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
 &1•i 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 agar•and 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 1—ml
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-
 turesT—As 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 g•of 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 cross—connections,  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:959—965.

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

-------

-------

-------

-------