 16g ATP per cell is
-------
238
used, it can be calculated that a minimum of 3 x 103 cells
would be required to evoke a quantitative response under
ideal conditions. If the maximum acceptable level is 500
cells per ml of finished water, then it is necessary to make
estimates based on this critical quantity, and therefore
concentration of the sample usually is required.
A recent report (Picciolo, et aj.. , 1978) compared
several methods of concentration which included centri-
fugation; in-line membrane filtration; radial- and tangen-
tial-flow, flat-surface membrane concentration; and hollow
fiber membrane filtration. Only the hollow fiber membrane
filtration procedure (modified for backwashing) provided
adequate concentration and recoveries. A commercially
available unit, with a filter area of 10 cm, was capable of
concentrating the sample 600-fold with up to 88 percent
recovery after backwashing steps had been performed. This
unit offers good potential for determining levels of bacteria
which one may find in tapwater. Unfortunately, the method
has yet to be tested extensively on actual tapwater.
(ii) Evaluation of ATP as an Indicator. Few studies
have as yet been performed on finished water, and more often
pertinent data obtained from raw water studies are used for
drinking water analyses. One study has been reported briefly
(Rexing, 1977) in which the luciferase ATP method was used
to replace the tedious microscopic counting of algae, and to
verify chlorination efficiency on drinking water. Raw water
was filtered through a 3.0 urn membrane filter which was then
submerged_in boiling Tris buffer. ATP levels-.were usually
about 10~ g per 1 and increased to about 10 g per 1
during a Cyclotella bloom. It was demonstrated that most of
the algae which passed through the treatment system were
killed and were, therefore, not positive for ATP; in this
instance, then, the method was not suitable for monitoring
total algal counts in finished water. A free chlprine
residual of 1.6 mg per 1 caused at least a 99 percent decrease
in the biomass, as measured by the luciferase ATP method.
Surface water of Lake Ontario was shown, in a limited study
(Afghan, et al., 1977), to harbor an average of 5 x 10 g
ATP per 1.
Although no extensive studies have been made of the ATP
content of drinking water, some comparisons between bacterial
estimates calculated from a model plate count system, and
those from an ATP system, have been made (Picciolo, et. al.,
1977a) . In this study, E. coli cultures were diluted~"in~~
-------
239
0.9 percent NaCl or dechlorinated tapwater. The results were
linear from ICTto 10y bacteria per ml by both methods, and
the ratio between plate counts and counts estimated by ATP
content did not differ from 1 at the 5 percent confidence level
However, when treated effluents from a pilot sewage treatment
plant were analyzed, the luciferase-measured ATP indicated
bacterial levels of 1.8 times the number obtained by viable
plate counts. This was likely attributable to residual ATP
in viable cells which were debilitated to the point of being
unable to undergo cell division under the culture conditions
used.
A limited study on the ATP content of a municipal
tapwater in Canada (J. Cairns, unpublished data^gshowed 11 x
10 Q g per 1 (range of five values was 5.3 x. 10 to 16 x
10 g per 1) and after passing through a 0.6 urn polycar- .,
bonate (Nuclepore) filter, the ATP content was reduced to
3.7 x 10 per'l (range was 2.5 x 10 to 6.3 x 10 g peg
1). While this would indicate the potential of about 10
bacteria per ml, it has been shown that chlorine-stressed
coliforms, thermo-tolerant coliforms, and total hetero-
trophic bacteria are much more reduced in ability to grow^on
their respective media than the decrease in ATP would indi-
cate (Cairns, et al., 1979). Thus, it is improbable that
this number of bacteria would be found by standard colony
counts.
These results point out a problem which has yet to be
resolved: that is, the effect of stress on the ATP content
of organisms and on their ability to divide. It appears
that in some stress situations, organisms lose their ability
to divide, and the ATP decreases to a minimum level which
may be a maintenance or survival value (Dawe and Penrose,
1978). A systematic study should be made of the effects
of chlorine on the total microbial population in water, so
that the interpretation of data will more closely reflect
actual conditions. ,
(iii) Potential Applications of the ATP Method.^ The
luciferin-luciferase assay has been shown to be a reliable
nonspecific'method for the analysis of the viable bacteria
in water. Because of its potential to enumerate micro-
organisms in water, it may become useful in supplementing
total plate counts in water, especially when applied as an
on-line system to signal any changes in water quality at a
water treatment plant. It may also be employed_during_
emergencies when the rapid analysis of samples is required,
-------
240
rrrvsoe, ^ ~4- ' -, "." event of disruption of a water main or
gross bacteriologic contamination of samples.
_Presently, the major problem with the ATP assay is the
requirement for concentration in order to quantify the
bacteria in water. Progress in this area has been made with
the use of hollow fiber filters, which should enable the
technique to be used routinely in the near future. A system
of semi-automatic filtration and analysis, both in one
apparatus (Picciolo, et al., I977b), would seem to lead the
way to rapid and accurate determinations. Some studies have
been initiated on the use of immobilized enzymes in the
performance of the ATP assay (Lee, et al., 1977). The use of
immobilized enzymes may increase the potential for auto-
mation and economy of reagents.
The combination of flow concentration and immobilized
enzyme_reactors in close proximity to phototubes could
significantly increase the number of samples processed and
lower the cost per assay. it may be possible to have an
automated sampling system which monitors raw water input and
finished water output, and which alerts the waterworks
operator to any unusual changes in values for either of the
two. Long-range trends in the quality of water should be
monitored closely by the use of retrospective data analyses,
and any trends noted in the deterioration of water qualitj
should evoke appropriate corrective steps.
(iv) Summary. ATP is a compound present in all living
cells and is degraded rapidly upon cell death. The luci-
ferin-luciferase assay for ATP can be used for analyzing the
approximate numbers of microorganisms in water. The reaction
is valid between about 1CT to 1CT bacterial cells and gener-
ally requires large concentration factors for the low densi-
ties of bacteria encountered in drinking water. Concentra-
tion can be achieved by means of hollow fiber filters to
about 600 times with up to 88 percent recoveries. The
principal applications of the ATP analysis are in monitoring
?i?S °on^en^ fnd removal in raw water, and in determining
total bacterial numbers and disinfection efficiencies in
finished water. other potential uses are analyzing samples
during main breaks and monitoring long-term changes in
treatment and in the distribution system.
c' Fecal Sterols. Coprostanol or coprosterol (5&-
cholestan-3£-ol) is the major fecal sterol of man, comprising
-------
241
40 - 60 percent of the total neutral sterols excreted.
Studies to date, identifying it in sewage and surface waters,
have lent support to proposals that it be considered for use
as a fecal marker of water pollution especially pollution
of the type associated with industrial or wastewater treated
discharges (Murtaugh and Bunch, 1967; .Smith, et al., 1968;
Smith and Gouron, 1969; Kirchmer, 1971; Tabak, et al.,
1972; Dutka, et a^L. , 1974; and Wun, et _a_l. , 1976). If
coprostanol were confirmed to originate only in the feces of
man and higher animals, as is generally believed the case,
then it would be a highly specific chemical indicator. And,
being a chemical indicator, coprostanol analysis would have
the advantage over bacterial indicators of requiring less
preparation and processing time. Furthermore, various
purification processes applied to drinking water or simple
passage through soil in the case of groundwater, can remove
fecal organisms without removing some of the associated
fecal material. Testing for coprostanol could be useful in
this regard, especially in connection with drinking water
analyses. However, more definitive information is needed
before coprostanol analysis of water quality could be per-
formed routinely.
(i) Cholesterol Reduction to Coprostanol. The reduction
of cholesterol to coprostanol and the various factors affecting
such a transformation in biological systems have been studied
by many workers. Earlier investigations have been documented
in several reports (Gould and Cook, 1958; Hoffman, 1970;
Kirchmer, 1971). It has been shown that the hydrogenation
of cholesterol to coprostanol is mediated by microorganisms
in the large intestine of man and higher animals. Apparently,
this is the only source of the 5^-stanol. Recently, an
anaerobic gram-positive eubacterium, having an absolute
requirement for A -3^ -hydroxy steroids, was isolated from
human feces and from rat intestinal contents (Eyssen, et
al., 1973; and Sadzikowski, et al., 1977). This diplobacillus
Feduced the 5, 6-double bond of cholesterol, campesterol,f-
sitosterol, and stigmasterol, yielding the corresponding 5p-
saturated derivatives. When cholestanone (4-cholesten-3-
one) and coprostanone were incubated with this organism, the
carbonyl and the 3-oxo groups of these molecules were reduced
to a 3P-hydroxyl group (Eyssen, et al.., 1973). Using dual
labeled cholesterol as a tracer, Parmentier and Eyssen
(1974) concluded that the major pathway for the conversion
of cholesterol to coprostanol in the animal intestine involved
the intermediate formation of 4-cholesten-3-one followed by
a reduction. ,
-------
242
(ii) Coprostanol Removal During Sewage Treatment. A
wastewater plant which employs an activated sludge process
will produce an effluent low in or free of coprostanol
(Murtaugh and Bunch, 1967; Smith, et al., 1968; and Smith
and Gouron, 1969). This removal was believed to be mainly
physical (Smith, et al., 1968). However, by comparing
coprostanol degradation rates in chlorinated versus un-
chlorinated sewage effluent, Kirchmer (1971) concluded that
coprostanol removal by the activated sludge process was due
principally to biodegradation. C. K. Wun and coworkers-
(unpublished data) detected unusually high levels of copro-
stanol in the activated sludge. They found coprostanol
breakdown to occur at a much slower rate when sewage was
combined with sterilized (membrane-filtered) fresh water and
seawater than when combined with unsterilized (unfiltered)
water. It is likely that both physical and biological
processes are at work removing coprostanol during activated
sludge treatment.
(iii) Environmental Effects on Coprostanyl Degra-
dation. The rate of coprostanyl ester degradation is essen-
tially the same as that of free coprostanol. Coprostanyl
esters are present in large amounts in feces (approximately
30 percent of the total coprostanol), and the ratio of the
steryl esters to the total sterols was found to be 10-15
percent in sewage samples (Kirchmer, 1971). Using radio-
isotope labeled coprostanyl palmitate, Wun, Walker, and
Litsky (unpublished data) found that the ester bond was
rapidly hydrolyzed resulting in 80 percent degradation
during one to three days of any environmental exposure.
This rapid hydrolysis may occur under aerobic conditions in
the sewer system prior to the treatment plant, and may
account for the low ratio of steryl esters to sterols found
in raw sewage samples.
(iv) Agreement between Coprostanol and Other Accepted
Criteria. In an attempt to establish fecal sterol criteria
and standards, efforts have been made to correlate cop-
rostanol concentrations with other accepted water quality
standards. Tabak and coworkers (1972) reported that copro-
stanol concentrations in non-chlorinated polluted water
correlated well with fecal coliform densities in the same
locations. Other investigators obtained contrasting results
(Kirchmer, 1971; Dutka, et al., 1974; Dutka and El-Shaarawi,
1975; and Wun, et al., 1976J7 Kirchmer (1971) postulated
that fluctuations in bacterial populations (from chlori-
nation and from additional coliforms released in waterfowl
feces) might explain why no clear relation could be shown
-------
243
between coprostanol and bacterial densities. Singley and
coworkers (1974) reported a linear relationship with a
coefficient of determination (r ,) of 0.977 between copro-
stanol and total organic carbon (TOC) in sewage samples.
Results of correlating the coprostanol, concentration with
biochemical oxygen demand (BOD) and chemical oxygen demand
(COD) were less satisfactory.
(v) Significance of Coprostanol in Water. The unique
characteristics of coprostanol recommend its use as a chemi-
cal index for water quality determination. It appears that
the only source of this chemical is the feces of man and
higher animals. It is biodegradable and can be removed from
sewage by adequate treatment. Studies have also shown that
coprostanol concentrations correlate closely with the extent
of fecal pollution (Murtaugh and Bunch, 1967; Smith and
Gouron, 1969; Kirchmer, 1971; Tabak, et aJL. , 1972; and Wun,
et al., 1976). Coprostanol, as opposed to coliforms, remained
unaffected by wastewater chlorination, heat, and toxic
industrial discharges; hence, coprostanol quantitation was
seen to overcome many flaws inherent to coliform analyses
(Kirchmer, 1971; and Tabak, et al., 1972). The relative
stability of coprostanol is especially significant in view
of current trends of rapid industrial - development and the
increasing emphasis on disinfection of raw and treated
wastewater.
Because of its tolerance of environmental rigors,
coprostanol might also be used for monitoring the source,
course, and extent of fecal pollution in the ocean or
brackish waters where bacteriological evidence is often
doubtful. It also offers promise as a method for analyzing
the distribution and mixing dynamics which occur when, for
example, sewage effluent enters river water, or when sewage
polluted surface water infiltrates into groundwater. Exami-
nations of water that reveal the presence of fecal sterols,
but no fecal bacteria, still indicate fecal pollution.
Conversely, when indicator bacteria are present, but fecal
sterols are not detected, this may indicate bacterial
regrowth or infiltration, rather than fecal pollution,
although degradation of the sterols must also be considered.
(vi) Coprostanol Analysis. The routine procedure for
estimating coprostanol levels in water consists of hexane
extraction, clean-up by thin layer chromatography (TLC), and
-------
244
subsequent analysis by gas-liquid chromatography (GLC) (Mur-
taugh and Bunch, 1967; Smith, et al., 1968; Smith and Gouron,
1969; Kirchmer, 1971; Tabak, et. al. , 1972). The TLC clean-
up procedure has been considered necessary for all surface
water samples regardless of concentration (Murtaugh and
Bunch, 1967; and Tabak, et aJL. , 1972). However, processing
samples through an alumina micro-column affords a more
efficient clean-up procedure (Dutka, ^t a^., 1974). The TLC
clean-up can also be omitted if a sample containing > 1 ug
of coprostanol per 1 is washed instead with acetonitrile
saturated with hexane (Kirchmer, 1971).
In earlier studies of coprostanol quantitation, analysis
included mild alkaline hydrolysis and/or preparation of
trimethylsilyl ether derivatives (Murtaugh and Bunch, 1967;
Tabak, et al., 1972). Results reported by Kirchmer (1971)
suggest that omission of these time-consuming processes does
not significantly affect coprostanol identification and
quantitation. Later findings led to elimination of the
saponification step, resulting in analysis of only the free
sterol (Dutka, et al., 1974; Wun, et al., 1976; ,1978).
Although the hexane liquid/liquid partitioning proce-
dure was satisfactory for coprostanol extraction, it required
an excess of reagents and was often further complicated by
emulsion problems (Murtaugh and Bunch, 1967). In addition,
much of the coprostanol present in heavily polluted areas is
bound to particulate matter (Kirchmer, 1971; Wun, et al.,
1978). Switzer-House and Dutka (1977; 1978) have "shown
that, in most samples, > 70 percent of the fecal sterols
are^associated with the pellet after centrifugation. Hexane's
immiscibility with water may protect the fecal sterols,
trapped within aggregates, from contact with the solvent.
Under such circumstances, fecal sterols are unlikely to be
extracted completely (Wun, - et ai., 1978). A column method
has-been developed which employs XAD-2 polystyrene polymeric
resin for extracting coprostanol from water (Wun, et al.,
1976; 1978). This method has been shown to be effective and
avoids the excessive use of chemicals. However, adsorption
and desorption are pH-dependent; the solvent must be dried
repeatedly due to traces of water in the acetone eluate.
Recently, the column technique has been improved by using
XAD-1 non-ionic resin chemically similar to, but more
porous than, XAD-2. Results indicate that XAD-1 resin
eliminates the need for adjusting pH. Rapid drying of the
eluate may be accomplished by means of a basic methanol
-------
245
partial clean-up step added to the new column procedure.
This entails displacing water in the column with methanol
prior to elution (Wun, Kho, Walker, and Litsky, unpublished
data). The use of benzene during elution, however, is
undesirable for routine testing and a more acceptable solvent
to remove the fecal sterols from the resin is now being
sought.
(vii) Comparison of Coprostanol, Cholesterol, and
Coprostanone as Indicators. Three of the neutral steroids
found in the feces of man and higher animals namely,
cholesterol (5-cholestan-3f-ol), coprostanol (S^-cholestan-
3^-ol), and coprostanone (5^-cholestan-3-one) have been
suggested as potentially useful chemical indicators of fecal
pollution in water (Murtaugh and Bunch, 1967; Smith, et
al., 1968; Tan, et al., 1970; and Dutka, et a_l. , 1974T7
Although cholesterol constitutes a major component of the
fecal sterols, it is also present in eggs, milk, lard, and
wool grease (Murtaugh and Bunch, 1967). Except in specific
instances, this sterol does not specifically indicate fecal
pollution. In addition, cholesterol seems to be widely
distributed at its limit of solubility in seawater (Smith
and Gouron, 1969). Coprostanone, on the other hand, is
usually a minor component of feces, being present at about
one-tenth the concentration of coprostanol. It was suggested
that coprostanol may be readily oxidized to coprostanone in
the environment and that this ketone is, in effect, a more
stable chemical indicator (Dougan and Tan, 1973). However,
environmental contaminants interfered less in the detection,
by means of gas liquid chromatography, of coprostanol,
rather than coprostanone (Wun, Walker, and Litsky, unpub-
lished data).
(viii) Conclusion. Coprostanol satisfies the generally
accepted criteria for a good indicator of fecal pollution.
To employ this fecal sterol for the practical assessment of
water quality, however, more work is needed. Although this
chemical appears to originate from the intestine of man and
higher animals, it has been found to occur in the feces of
chickens and in tissues of other animals (Kirchmer, 1971);
also, seaworms obtained from sewage-polluted seawater may
excrete large amounts of coprostanol in their feces (Wun,
Walker, and Litsky, unpublished data). High concentrations
of coprostanol have been shown to absorb onto sewage partic-
ulates, as do viruses. Hence, it is likely that by ingesting
-------
246
and concentrating these contaminated particulates, various
aquatic organisms have been responsible for the high levels
encountered. Whether lower animals concentrate coprostanol
exclusively by ingestion, or whether they also harbor micro-
organisms which can transform cholesterol to coprostanol,
remains to be learned. Two procedures, hexane liquid/liquid
partitioning and polystyrene polymeric resin column extrac-
tion, are presently available for coprostanol isolation from
waters. Gas-liquid chromatography is commonly used for
quantifying this fecal sterol.
Coprostanol is biodegradable* nonetheless, Dutka and
coworkers (1974) recovered high concentrations of this
sterol at a substantial distance from fecal sources, sug-
gesting a possible resistance to biodegradation. More
studies are needed to determine how coprostanol is degraded
and how various environmental factors affect decomposition
rates. Also needed are studies to ascertain coprostanol
persistence and baseline levels in surface waters and in
sediments.
(ix) Summary. Coprostanol is the major fecal sterol
of humans, comprising 40 to 60 percent of the total neutral
sterols excreted. This compound can be concentrated from
water and quantified by gas-liquid chromatography in a
relatively short time. It is considered a good indicator of
fecal pollution because it can be biodegraded during adequate
sewage treatment, but is unaffected by chlorination, heat,
and toxic industrial discharges; hence it overcomes some
inherent problems of conventional bacterial indicators.
Further testing of this method is required to obtain more
baseline levels in the environment and to ascertain the
factors affecting coprostanol persistence.
4. Potentials for Mechanization, Automation,
and Shorter Read-Out Times
As water becomes increasingly more subject to reclama-
tion and pollution, and owing to the limited resources
available for microbiologic analyses of water, the direction
taken in analytic procedures inevitably will be toward
large-scale economies. Some of the more successful of these
approaches are those that automate, simplify, or otherwise
modify the methodology to enable the work to be more precise
and accurate, and the results to be obtained more quickly.
Increasing the productivity of a method often means that
more samples can be taken, thus, increasing the degree of
-------
247
certainty of results. The methods presented here range from
improved versions of established methods to completely new
approaches to water microbiology.
a. Radioactive Isotope Methods. Methods in chemistry,
especially clinical biochemistry, have evolved rapidly
during the last 20 years. Autoanalyzers as well as chroma-
tographic and electrophoretic equipment have become an
integral part of most diagnostic laboratories, enabling
rapid and accurate chemical analyses. Radioimmundassay has
become an essential, sensitive tool for diagnosing various
viral diseases and assaying simple molecules, including
hormones.
However, progress in microbiology over the past 30
years has been slow. Diagnostic procedures have not changed
significantly. Most of the conventional methods are based
on the physiological properties of the various bacterial
strains, especially their ability to ferment certain car-
bohydrates or to metabolize specific amino acids. The
sanitary quality of drinking water is still usually deter-
mined by counting eoliform bacteria, which serve as indica-^
tors of fecal contamination [See Sections C.I.b to c].
Methods currently practiced in some countries, for
detecting coliforms, entail membrane filtration followed by
the fermentation of carbon sources such as lactose [See
Sections D.2 and 4]. Biochemical changes become apparent
only after a prolonged (minimum 24 to 48 h) incubation to
permit the growth of large numbers of the-indicator (it
takes 1.7 x 10 coliforms to produce 1 mm of gas Levin,
1963). Thus, contaminated samples of drinking water may
already have been consumed before public health officials
are apprised of the danger and can institute corrective
action. More rapid eoliform detection methods are needed so
that, ideally, bacteriological results would be obtainable
within 6 h or less. There is also a need for semi- or
fully-automated systems for rapid on-line eoliform detection
in water treatment plants or at strategic monitoring points
in the distribution system [See also Sections C.4.c to g3.
Radioactive isotopes have been used, extensively, to
study biochemical changes, including bacterial fermenta-
tions. These techniques are much more sensitive than the
conventional bacteriological procedures; therefore, it is
not surprising that labeling compounds with C to help
detect, quantify, or identify contamination indicators was
-------
248
first proposed more than twenty years ago. Unfortunately,
despite its merits, this approach has not received the
attention which it deserves and is not widely employed for
routine microbiological monitoring. There are two appli-
cable radioactive isotope methods, one based on fermen-
tation; the other, based on isotope incorporation.
(i) Fermentation Method. This approach makes use of
known biochemical properties of the various bacterial strains.
Levin and coworkers (1956) used C-labeled lactose broth to
utilize the coliform's ability to metabolize lactose, unlike
soil organisms which show no ^-galactosidase activity. They
reported that this method gave presumptive evidence of
coliforms in water in 1 h. Other carbon sources including
C C]glucose and [ C]sodium formate have also been tried,
but since these reactions lack specificity, they cannot be
used to identify or quantify coliforms.
The sensitivity of the assay can be markedly increased
by including additional nutrients in the medium, thereby
enhancing bacterial multiplication which in turn contributes
to the fermentation process. As in the standard procedures,
inhibitors can be added to the isotope-containing medium,
rendering it selective. Thus, adding bile salts prevents
the growth of soil bacteria but permits that of coliforms.
In addition, raising the temperature from 37 to 42°C (as
has been suggested)'increases the uptake of labeled carbo-
hydrate, along.with other enzymatic reactions, thereby
accelerating CO~ production.
The radioisotope procedure was modified further with
the introduction of membrane filters for concentrating
bacteria in large volumes of water. This modification
permitted the detection of two to ten coliforms within 6 h
(Bachrach and Bachrach, 1974; Reasoner and Geldreich, 1978).
The radioisotope procedure has limitations, however, in that
it does not permit the exact quantitation of an indicator.
The rate at which coliforms ferment carbohydrates may be
affected by their previous environmental exposure or han-
dling. Starvation, lack of nutrients, or., toxicity of membrane
filters may decrease the initial rate of.. . CO? release,
producing a nonlinear CO2 curve when [ C]lactose is the
carbon source. This may result from a lag in the induction
of ^-galactosidase production [See also Section C.4.g].
-------
249
(i.l) Detection Techniques. Two techniques for col-
lecting and measuring radioactive CO,, are: (1) trapping
CO,, on filter(lpapers or by hyamine injection, and counting
with a Geiger-Muller or a scintillation counter; and (2)
continuous monitoring of the CO,., by flushing the cultures
with gas, and counting with an analogue recorder connected
to an ionization chamber (Deland and Wagner, 1969).
(i.1.1) Trapping CO_. ..Various means have been
employed to trap the released CO_. The reaction is usu-
ally carried out in a sealed test tube or container which is
shaken in a water bath. Filter papers containing either
barium hydroxide (Levin, 1963) or potassium hydroxide are
attached to the lid of this container. The reaction is
usually stopped by injecting acid, which also releases the
radioactive CO_ trapped in the solution (in the form of
carbonate when the pH of the medium is high). The filter
papers containing the radioactive CO- are then transferred
to scintillation vials and assayed for radioactivity by
liquid scintillation spectrometry (Bachrach,,and Bachrach,
1974), or dried and counted with a Geiger-Muller counter.
An alternative method, which may be more precise, uses
rubber stoppers to support center wells (K-882320, Kontes
Glass Co., Vineland, New Jersey, U.S.) into which hyamine is
injected (Bachrach and Bachrach, 1974). The reaction is
stopped in the same manner as above, and the center wells
are transferred to scintillation vials and counted. In
principle, this radioisotope technique can be modified to
monitor potable water quality automatically, as suggested by
Pijck and Defalque (1963) arid Reasoner and Geldreich (1978).
(i.l.2) Continuous Monitoring of CO_. This approach
has the advantage of being automatic and permitting con-
tinuous monitoring of radioactive CO_ (Morello, 1975). It
should, however, be pointed out that accurate results can be
obtained only when bacteria are grown in an acidic medium to
prevent trapping the CO,, in the medium.
There are several commercially available instruments
which afford continuous automatic monitoring. The best
known is the BACTEC system (Johnston Laboratories, Cockeys-
ville, Maryland) which represents an important advance in
the field of automation for microbiology. It was originally
designed to detect bacteria, automatically, in blood cul-
tures and to test for sterility, but it may easily be used
to monitor water quality.
-------
250
With the BACTEC system, the sample is injected into a
culture medium containing a C-labeled substrate. There-
after, the culture vial is attached to a cylinder of gas
(containing 10 percent CO-), connected to an ionization
chamber, and incubated at the required temperature. At the
start of the cycle, needles, which have not yet descended
into the culture vial, are heated to sterility; and the
pressure in the ionization chamber is lowered to about 1/2
atm. The needles then descend into the culture vial; gas
from the cylinder is forced through a sterilizing filter
into the vial and then swept out into the ionization chamber,
where any radioactive CO- is counted.
(ii) Isotope Incorporation Method. The fermentation
methods just described are quite sensitive in that, a single
coliform bacterium may metabolize many molecules of the
radioactive carbohydrate; the amount of evolved radioactive
CO2 is a function of cell number, incubation time, and
temperature. Yet, these methods do have drawbacks. "In-
activated" coliforms, which are unable to multiply (due to
chemical or physical water treatment), may still ferment the
carbohydrate. To avoid inclusion of these in the results,
alternative procedures have been proposed. Incorporation
methods measure bacterial growth, usually to the exclusion
of inactivated microorganisms. Khanna (1973) used a radio-
metric assay, based on the synthesis of macromolecules,
wherein he incubated bacterial cells with P and measured
the amount of isotope incorporated into cellular nucleic
acids. At the end of incubation, cells were collected on
Whatman No. 40 filter paper, washed, dried, and their radio-
activity determined. This procedure has been used for
quantitating E. coli cells.
Radioactive phosphorus is not the only isotope which
can be used for incorporation studies. Radioactive sulfur
has been used in attempts to detect extraterrestrial life
(Levin, 1968). This isotope may easily be used to quanti-
tate coliforms in water. Using radioactive phosphorous or
sulfur has the advantage that these isotopes may be obtained
at high specific activity (unlike radioactive amino acids,
purines, or pyrimidines), facilitating labeling and quan-
titation of bacteria. However, they may present a health
hazard because they emit relatively high-energy @ particles;
in addition, their short half-lives may lead to practical
difficulties. Since sulfur isotopes would be incorporated
into bacterial proteins and phosphorous isotopes, into
-------
251
nucleic acids, they would permit the detection of bacterial
growth, but could not be used, as such, for the identifi-
cation of specific bacteria.
This procedure clearly requires modification before it
can be applied directly to water quality control. Soil
bacteria behave like coliforms in that, both groups in-
corporate the radioactive elements. Therefore, inhibitors
should be added to the growth medium to permit selective
growth. Bile salts and related compounds probably would
permit the growth of coliforms while inhibiting the multi-
plication of any bacteria not in the family Enterobacteria-
ceae.
Tritium-labeled amino acids, carbohydrates, or purines
are usually not suitable for incorporation studies. These
isotopes cannot be obtained at sufficiently high specific
activities? therefore the assay is not very sensitive. On
the other hand, iodinated amino acids may be used and would
probably permit the detection of bacteria even in small
numbers.
(iii) Application. Both the fermentation and incor-
poration methods provide means of testing for sterility and-,
as such, could be applied to water samples after chlorina-
tion or other treatment.
As mentioned earlier, radioactive lactose may be used
to differentiate between coliforms and soil bacteria. Other
radioactive carbohydrates may similarly be used and may
help in classifying indicator types. Related biochemical
properties of specific bacterial strains should also be
considered. 13. coli, for example, shows lysine decarboxy-
lase activity, and releases radioactive CO,, when incubated
with [1- C]lysine (labeled at the carboxyl group). In this
way, IS. coli may be differentiated from other Enterobac-
teriaceae; when this test is run concurrently with [ C]
lactose or [ C]glucose incubations, total coliform or
total bacterial levels, respectively, may be ascertained.
Incorporation studies may be used to test the sensi-
tivity of the organisms to various agents. Any decrease in
isotope incorporation in the presence of the drug would be a
sensitive indication of growth inhibition. By monitoring
isotope incorporation, in parallel cultures incubated aero-
bically and anaerobically, one may identify aerobic and
anaerobic bacteria.
-------
252
(iv) Conclusions. The radioactive isotope method,
applied either to bacterial fermentation or isotope in-
corporation, is a useful tool for the early detection,
quantitation and identification of microorganisms in water.
It can be automated and applied as an on-line system to
signal any changes in water quality. The fermentation of
radioactive substrates offers a means of differentiating E.
coli from other coliforms, and coliforms from soil bacteria,
but it does not necessarily distinguish between viable and
inactivated organisms. Enzymatic reactions can be accel-
erated by certain modifications to increase radioactive
CO output, although initial rates of CO production are
no always linear due to metabolic lag periods or previous
environmental exposure. Isotope incorporations permit easy,
accurate coliform quantitation, but do not identify specific
bacteria. The use of membrane filters permits detection of
sparse bacterial numbers in large volumes of water. The
procedures are sensitive and permit the detection of two to
20 coliforms within 6 h.
(v) Summary. There are two radioisotope methods
applicable to drinking water microbiology: one based on
fermentation, the other based on isotope incorporation. The
fermentation method utilizes the ability of bacteria to
metabolize specific C-labelled compounds, whereby the
CO 2 released can be trapped and quantified. A judicious
selection of substrates, inhibitors, and reactions will
enable detection of specific groups such as coliforms.
b. Impedance Methods. The impedance technique offers
potential as an automated, rapid, and non-specific indicator
of viable microorganisms, analogous in several ways to the
use of ATP [See Section C.S.b]. Both detect only viable
microorganisms, and the response of both is a function of
the number of viable organisms present. No work, however,
has been published on the use of impedance for micro-
biological monitoring of drinking .water, although this
technique has been applied to microbiological problems in a
number of other areas.
The impedance technique has been applied to several
problems of practical importance in clinical microbiology,
food microbiology, wastewater management, and environmental
monitoring. Using this method, clinical microbiologists can
screen rapidly for: microorganisms' tolerance of, or sus-
ceptibility to, antibiotics (Ur and Brown, 1974, 1975a;
-------
- ; 253
Cady, 1978); clinically significant populations (< 10 per
ml) in urine samples as a means of diagnosing bacteriuria
(Cady, 1978); evidence of infection in blood samples (Hadley
and Senyk, 1975); and pathogen characterizations and identi-
fications, such as for Neisseria gonorrhoeae, by simul-
taneous impedance responses to multiple substrates and media
(Hadley and Senyk, 1975; Cady, 1978). The impedance tech-
nique can be used to measure the rate of microbial sulfate
reduction in pure cultures of anaerobic sulfate reducing
bacteria and in estuarine sediment samples (Oremland and
Silverman, 1979).
(i) Principles Underlying the Use of Impedance. When
microorganisms in an aqueous medium metabolize, they gener-
ate a complex series of reactions whereby higher molecular
weight substrates are converted by extracellular enzymes .to
lower molecular weight compounds. These, in turn,, are
metabolized at or within cell surface structures, or they
are metabolized internally. Also, other lower molecular
weight organic and inorganic compounds are transported into
the cell, often accompanied by an influx or efflux of inor-
ganic ions. Eventually, metabolic products of lower molec-
ular weight, including gases, accumulate in the aqueous
environment outside the cell. As a result of microbial
growth, electrically charged molecules and ions in the water
will change in species and increase in concentration over-
time .
If an electric current is passed through the solution
.between two electrodes, a decrease in resistance (with
direct current) or impedance (with alternating current) will
occur as ions and electrically charged molecules increase
in concentration and relative mobility. Reactions also will
occur at or near the surface of the electrodes when charged
molecules or ions are attracted to or adsorbed on the elec-
trode surface, causing the electrode to act as a capacitor.
Thus, the entire system can best be visualized as a circuit
containing both a resistor and a capacitor in series. These
concepts are discussed more extensively by Cady (1978).
Other discussions on principles and applications in micro-
biology may be found in Cady (1975), Cady and coworkers
(1978), Ur and Brown (1974; 1975a,b), Hadley and Senyk
(1975), Munoz and Silverman (1979), Silverman and Munoz
(1979), and Oremland and Silverman (1979).
A certain minimum number of growing microorganisms must
be present before a detectable impedance signal will be
-------
254
delivered. When numbers lower than the minimum are present,
there will be a lag period of no detectable change in imped-
ance signal, until populations reach the minimum detectable
size among the bacteria, this number is generally 10 to
10 per ml. The lag period, or detection time, is inversely
proportional to the number of viable microorganisms present
at zer-o timegand can be as short as one hour for an inoculum
of 10 to 10 cells, or as long as 18 h for one cell (Cady,
et al../ 1978; Ur and Brown, 1974, 1975a; and Silverman and
Munoz, 1979) .
(ii) Instrumentation. Two commercial instruments are
available at present for measuring impedance changes gener-
ated by microbial metabolism. The Bactometer is available
from Bactomatic, Inc., Palo Alto, California, U.S., and the
Bactobridge (formerly called the Strattometer) can be
obtained from T.E.M. Sales, Ltd., Crawley, Sussex, U.K., or
from Koniak and Partners, Ltd., Geneva, Switzerland.
The Bactobridge, described by Ur and Brown (1974),
employs a matched pair of borosilicate glass capillary cells
30 mm in length and 2 mm internal diameter, plated at both
ends with a gold coating. One cell, used as the reference,
is filled with a sterile nutrient medium, while the sample
cell contains nutrient medium inoculated with the sample.
Both cells are inserted into the Bactobridge where, in
order to avoid extraneous effects on impedance, a constant
temperature is maintained to within 0.1°C. A 10 KHz sinu-
soidal alternating current, maintained below 0.5 volts, is
passed through both cells in a bridge circuit. The bridge
circuit is balanced initially using a 10 turn 1 Kohm poten-
tiometer. Thereafter, any impedance imbalance in the bridge
circuit, due to microbial metabolism, in the sample tube is
amplified, rectified, and recorded continuously on a chart
recorder. The only signal obtained, therefore, is that
which relates to the activity of the microorganisms. The
use of the Bactobridge in microbiology was described by Ur
and Brown (1974; 1975a,b).
The Bactometer, described by Cady (1975; 1978) and
Cady, et al. (1978), detects growth by measuring the in-
crease with time in the electrical impedance ratio, R ,
between an inoculated sample vial and an uninoculated sterile
reference vial (both vials contain nutrient medium), accord-
ing to the relation:
RZ ~ Zref/ CZref
Sample11'
where Z
and samp
__.= and Z , are the impedance of the reference
L. C5 J. . _ ScUll MX 6 . . - _ 1 .
pie vials, respectively. In theory, the impedance
-------
255
ratio at zero time should be 0.5000 With identical solutions
in the sample and reference vials. As cells grow and metab-
olize and the impedance of the inoculated sample declines
accordingly, the R will approach 1.0000. In practice,
initial impedance ratios ranged from 0.4500 to 0.5500
(Silverman and Munoz, 1979). More detailed descriptions of
the Bactometer and the theory and practice of impedance
ratio measurements of microbial growth and metabolism have
been published elsewhere (Cady, 1975; Cady, 1978; and Hadley
and Senyk, 1975).
The Bactometer model 32 accepts up to 32 sample/
reference pairs in four modules of eight pairs each. The
instrument generates an alternating current of either 400 or
2,000 Hz, passes it through the electrodes of a sample/ref-
erence pair every 3 sec, reads the impedance ratio, then
indexes automatically to the next pair. A complete cycle of
32 sample/reference pairs is read every 96 sec. The impedance
ratio output for each pair is recorded on individual channels
of a 32-channel strip chart recorder and displayed simulta-
neously on the face.of the instrument. The Bactometer also
converts the impedance ratio data into a form readily stored
in a computer for subsequent processing. A variety of
disposable sample/reference tubes is available, with total
volumes ranging from 2 ml per tube up to 100 ml or more,
complete with stainless steel electrodes.
(iii) Evaluation of Impedance in Microbiology. All
species of bacteria, yeasts, and fungi tested, that produce
visual evidence of growth in broth cultures, generated a
detectable impedance signal. Included are many species
important in public health, such as Escherichia coli,
Streptococcus faecalis, Klebsiella pneumoniae, Pseudomonas
aeruginosa, Salmonella enteritidis, Staphylococcus aureus,
and Shigella species. Lists of microbial species that have
been detected by the impedance technique were published by
Cady (1978) and Cady and coworkers (1978).
A selective medium was introduced into multiple sample
units of the Bactometer 32 to provide a rapid, single-step
most probable number method for enumerating fecal coliforms
in sewage treatment plant final effluents (Munoz and Silver-
man, 1979). A linear relation was found between the log-,,-,
of the number of fecal coliforms in the inocula of sewage
treatment plant effluents and the detection time by the
impedance technique. This permitted the preparation of
standard curves of loglo thermo-tolerant coliforms, plotted
-------
256
against detection time, for determining the number of fecal
coliforms in wastewater directly from the detection times.
Thermo-tolerant coliform counts obtained with this method
were in excellent agreement with counts obtained by the
standard MPN procedure (Silverman and Munoz, 1979).
Detection times determined for varying numbers of
microorganisms in frozen vegetables correlated with numbers
determined by the standard plate count method. This enabled
detection times by the impedance technique to be employed in
screening frozen vegetables for unacceptable levels of
bacterial contamination (Cady, 1978).
(iv) Application of the Impedance Technique to Drinking
Water Analysis. One problem common to all methods for
detecting and enumerating microorganisms in drinking water
is that of low numbers per unit volume. With respect to the
impedance technique, strategies can be employed to overcome
the problem. By accepting multiple samples, the Bactometer
32 should be able to perform a single-step MPN enumeration,
as was demonstrated for thermo-tolerant coliforms in waste-
water (Munoz and Silverman, 1979). Sample bottles and vials
of up to 100 ml or more capacities are available, so that it
should be possible to use larger sample volumes (i.e.,
greater than the 1 ml maximum in agar plating techniques)
for relating detection times to numbers of microorganisms
in a sample. Alternatively, methods for concentrating
organisms could be employed; for example, collecting them on
membrane filters, incubating the filters in an appropriate
medium, and relating numbers to detection time.
Specificity can be incorporated as part of the impedance
technique by the judicious use of selective media, tempera-
tures, antibiotics, etc., when particular genera or species
are sought. The impedance technique should prove useful
particularly in cases where sample turbidity precludes the
use of^membrane filter techniques or standard plate counts.
Turbidity will not affect microbial generation of an impedance
signal, nor interfere with its detection. Finally, the
instrumentation, commercially available, provides automated
and simultaneous analyses of multiple samples, affording
more frequent microbiological monitoring. In addition, its
capacity for computer data storage and processing provides
users with the opportunity for obtaining retrospective
analyses of long-term trends in microbial types and numbers
within a water supply.
-------
257
(v) Summary. Metabolic processes and products of
bacterial growth measurably decrease the resistance of an
electrical current passing through the growth medium. In
this way, the impedance method can be used to detect all
viable organisms present in a sample. Selection of appro-
priate media and growth conditions will enable detection of
a specific group of organisms or even one species. The
method is extremely sensitive and can yield results in as
little as one hour for highly polluted water, or in up to 18
h or longer for a single cell. Although the impedance
method has been employed with several bacterial species of
public health and sanitary significance, it has not yet been
extensively tested for use with drinking water.
c. Automated Sampling, Plating and Incubation. Even
though much is now known about water pollution, the problem
continues to grow with industrialization and urbanization of
society. The bacteriological control of drinking water-
quality is indispensible in this context; but methods
presently used for the purpose have shortcomings, in that
sampling is periodic, transporting of samples imposes addi-
tional quality control problems, and results represent a
static image of a dynamic system. Variations in the bac-
teriological quality of a water supply may result from human
activity, but also from that of wildlife [See Section B.2],
and from climatic [See Section B.3] and geologic [See Section
A] changes. An ideal method would take into account chang-
ing conditions and make possible the determination of a most
probable contamination risk.
A logical solution may lie in the application of continuous
sampling at selected sites at a water source, at the treatment
plant, and along the distribution network [See also Sections
4.a to e]. Although continuous sampling is attainable with
conventional manual methods, particularly the membrane
filtration technique [See Sections C.l.b to c, C.4.f, C.4.h,
D.2, and D.4] greater potential could be realized with the
development of automated sampling and analytic equipment.
Since these may be installed at a sampling site, the use of
such devices would obviate the need for transport.
Several conventional techniques could also serve as a
basis for automated procedures, although some obviously
would lend themselves more readily than others to automation,.
Thus, the multiple tube fermentation technique [See Sections
C.l.b to c, D.3, and D.5] would require extensive adaptation
for use with automated equipment, which would not be practical.
The membrane filtration technique, on the other hand, holds
-------
258
more promise. This technique has been simplified for use as
a manual procedure, but not as applied to automation.
A^laboratory prototype model for automated sampling and
bacteriological analysis has been developed and was ex-
hibited at the Symposium on Rapid Methods and Automation in
Microbiology, held in Stockholm in 1973. The apparatus is
divided into three components: a vessel through which water
is pumped, continuously, from the source to be tested; a
compartment for storing 24 culture plates of a suitable
solid medium; and an incubation chamber. The prototype
works on a direct current of 6 V. It is equipped with a
transformer and can run on the main supply. it also is
connected to a battery for auxiliary power, should the main
current fail.
Water, entering continuously into the first container,
is aspirated at preset intervals of 15, 30, or 60 min for
inoculation onto the culture plates. As each culture plate
is transferred from the storage compartment to the incu-
bation chamber, the lid is removed, and a measured dose (may
be adjusted to any volume between 0.2 and 2.0 ml to permit
testing of up to 48 ml of water in each series) of sample
water is swept evenly over the medium. A rinse cycle with
fresh water follows each inoculation to avoid any bacterial
transfer to consecutive samples. The inoculated plate
passes to the incubation chamber, which may be set at 22,
30, or 37°C.
This model was initially devised for use in the event
of^a catastrophe or war. However, the advantages to be
gained from such a system during normal times are many.
Soon-to-be-published results from pilot studies using this
prototype have demonstrated the efficacy of automated bac-
teriological monitoring of water from different sources. An
industrial adaptation of the prototype, incorporating the
latest developments in electronics and automation, might
enable a wide variety of individual water analyses.
Work is now in progress to increase the versatility of
the model. For example, ways are being sought to incor-
porate the detection of anaerobic sporeformers, such as
Clostridium perfringens, and to permit dilution of samples
of highly contaminated water or concentration of samples for
analysis of finished water.
-------
259
d. Automated Plate Counting. One of... the most slow,
tedious, and routine laboratory procedures practiced today
is the counting of certain groups of organisms or total
numbers of colony-forming units for water quality assess-
ments or ecological investigations. Raw, as well as finished,
water contains indigenous and exogenous microorganisms in
constantly changing numbers [See Section A] and studies of
the bacteriological quality or population dynamics of a
water source usually entail extensive quantification. _
Depending on the purpose of the investigation and the micro-
bial composition and density of the sample, counts generally
are taken on spread plates, pour plates, or membrane filters
[See Sections C.I and D].
Automated colony counting instruments, recently de-
veloped and made available, promise to reduce the monotony
that comes with manual counting and help greatly in micro-
biological testing by the spread plate, pour plate, and
membrane techniques if the operational characteristics are
fully understood.
(i) Principle of Automated Counting. The instrument
usually consists of a viewing monitor and an electronic
counter. A high resolution television (tv) -camera functions
as a rapid scanning light detector that recognizes signals
from a bright field/dark field light source.moving hori-
zontally in parallel lines over the culture plate. The
camera is connected to an electronic circuit with a counter
and a digital readout.. Usually, a tv monitor is also
connected to present an enlarged picture of the culture
plate. This magnified view makes it easier for the operator
to fix the electronic window precisely on the area to be
counted.
The position and size of the window, as well as its
shape (circular, square, or rectangular) are all adjustable.
Adjustments for colony size and optical density (referred
to as sensitivity) also are manually set so that any object
exceeding that threshold setting will be counted. Appearing
on the monitor screen will be an illuminated marker dot
superimposed over the object to be counted, enabling the
operator to check for a correct threshold setting that will
tune out smaller particles and debris and count all colonies.
The automated counter usually can be connected to a computer
for easy data handling and storage.
(ii) Sensitivity and Accuracy. What will constitute
the smallest detectable size depends on the magnification of
-------
26.0
the camera lens, the number of lines in the scanning pattern,
the spot diameter, and the level of contrast between the
colony and the plate. The contrast level, in turn, is
determined, in large measure, by the type of illumination
employed.
Since what is recognized by the counter is the dif-
ference in light intensity between the background and the
colony, the opacity of the agar and, therefore, the type of
illumination applied is of major importance.
There usually are three types of illumination from
which to choose. Reflected light is used with.membrane
filters or spread plates that contain opaque or colored agar
media such as blood or chocolate agar. Membrane filters
also can be counted using reflected light, but only un-
gridded filters are acceptable since grid marks may inter-
fere with detection and produce inaccurate counts. Gridded
filters serve only as a guide in manual counting and should,
if possible, be avoided when using the automated counting
device.
Transmitted light is preferable when the medium used is
relatively transparent (as with pour plates) and the con-
trast normal. Colonies near the bottom of the dish that are
low in contrast may not be counted, but applying an underlay
of sterile agar usually will overcome this effect. Use of
transmitted light also permits counting of plaques [See
Section C.2.a], lipolytic zones, and the like, when cleared
spots contrast with the agar by > 20 percent. Detection is
possible at lower contrast with an auxiliary dark field illumi-
nator .
With an ordinary 50 mm lens, objects measuring about
0.2 mm can be counted if the contrast level approaches 100
percent. If the contrast is low, the colony must be at
least 0.4 mm in diameter. This was demonstrated in an
earlier evaluation (Goss, et al., 1974) in which colonies
could be counted manually alTter 18 h, having attained a mean
diameter of 0.17 mm; but they were detectable by automated
counting only after 48 h, when they had reached a mean
diameter of 0.31 mm. With a 75 mm lens, colonies 0.1 to 0.2
mm in diameter (approximately the limit of resolution for
the human eye) are readily detectable.
(iii) Correcting for Variables. Automated counts
would tend to be somewhat lower than manual counts without
-------
261
application of a correction factor.. This calibration adjust-
ment takes into account distortions or error from colonies
that overlap or that are masked at the periphery of the
plate.
Overlapping or clustered colonies separated by < 0.3 mm
will be counted as multiple or single units depending on
their orientation to the scan lines. It is possible, how-
ever, to calculate the probable frequency of colonies spaced
< 0.3 mm apart if the original sample was homogenous and
distributed uniformly over the medium. This is because the
frequency with which colonies cluster increases in propor-
tion to the bacterial density of the sample (Goss, et al.,
1974; La Grange, et al., 1977; Miller, et al., 1975). Once
determined by experimental manipulation and set on the
machine, this correction factor will adjust the count auto-
matically, but it should be applied only within a range of
50 to 500 counts since densities above or below are subject
to variation from other factors.
(iv) The Spiral Counter. A special type of automated
counter has been developed for use with the spiral plate
method. Here, a small volume (35 ul) of sample is deposited
on the surface of a rotating agar plate in decreasing amounts
from the center to the edge, forming an Archimedes spiral.
The method enables the counting of a single plate for samples
having a 10,000-fold range of bacterial concentrations
(Gilchrist, et al., 1973). Counting is accomplished with a
laser electroEic~~counter, designed to record a preset number
of colonies that have developed from the edge to the center
of transparent plates. If the plate contains more than_the
preset number, the counter stops and displays the area in
which the count is made. The count per milliliter is then
determined from a calibration curve (Donnelly, et al.,
1976). This method is of potential value for counting
colony-forming units from different quality waters where the
total count exceeds 1,000 per ml.
(v) Agreement with Manual Counting. Good correlations
have been obtained in published comparisons using pure
cultures (Brusick, 1978; Fruin and Clark, 1977; Goss, et
al., 1974), mixed flora of human origin (Goss, et al.,
T9~74), cultures from dairy products (Fruin and Clark, 19 //;
LaGrange, et al., 1977), and cells forming plaques from
antibody p?odllction (Katz, et al. , 1977). There apparently
have not been any published accounts comparing organisms
-------
262
from raw water or slower growing organisms. Nevertheless,
the 14th edition of Standard Methods (American Public Health
Association, 1975, p. 911) states with regard to automatic
counters: "Their use is acceptable if evaluation in par-
allel with manual counting gives comparable results."
(vi) Advantages and Disadvantages of Automated Count-
ing. The major advantage of automated counting is the
savings afforded in time and labor. Counting time, includ-
ing manual adjustments and allowance for any operator error,
is less than one-third that of manual counting. Automated
counting also minimizes manual count variations due to
carelessness, fatigue, or other factors leading to human
error or to differences in technique. Several unpublished
studies have been made on this subject, including one by the
American Public Health Association, in which it was found
that for 95 percent of the manual counts, differences between
technicians counting the same plates varied by + 24 .percent.
Scratches, irregularities in the agar, air bubbles,
dust particles, and particles from the sample may be counted
with both transmitted and reflected light, giving mistakenly
high numbers. When illumination is by transmitted light,
stacking ridges and marks such as finger prints on the petri
dish can interfere with the count. Dust particles on the
camera lens also may interfere. These problems are most
evident with plates containing few colonies (Fruin and
Clark, 1977), very small colonies, or with plates requiring
ad3ustment to a high sensitivity. Adjusting the colony size
or re-adjusting the sensitivity control to tune out smaller
particles and irregularities usually will overcome these
interferences.' This is not always possible, however, with
small colonies. Instead, a zero count can be taken for the
freshly poured, spread, or filtered plates before incu-
bation, and then subtracted from the final count. Here, it
is necessary to put the plates in the same fixed position so
that any scratches or elongated particles hold the same
orientation, as before, to the scanning lines.
The automated counter does not distinguish between
colonies of different color. Therefore, only if the back-
ground flora are quite minimal, as with the thermo-tolerant
test CSee Section C.l.c], should automated counting be
used. Otherwise, with tests such as that for total coli-
forms [See Section C.l.b], where background growth is more
developed, automated counting is of limited value.
-------
263
Plates with mold colonies, filamentous bacteria, or
surface spreaders are not well suited for use with_automated
equipment and should be counted manually. Anaerobic cul-
tures incubated on spread plates may be a problem, since
many of these colonies normally have a very low contrast,
along with a tendency to spread at times.
- (vii) Applications of Automated Counting. Automated
counting is useful for all types of studies with pure cul-
tures, and for many others using mixed cultures where the
total number of colonies is to be counted. Differentiation
of colonies by aspects other than size, however, is less
applicable. . . -
Automated counters can be used with spread plates and
pour plates and with ungridded membrane filters (when tests
are highly selective). Since cleared zones also can be
counted when the dark field illuminator is used, automated
instruments probably can be employed for enumerating coli-
phages and cyanophages from different quality waters [See
Section C.2.a]; or they might serve in studies of bio-
chemical differentiation, for example, with quantification
of cleared zones of lipolytic organisms. An auxiliary
camera mounted on a photomicroseope enables objects to be
counted automatically. The automated counter also can be
connected to a computer programmed for statistical compi-
lations to speed handling of the numerous data entries.
(viii) Summary. Although the applications of auto-
mated counters are many, it is still the operator and not
the machine who make the judgments. The automated counter
is no more than a useful tool in the laboratory that can
help to release the technician from boring, repetitive, and
time-consuming work. The high resolution television cameras
with hiqh quality lenses are used to count colonies as small
as 0.1 mm in diameter, and can be used with transmitted or
reflected light. The major advantages over manual methods
are the greater speed and lower cost of automated counting.
The major disadvantage is the inability of the device to
differentiate between colonies and other objects in the same
size range or to distinguish colors. Advances in data
processing methods speed up the handling of the voluminous
data generated.
e. Automatic Enzymatic Methods. Current techniques
for the bacteriological monitoring of finished water have
-------
264
been of value in combating waterborne epidemics, but are not
without shortcomings. Some of these shortcomings may be
overcome by a recently developed, automated method which is
based upon the Technicon auto-analyzer. Current water
sampling and analytical practices entail a series of dis-
jointed, lengthy steps including sample transport and prep-
aration, as well as multi-phased laboratory procedures.
This not only invites mistakes, but permits only inter-
mittent sampling and analysis. Moreover, the coliform test,
on which monitoring of finished water is based, has proved
liable to misinterpretation.
The major advantages of the new apparatus are its
automated and continuous performance, its ability to be
installed at the water treatment site, and its increased
sensitivity to the detection of J3. coli (an indicator of
well established significance). This selectivity for E.
coli is accomplished by the introduction of a glutamate
solution under strictly controlled conditions of pH and
temperature. 15. coli was shown to carry on continuous
biosynthesis
acid decarboxylase (GAD), and
surveys of various bacterial species encountered in water
indicate that GAD production is limited to E. coli, Shigella,
Proteus rettgeri, P_. hauseri, and Clostridium species (Leclerc,
1967; Leclerc and Catsaras, 1967). Shigella is only rarely
recovered from water and poses no problems of interference.
Prom the authors' experience, P_. rettgeri and P. hauseri are
always ^ associated with E_. coli and occur in polluted water.
Clostridium species are unable to survive aerobic conditions
imposed in this analytical method. Therefore, the test is
highly specific for E. coli and exclusively so for fecal
contamination.
(i) Sampling and Culturing Mechanism. The entire
apparatus is shown schematically in Figure C.4.e-l. It
operates on the basis of continuous and sequential flow by
means of a peristaltic pump. A 3-way valve (constructed in
a T-shape to eliminate problems of contamination arising
from dead volume) enables attached silicone feeder tubes to
separately draw up water sample and culture medium [See
Figure C.4.e-2], after which these are mixed.
To compensate for the low volume of flow dictated by
this model, a hollow fiber ultrafiltration unit was designed
to reduce an initial 100 ml volume to 5 ml (Trinel and
Leclerc, 1976). The Technicon auto-analyzer can, in this
way, accommodate 120 water samples of 100 ml each (or 12 1)
daily, accomplishing analysis of 100 ml .every 12 min.
-------
FIGURE C.4.e-l
AUTOMATIC APPARATUS FOR CONTINUOUS TESTING FOR E. COLI IN WATER: SCHEMATIC DIAGRAM
Culture medium ,
_Water_sainp3le____ »
sterile
air.
Water minus C02
Acetate
buffer
'0.1M PH3,4
Acetate, buffer-- °-1M 'PH 3-9
* glutamie -«cid~
Air minus-
H2S04 N
_ Coloured re'agent
wwv
Incubation
MMT?
45C
IJe-
bubbler
Jle-
bubbler
S/\M
Recording Colorimeter
to
en
en
-------
FIGURE C.4.e-2
SEQUENTIAL DIAGRAM OF INTRODUCTION OF SAMPLES, CULTURE MEDIUM AND DECONTAMINATING SOLUTION
A & B: PNEUMATIC LIFTERS ALTERNATELY CLOSING OFF THE TUBES SUPPLYING CULTURE MEDIUM,
WATER SAMPLE AND DECONTAMINATING SOLUTION
Pump
Culture medium
_Sample
Decontaminating
Solution
SterHe-«ij!
P
M^^MHBM
MMMMM
«
77
/(
h
It
i
_r
-i
\
I/
n
t
'ft
u.
B
_^
* .
'//
'/*
'//
Pt
1
^n.i
X
'/
y<
«
PI
#
i>y
(/JL
11
^
//
zd
»
i
Ineubatinr
1
i
P: Programmer
.: T-valves
.circuit
en
-------
267
4' Soil and water coliforms (e.g., Citrobacter and
Enterobacter) do not grow well in this medium, and the
nrowth of aquatic bacteria (e.g., Pseudomonas, -Aeromonas,
ISlciSetobacter) is greatly inhibited. Research by Tnnel
and Leclerc (1976) indicated that *iis medium was^more^
sensitive and efficient^ promoting the growth of E. coli
than conventional techniques.
Phthalate HCl-buffer, a disinfectant which is intro-
duced between sample volumes, both sterilizes passages and
Sepa-rates" one watSr sample from another, thereby Preventing
any cross contamination of consecutive samples. Yet, once
diffused into the medium, phthalate buffer takes on the pH
of the medium and imparts no harmful effects to ^ubating
cultures. The continuous circulation of microorganisms
ShrougSout the system makes the Jf* f*?*?^ 4?
critical one. Phthalate HCl-buffer (0.047 M at pH 2.4)
appears, from results of experiments conducted in our labora
tory, to offer an effective means of decontamination.
The mixture of sample and medium next enters a coil of
1
.
air is bubbled through to provide oxygen for
growth.
(ii) Glutamic Acid Decarboxylase (GAD) Detection. ^As
the culture medium leaves the incubation coils, all gaseous
fermentation products are bubbled off and the remaining
iif «ssr.ss:
than one E. coli per 100 ml.
(ill)- Summary. An apparatus has been 'devised which is
particularly suited to the on-site (i.e., treatment plant,
Reservoir, or water main) sampling and analysis of water
-------
TABLE C.4.6-1
BLB FORMULATION
268
Meat extract (Liefoig)
"Bacto-tryptone" (Difco)
Lactose
Yeast extract (Difco)
Sodium chloride
Monopotassium phosphate
Disodium phosphate ' 12
Distilled water
3.0 g
10.0 g
4.0 g
3.0 g
5.0 g
6.0 g
19.7 g
q.s. 1000 ml
-------
269
supplies. It operates continously and automatically, thereby
eliminating manual procedures and providing for uninter-
runted monitoring. Daily water volumes of 10 to 20 1 may be
analyzed within 12 to 13 h using ultrafiltration. _The
technique employed is highly sensitive to and specific for
E. coli, as measured by the presence of carbon dioxide
eVolved from the biosynthesis of GAD.
f. Rapid Thermo-Tolerant Coliform MF Method. Tem-
porary disruptions to the water supply (e.g., water treat-
ment plant failure, line breaks in a distribution network,
or other natural or man-made disasters) create an urgent
need for methods that permit rapid assessment of the^sani-
tary quality of the water. The 7-h -membrane filter fecal
(thermo-tolerant) coliform test (M-7h FC) provides results
equivalent to those obtained using the conventional 24-h
thermo-tolerant coliform procedure employed in the U.S. LSee
Section C.l.c] and can thus serve as a rapid, preliminary
read-out while awaiting the/24-h test results. Accordingly,
the M-7h FC test offers a means of detecting gross con-
tamination of potable water in emergency situations, per-
mitting more rapid responses and decisions by public health
or other officials.
*
(i) Vprinciples Applied to this Method. The M-7h FC
test uses a lightly buffered lactose-mannitol based medium
that contains an acid-sensitive indicator system (Reasoner,
et al., 1976). To lessen the adverse effect of 'higher
temperatures upon recovery of thermo-tolerant coliform
organisms which may have been stressed by disinfectants, the
incubation temperature has been reduced to 41.5 C (Van
Donsel, et al., 1969). During 7 h of submerged incubation,
thermo-toTerant coliforms ferment lactose to acid, causing
the color of the indicator to change from purple to yellow.
The thermo-tolerant coliform colonies typically appear
yellow against a purple background; all yellow colonies are
counted as thermo-tolerant coliforms.
(ii) Medium Preparation and Procedure. The medium may
be obtained from Difco or prepared in the laboratory pee
Table C.4.f-l]. It can be prepared in bulk quantity by
mixing the dry ingredients and storing in a clean jar in a
dry, warm place, or in a dessicating chamber to prevent
absorption of moisture from the air. The medium is_prepared
for use by adding the appropriate quantity of the mixed dry
-------
TABLE C.4.f-l
M 7-h FC
MEDIUM FORMULATIONC
27.0
Proteose peptone No. 3
Yeast extract
Lactose
D-Mannitol
Sodium chloride
Sodium lauryl sulfate
Sodium desoxycholate
Brom cresol purple, free acid
Phenol red, water soluble
Agar
Distilled water
Final pH
5.0 g
3.0 g
10.0 g
5.0 g
7.5 g
0.2 g
0.1 g
0.35 g
0.3. g
15.0 g
1,000 ml
7.3
All ingredients are manufactured by Difco
Labs., Inc., except sodium chloride, brom
cresol purple, and phenol red which are
manufactured by Fisher Scientific Co.
-------
271
medium to the Required volume of laboratory quality dis-
tilled water. f The medium is then heated to boiling to
dissolve the agar. The final pH of the prepared medium
should be adjusted to 7.3 +0.1 using sterile 0.1 N NaOH or
0.1 N HC1. Plates are prepared by dispensihg 5 to 6 ml of
the melted M-7h FC agar into small (50 x 12 mm or 60 x 1.5
mm) petri dishes and allowing the agar to solidify. The
prepared plates can be stored in the dark at 2 - 8°C for up
to two weeks.
Water samples are tested by passing an appropriate
volume (up to 100 ml) through a membrane filter [See Section
C.l.c]. Membrane filters are then placed on the agar plates
(taking precautions to avoid trapping any air bubbles), put
inside watertight plastic bags, inverted, and submerged in a
41.5°C water- bath for 7 h. Plates with 20 - 60 yellow
colonies are counted using a binocular wide-field dissecting
microscope. Colonies are recorded as numbers per 100 milli-
liters. ' .
(iii) Accuracy and Verification of Results. Results
from a study of thermo-tolerant coliform differentiation
(Geldreich, 1975) showed that 94.£ percent of .the yellow
colonies from the 7-h medium verified as thermo-tolerant
coliforms; when the same samples were analyzed by the 24-h
thermo-tolerant coliform method, 93.7 percent of the blue
colonies verified as thermo-tolerant coliform. These data
indicate that both media measured essentially the same
population of bacteria. Both the 24-h and the 7-h tests
underestimate the actual thermo-tolerant coliform density.
(iv) Summary. A 7-h membrane filtration method for
thermo-tolerant coliforms (M-7h FC) has been developed that
is suitable for determining gross contamination of potable
water in emergency situations. It gives results equivalent
to the conventional 24-h thermo-tolerant coliform test rou-
tinely used in North America. After 7 h of incubation,
thermo-tolerant-. coliforms appear yellow against a purple
background. About 94 percent of the yellow colonies can be
verified as thermo-tolerant coliforms.
g. Presence Absence Test. Routine analyses of drink-
ing water samples from municipal distribution systems tradi-
tionally have been designed to quickly signal the presence
and numbers of total coliforms. Initially, most labora-
tories used the most probable number (MPN) procedure for
-------
272
this function [See Section D.3.]. However, MPN analyses
are rather time-consuming, and the incubation period can
extend up to four days before numerical results are avail-
able .
The membrane filter (MF) technique [See Section D.2]
was introduced in the late fifties as an alternative to the
MPN procedure. This technique yields results approximately
equivalent to the MPN procedure, but it does so within a 24-
"h period and requires less preparation. In addition, numer-
ical results from the MF method have been more consistent
among samples from the same location.
Sudden, unexplainable instances of pollution occurring
in a distribution system and leading to disease outbreaks
are relatively rare. Aside from main breaks or equipment
failure in the treatment plant, water quality deterioration
in a distribution system is more likely attributable to a ~
gradual process of microbial degradation, engendering bac-
terial slimes, corrosion, or encrustation [See Sections F.2
to 3 and G.2 to 4]. Distribution systems may, in fact,
exert their own influences upon finished water, producing
conditions whereby slower growing debilitated bacteria or
pollution indicators of non-fecal origin may become estab-
lished. This microbial activity will produce sediments and
tubercles that reduce flow rates [See Section G.3], and will
create a need for increased chlorination to prevent regrowth
and to afford adequate disinfection.
The presence-absence (P-A) test is a simple, inexpen-
sive procedure for the rapid, qualitative detection of fecal
indicators (e.g., Escherichia coli), as well as indicators
of marginal water quality (i.e., slower growing, debilitated
bacteria and water bacteria which cause sliming and fouling
of distribution systems). The test, as such, would be
advantageous to laboratories with,limited time and resources.
This method, in use for over ten years by the Ontario Minis-
try of the Environment Laboratories (Canada), supplies
results satisfying the criteria enumerated above when used
on drinking water samples submitted daily or weekly. Since
90 percent of the samples are ordinarily free of pollution
indicators, there is no need to provide quantitative results
for each sample.
(i) Procedure for Conducting the P-A Test. Fifty or
100 ml of drinking water samples are added to double or
triple strength MacConkey (MacC) broth (respectively) con-
tained in P-A bottles with inverted durham tubes. Samples
-------
273
and medium are mixed so as not to permit any stratification
of the concentrated medium which would inhibit growth. The
P-A bottles are incubated at 35°G for four days and checked
every 24 h for acid production (with or without gas), as
evidenced by the color yellow. If, after three or four days
of incubation, only a slight acid reaction is apparent,
cultures should be checked for the presence of Staphylo-
coccus (see below).
Pollution indicator types are determined by inoculating
presumptive-positive cultures into confirmatory media,
diagrammed in Figure C.4.g-l and described previously (Clark,
1969; Clark and Vlassoff, 1973). Most laboratories will
find that the initial confirmatory tests provide sufficient
information for assessing the sanitary quality of the water.
However, further taxonomic tests may be applied to isolated
cultures if more definitive information is required.
Tests for confirming fecal and total coliforms are
easily performed. Only fecal coliforms produce gas in the
EC broth at 44.5°C; but both fecal and total coliforms
produce gas in the EC broth at 35°C; and both produce lactose
fermenting colonies on MacC agar. If lactose fermenting
colonies are evident on MacC agar, but gas is not seen in
either of the EC broth tubes, the organisms are anaerogenic
coliforms (debilitated coliforms which have lost the ability
to produce gas from lactose fermentation), or aeromonads.
Aeromonas can be readily distinguished from anerogenic
coliforms by streaking on a non-carbohydrate medium such as
Nutrient-Gelatin agar, and performing an oxidase test on 24
h.colonies. If a good growth of only non-lactose fermenters
appears on MacC agar, these colonies should be identified
further with taxonomic tests;
Gram-positive organisms, including Staphylococcus,
Micrococcus, and Bacillus sp. are frequently present when
gram-negative organisms are absent from any of the other
confirmatory media. When grown on Mannitol-Salt agar,
Staphylococcus produces convex, yellow colonies surrounded
by a yellow zone. Micrococcus types usually produce white
or colorless colonies with no evidence of mannitol utili-
zation. Some Bacillus species which produce mucoid, yellow
colonies may be distinguished with a Gram stain.
Although Pseudomonas aeruginosa grows vigorously in P-A
bottles and may dominate in mixed cultures, its presence may
go unnoticed because of its inability to ferment lactose. It
-------
274
will, however, be noticed if, for coliform confirmation, the
organism is streaked onto MacC agar plates. 1?. aeruginosa
can be detected and differentiated using Drake's broth
(Drake, 1966), incubated at 41.5°C. Each day for four days,
tubes are checked for any fluorescent growth. Cultures from
positive tubes are inoculated onto Skim Milk agar, or onto
Milk agar which is prepared in the laboratory according to
the Brown and Foster modification (American Public Health
Assoc., 1975), and incubated at 35°C for 24 h. P_. aeru-
ginosa hydrolizes casein, producing typical green colonies.'
Fecal streptococci, when present alone in P-A bottles,
produce acid with no gas, but the same is also true for
anaerogenic.coliforms or aeromonads. To confirm the presence
of fecal streptococci, cultures are inoculated into Ethyl
Violet Azide (EVA) broth and onto Enterococcus (Entero)
agar.
Clostridium perfringens reacts in either of two ways in
P-A bottles: it may produce abundant gas and acid in the
MacC broth, thus resembling a coliform reaction; or, it may
produce abundant gas, but confer a bleached, pale, greenish
tinge to the medium. A stormy fermentation in Skim Milk
broth is usually conclusive for C_. perfringens, but a doubt-
ful reaction should be checked with a Gram stain.
All media described in this section (except Drake's
broth) may be obtained commercially from Difco. Use of the
various confirmatory and taxonomic test media is not rigid.
All laboratories have their own favorite media and proce-
dures for confirming and identifying the pollution indi-
cators described above. Even Lauryl Tryptose or Lactose
broth may be substituted for MacC broth if a Brom-Cresol-
Purple indicator is added at the rate of 0.01 g per 1;
although Lauryl Tryptose broth is somewhat more inhibitory
than MacC broth, requiring sometimes an extra 24 h to develop
a presumptive-positive reaction.
(ii) Interpreting Quantitative and Qualitative Results.
Since most laboratories cannot run quantitative analyses for
all the indicator bacteria, they may achieve an acceptable
compromise by conducting a quantitative analysis for total
coliforms with the MF technique; and by testing for the
presence of all other indicator bacteria with the P-A method.
When the P-A method is employed in conjunction with the MF
coliform method, 50 ml sample volumes are used in each test.
The P-A procedure, when applied by itself, calls for 100 ml
-------
275
sample volumes. The following is a recommended procedure
for handling results:
1. If the MF plate has a sheen colony count of
greater than ten per 100 ml and the P-A bottle
shows an acid reaction with or without gas,
then the MF count is taken as the final result
and the P-A bottle is discarded.
2. If the MF plate produces no sheen colonies,
the P-A bottle should be incubated up to four
days before discarding. If acid with or without
gas occurs, confirmatory tests should ensue to
identify the indicator group present.
3. If the MF plate has a sheen colony count, but
the 24 h P-A bottle shows little or no reaction,
several sheen colonies should be picked off and
confirmed. (This type of result is frequently
presumptive for Aeromonas organisms.) The P-A
bottle should be kept for the remainder of the
incubation period in case later presumptive
results need confirming.
The above procedures will provide a quantitative result
for total coliforms and a qualitative result for all the other
indicator bacteria. At least 5 percent of the samples
showing no total coliforms by the MF technique will be
demonstrated by the P-A test to contain them. Thus, by
substituting the P-A test for :routine analyses and by
applying the MF total coliform test only to random portions
(15 - 25 percent) of sample blocks, quantitative data will
be available if, at any time, the P-A test results should
indicate pollution in part or all of the distribution system.
Moreover, since many water suppliers find no total coliforms
in 95 percent or more of water samples tested throughout the
year, the money expended to run quantitative analyses for
each sample, may not be justifiable, particularly when the
P-A method can provide more information at less cost.
(iii) Generic Distribution of Organisms Isolated from
Raw and Finished Water by the P-A Test. Table C.4.g-l lists
organisms recovered from raw and finished surface water
samples. Cultures isolated from presumptive positive P-A
bottles were identified or characterized according to the
scheme in Figure C.4.g-l. Usually 85 percent of these
presumptive-positive cultures from both raw and finished
-------
TABLE C.4.g-l
IDENTIFICATION AND RELATIVE FREQUENCY OF CULTURES FROM
RAW AND DRINKING WATER SAMPLES
Raw Water*
Drinking Water**
Identification Frequency
Escherichia coli 40%
Klebsiella pneumoniae 17%
Enterobacter cloacae '14%
Aeromonas sp. 11%
Enterobacter 'agglomera.ns 5%
Citrobacter freundii ' 5%
Enterobacter aerogenes 2%
Enterobacter hafniae 2%
Proteus sp., Providencia sp.
C. diversus, K. ozaenae,
Salmonella sp., Shigella sp., < 2%
Ungrouped cultures each
Identification Frequency
Enterobacter cloacae 26%
Escherichia coli 24%
Klebsiella pneumoniae 16%
Aeromonas sp. 14%
Enterobacter agglomerans 6%
Citrobacter freundii 5%
Enterobacter aerogenes 2%
Ungrouped cultures ' 2%
Proteus sp., Providencia sp.,
E. hafniae, K. ozaenae, -
Serratia sp., C. diversus, < 2%
Salmonella sp., Shigella sp., each
Yersinia sp.
* Number of cultures identified from raw water samples were 413.
** Number of cultures identified from drinking water samples were 1,376,
fo
»j
(Ti
-------
FIGURE C.4.cr-l
Schane for Identification of Cultures from P-A Test
Pr*««nc«-Abune« (P-A) Bottlt j
2X/3X MocC Brolh
[ COLIFORMS/AEROMONAS I
^^^^^^^^MH^J
PSEUOOMONAS
FECAL
STREPTOCOCCI
CLOSTRIOIUM
Skim
Milk
Tub*
P
/
A
C
0
N
F
I
R
M
A
T
0
R
Y
T
A
X
0
N
0
M
I
C
T
E
S
T
T
E
S
T
S
T
E
S
T
S
rs>
-------
278
water sources proved to be coliforms. Other indicators were
recovered less often, their presence being more a function
of water source and nature of pollution. Most, but not all,
of the colonies isolated from MacC agar plates were lactose
fermenters. Those few which didn't ferment lactose, but
which predominated on a culture plate, were picked along
with the others for identification.
The use of Enterotubes and Oxyferm tubes (Hoffman-La
Roche, Ltd.), according to the scheme in Figure C.4.g-l,
afforded a more extensive determination of coliform genera.
E. coli was most often isolated from raw water samples, but
d~eclined the most following water treatment. Alternately,
isolation frequencies for other coliforms varied little from
raw to finished water sources. While Enterobacter cloacae
was more frequently isolated following water treatment,
other Enterobacteriaceae were isolated as often from one
source as from another.
Though IS. coli was most often isolated from raw water
samples, it ₯howed the greatest reductions following water
treatment. .E. coli [See Section C.l.c] has been recog-
nized as the indicator most closely associated with and
representative of fecal pollution in drinking water; hence,
its presence signifies the potential presence of pathogenic
organisms such as Salmonella [See Section C.l.c]. None-
theless, the results in Table C.4.g-1 demonstrate that
though the presence of IS. coli indicates fecal pollution,
its absence does not necessarily indicate water safety after
treatment. In contrast, any of the other coliform genera
can be used to measure water treatment efficiency, but not
to measure fecal pollution.
Klebsiella pneumoniae [See Section C.2.d] has also been
isolated on a number of occasions from drinking water samples
and is sometimes the organism responsible for positive fecal
coliform tests. Its presence was therefore anticipated,
but not in the relatively high numbers and unvaried iso-
lation frequencies noted between raw and drinking water
samples. From these results, and in view of the evidence
that K. pneumoniae is more often associated with pulp and
paper mill and other industrial effluents, its use as an
indicator of fecal pollution is not recommended.
Recovery of Aeromonas [See Section C.2.h], also identi-
fied in these tests, signifies a deterioration in water
treatment or along distribution lines, and it should, there-
fore, be included as an indicator of water treatment effi-
ciency. However, this organism is often indistinguishable
-------
279
from coliform colonies when the MF technique is used for
analysis. Unless Aeromonas is differentiated, the water
will be considered contaminated with fecal pollution, whereas
slime deposits in water mains may instead be supporting
entrenched Aeromonas populations. This organism is asso-
ciated with nutrient-rich waters which receive heavy soil
run-off's and it is also found as the predominant organism
in certain groundwater supplies.
These generic tabulations should not be interpreted to
represent all drinking and raw water sources; nor should
they be taken to reflect year-round populations. More
pollution indicators are detected in summer and fall months
when distribution systems carry warmer water which may
afford opportunities for regrowth. At other times, pollu-
tion indicators will be detected more frequently because of
periodic treatment deficiencies at an individual facility.
In the event of any such abnormalities, waterworks operators
should be alerted so that they may step up disinfection.
(iv) Understanding P-A Test Results. P. aeruginosa [See
Sections B.l.a(ix) and C.2.c] and £>. aureus are both poten-
tial pathogens and their detection is cause for concern,
especially since they may be established along portions of
the distribution system. Most healthy individuals may not
be affected by ingesting small numbers in drinking water,
but there may be those who are susceptible to infections by
these bacteria.
When fecal streptococci are isolated in association
with total or fecal coliforms, the water is probably con-
taminated with fecal material. However, isolation of fecal
streptococci by themselves may indicate marginal treatment
of the water supply. Because many species in this group
tend to be host specific, proper interpretation may call for!
species identification.
Staphylococcus epidermidis and Micrococcus species were
isolated (Clark, unpublished data) when no other pollution
indicators were detected in water samples. They were found
to be protected from disinfection by slime deposits in water
mains and reservoirs. Their detection may have no sanitary
significance, but rather, should signal that some portion
of the,distribution system needs attention due to deterio-
rating conditions.
Total coliforms are employed in North America as the
primary indicator of finished drinking water quality [See
Sections D.2.g and D.S.f]. However, it is useful to make a
-------
280.
distinction between aerogenic and anaerogenic coliforms when
discussing water of deteriorating quality. Among the coli-
form genera, variability exists within each genus, ranging
from no fermentation of lactose to fermentation of lactose
at 44.5°C with the production of gas. Depending on the
length of time and degree of treatment that intestinal
organisms have had in water, their physiological reactions
could be altered so that they no longer respond exactly as
typical strains. The detection of aerogenic coliforms
sometimes indicates that the water supply received no treat-
ment, whereas the detection and differentiation of anaero-
genic coliforms may indicate unsatisfactory water treatment
or insufficient chlorine residuals. The significance of
anaerogenic coliforms has not been fully investigated because
some MPN procedures do not use a pH indicator and thus, rely
mainly on gas production to detect coliforms. MF techniques
fail to distinguish non-lactose fermenting variants of the
coliform genera, and do not differentiate between aerogenic
and anaerogenic variants.
Clostridium perfringens has been shown to occur more
frequently in distribution systems during winter and spring
when seasonally colder waters reduce the efficiency of
chlorine disinfection and aid survival of C!. perfringens in
the distribution system. This organism has been shown to
build up in sand filters whereupon it may detach and be
passed along through the distribution system.
When good quality water is delivered through well
maintained and disinfected distribution systems, water
samples taken from anywhere along the distribution route
will be free of the pollution indicators described here.
However, if good quality water passes through pipelines and
reservoirs which have been allowed to deteriorate, pollution
indicators will continue to appear in water samples; albeit,
they may be of genera quite different from those usually
associated with fecal sources. The point to be emphasized
is that because of certain ecological factors in distri-
bution systems, particular types of organisms, whether gram-
positive or -negative, may become the resident flora. By
using simple, rapid, and reliable test procedures to deter-
mine which pollution indicators are becoming established,
remedial action can be, taken before a crisis develops (Ontario
Ministry of the Environment, 1976).
(v) Summary. The P-A test provides a simple, ef-
fective means of assessing water quality for the routine
surveillance of distribution systems. A full range of
-------
281
taxonomic tests would not be required on a routine basis, as
the confirmation tests specified would be sufficient to
assign most of the organisms to one or more indicator
groups. Differentiation of the coliform genera, other than
E. coli, would not likely be very helpful in determining
whether the water was safe to drink.. However, if coliforms,
aeromonads, or any of the other indicators are detected,
with any degree of frequency, in consecutive samples or
sample blocks, this signifies problems within the distri-
bution system. Either the walls of water mains are not
being adequately protected from slime deposits, or factors
such as warmer water and available nutrients are inducing
regrowth of such organisms within the distribution system.
h. Membrane Filter Procedure for U.S. Standard Plate
Count. The method presented here produces results closely
parallel to those achieved with the standard plate count
(SPC) used in the U.S. [See Section C.I.a]. The membrane
SPC (M-SPC) method is offered as a substitute for the SPC
and, as such, would be used (in conjunction with standard
coliform analyses) to measure changes in the bacteriologic
quality of potable water passing through water supply
distribution systems or point-of-use treatment devices.
(i) Advantages of Using Membrane Filters. The M-SPC
procedure would have a marked advantage for the enumeration
of bacteria in potcble waters by allowing the examination
of sample volumes larger than one milliliter, thereby
drastically increasing the statistical acceptability of
results heretofore obtained with the SPC method. A membrane
filter procedure would have the additional benefits of
savings in time, incubator space, equipment, materials, and
labor (.Clark, ejt aJL., 1951). Also, colony pigmentation
develops more readily on the membrane filter, and individual
colonies may be easily picked- for further identification.
The only factors potentially restrictive of this procedure
are sample turbidity and spreading of colonies on the filter
surface. Colony density on the filter can be controlled by
adjusting sample volumes appropriately.
(ii) Medium Preparation and Procedure. The M-SPC
medium utilizes a general enriched formulation consisting
of three conventional nutrients: peptone, gelatin, and
glycerol. It may be obtained, in dehydrated form, from
Difco (glycerol must foe ordered separately), or prepared in
the laboratory [See Table C.4.h-l]. Rehydrate the gel-
atin, peptone, and agar in distilled water« then heat to
-------
TABLE C.4.h-l
M-SPC MEDIUM FORMULATION
282
Gelatin
Peptone
Agar
Glycerol
Distilled water
25..0 g
20.0 g
15.0 g
10.0 ml
1000 ml
-------
boiling. Cool to 45 - 50°C and adjust to pH 7.0 with 1 N
NaOH. Then, add glycerol and autoclave for 5 min at 121°C.
Pour medium into petri plates with tight fitting lids [See
Section C.I.b], and store between 2 and 10°C.
An appropriate volume of water sample (100 - 500 ml or
more) is passed through a membrane filter [See Section
. C.l.b], the filter is then placed (talcing precautions to
avoid trapping any air bubbles) on an M-SPC agar plate.
Plates with filters are incubated at 35°C for 48 to 72 h to
achieve maximum colony size. An incubation temperature of
35°C is selected, to conform with procedures specified in
the U.S., for determining a "standard" plate count in potable
water (American Public Health Assoc., 1976). Membranes
selected for counting should contain between 20 and 200
bacterial colonies. All colonies are counted with the aid
of a wide-field dissecting microscope, 10 - 15X magnifi-
cation, and recorded as numbers per one milliliter.
(iii) Agreement Between the SPC and M-SPC Methods.
The mean SPC/M-SPC ratio for 164 drinking water samples was
1.099, indicating that the two procedures very closely
approximate one another (Taylor and Geldreich, in press).
In'point of fact, the geometric mean ratio was 0.501, re-
vealing the inclusion of a few unusually high ratios in the
field data. That is, 120 of the individual samples regis-
tered ratios of less than one, indicating that the M-SPC
procedure provided higher counts than the SPC procedure
approximately 75 percent of the time. It remains to be seen
whether results of the two methods will correlate as closely
when applied to a greater number and variety of field samples.
(iv) Summary. A membrane filtration method has been
developed that gives results closely parallel to those
achieved with the U.S. standard plate count. The method is
designed to replace the pour plate technique, and as such,
to measure changes in the distribution system or in passage
through point-of-use water treatment devices. In addition
to increased volumes of water that can be tested, the method
allows savings of time, space, and equipment. Colony
pigmentation develops better on the membrane filter} and the
colonies may be easily picked for identification.
i. Epifluorescence Total Cell Count. The direct
counting of bacteria with a microscope is a technique with a
-------
284
long "history. As traditionally practiced, it has entailed
concentrating bacteria by filtering 2 ml of sample water
through a membrane. Bacteria trapped on the filter are
stained. Then, the filter is dried and rinsed; finally,
what are supposed to be bacteria are counted with the aid of
transmitted light. Direct counting has never been popular
because most bacteria in environmental waters are so small
that identifying them from small dots required an act of
faith. However, recent improvements in both membrane filters
and microscopes have resulted in easier and more repro-
ducible methods of direct counting. The epifluorescence
direct count method is a rapid and simple way to determine
total numbers of bacteria in water [See Sections C.3.b and
C.4.b].
By employing direct counting methods, investigators
have observed high numbers of bacteria in all water sources
sampled: the-Great Lakes and coastal oceanic waters may
have about 10 bacteria per ml (Zimmerman and Meyer-Reil,
1924); mountain lakes or open ocean waters contain about
10 cells per ml; even distilled water^ stored for a month in
the laboratory, may have as many as 10 cells per ml (Daley
and Hobbie, 1975). Sewage polluted water has been found to
contain as many as 10 cells per ml (Straskraba and Stras^
krabova, 1969).
Bacteria will grow in any suitable water, including
most distilled water. Rather high numbers of bacteria, over
10 per ml, have been found in drinking water (although no
controls were run for presence of residual chlorine or
effects from storage) [See Section G.6]. Bacteria are
capable of growth during distribution where finished water
does not contain a sufficient chlorine residual. Though
generally not hazardous to health, these organisms may
cause a number of technical and nuisance problems [See
Sections C.4.g, F, and G.2 to 4] and may not be enumerated
efficiently using plate count agar.
Most bacteria seen with epifluorescence do not grow
well under laboratory conditions, although up to 90 percent
of these same organisms are, according to autoradiographic
studies, actively taking up substrates (such as sugars and
amino acids) from solution (Hoppe, 1976). Therefore, direct
counting using epifluorescence can remedy deficiencies in
total bacterial enumerations, though it is not specific for
pathogens or indicator organisms, and should only be used to
supplement or replace non-specific plate count techniques.
-------
285
(i) Materials for Direct Counting with Fluorescence
Microscopy. A variety of dyes have been successfully used
for staining samples prior to counting; they all provide
adequate contrast, but none is specific for bacteria.
Fluorescein isothiocynate (FITC) has been used (Fliermans
and Schmidt, 1975), as has euchrysine (Jones, 1974), but
acridine orange (AO) is the most widely, accepted (Hobbie,
e_t al., 1977). The usual final concentration is 0.01 percent
AO. :;
Polycarbonate membranes, such as those produced by
Nuclepore, have a very flat surface that retains all the
bacteria on top of the filter. Although cellulose filters
also retain all of the bacteria, many are trapped inside the
filter where they cannot be counted. The polycarbonate
membranes used in this procedure are 25 mm in diameter and
have straight-through cylindrical pores with a diameter of
0.2 urn. A 0.1 urn porosity filter can also be used but at
this diameter, the filtering rate is greatly reduced. The
filter is stained before use by soaking 2 to 24 h in a
solution of 2 g irgalen black (a competitive stain) in one
liter of 2 percent acetic acid (Hobbie, et aJL. , 1977).
A variety of microscopes have been used with success.
Most manufacturers now make an epifluorescence attachment
and have light filters especially for AO (e.g., Zeiss,
Leitz, American Optical, Olympus, Nikon). A quartz halogen
lamp is satisfactory, but a mercury lamp may give a higher,
intensity if necessary.
(ii) Method for Direct Counting with Fluorescence
Microscopy. A few drops of a fluorescent dye are added to a
water sample. It is convenient to place a 2.0-ml subsample
in a small test tube and to add 0.2 ml of 0.1 percent AO (in
distilled water). After the sample is held for one minute,
it is filtered through a polycarbonate membrane. Filtering
less than 2 ml does not permit an even distribution over the
filter surface; therefore, dilution water may be added so
that the total quantity filtered is 2 ml. When bacterial
numbers are low, larger sample volumes can be filtered.
Placing the polycarbonate membrane on top of a cellulose
membrane filter during filtration helps to produce a more
uniform distribution of cells.
While still damp, the membrane is placed on a glass
slidej and immersion oil is added- Thenx, a cover slip is put
on top. and more immersion oil added. Bacteria are counted,
-------
j 286
I
using epi-illumination, within jone hour or several hours
(depending on type of water sampled) after preparation.
The epi-illuminated microscope allows a wet filter to
be counted because the illuminating light comes from above;
it also provides very intense light because the objective
lens of the microscope acts as the condenser. As a rough
guide, 2 ml of most raw water will yield about 90 bacteria
per field when a 90X or 100X objective is used. Usually,
one quarter of each of ten randomly chosen fields is counted,
for a total of 200 to 400 cellsL Because bacteria will
readily grow in every kind of water, including distilled
water, the numbers found by this technique are likely to be
large and may reach 10 cells per ml.
With this method, bacteria;are seen as fluorescent
spots against a dark background. In some cases, dividing
cells, rods, or spirilla can beiseen. However, indicators
or pathogens cannot be separated from the abundant naturally
occurring bacteria. In most cases, bacteria can be dif-
ferentiated from debris because!bacteria take up more stain
(when kept at a relatively low concentration, AO interacts
with the nuclear material of bacteria) and appear more
regular in shape. That these small brightly fluorescing
foci are bacteria was verified by the good agreement of
their numbers with scanning eleqtron microscope counts
(Bowden, 1977) and with indirect measures of bacteria such
as the lipopolysaccharide content of the sample (Watson, et
al., 1977)(see Section c.3.a).
The water sample (and the Ao too) can be preserved with
formaldehyde at a final concentration of 2 to 4 percent and
stored for at least two months.I Permanent slides can be
produced with the Zimmermann (1977) technique.
(iii) Potential Applications for Use with Drinking
Water. Epifluorescent direct counts have not been used on
finished drinking water. However, this method offers poten-
tial for use in tracing breaks in supply systems, or in
monitoring the overall quality of water supply sources.
Pollution in a water supply may jcause an increase in total
cell numbers of 10 to 100 times.;
F '
^By conducting a series of studies of bacterial numbers
at different points in the treatment and distribution system,
waterworks microbiologists could determine general levels
and employ epifluorescence counting for quick assessment of
the quality of water. Likewise,] this technique could be
-------
287
used for rapid tracing of breaks in lines and other similar
emergencies. Further evaluation of this type of application
is recommended. ,
Another possible approach could be based, not on total
numbers of bacteria, but on the size of individual cells.
Most bacteria in unpolluted waters are tiny, 0.2 to 0.4 um
in diameter, whereas bacteria in extremely productive or
organically polluted waters grow to diameters of up to 1 um.
Whether a richer nutrient source brings about increased
sizes in the initial flora, or whether different organisms
predominate under polluted conditions, is not yet known.
(iv) Summary. Bacteria in water can rapidly be counted
(within 15 minutes of sampling) by staining with a fluores-
cent dye such as acridine orange, concentrating by filtra-
tion through a polycarbonate membrane, and viewing with an
epifluorescent microscope. Because the technique is not
specific for indicator or pathogenic organisms, it can be
used only to supplement or replace non-specific plate count
methods used for investigating the quality of finished
water. It may be valuable for monitoring the quality of
supply waters or for obtaining rapid analyses of line breaks
or other emergency situations.
5. Summary
The indicator systems in this section are extremely
varied, ranging from fecal indicators and pathogens to
biomass estimates, including new, faster, and automated
methods to increase speed and sampling efficiency. It is
clear that not all of these organisms or techniques have the
same health and sanitary significance, nor are they all
applicable .to the same types of waters.
It seems evident that the "classical" fecal indicator
organisms have served rather well in preventing transmission
of human disease by drinking water and shall continue to be
used for some time to come. Total coliform enumeration, for
example, is of limited value in the determination of raw
water quality, but is a sensitive method for determining
finished water quality and treatment adequacy. Total coli-
forms in drinking water do not necessarily reflect fecal
contamination, but their presence should alert the respon-
sible agency to a possible problem, and action can then be
taken to find the cause and correct it as necessary. Thermo-
tolerant coliforms (fecal coliforms) or Escherichia coli and
Clostridium perfringens, on the other hand, can be employed
for both raw and finished drinking water. The presence of
-------
288
any of these organisms in finished water should be an im-
mediate cause for concern and should be followed by resam-
pling and corrective measures. Their level in the raw water
is generally considered valuable in determining the type and
degree of treatment required. Obyiously, the greater the
degree of contamination of the raw water, the more safe-
guards (perhaps both intensified treatment and stepped-up
testing of the finished water) will be required to protect
the consumer. Clostridium perfringens is a particularly
hardy organism because of its ability to form disinfectant-
resistant spores; it may thus bejof value in determining
the adequacy of water treatment and filtration. Fecal
streptococci are of limited value:in monitoring finished
water quality, except when gross contamination occurs, but
are used extensively for raw water. The judicious use of
this indicator can greatly assist!in determining the source,
distance, and time of pollution o£ the raw water, when the
studies are done in conjunction with other indicators such
as thermo-tolerant coliforms. i
i
Since this group of fecal pollution indicators is so
important and routinely used, anyjprocedural modification
that reduces cost, speeds up the analysis, or reduces the
labor involved is considered an improvement. For this
reason, the 7-h membrane filtration fecal (thermo-tolerant ) colifonr.
test appears to be a valuable method, especially for emer-
gency use^such as when finished drinking water is found to
be polluted as a result of main breaks or cross contami-
nation with sewage pipes. The presence-absence test is
applicable to finished water only; it quickly and cheaply
screens out all coliform-free samples, so that the positive
samples can be given more detailed analysis. Both the
automated sampling and plating methods and the automated
plate counting methods could be applied in specific settings
for these organisms. At present,[they are not widely used,
but further improvements in automation and data handling, as
well as the increased costs for personnel to carry out
routine analyses, will eventually[make them more attractive.
Other methods presented here, that show potential for thermo-
tolerant coliform indicators, are radioisotope methods,
impedance methods, and enzymatic methods. Each of these has
been tested to a limited extent, to determine levels of-
coliforms or Escherichia coli andimay have more widespread
application in the future. They represent the application
of physical and chemical technology to the field of drinking
water microbiology, an endeavor that must be encouraged*
The analysis of fecal sterols is emerging as a promis-
ing indicator that eliminates some of the drawbacks of fecal
-------
289
bacteria indicators. The fecal sterols display many adsorp-
tion and elimination properties similar to viruses and
bacteria, but additionally, are not affected by most disin-
fection processes, by elevated temperatures, nor by toxic
effluents. Further work is required to determine their
ultimate applicability.
Certain other organisms are considered here, as indi-
cators, because of their applicability in certain problem
areas and their properties as pathogens or opportunistic
pathogens of susceptible humans. Pseudomonas, Klebsiella,
and Aeromonas are common opportunistic pathogens in water
and are able to multiply under relatively .low nutrient
conditions, sometimes causing problems in the distribution
system. It is worthwhile to monitor for these organisms,
occasionally, or when there is a known problem. Their
health significance in low numbers in finished water is not
well established, but their presence indicates potential
problems in the treatment or distribution system and should
stimulate more intensive sampling and investigation. The
presence of Klebsiella in raw waters is often the result of
its growth in carbohydrate-enriched waters, and its effect
as a positive bias on total coliform and thermo-tolerant
coliform counts is well known. Vibrios are becoming in-
creasingly important as pathogens in the marine and estuarine
environment, as is Candida albicans in the fresh-water
environment; both warrant further study in reference to
drinking water quality. Bifidobacteria show extremely^good
promise as a fecal indicator and are receiving increasing
attention: the problem at this time is the lack of adequate
methods for measuring them.
In addition to the presence of fecal material and human
pathogens in water, there is concern for the general bac-
terial population in tapwater. General bacteria are usually
present in much larger numbers, in both raw and finished
waters, than organisms of health and sanitary significance,
and they may have important consequences. First, when
present at levels greater than 500 per ml, they greatly
diminish the ability to detect total coliforms in the water,
thereby adversely affecting the accuracy of that method.
Second, increases, particularly sudden ones, in total bac-
terial colony counts may signal a change in water treatment
efficacy, distribution problems, or other sources of con-
tamination even before the traditional fecal indicators can
signal such a change; They are also useful to determine
bacterial growth in point-of-use water treatment devices.
Plate count methods vary as to procedure, media, and incu-
bation temperature; a new membrane filter method is described
-------
. [ 290
that is more versatile and economical than some of the pour-
plate or spread-plate methods. Tyo other possible .mechanisms
exist to determine numbers of bacteria in water: direct
counting and biochemical measurements. The epifluorescence
method has the distinct advantage[of being very fast and can
be particularly useful in emergency situations. Among the
biochemical methods, measurement of ATP is probably the one
most often used, although concentration methods are required
for low bacterial levels in finished drinking water. The
Limulus lysate assay is much more|sensitive, and can easily
detect the low levels of lipopolysaccharides derived from
the cell walls of gram-negative bacteria in tapwater.
Further research is required to ascertain how these methods
relate to the more common plate count methods.
The problem of an adequate indicator system to signal
the presence of human enteric viruses in water* is one that
has not been resolved. Bacteriological indicators of fecal
pollution are not always adequate for this purpose because
of the greater ability of viruses to resist environmental
extremes and disinfection processes. Concentration, elution,
and cell culture techniques do not exist for all of the
viruses known to be transmitted iri water, so it is impos-
sible to detect them all even if it were economically
feasible to do so. Vaccine polio^iruses are a logical
enteric virus indicator because of their relative abundance
in waters in areas where trivalent oral polio vaccine is
used. Present data leave some doubt, however, as to the
adequacy of polioviruses for this[purpose, for they are not
detected in waters with the frequency that would be pre-
dicted on the basis of the numbers of persons excreting
them. Coliphages, on the other hand, are found in much
larger numbers and are not dependent upon the use of vac-
cines in the population. The idea'l host for this method,
and the actual relationship between the occurrence and
survival of these bacterial viruses and human viruses is
still open to question. Thus, until the basic question of
virus indicators is settled, routine analysis of drinking
water for fecal pollution indicator bacteria and the main-
tenance of sound water treatment and disinfection prac-
tices must be relied upon to ensure the virological quality
of the water. !
. j
The list of indicator systems considered here is not
exhaustive; there are many others |that could have been
included. Some of those selected,! however, are among those
most firmly established, others arje among the most promising,
and still others are among the most interesting. It becomes
-------
291
evident very early that even though the field of water
microbiology is by no means new, there is still no best
indicator system for every situation, nor is there ever
likely to be one. Each situation may require one or several
microbiological tests that can be selected from the ever-
changing array of indicator systems.
6. Recommendations
1. Colony counts should be used in conjunction
with total coliforms to assess finished water
quality and to monitor for any changes in water
quality along the distribution system. A limit
of 500 colonies per ml at 35°C should be con-
sidered for finished water, to reduce problems
with detection of total coliforms due to inter-
ference from non-coliforms. ^
2. Membrane filtration media for total coliform
enumeration should be developed to replace those
requiring the dye, '.basic fuchsin, in view of
the impending shortage of this dye.
3. Membrane filter methods for detecting Esche-
richia coli should be improved to permit faster
and simpler enumerations with disinfected
water.
4. Better selective media and methods should be
developed for recovery of bifidobacteria in raw
and finished water. Bifidobacterium should be
studied, extensively, in its role as a fecal
indicator.
5. Studies should be carried out to determine
sources of Candida albicans (e.g., humans,
birds, animals, etc.). Also needed are studies
to determine the survival rates of C_. albicans
in water treatment and disinfection processes.
In addition, the relationship, if any, between
densities of C. albicans and classical indi-
cators should be determined. Finally, the role
of sediments in concentrating C. albicans in
drinking water collection basins and other
receiving waters should be evaluated.
-------
292
6. The significance to human health of vibrios
present in raw source waters should be deter-
mined . ;
F
7. Attention should be giveti to the use of immo-
bilized enzymes and other rapid methods for
determining adenosine triphosphate (ATP) levels.
8. Studies should be carried out to determine the
mechanism by which coprostanol is degraded in
water and what the effects are from environ-
mental factors. Coprost[anol persistence and
baseline levels in surface waters and sediments
should be determined.
I
9. The impedance method should be further refined
to make it sufficiently sensitive and fast for
on-line analysis of finished water.
* t
10. Automated sampling and plating devices should
be improved to enable detection of anaerobic
sporeforming organisms. | Concentrators for
these devices need to be| developed so that
assays may be sensitive enough for use with
finished water. '
-------
293
D. TESTING AND STANDARDS
Tests to assess the microbiological quality of drinking
water rely largely on the recovery or enumeration of coliform
bacteria, thermo-tolerant coliform bacteria, or Escherichia
coli and, in some instances, aerobic heterotrophic bacteria.
Formal standards of drinking water quality usually are
based on numerical values derived from specified test pro-
cedures. While it may be assumed that coliform or thermo-
tolerant coliform bacteria are identical and unaffected by
national boundaries, differences in test procedures may
produce differences in results because, in part, the defini-
tions are functional and depend on the analytical procedures
used. To evaluate these differences, a survey was conducted
among the participant nations to identify analytical techniques
and water quality standards. To simplify the data gathering
and reporting and to make possible direct comparisons in
definitions and techniques, the information has been summarized
in a series of tables [Tables D-l to D-6]. Although these
tables are largely'self-explanatory, some general comments
are necessary.
The questionnaire was prepared in English and was keyed
to the practices used in the U.S. as described in Standard
Methods (American Public Health Association, 1976). This
led to some difficulties of interpreting the questions and
may have resulted in inappropriate responses. One major
area in which this problem consistently occurred was that of
quality assurance. The term quality assurance usually is
understood as an assessment, statistical or otherwise, of
the quality of the product. What was not clearly understood
was the nature of the product. Thus, some respondents dealt
with water quality as indicated by standards, while others
responded in terms of quality control within the laboratory.
Although it was intended that this latter subject be addressed,
it frequently was not. Irrespective of the information
provided herein, it is clear that control of personnel,
procedures, materials, etc. in the laboratory, is of paramount
importance, but often is not dealt with officially. Excellent
-------
294
guides to quality assurance practices in health laboratories
in general (Inhorn, 1978) and in drinking water laboratories
(Environmental Protection Agency, 1978) are available.
An additional difficulty lies in the occasional absence
of a formal national methodology or in the failure of some
laboratories to conform to the national methodology. This
problem is compounded by the continuing evolution of laws
and regulations so that changes in procedures are constantly
underway. Hopefully, the tables summarize the officially
accepted techniques whether or not[they are in current use.
1- Sampling and Sample Storage
Table D-l summarizes the data obtained for sampling
frequency, sample size, and sample'storage. Generally,
sampling of both raw and treated water is I not required (except in*
the Netherlands, Norway, "and Sweden);and primary control of
bacteriological quality of water is attempted through col-
lecting and analyzing samples from (distribution systems.
Typically, the required frequency of sampling is a function
of the population served or the volume of water distributed
by a water system. The required sample size varies from 50
to 800 ml with the average being about 200 ml. Although many
European countries require refrigeration of water samples
not analyzed within 3 to 4 h, this is not mandatory
Canada unless sample transit times exceed 24 h, nor
U.S., where ambient water temperature is acceptable provided
no sample awaiting analysis exceeds a 30 h limit. Replies to
tne questionnaire indicated a consensus of opinion that a
temperature above freezing, but less than 10°C.probably is
adequate. Despite the recognition ;that holding samples,
even though they may be chilled.may affect the bacterio-
logical content of the samples [See! Sections E.I and G.6],
storage of up to 36 h (Sweden) is permitted. Problems of
sample transport clearly influence the maximum holding time
and every effort must be.made to keep the time to a minimum.
in
nor in the
2.
I
Total Coliform Testing, Membrane Filter Technique
Table D-2 summarizes the data Obtained with respect to
the definition of total coliform bacteria and the analytical
details used for the membrane filter technique. Two types
of definitions are used: one is based on specific phy-
siological and morphological characteristics of the organisms
(aerobic or facultatively anaerobic, rod-shaped, gram-nega-
tive, nonsporeforming); the other definition is based di-
rectly on the analytical technique ijised (organisms producing
a typical sheen on a defined medium). Some definitions,
-------
TABLE D-l
SAMPLING AND SAMPLE STORAGE, DRINKING WATER ANALYSIS
1. Frequency of required
sampling
a. Raw water
Canada
As required
by control
agency
Denmark FRG
As required1 Voluntarily.
by producer
France Greece Israel
As required1 Not specified1 As ^
required
Italy
b. Treated water
See c.
See c.
See c.
See c.
Not specified As ,
required
c. Distribution
system
d. Other
2. Sample size, ml
At least 4x/
month, depend-
ing on popula-
tion served
(Table & figure
available)
_> 100
At least lx/
year, depend-
ing on popu-
lation
served
_> 500
Undi s infected :
1/30,000 nT
distributed;
Disinfected:
1/15,000 mj
distributed
about 2501
At least 3x/
year; up to
Ix/day for
large
supplies
_> 350
At least lx/ As
month, depend- required
ing on popula-
tion served
(Table avail-
able)
200 100, 200
200
to
Ul
_->
-------
TABLE D-l -- Continued
Netherlands
Norway ,
Spain
Sweden
UK
US
1. Frequency of required
sampling
a. Raw water
b. Treated water
lx/4 weeks
Groundwater
Ix/week;
surface water
Ix/day
c. pistribution
system
d. Other
2. Sample size, ml
Ix/week
250-500
As required
Large systems
(> 10,000)
wk; medium
terns (5,000-
10,000) 2x/w]
small systems
(1,000 - 5,000)
Ix/wk, (< 1,000)
Ix/month
Same as for
treated water
Large sys- Groundwater
terns Ix/day? Ix/year; sur-
small sys- face water
terns Ix/month 4x/year
Large systems
(> 10,000)
l-2x/week
ems See a.
3x/
sys-
0-
/wk;
Large systems
OL 4,000) lx/
week; small
systems lx/
month
Large systems N<
5x/week; small
systems lx/
week at least
See a.
Not required
2 x 250
200-300
800
At least lx/
week, depend-
ing on popula-
tion served
(Table avail-
able)
228
Not required
Not required
At least lx/3
months, depend-
ing on popula-
tion served
(Table avail-
able)
MF 100
MPN4 5 x 10 or
5 x 100
tO
U9
in
-I
W
-------
TABLE D-l Continued
, , _ , - -~ ~ -
r.
3. Maximum storage time,
hours
4. Storage temperature,
«c
Canada Denmark
24 24
Refrigerate 0-5
if stored
> 24 hours
PRG France Greece
3, if refri- 8 12
gerated
24-30 h
about 4 4-6 about 4
Israel2' Italy
3, 0.5 6
ambient, 4
4-10
M
CTl
>
-------
TABLE D-l -- Continued
3. Maximum storage time, 24
hours
4. Storage temperature, 0-4
°C
3, if refri-
gerated 30
2-10
2-4
4 h, if refri-
gerated at <
10°C 36 h
< 10
30
ambient 0-10
MPN .. Most probable number or multiple tube dilution
procedure.
ro
vo
a\
I
tD
-------
TABLE D-2
TOTAL COLIFORM TESTING, MEMBRANE FILTER TECHNIQUE, DRINKING WATER ANALYSIS
Canada
1. Total coliform Aerobic and
bacteria, define facultative
anaerobic,
gram-negative,
non-sporeform-
ing cyto-
chrome-oxidase
negative, rod-
shaped bacteria
producing
colony with
golden green
metallic sheen
in 24 hours on
Endo-type
medium contain-
ing lactose
Denmark
Bacteria pro-
ducing yellow
colonies at
37°C & bio-
chemically &
morphologi-
cally similar
to E. coli
FRG
Gram-negative ,
cytochrome-
oxidase nega-
tive, non-
sporeforming
rod-shaped bac-
teria producing
acid & gas
from lactose
or producing
red colonies
on Endo agar
after 24 h
at 37°C
France
Ferment lac-
tose with
formation of
gas (See ISO
definition)
Greece
Not legally
specified
(See defini-
tion in
Standard
Methods,
American
Public Health
Assoc., 1976)
Israel
Not legally
specified
(See defini-
tion in
Standard
Methods ,
American
Public Health
Assoc., 1976)
Italy
Not
appli-
cable
2. Samples ,
a. Volume filtered
filter, ml
b. Number of
replicated/sample
c. Are diluted
samples analyzed
> 100
100
100
100 or 50
100, 50
Not required Not required Not required Not required Not required
Not routinely Not
routinely .
Routinely,
if neces-
sary
Not routinely Not routinely
100
Not required
Not routinely
-------
TABLE D-2 Continued
Netherlands
1. Total coliform Facultative
bacteria, define anaerobic,
gram-negative /
non-sporeform-
ing rod, produc-
ing acid and
gas in lactose
medium at 37"C
in 48 hours
Norway
Dark red
colonies with
or without
fuchsin sheen
Spain
Not legally
specified
(dark colo-
nies with
metallic
sheen)
-
Sweden UK
Not Gram-negative ,
applicable oxidase nega-
tive non-spor-
ing rods grow-
ing aerobically
on agar contain-
ing bile salts
and ferment-
ing lactose
within 48
hours at 37°C
producing acid
and qas
US
Organisms
producing
golden green
metallic sheen
2. Samples
a
b.
Volume filtered/
filter, ml
100
100
100
Number of Not required
replicated/sample
Are diluted
samples analyzed
Not routinely
Not routinely Yes, if
highly con-
taminated
50 or 100
100
Not required Not required
Yes, if
highly con-
taminated
Not routinely
to
to
-o
to
-------
TABLE D-2 Continued
. '
. _
.'
Canada
3. Dilution water, if
used, specify formula
4. Membrane filter
a. Filter 47
diameter, um
b. Pore 0-45
diameter, um
c. sterilization
Autoclave - time 1.0 min at
& temperature 121*C
__
Denmark
Phosphate
buffer (3 g
K,HP04 & 1 g
A7
0.45
Purchase
sterile
^
FRG
Normal
saline or
tap water
50
0.45
Purchase
sterile or
boil for 20
min in dis-
tilled water
France Greece1 Israel Italy
Distilled Ringer's
water or solution l/4x
Ringer's
l/4x
47 47 47
0.45 0.45 0-45
Purchase Purchase Purchase
sterile or sterile U*^1" °r
autoclave or u.v.
boil for
20 min
to
VO
-------
TABLE D-2 Continued
Netherlands
3. Dilution water, if 0.1% peptone
used, specify water
formula
4. Membrane filter
a. Filter 47
diameter, urn
b. Pore 0.45
diameter, um
c. Sterilization
Autoclave - time Purchase
& temperature sterile or
boil
Norway Spain
Not specified Phosphate
water (1.25
ml/1 of 34%
solution of
KH2P04)
47 47-50
0.45 0.45
Boil for 10 min at
5-10 min 120"C
Sweden UK
Ringer's
solution l/4x
47
0.45
Boil in
sterile dis-
tilled water
10-15 min
US
Phosphate buf-
fered dis-
tilled water
or 0 . 1% pep-
tone water
47
0.45
10 min at
121°C
M
to
f
a
-------
TABLE D-2 Continued
Canada
Denmark
FRG
France
Greece
Israel
Italy
5. Media
Single :step
procedure
Liquid medium
M-Endo broth
0.4% Teepol
broth or
M-Endo broth
M-Endo broth
Solid medium
LES-Endo agar
Teepol agar
(Burman)
Endo agar
Chapman
modified by
Buttiaux
b. Enrichment
procedure
Enrichment
liquid medium
Enrichment
solid medium
Final liquid
medium
Final solid
medium
c. Sterilization
time &
temperature
Lauryl tryp-
tose broth
M-Endo broth
LES-Endo agar
Lauryl tryp-
tose: 121*C
for 15 min
Endo broth or
"agar: heat to
boiling
110-C for
15 min
120eC for
20 min
Steaming for
30 min on 3
consecutive
days or
boiling --
Heat to
boiling
10
-------
TABLE D-2 Continued
Netherlands
Norway
Spain
Sweden
UK
US
5. Media
a. Single step
procedure
Liquid medium
Solid medium
b. Enrichment
procedure _
0.4% en-
riched Teepol
agar
M-Endo broth
or Sartorius
dehydrated
Endo pads
M-Endo agar or
Endo LES agar
M-Endo broth
0.4% Teepol
broth
M-Endo broth
LES-Endo agar
Enrichment liquid
medium
Enrichment solid
medium
Final liquid
medium
Final solid
medium
Sterilization -
time 6
temperature
121 »C for
IS min
Heat to
boiling and
cool imme-
diately to
45-50*C
Heat to
boiling
Steaming for
30 min on 3
consecutive
days
Lauryl tryp-
tose broth
M-Endo broth
LES-Endo agar
Lauryl tryp-
tose: 121*C
for 15 min
Endo broth or
agar: heat to
boiling
(0
to
-------
TABLE D-2 Continued
Canada Denmark FRG Prance " Greece1 Israel
Italy
6. Incubation
a. Single step
procedure
Time (hours), 22-24. 24 & 48 20 + 4. (but 24(21-25), 4 hr at 30- 20-24,
temperature (*C) 35 + 0.5* 37 + 0.1'C observe up to 37* then 14-16 hr 35 + 0.5
T^ * ~W / f
37 + 0.5*C
b. Enrichment.
procedure
. Enrichment: time, 1.5> hr,
temperature 35 + 0.5*
Final steps time, 20-22 hr,
temperature 35+0.5*
u>
o
o
-------
TABLE D-2 Continued
Netherlands
Norway
Spain
Sweden
UK
US
6. Incubation
a. Single step
procedure
Time (hours), 6 hr at 25"
temperature (*C) then 12 hr
at 37*
24 + 3,
37 + 0.5*
24
35 -f 0.5»
4 hr at 30*
then 14 hr
at 35' or 37'
22-24,
35 -f 0.5*
b. Enrichment
procedure
Enrichments time,
temperature
Final step: time,
temperature
1, 5-2 hr,
35 + 0.5*
20-22 hr,
35 + 0.5*
o
to
-------
TABLE D-2
Continued
7. Counting and
reporting
a . Appearance of
coliform colony
b. Visual aids used
Canada Denmark
"Golden green, Yellow
metallic
sheen
10-15 x bino- ~
cular wide-
FRG France
Flat, red or Yellow or
with metallic orange or
sheen brick with
yellow halo
in medium
Hand lens
(8x)
Greece1 Israel Italy
Teepol: Yellow Metallic or
Endo: Metallic dark sheen
sheen
Stereoscopic Hand lens
microscope
Units of
reporting results
ing microscope
No/100 ml
No/100 ml
Colifqrms
present or
absent /100
ml; number/
100 ml
No/100 ml
No/100 ml
No/100 ml
-------
TABLE D-2 Continued
Netherlands
Norway
Spain
Sweden
UK
US
7. Counting and
reporting
Appearance of
coliform colony
Yellow
Dark red, with Dark, metal-
or without lie sheen
metallic sheen
Yellow
Golden green,
metallic sheen
b. Visual aids used
c. Units of No/100, ml
reporting results
Hand lens
lOx
No. of No/100 ml
colonies and
average of
2 replicates/
100 ml
Hand lens
No/100 ml
10-15 x bino-
cular wide-
field dissect-
ing microscope
No/100. ml
to
o
u
-------
TABLE D-2 Continued
Canada
Denmark
FRG
France
Greece
Israel Italy
8. Coliform verification
a. If required,
frequency
Not required Not required
Yes, all
samples
Not required Not required Not required
b. Verification
procedure
Cytochrome
oxidase; lac-
tose fermen-
tation at
44 + 0.5*C in
20 + 4 hr;
utilization of
citrate
Treated
water: check
yellow colony
for fermenta-
tion of glu-
cose and lac-
tose, oxidase
negative, and
gram-negative
Transfer to
glutamate or
MacConkey
broth (48 hr
at 37')?
transfer posi-
tive acid and
gas to MacCon-
key agar; typi-
cal , gram-nega-
tive = coliform
Transfer
atypical
colonies to
MacConkey or
violet red,
bile agar;
red colonies
= coliforms
Adjustment of
results per
verification
No
Yes, only
verified
results
reported
No
NO
Yes, add num-
ber of veri-
fied colonies
-------
TABLE D-2 Continued
Netherlands
Norway
Spain
Sweden
UK
US
8. Coliform verification
a. "If required,
frequency
b. Verification
procedure
c. Adjustment of
results per
verification
Yes, if Not required
yellow colo- except for new
nies present sample sites
with atypical
colonies
Transfer 5
colonies to
2% brilliant
green bile
broth; gas
= coli forms
No, no coli- . .
forms permitted
Not required.
Check dark
colonies .with
metallic
sheen for
E. coli (See
Table 4)
_.
Yes, for
treated water
^ ' - '
Transfer to
1% lactose-
peptone and
incubate at
37° for
48 hr; trans-
fer to nutri-
ent agar;
acid and gas,
oxidase nega-
tive = coli-
form
Yes,, reduce
counts accord-
ing to results
Yes, 5 or
more colonies/
filter with
> 5 colonies
Transfer to
lauryl tryptose
broth; incu-
bate 24-48 hr
at 35 + 0.5*;
transfer to
brilliant green
bile broth.
24-48 hr at
35 + 0.5";
gas = coliform
Yes, but pro-
cedure not
specified
10
" O
-------
TABLE D-2 Continued
Canada
Denmark
FRG
France
Greece
Israel
Italy
Water quality
standard,
specify
< 10/100 ml,
90% of samples
in 30 days
negative, no
coliforms in
2 consecutive
samples, no
fecal coli-
forms
Coliforms
should be
absent in
100 ml
Coliforms
should be
absent in 100
ml (recom-
mended) ; no
E. coli/100 ml
"(mandatory)
Untreated
water: coli-
forms should
be absent;
no E. coli;
treated
waten no
coliforms
Unchlorinated:
50% of samples/
month < 1/100
ml; 80% < 2/100
ml; none > 10/
100 ml;
chlorinatedi
< 1/100 ml
See
Table 3
Membrane-filter technique not legally specified
to
o
-------
TABLE D-2 Continued
1 Netherlands Norway
1 9. Water quality No coliforms/ Undisinfected
standard, 100 ml surface water:
specify good < 1/100
ml; not fit
for use > 30/
100 ml; disin-
fected surface
water; good
< 1/100 ml;
not fit for
use > 2/100 ml;
in a year 80%
of samples
should be good
Spain
No E. coli/
100~mlT~~
coliforms:
good < 2/
100 ml
tolerable
> 2 - < 10/
100 ml
unsuitable
> 10/
100 ml
Sweden UK
Treated water:
no coliforms/
100 ml; in
practice, the
standard is
not always
obtainable in
the distribu-
tion system,
and the fol-
lowing stand-
ards apply:
in a yr, 95%
of samples
coli form free;
no samples >
10/100 ml; no
coliforms in
2 successive
samples
US
1/100 ml as
arithmetic
monthly mean,
4/100 ml in
< 1 sample if
< 20 samples/
month; 4/100
ml in 5% of
samples if
> 20 samples/
" month
Ul
o
-------
304
additionally, include the negative cytochrome-oxidase reac-
tion (Canada, FRG, and U.K.). Other definitions are derived
from liquid enrichment culture, procedures that specify the
production of gas or acid and gas from a medium containing
lactose (FRG, France, Netherlands, and U.K.)., Irrespective
of the formal definition, all countries generally have a
common understanding of the term total coliform bacteria.
Sample size is relatively uniform at 100 ml, although
50-ml samples are occasionally specified or permitted. With
two exceptions (Norway and Spain), analysis of replicate
sample portions is not required and diluted samples are not
routinely analyzed. In the event that sample dilutions were
necessary, as would occur with highly,contaminated raw or
untreated waters, there is no consensus with regard to type
of diluent used. Distilled water, tapwater, normal saline,
Ringer's solution at 1/4 normal strength, 0.1 percent pep-
tone water, or phosphate buffered distilled water are all
used.
The membrane filter itself is relatively standardized
at a diameter of 47 or 50 mm with a pore diameter of 0.45
um. It is most commonly purchased sterile; if not, it is
boiled for 5 to 20 min or sterilized by autoclaving for 10
min at 121°C.
With only one exception (Canada), every country uses a
single-step rather than a two-step (enrichment and transfer
to final medium) procedure. The single-step medium may be
liquid (M-Endo broth or 0.4 percent Teepol broth) or solid
(LES Endo agar, Endo agar, Chapman agar, or 0.4 percent
Teepol agar). In the two-step procedure, which is seldom
used in drinking water testing, lauryl tryptose is used as
the enrichment medium. Media are sterilized by steaming for
30 min on three consecutive days (Greece and U.K.), heating
to boiling (for M-Endo broth) or autoclaving at 120 to 121°C
for 15 or 20 min. Incubation, most commonly, is for 20 to
24 h at 35 or 37°C, although several countries use two-step
incubation (Greece and U.K. 4 h at 30°C, then 14 to 16 h
at 35 to 37°C; Netherlands ~ 6 h at 25°C, then'12 h at
37°C).
The specified appearance of the coliform colony is a
function of the medium used. For Endo-type media, a metallic
sheen and/or a red color is required; with Teepol media,
coliform colonies appear yellow; and with Chapman agar,
coliforms appear yellow to orange. Colonies on membranes
-------
-------
305
may be examined with or without magnification. Results
uniformly are reported as a number per 100 ml.
Verification of coliform colonies is interpreted by the
use of additional tests such that the colony identified as
a coliform on the membrane filter conforms to the above
definition of coliform bacteria, except where the practice
is to confirm the colony as E. coli which has a different
significance. The simplest procedure for verifying coli-
forms involves transfer to lactose-containing media (lauryl
tryptose broth and/or brilliant green bile broth) and the
demonstration of gas after up to 48 h incubation at 35 to
37°C. Lauryl tryptose broth and brilliant green bile broth
are not used for coliform verification in the U.K., on the
premise that an inhibitory agent is unnecessary where the
inoculum is from a'single (and presumably pure) colony.
More involved techniques include determination of cytochrome-
oxidase activity and citrate utilization as well as gram
stain. Verification usually is not required.
The criterion for the bacteriological acceptability of
water is the absence of coliform bacteria (in 100 ml) [See
also Section C.l.b]. Because of procedural and statistical
complications, this is variously stated as < 10 coliforms
per 100 ml, < 1 coliform per 100 ml, etc. Most countries
include in their standard the sampling frequency, how fre-
quently coliforms are found, and an absolute coliform number.
The membrane filter procedure is not formally specified in
several countries (Greece, Israel, Italy, and Sweden) and
appears not to be used officially in Italy and Sweden
(Sweden has proposed its use as an official procedure).
3. Total Coliform Testing, Multiple Tube Technique
Table D-3 parallels Table D-2 for total coliform testing,
but by the older multiple tube technique rather than the
newer membrane filter procedure. The definitions of coli-
form bacteria are comparable to those given in Table D-2,
but are restricted to organisms producing gas or acid and
gas from a medium containing lactose.
The simpler procedures require testing of five 10-ml or
five 100-ml portions of the sample (Canada and U.S.). All
other countries use various combinations of sample volumes
and tubes per dilution such as ten 10-ml, five 1-ml and one
0.1-ml portions (France); five 10-ml, one 1-ml, and one 0.1-
ml portions (Israel); and five 100-ml, five 10-ml, and five
-------
-------
TABLE D-3
TOTAL COLIFORM TESTING, MULTIPLE TUBE TECHNIQUE, DRINKING WATER ANALYSIS
Canada
Denmark
FRG"
France
Greece
Israel
Italy
1. Total coliform
bacteria, define
See Table 2 Bacteria pro- See Table 2
dueing gas
from lactose
in 48 h at
37"C that
have bio-
chemical
& morphologi-
cal simi-
larity to
E. coli
See Table 2 See Table 2
See Table 2
Organisms pro-
ducing gas
from lactose
in lactose
broth in
24/48 hours
at 37'C and
in brilliant
green lac-
tose bile
broth in
24/48 hours
at 37°C
2. Samples
a. Volume/tube, ml 10 or 100
b. Tubes/sample 5
c. Are diluted No
samples analyzed
(1 x 50 ml
(5 x 10 ml
(5 x 1 ml
100
Not routinely Yes, if ne-
cessary
(10 x 10 ml
(5 x 1 ml
(1 x 0.1 ml
(1 x 0.01 ml
(1 x 0.001 ml
Not specified (
Yes
(10,1, 0.1 ml
in either a
5-tube or
3-tube repli-
cate series
(5 x 10 ml (5 x 10 ml
(1 x 1 ml (5x1 ml
(1 x 0.1 ml (5 x 0.1 ml
Not routinely No
No
u>
o
ON
-------
TABLE D-3 Continued
Netherlands Norway
1. Total coliform See Table 2 Aerobic
bacteria, define facultatively
anaerobic.
gram-negative,
non-sporeform-
ing rod-shaped
bacteria pro-
ducing acid
& gas from
lactose at
37 *C in 48
hours
Spain
Organisms pro-
ducing acid &
gas from lac-
tose in 48 h
at 37*C
Sweden UK
Same as for See Table 2
Norway
except incu-
bation tem-
perature of
35'c
US
Aerobic and
facultative
anaerobic.
gram-negative
non-sporeform-
ing rod-shaped
bacteria
producing gas
at 35'C in
48 hours
2. Samples
a. Volume/tube, ml
b. Tubes/sample
c. Are diluted
samples analyzed
(100 ml or
(5 x 10,
(1 x 50 ml
No
(5
(5
(5
x 10 ml
x 1 ml
x 0.1 ml
(1
(5
(5
x 50 ml
x. 10 ml
x 1 ml
(5
(5
(5 x 1 ml
x 100 ml
x 10 ml
(1
(5
50 ml
10 ml
Not routinely Not routinely Not routinely
(5 x 1 ml
Yes, if
highly
contaminated
(5 x 100 ml
( or
(5 x 10 ml
(
No
o
CTl
tt
-------
TABLE D-3 Continued
Canada
Denmark
FRG
France
Greece
Israel
Italy
3. Dilution water, If
used, specify
formula
Phosphate
buffered
distilled
water
Normal saline
or tap water
Distilled Ringer's
water or solution
Ringer's l/4x l/4x
w
o
-------
TABLE D-3 ~ Continued
Netherlands Norway
Spain
Sweden
UK
US
3. Dilution water,
used, specify
formula
if
0.1% peptone Not specified
Phosphate Ringer's
buffered dis- solution
tilled water l/4x
u>
o
-------
TABLE D-3 Continued
Canada Denmark
4. Media
a. Presumptive test Lactose or MacCorikey
lauryl broth
tryptose
broth
FRG2
Lactose broth
(double
strength ) ,
Endo agar
France2 Greece Israel
Lactose broth Minerals modi- Lactose
fied glutamate broth
or MacConkey's
broth
Italy
Lactose
broth
b. Is double Yes Yes
strength required
for 10 ml
portions
c. Confirmed test Brilliant MacConkey
green bile broth
broth
d. Completed test, EMB or Endo
if used, specify agar, lac-
tose broth,
nutrient
agar
e. Sterilization - 15 min, 15 min
time and 121 *C
temperature
Yes Yes
Endo agar, Brilliant
tryptophane breen,bile
broth, citrate broth
agar, sugar
broths, cyto-
chrome oxidase
Yes
Yes
Yes
20 min,
0.8 atm
20 min,
120*C
MacConkey MacConkey Brilliant
agar, lactose agar or green bile
broth brilliant broth
green broth
12-20 min,
115*C
15 min,
121 *C
15 min,
121*C
u
o
00
-------
TABLE D-3 Continued
Netherlands
Norway
Spain
Sweden
UK
US
4. Media
a. Presumptive test
b. Is double
strength required
for 10 ml
portions
c . Confirmed test
i
d. Completed test,
if used, specify
e. Sterilization -
time and
temperature
Minerals
modified
glutamate
(oxoid)
Yes
Brilliant
green bile
broth
15 min,
121°C
Lactose-pep-
tone broth
with brom
cresol purple
No, triple
(3x)
Brilliant
green bile
broth
Brilliant
green bile
broth
15 min,
110'C
Lactose
broth with
brom cresol
purple
Yes
Not required
but identify
coliforms
isolated on
EMB agar
20 min
Lactose broth
with brom
cresol purple
Yes
Brilliant
green bile
broth or
Endo agar
Brilliant
green or
Endo agar
15 min,
110°C
Improved for-
mate lactose
glutamate
broth or
minerals
modified glu-
tamate (oxoid)
Yes
Brilliant
green bile
broth or lac-
tose ricino-
ieate broth
MacConkey
agar, nutri-
ent agar
-
IFLG 10 min,
115*C
BGB 15 min,
115°C
LRB 20 min,
115'C
Lauryl tryptose
broth
Yes
Brilliant
green bile
broth
EMB agar,
lauryl tryp-
tose broth.
nutrient agar
15 min,
121°C £
00
tn
-------
TABLE D-3 Continued
5 . Incubation
a. Presumptive test,
time and
temperature
b. Confirmed test,
time and
Canada
48 + 3,
35 + 0.5°
48 + 3,
35 + 0.5°
Denmark
48 h
37 + 0.1'C
24
44°C
FRG2
20 + 4,
37 + 0.5°
Negative at 20
+ 4 h at 44 +
2 '""'
France Greece
48, 30 + 1° 48, 37°
(or 37"T
48, 37° 24/48, 37°
Israel Italy
24/48, 35° 24/48, 37°
24, 35" 24/48, 37°
temperature
6. Counting and
reporting
a. Define positive
tubes
Presumptive
Confirmed
b. Units of
reporting
results
1-c
Gas in any Acid & gas
amount
Gas in any Acid & gas
amount
MPN/100 ml MPN/100 ml
Acid & gas Gas
See Item 7
Gas
Coliforms MPN/100 ml
present or ab-
sent/100 ml;
coliform titer
Acid & gas Gas in any Gas in any
amount amount
Typical colo-. Red and
ny gram-nega- dark pink
tive: coli- colonies
form
MPN/100 ml, MPN/100 ml MPN/100 ml
% tubes posi-_
tive
-------
TABLE D-3 -- Continued
Netherlands Norway
5 . Incubation
a. Presumptive test, 48, 37° 24/48 + 3,
time and 37 + oT5°
temperature
b. Confirmed test, 48, 37° 24/48 + 3,
time and , 37 + oT5°
temperature ~
6. Counting and
reporting
a. Define positive
tubes
Presumptive Acid & gas Acid & gas
Confirmed Gas Gas
b. Units of MPN/ml MPN/100 ml
reporting or 100 ml,
results present or
absent in
- 100 ml
Spain Sweden UK US
-
48 h 48, 35 + 0.3° 48, 37* 24/48 + 3,
30 + 1°C ~ 35 + 0-5o
_
-------
TABLE D-3 Continued
7. Completed test
(coliform
.verification)
a. Required
frequency
Canada Denmark FRG2 France2 Greece
As NO Yes, all See footnote Yes, for
required samples chlorinated
samples
Israel Italy
No No
b. Procedure
Adjustment of
results per
completion
Streak on
EMB or Endo
agar, pick
colonies to
lactose broth
and nutrient
agar; gas,
non-spore-
forming,
gram-
negative =
coliform
Yes
Streak on Endo
agar, pick to
lactose broth,
tryptophane
broth, citrate
agar; gas, in-
cubation at 44
+ 0.5"C, & in-
dole produced,
citrate & cyto-
chrome oxidase
(_) = E;. coli
Yes
Gas from lac-
tose, gram-
negative rods
Yes
CO
i-1
o
-------
TABLE D-3 Continued
7. Completed test
(coliform
verification)
a. Required
frequency
b. Procedure
Yes, for new
sources or as
necessary
Gas from lac-
tose at 44"C
and indole
positive
As required
Identify coli-
forms iso-
lated on EMB
dole, urease,
& citrate
utilization
tests); gram
(-) rods
Yes, if re-
sults diffi-
cult to
interpret
Gram and
spore stain
of colonies
from Endo
agar, gas
from lactose
Yes, for
treated water
Typical colo-
nies on
MacConkey
-agarv IMViC
.and oxidase
negative
Yes on 10% of
positive sam-
ples or lx/3
months
Same as Canada
Adjustment of
results per
completion
Yes
Yes
Yes
Yes
to
H
O
W
-------
TABLE D-3 Continued
Canada
Denmark
FRG
Prance
Greece
Israel
Italy
8. Water quality
standard
90% of sam-
ples in
consecutive
30 days
negative;
no coli-
forms in 2
consecutive
samples; no
sample with
MPN > 10/
100 ml
Coliforms
should be
absent in
100 ml
Coliforms
should be
absent in
100 ml; no
E. coli/100
ml
See Table 2,
#9
See Table 2,
19
Good:
< 2/100 ml;
repeat
sampling:
3-10/100 ml
(if posi-
tive, check
for fecal
coliforms);
> 10/100 ml,
repeat
sampling
immediately
Coliforms
and thermo-
tolerant
coliforms
should be
absent in
100 ml
u
H
H
I
-------
TABLE D-3 -- Continued
Netherlands
Norway
Spain
Sweden
UK
US
8.
Water quality
standard
See Table See Table 2, See Table 2
2, #9 f9
Clean water;
suitable < I/
100 ml; un-
suitable >
10/100 ml
See Table 2,
#9
10 ml sample:
< 10% tubes
pos itive/month
100 ml samples:
< 60% tubes
positive/month
fr°m tables in Standard Methods (American Public Health Assoc., 1976). 14th
2
Method of choice = Membrane filter technique.
3Unofficial confirmation uses brilliant green bile broth. Official procedure requires confirmation of E. coli
gas in bile broth and indole production, both at .44 + 0.5' or isolation and identification of all colTfbHS"
(by streaking on EMB and incubating at 30" for 24 or TS hours followed by IMViC tests) coiitorms
-------
312
1-ml portions (Sweden). If diluted samples are required,
the diluents mentioned above are used.
For the presumptive test, lactose or lauryl tryptose
broth are most commonly used, although several countries use
chemically defined media (minerals modified glutamate broth
Greece and Netherlands; improved formate lactose glutamate
broth U.K.). All countries except Norway (which uses
triple-strength) use double-strength presumptive medium for
sample portions of 10 ml or more. For the confirmed test,
brilliant green bile broth is most common except where Endo
agar (FRG), MacConkey agar (Greece and Israel), or lactose
ricinoleate broth (U.K.) are used. EMB or Endo agar, lactose
or lauryl.tryptose broth, and nutrient agar are used for the
completed test. Media are sterilized by autoclaving at
temperatures as high as 121°C for 15 min (Canada, Israel,
Italy, Netherlands, and U.S.) and 120°C for 20 min (France
and Spain) and at temperatures as low as 110°C for 15 min
(Norway and Sweden).
With the presumptive test, samples are usually incu-
bated for up to 48 h at 35 or 37°C; in the FRG, only a 24 h
incubation at 37°C is required; in France and Spain incu-
bation may be at 30°C. Likewise for the confirmed test,
incubation is typically for 48 h at 35 or 37°C except for
Israel (24 h at 35°C).
A presumptive-positive tube is defined as one in which
gas has been produced; if the medium contains a pH indicator,
gas and ac,id production both are required for a positive
result (FRG, Greece, Netherlands, Norway, Sweden, and U.K.).
Gas production is the characteristic of a positive confirmed
test in all countries using a liquid confirmation medium.
Results are reported either as coliforms present or absent,
percentage, of tubes positive, or most often, as the most
probable number (MPN) per 100 ml.
The completed test is used for coliform verification
and is comparable to verification in the membrane filter
procedure. It is not typically required, at least_not for
all positive samples. As with membrane filter verification,
it is sometimes defined in terms of E. coli presence (FRG,
France, Norway, and U.K.) by such tests as indole production
at 44°C, gas production from lactose at 44°C, or the IMViC
series.
The water quality standards are comparable to those
described above under the membrane filter technique [See
also Section C.l.b].
-------
-------
313
4. Thermo-Tolerant (Fecal) Coliform Testing,
Membrane Filter Technique
Table D-4 summarizes the data obtained for the defi-
nition of thermo-tolerant (fecal) coliform bacteria and the
analytical details used for the membrane filter technique.
The term "thermo-tolerant" is defined as that portion of the
coliform group capable of producing gas from lactose within
24 h at a temperature around 44.5°C. Several countries use
the term "fecal" coliforms to convey the same information.
Others, the U.K. and the FRG for example, do not base analyses
for sanitary significance on the demonstration of fecal
origin by the ability of coliform bacteria to ferment lactose
at an elevated temperature, but rather, test specifically
for 1C. coli.
Where the membrane filter method is used, analytical
details are identical to those given in the section on total
coliform testing except that the medium is M-FC broth (Teepol
agar for Denmark) and incubation is at 44.0 or 44.5°C for
24 h; or, in the case of the Netherlands, incubation is 6 h
at 25°C, then 12 h at 44°C; and for Denmark and the,U.K.,
incubation is 4 h at 30°C, then 14 to 18 h at 44°C. Thermo-
tolerant coliform colonies will appear blue (or yellow on
Teepol agar).
The total coliform test appears to serve as a primary
criterion in determining the microbiological quality of
finished drinking water. However, tests for thermo-tolerant
coliforms, or for E. coli, are specified in various coun-
tries for evaluation of raw water, of untreated water in
distribution, and even for finished water [See Section
C.l.c].
5. Thermo-Tolerant (Fecal) Coliform Testing,
Multiple Tube Technique
Table D-5 parallels Table D-4. The generalizations
made above are equally applicable here. The definitions of
thermo-tolerant coliform bacteria are comparable to those
given in Table D-4, although there are some added analytical
steps that are more specific for IS. coli than for thermo-
tolerant coliforms as a whole (gas production at 44°C and
negative oxidase and urease reactions).
Except for Italy and Sweden, which make direct inocu-
lations, those countries using the tube technique use it as
-------
-------
TABLE D-4
THEPMO-TOLERANT (FECAL) COLIFORM TESTING, MEMBRANE FILTER TECHNIQUE, DRINKING WATER ANALYSIS
1. Thermo-tolerant
coliform, define
Canada
Denmark
2. Samples
a. Volume filtered/
filter, ml
b. Number of replicates/
sample
c. Are diluted samples
analyzed
3. Dilution water, if used
That portion of Facultative
coliform group anaerobic,
producing gas
from lactose
at 44, 5°C in
24 hours
See Table 2
lactose in
24 h at 44"C
See Table 2
FRG
France
Greece
Israel
Italy
See
See 0 Not Not That portion
footnote1 footnote2 applicable applicable of coliform
Table 2 See Table 2
100
Not required
No
u>
(-
I
-------
TABLE D-4 Continued
Netherlands
1 . Thermo-tolerant Facultative
coliform, define anaerobic,
gram-negative
non-sporogenic
bacteria grow-
ing at 37° and
44°C in lactose
medium and pro-
ducing gas and
acid in 48 hours
Norway Spain Sweden UK
That portion See foot- Not appli- see footnote3
01 conform note cable see Table 2
group growing
at 44 + 0.2°C
""""
US
Not required
2.
Samples
a
See Table 2
See Table 2
Volume filtered/
filter, ml
Number of replicates/
sample
Are diluted samples
analyzed
3. Dilution water, if used
i
Cd
-------
TABLE D-4 Continued
Canada
Denmark
FRG
.France
Greece
Israel
4. Media
a. Medium
b. Sterilization - _time
and temperature
Incubation - time and
temperature
Counting and -reporting
a. Appearance of fecal
coliform colony
M-FC broth
*
Heat to boiling,
cool promptly
24, 44.5- +
0.2°C
Blue
4 h, 30°C
then 14-18
h, 44°C
Italy
M-FC broth
Heat to
boiling
24 + 2,
44.5°C
Blue
b.
c.
Visual aids used
Units of reporting
results
See Table 2
No/100 ml
No
No/100 ml
to
M
I
-------
TABLE D-4 Continued
4. Media
a. Medium
b. Sterilization - time
and temperature
5. Incubation - time and
temperature
6. Counting and reporting
a. Appearance of fecal
coliform colony
b. Visual aids used
c. Units of reporting
results
Netherlands
6 hr, 25°
then 12 hr,
44°
Norway
Same as in
Table 2
24 + 3,
44 +0.2°
Dark red
Lens and lamp
No/100 ml
Spain
Sweden
UK
US
4 h at 30*C
then 14 h
at 44"C
cn
w
-------
TABLE D-4 Continued
Canada
Denmark
FRG
France Greece Israel
Italy
7. Thermo-tolerant coliform
verification
a. If required, :
frequency
No thermo-
tolerant
coliforms/
100 ml
b. Procedure
w
>->
a\
-------
TABLE D-4 Continued
Netherlands
7. Thermo-tolerant coliform
verification
a. If required, «
frequency
Norway
Spain
Sweden
Yes, if yellow Not routinely
colonies present
UK
Yes, for treated
water
US
b. Procedure
See Table 2
By comparison
with multiple
tube method
Transfer to 1%
lactose peptone
and peptone
water; incubate
24 hours at 44°;
gas and indole
production =
E. coli
W
-------
I
8. Water quality standard
TABLE D-4 Continued
Canada
Denmark
FRG
France
Greece
Israel
Italy
Raw water:
90% of samples
in 30 days
< 100 fecal
coliforms/
100 ml; com-
plete treat-
ment required
No thermo-
tolerant
coliforms/
100 ml
(jj
M
I
-------
TABLE D-4 -- Continued
Netherlands
Norway
Spain
Sweden
UK
US
8. Water quality
standard
No thermo-stable
coliforms/100 ml
Treated water:
no E_. coli/100 ml
In practice
the standard
is not always
obtainable
in the distri-
bution system
and the fpl-.. -
lowing stan-
dards apply:
In a year, 95%
of samples
should not con-
tain E. coli;
no sample should.
contain > 2 E_.
coli/100 ml;
no sample
should contain
1 or 2 E. coli/
100 ml if coli-
forms > 3/100 ml
10
I-1
I
-------
TABLE D-4 Continued
Footnotes t
E. coli identified by negative cytochrome oxidase reaction, lactose fermentation at 37*C in 20 + 4 h; indole formation, lactose
fermentation at 44 + 0.5eC in 24 hours; inability to utilize citrate. Thermo-tolerant colifornTanalysis not legally required.
2 ' "
Term thermo-tolerant to be used (ISO definition). E. coli identified by gas production in brilliant green bile broth at 44 +
0.5e in 48 hours and indole formation at 44 + 0.5 *C or by picking colonies from EMB agar and making IMViC tests or by MF ~~
technique (Table 2} incubated at 44-45 "C.
Thermo-tolerant coliform test not used but E. coli is identified. IS. coll is defined as a coli form fermenting lactose
with production of acid and gas at 44"C in "24 hours and giving IMViC reactions ++--
-------
-------
TABLE D-5
THERMO-TOLERANT COLIFORM TESTING, MULTIPLE TUBE TECHNIQUE, DRINKING WATER ANALYSIS
Canada Denmark FRG Prance Greece
1. Thermo-tolerant coli- See Table 4 See Tables Not See Tables Gram-negative
form bacteria, defi- 3 & 4 . appli- 3 & 4 rods producing
nition ^able gas from lac-
tose and in-
dole at 44°C,
oxidase and
urease negative
Israel
That portion
of coliform
group producing
gas in EC me-
dium at 44.5°C
in 24 hours
Italy
That portion
of coliform
group produc-
ing gas from
lactose in
brilliant
green bile
broth in 24
hours at 44°C
2. Samples
a. Direct inoculation,
ml and number of
tubes-
b. Transfer from
positive presump-
tive total coliform
tubes
3. Dilution water, if
used, specify formula
Yes, from
all
Yes, from
all
See Table 3
Yes
Yes
(5 x 10 ml
(5 x 1 ml
(5 x 0.1 ml
-------
TABLE D-5 Continued
Netherlands
Norway
Spain
Sweden
UK
US
1. Thermo-tolerant coli-
form bacteria, defi-
nition
See Table 4
See Tables
3 & 4
See Tables
3 & 4
See Tables Not required
3 & 4
Samples
- a
Direct inoculation,'
ml and number of
tubes
b. Transfer from
positive presump-
tive total coliform
tubes
Dilution water, if
used, specify formula
Yes
(5 x 10 ml
(5 x 1 ml
(5 x 0.1 ml
Yes
Not specified
5 x 10 ml,
5x1 ml,
5 x
0.1 ml
Phosphate
buffered dis-
tilled water
vo
u
-------
TABLE D-5 Continued
f
4 . Media
a. Medium
b. Sterilization -
time and
temperature
5. Incubation - time
and temperature
- Canada Denmark FRG France
EC medium MacConkey Peptone
broth water,
brilliant
green bile
broth
15 min, 15 min, 20 min,
121"C 110°C 120°C
24+2, 24 h, 24/48,
44. ₯ +0.2° 44°C 44 + 0.5° '
Greece Israel
Peptone water, EC medium
brilliant
green bile
broth
(a) 15 min, 15 min, 121°C
110"C
(b) 15 min,
121'C
24/48, 44" 24+2,
44.5 + 0.2°
Italy
Lactose broth
(presumptive )
brilliant
green lactose
bile broth
15 min, 121°C
24, 44.5"
to
o
-------
TABLE D-5 Continued
Netherlands
Norway
Spain
Sweden
UK
US
4. Media
a. Medium
2% brilliant Lactose broth
green bile with brom
broth cresol purple
Lactose broth
with brom
cresol purple
b. Sterilization
time and
temperature
15 min, 121°C 15 min,
110 °C
15 min,
110°C
5. Incubation - time
and temperature
48, 44°
24/48 + 3 h
37 + 0.5°C
24/48 + 3 h
44 + 0.2"C
48, 44°
24, 44°
Ul
to
o
W
-------
TABLE D-5 Continued
Canada Denmark
6. Counting and reporting
a. Define positive Gas Acid & gas
tubes
b. Units of MPN/100 ml MPN/100 ml
reporting
results
7. Water quality See Table 4 See Table 4
standard
FRG France
Indole
positive,
gas
MPN/100 ml
See
Table 2
Greece
Indol
positive,
gas
MPN/100 ml
Unchlor inated :
none in succes-
sive samples;
chlorinated;
no sample
Israel
Gas
MPN/100 ml
No thermo-
tolerant
coliforms
should be
present
Italy
Gas
Number/100 ml
No thermo-
tolerant
coliforms/
100 ml
> 2/100, and
no sample
> 1-2/100
if coliform
MPN > 3/100 ml
-------
TABLE D-5 Continued
Netherlands
Norway
Spain
Sweden
UK
US
6. Counting and reporting
a. Define positive
tubes
Gas
b.
Units of
reporting
results
7.
Water quality
standard
MPN/100 ml
or present
or absent
No thermo-
tolerant coli-
forms should
be present
Gas & acid
presumptive
test
Gas & in-
dole com-
pleted test
MPN/100 ml
No thermo-
stable coli-
forms/100 ml
Acid and gas
MPN/100 ml
For well
water: good
quality =
< 2/100 ml;
doubtful =
2-9/100 ml;
unsuitable =
> 10/100 ml
Not legally specified but recommended by the Greek Microbiological Society.
Not required for drinking water but used for well water and water for individual consumption.
oo
to
-------
322
a confirmatory test following the standardized total coliform
presumptive test. Brilliant green bile broth or EC medium
are usually used with incubation for 24 h at 44.5°C (Canada,
Israel, Italy) or 44°C (U.K.); or an incubation of up to 48
h at 44°C (Prance, Greece, Netherlands, and Sweden). Gas
from lactose and indole production (France, Greece, and the
U.K.) is the typical positive reaction [See also Section
C.l.c]. In Sweden, samples are incubated in lactose broth
at 44°C and a confirmatory step normally is not done.
6. Colony Count
Table D-6 summarizes the data obtained on the defi-
nition of the colony count and the analytical details used.
The term colony count is not uniformly used. Other terms
include standard plate count, microorganism or total micro-
bial count; aerobic or mesophilic viable bacteria, and
others. Irrespective of the term employed, what is under-
stood is the number of bacterial colonies produced per ml of
water under defined conditions of medium and time, tempera-
ture of incubation, and magnification for counting. There
is considerable diversity of these defined conditions as
will be seen below.
Where the procedure is formally used, it is always on
the basis of a pour plate technique using, at most, a 1ml
sample, typically with duplicate plates made per sample
volume. As already indicated, there is no uniformity in
selecting dilution water. '
The media generally used are based on extracts of beef
or yeast and various peptone breakdown products, solidified
by agar. This variety of natural organic substances leads
to nonuniformity from medium manufacturer to manufacturer
and even from lot to lot when prepared by the same manu-
facturer. When medium diversity (planned or unplanned) is
coupled with different incubation times and temperatures
varying from 48 to 72 h at 20°C (FRG, France, Norway, and
U.K.) to 48 h at 22°C (Sweden), to 48 h at 32°C (Israel), to
24 to 48 h at 35 or 37°C (Canada, France, Greece, Italy,
Netherlands, and U.K.) it becomes clear that results
among different countries cannot be directly comparable. On
the other hand, it is likely that any of these national
methods would identify gross differences in water quality.
This suggests that, unless there is a considerable collec-
tion of data within a country, the colony count is difficult
to use as part of the water quality standard. That only
Canada, the FRG, Norway, and Sweden incorporate such a count
into formal standards supports this point of view [See also
Section C.I.a].
-------
-------
TABLE D-6
COLONY COUNT, DRINKING HATER ANALYSIS
Canada
1. Colony count See ,
footnote
a. Define Not legally
defined,
only by the
analytical
method
Denmark
Number of
colonies/ml
developing on
defined media
at 21 "C &
37 "C
FRG
-
Number of
colonies/ml
developing at
fixed tem-
perature and
time on de-
fined medium
France
Number of
colonies/ml
developing at
either 37" in
24 hours or
20-22"C in 72
hours on de-
Greece
Number of
colonies/ml
developing on
nutrient agar
at 37*C in
24 hrs
Israel
Number of
colonies/ml
developing on
defined
medium at
32'C in 72
hours
Italy
Number of
colonies/ml
developing on
trypticase
glucose
extract agar
b. Procedure
2. Samples
a. Volume/plate, ml
b. Number of
replicates/sample
c. Are diluted
samples analyzed
3. Dilution water,
if used
Pour plate Pour plate
Pour plate
fined medium
Pour plate
Pour plate
Pour plate
Pour plate
As required
2/dilution
As required
Phosphate
buffered
distilled
water
1 & 0.1
2
Yes
Phosphate
buffered
distilled
water (See
Table 2)
1
2
As required.
Sterile tap
water
1
2/dilution
As required
Distilled
water or
Ringer ' s
solution
l/4x
1, 0.1 or
1 ml of
dilutions
2
As required
Ringer ' s
solution
l/4x
As required Not required
Phosphate
buffered dis-
tilled water
to
i
-------
TABLE D-6 Continued
Netherlands
1. Colony count
a. Define Number of
colonies/ml
developing on
an agar medium
under defined
conditions
Norway
Colony form-
ing units
under defined
conditions
Spain
Number of
colonies/ml
developing
on an agar
medium under
defined
conditions
Sweden
Number of colo-
nies visible
under 1.5x
magni f ication
after 48 hours
at 22 °C on de-
fined medium
UK
Number of
organisms
growing
aerobically
and forming
colonies
under defined
conditions
US
Not
required
b. Procedure
2. Samples
a. Volume/plate, ml
Pour plate
Pour plate
Pour plate
Pour plate
Pour plate
b. Number of
replicates/sample
c. Are diluted
samples analyzed
Dilution water,
if used
As required
0.1% peptone
water
As required
0.9% NaCl
Phosphate
buffered dis-
tilled water
or Ringer's
solution l/3x
As required
Phosphate
buffered dis-
tilled water
1-2/dilution
As required
Ringer's
solution
l/4x
u>
M
w
-------
TABLE D-6 Continued
4 . ;Media
a. Medium
b. Sterilization -
time and
temperature
5 . Incubation - time
and temperature
6. Counting and reporting
a. Visual aids used
b. Units of
reporting results
Canada
Tryptone
glucose
extract
agar or
tryptone
glucose
yeast agar
15 min,
121°C
48 + 3,
35 + 0.5°
Quebec
colony
counter
Standard
plate
count/ml
Denmark
21 °C count -
Kings agar
B
37 °C count -
Plate
count agar
15 min,
121°C
48 h, 37"C
72 h, 21°C
None
No/ml (37°C)
& no/ml total
& fluorescent
FRG
Meat ex-
tract - pep-
tone agar
20 min,
120°C
44 + 4,
20 + 2"
Hand lens
(8x)
No/ml
France
Yeast extract
agar
20 min,
118°C
24+1,
37 + 1°
72 + 4,
20 - 22°
Hand lens
(2-4x)
No/ml (37°)
and no/ml
(20°)
Greece Israel
,
Nutrient agar Tryptone glu-
cose yeast
agar
15 min, 15 min,
121°C 121'C
24, 37° 48, 32°
Hand lens Quebec colony
counter or
hand lens
(9x)
No/ml No/ml
Italy
Trypticase
glucose
extract
agar
15 min,
121°C
24/48, 37°
None
No/ml
u
M
*»
1
(21'C)
-------
TABLE D-6 Continued
4. Media
a. Medium
b. Sterilization -
time and
temperature
5. Incubation - time
and temperature
Netherlands
Tryptone glu-
cose yeast
agar
15 min, 121 *C
48, 37
72, 22*
Norway Spain
^
Meat pep- Nutrient
tone agar agar
without NaCl
15 min, 20 min,
121'C 120'C
72 + 3, 24 h, 37*C
20 + 2°
Sweden UK US
Meat peptone Yeast extract
agar agar
20 min, 120'C 20 min,
115'C
48, 22" 24, 37"
72, 20-22'
6. Counting and reporting
a. Visual aids used Automatic Hand lens
colony counter
None
Hand lens
Hand lens
b. Units of No/ml
reporting results
No/ml (aver- No/ml
age of two
replicates)
No/ml
No/ml
to
N>
a
-------
TABLE D-6 Continued
Canada
Denmark
PRG
France
Greece
Israel
Italy
7.
Water quality standard < 500/ml
based on
geometric
mean of
> 10 month-
ly samples
for conven-
tional
treatment
Recommenda-
tions: 37'C
< 10/ml, 21 *C
< 100/ml &
_< 5 fluores-
cent colo-
nies/ml
Recommenda-
tions: Non-
disinfected
water < 100 /
ml; disin-
fected water
< 20/ml
Not specified Not specified Not specified Not specified
CJ
to
01
I
-------
TABLE D-6 Continued
Netherlands - Norway
7. Water quality standard Not specified Undisinfected
surface water:
good < 100 ml,
doubtful 100-
500/ml; dis-
infected sur-
face water;
good < 10/ml,
doubtful 10-
100/ml
Spain
Clean water:
suitable
< 100/ml;
less suitable
> 100/ml
Sweden UK US
Clean water: Not specified
suitable
< 100/ml; less
suitable
_> 100/ml
Uses procedure given in Standard Methods for the Examination of Water
Wastewat
14th edition, pages 908-913.
ia
to
Ul
w
-------
326
7. Miscellaneous
The questionnaire on -laboratory practices and bacterio-
logical standards also included a seventh set of questions,
in response to which came a tremendous diversity of replies
that could not be presented as tables.
The first of these questions asked about the kinds of
laboratories (national, provincial, municipal, university,
water utility, and nongovernmental or commercial) and their
numbers in each country. Because of different national
administrative arrangements, it is impossible to generalize
about the kinds of laboratories analyzing water. The
number of laboratories involved relates both to the role
played by the central government in water quality control
and, more importantly, fo the size of the country. In
Israel and Greece, for example, there are fewer than ten
laboratories; whereas in the U.S. there are thousands of
water laboratories, many of them at water utilities.
Other questions dealt with control of water labora-
tories and the use of standards of laboratory quality. In
most countries/there is no formal system of technical super-
vision or direction for water laboratories, nor are there
formal standards by which the acceptability or general
quality of laboratory services are judged. It is assumed,
however, -that internal and inter laboratory quality control
are achieved through good laboratory practices. In Sweden,
laboratories are controlled by the National Board of Health
and Welfare and each analyst is required to have a personal
authorization although, in practice, authorization is
required only of the supervisor. Comparable authorization
for the laboratories is under consideration. In the U.S.,
each laboratory must be approved by the state in which it is
located.
Responses to questions concerned with interpreting
laboratory results and acting on laboratory findings were
varied: interpretation of results and initiation of cor-
rective action are carried out by the water analyst, a water
engineer, or a health authority. No consistent pattern was
evident.
8. Summary and Conclusions
Bacteriological assessment of drinking water quality is
typically based on the coliform group of bacteria. Less
-------
-------
327
frequently, standards are based on thermo-tolerant coliform
bacteria or even E. coli. Still less often, a count of
aerobic heterotrophic organisms is required.
The analysis for coliform or thermo-tolerant coliform
bacteria may be conducted either by the multiple tube tech-
nique, with results reported as a most probable number (MPN)
per 100 ml, or by the membrane filter procedure^which yields
a colony count. The least methodological diversity exists
for the total coliform test even though there is no uni-
versally accepted medium or procedure. It may be concluded,
however, that the specified procedures, when correctly
followed, will provide closely comparable results.
Tests for thermo-tolerant coliforms are more varied,
most notably in the time and temperature of incubation. The-
significance of these differences probably is slight.
The least uniform test is that made for the colony
count. Here, differences in media and incubation procedures
may yield markedly different results and comparisons across
national boundaries may be impossible, although the extreme
case of a very high count would be identified irrespective
of the procedure used.
Despite the existence of formal standards of water
quality, there are seldom formal standards of laboratory
quality; that is, formal laboratory quality control programs,
such as in the U.S., seldom exist. It may be concluded that
this subject is most in need of attention.
9. Recommendations
1. Greater efforts should be made at the national
and international level to obtain maximum
comparability of data by the use of uniform
analytical methods. The procedures of the
European Economic Community appear most suitable
for universal adoption.
2. The absence of formal laboratory quality as-
surance programs should be remedied. Proce-
dures to establish quality assurance should be
developed by national laboratories and enforced
at all water laboratories. Likewise, the
responsible national body should insure that,
without exception, all laboratories within, a
country are using the adopted methods.
-------
328
An adequate water testing program, using uniform
methodologies that are under close quality as-
surance, cannot guarantee distribution of safe
water to all consumers. Laboratory analyses
merely assess the quality of water being dis-
tributed. It is essential, therefore, to have an
integrated water program wherein the water author-
ity takes action to correct deficiencies exposed
by laboratory testing. The procedures should be
formalized in such a way that laboratory infor-
mation can be transmitted to the appropriate
authority and corrective action instituted.
-------
329
E. MICROBIOLOGY OF WATER TREATMENT
AND DISINFECTION
ever increasing need for potable water has led to
reliance on low quality surface water sources which neces-
sitate intensified and improved treatment practices.^
Waterborne disease outbreaks resulting from insufficient
treatment of microbiologically contaminated water have been
well documented. Therefore, in order to prevent the trans-
mission of diseases through water, it is of the utmost
importance that treatment practices be properly chosen and
carefully applied from the raw source water to the final
product.
This section will evaluate several standard processes
used in the treatment of drinking water [See Table E.
Intro-1] in terms of their effects upon the microbiology of
the water. Emphasis will be placed on enteric bacteria, in
their role as indicators, and on viruses. Estimates will be
made, on the basis of recent pilot studies and experiments,
of the overall safety to consumers achieved through the
application of these respective treatment procedures. Not
included here are microbiologic processes for nitrification
of water from surface sources [See the report of Project
Area II: Advanced Treatment Technology] and for denitri-
fication of groundwater.
1. Storage Reservoirs
The storage of raw water in large, open reservoirs can
provide both quantity and improved quality of the finished
water. Storage reservoirs serve as an important supply
.under drought conditions and also as a safeguard against
polluting substances which may be present in a river. If
toxic materials are spilled or discharged into the river,
the intakes can be closed until the harmful substances have
passed downstream. Provided that the storage reservoir is
of sufficient volume, it will serve as a buffer against any
sudden deterioration of the influent quality. This buf-
fering action is particularly important where contamination
may be intermittent.
-------
330
TABLE E.INTRO-1
SCHEME OF TREATMENT
FOR THE PREPARATION OF DRINKING WATER
Treatment 1
2
3
4
5
6
Reservoir storage
Dune infiltration
Coagulation
Sand filtration
Activated carbon
Disinfection
-------
331
a. Physical and Biological Factors. Surface waters
may be purified by a treatment process consisting of several
stages, the first of which is storage. Storage_immensely
improves the physical, chemical, and microbiological quality
of the source water which is particularly important when
this is a river polluted with sewage effluent. Loss or
turbidity is perhaps the most striking improvement_in, that
particulate matter, along with adhering microorganisms, will
sediment out.
Sunlight has a lethal effect on bacteria and other
microorganisms, and high coliform mortalities have been
shown to occur in seawater during daylight hours (Gameson
and Saxon, 1967). Radiation of the shortest wavelength _
(ultra-violet and .the blue end of the visible spectrum) is
bactericidal and its lethal potency is approximately pro-
portional to its intensity. Except when algal and cyano- _
bacterial blooms interfere, sunlight penetration is enhanced
by the reduced turbidity obtained from sedimentation.
It is well known that higher temperatures increase
microbial metabolism, leading to accelerated die-off rates,
of indicators and pathogens. A rise in water temperature
will also increase predator activity. Certain protozoa are
particularly active bacterial feeders, as can ^seen at a
sewage effluent outfall where ciliates abound and help
reduce the vast numbers of bacteria. Among the somewhat
larger animals, rotifers and members of the genus Daphnia
are important predators of bacteria. The latter frequently
graze on and remove large numbers of E. coll;from reservoir
water [See Section A.2]. The minute Bdellovibrio bacterio-
vorus can grow on gram-negative bacteria by penetrating the
cell wall and multiplying within the cell. It is highly . .
motile, can rapidly clear dense suspensions of gram-negative
bacteria, and can grow equally well on live or_dead cells
(Guelin, et al. , 1967). iBacteriophaaes; (bacterial viruses)
cause lysTi £f bacteria cells, but are only ,able to develop
on actively growing bacteria. Since intestinal bacteria,
Such as E. coll, are not capable of multiplying in reservoir
water, death~~from coliphage infection may be minimal LSee,
Section C.2.a].
Although lethal to some microbes, antibiotics (produced
by actinomycetes and fungi) and toxins (produced by cyano-
bacteria) probably play a minor role in reducing numbers of
enteric bacteria whose populations are, in most cases,
already greatly diminished by the lack of available nu- _
trients, competition, and predation. The stresses exerted
-------
332
by an entrenched and well adapted aquatic community render
the reservoir an unfit environment for enteric bacteria.
According to Holden (1970), "it is impossible to assign
the exact importance of each factor in the self-purification
of water," because of the large number of agents operating
simultaneously. it is obvious that different climates will
have varying influences areas of high temperature com-
bined with sunlight being strongly bactericidal in clear
water.
b- Effect of Storage on Indicator Organisms. Thermo-
tolerant coliforms (i.e., E. coli and other fecal coliforms)
are the prime indicators of fecal pollution [See Section
C.I.a]. E. coli is present in the intestine of man, animals,
and birds and outnumbers the pathogens by millions to one in
sewage. Its reduction can be traced through the various
stages of water treatment, but is largely accomplished
during storage. The following pertains to the effect of
storage as carried out by the Metropolitan Water Division of
the Thames Water Authority, U.K., and special reference will
be made to these storage reservoirs.
Approximately two-thirds of London's water supply is
obtained_from the Thames River and one-sixth from the Lee
River, with the remainder derived from underground sources.
Numbers of E. coli are regularly monitored in both these
rivers^at the intakes to many of the storage reservoirs.
Water is stored in the large reser.voirs of the Thames Valley
for periods of seven to ten weeks depending on demand,
whereas in the Lee Valley the retention period is consid-
erably less. The percentage reductions of E. coli during
storage were computed from quarterly figures for the years
1971 to 1976 (Metropolitan Water Board, 197l-1973b- Thames
Water Statistics, 1976). Though they varied by season,
these losses remained considerable throughout.
During spring and summer, many 100 ml samples collected
from reservoir outlets contained no E. coli, whereas only
three reservoir samples taken in autumn and winter were free
of E. coli. Storage reservoirs in the Thames Valley reduced
! coli bv 95-7 to 99.8 percent in the spring, and by 90.7
to 99.7 percent in the summer. At these seasons, sunlight,
temperature, and biological factors would be exerting their
maximum influence. in the autumn, E. coli reductions during
storage varied from 77.8 to 98.1 percent and in the winter,
from 85.7 to 98.2 percent. It is noteworthy that the lowest
-------
333
reductions for each season occurred in a reservoir which
utilizes mixing jets to prevent thermal stratification.
Storage reservoirs in the Lee Valley exhibited smaller
seasonal fluctuations in E_. coli removal, most likely because
of their shorter retention times as compared to those in the
Thames Valley. However, as in the previous cases, the
lowest removal (84.3 percent) occurred in the autumn; winter
reductions were > 91.8 percent, and spring and summer storage
removed > 95.2 and 93.4 percent, respectively.
The storage-induced decline in bacterial numbers was
studied in a small reservoir in the Lee Valley (Poynter and
Stevens, 1975). The reservoir was filled with a mixture of
water from the Thames and Lee Rivers, and then closed. The
water was sampled weekly during the autumn of 1974 and
showed the greatest decline in E. coli, 95.8 percent, within
the first week. In the following eight weeks, low numbers
of E. coli were isolated, suggesting that feces from birds
and~animals and/or rainfall runoff from reservoir banks were
the contaminating source. When a sample of this reservoir
water was inoculated in vitro with strains of E. coli (re-
at a
cently isolated from the Thames River and from feces.
level of 260 IS. coli per 100 ml, and stored in the dark at
15 to 16°C "the rate of decline was logarithmic, ending
with near extinction (E. coli absent from 100 ml) within 11
days.
Many coliforms present in feces can survive better than
E. coli outside the body. Many are widely distributed in
iTature and may gain access to water from non-fecal sources.
Under certain circumstances, they may multiply on decaying
vegetation, wood, or other materials [See Section C.2.d].
Three weeks after copper sulfate was added to a storage
reservoir to control a spring growth of diatoms, there was
evidence of coliform growth in the secondary slow sand
filter beds (Burman, 1961). Coliforms are usually reduced
considerably, but less so than E. coli, during storage. In
the previously described study of the small storage reser-
voir in the Lee Valley (Poynter and Stevens, 1975), coli-
forms decreased by 85 percent in the first week, but re-
quired two weeks to be reduced by 90 percent (the decline of
E_. coli was almost 96 percent by the end of the first week).
During the remaining seven weeks, coliform reductions fluc-
tuated between 92 and 99.5 percent.
The term fecal streptococci [See Section C.l.d] applies
to the Streptococcus faecalis-faecium-durans group (entero-
cocci) and the S_. bovis-equinus group (i.e. , they all belong
to Lancefield's serological group D). The reported die-off
-------
334
rates of fecal streptococci and fecal coliforms in natural
waters (Geldreich and Kenner, 1969; Kjellander, 1960;
McPeters, et al., 1974) can be summarized as S. bovis-equinus
group > fecal coliforms > S^. faecalis-faeciunTgroup. ~S~.
faecalis is abundant in human feces, whereas S. faecium is
widespread in both man and animals. Thus, in~a storage
reservoir where the water source is a sewage polluted river,
streptococci may be isolated in the. absence of E. coli.
However, when water samples from the Lee River were stored
in clear glass bottles of 2.8 1 capacity, in the dark or in
the light, the observed die-off rates of fecal streptococci
were similar to or greater than those of E. coli (Burman,
£t al'j 1978). in these samples, the major source of fecal
pollution was considered to be treated sewage effluents.
Clostridium perfringens [See Section C.l.e] is present
in feces in much smaller numbers than either E. coli or the
fecal streptococci. It differs from these organisms in that
it produces spores which are resistant to adverse conditions
and, therefore, better able to survive for extended periods.
The presence of C. perfringens in a source water indicates
fecal contamination and, in the absence of coliforms, sug-
gests that the contamination occurred at some remote time.
The die-off rate of these organisms in a storage reservoir
would be slow, although sedimentation may account for
apparent reductions of spores adhered to particulates.
Storage may also act to reduce considerably the number
of^bacteria able to form colonies on yeast extract agar at
37°C (and, to a certain extent, of those which form colonies
at 22 C). The Lee Valley reservoir study (Poynter and
Stevens, 1975) showed a 50 percent decline the first week,
and a 72 percent decline the second week, in colony counts
on plates incubated at 37°C for 24 h. However, due to the
incidence of algal blooms there may be a considerable in-
crease in colony counts at both 37°C and 22°C, which may
occur during or after the algal blooms as was shown in the
Dutch storage reservoir "De Grote Rug" (Kool, 1977). Peaks
of bacterial numbers also coincided with peaks of algal
activity in a Thames Valley reservoir, U.K. (Price and
yaladon, 1970). Thus, under certain conditions, bacterial
increases, rather than decreases, will take place in a
reservoir.
c.
- Effect of Storage on Bacterial Pathogens. All
drinking water supplies should be free from pathogens, such
as those responsible for typhoid, paratyphoid, dysentery,
-------
335
cholera, bacterial or viral gastroenteritis, and infectious
hepatitis [See Section B] . For practical purposes, typhoid
fever, cholera, and infectious hepatitis are regarded as
exclusively human diseases, while the other diseases may be
Sf human or animal origin. The agents of all these diseases
are- discharged in the excreta (feces or urine) of patients
or carriers. Thus, if these .diseases are prevalent within a
population, the causative agents will be present in sewage;
and if this sewage is discharged to a river which is used to
supply a reservoir, the presence of such pathogens can be
expected.
The effect of storage on Salmonella typhi (the agent
of typhoid fever) was investigated some 70 years ago by
Houston. His "Reports on Research Work" (Houston, 1908-
1911) dealt with several experiments, the first of whicn
involved 18 samples of river water, six each from the Thames,
the Lee, and the New Rivers. Each of the 4 1 samples was
inoculated with a single strain or a mixture of different
strains (27 in three of the experiments) of S. typhi, some
of which had been recently isolated from typhoid patients.
The inoculated river water samples were stored in the dark
at temperatures ranging from 8 to 20°C. The numbers of S.
typhi organisms added to the samples ranged from eight to 4O
million per ml, and counts were made at weekly intervals.
The decline was especially rapid in the first week, ,the
reduction amounting to more than 99.9 percent in practically
all experiments. However, S. typhi could still be^recovered
from 100 ml of water, in some experiments, during the eighth
week of storage. Later, Houston carried out further experi-
ments on much larger volumes (1.6 mJ) of Thames River water.
He noted that temperature greatly affected the viability of
S. typhi, survival being considerably longer at low^tempera-
turesTAs a result of his work, he concluded that an
adequately stored water is a safe water."
Experiments were carried out in the U.S. (Jordan, et
al., 1904) in which river, lake, or canal water was inocu-
Ta~ted with S. typhi. The inoculated water was enclosed in
permeable sacs made of celloidin and parchment to allow
dialysis, and these sacs were suspended in flowing_river,
lake, or canal water. Under these conditions, designed to
simulate those in nature, the majority of S. typhi perished
within three to four days. The experimental results of
Thresh and coworkers (described by Holden, 1970) are in
general agreement with those of Houston, especially with
regard to the sudden initial decrease of £. typhi in pol-
luted waters. Though samples contained only five organisms
-------
336
per ml of river water, both £3. typhi and S. paratyphi de-
creased considerably in the first four days and were absent
from 1 1 after three weeks' storage in the dark at 10 to
15°C.
While typhoid fever, which is exclusively a human
disease, rarely occurs in the U.K., Salmonella-induced gas-
troenteritis is comparatively widespread among humans and
animals [See Section B.l.a(i)]. Storage experiments were
carried out by Fennell and coworkers (1974) on the survival
of S. typhimurium in reservoir water; Two outlet samples of
reservoir water (pH 7.3) were tested; one was naturally
contaminated with 25 salmonellae per 1 and the other was
inoculated with S^. typhimurium (isolated from the reservoir
during the previous week), to yield a count of 120 per 1.
Ten 1 volumes of water were stored at approximately 5°C in
the dark. At the end of the first 'week of storage, the
reduction was approximately 50 percent and after two weeks,
75 percent. By the middle of the third week, salmonellae
were difficult to isolate and at the end of the third week,
the count was only one per 1. Storage at the low tempera-
ture of 5°C (at pH 7.3) probably accounted for the compara-
tively small reduction (50 percent) in the first week.
When inlet water with a pH of 5.6 was inoculated with 120
S_. typhimurium per 1, death occurred more rapidly, reaching
a 95 percent reduction by the end of the first week, and a
99 percent reduction after two weeks. After 17 days of
storage, no salmonellae could be isolated. In contrast to
these results with J3. typhimurium, Cohen (1922) demonstrated
that the death rate of J3. typhi was minimal within a pH '
range of 5.0 and 6.4. A change in pH from 5.4 to 3.8 accel-
erated the death rate by almost a hundredfold, while a
similar change in pH on the alkaline side increased the
death rate about four to fivefold. It is clear from these
examples that both pH and temperature are important to the
survival of Salmonella [See Section B.l.a(i)].
Gulls feeding on domestic sewage, either at the sewage
works, at effluent outfalls, or at refuse disposal sites and
subsequently excreting..Salmonella elsewhere, represent a
widespread problem. Muller (1965) collected more than 1,000
fecal samples from gulls in Hamburg around the area of the
biggest sewage works, on bridges and pontoons in the port of
Hamburg, and in the streets of the city; 78, 66, and 28
percent of these samples, respectively, were positive for
Salmonella. In such widely separated areas as Hamburg,
Venice,and Tanganyika, approximately 30 percent of pigeon
feces contained Salmonella, whereas 16 percent of duck feces
-------
337
in Hamburg carried the pathogen (Muller, 1965). During a
thirteen-month survey of a storage reservoir in Yorkshire,
U.K. (Fennel 1/est al. , 1974), Salmonella was isolated from
52 out of 111 water samples from the reservoir, but never
from the inlet water of the catchment area. The number of
roosting gulls was at a maximum in December and the bacterio-
logical .quality of the reservoir water deteriorated markedly
during the winter. There was a close correlation between
the number of gulls and the degree of contamination with
Salmonella, E. coli, and fecal streptococi all of which are
excreted in,large numbers by gulls. There was no correla-
tion between the number of gulls and presence of Clostri-
dium perfringens, since these organisms occur only in small
numbers in gull feces. The presence of large birds, in high
numbers on open storage reservoirs may pose a serious problem
either from direct fecal discharge and/or from rainfall run-
off along contaminated banks.
There have been few reports in the literature of the
isolation of Shigella from river water, sewage, and sewage
effluent; and this has generally been associated with the
organism's poor stability outside the human body. However,
the failure to isolate Shigella from polluted water may
instead be due to unsuitable selective isolation techniques.
Shigeila has a longer, generation time than many of the
competing organisms/ and this complicates the use of an
enrichment medium CSee Section B.1.a(ii)].
Cholera has been practically non-existent within the
U.K. for more than 70 years, but prior to that period epidemics
occurred at intervals. Houston (1909) carried out a series
of 18 experiments on the survival of Vibrio cholerae in raw
river water (six each from the Thames, the Lee, and the New
Rivers), similar to those he conducted with S^. typhi. The
initial inoculations numbered from 70,000 to 13 million per
ml of water, and samples were stored in the dark at tem-
peratures ranging from 7 to 18°C. The cholera vibrios died
out very rapidly; and in all 18 experiments, there was a
reduction of at least 99.9 percent within one week. None
could be isolated from 100 ml of water after three weeks,
and negative results were obtained after two weeks in over
half the experiments. Survival,was longest in samples
receiving the most massive inocula. The survival of a very
small fraction of, organisms, which is.more resistant, is not
uncommon and may be correlated with the initial number of
cells in the inoculum. Houston interpreted his results as
indicating that cholera vibrios were "even less hardy" than
-------
338
typhoid organisms and that the storage time required to
protect reasonably well against typhoid fever was more than
sufficient for cholera [See Section B.l.a(vii)]. Cholera is
endemic in India, and cholera vibrios have been found to
perish only a few hours after entering certain rivers, such
as the Ganges and Jumna (Holden, 1970). This is due to
factors such as high temperature and the presence of an
established predator population sustained by frequent fecal
inputs. Conversely, the survival of enteric organisms is
now known to be more prolonged in pure than in polluted
d* Effect of Storage on Enteroviruses. Reservoir
storage causes a marked reduction in virus levels. From
December 1974 to March 1975, water samples from the river
intakes to the storage reservoirs in the Thames and Lee
Valleys were compared with the outlet water from these re-
servoirs (Slade, 1977). virus numbers in the River Thames
ranged from 12 to 49 plaque-forming units (PFU) per liter
and in the Lee River, 0.8 to 11.8 PFU per liter. Large '
numbers of enteroviruses are frequently isolated during the
Ti Lm0nthS'".the ThamQS River normally has more viruses
££a ,6/ ^S RlVSr' ?hS V±rUS numbers' after storage in the
Thames Valley reservoirs, varied between 0.00 and 9.73 PFU
1*93 J5Sr" J2£ ^^ reSUlt WaS h±gh and no other exceeded
J..y3 PFU. Numbers after storage in the Lee Valley reser-
voirs ranged from 0.00 to 1.00 PFU per liter.
Temperatures of the stored water in the Slade study at
the time of sampling ranged from 5 to 7°C, and the results
v?" kV°m?are?-t0 ^ose of Poynter (1968) who also studied
virus inactivation in water from the Lee River. He found
that at a temperature of 5 to 6°C, more than nine weeks of
storage were required for 99.9 percent inactivation of polio-
virus _type 3; but at temperatures of 15 to 16°C, 95 percent
inactivation occurred in seven days and 99.75 percent in
d^fl* =an 15^JaY? CSee a}so Section B.l.b]. These results,
which are sufficiently similar to the results obtained for
enteropathogenic bacteria, uphold the value of storage
reservoirs as a highly desirable method of safeguarding
water supplies when the source is a polluted river.
As a result of his research, Houston concluded that 30
days^of storage, prior to filtration, should suffice to
provide a safe water. However, this is not always practical
and it is now generally agreed that inflow-outflow arrange-
ments which ensure at least ten days of retention in the
-------
339
storage reservoir are practical methods of eliminating
between 75 and 99 percent of the excremental organisms"
(Holden, 1970). In temperate regions, the greatest reduc-
.tions occur in spring and summer, and the least reductions
in autumn and winter, especially during prolonged cold
weather. At a water temperature of 4 to 8°C, a retention
time of about 75 days was required for a 99.9 percent reduc-
tion of enteric viruses (Kool, 1979).
e. Recreational Use of Reservoirs. In some developed
countries, there is now an increasing awareness of the
amenities (connected.; with open storage reservoirs. They may
be stocked with fish, such as trout, and so attract fisher-
men. They may also be used for boating and by sailing
clubs. Power-driven boats (apart from those involved in
rescue) should be excluded to avoid any accidental spillage
of oil or other toxic substances. Swimming (or any total
immersion) should likewise be prohibited. Large numbers of
people visiting a reservoir would dictate the need for
adequate nearby toilet facilities. Provided that these and
other safeguards are strictly carried out, storage reser-
voirs may serve as areas for recreation, as well as for
storage and purification.
f. Summary. All the evidence indicates that storage
reservoirs are invaluable as a primary stage in the treat-
ment of fecally polluted river water. However, the advan-
tages of storage may be lessened (or even lost) where the
water is subsequently contaminated by waterfowl, such as
gulls, or is used as a recreational facility without proper
precautions.
2. Coagulation
This section will present and discuss data on the
removal of bacteria, enteric viruses, and parasitic cysts
by coagulation with aluminum and iron salts followed by
sedimentation. Most studies cited here report effects of
coagulation with raw surface water or secondary effluent,
but others, in which raw wastewater' was tested, were also
included in order to provide more complete coverage.
Coagulant dosages, stated for compounds in which the
water of hydration obviously was present, were reduced
accordingly; otherwise, the concentrations are shown as
reported. For example,- in studies where aluminum sulfate
had been the coagulant used, the concentration indicated
-------
340
was reduced for this report by a factor of 0.486 (18 waters
of hydration) to convert to A12(SO4)3< In cases where
ferric sulfate or ferric chloride was used without desig-
nation of the water of hydration, the concentration reported
in the study was used here. However, some studies may have
neglected to include the waters of hydration in reporting
concentrations of ferric compounds, in which case, the
actual concentration of active compound would have been less
than that reported here.
a. Removal of Bacteria. Data on the removal of bac-
teria by coagulation and sedimentation have been taken
primarily from early studies since this subject has drawn
little attention in the current literature [See Table E.2-1],
Gehm (1944) obtained reductions in colony counts of 97
percent from raw sewage to which 60 mg per 1 of Fed. had
been added, although only 77 percent of B. coli (now E^
coli) were removed. When 40 mg per 1 of FeCl_ was applied,
colony counts were reduced by about 20 percent, although 50
percent of the E_. coli were removed under similar conditions,
Early studies (Streeter, 1927, 1929) reported
average bacterial removals obtained at full-scale water
treatment plants on the Ohio River and Lake Erie [See Table
E.2-*!]. Reductions at the Ohio River plants averaged 83
percent for both colony counts and _E. coli, but only 46
percent for colony counts and 76 percent for ft. coli at the
Lake Erie plants. Turbidity was reduced by about the same
percentages as were bacteria. However, only three of the
seven Ohio River plants employed flocculation, of which only
one allowed more than 6 min of detention time (20 min). Of
the four Lake Erie plants, only two used flocculation.
Later studies (Cummins and Nash, 1978) made at the
Cincinnati, Ohio treatment plant showed that when 12.6 mg
per 1 of A12(SO^)3 were added to Ohio River water and then
allowed to settle for 48 h in open reservoirs, 97 percent of
total coliforms were removed [See also Section E.lj. When
the effluent water was coagulated with 6.8 mg per 1 of
Fe^tSO^K and allowed to settle for 4 h, only 42 percent of
the initial 2,400 coliform bacteria per 100 ml were removed.
The latter procedure more nearly typifies actual treatment
plant operations, but the former offers evidence of the
advantages to be had from longer sedimentation with its
increased opportunities for bacterial die-off.
Results vary widely in the few available reports on
removal of bacteria by coagulation and sedimentation. The
-------
TABLE E.2-1
REMOVAL OF BACTERIA BY COAGULATION WITH ALUMINUM SULFATE AND SEDIMENTATION
Hater
Source
Indicator System
(initial cone, per ml)
Conditions of Coagulation
Coagulant Turbidity Temp
Dose (mg/ (TU) (*C)
. Final
PH
Removal l£pe b Reference and Remarks
study
Indi- Tur-
cator bidity
1 or ppm) (%)
River
River
River
River
River
River
River
River
Colony count (2951)
JB. coli, index (3526)
Total coliform (840)
Colony count (1675)
Total coliform,
coliform index (5943)
Total coliform, coli?
form index (2.5 x 10 )
Total bacterial
count, MPN (NS)
Total bacterial
10.5
10.5
12.6
20
20
20
25
25
168
168
14
8
8
NS
40-135
i
140-255
14
14
22
NS
NS
NS
5
15
NSC
NS
NS
7.6
7.6
7.6
6.7-7.4
6.7-7.4
83 90 . F
83 90 F .
97 96 F
61 40 P
74 40 P
81 NS P
98.7 1-5 TU L
99.3 1-5 TU L
Streeter (1927) avg. of
plants on Ohio River
Streeter (1927) avg. of
plants on Ohio River
Cummins & Nash (1978) 48
sedimentation
Mallmann & Kahler (1348)
Mallmann & Kahler (1348)
Mailman & Kahler (1948)
Chang, et al. , (1958a)
*
Ghana, et al.. (1958a)
7
7
h
count, MPN (NS)
U)
*»
-------
TABLE E.2-1 -- Continued
Water
Source
River
River
River
River
River
" Inc
(initj
Total
count
Total
Total
Total
licator System Conditions of Coagulation Removal gf?b ^f erence
Lai cone, per ml) ,,, fatuay
and Remarks -^
Coagulant Turbidity Temp. Final Indi- Tur-
Dose (mg/ - (TU) (°C) pH cator bidity
1 or ppm) (%)
bacterial 25
MPN (NS)
coliform, MPN (NS) 25
coliform, MPN (NS) 25
coliform. MPN (NS) 25
Colony count (475- NS
18,000; 5500 avg.)
16-240 25 6.7-7.3 99.8 1-5 TU L
40-135 5 6.7-7.4 98.8 1-5 TU L
140-255 15 6.7-7.4 99.4 1-5 TU L
16-240 25 6.7-7.3 99.8 1-5 TU L
10-800 NS NS 89 99 L
225 avg.
Chang/
Chang,
Chang,
Chang,
et
et
et
et
al. ,
al. ,
al. ,
al. ,
(1958a)
(1958a)
(1958a)
(1958a)
Calvert (1939) avg. 9 tests,
15 min sedimentation
Lake Colony count (1358) 12.1 40 13 NS 46 72
Lake E. coli, index (818) 12.1 40 13 NS 76 72
Streeter (1929) avg. of 4
plants on Lake Erie .
Streeter (1929) avg. of 4
plants on Lake Erie
aTurbidity removal is expressed as a percentage unless otherwise stated.
bL - laboratory; P - pilot plant; F - full scale.
CNS - not stated.
-------
343
widely quoted laboratory study by Chang and coworkers (1958a)
reported bacterial removals of 99 percent or more, but these
figures have not been substantiated in other reported data.
Streeter (1927, 1929) reported bacterial removals of 46 to
83 percent at full-scale treatment plants. Cummins and
Nash (1978) reported total coliform removals of 42 percent.
The lack of recent reports on the removal of bacteria
by this process is probably due to the widespread reliance
on prechlorination dating from the late 1930's and early
1940's, with the result that attention was diverted from the
capability of coagulation for removing bacteria. With the
recent interest in decreasing prechlorination of water to
reduce trihalomethane concentrations, the potential that
coagulation offers for removal of organisms takes on new
importance. Additional studies are needed to determine the
effects that variations in pH, coagulant type and dosage,
temperature, and other water characteristics would have on
the removal of different bacteria.
b. Removal of Viruses. Poliovirus 1 and coxsackie-
virus A2 are the only enteric viruses for which data on
removal by coagulation and sedimentation are available [See
Table E.2-2]. Several studies conducted at large pilot
plants reported relatively high reductions from influent
containing low virus titers CTable E.2-23. These studies
indicate that viruses present in relatively low concentra-
tions should be reduced by approximately 90 percent or more
with coagulation. Chang and coworkers (1958a,b) reported
that variations in the type of coagulant and temperature had
minor effects on the removal of coxsackievirus A2, although
an increased coagulant dose did increase removal. Data from
other investigators (Guy, et al., 1977) support the obser-
vation that there is little difference in virus removals
between the use of Fe2(SO4)3, FeCl3, or A12(SO4)3.
Although the coagulation process is relatively non-
specific in its capacity to remove small particles, prudence
dictates that generalizations based on tests made with only
poliovirus 1 and coxsackievirus A2 be validated with studies
using a spectrum of the enteric viruses.
c. Removal of Parasitic Protozoa and Metazoa. Cysts
of protozoa and eggs of metazoa are heavier than bacteria
and viruses and will settle in quiescent water. Helminth
eggs are quite heavy and settle rather rapidly. The data
-------
TABLE E.2-2
REMOVAL OP ENTERIC VIRUSES BY COAGULATION AND SEDIMENTATION
Coagulant
Type of
Hater
Virusa
Conditions of Coagulation Removal Type
. Reference
Type initial Coagu- Turbidityb Temp. pH virus Tur- "^ aUU RemdiK"
cone. lant («C) (%) bidityc
PFU/ml Dose Start End
(mg/ 1
or PPM)
A12(S04)3
A12(S04)3
A12(S04)3
A12(S04)3
A12(S04)3
A12(S04)3
Distilled
River
River
River
Distilled
Activated
sludge
effluent
Polio 1 3-7xl04 10
(Sabin)
Coxsackie NS 25
A2
Coxsackie NS 25
A2
Coxsackie NS 25
A2
f
Coxsackie 2.25xl05 40
A2
Polio 1 596 56
(Vaccine)
50 mg/1 Room NSe 6.8 90 97 L
clay
16-240 25 NS 6.7-7.3 99 1-5 TU L
140-255 15 NS 6.7-7.4 95 1-5 TU L
40-135 5 NS 6.7-7.4 96 1-5 TU L
0.4 ml 25 6.2 6.2 86 NS L
2
/U10 FTU 29 7.1 6.8 63 A^2.5FTU P
Thorup, et
al., (1970)
Chang, et al.
(1958a)
Chang, et al.
(1958a)
Chang, et al.
(1958a)
Chang, et al.
(195Bb)
Wolf, et al.,
= .44 - 1
t
t
9
(KFVISED FFOM SPROUL, 1979)
u>
-------
TABLE E.2-2 Continued
Coagulant
A12(S04)3
A12(S04)3
A12(S04)3
ff *T «J
A12(S04),
& «*O
A1-(S04)3
+ Calgon
WT 3000
FeClg
PeCl3
Type of
Water
Distilled
Distilled
Distilled
Activated
sludge
effluent
Activated
sludge
effluent
River
Distilled
Virus3
Type
Coxsackie
A2
Coxsackie
A2
Coxsackie
A2
Polio 1
(Vaccine)
Polio 1
(Chat)
Coxsackie
A2
Coxsackie
A2
Initial
cone.
PFU/ml
2.25xl05
2.25xl05
,2.25xl05
113
130
NS
4.5xl03
Coagu-
lant
Dose
(mg/ 1
or PPM)
60
80
100
NS
76 +
0.2
15
20
Conditions of Coagulation Removal
b
Turbidity Temp. pH Virus Tur-
(°C) (%) bidityc
Start End
0.4 ml 25 6.2 6.2 96 NS
sio2
0.4 ml 25 6.2 6.2 97 NS
Si°2
0.4 ml 25 6.2 6.2 99 NS
sio2
3.4 FTU 22 7.3 6.9 99.7+ 1.5 FTU
4.6 24 7.3 6.9 95 10
(avg.)
5-10 25 NS 8.1-8.4 95 0.1 TU
0.4 ml 25 6.2 6.2 97 NS
sio2
Type , Reference
Study and Remarks
1
L Chang, et al.,
(1958b)
L Chang, et' al.,
(1958b)
L Chang, et al.,
(1958b)
P Wolf, et al.,
(1974) ~~
A1:P = 7.1
P Anon (1977)
A1:P = 1.7:1
L Chang, et al. ,
(1958a)
L Chang, et al.,
(1958b)
Or
-------
TABLE E.2-2 Continued
Coagulant
FeCl,
3
FeCl,
J
FeCl,
Fe0(SO,.),
2 43
Fe2(S04)3
Type of Virus3 Conditions of Coagulation Removal
Type Initial Coagu- Turbidity Temp. pH Virus Tur-
conc. lant CO (%) bidityc
PFU/ml Dose Start End
(mg/ 1
or PPM)
3f
Distilled Coxsackie 4.5x10 40 0.4 ml 25 6.2 6.2 98 NS
A2 Si02
Deminer- Polio 1 1.4xl06 60 25 & 500 15-17 5-8 NS 48-99.7 NS
alized (Mahoney) mg/1 clay
Deminer- Polio 1 10^*^ ~ 66 45-178 13-21 7- 6.6-6.9 99.7- 0.6-
alized (Mahoney) 10°*J/ 7.7 99.999 3.8 TU
i
River Polio 1, 9. 2x10 ' 40 NS NS NS NS 99.8 NS
2, 3
(Glaxo
oral
vaccine)
River Naturally .004^ 40 NS NS NS NS .No NS
occurring virus
(mainly detected
coxsackie > 88.3%
B3 & B5)
Type - Reference
Study and Remarks
L Chang, et al. ,
(1958b)
P Foliguet &
Doncoeur
(1975) poorer
removal with
low clay
P Foliguet &
Michelet
(1969)
P Guy, et al. ,
(19771
P Guy, et al. ,
(1977T~"
u
a\
-------
TABLE E.2-2 ~ Continued
Footnotes;
Experimentally inoculated, except as indicated.
Expressed as turbidity units (TU) except where indicated.
°Expressed as a percentage except where indicated.
L - laboratory; P - pilot plant.
eNS - not stated.
LD_.j per ml.
^MPN cytopathic units per ml.
hTCD50 per ml.
""Infectious particles (single concentrated inoculum).
'infectious particles per ml.
ui
*»
-4
-------
348
presented in Table E.2-3 indicate that the removal of para-
sitic protozoan cysts by coagulation and sedimentation
should be 90 percent or more under optimum conditions.
d. Conclusions. The few reports available on removal
of bacteria indicate that reductions will vary from 50 to 97
percent with coagulation carried out on a plant scale.
Removal of parasitic cysts should exceed 90 percent, although
this also is based on only a few reports. Additional
studies, therefore, are needed to verify the removal,effi-
ciencies obtained with bacteria and parasitic cysts, espe-
cially as these are affected by variations in pH, coagulant
type and dosage, temperature, and different quality waters.
The application of sufficient quantities of either
ferric or aluminum salts should result in 90 to 99.999
percent reductions in enteric viruses. However, these
predictions are based only on results from coagulation
studies with poliovirus 1 and coxsackievirus A2 and should
be confirmed by tests made with other representative enteric
viruses.
3. Sand Filtration
One of the treatment processes often used for water
purification is filtration through sand, soil, or dunes
(Baker, 1934; Burman, 1962; Burman, 1978; Graveland and
Hrubec, 1976; Kool, 1979; Robeck, et. a.1. , 1962). Filtration
is a process whereby raw water is passed through a porous
matrix to remove suspended and colloidal impurities and,
frequently, to change the chemical characteristics of the
water and reduce the levels of microorganisms present [See
also Sections E.3 to 4 and G.2].
Removal of constituents from the water is accomplished
by a combination of physical and biological processes, the
most important of which are mechanical straining, sedimen-
tation, adsorption, electrostatic binding, and microbial
activity. Filtration through sand, as it is practiced today
in drinking water treatment, can be divided into three
types: rapid sand filtration, slow sand filtration, and
dune or bank infiltration.
James Simpson is acknowledged as the first to introduce
slow sand filtration into Europe, at the Chelsea Waterworks
Company (one of the suppliers of London's drinking water at
the time) in 1829. A cholera outbreak [See Section B.l.a(vii)]
-------
TABLE E.2-3
REMOVAL OP PARASITIC CYSTS BY COAGULATION WITH ALUMINUM SULFATE AND SEDIMENTATION
Type of Water
Gravel pit
water
Gravel pit
water
Gravel pit
water
Gravel pit
water
Clear water
Clear water
Clear water
Activated
sludge
effluent
NS
Organism
Species Initial
Cone.
(cysts/ml)
Giardia NS
muris
Giardia NS
muris
Giardia NS
muris
Giardia NS
muris
Entamoeba 135
histolytica
Entamoeba 150
histolytica
Entamoeba 150-230
histolytica
Entamoeba NS
histolytica .' .
Entamoeba , 1
histolytica
Conditions of Coagulation
Coagulant Tur- Final Cysts (%) Tur- Studj
Dose (mg/ bidity pH bidity
1 or ppm) (TU) (TU)
5 9 6.5 96 3.5 L
10, 25, 9 6.5 > 99.4 for 0.5-1 L
50 each dose
25 NS 5 58 0.5 L
25 NS 6-9 > 90 0.2 L
30 "high" NS 33 NS L
40 "high" NS 40 NS L
60 "high" NS 55 NS L
1000 NS NS Effective NS L
removal
(cultures
were neg)"
NS NS NS 99.2 NS P
Reference and Remarks
r
Arozarena (1977)
. . -
Arozarena (1977)
Arozarena (1977)
Arozarena (1977)
Spector, et al., (1934)
Spector, et al. , (1934)
Spector, et al. , (1934)
(avg. of three tests)
Cram (1943)
Anon. (1944), as cited by
Logsdon, et al., (1978)
-------
350
in Germany in 1892, involving water from the Elbe River,
demonstrated the efficiency of this treatment in removing a
pathogen. The outbreak occurred in Hamburg, where water of
the Elbe was drunk untreated, but not in nearby Altona,
where the water was drunk only after slow sand filtration.
This section discusses the physical and biological
processes that are involved during sand filtration, the
problems that may arise during operation, and the capacities
of such a treatment for removing organisms. The concept of
this presentation is based on a paper by Huisman and Wood
(1974).
a. Physical Processes. Particles of suspended matter
too large to pass through the interstices between the sand
grains are removed by the action of mechanical straining.
This process takes place almost entirely at the surface of
the filter, where water first enters the pores of the filter
bed. As filtration continues, pore openings become smaller
from the deposition of suspended matter, enabling straining
of progressively smaller particles. A mat eventually de-
velops as more material accumulates on top of the filter
bed, and this also increases the efficiency of the straining
process. When flocculation occurs within the filtration
bed, the aggregated particles may be retained at greater
depths of the bed. Bacteria, viruses, other microorganisms,
colloidal material, and other substances smaller than the
pore spaces are not, however, removed by mechanical strain-
ing.
Sedimentation is largely responsible for the removal
of colloids, small particles of suspended matter, and bac-
teria. Here, the interstices between the sand grains serve
as minute sedimentation basins, along the sides of which
suspended particles settle. Smaller and lighter particles
are only partly removed, although flocculation accompanying
further downward movement of the water will slightly in-
crease sedimentation efficiency with depth. Truly colloidal
matter cannot, however, be removed in this way.
Adsorption can take place in either of two ways. The
simplest is when suspended particles collide with and adhere
to the sticky gelatinous coating formed on the sand grains
by previously deposited bacteria and colloidal matter. Of
greater influence is the active promotion of adsorption by
physical attraction between two particles of matter (Van
der Waals forces), and still more important is the electro-
static attraction between opposite electrical charges
(Coulomb forces) [See Section A.l.e].
-------
351
b. Biological Processes. Biological activity is the
action of microorganisms living on and in the filter bed
which produce chemical and physical, as well as biological,
changes in the quality of water. The protozoa, metazoa, and
bacteria inhabiting the filter bed include many organisms
(often in quite high numbers) not commonly detectable in the
raw water [See Section A]. The intensity of their activity
is a function of infiltration rate; that is, the slower the
flow, the longer the detention and the more important
becomes the role of the filter bed flora in the treatment
process. Biological activity, therefore, counts heavily in
slow sand filtration and artificial recharge. In these
systems, the greatest biological activity occurs near the
filter surface (Kool, 1976). Their influence on the physical,
chemical, and biological properties of the water arises
mainly from activities of nutrient utilization and predation.
(i) Nutrient Utilization. Many bacteria and fungi in
the filters are able to derive needed energy and nutrients
from the numerous organic compounds, of various origins,
that are dissolved in the raw water. These compounds are
incorporated into cell mass, degraded to intermediate
compounds, or mineralized entirely to carbon dioxide, water,
and inorganic salts. Some organic compounds are less
readily utilized than others by microorganisms in the filter
bed. On the other hand, some bacteria can adapt rather
quickly to new compounds* to which they were not previously
exposed. This has been demonstrated with various phenols.
Within a few days after exposure to 8 to 10 mg per liter of
phenols added to the raw water, organisms in the slow sand
filter had adapted so as to be able to metabolize and remove
all of the phenol (Metropolitan Water Board, 1971-1973a).
Organic compounds- present in raw water are .usually pollutants
and action upon them by the filter bed flora inevitably
results in improved water quality (Zwaagstra, 1978).
(ii) Predation. The development of a large mixed
microbial flora in response to organic,nutrients present in
the raw water results in the growth of predators. The
antagonism between predators and nutrient utilizers does
not, however, reduce the activity of nutrient utilizers.
On the contrary, it is more likely to maintain an ecological
-equilibrium that stabilizes the activity of all the organisms
in the community (Waksman, 1937).
-------
352
Predators may include bacteriophages, the bacterium
Bdellovibrio bacteriovorus, protozoa (including amoebae,
flagellates, and ciliates), and metazoa such as rotifers and
the larger oligochaetes (Metropolitan Water Board, 1971-
1973a; Reijnen, 1973) [See also Sections A.2, B.3, and
E.I]. Bacteriophages, like other viruses, require living
and growing host cells, in which to replicate, and this
results in the lysis of the host cell. The tiny bacterium
Bdellovibrio makes use of its very active motility to embed
itself in the host bacterium where it multiplies, causing
lysis of the host cell. It has been found in sewage effluent,
surface waters, filter beds, and soil (Reijnen, 1973). It
is not known whether protozoan predation significantly alters
bacterial numbers and activity in slow sand filters (Dor-
nedden, 1930; Reijnen, 1973).
All of these predators feed primarily on the sapro-
phytic microorganisms growing in the filter bed, but also
will feed on intestinal bacteria and other organisms of
public health significance that may be present and thereby
improve the bacteriological quality of the raw water.
(iii) Effect of Temperature. Every microorganism
grows best at its own optimum temperature [See Section B.3]?
mixed communities in an aquatic environment probably exhibit
optimum activity at around 20 to 30°C. At higher tempera-
tures, there is less diversity of species, whereas lower
temperatures retard metabolic rates and hence, growth and
activity. It follows that during prolonged cold weather,
slow sand filters will be less effective, as is evidenced in
temperate regions where the chemical and bacterial quality
of the filtrates changes with the seasons. This suggests
that slow sand filters would be less effective in countries
with extremely cold and prolonged winters, not* to mention
problems of cleaning beds during freezing weather.
Temperature exerts a great influence over the fate of
pollutants, as in the case of ammonia. Ammonia can be
utilized by some bacteria as an energy source. Nitro-
somonas oxidizes ammonia to nitrite and Nitrobacter oxidizes
nitrite to nitrate [See Section A]. This process is depen-
dent on temperature and slows rapidly below 6°C (Huisman,
unpublished data) even after only a short period of exposure;
for example, when beds are drained for cleaning during a
cold night, the ammonia oxidizing capacity takes several
weeks to recover.
-------
353
c. Rapid Sand Filtration. Flow rates during filtra-
tion are usually measured linearly as tlie vertical flow of
water from the top, down through the sand. Rapid sand
filters are ordinarily filled with medium to coarse sand
(grain size 0.5 to 1.5 mm or larger) and have vertical flow
rates of 5 to 10 m per h for free surface filters, and up to
approximately 20 m per h for pressure filters enclosed in
watertight steel cylinders. Impurities in the water are
deposited deep in the bed of a rapid sand filter because of
the high flow rates and the large size of the sand grains.
The rapid clogging that results from the high flow rates
necessitates cleaning at intervals of hours to a few days
(Graveland and Hrubec, 1976).
d. Slow Sand Filtration. With slow san'd filtration,
water is passed down by gravity through a constructed layer .
of fine sand (usual grain size of 0.15 to 0.4 mm). Fil-
tration rates vary from about 0.1 to' 0.4 m per h, although ,
rates of up to 0.8 m per h have been used experimentally
(Metropolitan Water Board, 1971-1973a). After a few weeks
to a few months of operation (depending on the quality of
the raw water), particulates that have accumulated on and in
the bed will begin to reduce the flow so that filtration
rates cannot be maintained. However, because the sand is
fine, clogging takes place only at the very top of the sand,
and cleaning is accomplished simply by removing the top 1 to
3 cm of sand.
' e. Artificial Recharge. With artificial recharge,
water is passed through naturally occurring dunes or river
banks at flow rates lower than those obtained with slow sand
filtration. For example, in the Netherlands, water is
infiltrated through dunes at velocities of up to 0.2 m per
day and through river banks at rates of 0.1 to 0.5 m per day
(Graveland and Hrubec, 1976). At times, detention of the
infiltrated water may be very prolonged, and anaerobic
conditions may develop, leading to the microbial production
of off flavors and odors. As with slow sand filtration,
particulate matter gradually accumulates on and in the
filter bed, reducing flow rates; hence, beds must be cleaned
periodically (every two to five years for dune infiltration)
by the same procedure carried out for slow sand filters.
f. Characteristics of the Raw Water. The biological.
nature of dune, bank, and slow sand filtration implies that
-------
354
some organic content is required in the water to encourage
the^necessary bacterial population to bring about purifi-
cation. The methods are, therefore, unlikely to be satis-
factory with waters that contain no organic matter and where
all turbidity is due to mineral matter. Raw waters high in
turbidity due to suspended mineral matter and low in dis-
solved organic substances are unlikely to be much benefitted
by these processes. Neither is it advisable, in the case of
dune and slow sand filtration, to filter highly turbid
waters without some preliminary treatment to avoid the need
for frequent cleaning. Dune and slow sand filtration can
best be employed in a dual arrangement of preliminary fil-
tering through a conventional rapid sand filter. The higher
quality water obtained from this primary filtration extends
the cleaning interval required for subsequent slow sand
filters and should enable faster filtration rates without
significant silt penetration.
9- Algal Growth. Algae can cause filtration problems.
Growth of unicellular algae in storage reservoirs [See Sec-
tion E.I] results in an increase in particulate matter which
has to be removed either on the primary or secondary filters,
or both. From the filtration management standpoint, this
requires more frequent cleaning and backwashing of filters,
particularly the primary ones. Algae can also grow in
uncovered filters or filter beds which are exposed to light,
especially filamentous algae during the summer months (Metro-
politan Water Board, 1967-1968). Even when a good quality
water is infiltrated, .-algal, growth is supported if the
water contains a high phosphate load, especially with a
rapid infiltration rate (Hrubec,1975). Where filamentous
algae are a problem, they must be removed promptly when a
filter is drained for cleaning, lest they begin to compost
and raise temperatures sufficiently to support growth of
undesirable organisms such as E. coli (Metropolitan Water
Board, 1969-1970a). ~
The temperature and the phosphate content of raw water
usually are not subject to control, but all algal growth on
filters can be prevented by excluding light. A roof or
shade may suffice; however, a completely enclosed building
over the filters offers the further advantages that it
permits the filters to be cleaned even during extremely
cold or wet weather and that it excludes birds that might
pollute the filters (Graveland and Hrubec, 1976; Metro-
politan Water Board, 1969-1970b and 1971-1973a) [See Section
F.l.b].
-------
355
h. Maintenance of a Steady State. It should be clear
from the foregoing that dune, bank, and slow sand filtra-
tion are unique ecological systems of which the microbial
types and distribution are subject to change with changing
conditions, especially changes in filtration rate and raw
water quality.
Sudden changes in filtration rates can cause deterio-
ration in efficiency [See also Section G.2]. This is because,
with very slow filtration, most microbial degradation and
removal of nutrients takes place in the upper layers; whereas,
with faster rates, nutrients penetrate deeper and conse-
quently, microbial activity occurs at greater depths. Any
abrupt rate change would remove nutrient deposition away
from access by the established filter bed flora.
Another cause of perturbation is leaving the filter bed
full of water when it is not in operation. At very slow
flow rates, or when there is no flow, the demand inherent
to the system can use up all available oxygen, causing a
drastic change in the microbial flora and resulting in the
production of some objectionable compounds.
When a bed is drained for cleaning, the bacterial
micro-environment changes completely. Instead of receiving
a steady flow of water containing nutrients, the interstices
of sand become filled with air. Bacterial metabolism becomes
more oxidative, and bacterial gums and other attachment
substances are used as nutrient sources. Bacteria that
previously were attached to the sand grain surface tend to
be released so that when filtration recommences, they are
washed out into the filtrate (Metropolitan Water Board,
1969-1970a). The longer the interruption in filtration and
the higher the temperature, the greater the contamination of
the filtrate. It is important, therefore, that beds be kept
out of use for cleaning for as short a period as possible,
especially in the summer months, and that former filtration
rates be resumed as soon as possible. Beds returned to use
at their maximum filtration rates within 24 h show no deterio-
ration in efficiency (Metropolitan Water Board, 1971-1973a).
i. Removal of Microorganisms by Dune, Slow, and Rapid
Sand Filtration. Sand filtration of water removes micro-
organisms by physical and biological processes as has already
been discussed. Filtration and adsorption were studied to
determine their influence on virus retention (Goldschmid,
et al., 1972; Lefler and Kbtt, 1974). Adsorption on sand
-------
356
was found to be strongly enhanced in the presence of
cations, such as those of calcium and magnesium, and. fresh
sand was demonstrated less retentive of viruses than con-
ditioned sand. Viruses could be eluted from the particles
by raising the pH or by reducing the ionic strength of the
water (Goldschmid, et a^., 1972; Lefler and Kott, 1974;
Nestor and Costin, 1971; Wellings, et. a^. , 1975) [See Section
A.I.6].
A dune infiltration system in the Netherlands was
examined for its capacity to remove bacteria and viruses
(Hoekstra, 1978; Kool, 1978). Tests in which raw surface
water was infiltrated through sand dunes produced the fol-
lowing results: > 99.99 percent of the colony count (at 22
and 37°C), > 99.9 percent of coliphages, and > 99 percent of
thermo-tolerant and total coliforms were removed. The
somewhat lower percentages for coliphage and coliform removal,
compared to that of the colony count, were due to a rela-
tively small number of coliphages and coliforms present in
the raw water before filtration. None of these organisms
were detected in samples of 1,000 and 100 ml after filtra-
tion. Therefore, it is likely that coliphages and coliforms
would exhibit the same high removal rates as those obtained
for the colony counts. These high removals of > 99.99
percent are in agreement with Robeck and coworkers (1962)
who observed that seepage through unsaturated sand reduced
poliovirus levels by > 99.99 percent. The capacity of slow
sand filtration to remove viruses was investigated by Robeck
and coworkers (1962) who obtained a 96 percent removal of
poliovirus. Poynter and Slade (1977) also showed that this
process was efficient for removing enteroviruses.
Factors that adversely affect removal of microorganisms
during sand filtration are low temperatures, high flow
rates, insufficient sand depth, and filter immaturity. The
importance of temperature was demonstrated by Burman (1962)
who reported that during cold weather, only 41 percent of
13. coli and 88 percent of total coliforms were removed by
slow sand filtration. -Poynter and Slade (1977) reported that
the lowest poliovirus reduction of 98.25 percent was obtained
at 5°C and at a flow rate of 12.0 m per day; whereas, at a
standard rate of 4.8 m per day and at 11°C, poliovirus was
reduced by 99.999 percent [See also Sections B.l.b, B.l.d,
and C.2.b].
Bacteria are not retained quite as efficiently as are
the enteric viruses in slow sand filters. Total coliforms
were reduced by 96.3 to 99.5 percent, and slightly lower
-------
357
reductions were obtained for 1C. coli and colony counts at
22°C (reductions of the colony count at 37°C were lower
yet). Results from a 12-year study of slow sand filtration
showed that reductions in total coliforms and colony counts
varied from 70 to 98 percent when an infiltration rate of 5
m per day was used (Hoekstra, 1978) [See also Sections C.I
and D].
Only marginal reductions of microorganisms are possible
with rapid sand filtration. Robeck and coworkers (1962)
showed that not more than 50 percent of polioviruses could
be removed with this step, and in other studies removal of
enteric viruses and bacteriophages fluctuated between 0 and
98 percent (Berg, et. aL. , 1968; Guy, et a.L. , 1977).
Less information is available on the removal during
infiltration of pathogenic protozoa such as Entamoeba histo-
lytica and Giardia lamblia, and metazoan parasites such as
Ascaris lumbricoides and Schistosoma. Results of Baylis
(1936) and a review in the report of the Safe Drinking Water
Committee (1977) indicate that most of these organisms can
be removed by sand filtration, although Rivas (1967) re-
ported that, in some cases, sand filters do not remove
Schistosoma cercariae. Nevertheless, mechanical means of
clarifying water, such as sand filtration, are important as
antiparasitic measures, especially when one considers that
the protozoan cysts are resistant to usual residual chlorine
concentrations maintained in the distribution system [See
Section B.l.c].
j. Conclusion. When the available data on the various
sand filtration processes are compared, it becomes clear
that artificial recharge in the ground, through river banks
and in sand dunes, is one of the most effective treatments
for removing microorganisms from raw water. Slow sand
filtration is also well suited for removing enteric viruses
and is slightly less so for bacteria.
The efficiency of water purification in slow sand
filters is a function of raw water quality and is adversely
affected by low temperatures, excessive flow rates, inade-
quate sand depths, and immaturity of the filter. The ^
mechanisms by which bacteria and viruses are removed in the
sand are not yet clear, but both organisms seem to respond
in the same way to the cleaning process, whereby the viruses
are removed more efficiently than the bacteria. Rapid sand
filtration is not efficient enough to serve alone for the
-------
358
removal of viruses and bacteria, thus, it should be used
only in combination with other treatments, such as slow sand
filtration and flocculation.
Areas with the needed soil configurations or the requi-
site land area and sand to build filters, would find arti-
ficial recharge or slow sand filtration a very simple,
practical, reliable, and inexpensive method for removing
microorganisms from raw water. The level of efficiency
falls, however, in cold climates.
4. Activated Carbon Filtration
Activated carbon filtration has proved to be extremely
useful for the treatment of drinking water and is almost
indispensable with water from surface sources. However, the
microbiological quality of water treated by activated carbon
filtration has often failed to meet legal requirements
because of high colony counts measured by the German Stan-
dard Procedures. Bacteria can actively contribute to the
improvement of water quality during activated carbon fil-
tration [See the reports of Project Area II: Advanced
Treatment Technology]. The microbiology of activated carbon
filtration for the treatment of drinking water is poorly
understood because testing, to date, has been done accord-
ing to German Standard Methods that were devised to guar-
antee safe water rather than, to characterize the microbial
community.
The present discussion addresses the numbers, kinds,
and metabolic activities of bacteria in activated carbon
filters for drinking water. The numbers of microorganisms
can be determined accurately only by separating them from
the carbon (e.g., with a standard household mixer) and
culturing them on a rich medium incubated at 25°C for seven
or more days to accommodate the long generation times of
some of them. Bacteria, in a study by Klotz and coworkers (1976),
were identified according to Bergey's Manual of Determina-
tive Bacteriology ('Buchanan and Gibbons, 1974). The metabolic
activities of the bacteria were assessed by comparison of
a sterile and a nonsterile activated carbon filter. Biocidal
substances were not effective, so activated carbon filters
were autoclaved and operated with filter-sterilized water.
The activities of the bacteria could not be determined solely
on the basis of CO2 production or Q£ consumption during the
passage of.the water through the filter beds because some
purely chemical oxidation of organic substances adsorbed to
active carbon also takes place.
-------
359
a. Numbers and Kinds of Microorganisms. It proved
impossible in practice to produce, by means such as the use
of silver-impregnated carbon or frequent back-washing of
filters, water that had low enough colony counts to meet
German Standards after activated carbon treatment. Colony
counts in_the activated carbon filters examined -ranged from
10 to 10 bacteria per g of wet material and were stable
for several years if the treatment processes remained
unchanged. Bacterial numbers were a function of the 'treat-
ment processes, especially those affecting the degradability
of organic substances. Bacterial numbers were minimally
affected by the use of different types of activated carbon
and apparently not at all by seasonal variations, including
changes of temperatures between 5° and 25°C.
The.activated carbon filtrate gives colony counts of
about 10 per ml of water, but is otherwise satisfactory
when examined according to the German Standard Methods.
Activated carbon filters can be compared to chemostats in
that the rate of flow has a direct influence on microbial
development.
Although numbers of bacteria found on the activated
carbon and in the filtrate were relatively constant, the
proportions of organisms in each of the two phases are
governed by events occurring on the boundary surfaces of the
water-carbon system. Examination by scanning electron
microscopy reveals that the space on the carbon surface is
only partially occupied by bacteria. In the middle range of
concentrations, the adsorption of bacteria is governed by
the isotherm equation of Freundlich. The equilibrium of
adsorption in a beaker in the laboratory is reached only
after 24 h and is influenced by the ionic composition of the
aqueous phase and by whether the bacteria are dead or alive,
but evidently not by variations of temperatures normally
encountered in drinking water treatment.
A great variety of bacteria, among which* 26 species
from 12 genera could be identified, was found in the acti-
vated carbon filtrate. As expected, most species identified,
was found in the activated carbon filtrate. As expected,
most species identified belonged to the genus Pseudomonas,
which is characterized by a very adaptable metabolism [See
also Sections A.2.a to b, B.l.a(ix), C.2.C, F.2.a, and F.3].
Next most prevalent was the genus Bacillus. Fungi and
yeasts are encountered only rarely and do not seem to be of
any importance.
-------
360
b. Microbial Metabolism on Activated Carbon and Its
Effects on Water Quality. Activated carbon adsorbs organic
substances and desorbs them slowly, so they serve as a
sustained source of nutrients for bacteria associated with
the carbon. The resulting stimulation of bacterial growth
apparently increases with increasing diameter of the carbon
grains, negating the adverse effect of the reduction in
surface area that goes with larger grain size. Adsorption
increases the local concentration of the organic substances,
which enhances the probability that they will be degraded.
Even substances toxic to the bacteria are more susceptible
to biological degradation once they have been adsorbed to
the carbon.
Most of the organic substances removed in treating
water by activated carbon filtration are removed by adsorp-
tion rather than biological degradation. In a recent study,
the total reduction of organic substances (measured by COD,
KMnO4~reaction, and UV-adsorption) by non-sterile activated
carbon filtration was found to be 80 percent (while the
material still had very nearly its initial high adsorptive
capacity) of which the part attributable to bacterial
activity was approximately 4 percent. As the adsorptive
capacity of the carbon declines with use, the part of removal
of organics attributable to bacterial activity rises con-
siderably. The changes of CU and CO2 content that take
place during filter passage are mainly produced by bacteria;
only one-third is caused by purely chemical reactions.
*
Bacteria act most efficiently upon the most easily
degraded substances: 60 percent of the removal of easily
degradable substances (expressed as BOD_) is effected by
bacteria, whereas only 17 percent of the more resistant
substances (expressed as BOD-g) is degraded by bacteria. It
is important that the more easily degradable substances be
eliminated by bacteria, for the biologically active carbon
filter quite effectively eliminates substances which can
give rise to aftergrowth within the distribution system [See
Section F.2]. In addition, bacterial activity produces
something of a regeneration of the carbon, thus prolonging
the period of its usefulness. In general, the microbio-
logical degradability of organic substances will be dimin-
ished by chlorination and enhanced by ozonation [See Section
P.2.a].
c. Conclusions. The bacterial activity that takes
place in activated carbon filters is supported by organic
substances adsorbed to the carbon (Klotz, 1978; Klotz, et
al. , 1975, 1976a, 1976b; Werner, 1978, 1979). The more
-------
361
readily degradable substances are used by the bacteria in.
the filters, extending the useful life of the carbon and
reducing the capacity .of the treated water to support micro-
bial aftergrowth during distribution. A suggested adverse
effect on consumer health, from bacteria and their_deriva-
tives in the treated water, has yet to be substantiated.
5. Disinfection
a. Chlorination. Since the beginning of the 20th
century, chlorine has been one of the most widely used^and
thoroughly studied of chemical oxidants employed in drinking
water disinfection. This paper discusses the various reac-
tions between chlorine and water constituents, in terms of_
the important parameters governing these reactions and their
ultimate effects on microorganisms. Also described is the
mechanism by which chlorine acts upon the biota present in
water.
(i) Chlorine Dissociation in Water. Chlorine, when
added to water in the gaseous state or as sodium hypochlo
rite, initially hydrolizes to form hypochlorous acid:
2. NaOCl
-> HOCl + H + Cl
-> HOCl + Na+ + OH
At pH 5, all chlorine occurs as hypochlorous acid, but as
the pH increases, more hypochlorous will dissociate forming
hydrogen and hypbchlorite ions. Below pH 5, and at a con-
centration of > 1,000 mg per 1, chlorine may be present in
the elemental state (Cl,) (White, 1972). The chemical
species thus formed Cl,, HOCl, and .OCI~~ are called free
residual chlorine and their occurrence in water can be
summarized as:
C.I + HOCl at pH < 6 and HOCl + OC1~ at pH > 6 .
£
Temperature affects the equilibrium of the reaction, as
expressed by the dissociation constant (,KD) according to the
following relation:
KD = 2.6 x 10"8 mole per 1 at 10 °C
KD = 3.7 x 10~8 mole per 1 at 25°C
Thus, a rise in temperature increases the dissociation of
hypochlorous acid .
-------
362
Depending on the level and nature of pollution in a
water supply, chlorine will react partially or totally, by
oxidation or substitution, with organic and inorganic con-
stituents of the water. In the case of iron, for example,
chlorine reacts with ferrous ion and converts it to the
ferric form:
2FE
HOC1
Fe(OH)
5H
Soluble manganese compounds are similarly oxidized:
Mn
HOC1
HO
-> Mn02 + Cl + 3H
Both of these metal compounds will precipitate out. With
anions such as nitrites, sulfides, and sulfites, the fol-
lowing reactions take place:
+ HOC1
HS~ + HOC1 -
> NO,
°
+ Cl + H
Cl
SO,
+ HOC1
-> SO
4
+ Cl + H
At pH's > 8, chlorine occurs principally as hypochlorite
ions and gives the following reactions:
CN + OC1
S + 40C1"
-> CNO H- Cl"
-> SO,
+ 4C1"
Ammonia reacts with chlorine to produce inorganic
chloramines according to a sequence of reactions that begins
(at pH's up to 8.3) with the formation of monochloramines:
HOC1
The second stage proceeds as an equilibrium reaction between
mono and dichloramine:
NHLC1 + HOC1
NHC1
. * *.* .b,i.-w -*- ry * LA ry Vx
Finally, trichloramine is formed in the third stage:
NH_C1 + HOC! > NC1_ + H0O
2 3 2
These reactions are dependent on pH, temperature, contact
time, and the initial ratio of chlorine to ammonia. Chlora-
mines will convert to form elemental nitrogen if the ratio
-------
363
of chlorine to nitrogen exceeds 7.6:
2NH
3C1
6H
and in some cases, nitrates will, instead, be the product
formed.
Organic chemicals with functional groups similar to
ammonia (e.g., urea, amino acids, proteins, and other amines)
react with chlorine to produce organic chloramines:
HOC1
RNHC1 + HOC1
RNHC1
> RNC1
Chlorine linked with nitrogen is called combined residual _
chlorine. The sum of free residual chlorine (HOC1 and OC1 )
and combined residual chlorine determines the total re-_
sidual chlorine; it is this total residual chlorine^that
constitutes the germicidal potential available in chlorine
disinfection.
All of these reactions may occur in succession during
water treatment [See Figure E.S.a-l]. Zone 1 of the curve
in Figure E.5.a-l occurs with low quantities of chlorine
added to water. Here, chlorine will combine immediately
with reduced mineral compounds. Higher chlorine inputs
result in the formation of chloramines and some organo- _
chlorinated compounds (Zone 2). Further addition of chlorine
causes a loss in total residual chlorine due to the oxi-
dation of chloramines (Zone 3). When sufficient chlorine has
been added to react with all the chlorine-consuming chemicals
in the water, the breakpoint will be reached, during and
after which chlorine forms free residual chlorine, the same
as would occur in pure water (Zone 4). Reactions with other
chemical constituents of a particular water source that may
lead to the formation of toxic or even carcinogenic com-
pounds are discussed in Project Area IIA of this Pilot
Study.
(ii) Parameters Affecting the Disinfection Efficiency
of Chlorine. How efficiently chlorine will disinfect depends
on the microbial composition of the water, the chlorine^dose
applied, the contact time, and the chemical, form taken by
the chlorine. The chemical form of chlorine, in turn, _
depends on the pH and temperature of the water, along with
the amounts and types of dissolved organic and inorganic
substances present.
-------
364
FIGURE E.5.a-l
DIAGRAMMATIC REPRESENTATION OF COMPLETED
BREAKPOINT REACTION
c:
o
CO
O
5.
CHLORINE APPLIED M6/LITER OF WATER
-------
3G5
The following discussion takes into account only that
chlorine which occurs as a residual during and after break-
point chlorination. These chemical forms exert different
germicidal effects. Feng (1966) proposed the following
order of disinfecting efficiency:
C1
HOC1 > OC1
NHC1
R-NHC1
Morris (1967) measured the effects of HOC1, OC1 , and NH2C1
on various microorganisms and constructed a table of le-
thality coefficients for making comparisons [See Table
E.5.a-l]. His results reiterated that HOC! > OC1 > NI^Cl.
These studies confirmed earlier results of Butterfield and
coworkers (1943) who found that at pH 10.7, when most of the
chlorine would dissociate to form hypochlorite ion, a 60-
fold increase in contact time was required to inactivate IS..
coli and S^. typhi over that required at pH 7, when the
greater fraction of chlorine would be.HOCl. In similar
studies, it was found that increasing the pH from 7 to 8.5
for poliovirus 1 (Weidenkopf, 1958) and from 8.8 to 9 for
adenovirus 3 (Clarke, et al., 1956) extended the inacti-
vation time sixfold.
Although_the bactericidal and virucidal superiority of
HOC1 over OC1 has since been demonstrated by others (Engel-
brecht, et al., 1978; Scarpino, et al., 1974), inconsistent
results were obtained with poliovirus (Scarpino, et al.,
1974) Which were thought to be due either to direct effects
of pH (Morris, 1970) or to the presence, in the medium, of
mineral ions (Engelbrecht, et al., 1978; Kuzminski, et al.,
1970). Indeed, Engelbrecht and coworkers (1978) have shown
that the presence of potassium chloride accelerated polio-
virus 1 inactivation by hypochlorous acid. Because, unlike
the poliovirus, E_. coli inactivation by hypochlorite ions
was not enhanced in the presence of potassium chloride, it
would seem likely that it is the potassium choride, more
than the hypochlorite ions, that was responsible for pro-
ducing greater poliovirus inactivation.
It is commonly >agreed that chloramines are less effec-
tive disinfectants than free chlorine against bacteria
(Butterfield, 1948; Esposito, et al. , 1974; Siders, et al. ,
1973), viruses (Esposito, et al., 1974; Lothrop and Sproul,
1969; Siders, et a_l. , 19737~[See Table E.5.a-2], and proto-
zoan cysts (Chang, 1971b; Stringer, et a^., 1975). More-
over, the organic chloramines disinfect even less efficiently
than the inorganic chloramines (Feng, 1966; Kjellander and
Lund, 1965; Nusbaum, 1952). Hence, for optimal disinfection,
-------
366
TABLE E.5.a-l
LETHALITY COEFFICIENTS FOR DIFFERENT MICROORGANISMS
BASED ON TREATMENT WITH FREE AND
COMBINED CHLORINE AT 5°C
Oxidant
HOC1
OC1~
NH2C1
Enteric
Bacteria
20
0.2
0.1
Protozoan
Cysts
0.05
0.0005
0.02
Virus
1.0
0.02
0.005
Bacterial
Spores
0.05
0.0005
0.001
From Morris, 1967.
-------
TABLE E.5.a-2
DOSAGES OF VARIOUS CHLORINE SPECIES REQUIRED TO INACTIVATE 99% OF ESCHERICHIA COLI "AND POLIOVIRUS 1
Test
Microorganism
E . coli
Poliovirus 1
t
Chlorine
Species
Hypochlorous
acid (HOC1)
Hypochlorite
ion (OC1~)
Monochloramine
(NH2C1)
Dichloramine
(NHC12)
Hypochlorous
acid (HOC1)
Hypochlorite
ion (OC1~)
Monochloramine
(NH2C1)
Dichloramine
(NHC12)
Concen-
tration
mg/1
0.1
1.0
1.0
i.o
1.2
1.0
1.0
0.5
1.0
0.5
10
10
100
100
Contact
Time
min
0.4
0.92
175.0
64
33.5
5.5
1.0
2.1
2.1
21
90
32
140
50
c.ta
0.04
0.92
175.0
64.0
40.2
5.5
1.0
1.05
2.1
10.5
900
320
14,000
5,000
pH
6.0
10.0
9.0
9.0
9.0
4.5
6.0
6.0
6.0
10.0
9.0
9.0
4.5
4.5
Tempera-
ture "C
5
5
5
15
25
15
0
5
5
5
15
25
5
15
References,
4
Scarpino, et al., 1974
Scarpino, et al., 1974
Siders, et al. , 1973
Siders, et al., 1973
Siders, et ;al . , 1973
Esposito, et al., 1974
Weidenkopf, 1958
Engelbrecht, et al., 1978
Scarpino, et al., 1974
Engelbrecht, et al., 1978
Siders, et al., 1973
Siders, et al. , 1973
Esposito, et al., 1974
Esposito, et al., 1974
Concentration of compound multiplied by contact time (mg/1) (min).,
-------
368
free chlorine should be present in the water. Free chlorine,
inorganic chloramines, and organic chloramines all have
their own characteristic oxidation potentials, which also
affects the efficiency of disinfection [See Table E.5.a-3].
Chlorine may be applied at the end of the treatment
process so that a residual will be maintained throughout the
distribution system. In this way, aftergrowth is prevented
[See Section F.2.a] and as long as the residual remains
detectable, it ensures that no contamination has occurred
during distribution. As yet, no alternative has been found
to provide an equivalent measure of protection during dis-
tribution [See also Project Area IIA].
(iii) Effects of Chlorine Dose and,Contact Time on
Disinfection Efficiency. The relationship of the concen-
tration of free available chlorine to the time required for
destroying a certain portion of microorganisms is expressed
by the empirical equation:
where c ~ disinfectant concentration, t = the time required
for obtaining a constant percentage destruction of micro-
organisms, n = the Van't Hoff order of the reaction (or
coefficient of dilution), and k = a calculated constant that
represents an organism's relative resistance to a given
disinfectant. For example, Fair and coworkers (1968) cal-
culated values for an elimination of 99 percent of an initial
population by HOC1 (for which n = 0.86):
0 86
C t = 6.3 for coxsackievirus A2
O 86
C * t = 1.2 for poliovirus 1
O ft 6
C ' t = 0.098 for adenovirus 3
C°*86t = 0.24 for E. coli
p
When n is > 1, the efficiency of the disinfectant decreases
rapidly with dilution. When n < 1, the efficiency o£ dis-
infection is based principally on contact time. When n = 1,
the efficiency of disinfection is based equally on con-
centration and contact time.
It is difficult to review the literature for this type
of information because of insufficient reporting of experi-
mental conditions. However, from more recently published
-------
369
TABLE E.5.a-3
THE OXIDATION-REDUCTION POTENTIAL OF CHLORINE IN WATER
Reaction
Redox Potential at 25°C
C12 + 2e H= 2C1
HOC1 + H+ + 2e~ - Cl"* + 2H2O
OC1~ + 2H^O + 2e~ Cl" + 2OH~
+1.39 V
+ 1.49 V
+ 0.94 V
-------
370
accounts in which data were expressed as concentration x
contact time (C x t) [See Table E.5.a-2], it can be seen
that the different forms of chlorine vary markedly in their
abilities to inactivate _E. coli and poliovirus 1.
(iv) Sensitivity of Microorganisms to Chlorine. Under
ideal conditions, the destruction of a microbial population
theoretically can be predicted on the basis of Chick's law
(Morris, 1975), given that one disinfectant unit acts on one
single target site. This relation is written:
N = N C
o
-at
is
where N is the number of surviving microorganisms; N
the initial number of microorganisms; a is the slope°of the
inactivation curve; and t is the contact time. However, in
practice, discrepancies between laboratory experiments and
field assays have been noted with vegetative bacteria
(Carson, et al., 1972), bacterial spores (Bond, et al.,
1973), E. histolytica (Stringer, et al., 1975), and~Tab-
oratory and indigenous enteroviruses~TBlock, et al., 1978).
It would seem possible, therefore, that factors such as
aggregation, adsorption, and microbial resistance to a dis-
infectant, acquired over time, might limit the usefulness of
calculations when applying Chick's law.
Bacteria may aggregate in the presence of naturally
occurring polysaccharide polymers (Harris and Mitchell,
1975) or they may adsorb to hydroxides in the water (Carlson,
et al., 1975). Moreover, viral particles have been shown to
aggregate spontaneously. in distilled water (Floyd and Sharp,
1977) or in the presence of di- or trivalent ions (Floyd
and Sharp, 1978). This clumping of bacteria (Carlson, et
al*, 1975; Poynter, et al., 1973) and viruses (Berg, 1964)
limits diffusion of the oxidant, thereby decreasing the
disinfection rate (Sharp, et al., 1975), which could explain
the persistence of infectious particles in treated water
(Engelbrecht, et al.., 1978; Lovtsevich, 1973).
Turbidity also limits disinfectant action because
microorganisms will adhere to suspended matter (Hoff, 1978).
E. coli persisted several hours in turbid water to which
chlorine had been added (Tracy, et al., 1966). Likewise,
adding bentonite decreased the chlorine inactivation of
coliphage MS2 (Stagg, et ail. , 1977), although decreased
disinfectant efficiency was not observed when kaolinite and
aluminum oxide were tested with coliphage T7 and poliovirus
1 (Boardman and Sproul, 1977) [See also Sections C.2.a to
-------
371
Beside, the enhanced resistance to disinfectants
afforded bacteria and viruses that are adsorbed to suspended
matter or aggregated in .clumps, sensitivity to chlorine
differs from one microbial group to another and, in some
cases, within the life cycle of one species. Indicators and
pathogens most commonly isolated from water (e.g., E. coli,
Klebsiella, Salmonella, Shigella, and Pseudomonas) "ihow""T"-
similar resistance to chlorine. Salmonella species ["See
also Section B.l.a(i)] are not exceptionally resistant to
chlorine and should be controlled effectively by proper ;
application of disinfectants. The same holds true for
Francisella tularensis (Gotovskaia and Mogaram, 1945) [See
Section B.I.a(v)] and most of the other bacterial pathogens.
However, there are other microbes whose resistance'to
chlorine exceeds that of the coliform group; these include
spores of clostridia and of aerobic bacteria, enteroviruses,
yeasts, acid fast bacilli, and cysts of protozoan pathogens.
It is well established that enteroviruses are more resistant
than E. coli to chlorine^ (Butterfield, 1948; Engelbrecht,
.et al.,.1978; Weidenkopf, 1958) [See Table E.5.a-3] although
there are no data for the hepatitis A virus. Yet, even
among the enteric viruses, there is variation in resistance
to chlorine. When water of zero chlorine demand and 0.5 mg
per 1 free residual chlorine was held at 2°C, the time
needed to inactivate 25 human enterovirus types ranged from
2.7 min for reoviruses to > 120 min for coxsackievirus A
(Liu, 1978). Grouped according to their resistance to
chlorine, viruses compare with E. coli'in the following
order: poliovirus > coxsackievirus > coliphage MS,, > E.
coli. Moreover, viruses may be gradually selected for~
resistance to chlorine after repeated sublethal exposure
(Bates, £t aJL. , 1977). In similar studies, Shaffer (1978)
found poliovirus, all of serotype 1, to exhibit varying
sensitivity to chlorine.
A review by Engelbrecht and Greening (1978) compared
data^which, although collected under different experimental
conditions, nevertheless serve to indicate that Mycobac-
terium fortuitum, M. phlei, and Candida parapsilosis [See
Sections A.2.a and C.2.f] are more resistant than poliovirus
to chlorine.
Data on survival of Entamoeba histolytica [See Section
B.l.c(i)] exposed to free chlorine, ammonia, and chloramines
are summarized in Figures E.5.a-2 and 3. The parameter of
cysticidal residual used" in these calculations is defined as
that concentration required to kill 99.999 percent of cysts.
From the curve in Figure E-.5.a-2, it is clear that ordinary
-------
FIGURE E.5.a-2
TIME-CONCENTRATION RELATIONSHIP IN DESTRUCTION OP CYSTS OF ENTAMOEBA HISTOLYTICA BY
HOC1 AT 3* AND 23*C
100
U)
-4
M
Cysticidal residual* in ppm as HOC1
Concentration required to kill 99.999 percent of cysts,
-------
FIGURE E.5.a-3
373
TOO
EFFECT OF TEMPERATURE ON CYSTICIDAL EFFICIENCY
OF NH2C1 and NHC1
o
«
4J
I
f
1
,2.3nR 5.575
E = 2,750 X 4.575
= 12,500 calories
1/T X 10
Concentration required to kill 99.999 percent .of cysts,
-------
374
are
chlorination practiced in water treatment is inadequate to
kill cysts (taking into account pH and temperature effects),
Chlorine concentrations needed to kill E. histolytica cysts
in clarified water in 20 to 30 min can exceed 9 ppm, de-
pending on pH, temperature, and ammonia concentration.
Likewise, Giardia lamblia cysts [See Section B.l.c(ii)]
not destroyed by chlorination at dosages and contact times
commonly used in water treatment. Moreover, Acanthamoeba
and Naegleria cysts [See Sections B.l.c (iii) and (iv)J are
botli highly resistant to chlorine (Dejonckheere and van de
Voorde, 1976), which may account for their widespread
presence in U.S. (and probably other) water supplies.
Infective larvae of hookworms and threadworms [See Section
B.l.c(v)3 are more resistant to disinfection than enteric
bacteria, viruses, or even protozoan cysts.
(v) The Action of Chlorine on Microorganisms. Chlorine
acts on bacteria by changing membrane permeability, causing
a loss of cellular material including potassium and UV-
absorbing material (probably nucleic acids and protein).
Also altered is the zeta potential of the membrane (Ven-
kobachar, et al., 1977), whereby ozone acts upon those
membrane-linked respiratory enzymes containing sulfhydryl
groups (e.g., aldolase, triosephosphate-dihydrogenase, gluta-
mate decarboxylase) (Green and Stumpf, 1946; Knox, et al.,
1948). Finally, chlorine reacts with nucleic acidsTBocharov
and Kulikovski, 1971; Fetner, 1962; Rosenkranz, 1978; Shih
and Lederberg, 1976) of both bacteria and viruses.
Using coliphage f2, Dennis and coworkers (1979) de-
monstrated that chlorine reacted first with RNA (incorpora-
tion of chlorine being pH-dependent). In another study,
chlorine was found to also react with the protein capsid of
coliphage f2 (Olivieri, et al., 1975). In the case of
enteric viruses, it probaTSly is their lack of enzymes and
other sensitive systems that accounts for their greater, :
resistance to chlorine.
(Vi) Summary. it is certainly true that chlorine alone
cannot produce safe drinking water. However, the chemistry
and microbiology of chlorine disinfection of water have been
studied for many years and'have resulted in a rich base of
scientific literature. Used appropriately, chlorine is
known to be effective against a wide variety of undesirable
or pathogenic organisms in water. Other disinfectants may
be superior in certain specific applications, but chlorine
-------
375
remains the unchallenged standard for many purposes in
drinking water treatment and distribution.
b. Ozone Treatment. Ozone (0_), an allotropic form of
oxygen, is a potent biocide used extensively throughout
Europe in drinking water treatment. Schonbein coined the
name ozone in 1839, from the Greek word ozein, to smell,
which is appropriate because ozone was first recognized in
1783 by the Dutch physician Van Marum, on the basis of a
peculiar odor given off by an electrical machine. However,
it was not until 1865, that Soret demonstrated the existence
of a triatomic molecule and in 1873, Fox described the
bactericidal properties of ozone.
In 1976, 200 water treatment facilities in Europe
disinfected with ozone, compared to 20 in Canada, and only
three in the U.S. (International Ozone Institute,. 1976).
Ozone has received more attention in the U.S., though, as a
disinfectant of urban wastewaters and is used for this
purpose at ten U.S. facilities.
Ozone's disinfecting capability is at least equal to
that of chlorine dioxide (Holluta and Unger, 1954) [See
Section E.S.c] and is greater than that of chlorine (Fluegge,
et. al , 1978; Hunger,. et al., 1977; Traenhart and Kuwert,
1975) [See Section E.2D, bromine, and iodine (Kruse, et
al., 1970) [See Sections B.l.c(i) and E.S.d]. The powerful
germicidal properties of this agent are due to its high
oxidation-reduction (redox) potential that results from the
different 'chemical forms of ozone formed in water, particu-
larly the primary and secondary free radicals. It follows
that, for all types of microorganisms, the specific coeffi-
cient of lethality, as defined by Morris (1967), is always
higher for ozone than it is for free chlorine or chloramines.
(i) Factors Affecting the Disinfection Efficiency of
Ozone. The efficiency with which pzone can be expected to
disinfect water is influenced, as is true with other chem-
ical oxidants, by: (1) the quality of the water to be
treated; (2) the amount of disinfectant applied and the
'contact time; and (3) the types and densities of microorgan-
isms in the water.
(i.l) Quality of the Water to be Treated. A discus-
sion of water quality must address the parameters of pH,
-------
376
temperature, suspended matter, and chemical composition.
Ozone's bactericidal and virucidal capabilities are far less
dependent on pH than are those of chlorine (Diaper, 1972;
Kinman, 1975; Traenhart and Kuwert, 1975). Different authors
have variously reported ozone to be effective as a disin-
fectant between pH 5.8 to 8 (Suchkov, 1964), pH 6 to 8
(Evison, 1977), and pH 5.6 to 9.8 (Farooq, e_t al. , 1977a) .
Part of the ozone decomposes into secondary oxidants in
water at a rate that increases with higher pH's. At higher
decomposition rates there is less ozone residual. This
results in a reduced microbial inactivation rate for a
given ozone dosage. The effect of pH seems to be entirely
through its influence on the ozone residual, such that, if
the ozone dose is adjusted to produce a .standard residual,
the microbial inactivation rate is the same at different
pH's (Farooq, e_t aJL. , 1977a) .
Higher temperatures cause greater ozone inactivation
of microorganisms, as predicted by the Van't Hoff-Arrhenius
theory (Fair, e_t ajL. , 1968), possibly due to the increased
rate of disinfectant diffusion into microbial cells and
greater reactivity with cellular substrates. However, the
solubility of ozone in water decreases with increased temper-
atures and one might, therefore, expect the need for greater
amounts of ozone with higher temperatures. Yet, between 6
and 30°C (Kinman, 1975), or 5 and 25°C (Evison, 1977), the
inactivation rate of ozone remained nearly the same for
colony counts and for phage 185. In fact, Farooq and co-
workers (1977b) found that, for a given quantity of in-
jected ozone and within a range of 9 to 37°C, an increase in
temperature both enhanced inactivation and increased the
residual ozone concentration.
Although some authors claim that turbidity has no
effect on ozone disinfection (Cerkinsky and Trahtman, 1972;
Gevaudan, et al., 1971), Block (1977) found that a popu-
lation of polTovirus 1, 90 percent of which was adsorbed to
kaolin, was more difficult to inactivate with ozone when the
turbidity reached 9 Jackson turbidity units (JTU).
Ozone, upon injection into water, may follow any (or a
sequential combination) of four pathways: (1) no diffusion
in the water; (2) diffusion and autocatalytic oxidation; (3)
diffusion and immediate reaction with organic matter; or (4)
diffusion and temporary maintenance (a few minutes) of an
active residual. The oxidizing action of ozone is dissi-
pated rapidly in the presence of organic matter, but a weak
-------
377
microbial inactivation is obtained with no measurable re-
sidual of ozone (Block, 1977; Block, et aJL. , 1977; Burleson,
ejt al. , 1975; Coin, et'al_., 1967; Farooq, 1976; Katzenelson,
et al., 1974; Majumdar, et al., 1974; Mercado-Burgos, et
al. , 1975) [See Figure E.6-1, Zone ij. According to Fa~rooq
and coworkers (1976), inactivation of microorganisms in the
absence of residual ozone could be the result of collisions
between microbes and ozone bubbles that are covered with a
liquid film of high ozone content. The same authors found
that the presence of ozone bubbles,alone increased disin-
fection by 90 percent. Block (1977) showed that in the
absence of residual ozone, ,inactivation was proportional to
the ozone concentration in the gas bubbles.
Beyond a critical ozone dose injected at the break-
point, a dissolved ozone residual is formed [See Figure
E.5.b-l, Zone 2] which can be measured by conventional
methods. Inactivation of microorganisms in Zone 2 is rapid
and considerable, and is a function of the concentration of
dissolved residual ozone and contact time (to be discussed
later). Disinfection, being almost instantaneous once the
ozone demand is satisfied, is reflected by a sharp drop in
the survival curve, referred to as the all-or-nothing effect
(Fetner and Ingols, 1956; Pavoni, £t al., 1972; Tonelli
and Ho, 1977). This effect is probably due to a sudden
increase in the redox potential of the solution, causing
immediate catalysis induced toy the residual ozone present.
A similar discontinuity is' not evident in the curve for
organic matter [See Figure E.5.b-1], possibly because only
minor changes occur in it (Block, et al., 1976), or because
microorganisms comprise only a miniscule portion of the
total organic matter that is.measured in the system (Block,
1977). At the breakpoint, low concentrations of organic
matter do not interfere with the inactivation of microor-
ganisms by ozone unless the organism actually is imbedded
within, and therefore physically protected by, the organic
matter (Evison, 1977; Gevaudan, et al., 1971).
(i.2) Ozone Dosages and Contact Time. Inactivation of
microorganisms occurs extremely rapidly once an ozone re-
sidual is present. Inactivation also relates directly to
the redox potential, wherein ozone, with a redox of 350 mv .
(Fetner and Ingols, 1956) or 400 mv (Block, 1977; Hugues and
Plissier, 1977), applied for 2 to 4 min has been demon-
strated to inactivate >_ 99.9 percent of the microbial popu-
lation. This compares quite favorably to chlorine, which
-------
378
FIGURE E.5.b-l
DIAGRAMMATIC REPRESENTATION OF COMPLETED BREAKPOINT REACTION
.For OZONE
organic matter
microorganisms
4_ Redox
_ residual ozone
residual
ozone
(mg/l)
0.5 1£>A 1.5
break-point
2.0
ozone injected
(rag/1) *
-------
379
requires a redox of 500 mv and a contact time of 60 min for
an equivalent inactivation (Lund, 1963).
Tables E.5.b-l and 2 summarize data from various studies
on inactivation of bacteria and viruses by ozone. There
are a great many apparent discrepancies in these tables,
which might at first glance be attributed to differences in
experimental conditions. However, if one recalculates the
results on the basis of the product of ozone concentration
(in mg per 1) and time (in sec) required to cause 99.9
percent inactivation of a microbial agent in water that is
free of organic matter, one finds that the discrepancies are
small. For example, this product is either 3.0 (Block,
1977; Burleson, et aJL. , 1973) or 2.5 (Farooq, et al.. , 1976)
for E. coli and.either 1.8 (Evison, 1977) or 1.9 (Block,
1977J for poliovirus 1. ,
In every case, the ozone residual dissolved in water
remained the one critical parameter of disinfection effi-
ciency. Knowing this, water treatment authorities can
ensure the elimination of both micropollutants and aggre-
gated or adsorbed microorganisms simply by applying ozone
dosages higher than those applied in the laboratory.
(i.3) Sensitivity of Microorganisms to Ozone. Dif-
ferent organisms exhibit different sensitivities to ozone
and many authors have tried to use this observation as a
basis for classifying groups of organisms. Although some
generalities can be drawn from such an exercise' for
example, that poliovirus 1 is more resistant than 1C. coli to
ozone (Block, 1977; Evison, 1977; Farooq, 1976; Katzenelson,
et al., 1974), or that bacterial spores are highly resistant
to ozone treatment (Broadwater, et al., 1973; Burleson, et
al., 1976; Haufele and Sprockoff, 1973; Leiguarda, et al.,
1949), ambiguities emerge under closer scrutiny. Contra-
dictory results have been obtained by several.authors
studying the relative sensitivities of three poliovirus
types (Coin, et al., 1964, 1967; Evison, 1977; Snyder and
Chang, 1974).
Other factors are likely involved in influencing the
sensitivity of organisms to ozone. These might include:
different techniques for preparation of stock cultures, age
differences of the cultures (Wuhrmann and Meyrath, 1955),
and aggregation phenomena (demonstrated experimentally
using ultrasonication to change the inactivation kinetics:
Burleson and Pollard, 1976; Dahi, 1977; Katzenelson, et
al., 1974). Furthermore, relative sensitivity to ozone
-------
TABLE E.5.b-l
SUMMARY OF DATA FROM STUDIES OF BACTERIAL INACTIVATION BY OZONE IN WATER
Ozone (mg/1)
Water
Distilled water
Distilled water
Distilled water
Distilled water
Deionized water
Pure water
Pure water
Tap water
Tap water
Tap water
River water
Phosphate buffer
Phosphate buffer
Phosphate buffer
Bacterium
E. colia
E. colia
E. colia
E. colia
E. colia
E. colia
S. faecalis0
E. coli
S. faecalis
E. coli
E. coli
E. coli
E. coli
E. coli
Injected
-b o
0
0
0
0
0
0
1.0 0
1.0 0
0
0
0
- 0
0.4 to 0.5
Residual
to 0.02
.05 to 0.16
.07
.3
.19
.01
.01
.02
.02
.3
.04
.20
.25
Contact Inactivation
Time (sec) (%)
180
10
3
300
60
15
150
150
600
240
15
10
60
95 to 99.9
100
99.99
100
100
99.99
9.9.99
99.9
99.9
93
100
99.9
99.9 ro
o
99.9
-------
TABLE E.5.b-l Continued
Water
Buffer
Buffer
Wastewater
Wastewater
Escherichia coli .
Ozone (mg/1) .
.._._,..., fonirir'l' Tn?i(~"h i vat" 1 on
Bacterium Injected Residual Time (sec) (%)
E. coli 0.006 - 420 99
»
E. coli - 0.05 . 12 99
E. coli - 0.01 75 99.9
E. coli - 0.30 15 99.9
E. coli - 0.009 40 99.9
E. coli - 0.01 90 99.99
E. coli - 0.1 5 99.9
E. coli and - 0.13 to 0.2 60 99.9
S. faecalis
v> '
Data not reported by author.
c
Streptococcus faecalis.
00
'. . ' . : H
-------
TABLE E.5.b-2
SUMMARY OF DATA FROM STUDIES OF VIRAL INACTIVATION BY OZONE IN WATER
Ozone (mg/1)
Water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Distilled water
Tap water
Tap water
Virus Tested Injected Residual
Poliovirus 1 1
Poliovirus 3 -a
Poliovirus 1.27
Poliovirus 1 1.27
Poliovirus 1 1
Poliovirus 1
Poliovirus 1,2,3
Poliovirus 1
Coliphage T_
Coliphage T..
Coliphage f_ -
Coliphages T. , 2
T2, and T^
Poliovirus 1 1.0
Poliovirus -
0.1
0.22
0.23
0.23
0.25
0.3
0.4
1.0
0.09
0.5 to 0.55
-
0.01
0.05
Contact
Time (sec)
60
180
150
150
240
10
180 to 240
120
10
300
15
120
150
_
Inactivation
(%)
99.99
100
100
100
100
99.5
99.99
99.99
99.5
100
100
100
w
99.9 S
99.9
-------
TABLE E.5.b-2 Continued
Ozone (mg/1)
Water Virus Tested Injected Residual
Contact
Time (sec)
Inactivation
(%)
Surface water
Surface water
' Dnieper river
water
Poliovirus 3
Coliphage T..
Poliovirus 3 &
coxsackievirus B
4 to 5
0.22 180
0.5 to 0.55 300
0.2 900
100
100
99.7 to 99'.9
Seine river water
Phosphate buffer
Buffer
Buffer
Wastewater
Wastewater
Wastewater
- '
-
-
Poliovirus 1,2,3
Poliovirus 1
Poliovirus 1
Coliphage
Poliovirus 1.38
Poliovirus 1
Coliphage f~ 15
Poliovirus
(purified)
Poliovirus
Poliovirus 2 and 1.13
coxsackievirus B3
0.4
0.25
0.03
0.18
0.20
0.30
0.015
0.045 to
0.45
0.15
0.28
180 to 240
20
60
240
150
20
300
120
-
60
99.
99
99.
99.
100
99
100
100
100
100
99
9
95
U)
CXI
CO
-------
TABLE E.5.b-2 Continued
Water
-
-
-
-
Ozone (mg/1)
Virus Tested Injected Residual Time (sec) (%)
Poliovirus - 0.45 120 100
Poliovirus - 0.5 240 100
Poliovirus 3 0.7 to 1.2 10 99.99
Enterovirus - 0.2 - 99.9
Data not reported by author.
00
-------
385
varies with contact time and ozone concentration. That is,
at disinfection levels necessary to obtain a 99 percent
inactivation, poliovirus 1 was more sensitive than coli-
forms; yet, these two entities showed the same sensitivity
when a 99.99 percent inactivation was sought (Ghan, et al./
1976).
The initial density of a microbial population does not
appear to affect inactivation kinetics as long as increased
numbers do not increase the ozone demand in the water (Block,
1977). On the other hand, the presence of organic matter
does interfere with disinfection efficiency (Farooq, 1976).
(ii) Action of Ozone on Microorganisms. Ozone inacti-
vation of bacteria can be described as an oxidation reaction
(Bringmann, 1954), both the reaction and the inactivation
being rapid. The ozone acts on at least three sites on the
bacterium: the cell wall, the cytoplasmic membrane, and the
nuclear apparatus.
Ozone acts on the cell wall, causing bacterial cell
lysis and death which adds soluble COD to the water (Rosen,
et al., 1974). Cell lysis, however, probably is not the
primary means of inactivation, but a consequence of either
high ozone levels (at the gas-liquid interface) or prolonged
ozonation (Perrich, 1976). The cytoplasmic membrane proba-
bly is affected first (Christensen and Giese, 1954) by
reactions of ozone with glycoproteins, glycolipids, (Scott
and Lesher, 1963; Smith and Bodkin, 1944), or the amino
acids tryptophan (Goldstein and McDonagh, 1975) or cysteine
(Mudd and Freeman, 1977). Bacterial death also may result
from changes in permeability of the cytoplasmic membrane.
Ozone disrupts bacterial respiration by reacting with the
sulfhydryl groups of certain enzymes. Bacteria exposed to
ozone were shown to lose their ability to degrade sugars and
produce gases (Vronchinskii, 1963). Ozone reacts with
glucose-6-phosphate dehydrogenase as well as with other
enzyme systems (Chang, 1971b).
In addition, ozone can act on the nuclear material and
provoke mutations. Christensen and Giese (1954) and Scott
and Lesher (1963) demonstrated that ozone affects both
purines and pyrimidines. More recently. Prat and coworkers
(1968) showed that pyrimidine bases of nucleic acids in E_.
coli undergo modifications when the bacteria are treated
with ozone (thymine being more sensitive than cytosine or
uracil). Sublethal exposure to ozone can cause nucleic acid
mutations similar to those caused by x-rays, without impairing
growth and division of the cell (Hamelin and Chung, 1974,
1975).
-------
386
(iii) Summary. Ozone is effective against both bac-
teria and viruses in water. As a principal disinfectant
(and disregarding cost), it appears to be faster and more
effective than chlorine as well as less influenced by water
quality; its only relative shortcoming is that active ozone
residuals cannot be maintained in water for significant
periods of time [See also Section F.2.a and Project Area
IIA]. Ozone also has been used, with relative success,
prior to slow sand filtration [See Section E.3] to render
compounds more biodegradable.
c. Chlorine Dioxide. The history of chlorine .dioxide,
its physical and chemical properties and reactions, methdds
for generation and measurement, and its uses in water treat-
ment have been reviewed in several recent publications
(Gall, 1978; Miller, et al., 1978; Rosenblatt, 1975; Sussman
and Rauh, 1978; White, 1972). Chlorine dioxide has been
used primarily as an oxidant in industrial processes for
bleaching of pulpwood, textiles, flour, fats, oils, and
waxes. Its use in the treatment of drinking water has been
extremely limited, however, especially in the U.S.; and when
used at all, chlorine dioxide has been applied mainly to
improve the organoleptic properties of the water rather
than to disinfect. Recently, chlorine dioxide has been ,
viewed with increased interest as a possible alternative to
disinfection with chlorine, against which arguments have
been raised concerning its capacity to generate toxic and
carcinogenic compounds [See Section E.S.a].
(i) Bactericidal Efficiency. Of results reported in
earlier studies on the bactericidal efficiency of chlorine
dioxide, only some have been upheld by later work.
Ridenour and coworkers (1949) showed that chlorine
dioxide was more effective than chlorine for inactivating
endospores of Bacillus subtilis, JB. megaterium, and 13.
mesentericus. Berndt and Linneweh (1969) also reported that
chlorine dioxide was a more efficient sporicide than chlo-
rine. In addition, Ridenour and Ingols (1947), and Ridenour
and Armbruster (1949) found chlorine dioxide to be an ef-
fective bactericide against indicators (Escherichia coli,
Enterobacter aerogenes), enteric pathogen's(SalmonelTa
typhi, jS. paratyphi B, and Shigella dysenteriae) [See Sec-
tions B.l.a(i) and (ii) ], and other pathogens '(Staphylococcus
aureus, Pseudomonas aeruginosa) [See Sections B.l.a(ix)and
F.2 to 3]. Their results indicated that the disinfection
-------
387
efficiency of chlorine dioxide increased at higher pH1s and
that it was a more efficient bactericide at pH 9.5 than
chlorine. The work of Trakhtman (1946), and Bedulevich and
.coworkers (1953) also indicated that the bactericidal ef-
fectiveness of chlorine dioxide was equal to or greater than
that of chlorine against E. coli, J3. typhi, and £3. para-
typhi. However, their results indicated that its effec-
tiveness decreased rather than increased under alkaline
conditions. Early studies by McCarthy (1944; 1945)indi-
cated that chlorine dioxide was an effective bactericide^in
water low in organic matter content, but was less effective
when organic matter concentrations were higher.
White (1972) reviewed several of these early studies
and pointed out that the lack of methods for preparing pure
chlorine dioxide free of residual chlorine, and the lack of
analytical procedures for measuring it in the presence of
other oxychloro chemical species, would have tended to
produce errors in which analytic results showed mistakenly
high chlorine dioxide levels. Hence, the disinfection
efficiency of chlorine dioxide would, in fact, be higher
than had appeared in these earlier tests.
Later, Granstrom and Lee (1958) developed improved
methods for preparing and analyzing chlorine dioxide.
Benarde and colleagues (1965), using these improved tech-
niques, compared the bactericidal efficiency of chlorine and
chlorine dioxide at pH 6.5 and 8.5 in a disinfectant demand-
free buffered system. At pH,6.5, chlorine was slightly more
effective than chlorine dioxide when compared on the basis
of weight. However, at pH 8.5, chlorine dioxide was even
more effective than it was at pH 6.5, whereas chlorine
efficiency declined at the higher pH, reflecting a decrease
in HOC1 (the more effective disinfecting species) and an
increase in OC1 (the weaker disinfecting species) at higher
pH's.. Cronier and colleagues (1978) also showed that within
a pH range of 4.5 to 9.0, the bactericidal efficiency of
chlorine dioxide increased as pH increased.
In other studies, Benarde and coworkers (1965) used
sewage effluent to demonstrate that chlorine dioxide was a
more efficient disinfectant than chlorine in the presence of
high levels of organic material. An initial dose of 2 mg
per 1 of chlorine dioxide obtained > 99 percent reduction of
E. coli in sewage after 5 min of exposure, whereas 5 mg per
T of chlorine 'destroyed only 90 percent of E. coli after the
same length of exposure. Moffa and coworkers (1975) re-
ported similar results when they compared the bactericidal
-------
388
action of chlorine and chlorine dioxide applied as disin-
fectant to combined sewer overflows.
(ii) Virucidal Efficiency. Ridenour and Ingols (1946)
reported that chlorine dioxide was as effective as chlorine
against a mouse-adapted strain of poliovirus 1. Hettche and
Schulz-Ehlbeck (1953) reported that 0.08 mg per 1 of chlorine
dioxide was as virucidal for poliovirus as was 0.15 mg per 1
of ozone [See Section E.S.b] or 0.25 mg per 1 of free
chlorine. An increase in pH from 5.7 to 8.5 enhanced in-
actiyation of poliovirus 3 (Warriner, 1967); and chlorine
dioxide inactivated poliovirus (Mahoney)-4.6 times faster at
pH 9.0 than at pH 7.0 (Cronier, et _al. , 1978). Chlorine
dioxide at pH 7.0 and free residual chlorine at pH 6.0
showed approximately equivalent virucidal efficiencies. As
with chlorine [See Section E.2]; however, enteric viruses
were found more resistant than E_. coli to chlorine dioxide '
(reviewed by Allen, 1979) [See also Sections B.l.b and
C.2.b].
(iii) Cysticidal Efficiency. No reports appear to be
available on the efficiency of chlorine dioxide for disin-
fecting -v . protozoan cysts, notably those of Entamoeba
histolytica and Giardia lamblia [See Section B.l.cTi~
(iv) Mechanism of Action. Ingols and Ridenour (1948)
suggested that, following adsorption onto and penetration of
the^cell wall, chlorine dioxide exerted its bactericidal
action by reacting with intracellular enzymes containing
sulfhydryl groups. Benarde and coworkers (1967) showed that
the incorporation of C -labelled amino acids [See also
Section C.4.a] stopped within a few seconds after exposure
of E' coli to chlorine dioxide, indicating interference with
protein synthesis. Olivieri (1968) showed that the inhi-
bition of protein synthesis in bacteria exposed to chlorine
dioxide was dose-dependent. Examination of cell extracts
showed the site of action to be localized in the portion of
the cell containing enzymes. The ability of ribosomes to
function in protein synthesis was not affected. Studies on
the effects of chlorine dioxide on bacterial and viral
nucleic acids apparently have not been performed.
(v) Use of Chlorine Dioxide in Water Treatment.
Miller and coworkers (1978) reported that chlorine dioxide
-------
389
is used for some purpose in at least 607 water treatment
plants throughout the world. Of these, 495 are in Europe,
84 are in the U.S., and ten are in Canada. In the U.S.,
chlorine dioxide is used mainly for purposes other than
disinfection (i.e., taste, odor, and color control; manga-
nese and iron removal). It is used as the sole disinfectant
at only one treatment plant in the U.S. In Europe, chlorine
dioxide is used frequently for disinfection, but nearly
always in combination with some other disinfectant, usually
ozone or chlorine.
(vi) Conclusions. The available evidence indicates
that chlorine dioxide is effective for disinfection against
bacteria and viruses. When compared to chlorine on the
basis of weight, in mg per 1, its disinfection capability is
equivalent to free residual chlorine at pH 6 to 7 and is
appreciably more efficient than free residual chlorine at
higher pH's, which are encountered more commonly in actual
water treatment plant operations. An additional advantage
is that it does not react with ammonia, as does free re-
sidual chlorine, to form chloramines, which are less effec-
tive as germicides. Organic compounds normally present in
water exhibit a high chlorine demand with free residual
chlorine, but appear to be less reactive with chlorine
dioxide. Thus, biocidal capability is maintained more
effectively with chlorine dioxide than with chlorine.
The bactericidal and virucidal effectiveness of chlorine
dioxide is well established; research is needed to determine
its ability to inactivate the cysts of waterborne pathogenic
protozoa, especially those of E. histolytica and G. lamblia.
Additional information on the mode and site of its disinfec-
tant activity also is needed. Other research needs not
related directly to disinfection include: (1) developing
improved methods for measurement, particularly in the
presence of other disinfectants; and (2) evaluating health
effects of its major end products, chlorite and chlorate,
and of compounds formed by its reactions with other organic
compounds in water.
d. Iodine. Historically, iodine has been used pri-
marily in homes and hospitals as an antiseptic for skin
surfaces, wounds, surgical instruments, etc. Iodine has
been studied as a disinfectant for water for nearly 30
years.
The chemistry of iodine, methods for measurement, and
use in water treatment were reviewed by White (1972). In
-------
390
the pH range 5 to 8,'iodine dissolved in water is present
predominantly as diatomic iodine (I2) and hypoiodous acid
(HIO). Both of these chemical species are considered to be
effective biocides, although more recent evidence suggests
that the hydrated iodine cation (H^OI ) may be the active
species (Cramer, et al., 1976). AE higher pH levels, the
ineffective biocidal species hypoiodite (IO~) and iodate
(IO~) are formed. White (1972) indicated that formation of
these inefficient disinfecting species would not be a prob-
lem below pH 8.4. The relative amounts of titrable iodine
existing as 1^ and HIO in aqueous solutions depends on the
pH, iodine concentration, and to a lesser extent, tempera-
ture. Although it is true that elemental iodine and hypoio-
dous acid are the two most powerful disinfecting species of
iodine, their relative efficiency varies considerably de-
pending on the microorganism exposed. Elemental iodine was
more effective than hypoiodous acid for inactivation of
Bacillus metrens spores and Entamoeba histolytica cysts [See
Section B.l.c], whereas hypoiodous acid was more effective
than elemental iodine for inactivation of enteroviruses and
Escherichia coli (White, 1972). Although there is a con-
siderable body of literature on the biocidal efficiency of
iodine, comparison of the results of different investiga-
tions is virtually impossible because of the lack of con-
sistency in experimental conditions such as temperature and
pH, methods for measuring and reporting free iodine, and
microorganisms used.
(i) Bactericidal Efficiency. Chambers and coworkers
(1952) studied the bactericidal efficiency of iodine under
closely controlled conditions of temperature, pH, free
iodine concentration, and exposure time. Bacterial species
studied were Aerobacter (now Enterobacter) aerogenes,
Salmonella paratyphi, S. schottmuelleri, S.Typhimurium, S.
flexneri, Shigella dysenteriae, S. sonn'eiT and Streptococ~
cus faecalis, as well as two strains of 12. coli and three
strains of S_. typhosa. The results were reported as that
free iodine concentration required to kill all bacteria
exposed (y*
-------
391
were interested in practical applications, they used tap
water rather than well-defined, iodine-demand-free buffer
systems. They determined the effects of color, turbidity,
and nitrogenous material, as well as temperature and pH, on
the bactericidal efficiency of iodine. At 25°C and pH 8.1
to 8.5, iodine concentrations of 2 to 5 mg per 1 reduced E_.
coli by approximately 6 logs in 10 min. They found in
further studies that S_. typhosa, S^. schottmuelleri, Shigella
dysenteriae, and mixed coliforms present in sewage were
about as iodine-sensitive as 12. coli, whereas Vibrio cholerae
was more sensitive. The bactericidal efficiency of iodine
was approximately the same over a pH range of 4.5 to 8.1.
Low levels of ammonia and urea nitrogen (5 mg per 1), and
turbidity from clays (50 to 500 mg per 1) had no measurable
effect on disinfection efficiency, but high concentrations
of fine loess (165 to 245 mg per 1) interfered with bac-
tericidal action. Berg (1966) indicated that the bacteri-
cidal efficiency of I was about one-fifth that of hypo-
chlorous acid, but about ten times that of the hypochlorite
ion.
(ii) Virucidal Efficiency. Chang and Morris (1953)
studied the virucidal efficiency of aqueous iodine using
mouse-adapted poliovirus 1. Because of technical problems,
they were able to conclude only that iodine did have viru-
cidal properties and that it was effective against viruses
at the level needed for inactivation of
-------
392
that concentrations of iodine needed to inactivate all
viruses at a pH between 5 and 8.5 were equal to or less than
the concentrations of free residual chlorine needed for
equivalent inactivation.
(iii) Cysticidal Efficiency. Much of the research on
the cysticidal effects of iodine resulted from a military
need for emergency disinfection of small water supplies.
Amebic dysentery was endemic in many areas of military
operation, and cysts of Entamoeba histolytica [See Section
B.l.c] had been shown to be highly resistant to chlorine;
disinfection studies directed at inactivation of this
organism led to the development of the globaline tablet,
presently used by the U.S. Armed Forces for disinfection of
individual drinking water supplies (Chang, 1958? Chang and
Morris, 1953; and Morris, ejt al. , 1953). I2 was two to
three times as cysticidal as HOI, so the globaline tablet
was designed to release 8 mg per 1 of free iodine, which was
sufficient to inactivate 5 logs of E_. histolytica cysts in
10 min at pH ^ 8.0.
Stringer and coworkers (1975) essentially confirmed the
earlier results indicating that ~L was the more active
cysticidal species and that cyst Inactivation rates were
very slow at higher pH's. However, they found hypochlorous
acid to be a more effective cysticide than 1^.
Information on Giardia lamblia cyst inactivation by
iodine is not available.
(iv) Mechanism of Action. Historically, iodine has
been considered to act like chlorine as a general cellular
poison exerting its lethal effect by oxidation of sulfhydryl
groups on enzymes or proteins (Dunn, 1952). Brandrick and
coworkers (1967) showed that radioactive elemental iodine
reacted with E. coli and Staphylococcus aureus by both
oxidation-reduction and halogenation. Brammer (1963), Hsu
(1964), and Hsu and coworkers (1966) showed that viral RNA
from viruses and transforming DNA from bacteria inactivated
by iodine remained active, indicating that the biocidal
activity resulted from reactions with protein rather than
nucleic acid components. Berg and coworkers (1964) reported
that the kinetics of coxsackievirus A9 inactivation were
consistent with a "single hit" hypothesis and inferred that
inactivation resulted from reaction with a single iodine
molecule. The failure of iodine to inactivate viral nucleic
-------
393
acids has been cause for concern, although evidence of an
actual hazard for man posed by these entities has not been
provided.
(v) Use in Water Treatment. Although much of the
research on the biocidal properties of iodine has been
directed toward its potential use for drinking water disin-
fection, actual field studies have been very limited. ^Black
and cowqrkers (1965) found iodine was an effective disin-
fectant'in the water supplies at three Florida correctional
institutions*" serving a population of about 700 persons.
Although coliform bacteria were usually present in the un-
treated water, fewer than 1 percent or rne samples treated
with iodine were positive for coliforms. They concluded
that 1 mg per 1 of iodine rendered the water supply safe
under a variety of conditions including pH levels up to 7.5.
(vi) Conclusions. The available evidence indicates
that iodine is an effective disinfectant for water. Iodine
is probably somewhat less efficient than hypochlorous acid,
but maintains peak disinfectant activity even in mildly
alkaline water and is less subject to demand because it is
less chemically reactive. The main deterrents to its use
appear to be high cost and concern for yet-to-be-demonstrated
adverse effects of long-term exposure to iodine on thyroid
function (Kinman, et al., 1970). Although interest in
iodine as a disinfectant has been restimulated as a result
of the concern about chloroform and other trihalomethanes
formed by water chlorination, studies on the formation of
iodine-containing trihalomethanes apparently have not been
published. Needed research on iodine as a disinfectant for
' drinking water principally concerns health effects and
chemistry, rather than microbiology.
6. Summary
Pathogenic microorganisms are often present in waters
that are contaminated with domestic sewage. It is essential
to consumer safety that waterworks produce safe drinking
water even from potentially contaminated raw water despite
the demand for ever-increasing quantities. In this con-
nection, the different treatment processes applied in the
preparation of drinking water have been evaluated from the
standpoint of their capacities to remove microorganisms.
The most generally applied treatment processes, which have
-------
394
been discussed in the previous sections, include: storage
in reservoirs, coagulation, rapid and slow sand filtration,
dune infiltration and underground storage, activated carbon
treatment/ chlorination, and ozonation.-
These processes can be evaluated on the basis of their
abilities to remove bacterial indicators of fecal contami-
nation (viz., thermo-tolerant coliforms) and pathogenic
agents that are more resistant than indicator bacteria
(viz., polioviruses of the enterovirus group). A rough
estimate of the removal percentage has been made for each
treatment process, based on the data presented in Topic E
and the cited literature [See Table E.6-1]; these must be
considered only as guideline approximations.
The tabulations show that, depending on the quality of
the raw water, a combination of different treatment pro-
cesses can produce hygienically good finished water. A
proper combination of treatments will result in a high
capacity to remove bacteria and viruses. It, should be noted
that the removal or inactivation estimates for viruses are
based on the polioviruses, which are only three of the more
than 100 types of enteric viruses that may be present in
water. However, in practice, no outbreak of enteric virus
disease has ever been attributed to properly treated water.
Less information is available concerning parasitic
cysts and metazoan parasites. Most of these organisms are
supposed to be removed by sand filtration processes, but in
some cases, sand filters did not effectively remove Schis-
tosoma cercariae from the water. Thirty to 90 percent of
the parasitic cysts can be removed by coagulation. Disin-
fection with chlorine can be accomplished only by doses much
higher than those used in the preparation of drinking water.
The evaluation of the treatment processes with respect to
removal of these organisms is difficult because insufficient
data are available. In practice, however, outbreaks of
parasitic diseases have not been reported in connection with
drinking water that has been adequately treated; coagulation
and filtration steps are essential.
The data presented here and the experience of water-
works over many years lead to the conclusion that a proper
choice of different combinations of treatment processes,
depending on the raw water quality, will result in the
production of finished drinking water that meets the micro-
biological standards. Further research is needed concerning
the behavior or parasitic cysts, metazoan parasites, and
-------
TABLE E.6-1
ESTIMATION OF THE PERCENTAGE OF MICROORGANISMS
REMOVED BY VARIOUS WATER TREATMENT PROCESSES
Residual
Disinfectant
Treatment Process (mg/)
Open reservoir
storage
Coagulation
Sand filtration
(a) Rapid
(b) Slow
Dune infiltration
and underground
storage
...
Activated carbon
Chlorination
(a) Breakpoint 2 - 3 .
(b) Post 0.1 - 0.5
Ozonization 0.3-0.4
Removal Percentage
Contact Fecal
Time Coliform
(min) Bacteria Poliovirus
95 - 99.9 99 - 99.9
50-97 70 - 99.99
10-50 0 - 50
70 - 99 96 - 99.99
. . '--.'
99.99 99.99
10-50
1 - 2 99.99 99.9
15-30 99.9 99 .
2 - 4 99.999 99.99
U)
,. - 10
. ' . .- ui
LAverage detention t^ime 30-75 days
2- not considered.
-------
396
viruses, and to evaluate the threat to human health engen-
dered by the presence of extremely low levels of pathogens
in drinking water.
7. Recommendations
1.
2.
The variety and degree of treatment used in
preparing drinking water should be determined
by the quality of the raw water.
Where the source is variable in quantity or
quality, reservoir storage may be used to
buffer some of the fluctuations.
Disinfection is essential in the production
of safe drinking water.
Physical (e.g., rapid sand filtration) and
biological (e.g., slow sand filtration) pro-
cesses should be employed more extensively prior
to terminal disinfection so as to increase
disinfection efficiency and reduce the risk of
formation of toxic substances.
-------
397
F. DISTRIBUTION SYSTEMS
With the application of standard water treatment tech-
niques, water suppliers usually have no difficulty in pro-
ducing potable water supplies of a quality that will meet
the WHO European Standards for Drinking Water. During
distribution through the supplier's network and within the
consumer's plumbing system, this water is exposed to a
variety of hazards.
A water supply distribution system is not a totally
sealed entity, so the possibility of contamination from
external sources must be considered. It is easy for con-
sumers to make connections into the system and either acci-
dentally or unknowingly introduce hazardous situations.
Equipment and fittings of numerous types are necessary for
the effective use of a water supply, and the design of these
must be considered in relation to potential hazard.
The multitude of possible materials that coul?
for the construction of pipelines and joints, all _ the water
utilities' apparatus, and all the consumers fittings is
increasing at a rapid rate. The possible effects of all
thSse materials on the quality of the water need considers-
tion.
Some of the problems that have occurred from time to
time in distribution systems have been public health prob-
lems; more commonly, they lead to complaints of taste, odor,
color, or turbidity resulting from microbiological growth
which have no public health significance, but which result
In water quality that does not conform to national or inter-
national standards. This section examines some of these
problems and ways of overcoming them.
-------
398
1. Service Reservoirs
Service reservoirs are an essential part of any water
supply distribution system, but they are frequently the
weakest link in the system and are subject to several
hazards. Reservoirs of this nature are known by different
names in different countries, but the term is used here to
include any treated water storage facility imposed between
the water treatment works and;the consumer. They may be
underground concrete or brick structures, elevated tanks or
towers, or possibly even open reservoirs suitably protected.
a. Direct Contamination from Sewage. The siting of
sewers and service reservoirs in relation to each other is an
essential feature of engineering design. Reservoirs should
be on high ground so that surface water drains away from
them and all sewers should be at a lower level. If a reser-
voir is on a slope, construction of buildings on higher land,
with its attendant sewage and drainage hazards, should be
avoided. Where there is any possible risk of leakage from
local sewers, these should be laid with special precautions,
such as in cast iron or encased in concrete. Risks from
storm overflow should be avoided.
b. Contamination of Open Service Reservoirs. Once
water has been treated, it should be fully protected and not
see the light of day until it reaches the consumer's tap.
This implies that all service reservoirs, tanks, and water
towers should be covered, although there are known to be
many open service reservoirs still in use in several de-
veloped countries.
The first and most obvious source of contamination of
open reservoirs is from birds, including all kinds of water
fowl, especially gulls. There have been several studies of
pollution from gulls in different parts of the world, all of
which have shown that a high proportion of these birds are
carriers of salmonellae, especially in the more highly
developed areas where they are in closer association with man
(Jones, est a.1^. , 1978; Luttman, 1967; Metropolitan Water
Board Report, 1966 and 1970a).
Elevated tanks and water towers with roofs also often
provide attractive nesting sites for birds of all kinds, so
that access to roof spaces must be adequately protected from
small birds by netting which must always be kept in good
-------
399
repair. Ventilating shafts of covered reservoirs.must be
similarly protected.
Flying insects can also be a nuisance and these can be
excluded by covering the bird netting with fly screens. If
only one ventilating shaft is installed, in a building at
the side of a reservoir, the screening required will be
small in area and more readily maintained.
Many waters exposed to daylight can also support algal
growth of one sort or another; problems can occur, where they
have not been encountered before, due to the increasing need
in many developed countries to utilize lowland, nutrient-
rich rivers as sources of water supply. The algae can be
controlled only by excluding daylight with adequate roofing.
c. Contamination from Leakage. A very common cause .of
bacterial contamination of service reservoirs is rain
water penetrating through leaks in the roof. This is more
common in older, brick-roofed reservoirs than in modern
reinforced concrete structures. Leakage is enhanced by the
penetration of plant roots and for this reason trees should
not be planted on or alongside service reservoirs.
Contamination is frequently worse after heavy rain
following a dry spell. A similar quantity of rain on an
already saturated soil does not produce the same effect.
This is probably due in the one case to penetration of
polluted water through cracks in the dry soil and in the
other case to the moist soil acting as a bacterial filter.
Contamination of this type frequently includes Esche-
richia coli or thermo-tolerant coliforms as well as other
coliforms, presumably derived from birds, other wild animals,
or domestic pets. The extent to which this type of contam-
ination is detected will depend, considerably, on the method
of sampling and the rate of water turnover in the reservoir.
If sampling is carried out only once a week and the rate of
turnover is high, the pollution resulting from one heavy
thunderstorm can be gone before it is detected. Further-
more, as the contamination is from above, the highest con-
centration of polluting organisms is found near the surface
and may be undetectable at the bottom drawoff point. Dip
samples taken at the surface c'an, therefore, give results
totally different from those of samples taken from the
outlet main. Where this problem occurs, samples should be
taken more frequently and at different points or depths in
the reservoir.
-------
400
Pollution of this nature can only be prevented by major
repairs to the roof, either by complete reconstruction or by
applying a waterproof coating, covering, or membrane which
should extend down the walls below normal water level. In
some cases, where minimal numbers of coliforms persist over
a long period of time or occur relatively frequently during
the year, long-term major repair or reconstruction programs
may be necessary. The degree of risk involved must then be
assessed in relation to priorities for capital expendi-
tures .
<3. Growth in Bottom Deposits. Because service reser-
voirs contain relatively still water, any suspended matter
has an opportunity to settle as a sludge on the bottom.
This suspended matter may be derived directly from inade-
quately treated water, due either to incomplete removal of
silt, to filtration problems during algal blooms in the raw
or stored water, or to post-precipitation of aluminum hy-
droxide. It may be generated in the trunk mains to the
reservoir by the rusting or corrosion of the iron or steel
mains or from wastes produced by small animals, such as Gam-
marus anS Asellus, in the mains. It may also be produced~^y
the precipitation of calcium carbonate, due to the loss of
carbon dioxide, from high bicarbonate waters in the reser-
voir itself. Small animals may also live in this sludge and
further modify its properties. Easily recognizable con-
stituents of these sludges include calcite. crystals, rust
particles, amorphous siliceous material, Asellus fecal
pellets, and organic debris. :
Studies of these sludges have shown that they contain
an abundant microbial flora which is quite different from
that of the overlying water. Coliform organisms and E.
co-Li are quite common in these sludges, even where tHey
have never been found in routine examinations of the over-
lying water. Clostridium perfringens is also common, as are
various fungi adapted to an aquatic environment. Some
species of actinomycetes may be abundant, depending on the
nature of the sludge, the water, and the water temperature.
Protozoa usually are also abundant (Metropolitan Water Board
Report, 1970b and 1973a).
Although these sludges seem to provide an environment
for the development of particular ecological communities,
the$bacteria so produced do not seem to migrate in sig-
nificant numbers to the overlying water. It is possible
that certain by-products of microbial metabolism, notably
-------
401
the odoriferous substances produced by fungi, could diffuse
into the overlying water to cause a deterioration in quality,
especially in taste.
The engineering design of reservoirs should be such
that the water circulates along the floor and prevents
static conditions that lead to accumulation of sludge, but
this is not easy to arrange. Alternatively, the bottom
layer should be left undisturbed, but cleaned out periodi-
cally when the reservoir is drained. This usually needs to
be done at intervals of ten years or less.
2. Aftergrowth
Current methods of drinking water treatment in developed
countries normally achieve a satisfactory bacterial quality
in the water leaving the works and entering the distribution
system. However, under some circumstances, growth of bac-
teria can occur in the distribution system after the water
has left the works and this is usually referred to as after-
growth.
a. Influence of Nature of Raw Water and Methods
of Treatment on Aftergrowth. In order to grow, most common
bacteria require the basic minerals that all living organisms
require and a source of organic carbon, There are very few
raw source waters used for public water supplies that do jiot
contain adequate minerals in solution for the growth of a
variety of bacteria, although different waters may favor
different species according to the-ir mineral composition.
However, source waters differ widely in their content of
organic carbon and, hence, in their ability to support
microbial growth. Many underground waters are very low in
organic .matter and unable to support significant- bacterial
growth even though they may be quite rich in minerals [see
Section A.I.]. ,
Surface-derived waters, on the other hand, are rela-
tively rich in organic matter and can support significant
microbial growth; this microbial growth is an essential
factor in the self-purification of rivers [see Section A.2].
If this organic matter is not removed during treatment, it
remains available and can support microbial growth thereafter,
provided that conditions are otherwise suitable for growth.
The major part of the organic matter in natural waters
consists of humic compounds that are relatively resistant
-------
402
to biodegradation and only slowly utilized for bacterial
growth, mainly by specialized groups of bacteria. The by-
products of the initial breakdown are then used by other
bacteria.
The more readily biodegradable carbohydrates, pro-
teins, and protein breakdown products are less abundant in
_, water, but when present, can support growth of numerous
bacteria. The various water treatment processes will remove
a proportion of this organic matter. Alum coagulation [see
Section E.2] processes bring about the inclusion of humic
substances in the alum floe. Slow sand filtration provides
suitable [see Section E.3] conditions for bacterial growth,
and biodegradation of organic matter; that which is not
readily biodegradable will not easily support microbial after
treatment.
These treatment processes are almost invariably fol-
lowed by an oxidation process, usually chlorination [see Section
E.S.a] and sometimes ozonation [see Section E.S.b.]. During
this oxidation process, which is primarily for disinfection,
any remaining organic compounds which may be relatively
resistant to biodegradation are liable to become partially
oxidized to intermediate products that are more readily usable
as bacterial nutrients and may thus, be a potential stimulus
to aftergrowth.
Ozone, in particular, has often been claimed to have
this effect. One of the major, advantages of ozone treat-
ment is that it oxidizes the humic compounds, which are
yellow-brown in color, and thus, produces a more attractive
looking water. This effect, coupled with the instability of
ozone and the inability to maintain a residual in the
distribution system, enables aftergrowth to occur. It has
been shown, however, that chlorination can give rise to a
similar aftergrowth if no residual is maintained (Metro-
politan Water Board Report, 1973b).
Whether the basic disinfection process is chlorination
or ozonation, aftergrowth of this nature can be limited by
maintaining a suitable chlorine residual in the distri-
bution system. In some countries and some areas, however,
consumers object to the taste of chlorine. Complete pre-
vention of aftergrowth requires a substantial and often
unacceptable chlorine residual unless, as a minimum, a rigid
line flushing program and treatment to minimize pipe cor-
rosion are also practiced. Attempts to resolve this problem
with chlorination alone are totally inadequate to prevent
growth on some materials which have a chlorine demand, or in
deposits of organic substances, particularly in protected
-------
403
situations such as pipe joints, where chlorine cannot pene-
trate effectively and where the organic material present may
bind the chlorine. Beulow and Walton (1971) reported that
an increase from 0.2 to 1.5 mg of chlorine per liter was neSis^rv to
improve the coliform quality of water in a distrioutiori
system. Reduction in colony counts could be expected to
require at least as great an increase in chlorine level.
Lee (1971) showed under experimental conditions that, in the
presence of nutrients, the same bacterial growth was ob-
tained with 0.5 mg per 1 of chlorine .(80 percent as free
chlorine)j as with no chlorine at all. Under the same condi-
tions, 1 mg per 1 suppressed the growth of bacteria.
It has-also been suggested, however, that if ozonation
precedes a biological treatment process such as slow sand
filtration, the bacteria will utilize the partially oxidized
organic matter in the filter, producing a further improve-
ment in the water quality and preventing aftergrowth in the
distribution system. Chlorination before distribution would
still be necessary. This principle has been applied in the
Netherlands, where ozonation is followed by filtration to
enable the aftergrowth, and the biodegradation associated
with it, to occur in the filter instead of in the distri-
bution system.
The bacteria that cause aftergrowth are usually those
which prefer lower temperatures. This is reflected in the
difference between bacterial counts using the two standard
methods of bacterial colony counting employed in the water
industry, namely, with incubation at 20 to 22°C and at 35 to
37°C. Aftergrowth is much more evident with incubation at
20 to 22°C and extension of incubation to seven days pro-
duces even greater differences. In lowland river-derived
waters which receive treated sewage effluent, it is not
unusual for colony counts (at 20 to 22°C) that are less than
100 per ml in water leaving the treatment works, to increase
to over 10,000 per nil in a distribution system where only a
low chlorine residual is maintained. These increases are
directly related to temperature and time, being greater in
warm weather and where 'the water has remained for long
periods in service reservoirs. The draft EEC Directive
relating to the Quality of Water for Human Consumption
recommends some very low colony count guide levels that
appear to have been formulated without regard for the prob-
lems of aftergrowth in the distribution system. in certain
areas, some drastic changes would be needed to achieve
these levels.
-------
404
The bacteria that predominate in aftergrowth are not
easy to characterize, for they are difficult to maintain in
pure culture. Furthermore, determination of their bio-
chemical properties often results in a classification of
bacteria into groups that are hot taxonomically precise-
However, a major part of the colony count is usually assign-
able to the rather ill-defined Flavobacterium group, Pseudo-
monas, Aeromonas, Arthrobacter, Caulobacter, Cytophaga, and
actinomycetes also form a small part of the population (Van
der Kooij and Zoeteman, 1978).
The bacteria from lowland, surface-derived water that
form colonies at 37°C are usually survivors of treatment,
consisting largely of the chlorine-resistant spores of the
aerobic sporeforming bacilli.
A technique for determining the concentration of growth
stimulating organic compounds in water, in order to be able
to quantify the aftergrowth potential of drinking water and
to measure the effect of different water treatment tech-
niques on these compounds, has been developed in the Nether-
lands at Keuringsinstituut vodr Waterleidingartikelen (KIWA).
Maximum colony counts of pure bacterial cultures developing
in water samples were expressed in terms of the concentra-
tion of assimilable organic carbon (Van der Kooij and
Zoeteman, 1978).
By using a medium containing citrate as the sole carbon
source. Van der Kooij (1977) also showed that treated water,
before entering the distribution system, contained a small
percentage (0.1 to 1 percent) of citrate-utilizing bacteria,
mainly Pseudpmonas and some Aeromonas, but that tap water
contained a higher percentage (1 to 10 percent and sometimes
more). It is suggested, therefore, that these organisms can
be used to give information on the efficiency of substrate
removal by water treatment.
b. Accumulation of Organic Debris in Dead Ends and
Other Protected Area's"! The causes and effects of accumu-
lation of sludges in service reservoirs are described in
Section l.d. Similar accumulations of debris occur in the
distribution system wherever there is little or no flow of
water, such as in dead ends or in any disused apparatus that
has not been disconnected. The bacteria and fungi that grow
in these situations, and their by-products, are continually
seeded into the water that flows past the dead end, causing
a deterioration in bacteriological and other quality charac-
teristics, especially taste and odor..
-------
405
A similar effect occurs, even in the absence of dead
ends, where a district is supplied from two sources that
are^fed into different ends of a system. At some inter-
mediate points where the pressures balance, there will be
little or no flow in the main, and a similar situation will
develop. In addition to the "dead end" effect, this situa-
tion contains the added hazard that changes in the supply,
caused either by repair work or the introduction of new
treatment works or extensions to existing works or mains,
may alter the pressure balance. The areas of little or no
flow will then shift and the accumulations of debris will
become stirred up, resulting in turbid or colored water.
These problems can be overcome by designing distri-
bution systems to avoid dead ends and by determining the
sections "where little or no flow occurs, then ens.uring that
adequate flow does occur at sufficiently frequent intervals
to keep them clean by means of valves that control the
direction and rate of flow. Where dead ends already exist,
these must be adequately flushed through hydrants situated
at their extremities. Furthermore, adequate monitoring of
the water quality in the distribution system should include
regular sampling from such dead ends, for they are likely to
contain the worst quality water. Most developed countries
have regulations, codes of practice, or make recommendations
about frequency of sampling in the distribution system, and
sampling points are defined in a variety of ways. All of
the matters discussed in Topic F should influence the choice
of sites and the frequency of sampling.
c. Interrelationships Between Organic Matter, Bac-
terial Growth, and Animals in the Distribution System. The
problems of animals in the distribution system are not
relevant to a report of microbiology, except insofar as the
two topics are interrelated. Animals will only become a
nuisance in distribution systems when they can multiply to
relatively large numbers. To do this, they require a source
of food which must usually be organic particulate matter",
either living or dead; unlike bacteria, they cannot live
directly on organic matter in solution. However, if the
organic matter in solution supports growth of bacteria and
fungi, these organisms can be used as food by protozoa, and
many animals can feed on bacteria, fungi, and protozoa.
Therefore, conditions that discourage the aftergrowth of
bacteria should discourage the growth of animals.
Animals are very rarely a nuisance in distribution
systems supplied with slow sand filtered water. The paucity
-------
406
of animals has been attributed to the greater efficiency of
slow sand filters in removing the animals' ova from the
water. However, the true explanation may" be quite different
Nearly all distribution systems contain a small nucleus of
animals, but they are generally unable to grow in slow sand
filtered water, probably because the bacteria on which they
feed have grown and remained in the filter bed, so that the
food chain leading to growth of animals in the distribution
system cannot get started.
d. Chemical Changes in Distribution Systems. Changes
in the physical and chemical characteristics of water during
distribution may be chemically induced, such as the leaching
of heavy metals and other materials from pipes, pipe joints,
or pipe linings. They may be, on the other'hand, micro-
biologically induced, such as the production of moldy,
musty, or earthy tastes resulting from the by-products of
the growth of fungi or actinomycetes on susceptible organic
materials in the distribution system. More commonly, how-
ever, there may be a complete interaction between micro-
bially induced changes and chemical reactions.
(i) Leaching of Heavy Metals. The problems associated
with leaching of lead from lead piping are well known.
There is now some concern over the possibility of leaching
of lead, and to a lesser extent of arsenic and cadmium, from
copper alloys containing lead and from solders and galva-
nizing. The copper alloys are used to overcome dezincifi-
cation corrosion problems with brass. The leaching of lead
from solder seems to be closely associated with the nature
of the flux used. Arsenic is a common ingredient, at very
low concentrations, in brass and cadmium and may be present
as an impurity in the zinc used for galvanizing. Although
these subjects are being investigated in various labora-
tories, little published information is yet available. The
use of lead and tin compounds as catalysts in production of
some plastic pipes is controlled by national and inter-
national standards. All these possibilities should be
borne in mind when investigating trace heavy metal problems.
(ii) Precipitation and Depositions. The information
in this and the following paragraphs (iii), (iv), (v), and
(vi) is derived mainly from two reviews by the Water Research
-------
407
Association in the U.K. (Ainsworth, Ridgway, and Gwilliam,
1978). Deficiences in treatment are probably the major
cause of dirty water. High levels of residual coagulant,
together with the materials that it is intended to remove,
namely, particulate matter, algae, and organic matter, will
cause such problems; silica and manganese may also contri-
bute. Gradual reductions in their concentrations through a
supply indicate that these substances are settling out in
the distribution system. Subsequent disturbance of these
loose deposits is a common cause of objectionably dirty
water.
Deposits formed by precipitation can also contain
appreciable amounts of lead, nickel, and sometimes zinc and
cadmium derived from waters with heavy metal concentrations
well within-the WHO recommended limits. These metals are
probably being concentrated in the precipitates by chelation
with organic matter and/or adsorption to hydrous iron and
manganese oxides. Adsorption increases with pH and is
reversible, so that pH changes could desorb heavy metals
from precipitated organic complexes.
(iii) Changes in Source Water. Accumulated deposits
of any kind are liable to undergo solution or disintegration
if water from a supply with different chemical characteris-
tics enters the distribution system. The growing tendency
to obtain water from different catchments, and to alternate
between different sources, has made this an increased pos-
sibility and probably the most important consideration when
investigating the quality of alternative water sources for
use in existing distribution systems. The most obvious
result is a persistent dirty water problem, but the pos-
sible release of precipitated heavy metals must also be
considered. .
(iv) Total Organic Carbon. Even in a well-run treat-
ment facility which relies on non-biological processes,
soluble organic matter will remain in the treated water from
a surface-derived supply at levels up to 5 mg per 1 of total
organic carbon (TOCJ. A decrease in TOG, which will be
greatest in lowland surface waters with initially high
concentrations, may occur during distribution. Residual
iron and aluminum, and iron produced by corrosion, will
coagulate some organic material and some will interact with
existing deposits and presumably be adsorbed. Further
removal occurs by microbial metabolization. This leads to
-------
408
decreases in dissolved oxygen, and more oxygen removal may
result from corrosion.
(v) Microbially Induced Changes. These have been
reviewed in detail by Hutchinson and Ridgway (1977). Some
of these have been discussed in other sections of topic F.
After some preliminary studies, these workers used a seven
day, 22 °C colony count in preference to a three day count,
largely for the sake of convenience and greater flexibililty
in sampling days. In a general survey of distribution
systems, they found, overall, that 40 percent of the samples
had a count exceeding 10 per ml and 20 percent exceeded
10 per ml. The higher counts were commonly associated with
dirty water problems. In lowland, surface-derived waters
there was a marked trend for an increased incidence of other
microorganisms, including fluorescent pseudomonads, anaero-
bic bacteria, sulphate-reducers, and, sometimes, the
organisms found in the 37°C colony count. Microbial activity
produced water quality changes such as depletion of dis-
solved oxygen, increases in total organic carbon, objec-
tionable tastes and odors, greater corrosiveness, and the
incidence of animals.
Local accumulation of biomass, whether in dead ends, at
the bottom of service reservoirs, in pipe joints, or on
unsuitable materials used in fittings, may result in anaero-
bic conditions. In these circumstances nitrate or sulphate-
reducing bacteria and anaerobic sporeforming bacilli may
grow (O'Connor, et al., 1975; Willis, 1957).
Nitrate-reducing bacteria are common in tap water, but
they are unable to reduce nitrate at the dissolved oxygen
concentrations normally found in tap water. Ammonia-
oxidizing bacteria that are capable of producing nitrites
may also be present. The reaction seems to be time-dependent
under fast flow conditions, a decrease' in nitrite concen-
tration was also shown to occur, presumably due to produc-
tion of nitrate by nitrite-oxidizing bacteria also present
in the water; however, techniques for detecting these
organisms are difficult (Victorin and Stenstrom, 1975).
Substances derived from otherwise harmless bacteria
may cause pyrogenic reactions in a patient when drinking
water containing high numbers of bacteria, due to after-
growth, is used for some highly specialized purpose, such as
in artificial kidney machines.
-------
409
(vi) Corrosion. Corrosion products in iron mains are
usually in the form of tubercles composed of a thin shell of
the very hard mineral, magnetite (Fe^O.). A mixture of
hydrated ferrous oxides and iron suifides is usually found
under this shell, while the surface in contact with the
water is a thin layer of goethite ( ~ Fe.O.OH) in con-
junction with substances adsorbed from the water, such_as
organics, silica, and manganese. The presence of suifides
indicates the involvement of sulfate-reducing bacteria in
the corrosion process, although the extent of this is still
open to conjecture [See Section G.4.b]. High corrosion
rates are best diagnosed by inspecting pipe interiors and
eliminating all other possible causes of discoloration
before assuming that bacteria are responsible.
3. Growth on.Materials
a. Ability of Various Microorganisms to Grow on Some
Materials Used in Distribution and Plumbing Systems.It
has been stated in Section 2.a that most common bacteria and
fungi require mineral salts and organic carbon for their
growth and that most water supplies contain adequate mineral
salts in solution. In the construction of distribution
systems, and in consumers' plumbing and fittings, many
organic carbon materials are used, and the number and variety
of these is rapidly increasing with the growth of the plas-
tics and synthetic chemicals industries. It is, therefore,
necessary to consider whether any of these materials are
capable of supplying organic carbon compounds, that support
microbial growth, to a sufficient extent to cause water
quality problems, either directly by increasing bacterial
numbers or indirectly from their by-products, inducing
objectionable taste, odor, or even toxicity (Burman and
Colbourne, 1979).
(i) Nature of Materials that Can Support Growth. The
range of materials containing organic carbon compounds that
have been employed in water supply systems includes natural
and synthetic rubbers, paints and coatings, pump and valve
gland packings, lubricants, sealants, jointing compounds,
and soldering fluxes. They also include a wide variety of
plastics such as polyethylene, polypropylene, polyvinyl-
chloride (PVC), polyamide (nylon), polytetrafluorethylene
(PTFE), acirylonitrile - butadiene - styrene (ABS), glass
fiber reinforced polyesters (GRP), acrylics, acetal copoly-
mers, polycarbonates, silicones, polyurethanes, polysulfides,
-------
410
and many others, all of which exist in a wide variety of
modifications with different characteristics for different
purposes? some traditional materials such as leather washers
and jute or hemp yarn are also present.
Deterioration in bacteriological quality of tap water
due to the growth of coliform organisms on vegetable tanned
leather washers was first described by Houston (1916). The
bacteriological problems caused by the use of jute yarn as a
caulking material for lead run joints in water mains has
been well documented (Windle Taylor, 1947). Any other
natural, unmodified product of plant or animal origin in
contact with water, including any cellulosic material,
linseed oil and other vegetable or animal oils and greases,
soaps, and shellac can be expected to support microbial
growth. Many mineral oils, greases, and waxes can also
support growth of a more limited range of microorganisms.
Natural rubber is in a different category. Natural latex
will support abundant growth of a variety of bacteria and
fungi, but many compounded rubbers are much more resistant
and do not support sufficient growth to have a significant
effect on water quality. However, long-term resistance to
biodeterioration is a different problem [See Section G.I].
With some notable exceptions, the basic high molecular
weight polymers used in plastics manufacture do not support
microbial growth. The exceptions are the soft polyurethanes
and polysulfides both of which can support considerable
microbial growth resulting in biodegradation of the product
(Evans and Levisohn, 1968; Jones and Campion-Alsumard, 1970;
Metropolitan Water Board Report, 1970c). However, nearly
all plastics contain a variety of other low molecular weight
carbon compounds essential for their manufacture or neces-
sary to provide properties for particular uses. These
include catalyzers, antioxidants, plasticizers, fillers,
extenders, pigments, sizes, bonding agents, lubricants, mold
release agents, etc. Excess monomer may still be present, .
due either to inadequate polymerization or to a deliberate
excess. Overheating in molding a fitting can also result in
partial degradation with formation of other low molecular
weight compounds. In the production or use of some polymers,
a volatile organic chemical may be formed that can support
microbial growth and must evaporate by adequate curing if
problems are to be avoided when in contact with water. The
production of acetic anhydride in some silicone jointing
compounds is a good example of this. The materials most
likely to cause problems are the plasticizers, including
phthalates and sebacates, oils and waxes, cellulosic fillers
-------
411
such as wood-flour, or cotton fibers, excess monomer, espec-
ially styrene, and bonding agents incorporated with various
fillers such as glass fiber or pigments.
Synthetic detergents that support microbial growth can
also be present in some products, either as residues from
precipitation of a polymer from an emulsion in a detergent,
or from the washing of fibers used in packing materials.
(ii) Nature of Microorganisms that Can Grow. To
maintain national and international standards for bacterio-
logical quality of water at the consumer's tap, the water
bacteriologist will be interested in any organisms which
affect his standard, quality tests. These will include
coliform organisms and any of the bacteria which can produce
colonies in the standard 37°C or 22°C colony counts. Many
of the materials under discussion have been shown to support
the growth of coliform;organisms and mixtures of organisms
capable of growth in the colony counts. Many of them can
also support the growth of Pseudomonas aeruginosa and various
fungi. Growth is often sufficiently abundant under test
conditions to form slimes on the material or clumps or
flakes of growth suspended in the water which are readily
visible to the naked eye. These slimes, however, often
consist of bacteria which do not grow readily in the stan-
dard colony counts. This is probably because the carbon
compounds on which they are growing, and to which they have
become adapted, are totally different from the carbon
sources contained in the standard colony counting media.
Numerous protozoa usually develop in these slimes, and
nematodes may eventually grow there as well.
b. Problems that Can be Caused by Microbial Growth.
Apart from failure to reach bacteriological standards, when
tested by standard national or international methods, there
are a number of additional problems that can arise. These
problems are all accentuated at warm temperatures and under
conditions of little or no flow. In modern buildings, for
aesthetic and architectural reasons, hot and cold water
pipes and central heating pipes are often installed in
common ducts, with the result that the cold water has the
opportunity to become warmed. In business premises unoccu-
pied on weekends, there is no flow of water, thus, cold
water temperatures can exceed 30°C and allow fungi and
bacteria to grow much faster. However, warm water alone will
not support the growth of organisms as described here.
-------
412
Vending machines for drinks are often installed at the end
of long lengths of narrow piping through warm factories
separate from the rest of the supply system so that the only
flow of water is that required for the cups of tea, coffee,
or soft drinks. Drinking water fountains are often installed
in similar situations and frequently give rise ^° G°mp^tS>
Unused apparatus is often disconnected and the branch pipe
left attached where there is no flow. These should always
be removed at the junction with the main pipe^r
(i) Taste and Odor. The growth of fungi and possibly
of actinomycetes on materials, especially in consumers
Dlumbinq systems, gives rise to musty, moldy, or earthy
tastes in the water. In the Metropolitan Water Division, .
Thames Water Authority, London, it has been found that
approximately 50 percent of all complaints refer/Jd^oof e
Laboratories are of bad taste. In approximately half of
these, there has been a significant increase in fungi, some-
times in conjunction with high bacterial colony counts or
with the presence of P. aeruqinosa (Burman and Colbourne,
1976). .
(ii) Visible Growth. Under test conditions, visible
growth is frequently observed on materials. This usually
becomes very bad before a consumer notices it. Clumps of
fungal growth on plasticized PVC tubing in dr ^k vendi ng_ _
machines sometimes detach and are discharged into the Brinks.
Accumulations of slimes have also been observed in tanks and
ciSternfconstructed of unsuitable materials. Plasticized
PVC capillary tubing used to supply water to dentists
drills has become blocked with visible growths of fungi.
The aerator or antisplash devices, again usually of plasti-
cized PVC, often fitted to kitchen taps, are a frequent
cause of complaint in which slime growths can readily be
observed.
(iii) Health Hazards. It has generally been assumed
that although the by-products generated by bacteria and
fungi growing in distribution and plumbing systems, can cause
complaints of deterioration in quality, they do not constitute
a health hazard. However, there are two possible health
hazards which should be considered.
First, the significance of P. aeruginosa must be con-
sidered. This is an bpportunist pathogen of ev!s *n*<;af S
and causes infection of wounds and burns. It is P^ticu
larly undesirable in water supplies to hospital burn units
-------
413
and to tonsillectomy wards (Lowbury, ert But. , 1970). P_.
aeruginosa contamination of ice served in drinks has been
incriminated in serious post-operative infections in ton-
sillectomy patients (Newson, 1968) . This organism may
rapidly multiply at low temperature and is undesirable in
cooling waters, particularly in dairies. By forming an
enterotoxin, P. aeruginosa can give rise to severe diarrhea
(Muller, 19 74T- There are currently no international stan-
dards for P. aeruginosa in drinking water, but they are
under discussion in a number of countries. German drinking
water legislation considers J^. aeruginosa as a pathogen, and
according to paragraph 1 of the German Drinking Water
Ordinance (Aurand, et al. , 1976),, drinking water has to be
free of pathogens. ~~The~absence of J?. aeruginosa in swimming
water will be regulated by a law, currently in .preparation;
a standard method for enumeration of JP. aeruginosa is speci-
fied. Water leaving the treatment works is normally free
from this organism, so its appearance at the consumer's tap
is an indication of deterioration in water quality [See also
Sections B.l.a(ix) and C,2.c].
Secondly, the significance of the growth of fungi must
be considered with regard to the possible production of
mycotoxins or of allergens. Active toxins (mycotoxins) have
been associated with massive growths of fungi on food
products and there is no evidence that growth of fungi in
water cannot produce similar effects.
Exposure to fungal spores inhalation in
industries has given rise to respiratory allergies. A
"bathing sickness" with fever and respiratory symptoms
following the taking of hot baths, has been described in
Sweden (Atterholm, et al. , 1977). This has been shown to be
produc.ed merely by breathing the air above hot water, through
volatile substances released from the water. It has been
associated with water from two areas, one a filtered, surface-
derived water and one an underground water. Although fungi
produce volatile by-products readily detected by their odor,
and released when water is heated, no abnormal counts of
fungi or other microorganisms have been obtained from this
water. However, similar reactions have resulted at sauna
baths where abundant growth of fungi has occurred on the .
wood: precipitating antibodies against an antigen extracted
from the fungal culture could be found in the patients' serum.
c. Methods of Testing Materials. The most reliable
methods for testing the effects of materials on the quality
-------
414
of the water with which they come into contact, are based on
attempting to simulate the worst conditions under which the
materials are likely to be used and to observe the effects
on the test water. This is achieved by immersing the mate-
rial in limited quantities of water kept at optimum tempera-
ture for microbial growth and changing the water at inter-
vals of three or four days.
The overall methods used in the U.K. have been described
by^urman and Colbourne (1979) and have been written as a
British Standard draft which has been circulated for public
comment. The microbiological methods have also been described
by Burman and Colbourne (1977). The testing procedure
involves consideration of three areas of study: (1) ability
to support microbial growth; (2) effect on organoleptic and
physical properties, especially taste, odor, color, and
turbidity; and (3) toxicity, including toxic metals.
In the PRG, a test has been developed using a strepto-
mycin-resistant mutant of I?, fluorescens, £. cepacia, or P.
putida in order to detect substances leached out of plastic
materials that will be used in drinking water supply (Muller,
et al_., 1979). This is in preparation for a standard method
and will be included in the FRG Standard Methods for water,
sewage, and sludge examination (1979).
Although this section deals with testing of materials as it relates
to microbiology, it would not be complete without references to the non-
microbiological test procedures.
(i) Testing for Microbial Growth. As the test involves
quantitative comparison with a control, it is necessary to
standardize the test procedure as far as possible. Con-
tainers ^must be specially cleaned and the test carried out
in premises free from atmospheric pollution by volatile
organic solvents. A sample having a specified surface area
is immersed in a standard volume of dechlorinated tap water
and inoculated with a mixture of microorganisms by adding
some polluted river water. This is incubated at 30°C and
the water is changed twice weekly by replacement with de-
chlorinated tapwater only. This process continues for six
to eight weeks. Negative controls containing glass samples,
and positive controls containing paraffin wax, are simi-
larly treated. Beginning the fourth week, the water is
tested before it is discarded, for 378C and 22eC colony
counts, coliform organisms, ]?. aeruginosa, and fungi; the
-------
415
sample and water are also observed for visible growth.
A material is considered to have passed the test if, in
three consecutive water samples, coliforms and P_. aerugi-
nosa are not detected, the colony and fungal counts are less
than ten times that of the glass control, and there is no
readily visible growth on samples or in the surrounding
water. This is a very brief summary of the procedure
described by Burman and Colbourne (1977). It is claimed,
after comparative trials in other laboratories, that the
overall result of passing or failing the test is repro-
ducible, although the group of organisms indicating failure
may vary on different occasions and in different labora-
tories.
There are two obvious non-standard factors in this
procedure: the mixture of .organisms in the inoculum and the
nature of the tap water. It has been shown that, provided
the inoculum contains coliform organisms, P^. aeruginosa, and
a mixture of bacteria and fungi within fairly wide limits,
its precise definition is not necessary. Soil extract,
however, is not satisfactory. Distilled water is not an
adequate replacement for tap water in the test, nor have
satisfactory results been obtained with synthetic tap waters.
Comparable results have been obtained with several surface-
derived, underground, and soft upland waters.
A simple, quantitative alternative to this procedure
has been developed; it replaces the 37° and 22°C colony
counts, the fungal counts, and the assessment of visible
growth by a more accurate and reproducible determination of
total microbial activity based on the measurement of dis-
solved oxygen reduction (Colbourne and Brown, 1979). The
test is carried out by a procedure similar to the growth
test previously described, but in closed containers filled
to the top with tap water saturated with air. This method
still requires counting coliform organisms and £. aerugi-
nosa, as these must be shown to be absent from 100 ml.
Inequalities due to variability in the initial inoculum have
been shown to disappear after the first three weeks of
testing and no significant variation has been observed with
several different types of tapwater (from surface-derived,
underground, and soft upland sources); however, distilled
water is unsuitable. This has become the recommended method
of preference in the U.K.
-------
416
(ii) Testing for Taste, Odor, Color, and Turbidity.
These are usually determined by soaking the test material in
chlorinated water and chlorine free water for 24 h at room
temperature. If any change in taste, odor, color, or tur-
bidity of the water is observed, the water is discarded and
a further 24 h soaking is conducted and tested. This is
continued up to a maximum of seven days as long as any
change is observed. The tastes and odors observed in these
relatively short-term tests are mainly those derived di-
rectly from the .ingredients used in the materials and not
secondary tastes resulting from bacterial or fungal growth.
The ratio of sample size to volume of water must be stan-
dardized, and the intensity and nature of any persistent
tastes or odors require interpretation, for many materials
in common use will produce some persistent detectable change
in taste or odor under these test conditions.
(iii) Toxicity. Determination of toxic hazards of
new materials is much more difficult. There is no one
standard test that can readily be used to assess even acute
toxicity of leachates from materials. Carcinogenic and
teratogenic properties are even more difficult, expensive,
and time-consuming to determine.
Where all the ingredients of a material are known,
useful information can be obtained from toxicology data
banks, but this requires that manufacturers disclose the
necessary details. Even where initial ingredients are
known, all the' subsequent reaction products may not be iden-
tified. The presence of toxic substances as impurities must
also be considered.
Some acute toxicity test procedures, notably cyto-
toxicity tests, have, however, been used as screening tests
and yielded occasional 'positive results with some materials.
The leaching of toxic metals is more easily determined by
the use of standard chemical analytical methods on leachates.
Here again, it is necessary to standardize the aggressivity
of the test water and to decide whether WHO standards should
be applied to the water under test conditions or whether a
dilution factor should be applied.
d. Standards and Regulations Relating to Materials in
Contact With Water.In the U.K., the National Water Council
tests water fittings to determine whether they conform to
the water bylaws and publishes lists of approved fittings.
The non-metallic materials used in fittings are tested for
-------
417
their effect on water quality by the methods described in
Section C and a list of approved materials is published.
The Regional Water Authorities normally require all fittings
and materials used on consumers' premises to be on these
lists in order to ensure conformity with the bylaws.
Many British Standards, and some international stan-
dards, contain clauses specifying that materials used for
fittings in contact with potable water shall be non-toxic;
shall not produce any taste, odor, color, or turbidity in
the water; and shall not support microbial growth. However,
there is no British Standard for testing these properties.
In response to a strong demand, a draft standard has been
prepared and is currently under consideration by a British
Standards Institution Committee.
In Germany, materials used in contact with drinking
water are subject to regulations in the German Food Act and
have to be tested by prescribed chemical and bacteriological
procedures (Anon., 1974).
In the Netherlands, KIWA (Keurings Instituut voor
Waterleiding-Artikelen, 1974) publishes a brochure present-
ing guidelines for the installation of domestic water supply
systems to prevent many of the problems described in Topic
F. These guidelines are currently being revised. KIWA
(1977) also publishes a brochure (in Dutch) on "Protection
against the penetration of foreign compounds in water dis-
tribution systems."
In the U.S., greater emphasis is given to regulating
the durability, mechanical performance, and toxicoiogical
properties of a material, rather than to the effects of
microbial growth. Durability, however, does imply resis-
tance to biodeterioration.
e. Growth on Atmospheric Volatile Organic Chemicals.
A rather special case of microorganism growth on materials
in contact with water is that involving water exposed to air
contaminated with volatile organic chemicals, the commonest
of which is ethanol. This is a common problem in some
industries, but it has also been reported in retail shops,
business premises (especially hairdressers), and even in
homes (Poynter and Mead, 1964).
Ethanol is the most common substrate because it is most
widely used; but any volatile organic chemical, such as
-------
416
methanol, acetone, acetic acids, styrene, benzene, petro-
leum, and paraffins can cause a problem. The nature of the
mixture of organisms that will grow depends on the chemical
nature and characteristics of the water. Mixtures of fungi
and pseudomonads are common, including P.. aeruginosa, and
sometimes aerobic sporeforming bacilli axe present.
The growth usually appears as a gelatinous slime; espe-
cially around the edges of water tanks, around float valves
discharging into the water, or hanging from running or
dripping taps. This gives the impression that the slime is
in the water supply, whereas it does not form until the
water is exposed to the contaminated atmosphere. These
slimes often become quite massive and include numerous
protozoa that feed on the bacteria, and. eventually nema-
todes.
The problem has occurred in the alcohol-distilling and
bottling industries, varnish and paint industries, in museums
using alcohol for preservatives, in buildings under con-
struction where adhesives containing volatile solvents are
used for laying floor tiles, in premises of hairdressers
using alcohol-based hair lacquers and in homes where hair
lacquers are used in unventilated bathrooms, in the printing
industry where solvents are used to clean type, in printing
and duplicating departments in offices, in roof spaces
following treatment of roof timbers with preservatives, in
buildings where petroleum products are stored in drums, in
rooms where glass reinforced polyester products containing
excess styrene are stored, and in numerous similar situ-
ations .
The fittings usuaLly involved are open tanks and float
valves, dripping taps, tea boilers, vending machines for
drinks, or any other exposed water surface. The problem can
occur in rooms adjacent to where the offending chemical is
used or produced, or even in rooms separated by a corridor.
The problem is always worse with poor ventilation. It can
only be overcome by a total exclusion of all exposed water
surfaces from areas where volatile organic chemicals are
likely to pollute the atmosphere, or by thorough ventilation
after use where these chemicals are used only intermittently
for short periods. It is important that buildings in which
volatile organic chemicals are to be handled, be correctly
designed in the first place, for it is much more expensive
to correct this problem after it has occurred.
The phenomenon usually is only a nuisance, making the
water unattractive for use, producing a taste or odor, or
-------
419
making it unsuitable for the purpose required; but in the
printing industry it has given rise to health hazards which
are now recognized by H.M. Factory Inspectorate in the U.K.
Industrial alcohol is used to clean type. Humidity in
printing rooms is maintained, to prevent paper curling, by
spraying water into the atmosphere, often from tanks located
in the printing room and supplied via float valves. Under
these conditions typical slimes develop in the tanks and the
resultant bacteria and/or fungi are sprayed into the atmo-
sphere where the staff are working. This produces an al-
lergic respiratory response that may occur after about 10
min or may be delayed for 2 to 4 h; due to progressive
sensitization, the intensity of the response increases with
subsequent exposure. This could be an allergy specific to
fungal spores, but it may be a response to the general
effect of any foreign protein sprayed into the atmosphere,
for the effect has been observed with slimes composed mainly
of a single bacterial species without any fungi.
4. Ba.cTc-Siphona.ge and Cross-Connect ions
All water supply authorities in developed countries can
quote examples of water supply contamination by back-si-
phonage due to inadequate plumbing, to equipment that pol-
lutes the water during use, or from illegal cross-connec-
tions between different supplies. This type of pollution
may introduce either chemical or microbiological hazards.
In . that water quality in distribution systems is not
being considered elsewhere in this report, it is appropriate
to consider pollution of this type under Topic F, Distri-
bution Systems.
a. Types of Back-Siphonage and Cross-Connections and
Risks Involved. Back-siphonage may be defined as siphonage
of liquid back against the direction of normal flow or
pressure. Reduction in pressure can occur for various
reasons within public water supply mains and in individual
plumbing systems, to the extent that pressure falls below
atmospheric pressure. If outlets are submerged below the
level of a contaminated liquid, or even just above it, the
liquid can be drawn back and pollute the supply, either to
other premises or to other outlets in the same premises.
Therefore, there are three simultaneous requirements for
back-siphonage to occur: the water outlet must be submerged
or just above the water surface; the controlling valve must
-------
420
be open; and the pressure at the outlet must be below that
of the free water level at the fitting.
Submerged outlets occur for a variety of reasons
such as faulty design or wrong positioning of float valves;
inadequate overflows; hoses and other attachments fitted to
taps and showers discharging below sink, bath, or shower
tray level so that they can be submerged in contaminated
water; garden, agricultural, or industrial hoses discharg-
ing into tanks, ponds, or even buckets; or leakage between
primary and secondary heating circuits. It must be recog-
nized that pressure changes will occur in the supply and
cannot be prevented. Sewer flushing systems used in some
countries also impose a special risk under some circum-
stances .
In multi-story buildings, negative pressures can occur
at higher levels due to discharges at lower levels. The use
of booster pumps on consumers' premises can create special
problems because of the reduced pressure that can arise on
the suction side of the pumps.
The risks to which consumers are exposed, in the event
of back-siphonage, will depend on the nature of the contami-
nant that is introduced and the frequency with which it is
likely to occur. In addition to threats to health, such
contamination may make water unacceptable for a variety of
other reasons. The seriousness of risk can, therefore, be
classified and the degree of protection against back-
siphonage can be chosen according to the class of risk. The
Water Research Centre in the U.K. carried out a survey of
back-siphonage risks in homes and reached the conclusion
that, although 85 percent of properties surveyed were at
risk in terms of the requirements laid down in Model Water
Bylaws, the probability of back-siphonage occurring was
really very low (Gilfillan, 1971).
Cross-connections may be defined as connections between
a piped public water supply and any other pipe carrying any
other water or liquid that is not obtained solely from the
public supply. They occur most frequently in industrial
premises, where various process waters are used in complex
plumbing systems, and they are especially significant where
high pressure hydraulic mains are concerned. However,
cross-connections can also occur in mains in the street and
even in homes. Most of them occur outside of the control of
the water authorities and are very difficult to prevent.
There are many recorded instances of illness among factory
staff, and sometimes neighboring consumers, due to cross-
connections .
-------
421
b. Regulations Concerning Back-Siphonage. Most de-
veloped countries have water bylaws, or regulations of some
description, that are designed to prevent contamination,
waste, or undue consumption of the public water supply.
Many of these regulations are specifically designed to
prevent back-siphonage. In order to make adequate regula-
tions and to rationalize existing ones, it is necessary to
have a thorough understanding of potential back-siphonage
situations, the degree of risk involved,, and methods of
preventing their occurrence. To this end, the Department
of the Environment in the U.K. appointed a committee which
produced its "Report of the Committee on Back-Siphonage in
Water Installations" in 1974 (Committee on Back-Siphonage in
Water Installations, 1974).
However, regulations or bylaws are useless without
adequate means of enforcement, including inspection of
installations and schemes for the approval of fittings and
materials, as discussed in Section 3.d of this report.
Surveys have shown that there are large numbers of premises
in which fittings do not comply with the provisions of
existing regulations, that requirements for inspection
following alteration or extension to existing plumbing are
largely ignored, and that present frequency of inspection is
inadequate to identify all the risks. It is recommended in
the Report (1974), that an inspection force be staffed with
one inspector per 25,000 people served. The number of
violations that would be discovered by routine regular
inspections would subject consumers to considerable expense
if immediate corrections were required. It is, therefore,
recommended that special attention be given to installations
where large numbers of people congregate, to industrial
premises, and to buildings with multiple occupancy.
c. Mechanical Devices and Hydraulic Safeguards. The
prevention of back-siphonage is dependent on the incorpor-
ation of adequate hydraulic safeguards during installation
and plumbing alterations and on the availability of reliable
mechanical devices. '
As a general principle, the degree of risk in each
situation should be assessed and a mechanical device or
hydraulic safeguard incorporated which is appropriate to the
risk as well as to the reliability of the device. The
cheapest, most effective, and most reliable means of protec-
tion is some form of air gap. Where the risk is greatest,
such as where a pipe discharges into a container in which
-------
422
the water becomes highly contaminated, an adequate air gap
should be maintained between the pipe and the top of the
container at which water.^ overflows. Where the risk is less,
the receiving container can be fitted with an overflow pipe;
the effective air gap. is the distance between the inlet pipe
and the overflow pipe. This type of air gap works only if
the overflow pipe does not become blocked. Stoppage of
pumps downstream can also cause reverse flow.
The first type of air gap is used in certain industrial
and research establishments; in baths, wash basins, and
sinks; in drinking fountains; in tanks receiving water from
other sources; on animal drinking troughs; and on overflow
pipes discharging into a water closet or urinal. The second
type of air gap is used for supplies to cold water cisterns
and water closet flushing cisterns and is controlled by a
float-operated valve.
A number of other devices are used in certain circum-
stances including complete separation of systems; for
example, at mixing taps for hot and cold water; non-return
and air break valves; air inlet valves at the fixed ends of
hand-held flexible hoses; lay-flat hoses that collapse when
sub-atmospheric pressure occurs, as commonly used for fire
hoses; air venting devices on hot water systems; and pipe
interrupters. International standardized test procedures
are needed for assessing the adequacy of devices to prevent
back-siphonage.
5. Main Laying and Repair
a. Codes of Practice to Prevent Contamination. Codes
of practice can be formulated to minimize contamination of
mains during installation and repair, and many codes exist
for this purpose, but the extent to which these are enforced
seems to vary considerably, both between countries and
between different water authorities in the same country.
Ideally, protection from contamination should start in
the pipe store by blocking off the ends of all pipes to
prevent access of small animals and birds. These end seals
should remain in position until the pipe is in the trench
and about to be joined. Pipe trenches should be deep enough
to leave a clear space below the pipe joint> so that dirt
from the trench is not incorporated into the joint. Pipes
should not be lowered into the trench and left standing in
the dirt and water at the bottom of the trench: they should
-------
423
be lowered directly.into the jointing position and joined
immediately. Surplus water should be continuously pumped
from the trench, and any open pipe ends should be blocked off
at the end of a day's work.
b. Problems with Jointing Materials. .There are a
number of jointing systems available, but these are mainly
variations of four basic types: (1) lead caulking; (2) push
fit with rubber rings; (3) bolted flanges; and (4) solvent
cement.
The traditional "socket and spigot" joint for iron
mains used a melted lead caulking method with a packing
material, usually jute yarn, to prevent lead flowing into
the pipe. The problems associated with bacterial contami-
nation of jute yarn have been well documented (Mackenzie,
et al., 1948). This material usually contains large numbers
of coliform organisms and supports their growth, which made
it difficult to obtain water free of coliform organisms from
new mains. Chlorination was unable to solve this problem
because the chlorine deviated before it could penetrate the
packed yarn; furthermore, once the initial high dose of
chlorine had dispersed, the remaining yarn could support
further growth.
This problem was greatly reduced by using yarn treated
with a mercurial biocide that sterilized the yarn and pre-
vented subsequent growth, although later leaching would
ultimately nullify this effect, and bacterial degradation
of the yarn could be expected. The use of mercury-based
biocides would not now be approved and in situations where
lead caulked joints are still needed, a biologically inert
yarn packing such as nylon or polypropylene usually is used.
Whatever yarn is used, however, it should be kept in a
closed container and cut to appropriate lengths with a clean
tool. It should not be laid on the ground or in the trench,
but inserted direct from the container into the joint.
The jointing system has now been very largely replaced
by various push fit type joints using specially molded
rubber rings. These newer types of joints are needed
because non-metallic pipes, which cannot be caulked with
lead, are increasingly in use. To assist the jointing
process, a lubricant is essential on the rubber rings.
Excess lubricant tends to get pushed into the socket, where
it is protected from chlorination and flushing water. The
-------
424
original lubricants were usually based on soft soap, which
is capable of supporting abundant microbial growth. Soft
soap is completely soluble in soft water and is, therefore,
readily flushed out when the main is charged with water,
except in the protected areas behind the rubber rings.
However, hard water produces precipitates of calcium salts
with this soap; these can remain on the joints, relatively
immune to flushing out with water, and can also support
microbial growth. A detailed study of this phenomenon was
made by the U.K. Water Research Association and resulted in
the introduction of an antibacterial lubricant, based on
cetrimide, that is now marketed and widely used under the
name Medlube. It is miscible with hard or soft water and
cannot support microbial growth. The bacteriological quality
of water from mains laid using this lubricant has shown
considerable improvement over that obtained with the soft
soap lubricant. Other newer lubricants are based on syn-
thetic detergents that are capable of supporting microbial
growth, but that are completely soluble in hard or soft
water and are, therefore, flushed out more readily.
The rubber rings themselves should be made of a ma-
terial that is unable to support microbial growth, as de-
scribed in Section F.3. Many rubber formulations, both
natural and synthetic, are also subject to long-term bio-
degradation brought about by species of Nocardia. The
phenomenon was first described by Leeflang (1963) and
subsequently investigated further in the U.K. by the Water
Research Centre, the rubber manufacturers, and their trade
associations. Difficulty is being experienced in formu-
lating materials that will not support microbial growth,
that will not biodegrade, and that have the required mechani-
cal and physical properties for joint rings [see Section G.I.a]
c. Cleaning, Disinfection, and Sampling Procedures.
Even where proper codes of practice are followed and there
is good supervision, further precautions, to ensure adequate
cleanliness of new mains, are necessary. Foam swabbing is
effective for this purpose. Cylindrical polyurethane foam
swabs, slightly larger in diameter than the main, are in-
serted at the beginning of new work. When the work is com-
pleted and all valves are open and the water supply is
turned on, the water pressure pushes the swab forward while
allowing some water to pass through. The flexibility of the
swab enables it to negotiate bends and valves and pushes
before it any dirt or debris, eventually discharging from
-------
425
the open end. It is surprising how much dirt is removed by
this process even in situations of supposedly clean main
installation. It is important to have a strictly controlled
system of issue and recovery of swabs to ensure that none
are left in the main when it is put into supply, for there
is nothing more certain to result in irate complaints than
fragments of polyurethane foam discharging from consumers'
taps. It should also be borne in mind that fragments of
foam left in the system can support microbial growth, for
polyurethanes can be biodegraded by various fungi.
It is not usual practice to carry out foam swabbing
after repairs to old mains; inner surfaces of old mains may
not be smooth and may cause excessive breaking up of the
foam or even stop the swab completely. Also, it is often
difficult to arrange a suitable discharge point in an
existing main.
Before putting a new or repaired main into supply, it
must be disinfected; most countries or water authorities
have their own recommended procedures for this. Hypochlo-
rite solutions using an injector or chlorine gas from a
mobile chlorination unit may be used for this purpose. The
section of main to be treated must be isolated by closing
all valves; a dose of 20 mg per 1 Cl, which must be achieved
at all outlet points, is usually recommended. This should
be left for 24 h in new mains, but in small repairs, 2 h is
usually adequate. The main is then flushed with supply-
water until the chlorine residual at the discharge points is
within acceptable limits, which may vary in different
s-ituations and in different countries, but may be as high as
2 mg per 1.
After repair, disinfection, and flushing of small mains
it is usual to put them straight back into supply. Samples
are then taken to ensure that they meet the recommended
bacteriological quality standards. With new mains, it is
advisable to leave them full of water for 24 h after flushing
before taking samples, to allow for possible aftergrowth in
joints. They should not be put into service until satis-
factory bacteriological results have been obtained. If this
procedure is carried out a substantial amount of time in
advance of their being required for supply, further bac-
teriological testing should be carried out before any con-
sumers are connected. ; '...-..
-------
426
d. Acceptable Standards. The WHO Standards for
Drinking Water recommend that coliform organisms should be
absent from 100 ml samples at all times and at all points in
the distribution system. Various national standards recom-
mend the action to be taken when this is not achieved. In
new mains, this is sometimes very difficult to achieve, for
the reasons given in Section F.3.b. Some water authorities,
therefore, permit some relaxation under these circumstances.
If the only detectable deterioration in bacterial quality is
an increase in coliform organisms, without any thermo-toler-
ant coliforms, Escherichia coli, or fecal streptococci in a
new main that is supplied with water free from any of these
organisms, it must be concluded that the coliforms are
multiplying on materials used in the construction of the
main, either on the jointing materials, lubricants, or mains
lining and as such would have no sanitary significance.
Under these circumstances, some temporary relaxation of
standards might be permitted, but the absence of fecal strep-
tococci should be required.
The problems of aftergrowth, discussed in Section F.2,
are liable to be accentuated in new mains and would make the
application of the colony count guidelines in the draft EEC
Directive relating to the Quality of Water for Human Con-
sumption very difficult to achieve, ,..,.-,
The significance of JP. aeruginosa may also be considered
in new mains. It is not a usual practice to look for this
organism, but some studies have revealed its presence in a
proportion of new mains. Its significance has been dis-
cussed in Section P.3.b(iii). It is desirable that inter-
national agreement should be reached regarding the status of
this organism in water supplies, but its absence from 100
ml at all points in the distribution system would be much
more difficult to achieve than the absence of coliform
organisms.
6. Summary
Service Reservoirs.
Service reservoirs, which are
defined as any facilities for storage of treated water
between the water treatment works and the consumer, are
probably the weakest link in the distribution system, but
are an essential part of it. Risks include direct contami-
nation from sewage, which should be avoidable by careful
-------
° 427
siting in relation to sewers. All such reservoirs should be
completely covered to exclude daylight and all openings
screened to prevent access by insects or birds,_as the
latter are a common cause of contamination. Rainwater
penetrating through soil cover on roofs and through cracks
in roofs is another common cause of contamination. This is
worse after heavy rain following a dry spell, when cracks
are present in the soil, and is accentuated by penetration
of roots from trees planted too near the reservoir.
Any particulate matter present in the water is liable
to settle in service reservoirs, and the accumulating sludge
will support the growth of a variety of bacteria. The
engineering design of reservoirs should be planned to avoid
static conditions on the floor as much as possible, and
periodic draining and cleaning is necessary at least once
every ten years.
b. Aftergrowth. Untreated water sources contain
variable amounts of organic matter which is removed or
degraded to varying extents by different water treatment
processes, including disinfection. This 'residual organic
matter can support microbial growth that may appear as
aftergrowth in the distribution system, especially when
water temperatures are high. This can be limited, to some
extent, by maintaining a high chlorine residual in the
distribution system or by treating with ozone prior to a
biological treatment process (e.g., slow sand filtration or
carbon filter treatment).
Particulate matter may accumulate in dead ends and at
any points where there is little or no flow between two or
more interconnected sources of supply. The bacteria that
grow, or tastes and odors derived from them, can be con-
tinually leaked into the water that flows past. Small
animals, such as Crustacea and oligochaetes, may also become
a problem in areas where bacterial aftergrowth has occurred.
Changes in the chemical quality of water during distri-
bution may be chemically induced, such as the leaching of
metals from pipes or fittings or from the ingredients in
some plastics. Deficiencies in treatment are a common cause
of dirty water, for example when high levels of coagulant
are in the system. Changes in source water, or mixing of
water from sources with different chemical characteristics,
can cause disintegration of accumulated deposits. Changes may
also be microbioly induced, such as the production of moldy
-------
428
tastes due to the growth of fungi on organic materials.
There are, as well, a number of common 'interactions between
microbial changes and chemical reactions, some of which may
result in the internal corrosion of iron and steel mains.
Local accumulations of biomass can produce anaerobic con-
ditions, resulting in growth of nitrate or sulfate-reducing
bacteria and their chemical by-products.
c- Growth on Materials. Many bacteria and fungi can
grow on a great variety of simple and complex organic com-
pounds, either of natural or man-made origin. It is, there-
fore, ^ important that materials used in distribution and
plumbing systems should not contain enough of such organic
compounds to cause either a direct microbial contamination
or a deterioration in quality, such as moldy taste. In
consequence, all new products intended for use in contact
with treated water supplies should be tested, by a reliable
method, for their ability to support microbial growth.
A special example of growth on materials is that asso-
ciated with water exposed to volatile organic solvents, the
commonest being ethanol. This gives rise to the growth of
massive bacterial and fungal slimes wherever a free water
surface is exposed to an atmosphere containing such vapors.
These situations can occur in domestic as well as business
and industrial premises.
d« Back-Siphonage and Cross-Connections. Reduction in
pressure can occur in public water supply mains as well as
in individual plumbing systems,, and this is one of the condi-
tions necessary for back-siphonage to occur, or for water to
move against the normal direction of flow or pressure. Most
water authorities have experienced problems from this cause/
and many^countries have bylaws or regulations designed to
prevent its occurrence. Cross-connections, which may be
defined as connections between a public water supply main
and any other pipe, usually occur outside the control of
water^authorities. Recorded instances are not infrequent,
sometimes with serious consequences.
Laying and Repair. Most countries have codes
^
of practice to minimize contamination during main instal-
lation and repair. Jointing systems have been a common
cause of problems, such as bacterial growth on jute yarn or
-------
429
on^the soft soap used for lubricating rubber ring joints.
*£r? °r.Polpr°Pylene Yarn should be used instead of jute
£ ^b£lcants whxch will not support microbial growth
should_be used as well. Cleaning of new mains after con-
struction may be facilitated by using foam swabs. Disin-
fection procedures are critical and codes of practice should
be -followed'for satisfactory results. Sampling should be
oT^effor^ h^bef" h",been flushed * has stood full
water tor 24 h, before being put into supply.
7. Recommendations
1.
Service reservoirs, which are defined as any
treated water storage imposed between the water
treatment works and the consumer, should be
situated where risks from sewage contamination
are minimal. Service reservoirs should be
completely covered to exclude daylight and all
openings screened to prevent access by insects
or birds. They should be maintained in a good
state of repair and drained periodically for
cleaning.
;
Factors such as the presence of organic matter
and the accumulation of debris in dead ends and
at points where there is little or no water
flow may lead to aftergrowth and possibly the
presence of small animals in the distribution
system. Dead ends should, therefore, be avoided
in the design of a distribution system, and all
disused apparatus should be disconnected at the
junction with the main pipe. In areas where
balanced pressures between different supply
sources cause little or no flow, the mains
should be flushed at intervals.
Distribution and plumbing systems should be
constructed only of materials which will not
support microbial growth. New products should
be assessed by a reliable test procedure before
they are used in contact with potable water.
It is essential that regulations to prevent
back-siphonage and cross-connections be en-
forced.
-------
430
5 Materials used in jointing systems and lubri-
cants in mains should be restricted to those
which have been tested by a reliable method and
found not to support microbial growth. Ade-
quate disinfection procedures after main
installation and repair are essential and codes
of practice should be strictly followed..
Sampling of the water should take place after
thorough flushing of .the main.
-------
431
G. TECHNOLOGICAL ASPECTS
In highly industrialized countries, drinking water of
good microbiological, chemical, and aesthetic qualities is ,
routinely achieved with available treatment processes.
Nevertheless, improved practices and equipment can exert
their own influences by introducing man-made materials upon
which microorganisms can become established. Reduced effi-
ciency of operation, increased costs, and overstressed
equipment are oftentimes the consequence. Such factors as
degree of pipe bends and placement of joints, type and .age
of pipes, valves, and fittings, changes in pressure all
combine with the constituents of a particular source water
and occasionally give rise to complications within the
system. Therefore, the growth of organisms within a dis-
tribution system, however harmless to man at the outset> may
ultimately threaten the quality of the finished water through
these indirect means. .
Filamentous iron bacteria are largely responsible for
corrosion of metal pipes and the concomitant deposition of
insoluble ferric hydroxide along pipe walls and well screens.
Such bacteria may also initiate pitting and tuberculation of
pipes. The formation of pits and tubercles favors the
survival of other organisms which can then attach to pipe'
walls and utilize dissolved substances present in the water.
Furthermore, variations in hourly demand for water can bring
about pressure changes sufficient to cause stripping of
metabolic deposits and the delivery of dirty water.
Water mains are made of cast iron (both cement-lined
and unlined), concrete, galvanized iron, polyvinyl chlo-
ride, and cement asbestos. Whether enzymatic digestion of
these and other materials occurs or not.is dependent upon
individual ingredients used in the manufacture of various
component parts, plus whatever organisms inhabit a-par-
ticular water source. . .
Many additional devices are employed as aids to house-
hold practices; for example, to soften water for dishwashers,
-------
432
washing machines, coffee percolators, steam irons, etc.
Such auxiliary equipment, installed beyond a household tap
or the water meter, may include ion exchangers which apply
phosphate, polystyrene filters (to remove small particles or
bacteria), and active carbon filters (to remove taste and
odor). These devices may give rise to bacteriological
problems, since the resins or the filters adsorb organic
matter which is a good nutrient source for bacteria.
Besides imparting to the water objectionable odor,
taste, frothing, color, and turbidity [See Section F.3.b]
certain slime-producing bacteria can significantly disturb
the hydrodynamics of turbulent flow through a pipe. A very
slight irregularity in the thickness of a pipe wall, caused
by growth of slimes, can interrupt the laminar stream flow
enough to increase the friction factor, encumbering the
energetics of the system with a greater resistance to flow.
Industrialized nations share a responsibility to under-
stand and protect against indirect and long-term effects
from new machinery and techniques, as they apply to old,
familiar microbes. The added expense, loss of efficiency,
and possible health hazards triggered by a plugged well
screen, a constricted pipe, or an obstructed filter are some
of the considerations with which this section is concerned.
1. Biodeterioration of Materials
a. Biodeterioration of Rubber Sealing Rings. Although
it is over half a century ago that microorganisms were first
claimed to be capable of destroying natural rubber (Sohngen
and Fol, 1914), the first evidence that this occurred in the
water industry was found in the Netherlands around 1950.
Rook (1955) isolated pink colonies of Streptomyces sp., from
rings from a water main, on pure latex agar. In pure cul-
ture, these organisms could bring about deterioration of
thin strips of vulcanized rubber.
These investigations were continued by Leeflang and
others on behalf of the Institute for Testing Waterworks
Materials, the Netherlands (Keurings Instituut voor Water-
leiding-Artikelen, 1961). Of 651 rings examined, 59 percent
were deteriorated on the surface in contact with the potable
water and only 15 percent showed deterioration on the soil
side. The majority of joint rings showed a loss of rubber
of less than 5 percent of the cross-sectional area. The age
of the rings varied from two to 23 years, but age was not
-------
433
the "sole determinant of the degree of 'deterioration. In
all, five or .six instances of hydraulic failure were ob-
served. It was concluded that the problem occurred when
natural rubber rings were exposed to a flow of. unchlorinated
potable water and, in the Netherlands, this was particularly
prevalent with dune waters. These conclusions have been
reported together with the results of exposure trials of
rubbers (Leeflang, 1963).
In 1963, reports of deterioration of rings manufactured
to BS 2494 (British Standards Institute, 1955) were ema-
nating from Australia through the Standards Association of
Australia. The problem was .associated, in particular,, with
rubbers from sewer lines which, aften ten to twelve
-------
.'434
examined showed serious deterioration and 40 percent slight
deterioration on the sewage, or water, side. Examination of
rubber rings after 25 years' service showed that, although
10 percent were free from excessive deterioration, others
were degraded by as much as 50 percent. The condition was
more evident in sewer pipes than in water lines.
Cundell and Mulcock (1973) continued this work and
found, in a survey of 583 natural rubber jointing rings
collected from domestic sewers in Christchurch, New Zealand,
that 36 percent showed deterioration about the mean water
line, 33 percent were deteriorated on the outside of the
ring in contact with the soil, and 6 percent were dete-
riorated on both sides. Furthermore,, three of the rings
showed deterioration in excess of 50 percent of the cross-
sectional area, with a further three having 10 to 50 percent
deterioration. It was found to be difficult to predict the
extent or future progression of attack. Signs of attack
usually were not apparent until four years after instal-
lation, and after six years' exposure, more rings were
attacked than unattacked. In that the majority of rings
have only been used in Christchurch since about I960, it is
not possible to predict their probable life. All rings
installed at this time in concrete sewers were found to be
in advanced state of decay, suggesting that the phenomenon
is progressive. However, it was claimed that no instances
of hydraulic failure of pipelines had been reported in
Christchurch to date. The authors remark that deterioration
of natural rubber rings is not readily appreciated by those
who lack firs,t-hand experience with the problem. For this
reason, it was claimed that the occurrence of the phenomenon
may be more prevalent throughout the world than is generally
believed by rubber and pipe manufacturers or pipe users.
In the U.K., the Malaysian Rubber Producers Research
Association defended the use of natural rubber for pipe
joint rings (Dickenson, 1965). It was suggested that the
problem in the Netherlands might be related to the atypical
potable water, derived from dunes, rich in dissolved min-
erals (notably phosphate), and distributed without chlori-
nation. Furthermore, the conditions within the joint space
of asbestos cement mains would lead to a high pH of the
water in contact with the rings which, it was claimed, would
favor the growth of Streptomyces sp., the organisms impli-
cated in deterioration.
In 1969, Dickenson claimed that the oxidative attack on
the rubber molecule by microorganisms was probably terminal
-------
435
and, in view of the low incidence of terminal sites in fully
vulcanized rubber, it was argued that low molecular weight
fragments were necessarily produced by primary, non-bio-
logical degradation. In view of the extreme slowness of
deterioration, it was suggested that the initial phase was
one of nonspecific chemical oxidation; in support of this,
it was claimed that chemical analysis of deteriorated rubbers
had revealed a notable absence of persistent antioxidant,
especially in the water environment. Other components of
natural rubber might be subject to microbiological attack
that could result in deterioration. These could include
plasticizers and antiozonants, especially paraffin wax.
Further work in the U.K., undertaken by the Water
Research Center (WRC), has concentrated on three aspects of
the problem: (1) the bacteria involved in biodeterioration;
(2) the incidence of deterioration in rubber rings recovered
after service in potable water pipelines; and (3) the effect
of rubber formulation on the,rate of deterioration. WRC
studies have confirmed many previous findings. Natural
rubber is the most prone to attack, which is normally re-
stricted to the surface in contact with the water. In
sewage pipes, the attack may be concentrated in the area of
the mean sewage flow. High levels of rubber weight loss (up
to 33 percent) have been observed without apparent leakage,
but the possibility of mechanical failure cannot be ignored.
Deterioration can occur in pipelines carrying ground-
water, surface water, or sewage. Unchlorinated waters, or
waters in which the chlorine residual has disappeared, would
appear to support higher rates of attack than those in which
a residual may still be detected. In certain instances,
several rings showing deterioration have been recovered from
the same section of pipeline, suggesting that the phenom-
enon may be widespread.
Rings less than four years old do not show attack, but
there is an increasing tendency for rings older than this to
show deterioration. This is most marked in rubber rings
recovered from pipelines laid between 1950 - 1960. The
timing would indicate that this might be associated with
changes in rubber formulations which were necessary to
protect the health of workers in the rubber industry.
The incidence of attack does not seem to be specifi-
cally associated with ring diameter. However, there is a
distinct tendency for deterioration to be most marked in;
asbestos-cement pipes and least in iron mains. Initially,
-------
436
this was considered to reflect the preference of the actino-
mycetes, which may attack rubber, for the alkaline condi-
tions found in ..the joint spaces of asbestos-cement pipes.
However', laboratory studies have shown that the formulation
of some rubber rings used for the assembly of asbestos-
cement pipes could make the rings more susceptible to deterio-
ration.
Examination of large numbers of rubber rings has indi-
cated that roughening and surface erosion are often symp-
tomatic of biological attack. Two selective media, Kuster
and Williams agar (Kuster and Williams, 1964) and modified
DST (oOrchard and Goodfellow, 1974), were used for the
examination of both deteriorated arid .undeteriorated sections
of the same ring/ it was shown that actinomycetes are a
significant factor in deterioration. Irrespective of the
origin of the rubber ring, three morphological types were
predominant; these have been identified as Streptomyces
litmanii, Streptomyces fulvoviridis, and Nocardia asteroides
(Hutchinson, et al.,1975).These were clearly the types
described as white and orange-pink streptomycetes by other
authors (Dickenson, 1969,; Leeflang, 1963; Rook, 1955).
Numbers as large as 10 to 10 per gof deteriorated rubber
have been recorded. The ability of these isolates to degrade
rubber has been confirmed by inoculation of a rubber latex
film supported on an aqueous agar base. The organisms grow
and alter the physical properties of the rubber, confirming
their ability to degrade the polymer. The ability of these
actinomycetes to use certain rubber ingredients as sole
carbon sources has also been demonstrated.
Other workers (Berridge, 1951; Thaysen, et. aJL. , 1945)
have commented on the involvement of sulfur-oxidizing bac-
teria, which are capable of producing significant levels of
sulfuric acid when growing on excess sulfur present in
sulfur-vulcanized rubbers. This strong acid could cause a
chemical deterioration of the rubber. Although these
organisms were found to be associated with deteriorated
rubber, it was considered that their presence was more
directly related to the sewage environment, in which reduced
sulfur compounds, namely hydrogen sulfide, were freely
formed (Hutchinson, ejt a^L. , 1975).
The importance of rubber formulation in controlling the
rate at which a ring is attacked was clearly shown in a
series of accelerated aging tests performed at the WRC
using thin strips of rubber. These showed that unprotected
natural rubber is rapidly degraded, but can be protected by
-------
437
the use of antioxidants, especially those which are insoluble
in water as reported earlier by Derham (1975) and Dickenson
(1969). Chlorinated waxes and aromatic processing oils also
have a beneficial effect in improving the resistance of
natural rubber, but to a lesser extent than that afforded by
antioxidants. The presence of relatively high concentra-
tions of organic accelerators, such as thiocarbamates and
thiurams, has a beneficial effect on the resistance of
natural rubber to biodeterioration, as was found by
Cundell and Mulcock (1973). However, some ingredients,
especially plasticizers, may enhance the rate of deterio-
ration and in separate tests, some, especially sodium
octonoate, were found actively to support the growth of
rubber-degrading bacteria.
Although changes in the formulation of natural rubber may
delay the onset of microbial attack, they will not prevent
it from eventually happening. Whether careful selection of
ingredients (especially antioxidants) that are currently
acceptable to the rubber industry will protect rubber jointing
rings for the 50 years' service life normally expected by
water engineers is not clear from the short-term trials
that have been done. The demands of the water industry for
rubber formulations which are not toxic and do not permit
the growth of microorganisms is a further complication in
the selection of satisfactory rubber seals.
b. Biodeterioration of 'Sealants and Mastics. A sealant
is intended to maintain a seal between the sides of a joint
which may be subject to some degree of movement. In the
water industry, sealants are primarily to control the move-
ment of water, preventing both loss and contamination of
water held in various concrete and steel tanks. In addition
to good mechanical performance, sealants used in water
supply must show minimal toxicity for mammals and must not
support the growth of microorganisms. The National Water
Council approvals procedure [See Section F.3.dJ will protect
the consumer from deterioration in water quality caused by
these two factors. Additionally, the service life of the
sealant must not be'limited by microbial degradation.
WRC has assessed, for signs of microbial attack,
sealants which have lost their mechanical properties. The
following indices and microorganisms have been studied:
colony counts, fluorescent pseudomonads, actinomycetes
(including Nocardia sp.), sulfur-oxidizing bacteria, sul-
fate-reducing bacteria, and fungi. In strong contrast to
the findings with rubber sealing rings, no specific micro-
bial cause for deterioration could be found. Invariably,
-------
.438
the specific organisms likely to cause microbial deterio-
ration (pseudomonads, actinoniycetes, and fungi) have been
found in no higher numbers than are present in neighboring
unattacked material, the surrounding water environment, or
in the case of filters, no more than is present in the
adjoining sand. However, the Metropolitan Water Board
(1970c) has shown that an ascomycete, possibly Herpotri-
chiella sp., was associated with disintegrated polysulfide
sealants.
It is possible that sealants, because their formu-
lations differ greatly (especially,with respect to their
"higher levels of plasticizer) from those of rubber sealing
rings, are more prone to a generalized attack by a variety
of microorganisms that are capable of exploiting a less
restrictive environment. At present, it is not possible to
say whether the service life of sealants is primarily con-
trolled by physical-chemical aging accentuated by chemical
leaching, or whether microbial activity can be considered a
limiting factor. However, it must be accepted that the
working life of a sealant will be considerably less than
that of a rubber ring and is unlikely to exceed ten to
fifteen years.
2. Microbial Growth On Resins and Filter Media
In a technologically developed country, the drinking
water is relatively free of bacteriological and aesthetic
problems. However, only a small proportion of this domestic
water is actually ingested by consumers as a beverage or in
food; the remainder, probably over a hundred times the
amount that is ingested, is used for hygienic, cleansing, or
technical purposes. Whereas this water may be of good
quality from the standpoint of health, it may have proper-
ties that make it unsuited to its alternative uses. There-
fore, it may be necessary to eliminate certain ions from the,
water or to add special preparations so as to minimize
technical difficulties.
In a household, filtration or ion exchange devices may
be used to eliminate taste and odor-producing substances or
bacteria, or to soften hard water in order to improve the
performance of hot water equipment. Water can also be
softened by adding substances that contain phosphorus, such
as polyphosphates or ortho-phosphates, in order to prevent
corrosion or encrustation within the pipes. This equipment
may be installed behind the tap in an apartment or in the
-------
439
basement of homes. Alternately, filters or small ion ex-
change boxes may not be connected to the distribution system,
but may instead, be used separately as adjuncts to a specific
household task. These small devices are often advertised
for use while traveling and camping, especially to areas
where sanitary conditions are suspected of being poor and
disease is endemic. On the whole, these apparati are un-
necessary in households of the developed countries and can
be ineffective and even dangerous when used in areas where
there is epidemic disease. Although useful in their spe-
cialized technical roles, they are apt to create bacterio-
logical problems in the process.
a. Filters. There is usually no need to use filters
for removal of bacteria in households of developed countries.
This is already accomplished by the municipal supplier.
However, filters (mainly of activated carbon or some plastic
compound) are advertized as a means of eliminating objec-
tionable tastes, odors, colors, and turbidity in^the water.
It is well known that bacteria, along with organic matter,
can collect within filter pores providing a suitable en-
vironment for microbial growth. Periodic cleaning of the
filter becomes necessary to avoid large swells in bacterial
populations emanating from the filter. Since publically
supplied drinking water normally is free of pathogens, those
bacteria which multiply in the filter are more a nuisance
than a health hazard. Nevertheless, if there are legal
codes limiting the acceptable colony count levels, this can
create administrative problems.
b. Ion Exchangers. Ion exchange resins can produce
bacteriological problems not only from attachment of bac-
teria and organic matter to the resin, but also from com-
ponents of the resin itself that can serve as a nutrient
source. Frequently, newly purchased ion exchange resins may
already have defects in the individual resin granules. As a
rule, these defects will increase during the life of the
exchanger. Bacteria can very easily collect within these
surface irregularities and be protected from disinfection,
backwashing, or other cleaning attempts. Similar problems
may be seen with filters and can lead to high microbial
loads in the product [See Section E.4].
c. Portable Filters. As has already been mentioned,
portable filters are intended to produce good quality water
in areas where it is otherwise unobtainable. However,
-------
440
portable filters are also advertised for use at camp sites
or during times of emergency. Therefore, they should be
designed so as to be capable of removing not only pathogenic
bacteria, but also viruses, protozoa, and helminthic eggs
that may be present in contaminated water. The small size
of viruses dictates the need for extremely small pore
diameters, which greatly reduces potential filtration speeds.
Often, after a short period of use, the filter becomes
clogged and will stop filtering altogether. Thus, effective
portable filters are likely to be inconvenient, and con-
venient portable filters are likely to prove unsafe.
d. Waj-er Softeners that Add Phosphorus. Unlike ion
exchangers and filtration units, phosphate-dosing equipment
adds small amounts of phosphate to the drinking water.
Legislation in some countries specifies an upper limit for
phosphate that can be present in the drinking water. Dosing
with phosphate is useful only if the water is not too hard
and the temperature of the water is about 60°C. Phosphate
should not be added to a cold water supply because at low
temperatures it encourages the growth of bacteria and (if
light is available) algae, especially if there is any
degree of stagnation. There is also the danger that phos-
phate preparations will not be handled properly in transit
from producer to consumer,1 this too can result in contami-
nation of water with bacteria. Provision of nutrients to
bacteria already present in the water, plus introduction of
new bacterial species, possibly pathogens, are potential
problems connected with the use of phosphate preparations.
It is, therefore, recommended that such preparations be
divided into small portions and durably packaged before
sale.
e. Conclusion. Small-scale water treatment apparatus
for use in the field is frequently unnecessary or unreliable.
Some devices fail to remove the microbes that they are
supposed to take out of water, whereas others encourage
microbial growth so that colony counts are increased in the
treated water. The latter problem is also encountered when
filter media, such as activated carbon [See Section E.5] or
ion exchange resins, are used on a large scale; however, it
is less of a nuisance when the treatment is being done by
professionals and will be followed by disinfection.
-------
4.41
3. Hydraulic Effects of Microbial Growth
Water from all public supplies must be delivered under
pressure, whether its destination is a tap an the tallest
skyscraper or the last tap at the end of a large distri^-
bution system. Pressure is necessary not only to supply
people quickly with a desired volume of water,'but also to
maintain hygienic conditions. .Flowing water seldom permits
bacterial growth, and a properly functioning distribution
system will maintain an adequate flow to ensure against
microbiological problems.
a. Types and Diameters of Pipes. Often, the type of
pipe used, or its diameter, will have an effect on bacterial
growth, especially at locations in the home where a con-
tinuous flow of water is not guaranteed because of infre-
quent usage. Comparative studies with plastic,- rubber,
copper, and steel pipes showed that colony counts increased
from one to 14,000 per ml in the rubber .pipe; whereas steel
and plastic pipes had colony counts of about 30, and the
copper pipe showed a count of about zero after four days of
use. Plastic material must be tested for use with drinking
water because plastic paints and covers used in closed
reservoirs very often lead to problems with bacterial -
growth.
^In some buildings, especially in schools, the diameters
of pipes in the distribution system are calculated for
emergency cases (e.g., fire) and not for routine drinking
water supply. This means that the amount of water being
regularly used is small in comparison to the pipe diameters.
The result can sometimes be stagnation and taste and odor
problems, especially if drinking water pipes are installed
near heating pipes where elevated temperatures, will promote
bacterial growth. Streptomyces species are especially
adapted to taking advantage of such situations, imparting an
earthy taste and odor to the water.
b. Microbial Oxidation of Manganese in Pipes* Certain
bacteria, including Pseudomonas sp., Hyphomicrobia, and.
Sphaerotilus discophorus can form oxides of manganese in
water containing > 0.01 to 0.05 mg per 1 or more-of manga-
nese within_a pH range of 6 to 7.5. These manganese.oxides
adhere to pipe walls forming ripples across the direction
of flow, which invariably leads to. losses in pressure. It
has been shown (Schweisfurth and Mertes, 1962) that, at hign
-------
442
flow rates, the hydrodynamic effects in pipes become sig-
nificant, leading to a substantial drop in pressure; where-
as, at low flow rates, pressure changes are inconsequential.
Pipes in which such a build-up has occurred, can be restored
to good operating condition with mechanical cleaning, albeit
the process is cumbersome and time-consuming. Alternatively,
chlorination can be used to inhibit microbial activity; with
chlorine, oxidation also takes place, but the end product ,
Mn (IV) does not adhere to pipes and can be removed by fil-
tration.
c. Microbial Growth from Improper Installation of
Household Equipment"! The installation of dishwashers,
washing machines, ion exchangers, or filtration units must
take into account pressure differentials between the house-
hold water supply and equipment capacities. If some means
of compensation is not also installed, back-siphonage and
contamination can result. Rubber tubes are often added to
the tap to extend use to the bottom of bathtubs or sinks.
Turning water outlets in different locations of the house on
or off can sometimes be sufficient to create hydraulic
pressure changes and to flush contaminated bathing or wash-
ing water into the distribution system. Such events have
been responsible for the occurrence of enteric disease in
inhabitants of one household or several households along an
entire street.
d. Microbial Growth from Cross-Connections. Indus-
trial buildings and ships,in particular [See Section G.5],
are known to have two or more distribution systems: one for
drinking water supply and another to supply low quality.
water for other uses, such as cooling or fire protection.
If these industries are situated near a large river or
harbor, they very often use raw river or harbor water for
these latter purposes. Normally, the two systems must be
separated from each other by special instrumentation to
ensure that no mixing occurs. In some cases, pressure in
the water pipes used for fire protection is higher than that
which is in the drinking water pipes and small amounts of
contaminated water may continuously drip into the drinking
water system. If this "inoculum" is derived from polluted
river water, pathogens may be not only introduced, but may
find conditions suitable for multiplying. If at any time>
the slide valve is opened, the danger of contamination is
greatly increased. Over the last 30 years, outbreaks of
typhoid fever, paratyphoid-B fever, and salmonellosis have
-------
443
resulted from such cross-connections and pressure differences
in distribution systems of this kind. Therefore, two systems
handling different quality water and operating side-by-side
must be carefully controlled and supervised.
4. Microbially Mediated Chemical Cycles
a. Microbial Manganese Oxidation Affecting Wells and
Water Distribution Systems.Manganese-oxidizing micro-
organisms (bacteria, fungi, and very rarely protozoa) re-
quire a treatment separate from that for "iron bacteria"
because only a few of the rare species of chlamydobacteria
oxidize or precipitate both iron and manganese under natural
conditions or in the laboratory. The references to reports
of both iron and manganese oxidation by Siderocapsaceae are
incorrectly quoted in most cases (Schweisfurth, 1973).
Enzyme-mediated iron or manganese oxidation cannot be proved
solely on the basis of microchemical detection of Fe (III)
or Mn (IV) at the sheaths of the surfaces of these bacteria.
By contrast to Fe (II), Mn (II) is oxidized by O , at
a rate that is of practical significance .in the field of
drinking water, only above a pH value of approximately 8 to
8.5. Correspondingly, reduction of Mn (IV) already occurs
at Eh values near or below + 200 mV; this means that the
microbiological oxidation of Mn (II) can only take place
above this Eh value (Schweisfurth, 1972).
Depending on the geological and hydrological conditions
in the aquifers or reservoirs, 'Mn (II) is formed from Mn
(IV) by reduction or dissolved from manganese carbonate by
reaction with CO2 or HCO^ . In the case of waters with little
or no dissolved oxygen, the Mn (II) penetrates into the wells
and into the raw water, where, in the presence of O^, micro-
bial oxidation starts, as well as the formation of MnO "
(= MnO , x = 0.7) and the formation of deposits in the
wells anct water conduits. The problems caused by these
deposits include clogging of slots in well pipes; increased
turbulence and thus, a loss of flow velocity in conduits; damage
to equipment for measuring water flows; bla^ck-colored
water; stains in laundry; disturbances in food handling
establishments; accumulation of heavy metals such as arsenic,
lead, zinc, and copper (Basco and Szalay, 1978); and some-
times an increase in the colony count of the water. Pre-
vention is based on elimination of Mn (II) from raw water
if limit values of 0.01 to 0.05 mg per 1 are exceeded.
Based on the conditions under which Mn (II) and Mn (IV) are
-------
444
stable (Hem, 1963-1965), Mn (II) can only be oxidized,
within^the Eh and pH ranges common for drinking water, by
microbiological means. Various bacteria (Sphaerotilus
discophorus, Pseudomonas, Arthrobacter, and Hyphomicrobium
species), as well as fungi, induce manganese oxidation
(Schweisfurth, 1978). Apparently only in the case of
Hyphomicrobium manganoxidans is there an interdependence
between the multiplication of microorganisms and manganese
oxidation (Schweisfurth, et al., 1978).
The removal of manganese from raw water 'at a treat-
ment plant can be accomplished either by chemical or
microbiological means if conditions are adjusted. The
chemical manganese oxidation is carried out by aeration, pH
values, that is induced by the MnOi+x already formed. During
the microbiological removal of manganese, the medium of a
filter serves as a support for the growth of microorganisms
(mostly gram-negative, rod-shaped bacteria and in rare cases,
sphaerotilus discophorus and manganese-oxidizing fungi). It
is essential that the Eh value of the water be sufficiently
high (more than + 200 mV); otherwise, manganese oxidation
does not start, even after an inoculation with backwash
sludge from another plant. If there is no removal of manganese
in a microbial water treatment plant, no manganese-oxidizing
microorganisms will be detectable by culture.
Formation of a high proportion of manganese-oxidizing
growth on the inner surfaces of water pipes has not been
shown to be directly related to corrosion.
b- Iron and .Sulfur Bacteria Corroding Well Casings
and Other Structures.Bacteria may take part in the cor-
rosion of iron well pipes and drinking water conduits. In
each case, corrosion may start from either the inside or the
outside. Except for the coating of pipes against exterior
corrosion, protective measures such as bituminizing or
plastic coating afford only a partial protection.
In this discussion, the term "iron bacteria" will
include: (1) bacteria suspected of being capable of drawing
an energetic benefit from the oxidation of Fe to Fe
(Gallionella, possibly chlamydobacteria such as Leptothrix);
and (2) bacteria utilizing the organic ligand from complex-
bound Fe , following which Fe is subject to oxidation by
O2 at pH values above 5, such as are found in drinking
-------
445
water. The latter bacteria, which ar|s+mpstly rod-shaped,
may also precipitate complex-bound Fe .
The term "sulfur bacteria" will include: (1) sulfur-
oxidizing bacteria that, being autotrophic, will generally
oxidize reduced sulfur compounds to SO . , mainly under
aerobic conditions; and (2) sulfate-reducing bacteria that,
at a locally or generally low Eh value and in the absence of
O2, form hydrogen sulfide using both organic hydrogen donors
and molecular hydrogen.
Microbially mediated corrosion (i.e., a dissolution of
elemental iron and other metals) will occur as a consequence
of: (1) absorption of nutrients, as well as oxygen, by the
colonies of microorganisms that have accumulated at the
metal surface; (2) liberation of corrosive metabolites, such
as organic acids and other complex-forming compounds; (3)
production of sulfuric acid from S or elemental sulfur;
and (4) inclusion of sulfate-reducing bacteria in the
cathodic process under-anaerobic conditions. Both natural
and artificial anti-corrosive coatings are also subject to
microbial attacks. Although the types of corrosion men-
tioned in (1) and (2) are assumed to occur only under condi-
tions in which nutrients are plentiful (as is true in cool-
ing systems), they may also play a role in interior cor-
rosion within drinking water distribution systems.
The chemical and microbiologic processes involved have
been summarized by Chantereau (1977). Subtle heterogeneity
in the system can lead to the establishment of foci of
anodic and cathodic depolarization, enabling electrolysis
and metabolic activity of iron bacteria and, given a sulfur
source, of sulfur bacteria as well [See Figure G.4.b-l],
As iron from the pipe is being eroded in the form of Fe
ions, various derivative substances are being deposited on
the inner surface of the pipe, sometimes to an extent that
occludes the bore of the pipe substantially [See Figure
G.4.b-2]. Maturation of the deposits leads to the presence
of an anaerobic zone nearest the pipe surface and an aero-
bic zone in contact with the water, as well as an inter-
mediate zone of low oxygen activity: this allows iron
bacteria and both sulfur-reducing and sulfur-oxidizing
bacteria to participate in the corrosion and deposition
process [See Figure G.4.b-3]. .
In addition to their participation in corrosion pro-
cesses, the iron bacteria have both favorable and unfavor-
able effects on the procurement and treatment of raw water
-------
446
FIGURE G.4.b-l
CYCLE OF BIOLOGICAL CORROSION
2Fe(OH)3
ANAEROBIC SIS
A2
T Chemoorganotrophism
chemolitho-
H S trophism
SULPHATE-REDUCING K
BACTERIA
VRON BACTERIA
3Fe(OH)2
CH3COOH
\\\\\\\\\\
\Y\YAV\v\V\VW
QH
Y\V\Y\\
SULPHUR BACTERIA \
Cathodic
depolarisation
Anodic
depolcrisation
From Chantereau, 1977.
-------
FIGURE G.4..b-2
FORMATION OF DIFFERENT FORMS
OF IRON DEPOSITS IN WATER PIPES
447
Polarisation
hydrated Fe**bx/de
Accelerated formation of hydrated Fe oxide
(Depolarisation)
Corrosion Beginning formation
of anaerobic zone
Iron bacteria
Sul fat e- reducing
bacteria
From Chantereau, 1977.
-------
448
FIGURE G.4.b-3
SCHEMATIC REPRESENTATION OF A SECTION
OF A SO-CALLED RUST KNOB
0
Sulphate-re-
ducing bacteria
Hydrogen
From Chantereau, 1977.
-------
449
for the production of drinking water (Glathe and Ottow,
1972; Kullraann and Schweisfurth, 1978; Schulze and Schweis-
furth, 1975). During procurement, they will accelerate
Fe precipitation when oxygen has entered the aquifer and
the groundwater, and will thus, reduce the water-permeable
cross-section of all structures (in the soil and in slotted
well-pipes), thereby'eventually resulting in failure of the
well [See also Sections A.l.b to c and A.2.a(ii)]. On the
other hand, the iron bacteria certainly take part in the
breakdown of organic substances during the process of iron
removal from waters, which opposes the precipitation of
Fe . These bacteria are also found where iron is pre-
cipitated within distribution networks [See Section F.2.d(vi)]
The problems caused by iron and sulfur bacteria have
been described as briefly as possible in this discussion.
In addition to the references cited above, valuable and much
more extensive information will be found in Iverson (1975),
Lovelock and Gilbert (1975), and Miller (1971).
5.' Drinking Water Supply for Ships
Many countries have no regulations, for ensuring good
quality drinking water aboard ships (Muller, 1976; Goethe,
et aJL. , 1969) although a disproportionately high frequency
of waterborne infections are known to occur on these vessels
compared with incidences on land. A serious problem with
shipboard water lies in the fact that many ports only have
contaminated water (sometimes 'containing helminth eggs,
protozoa, bacteria, and/or viruses) available to fill ships'
tanks and because this water does not undergo any treatment
or purification during storage.
Being confined on a ship means that both passengers_and
crew cannot avoid infection if water supplies are contami-
nated with pathogens^ hence, outbreaks aboard ships tend to
be epidemic. These not only include common waterborne
diseases (e.g., typhoid fever, paratyphoid B fever, other
salmonellosis, cholera, and dysentery), but also infections
caused by large concentrations of opportunistic pathogens
(e.g., P. aeruginosa, as well as Clostridium, Bacillus, and
Staphyl'ococcus species). This latter group,, if incorporated
into food through the addition of contaminated water, may
become enriched in the process (Muller, 1974).,
-------
450
a. Occurrence of Contamination Aboard Ships. Investi-
gators have found that the bacteriological quality of ships'
drinking water is, on the average, inferior to that of land
water. For example, from 1950 to 1960, 25 percent of the
ships docking at the Hamburg harbor had water which tested
positive for E. coli (Muller, 1961). In 1970, the Japanese
found high total bacterial and coliform counts in drinking
water from nearly 50 percent of their vessels (Hayashi, eit
al., 1975). They later reduced the proportion to 27 percent
in 1974 by adding sterilization and filtration equipment.
b. Why Ships' Water Becomes Contaminated. Observa-
tions made over a 20-year period have revealed several areas
of neglect which are largely responsible for contamination
of water supplies on ships. They are listed in order of
their frequency of occurrence:
- Drinking water kept in open tanks during con-
veyance aboard lighters.
- Using the same bucket for drinking water as well
as for river water.
- Tracking dirt frpm shoes into tanks during clean-
ing operations. *
- Using river or sea water to clean tanks.
- River or sea water seeping into improperly closed
tanks.
Using the same hoses and pumps for both drinking
water and river water.
c. Water Distribution Systems on Ships. Ships possess
either a single, double, or triple distribution system.' The
single system, theoretically, provides potable water throughout
the entire distribution network. Large, recently built
ships are likely to have this arrangement, now that equipment
for distilling salt water has provided an alternative to
refilling tanks from water supplies of foreign harbors.
The double system conveys finished water for drinking
and washing through one network, while river or sea water-
travels through a separate pipeline to water closets, ship
cleaning, and fire protection facilities.
-------
451
The triple system channels river water into- showers and
sinks for hand and dish washing; sea water into a network
for fire protection, ship cleaning, and toilets; and drinking
water into only a few taps, primarily for that purpose
alone.
d. Types of Disinfection Practiced on Ships. Disin-
fection can be accomplished by physical or chemical means.
These processes include boiling; filtration; or applying
chlorine, silver, ozone,, or ultraviolet irradiation.
The cost and bulky apparatus needed for boiling make
this process impractical for use on ships. As for filters,
they may become defective or clogged and they require care-
ful maintenance, backwashing, and bacteriological monitoring
of the filtrate (Mu'ller, 1975) all of which require exper-
tise not commonly available on board a ship.
The effectiveness of chlorine disinfection depends on
maintaining an active residual in spite of possible high
loads of organic matter; this can give rise to problems of
taste and odor. Furthermore, automatic devices which utilize
gaseous chlorine, and electrolytic processes for deriving
chlorine from sodium chloride, can malfunction on a stormy
sea.
Relatively small quantities of silver are bactericidal
or bacteriostatic. If the dose is high enough and contact
time long enough, adequate disinfection can result. More-
over, silver is easy to handle and maintains its potency for
long periods of time.
Ozone has not, as yet, been employed on ships because
of fire and explosion hazards associated with it, as well as
the sensitivity of the production mechanism. Neither does
it offer prolonged disinfection since, being toxic, it must
be removed from the water.
Ultraviolet irradiation disinfects quite well, but only,
when applied to relatively clear,^colorless water already
somewhat low in microorganisms (Muller, 1972). Unfortu-
nately, a ship's movement during stormy weather can stir up
sediment in the water tank and create problems of turbidity
which prevent the ultraviolet light from adequately pene-
trating the water.
Distillation is the method of choice for providing
drinking water on ships, although it too has its problems.
-------
452
One great advantage of this technique is the availability of
source water, once a ship is on the open seas. This limit-
less capacity obviates the need for strict conservation or
priority uses, practices which may endanger the quality of
the water.
Inasmuch as vacuum distillation may be done in the
temperature range of 25 to 40°C, it will not protect against
pathogens and should not be considered a substitute for
boiling. A properly operated still will produce water with
a chloride content below 6 mg/1, but the chloride test does
not indicate microbiological purity. It is recommended
(Muller, 1968) that distillation be used in conjunction with
chlorine or silver, but that highly polluted waters (such as
one may find in a harbor) not be used at-all in this pro-
cess .
e. Bacterial Standards for Drinking Water on Ships.
All water used on ships for drinking, food preparation, and
hand washing should conform to drinking water standards [See
Section D]. This means that 100 ml of water sample should
contain no JS. coli or total coliforms. Because it is
situated so close to polluted surface waters, ships' drink-
ing water should, in addition, be monitored for total bac-
teria, in which no more than 1,000 per ml are an acceptable
number. Pseudomonads (along with coliforms) are typical
inhabitants of polluted surface waters. They can grow
rapidly in low nutrient water and at low temperatures (e.g.,
in ice water). Because of these attributes and because P.
aeruginosa is a potential pathogen, a test for these ~~
organisms should be included among microbiological analyses
of ships' water. Culturing on gelatin pour plates will
reveal any pseudomonads as green fluorescent colonies.
f. Emergency Drinking Water Storage. In case of
shipwreck"consideration is first given to obtaining enough
water. Water quality and safety, therefore, become issues
of secondary importance.
Provision of sufficient amounts of potable water is one
way of combating deficiencies during emergencies. The
International Commission (International Ship Safety Treaty)
specifies that quantities of 3 1 per person be stored on
lifeboats and 1.5 1 per person on life rafts. These sup-
plies are usually stored in plastic tanks (though formerly
wooden or metal tanks were used) and fastened under the
-------
453
thwarts. Water may also be contained in sterilized plastic
bags or tins which, due to expense and number of processing
steps, are usually reserved for life rafters. To avoid
contamination after filling with good quality water, silver
compounds are added, thereby permitting storage for three
years or longer (Muller, et, al. , 1977). Also available are
several small devices for converting sea water to drinking
water, such as ion exchangers or distillers, which may be
installed on life boats and life rafts.
6. Water in Containers
Microbiologically safe, piped water is available to
most people in the industrialized nations a majority of the
time. In these countries, water is put into containers and
sealed for two principal reasons: (1) to have it available
in the event of an emergency in which normal supply is
interrupted; and (2) to allow water from a specific source
to be distributed commercially over a wide area. The main-
tenance of water quality during prolonged storage in a
container presents some special problems (Muller, 1969;
Sepilli, et al., 1965).
a. Water for Emergencies. Water stored in containers
for emergencies might be prepared in either of two ways:
first, water of adequate bacteriologic quality might be
sealed in containers just as it comes from the public supply.
Second, water may be sealed in containers (tins, plastic
bags, or glass vessels) and sterilized. Depending upon the
level of organic matter present, water that has not been
sterilized is likely to permit bacterial growth to reach
colony counts of millions per milliliter unless some long-
acting disinfectant such as silver salts is added. Steril-
ized water will remain bacteriologically acceptable for a
virtually unlimited period of time, depending only on the
integrity of the container.
The container is, therefore, the key to successful
storage of water for emergencies. Tins sometimes oxidize,
especially at the edges, so that metallic (tin or iron) ions
or products become dissolved in the water. The resulting
color, turbidity, and astringent o'r metallic taste impair
the organoleptic quality of the water. Plastic bags appear
to offer some comparative advantages, in that they are light
in weight, and visible changes in water quality will be
perceptible without opening the container. However, they.
-------
454
are more susceptible than tins to physical damage and occa-
sionally impart off flavors (described as musty, foul, or
not fresh) to the water if appropriate polymers are not
chosen [See also Section F.3.b].
t>- Commercially Distributed Water. For centuries,
people have gone to certain wells and springs to drink or
bathe in the waters as cures for, or to mitigate,, symptoms of
rheumatism and biliary or renal concretions, or to stimulate
intestinal motility. More recently, water from such sources
has been put into bottles (first of glass, but now more
often of plastic) and sent all over the world. In addition
to the renowned sources, commercially distributed water may
simply derive from wells in which mineral content is known
to be high (mineral water) or come from wells or public
systems and then be supplemented with salts and carbon
dioxide gas (bottled drinking water). Water bottled without
carbon dioxide "still bottled water" may be preferred by
those whose stomachs are upset by carbonation. Bottled
water is favored in Europe because people are sensitive to
the tastes and odors of chlorine and other substances present
in tap water or because tap water is thought to be hazardous
to health. Bottled water has become an item of great
economic importance, partly because advertising has led
people to believe it to be more healthful than tap water.
Nevertheless, water cannot be assumed safe simply because it
comes from a bottle. It must produce low bacterial colony
counts^and be free of fecal indicators, pathogens, and
potential infectious agents as well as toxic chemicals. The
bacteriologic quality of commercially distributed water
should remain as good as that of tap water [See Section D]
even through prolonged periods of storage at room tempera-
ture.
c. Bacteriological Quality of Water in Containers.
Water in containers may contain bacteria which were present
in the source water [See Section A] or may be contaminated
in the process of bottling, through contact with contami-
nated containers or closures, or by contaminated air [See
also^Section F.3.d]. This means that the bacteriological
quality of both the source water and the product in the
final container must be continually monitored (Geldreich,
et al., 1975). Water that has just been sealed into a
container should be free of coliforms and E. coli and should
have a colony count no higher than that of the source water.
Sporeformers sometimes associated with food poisoning and
-------
455
enteric diseases (e.g., Clostridium perfringens and Bacillus
subtilis) should also be absent. Pseudomonas species have
been a problem when certain plastic containers or caps with
plastic inserts are used, and high counts of ^. aeruginosa
II See Section B.l.a(vii)] have been found in many samples of
water bottled without carbonation (Muller, 1974; Soenke,
1974). Bacterial activity in the product may also give rise
to toxicologic problems; given a high initial nitrate level,
bacterial nitrosification may yield levels of nitrite which
are hazardous to newborn and young children, for whom bottled
water is especially recommended.
d. Storing Water in Containers. Whether publically
supplied or commercially packaged, drinking water will
invariably harbor a few indigenous microbes whose numbers,
under entirely normal circumstances, will increase as a
function of time, temperature, and available nutrients.
However, when water is stored in containers, initial counts
of ten to twenty colonies per ml may soon grow to popula-
tions numbering in the millions per ml. Later, after three
weeks to three months of storage, colony counts should
decrease as a result of nutrient depletion. Such a swell in
bacterial numbers can be prevented during bottling by the
addition of disinfectants, such as silver salts, or by
filter sterilization. Carbon dioxide in the water will help
to inhibit growth of the indigenous flora, but will not kill
pathogens or fecal indicators such as 1C. cold.. Therefore,
when analyzing samples it is important~to look beyond total
colony counts which, especially in the case of carbonated
water, may prove misleading. Along with the standard fecal
indicators [See Sections C.l.c to d], tests for pathogens
(such as P^. aeruginosa) [See Section C.2.c] and potential
pathogens (such as aerobic and anaerobic sporeformers) [See
Section C.l.e] should be routinely employed.
7. Summary
Bacteria are classified, and subsequently identified,
on the basis of the limited circumstances under which they
can grow and the nutrients they can use. Although indi-
vidual bacterial genera and species may be extremely limited
in versatility, there always seems to be one or more species
capable of growing in any given aqueous environment regard-
less of adverse conditions, extremes of temperature, or
apparent absence of nutrients. Growth of such organisms in
drinking water treatment, storage, and distribution facili-
ties often proves detrimental to the facilities or to the
quality of the water.
-------
456
Facilities are structurally affected by deterioration
of materials and by some of the chemical reactions that can
lead to mobilization of iron in water mains. The function-
ing of facilities is affected when deposits of microorganisms
and their products interfere with hydraulic conduction
through water pipes or coat the active surfaces of filter
media. Water quality may be degraded by inappropriate
storage or by treatment in small-scale apparatus, the
design of which is not microbiologically sound.
The importance of sanitation, in limiting the micro-
bial inoculum that may give rise to these technological
problems, must not be minimized. However, it must also be
recognized that drinking water is almost never sterile, that
organisms which may cause problems will eventually be en-
countered in every situation, and that only active measures
against such organisms can be expected to succeed in the
long term.
8. Recommendations
Construction materials used in treatment and
distribution systems should be selected (among
other factors) according to their resistance to
biodeterioration. Those components that are
prone to microbially-induced deterioration should
receive appropriate servicing and be replaced as
they near the end of their life expectancies.
Devices used in the home, primarily for the
sake of convenience, as adjuncts to household
tasks are apt to create bacteriological problems
and should be avoided'or professionally con-
trolled.
Water, destined for public consumption, that is
supplied in non-standard form (e.g., aboard
transport conveyances or packaged commercially)
should be subject to the same microbiological
standards of quality enforced for publicly
supplied water.
-------
457«,
H. SUMMARY
The most potentially significant constraint upon the
scope of the microbiology project is that it was to address
the drinking water supply problems of industrialized nations.
As was stated in the introduction to .this report, indus-
trialized nations usually have large quantities of water or
industrialization would not have been possible. Even so,
the quantities of water may not be adequate to meet antici-
pated needs and the quality of available water may be much
degraded by prior use. It is well, at this point, to ask
what distinctive features of water supply in industrialized
nations are significant to microbiology.
>As major feature is high daily per capita water use.
Daily consumption of beverage water is physiologically
determined: the body's water losses must be replaced, so
the total daily use of water, by ingestion, is determined by
the climate and the size of the population in both indus-!
trialized and developing countries. Other uses of water in
the home, which tend to be larger in industrialized nations
than in developing countries, include flushing of water-
carriage toilets and cleansing of food, clothing, the home,
its contents, and its inhabitants. Outside-the-home uses
that contribute to the high daily per capita demand in
industrialized nations include a variety of applications in
commerce and industry, as well as the occasional use of
water for fighting fires.
Where the volume of water used exceeds the supply, a
yreuse factor" (e.g., perhaps 25 percent of the water avail-
able at some point in a certain river has already been used
somewhere upsteam) may be calculated. Some classes of water
reuse are of far greater microbiologic concern than others.
For example, human feces, and therefore, the water used to
eliminate them from a household, are the most significant
sources of infectious agents transmissible through water.
By contrast, the direct disposal of human waste into water-
ways, that occurs in some parts of the world (either from
ships and boats on the water or from unsewered communities
the banks), may decrease the calculated reuse factor,
-------
458
but increase the risk of waterborne disease. Other micro-
biologically significant components of used water include
substances which may support the growth of organisms in the
water, and heat. Project Area V of this CCMS Pilot Study
addresses the broader aspects of water reuse, but reuse must
also be considered in the context of drinking water micro-
biology because the majority of the infectious agents that
might be transmitted by drinking water derive from the
human intestines and are carried by wastewater, in some
proportion, into waters that may serve as drinking water
sources. The problem of preventing waterborne infectious
diseases would be greatly mitigated if an appropriate, non-
polluting alternative to the water carriage toilet could be
developed; such a device would be accepted only if it met
the esthetic standards now prevalent in industrialized
nations. Otherwise, the protection of source waters demands
that wastewater discharges be carefully supervised and that
wastes be treated and disposed of in a manner that permits
water to be reclaimed as necessary. Many features of this
problem are considered in Project Area V and, in the speci-
fic context of groundwater protection, in Project Area VI.
Topic A of the microbiology report concerns Raw Water.
Every community or water supplier should, and ordinarily
will, select the purest available source for use in drinking
water production. Groundwater from a deep, protected aqui-
fer may contain only a few microorganisms, none of which is
of intestinal origin. Water of this kind may be used in
both private and public supplies with little or no treat-
ment. Even then, chlorination may have to be used in public
systems to avoid problems in distribution. However, really
well-protected aquifers in some areas may lie at such a
depth that they are not accessible to private users. The
depth of an aquifer may be less of a problem where public
water supplies are concerned, but the quantities of water
available must be adequate to the community's needs, or less
pure raw water will have to be used.
Groundwater is frequently discharged to surface water-
ways after use (and, one would hope, after treatment), but
increasing efforts are being made to dispose of wastewater
in such a way that the groundwater will be recharged. The
success and safety of this method of water reclamation
depend greatly upon the ability of soil to accomplish micro-
biologic purification of the water between the point of
application and the aquifer or, at least, the point of
abstraction of water from the aquifer. The efficiency of
-------
459
purification will depend on the rate of application, on the
character of the soil, and on whether recharge is undertaken
by vertical infiltration under the influence of gravity or
by pressure injection directly into the aquifer. This
artificial recharge must be done with extreme caution, for
the microbiologic purification of water by soil is not a
well understood process and once contamination of a ground-
water source has occurred, it is extremely difficult to
correct. This is an area where a great deal of research is
needed; until results are available, it will be prudent to
treat wastewater, to be used for groundwater recharge, to a
degree that does not demand too much of the purifying capa-
city of the soil. The introduction of viable organisms to
an aquifer is not the only microbiologic concern in ground-
water recharge. Some microbially-induced chemical trans-
formations can lead to degradation of groundwater and toxins
liberated upon the death of microbial contaminants may also
prove significant.
Surface sources of raw water are more difficult to
protect from microbial contamination and will frequently
support the growth of some of the organisms that may be
introduced. Airborne contamination and runoff from adjacent
land surfaces, in addition to discharged wastewater, may
contribute undesirable organisms or nutrients to surface
waters. Surface water quality may also be directly affected
by human activities such as recreational swimming and boating,
residence in houseboats, or navigation on waterways for
commercial transport of goods. Under optimal conditions,
microbialgcounts in lakes, extremely rich in nutrients, may
exceed 10 per ml. Many organisms in surface waters are
capable of inducing undesirable chemical transformations.
Photosynthetic bacteria and cyanobacteria (formerly called
blue-green algae) can grow to high levels, given favorable
conditions and adequate light. These bacteria, as well as
the true algae, can induce off-flavors in water, bind large
quantities of dissolved oxygen under certain circumstances,
and physically interfere with water purification. Toxins
produced by some cyanobacteria may also be significant to
human health. It is clear that surface waters which serve
as sources of drinking water need more rigorous protection
than many of them get, but it is also clear that more research
will be needed before the requisite protective measures are
well understood.
-------
460
Topic A also includes a survey of the microbiologic
quality of at least some of the raw water in nine different
countries. Not surprisingly, raw water quality is generally
better where the source is groundwater and where it origi-
nates in less densely populated areas. Even in these situa-
tions, raw water quality is seldom so consistently excellent
that one-step treatment can safely be trusted. Where raw
water quality is poorer, more intensive treatment methods
are generally employed. Finished water of adequate quality
and safety can evidently be produced by a series of unit
processes ending with disinfection; but the community served
is, obviously, more vulnerable to any kind of event that
might even temporarily interrupt treatment of the water
before'distribution.
Recommendations Concerning Raw Water
1. More research should be done on the microbial
ecology of groundwater.
2. The microbiologic quality of source waters
should routinely be tested, especially in those
instances where the quality of the raw water is
judged so good as to require only minimal
treatment.
3. Laws to protect both groundwater and surface
water sources should be more restrictive and
specific, and should include adequate provision
for enforcement.
4. Means of conveying good quality raw water to
areas where source water quality is deficient
should be investigated further.
Topic B surveys the specific Pathogens that may be
transmitted through drinking water. Generally, these are
infectious agents including bacteria, viruses, protozoa, and
metazoa. The proximate source of these agents in water may
vary, but most of them ultimately derive from the human
intestines. Intestines of other warm-blooded animals are an
alternate source of some waterborne agents infectious to
humans and only a very few of these agents are apparently
capable of living free in the environment for extended
periods of time. Although the agents discussed are all of
concern to human health, the majority of 'the infections
caused by them are mild or asymptomatic. This fact tends
-------
461
to complicate the investigation of waterborne disease out-
breaks, in that the majority of infections may go unrec-
ognized unless intensive and wide-ranging laboratory test-
ing, frequently of fecal specimens, is undertaken.
Although quite long periods of persistence in water
have been reported for some pathogens, the aqueous environ-
ment is not generally a favorable one for human infectious
agents. Physical and chemical factors, as well as compe-
tition or predation by better-adapted indigenous microbial
species, combine to kill or inactivate pathogens in water
with the passing time. Other things being equal, higher
temperatures and longer detention times favor the destruc-
tion of greater quantities of waterborne pathogens. How-
ever, it is important to note that most time-dependent
death processes occur as logarithmic functions of time, so
that the level of the dying agent, theoretically, never
reaches absolute zero. This raises the dual questions of
how sensitive a method must be for detecting a waterborne
pathogen, and what significance to human health may be
represented by some small residual level of a pathogen in
water. These two questions may or may not be intimately
interrelated. It is often argued that, as difficult as
pathogens frequently are to detect in water, there is no
need to develop methods to detect them at levels below which
they represent a threat to health. However, there are those
who believe that no level of an infectious agent is so low
as to be insignificant to human health and that sample-to-
sample variation offers the possibility that water sent to
the laboratory for testing may> contain less contaminant than
water that someone drinks from the same source, so there is
no limit to the desired level of sensitivity in testing for
waterborne pathogens.
Further research is needed both to determine the relative
threat to human health that is presented by different levels
of a pathogen in water and to develop more sensitive methods
for detecting waterborne pathogens. However, there is also
a need for simpler methods of detecting waterborne patho-
gens, in that more samples, tested by more laboratories, may
eventually produce a more useful body of information, to aid
in producing a safer supply of water, than would the testing
of a few very large water samples by the few laboratories
that are equipped to deal with them. Even if these were
available, tests for pathogens could not afford a basis for
routine quality control in drinking water supply. Given the
problems inherent in detecting waterborne pathogens, it is
not surprising that microbiologic testing of water is focused,
instead, upon indicators.
-------
462
Recommendations Concerning Pathogens
1. Even,where water is protected, as is true of
many groundwater sources, disinfection is
recommended. This is especially important
where finished water is to be stored rather
than being used immediately.
2. Aftergrowth of opportunistic pathogens during
distribution of finished water should be prevented
by maintaining a chlorine residual throughout
the distribution network, especially in large
systems.
3. Testing for pathogens within the distribution
water is appropriate: (a) after contamination
is found to have occurred; (b) to trace the
source of an outbreak; and .(c) in analyzing
disinfection efficiency.
4. Given the lack of correlation between viruses
and the bacterial indicator systems, more
research on the antiviral effectiveness of
various water treatment processes, is needed.
5. Mapping of waterborne outbreaks should be con-
ducted in conjunction with epidemiblogical sur-
veys of the population served by the water sup-
piy.
Topic C of this report deals with Indicator Systems,
which are microbiologically-based quality control methods.
Indicator systems already established in use are generally
based upon enumeration of viable organisms on a selective
basis. Some, but not all, of these indicators are supposed
to correlate with the occurrence of fecal contamination and,
implicitly, with the presence of enteric pathogens. The
closeness of correlation between different established
viable indicator systems and fecal contamination varies
greatly. Where correlations are low, it is usually either
because the organisms measured may include some that are not
of fecal origin at all or sane that are capable of
proliferation in the environment outside of the body.
Another potential liability, where disinfection is practiced,
is that the indicator organisms may be more sensitive to the
disinfectant than are some of the enteric pathogens which
might be present. In addition to indicators of fecal contami-
nation, there are indicator systems that gauge water quality
-------
463
and others that serve to signal the presence of specific
pathogens. Finally, there is always reason to wish for
indicator systems that are more rapid or simpler to apply;
such systems could permit more replication of tests.
These considerations led to the inclusion in Topic C
of a survey of proposed alternate indicator systems based
upon viable microbes, including coliphages and animal viruses,
as well as a number of groups of bacteria. Some of these
may eventually serve special purposes on a regular basis,
but none is presently ready to supplant the "established"
indicators for routine quality control in drinking water
treatment and distribution. Other indicator systems sur-
veyed are not based on determining numbers, of viable organ-
isms, or at least may not require incubation through many
microbial generation times before results are obtained.
Some of these alternate systems seem to offer significant
potential for continuous monitoring of water quality in
situations where rapidly obtained results will permit prompt
remedial action. No .system considered would obviate the
need for proper sampling techniques or for adequately trained
laboratory personnel.
Topic D surveys Testing and Standards for drinking
water in various countries.For the time being, microbib-
logic quality control of drinking water, in most of the
countries surveyed, is based upon the coliform group.
However, in some countries, thermo-tolerant coliforms or
Escherichia coli (based on a working definition of the
species)are determined instead or in addition to the coli-
forms. Procedures for both sampling and testing are seen to
vary from country to country, but the differences are not so
great that the coliform test results cannot be compared.
The general intention in every case seems to be that coli-
forms (or thermo-tolerant coliforms or 12. coli, as the case
may be) should be absent from samples of finished drinking
water taken at the treatment plant or, in many instances,
throughout the distribution system. Unfortunately, the
survey that was done did not yield adequate bases for com-
paring methods of laboratory quality control; nor was it
possible to determine how corrective action is taken in the
event that indicators are found in finished drinking water.
Efforts being made by several organizations to standardize
analytic procedures in water microbiology are certainly to
be commended. Whatever the analytic procedures used, it
seems clear that the ability of laboratory microbiology to
contribute to the safety of drinking water depends less on
how standards are written than on the dedication with which
they are applied or enforced.
-------
464
Topic E deals with the Treatment Methods used in pro-
ducing finished drinking water. The emphasis in this case
is on how various unit processes affect pathogens and indi-
cator systems. However, much information on other aspects
of some of the same treatments can be found in the reports
of Project Area II: Advanced Treatment Technology. The
variety and degree of treatment used in preparing drinking
water should be, and usually are, determined by the quality
of the raw water that is available. Where the source is
variable in quantity or quality, reservoir storage may be
used to buffer some of the fluctuations. The primary function
of the reservoir may be storage, but considerable changes,
for better or worse, can result from holding water in a
reservoir. This depends on whether the reservoir is managed
so as to minimize opportunities for contamination or growth
of noxious organisms and to make use of the water's tendency
for self-purification; at best, storage of water in a reser-
voir can serve as a treatment step, arid is regarded as such
in this report. Physical treatments, such as coagulation
and flocculation or various versions of sand filtration,
serve to remove suspended matter including many microbial
cells. These treatments are especially important in re-
moving protozoan cysts and metazoan eggs, as well as in
eliminating suspended matter that might interfere with
disinfection. Slow sand filtration, and often activated
carbon treatment, have an important biological component.
Activated carbon treatment is intended, primarily, to remove
impurities that are dissolved, rather than suspended, in
water. To the degree that the substances removed might have
served as substrates for microbial growth, the growth becomes
less likely to take place in the water, but more likely on
the carbon surface. This will not necessarily produce a
health hazard, but it can lead to disinfection problems and
to a decrease in water palatability. The microflora in slow
sand filters can effect important reductions in biodegrad-
able dissolved substances, especially in water pre-treated
with ozone.
The ultimate defense against carry-over of pathogenic
bacteria and viruses into finished water is, ordinarily,
disinfection. If pathogens were unlikely to have been
present in the raw water, disinfection may be done solely to
suppress opportunistic organisms and to avoid technical
problems during distribution of the water. This requires
use of a disinfectant such as chlorine, a residual of which
can be maintained throughout the distribution network. On
the other hand, disinfection may be needed to kill large
numbers of microorganisms, possdbly including pathogens.
-------
465
Ozone is seen to be a major alternative to chlorine in this
application; it is already well established as a primary
drinking water disinfectant in many areas. Other disin-
fectants are also surveyed) and some of these may eventually
capture a portion of the disinfection .market. It is impor-
tant to note that no disinfectant can make good water from
bad, and that disinfection may fail if water has not first
been treated in a manner appropriate to its original qual-
ity, so that the disinfectant has only to act upon reason-
able numbers of microorganisms.
Topic F addresses the problem of maintaining finished
water quality during Storage and Distribution. This repre-
sents a special challenge from both the standpoints of
quality and safety. On the one hand, the quality of fin-
ished water at the treatment plant must be assumed to be the
best that can be achieved with the means available, so that
storage of finished water, for example, can maintain or
degrade quality, but cannot improve it as in the case of raw
water storage. On the other hand, the epidemiological
record shows that cross-contamination and back-siphonage, by
introducing raw sewage or otherwise polluted water into
finished water in distribution, have been relatively impor-
tant among causes of the rare outbreaks of disease asso-
ciated with public water supplies. The most general prob-
lems are those of avoiding growth of organisms present in
the finished water (for example, by maintaining an active
level of chlorine in water throughout the distribution
network) and preventing contamination of the finished water
from external sources (for example, by covering service
reservoirs in which finished water is stored). The mate-
rials and the manner of construction of the facilities are
critical at every stage. Water contact surfaces in reser-
voirs and in mains all too frequently include materials
which may support microbial growth. The joints that have
been well designed to exclude contaminants from without are
sometimes found to present favorable conditions for micro-
bial gtowth within the system. Older materials used in
constructing water distribution systems all have disadvan-
tages, including increasingly high costs of production and
installation and, in some instances, exceedingly short
service lives when in contact with the water of some com-
munities. Newer materials and joint designs appear to offer
important advantages; but the testing of these cannot always
include all of the conditions to which they will be sub-
jected in use at various places. Thus, unforeseen diffi-
culties are always possible, even under what may be de-
scribed as routine conditions.
-------
466
Butt, C.G., Baro, C. and Knorrv R.W. (1968) Naegleria sp.
identified in amoebic meningoencephalitis. American
Journal of Clinical Pathol. 50:568-576. ~
Butterfield, C.T. (1948) Bactericidal properties of chlora-
mines and free chlorine in water. U.S. Public Health
Rep. 63:9340-9349.
Butterfield, C.T., Wattie, E., Megregian, S. and Chambers,
C.W. (1943) Influence of pH and temperature on the
survival of coliforms and enteric pathogens when exposed
to free chlorine. Public Health Rep. 58:1837-1866.
Butzler, J.P., Dekeyser, P., Detrain, M. and Dehaen, F.
(1973) Related Vibrio in stools. Journal of Pediatrics
82:493-495. ~
Cabelli, V.J. (1977) Indicators of recreational water
quality. In Bacterial Indicators/Health Hazards
Associated with Water ASTM STP 635, pp. 222-238 (Edited
by Hoadley, A.W. and Dutka, B.J.), American Society for
Testing and Materials, Philadelphia.
Cabelli, V.J. (1979) Evaluation of recreational water
quality, the EPA approach. In Biological Indica-
tors of Water Quality (Edited by James, A. and Evison,
L.M.)Pub. Wiley,London.
Cabelli, V.J., Dufour, A.P., Levin, M.A. and Haberman, P.W.
(1976) The impact of pollution on marine bathing beaches:
An epidemiological study. In Proc. Symposium on Middle
Atlantic Continental Shelf and the New York Bight,
Special Symposium # 2 (M. Grant Gross, editor). The
American Society of Limnology and Oceanography, Inc.,
New York City, pp. 424-432.
Cady, P. (1975) Rapid automated bacterial identification by
impedance measurements. New Approaches to the Iden-
tification of Microorganisms pp. 73-99 (Edited by
Heden, C.G. and Illeni, T.) John Wiley & Sons, Inc.,
New York.
Cady, P. (1978) Progress in impedance measurements in
microbiology. Mechanizing Microbiology, pp. 199-239
(Edited by Sharpe, A.N. and Clarke, P.S.) Charles C.
Thomas, Publisher, Springfield, Illinois, U.S.A.
-------
467
pipes may be degraded by microbial action, either because
the organisms were able to use the material of the pipe as
substrate or because microbial metabolism caused minerals to
be eroded from or deposited on the inner surfaces. Micro-
bial cells themselves, and the slimes associated with some
of them, are able to coat resins and filter.media or the
interiors of pipes so as to exert a direct adverse physical
effect upon the function of the facility. It is not sur-
prising that resin function would be extremely susceptible
to microbial growth, given the fact that normal function of
the resin depends upon intimate interaction between the
resin surface and the water; however, it is also true that a
thin microbial slime coat can significantly interfere with
the hydraulic conductivity of a water main, even though the
deposit obstructs very little of the inside of the pipe.
Another set of technological problems involves the
storage of water aboard ships, and in containers for commer-
cial distribution or for use in emergencies. In a way, one
might assert that drinking water in these contexts needs to
be even purer than that in public supplies, for these classes
of stored water will ordinarily be used in exactly the
condition that the consumer receives them. Problems asso-
ciated with ships' water supplies are discussed in detail;
some of these problems are shared with supplies of drinking
water aboard all classes of public conveyances, but they may
be more extreme with ships because longer periods of storage
and greater volumes of water are involved, and because many
-opportunities exist for cross-contamination from wastewater,
water from the ship's bilge, .and wash water derived from the
often-polluted water in which the ship floats. Water sealed
in containers also requires great care, as to the initial
quality of the water, the use of preservatives (if any), and
the selection of a container. Obviously, those who must use
packaged water in times of disaster will have no opportunity
at that point to reject that which, on the basis of off-
flavor, odor, or appearance, might be suspected of being
toxic. On the other hand, those who regularly drink bottled
water in their daily lives place their trust in the safety
of the commercial product and are therefore vulnerable to
any lapse on the part of the bottler or distributor.
This report necessarily emphasizes water treatment and
control measures for routine use. However, it must be
recognized that emergencies do arise and that plans for ,
dealing with them should be made beforehand, as much as
possible. Causes of emergencies, in what may be descending
order of likelihood, include undetected deterioration in the
-------
468
physical apparatus, human error, power failures, adverse
weather, willful mischief, earthquakes, and war. Both
apparatus at the treatment plant and in the distribution
network may be subject to deterioration or sudden malfunc-
tion. Human errors might include such events as construc-
tion machinery breaking water mains. Loss of electrical
power could inactivate pumps, ozone generators, and vital
control apparatus. Adverse weather can cause power fail-
ures; or extreme cold, floods, or windstorms may directly
interfere with water treatment or distribution. Willful
mischief would include any malicious act by which one or a
few persons abused a water system in an effort to disrupt
society. Earthquake prediction seems to be progressing, but
is still not very useful for protecting water supplies.
Finally, if war occurs, water systems may be disrupted
incidentally to general bombardment, overtaxed through
excessive water demand for firefighting, or directly tar-
geted as a vehicle for biological warfare.
All communities are vulnerable to emergencies affecting
their water distribution systems. However, there are po-
tential differences, involving water sources and treatment,
in susceptibility to emergencies. A community that has
relatively low-quality water must use a complex treatment
scheme and is vulnerable from that standpoint. On the other
hand, a community that derives very pure source water from a
deep aquifer will have no water at all if it loses its
pumping capacity. Large water supply systems probably
present more points of vulnerability than small systems, and
communities that derive their water from distant sources are
especially at risk.
If treatment is interrupted, but distribution is main-
tained, microbiologic safety can be achieved by drawing
water from the tap and boiling it. Otherwise, any available
water that does not contain acute toxicants may have to be
boiled and used. Restoration of treatment and distribution
services are likely to require extensive flushing and use of
large quantities of chlorine to restore a system to normal;
plans for such action should be made in advance, and key
personnel should learn their tasks. Large communities,
where great numbers of people might be unable to supply
themselves with water in the event of a system stoppage,
should consider storage of water for emergencies in mod-
erate-sized containers at well-distributed and marked loca-
tions. .
Industrialized nations in general share a relatively
high level of public health, as measured by long life
-------
469
expectancies and low child mortalities, which is at least
partly a tribute to the technical and institutional success
of drinking water supply in these countries. Even private
water supplies in these countries are often monitored on a
limited basis, so that relatively few inhabitants use water
that has not been safety tested in some manner. It is,
perhaps, noteworthy that the primary safety criteria ap-
plied, even to this day when many other aspects of drinking
water safety are under scrutiny, are based upon microbio-
logic indicator systems. This is reasonable, for a great
part of the public health gains that have been achieved in
industrialized nations have resulted from reduced incidence
of infectious diseases through sanitation. Hardly any
aspect of drinking water microbiology would not be likely to
benefit from further research, but it is important to note
that presently available treatment techniques, monitoring
methods, and other features of current drinking water supply
practice are serving- their purposes remarkably well. In a
general sense, standards presently in effect must never be
relaxed because the populations of industrialized nations,
accustomed to a high level of sanitation, are likely to be
quite vulnerable to any abrupt lapse in established drinking
water practice.
Change is inevitable, however, and new classes of
chemical contaminants are being identified in wastewaters
and in some raw waters from which drinking water must be
produced. Aspects of these problems are discussed in Project
Areas I and IV. Research is needed both on effective and
feasible methods for protecting source waters and on modi-
fied disinfection methods that will produce fewer -undesirable
disinfectant derivatives while serving the original purpose
of disinfection, which is to kill as many microorganisms as
possible in the water. The task of protecting source waters,
from a microbiologic standpoint, will be aided when more
research results are available regarding detection methods
for waterborne pathogens, as well as the probability of
infection as a function of ingesting different quantities of
waterborne pathogens. Indicator systems that are intended
to signal fecal or other microbiologic contamination of
water might be further refined and standardized, but it
seems likely that monitoring the adequacy of water treatment
and disinfection could better be based on the development
and application of a separate set of indicator systems.
These, and indicator systems designed to detect recontami-
nation of finished water in distribution, probably stand to
be most improved by automation or modification to afford
shorter readout times. In this age of dramatically improved
international communication, it seems clear that more standardi-
zation of criteria for water quality and safety will ensue.
-------
470
If primary disinfection procedures must be modified out
of concern for interactions between the disinfectants and
chemical contaminants of water, further research will be
needed on the adequacy of alternative disinfectants. At the
same time, it will be very important to determine and attempt
to utilize the antimicrobial effects inherent in all of the
other unit processes employed in water treatment. Research
to aid in protecting the quality of finished water during
storage and distribution will, assuredly, focus on the
development of low-cost, durable materials that are inert to
the micrqflora in the water, but there are also many other
research needs in this area. To the degree that microbial
growth is capable of creating technologic problems, which
have been enumerated previously, it is important that research
contribute more to the understanding of these microbiologic
processes, for it may be that the organisms cannot be entirely
suppressed, but only minimized. Finally, further research
on the evaluation and maintenance of water quality in closed
containers is still needed.
Many new concerns about drinking water safety have been
raised in recent years. Because these are generally chem-
ical in nature and may be associated with such dire effects
as cancer, they have tended to overshadow the microbiology
of drinking water. Under the circumstances, it seems fitting
to close by pointing out that: (1) the primary criteria of
drinking water safety and quality are based upon microbio-
logic indicator systems; and (2) in any major lapse in
drinking water treatment and distribution practices, the
most immediate consequences to consumer health are more
likely to be caused by pathogenic microorganisms than by
chemicals.
-------
471
GENERAL RECOMMENDATIONS
Every public water supply should begin with the highest
quality raw water that is available in quantities
sufficient to meet the community's needs. Efforts to
protect and improve the quality of source waters are
important; both waste discharges and non-point sources
of pollution should be considered in attempts to prevent
or alleviate contamination. Where possible, water to be
used for irrigation or for industrial purposes other
than food, drug, or cosmetic manufacture, should usually
be drawn from less pure sources than those from which
the public supply derives.
Disinfection is necessary, but not always sufficient, to
ensure the safety of drinking water from virtually any
source.
Complete treatment of drinking water, including at
least coagulation and sedimentation with sand filtration
or alternatively dual filtration including effective
slow^sand filtration, followed in all cases by disin-
fection, is essential in all cases where source waters
are unprotected and is highly desirable even with pro-
tected sources.
Sound engineering practice is required to produce safe
drinking water; microbiologic laboratory testing per-
formed on a routine basis, according to standardized
methods, and by properly trained and supervised staff '
is an important basis for assessment of drinking water
quality. Indicator systems for use in these tests
may be selected for any one of the following purposes:
(i) to signal fecal contamination; (ii) to detect any
abnormal and probably undesirable conditions that may
occur; or (iii) to warn of the probable presence of
specific pathogens. Larger waterworks, at least,
should develop microbiologic quality control procedures
and baseline data for all stages of treatment from raw
water through distribution. Prompt corrective action
-------
472
8.
9.
should be taken when norms are exceeded. At least one micro-
biology laboratory in each country should have the
ability to detect waterborne pathogens, either for spot-
checking or for investigating outbreaks, both from
water samples and from clinical specimens. Results of
tests for both indicators and pathogens should be re-
corded through a focal national center and shared on an
international basis.
Innovative indicator systems, capable of signaling
fecal contamination, problems in treatment efficiency,loss of
integrity of the distribution system, and perhaps the
presence of pathogens, should continue to be sought.
Rather than try to find a single indicator system that
will serve all of these disparate functions at once,
emphasis should be placed on individual .systems offering
convenience and economy that will allow more frequent
testing.
In addition to the well-established research on water-
borne bacterial pathogens, considerably more research
is needed concerning viruses and protozoa transmissible '
by drinking water, in the areas of the dose-dependence
of peroral infectivity and pathogenicity, the detection
of these agents in water, and their removal or destruc-
tion by water treatment and disinfection processes.
Monitoring of raw water quality on the basis of appro-
priate indicator systems is desirable in all cases and
essential in those instances where the usual purity of
the raw water is such that less than complete treatment
is used. At least one indicator system that is directly
correlated with fecal contamination should be included;
the choice of other indicator systems to signal other
kinds of problems should be made on the basis of knowl-
edge about local conditions.
Materials of which water treatment and distribution
facilities are constructed should be pretested for
chemical and biological stability. Testing methods, as
well as results, should be shared internationally as
much as possible; however, it is also best to test
materials, before use in a given system, with the very
water with which they will, in fact, be in contact.
Finished water in distribution, in both public and
semipublic systems, should be sampled at representative
locations and tested microbiologically with a frequency
-------
473
that depends on the size of the population served.
Private water supplies should be tested at least annually,
In all instances, the presence of coliforms, thermo-
tolerant coliforms, or _E. coli in a 100-ml sample
should be treated as unacceptable, or at the very least,
undesirable.
10. To minimize aftergrowth or other technological problems
and to provide a means of determining whether cross-
contamination has occurred, water in public supply
distribution should wherever possible contain a measur-
able residual level of disinfectant (e.g., chlorine) at
all points.
11. Inasmuch as distribution systems are a potential source
of problems in all water supplies, every system should
be under continuous surveillance. Where problems are
identified, they should either be eliminated by modifi-
cation of the system or be mitigated by routine main-
tenance procedures.
12. Procedures for the installation and repair of water
mains should be established beforehand and applied
diligently when needed. Plans for dealing with emer-
gencies should be made and communicated, in advance, to
those responsible for implementing them.
13. Means are needed to control crossconnections, to
ensure both that the consumer does not degrade publicly-
supplied water to the detriment of his own health and
that his use of the water does not cause contamination
that threatens the health of others. To achieve these
objectives, plumbing codes may need revisions and better
enforcement with respect to water supply systems in
buildings, to attachment devices that use water, and to
point-of-use treatment units attached to the consumer's
tap. .
14. More intensive research and epidemiologic surveys are
needed to determine the true health effects of microbes
and their products in finished drinking water. For this
purpose, closer cooperation and communication are needed
among practicing physicians and veterinarians, public
health authorities, and water microbiologists. Any
proposed change in treatment, distribution, or quality
control practice should be evaluated from the standpoint
of probable impact on public health, as far as possible,
before implementation.
-------
474
REFERENCES
Adamek, R., Daubner, I., Johnova, V. and Trzilova B. (1977)
Preliminary tests of the influence of wastewaters from
sugar factories upon the numbers of intestinal bacteria
in water. Ztl.u 'BaKteriol. Hyg. Abt. I, orig. B
164, 340-351.
Adams, M.H. (1959) Bacteriophages.
New York, 592 pages.
Interscience Publ., Inc.,
Afghan, B.K., Ryan, J.F., and Tobin, R.S. (1977) Improved
method for quantitative measurement of adenosine
triphosphate in lake waters, activated sludges and
sediments. Environmental Analysis, pp. 29-45 (Edited
by Ewing, G.W.) Academic Press, New York.
Ahearn, D.G. (1974) Identification and ecology of yeasts of
medical importance. In Opportunistic Pathogens, pp.
129-146 (Edited by Prier, J.E. and Friedman, H.)
University Park Press, Baltimore.
Ahearn, D.G. (1978) Medically important yeasts. Ann. Rev.
Microbiol. 32:59-68.
Ahmed, K.Y. (1975) Survival of bacteria in turbid and
polluted waters. M.S. Thesis, New Mexico State Univ.,
39 pages. Available from National Technical Infor-
mation Svc., 5285 Port Royal Rd., Springfield, Va.
(Rept. #PB-268801).
Ainsworth, R.G., Ridgway, J. and Gwilliam, R.D. (1978)
Corrosion products and deposits in iron mains. Proc.
Conf. Water Distribution Systems, 1978. Paper 8,
Session 2,Publ. Water Research Centre, Medmenham.
Alexander, M; (1971) Microbial Ecology.
Sons, New York, 467 pages.
John Wiley &
-------
475
Alexopoulos, C..J. . - _.! (1962) Introductory Mycology, 2nd
edition. John Wiley & Sons, New York, 613 pages.
Allen, J.M. and Geldreich, E.E. (1975) Bacteriological
criteria for groundwater quality. Ground Water 13:1-7.
Alonso, J.M., Bercovier, H., Servan J., Bourdin, M. and
Mollaret, H.H. (1976) Contribution a 1'etude epidemio-
logique des infections a Yersinia enterocolitica. II.
Enquete ecologique. Med. Mai. Infect. 6:434-441.
American Public Health Association (1976) Standard Methods
for the Examination of Water and Wastewater, 14th
edition. American Public Health Assoc., Washington,
D.C., 1193 pages.
Andre, D.A., Weiser, H.H. and Malaney, G.W. (1967) Sur-
vival of bacterial enteric pathogens in farm pond
water. J. Am. Waterworks Assoc. 59:503-508.
Annapurna, E. and Sanyal, S.L. (1977) Enterotoxicity of
Aeromonas hydrophila. Journal of Medical Microbiol.
10:317-323. ~
Anon (1974) Gesetz zur Gesamtreform des Lebensmittelrechtes.
15.8.1974, BGesbl. I, S. 1945.
Appleton, H. and Buckley, M. (1977) Virus-like particles in
winter vomiting disease. Lancet, Feb. 19, 409-411.
Appleton, H. and Pereira, M.S. (1977) A possible virus
aetiology in outbreaks of food-poisoning from cockles.
Lancet, Apr. 9, 780-781.
Ator, L.L. and Starzyk, M.J. (1976) Distribution of group D
streptococci in rivers and streams. Microbios 16:91-
104.
Atterholm, IfcJ ' Hallberg, T. , Ganrot-Norlin, K. and
Ringertz, O. (1977) Unexplained acute fever after a
hot bath. Lancet, II, 684-686.
Aurand, K. , Hasselbarth, U. , Muller, G.j Schumacher,
W., Steuer, W. (1976) Die Trinkwasserverordnung Erich
Schmidt Verlag, Berlin, 255 pages.
Baars, J.K. (1957) Travel of pollution and purification en
route in sandy soils. Bulletin of the World Health
Organization 16:727-747.
-------
476
Bachrach, U. and Bachrach, Z. (1974) Radiometric method for
the detection of coliform organisms in water. Appl.
Microbiol. 28:169-171.
Bachrach, U. and Friedman, A. (1971) Practical procedures
for the purification of bacterial viruses. Appl.
Microbiol. 22:706-715.
Baqley, S.T. and Seidler, R.J. (1978) Comparative patho-
genicity of environmental and clinical Klebsiella.
Health Lab. Sci. 15:104-111.
Baker, M.N. (1934) Sketch of the history of water treatment.
J. Am. Wat. Wks. Ass. 26:902-938.
Barua, D. (1970) Principles and practice of cholera control.
Public Health Papers, World Health Organization 40:29-
_ .
Basco, J. and Szalay, A. (1978) Accumulation of arsenic,
lead, and other heavy elements in the iron-manganese
oxide-hydroxide precipitation in the pipelines of city
waterworks. The Science of the Total Environment
9:271-278.
Bates, R.C., Shaffer, P.T.B., and Sutherland, S.M. (1977)
Development of poliovirus having increased resistance
to chlorine inactivation. Appl. Environ. Microbiol.
34:849-853.
Baylis, J.R. (1936) The efficiency of rapid sandfilters in
removing the cysts of the amoebic dysentery organisms
from water. Public Hlth. Rep. 51:1567-1573.
Becker, K. (1969) Die Verbreitung der Leptospirosen bei
Ratten in einer Grobstadt (Berlin). Schriftenreihe
Wa-Bo-Lu ; 32:49-66.
Bedulevich, T.S., Svetlakova, M.N., and Trakhtman, N.N.
(1953) New data on the use of chlorine dioxide in water
purification. Gig. Sanit. 19:14-17.
Bell, J.F. and Stewart, S.J. (1975) Chronic shedding tula-
remia nephritis in rodents: Possible relation to
occurrence of Francisella tularensis in lotic waters.
Journal of Wildlife Pis. 11:421-430.
-------
477
Benarde, M.A., Israel,, B.M., Olivieri, V.P., and Granstrom,
M.L. (1965) Efficiency of chlorine dioxide as a bac-
tericide. Appl. Microbiol. 13:776-780. ;
Benarde, M.A., Snow, W.B., Olivieri, V.P., and Davidson, B.
(1967) Kinetics and mechanism of bacterial disinfec-
tion by chlorine dioxide. Appl. Microbiol. 15:257-265.
Benson-Evans, K. and Williams, P.F. (1975) Algae and bryo-
phytes. In Ecological Aspects of Used-Water Treat-
ment 1:153-202 (Edited by Curds, C.R. and Hawkes, H.A.)
Academic Press, New York.
Berg, G. (1964) Devitalization of microorganisms by iodine.
Virology 22:469-479.
Berg, G. (1966) Virus transmission by the water vehicle.
III. Removal of viruses by water treatment procedures.
Health Lab. Sci. 3:170-181.
Berg, G. (1967) Summary. In Transmission of Viruses by
the Water Route, pp. 471-472 (Edited by Berg, G.)
Interscience Publishers, New York, N.Y.
Berg, G. (1973) Removal of viruses from sewage effluents and
waters. Bull. World Health Organ. 49:451-469.
Berg, G. (1978) The indicator system. In Indicators of
Viruses in Water and Food, pp. 1-14 (Edited by Berg,
G.) Ann Arbor Science, Ann Arbor.
Berg, G. and Metcalf, T.M. (1978) Indicators of viruses in
water. In Indicators of Viruses in Water and Food,
pp. 267-296 (Edited by Berg, G.) Ann Arbor Science, Ann
Arbor.
Berg, G., Chang, S.L., and Harris, E.K. (1964) Dynamics of
the devitalization of enteroviruses by elemental iodine.
Virology 22:469-481.
Berg, G., Dean, R.E., Dahling, D.R. (1968) Removal of polio-
virus 1 from secondary effluent by lime flocculation and
rapid sand filtration. J. Am. Wat. Wks. Ass. 60:193-
198. "~~~
B,erg, G. , Bodily, H.L., Lennette, E.H. , Melnick, J.L. and
Metcalf, T.G. (Editors) (1976) Viruses in Water.
American Public Health Assoc., Washington, D.C., 256
pages.
-------
478
Berg, G.f Dahling, D.R., Brown, G.A. and Berman, D. (1978)
Validity of fecal coliforms, total coliforms, and fecal
streptococci as indicators of viruses in chlorinated
primary sewage effluents. Appl. Environ. Microbiol.
36:880-884.
Berndt, H. and Linneweh, H.J. (1969) Comparative quantita-
tive studies on the bactericidal activity of chlorine
dioxide and chlorine. Arch. Hyg. Bakteriol. 153:41-46
(from Chem. Abst. 71, 29528e, 1969).
Berridge, N.J. (1951) The deterioration of milking rubbers.
I. The effect of microorganisms. J. Dairy Res. 18:
246-250.
Beulow, R.H. and Walton, G. (1971) Bacteriological quality
vs. residual chlorine. Journal of the American Water
Works Assn. 63:28-35.
Bhoonchaisri, P. (1979) An evaluation of differential media
used to enumerate Bifidobacteria. M.Sc. Thesis, Univer-
sity of Newcastle upon Tyne.
Bishop, R.F., Davidson, G.P., Holmes, I.H. and Ruck, B.J.
(1973) Virus particles in epithelial cells of duodenal
mucosa from children with acute nonbacterial gastro-
enteritis. Lancet, (ii), 1281-1283.
Bisson, J.W. and Cabelli, V.J. (1979) Membrane filter method
for Clostridium perfringens. Applied and Environmental
Microbiol. 37:55-56.
Bissonnette, G.K., Jezeski, J.J., McFeters, G.A. and Stuart,
D.G. (1975) Influence of environmental stress on enu-
meration of indicator bacteria from natural waters.
Appl. Microbiol. 29:186-194.
Bissonnette, G.K., Jezeski, J. J., and McFeters, G.A. U^<"/;.
Evaluation of recovery methods to detect coliforms in
water. Appl. Env. Microbiol. 33:590-595.
Black, A.P., Kinman, R.N., Thomas, W.C. Jr., Freund, G., and
Bird, E.D. (1965) Use of iodine for disinfection.
J. Am. Water Works Assoc. 57:1401-1421.
Block, J.C. (1977) Contribution "a 1'etude de la disinfection
des eaux d1alimentation par I1ozone. These Doctorat
d'Etat es-Sciences Pharmaceutiques, Nancy, France.
-------
479
Block, J.C., Morlot, M., and Foliguet, J.M. (1976) Problemes
lies 'a 1'evolution du caractere d'pxydabilite' de cer-
tains corps organiques presents dans 1'eau traitee par
1'ozone. Techno. Sc. Municipales, L'eau 71:29-34.
"Block, J.K., Foliguet, J.M., Hartonann, P., and Joret, J.C.
(1977) Etude comparative de I1action de I1ozone:
Etudes comparatives sur les.enterovirus et les bac-
teries. Unpublished paper presented at the 3rd World
Ozone Congress (Paris, France, 1977), sponsored by the
International Ozone Institute (now Association),
Cleveland, Ohio.
Block, J.C., Joret, J.C., Holland, D., Richard, Y. and
Foliguet, J.M. (1978) Elimination des bacteries et des
virus au cours de la prechloration des eaux d1alimenta-
tion: Resultats preliminaires sur une station indus-
trielle. . Proceedings of the colloque "Eaux d1Alimen-
tation et Sante Publique", Lille, France, in press.
Boardman, G.D. and Sproul, O.J. (1977) Protection of viruses
during disinfection by adsorption to particulate matter.
J. Wat. Poll. Control Fed. 49:1857-1861.
Bohm, A., Dreher, M., Gran, H., Hirmer, R., Lehrnbecher,
M., Metz, H., Moritz, P. and Schmidt, H.W. (1978) On a
dysentery epidemic (Shigella sonnei) in Ismaning.
Off. Gesundh.-Wesen 40:643-65T:~
Bplivar, R. , Conklin, R.H., Vollett, J.J., Pickering, L.K.,
DuPont, H.L., Walters, D.L. and Kohl, S. (1978) Rota-
virus in travelers' diarrhea: Study of an adult stu-
dent population in Mexico. Journal of Infectious Pis.
137:324-327.
Bond, W.W. , Favero, M.S. and Korber, M.R. (1973) Bacillus -, Sp
ATTC 27380: a spore with extreme resistance to dry
heat. Appl. Microbiol. 26:614-616.
Bonde, G.J. (1962) Bacterial Indicators of Water Pollution.
Teknisk Forlag, Copenhagen.
Bonde, G.J. (1977) Bacterial indication of water pollu-
tion. In Advances in Aquatic Microbiology, Vol. 1
(Edited by Droop, M.R. and Jannasch, H.W.) pp. 273-364,
Academic Press, New York.
-------
480
Bonde, G., Grunnet, K. and Kristensen, K.K. (in press)
Examination as to consequences of use of infiltration
ponds.
Boring III, J.R., Martin, W.T. and Elliott, L.M. (1971)
Isolation of Salmonella typhi-murium from municipal
water, Riverside, California, 1965. Amer. J. Epidemiol.
93:49-54.
Bottone, E.J. (1977) Yersinia enterocolitica; A panoramic
view of a charismatic microorganism. Critical Reviews
in Microbiol. 5:211-241.
Bowden, W.B. (1977) Comparison of two direct-count tech-
niques for enumerating aquatic bacteria. Appl. Environ.
Microbiol. 33:1229-1232.
Brammer, K.W. (1963) Chemical modification of viral ribo-
nucleic acid. II. Bromination and iodination. Biochem.
Biophys. Acta 72:217-229.
Brandrick, A.M., Newton, J.M., Henderson, G. and Vickers,
J.A. (1967) An investigation into the interaction
between iodine and bacteria. Appl. Microbiol. 30:484-
487. .
Breidenbach, A.W. (1976) Toward a common language: Keynote
address. In Environmental Modeling and Simulation, pp.
3-6 (Edited by Ott, W.R.) EPA 600/9-76-016, U.S.
Environmental Protection Agency, Washington, D.C.
Brenner, D.J., Steigerwalt, A.G., Falcao, D.P., Weaver,
R.E. and Fanning, G.R. (1976) Characterization of
Yersinia enterocolitica and Y. pseudotuberculosis by
DNA hybridization and by biochemical reactions.
Int. J. Syst. Bacteriol. 23:205-216.
Brezenski, F.T. and Russomanno, R. (1969) The detection and
use of salmonellae in studying polluted tidal estuaries.
J. Water Pollution Control Fed. 41:725-737.
Bringman, G. (1954) Determination of the lethal activity of
chlorine and ozone on E_. coli. Z. Hyg. Infect. 129:333-
337.
Brisou, J. (1975) Les leuvres et les champignons du milieu
marin. Bull. Soc. Fr. Mycol. Med. 4:159-162.
-------
481
British Department of Health and Social Security (1969) The
Bacteriological Examination of Water Supplies, Reports
on Public Health and Medical Subjects, 4th ed., No. 71,
Her Majesty's Stationery Office, London.
British Standards Institute (1955) Rubber Joint Rings for
Gas Mains, Water Mains and Drainage Purposes. British
Standard BS 2494.B.S.I., London.
Broadwater, W.T., Hoehn, R.C., King, P.H. (1973) Sensitivity
of three selected bacterial species to ozone. Appl.
Microbiol. 3:391-393.
Brodsky M.H., and Ciebin, B.W. (1978) Improved method for
recovery and enumeration of Pseudomonas aeruginosa from
water using membrane filters. Appl. Environ. Microbiol.
36:36-42. ~~~~'
Brown, O.R. (1973) Inhibition of Escherichia coli on cellu-
lose acetate membrane filters^MicrobTos 7:235-240.
Brusick, D.J. (1978) Automated plate counting assays in the
Ames Salmonella/microsome mutagenicity test. Pharm.
Technol. 2:37-40 and 65.
Buchanan, R.E. and Gibbons, N.E. (editors) (1974) Sergey's
Manual o£ Determinative Bacteriology. The Williams and
Wilkins Co., Baltimore, 1246 pages.
Bucharov, D.A. and Kulikovskii, A.V. (1971) Structural and
biochemical changes of bacteria after the action of
some chlorine containing preparation, Report 1. Tr.
Vses. Nauchno Issled. Inst. Vet. Sanit. 38:165-1/0.
Buck, J.D. (1975) Distribution of aquatic yeasts. Effect of
incubation temperature and chloramphenicol concentra-
tion on isolation. Mycopathologia 56:73-79.
Buck, J.D. (1978) Comparison of in situ and in vitro sur-
vival of Candida albicans in seawater. Micro. Ecol.
4:291-302. ^
Buck, J.D. and Bubucis, P.M. (1978) Membrane filter pro-
cedure for enumeration of Candida albicans in natural
waters. Appl. Environ. Microbiol. 35:237-242.
Bundesen, H.N., Connolly, J.I., Rawlings, I.D., Gorman,
A.E., McCoy, G.W. and Hardy, A.V. (1936) Epidemic amebic.jn
dysentery. National Inst. Health Bull. No. 166, 187
pages.
-------
482
Burasf N. (1976) Concentration of enteric viruses in waste-
water and effluent: A two year survey. Water Research
10:295-298.
Burke, J.A. (1975) Giardiasis in childhood.
Journal Dis. Child. 129:1304.
American
Burleson, G.R. and Pdllard, M. (1975) Inactivation of .vege-
tative cells and spores of bacteria by ozone and
sonication. Second Intl. Symposium on Ozone Technology
(held in Montreal,Canada 1975),International Ozone
Institute (now Association), pp. 445-454, Cleveland,
Ohio.
Burleson, G.R., Murray, T.M. and Pollard M. (1975) Inacti-
vation of viruses and bacteria by ozone with and
without sonication. Appl. Microbiol. 29:340-344.
Burman, N.P. (1961) Some observations on coli-aerogenes
bacteria and streptococci in water. J. Appl. Bact.
24:368-376.
Burman, N.P. (1962) Bacteriological control of slow sand
filtration. Effl. and Water Treatment J. 2:674-677.
Burman, N.P. (1978) Slow sand filtration. H2O 11:348-351.
Burman, N.P. and Colbourne, J.S. (1976) The effect
of plumbing materials on water quality. J. Inst.
Plumbing 3:12-13.
Burman, N.P. and Colbourne, J.S. (1977) Techniques
for the assessment of growth of microorganisms on
plumbing materials used in contact with potable water
supplies. Journal of Applied Bacteriology 43:137-144.
Burman, N.P. and Colbourne, Jennifer S. (1979) Effect of
nonmetallic materials on water quality. J. Inst. Wat.
Engrs. Sci. 33:11-18.
IT
Burman, N.P., Stevens, J.K. and Evans, A.W. (1978)
Streptococci as indicators in water supplies. In
Streptococci, 335-347 Soc. Appl. Bact. Symposium
Series No. 7, ed. F.A. Skinner and L.B. Quesnel.
Academic Press, London.
Butt, C.G. (1966) Primary amoebic meningoencephalitis.
New England Journal of Medicine 274:1473-1476.
-------
483
Conditions in water distribution cannot always be
counted on to remain routine. Perturbations of the system
may occur through: (1) necessary expansion of the distri-
bution network because of growth of the community; (2) use
of large volumes of water to fight fires; (3) natural or
manmade disasters that disrupt the integrity of the network;
and (4) errors by users, beyond the direct control of a
water authority, that result in back-contamination of the
water in public distribution. If appropriate designs and
materials have been used in constructing a distribution
system, water quality can be protected, in most instances,
by properly organized maintenance and surveillance. How-
ever, regulations of users connected to the system, as well
as the development of effective plans for dealing with
emergencies, are important further aspects of operation.
Recommendations Concerning Storage and Distribution
1. Careful consideration must be given to the
siting of service reservoirs.
2. Dead ends in pipes must be avoided and disused
apparati disconnected.
3. Distribution and plumbing systems -should consist
of materials that will not support microbial
growth. New products should be tested for
their ability to support microbial growth
before they are accepted or rejected.
4. Regulations to prevent back-siphonage and
cross-connections need enforcement.
5. Adequate disinfection procedures for the
construction and repair of water mains are
needed. Installers should be instructed to
follow the installation codes, exactly. .:
Topic G discusses Technologic Problems in drinking
water microbiology. A pervasive theme in drinking water-
microbiology is the avoidance, suppression, or destruction
of microorganisms in water. As Topic G shows, some micro-
organisms have what might be regarded as a certain retal-
iatory capacity. Microorganisms have adapted to such a
variety of aquatic environments outside of water systems
that it is probably not surprising to find them so firmly
entrenched in much of this manmade system as well. They may
cause problems both in treatment and in distribution. .Water
-------
484
Cady, P., Dufour, S.W., Shaw, J., and Kraeger, S.J. (1978)
Electrical impedance measurements: Rapid method for
detecting and monitoring microorganisms. J. Clin.
Microbiol. 7:265-272.
Cairns, J . Jr. (1977) Quantification of biological in-
tegrity. In The Integrity of Water (Edited by Ballen-
tine, R.K., and Guarrain, L.J.) U.S. EPA, Office of t
Water and Hazardous Materials, Washington, D.C., pp.
171-187.
Cairns, J.E., Nutt, S.G., and Afghan,w B.K. (1979) Applicationt
the ATP assay for monitoring the disinfection of waste-
water streams. Proceedings of the International
Symposium on Analytical Applications of Bioluminescence
and Chemiluminescence. Brussels, 1978 (in press).
Calabra, J.F., Cosenza, B.J. and Kolega, J.J. (1972) Recovery
of gram-negative bacteria with hektoen agar. J. Water
Poll. Cont. Fed. 44:491-493.
Calkins, J., Buckles, J.D. and Moeller, J.R. (1976) The
role of solar ultraviolet radiation in "natural" water
purification. Photochem. Photobiol. 24:49-57.
Callaghan, P. and Brodie, J. (1968) Laboratory investi-
gation of sewer swabs following the Aberdeen typhoid
outbreak of 1964. J. Hyg. 66:489-497.
Carlson, S., Hasselbarth, U. and Langer, R. (1975) Abtotung
aggregierter Keime bei der Wasserdesinfektion durch
Chlor. Zbl. Bakt. Hyg., i Abt. Prig. 6161:233-247.
Carmichael, W.W. and Gorham, P.R. (1978) Anatoxins from
clones of Anabaena flos-aquae isolated from lakes of
Western Canada. Mitt. Internatl. Verein. Limnol.
21:285-295.
Carrington, E.G. (in press) The fate of pathogenic micro-
organisms during wastewater treatment and disposal.
Carson, L.A., Favero, M.S., Bond, W.W. and Petersen, N.J.
(1972) Factors affecting comparative resistance of
naturally occurring and subcultured Pseudomgnas aeru-
ginosa to disinfectants. Appl. Microbiol. 23:863-869.
-------
485
Carson, L.A., Petersen, N.J., Favero, M.S., Doto, L. ,
Collins, P.E. and Levin, M.A. (1975) Factors influ-
encing detection and enumeration of P. aeruginosa by
most probable number .and membrane filtration techniques.
Applied Microbiol. 30:935-942.
Caselitz, F.H. (1966) Pseudomonas-Aeromonas und ihre human-
medizinische Bedeutung. Jena, Fisher Verlag.
Cash, R.A., Music, S.I. and Libonati, J.P. (1974) Response
of man to infection with' Vibrio cholerae. I. Clinical
and serologic and bacteriologic responses to a known
inoculum. Journal Infect. Pis. 129:45-52.
Center for Disease Control (1972) Morbidity and Mortality
Weekly Rept. 21:290.
Center for Disease Control (1975a) Yersinia enterocolitica
infections Georgia, Utah. Morbidity and Mortality
Weekly Report 24:147-148.
Center for Disease Control (1975b) Viral hepatitis reporting,
Morbidity and Mortality Weekly Report 24(19):165-166.
Center for Disease Control (October 1, 1976a) Endemic
giardiasis Northern Canada. Morbidity and Mortality
Weekly" Report 25:307.
Center for Disease Control (1967b) Waterborne disease
outbreaks, 1975. Foodborne and Waterborne Disease
Outbreaks Annual Summary 1975, (CDC) 76-8185, pp. 55-
58, U.S. Department of Health, Education and Welfare,
Atlanta, Georgia.
Center for Disease Control (1977a) Reported morbidity and
mortality in the United States, 1976. Morbidity and
Mortality Weekly Report Annual Summary 1976 25(53):2.
Center for Disease Control (1977b) Waterborne disease
outbreaks, 1976. Foodborne and Waterborne Disease
Outbreaks Annual Summary 1976, (CDC) 78-8185, pp. 59-
64, U.S. Department of He,alth, Education and Welfare,
Atlanta, Georgia.
Center for Disease Control (May 27, 1977c) Waterborne
giardiasis-outbreaks ~ Washington, New Hampshire.
Morbidity and.Mortality Weekly Report 26:169 and 175.
-------
486
Center for Disease Control (May 19, 1978a) Intestinal
parasite surveillance United States, 1976.
Morbidity and Mortality Weekly Report 27:167-168.
Center for Disease Control (June 23, 1978b) Waterborne
Campylobacter gastroenteritis Vermont, Morbidity
and Mortality Weekly Report 27:207.
Center for Disease Control (October 13, 197-8c) Vibrio
cholerae infection Louisiana. Morbidity and Mor-
tality Weekly Report 27:341-342 and 402.
Center for Disease Control (1978d) Viral gastroenteritis
Pennsylvania. Morbidity and Mortality Weekly Report
27(42):403-404. ,
Center for Disease Control (November 10, 1978e) Legion-
naires' Disease United States. Morbidity and
Mortality Weekly Report 27:439-441.
Cerkinsky,_ S.N., and Trahtman, N. (1972) The present status of
research on the disinfection of drinking water in the
U.S.S.R. Bull. World Health Organ. 46:277-283.
Cerva, L. and Novak, K. (1968) Amoebic meningoencephalitis:
Sixteen fatalities. Science 160:92.
Chambers, C.W. , Kabler, P.W., Malaney, G. , and Bryant, A.
(1952) Iodine as a bactericide. Soap and Sanit.
Chem. 28:149-163.
Chang, S.L. (1948) Some epidemiological and'biological
problems in waterborne amoebiasis. In Symposium on
Limnol. Aspects of Water Supply and Waste Disposal,
American Acad.for the Advancement of Sci. Proc., pp.
16-30.
Chang, S.L. (1950) Kinetics in the thermodestruction of
cysts of Endamoeba histolytica in water. American
Journal of Hygiene 52:82-90.
Chang, S.L. (1958) The use of active iodine as a water
disinfectant. J. Am. Pharm. Assn. Sci. Ed. 48:417-423.
Chang, S.L. (1968) Waterborne viral infections and their
prevention. Bull. World Health Organ. 38:401-414.
-------
487
Chang, S.L. .(1971a) Current Topics in Comparative Pathg-
biology, Vol. 1, pp.201-254 (Edited by Cheng, T.C.)
Academic Press, New York and London.
Chang, S.L. (1971b) Modern Concept of disinfection. J.
Sanit. Eng. Div., Am. Soc. Civ. Eng. SA5:689-707.
Chang, S.L. (1978) Pathogenesis of pathogenic Naegleria
amoeba. In Folia Parasitologica (Fourth International
Congress of Parasitology, August 19-26, 1978) Warsaw,
Poland.
Chang, S.L., and Morris, J.C. (1953) Elemental iodine as a
disinfectant for drinking water. Ind. and Eng. Chem.
45:1009-1012.
Chang, S.L., Buckingham, M. and Taylor, M.P. (1948) Studies
on Leptospira icterohaemorrhagiae; survival in water
and sewage;' destruction in water by halogen compounds,
synthetic detergents and heat. Journal of Infectious
Pis. 82:256-266.
Chantereau, J. (1977) Corrosion bacterienne, Bacteries
de la Corrosion. Technique et Documentation, Paris,
246 pages.
Cherry, W.B., Hanks, J.B., Thomason, B.M., Murlin, A.M.,
Biddle, J.W. and Groom, J.M. (1972) Salmonellae as an
index of pollution of surface water. Appl. Microbiol.
24:334-340.
Chin, T.P.Y., Mosley, W.H., Robinson, A. and Gravelle, C.R.
(1967) Petection of enteric viruses in sewage and
water. Relative sensitivity of the method. In Trans-
mission of Viruses by the Water Route, pp. 389-4TTO
(Edited by Berg, G.) John Wiley and Sons, New York.
Christensen, F. and Giese A.C. (1954) Changes in absorption
spectra of nucleic acids and their derivatives fol-
lowing exposure to ozone and ultra violet radiation.
Arch. Biochem. Biophys. 51:208-216.
Clark, H.F., Geldreich, E.E., Jeter, H.L. and Kabler, P.W.
(1951) The membrane filter in sanitary bacteriology.
Public Health Reports 66:951-977, Washington, D.C.
Clark, J.A. (1969) The detection of various bacteria indica-
tive of water pollution by a presence-absence (P-A)
procedure. Can. J. Microbiol. 15:771-780.
-------
488
Clark, J.A. and Vlassoff, L.T. (1973) Relationships among
pollution indicator bacteria isolated from raw water
and distribution systems by the presence-absence (P-A)
test. Health Lab. Sci. 10:163-172.
Clarke, N.A. and Chang, S.L. (1959) Enteric viruses in
water. J. Am. Water Wks. Ass. 51:1299-1317.
Clarke, N.A., Stevenson, R.E. and Kabler, P.W. (1956) The
inactivation of purified type 3 adenovirus in water by
chlorine. Am. J. Hyg. 64:314-319.
Clarke, N.A., Berg, G. , Kabler, P.W. and Chang, S.L. (1964)
Human enteric viruses in water: Source, survival and
removability. Adv. Water Pollut. Res. 2:523-541.
Claudon, D.G., Thompson, D.I., Christenson, E.H., Lawton,
G.W. and Dick, E.G. (1971) Prolonged Salmonella con-
tamination of a recreational lake by runoff waters.
Appl. Microbiol. 21:875-877.
Clausen, E.M., Green, B.L. and Litsky, W, (1977) Fecal
streptococci: Indicators of pollution. In Bacterial
Indicators/Health Hazards Associated with Water ASTM
STP 635:247-264 (Edited by Hoadley, A.W. and Dutka,
B.J.) American Society for Testing and Materials,
Philadelphia.
Cleary, E.J. and Warner, D.L. (1970) Some considerations in
underground wastewater disposal. Journal of the
American Water Works Assoc. 62:489-498.
Cohen, B. (1922) The effects of temperature and hydrogen
ion concentration upon the viability of Bact. coli and
Bact. typhosum in water. J. Bact., 7(2):183-230.
Coin, L., Hannoun, C. and Cornelia, C. (1964) Inactivation
par I1ozone du virus de la poliomyelite present dans
les eaux. La Presse Medicale 72:2153-2156.
Coin, L., Gomella, C., Hannoun, C. and Trimoreau, J.C.
(1967) Inactivation par 1'ozone du virus de la polio-
myelite present dans les eaux. La Presse Medicale
38:1883-1884. ~
Colbourne, J -S. s. and Brown, D.A. (1979) Dissolved
oxygen utilization as an indicator of total microbial
activity on non-metallic materials in contact with
potable water. Journal of Applied Bacteriology 47:223-
231.
-------
489
Colwell, R.R. and Kaper, J. (1977) Vibrio species as bac-
terial indicators of potential health hazards associated
with water. In Bacterial Indicators/Health Hazards
Associated with Water, pp. 115-25 (Edited by Hoadley,
A.W. and Dutka, B.J.) STP 635 American Society for
Testing and Materials, Philadelphia.
Colwell, R.R., Kaper, J. and Joseph, S.W. (1977) Vibrio
cholerae, Vibrio parahaemolyticus, and other vibrios:
Occurrence and distribution in Chesapeake Bay. Science
198:394-396.
Colwell, R.R., Austin, B. and Wan, L. (1978) Public health
considerations of the microbiology of "potable" water.
In Evaluation of the Microbiology Standards for Drink-
ing Water (Edited by Hendricks, C.W.) EPA 570/9-78-00
C, Environmental Protection Agency, Washington, D.C.
Commission of the European Communities (1975) Council direc-
tive of May 16, 1975 concerning the quality of sur-
face water intended for the abstraction of drinking
water in the member states. Official Journal of the
European Communities 194:26-31. ^
Commission of the European Communities (1977) Bacterio-
logical analyses of drinking water, Lyons Technical
Seminar 16 to 20 June 1975. Environment and Quality
of Life, EUR 5694 e, Office for Official Publications
of the European Communities, Luxembourg, 106 pages.
Commission of the European Communities (1978) Proposal to a
council directive concerning measuring methods and
frequency for sampling and analysis of parameters as
to quality of surface water intended for abstraction of
drinking water. Official Journal of the European
Communities C-208, 1-11.
Commission of the European Communities (1979) Council direc-
tive of December 17, 1979 concerning protection of
groundwater as to pollution caused by special dangerous
chemical compounds. Official Journal of the European
Communities.
Committee on Back-Siphonage in Water Installations (1974)
Report of the Committee on Back-Siphonage in Water
Installations. H.M. Stationery Office, London.
-------
490
Cooney, M.K., Hall, C.E. and Fox, J.P. (1972) The Seattle
virus watch. III. Evaluation of isolation methods and
summary of infections detected by virus isolations.
Amer. Journal of Epidemiol. 96:286-305.
Craig, C.E. and Faust, E.G. (1964) Clinical Parasitology, 7th
Edition, Lea and Febiger, pp. 90-94, Philadelphia.
Cramer, W.M., Kawata, K., and Kruse, C.W. (1976) Chlorina-
tion and iodination of poliovirus and f2. J. Water
Poll. Control Fed. 48:61-76.
Craun, G.F. (1976) Microbiology-Waterborne outbreaks.
Water Pol. Control Fed. 48:1378-1397.
J.
Craun, G.F. (June 1977) Waterborne outbreaks. Journal of
the Water Pollution Control Fed., Literature Review,
pp. 1268-1279.~~~~~~
Craun, G.F. (1978) Impact of the coliform standard on the
transmission of disease. In Evaluation of the Micro-
biology Standards for Drinking Water, pp. 21-35 (Edited
by Hendricks, C.W.) EPA 570/9-78-OOC, U.S. Environ-
mental Protection Agency, Washington, D.C.
Craun, G.F. (1979) Waterborne disease outbreaks in the
United States. J. Env. Hlth. 41:259-265.
Craun, G.F. and McCabe, L.J. (1973) Review of the causes of
Waterborne disease outbreaks. J. Am. Water Works
Assoc. 65:74-83.
Craun, G.F., McCabe, L.J. and Hughes, J.M. (1976) Waterborne
disease outbreaks in the U.S. 1971-1974. J. Am. Water
Works Assoc. 68:420-424.
Cronier, S., Scarpino, P.V., and Zink, M.L. (1978) Chlorine
dioxide destruction of viruses and bacteria in water.
In Water Chlorination Environmental Impact and Health
Effects, pp. 651-657, vol. 2 (Edited by Jolly, R.L.,
Gorchev, H., and Hamilton, D.H., Jr.) Ann Arbor Sci.,
Ann Arbor, Mich.
Cullimore, D.R. and McCann, A.E. (1977) The identification,
cultivation and control of iron bacteria in ground-
water. In Aquatic Microbiology (Edited by Skinner,
F.A. and Shewan, J.M.) Academic Press, London, New
York, San Franciso, 369 pages.
-------
491
Cundell, A.M. and Mulcock, A.P. (1973) Microbiological
deterioration of natural rubber pipe-joint rings.
Mater und Organismen 8:165-175.
Cvjetanovic, B., Grab, B. and Uemura, K. (1978) Dynamics of
acute bacterial diseases. Epidemiological models and
their application in public health. Bull. World Health
Organ. 56 Suppl. no. 1, 143 pages.
Dahi, E. (1977) Impact of ultrasonic disaggregation of
microorganisms in ozone water sterilization methods.
Unpublished paper presented at the 3rd World Ozone
Congress (Paris, France, 1977), sponsored by the Inter-
national Ozone Institute (now Association), Cleveland,
Ohio.
Daley, R. (in press) Direct epifluorescence enumeration of
native aquatic bacteria: Uses, limitations and com-
parative accuracy. In J.W. Costerton and R. Colwell
(eds.) Native Aquatic Bacteria Enumeration, Activity
and Ecology, AST Publication.
Daley, R.J. and Hobbie, J.E. (1975) Direct counts of aquatic
bacteria by a modified epifluorescence technique.
Limnol. and Oceanog. 20:875-882.
Danielsson> D., Laurell, G., Nordbring, F. and Sandier, O.
(1968) A membrane filter method for the demonstration
of bacteria by the fluorescent antibody technique.
Acta Path, et Microbial. Scandinav. 72:118-124.
Daubner, I. (1972) Drittes kapitel. In Mikrobiologie des
Wassers, pp. 53-104, Inst. f. Limnologie, Slowakischen
Academie der Wissenschaften, Bratislava, 440 pages.
Daubner, I. (1975) Changes in the properties of V. coli
under the influences of water environments. Verh.
Int. Ver. Limnol. 19:2650-2657.
Davenport, C.V., Sparrow, E.B. and Gordon, R.C. (1976) Fecal
indicator bacteria persistence under natural conditions
in an ice-covered river. Appl. Environ. Microbiol.
32:527-536.
Davis, E.M. and Gloyna, E.F. (1972) Bacterial die-off in
ponds. J. Sanit. Eng. Div. Am. Soc. Civ. Engrs. 98:59-
69. ~ r~" '
-------
492
Dawe, L.L., and Penrose, W.R. (1978) "Bactericidal" property
of seawater: Death or debilitation? Appl. Environ.
Microbiol. 35:829-833.
DeJonckheere, J. and van de Voorde, H. (1976) Difference in
destruction of cysts of pathogenic and nonpathogenic
Naegleria and Acanthamoeba by chlorine. Applied and
Environmental Microbiol. 31:294-297.
Deland, F.H., and Wagner, H.N. Jr. (1969) Early detection of
bacterial growth with carbon-14-labelled glucose.
Radiology 92:154-155.
De Man, J.C.f Rogosa, M. and Sharps, M.E. (1960) A medium for
the cultivation of Lactobacilli. J. Appl. Bact. 23
De Mol, P., and Bosmans, E. (1978) Campylobacter enteritis in
Central Africa. Lancet 1:604.
Dennis, W.H. Jr., Olivieri, V.P. and Kruse, C.W. (1979)
Mechanism of disinfection: Incorporation of Cl-36
into f~ virus. Water Research 13:363-369.
Derham, C.J. (1975) Physical properties of pipe sealing
rings a comparison between natural and synthetic
rubbers. Natural Rubber Technol. 6:41-51.
Diaper, E.W.J. (1972) Practical aspects of water and waste-
water treatment by ozone. In Ozone in Water and Waste-
water Treatment (Edited by Evans, F.L.) Environ.
Protection Agency, Cincinnati.
Dickenson, P.B. (1965) Natural rubber and its use in under-
ground pipe sealing rings. Rubber Dev. 18:85-90.
Dickenson, P.B. (1969) Natural rubber and its traditional
use in underground pipe sealing rings. J. Rubber Res.
Inst., Malaya 22:165-175.
Dillenberg, H.W. and Dehnel, M.K. (1960) Toxic waterbloom
in Saskatchewan. Canadian Medical Assoc. Journal
83:1151-1154.
DiLuzio, N.R. and Friedmann, T.J. (1973) Bacterial endo-
toxins in the environment. Nature 244:49-51.
-------
493
Dominowska, C. and Malottke, R. (1971) Survival of Yersinia
in water samples originating from various sources.
Bull. Inst. Mar. Med. Gdansk 22:173-182.
Donnelly, C.B.f Gilchrist, J.E., Peeler, J.T. and Campbell,
J.E. (1976) Spiral plate count method for the exami-
nation of raw and pasteurized milk. Appl. Environ.
Microbiol. 32:21-27.
Dornedden, H. (1930) Beitrage zur Trinkwassereinigung durch
die langsame sandfiltration. G.W.F. 73:319-325.
Dougan, J. and Tan, L. (1973) Detection and quantitative
measurement of fecal water pollution using a solid-
injection gas chromatographic technique and fecal
steroids as a chemical index. J. of Chromatogr.
86:107-116.
Drake, C.H. (1966) Evaluation of culture media for the
isolation and enumeration of Pseudomonas aeruginosa.
Health Lab. Sci. 3(1):10-19.
Drake, J.F. and Tsuchiya, H.M. (1976) Predation on Esche-
richia coli by Colpoda steinii. Appl. Environ.
Microbiol.31s 870-874.
Drasar, B.S. (1974) Some factors associated with geographi-
cal variations in the intestinal microflora. In
Society for Applied Bacteriology Symposium Series No.
3. The Normal Microbial Flora of Man, pp. 187-196
(Edited by Skinner, F.A. and Carr, J.C.) Academic
Press, London.
Druce, R.G. and Thomas, S.B. (1970) An ecological study"'
of the psychrotrophic bacteria of soil, water,
grass, and hay. Journal of Applied Bacteriology
33:420-435.
Dudly, R.H., Hekimian, K.K. and Mechalas, B.J. (1976) A
scientific basis for determining recreational water
quality criteria. J. Water Poll. Control Fed. 12:
2761-2777.
Dufour, A.P. (1977) Escherichia coli; The fecal coliform.
In Bacterial Indicators/Health Hazards Associated with
Water, ASTM STP 635 (Edited by Hoadley, A.W. and Dutka,
B.J.) American Society for Testing and Materials,
Philadelphia, Pennsylvania.
-------
494
Dufour, A.P. and Cabelli, V.J. (1976) Characteristics of
Klebsiella from textile finishing plant effluents.
J. Water Pollution Control Fed. 48:872-879.
Duncan, D.W. and Razzell, W.E. (1972) Klebsiella biotypes
among coliforms isolated from forest environments and
farm produce. Appl. Microbiol. 24:933-938.
Dunn, C.C. (1952) Mechanism of action of some germicidal
inhibitory agents. Am. Brewing 85:25-30.
Dupont, H.L., Hornick, R.B. and Pankins, A.T. (1969) The
response of man to virulent Shigella flexneri 2A.
J. Infect. Pis. 119:296-299.
Dupont, H.L., Hornick, R.B., Snyder, M.J., Libonati, J.P.
and Formal, S.B. (1972) Immunity in shigellosis. II.
Protection induced by oral live vaccine or primary
infection. J. Infect Pis. 125:12-16.
Dutka, B.J. (1973) Coliforms are an Inadequate Index of
Water Quality. J. Environ. Health 36:39-46.
Dutka, B.J., Chau, A.S.Y. and Coburn, J. (1974) Relationship
between bacterial indicators of water pollution and
fecal sterols. Water Res. 8:1047-1055.
Dykes, A.C., Schultz, M.G. and Juranekm, D.D. (1977) Out-
break of suspected waterborne giardiasis in the state of
Washington. Public Health Service-CPC-Phoenix, EPI 76-
82-3.
Dykes, D.R., Bry, T.S., and Kline, C.H. (1967) Water Manage-
ment. Env. Science & Tech. 1:780-784.
Eijkman, C. (1904) Pie Garungsprobe bie 46 als Hilfmittel
bei der Trinkwasseruntersuchung. Zentralblatt fur
Bakteriologie, Abth. I. Prig. 37:742.
Engelbrecht, R.S.. and Greening, E.G. (1978) Chlorine-
resistant indicators, 243-266. In Indicators of
Viruses in Water (Edited by Berg, G.) Ann Arbor Science
Publisher,Inc., Ann Arbor, Mich., 424 pages.
Engelbrecht, R.S., Severin, B.F., Masarik, M.T., Farooq,
S., Lee, S.H., Haas, C.N. and Lalchandani, A. (1977)
New microbial indicators of disinfection efficiency.
EPA-600/2-77-052, U.S. Environmental Protection Agency,
Cincinnati.
-------
495
Engelbrecht, R.S., Weber, M.J., Schmidt, C.A. and Salter,
B.L. (1978) Virus sensitivity to chlorine disinfec-
tion of water supplies. Final Report, EPA-600/
2-78-123, 43 pages.
Environmental Protection Agency (Wed., Dec. 24, 1975) Water
Programs National Interim Primary Drinking Water
Regulations, Part IV, Federal Register 40(248):59566-
59588. - .
Environmental Protection Agency (1978) Human Viruses in the
Aquatic Environment; A Status Report With Emphasis
on the EPA Research Program, EPA 570/9-78-006, Environ-
mental Research Center, Cincinnati, Ohio.
Esposito, P., Scarpino, P.V., Chang, S.L. and Berg, G.
(1974) Destruction by dichloramine of viruses and
bacteria in water. Abstract G-99. Abstracts of the
annual meeting, Am. Soc. for Microbiol., Washington,
D.C.
Evans, D.M. and Levisohn, I* (1968) Biodeterioration
of polyester-based polyurethanes. Int. Biodeterior.
Bull. 4:89-92.
Evans, T.M., Shillinger, J.E., and Stuart, D.G. (1978)
Rapid determination of bacteriological water quality by
using Limulus lysate. Appl. Environ. Microbiol. 35:376-
382.
Evison, L. (1977) Disinfection of water with ozone: Com-
parative studies with enteroviruses, phage and bacteria,
Unpublished paper presented at the 3rd World Ozone
Congress (Paris, France, 1977), sponsored by the Inter-
national Ozone Institute (now Association), Cleveland,
Ohio.
Evison, L.M. and James, A. (1975) Bifidobacterium as an
indicator of faecal pollution in water. Prog. Wat.
Tech. 7(2);57-66.
Evison, L.M. and Morgan, S. (1978) Further studies on
Bifidobacteria as indicators of faecal pollution in
water. Prog. Wat. Tech. 10(5/6);346-350.
Eyssen, H.J., Parmentier, G.G., Compernolle, F.C., DePauw,
G. and Pressens-Denef, M. (1973) Biohydrogenation of
sterols by Eubacterium ATCC 21408 Nova Species.
Env. J. Biochem. 36:411-421.
-------
496
Fair, G.M., Geyer, J.C., and Okun, D.A. (1968) Water and
Wastewater Engineering, Vol. 2, pp. 31/1-31/29, John
Wiley and Sons Publishers, New York.
Fair, J.P. and Morrison, S.M. (1967) Recovery of bacterial
pathogens from high quality surface waters. Water
Resour. 3:799-803.
Parooq, S. (1976) Kinetics of inactivation of yeasts and
acid-fast organisms with ozone. Thesis of doctorate
University of Illinois at Urbana-Champaign (Report
Number 76-24,077).
Farooq, S., Engelbrecht, R.S. and Chian, E.S.K. (1977) The
effect of ozone bubbles on disinfection. Prog. Water
Technol. 9:157-166. ~
Parooq, S., Chian, E.S.K. and Engelbrecht, R.S. (1977a)
Basic concepts in disinfection with ozone. J. Water
Pollution Control Fed. 49:1818-1831.
Farooq, S., Engelbrecht, R.S. and Chian, E.S.K. (1977b)
Influence of temperature and UV light on disinfection
with ozone. Water Res. 11:737-741.
Fattal, B. and Nishmi, M. (1977) Enterovirus types in
Israel sewage. Water Research 11:393-396.
Feachem, R.G. (1977) Infectious disease related to water
supply and excreta disposal facilities. Ambio. 6:55-
58.
Feachem, R.G., Bradley, D.G., Garelick, H., and Mara, D.D. (1978). Health
aspects of Excreta and Wastewater Management. The World Bank
Washington, B.C. '
Federal "jjistry of Health (1969) Evaluation and treatment
of accidents with mineral oil in the country under the
aspects of water protection. Water and Mineral Oil
Work Group, Federal Ministry of Health, Bad Godesberg,
1J8 pages. y'
PO^V*. .p-lu",,", interior (1975) Bank filtration.
Report of the Federal Ministry of the Interior. Com-
S1* M°? Water SuPPlv and Bar* Filtrate. Committee of
Water Maintenance and Bank Filtration, Bonn, 182 pages.
Felsenfeld, O. (1967) The Cholera Problem. Warren H. Green,
Inc., St. Louis, Missouri, 165 pages.
Feng, T.H. (1966) Behavior of organic chloramines in dis-
infection. J. Water Poll. Control Fed. 38:614-628.
-------
497
Fennell, H., James, D.B. and Morris, J. (1974) Pollution of
a storage reservoir by roosting gulls. J. Soc. Wat.
Treat. Exam. 23(1):5-24.
Fetner, R.H. (1962) Chromosome breakage in Vicia faba by
monochloramine* Nature 196:1122-1123.
Fetner, R.H. and Ingols, R.S. (1956) A comparison of the
bactericidal activity of ozone and chlorine against
Escherichia coli at 1°C. J. Gen. Microbiol. 15:381-
385.
Finkelstein, R.A. (1973) Cholera. CRC Critical Reviews
in Microbiol.. 2:553-623.
Fjerdingstad, E. (1956) Bacteriological investigations of
mine water from lignite pits in Denmark. Schweitz* Z.
Hydrobiol. 18:215-238.
Fjerdingstad, E. (1969) Bacteriological corrosion of con-
crete in water. Water Research 3:21-30.
Flewett, T.H., Bryden, A.S. and Davies, H. (1975) Epidemic
viral enteritis in a long-stay children's ward. Lancet
Fliermans, C.B. and Schmidt, E.L. (1975) Fluorescence micro-
scopy: Direct detection, enumeration and spatial
distribution of bacteria in aquatic systems. Arch.
Hydrobiol. 76:33-42.
Floyd, R. and Sharp, D.G. (1977) Aggregation of polio- and
reovirus by dilution in water. Appl. Environ. Microbiol.
33:159-167.
Floyd, R. and Sharp, D.G. (1978) Viral aggregation: Effects
of salt on the aggregation of poliovirus and reovirus
at low pH. Appl. Environ. Microbiol. 35:1084-1094.
Fluegge, R.A.f Metcalf, T.G. and Wallis, C. (1979) Virus
inactivation in wastewater effluents by chlorine, ozone
and ultraviolet light. Prog, in Wastewater Disinfec-
tion Technology, U.S. Environmental Protect. Agency,
Office of Research and Development, EPA 600/9-79-018,
pp. 223-232.
-------
498
Foliguet, J.M., Lavillaureix, P. and Schwartzbrod, L. (1973)
Virus et eaux II. Mise en evidence des virus dans le
milieu hydrique. Rev. Epidem. Med. Soc. et Sante
Publigue 21:185-259.
Fowler, M.C. and Carter, R.F. (1965) Acute pyogenic menin-
gitis probably due to Acanthamoeba species: A pre-
liminary report. British Medical Journal 2:740-742.
Fruin, J.T. and Clark, W.S. (1977) Plate count accuracy:
Analysts and automatic colony counter versus a true
count. J. Food Prot. ,40:552-554.
Fuchs, G.H.P. and Burger, G. (1974) Standardisierungs-
probleme bei serodiagnostischen Leptospirose-
Untersuchungen. Zbl. Bakt. Hyg., I. Abt. Prig. A.
228:378-387.
Gall, R. (1978) Chlorine dioxide. An overview of its prepa-
ration, properties and uses. In Ozone/Chlorine Dioxide
Oxidation Products of Organic Materials pp. 356-382
(Edited by Rice, R.G. and Cotruvo, J.A.) International
Ozone Institute, Inc., Cleveland, Ohio.
Gallagher, T.P. (1965) Public Health Service Report on
pollution of interstate waters of the Red River of the
North (Minnesota, North Dakota), Robert A. Taft Sani-
tary Eng. Ctr., Cincinnati, Ohio, 175 pages.
Gameson, A.L.H. and Saxon, J.R. (1967) Field studies on
effect of daylight on mortality of coliform bacteria.
Water Res. 1, pp. 279-295.
Geldreich, E.E. (1970) Applying bacteriological parameters
to recreational water quality. Journal of the American
Water Works Association 62:113-120.
Geldreich, E.E. (1972) Waterborne pathogens. In Water
Pollution Microbiology, pp. 207-241 (Edited by Mitchell,
R.)John Wiley and Sons, New York.
Geldreich, E.E. (1975) Handbook for Evaluating Water Bac-
teriological Laboratories pp. 1-195.EPA-670/9-75-006,
U.S. Environmental Protection Agency, Cincinnati, Ohio.
Geldreich, E.E. (1978) Bacterial populations and indicator
concepts in feces, sewage, stormwater and solid wastes.
In Indicators of Viruses in Water and Food, pp. 51-98
(Edited by Berg, G.) Ann Arbor Science, Ann Arbor.
-------
Geldreich, E.E. and Clarke, N.A. (1966) Bacterial pollution
indicators in the intestinal tract of freshwater fish.
Appl. Microbiol. 14:429-437.
Geldreich, E.E. and Kenner, B.A. (1969) Concepts of fecal
streptococci in stream pollution. Journal of the
Water Pollution Control Federation 41:R336-352.
Geldreich, E.E., Clark, H.F., Kabler, P.W., Huff, C.B. and
Bordner, R.H. (1958) The coliform group. II. Reac-
tion in EC medium at 45°C. Applied Microbiology 6:347-
348. , ~ ~
Geldreich, E.E., Kenner, B.A. and Kabler, P.W. (1964) Occur-
rence of coliforms, fecal coliforms, and streptococci
on vegetation and insects. Appl. Microbiol. 12:63-69.
Geldreich, E.E., Nash, N.D., Reasoner, D.J. and Taylor,
R.H. (1972) The necessity for controlling bacterial
populations in potable water: Community water supply.
J. American Water Works Assoc. 64:596-602.
Geldreich, E.E., Nash, H.D., Reasoner, D.J. and Taylor,
R.H. (1975) The necessity of controlling bacterial
populations in potable waters: Bottle water and emer-
gency water supplies. J. Am. Water Works Ass. 67:117-
124. ~
Gelfand, H.M., Fox, J.P. and Leblanc, D.R. (1957) The en-
teric viral flora of a population of normal children in
Southern Louisiana. Amer. Journal of Trop. Med.
Hygiene 6:521.
Gellman, I. (1975) Further studies of the sanitary signifi-
cance of Klebseilla pneumoniae occurrence in mill
effluents and surface waters. Tech. Bull. 279.
National Council of the Paper Industry for Air and
Stream Improvement, Inc., New York.
Gentile, J.H. and Maloney, T.F. (1969) Toxicity and environ-
mental requirements of a strain of Aphanizomenon
flos-aquae (L.) Ralfs. Canadian Journal of Microbiology
15:165-173.
499
Gerba, C.P., Wallis, C. and Melnick, J.L. (1975a) Viruses
,in water: The problem, some solutions. Environmental
Science and Technol. 9:1122-1126.
-------
500
Gerba, C.P., Wallis, C. and Melnick, J.L. (1975b) Fate of
wastewater bacteria and viruses in soil. Journal of
the Irrigation and Drainage Div., Proceedings of the
American Society of Civil Engineers, 101, IR 3:145-155.
Gevaudan, P., Bossy, G. , Gulian, C. and Sanchez, Y. (1971)
Action de I1ozone sur le virus de la poliomyelite
present dans 1'eau. Terres et Eaux 67:25-28.
Ghan, H.B., Chen, C.L. and Miele, R.P. (1976) The signif-
icance of water quality on wastewater disinfection
with ozone. Forum on Ozone Disinfection, Proc. of a
forum held in Chicago, 111., 1976 (Edited by Fichtman,
E.G., Rice, R.G. and Browning, M.E.) International
Ozone Institute (now Association), Cleveland, Ohio, pp.
46-65.
Gilchrist, J.E., Campbell, J.E., Donnelly, C.B., Peeler,
J.T. and Delaney, J.M. (1973) Spiral plate method for
bacterial determination. Appl. Microbiol. 25:244-255.
ft
Gilcreas, F.W. and Kelly, S.M. (1955) Relation of coliform-
organism test to enteric-virus pollution, J. Am. Water
Works Assoc. 47:683-694.
Gilfillan, D.J. (1971) Report on the investigation of
back-siphonage risks in domestic properties, Winter
1970/71. Report T.P. 82, Water Research Association,
Medmenham, England.
Glathe, H. and Ottow, J.C.G. (1972) Okologische ,und phy-
siologische Aspekte zum Mechanismus der Eisenoxydation
und Ockerbildung Eine Ubersicht. Zbl. Bakt. II.
Abt. 127:749-769.
, »
Godlewska-Lipowa, W.A. and Jablonska, I. (1972) Spatial
differentiation abundance of bacteria in the water of
Mikolajskie lakes. Ekologia Polska 20:367-371.
Goethe, H., Zorn, E. and Muller, G. (1969) Hygienische
Probleme der Trinkwasserversorgung an Bord von See-
schiffen unter Besonderer Berucksichtigung von Vakuum-
verdampferanlagen und Ultraviolettwasserentkeimung.
Stadtehygiene 20:1-3.
Goldschmid, J., Zohar, D., Argaman, Y. and Kott, Y. (1973)
Effect of dissolved salts on the filtration of coliform
bacteria in sand dunes. In Advances in Water Pollution
Research, pp. 147-157 (Edited by Jenkins S.H.) Pergamon,
Oxford, U.K.
-------
501
Goldstein, B.D. and McDonagh, E.H. (1975) Effect of ozone on
cell membrane protein fluorescence. I. In vitro studies
utilizing the red cell membrane. Environ. Res. 9:179-
186. ~ : r-
Gorham, P.R. (1962) Laboratory studies on the toxins pro-
duced by waterblooms of blue-green algae. American
Journal of Public Health 52:2100-2105. "
Gorham, P.R. (1965) Toxic waterblooms of blue-green algae.
In Transactions of the Third Seminar on Biological
Problems in Water Pollution (August 13-17, 1962) U.S.
Public Health Service Bulletin 999-WP-25, pp. 37-44,
Cincinnati, Ohio.
Goss, W.A., Michaud, R.N. and McGrath, M.B. (1974) Evalu-
ation of an automated colony counter. Appl. Microbiol.
27:1:264-267. .
Gotovskaia, S.N. and Mogaram, M.E. (1945) Study of the
bactericidal action of chlorine on B. tularense in
drinking water. Zh. Mikrobiol. 7(8):50-55.
Gould, R.G. and Cook, R.P. (1958) The metabolism of chole-
sterol and other sterols in the animal organism. In
Cholesterol, Chemistry, Biochemistry and Pathology
(Edited by Cook, R.P.) pp. 289-292, Academic Press, New
York.
Goyal, S.M., Gerba, C.P. and Melnick, J.L. (1978) Prevalence
of human enteric viruses in coastal canal communities.
Journal of the Water Pollution Control Fed. 50:2247-
2256. ~
Granstrom, M.L. and Lee, G.F. (1958) Generation and use of
chlorine dioxide in water treatment. J. Am. Water
Works Assoc. 50:1453-1466. "~~
Graveland, A. and Hrubec, J. (1976) Technieken voor de
bereiding van drinkwater. in Gezond drinkwater
(Editor Samenstelling) B.C.J. Zoeteman, Staatsuit-
geverij 's-Gravenhage, the Netherlands.
Graves, I.L. and Oppenheimer, J.R. (1975) Human viruses in
animals in West Bengal: An ecological analysis.
Human Ecology 3:105-130.
-------
502
Green, D.E. and Stumpf, P.K. (1946) The mode of action of
chlorine. J. Am. Water Works Ass. 38:1301-1305.
Grew, N., Gohd, R.S., Arguedas, J. and Kato, J.I. (1970)
Enteroviruses in rural families and their domestic
animals. Amer. Journal of Epidemiol. 96:518-526.
Guelin, A., Lepine, P. and Lamblin, D. (1967) Pouvoir
bactericide des eaux polluees et role de Bdellovibrio
bacteriovorus. Ann. Inst. Pasteur. 113:660-665.
Guy, M.D., Mclver, J.D. and Lewis, M.J. (1977) The removal
of virus by a pilot treatment plant. Water Res. 11:
421-428.
Gyllenberg, H. and Niemela, S. (1959) A selective method
for the demonstration of bifidobacteria (Lactobacillus
bifidus) in materials tested for faecal contamination.
Maataloustieteellinen Aikakauskirja. 31:94-97.
Hadley, W.K. and Senyk, G. (1975) Early detection of micro-
bial metabolism and growth by measurement of electrical
impedance. Microbiology 1975 (Edited by Schlessinger,
D.) pp. 12-21, American Society for Microbiology,
Washington, B.C.
Hajna, A.A. and Perry, A.C. (1943) Comparative study of
presumptive and confirmative media for bacteria of
coliform group and for fecal streptococci. American
Journal of Public Health 33:550-556.
Hamelin, C. and Chung, Y.S. (1974) Resistance \ 1'ozone
chez Escherichia coli. II. Relations avec certains
mecanismes de reparation de 1'ADN. Molec. Gen. Genet.
129:177-184.
Hamelin, C. and Chung, Y.S. (1975) The effect of low concen-
trations of ozone on Escherichia coli chromosome.
Mutation Res. Netherl. 28:131-132.
Hamilton, R.D. and Holm-Hansen, O. (1967) Adenosine tri-
phosphate content of marine bacteria. Limnol., Oceanog.
12:319-324.
Hanes, N.B. Sarles, W.B. and Rohlich, G.A. (1964) Dissolved
oxygen and survival of coliform organisms and entero-
cocci. J. Am. Water Works Assoc. 56:441-446.
-------
503
Hanson, P.G., Standridge, J. , Jarrett, F. and Maki, D.G.
(1977) Freshwater wound infection due to Aeromonas
hdrohila. Journal of the American Medical Assoc.
hydrophila.
238:1053-10
1054"
Hardy, A.V. and Spector, B.K. (1935) The occurrence of
infestation with Endamoeba histolytica associated with
waterborne epidemic disease's"! Public "Health Report
50:323-329.
Harris, R.H. and Mitchell, R. (1975) Inhibition of floc-
culation of bacteria by biopolymers. Water Research
9:993-999. .
Harvey, R.W.S., Price, T.H., Foster, D.W. and Griffiths,
W.C. (1969) Salmonellas in sewage. A study in latent
human infections. J. Hyg. (Brit.) 67:517-523.
Harvey, S., Greenwood, J.R., Pickett, M.J. and Mah, R.A.
(1976) Recovery of Yersinia enterocolitica from streams
and lakes of California. Appl. Environ. Microbiol.
32:352-354. ~
Haufele, A. and Sprockhoff, H.V. (1973) Ozon als Desin-
fektionsmittel gegen vegetative Bakterien, Bazillen-
sporen, Pilze und Viren in Wasser. Zbl. Bakt. Hyg.,
I Abt. Prig. B. 157:53-70.
Hayashi, M., Suzuki, T. end Saito, H. (1975) An investi-
gation of drinking water on ships. Yokohama-shi
Eisei Kenkyusho Nempo 15:35-38.
Heath, C.W., Alexander, A.D. and Galton, M.M. (1965) Lepto-
spirosis in the United States. Analysis of 483 cases
in man, 1949-1961. New England Journal of Medicine
273:915-922.
Heathman, L.S., Pierce, S.O., and Kabler, P.W. (1936)
Resistance of Various Strains of £3. typhi and Coli-
Aerogenes to Chlorine and Chloramine. Publ. Health
Rpts. 51:1367. ~
Hegner, R. (1930) Host-parasite specificity in the genus
Giardia. In Problems and Methods of Research in
Protozoology (Edited by Hegner, R. and Andrews, A.J.)
MacMillan, New York.
"V
-------
504
Heiberg, B. (1935) On the classification of Vibrio cholerae
and the cholera-like vibrios. A. Buschk, Kopenhagen,
181 pages.
Hem, J.D. (1963-1965) Chemical equilibria and rates of
manganese oxidation. Geological Survey Water Supply
Paper, pp. 1667 A-D, U.S. Government Printing Office,
Washington, D.C.
Henderson, J.M. (1968) Enteric disease criteria for recrea-
tional waters. J. of Sanitary Eng. Div., Amer. Soc.
'Civil Eng. 94:1253-1276.
Hendricks, C.W. (1978) Exceptions to the coliforms and the
fecal coliform tests. In Indicators of Viruses in
Water and Food, pp. 99-145 (Edited by Berg, G.) Ann
Arbor Science, Ann Arbor.
Hettche, O. and Schulz-Ehlbeck, H.W. (1953) Epidemiology
and prophylaxis of polymyelitis with respect to the
role' of water in its transmission. Arch. Hyg. Bakt.
137:440-449.
Highsmith, A.K., R. , Feeley, J.TC., Skaliy, P.
Wells, J.C., . and Wood,.T. (1977a) Isolation of
Yersinia enterocolitica from well water and growth in
distilled water. Appl. Env. Microbiol. 34:745-750.
Highsmith, A.K., Feeley, J.C. and Morris, G.K. (1977b)
Isolation of Yersinia enterocolitica from water. In
Bacterial Indicators/Health Hazards Associated with
Water, pp. 265-274 (Edited by Hoadley, A.W. and Dutka,
B.j.) ASTM STP 635, Amer. Society for Testing and
Materials, Philadelphia.
Hills, D.A. (1967) The degradation of natural rubber pipe-
joint rings. Rubber J. 149:12-17.
Hirn, J. and Raevuori, M. (1978) Tryptose-sulfite-cycloserine-
egg yolk agar, membrane filter method for the detection
of Cl. perfringens in Water. Water Research 12:451-
454T"
Hoadley, A.W. (1977) Potential health hazards associated
with Pseudomonas aeruginosa. In Bacterial Indicators/
Health Hazards Associated with Water, pp. 80-114
(Edited by Hoadley, A.W. and Dutka, D.J.) American
Society for Testing and Materials, Philadelphia, 356
pages.
-------
505
Hoadley, A.W. and Cheng, C.M. (1974) The recovery of indi-
cator bacteria on selective media. J. Appl. Bacteriol.
37:45-57. ' ' ~
Hoadley, A.W. and Dutka, B.J. (Editors) (1977) Bacterial
Indicators/Health Hazards Associated with Water, ASTM
STP 635, American Society for Testing and Materials,
Philadelphia.
Hobbie, J.E., Daley, R. J. , and Jasper, S. (1977) Use of Nucle-
pore filters for counting bacteria by flourescence
microscopy. Appl. Environ. Microbiol. 33:1225-1228.
Hoehn, R.C., Randall, C.W., Bell, F.A., Jr. and Shaffer,
P.T.B. (1977) Trihalomethanes and viruses in water
supply. Journal of Environmental Engineering Div.
ASCE 103:803-814.~~
Hoekstra, A.C.- (1978) Desinfectie van Haags drinkwater en
zuiverende werking van de langzame zandfilter, periode
1966-1978. Report of DWL's-Gravenhage, 5 pages.
Drinkwaterleiding van 's-Gravenhage, 's-Gravenhage, The
Netherlands.
Hoff, J.C. (1978) The relationship of turbidity to disin-
fection of potable water. In Evaluation of the Micro-
biology Standards for Drinking Water (Edited by Hen-
driks, C.W.) EPA-570/9-78-OOC, U.S. EPA, Washington,
D.C., 235 pages.
Hoffman, A. (1970) Studies on the suitability of copro-
stanol as an indicator of fecal pollution in water.
University of Massachusetts. Ph.D. Thesis.
Holden, W.S. (Editor) (1970) Water Treatment and Examination,
pp. 197, 254, 248, and 330.Churchill,London.
Hollister, A.C., Jr., Beck, M.D., Gittelsohn, A.M. and
Hemphill, E.G. (1955) Influence of water availability
on Shigella prevalence in children of farm labor
families. American Journal of Public Health 45:354-
362. :
Holluta, J. and Unger, U. (1954) Die Keimtotung von Bact.
E. coli durch Chlordioxyd und Ozon. Vom Wasser 21:129-
-.
-------
506
Holm-Hansen, O. and Booth, C.R. (1966) The measurement of
Adenosine triphosphate in the ocean and its ecological
significance. Limnol.f Oceanog. 11:510-519.
Hopla, C.E. (1974) The ecology of tularemia. In Advances
in Veterinary Science and Comparative Medicine, pp.
53 (Edited by Brandly, C.A. and Cornelius, C.E.)
Academic Press, New York and London.
'25-
Hoppe, H.G. (1976) Determination and properties of actively
metabolizing heterotrophic bacteria in the sea, in-
vestigated by means of micro-autoradiography. Marine
Biology (Berlin) 36:291-302.
Hoppe, R.C. (1976) Determination and properties of actively
metabolizing heterotrophic bacteria in the sea, in-
vestigated by means of microautoradiography. Marine
Biol. 36:291-302.
Hornick, R.B., Greisman, S.E. and Woodward, T.E. (1970)
Typhoid fever: Pathogenesis and immunologic control.
New England J. Med. 283:686-691.
Horsfall, F.L. and Tamm, I. (Editors) (1965) Viral and
Rickettsial Infections of Man, p. 574, Lippincott,
Philadelphia.
Horster, L. (1943) Die bedeutung der ambenruhr fur Nord-
dafrika und ihre Bekampfung. Deut. Tropenmed. Ztschr.
47:299-305.
Horwitz, M.A., Hughs, J.M. and Craun, G.C. (1976) Outbreaks
of waterborne diseases in the United States, 1974.
Journal of Infectious Pis. 135:588-593.
Houston, A.C. (1908 to 1911) First through Seventh Reports
on Research Work, Metropolitan Water Board, Thames
Water Authority, London.
Houston, A. (1916) The Growth of Microbes on Leather Washers,
12th Research Report, Metropolitan Water Board, London.
Hsu, Y. (1964) Resistance of infectious RNA and 'transforming
DNA to iodine which inactivates f2 phage and cells.
Nature 203:152.
Hsu, Y., Shigeko, N. and Kruse, C.W. (1965) Some bacteri-
cidal and virucidal properties of iodine not affecting
infectious RNA and DNA. Am. J. Epidemiol. 82:317-328.
-------
507
Hugh, R. and Gilardi, G.L. (1974) Pseudomonas. In Manual
of Clinical Microbiology, Second Edition, pp. 250-269
(Edited by Lennette, E.H., Spaulding, E.H. and Truant,
J.P.) American Society for Microbiology, Washington,
D.C. ,
Hughes, J.M. and Merson, M.H. (1975) Outbreaks of water-
borne disease in the U.S., 1973. Journal of Infectious
Diseases 132:336-339.
Hugues, B. -and Plissier, M. (1977) Inactivation comparee de
poliovirus 1, Escherichia coli, Salmonella paratyphi B
par le chlore et par I1ozone. Convention de recherche
No. 74-47 avec le Ministere de la Qualite de la Vie et
de 1'Environnement Operation No. 5600 74-069.
Huisman, L. and Wood, W.E. (1974) Slow sand filtration.
World Health Organization, Geneva.
Hunter, C.A. and Ensign, P.R. (1947) An epidemic of diarrhea
in a newborn nursery caused by Pseudomonas aeruginosa.
Amer. Journal of Public Health 37:1166-1169.
Hutchinson, M. and Ridgway, J.W. (1977) Microbiological
aspects of drinking water supplies. In Aquatic Micro-
biology. Society for Applied Microbiology, Symposium
Series No. 6 (Edited by Skinner, F.A. and Shewan, J.M.)
pp. 179-218, Academic Press, London.
Hutchinson, M., Ridgway, J.W. and Cross, T. (1975) Bio-
deterioration of rubber in contact with water, sewage
and soil. In Microbial Aspects of the Deterioration
of Materials. Society for Applied Bacteriology, Tech-
nical Series No. 9 (Edited by Lovelock, D.W. and
Gilbert, R.J.) p. 261, Academic Press, London.
Hvid, N. (1955) Microbiological and hygienic studies on
underground water in Sjaelland. Thesis. Danish
Science Press Ltd., Copenhagen, 194 pages.
Ilvovitch, M. (1977) World water resources present and
future. Ambio. 6:13-21.
Ingols, R.S. and Ridenour, G.M. (1948) Chemical properties
of chlorine dioxide in water treatment. J. Am. Water
Works Assoc. 40:1207-1227.
International Ozone Institute (now Association) (1976)
Ozone News 1:1-6.
-------
508
Internationaler Schiffsicherheitsvertrag (Solas) 60, Kap. 3,
Regel 11, Abs. A (XIII, XVII) International Treaty for
Safety on Ships.
Iversen, W.P. (1975) Anaerobic corrosion: Metals and
microbes in two worlds. Dev. Ind. Microbiol. 16:1-10.
Jakubowski, W., Hoff, J.C., Anthony, N.C. and Hill, W.F.,
Jr. (1974) Epoxy fiberglass adsorbent for concentrating
viruses from large volumes of potable water. Applied
Microbiol. 28:501-502.
Jenkins, S.H. and Mendia, L. (Editors) (1967) Advances
in Water Pollution Research, Water Pollution Control
Fed., Washington, D.C., 421 pages.
Jepsen, A. (1972) Jord/Vand Hygiejne.
tensen, Copenhagen, 223 pages.
A/S Carl Fr. Mor-
Joklik, W.K. and Willett, H.P. (Editors) (1976) Zinsser
Microbiology, Appleton-Century-Crofts, New York, 1223
pages.
Jones, E.B.G. and Campion-Alsumard, T. Le (1970). The
biodeterioration of polyurethane by marine fungi,
Int. Biodeterior. Bull. 6:119-124.
Jones, F., Smith, P. and Watson, D.C. (1978) Pollution of a
water supply catchment by breeding gulls and the poten-
tial environmental health implications. J. Inst. Wat.
Engrs. Sci. 32:469-482.
Jones, J. and Schmitt, J.A. (1978) The effect of chlori-
nation on the survival of cells of Candida albicans.
Mycologia 70:684-689.
Jones, J.G. (1974) Some observations'on direct counts of
freshwater bacteria obtained with a fluorescence micro-
scope. Limnol.. Oceanogr. 19:540-543.
Jordan, E.G., Russell, H.L. and Zeit, F.R. (1904) The
longevity of the typhoid bacillus in water. Journal
of Infectious Pis. 1:641-689.
Jorgensen, J.H., Lee, J.C., and Pahren, H.R. (1976) Rapid
detection of bacterial endotoxins in drinking water and
renovated wastewater. Appl. Environ. Microbiol. 32:347-
.351.
-------
509
Kampelmacher, E.H. and van Noorle Jansen, L.M. (1973)
Salmonella and thermo-tolerant E. coli in the Rhine
and Meuse at their points of entry into the Nether-
lands. Water 6:199-200.
Kampelmacher, E.H. and van Noorle Jansen, L.M. (1976)
Salmonella effluent from sewage treatment plants,
wastepipe of butcher's shops and surface water in
Walcheren. Zentralbl. Bakteriol. Hyg. Abt. I, orig. B.
167:307-319.
Kaneko, T. and Colwell, R.R. (1978) The annual cycle of
Vibrio parahaemolyticus in Chesapeake Bay. Micro-
biol. Ecol. 4:135-155.
Kaper, J., Lockman, H., Colwell, R.R. and Joseph, S.W.
(1979) Ecology, serology, and enterotoxin production of
Vibrio cholerae in Chesapeake Bay. Applied & Environ.
Microbiol. 37:91-103.
Kapikian, A.S., Wyatt, R.G., Dolin, R., Thornhill, T.S.,
Kalica, A.R. and Chanock, R.M. (1972) Visualization by
immune electron microscopy of a 27-nm particle asso-
ciated with acute infectious nonbacterial gastroen-
teritis. J. Virology 10:1075-1081.
Kapperud, G. (1977) Yersinia enterocolitica and Yersinia-
like microbes isolated from mammals and water in Norway
and Denmark. Acta path, microbiol. scand. Sect. B,
85:129-135.
Kapperud, G. and Jonsson, B. (1976) Yersinia enterocoli-
tica in brown trout (Salmo trutta L.) from Norway.
Acta path, microbiol. scand. Sect. B, 84:66-68.
Kaschkin, P.N. (1974) Some aspects of the candidosis problem.
Mycopathol. Mycol. Appl. 53:173-181.
Kathe, J. (1945) Witterungsverhaltnisse und Mause in ihrer
Bedeutung fur die Epidemiologie des Schlamm- und
Feldfiebers. Zbl. Bakt., I. Abt. Orig. 151:481-489.
Katz, D.H., Faulkner, M., Katz, L.R., Lindh, E., Leonhardt,
C.C., Herr, K. and Tung, A.S. (1977) A rapid, semi-
automated counting procedure for enumeration of anti-
body-forming cells in gel and nucleated cells in sus-
pension. J. Immunol. Meth. 17:285-292.
-------
510
Katz, J. (1972) Experimentation With Human Beings, Russel
Sage Foundation, New York, NY.
Katzenelson, E. (1967) A rapid method for quantitative
assay of poliovirus from water with the aid of fluorescent
antibody technique. Archives of Virology 50:197-206.
Katzenelson, E. (1978) Survival of viruses. In Indicators
of Viruses in Water and Food, pp. 39-50 (Edited by
Berg, G.) Ann Arbor Science, Ann Arbor.
Katzenelson, E. and Kedmi, S. (1979) Unsuitability of
polioviruses as indicators of virological quality of
water. Applied and Environmental Microbiology 37(2):
343-344.
Katzenelson, E., Kletter, B. and Shuval, H.I. (1974) Inactivation
on kinetics of viruses and bacteria in water by use of
ozone. J. Am. Water Works Assoc. 66:725-729.
Kedmi, S. and Katzenelson, E. (1978) A rapid quantitative*
fluorescent antibody assay using tragacanth gum.
Archives of Virology 56:337-340.
Keet, E.E. (1974) Yersinia enterocolitica septicemia.
Source of infection and incubation period identified.
New York State Journal Med. 74:2226-2229.
Kenner, B.A. (1978) Fecal streptococcal indicators. In
Indicators of Viruses in Water and Food, pp. 147-169
(Edited by Berg, G.), Ann Arbor Science, Ann Arbor.
Rennet, A.C. (1966) Attack of microorganisms on underground
flexible joints. N.Z. Ind. Bull. 22:26-28.
Kermack, W.O. and McKendrick, A.G. (1927) A contribution to
the mathematical theory of epidemics, [Part 1], Proc.
Roy. Soc. A 111:700-721.
Keurings Instituut voor Waterleiding-Artikelen (1961)
Testing of Rubber Rings for the Water Industry, KIWA,
The Hague, 160 pages.
Keurings Instituut voor Waterleiding Artikelen (KIWA)
(1974) Richtlijnen voor de aanleg van drinkwater-
installaties in woningen. Rapport uitgebracht door
de commissie aanleg binnenleidingen van het KIWA
Augustus 1974 Keuringsinstituut voor waterleiding-
artikelen N.V. Keurings Instituut voor Waterleiding
Artikelen, The Hague
-------
511
Keurings Instituut voor Waterleiding-Artikelen (KIWA) (1977)
Beveiliging tegen het binnendringen van vreemde stoffen
in waterleidingen juli 1977, 4e (herziene) druk. The Hague.
Khanna, P. (1973) Enumeration and differentiation of water
bacteria with phosphorous 32. J. Water Poll. Contr.
Fed. 45:262-268.
Kibbey, H.J., Hagedorn, C. and McCoy, E.L. (1978) Use of
fecal streptococci as indicators of pollution in soil.
Applied and Environmental Microbiology 35(4):711-717.
King, E.O. (1962) The laboratory recognition of Vibrio
fetus and a closely related Vibrio isolated from cases
of human vibriosis. Annual New York Academy of Sciences
98:700.
Kinman, R.N. (1975) Water and wastewater disinfection with
ozone: A critical review. Critic. Rev. Environ.
Contr. 5:141-152.
Kinman, R.N., Black, A.P. and Thomas, W.C. Jr. (1970)
Status of water disinfection with iodine, pp. 11-35. In
Proc. National Specialty Conference on Disinfection.
Amherst, Mass. July, 1970. American Society of Civil
Eng., New York.
Kirchmer, C.J. (1971) 5 -cholestan-3 -ol: An indicator of
pollution. University of Florida. Ph.D. Thesis.
Kittrell, F.W. and Furfari, S.A. (1963) Observations of
coliform bacteria in streams. J. Water Pollut. Control
Fed. 35:1362-1385.
Kjellander, J. (1960) Enteric streptococci as indicators of
fecal contamination of water. Act a Pathologica: et
Microbiologica Scandinavica Suppl. 136:1-133.
Kjellander, J. and Lund, E. (1965) Sensitivity of Esche-
richia coli and poliovirus to different forms of
combined chlorine. J. Am. Wat. Wks. Ass. 57:893-900.
Klein, P.A. and Wu, S.Y. (1974) Stress: A factor to be
considered in heterotrophic microorganisms in enumer-
ation from aquatic environments. Appl. Microbiol.
27:429-431.
Klock, J.W. (1971) Survival of coliform bacteria in waste-
water treatment lagoons. Journal of the Water Pollu-
tion Control Fed. 43:2071-2083.
-------
512
Klotz, M. (1979) Mikrobiologische Untersuchungen zur Trink-
wasseraufbereitung mit Aktivkohle. Quantitative Er-
fassung und Bewertung. Diplom-Arbeit, Universitat des
Saarlandes, Saarbrucken, FRG.
Klotz, M.f Werner, P. and Schweisfurth, R. (1975) Unter-
suchungen zur Biologie in Aktivkohlefiltern. In
Heft 9:7. Vortragsreihe mit Erfahrungsaustausch
dber spezielle Fragen der Wassertechnologie-Adsorption,
pp. 270-282, Engler-Bunte Institut der Universitat
Karlsruhe, Karlsruhe, FRG.
Klotz, M., Werner, P. and Schweisfurth, R. (1976a) Mikro-
biologische Untersuchungen der Aktivkohlefiltration zur
Trinkwasseraufbereitung. In Forschung und Entwicklung
in der Wasserpraxis (Wissenschaftliche-he Berichte '-fiber'
Untersuchungen und Planungen der Stadtwerke Wiesbaden
AG), Band 3, pp. 73-82, Stadtwerke Wiesbaden AG, Wies-
baden, FRG.
Klotz, M., Werner, P. and Schweisfurth, R. (1976b) Investigations
concerning the microbiology of activated carbon filters. In
Translation of Reports on Special Problems of Water Technology.
Volume 3. Adsorption. Conference: Jfeia in Karlsruhe FRG 1975,
pp. 312-330, Municipal Environmental Research Laboratory,
Office of Research and Development, US""Hivironmental
Protection Agency, Cincinnati, Ohio EPA-600/9-76-030,
December 1976.
Knittel, M.D., Seidler, R.J., Eby, C. and Cabe, L.M. (1977)
Colonization of the botanical environment by Kleb-
siella isolates of pathogenic origin. Appl. Environ.
Microbiol. 34:557-563.
Knox, W.E., Stumpf, P.K., Green, D.E. and Auerbach, V.H.
(1948) The inhibition of sulfhydryl enzymes as the
basis of the bactericidal action of chlorine. J.
Bacteriol. 55:451-458.
Koff, R.S., Chambers, T.C., Culhane, P.O. and Iber, F.L.
(1973) Underreporting of viral hepatitis. Gastroen-
terology 64:1194-1195.
Konst, H., McKercher, P.O., Gorham, P.R., Robertson, A.
and Howell, J. (1965) Symptoms and pathology produced
by Microcystis aeruginosa NCR-1 in laboratory and
domestic animals. Canadian Journal of Comparative
Medicine and Veterinary Science 29:221-228.
-------
513
Kool, H.J. (1976) The use of adenosine triphosphate (ATP)
as an indicator of biological activity and biomass in
ecological studies. Hydrobiol. Bull. 10:145-154.
Kool, H.J. (1977) Het gedrag van heterotrofe bacterien in
relatie tot net fytoplankton in het spaarbekken De
Grote Rug en drie model reservoirs. RID rapport CBH
78/09, pp. 1-16, Natl. Inst. for Water Supply, P.O. Box
150, 2260 AD Leidschendam, the Netherlands.
Kool, H.J. (1978) Treatment processes in public water supply
for the removal of microorganisms. Chapter 17. In
Biological Indicators of Water Quality, pp. 17-1-17-31
(Edited by James, A. and Evison, L.E.) John Wiley and
Sons Ltd., Chichester, UK.
Koshi, G. and Lalitha, M.K. (1976) Edwardsiella tarda in a
variety of human infections. Indian. Journal of
Medical Research 64:1753-1759.
Kott, Y. (1966) Estimation of low numbers of Escherichia
coli bacteriophages by use of the most probable number
method. Appl. Microbiol. 14:141-144.
Kott, Y., Ben-Ari, H., and Vinokur, L. (1978a) Coliphages
survival as viral indicator in various wastewater
quality effluents. Prog. Water Technol. 10(1/2):337-
346.
Kott, Y., Ben-Ari, H., and Betzer, N. (1978b) Lagooned,
secondary ' effluents as water source for extended
agricultural purposes. Water Res. 12:1101-1106.
Kovacs, G. (1977) Human interaction with groundwater.
Ambio. 6:22-26.
Kristensen, K«» K. (1974) Vandhygiejne, foreliggende
viden og forskningsbehov. Publ. by Water Quality
Institute, Agern alle 11, Horsholm, Denmark. 293
pages.
Kristensen, K.K. (1977) Forekomst og overlevelsesevne
af Yersinia enterocolitica og Yersinia pseudotuber-
culosis i spild vand, slam og overfladevand med til-
h^rende fauna. Vandkvalitetsinstituttet, ATV, 2970
HjzJrsholm, Denmark.
Kristensen, K.K. (1978) Endotoxiner-et omgivelseshygiejnisk
problem. Vand. 2:54-56.
-------
Krogstad, 0. (1974) Survival of Yersinia enterocolitica
in water. Proc. 12th Nordic Vet. Congress Reykjavic,
p. 256. A/S Carl Fr. Mortenson, Copenhagen.
Kruse, C.W., Hsu, Y.C., Griffiths, A.C. and Stringer, R.
(1970) Halogen action on bacteria, viruses, and proto-
zoa. In Proceedings of the National Specialty Con-
ference on Disinfection, pp. 113-136. Amer. Society of
Civil Eng., New York.
Kubota, Y. and Liu, P.V. (1971) An enterotoxin of Pseudo-
monas aeruginosa. J. Infect. Pis. 123:97-98.
Kullmann, K.H. and Schweisfurth, R. (1978) Eisenoxydierende,
stabchenformige Bakterien. II. Quantitative Unter-
suchungen zum Stoffwechsel und zur Eisenoxydation mit
Eisen (Il)-Oxalat. Z. Allg. Mikrobiol. 18:321-327.
Kullmann, K.H. and Schweisfurth, R. (1978) Eisenoxydierende,
stabchenformige Bakterien. III. Quantitative Unter-
suchungen zum Effekt von Eisen auf das Wachstum.
Z. Allg. Mikrobiol. 18:427-435.
Kuster, E. and Williams, S.T. (1964) Selection of media for
isolation of streptomycetes. Nature, Lond. 202:928.
Kuzminski, L.N., Feng, T.H., and Liu, C.C. (1970) Effect of
inorganics on disinfection by halogens. In Proceedings
of the National Specialty Conference on Disinfection,
pp. 85-111, American Society of Civil Engineers, New
York (no editor).
LaBelle, R.L. and Gerba, C.P. (1979) Influence of pH,
salinity, and organic matter on adsorption of enteric
viruses to estuarine sediment. Applied and Environ-
mental Microbiol. 38:93-101.
Labonde, J. and Festy, B. (1979) Bilan d'une recherche sys-
tematique de Pseudomonas aeruginosa dans les eaux de
consommation. Revue de 1'Institut Pasteur de Lyon
12:507-517.
Laboratory Centre for Disease Control (1978) Campylobacter
enteritis Ontario. Canada Diseases Weekly Report
4:57-58. ~
LaGrange, W.S., Moon, N.J. and Reed, W. (1977) The 3M
Automatic colony counter in the bacterial evaluation of
manufacturing-grade milk. J. Food Prot. 40:11:744-746.
514
-------
515
Lance, J.C.f Gerba, C.P. and Melnick, J.L. (1976) Virus
movement in soil columns flooded with secondary sewage
effluent. Applied and Environmental Microbiol.
32:520-526."
Lanyi, B.J., Szita, J., Ringelhann, B. and Kovach, K.
(1959) A waterborne outbreak of enteritis associated
with Escherichia coli serotype 124:72:32. Acta micro-
biol. acad. sci. Hung 6:77-84.
Lassen, J. (1972) Yersinia enterocolitica in drinking
water. Scandinavian Journal ot Infectious Diseases
4:125-127.
Leclerc, H. (1967) Mise en evidence de, la decarboxylase de
glutamique chez les bacteries a I1 aide d ' une
1'acide glutamique
technique automatique.
Ann. Inst. Pasteur 112:713-731.
Leclerc, H. and Catsaras, M. (1967) Un caractere constant
chez 1'Escherichia coli; La presence d'une decarboxy-
lase de 1'acide glutamique. Bull. Ass. Dipl. Microbiol.
Nancy 107:39-49.
Leclerc, H., Mizon, F., Boniface, B. and Boniface, M.
:teries resistantes aux antibio-
(1977a) Eaux etxbacteries
tiques: Etude ecologique.
Pasteur) 128 Bs253-265.
Ann. Microbiol. (Inst.
Leclerc, H., Mossel, D.A.A., Trinel, P.A. and Gavini, F.
(1977b) Microbiological monitoring a new test for
fecal contamination. In Bacterial Indicators Health
Hazards Associated with Water ASTM STP 635, pp. 21-31
(Edited by Hoadley,~A.W. and Dutka, B.J.) American
Society for Testing and Materials, Philadelphia.
Lee, P.K. (1972) Role of crucian carp in transmission of
enteric pathogens. J. Cath. Med. Col. (Korea) 23:
379-386.
Lee, S.H., O'Conner, J..T. and Banerji, S.K. (1977) Bio-
logically mediated deterioration of water quality in
distribution systems. Paper 3B-2, Water Works Asso-
ciation, 5th Water Quality Technology Conference,
Kansas City, Dec 4-7, 1977. American Water Works
Association, Denver, 14 pages.
Lee, Y., Jablonski, I. and DeLuca, M. (1977) Immobilization
of firefly luciferase on glass rods: Properties of the
immobilized enzyme. Analytical Biochem. 80:496-501.
-------
516
Leeflang, K.W.H. (1963) Microbiologic degradation of rubber.
Journal of the American Water Works Association 55:1523-
1535. ~~~
Leffler, E. and Kott, Y. (1974) Enteric viruses behaviour
in sand dunes. Israel J. of Techn. 12:298-304.
Leiguarda, R.H., Peso, O.A. and Palazzolo, A.Z. (1949)
Bactericidal action of ozone. An. Assoc. Quim. Argent.
37:165-176.
Lennette, E.H., Spaulding, E.H. and Truant, J.P. (1974)
Manual of Clinical Microbiology, 2nd edition, American
Society for Microbiology, Washington, D.C., 970 pages.
Levin,^G.V. (1963) Rapid microbiological determinations
with radioisotopes. Advances in Applied Microbiology
. (Edited by Umbreit, W.W.) 5:95-133, Academic Press,
New York, USA.
Levin, G.V. (1968) Experiments and instrumentation for
extraterrestrial life detection. Advances in Applied
Microbiology (Edited by Umbreit, W.W. and Perlman, D.)
Academic Press, New York, USA, 10:55-71.
Levin, G.V., Harrison, V.R., Hess, W.C. and Gurney, H.C.
(1956) A radioisotope technique for the rapid detec-
tion of coliform organisms. Am. J. Public Health
46:1405-1414.
Levin, J. and Bang, F.B. (1964) A description of cellular
coagulation. In Limulus Bull., Johns Hopkins Hosp.
115:337.
Levin, M.A. (1977) Bifidobacterium as water quality indi-
cators. In Bacterial Indicators/Health Hazards Asso-
ciated with Water ASTM STP 635 (Edited by Hoadley,
A.W. and Dutka, B.J.) pp. 131-138.
Levin, M.A., and Cabelli, V.J. (1972) Membrane filter tech-
nique for enumeration of Pseudomonas aeruginosa.
Appl. Microbiol. 24:864-870.
Levin, M.A., Fischer, J.R. and Cabelli, V.J. (1974) Quan-
titative large-volume sampling technique. Appl.
Microbiol. 28:515-517.
-------
517
Levine, M.H., Dupont, H.L. and Formal, S.B. (1973) Patho-
genesis of Shigella dysenteriae (Shiga) dysentery.
J. Infect. Pis. 127;261-270.
Levy, B.S., Mature, J. and Washburn, J.W. (1977) Intensive
hepatitis surveillance in Minnesota: Methods and
results. Am. J. Epidemiol. 105:127-134.
Liao, S.J., Berg, P.P. and Bonchard, R.J. (1954) Epidemi-
ology of infectious hepatitis in an urban population
group. Yale J. Biol. Med. 26:512-526.
Lin, S. (1977) Evaluation of methods for detecting coli-
forms and fecal streptococci in chlorinated secondary
sewage effluents. In Symposium on the Recovery of
Indicator Organisms Employing Membrane Filters.
(Edited by Bordner, R.H. , Frith, C.H., and Winter, J.A.)
U.S. E.P.A., Cincinnati, Ohio, U.S.A., pp. 113-132.
Liu, O.C. (1973) Effect of chlorination on human enteric
viruses in partially treated water from the Potomar
River estuary. Study report, U.S. Army Engineer Divi-
sion, North Atlantic Philadelphia, 118 pages.
Lloyd, B.C. (1973) The construction of a sand profile
sampler, its use in the study of the Vorticella popu-
lations and the general interstitial microfauna of slow
sandfilters. Water Res. 7:963-973.
Lothrop, T.L. and Sproul, O.J. (1969) High-level inacti-
vation of viruses in wastewater by chlorination.
J. Wat. Poll. Control Fed. 41:567-575.
Lovelock, D.W. and Gilberg, R.J. (Editors) (1975) Micro-
bial Aspects of the Deterioration of Materials.
Society for Applied Bacteriology, Technical Series 9,
Academic Press, London.
Lovtsevich, E.L. (1973) Certain aspects of the processes of
enterovirus inactivation with chlorine (in Russian)".
Gig. Sanit. 10:11-15.
Lowbury, E.J.L., Thorn, B.T., Lilly, H.A., Rabb, J.R., and
Whittall, K. (1970) Sources of infection with Pseudo-
monas aeruginosa in patients with tracheotomy. Journal
of Medical Microbiology 3:39.
-------
518
Lund, E. (1963) Effect of pH on the oxidative inactivation
of poliovirus. Arch. Ges. Virusforsch. 12:632-647.
Lund, E. (1978) Human pathogens as potential health hazards
in the reuse of water. Ambio. 7:56-61.
Luttmann, U. (1967) Investigation into the biology of free-
growing Salmonellas in the Weser-Leine region during
1964-7. Gesundheitswes. Desinfekt. 59:139, 159 and
180.
Lycke, E., -Blomberg, J., Berg, G., Eriksson, A. and Madsen,
L. (1978) Epidemic acute diarrhea in adults associated
with infantile gastroenteritis virus. Lancet 1056-
1057.
Mack, W.N. (1977) Total coliform bacteria. In Bacterial
Indicators/Health Hazards Associated with Water ASTM
STP 635, pp. 59-64 (Edited by Hoadley, A.W. and Dutka,
B.J.) American Society for Testing and Materials,
Philadelphia.
Mackenthun, K.M. and Ingram, W.M. (1967) Biological Asso-
ciated Problems in Freshwater Environments. US Dept.
of the Interior, FWPCA, Washington, B.C., 287 pages.
MacKenzie, E.F.W., Taylor, E.W. and Gilbert, W.E. (1948)
Recent experiences in the rapid identification of
Bacterium coli Type 1. Journal of General Microbiology
2(2):197-204.
Madeley, C.R. (1979) Viruses in the stools,
32:1-10.
J. Clin. Path.
Mahnel, H. (1977) Untersuchengen zur Virusinaktivierung in
Trink und Oberflachenwasser. Ein Beitrag zur Dekon-
tamination von trinkbaren Wassern mit Feldangleichen
Methoden. Zbl. Bakt. Hyg. 1, Abt. Prig. B 165:527-538,
Mair, I.W.S., Natvig, K. and Johannessen, T.A. (Sept.,
1973) Otolaryngological manifestations of tularemia.
Arch. Otolaryngol. 98:156-158.
Majumdar, S.B., Ceckler, W.H. and Sproul, O.J. (1974)
Inactivation of poliovirus in water by ozonation.
J. Water Poll. Control Fed. 8:2048-2053.
Marier, R. (1977) The reporting of communicable diseases.
Am. J. Epidemiol. 105:587-590.
-------
519
McCabe, L.J. (1977) Epidemiological consideration in the
application of indicator bacteria in North America.
In Bacterial Indicators/Health Hazards Associated with
Water, pp. 15-22, ASTM STP 635 (Edited by Hoadley, A.W.
and Dutka, B.J.) American Society for Testing and
Materials, Philadelphia.
McCoy, J.H. (1962) The isolation of salmonellae. J. Appl.
Bacteriol. 25:213-224.
McFeters, G.A. and Stuart, D.G. (1972) Survival of coli-
form bacteria in natural waters: Field and laboratory
studies with membrane-filter chambers. Appl. Micro-
biol. 24:805-811.
McFeters, G.A., Bissonnette, G.K., Jezeski, J.J., Thomson,
C.A. and Stuart, D.G. (1974) Comparative survival of
indicator bacteria and enteric pathogens in well water.
Appl. Microbiol. 27:823-829.
McCarthy, J.A. (1944) Bromine and chlorine dioxide as water
disinfectants. J. New Eng. Water Works Assoc. 58:55-
68.
McCarthy, J.A. (1945) Chlorine dioxide for the treatment of
water supplies. J. New Eng. Water Works Assoc. 59:252-
261.
Melnick,.J.L. and Ledinko, N. (1951) Social serology:
Antibody levels in a normal young population during an
epidemic of poliomyelitis. Amer. Journal of Hygiene
54:354-382.
Melnick, J.L., Gerba, C.P. and Wallis, C. (1978) Viruses in
water. Bulletin of the World Health Organization
56(4) :499-5W;
Mendis, N.M., De La Motte,P-U., and Gunatillaka, P.D. (1976)
Protected infection with Salmonella bareilly in a
maternity hospital. J. Tropical Med. Hyg. 79:142-150.
Mercado-Burgos, N. , Hoehn,R-C., and Holliman, R.B. (1975)
Effect of halogens and ozone on Schistosoma ova.
J. Water Poll. Control Fed. 47:2411-2419.
Metcalf, T.G. (1961) Use of membrane filters to facilitate
the recovery of virus from aqueous suspensions.
Applied Microbiol. 9:376-381.
-------
520
Metropolitan Water Board Report (1966) Pollution by gulls.
Report on the results of the bacteriological, chemical
and biological examination of the London Waters 42:20-21.
Metropolitan Water Board (1967-1968) Thirty-third report on
the results of the bacteriological, chemical and bio-
logical examination of the London waters. Ecological
studies of slow sand filter algae 43:61-65.
Metropolitan Water Board (1969-1970a) Thirty-fourth report:
Aerobic sporing bacilli, fungi and actin omycetes in
slow sand filters. Escherichia coli in "blanket weed"
44:16-23.
Metropolitan Water Board '(1969-1970b) Experimental studies
of slow sand filtration 44:83-86.
Metropolitan Water Board Report (1970a) Gulls on filter
beds. Report on the results of the bacteriological,
chemical and biological examination of the London
waters 44:24-25.
Metropolitan Water Board Report (1970b) Silt deposits in
service reservoirs. Report on the results of the
bacteriological, chemical and biological examination of
the London waters 44:39-42.
Metropolitan Water Board Report (1970c) Biodegradation of
[jointing compounds. Report on the results of the
bacteriological, chemical and biological examination of
the London waters 44:42-47.
Metropolitan Water Board (1971-1973a) Thirty-fifth report:
Slow sand filtration 45:27-53.
Metropolitan Water Board (1971-1973b) Reports on Research
Work Bacteriological, chemical, and biological exami-
nation of London water. Metropolitan Water Board
45:147.
Metropolitan Water Board Report (1973a) Silt deposits in
service reservoirs. Ibid. 45:26.
Metropolitan Water Board Report (1973b) Prefiltration
ozonation. Report on the results of the bacteriological,
chemical and biological examination of' the London
waters 45:49-53.
-------
521
Mikkelsen, T. (1977) Palaeozoiske principper i moderne
medicin. Dansk Kemi 60:76-82.
Miller, G.W., Rice, R.G., Robson, C.M., Scullin, R.L., Kuhn,
W., and Wolf, H. (1978) An assessment of ozone and
chlorine dioxide technologies for treatment of munici-
pal water supplies. U.S. Environmental Protection
Agency Final Report EPA-600/2-78-147.
Miller, J.D.A. (Editor) (1971) Microbial Aspects of^Metal-
lurgy. Medical and Technical Publishing Co., Chiltern
House, Aylesbury, UK.
Miller, L., Snyder, P., Trout, J.R., and Sousner, M. (1975)
Improvement in accuracy of automatic colony counter
through use of calibration factors. Paper Q27. Ab-
stracts of the Annual Meeting, American Society for
Microbiology, p. 209. Presented at 16th meeting of ASM-,
N.Y.
Mitchell, D.O. and Starzyk, M.J. (1975) Survival of Salmonella
and other indicator microorganisms. Can. J. Microbiol.
21:1420-1421.
Mitchell, R.I. (1972) Ecological control of microbial
imbalances. In Water Pollution Microbiology, pp. 273-
288 (Edited by Mitchell, R.) Wiley Interscience, New
York.
Moffa, P.E., Tifft, E.G., Richardson; S.L. and Smith, J.E.
(1975) Benchscale high rate disinfection of combined
sewer overflows with chlorine and chlorine dioxide.
U.S. Environmental Protection Agency Final Report EPA-
670/2-75-021.
Moore, B.E., Sagik, B.P. and Sorber, C.A. (1977) An as-
sessment of potential health risks associated with
land disposal of residual sludges. Presented at the
Third National Conference on Sludge Management Dis-
posal and Utilization, December 14-16, 1976 (Miami,
Florida) 108 Information Transfer, Inc., Rockville,
Maryland.
Mohadjer, S. and Mehrabian, S. (1975) Studies on the sur-
vival of Shigella flexneri in river and tap water.
Arch. Roum. Exp. Microbiol. 34:307-312.
-------
522
Mohrmann, R. (1927) Die Hannoversche Typhusepidemie Im
Jahre 1926. Veroffentl. a. d. Geb. d. Med.-Verwalt
24:393-444.
Mollaret, H.H. (1976) ContributionNa 1'etude epidemiologique
des infections^ Yersinia enterocolitica. III. Bilan
provisoire des connaissances. Med. Mai. Infect. 6:442-
448.
Monnet, P., Buttiaux, R., Papavassiliou, J., Nicolle, P.,
and Le Minor, S. (1954) Escherichia coli des types
0111 et 055 dans des eaux d1alimentation. Ann. Inst.
Pasteur. 87:347-349.
Moore, G.T., McGuire, C.M., Mollohan, C.S., Gleason, N.N.,
Healy, G.R. and Newton, L.H. (1969) Epidemic giar-
diasis at a ski resort. New England Journal of Med.
281:402-404.
Morello, J.A. (1975) Automated radiometric detection of
bacteria. Automation in Microbiology and Immunology
(Edited by Heden, C.G. and Illeni, T.) pp. 35-46, John
Wiley & Sons, New York, USA.
Morris, G.K. and Feeley, J.C. (1976) Yersinia enterocolitica;
A review of its role in food hygiene. Bull. World
Health Organization 54:79-85.
Morris, J.C. (1967) The future of chlorination. J. Am.
Wat. Wks. Ass. 58:1475-1482.
Morris, J.C. (1970) Disinfectant chemistry and biocidal
activities. In Proceedings of the National Specialty
Conference on Disinfection, Massachusetts, pp. 609-633,
American Society of Civil Engineers, New York.
Morris, J.C. (1975) Aspects of the quantitative assessment
of germicidal efficiency, pp. 1-10 in Disinfection;
Water and Wastewater. (Edited by Johnson, J.D.) Ann
Arbor Science Publishers, Inc., Ann Arbor, Mich.
Morris, J.C., Chang, S.L., Fair, G.M. and Conant, G.H., Jr.
(1953) Disinfection of drinking water under field
conditions. Ind. and Eng. Chem. 45:1013-1015.
Mosley, J.W. (1967) Transmission of viral diseases by
drinking water. In Transmission of Viruses by the
Water Route, pp. 5-23 (Edited by Berg, G.), Interscience
Publishers, New York.
-------
523
Mosley, W.H. (1970) Principles and practice of cholera
control. Public Health Papers, World Health Organi-
zation 40:23-27.
Mudd, J.B. and Freeman, B.A. (1977) Reaction of ozone.with
biological membrane. In Biochemical Effects Envi-
ronmental Pollutants, pp. 97-133 (Edited by Duk, S.L.)
Ann Arbor Sci. Publishers, Ann Arbor, Michigan.
Muller, G. (1961) Probleme einer bakteriologisch einwand-
freien Trinkwasserversorgung auf kleineren Schiffen.
Stadtehygiene 12:113-114.
Muller, G. (1964) Welche Konsequenzen ergeben sich aus
den Erfahrungen der Hamburger Flutkatastrophe fur
die hygienische Trinkwassseruntersuchung? Arch.
Hyg. 148:321-327.
Muller, G. (1965) Salmonella in bird faeces.
5003:1315.
Nature, pp,
Muller, G. (1967) Das Vorkommen von agglutinierenden E.
coli Typen inltink- f wid Oberflachenwassern. Z. Bakt,
Parasitenkunde, Infektionskrankheiten und Hygiene I
Orig. 203:464-468.
Muller, G. (1969a) Die gesundheitliche Bedeutung der Ratten.
Schriftenreihe Wa-Bo-Lu 32;9-18.
Muller, G. (1969b) Wasserversorgung. In Das Offentliche
Gesundheitswesen, Vol. 3, pp. 353-390 (Edited by Trub,
C.L.P., Daniels, F. and Posch, F.) Georg Thieme Verlag,
Stuttgart.
Muller, G. (1974) Vorkommen und Bedeutung technisch uner-
wunschter Bakterien in der Trinkwasserversorgung.
Zbl. Bakt. Hyg. I.Abt. Orig. A 227:50-55.
Muller, G. (1976) Die Uberwachung der Wasserversorgung von
Schiffen nach der Trinkwasser-Verordnung. ,;In Die
Trinkwasser-Verordnung, Einfuhrung und Erlauterungen
fur Wasserversorgungsunternehmen und Uberwachungsbe-
horden, pp. 57-64 (Edited by Aurand, K., Hasselbarth,
U., Muller, G., Schumacher, W., Steuer, W. and Schmidt,
E.) Verlag Berlin.
Muller, G. (1977a) Non agglutinable cholera vibrios (NAG) in
sewage, river water and seawater. Zbl. Bakt. Hyg., I.
Abt. Orig. B 165:487-497.
-------
.524
Muller, G. (1977b) Bacterial indicators and standards for
water quality in the Federal Republic of Germany. In
Bacterial Indicators/Health Hazards Associated with
Water. ASTM STP 635, pp. 159-167 (Edited by Hoadley,
A.W., Dutka, B.K.) American Society for Testing and
Materials, Philadelphia.
Muller, G» und Herzel, F. (1973) Chemische, mikro-
skopische und bakteriologische'_ Untersuchungen an lone-
naustauscherharzen. Z. Bakt. Hyg. Abt. I. Prig. B
156:S. 524.
Muller, G. , Goethe, H. and Herrmann, R. (1968) Bakterio-
logische Untersuchungen an Verdampferanlagen zur Trink-
wasserversorgung auf Seeschiffen. Zbl. Bakt. I. Abt.
Prig. 207:409-418.
Muller, G., Goethe, H. and Herrmann, R. (1972) Bakterio-
logische Untersuchungen zur Trinkwasserdesinfektion
durch UV-Bestrahlung an Bord von Seeschiffen. Mittlg.
Zbl. Bakt. Hyg. I. Abt. Prig B 156:361-372.
Muller, G., Kempf, Th., Goethe, H. and Herrmann, R. (1977)
Trinkwasserkonserven ffir Rettungsboote. Forum Umwelt-
hygiene 28:33-34.
Muller, Gertrud, Althaus, H., Hamburger, B«* Schubert,
R.W. and Selenka, F. (1979) Mitteilung uber ,-ine Methode
zur Prufung von Kunststoffen im Trinkwasserbereich mit
Hilfe von Bakterien. (Bundesgesundheitsblatt, in
press).
Mundt, J.O. and Graham, W.F. (1968) Streptococcus faecium
variety casseliflavus, nov.var. Journal of Bacteriology.
95(6):2005-2009.
Hunger, S.F., Heyward, A.A and Swartz, R.G. (1977) Inac-
tivation of virus by chlorination or ozonation in four
wastewater treatment system effluents. Paper Q47.
Abstracts of the Annual Meeting, Am. Soc. Microbiol.,
New Orleans, p. 269. Washington, B.C.
Munoz, E.F. and Silverman, M.P. (1979) Rapid, single-step
most probable number method for enumerating fecal
coliforms in effluents from sewage treatment plants.
Appl. Environ. Microbiol. 37(3):527-530.
-------
525
Murray, C.R. and Reeves, E.B. (1977) Estimated Use of Water
in the United States in 1975, Geological Survey Circular
765. US Geological Survey, Department of the Interior,
Arlington, Virginia.
Murray, R.G.E. (1965) Location of mucapeptide of selections
of the cell wall of E. coli and other gram-negative
bacteria. Can. J. Mlcrobiol. 11:547-560.
Murtaugh, J.J. and Bunch, R.L. (1967) Sterols as a measure of
fecal pollution. J. Water Pollut. Contr. Fed. 39:404-
409.
>
Naemura, L.G. and Seidler, R.J. (1978) Significance of low-
temperature growth associated with the fecal coliform
response, indole production and pectin liquifaction in
Klebs'iella. Appl. Environ. Microbiol. 35:392-396.
Nagington, J., Watson, P.G., Playfair, T.J. , McGill, J.,
Jones, B.R., and Steele, A.D.M. (1974) Amebic infection of
the eye. Lancet, December 28, pp. 1537-1540.
Nath, S. (1974) Histopathological and radiological appearance
of duodenum in giardiasis. Indian Journal of Medical
Research 62:719.
Neefe, J.R-. and Stokes, J. (1945) An epidemic of infec-
tious hepatitis apparently due to a waterborne agent.
Journal of the American Medical Assoc. 128:1063-1075.
tt
Nemedi, L. and Lany, B. (1971) Incidence and hygienic
importance of Pseudomonas aeruginosa in water. Acta
Microbiol. Acad. Sci. Hung. 18:319-326.
Nestor, I.,, and Costin, L. (1971) The removal of Coxsackie
virus from water by sand obtained from the rapid sand
filters of water plants. J. Hyg. Epidenr. Microbiol.
Immun. 15:129-136..
Newson, S.W.B. (1968) Hospital infection from contaminated
ice. Lancet 11:620.
Newton, W.L. and Jones, M.F. (1949) Effect of ozone in
water on cysts of Entamoeba histolytica. Amer. Journal
of Tropical Med. 29:669-681.
New Zealand Department of Scientific and Industrial Research
(1969) Degradation of Rubber Ring Pipe Joints. NZ DSIR
Ann Rept., 1969, pp. 19-20.
-------
526
Nilehn, B. (1973) The relationship of incubation tempera-
ture to serum bactericidal effect, pathogenicity and
in vivo survival of Yersinia enterocolitica. In Con-
tributions to Microbiology and Immunology, Vol. 2, pp.
85-92 (Edited by Winblad, S.) Karger, Basel.
Norwegian Institute of Water Research (1980) Bacteriological
relations in Norwegian and foreign raw water sources,
Report number 0-78029. Norwegian Institute of Water
Research, Box 333, Blindern, Oslo 3, 74 pages.
Nusbaum, I. (1952) Sewage chlorination mechanism-. A survey
of fundamental factors. Water Sewage Works 99:295-297.
O'Connor, J.T., Hash, L. and Edwards, A.B. (1975) Deter-
ioration of water quality in distribution systems.
Journal of the American Water Works Assn. 67:113-116.
Offut, A.C., Poole, B.A. and Fassnacht, G.G. (1955) A
waterborne outbreak of amebiasis. Amer. Journal of
Public Health 45:486-491.
Oger, C., Catteau, M. and Leclerc, H. (1976) Etude, en
laboratoire, de la survie des Salmonella dans une eau
;polluee. Microbia. 2:3-9.
Olivieri,.V.P. (1968) Chlorine dioxide and protein syn-
thesis. Master's Thesis. West Virginia Univ., Morgan-
town, West Virginia.
Olivieri, V.P., Kruse, C.W., Hsu, Y.C., Griffiths, A.C. and
Kawata, R. (1975) The comparative mode of action of
chlorine, bromine and iodine in f. bacterial virus,
145-162 in Disinfection; Water and Wastewater (Edited
by Johnson, J.D.) Ann Arbor Science Publishers, Inc.,
Ann Arbor, Mich.
Ontario Ministry of the Environment (1976) The presence-,
absence procedure. Handbook of Analytical Methods for
Environmental Samples. Laboratory Services Branch,
Ministry of the Environment, P.O. Box 213, Rexdale,
Ontario, Canada.
Opara, A.A. (1978) The role of anaerobic faecal bacteria as
indicators of water pollution in hot climates. Ph.D.
Thesis, University of Dundee.
-------
527
Orchard, V.A. and Goodfellow, M. (1974) The selective
isolation of Nocardia from soil using antibiotics.
J. Gen. Microbiol^ 8^:160-162.
Oremland, R.S. and Silverman, M.-P. (1979) Microbial sulfate
reduction measured by an automated electrical impedance
technique. Geomicrobiol. J. (in press).
0rskov, I. (1974) Klebsiella. In Bergey's Manual of Deter-
minative Bacteriology 8th ed., pp. 321-324 (Edited by
Buchanan, R.E. and Gibbons, N.E.) Williams and Wilkins,
Baltimore, Maryland.
Padchenko, I.K. and Stolyarchuk, N.G. (1969) On the pos-
sible circulation of G. lamblia in nature. Prog.
Protozool. 3:311-312.
Pai, C.H., Sorger, R.T., Lackman, L., Sinai, R.E. and Marks,
M.I. (1979) Campylobacter gastroenteritis in children.
Journal of Pediatrics 94:589-591.
Palfi, A.B. (1971) Virus content of sewage in different
seasons in Hungary. Acta Microbiologica Academiae
Scientiarum Hungaricae 18:231-237.
Palmer, C.M. (1962) Taste and odor algae. In Algae in Water
Supplies, Chapter IV, pp. 18-21, US Dept. of Health,
Education and Welfare, Public Health Service, 88 pages.
Panigrahy, B.
irmiy, D., Grimes, J.E., Rideout, J.E., Simpson, R.B., and
Grumbles, L.C. (1979) Apparent increased occurrence of
chlamydiosis (psittacosis), salmonellosis, and giar-
diasis. Journal of the Amer. Veterinary Med. Assoc.
175:359-361.
Parmentier, G.G. and Eyssen, H.J. (1974) Mechanism of
biohydrogenation of cholesterol to coprostanol by
Eubacterium ATCC 21408. Biochim. Biophys. Acta 348:
279-284.
Pavoni, J.L., Tittlebaum, M.E., Spencer, H.T., Fleischman,
M., Nebel, C., and Gottschling, R. (1972) Virus re-
moval from wastewater using ozone. Water Sew. Works
119:59-67.
Perrich, J.R. (1976) Inactivation of Escherichia coli by
ozonation in an aqueous reactor system. Thesis for the
degree of Doctor of Philosophy, Univ. of Louisville,
Kentucky.
-------
528
Pfeiffer, K.R. (1973) The homestead typhoid outbreak. J_._
Amer.'Water Works Assoc. 65:803-805.
Philips, B.P., Wolfe, P.A.f Rees, C.W., Gordon, H.A.,
Wright, W.H. and Reyniers, J.A. (1955) Studies on the
amoeba bacteria relationship in amoebiasis. Compara-
tive results of the intracecal inoculation of germ-free
non-contaminated and conventional guinea pigs with
Entamoeba histolytica. American Journal of Tropical
Med. Hyg. 4:675-692. ~~~
Picciolo, G.L., Thomas, R.R., Deming, J.W., Chappelle, E.W.
(1977a) Environmental applications of the firefly
luciferase ATP assay flow techniques for monitoring the
wastewater effluent and of drinking water supplies. In
Proceedings of the Second Bi-Annual ATP Methodology
Symposium, pp. 547-564, SAI Technology, 4060 Sorrento
Valley Blvd., San Diego, California.
Picciolo, G.L., Chappelle, E.W., Thomas, R.R. and McGarry,
M.A. (1977b) Performance characteristics of a new
photometer with a moving filter-tape for luminescence
assay. Appl. Environ. Microbiol. 34:720-724.
Picciolo, G.L., Chappelle, E.W., Deming, J.W., Thomas,
R.R., Nibley, A. and Okrend, H. (1981) Firefly
luciferase ATP assay development for monitoring
bacterial concentrations in water supplies.
Environmental Protection Agency (EPA-600/32-81-040
Cincinnati, Ohio.
Pijck, J. and Defalque, A. (1963) Le-depistage precoce de
microorganismes coliformes par voie radiometrique.
J. Pharm. Belg. 18:180-190.
Pipes, W.O. (Editor) (1978) Water Quality and Health Sig-
nificance of Bacterial Indicators of Pollution, Work-
shop Proceedings, National Science Foundation and
Drexel University, Philadelphia.
Pollitzer, R. (1967) History and Incidence of Tularemia
in the Soviet Union; A Review. The Institute of
Contemporary Russian Studies, Fordham Univ., Bronx, New
York, 366 pages.
Popp, L. (1974) Salmonella and natural purification of
polluted waters. Zentralbl. Bakteriol. Hyg. ^^ orig.
8158:432-445.
-------
529
Portnoy, B.L., Mackowiak, P.A., Caraway, C.T., Walker, J.A.,
McKinley, T.W. and Klein, C.A. (1975) Oyster-associated
hepatitis. Journal of the American Medical Association
233(10):1065-1068.
Postgate, J.R. (1976) Death in macrobes and microbes. In
Twenty-sixth Symposium of the Society for General
Microbiology. The Survival of Vegetative Microbes, pp.
1-18 (Edited by Gray, T.R.G. and Postgate, J.R.) Cam-
bridge Univ. Press, Massachusetts.
Poynter, S.F.B. and Mead, G.C. (1964) Volatile organic
liquids and slime production. Journal of Applied Bac-
teriology 27:182-195.
Poynter, S.F.B. and Slade, J.S. (1977) The removal of
viruses by slow sand filtration. Prog. Water Tech.
9:75-88.
Poynter, S.F.B. and Stevens, J.K. (1975) The effects of
storage on the bacteria of hygienic significance.
Proceedings of a Symposium, The Effects of Storage
on Water Quality, Water Research Centre, Medmenham
Laboratory, Henley Road, Medmenham, Marlow, Bucks,
Henry Ling Ltd., Dorchester, 90 pages.
Poynter, S.F.B., Slade, J.S. and Jones, H.H. (1973) The
disinfection of water with special reference to viruses.
Water Treatment Examination 22:194-208.
Prat, R., Nofre, C. and Cier, A. (1968) Effets de 1'hypo-
chlorite de sodium, de I1ozone et des radiations
ionisantes sur les constituants pyrimidiques de
Escherichia coli. Ann. Inst. Pasteur 114:595-607.
Price, A.T. and Valadon, L.R.G. (1970) A bacteriological
study of a new Thames Valley Reservoir. Microbios.
2:269-271.
Ptak, D.J., Ginsburg, W. and Willey, B.F. (1973) Identifi-
cation and incidence of Klebsiella in chlorinated water
supplies. J. American Water Works Assoc. 65:604-607.
Raj, H. (1966) Enrichment medium for selection of Salmonella
from fish homogenate. Appl. Microbiol. 14:12-20.
Reasoner, D.J. (1977) Microbiology: Detection of bacterial
pathogens and their occurrence. Literature Review.
Journal of the Water Pollution Control Fed. 49:1279-
-------
530
Reasoner, D.J. (1978) Microbiology: Detection of bacterial
pathogens and their occurrence. Literature Review.
Journal of the Water Pollution Control Fed. 50:1382-
1395. ~~~ " ~~
Reasoner, D.J. and Geldreich, E.E. (1978) Rapid detection
of waterborne fecal coliforms by CO release.
Mechanizing Microbiology (Edited by SKarpe, A.N. and
Clarke, D.S.) pp. 120-139, C.A. Thoraan, Springfield,
Illinois, U.S.A.
Reasoner, D.J., Blannon, J.C. and Geldreich, E.E. (1976)
The seven hour fecal coliform test. In Proc. Am.
Water Works Assoc. Water Quality Technol. Confr
"- ~ ~~ ~ .4 _ " ~ - -- -^^ -r VT^- + * *. j- m Jr Cl t^G X7
No. 3B-5 (held in San Diego, Calif.) Am. Water Works
Assoc., Denver, Colorado.
Reijnen, G.K. (1973) De werking van een langzame zandfilter
met betrekking tot de reduktie van het kiemgetal en de
afname van de hoeveelheid organische stof. Keurings-
instituut voor Waterleidingartikelen (KIWA) city
rapport nr. SW-156. Rijswijk, The Netherlands, 24
pages.
Rennie, R.p. and Duncan, I.B.R. (1974) Combined biochem-
ical and serological typing of clinical isolates of
Klebsiella. Appl. Microbiol. 28:534-539.
Resnick, G. and Levin, M.A. (1977) Enumeration of bifido-
bacterium in aquatic and fecal samples. Paper Q9 p.
263. Abs. Ann. Meeting (1977) Am. Soc. Microbiol.
Rexing, D.J. (1977) ATP analysis of a potable water supply.
In Proceedings of the Second Bi-Annual ATP Methodology
Symposium, pp^ 391-395, SAI Technology,
Valley Blvd., San Diego, California.
Ridenour, G.M. and Armbruster, E.M. (1949) Bactericidal
effects of chlorine dioxide. J. Am. Water Works
Assoc. 41:537-550. ~~
Ridenour, G.M. and Ingols, R.S. (1946) Inactivation of
poliomyelitis virus by free chlorine. Am. J. Pub.
Hlth. 36:639-644.
Ridenour, G.M. and Ingols, R.S. (1947) Bactericidal proper-
ties of chlorine dioxide. J. Am. Water Works Assoc.
39:561-567.
-------
531
Ridenour, G.M., Ingols, R.S. and Armbruster, E.H. (1949)
Sporicidal properties of chlorine dioxide. Water
and Sewage Works 96:279-283.
Ridgway, J., Ainsworth, R.G. and Gwilliam, R.D. (1978)
Water quality changes chemical'and microbiological
studies. Proc. Conf. Water Distribution Systems, 1978.
Paper 4, Session 1. Publ. Water Research Centre, Med-
menham.
Ritchie, L.S. and Davis, C. (1948) Parasitological findings
and epidemiological aspects of epidemic araebiasis
occurring in occupants of the Mantetsu Apartment
Building, Tokyo, Japan. Amer. Journal of Trop. Med.
28:803-816.
Rittenberg, S.C. (1972) The obligate autotroph the ,
demise of a concept. Antonie van Leewenhoeck* 38:437-
458.
Rivas, M.G. (1967) Influence of water and sewage treatment
processes on Schistosomes. Med. Bull. Standard Oil Co
27:280-293.
Robeck, G.G., Clarke, N.A., Dostal, K.A. (1962) Effective-
ness of water treatment processes in virus removal.
J. Am. Wat. Wks. Assoc. 54:1275-1292.
Rodger, S.M., Schnagl, R.D. and Holmes, I.H. (1975) Bio-
chemical and biophysical characteristics of diarrhea
viruses of human and calf origin. Journal of Virol.
16:1229-1235.
Rook, J.J. (1955) Microbiologic degradation of vulcanized
rubber. Appl. Microbiol. 3:302-309.
Rosen, H.M., Lowthier, F.E. and Clark, R.G. (1974) Get
ready for ozone. Water and Wastes Engineering 7:25-
31.
Rosenberg, M.L., 'Koplan, J.P., Wachsmuth, I.K., Wells,
J.G., Gangarosa, E.J., Guerrant, R.L. and Sack, D.A.
(1977) Epidemic diarrhea at Crater Lake from entero-
toxigenic Escherichia coli. A large waterborne out-
break. Ann. Intern. Med. 86:714-718.
-------
532
Rosenblatt, D.H. (1975) Chlorine and oxychlorine'species
reactivity with organic substances. In Disinfection
Water and Wastewater, pp. 249-276 (Edited by Johnson,
J.D.) Ann Arbor Sci.f Ann Arbor, Mich.
Rosenkranz, H.S. (1973) Sodium hypochlorite and sodium
perborate: Preferential inhibitors of DNA polymerase-
deficient bacteria. Mutat. Res. 21:171-174.
Rudolphs, W., Falk, L.L. and Ragotzkie, R.A. (1950a) Litera-
ture review on the occurrence and survival of enteric,
pathogenic and relative organisms in soil, water,
sewage and sludges, and on vegetation. I. Bacterial
and viral diseases. Sewage Ind. Wastes 22:1261-1281.
Rudolfs, W., Falk, L.L. and Ragotzkie, R.A. (1950b) Lit-
erature review on the occurrence and survival of
enteric, pathogenic and relative organisms in soil,
water, sewage and sludges, and on vegetation. II
Animal parasites. Sewage Ind. Wastes ??. 1417-1427.-
Ruys, A.C. (1940) On the behavior of El. typhi in surface
water. Antonie van Leeuwenhoek J. Microbiol. Serol
7:93-103. ~ - ^^
Sadzikowski, M.R., Sperry, J.G. and Wilkins, T.D. (1977)
Cholesterol-reducing bacterium from human feces.
Appl. and Environ. Microbiol. 34:355-362.
Safe Drinking Water Committee (1977) Drinking Water and
Health. National Academy of Sciences, Washington,
D.C., 939 pages.
Sarrette, B.A., Donglot, C.D. and Vilagenes, R. (1977) A
new and simple method for recuperation of enteroviruses
from water. Water Research 11:355-358.
Scarpino, P.V., Berg, G., Chang, S.L., Dahling, D., and
Lucas, M. (1972) A comparative study of the inacti-
vation of viruses in water by chlorine. Water Res.
o:959965.
Scarpino, P.V., Lucas, M., Dahling, D.R., Berg, G. and
Chang, S.L. (1974) Effectiveness of hypochlorous acid
and hypochlorite ion in destruction of viruses and
bacteria, pp; 359-368. In Chemistry of Water Supply.
Treatment and Distribution (Edited by Rubin, A.J.), Ann
Arbor Science Publishers, Ann Arbor, Michigan.
-------
533
Schantz, E.J., Ghazarossian, V.E., Schnoes, H.K. , Strong,
F.M., Springer, J. P., Pezzanite, J.O. and Clardy, J.
(1975) The structure of saxitoxin. Journal of the
American Chemical Society 97:1238-1239.
Schothorst, M. van (1976) Resuscitation of injured bacteria
in foods. In Inhibition and Inactivation of Vegetative
Microbes, Soc. Appl . Bacteriol. Symp. Ser. 5, pp. 317-
328 (Edited by Skimmer, F.A. and Hugo, W.B.) Academic
Press, London.
Schubert, R.H.W. (1967a) Die Pathogenitat der Aeromonaden
fur Mensch und Tier. Arch. Hyg . 150:709-716.
Schubert, R.H.W. (1967b) Das Vorkommen der Aeromonaden in
oberirdischen Gewassern. Arch . Hyg . 150:668-708.
Schubert, R.H.W. (1967c) Das Schicksal der Aeromonaden in
Klaranlagen, Faulturmen und in der Erde. Arch. Hyg.
151:243-258.
Schubert, R.H.W. (1975a) Der Nachweis von Sporen der Bacillus
species im Rahmen der hygienischen Wasserbeurteilung .
Zbl. Bakt. Hyg. I Abt . Original B 160:155-162.
Schubert, R.H.W. (1975b) Die Relation von aerogenen zu
anaerogenen Aeromonaden der "Hydrophila-Punctata-
Gruppe" in Fliessgewassern . in Abhangigkeit von der
Abfallstof fbelastung. Zbl. Bakt. Hyg., I. Abt. orig.
B 610:237-247.
Schubert, R.H.W. (1976) Der Nachweis von Aeromonaden der
"Hydrophila-Punctata-Gruppe." In Rahmen der Hygien-
ischen Trinkwasserbeurteilung. Zbl. Bakt. Hyg., I
Abt. Orig. B 161:482-497.
Schubert, R.H.W. and Blnm , u. (1974a) Zur Frage der Erwei-
terung der hygienischen Wasserkontrolle auf den Nach-
weis von Pseudomonas aeruginosa. Unpublished Manuscript.
Schubert, R.H.W. and Blum U. (1974b) Zur Frage der Eignung
' der Malachitgrun - Bouillon nach Habs und Kirchner als ,_
Anreicherungsmedium -f ur Pseudomonas aeruginosa aus dem
Wasser. Zbl. Bakt. Hyg., I. Abt. Orig. B 158:583-587.
Schubert, R.H.W. and_Schafer, E. (1971) Vergleichende
Untersuchungen uber die Eliminierung von Viren, Sal-
monellen und Aeromonaden im Vorfluter. GWF Wasser/
Abwasser 112:401-403.
-------
Schubert, R.H.W. and Scheiber, P. (1975) Das Vorkommen von
Pseudomonas aeruginosa im Grundwasser, Oberflachen-
wasser und in Wasserversorgungsleitungen in tropischen
Gebieten. GWF Wasser/Abwasser 116:413-415.
Schubert, R.H.W., Esanu, J., and Esanu, F. (1975) Uber den
Abbau von Acetamid als taxonomisches Merkmal bei
Species der Gattung Pseudomonas. Zbl. Bakt. Hyg.,
I. Abt. Prig. A 233:342-346. ~^
Schulze, H. and Schweisfurth, R. (1975) Eisenoxydierende
stSbchenformige Bakterien. I. Untersuchungen unter
Verwendung von Fe(II)-Oxalat im Medium. Z. Allg.
Mikrobiol. 15:605-614.
Schweisfurth, R. (1972) Manganoxydierende Mikroorganismen in
Trinkwasserversorgungsanlagen. GWF Wasser/Abwasser
113:562-572. ~ '
Schweisfurth, R. (1973) Manganoxydirende Bakterien. Litera-
turubersicht (Bestandsaufnahme beschriebener Arten)
Zbl. Bakt. I. Ref. 233:257-270.
Schweisfurth, R. (1974) Crenothrix polyspora Cohn as an
indicator of organic pollution of groundwater. Proc.
Soc. Ecol., pp. 161-165 (Edited by Muller, P. and Junk,
W.) Erlangen.
Schweisfurth, R. and Mertes, R. (1962) Mikrobiologische und
chemishe untersuchungen uber Bildung und Bekampfung von
Manganschlamm-Ablagerungen in einer Druckleitung fur
Talsperrenwasser. Arciv. fur Hygjen und Bakterioloqie
146:401-417. ~ ~
Schweisfurth, R. and Ruf, A. (1976) Studies of the occur-
rence of methane-oxidizing bacteria in drinking water
plants. GWF Wasser/Abwasser 117:313-318.
Schweisfurth, R., Eleftheriadis, D., Gundlach, H., Jacobs,
M. and Jung, W. (1978) Microbiology of the precipita-
tion of manganese. In Environmental Biogeochemistry
and Geomicrobiology, Vol. 3, pp. 923-928 (Edited by Krumbein,
W.E.) Ann Arbor Science Publishers, Ann Arbor, Michigan.
Scott, D.B.M. and Leshier, E.G. (1963) Effect of ozone on
survival and permeability of E. coli. J. Bacteriol.
85:567-576. ~ ~
-------
535
Sebald, M. and Veron, M. (1963) Annales de 1'Institut
Pasteur , Lille 105:897.
Seidler, R.J., Morrow, J.E. and Bagley, S.T. (1977) Kleb-
siella in drinking water emanating from redwood tanks.
Appl. Environ. Microbiol. 33:893-905.
Seigneurin, R., Magnin, R., and Archard, M.L. (1955) Types
d'Escherichia coli dans des eaux d1alimentation.
Ann. Inst. Pasteur 89:473-474.
Sepilli, L., Scassellati, R., Sforzolini, M. and Zavka, M.
(1965) Prime osservazioni sulla possibilita di imbot-
tigliamento di acque potabile in recipient! a perdere.
Sanita Publ. 26:1133.
Sepp, E. (1971) The use of sewage for irrigation. A litera-
ture review. State of California Department of Public
Health, Bureau,of Sanitary Engineering, pp. 1-41.
Shaffer, P.T.B. (1978) Disinfectant resistant organisms,
pp. 101-102. In Evaluation of the Microbiology Stan-
dards for Drinking Water (Edited by Hendricks, C.W.)
EPA-570/9-78-OOC.
Sharp, D.G., Floyd, R. and Johnson, J.D. (1975) Nature of
the surviving plaque-forming unit of reovirus in water
containing bromine. Appl. Microbiol. 29:94-101.
Shaw, P.K., Brodsky, R.E., Lyman, D.O., Wood, B.T., Hibler,
C.P., Healy, G.R., Macleod, K.I.E., Stahl, W. and
Schultz, M.G. (October 1977) A community-wide outbreak
of giardiasis with evidence of transmission by a
municipal water supply. Annals of Internal Med.
87:426-432.
Shehabi, A. (1976) Bacteriological pollution in the drink-
ing water of Jordan. Environmental Conservation
3:197-199.
Sherry, J.P., Kuchma, S.R., Zarzour, J. and Dutka, B.J.
(1979) Occurrence and significance of Candida albicans
in Lake Ontario bathing beaches. Environment Canada
Scientific Series No. 98. Inland waters Directorate,
Burlington, Ontario, Canada.
Shih, K.L. and Lederberg, J. ;(1976) Effects of chloramine
on Bacillus subtilis deoxyribonucleic acid. J. Bac-
teriol. 125:934-945.
-------
536
Shillinger, J.E. and McFeters, G.A. (1978) Survival of
Escherichia coli and Yersinia enterocolitica in stream
and tap water. Abstracts of the Annual Meeting, Amer.
Soc. for Microbiol. Abstract #79 , American Society for
Microbiol.
Shotts, E.B. and Rimler, R. (1973) Medium for the isolation
of Aeromonas hydrophila. Applied Microbiology 26:550-
Shrewsbury, J.F.D. and Barson, G.J. (1952) Note on abso-
lute viability in water of Salmonella typhi and dysen-
tery bacilli. Brit. Med. J. 1:954.
Shuval, H.I. (editor) (1977) Water Renovation and Reuse.
Academic Press, New York, San Francisco, London, 463
pages .
Shuval, H.I. and Katzenelson, E. (1977) Detection and
inactivation of enteric viruses in wastewater. Environ
mental Protect. Technol. Series EPA 600/2-77-095"
Siders, D.L., Scarpino, P.V., Lucas, M., Berg, G. and Chang,
S.L. (1973) Destruction of viruses and bacteria in
water by monochloramine. Abstracts of the Annual
Meeting Am. Soc. for Microbiol. (E 27).
Silverman, M.P. and Munoz, E.F. (1979) Automated electrical
impedance technique for rapid enumeration of fecal
coliforms in effluents from sewage treatment plants.
Appl. Environ. Microbiol. 37(3) : 521-526.
Simard, R.E. and Blackwood, A.C. (1970) Yeasts from the St.
Lawrence River. Canadian Journal of Microbiol.
17:197-203. ~ -
Simudi, U. , Ashino, K. and Kaneko, E. (1971) Bacterial
flora of phyto- and zooplankton in the inshore water of
Japan. Journal of Microbiol. 17:1157-1160.
Singh, B.N. (1953) A new system of classifying amoeba based
on their nuclear division and possible pylogenetic
relationships. Symp. Organic Evolution, Bull. No. 7,
pp. 178-183, National Inst. Sci. New Delhi, India.
-------
537
Sinqley, J.E., Kirchmer, C.J. and Miura, R. (1974 Analysis
of coprostanol, an indicator of fecal contamination.
Environmental Protection Technology Series. U.S.
Environmental Protection Agency, Washington, D.C., hPA-
660/2-74-021, 114 pages.
Singley, J.E., Hoadley, A.W. and Hudson, H.E. (1975) A Bene-
bingj.ey,_ ^st'Evaluatjon o£ Dr inking Water Hygiene Problems,
U.S. Environmental Protection Agency, NTIS Report No.
PB 249-891, Washington, B.C.
Skirrow, M.B. (1977) Campvlobacter enteritis; A "new-
disease. British Medical Journal 2:9-11.
Slade, J.S. (1977) Enteroviruses in partially Purified
water. Journal Soc. Water Treatment Exam. 31:219-224.
Sladek, K.J., Suslavich, R.V. , Sohn, B.I. and Dawson, F.W.
(1975) Optimum Membrane Structures for Growth of ^oil-
form and Fecal Coliform Organisms. Appl . Microbiol.
30:685-691.
Smibert R
Smibe
(1978) The genus Campylobacter. In Annual
Microbiology, Vol. 32, pp 673-709 (Edited by
Starr, M.P.) Annual Reviews Inc., Palo Alto, Cali-
fornia.
Smith, L.L. and Gouron, R.E. (1969>S^rol metabolism. VI.
Detection of 5 -cholestan-3 -01 in polluted waters.
Water Res. 3:141-148.
Smith, L.L., Kohler, D.P. , Hempel, J.E. and van Lier, J.E.
(1968) Sterol metabolism. IV. Microbial disposition
of 5 -cholestan-3 -ol. Lipids 3:301-396.
Smith, W.W. and Bodkin, R.E. (1944) The influence of hydro-
gen ion concentrations on the bactericidal action of
ozone and chlorine. J. Bacteriol. 47:445-449.
Snyder, J.E. and Chang, P.W. (1974) Relative r?si^ance of
eight human enteric viruses to ozonation in Saugatucket
River water. In Aquatic Applications of Ozone Proc.
of workshop held in Boston, Massachusetts, 1974) Inter
national Ozone institute (now Association) Cleveland,
Ohio, pp. 29-42.
-------
538
Sobsey, M.D. (1976) Methods for detecting enteric viruses
in water and wastewater. in Viruses in Water (Edited
by Berg, G., Bodily, H.L, Lennette, E.H., Melnick, J.L.
and Metcalf, T.G.) pp. 89-127, American Public Health
Association, Washington, D.C.
Soenke, H. (1974) Untersuchungen uber bakterielle Indi-
katoren und deren Verhalten in Tafelwassern. Disser-
tation Freie Universitat Berlin, Fachbereich Zahn-Mund
und Kieferheilkunde (unpublished).
Sohngen, N.L. and Pol, J.G. (1914) The decomposition of
by microbes- Centr. Bakteriol. Parasitenk.
Sorber, C.A., Malina, J.F. and Sagik, B.P. (I972a) Virus
rejection by the reverse osmosis ultrafiltration
process, Water Res. 6:1377-1388.
Sorber, C.A., Malina, J.F. and Sagik, B.P. (1972b) Quanti-
tative procedure for evaluating the performance of
water and wastewater processes at naturally occurring
virus levels. Environ. Science & Technology 6:438-448,
Sotelo-Avila, C., Taylor, P.M. and Ewing, C.W. (1947) Pri-
mary amoebic meningoencephalitis in a healthy 7-year-
old boy. Journal of Pediatrics 85:131-136.
Sphar, L.L. (1966) Microbiologic degradation of rubber
gaskets in sanitary sewer service. J. Wat. Pollut.
Control Fed. 38:641-644. ~
Spino, D.F. (1966) Elevated temperature technique for the
isolation of Salmonella from streams. Appl. Microbiol,
14:591-596.
Stagg, C.H., Wallis, C. and Ward, C.H, (1977) Inactivation
of clay-associated bacteriophage MS 2 by chlorine.
Appl. Environ. Microbiol. 33:385-391.
Steel, T.W. and McDermott, S.. (1978) Campylobacter enteritis in
South Australia. Medical Journal of Australia 2:404-
406. -"
Straskraba, M. and Straskrabova, V. (1969) Eastern Euro-
pean Lakes, pp. 65-97. In Eutrophication; Causes.
Consequences, Corrections.Proceedings of a Symposium,
Natl. Acaa. sci., Washington, D.C., 661 pages.
-------
539
Stringer, R.P., Cramer, W.M., and Kruse, C.W. (1975) Com-
parison of bromine, chlorine, and iodine as disinfec-
tants for amoebic cysts, pp. 193-210. In Disinfection
Water and Wastewater (Edited by Johnson, J.D.) Ann
Arbor Sci., Ann Arbor, Mich.
Strong, R.P. (1944) Stitt's Diagnosis, Prevention and
Treatment of Tropical Diseases, 7th Edition. Blakiston
Co., Philadelphia.~
Suchkov, B.P. (1964) Decontamination of drinking water
containing agents which can cause intestinal diseases
and enteroviruses by ozonation. Gig, i Sanit. 29:22-
29.
Sussman, S.J. and Rauh, J.E. (1978) Use of chlorine dioxide
in water and wastewater treatment, pp. 344-355. In
Ozone/Chlorine Dioxide Oxidation Products of Organic
Materials (Edited by Rice, R.G. and Cotruvo, J.A.)
International Ozone Institute, Inc., Cleveland, Ohio.
Switzer-House, K.D. and Dutka, B.J. (1977) Fecal sterols:
Their distribution in water and resultant significance.
Water Pollution Res. Canada 13:175-182.
Switzer-House, K.D. and Dutka, B.J. (1978) Fecal sterol
studies: Sample processing and microbial degradation.
Environment Canada Scientific Series, No. 89 N.W.R.I.
Canada Centre for Inland Waters, Burlington, Ontario,
Canada.
Szmuness, W., and Prince, A.M. (1971) Epidemiologic pat-
terns of viral hepatitis in Eastern Europe in light of
recent findings concerning the serum hepatitis antigen.
J. Infect. Pis. 123:200-212.
Tabak, H.H., Bloomhuff, R.N. and Bunch, R.L. (1972) Copro-
stanol: A positive tracer of fecal pollution. In
Development in Industrial Microbiology, Vol. 13, pp.
296-307. Publication of the Society for Industrial
Microbiology, American Institute of Biological Science,
Washington, D.C.
Tabor, W.A. (1976) Wastewater microbiology. Ann.
Rev. Microbiol. 30:263-277.
Tan, L., Clemence, M. and Gass, J. (1970) Gas chromatographic
analysis of fecal pollution sterols on a single com-
bined packed column. J. Chromatogr. 53:209-215.
-------
540
Tannock, G.W. and Smith, J.M.B. (1971) Studies on the
survival of Salmonella typhimurium and Salmonella
bovismorbificans on pasture and water. Aust. Vet. J.
47:557-559.
Taylor, F.B., Eagen, J.H., Smith, H.F.D. and Coene, J.F.
(1966) The case for waterborne infectious hepatitis.
American Journal of Public Health 56:2093-2105.
Taylor, R.H. and Geldreich, E.E. (in press) Standard plate
count methodology: A new membrane filter procedure for
potable water and swimming pools. J. Am. Water Works
Assoc., Denver, Colorado.
Terney, O. (1973) Om kemolitotrofe bakterier. En sammen-
ligning af kemolitotrofe bakteriers aktivitet i naturens
kredsljz>bsprocesser. Report from the University of
Copenhagen, Hygienic Institute, Copenhagen, 90 pages.
Thames Water Statistics (1976) Vol. 2, p. E.2.1.
Thaysen, A.C., Bunker, H.J. and Adams, M.E. (1945) Rubber
acid damage in fire hoses. Nature, Lond. 155:323.
Thornhill, T.S., Wyatt, R.G., Kalica, A.R., Dolin, R.,
Chanock, R.M. and Kapikian, A.Z. (1977) Detection by
immune electron microscopy of 26-27-nm virus-like
particles associated with two family outbreaks of
gastroenteritis. J. of Inf. Pis. 135:20-27.
Tobin, R.S. and Dutka, B.J. (1977) Comparison of the sur-
face structure, metal binding and fecal coliform
recoveries of nine membrane filters. Appl. Environ.
Microbiol. 34:69-79.
Tobin, R.S., Ryan, J.F. and Afghan, B.K. (1978) An improved
method for the determination of adenosinetriphosphate in
environmental samples. Water Research 12:783-792.
Tolsma, D.D. and Bryan, J.A. (1976) The economic impact of
viral hepatitis in the United States. Publ. Health
Rep. 91:349-353.
Toma, S. (1973) Survey on the incidence of Yersinia entero-
colitica in the province of Ontario. Can. J. Public
Health 64:477-487. ~
-------
541
Toma, S. and LaFleur, L. (1974) Survey on the incidence of
Yersinia enterocolitica infection in Canada. Appl.
Microbio'l. 28:469-473.
Tracy, H.W., Camarena, V.M. and Wing, F. (1966) Coliform
persistence in highly chlorinated waters. J. Am.
Water Wks. Ass. 58:1151-1159.
Traenhart, O. and Kuwert, E. (1975) Vergleichende Unter-
suchungen uber die Wirkung von Chlor und Ozon auf
Polioviren bei der Trinkwasseraufbereitung der Stadt
Essen. Zbl. Bakt. Hyg. I. Abt. Prig. B 160:305-341.
Trakhtman, N.N. (1946) Chlorine dioxide in water disin-
fection. Gig. Sanit. 11:10-13.
Trinel, P.A. and Leclerc, H. (1976) Essais de concentration
de bacteries en suspension aqueuse par ultrafiltration,
Ann. Microbiol., Inst. Pasteur 127B:201-212.
Trinel, P.A. and Leclerc, H. (1977) Automatisation de
I1analyse bacteriologique/de 1'eau: Description d'une
nouvelle methode de colimetrie. Ann. Microbiol., Inst,
Pasteur 128A:419-432.
Turner, L.H. (1973) Leptospirosis.
1:537-540.
British Medical Journal
Ur, A. and Brown, D.F.J. (1974) Rapid detection of bac-
terial activity using impedence measurements. Biomed.
Eng. 9:18-20.
Ur, A. and Brown, D.F.J. (1975a) Monitoring of bacterial
activity by impedence measurements. New Approaches to
the Identification of Microorganisms (Edited by Heden,
C.G. and Illeni, T.) pp.61-71, John Wiley & Sons,
Inc., New York.
Ur, A. and Brown* D.F.J. (1975b) Impedence monitoring of
bacterial activity. J. Med. Microbiol. 8:19-28.
Vajdic, A.H. (1970) The isolation of virus (bacteriophage)
from water and wastewater by millipore filter method,
Water Management in Ontario (Ontario Water Resources
Commission) 37:1-22.
Van der Kooij, D. (1977) The occurrence of Pseudomonas sp.
in surface water and in tap water as determined on
citrate media. Antonie van Leeuwenhoek. 43:187-197.
-------
542
Van der Kooij, D. and Zoeteman, B.C.J. (1978) Water quality
in distribution systems. Int. Wat; Supply Ass. Cong.,
Kyoto, Special subject 5.
Van Dijck, P. and Van der Voorde, H. (1976) Sensitivity of
environmental microorganisms to antimicrobial agents.
Applied and Environmental Microbiol. 31:332-336.
Van Donsel, D.J. and Geldreich, E.E. (1971) Relationships of
Salmonella to fecal coliforms in bottom sediments.
Water Res. 5:1079-1087.
Van Donsel, D.J., Twedt, R.M. and Geldreich, E.E. (1969)
Optimum temperature for quantitation of fecal coliform
in seven hours on the membrane filter. Bacteriol. Proc,
Abstract No. G46, 25, Am. Society for Microbiol.,
Washington, D.C.
Van Uden, N. and Buckley, H. (1970) Candida Berkhout. In
The Yeasts A Taxpnomic Study, pp. 893-1087 (Edited
by Lodder, J. ) North-Holland, Amsterdam.
Vaughn, J.M. and Metcalf, T.G. (1975) Coliphages as indi-
cators of enteric viruses in shellfish and shellfish
raising estuarine waters. Water Research 9:613-616.
Venkobachar, D. , Lyengar, L. and Rao, A.V. (1977) Mech-
anism of disinfection effect of chlorine on cell
membrane functions. Water Research 11:727-729.
Verstraete, W. and Voets, J.P. (1976) Comparative study of
E. coli survival in two aquatic ecosystems. Water Res.
Victorin, Ko and Stenstrom, T. (1975) Nitritbildning i
vattenledningsvatten (Formation of nitrite s in water .
distribution systems). Report; Statens Naturvardsverk,
Omgiving Omgivningshygieniska Avdelningen (Report,
.1 Swedish Environment Protection Board Res
Nationa
Laboratory)
Research
Stockholm.
Viswanathan, R. (1957) Infectious hepatitis in Delhi, 1955-
1956: A critical study. I. Epidemiology. Ind.
Journal of Medical Res., Supplementary Number, XLV, 1-
~
Vlassoff, L.T. (1977) Klebsiella. In Bacterial Indicators/
Health Hazards Associated with Water ASTM STP 635
(Edited by Hoadley, A.W. and Dutka, B.J.) Amer. Soc.
for Testing and Materials, Philadelphia, Pennsylvania.
-------
543
Volker, H., Schweisfurth, R. and Hirsch, P. (1977) Mor-
phology and ultrastructure of Crenothrix polyspora
Cohn. J. Bacteriology 131;306-313.~~
Vrochinskii, K.K. (1963) Experimental data on water decon-
tamination with ozone. Hyg. Sanit. 28;3-9.
Waksman, S.A. (1937) Associative and antagonistic effects
of microorganisms. I. Historical review of antago-
nistic relationships. Soil Sci. 43:51-68.
Wallis, C.f Stagg, C.H. , and Melnick, J.L. (1974) The
hazards of incorporating charcoal filters into domestic
water systems. Water Res. 8:111-113.
Warren, Charles E. (1971) Biology and Water Pollution
Control. W.B. Saunders Co., Philadelphia, 434 pages.
Warriner, T.R. (1967) Inaktivering av poliovirus med klor-
dioxid. Vatten 67:284-290.
Watson, S.W., Novitsky, T.J., Quinby, H.L., and Valois,
F.W. (1977) Determination of bacterial number and
biomass in the marine environment. Appl. Environ.
Microbiol. 33:940-946.
Weidenkopf, S.J. (1958) Inactivation of type 1 poliomyelitis
virus with chlorine. Virology 5:56-67.
Wellings, P.M., Amuso, P.T., Chang, S.L. and Lewis, A.L.
(1977) Isolation and identification of pathogenic
Naegleria in Florida lakes. Applied and Environmental
MicrobioT. 34:661.
Wellings, F-M., Lewis, A.L., Mountain, C.W.
and Pierce, L. V. (1975) Demonstration of virus
in groundwater after effluent discharge onto soil.
Appl. Microbiol. 29:751-757.
Werner, P. (1978) Mikrobiologische Untersuchungen des
Wassers von Trinkwasser versorgungsanlagen unter
besondere Berucksichtigung der Aktivkohlefiltration.
Diplom-Arbeit, Universitat des Saarlandes, Saarbrucken,
FRG.
H
Werner, P. (1979) Okologisch-mikrobiologische Untersuchungen
an Aktivkohlefiltern in Zusammenhang mit Trinkwasser-
aufbereitung. Dissertationsarbeit, Universitat des
Saarlandes, Saarbrucken, FRG.
-------
544
Wenyon, C.M. (1947) Discussion on amoebiasis. Royal Soc.
Trop. Med. Hyg. 41:35-39.
White, G.C. (1972) Handbook of chlorination. Van Nostrand
Reinhold Company Publishers, New York, 744 pages.
Willis, A.T. (1956) Anaerobes as an index of faecal pol-
lution in water. Journal of Applied Bacteriology
19:105-107.
Willis, H.T. (1957) Anaerobic bacilli in a treated water
supply. Journal of Applied Bacteriology 20:61-64.
Wilson, G.S. and Miles, A. (1975) Topley and Wilson's
Principles of Bacteriology, Virology and Immunity, 6th
Edition. London, 2706 pages.
Wilson, W.J. and Blair, E.M. (1925) Correlations of the
sulphite-reduction test with other tests in the
Bacteriological examination of water. J. Hyg. 24:111-
119.
Windle, Taylor E. (1947) Bacteriology of water samples from
water mains. Journal of Applied Bacteriology 10:34-38.
Windle, Taylor E. (1958) The Examination of Water and
Water Supplies, 7th ed., Churchill, London.
Winner, H.I. and Hurley, R. (1964) Candida albicans. Little,
Brown, and Co., Boston.
Woode, G.H., Bridger, J.C., Jones, J.M., Flewett, T.H.,
Bryden, A.S., Davies, H.A. and White, G.B.B. (1976)
Morphological and antigenic relationships between
viruses (rotaviruses) from acute gastroenteritis of
children, calves, piglets, mice, and foals. Infection
and Immunity 14:804-810.
Woratz, H. and Bosse, G. (1968) The viability of Salmonellas
in manganese and iron-containing slime in an urban
water supply. Gesundhwes. Desinfekt. 60:97.
World Health Organization (1977) Advances in Viral Hepatitis,
Technical Report Series No. 602, Geneva^Switzerland.
Wuhrmann, K. (1972) Stream purification. In Water Pollution
Microbiology, pp. 119-151 (Edited by Mitchell, R.) John
Wiley and Sons, New York.
-------
545
Wuhrmann, K. and Meyrath, J. (1955) The bactericidal action
of ozone solution. Schweiz. A. Allgem. Pathol. Bak-
teriol. 18:1060-1069.
W.un, C.K., Walker, R.W. and Litsky, W. (1976) The use of
XAD-2 resin for the analysis of coprostanol in water.
Water. Res. 10:955-959.
Wun, C.K., Walker, R.W. and Litsky, W. (1978) An improved
polystyrene polymeric XAD-2 resin column extraction of
5 -cholestan-3 -ol from polluted water. Health Labora-
tory Science 15:67-73.
\
Yoshimizu, M., Kamiyana, K., Kimura, T. and Sakai, M. (1976)
Studies on the intestinal microflora of salmonids. IV.
The intestinal microflora of freshwater salmon. Bul-
letin Jap. Soc. Sci. Fisch. 42:1281-1290.
Zimmermann, R. (1977) Estimation of bacteria number and
biomass by epifluorescence microscopy and scanning
electron microscopy, pp. 103-120. In G. Rheinheimer
(ed.) Microbial Ecology of a Brackish Water Environ-
ment. Springer-Verlag, Berlin, 291 pages.
Zimmerman, R. and Meyer-Reil, L. (1974) A new method for
fluorescence staining of bacterial populations on
membrane filters. Keil. Merresforsch. 30:24-27.
Zo Bell, C.E'. and Grant, C.W. (1943) Bacterial utilization
of low concentrations of organic matter. Journal of
Bacteriol. 43:555-564.
Zoeteman, B.C.J. (1972) The suitability of iodine and iodine
compounds as disinfectants for small water supplies.
Tech Paper No. 2. World Health Organization Intnl.
Ref. Centre for Community Water Supply. WHO File Ref.
W2/181/2.
Zwaagstra, J. (1978) Proeven met koolfilters in langzame
zandfilters. Rapport PWN X.001.2"628.16. Provinciaal
Waterleidingbedrijf van Noord-Holland, Bloemendaal, The
Netherlands.
Zwadyk, P. (1976) Vibrionaceae. In Zinsser Microbiology
(Edited by Joklik, W.K. and Willett, H.P.) pp. 583-592,
Appleton-Century-Crofts, New York, 1223 pages.
tf.S. OOVEEHMEHT IHINTIHS OPPIOE ! 1984 0-421-082/538
-------
-------
-------
------- |